Methods in

Molecular Biology 1571

Avraham Rasooly
Ben Prickril Editors

Biosensors
and Biodetection
Methods and Protocols
Volume 1: Optical-Based Detectors
Second Edition
METHODS IN MOLECULAR BIOLOGY

Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK

For further volumes:

http://www.springer.com/series/7651
Biosensors and Biodetection

Methods and Protocols Volume 1:

Optical-Based Detectors
Second Edition

Edited by

Avraham Rasooly
National Cancer Institute
National Institutes of Health
Rockville, MD, USA

Ben Prickril
National Cancer Institute
National Institutes of Health
Rockville, MD, USA
Editors
Avraham Rasooly Ben Prickril
National Cancer Institute National Cancer Institute
National Institutes of Health National Institutes of Health
Rockville, MD, USA Rockville, MD, USA

ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-6846-6 ISBN 978-1-4939-6848-0 (eBook)
DOI 10.1007/978-1-4939-6848-0
Library of Congress Control Number: 2017932742

© Springer Science+Business Media LLC 2009, 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations
and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.
The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Humana Press imprint is published by Springer Nature

The registered company is Springer Science+Business Media LLC
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
This book is dedicated to the memory
of Jury Rasooly Ph.D., Malkah Rasooly,
and Ilan Rasooly.
Preface

Biosensor Technologies

A biosensor is defined by the International Union of Pure and Applied Chemistry (IUPAC)
as “a device that uses specific biochemical reactions mediated by isolated enzymes, immu-
nosystems, tissues, organelles or whole cells to detect chemical compounds usually by
electrical, thermal or optical signals” [1]; all biosensors are based on a two-component
system:
1. Biological recognition element (ligand) that facilitates specific binding or biochemical
reaction with the target analyte
2. Signal conversion unit (transducer)
Since the publication of the first edition of this book in 2009, “classical” biosensor
modalities such as electrochemical or surface plasmon resonance (SPR) continue to be
developed. New biosensing technologies and modalities have also been developed, includ-
ing the use of nanomaterials for biosensors, fiber-optic-based biosensors, genetic code based
sensors, field effect transistors, and the use of mobile communication device-based biosen-
sors. Although it is impossible to describe the fast-moving field of biosensing in a single
publication, this book presents descriptions of methods and uses for some of the basic types
of biosensors while also providing the reader a sense of the enormous importance and
potential for these devices. In order to present a more comprehensive overview, the book
also describes other biodetection technologies.
Dr. Leland C. Clark, who worked on biosensors in the early 1960s, provided an early
reference to the concept of a biosensor by developing an “enzyme electrode” for glucose
concentration measurement using the enzyme glucose oxidase (GOD) [2]. Glucose moni-
toring is essential for diabetes patients, and even today the most common clinical biosensor
technology for glucose analysis is the electrochemical detection method envisioned by Clark
more than 50 years ago. Today, glucose monitoring is performed using rapid point of care
biosensors made possible through advances in electronics that have enabled sensor minia-
turization. The newest generation of biosensors includes phone-based optical detectors with
high-throughput capabilities.

The Use of Biosensors

Biosensors have several potential advantages over other methods of biodetection, including
increased assay speed and flexibility. Rapid, real-time analysis can provide immediate inter-
active information to health-care providers that can be incorporated into the planning of
patient care. In addition, biosensors allow multi-target analyses, automation, and reduced
testing costs. Biosensor-based diagnostics may also facilitate screening for cancer and other
diseases by improving early detection and therefore improving prognosis. Such technology
may be extremely useful for enhancing health-care delivery to underserved populations and
in community settings.

vii
viii Preface

The main advantages of biosensors include:

Rapid or real-time analysis: Direct biosensors such as those employing surface plasmon
resonance (SPR) enable rapid or real-time label-free detection and provide almost
immediate interactive sample information. This enables facilities to take corrective
measures before a product is further processed or released for consumption.
Point of care detection capabilities: Biosensors can be used for point of care testing. This
enables state-of-the-art molecular analysis without requiring a laboratory.
Continuous flow analysis: Many biosensors are designed to allow analysis of bulk liquids. In
such biosensors the target analyte is injected onto the sensor using a continuous flow
system immobilized in a flow cell or column, thereby enhancing the efficiency of analyte
binding to the sensor and enabling continuous monitoring.
Miniaturization: Increasingly, biosensors are being miniaturized for incorporation into
equipment for a wide variety of applications including clinical care, food and dairy
analyses, agricultural and environmental monitoring, and in vivo detection of a variety
of diseases and conditions.
Control and automation: Biosensors can be integrated into online process monitoring
schemes to provide real-time information about multiple parameters at each production
step or at multiple time points during a process, enabling better control and automation
of biochemical facilities.

Biosensor Classification
In general biosensors can be divided into two groups: direct recognition sensors in which
the biological interaction is directly measured and indirect detection sensors which rely on
secondary elements (often catalytic) such as enzymes or fluorescent tags for measurements.
Figure 1 illustrates the two types of biosensors. In each group there are several types of

A B

Recognition Element Recognition Element

Transducer

Output Interface

Fig. 1 General schematic of biosensors. (A) Direct detection biosensors where the recognition element is
label-free; (B) indirect detection biosensors using “sandwich” assay where the analyte is detected by labeled
molecule. Direct detection biosensors are simpler and faster but typically yield a higher limit of detection
compared to indirect detection systems
 Preface ix

optical, electrochemical, or mechanical transducers. Although the most commonly used

ligands are antibodies, other ligands are being developed including aptamers (protein-
binding nucleic acids) and peptides.
There are numerous types of direct and indirect recognition biosensors, and choice of a
suitable detector is complex and based on many factors. These include the nature of the
application, type of labeled molecule (if used), sensitivity required, number of channels (or
area) measured, cost, technical expertise, and speed of detection. In this book we describe
many of these detectors, their application to biosensing, and their fabrication.
The transducer element of biosensors converts the biochemical interactions of the
ligand into a measurable electronic signal. The most important types of transducer used
today are optical, electrochemical, and mechanical.

Direct Label-Free Detection Biosensors

Direct recognition sensors, in which the biological interaction is directly measured in real
time, typically use non-catalytic ligands such as cell receptors or antibodies. Such detectors
typically measure directly physical changes (e.g., changes in optical, mechanical, or electrical
properties) induced by the biological interaction and do not require additional labeled
molecules (i.e., are label-free) for detection. The most common direct detection biosensors
are optical biosensors including biosensors which employ evanescent waves generated when
a beam of light is incident on a surface at an angle yielding total reflection. Common
evanescent wave biosensors are surface plasmon resonance (SPR) or resonant mirror sensors.
Other direct optical detectors include interferometric sensors or grating coupler. Nonoptical
direct detection sensors are quartz resonator transducers that measure change in resonant
frequency of an oscillating piezoelectric crystal as a function of mass (e.g., analyte binding)
on the crystal surface, microcantilevers used in microelectromechanical systems (MEMS)
measuring bending induced by the biomolecular interactions, or field effect transistor (FET)
biosensors, a transistor gated by biological molecules. When biological molecules bind to
the FET gate, they can change the gate charge distribution resulting in a change in
conductance of the FET.

Indirect Label-Based Detection Biosensors

Indirect detection sensors rely on secondary elements for detection and utilize labeling or
catalytic elements such as enzymes. Examples of such secondary elements are the enzyme
alkaline phosphatase and fluorescently tagged antibodies that enhance detection of a sand-
wich complex. Unlike direct sensors, which directly measure changes induced by biological
interaction and are “label-free,” indirect sensors require a labeled molecule bound to the
target. Most optical indirect sensors are designed to measure fluorescence; however, such
sensors can also measure densitometric and colorimetric changes as well as chemilumines-
cence, depending on the type of label used.
Electrochemical transducers measure the oxidation or reduction of an electroactive
compound on the secondary ligand and are one common type of indirect detection sensor.
Several types of electrochemical biosensors have been developed including amperometric
devices, which detect ions in a solution based on electric current or changes in electric
current when an analyte is oxidized or reduced. Another common indirect detection
biosensor employs optical fluorescence, detecting fluorescence of the secondary ligand via
CCD, PMT, photodiode, and spectrofluorometric analysis. In addition, visual measurement
such as change of color or appearance of bands (e.g., lateral flow detection) can be used for
indirect detection.
x Preface

Indirect detection can be combined with direct detection to increase sensitivity or to

validate results; for example, the use of secondary antibody in combination with an SPR
immunosensor. Using a sandwich assay, the analyte captured by the primary antibody is
immobilized on the SPR sensor and generates a signal which can be amplified by the binding
of a secondary antibody to the captured analyte.

Ligands for Biosensors

Ligands are molecules that bind specifically with the target molecule to be detected. The
most important properties of ligands are affinity and specificity. Of the various types of
ligands used in biosensors, immunosensors—particularly antibodies—are the most common
biosensor recognition element. Antibodies (Abs) are highly specific and versatile and bind
strongly and stably to specific antigens. However, Ab ligands have limited long-term stability
and are difficult to produce in large quantities for multi-target biosensor applications where
many ligands are needed.
Other types of ligands such as aptamers and peptides are more suited to high-
throughput screening and chemical synthesis. Aptamers are protein-binding nucleic acids
(DNA or RNA molecules) selected from random pools based on their ability to bind other
molecules with high affinity. Peptides are another potentially important class of ligand
suitable for high-throughput screening due to their ease of selection. However, the affinity
of peptides is often lower than that of antibodies or aptamers, and peptides vary widely in
structural stability and thermal sensitivity.

New Trends in Biosensing

While the fundamental principles and the basic configuration of biosensors have not
changed in the last decade, this book expands the application of these principles using
new technologies such as nanotechnology, integrated optics (IO) bioelectronics, portable
imaging, new fluidics and fabrication methodologies, and new cellular and molecular
approaches.
Integration of nanotechnology: There has been great progress in nanotechnology and nano-
material in recent years. New nanoparticles have been developed having unique electric
conductivity, optical, and surface properties. For example, in several chapters new
optical biosensors are described that integrate nanomaterials in SPR biosensor config-
urations such as localized surface plasmon resonance (LSPR), 3D SPR plasmonic
nanogap arrays, or gold nanoparticle SPR plasmonic peak shift. In addition to SPR
biosensors, nanomaterials are also applied to fluorescence detection utilizing fluores-
cence quantum dot or silica nanoparticles to increase uniform distribution of enzyme
and color intensity in colorimetric biosensors or to improve lateral flow detection. In
addition to optical sensors, gold nanoparticles (AuNPs) have been integrated into
electrochemical biosensors to improve electrochemical performance, and magnetic
nanoparticles (mNPs) have been used to improve sample preparation. Nanoparticle-
modified gate electrodes have been used in the fabrication of organic electrochemical
transistors.
Bioelectronics: Several chapters described the integration of biological elements in electronic
technology including the use of semiconductors in several configurations of field effect
transistors and light-addressable potentiometric sensors.
 Preface xi

Application of imaging technologies: The proliferation of high-resolution imaging

technologies has enabled better 2D image analysis and increases in the number of
analytical channels available for various modalities of optical detection. These include
two-dimensional surface plasmon resonance imaging (2D-SPRi) utilizing CCD cameras
or 2D photodiode arrays. The use of smartphones for both fluorescence and colorimet-
ric detectors is described in several manuscripts.
Integrated optics (IO): Devices with photonic integrated circuits are presented which inte-
grate several optical and often electronic components. Examples include an integrated
optical (IO) nano-immunosensor based on a bimodal waveguide (BiMW) interferomet-
ric transducer integrated into a complete lab-on-a-chip (LOC) platform.
New fluidics and fabrication methodologies: Fluidics and fluid delivery are important com-
ponents of many biosensors. In addition to traditional polymer fabricated microfluidics
systems, inkjet-printed paper fluidics are described that may play an important role in
LOCs and medical diagnostics. Such technologies enable low-cost mass production of
LOCs. In addition, several chapters describe the use of screen printing for device
fabrication.
Cellular and molecular approaches: Molecular approaches are described for aptamer-based
biosensors (aptasensors), synthetic cell-based sensors, loop-mediated DNA amplifica-
tion, and circular strand displacement for point mutation analysis.
While “classic” transducer modalities such as SPR, electrochemical, or piezoelectric
remain the predominant biosensor platforms, new technologies such as nanotechnology,
integrated optics, or advanced fluidics are providing new capabilities and improved
sensitivity.

Aims and Approaches

This book attempts to describe the basic types, designs, and applications of biosensors and
other biodetectors from an experimental point of view. We have assembled manuscripts
representing the major technologies in the field and have included enough technical
information so that the reader can both understand the technology and carry out the
experiments described.
The target audience for this book includes engineering, chemistry, biomedical, and
physics researchers who are developing biosensing technologies. Other target groups are
biologists and clinicians who ultimately benefit from development and application of the
technologies.
In addition to research scientists, the book may also be useful as a teaching tool for
bioengineering, biomedical engineering, and biology faculty and students. To better repre-
sent the field, most topics are described in more than one chapter. The purpose of this
redundancy is to bring several experimental approaches to each topic, to enable the reader to
choose an appropriate design, to combine elements from different designs in order to better
standardize methodologies, and to provide readers more detailed protocols.

Organization
The publication is divided into two volumes. Volume I (Springer Vol. 1571) focuses on
optical-based detectors, while Vol. II (Springer Vol. 1572) focuses on electrochemical,
bioelectronic, piezoelectric, cellular, and molecular biosensors.
xii Preface

Volume I (Springer Vol. 1571)

Optical-based detection encompasses a broad array of technologies including direct and
indirect methods as discussed above. Part I of Vol. I describes various optical-based direct
detectors, while Part II focuses on indirect optical detection. Three types of direct optical
detection biosensors are described: evanescent wave (SPR and resonant waveguide grating),
interferometers, and Raman spectroscopy sensors.
The second part of Vol. I describes various indirect optical detectors as discussed above.
Indirect directors require a labeled molecule to be bound to the signal-generating target.
For optical sensors such molecules emit or modify light signals. Most indirect optical
detectors are designed to measure fluorescence; however, such detectors can also measure
densitometric and colorimetric changes as well as chemiluminescence, and detection
depends on the type of label used. Such optical signals can be measured in various ways as
described in Part II. These include various CCD-based detectors which are very versatile,
inexpensive, and relatively simple to construct and use. Other optical detectors discussed in
Part II are photodiode-based detection systems and mobile phone detectors. Lateral flow
systems that rely on visual detection are included in this section. Although lateral flow
devices are not “classical” biosensors with ligands and transducers, they are included in
this book because of their importance for biosensing. Lateral flow assays use simple immu-
nodetection (or DNA hybridization) devices, such as competitive or sandwich assays, and
are used mainly for medical diagnostics such as laboratory and home testing or any other
point of care (POC) detection. A common format is a “dipstick” in which the test sample
flows on an absorbant matrix via capillary action; detection is accomplished by mixing a
colorimetric reagent with the sample and binding to a secondary antibody to produce lines
or zones at specific locations on the absorbing matrix.

Volume II (Springer Vol. 1572)

Volume II describes various electrochemical, bioelectronic, piezoelectric, and cellular- and
molecular-based biosensors.
In Part I of Vol. II, we describe several types of electrochemical and bioelectronic
detectors. Electrochemical biosensors were the first biosensors developed and are the most
commonly used biosensors in clinical settings (e.g., glucose monitoring). Also included are
several electronic/semiconductor sensors based on the field effect. Unlike electrochemical
sensors, which are indirect detectors and require labeling, electronic/semiconductor bio-
sensors are label-free.
In Part II we describe “mechanical detectors” which modify their mechanical properties
as a result of biological interactions. Such mechanical direct biosensors include piezoelectric
biosensors which change their acoustical resonance and cantilevers which modify their
movement.
Part III describes a variety of biological sensors including aptamer-based sensors and
cellular and phage display technologies.
Part IV describes several microfluidics technologies for cell isolation. In addition, a
number of related technologies including Raman spectroscopy and high-resolution micro-
ultrasound are described.
The two volumes provide comprehensive and detailed technical protocols on current
biosensor and biodetection technologies and examples of their applications and capabilities.

Rockville, MD Avraham Rasooly

Rockville, MD Ben Prickril
 Preface xiii

References
1. International Union of Pure and Applied Chemistry (1992) IUPAC compendium of chemical
terminology, 2nd edn (1997). International Union of Pure and Applied Chemistry, Research Triangle
Park, NC
2. Clark LC Jr, Lyons C (1962) Electrode systems for continuous monitoring in cardiovascular surgery.
Ann N Y Acad Sci 102:29–45
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix
1 Localized Surface Plasmon Resonance (LSPR)-Coupled Fiber-Optic
Nanoprobe for the Detection of Protein Biomarkers . . . . . . . . . . . . . . . . . . . . . . . . 1
Jianjun Wei, Zheng Zeng, and Yongbin Lin
2 Ultra-Sensitive Surface Plasmon Resonance Detection by Colocalized
3D Plasmonic Nanogap Arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Wonju Lee, Taehwang Son, Changhun Lee, Yongjin Oh,
and Donghyun Kim
3 Two-Dimensional Surface Plasmon Resonance Imaging System
for Cellular Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Tanveer Ahmad Mir and Hiroaki Shinohara
4 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings. . . . . . . . . 47
Christophe Caucheteur, Clotilde Ribaut, Viera Malachovska,
and Ruddy Wattiez
5 Biosensing Based on Magneto-Optical Surface Plasmon Resonance . . . . . . . . . . . 73
Sorin David, Cristina Polonschii, Mihaela Gheorghiu,
Dumitru Bratu, and Eugen Gheorghiu
6 Nanoplasmonic Biosensor Using Localized Surface Plasmon
Resonance Spectroscopy for Biochemical Detection . . . . . . . . . . . . . . . . . . . . . . . . . 89
Diming Zhang, Qian Zhang, Yanli Lu, Yao Yao,
Shuang Li, and Qingjun Liu
7 Plasmonics-Based Detection of Virus Using Sialic Acid Functionalized
Gold Nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Changwon Lee, Peng Wang, Marsha A. Gaston, Alison A. Weiss,
and Peng Zhang
8 MicroRNA Biosensing with Two-Dimensional Surface
Plasmon Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Ho Pui Ho, Fong Chuen Loo, Shu Yuen Wu, Dayong Gu,
Ken-Tye Yong, and Siu Kai Kong
9 Gold Nanorod Array Biochip for Label-Free, Multiplexed
Biological Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Zhong Mei, Yanyan Wang, and Liang Tang
10 Resonant Waveguide Grating Imager for Single Cell Monitoring
of the Invasion of 3D Speheroid Cancer Cells Through Matrigel . . . . . . . . . . . . . 143
Nicole K. Febles, Siddarth Chandrasekaran, and Ye Fang

xv
xvi Contents

11 Label-Free Biosensors Based on Bimodal Waveguide

(BiMW) Interferometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Sonia Herranz, Adrián Fernández Gavela, and Laura M. Lechuga
12 DNA-Directed Antibody Immobilization for Robust Protein Microarrays:
Application to Single Particle Detection ‘DNA-Directed Antibody
Immobilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
€ u , Fulya Ekiz Kanik, Elif Seymour,
Nese Lortlar Unl€
John H. Connor, and M. Selim Unl€ € u
13 Reflectometric Interference Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Guenther Proll, Goran Markovic, Peter Fechner, Florian Proell,
and Guenter Gauglitz
14 Hypermulticolor Detector for Quantum-Antibody Based
Concurrent Detection of Intracellular Markers for HIV Diagnosis . . . . . . . . . . . . 221
Annie Agnes Suganya Samson and Joon Myong Song
15 Low-Cost Charged-Coupled Device (CCD) Based Detectors for Shiga
Toxins Activity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Reuven Rasooly, Ben Prickril, Hugh A. Bruck, and Avraham Rasooly
16 Smartphone-Enabled Detection Strategies for Portable
PCR–Based Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Aashish Priye and Victor M. Ugaz
17 Streak Imaging Flow Cytometer for Rare Cell Analysis . . . . . . . . . . . . . . . . . . . . . . 267
Joshua Balsam, Hugh Alan Bruck, Miguel Ossandon, Ben Prickril,
and Avraham Rasooly
18 Rapid Detection of Microbial Contamination Using
a Microfluidic Device. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Mustafa Al-Adhami, Dagmawi Tilahun, Govind Rao,
Chandrasekhar Gurramkonda, and Yordan Kostov
19 Resonance Energy Transfer-Based Nucleic Acid Hybridization
Assays on Paper-Based Platforms Using Emissive
Nanoparticles as Donors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Samer Doughan, M. Omair Noor, Yi Han, and Ulrich J. Krull
20 Enhanced Performance of Colorimetric Biosensing on Paper
Microfluidic Platforms Through Chemical Modification
and Incorporation of Nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Ellen Flávia Moreira Gabriel, Paulo T. Garcia, Elizabeth Evans,
Thiago M.G. Cardoso, Carlos D. Garcia,
and Wendell K.T. Coltro
21 A Smartphone-Based Colorimetric Reader for Human C-Reactive
Protein Immunoassay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
A.G. Venkatesh, Thomas van Oordt, E. Marion Schneider,
Roland Zengerle, Felix von Stetten, John H.T. Luong,
and Sandeep Kumar Vashist
 Contents xvii

22 A Novel Colorimetric PCR-Based Biosensor for Detection

and Quantification of Hepatitis B Virus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
Li Yang, Mei Li, Feng Du, Gangyi Chen, Afshan Yasmeen,
and Zhuo Tang
23 CCD Camera Detection of HIV Infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
John R. Day
24 “Dipstick” Colorimetric Detection of Metal Ions Based
on Immobilization of DNAzyme and Gold Nanoparticles
onto a Lateral Flow Device. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
Debapriya Mazumdar, Tian Lan, and Yi Lu
25 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses. . . . . . . . . . . . . . . . 407
Katie A. Edwards, Ricki Korff, and Antje J. Baeumner
26 Development of Dual Quantitative Lateral Flow Immunoassay
for the Detection of Mycotoxins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
Yuan-Kai Wang, Ya-Xian Yan, and Jian-He Sun

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
Contributors

MUSTAFA AL-ADHAMI  University of Maryland Baltimore County, Baltimore, MD, USA

ANTJE J. BAEUMNER  Department of Biological and Environmental Engineering, Cornell
University, Ithaca, NY, USA; Institute for Analytical Chemistry, Chemo- and Biosensors,
University of Regensburg, Regensburg, Germany
JOSHUA BALSAM  Center for Devices and Radiological Health, FDA, Silver Spring, MD,
USA
DUMITRU BRATU  International Centre of Biodynamics, Bucharest, Romania
HUGH A. BRUCK  University of Maryland College Park (UMCP), College Park, MD, USA
THIAGO M.G. CARDOSO  Instituto de Quı́mica, Universidade Federal de Goiás, Goiânia,
GO, Brazil
CHRISTOPHE CAUCHETEUR  Université de Mons, Mons, Belgium
SIDDARTH CHANDRASEKARAN  Biochemical Technologies, Corning Research and Development
Corporation, Corning Incorporated, Corning, NY, USA; Department of Biomedical
Engineering, Cornell University, Ithaca, NY, USA
GANGYI CHEN  Natural Products Research Center, Chengdu Institution of Biology, Chinese
Academy of Science, Chengdu, China
WENDELL K.T. COLTRO  Instituto de Quı́mica, Universidade Federal de Goiás, Goiânia,
GO, Brazil; Instituto Nacional de Ciência e Tecnologia de Bioanalı́tica (INCTBio),
Campinas, SP, Brazil
JOHN H. CONNOR  Microbiology Department, Boston University School of Medicine, Boston,
MA, USA
SORIN DAVID  International Centre of Biodynamics, Bucharest, Romania
JOHN R. DAY  Illumina, Inc., San Diego, CA, USA
SAMER DOUGHAN  Chemical Sensors Group, Department of Chemical and Physical Sciences,
University of Toronto Mississauga, Mississauga, ON, Canada
FENG DU  Natural Products Research Center, Chengdu Institution of Biology, Chinese
Academy of Science, Chengdu, China
KATIE A. EDWARDS  Department of Biological and Environmental Engineering, Cornell
University, Ithaca, NY, USA
ELIZABETH EVANS  Department of Chemistry, , Clemson University, Clemson, SC, USA
YE FANG  Biochemical Technologies, Corning Research and Development Corporation,
Corning Incorporated, Corning, NY, USA
NICOLE K. FEBLES  Biochemical Technologies, Corning Research and Development
Corporation, Corning Incorporated, Corning, NY, USA; NanoScience Technology Center,
Department of Mechanical, Materials and Aerospace Engineering, University of Central
Florida, Orlando, FL, USA
PETER FECHNER  Biametrics GmbH, Tuebingen, Germany
ELLEN FLÁVIA MOREIRA GABRIEL  Instituto de Quı́mica, Universidade Federal de Goiás,
Goiânia, GO, Brazil
PAULO T. GARCIA  Instituto de Quı́mica, Universidade Federal de Goiás, Goiânia, GO,
Brazil
CARLOS D. GARCIA  Department of Chemistry, Clemson University, Clemson, SC, USA

xix
xx Contributors

MARSHA A. GASTON  Department of Molecular Genetics, Biochemistry and Microbiology,

University of Cincinnati, Cincinnati, OH, USA
GUENTER GAUGLITZ  Institute of Physical and Theoretical Chemistry,
University of Tuebingen, Tuebingen, Germany
ADRIÁN FERNÁNDEZ GAVELA  Nanobiosensors and Bioanalytical Applications Group,
Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC, The Barcelona
Institute of Science and Technology, Bellaterra, Barcelona, Spain
EUGEN GHEORGHIU  International Centre of Biodynamics, Bucharest, Romania; University
of Bucharest, Bucharest, Romania
MIHAELA GHEORGHIU  International Centre of Biodynamics, Bucharest, Romania
DAYONG GU  Department of Electronic Engineering, Center for Advanced Research in
Photonics, The Chinese University of Hong Kong, N.T. Hong Kong SAR, China
CHANDRASEKHAR GURRAMKONDA  University of Maryland Baltimore County, Baltimore,
MD, USA
YI HAN  Chemical Sensors Group, Department of Chemical and Physical Sciences,
University of Toronto Mississauga, Mississauga, ON, Canada
SONIA HERRANZ  Nanobiosensors and Bioanalytical Applications Group, Catalan
Institute of Nanoscience and Nanotechnology (ICN2), CSIC, The Barcelona Institute
of Science and Technology, Bellaterra, Barcelona, Spain
HO PUI HO  Department of Electronic Engineering, Center for Advanced Research in
Photonics, The Chinese University of Hong Kong, N.T. Hong Kong SAR, China
FULYA EKIZ KANIK  Electrical and Computer Engineering Department, Boston University,
Boston, MA, USA
DONGHYUN KIM  School of Electrical and Electronic Engineering, Yonsei University,
Seoul, Republic of Korea
SIU KAI KONG  Department of Electronic Engineering, Center for Advanced Research in
Photonics, The Chinese University of Hong Kong, N.T. Hong Kong SAR, China
RICKI KORFF  Department of Biological and Environmental Engineering, Cornell
University, Ithaca, NY, USA
YORDAN KOSTOV  University of Maryland Baltimore County, Baltimore, MD, USA
ULRICH J. KRULL  Chemical Sensors Group, Department of Chemical
and Physical Sciences, University of Toronto Mississauga, Mississauga, ON, Canada
TIAN LAN  Glucosentient Inc., Champaign, IL, USA
LAURA M. LECHUGA  Nanobiosensors and Bioanalytical Applications Group, Catalan
Institute of Nanoscience and Nanotechnology (ICN2), CSIC, The Barcelona
Institute of Science and Technology, Bellaterra, Barcelona, Spain; CIBER-BBN,
Campus UAB, Ed-ICN2, Bellaterra, Barcelona, Spain
CHANGWON LEE  Department of Chemistry, University of Cincinnati, Cincinnati,
OH, USA
WONJU LEE  School of Electrical and Electronic Engineering, Yonsei University, Seoul,
Republic of Korea
CHANGHUN LEE  School of Electrical and Electronic Engineering, Yonsei University, Seoul,
Republic of Korea
SHUANG LI  Biosensor National Special Laboratory, Key Laboratory for Biomedical
Engineering of Education Ministry, Department of Biomedical Engineering, Zhejiang
University, Hangzhou, China
MEI LI  Natural Products Research Center, Chengdu Institution of Biology, Chinese
Academy of Science, Chengdu, China
 Contributors xxi

YONGBIN LIN  Center for Applied Optics, University of Alabama at Huntsville,

Huntsville, AL, USA
QINGJUN LIU  Biosensor National Special Laboratory, Key Laboratory for Biomedical
Engineering of Education Ministry, Department of Biomedical Engineering, Zhejiang
University, Hangzhou, China
FONG CHUEN LOO  Department of Electronic Engineering, Center for Advanced Research
in Photonics, The Chinese University of Hong Kong, N.T. Hong
Kong SAR, China
YANLI LU  Biosensor National Special Laboratory, Key Laboratory for Biomedical
Engineering of Education Ministry, Department of Biomedical Engineering, Zhejiang
University, Hangzhou, China
YI LU  Department of Chemistry, University of Illinois at Urbana-Champaign, Urbana,
IL, USA
JOHN H.T. LUONG  Innovative Chromatography Group, Irish Separation Science Cluster
(ISSC), Department of Chemistry and Analytical, Biological Chemistry Research Facility
(ABCRF), University College Cork, Cork, Ireland
VIERA MALACHOVSKA  B-SENS, Mons, Belgium
E. MARION SCHNEIDER  Sektion Experimentelle Anaesthesiologie, University Hospital Ulm,
Ulm, Germany
GORAN MARKOVIC  Biametrics GmbH, Tuebingen, Germany
DEBAPRIYA MAZUMDAR  ANDalyze Inc., Champaign, IL, USA
ZHONG MEI  Department of Biomedical Engineering, University of Texas at San Antonio,
San Antonio, TX, USA
TANVEER AHMAD MIR  Graduate School of Science and Engineering for Research, University
of Toyama, Toyama, Japan; Graduate School of Innovative Life Sciences for Education,
University of Toyama, Toyama, Japan; Institutes for Analytical Chemistry, Chemo- and
Biosensor, University of Regensburg, Regensburg, Germany; Institute of BioPhysio Sensor
Technology (IBST), Pusan National University, Busan, South Korea
M. OMAIR NOOR  Chemical Sensors Group, Department of Chemical and Physical Sciences,
University of Toronto Mississauga, Mississauga, ON, Canada
YONGJIN OH  School of Electrical and Electronic Engineering, Yonsei University, Seoul,
Republic of Korea
THOMAS VAN OORDT  Hahn-Schickard, Freiburg, Germany
MIGUEL OSSANDON  National Cancer Institute, NIH/NCI, Rockville, MD, USA
CRISTINA POLONSCHII  International Centre of Biodynamics, Bucharest, Romania
BEN PRICKRIL  National Cancer Institute, National Institutes of Health, Rockville, MD,
USA
AASHISH PRIYE  Artie McFerrin Department of Chemical Engineering, Texas A&M
University, College Station, TX, USA
FLORIAN PROELL  Institute of Physical and Theoretical Chemistry, University of Tuebingen,
Tuebingen, Germany; Biametrics GmbH, Tuebingen, Germany
GUENTHER PROLL  Institute of Physical and Theoretical Chemistry, University of Tuebingen,
Tuebingen, Germany; Biametrics GmbH, Tuebingen, Germany
GOVIND RAO  University of Maryland Baltimore County, Baltimore, MD, USA
REUVEN RASOOLY  Western Regional Research Center, Agricultural Research Service,
U.S. Department of Agriculture, Albany, CA, USA
AVRAHAM RASOOLY  National Cancer Institute, National Institutes of Health Rockville,
MD, USA
xxii Contributors

CLOTILDE RIBAUT  B-SENS, Mons, Belgium

ELIF SEYMOUR  Biotechnology Research Program Department, ASELSAN Research Center,
Ankara, Turkey
HIROAKI SHINOHARA  Graduate School of Science and Engineering for Research, University
of Toyama, Toyama, Japan; Graduate School of Innovative Life Sciences for Education,
University of Toyama, Toyama, Japan
TAEHWANG SON  School of Electrical and Electronic Engineering, Yonsei University, Seoul,
Republic of Korea
JOON MYONG SONG  College of Pharmacy, Seoul National University, Seoul, South Korea
FELIX VON STETTEN  Hahn-Schickard, Freiburg, Germany; Laboratory for MEMS
Applications, Department of Microsystems Engineering—IMTEK, University of Freiburg,
Freiburg, Germany
ANNIE AGNES SUGANYA SAMSON  College of Pharmacy, Seoul National University, Seoul,
South Korea
JIAN-HE SUN  Key Laboratory of Veterinary Biotechnology, School of Agriculture and
Biology, Shanghai Jiao Tong University, Shanghai, China
LIANG TANG  Department of Biomedical Engineering, University of Texas at San Antonio,
San Antonio, TX, USA
ZHUO TANG  Natural Products Research Center, Chengdu Institution of Biology, Chinese
Academy of Science, Chengdu, China
DAGMAWI TILAHUN  University of Maryland Baltimore County, Baltimore, MD, USA
VICTOR M. UGAZ  Artie McFerrin Department of Chemical Engineering, Texas A&M
University, College Station, TX, USA; Department of Biomedical Engineering, Texas
A&M University, College Station, TX, USA
NESE LORTLAR U € NLU€  Biomedical Engineering Department, Boston University, Boston, MA,
USA; Faculty of Medicine, Bahcesehir University, Istanbul, Turkey
M. SELIM U€ NLU€  Biomedical Engineering Department, Boston University, Boston, MA,
USA; Electrical and Computer Engineering Department, Boston University, Boston,
MA, USA; Microbiology Department, Boston University School of Medicine, Boston
University, Boston, MA, USA
SANDEEP KUMAR VASHIST  Hahn-Schickard, Freiburg, Germany; Laboratory for MEMS
Applications, Department of Microsystems Engineering—IMTEK, University of Freiburg,
Freiburg, Germany; Immunodiagnostics Systems, Liege, Belgium
A.G. VENKATESH  Department of Electrical and Computer Engineering, Jacobs School of
Engineering, University of California San Diego, San Diego, CA, USA
PENG WANG  Department of Chemistry, University of Cincinnati, Cincinnati, OH, USA
YANYAN WANG  Department of Biomedical Engineering, University of Texas at San Antonio,
San Antonio, TX, USA
YUAN-KAI WANG  Key Laboratory of Veterinary Biotechnology, School of Agriculture and
Biology, Shanghai Jiao Tong University, Shanghai, China
RUDDY WATTIEZ  B-SENS, Mons, Belgium
JIANJUN WEI  Department of Nanoscience, Joint School of Nanoscience and
Nanoengineering (JSNN), University of North Carolina at Greensboro, Greensboro,
NC, USA
ALISON A. WEISS  Department of Molecular Genetics, Biochemistry and Microbiology,
University of Cincinnati, Cincinnati, OH, USA
SHU YUEN WU  Department of Electronic Engineering, Center for Advanced Research in
Photonics, The Chinese University of Hong Kong, N.T. Hong Kong SAR, China
 Contributors xxiii

YA-XIAN YAN  Key Laboratory of Veterinary Biotechnology, School of Agriculture and Biology,
Shanghai Jiao Tong University, Shanghai, China
LI YANG  Natural Products Research Center, Chengdu Institution of Biology, Chinese
Academy of Science, Chengdu, China
YAO YAO  Biosensor National Special Laboratory, Key Laboratory for Biomedical
Engineering of Education Ministry, Department of Biomedical Engineering, Zhejiang
University, Hangzhou, China
AFSHAN YASMEEN  Natural Products Research Center, Chengdu Institution of Biology,
Chinese Academy of Science, Chengdu, China
KEN-TYE YONG  Department of Electronic Engineering, Center for Advanced Research in
Photonics, The Chinese University of Hong Kong, N.T. Hong Kong SAR, China
ZHENG ZENG  Department of Nanoscience, Joint School of Nanoscience and
Nanoengineering (JSNN), University of North Carolina at Greensboro, Greensboro,
NC, USA
ROLAND ZENGERLE  Hahn-Schickard, Freiburg, Germany; Laboratory for MEMS
Applications, Department of Microsystems Engineering—IMTEK, University of Freiburg,
Freiburg, Germany
QIAN ZHANG  Biosensor National Special Laboratory, Key Laboratory for Biomedical
Engineering of Education Ministry, Department of Biomedical Engineering, Zhejiang
University, Hangzhou, China
PENG ZHANG  Department of Chemistry, University of Cincinnati, Cincinnati, OH, USA
DIMING ZHANG  Biosensor National Special Laboratory, Key Laboratory for Biomedical
Engineering of Education Ministry, Department of Biomedical Engineering, Zhejiang
University, Hangzhou, China
 Chapter 1

Localized Surface Plasmon Resonance (LSPR)-Coupled

Fiber-Optic Nanoprobe for the Detection of Protein
Biomarkers
Jianjun Wei, Zheng Zeng, and Yongbin Lin

Abstract
Here is presented a miniaturized, fiber-optic (FO) nanoprobe biosensor based on the localized surface
plasmon resonance (LSPR) at the reusable dielectric-metallic hybrid interface with a robust, gold nano-disk
array at the fiber end facet. The nanodisk array is directly fabricated using electron beam lithography (EBL)
and metal lift-off process. The free prostate-specific antigen (f-PSA) has been detected with a mouse anti-
human prostate-specific antigen (PSA) monoclonal antibody (mAb) as a specific receptor linked with a self-
assembled monolayer (SAM) at the LSPR-FO facet surfaces. Experimental investigation and data analysis
found near field refractive index (RI) sensitivity at ~226 nm/RIU with the LSPR-FO nanoprobe, and
demonstrated the lowest limit of detection (LOD) at 100 fg/mL (~3 fM) of f-PSA in PBS solutions. The
SAM shows insignificant nonspecific binding to the target biomarkers in the solution. The control
experimentation using 5 mg/mL bovine serum albumin in PBS and nonspecific surface test shows the
excellent specificity and selectivity in the detection of f-PSA in PBS. These results indicate important
progress toward a miniaturized, multifunctional fiber-optic technology that integrates informational com-
munication and sensing function for developing a high-performance, label-free, point-of-care (POC)
device.

Key words Fiber optics, Protein biomarker biosensors, Nanofabrication, Au nanodisk array, Localized
surface plasmon resonance (LSPR), Signal transduction

1 Introduction

Recent advances of biomarker detection have been made in optical

fluorescence [1], light scattering [2], surface enhanced Raman
spectroscopy (SERS) [3], electrochemical [4], functional quartz
crystal microbalance [5], microcantilevers [6], and surface plasmon
resonance (SPR) imaging [7] sensors. Harnessing the advances in
biological ligand interactions, the fiber-optic (FO) technology
incorporating localized surface plasmon resonance (LSPR) nanop-
robe sensing may provide an alternative tool for effective biomarker
diagnosis via a compact, label-free format that does not require a

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_1,
© Springer Science+Business Media LLC 2017

1
2 Jianjun Wei et al.

dedicated laboratory facility or highly trained personnel. Further-

more, there is a need to develop advanced LSPR-FO biosensors
that may avoid the use of bulky optics and high-precision mechanics
for angular or wavelength interrogation of metal films in contact
with analytes, and provide high-performance, e.g., enhanced stabil-
ity and high RI sensitivity, and overcome unwanted doping or weak
adhesion.
Surface plasmon resonance (SPR) is the resonance oscillation of
conduction electrons at the interface between a negative and a
positive permittivity material excited by an electromagnetic radia-
tion, e.g., light. The surface plasmon polaritons (SPPs) launched
upon the radiation can be propagating along the metal-dielectric
interface and decay evanescently at the normal direction for a flat
surface. Surface plasmons [8] (SPs) are very sensitive to the near
surface refractive index (RI) changes and well suited to the detec-
tion of surface-binding events. The basic methodology of SPR
sensing is based on the Kretschmann configuration (Fig. 1) where
a prism is used for the light-SP coupling at the surface of a thin
metal film. The probe light undergoes total internal reflection on
the inner surface of the prism. At a defined SPR angle, an evanes-
cent light field travels through the thin gold film and excites SPs at
the metal-dielectric interface, reducing the intensity of the reflected
light at the resonance wavelength or changing the phase of the
incident light. The intensities of scattered and transmitted light
fields are used to determine the thickness and/or dielectric con-
stant of the coating [9]. The control variables for SPR sensor
applications, i.e., the wavelength of incident light, the thickness of
the metal film, the physical and optical properties of the prism, and
the RI of the medium near the metallic interface have been well
studied [10]. However, the advantages of averaging over a large
surface area and the challenges of miniaturizing the optics limit the
integration of SPR-based sensing.

Fig. 1 A conventional surface plasmon resonance (SPR) configuration and setup

for biological sensing
Localized Surface Plasmon Resonance (LSPR)-Coupled Fiber-Optic. . . 3

Fig. 2 Schematic diagrams illustrating (a) a localized surface plasmon [18], (b) a
configuration of representative experimental setup and procedure for LSPR
sensing [19]

LSPR is caused by resonant surface plasmons localized in nano-

scale systems when light interacts with particles much smaller than
the incident wavelength (Fig. 2a). Similar to the SPR, the LSPR is
sensitive to changes in the local dielectric environment near the
nanoparticle surfaces. Usually, the sense changes are measured in
the local environment through an LSPR wavelength shift of the
scattering light via reflection or transmission (Fig. 2b). Nanoscale
LSPR makes possible for the development of a portable device for
point-of-care (POC) detection regarding its requirements, such as
robust to handle, small volume sample, and little to no sample
pretreatment, label-free and rapid response time, and compact size.
Incorporation of the LSPR to fiber optics (FO) offers a few
advantages in terms of avoiding the use of bulky optics and high-
precision mechanics for angular or wavelength interrogation of
4 Jianjun Wei et al.

metal films in contact with analytes, including immobilization of Au

or Ag nanoparticles (NPs) to optical fiber probes for LSPR detec-
tion [11, 12]. It may allow realizing an optical communication and
analytical tool for a wide spectrum of applications. However, more
controllable and stable LSPR nanoscale systems for fiber optic
integration are desirable to develop robust, reliable, portable
LSPR biosensors.
In this work, the progress of developing a miniaturized LSPR-
FO probe and demonstration of sensitive, label-free detection of a
protein cancer biomarker, free prostate-specific antigen (f-PSA) are
reported, which involves three major steps: (1) a low-cost lift-off
process adapted to fabricate gold nanodisk arrays at the end of tip-
facet, providing a very stable, robust, and clean LSPR fiber tip
probe, (2) the probe was functionalized via a facile self-assembled
monolayer (SAM) of alkanethiolates on the gold nanodisk array to
attach a capture ligand, anti-PSA antibody, as a selective immuno-
assay for the detection of the f-PSA, (3) the FO-based sensing
technology was used as a powerful analytical tool by integrating
the LSPR nanoprobe to communicative fiber optics. The sensing
principle and configuration of the LSPR nanoprobe is shown in
Fig. 3. The white light guided in the optical fiber using as incident
light to the gold nanodisk arrays at the end of the fiber tip surface
for excitation of the LSPR. The reflectance of the light scattering
from the nanodisk arrays was recorded before and after the
biological binding. The correlation of the changes of reflectance
spectra to the binding of the analytes to the nanodisk arrays,
corresponding to the analyte concentrations in samples, was estab-
lished for detection. The reported label-free LSPR fiber biosensor
may allow an alternative approach for direct discrimination of the

Au Nanodisk
140
Reflectance Intensity

120
100 After
White Binding
Cladding

80
Light Antigen 60
40
Optical Fiber Tip 20

Core 0
500 600 700 800 900
LSPR Reflectance (nm)
LSPR Reflectance
Fig. 3 A diagram illustrating the principle of LSPR coupling on fiber optic probe for biosensing. The nanodisk
arrays are fabricated on optical fiber tip end. The LSPR reflectance is recorded with a white incident light
guided in the fiber
 Localized Surface Plasmon Resonance (LSPR)-Coupled Fiber-Optic. . . 5

cancer biomarkers, and potentially developing a miniaturized,

point-of-care (POC) device for early disease diagnosis.
This work suggests that: (1) as an emerging technology, the
LSPR-FO sensor has ultra-high sensitivity in molecular adsorption
detection; (2) the LSPR Au nano-array fabricated at the end facet of
the fiber tip for sensing is very robust and reusable; (3) the primary
resonant wavelength can be tuned to a desired range (e.g., NIR) by
tailoring the nano-array structure to enhance the sensitivity; (4)
tailored surface functionalization harnessing the advances in
biological ligand interactions (e.g., immunoassays) enables signal
amplification and a label-free, selective detection; and (5) the target
molecules are immobilized by dipping the fiber-optic probe in
sample/reagent solution, contrary to pouring of sample solution
in conventional methods (ELISA, Electrochemiluminscence,
Radioimmunoassay-RIA), resulting in drastic reduction of the
amount of sample/reagents needed and decreases the washing
time of probes.

2 Materials and Equipment

1. Single-mode optical fiber for 633 nm wavelength was pur-

chased from Newport Corporation.
2. Electron beam resist (ZEP-520A), thinner (ZEP-A), developer
(ZED-N50), resist remover (ZEDMAC) were purchased from
ZEON Corporation, Japan, and used without further
purification.
3. 11-Mecaptoundecanoic acid (HSC10COOH, 99%),
8-Mercapto-1-Octanol (HSC8OH, 98%), N-(3-Dimethyla-
minopropyl)-N0 -ethylcarbodiimide hydrochloride (EDC),
N-Hydroxysuccinimide (NHS), and glycine were purchased
from Sigma-Aldrich (Milwaukee, WI) and used without
further purification.
4. Mouse anti-human PSA monoclonal antibody (capture mAb),
human free-PSA, and ELISA kits for CA125 were obtained
from Anogen-YES Biotech Laboratories Ltd. (Mississauga,
Canada).
5. The 5 ng/mL free-PSA standard solution was used for prepa-
ration of free-PSA solutions with lower concentrations
obtained using sample dilutant provided in the ELISA kit.
The 5 ng/mL free-PSA standard solution was prepared in a
protein matrix solution according to the World Health Orga-
nization (WHO) standard [13] by the vendor.
6. Chrome etchant was obtained from Microchem GmbH,
Germany.
6 Jianjun Wei et al.

7. The fiber clamp for hoisting fiber for vibration was obtained
from Newport Corporation, CA, USA.
8. Ultra clean convection oven (Ultra-clean 100) was obtained
from Thermo Fisher Scientific, OH, USA.
9. Nano pattern generation system (NPGS) was obtained from JC
Nabity Lithography Systems, MT, USA.
10. Field emission scanning electron microscope (FESEM) was a
LEO 1550 model.
11. Three cathode vacuum sputter system (Denton Discovery-18
Sputter System) was obtained from Denton Vacuum LLC, NJ,
USA.
12. Vacuum thermal evaporate system was donated to University
of Alabama in Huntsville by US Army.
13. Reactive Ion Etching (RIE) system (Plasma-Therm 790) was
obtained from Plasma-Therm, FL, USA.
14. Optical fiber coupled Tungsten Halogen light source (LS-1)
was obtained from Ocean Optics, FL, USA.
15. Mini-spectrometer (USB2000-VIS-NIR) was obtained from
Ocean Optics, FL, USA.
16. 2  2 Single-mode fiber optic couplers for 633 nm wavelength
were obtained from Newport Corporation, CA, USA.

3 Methods

3.1 Fabrication of 1. Typical semiconductor fabrication technologies [14, 15] with

LSPR-FO Nanoprobe lift-off method are utilized in the fabrication development for
Au nanostructures on fiber end face, as schematically illustrated
in Fig. 4. There are four main technological steps: (1) deposi-
tion of positive electron beam resist (ZEP520A, Zeon Chemi-
cals, Japan) on the fiber end face with uniform thickness
(Fig. 4a–c), (2) nano-patterning on the E-beam resist using
EBL method (Fig. 4d), (3) vacuum deposition of functional
metallic materials over the e-beam resist using cleanroom

Fig. 4 A schematic flow-chart displays the fabrication procedure for nanodisk arrays at the optical fiber end
facet. (a) optical fiber tip, (b) 2 nm Cr deposition, (c) resist deposition, (d) EBL process, (e) gold film deposition,
(f) lift-off process, (g) Cr etch to get the nanodisk arrays
 Localized Surface Plasmon Resonance (LSPR)-Coupled Fiber-Optic. . . 7

thermal evaporation (Fig. 4e), and (4) nano-pattern transfer

using standard liftoff method (Fig. 4f, g).
2. An optical fiber for single-mode wavelength of 633 nm is
employed in this work, which has a core diameter of 4 μm, a
cladding diameter of 125 μm, and a polymer buffer coating
diameter of 250 μm (Fig. 4a). Compared to standard optical
fiber operated at single-mode wavelength of 1310 nm, the
advantage of using this small core fiber is that it provides
more spectral stability in the wavelength ranges of
600–750 nm, in which locate the resonance peak of our fabri-
cated fiber tip LSPR sensors. Small core size also means higher
fabrication yield and less nanoparticles involved in the sensing,
thus requires smaller amount of target samples.
3. The preparation of optical fiber tip includes stripping off the
polymer buffer layer 4 cm from the end and cleaving the end
face with a fiber cleaver, followed by cleaning with acetone and
isopropyl alcohol (IPA) rinse for 5 min.
4. Two nanometers of Cr film is firstly deposited on the fiber end face
using the vacuum sputtering method [16] to provide a conductive
layer for the e-beam resist in the EBL process (Fig. 4b).
5. A simple and unique wet resist coating method called “dip and
vibration” technique has been developed in a nanofabrica-
tion lab to deposit e-beam resist on the optical fiber tip
(Fig. 4c), and its procedure is schematically represented in
Fig. 5. The optical fiber is dipped into the diluted e-beam resist
(ZEP520A diluted with ZEP thinner at a ratio of 1:3) for 10 s
(Fig. 5a). Then, it is removed from the resist and hoisted into a
vertically upward position using a Newport fiber clamp, with
~15 mm of fiber tip outside the clamp at the top (Fig. 5b, c).

Fig. 5 Illustrations of the procedure called “dip and vibration” technique, (a) dip in the diluted e-beam resist,
(b) after dip, (c) vibration, and (d) after bake
8 Jianjun Wei et al.

The fiber tip is then vibrated manually by pulling the fiber tip to
one side and then releasing to get rid of excessive resist by
means of cantilever beam free vibration. The vibration fre-
quency and strength is controlled by the length of fiber tip
outside the fiber clamp and the initial displacement of the fiber
tip. The thickness of resist on the optical fiber tip is dependent
on the ZEP dilution ratio and vibration strength. The iterative
method is used to optimize the vibration process. The initial
fiber tip displacement for the vibration is 2–5 mm, and the fiber
tip length is 15–25 mm. The ZEP dilution ratio used is from
20% to 40%, diluted in ZEP thinner. The resulted e-beam resist
layer thickness on the fiber tip is 100–200 nm, measured by
SEM observation. The fiber tip is held vertically upward and
baked in a 120 C oven for 60 min (Fig. 5d).
6. The EBL process based on the Nano Pattern Generation Sys-
tem (NPGS) and a field emission scanning electron microscope
(FESEM) is used to create nanodot array pattern on the e-beam
resist on the fiber end face (Fig. 4d). A Newport fiber clamp is
used to hold the fiber vertically on the translation stage in the
SEM chamber. The EBL voltage is 30 kV and the exposure
dose is 70 μC/cm2. The fiber tip is developed by dipping in
resist developer (ZEP N50) for 1 min. The fiber tip is then
rinsed in DI water and baked in the 120 C oven for 10 min for
dehydration.
7. An oxygen plasma descum procedure in a reactive ion etcher
(RIE, 25 watts power for 3 min) is used to remove the thin
residual layers of photoresist following the photoresist devel-
opment step (see Note 1).
8. The deposition of 40 nm Au overlay over the patterned area is
carried out by using the standard thermal evaporation coating
technique (Fig. 4e). The 2 nm Cr layer previously deposited as
a conductive layer for the EBL process is now served as an
adhesive layer for Au overlay. To lift off the e-beam resist, the
fiber tip is dipped in the ZEP remover for 10 min, followed by a
1-min ultrasonic bath to assist the lift-off process (Fig. 4f). The
fiber tip is rinsed in DI water and checked under an optical
microscope to make sure that the resist layer on the fiber end
face has been removed (see Note 2).
9. The fiber tip is dipped into the Cr remover solution for 30 s to
remove the Cr layer that is not covered by the Au nanoparticles
(Fig. 4g). The fiber tip is rinsed again in DI water and baked in
the 120  C oven for 10 min for dehydration. Figure 6 shows
the scanning electron micrographs of a gold nanodisks array on
an optical fiber tip end facet.
10. Finally, the Au nanodisk array on fiber end face is annealed at
530  C for 5 min and ready for next usage (see Note 3).
 Localized Surface Plasmon Resonance (LSPR)-Coupled Fiber-Optic. . . 9

Fig. 6 Images of (a) overview of the optical fiber tip end, (b) SEM images of gold nanodisk arrays on the optical
fiber facet after the lift-off process

Fig. 7 A schematic diagram illustrating optical setup for the fiber tip LSPR sensor based on reflection spectrum
measurement

3.2 Optical 1. In this optical setup (Fig. 7), a 2  1 single-mode optical fiber
Measurement coupler for 633 nm wavelength is used, which has three light
connection ports. In Fig. 7, port (a) and port (b) are two
connections on one side of optical fiber coupler, and port (c)
is connection on the opposite side. Port (a) is connected to the
tungsten halogen white light source; port (b) is connected to a
mini-spectrometer; and port (c) is connected to the LSPR-FO
probe using a fusion splicer. The white light (450–950 nm) is
launched from port (a), propagated to port (c), and reflected
from the end facet with the Au nanodisks arrays. The reflection
light propagated to port (b), where the spectrum of the
reflection light is measured by the mini-spectrometer, which
is connected to a computer for data acquisition and processing.
Figure 8 shows the photograph of the Optical setup for
the fiber tip LSPR sensor based on reflection spectrum
measurement.
2. During fiber optic LSPR detection, the light is launched and
guided to the LSPR fiber tip facet. The light wave propagates
along the core in the center of the optical fiber tip where the
10 Jianjun Wei et al.

Fig. 8 Photograph of the optical components (optical fiber, fiber coupler, mini-
spectrometer, light source) and setup for the fiber tip LSPR sensor. Computer
connection to the mini spectrometer via USB cable is not shown

nanodisk array is located. The guided white light interacts with

the Au nanodisk array and excites the localized surface plasmon
giving raise to enhanced scattering around the resonance wave-
length. The LSPR wavelength strongly depends on interface
refractive properties [16]. The light reflected from the LSPR
interface is guided in the fiber and propagated to the mini-
spectrometer, and the spectrum of the reflected light is
recorded as a function of wavelength. The detection of RI
change at the interface is accomplished by observing the pri-
mary resonance peak shift in the spectrum.
3. Before starting the experiment, the sensor tip must be cleaned
of any impurities or other contaminants. The sensor tip is
rinsed with ethanol to clean the tip. A baseline wavelength is
achieved if the sensor tip returns to this wavelength three times
after being washed in deionized water. If the tip spectra do not
return to this line, acetone and isopropyl alcohol (IPA) dip for
5 min may be used to further clean the tip of lingering
impurities.
4. In order to get the LSPR spectrum response due to Au nanos-
tructures on the fiber end facet, a reference spectrum and a dark
spectrum need to be recorded. The reference spectrum is
acquired without fiber LSPR probe on port (c), and the fiber
end face of port (c) is perpendicularly cleaved. The dark spec-
trum is obtained by turning off both the tungsten halogen light
source and room light. The measured reflection spectra (Mλ) of
the fiber tip sensor probe is obtained by the equation:
Mλ ¼ (Sλ  Dλ)/(Rλ  Dλ)  100 % ,
where Sλ is the sample intensity, Dλ is the dark intensity, and Rλ
is the reference intensity at wavelength λ (the intensity defines
as photon counts).
 Localized Surface Plasmon Resonance (LSPR)-Coupled Fiber-Optic. . . 11

3.3 LSPR-FO 1. The nanoprobe sensitivity is used to determine if the sensor tip
Nanoprobe Sensitivity in experimentation would be sensitive enough to determine
Characterization small wavelength shifts accurately.
2. The solvents used in determining the bulk RI sensitivity of the
LSPR tip are acetone, methanol, ethanol, isopropyl alcohol,
and water of different RIs.
3. The LSPR-FO nanoprobe is dipped in the various solvents, and
the spectra of the reflected light were recorded, respectively, as
shown in Fig. 9a (see Note 4).
4. To determine the peak wavelength in the LSPR reflection
spectrum, a Matlab program was created to fit the data to an
eighth-order polynomial function over a range of 80 nm [17],
centered at the wavelength of maximum reflection in the raw
data.
5. The calculated LSPR wavelength red shifted as the RI of the
solvent increased. A linear relationship between the LSPR peak
wavelength and the bulk RI of the medium is obtained (Fig. 9b
black dots), with the sensitivity of 226 nm/RIU (RIU, refrac-
tive index unit) used in this study.
6. The light intensity-based RI sensitivity is investigated as well
(Fig. 9b red diamonds). The return light intensities at the
LSPR peak wavelength are plotted against the bulk RI of the
media. The LSPR peak intensity is seen to be linearly propor-
tional to the RI, and the gradient of the line is 123 per RIU.
The return light intensity of a single-wavelength laser can be
used as a probing light for the refractive index change due to

Fig. 9 Sensitivity measurements of the LSPR-FO nanoprobe. (a) Measured reflectance spectra for the LSPR-FO
probe in various solvents of different refractive index. (b) Correlation of the LSPR peak wavelengths (left axis)
and LSPR peak intensity (right axis) with the refractive index (RI) of the solvents, the linear fit of the changes
vs. the RI gives the sensing sensitivity. In the inset equation, x and y represent the values of x and y axis. R2 is
the coefficient of determination in the linear fitting
12 Jianjun Wei et al.

biochemistry binding events, and thus eliminate the need for a

broadband light source and optical spectrometer for an LSPR
biosensing.

3.4 Functionalization 1. To achieve the immobilization of mAb at the LSPR-FO sensor

of the LSPR-FO surface, the first step is to incubate a cleaned gold nanodisck
Nanoprobe fiber tip in a mixture of 1 mM HS(CH2)10COOH and
HS(CH2)8OH with mole ratio 1:5 in absolute ethanol
solution for 1 h. A mixed self-assembled monolayer is formed
at the gold nanodisck surfaces.
2. The self-assembled monolayer (SAM) tethered with carboxylic
acid groups then is activated by incubation in a pH 7.0, 10 mM
phosphate buffer solution containing 0.5 mM of EDC/NHS,
respectively, for 1 h.
3. The activated SAM is rinsed with the distilled water and imme-
diately moved to a freshly prepared 10 mM PBS containing
10 μg/mL of the detector mAb for 1 h incubation.
4. The fiber probe is finally rinsed with the PBS and followed by
dipping in a 0.2 M glycine PBS solution for 10 min to deacti-
vate the remaining active sites at the SAM.

3.5 Detection of f- 1. During the experiments, PBS (containing 0.05% tween-20) is

PSA Biomarker used as a running buffer to help minimize the nonspecific
adsorption of f-PSA at the fiber tips.
2. In order to test binding of PSA to the anti-PSA mAb modified
surfaces, five tenfold-dilutions of a 5 ng/mL concentration of
f-PSA are prepared in PBS solution for the desired concentra-
tions (i.e., 5 ng/mL, 0.5 ng/mL, 0.05 ng/mL, 5 pg/mL,
0.5 pg/mL, 0.05 pg/mL). The sensor tip is placed in the
f-PSA PBS for 10 min in the sequence at the lowest concentra-
tion of 50 fg/mL to measure the reflectance spectrum.
3. After each measurement, the fiber tip is rinsed thoroughly with
DI water and dried in air. All spectra are obtained after the fiber
tip was cleaned and dried in air. Figure 10a shows the represen-
tative reflectance spectra of f-PSA detection sensing various
f-PSA concentrations from 0 to 0.5 ng/mL.
4. Figure 10b shows the dependence of the wavelength peak shift
on the f-PSA concentration. The peak shift for each point is
obtained by averaging three measurements. A wavelength shift
of twice the largest standard deviation (1.2 nm) is used to
determine LOD of the fiber LSPR sensors, which corresponds
to 100 fg/mL of f-PSA in PBS solution.
5. Control experiments have been designed and carried out to
evaluate the specificity/selectivity of the immunoassay using
the LSPR-FO devices. Specifically, the binding between the
detector mAb and bovine serum albumin (BSA), similar to
 Localized Surface Plasmon Resonance (LSPR)-Coupled Fiber-Optic. . . 13

Fig. 10 Detection of protein biomarker. (a) The representative reflectance spectra of the anti-PSA mAb
modified LSPR-FO nanoprobe in various concentrations of f-PSA (from 0 up to 0.5 ng/mL). Insert plot shows
spectra normalized to peak intensity to show the LSPR reflectance peak shifts. (b) Correlation of f-PSA
concentration to the reflectance light of resonance peak wavelength shift

human serum albumin, at concentrations of 5 mg/mL has

been evaluated. Additionally, to evaluate the nonspecific bind-
ing to the SAM, the LSPR-FO devices modified only with the
mix SAM of HSC10COOH and HSC8OH SAM without
antibody attached are tested by following the f-PSA detection
process, as shown in Fig. 10b (blue diamond markers).

4 Notes

1. The optional step after the oxygen descum is to dip the fiber tip
in the chrome remover solution for 60 s to remove the thin Cr
layer on the nanodisk area. This step will eliminate the Cr
adhesion layer under the Au nanodisk, avoiding the undesired
side effects of Cr layer under Au nanodisk in the LSPR sensing
process, such as spectrum broadening and dephasing. How-
ever, this step will weaken the adhesion strength between the
Au nanodisks and the fiber end glass substrate, and will reduce
the reusability of the fiber LSPR probe.
2. If the lift-off process is not complete, ZEP remover and ultra-
sonic bath may be used to further lift off the e-beam resist.
3. The fiber probe is inserted horizontally into the oven through
an access hole on the sidewall of a high-temperature oven. In
order to protect the polymer buffer on the fiber, only
10–15 mm of fiber probe needs to be inside the oven.
4. Note that there is a noisy region in all spectra from ~760 to
800 nm, which is caused by the transmission minimum for the
optical fiber coupler (single mode at 633 nm) in the same
wavelength range.
14 Jianjun Wei et al.

Reference
1. Mukundan H et al (2009) Optimizing a wave- 9. Leskova TA, Maradudin AA, Zierau W (2005)
guide-based sandwich immunoassay for tumor Surface plasmon polariton propagation near an
biomarkers: evaluating fluorescent labels and index step. Optics Commun 249(1–3):23–35
functional surfaces. Bioconjug Chem 20 10. Jung LS et al (1998) Quantitative interpreta-
(2):222–230 tion of the response of surface plasmon reso-
2. Liu X et al (2008) A one-step homogeneous nance sensors to adsorbed films. Langmuir 14
immunoassay for cancer biomarker detection (19):5636–5648
using gold nanoparticle probes coupled with 11. Mitsui K, Handa Y, Kajikawa K (2004) Optical
dynamic light scattering. J Am Chem Soc 130 fiber affinity biosensor based on localized sur-
(9):2780–2782 face plasmon resonance. Appl Phys Lett 85
3. Sekhar PK, Ramgir NS, Bhansali S (2008) (18):4231–4233
Metal-decorated silica nanowires: an active sur- 12. Chang T-C et al (2012) Using a fiber optic
face-enhanced raman substrate for cancer bio- particle plasmon resonance biosensor to deter-
marker detection. J Phys Chem C 112 mine kinetic constants of antigen–antibody
(6):1729–1734 binding reaction. Anal Chem 85(1):245–250
4. Mani V et al (2009) Ultrasensitive immunosen- 13. 1985. http://apps.who.int/iris/bitstream/
sor for cancer biomarker proteins using gold 10665/38405/1/WHO_TRS_725.pdf.
nanoparticle film electrodes and multienzyme- 14. Nishi Y, Doering R (2000) Handbook of semi-
particle amplification. ACS Nano 3 conductor manufacturing technology. CRC
(3):585–594 Press, Boca Raton, FL
5. Henne WA et al (2006) Detection of folate 15. May GS, Spanos CJ (2006) Fundamentals of
binding protein with enhanced sensitivity semiconductor manufacturing and process
using a functionalized quartz crystal microbal- control. John Wiley & Sons, New York, NY
ance sensor. Anal Chem 78(14):4880–4884
16. Obando LA, Booksh KS (1999) Tuning
6. Baker GA, Desikan R, Thundat T (2008) dynamic range and sensitivity of white-light,
Label-free sugar detection using phenylboro- multimode, fiber-optic surface plasmon reso-
nic acid-functionalized piezoresistive micro- nance sensors. Anal Chem 71(22):5116–5122
cantilevers. Anal Chem 80(13):4860–4865
17. Wu H-J et al (2012) Membrane-protein binding
7. Lee HJ, Nedelkov D, Corn RM (2006) Surface measured with solution-phase plasmonic nano-
plasmon resonance imaging measurements of cube sensors. Nat Methods 9(12):1189–1191
antibody arrays for the multiplexed detection
of low molecular weight protein biomarkers. 18. Willets KA, Van Duyne RP (2007) Localized
Anal Chem 78(18):6504–6510 surface plasmon resonance spectroscopy and
sensing. Annu Rev Phys Chem 58(1):267–297
8. Raether H (1988) Surface plasmons on
smooth and rough surfaces and on gratings. 19. Zhao J et al (2006) Localized surface plasmon
Springer tracts in modern physics. Springer, resonance biosensors. Nanomedicine 1
Berlin (2):219–228
 Chapter 2

Ultra-Sensitive Surface Plasmon Resonance Detection

by Colocalized 3D Plasmonic Nanogap Arrays
Wonju Lee, Taehwang Son, Changhun Lee, Yongjin Oh,
and Donghyun Kim

Abstract
Ultra-sensitive detection based on surface plasmon resonance (SPR) was investigated using 3D nanogap
arrays for colocalization of target molecular distribution and localized plasmon wave in the near-field.
Colocalization was performed by oblique deposition of a dielectric mask layer to create nanogap at the side
of circular and triangular nanoaperture, where fields localized by surface plasmon localization coincide with
the spatial distribution of target molecular interactions. The feasibility of ultra-sensitivity was experimen-
tally verified by measuring DNA hybridization. Triangular nanopattern provided an optimum to achieve
highly amplified angular shifts and led to enhanced detection sensitivity on the order of 1 fg/mm2 in terms
of molecular binding capacity. We confirmed improvement of SPR sensitivity by three orders of magnitude,
compared with conventional SPR sensors, using 3D plasmonic nanogap arrays.

Key words Surface plasmon resonance, Localized surface plasmon resonance, Surface plasmon reso-
nance detection, Nanogap arrays, Nanoapertures, Colocalization, DNA hybridization

1 Introduction

In recent years, diverse optical techniques have attracted tremen-

dous interests for ultra-sensitive real-time detection of various
phenomena involving biomolecular interactions. Most of these
techniques have been based on fluorescence. Fluorescence-based
sensing, however, suffers from fluorescence interference or chemo-
toxicity issues. In contrast, surface plasmon resonance (SPR) bio-
sensing has been widely investigated as a representative label-free
technique, by which specific molecular interactions can be moni-
tored in real time and kinetic characteristics related to molecular
binding are conveniently measured on a quantitative basis.
Surface plasmon (SP) is a collective logitudinal oscillation of
electrons existing at the metal/dielectric interface. The oscillation
of electron can be coupled with a TM-polarized incident light.

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_2,
© Springer Science+Business Media LLC 2017

15
16 Wonju Lee et al.

By tuning angle of incidence or incident wavelength, the momen-

tum matching condition between incident light and SP can be
satisfied. At resonance, due to energy transfer from incident light
to a longitudinal surface wave, a narrow dip in the reflection char-
acteristics is observed with respect to wavelength or angle of inci-
dence. A small change of dielectric medium refractive index
induced from a biochemical interaction affects momentum match-
ing condition, followed by the shift of resonance dip in the reflec-
tion characteristics [1].
Despite broad uses of SPR sensors, conventional SPR sensors
have relatively poor detection limit on the order of 1 pg/mm2 in
binding capacity [2]. As an optical sensing technique other than
SPR, metal nanoparticle-based surface-enhanced Raman scattering
and microresonator-based whispering gallery modes have emerged
for ultra-sensitive label-free detection [3, 4].
In this chapter, we describe self-aligned colocalization using
three-dimensional plasmonic nanogap arrays for ultra-sensitive SPR
biosensors [5]. Note that colocalization indicates the spatial overlap
between the area of localized electromagnetic field and molecular
interaction. Colocalization is preferred in terms of sensitivity
enhancement per molecule, because much smaller number of mole-
cules are involved at the interaction. In previous studies, use of
plasmonic nanostructures was investigated to localize SP and eva-
nescent near-field to enhance detection sensitivity [6, 7]. Random
and periodic nanostructure was also employed for specific and
nonspecific detection based on enhanced surface plasmon reso-
nance [8, 9]. It was also shown that when biomolecules are spatially
aligned for colocalization with the localized field that is defined by
2D linear nanopattern arrays, SPR signals can be effectively ampli-
fied, which leads to efficient improvement of detection sensitivity
[10–12]. Moreover, theoretical investigation of nonspecific, non-
colocalized, and colocalized detection models was performed
based on silver nanoislands, which confirmed possibility of further
amplification of optical signature [13]. Meanwhile, SPR sensor
characteristics using grapheme-related materials were also studied
[14, 15].
Here, we describe use of 3D plasmonic nanogap arrays, illu-
strated in a schematic diagram of Fig. 1, for colocalized detection of
bio-interactions of target molecules to achieve even more dramatic
enhancement of sensitivity in SPR biosensors. In particular, circular
and triangular nanoholes were developed lithographically and 3D
nanogaps were then formed for colocalization by the shadowing of
obliquely evaporated dielectric mask layer on the metallic nanos-
tructures. This enables target molecules to directly access the
underlying metal film. Oblique evaporation was previously used
to fabricate molecular electronic devices based on nanoscale gap
structures [16]. Since localized near-field is formed at the ridge of
the nanostructure, localized fields and nanogap where target
 Ultra-Sensitive Surface Plasmon Resonance Detection. . . 17

Fig. 1 (a) Schematic illustration of 3D triangular nanogap aperture arrays that

may be produced by oblique evaporation for colocalized biomolecular
interaction. (b) Lateral profile across the solid line (A–B) in (a)

molecules can bind are self-aligned for colocalization. 3D plasmo-

nic nanostructures induce colocalization area to be much smaller
than what 2D linear grating patterns may produce. Therefore,
detection of much weaker interactions and/or those involve a
smaller number of molecules would be feasible, which allows
extreme sensitivity enhancement in SPR sensing.

2 Materials

2.1 Nanogap 1. A glass substrate (SF10) with a 2-nm-thick chromium adhesion

Fabrication layer and a 40-nm-thick gold film.
2. Electron beam (VEGA II LSH; TESCAN, Brno, Czech).
3. Scanning electron microscope (Elphy Quantum; Raith, Dort-
mund, Germany).
4. A polymethylmethacrylate (PMMA) photoresist (AR-N
7520.18; ALLRESIST, Strausberg, Germany).
5. Spin-coater (ACE-200; DONG AH Trade Corp, Seoul, Korea).
6. A dielectric mask layer using ITO or SiO2.
7. Developer (AR 300-47; ALLRESIST, Strausberg, Germany).
8. Remover (AR 300-70; ALLRESIST, Strausberg, Germany).
18 Wonju Lee et al.

Fig. 2 (a) Schematics of optical setup for a SPR sensor based on colocalization
using plasmonic nanogap arrays and (b) photograph of the experimental setup

2.2 Optical Setup A schematic illustration of the optical setup for colocalized SPR
detection is presented in Fig. 2 (see Note 1).
1. He-Ne laser (36 mW, λ ¼ 632.8 nm, Nominal beam diameter:
650 μm; Melles-Griot, Carlsbad, CA).
2. Glan-Thompson Linear Polarizer (Thorlabs, Inc., Newton, NJ).
3. Chopper and controller (SR540; Stanford Research Systems,
Sunnyvale, CA).
4. Two concentric motorized rotation stages (URS75PP and
ESP330; Newport, Irvine, CA) (see Note 2).
5. Low-noise lock-in amplifier (Model SR830; Stanford Research
Systems, Sunnyvale, CA).
6. A p-i-n photodiode (818-UV; Newport, Irvine, CA).
7. Lab ViewLabVIEW (National Instruments, Austin, TX).
8. Index-matching gel (n ¼ 1.725; Cargille Laboratories, Cedar
Grove, NJ).
 Ultra-Sensitive Surface Plasmon Resonance Detection. . . 19

2.3 DNA Preparation 1. HPLC purified 24-mer sequence length capture probe and
target oligonucleotides (IDT, Coralville, IA).
2. The sequence of single-stranded probe DNA (p-DNA) was 50 -
TTT TTT CGG TAT GCA TGC CAT GGC-3 modified with
thiol at 50 .
3. The sequence of single-stranded target DNA (t-DNA) was 50 -
GCC ATG GCA TGC ATA CCG AAA AAA-30 .
4. Plasma cleaner (Harrick Scientific Products, Pleasantville, NY).
5. Micropump (KD Scientific, Holliston, MA).
6. Acetone (1009-4404; DAEJUNG CHEMICAL & METALS
CO., Siheung, Korea).
7. Phosphate Buffered Saline (PBS) buffer (pH ¼ 7.4, BP3991;
Fisher Scientific, Pittsburgh, PA).

3 Methods

3.1 Nanogap To make linear nanograting-based 2D nanogap arrays, the overall

Fabrication fabrication steps are explained below.
1. A 40-nm-thick gold film was evaporated on an SF10 glass sub-
strate with a 2-nm-thick chromium adhesion layer.
2. Nanograting array patterns were defined by electron beam
(e-beam) lithography using PMMA photoresist that was spin-
coated on the gold film.
3. Metallic nanograting arrays were transferred by a lift-off process
after gold evaporation (see Fig. 3a).
4. 2D nanogap arrays were created on nanograting arrays by
oblique evaporation of a dielectric mask layer using ITO or
SiO2 (see Fig. 3b).

Fig. 3 (a) SEM image of fabricated linear nanograting arrays and (b) magnified image of 2D nanogap arrays
20 Wonju Lee et al.

Fig. 4 Fabrication procedure to implement 3D plasmonic nanogap arrays

To create 3D nanogap arrays, the overall fabrication process is

presented in Fig. 4.
1. A 40-nm gold film was evaporated on an SF10 glass with a 2-nm
chromium adhesion layer.
2. Circular and triangular nanostructures with a 2-μm array period
and an aperture size of 600 nm were defined by electron beam
lithography.
3. 20-nm-thick nanohole arrays were formed after a lift-off process
that involves e-beam lithography, evaporation of gold, and
removal of polymer resist.
4. 5-nm-thick ITO layer was obliquely deposited with the deposi-
tion angle varied at 30 , 45 , and 60 to adjust the gap area
differently (see Note 3).

3.2 Optical Setup A He-Ne laser beam (36 mW, λ ¼ 632.8 nm, nominal beam
diameter: 650 μm, Melles-Griot, Carlsbad, CA) is TM-polarized
by linear polarizer and temporally modulated by optical chopper
(SR540; Stanford Research Systems, Sunnyvale, CA), which alters
the state of light on and off at a specific frequency. The chopping
frequency is used as the reference input of a lock-in amplifier
composed with usually set to 0.6 kHz lock-in amplifier and feed-
back system. Among the signals captured at a photodiode, noise
signal at frequency other than the chopping can be removed. Two
motorized rotation stages (URS75PP; Newport, Irvine, CA) are
employed; one is for rotating a prism on which a plasmonic nanos-
tructured sample is located, and the other is for rotating the pho-
todiode to keep laser alignment to photodiode. The angle of the
stage on which the nanostructure sample located can be corrected
by using the calibration constant and the other stage with the
 Ultra-Sensitive Surface Plasmon Resonance Detection. . . 21

photodiode follows the former stage. The nanostructured sample is

located on the SF10 prism with index matching gel (n ¼ 1.725;
Cargille Laboratories, Cedar Grove, NJ), and covered with flow cell
system head. Using one motor controller (ESP330, Newport), two
stages are controlled at the same time with a minimum angle
increment of 0.0002 , uni-directional repeatability of 0.001 , abso-
lute accuracy of 0.008 . A p-i-n photodiode (818-UV, Newport)
is used to detect refractance change by the angle scanning. The
value measured by the diode is used as input for the feedback
system of lock-in amplifier. Therefore, the signal from photodiode
is noise removed due to the phase-sensitive detection. The stage
and chopper controller are connected to PC and controlled by
LabVIEW, so whole system is fully automated.

3.3 Estimation of A 2D nanogap area GΛ,2D produced by linear nanograting per unit
Colocalization Areas grating length with the thickness dg can be calculated as
3.3.1 2D Nanogap G Λ, 2D ¼ d g ð1 þ tan θeva Þ ð1Þ
Production
where θeva is the oblique evaporation angle (see Fig. 1b) in
Eq. 1, the first term dg and the second term dgtan θeva come from
vertical nanograting edge and horizontal surface which are not
evaporated with dielectric, respectively. For example, when the
nanograting with dg ¼ 20 nm and θeva ¼ 30 produces a nanogap
with an area of 31.56 nm times the grating length.

3.3.2 Nanogap Reduction 3D nanoaperture arrays can reduce the colocalization area, in which
of the Colocalization Area target molecular binding overlaps with localized fields to amplify
optical signatures of the target interaction. A 3D nanogap area
defined by an equi-triangular aperture with a side length (L) is
calculated as
 
d g tan θeva
G Λ, 3D ¼ Ld g  ð1 þ tan θeva Þ þ pffiffiffi ð2  tan θeva Þ ð2Þ
3L
where the direction of oblique evaporation is parallel to a side
of a triangular pattern. The first term Ldg represents the area open
to the side, while the second term Ldg tan θeva is the area formed at
nanogap surface. The last term in Eq. 2 is the correction due to the
triangular shape. If L is much longer than dg, the above equation
can be simplified as

G Λ, 3D  Ld g  ð1 þ tan θeva Þ: ð3Þ

In the case of using circular nanoholes, the nanogap area with a
nanohole diameter (ϕ) is changed to
22 Wonju Lee et al.

 
πϕ 1
G Λ, 3D  d g  1 þ tan θeva : ð4Þ
2 2
Using 3D nanogap arrays, the number of molecular interac-
tions can be adjusted by changing the size to period ratio L/Λ.

3.4 Numerical In this section, the effectiveness of colocalization is evaluated from

Studies numerically calculated near-field distribution and the experimen-
tally determined nanogap area. Electromagnetic field distribution
at the nanogap was theoretically calculated based on rigorous cou-
pled wave analysis (RSoft DiffractMOD™). Numerical calculation
with DNA molecules can be done as below.
1. DNA molecules are assumed to be uniformly distributed on the
nanogap.
2. The thickenss of hybridized DNA layers is modeled to 8.16 nm,
which corresponds to the length of oligonucleotides.
3. The refractive indices of ssDNA and dsDNA are set to 1.449 and
1.517 [17].
As shown in Fig. 5, it is clear that fields are highly localized at
nanogap. The presence of side lobe peaks was also observed on the
opposite side of nanogaps, which cannot be neglected compared
with the main peaks of localized fields. The suppression ratio,
defined as the intensity ratio of the main peak to that of side lobe,
in circular and triangular nanopatterns was estimated to be
2.3–2.5 dB. In the case of colocalization, optical signature detected
at the main peak can be dominant because of spatial selectivity of

Fig. 5 Numerically calculated near-field intensity distribution |E |2 of (a) circular and (b) triangular nanopatterns
with ϕ, L ¼ 600 nm at Λ ¼ 2 μm. A 5-nm-thick ITO layer was assumed to be deposited on the nanopattern
with θeva ¼ 30
 Ultra-Sensitive Surface Plasmon Resonance Detection. . . 23

Fig. 6 SEM images: (a) circular and (b) triangular nanopattern arrays. (c) A magnified image of a triangular
nanogap pattern. A very narrow nanogap is clearly visible at the side ridge of a triangular nanoaperture.
Each inset of (a) and (b) shows the nanogap created on a circular and triangular aperture (scale bars: 2 μm for
(a) and (b), 200 nm for (c), insets of (a) and (b))

target binding in the nanogap. Nonspecific binding that may be

present is estimated to be insignificant.
SEM images of fabricated 3D nanogap arrays formed on circu-
lar and triangular patterns were presented in Fig. 6, in which the
existence of nanogap at the ridge of nanoaperture can be corrobo-
rated. When dg is fixed, the angle of oblique evaporation affects the
underlying nanogap size, which is directly related to the amount of
target molecules, i.e., DNA immobilization and eventually hybri-
dization. Each colocalization area of experimentally fabricated cir-
cular and triangular nanopatterns was estimated from the SEM
images in Fig. 6a, b to be 24,600  7000 and 19,300  4600 nm2
without considering the sidewall area of the apertures. This is in
good agreement with the numerically estimated gap area (see
Note 4).

(a) Before immobilization of p-DNAs, first, sample chips were

cleaned in acetone for 5 min and in ethanol.
(b) The sample chip surface was treated in plasma cleaner (Harrick
Scientific Products, Pleasantville, NY).
24 Wonju Lee et al.

3.5 Experimental (c) 1 ml Phosphate buffered saline (PBS) buffer (pH ¼ 7.4)
Measurements of DNA including 4 μM thiol hexane labeled probe ssDNA
Hybridization (HSssDNA) was injected to flow channel by a sylinge pump.
The sample chip was immersed in the buffer at 4  C for 4 h to
3.5.1 DNA Preparation immobilize the p-DNAs at surface (see Note 5). In order to
obtain SPR angle shift, the reference SPR angle was measured
and smoothed by 50th order Fourier filter.
(d) For DNA hybridization, PBS buffer including 4 μM t-DNAs
was injected on the p-DNA-immobilized sample chip surface
for 120 min by a fluidic channel of a micropump (KD Scien-
tific, Holliston, MA) at a volume flow rate of 150 μL/hr. The
sample chip was incubated to improve the efficiency of hybri-
dization through mild heating. For 2 h, the solution of t-
DNAs is warmed to 65  C and slowly cooled down to room
temperature. Finally, the SPR angle shift was obtained by
comparing SPR angles before and after the DNA
hybridization.
SPR shifts were measured using linear nanogratings and 3D
nanopattern arrays to analyze the sensor performance, as shown in
Figs. 7 and 8 for DNA hybridization. Two metrics were set to
evaluate the sensor performance: (1) sensitivity enhancement fac-
tor, i.e., resonance shift with respect to the shift obtained from
conventional SPR detection, and (2) molecular binding capacity
measured by colocalization in the nanogap.

3.5.2 Sensitivity First, resonance angle shifts were measured based on colocalization
Enhancement using plasmonic nanogap arrays, such as nanogratings, circular and
triangular nanopatterns, compared to non-colocalized SPR detec-
tion without nanogap structures. In the non-colocalized detection,
DNA hybridization occurs on the overall metal surface of nanos-
tructures. This leads to inefficient detection of target molecules,
majority of which may be distributed in the region with weak near-
fields.
Figure 7a shows raw data of measured SPR angle shifts θSPR in
various platforms. In case of non-colocalized detection, resonance
angle shifts made on nanostructures were detected to be slightly
larger than that of the thin film control. The larger resonance shift
is associated with the increase of the total area available for DNA
hybridization due to the presence of nanopatterns. For colocalized
detection, SPR angle shift tends to be larger with an increase of the
ITO-evaporation angle (θeva). A larger θeva enlarges the nanogap
surface area, which allows more probe-DNAs (p-DNAs) to immo-
bilize at the nanogap surface. Increased resonance shifts were
observed for colocalization by nanogratings because the nanogap
area by grating patterns is much larger than that of 3D nanogap
patterns and therefore induces more p-DNA targets to participate
in hybridization.
 Ultra-Sensitive Surface Plasmon Resonance Detection. . . 25

Fig. 7 (a) Measured SPR angle shifts as the evaporation angle (θeva) varies. (b)
Relative angle shifts (RAS) after normalization by the nanogap area. Reproduced
from Y. Oh et al. 2014 with permission from Elsevier

The angular shift made of 3D nanogap patterns may not be

clear in Fig. 7a. To consider the total area of nanogap to which
target molecules can bind for biomolecular interactions, a new
parameter was introduced as the resonance shift normalized by
the amount of molecular interactions, i.e., relative angle shift
(RAS) ¼ θSPR/nanogap area, where the amount of molecular
interactions is presumed to be proportionate to the nanogap area.
Mearsured RAS, corresponding to each case in Fig. 7a, is presented
in Fig. 7b. First, RAS3D increases by more than two orders based
on colocalization with 3D nanogap structures such as circular and
triangular patterns, compared to RAS of the non-colocalized con-
trol. In the case of using 2D grating-based nanogap arrays, RAS2D
is increased by one order of magnitude. Colocalization by 3D
triangular patterns produces the maximum RAS3D at θeva ¼ 60 ,
26 Wonju Lee et al.

Fig. 8 (a) Angular SPR shifts (ΔθSPR) measured as the array period (Λ) is varied.
3D nanogap arrays were created by triangular nanopatterns at θeva ¼ 60 and
the concentration of t-DNA was fixed at 4 μM. Inset shows a linear relation
between ΔθSPR and log(Nnp). (b) Angular shifts measured as the concentration of
t-DNAs (Ct-DNA) varied with the fixed p-DNA concentration at 4 μM. Inset also
shows a linear relation between ΔθSPR and log(Ct-DNA). Reproduced from Y. Oh et
al. 2014 with permission from Elsevier

which is 460 times larger than non-colocalized RAS measured at

thin film control. The result is associated with efficient localization
of SP at triangular nanopatterns as well as reduced nanogap area
that is optimized to the size of localized fields [18, 19]. Note that
the optimum evaporation angle for 3D nanogap arrays is not
directly proportional to the nanogap area. The evaporation angle
θeva may also affect the peak near-field intensity. The insets of Fig. 7.
 Ultra-Sensitive Surface Plasmon Resonance Detection. . . 27

show the resonance characteristics: A and B for non-colocalized

detection, C and D for nanogap-based colocalization. Compared to
the insets A, B, and D, 3D nanogap arrays presented in the inset C
keep the SPR curve from broadening due to lower plasmonic
damping and longer array period.

3.5.3 Molecular Binding To evaluate molecular binding capacity as the sensor performance,
Capacity Fig. 8 shows the isotherm characteristics measured in two ways
using the optimum triangular nanogap arrays in 3D at θeva ¼ 60 ,
which was determined to produce the maximum RAS as shown in
Fig. 7b. First, the array period was increased when the concentra-
tion of target DNA was fixed as 4 μM, which in effect decreases the
concentration of probe DNA. Secondly, the isotherm characteritic
was also measured by varying the concentration when the array
period and the concentration of probe DNA were fixed at 2 μm and
4 μM respectively.
To extract the detection sensitivity of colocalized SPR sensing
based on 3D nanogaps, resonance angle shifts (ΔθSPR) were
measured by varying the array period as shown in Fig. 8a, where
the amount of probe DNA is gradually reduced with an increase of
array period while the amount of target DNA remains constant.
The relation between ΔθSPR and the number of nanopatterns (Nnp)
was presented as
ΔθSPR ¼ 0:015 þ 0:030logN np : ð5Þ
In Eq. 5, ΔθSPR is proportional to log(Nnp), which means SPR
shifts are linearly increased when the amount of p-DNA is expo-
nentially increased. As Nnp is reduced, the detection and the nature
of resonance angle shifts change from being target-limited to
probe-limited with transition taking place at Nnp ¼ 3300~13,000
(Λ ¼ 5–10 μm), i.e., the limit of detection arises in the probe-
limited detection of DNA hybridization. If we assume surface
probe density to be 4  1012 molecules/cm2 and hybridization
efficiency 40%, a total effective nanogap area of 264,000 nm2 in a
650-μm-diameter beam spot measures approximately 4200 mole-
cules. This corresponds to the molecular binding capacity of
1.6 fg/mm2 based on the molecular weight of 7.4 kDa for a 24-
mer t-DNA. The results suggest that the detection sensitivity of 3D
nanogap-based SPR sensing, compared to that of conventional SPR
sensors, should be improved by more than three orders of
magnitude.
Figure 8b shows the measured angular shifts when the concen-
tration of target DNAs (Ct-DNA) was varied. In this case, p-DNA
concentration was fixed at 4 μM and other conditions remained the
same as in Fig. 8a. Surprisingly, the relation between ΔθSPR and Ct-
DNA is similar to what was measured from ΔθSPR and Nnp, which
implies that the amount of DNA hybridization is limited by
28 Wonju Lee et al.

p-DNAs (Nnp) or t-DNAs (Ct-DNA). On condition that a much

smaller number of DNAs participate in hybridization, 160% of
enhancement was measured based on the slope in the inset
ΔθSPR/log(Ct-DNA).

4 Notes

1. Overall procedure was automatically controlled by computer.

The size of a measurement chamber is roughly 10 mm
(L)  1.5 mm (W)  0.5 mm (D), and ambient temperature
was maintained at 19.6  0.2  C for measurement stability.
2. Two concentric motorized rotation stages were employed to
implement angle-scanning of θ/2θ measurement. The motor-
ized rotation stages were controlled with a minimum motion
increment of 0.0002 , absolute accuracy of 0.008 , uni-direc-
tional repeatability of 0.001 , and wobble at 20 μrad.
3. ITO provides chemical and optical stability to SPR sensors in
terms of material degradation [20]. During the evaporation,
chamber temperature was maintained at 200  C to enhance
adhesion by annealing.
4. The area of nanogap produced by a circular (triangular) nanoa-
perture with ϕ (L) ¼ 600 nm and dg ¼ 20 nm was numerically
calculated as listed in the table at 2-μm array period, i.e., a unit
area of 2  2 μm2. For 2D gratings, the array period was
400 nm. Area unit is nm2.

θeva 2D grating 3D circular aperture 3D triangular aperture

30 320,000 24,000 19,000

45 400,000 28,000 24,000

60 550,000 35,000 33,000

5. The estimated surface density of immobilized p-DNA is

4.0  1012 molecules/cm2 [21].

References
1. Kim D (2005) Effect of the azimuthal orienta- 3. Dasary SSR, Singh AK, Senapati D, Yu H, Ray
tion on the performance of grating-coupled PC (2009) Gold nanoparticle based label-free
surface-plasmon resonance biosensors. Appl SERS probe for ultrasensitive and selective
Optics 44:3218–3223 detection of trinitrotoluene. J Am Chem Soc
2. Campbell CT, Kim G (2007) SPR microscopy 131:13806–13812
and its application to high-throughput analyses 4. Vollmer F, Arnold S (2008) Whispering-gal-
of biomolecular binding events and their kinet- lery-mode biosensing: label-free detection
ics. Biomaterials 28:2380–2392 down to single molecules. Nat Methods
5:591–596
 Ultra-Sensitive Surface Plasmon Resonance Detection. . . 29

5. Oh Y, Lee W, Kim Y, Kim D (2014) Self- localized surface plasmon resonance biosen-
aligned colocalization of 3D plasmonic nano- sing. Opt Express 22:28412–28426
gap arrays for ultra-sensitive surface plasmon 14. Ryu Y, Moon S, Oh Y, Kim Y, Lee T, Kim DH,
resonance detection. Biosens Bioelectron Kim D (2014) Effect of coupled graphene
51:401–407 oxide on the sensitivity of surface plasmon res-
6. Byun KM, Kim S, Kim D (2005) Design study onance detection. Appl Optics 53:1419–1426
of highly sensitive nanowire-enhanced surface 15. Chung K, Rani A, Lee J-E, Kim JE, Kim Y,
plasmon resonance biosensors using rigorous Yang H, Kim SO, Kim D, Kim DH (2015) A
coupled wave analysis. Opt Express systematic study on the sensitivity enhance-
13:3737–3742 ment in graphene plasmonics sensors based
7. Kim K, Yoon SJ, Kim D (2006) Nanowire- on layer-by-layer self-assembled graphene
based enhancement of localized surface plas- oxide multilayers and their reduced anologues.
mon resonance for highly sensitive detection: ACS Appl Mater Interfaces 7:144–151
a theoretical study. Opt Express 16. Kubatkin S, Danilov A, Hjort M, Cornil J,
14:12419–12431 Brédas J-L, Stuhr-Hansen N, Hedegård P,
8. Moon S, Kim Y, Oh Y, Lee H, Kim HC, Lee K, Bjørnholm T (2003) Single-electron transistor
Kim D (2012) Grating-based surface plasmon of a single organic molecule with access to
resonance detection of core-shell nanoparticle several redox states. Nature 425:698–701
mediated DNA hybridization. Biosens Bioelec- 17. Elhadj S, Singh G, Saraf RF (2004) Optical
tron 32:141–147 properties of an immobilized DNA monolayer
9. Yu H, Kim K, Ma K, Lee W, Choi JW, Yun CO, from 255 to 700 nm. Langmuir
Kim D (2013) Enhanced detection of virus 20:5539–5543
particles by nanoisland-based localized surface 18. Fromm DP, Sundaramurthy A, Schuck PJ,
plasmon resonance. Biosens Bioelectron Kino G, Moerner WE (2004) Gap-dependent
41:249–255 optical coupling of single “bowtie” nanoanten-
10. Hoa XD, Kirk AG, Tabrizian M (2009) nas resonant in the visible. Nano Lett
Enhanced SPR response from patterned immo- 4:957–961
bilization of surface bioreceptors on nano-grat- 19. Haes AJ, Zou S, Schatz GC, Van Duyne RP
ings. Biosens Bioelectron 24:3043–3048 (2004) Nanoscale optical biosensor: short
11. Ma K, Kim DJ, Kim K, Moon S, Kim D (2010) range distance dependence of the localized sur-
Target-localized nanograting-based surface face plasmon resonance of noble metal nano-
plasmon resonance detection toward label-free particles. J Phys Chem B 108:6961–6968
molecular biosensing. IEEE J Sel Topics Quan- 20. Szunerits S, Castel X, Boukherroub R (2008)
tum Electron 16:1004–1014 Surface plasmon resonance investigation of sil-
12. Kim Y, Chung K, Lee W, Kim DH, Kim D ver and gold rilms coated with thin indium tin
(2012) Nanogap-based dielectric-specific colo- oxide layers: Influence on stability and sensitiv-
calization for highly sensitive surface plasmon ity. J Phys Chem C 112:15813–15817
resonance detection of biotin-streptavidin 21. Steel AB, Herne TM, Tarlov MJ (1998) Elec-
interaction. Appl Phys Lett 101:233701 trochemical quantitation of DNA immobilized
13. Yang H, Lee W, Hwang T, Kim D (2014) on gold. Anal Chem 70:4670–4677
Probabilistic evaluation of surface-enhanced
 Chapter 3

Two-Dimensional Surface Plasmon Resonance Imaging

System for Cellular Analysis
Tanveer Ahmad Mir and Hiroaki Shinohara

Abstract
Optical biosensors based on surface plasmon resonance (SPR) phenomenon have received a great deal of
attention in cellular analysis applications. Sensitive and high-resolution SPR imaging (SPRi) platforms are
very useful for real-time monitoring and measurement of individual cell responses to various exogenous
substances. In cellular analysis, mainstream SPR-based sensors have potential for investigations of cell
responses under ambient conditions. Evaluations that account only for the average response of cell
monolayers mask the understanding of precise cell-molecular interactions or intracellular reactions at the
level of individual cells. SPR/SPRi technology has attracted a great deal of attention for detecting the
response of cell monolayers to various substances cultivated on the gold sensor chip. To unleash the full
strength of SPRi technology in complex cell bio-systems, the applied SPR imaging system needs to be
sufficiently effective to allow evaluation of a compound’s potency, specificity, selectivity, toxicity, and
effectiveness at the level of the individual cell. In our studies, we explore the utility of high-resolution
2D-SPR imaging for real-time monitoring of intracellular translocation of protein kinase C (PKC), and
detection of neuronal differentiation in live cells at the level of individual cells. The PC12 cell line, which is
one of the most commonly used neuronal precursor cell lines for research on neuronal differentiation, was
chosen as a nerve cell model. Two dimensional SPR (2D-SPR) signals/images are successfully generated.
We have found that cells treated with the differentiation factor nerve growth factor (NGF) showed a
remarkable enhancement of SPR response to stimulation by muscarine, a nonselective agonist of the
muscarinic acetylcholine receptor.

Key words Nerve growth factor, Surface plasmon resonance, Surface plasmon resonance imaging,
Muscarine, Protein kinase C

1 Introduction

The ability to assess single-cell activity under physiological condi-

tions in a label-free format would be highly beneficial for evaluating
cellular and molecular binding dynamics. Conventional approaches
to evaluating single-cell activity with high precision and temporal
resolution are limited. The majority of currently available methods
for studies of single cells present difficulties such as the need for
fluorescent tagging and washing steps prior to readout. Fluorescent

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_3,
© Springer Science+Business Media LLC 2017

31
32 Tanveer Ahmad Mir and Hiroaki Shinohara

labeling may also alter membrane receptors and binding sites,

thereby compromising measurements of in vivo interactions. By
contrast, label-free methods directly provide additional information
due to the ability to evaluate the intrinsic biological and physical
characteristics of the individual cell.
SPRi biosensors are composed of three main components:
optics (lasers or light emitting diodes) that illuminate a gold
metal film, a high refractive index glass prism, and a charge-coupled
device (CCD) camera detector that records changes in reflectance
and captures images [1]. In general, SPR biosensors are based on
surface evanescent waves that propagate parallel to a metal surface.
The waves are generated under attenuated total reflection condi-
tions when energy from a photon of p-polarized light impinges a
thin Au metal sensor chip. Surface plasmons result from coupling of
oscillating modes of free electron density at a specific incident
angle. The plasmons possess a maximum intensity at the sensor
surface and decay exponentially away from the metal/dielectric
interface. This, in turn, gives SPR a sensing depth of only few
hundred nanometers. SPR generates a dip in the reflection intensity
curve in the cell area close to the substrate [2].
SPR experiments can be performed on a variety of commer-
cially available platforms. Several reports have been published
implementing one-dimensional SPR system for cell stimulation.
In addition, conventional 2D-SPR imagers have been implemented
for living cells [3]. However, the resolution of the 2D imaging
systems is not sufficient for the observation of cellular responses
at the individual cell level. To address this drawback our laboratory
applied a high-resolution 2D-SPR imager for real-time and label-
free monitoring of mammalian cells [4, 5].
The 2D-SPRi technique is ideally suited for single-cell analysis,
as cell stimulation with exogenous molecules and their interactions
are monitored by local refractive index changes. This allows real-
time, noninvasive, and tag-free evaluation of cellular reactions such
as binding dynamics, adhesion, spreading behavior, and morpho-
logical changes [6–9].
Figure 1 shows a 2D-SPR apparatus of Kretchmann configura-
tion (A) and its schematic representation (B). The 2D-SPRi (2D-
SPR 04A, NTT-AT Co. Corp., Japan) apparatus is equipped with a
collimator (to parallelize incident light) and a CCD camera with
four kinds of lens (1
, 2
, 4
, and 7
) for image magnification.
The optical system consists of a high-refractive index SF6 glass
prism (RI ¼ 1.72) deposited with a thin (50 nm) gold layer. To
avoid external light interference, the measurement system is assem-
bled in the temperature controlled (air conditioned) instrumental
box. A light emitting diode (640 nm LED) producing the incident
light beam passes through a collimator lens to illuminate the back
side of the gold sensor chip through the coupling prism. The light
beam from a collimated source passes through a polarizer filter and
 SPR Imaging for Cellular Analysis 33

b Cultured PC12 cell

Au film
(50 nm)
Glass
substrate

Reflection
P-polarizer Prism Prism light
Incident
light q
LED
Incident
angle
Collimator Magnification CCD
lens unit camera
(1,2,4,7X)

Fig. 1 (a) Shows a 2D-SPRi apparatus of Kretchmann configuration. (b) Schematic illustration of the used 2D-
SPR system equipped with collimator and a CCD camera with four kinds of lens (1
, 2
, 4
, and 7
) for
image magnification. A light emitting diode (LED) produces the incident light beam (670 or 770 nm) after
passing through a collimator lens that illuminates the back side of the sensor chip through the coupling prism.
The light beam from a collimated source passes through a polarizer filter and impinges on the sensor chip
containing cells at a specific angle near the SPR angle. The reflected light is passed through a narrow band-
pass filter, a 7
microscope objective collects the reflected light and forms reflectivity images that are
detected recorded by the digital CCD camera at a fixed angle
34 Tanveer Ahmad Mir and Hiroaki Shinohara

impinges on the sensor chip containing cells at a specific angle near

the SPR angle. The reflected light passes through a narrow band-
pass filter and a magnification objective lens (1
, 2
, 4
, or 7
)
collects the reflected light and forms reflectivity images that are
captured and stored by the cooled CCD camera. 2D-SPR recording
is based on the reflection intensity modulation of incident p-
polarized light beam at a given angle with the spatial capabilities
of imaging. The polarizing filter allows the recordings of both p-
polarized and s-polarized light, but s-polarized light is used only as
a reference signal to eliminate artifacts and to improve the image
contrast.
For cell-stimulation and intracellular signal analysis by 2D-
SPRi, the gold sensor chip surface is modified with poly-L-lysine
initially and is incubated overnight in a humid environment, and
PC12 cell suspension is inoculated onto the same sensor chip to
produce the mammalian cell sensor sample. The gold sensor slide
on which the PC12 cells are cultured is placed on the top of the
prism, and an index matching fluid is used between the prism and
the SPR chip to eliminate unwanted reflection. During measure-
ment, CCD camera visualizes whole sensor surface, and records the
reflected light images (SPR images) from the entire observable area.
Sequential changes in refractive index in each CCD pixel
correspond to the region that is measured (e.g., regions 1, 2, 3,
etc.). The interaction between ligands and cell membranes on the
SPR chip influences the index of refraction, thereby causing a shift
in reflectivity of incident light thus yielding a profile of refractive
indexes.
2D-SPRi system has the ability to observe both angle shifts in
the reflectivity curve, as well as to measure time-dependent cellular
interaction dynamics. In angle shift measurements, when cells are
cultivated onto the sensor chip, the SPR response signal is observed
as angle shifts from lower to higher angle, as a result of a change in
the index of refraction at the sensor surface. For time course
measurements the light reflectivity is measured at an optimum
angle close to the plasmon angle. The overall SPR response is
measured as a change in the light intensity reflected from the sensor
chip surface, and SPR images of cells are captured by the CCD
camera (Fig. 1). For data analysis, a reference image is collected
from the area without cells. Single-cell investigation on sensor chip
is done by simply controlling the cell density, and by adjusting the
high-resolution magnification lens unit. In our studies, the 7

magnification lens is adjusted for monitoring the activity of indi-
vidual cells. To analyze reflection intensity response of various cell
regions with 2D-SPRi at single cell level, from the total cells in the
observation area, a random selection for region of interest is made
and the reflection intensity percentage of the selected cells is aver-
aged. The overall reflection intensity change response is determined
as the percentage of reflected light intensity of P-polarized light by
 SPR Imaging for Cellular Analysis 35

subtracting cell region SPR response recorded before (reference)

from a cell region response recorded after exposing to stimuli.
Additionally, a threshold of reflection intensity is used to decide
whether the cell has responded or not.
The key to applying 2D-SPR imaging system for the study of
cellular analysis is the development of a sensitive, nondestructive,
label-free assay of nearly all living cells and tissues. Developing a
real-time cell-based sensor system seemed most valuable to cell
biologists to provide a technical means to study the dynamics of
intracellular signaling and neuronal differentiation rather than just
snapshots.
In this chapter, the utility of a 2D-SPR imager for studying
mammalian cells at individual cell level is demonstrated with two
biological systems: the influence of exogenous substances on intra-
cellular signal transduction response in normal PC12 cells, and the
effect of exogenous substances to observe the intensity modulation
of intracellular signaling as functional readout for neuronal differ-
entiation associated with intracellular PKC translocation.

2 Materials

1. PC12 cell line (cell no. RCB0009) derived from rat adrenal
gland is obtained from RIKEN BioResource Center.
2. Dulbecco’s Modified Eagle Medium (DMEM) is obtained
from GIBCO (Cat. No. 31600-034).
3. Horse serum is obtained from GIBCO (Cat. No. 16050-122).
4. Fetal bovine serum (FBS) is obtained from (ICN Biomedicals,
Cat. No. F4135).
5. Penicillin/streptomycin is obtained from GIBCO (Cat. No.
15070-063).
6. 0.05% Trypsin-EDTA is obtained from GIBCO (Cat. No.
25200).
7. NaCl is obtained from Wako Pure Chemicals Industries (Cat.
No. 191-01665).
8. NaOH is obtained from Kanto Chemical (Cat. No. 37184-00).
9. KCl is obtained from Wako Pure Chemicals Industries (Cat.
No. 163-03545).
10. CaCl2 2H2O is obtained from Wako Pure Chemicals Industries
(Cat. No. 031-00435).
11. NaHCO3 is obtained from Wako Pure Chemicals Industries
(Cat. No. 191-01305).
12. Glucose is obtained from Wako Pure Chemicals Industries
(Cat. No. 041-00595).
36 Tanveer Ahmad Mir and Hiroaki Shinohara

13. HEPES is obtained from Nacalai Tesque (Cat. No. 17546-34).

14. PBS without calcium and magnesium is obtained from Wako
Pure Chemicals Industries (Cat. No. 168-19323).
15. Dimethyl sulfoxide (DMSO) is obtained from Sigma (Cat. No.
D2650).
16. Nerve growth factor is obtained from Sigma (Cat. No.
N0513).
17. T-25 tissue culture flasks are obtained from IWAKI (Cat. No.
3103-025).
18. Hanks’ balanced salts (HBBS) is obtained from Sigma (Cat.
No. H6136).
19. 24-well plate is obtained from Sigma (product number:
Z707791).
20. Hemicytometer is obtained from Sigma (product number:
Z359629).
21. Bottle top filter is obtained from Corning® bottle-top vacuum
filters-(CLS430049).
22. flexiPERM (11
7
10 mm) is purchased from Greiner Bio
One (Cat. No 90034067.
23. 2D-SPR apparatus (2D-SPR 04A) coupled with a collimator,
four magnification lenses (1
, 2
, 4
, and 7
), and a cooled
charge-coupled device (CCD) camera is provided by NTT
Advanced Technology (NTT-AT.Co.Corp) Japan.
24. Light emitting diode (640 nm LED) provided by NTT
Advanced Technology (NTT-AT.Co.Corp) Japan, with 2D-
SPR 04A apparatus.
25. CCD camera provided by NTT Advanced Technology (NTT-
AT.Co.Corp) Japan, attached with 2D-SPR 04A apparatus.
26. 2D-SPR analysis software provided by NTT Advanced Tech-
nology (NTT-AT. Co. Corp) Japan, assembled on the com-
puter monitor connected with 2D-SPR 04A apparatus.
27. 50-nm gold layer deposited high-refractive index glass (SF6
chips, 18
17 mm) is obtained from BAS, Japan.
28. Index-matching fluid (n ¼ 1.72) obtained from Cargille
Laboratories.
29. Sterile bench installed in the clean culture room.
30. CO2 incubator installed in the clean culture room.
31. Inverted microscope (Olympus, cat. no. IX70) equipped with
appropriate filter and magnifier (Nikon) installed in the clean
culture room.
32. (+)-Muscarine chloride is obtained from Sigma (Cat. No.
M6532-5MG).
 SPR Imaging for Cellular Analysis 37

33. Phorbol-12-myristate-13-acetate (PMA) is obtained from

Sigma (Cat. No. P8139-1MG).
34. Staurosporine from Streptomyces sp. is obtained from Sigma
(Cat. No. S4400).

3 Methods

All the solution preparations and cell cultures were performed

under sterile conditions.

3.1 Cell Culture and 1. A 50-nm gold thin film-coated high-refractive index glass
Preparation of SPR (SF6) chip adhered with a square well of flexiPERM (11
7
Sensor Surface
10 mm) is used (see Note 1).
2. Prior to cell culture, SPR sensor slides are put into small cell
culture dishes and are exposed to ultraviolet (UV) sterilization
for 10 min.
3. Cryopreserved nerve model cells (PC12 cells) are harvested
and then transferred into a 15 ml centrifuge tube containing
5 ml DMEM medium supplemented with 10% horse serum
(HS) and 5% fetal bovine serum (FBS) (see Note 2).
4. T-25 flasks are filled with 5 ml of cell suspension and incubated
in a cell culture incubator for 3–4 days (see Note 3).
5. After a given incubation period, T-25 flask is taken into a sterile
clean bench. Old medium is removed and cells are thoroughly
washed with sterile Hanks’ balanced salts (HBBS) solution.
6. To detach the cells from T-25 flask surface, 1 ml of 0.05%
trypsin-EDTA is added and agitated gently for 1 min. Finally,
4 ml of medium (DMEM, 10% HS, 5% FBS) is added, the total
content is transferred into a 15 ml tube to stop the function of
trypsin-EDTA and centrifuged.
7. After centrifugation, supernatant is discarded and the pellet
aggregate is mechanically separated into single cells by gentle
trituration of the suspension using pipette.
8. To prepare a SPR sensor surface of cellular analysis, a homoge-
neous solution is diluted in culture medium so that the cell
suspension becomes 4.0
105 cells/ml. A total volume of
300 μl of medium containing cell suspension is poured onto
the sensor chip to prepare one sensor sample (see Note 4).
9. For cellular analysis using NGF-treated neuronal differentiated
PC12 cells, a homogeneous solution is diluted in culture
medium so that the cell suspension becomes 3.0
105. A total
volume of 300 μl of medium containing cell suspension is
poured onto the sensor chip to prepare one sample (see Note 5).
38 Tanveer Ahmad Mir and Hiroaki Shinohara

10. In neuronal differentiation study, sensor chip surface is first

modified with poly-L-lysine, PC12 cells are cultured onto the
coated sensor surface. After 24 h, normal DMEM medium is
exchanged with new DMEM containing 80 ng/ml NGF and
only 1% HS.

3.2 Preparation of 1. First, 1 g Dulbecco’s Modified Eagle Medium (DMEM) pow-

Working Solutions der is diluted with 84 ml sterile Millipore water and then added
0.37 g NaHCO3 while stirring gently. Then, 1 ml penicillin–-
streptomycin (1%), 10 ml horse serum (10% vol/vol), and 5 ml
FBS (5% vol/vol) are added using 0.22-mm top bottle filter (see
Note 6).
2. Hanks’ balanced salts (HBBS) is prepared by dissolving original
HBBS powder pack in cell culture-grade distilled, deionized
water at room temperature. The prepared solution is sterilized
using a 0.22-μm sterilizing filter, the solution is dispensed into
sterile containers, and stored from 2  C to 20  C.
3. Just before use, phorbol-12-myristate-13-acetate (PMA) is dis-
solved as 1 and 0.2 mg/ml solutions in Dimethyl sulfoxide
(DMSO) (see Note 7).
4. Just before use, staurosporine is dissolved as 1 mg/ml solutions
in Dimethyl sulfoxide (DMSO) (see Note 8).
5. Just before use, (+)-Muscarine chloride is dissolved in water (see
Note 9).
6. Just before use, KCl is dissolved in water.
7. Nerve growth factor (NGF) is dissolved in HBBS solution at
50–200 ng/ml and stored at 20  C.

3.3 2D-SPR Imaging 2D-SPR experiments are performed using a high-resolution 2D-
Experimental Setup SPR apparatus, as shown schematically in Fig. 1. A 50-nm gold-
coated high-refractive index glass chip, assembled with a square
wall-type reusable cell culture chamber system (flexiPERM) made
of silicone that can easily stick on the sensor chip due to its naturally
adhesive underside to all smooth surfaces without glue creating
growth chambers. PC12 cells are maintained on the sensor chips
for over 24 h in a humidified atmosphere containing 5% CO2 at
37  C. After taking out sensor chip from the incubator, the growth
medium is discarded, and cells are rinsed twice with Hanks’ solu-
tion (pH 7.4, 37  C). After washing, Hanks’ solution (240 μl) is
added to the sensor chip. To eliminate the unwanted reflection and
recording of reflection image of P-polarized parallel incident light,
the sensor chip is adjusted on the top of the prism using index
matching fluid. Thereafter, a volume of 30 μl of stimulant solution
(KCl and PMA) or reference solution for negative control (Hanks’
solution) is carefully injected onto the sensor slides using a manual
pipette. During the experiment, SPR curves (reflection intensity
 SPR Imaging for Cellular Analysis 39

changes) are monitored in real time and SPR images are recorded.
Offline data is analyzed on a computer connected with the mea-
surement system using 2D-SPR analysis software (2D-SPR analysis
software provided by NTT-AT Company with the 2D-SPR appara-
tus). For inhibitory response, PC12 cells are pre-incubated with
100 nM staurosporine (a PKC inhibitor) solution for 30 min prior
to PMA treatment.
In order to evaluate reflection intensity change response of
various cell locations with 2D-SPRI at the individual cell level,
from the total cell population on the chip observation region a
random sorting of cell regions of interest is done and the reflection
intensity percentage of the selected cell areas is averaged. The time-
dependent reflection intensity change responses are determined as
the percentage of reflected light intensity of P-polarized light to
that of s-polarized light at a single incident angle as a function of
time by simply subtracting cell region SPR responses recorded
before (negative control) and after exposing to stimuli. To deter-
mine a resonance angle, SPR angles of the sensor surface with cells
and the bare surface are measured by changing the incident angle
from 49 to 57 . For overall data analysis, a threshold of reflection
intensity is used to decide whether the cell has responded to stimuli
or not. For example, responded cells are determined by ΔRI ≧ 15%
over reflection intensity increase at 1 min from K+ stimulation,
ΔRI ≧ 5% over reflection intensity increase at 24 min from phor-
bol-12-myristate-13-acetate (PMA) injection, ΔRI ≧ 3% over
reflection intensity increase 20 min from muscarine injection,
respectively.

3.4 SPR Data 1. 2D-SPR software (originally developed by NTT-AT. Co,

Analysis Japan, and incorporated in the computer connected to the
2D-SPR apparatus) was used for experimental SPR analysis.
However, other imaging software (e.g., ImageJ) theoretically
can be used for the analysis.
2. At the time of cell culturing on the gold chip, the cell number
was controlled by checking the samples using phase contrast
microscopy before placing on the 2D-SPR measurement
platform.
3. During SPR measurement the CCD camera visualizes the
entire sensor surface. Selected areas of interest in the viewable
regions of the computer screen are encircled and selected for
analysis.
4. The circle size is not fixed and the actual size of the circle was
not measured.
5. As shown in Fig. 2a, b, areas corresponding to pixels containing
cells are marked as regions 1–5. An area without cells (control)
is marked as region 6.
40 Tanveer Ahmad Mir and Hiroaki Shinohara

Fig. 2 (a) 2D-SPR image of the sensor chip surface is recorded before K+ injection and after the injection of
Hanks’ solution (negative control). (b) 2D-SPR image of the sensor chip surface recorded after 100 mM KCl
injection, showing a robust increase in reflection intensity in PC12 cell regions. (c) Time course of the
reflection intensity changes at individual PC12 cell regions on KCl stimulation with the 2D-SPR imager. (d)
Representative columns indicating the percentages of SPR-responding cells according to concentrations of
KCl. Reprinted from ref. 6, with permission from Elsevier

6. The images of the encircled areas as well as signal patterns (SPR

curves) correspond to these encircled areas measured in real
time (results in Fig. 2c, regions 1–5).
7. The refractive index in each CCD pixel corresponding to the
chosen region is recorded and saved.
8. 2D-SPR analysis software automatically records the reflected
light images (SPR images), which can be exported to Excel.
9. Finally, the saved Excel data was used for (time course measure-
ment) graph plotting.
10. Because of the low magnification (7
) while the cells are clearly
visible, the cell volume/dimensions or cell organelles are not
measured, SPR imaging research is still at its infancy, and
improvement in the resolution of SPR imaging system is
greatly needed.
 SPR Imaging for Cellular Analysis 41

3.5 High KCl-Induced 1. Prior to experiment, the 2D-SPR imager is assembled. The
Response in PC12 sensor chip on which cells are adhered is fixed on the prism of
Cells Observed with the 2D-SPR imager.
2D-SPR Imager 2. The cell and blank areas on the sensor chip surface are observed
and then the measurement is started. SPR curve that is depen-
dence of reflectivity of the area of interest on incident angle is
first evaluated to get exact resonance angle and to obtain
suitable measurement angle.
3. SPR curve in terms of time course measurement after exposing
to stimulant is observed at the suitable measurement resonance
angle.
4. After stabilization of baseline, cells are first exposed to Hanks’
solution as reference. As expected, no change in reflection
intensity is observed after exposing to negative control solution
(see Note 10).
5. To monitor SPRi sensorgram for intracellular signal transduc-
tion pathway (mainly PKC translocation) via depolarization,
the cells are exposed to different concentrations of K+ solution
and sequential alterations of average reflected light intensity in
terms time course observation for each pixel of the selected
areas of interest on the monitor are obtained and analyzed
using the 2D-SPR analyzer software and images are captured
by the CCD camera.
6. After exposing cell to K+ solution, 2D-SPR images of PC12 cell
areas are gone from dim to bright and variation in cell-to-cell
reflection intensity change response is also observed. It is
speculated that this variation may be due to differences in the
physique of the respective cells (Fig. 2) (see Note 11).

3.6 PKC Agonist and 1. To confirm the results whether K+-induced reflection intensity
Antagonist-Induced change in 2D-SPR sensing is genuinely implicated to intracel-
Response in PC12 lular signal transduction (translocation of PKC) in PC12 cells,
Cells Observed with cells are exposed to Phorbol-12-myristate-13-acetate (PMA), a
2D-SPR Imager specific activator of Protein Kinase C (PKC).
2. SPR images of the PC12 cells recorded with the CCD camera
before and after stimulation of varying the concentrations of
Phorbol-12-myristate-13-acetate (PMA), the obtained SPR
curves showed a slow increase with fluctuation after gaining
phorbol-12-myristate-13-acetate (PMA) access to the cell inte-
rior (Fig. 3) (see Note 12).
3. To further validate the results and examine the inhibitory
effects, PC12 cells are pretreated with a potent inhibitor of
PKC (staurosporine) for 30 min, on the Phorbol-12-myris-
tate-13-acetate-induced response (see Note 13).
42 Tanveer Ahmad Mir and Hiroaki Shinohara

Fig. 3 (a) 2D-SPR image of the sensor chip surface obtained after the injection of Hanks’ solution for
background observation. (b) 2D-SPR image of PC12 cell regions obtained after phorbol-12-myristate-13-
acetate (PMA)stimulation (500 nM), which showed enhancement of the reflection intensity in cell regions. (c)
Reflection intensity change corresponding to the intracellular refractive index change, measured with the 2D-
SPR imager for PC12 cells adhered to a gold chip on 500 nM phorbol-12-myristate-13-acetate (PMA)
injection. After injecting 500 nM PMA, a slow elevation of reflection intensity was seen at cell regions. No
reflection intensity change is observed in non-cell regions. (d) Representative columns indicating the
percentages of SPR-responded cells in the presence (red) or absence (blue) of 100 nM staurosporine.
Reprinted from ref. 6, with permission from Elsevier

4. The representative columns of comparative graph for the

dependence of reflection intensity change induced by stimula-
tion with various concentrations of phorbol-12-myristate-13-
acetate (PMA) in the presence or absence of staurosporine
showed a clear effect on these activators or inhibitors on SPRi
generated signal (Fig. 3).

3.7 2D-SPRi 1. 2D-SPR imaging system is further applied to observe the

Observation of the intensity modulation of intracellular signaling as functional
Effect of Muscarine on readout for differentiation associated with intracellular PKC
Normal and NGF- translocation, PC12 cells are exposed to Muscarine (acetylcho-
Treated Differentiated linergic receptor agonist) before and after differentiation (see
PC12 Cells Note 14).
2. Compared to non-differentiated cells, NGF-treated differen-
tiated PC12 cells showed a remarkable enhancement in reflec-
tion intensity change signal (Fig. 4).
 SPR Imaging for Cellular Analysis 43

a
3
28
Reflection intensity / %

Cell regions (2-5) 2

Muscarine
(100μM) 4
24 5
Hanks’
c
1 8
20

Δ Reflection intensity / %
No cell region (1)
6
16
0 5 10 15 20 25 30 4
Time / min

b 2
54 1
Reflection intensity / %

Cell regions (1-4) 3 0

Muscarine 2 0 10 20 30 40 50
47
(100μM)
Hanks’ 4 Muscarine concentration. / μM

40
5

No cell region (5)

33
0 5 10 15 20 25 30
Time / min

Fig. 4 Time course of the reflection intensity change for undifferentiated (a) and differentiated (b) PC12 cells at
individual cell regions on muscarine stimulation with the 2D-SPR imager. After the application of muscarine,
slow and fluctuating elevation of reflection intensity was seen at cell regions. No reflection intensity change is
observed at the non-cell region. (c) Dependence of reflection intensity change at the undifferentiated cell
regions and the differentiated cell regions (2 days treatment, 4 days treatment) on the concentration of
muscarine as a stimulant. The PC12 cells are treated with 80 ng/ml NGF for 2 or 4 days. Reprinted from ref. 7,
with permission from Elsevier

4 Notes

1. After finishing the experiment, the medium is discarded and

the chips are washed carefully and thoroughly to remove sus-
pended dead cells or other material.
2. Cells are centrifuged at 500
g for 3 min, supernatant is
decanted, and the cell pellet diluted in pre-warmed culture
medium.
44 Tanveer Ahmad Mir and Hiroaki Shinohara

3. Usually, optimum confluency is obtained within this period.

4. The cell suspension is diluted to 4.0
105 cells/ml (for experi-
ments before differentiation) with culture medium and poured
onto a sensor chip placed in the sterilized culture dish. The cells
are evenly distributed by brief agitation of the gold sensor chip,
and the cells are allowed to adhere and spread typically over-
night to reach a stable baseline before being placed onto the
2D-SPR imaging platform for stimulant stimulation and obser-
vation of reflection intensity changes.
5. For the neuronal differentiation study, the cell suspension is
diluted to 3.0
105 cells/ml with culture medium and poured
onto poly-L-lysine coated sensor chips.
6. Bottles are labeled as follows: l DMEM l Sterile l Date l Initials.
Dry powder must be free-flowing. Do not use if powder is
caked or clumping, indicating accumulation of moisture.
7. Phorbol-12-myristate-13-acetate (PMA) solution is prepared
on the day of the experiment. Stock solutions are stored at
20  C. This compound is irritating to eyes, respiratory sys-
tem, and skin, so avoid contact with skin and eyes by wearing
suitable protective clothing.
8. Staurosporine solution is prepared on the day of the experi-
ment. Stock solutions are stored at 20  C. This compound is
irritating to eyes, respiratory system, and skin, so avoid contact
with skin and eyes by wearing suitable protective clothing.
9. Muscarine solution is prepared on the day of the experiment.
Stock solutions are stored at 20  C. This compound is irritat-
ing to eyes, respiratory system, and skin, so avoid contact with
skin and eyes by wearing suitable protective clothing.
10. Hanks’ buffer is used to avoid possible confounding effects of
the measuring buffer solution that may stem from the mechan-
ical disturbance caused by the turbulence during solution ejec-
tion by manual system.
11. It is shown in Fig. 2c, the reflection intensity of the PC12 cell
region is dramatically increased and individual cell regions
showed a similar response pattern; however, the reflection
intensity was uneven variation that may be borne from differ-
ences in physical condition of the particular cell.
12. It is shown in Fig. 3c, the extracellular administration of
Hanks’ solution stimulation (bottom line) failed to produce
any alteration in reflection intensity response curve. However, a
few minutes after phorbol-12-myristate-13-acetate (PMA), a
potent tumor promoter that can intercalate into membrane
and potently activate intracellular PKC by recruiting it to cell
membranes, thereby mimicking the mechanism of the natural
PKC activator, diacylglycerol exposure, the reflection intensity
 SPR Imaging for Cellular Analysis 45

response curve increased, accompanied by a conspicuous

increase of intracellular signal transduction. It is observed
that the signal transduction that might be associated with the
translocation of PKC is monitored for individual cells exhibit-
ing a similar slow elevating SPR curve. It is thought that, unlike
diacylglycerol, phorbol esters are not readily metabolized, and
exposure of cells to these molecules results in prolonged acti-
vation of PKC.
13. To further study whether the reflection intensity change signal
measured by the 2D-SPR imaging system with the stimulation
of these diacylglycerol surrogates indeed associated with the
activation of PKC, the effect of treating cells with an antagonist
of PKC, staurosporine, on the response to phorbol-12-myris-
tate-13-acetate (PMA) is determined and compared. As shown
in in Fig. 3d, Staurosporine-treated cells decreased the
response after treating with phorbol-12-myristate-13-acetate
(PMA) solution and lead reflective intensity decreases toward
the basal level.
14. The effect of nerve growth factor (NGF) on PC12 cell culture is
evaluated by observing reflection intensity change responses at
the individual cell level under the consideration that NGF tends
to decrease cell growth; proliferation; different responses could
be recorded in PC12 cells maintained under controlled circum-
stances in comparison with those incubated in the presence of
NGF. It is shown in Fig. 4, after Hanks’ solution injection no
reflection intensity change response is detected, whereas after
exposing the cells to muscarine solution a slow and fluctuating
elevation of reflectivity curves is obtained, with a similar
response pattern for both non-differentiated and differentiated
PC12 cells. However, a significant enhancement in reflection
intensity change response is observed in NGF-treated
PC12 cells. The obtained data demonstrates that the 2D-SPR
imager is not only useful for intracellular signal detection but
could also be useful to distinguish differentiated cells from non-
differentiated counterparts noninvasively, in real time and inde-
pendent of any fluorescent or radioactive probes.

References
1. Suzuki M, Ohshima T, Hane S, Iribe Y, Tobita T 4. Horii M, Shinohara H, Irebe Y, Suzuki M
(2007) Multiscale 2D-SPR biosensing for cell (2007) Application of high resolution 2D-SPR
chips. J Robot Mechatron 19:5519–5523 imager to living cell-based allergen sensing. In:
2. Homola J (2008) Surface plasmon resonance Viovy J-L, Tabeling P, Descroix S, Malaquin L
sensors for detection of chemical and biological (eds) The proceedings of lTAS 2007 conference.
species. Chem Rev 108(2):462–493 Chemical and biological microsystems society,
3. Yanase Y, Hiragun T, Ishii K, Kawaguchi T, Paris, pp 451–453
Yanase T, Kawai M, Sakamoto K, Hide M 5. Horii M, Shinohara H, Irebe Y, Suzuki M
(2014) Surface plasmon resonance for cell- (2011) Living cell-based allergen sensing using
based clinical diagnosis. Sensors 14 a high resolution two-dimensional surface plas-
(3):4948–4959 mon resonance imager. Analyst 136:2706–2711
46 Tanveer Ahmad Mir and Hiroaki Shinohara

6. Shinohara H, Sakai Y, Mir TA (2013) Real-time by 2D-SPR sensor: an approach to study neuro-
monitoring of intracellular signal transduction in nal differentiation. In: Proceeding of the 14th
PC12 cells by two-dimensional surface plasmon international meeting on chemical sensors. doi:
resonance imager. Anal Biochem 441:185–189 10.5162/IMCS 2012/P1.1.18
7. Mir TA, Shinohara H (2013) Two-dimensional 9. Mir TA and Shinohara H (2012) 2D-SPR bio-
surface plasmon resonance imager: an approach sensor detects the intracellular signal transduc-
to study neuronal differentiation. Anal Biochem tion in PC12 cells at single cell level. In:
443:46–51 Proceeding of the sixth international conference
8. Mir TA and Shinohara H (2012) Real-time on sensing technology. doi: 10.1109/ICSensT.
monitoring of cell response to drug stimulation 2012.6461763
 Chapter 4

Immunosensing with Near-Infrared Plasmonic

Optical Fiber Gratings
Christophe Caucheteur, Clotilde Ribaut, Viera Malachovska,
and Ruddy Wattiez

Abstract
Surface Plasmon resonance (SPR) optical fiber biosensors constitute a miniaturized counterpart to the
bulky prism configuration and offer remote operation in very small volumes of analyte. They are a cost-
effective and relatively straightforward technique to yield in situ (or even possibly in vivo) molecular
detection. They are usually obtained from a gold-coated fiber segment for which the core-guided light is
brought into contact with the surrounding medium, either by etching (or side-polishing) or by using
grating coupling. Recently, SPR generation was achieved in gold-coated tilted fiber Bragg gratings
(TFBGs). These sensors probe the surrounding medium with near-infrared narrowband resonances,
which enhances both the penetration depth of the evanescent field in the external medium and the
wavelength resolution of the interrogation. They constitute the unique configuration able to probe all
the fiber cladding modes individually, with high Q-factors. We use these unique spectral features in our
work to sense proteins and extra-cellular membrane receptors that are both overexpressed in cancerous
tissues. Impressive limit of detection (LOD) and sensitivity are reported, which paves the way for the
further use of such immunosensors for cancer diagnosis.

Key words Surface plasmon resonance, Optical fiber sensors, Bragg gratings, Immunosensing, Cells,
Proteins

1 Introduction

Biosensors are by definition a combination of a biological receptor

compound and a physical or physicochemical transducer. Optical
methods of transduction offer important advantages such as:
– Noninvasive, safe, and multi-dimensional detection (wave-
length, intensity, phase, polarization, etc.).
– Well-established technologies readily available from communi-
cation and micro-nano technologies industries (lasers, detectors,
passive components, etc.).

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_4,
© Springer Science+Business Media LLC 2017

47
48 Christophe Caucheteur et al.

– Visible and near-infrared optical frequencies coincide with a

wide range of physical properties of bio-related materials.
Biosensors are continuously developed to respond to the
demand for rapid, accurate, and in situ monitoring techniques in
different areas such as medical diagnosis, genomics, proteomics,
environmental monitoring, food analysis, and security. Label-free
optical biosensors yield direct and real-time observation of molec-
ular interaction, without requiring the use of labels since they sense
binding-induced refractive index changes. Among the different
optical biosensor configurations (measurements of absorbance,
reflectance, fluorescence, refractive index changes, etc.), those
using surface plasmon polaritons (SPPs) have been intensively
developed during the past three decades.
SPPs result from collective excitations of electrons at the
boundary between a metal and a dielectric. The strong sensitivity
of the plasmon-propagation constant to the permittivity of a nano-
scale layer of material (bioreceptors) deposited on the metallic
surface has led to the monitoring of surface plasmon resonance
(SPR) to detect (bio)chemical changes occurring between mole-
cules. As a result, biochemical reactions are measured with SPR by
monitoring their effective refractive index. In practice, as sketched
in Fig. 1, the so-called Kretschmann-Raether approach realizes this
by launching light beams from a high refractive index prism to a
thin metallic interface at an angle such that light is totally reflected
[1]. In doing so, an evanescent wave extends in the metal overlay.
When the component of the propagation constant of the light
along the interface matches that of a plasmon excitation of the
other side of the metal layer, part of the light couples to the
plasmon, which decreases the reflection (at θ0 in Fig. 1 where θc
denotes the critical angle of incidence). The interrogation is made
either by varying the wavelength and keeping the incidence angle
constant or by using monochromatic light and modifying the

Fig. 1 Sketch of the Kretschmann-Raether prism approach for plasmonic generation and operating principle of
its interrogation to measure binding of biomolecules on receptors grafted on the gold substrate
Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 49

angle. In both cases, the polarization state of the light has to be

parallel to the incidence plane so that the plasmon wave is normally
polarized to the interface. Biochemical reactions occurring above
the metal surface affect the effective index of the plasmon wave,
which is detected through a shift of the SPR (from θ0 to θ1). The
sensitivity to the surrounding refractive index often ranges in the
order of 10
6–10
7.
Numerous transducers have been developed, bringing addi-
tional assets with respect to the Kretschmann-Raether prism
approach, which is exploited in most commercial systems. In this
aspect, optical fiber-based sensors are particularly attractive. With
their lightweight, compactness, and ease of connection, they pro-
vide remote operation in very small volumes of analyte of the order
of microliter. And with their continuous development and optimi-
zation, they appear perfectly suited for in situ or even possibly
in vivo diagnosis. They also have the potential to assay different
parameters simultaneously.
To excite SPR from an optical fiber, light confined in the fiber
core has to be locally outcoupled and brought into contact with the
surrounding medium. In practice, this is achieved either by a geo-
metrical modification (polishing or etching of the cladding) so as to
expose the evanescent wave to the surrounding medium or by using
in-fiber gratings (refractive index modulations photo-imprinted in
the fiber core along the propagation axis). Hence, various architec-
tures are available [2]: etched multimode optical fibers, side-
polished, D-shaped, tapered, or U-bent optical fibers, long period
fiber gratings (LPFGs) and tilted fiber Bragg gratings (TFBGs).
Configurations based on cladding removal/decrease can be quite
easily achieved. The SPR is in this case spectrally manifested by a
broadband resonance (full width at half maximum (FWHM)
~20 nm or higher) in the transmitted amplitude spectrum. Opera-
tion in reflection mode is possible by using a mirror deposited on
the cleaved fiber end face beyond the sensing region. However,
these configurations considerably weaken optical fibers at the sen-
sor head and may prevent their use in practical applications, out of
laboratory settings. For this reason, large core fibers (unclad
200–400 μm core fibers) are the most spread in practice [3].
These configurations operate at visible wavelengths, which limits
the extension of the evanescent wave in the surrounding medium.
Indeed, its penetration depth is proportional to the operation
wavelength (λ) and usually ranges between λ/5 and λ/2, depending
on the mode order [4]. Hence, operation at near-infrared telecom-
munication wavelengths enhances the penetration depth, which in
turn improves the overall sensor sensitivity to large-scale targets
such as proteins or cells. Such operation can be easily achieved with
in-fiber gratings.
Gratings preserve the fiber integrity while providing a strong
coupling between the core-guided light and the cladding. LPFGs
50 Christophe Caucheteur et al.

consist in a periodic refractive index modulation of the fiber core of

a few hundreds of μm. They couple the forward-going core mode
into forward-going cladding modes. Their transmitted amplitude
spectrum is composed of a couple of wide resonances (FWHM
~20 nm) dispersed in a wavelength range of a few hundreds of
nm. TFBGs are short period (~500 nm) gratings with a refractive
index modulation slightly angled with respect to the perpendicular
to the optical fiber axis. In addition to the self-backward coupling
of the core mode, they couple light into backward-going cladding
modes. Their transmitted amplitude spectrum displays several tens
of narrow-band cladding mode resonances (FWHM ~200 pm or
even below) located at the left-hand side of the Bragg resonance (or
core mode resonance) corresponding to the core mode self-
coupling. According to phase matching conditions, every cladding
mode resonance possesses its own effective refractive index and the
maximum refractometric sensitivity is obtained when this effective
index tends to the surrounding refractive index value. TFBGs act as
spectral combs and constitute the only optical fiber configuration
able to probe simultaneously but distinctively all the cladding
modes supported by an optical fiber [5].
SPR optical fiber sensors can be obtained from the above-listed
structures surrounded by a noble metal (most often gold or silver).
Sheaths of thickness ranging between 30 and 70 nm are most
generally used. SPR generation is achieved when the electric field
of the light modes is polarized mostly radially at the surrounding
medium interface. The orthogonal polarization state is not able to
excite the SPR, as the electric field of the light modes is polarized
mostly azimuthally (i.e., tangentially to the metal) at the surround-
ing medium interface and thus cannot couple energy to the surface
Plasmon waves. Depending on the configuration, refractometric
sensitivities in the range [102–105 nm/RIU (refractive index
unit)] have been reported [6].
When comparing the sensor performances between different
configurations, it is not sufficient to compare only sensitivities (i.e.,
wavelength shifts), without considering the wavelength measure-
ment accuracy. It is more convenient to refer to the figure of merit
(FOM) of the device. The FOM corresponds to the ratio between
the sensitivity and the linewidth of the resonance, taking into
account that it is easier to measure the exact location of a narrow
resonance than a broad one [7]. Hence, in terms of experimentally
demonstrated FOM, TFBGs outclass all other configurations by
more than one order of magnitude [2]. This results from their
sensitivity close to 500 nm/RIU and the narrowness of their
resonances.
Plasmonic generation in gold-coated TFBGs has been pio-
neered in 2006 by the team of Prof. Jacques Albert at the Carleton
University of Ottawa, Canada [8]. Since then, our two research
groups have worked together on the spectral characterization of
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 51

these probes that are inherently polarization selective [9, 10] and
on their subsequent use for biochemical sensing [11, 12].
In our work, we use the well-acknowledged antibody-antigen
affinity mechanism to assess the unique protein detection and
quantification capabilities of gold-coated TFBGs and evaluate
their performances in terms of both sensitivity and limit of detec-
tion (LOD). We have conducted experiments on two kinds of
bioreceptors, e.g., proteins and extracellular membrane receptors
that are both overexpressed in the case of cancer cells. In particular,
we have quantified a certain type of cytokeratins that are secreted by
lung cancer tumors and cells suspensions. Using another functional
film on the same sensor, we have also detected extracellular mem-
brane receptors in native membranes of different human epithelial
cell lines. A differential diagnosis has been demonstrated between
two cell lines, with overexpressed membrane receptors (positive
control) and with a low number of these receptors (negative con-
trol). Such results make an important step forward toward the
demonstration of in vivo diagnosis.
In the remaining of this chapter, we will first describe the
operating principle of the sensor. Then, we will focus on its perfor-
mances in the quantification of proteins and cells in vitro.

2 Materials

2.1 Optical System to 1. Frequency-doubled Argon-ion laser emitting at 244 nm with a

Produce TFBGs mean output power of 50 mW (Newport SpectraPhysics).
2. Uniform phase mask with a pitch of 1090 nm (Coherent).
3. Automated one-axis translation stage (Physik Instrumente).
4. 5-cm-focal distance cylindrical lens in front of the phase mask
(Newport).
5. Diaphragm to spatially filter the UV laser beam (Newport).
6. Telecommunication-grade single-mode optical fiber (Corning
SMF-28) .

2.2 Gold Deposition 1. Sputtering chamber (LEICA EM SCD 500).

on the Optical Fiber 2. 99.99% purity gold foil used as the target in the sputtering
Surface chamber.

2.3 Optical System to 1. Optical vector analyzer (Luna Technologies OVA CTe).
Measure the Spectrum 2. In-line linear polarizer (General Photonics).
of Gold-Coated TFBGs
3. In-house developed microfluidic chamber to immerse gold-
coated TFBGs in controlled liquids.
4. Hand-held Abbe refractometer accurate to 10
4 RIU (Reichert
AR200).
52 Christophe Caucheteur et al.

2.4 Surface 1. Dithiolalkanearomatic PEG6-COOH (thiols) (SensoPath

Chemistry Technologies, N SPT-0014A6) was prepared at 2 mM in
absolute ethanol.
2. Phosphate buffer saline (PBS) (Sigma N P4417) was prepared
to yield 10 mM phosphate (PO4
), 2.7 mM potassium chloride
(KCl), and 137 mM sodium chloride (NaCl), at pH 7.4 in
milliQ water.
3. The mixture N-hydroxysuccinimide (NHS) (Sigma-Aldrich
56480) and N-(3-)dimethylaminopropyl)-N0 -ethylcarbodii-
mide hydrochloride (EDC) (Sigma-Aldrich E7750) was
prepared at a ratio 1:5 with a final concentration of NHS at
100 mM and 500 mM of EDC in milliQ water.
4. 1% (w/v) bovine serum albumin (BSA) (Acros Organics
134731000) was made in phosphate buffer solution at pH 7.4.
5. Polyclonal anti-Cytokeratin 7 antibodies (AbCK7) (Biorbyt
orb10410) have been diluted in phosphate buffer solution at
pH 7.4, at 20 μg/mL.
6. Cytokeratin 7 peptide (CK7pep) (Biorbyt orb72193) made of
23 amino acids, corresponding to a molecular weight of
2.65 kDa, was diluted in phosphate buffer solution at pH
7.14, at different concentrations.
7. Cytokeratin 7 full protein (CK7FP) (abcam 132933) which is
characterized by the association of 469 amino acids and a
molecular weight equal to 78 kDa, has been prepared at differ-
ent concentrations in PBS.
8. Fetal bovin serum (FBS) (Sigma F7524) was used to comple-
ment the phosphate buffer solution for the measurement in
complex media.

3 Methods

3.1 Operating The sensor is obtained from a single-mode optical fiber in which a
Principle of TFBGs photo-imprinted refractive index grating is formed over a short
section of the core, and that is further coated with a bilayer coating:
a very thin metal coating and a suitable biochemical recognition
binding layer.
The glass optical fiber is a 125-μm-thick cylindrical waveguide
made of two concentric layers, the core in the middle surrounded
by a cladding that is thick enough to prevent light propagating in
the core to interact with the fiber surroundings. The refractive
index of the core is slightly higher than that of the cladding to
allow for the guidance of light in the fiber core. Our experiments
are performed on 8 μm core single-mode optical fibers that guide
light into a single optical mode at wavelengths between 1300 and
1650 nm. Such fibers are widely available at low cost (less than 100
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 53

Fig. 2 Sketch of the light coupling mechanism for uniform FBGs (a) and tilted FBGs (b)

US$ per km) and are telecommunication-grade. In the same way, a

large quantity of equipment is available to characterize, manipulate,
and use such fibers and devices made from them.
As sketched in Fig. 2a, a uniform FBG is a periodic and perma-
nent refractive index modulation of the fiber core that is imprinted
perpendicularly to the propagation axis [13, 14]. It behaves as a
selective mirror in wavelength for the light propagating in the core,
reflecting a narrow spectral band centered on the so-called Bragg
wavelength λBragg ¼ 2neff,coreΛ where neff,core is the effective refrac-
tive index of the core mode (close to 1.45) and Λ is the grating
period. In practice, Λ ~500 nm to ensure that the Bragg wavelength
falls in the band of minimum attenuation of the optical fiber cen-
tered on 1550 nm. The Bragg wavelength is inherently sensitive to
mechanical strain and temperature, through a change of both neff
and Λ [13]. In practice, a change of temperature of +1  C yields a
Bragg wavelength shift of ~10 pm (at 1550 nm). Such change is
easily measured with standard telecommunication instruments
since the full spectral width of the reflected light from a typical
1-cm-long grating is of the order of 100 pm. However, our purpose
here is to measure events occurring on or near the fiber cladding
surface and the core mode reflection spectrum is inherently and
totally insensitive to such changes (because the penetration depth
of core-guided light into the cladding does not exceed a few micro-
meters). We use a small modification of the FBG to couple light
from the core to the cladding and still benefit from narrowband
spectral resonances that will reveal small changes at the cladding
boundary.
A TFBG corresponds to a refractive index modulation angled
by a few degrees relative to the perpendicular to the propagation
axis (Fig. 2b) [15]. In addition to the self-backward coupling of the
core mode at the Bragg wavelength Gold-coated tilted fiber Bragg
gratings (TFBGs):, the grating now redirects some light to the
cladding whose diameter is so large that several possible cladding
modes can propagate, each with its own phase velocity (and hence
effective index neff,clad). These possible modes of propagation cor-
respond to different ray angles in Fig. 2b. Again, there is a one-to-
one relationship between the wavelength at which coupling occurs
for a given cladding mode and its effective index. This relationship
54 Christophe Caucheteur et al.

Fig. 3 Typical transmission spectrum Gold-coated tilted fiber Bragg gratings

(TFBGs):transmission spectrum of a 10 tilted fiber Bragg grating measured in air
(using linearly polarized input light as described in Reference [5])

is expressed by a similar phase matching condition as before:

λiclad ¼ (neff,core + nieff,clad) Λ where the index “i” reflects the fact
that the fiber cladding can support many modes. Figure 3 shows a
TFBG transmission spectrum where each resonance corresponds to
the coupling from the core mode to a group of backward propagat-
ing cladding modes. As a result of phase matching, the spectral
position of a resonance now depends on the effective index of its
associated cladding mode, which in turn depends on the optical
properties of the medium on or near the cladding surface.
Therefore, spectral shifts of individual resonances can be used
to measure changes in fiber coatings and surroundings. Laffont and
Ferdinand were the first to demonstrate SRI sensing with TFBGs in
2001 [16]. They observed a progressive smoothing of the trans-
mitted spectrum starting from the shortest wavelengths as the SRI
increased from 1.30 to 1.45. Several data processing techniques can
be used to quantitatively correlate the spectral content with the SRI
value, either based on a global spectral evolution or a local spectral
feature change. The first method involves monitoring the area
delimited by the cladding mode resonance spectrum, through a
computation of the upper and lower envelopes as resonances grad-
ually disappear when the SRI reaches the cut-off points of each
cladding mode [16, 17]. Another technique tracks the wavelength
shift and amplitude variation of individual cladding mode reso-
nances as they approach cutoff [18]. Both techniques present
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 55

minimum detectable SRI changes of ~10
4 RIU (refractive index

unit). In terms of wavelength tracking, this result is obtained with a
sensitivity that peaks between 10 and 15 nm/RIU for the modes
near cutoff. In all cases, the Bragg wavelength provides an absolute
power and wavelength reference that can be used to remove uncer-
tainties related to systematic fluctuations (such as power level
changes from the light sources) and even temperature changes
(because it was demonstrated in Chen et al. [19] that all resonances
shift by the same amount when the temperature changes). The
TFBG thus allows inherently temperature-insensitive SRI measure-
ments and large signal-to-noise ratio. For biochemical experiments
however, it is usually necessary to improve the LOD levels to at least
10
5 RIU, by increasing the wavelength shift sensitivity while
keeping noise level down and spectral features narrow [20]. Fortu-
nately, it has been recently demonstrated that the addition of a
nanometric-scale gold coating overlay on the TFBG outer surface
considerably enhances the refractometric sensitivity through the
generation of surface Plasmon resonances (SPR) [8, 21, 22]. This
sensitivity increase can be easily achieved with gold-coated TFBGs.

3.2 Photo- 1. TFBGs are manufactured in the same way as conventional

Inscription of TFBGs FBGs, i.e., through a lateral illumination of the fiber core
using an interference pattern of ultraviolet (UV) light between
190 and 260 nm. The exposure of the fiber to bright interfer-
ence fringes for a few minutes locally and permanently increases
the mean refractive index of the core. The ultraviolet interfer-
ence pattern is reproduced inside the fiber much like in a
photographic process. Mass production of identical gratings is
straightforward when the phase mask technique is used to
generate the interference pattern. A phase mask is a diffractive
element made in a pure silica substrate that is optimized to
maximize the diffraction in the first (+1 and
1) diffraction
orders. An interference pattern at half the period of the mask is
therefore produced in the fiber core when the optical fiber is
located in close proximity behind the mask, as depicted in
Fig. 4. In order to obtain a TFBG, we usually rotate the
phase mask in the plane perpendicular to the incident laser
beam. We routinely produce TFBGs with tilt angles varying
between 4 and 10 , the larger angle being preferred to operate
when the SRI lies near the index of water and water solutions,
which is often the case in biochemical research, because then
the strongest cladding mode resonances are located near
1550 nm (when the Bragg wavelength is near 1610 nm),
where they are easier to measure. Their physical length is
typically 1 cm.
2. Prior to the manufacturing process, the single-mode fibers are
hydrogen-loaded to enhance their photosensitivity (capability to
change their refractive index when exposed to a light
56 Christophe Caucheteur et al.

Fig. 4 Sketch of the phase mask technique

interference pattern) to ultraviolet light. This is done in a vessel

under pressure (200 atm) and at moderate temperature (70  C) .
3.3 Gold Deposition 1. We use a sputtering chamber to extract gold from a target.
on the Optical Fiber 2. Two consecutive depositions are made in the same conditions,
Surface with the optical fibers rotated by 180 between both processes,
to ensure that the whole outer surface is covered by gold.
3. The optimum gold thickness has been found to be approxi-
mately 50 nm, for the narrowest, deepest SPR attenuation. In
spite of the two-step deposition that yields a slightly nonuni-
form thickness around the fiber circumference, we did not find
this nonuniformity to be detrimental for SPR generation in
practice. Also, as the vacuum is obtained in the chamber start-
ing from ambient air and not from an inert (Argon) starting
atmosphere, there is no need for a 2–3-nm-thick adhesion layer
of chromium or titanium between the silica surface and the
gold coating.

3.4 Surface The gold-coated TFBGs are functionalized for biosensing pur-
Functionalization poses. The chemistry involved in this process depends on the target
application. In our case study, it is based on the antigen/antibody
affinity. Whatever the analyte to be detected, a self-assembled
monolayer (SAM) is manufactured, prior to the biomolecules
immobilization.
1. For this, gold-coated TFBGs were first thoroughly rinsed with
ethanol and dried under nitrogen.
2. Gold-coated TFBGs were then immersed in a solution of thiols
(2 mM) dispersed in ethanol. Thiols incubations were usually
done during 18 h at room temperature in a 1-mm-thick capil-
lary tube sealed at both ends to prevent solvent evaporation. At
the end of the incubation, the functionalized gold-coated
TFBGs were removed from the tube and again rinsed with
ethanol (see Note 1).
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 57

3. The next step of the surface functionalization consisted in the

esterification of the carboxylate functions of the SAM. The
reaction was realized by an immersion of the gold-coated
TFBGs in a mixture of NHS/EDC (1:5 M/M) in H2O milliQ
for 2 h. The surface was then rinsed with milliQ water and dried
under nitrogen.
4. The gold-coated TFBGs were immediately immersed in a pri-
mary antibody solution (AbCK7 at 20 μg/mL in PBS at pH
7.4). After 2 h of incubation at room temperature, the gold-
coated TFBGs were rinsed with PBS (see Note 2).
5. A blocking step was realized by immersion of the optical fibers
in BSA solution 1% (w/v) in PBS at pH 7.4 for 1 h to avoid
nonspecific interaction.
6. Bioreceptors were finally rinsed with PBS and ready to be used
for Cytokeratin 7 proteins detection (see Note 3).
These four consecutive steps (grating manufacturing, gold
deposition, surface functionalization, and bioreceptors grafting)
yield the plasmonic probe sketched in Fig. 5. The illustration
below displays the photo-inscribed optical biosensor with antibo-
dies (orange semicircle) immobilized on the gold layer deposited
on the optical fiber’s cladding. The optical fiber is finally presented
here in a packaging (gray cylinder) in the presence of antigens (blue
sphere) in a bulk solution.

3.5 Interrogation of 1. Transmitted amplitude spectra of gold-coated TFBGs were

Gold-Coated TFBG recorded by means of an optical vector analyzer from Luna
Immunosensors Technologies, which was chosen for both its high measurement
resolution (1.25 pm) and fast scanning rate (~1 s to cover the
full TFBG spectrum). This analyzer gathers a narrowband

Fig. 5 Sketch of a gold-coated tilted fiber Bragg grating used for biosensing
purposes which displays the functionalized photo-inscribed optical fiber inserted
in a packaging
58 Christophe Caucheteur et al.

tunable laser source and a detector. It can work either in

transmission or in reflection.
2. A linear polarizer was placed behind the optical source to
control and orient the state of polarization (SOP) of the light
launched into the TFBG. It is worth mentioning that care was
taken to avoid polarization instabilities (use of short fiber
lengths, strong curvatures avoided and ambient temperature
kept constant to within 1  C).
Figure 6 displays the transmitted amplitude spectrum of a gold-
coated TFBG immersed in salted water (refractive index measured
close to 1.356 at 589 nm), which was recorded with a linear input
SOP optimized to maximize coupling to the SPR, corresponding
to the radial polarization, as further explained in the following. This
spectrum exhibits the typical SPR signature around 1551 nm,
which is due to the maximum phase matching of the cladding
mode to the surface plasmon mode of the gold water interface,
according to [8]. The Bragg wavelength appears at the right end
side, centered at 1602 nm. In the following, the Bragg wavelength
is used to remove any effect from surrounding temperature changes
by monitoring its shift. This intrinsic feature is very interesting in
practice as a change of 0.1  C induces an SRI change of 10
5, thus
susceptible to generate erroneous spectral modifications.
Figure 7 displays a zoom around the SPR signature for radially-
(EH and TM modes) and azimuthally polarized (HE and TE
modes) amplitude spectra that yield antagonist behaviors in liquids,

Fig. 6 Transmitted amplitude spectrum of an SPR-TFBG immersed in salted

water (radial polarization—SRI ¼ 1.358)
Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 59

Fig. 7 Transmitted amplitude spectra for two orthogonal SOPs (radial and
azimuthal polarizations) of a TFBG immersed in salted water (SRI ¼ 1.338)

as explained in [8]. It is obvious from this figure that EH and HE

modes come in pairs and to facilitate the following discussion, the
mode resonances are labeled as a function of their position with
respect to the most important one in the EH spectrum (mode 0 in
our numbering).
Figure 6 reveals that, for short wavelengths (mode
2), the
behavior is similar to that of bare TFBGs in air, with EH modes at a
longer wavelength than HE ones. Then, getting closer to the SPR
mode and for each cladding mode resonances pair, the EH mode
wavelength increases less than the HE one. The crossing point
occurs for the 0 mode. Past this mode, EH resonances appear on
the short wavelength side of HE ones. This peculiar behavior
results from the fact that EH modes begin to localize in the gold
sheath as they approach the SPR, which lowers their effective
refractive index (due to the small value of the gold refractive
index). HE modes are tangentially polarized at the gold boundary
and hardly penetrate it; therefore, they do not feel this effective
index decrease.
Beyond the crossing, the +2 EH mode now lags significantly
behind the HE mode, which points out to a strong localization of
the EH mode field in the gold layer and the strong influence of the
SPR (and hence of the SRI) on the mode resonance. The next
modes show an even stronger lag of the EH mode but the asso-
ciated resonance also becomes somewhat wider, because of the loss
of energy to the metal. In the case of the +4 mode, the wavelength
60 Christophe Caucheteur et al.

0
-10
-20
n=1.325
-30

Amplitude [dB]
0
-10
-20
n=1.340
-30
0
-10
-20
n=1.360
-30
1525 1530 1535 1540 1545 1550 1555 1560 1565 1570 1575 1580
Wavelength [nm]

Fig. 8 SPR signature in the transmitted spectrum of gold-coated TFBG for coarse
changes in SRI

separation becomes so great that the EH and HE modes are

completely dissociated. And further analysis of the modes in the
vicinity of the SPR becomes meaningless.
Gold-coated TFBGs were immersed in different refractive
index liquids to measure their refractometric sensitivity. Figure 8
shows the transmitted amplitude spectra of a 50-nm gold-coated
TFBG measured for three different refractive index values
measured by a hand-held Abbe refractometer (Reichert AR200
accurate to 10
4 RIU). For such large SRI changes, the SPR
location can be unambiguously located by following the strongest
attenuation in each spectrum. Therefore, the interrogation relies on
the tracking of the wavelength shift of the center of the envelope of
the most attenuated resonances. Using this technique, a linear
response is obtained, as depicted in Fig. 9 and the SRI sensitivity
is ~550 nm/RIU in the range between 1.32 and 1.42.
For high-resolution refractometric sensing over an SRI range
limited to 10
3 typically, accurate measurements of the SPR mode
are not possible from the radially polarized spectrum taken alone.
Indeed, the SPP is only revealed by its absence from the spectrum
and it cannot be reliably measured for wavelength shifts limited to a
few picometers. Hence, different methods have been developed to
track the SPR shift, mainly based on a comparison between both
orthogonally polarized amplitude spectra as reported in our previ-
ous works [9, 10]. In practice, modes slightly off the SPR are used,
because they combine relatively high sensitivity with a narrow
spectral width and they can be “followed” by a combination of
their changes in amplitude and wavelength. Indeed, as they stand
on the shoulder of the SPR envelope, a slight change of the SPP
location yields a modification of the peak-to-peak amplitude of
these modes, as shown below.
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 61

1560
wavelength SPR
Linear fit
1555

1550

Wavelength [nm]
1545

1540

1535
Slope=553nm/RIU
R=99.66%
1530
1.32 1.33 1.34 1.35 1.36 1.37
Surrounding refractive index [RIU]

Fig. 9 SPR wavelength shift as a function of the SRI value

-6

-8
Relative Transmission (dB)

-10

-12

-14

-16

-18
1529.4 1529.6 1529.8 1530.0 1530.2 1530.4 1530.6 1546.0 1546.2 1546.4 1546.6 1546.8 1547.0 1576.8 1577.0 1577.2 1577.4 1577.6
Wavelength (nm)

Fig. 10 Spectra measured during a biosensing experiment (a) with three spectral regions shown in detail: on
the short wavelength side of the SPR (b); near the SPR (+2 mode) (c); and on the long wavelength side (d)

Such high-resolution sensing is demonstrated in Fig. 10 with

spectra measured during a biochemical binding experiment, i.e., for
an SPR shift close to the detection limit (a full report on these
experiments can be found in [12]). By zooming in on resonances
62 Christophe Caucheteur et al.

near the SPR and a few nm away, it becomes clear that the presence
of a comb of resonances with widely different sensitivities allows for
very small wavelength shifts to be detected unambiguously with
high precision. Most often, the focus is made on the +2 mode
among the spectral comb, as it appears to be the most sensitive in
terms of both amplitude variation and wavelength shift.
Optical fiber devices have two important advantages over bulky
prism SPR configurations: light propagates essentially without loss
in short lengths of fibers, resulting in very high signal-to-noise
ratios, and interfacing devices to light sources and detectors con-
sists essentially of plugging connectors into widely available fiber
optic instrumentation, instead of having to carefully align optical
beams through imaging systems. In terms of experimentally
demonstrated FOM, TFBGs exhibit a value reaching 5000, which
surpasses all configurations by more than one order of magnitude
[2]. As shown in Fig. 11, this results from the fact that they exhibit
narrow resonance bands (FWHM ~0.1–0.2 nm) compared to even
the best possible theoretical value (~5 nm obtained by calculating
the reflection from the base of a prism in the Kretschmann-Raether
configuration), also keeping in mind that the experimental SPR
FWHM from other fiber configurations all exceed 20 nm and more.

Fig. 11 Comparison between the best theoretical SPR response for 50 nm gold on
silica in the Kretschmann-Raether configuration (thick blue line) and a measured
TFBG-SPR spectrum with the same thickness of gold (thin red line). The arrows
indicate the resonance to be followed in each case
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 63

3.6 Proteins and Up to now, and to the best of our knowledge, there have been a few
Cells Quantification reports where gold-coated TFBGs were used for biochemical sens-
with Gold-Coated TFBG ing. In [23], the probe was associated with aptamers and a mea-
Immunosensors surement of the dissociation constant was demonstrated. In [24],
the probe was used to measure the intracellular density of non-
physiological cells, namely human acute leukemia cells. In [25], the
sensor was integrated into cell culture equipment and was used for
real-time monitoring of cellular response to chemical stimuli
obtained by adjunction of trypsin, serum, and sodium azide. The
corresponding effects—detachment of cells from the surface,
uptake of serum, and inhibition of cellular metabolism,
respectively—were monitored through a shift of the SPR signature
in the transmitted amplitude spectrum of gold-coated TFBGs.
In the following, we summarize our main achievements toward
the demonstration of lab-on-fiber devices with gold-coated TFBGs
suited for cancer diagnosis [26, 27].

3.7 Sensing A diverse range of tumor markers has been associated with lung
Cytokeratins cancer. In this study, we focused on cytokeratins 7 (CK7) that are
particularly useful tools for diagnosis in oncology. In particular,
CK7 profile of lung tumor has proved to be a useful aid in the
differential diagnosis of carcinomas, since primary and metastatic
tumors present different profiles. In fact, primary lung tumors
express cytokeratin 7 (CK7+) while secondary tumors are deficient
in CK7 (CK7
). Moreover, it has been demonstrated that cytoker-
atin fragments can be released from malignant cells and conse-
quently CK fragments can be located in blood circulation and are
therefore easily accessible with an optical fiber properly modified.
The cytokeratin 7 antigen detection by the bioreceptor is based
on the specific chemical reaction with its corresponding antibody
(AbCK7) previously immobilized on the optical fiber surface. The
operating principle is based on the recognition of the cytokeratin 7
antigen epitope by the fragment antigen-binding (Fab), a region
present on the cytokeratin 7 antibody. In addition to the full protein
(CK7FP) detection made of 469 amino acids, corresponding to a
large biomolecule, we proposed here to monitor a protein fragment
called cytokeratin 7 peptide (CK7pep) made of only 23 amino acids
to present a comparative study depending on the size of the target.
In fact, the detection of small molecule antigens remains a challenge
in the SPR immunosensing. Most of the SPR biosensors are based on
large molecule detection since SPR response in the presence of small
mass suffers from various disadvantages not encountered with full
protein, such as low signal-to-noise ratio.
1. Detection of CK7 full protein.
(a) Gold-coated TFBGs, preliminary functionalized with
cytokeratin 7 antibodies, have been first connected to the
LUNA and held straight between two clamps, for trans-
mission recording (see Note 4).
64 Christophe Caucheteur et al.

(b) The first measurement consisted in the calibration of the

signal. The gold-coated TFBG was immersed in PBS, and
then the polarization controller allowed reaching the
modes s and p, which were the references.
(c) Detection started with the immersion of the gold-coated
TFBG in CK7 full protein solutions diluted in PBS. To
reach the sensitivity of the biosensor, initial measurements
were realized in a phosphate buffer with increasing
CK7FP concentrations from 1E
12 to 1E
6 g/mL.
(d) Recording of the transmission spectrum during experi-
ments put in evidence an evolution of the SPR-TFBG
starting from 3E
12 g/mL until it has reached a thresh-
old at 2E
8 g/mL, as illustrated in Fig. 12a. Such obser-
vations indicate that the biosensor was highly responsive
since our sensitivity, estimated lower than 1 pM, is very
good (see Note 5).
2. Detection of CK7 peptide.
(a) Gold-coated TFBG functionalized with AbCK7 was
plugged in transmission on the LUNA coupled with the
polarizer.
(b) Reference measurement (modes s and p) was performed in
phosphate buffer solution at pH 7.4.
(c) Then gold-coated TFBG was immersed in CK7 peptide
solutions at different concentrations, diluted in PBS at
pH 7.4.
(d) The amplitude monitoring has highlighted a similar trend
than the detection of CK7FP. Nevertheless, differences
between CK7pep and CK7FP detection have been noticed

-10 4

(a) (b)
Biosensor Response (dB)

3
-11
Amplitude (dB)

2
-12

-13

0
1E-12 1E-11 1E-10 1E-9 1E-8 1E-7 CK7 peptide CK7FP Control
Concentration CK7FP (g/mL) Samples in PBS

Fig. 12 Amplitude monitoring of the sensitive mode in the presence of CK7FP at different concentrations
diluted in PBS (a). Biosensor response obtained for optical fiber functionalized with AbCK7, in the presence of
CK7 peptide or CK7FP in comparison with OF functionalized without AbCK7 (b)
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 65

during experiment. First, instead of 3E
12 g/mL, second

the final amplitude measured between the maximal and the
minimal threshold has been determined at 1.0 +/
0.2 dB
against 2.8 dB for CK7FP, as illustrated in Fig. 12b 1E
9 g/
mL.
3. Detection of CK7 peptide in complex media.
(a) The gold-coated optical fiber biosensor was performed on
phosphate buffer complemented with 10% fetal bovine
serum (FBS), resulting in a complex buffer full of proteins
including albumin, enzymes, and others; as well as vita-
mins and nutrients. Objectives were to get close to physi-
ological conditions on the one hand, and to guarantee the
selectivity of the biosensor on the other hand. In the
presence of FBS in the media, the optical fiber ended up
with numerous compounds able to attach on the surface.
(b) The functionalized gold-coated TFBG was tested on CK7
peptide solution diluted in PBS complemented with 10%
FBS, and the SPR-TFBG spectra were recorded in
transmission.
(c) The amplitude monitoring, as presented in Fig. 13a,
clearly shows the evolution of SPR-TFBG signal as soon
as the CK7pep was injected until it has reached a minimum
threshold at 1E
7 g/mL. Figure 13b presents the average
output and errors for a three-fold repeated CK7pep detec-
tion on distinct optical fibers prepared on the exact same
conditions. All three fibers have presented the identical
amplitude evolution depending on the CK7 concentra-
tion, including an amplitude shift from 1E
9 g/mL,
which identifies that the lowest concentration detected is
estimated at 0.4 nM.

-6
(b)
3
-8
Biosensor Response (dB)
Amplitude (dBm)

1E-9 g/mL
2E-9 g/mL
-10 3E-9 g/mL
4E-9 g/mL
2
5E-9 g/mL
1E-8 g/mL
2E-8 g/mL
-12 3E-8 g/mL
4E-8 g/mL
5E-8 g/mL
6E-8 g/mL
7E-8 g/mL
-14 1E-7 g/mL
1
2E-7 g/mL
5E-7 g/mL
1E-6 g/mL

-16
(a) 0
1545.8 1546.0 1546.2 1546.4 1546.6 1546.8 1547.0 CK7 peptide diluted in PBS + serum Control PBS + serum
Wavelength (nm)

Fig. 13 Shift of the sensitive mode during incubation of the optical fiber functionalized with AbCK7 on CK7
peptide at different concentrations diluted in PBS + 10% FBS (a). Biosensor response obtained for optical fiber
functionalized with AbCK7 immersed in PBS + 10% FBS, in the presence or in the absence of CK7 peptide (b)
66 Christophe Caucheteur et al.

3.8 Sensing In this section, we report on the use of gold-coated TFBGs for
Transmembrane selective cellular detection through membrane protein targeting.
Receptors The focus is made on the epithelial growth factor receptor (EGFR),
which is a transmembrane receptor from the 4-tyrosine kinase
receptors family. It is an important biomarker and therapeutic
target that it is over-expressed by numerous cancer cells.
1. Surface functionalization.
The sensor surface selectivity was ensured by bio-
functionalization through a two-step approach.
(a) First, the clean waveguide surface was activated with the
Carboxylic acid-SAM formation reagent (cat. n: C488)
obtained from Dojindo (Japan).
(b) Then, monoclonal mouse immunoglobulin G (IgG)
raised against the human epidermoid carcinoma cell line
(Santa Cruz Biotechnology Inc. and American Type Cul-
ture Collection (USA)) were immobilized on the surface
through carbodiimide covalent biochemistry using Dojin-
do’s amine coupling kit (cat. n: A515-10). This antibody
was diluted in the Dojindo’s reaction buffer to 0.01 μg/
mL for efficient covalent immobilization during 30 min.
The manufacturer’s instructions were followed for each
step for both processes.
2. EGFR detection by SPR-TFBG.
Two cell lines were used: the first one with overexpressed
human epidermal growth factor receptors (EGFRs)—A431
cell line, (EGFR positive, EGFR (+)). The second cell line
was EGFR negative – OCM1 cell line, (EGFR (
)).
(a) The culture process was done at 37  C in a humidified
incubator with 5% CO2 atmosphere. No antibiotic was
used and cell cultures were free of mycoplasma and patho-
genic viruses. Cells suspensions were then prepared for
SPR experiments. For this, A431 and OCM1 cells from a
confluent monolayer were detached mechanically by a
gentle scraping of cells from the growth surface into ice-
cold phosphate buffered saline (PBS). Cells were then
washed three times in cold PBS by repeated centrifugation
(at 2500 RPM and 4  C for 3 min). Pellets were then again
suspended to a concentration of 2–5  106 cells/mL.
0.5 mL volumes were used for experiments. A handheld
automated cell counter from Millipore (Scepter 2.0,
PHCC00000) was used to count the cells.
(b) For immunosensing experiments, gold-coated TFBGs were
hold straight between two clamps and were then
approached toward the cells suspensions, using a vertical
translation stage. To demonstrate that a differential
response can be obtained between both cell lines, TFBGs
Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 67

were first immersed in RPMI-1640 cell culture media (Ros-

well Park Memorial Institute (RPMI) 1640, supplemented
with 10% FBS—refractive index ¼ 1.3366), until a stable
baseline was reached.
(c) In the first step, sensors were incubated in a suspension of
OCM1 during ~10 min. After this assay, a secondary
baseline of the media (RPMI-1640) was recorded.
(d) In the second step, sensors were incubated in a suspension
of A431 during the same period. Finally, the third media
baseline was recorded before, and after an extra rinsing
step with the media.
(e) The conditioning in media was 5 min and assays lasted
10 min. Figure 14 shows the processed data (evolution of
the peak-to-peak amplitude of the +2 mode) obtained
after such binding experiments. It clearly figures out that
an important response change is obtained with the A431
cell line (here at a concentration of 3  106 cells/mL), for
which EGFRs are overexpressed. This is not the case for
the other cell line for which the response remains compa-
rable to the background noise level obtained in RPMI-
1640 media. Error bars are the standard deviation
obtained for three experiments made in the same
conditions.

Fig. 14 Different assays presenting specific cell interactions. Dark gray bars
represent A431 cells, light gray bars represent OCM1 cells. Blank bars are
washing and rinsing steps for RPMI-1640 cell culture media supplemented
with FBS
68 Christophe Caucheteur et al.

Table 1
Sensor response as a function of the cells concentration, bringing an
estimate about the LOD

Cells conc. (106 ml1) Response (%)

0.5 N/A
1.1 N/A
2.0 6.6
2.9 20.0
3.5 23.3
5.1 90.0

(f) Different cells suspensions were used with various concen-

trations to estimate the LOD of the sensors. Table 1 sum-
marizes the obtained results in terms of sensor response,
which is here computed as the relative amplitude change
of the +2 mode with respect to its value measured in the
reference solution (RPMI medium). It shows that the
LOD is ~2  106 cells/mL, which is relevant with respect
to the target application. Obviously, the sensor response is
not linear with the cells concentration. The saturation has
not been properly measured but it is beyond
~5  106 cells/mL.
Optical fiber-based biochemical sensors fill an increasingly well-
defined niche as they do not require elaborate light management
schemes to probe molecules and materials. Light remains guided in
the fiber from the source to the detector, apart from localized probe
regions where it interacts with the immediate surroundings of the
fiber. This feature is the dominant factor in making such sensors less
expensive to fabricate and to use than conventional biosensors
(especially those based on SPR and resonant waveguide gratings).
However, this comes with two main inconveniences: a higher
(worse) limit of detection and the impossibility to scale up toward
massively parallel testing.
Our work features a robust biosensing platform with a simple
four-step protocol enabling label-free cells sensing. It addresses the
first issue by combining SPR effects with a grating-based approach
in a single sensor that keeps all the advantages of the fiber solution.
TFBGs provide a dense spectral comb of high Q-factor resonances
that probe the SPR envelope and allow an order of magnitude
improvement in the determination of the SPR shifts under bio-
chemical interactions at the fiber surface. We have shown that we
can separately excite resonances that enable SPR generation and
others that do not (which can thus serve as reference channels).
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 69

The sensor also provides absolute temperature information, from

the wavelength position of Bragg core resonance that does not
sense changes of the surrounding refractive index medium. This
additional information can be used to average out noise and to
eliminate many cross-sensitivity factors. We have also demonstrated
that our biosensing platform can be used to investigate the binding
of living human epithelial cells through specific interaction of
transmembrane proteins and target extracellular membrane recep-
tors in biological matrices.
The strength of our platform arises from its near-infrared
operating wavelength, thus allowing improved penetration depth
and full compatibility with telecommunication-grade equipment,
including switch matrix devices that allow automatic interrogation
of several tens of optical fibers in relatively rapid succession. Hence,
different fibers coated with different bioreceptors could be easily
interrogated to assay several parameters simultaneously, which
would solve the second aforementioned limitation. Additionally
for practical use, it could considerably enhance the reliability of a
diagnosis.
Much of the development required for the exact quantification
and deployment of practical sensors based on gold-coated TFBGs is
still in progress. Therefore, it is hoped that the achievements pre-
sented herein will stimulate further research in this area.

4 Notes

1. Sulfur atoms of the SAM attach within minutes on the gold

substrate; nevertheless, carbon chain requires hours to be
arranged via van der Waals forces, in the aim to form a packed
and stable monolayer.
2. Polyclonal antibodies were used instead of monoclonal to
improve the protein detection, since polyclonal antibodies
react with multiple epitopes of the protein.
3. SPR-TFBG experiments were realized in a constant volume
(500 μL) in a constant temperature.
4. Recording of the amplitude spectra was realized after the stabi-
lization of the signal.
5. Proteins detection measurements require at least triplicates
experiments to obtain statistics. Column bars in Figs. 11 and
12 represent the average of the amplitude shift while the error
bars correspond to the standard deviation of three experiments
realized in the same conditions.
70 Christophe Caucheteur et al.

References
1. Kretschmann E, Raether H (1968) Radiative 13. Othonos A, Kalli K (1999) Fiber Bragg gratings:
decay of non radiative surface plasmon excited fundamentals and applications in telecommuni-
by light. Z Naturforsch 23:2135 cations and sensing. Artech House, London
2. Caucheteur C, Guo T, Albert J (2015) Review 14. Erdogan T (1997) Fiber grating spectra.
of recent plasmonic fiber optic biochemical J Lightwave Technol 15:1277–1294
sensors: improving the limit of detection. 15. Erdogan T, Sipe JE (1996) Tilted fiber phase
Anal Bioanal Chem 407:3883 gratings. JOSA A 13:296–313
3. Pollet J, Delport F, Janssen KPF, Jans K, Maes 16. Laffont G, Ferdinand P (2001) Tilted short-
G, Pfeiffer H, Wevers M, Lammertyn J (2009) period fiber Bragg grating induced coupling to
Fiber optic SPR biosensing of DNA hybridiza- cladding modes for accurate refractometry.
tion and DNA-protein interactions. Biosens Meas Sci Technol 12:765–772
Bioelectron 25:864–869 17. Caucheteur C, Mégret P (2005) Demodula-
4. Baldini F, Brenci M, Chiavaioli F, Gianetti A, tion technique for weakly tilted fiber Bragg
Trono C (2012) Optical fiber gratings as tools grating refractometer. Photon Technol Lett
for chemical and biochemical sensing. Anal 17:2703–2705
Bioanal Chem 402:109–116 18. Chan CF, Chen C, Jafari A, Laronche A, Thom-
5. Albert J, Shao L-Y, Caucheteur C (2013) son DJ, Albert J (2007) Optical fiber refractom-
Tilted fiber Bragg grating sensors. Laser Pho- eter using narrowband cladding-mode
ton Rev 7:83–108 resonance shifts. Appl Optics 46:1142–1149
6. Sharma AK, Rajan J, Gupta BD (2007) Fiber- 19. Chen C, Albert J (2006) Stain-optic coeffi-
optic sensors based on surface plasmon reso- cients of individual cladding modes of single
nance: a comprehensive review. IEEE Sensors J mode fibre: theory and experiment. Electron
7:1118–1129 Lett 42:1027–1028
7. Offermans P, Shaafsma MC, Rodriguez SRK, 20. White IM, Fan XD (2008) On the performance
Zhang Y, Crego-Calama M, Brongersma SH, quantification of resonant refractive index sen-
Rivas JG (2011) Universal scaling of the figure sors. Opt Express 16:1020–1028
of merit of plasmonic sensors. ACS Nano 21. Shevchenko YY, Chen C, Dakka MA, Albert J
5:5151–5157 (2010) Polarization selective grating excitation
8. Shevchenko YY, Albert J (2007) Plasmon reso- of plasmons in cylindrical optical fibers. Opt
nances in gold-coated tilted fiber Bragg grat- Lett 35:637–639
ings. Opt Lett 32:211–213 22. Caucheteur C, Chen C, Voisin V, Berini P,
9. Caucheteur C, Shevchenko YY, Shao L-Y, Albert J (2011) A thin metal sheath lifts the
Wuilpart M, Albert J (2011) High resolution EH to HE degeneracy in the cladding mode
interrogation of tilted fiber grating SPR sensors refractometric sensitivity of optical fiber sen-
from polarization properties measurement. sors. Appl Phys Lett 99:041118
Opt Express 19:1656–1664 23. Shevchenko Y, Francis TJ, Blair DAD, Walsh R,
10. Caucheteur C, Voisin V, Albert J (2013) DeRosa MC, Albert J (2011) In situ biosensing
Polarized spectral combs probe optical with a surface plasmon resonance fiber grating
fiber surface plasmons. Opt Express aptasensor. Anal Chem 83:7027–7034
21:3055–3066 24. Guo T, Liu F, Liu Y, Chen NK, Guan BO,
11. Albert J, Lepinay S, Caucheteur C, DeRosa Albert J (2014) In situ detection of density
MC (2013) High resolution grating-assisted alteration in non-physiological cells with polar-
surface plasmon resonance fiber optic aptasen- imetric tilted fiber grating sensors. Biosens
sor. Methods 63:239–254 Bioelectron 55:452–458
12. Voisin V, Pilate J, Damman P, Mégret P, Cau- 25. Shevchenko Y, Camci-Unal G, Cuttica DF,
cheteur C (2014) Highly sensitive detection of Dokmeci MR, Albert J, Khademhosseini A
molecular interactions with plasmonic optical (2014) Surface plasmon resonance fiber sensor
fiber grating sensors. Biosens Bioelectron for real-time and label-free monitoring of cellu-
51:249–254 lar behavior. Biosens Bioelectron 56:359–367
 Immunosensing with Near-Infrared Plasmonic Optical Fiber Gratings 71

26. Ribaut C, Voisin V, Malachovska V, Dubois V, 27. Malachovska V, Ribaut C, Voisin V, Surin M,
Mégret P, Wattiez R, Caucheteur C (2016) Leclère P, Wattiez R, Caucheteur C (2015)
Small biomolecule immunosensing with plas- Fiber-optic SPR immunosensors tailored to
monic optical fiber grating sensor. Biosens target epithetial cells through membrane
Bioelectron 77:315–322. doi:10.1016/j.bios. receptors. Anal Chem 87:5957–5965
2015.09.019
 Chapter 5

Biosensing Based on Magneto-Optical Surface

Plasmon Resonance
Sorin David, Cristina Polonschii, Mihaela Gheorghiu, Dumitru Bratu,
and Eugen Gheorghiu

Abstract
In spite of the high analytic potential of Magneto Optical Surface Plasmon Resonance (MOSPR) assays,
their applicability to biosensing has been limited due to significant chip stability issues. We present novel
solutions to surpass current limitations of MOSPR sensing assays, based on innovative chip structure,
tailored measurements and improved data analysis methods. The structure of the chip is modified to contain
a thin layer of Co-Au alloy instead of successive layers of homogenous metals with magnetic and plasmonic
properties, as currently used. This new approach presents improved plasmonic and magnetic properties, yet
a structural stability similar to standard Au-SPR chips, allowing for bioaffinity assays in saline solutions.
Moreover, using a custom-designed measurement configuration that allows the acquisition of the SPR
curve, i.e., the reflectivity measured at multiple angles of incidence, instead of the reflectivity value at a
single-incidence angle, a high signal-to-noise ratio is achieved, suitable for detection of minute analyte
concentrations. The proposed structure of the MOSPR sensing chip and the procedure of data analysis
allow for long time assessment in liquid media, a significant advancement over existing MOSPR chips, and
confirm the MOSPR increased sensitivity over standard SPR analyses.

Key words Magneto-optic surface plasmon resonance, Magnetic alloys, Surface plasmon resonance
enhancement, Affinity biosensor, Angle-resolved surface plasmon resonance, Fixed-angle surface
plasmon resonance

1 Introduction

Surface Plasmon Resonance is a phenomenon that appears when a

plane-polarized light is coupled, under total internal reflection
(TIR) conditions, into a metal film at the interface between two
media. The TIR conditions are met when the incident light beam
intersects the interface between two adjacent media that have dif-
ferent refractive indices (n1 > n2) at an angle θ greater than the
critical angle θc defined by Eq. 1 derived from the Snell law,

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_5,
© Springer Science+Business Media LLC 2017

73
74 Sorin David et al.

Fig. 1 Principle of surface plasmon resonance in Kretschmann configuration

n2
θc ¼ ð1Þ
n1
where n2 is the refractive index of the upper medium and n1 is the
refractive index of the lower medium and the light travels from n1
toward n2.
The coupling of the light into the metal film is done via a
grating (Otto configuration) or a prism (Kretschmann configura-
tion) [1]. In the following, only the Kretschmann configuration is
considered.
Depending on the value of the incident light angle θ, the type
of the plasmonic metal, and the refractive indices of the two media,
the light interacts with the free electrons of the metal film and the
energy of the photons is transferred resonantly to these electrons. A
drop in the intensity of the reflected light is recorded by a suitable
detector (e.g., a photodetector array) placed in the path of the
reflected beam—Fig. 1 corresponding to the specific SPR angle.
The dependency of the intensity of the reflected light over the
incidence angle generates the SPR curve that has a dip minimum
corresponding to the SPR angle, θmin (Fig. 1). According to Eq. 2
[2], the SPR angle depends on λ (the wavelength of the incident
light), via n*m, the complex refractive index of the plasmonic layer
and the values of the refractive indices of the media n1 and n2.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
n∗2  n22
sin θmin ¼  ∗2m  ; n∗ ¼ nm þ i  k m : ð2Þ
nm þ n22 n21 m

Considering n*m and n1 constant (as part of the measuring

setup, for a given wavelength of the incident light), the SPR angle
Biosensing Based on Magneto-Optical Surface Plasmon Resonance 75

depends exclusively on n2—the refractive index of the upper

medium. As such, due to their high sensitivity to local changes in
the refractive index of the upper media [1], SPR biosensing assays are
currently among the methods of choice to assess bioaffinity interac-
tions at interfaces. For example, the minimum resolvable refractive
index change for the SPREETA sensor (Texas Instruments Inc.)
used in this chapter is 1  106 RIU (refractive index units) [3].
Nevertheless, there is a continuous strive to improve the sensi-
tivity of the existing SPR analysis. Possible avenues for improve-
ment include optimization of the plasmonic layer into a
multilayered structure, improvement in the sensitivity of photo-
detectors [1], or signal modulation [4, 5]. Promising signal modu-
lation approaches are based on the magneto plasmonic effects [6–8]
by employing the transverse magneto-optic Kerr effect (TMOKE).
The TMOKE occurs as a result of applying a magnetic field per-
pendicular to the propagation plane of the incident p-polarized
light versus a surface with magneto-optical properties. This leads
to modulation of the wave vector resulting in changes in the
intensity of the reflected light [9]. The process lays the foundation
of the novel magneto-optical SPR (MOSPR) method [10].
In the MOSPR approach the modulation of the reflectivity
(i.e., SPR) curve is accomplished by applying an alternating trans-
versal magnetic field, perpendicular to the propagation plane of an
p-polarized beam of light [9] incident via a prism coupler, onto a
sensor chip exhibiting both magnetic and plasmonic properties (so-
called MOSPR chip [4]). MOSPR response measured at a single
angle of incidence yields improved sensitivity to the refractive index
changes of up to two orders of magnitude, compared to classical
SPR assays, as substantiated by both theoretical and experimental
studies [2, 9–12].
Important drawbacks of the technique are given by the tech-
nologically challenging MOSPR chip fabrication involving metallic
multilayers or nanostructures (e.g., nanodiscs, nanocrystals) [6–8]
and the intrinsic magnetostrictive effect [5] that causes ferromag-
netic materials to change their shape or dimensions during the
process of magnetization and adds further stability concerns.
Indeed, this complex, multilayered structure of the MOSPR sens-
ing chips is prone to induce a lack of stability when chips are
exposed to (saline) liquid samples and, as a result, current highly
sensitive MOSPR assays are mostly gas sensors [10, 12, 13].
To overcome this barrier, a multilayered chip comprising a thin
film of amorphous Au-Co alloy, capped with a layer of Au (to allow
further functionalization) with optimized structure for improved
magneto-plasmonic properties, provides a powerful alternative.
This structure, recently proposed, exhibits both plasmonic and
magnetic properties, and proves excellent stability in (saline) liquid
media [14].
76 Sorin David et al.

In a further improvement, we used the optimized MOSPR

structure in an angle-resolved MOSPR bioassay (that allows the
evaluation of the reflectivity curve in a wider angular range, instead
of the reflectivity at a single (fixed) angle of incidence as employed
in most of the current SPR/MOSPR experiments [13]) and con-
firmed experimentally both increased sensitivity in comparison with
the current SPR/MOSPR approaches and a high stability (similar
to the one achieved with classic, gold-only, SPR chips) when
addressing saline liquid samples.

2 Materials

2.1 Surface 1. Bovine serum albumin (BSA), human IgG (HIgG), affinity
Chemistry isolated anti-human IgG (AHIgG), N-hydroxysuccinimide
(NHS), 1-ethyl-3-(dimethylaminopropyl) carbodiimide
(EDC), ethilenediaminetetraacetic acid (EDTA), and ethanol-
amine were purchased from Sigma–Aldrich (Germany).
2. (1-mercapto-11-undecyl)hexa(ethylene glycol)carboxylic acid
was purchased from Prochimia Surfaces, Poland (see Note 1).
3. Surfactant P20 was provided by GE Healthcare.
4. Matching oil Immersol 518 F, ne ¼ 1.518 from Carl Zeiss.
5. Aqua regia HCl:HNO3 3:1 mixture (see Note 2).
6. piranha solution H2SO4:H2O2 3:2 mixture (see Note 3) .

2.2 Preparation of 1. Au pellets (99.999%), Co (99.99%) pellets, and Ti sputter

Magneto-Optical target (grade 2) were purchased from Kurt J. Lesker USA.
Sensor Chip 2. N-BK7 glass wafers were purchased from Sydor Optics.
3. Kurt J. Lesker PVD 75 system with substrate heater, thermal
evaporation sources, and reactive sputtering source for metallic
layer deposition.
4. PSD–UV–plasma cleaner from Novascan Technologies, Inc.

2.3 Working 1. immobilization buffer: 10 mM acetate buffer pH 5.5.

Solutions 2. running buffer: HBS-EP buffer (10 mM HEPES, pH 7.4,
150 mM NaCl, 3 mM EDTA, 0.005% Surfactant P20).
Ultrapure water (Millipore) was used throughout the
preparations.

2.4 Measurement 1. Fluidic assembly.

Setup 2. SPR module based on SPREETA sensor model TSPR1K23
and custom data acquisition board with appropriate software.
3. Electromagnet with power supply for applying the oscillating
magnetic field.
4. PC for data saving and processing.
 Biosensing Based on Magneto-Optical Surface Plasmon Resonance 77

2.5 Fluidics 1. Injection pumps—Cavro XLP 600 Tecan Inc.

Assembly 2. PTFE tubes ID 0.3 mm and fittings from Upchurch Scientific.
3. Polyether ether ketone (PEEK) slab micromachined as flow
chamber.
4. PDMS silicone—Sylgard 184 from DowCorning as gasket
between the flow chamber and the (MO)SPR chips.

3 Methods

3.1 Fabrication of 1. Substrates of BK7 glass (0.3 mm  4 mm  11 mm) are

MOSPR Chips cleaned by successive sonication in water, acetone, and isopro-
panol and dried in a stream of nitrogen.
2. The substrates are exposed to ozone plasma for 10 min using a
PSD-UV plasma cleaner.
3. Substrates are loaded in the pressure chamber of the PVD 75
system.
4. The pressure chamber is pumped to a pressure of approxi-
mately 2  106 Torr.
5. Samples are heated to 150  C for degassing for 30 min using
the substrate heating facility of the PVD 75.
6. Samples were rotated at 10 rot/min to ensure uniform coating.
All the deposition processes were done at room temperature.
7. Thickness and deposition rate of the deposited layers are con-
stantly monitored by a quartz crystal microbalance (QCM)
inside the evaporation chamber of the PVD 75. The QCM
works by measuring the changes in the oscillation frequency
of a quartz crystal as a result of adsorption of materials on the
crystal surface. The PVD 75 QCM is calibrated in terms of
adsorbed material (taking into account the elastic properties of
the material) and position in relation to the substrate.
8. A layer of Ti (4 nm) is first sputtered on the glass substrates for
improving the adhesion of subsequent layers, at 5  103 Torr
Ar pressure and 8 W/cm2 (see Note 4).
9. Standard SPR chips are fabricated by evaporating 50 nm of Au
onto the Ti modified substrates (see Note 5).
10. The Au and Co layers are thermally evaporated at rates of
0.5 Å/s and 3 Å/s respectively (see Note 6).
11. The Au-Co-Au tri-layers consist in a 26 nm Au, 6 nm Co layer,
separated from the sensing surface by a 12 nm Au layer. An
intermediate layer of 2 nm Ti is inserted between Au and Co to
improve adhesion of the two metals.
12. The Au-Co alloy chips are fabricated by evaporating 30 nm
layer of alloy Au-Co followed by 15 nm of Au. No intermediate
Ti layer is used.
78 Sorin David et al.

13. The Au-Co alloy is thermally evaporated at a rate of 0.5 Å/s from
a single tungsten boat containing corresponding quantities of Au
and Co to yield an alloy of 10% Co—volumetric ratio (588 mg Au
and 27 mg Co). Alternative techniques for producing thin films
of magnetic materials and alloys do exist and may be used for a
better control of the film composition and structure. However,
they have limitations in respect to costs (see Note 7), or material
type (see Note 8) .

3.2 Chips 1. (MO)SPR are cleaned by immersion in piranha solution for

Functionalization and 10 min and washed with deionized water and ethanol.
Bioaffinity Assays 2. Cleaned (MO)SPR chips are functionalized by immersion in
ethanol solution containing 1 mM of 11-PEG-COOH thiol
for 3 h.
3. Functionalized chips are mounted in the flow chamber of the
measurement setup and further modified with the ligand
(HIgG) by amine coupling [15]:
(a) A mixture of 100 mM NHS and 400 mM EDC is passed
over the chip for 7 min to activate the carboxylic groups
of the thiol.
(b) The ligand (30 μg/ml) is injected for 30 min in 10 mM
acetate buffer at pH 5.5 to promote electrostatic pre-
concentration.
4. The modified surface is deactivated for 7 min with 1 M etha-
nolamine pH 8.6.
5. The deactivated surface is blocked and tested for nonspecific
adsorption by injecting a solution of 1 mg/ml BSA in HBS-EP
(10 min).
6. All the immobilization procedure is performed at a flow rate of
30 μl/min with HBS-EP as running buffer.

3.3 (MO)SPR For classical SPR measurements the response is derived from the
Measurements and SPR curve as a function the angle for which the minimum of the
Data Processing reflectivity is recorded [1]. The reflectivity (Rpp) is obtained by
normalizing the intensity of the measured reflected p-polarized
light versus a reference signal (e.g., obtained using an s-polarized
light or a solution with an SPR dip outside the measurement range
of the device). As pinpointed by Eq. 2, the SPR angle (θmin),
depends on the refractive index of the media in the immediate
vicinity of the sensing surface [1] and is determined most precisely
if the reflectivity dip is the sharpest. Changes at the sensing surface,
accompanied by a modification of the refractive index, cause a shift
in the SPR angle that is used for label-free assessment of biomolec-
ular surface interactions. This is the preferred mode of operation for
real-time measurements having a linear dependence on the refrac-
tive index changes and with the amount of molecules accumulated
 Biosensing Based on Magneto-Optical Surface Plasmon Resonance 79

Fig. 2 The magneto optical SPR concept and schematic setup: electromagnet for transverse oscillating
magnetic field application; tailored chips with magneto-plasmonic layers; top microfluidics; all system
components were custom developed and built around an integrated illumination and detection Kretschman
configuration as provided by a Spreeta module

on the surface [1]. Based on the full curve analysis, various data
processing methods were developed to improve sensitivity, robust-
ness against noise [16] and foster reduced time of analysis [17].
Figure 2 outlines the MOSPR concept. In brief, the measure-
ment setup, described in more detail elsewhere [18], comprises:
– A SPR module.
– An external electromagnet with power supply for applying the
alternating transverse magnetic field.
– A Flow Injection Analysis (FIA) type setup as fluidic assembly.
The SPR measurement module is based on a SPREETA sensor
model TSPR1K23 (Texas Instruments USA). The FIA comprises a
flow channel connected to an injection pump using PTFE tubes.
The flow channel is made by micromachining a polyether ether
ketone (PEEK) slab and a gasket of PDMS (Fig. 2). The data
acquisition board is designed to give online access to whole curve
measurements to allow calculation/extraction of the MOSPR data.
80 Sorin David et al.

In the Kretschmann configuration [2] of the SPR module, the

backside of the chip is illuminated using an embedded divergent
p-polarized beam of light in a continuous range of incidence angles,
spanning the 61–73 angular domain. The reflected light is assessed
by a linear photo detector array comprising 128 pixels, each pixel
measuring the intensity of the reflected light corresponding to a
different angle of incidence. This allows for virtually simultaneous
acquisition of angle specific reflectivity data, the whole SPR curve
being measured in ~5 ms. The wavelength used is in the near infra-
red region of the spectrum (840 nm) and provides a rather sharp
dip and large penetration depths for standard SPR Au chips
measured in aqueous media. The relation between the individual
pixels of the photo detector array and the corresponding angle of
incidence θ, important especially when comparing SPR experimen-
tal data with numerical simulations provided by the Transfer Matrix
approach, can be derived using the manufacturing parameters of
the sensor. In the case of the TSPR1K23 sensor geometry, we find
[14, 19]:
θ ¼ U max  ðU max  U min Þ=PixelRange  PixelPosition, ð3Þ

where Umax ¼ 73.427 and Umin ¼ 60.425 are, respectively, the

maximum and the minimum angle achieved in the SPREETA
sensor, PixelRange ¼ 128 is the total number of pixels of the
photodetector array and PixelPosition is the value of the position
of the individual pixel on the detector array, for which the angle θ is
calculated.
For attaching tailored sensor chips to the SPREETA prism, a
refractive index matching oil is used on the glass window of the
sensor priory cleaned with aqua regia [20].
Calibration measurements are performed for each individual
chip and relate the SPR responses (i.e., the variations of the
measured SPR angle) to the effective refractive index values and
involve assessment of water and glycerol solutions of various con-
centrations with known refractive index. The calibration process
comprises injections in the measurement flow channel of solutions
of known refractive index. The SPR response is recorded and the
SPR angle is related to the refractive index of the calibration solu-
tion. Using several solutions with different refractive indices in the
domain of interest, one can derive specific calibration curves for
the SPR angle as a function of the refractive index of the solution in
the flow channel. Further, the SPR analytic responses may be
expressed as refractive index variations versus time (i.e., sensor-
grams) or as concentrations of analyte (i.e., calibration curves).
MOSPR measurements are performed using tailored MOSPR
sensor chips and an electromagnet placed transversal on the outside
of the measurement chamber actuated with a sinusoidal voltage.
The voltage is applied using a custom-built alternative current
 Biosensing Based on Magneto-Optical Surface Plasmon Resonance 81

source with a frequency domain of 0.1–100 Hz and a maximum

current of 1 A. Field strengths of up to 5 mT (demonstrated as
being sufficient for saturation of the magneto-optical chips [14])
were achieved using 120 mA, at the frequency of 0.3 Hz, current
for electromagnet actuation.
In the typical experiment, reflectivity values are recorded when
there is no magnetic field (Rpp) as well as for each direction of
magnetization (M+) and (M) when electromagnet actuation is
performed.
Applying the alternating magnetic field did not create signifi-
cant heating of the electromagnet nor had an effect over the SPR
measurement on a classical gold surface, proving that there is no
crosstalk between the magnetic field and the electronic and optic
components of the SPREETA chip.

3.3.1 Extracting the The MOSPR signal is calculated from the reflectivity curves
MOSPR Data recorded in different states of magnetization of the MOSPR chip
and is defined as ΔRpp/Rpp [11]. Here, Rpp is the reflectivity
(standard for SPR assays), obtained when not applying a magnetic
field and ΔRpp represents the difference between the reflectivity
measured at one direction of magnetization (M+) and the reflectiv-
ity measured at the opposite direction of magnetization (M).
Figure 3a indicates the reflectivity curves in respect with magneti-
zation as well as the corresponding ΔRpp, Rpp, Δθ (the difference
between the SPR angle measured at M+ and M respectively) and
θmin (the SPR angle measured with no magnetic field) data.
The MOSPR response is derived from the reflectivity values
corresponding to a selected angle of incidence, as the variation of
the reflectivity Rpp while switching the direction of the in plane
magnetization of the sensor [10]:
ΔRpp Rpp ðM þÞ  Rpp ðM Þ
¼ ð4Þ
Rpp ð0Þ Rpp ð0Þ
where Rpp(M+) and Rpp(M) represent the reflectivities for the
direction of the magnetization perpendicular to the propagation
plane of the incident p-polarized light in both ways and Rpp(0)
represents the reflectivity without magnetization. In real measure-
ments, ΔRpp is obtained from the amplitude of the oscillations of
the measured reflectivity (Fig. 3a). The MOSPR curve is defined as
the variation of the ΔRpp/Rpp versus the angle of incidence of the
light.
As shown in Fig. 3b, for noise-free data a change in the refrac-
tive index of the media that yields a small reflectivity change in the
SPR signal, determines a larger change in MOSPR signal. To this
effect, an intermediary smoothing step is applied to the raw data.

3.3.2 Fitting Function of Reflectivity data for the entire angular range (SPR curves) are
SPR Data collected and fitted with a rational polynomial equation to achieve
smoothing.
82 Sorin David et al.

Fig. 3 (a) Representative parameters of the measured SPR curves used to derive the MOSPR data: Rpp the
reflectivity value for a selected incidence angle; ΔRpp corresponds to reflectivity differences between M+ and
M magnetization states, Δθ (the difference between the SPR angles for M+ and M magnetization states)
and θmin (the SPR angle measured with no magnetic field). (b) Schematic plot showing ideal SPR (black) and
MOSPR (blue) curves, as well as their variations corresponding to refractive index n (dotted lines) and n + Δn
(continuous lines), respectively

For fitting the experimental reflectivity data the following func-

tion is proposed [14, 21]:
a0 þ a1 x þ a2 x 2 þ a3 x 3 þ a4 x 4 þ a5 x 5 þ a6 x 6
f ðx Þ ¼ ð5Þ
b2x 4 þ b1x 2 þ 1

Here, x denotes the pixel position value and is related to the

angle of incidence via Eq. 3, and a0  a6, b1 and b2 are the
polynomial coefficients of the fitting function.
The good match of the rational polynomial function to the
theoretical curve is presented in Fig. 4. A theoretical SPR curve is
generated using the Transfer Matrix approach considering the liter-
ature optical constants of the standard structure comprising a glass
prism (BK7 ε0 ¼2.2801), covered with 4 nm Ti adhesion layer
(ε0 ¼ 1.7982; ε00 ¼ 20.0613) and 50 nm gold layer (ε0 ¼ 30.2270;
ε00 ¼2.1235) corresponding to λ ¼ 840 nm. The theoretical curve is
sampled to 128 points similar to the data received from the SPR
detector (128 pixels). A χ 2 ¼ 8.9  109 residual error has been
obtained for the fit, proving relatively insensitive to noise.
The fitted SPR curves provide corrected reflectivity values for
each incident angle and are used to calculate the MOSPR curves.
These smoothened MOSPR curves are used for further processing.
 Biosensing Based on Magneto-Optical Surface Plasmon Resonance 83

Fig. 4 Theoretical SPR curve for 50 nm layer of gold on glass with water on top
(black dotted line) and the fit with the phenomenological, polynomial function (red
line)

3.3.3 Sensitivity The MOSPR responses are compared with SPR responses in terms
Comparison of MOSPR and of sensitivity and signal-to-noise ratio when analyzing the reflectiv-
SPR Sensors ity at a given, single angle of incidence.
According to Eq. 4 the MOSPR signal is derived from
ΔRpp(θ)/Rpp(θ) data using the value of the reflectivity
corresponding to a selected angle of incidence θs, chosen to
yield the highest sensitivity. At this angle, the MOSPR curve (i.e.,
ΔRpp/Rpp vs. incidence angle) presents the steepest slope [10]. We
derive the θs angle by analyzing the derivative of the fitted SPR
curve. Time series of full SPR curves are recorded, and then fitted
with Eq. 5 and the reflectivity values (corresponding to θs) are
plotted over time. Further, frequency (Fourier) analysis is per-
formed to extract the amplitude of oscillations (corresponding to
ΔRpp) and their mean (corresponding to Rpp). MOSPR and SPR
reflectivity responses derived from raw measurements directly from
the detector corresponding to θs are presented in Fig. 5a while
reflectivity responses calculated from fitted SPR curves are given
in Fig. 5b. The SNR is calculated as the ratio between the response
(SPR, MOSPR) and the corresponding root mean square deviation
(RMSD  calculated on 100 data points). The comparison shows
that fitting the curves (and hence reducing the noise) improves
greatly the sensitivity (Fig. 5b). The sensitivity increase for the
84 Sorin David et al.

Fig. 5 (a) Signal-to-noise ratios for Au—SPR (red), Au-Co alloy—MOSPR (black) and Au-Co-Au tri-layer—
MOSPR (blue) derived from raw reflectivity data for different refractive index solutions. (b) Signal-to-noise ratio
for Au—SPR (red), Au-Co alloy—MOSPR (black) and Au-Co-Au tri-layer—MOSPR (blue) for solutions with
different refractive indices—reflectivity values derived from fitted SPR curves

Au-Co alloy chip in respect to gold is 1.5-fold for raw reflectivity

approach and reaches 2.5 times when using the reflectivity values
derived from fitted SPR curves. In summary, Fig. 5a, b compare the
SNR curves derived: (1) directly from the raw reflectivity data, and
(2) when using the fitted SPR data. This highlights the advantage of
our method in terms of improving on the SNR for both SPR and
MOSPR assays.

3.4 MOSPR Chip Stability of the MOSPR chips in physiological solutions is tested by
Stability in Liquid exposure to a saline solution (e.g., HEPES buffer solution—HBS)
Media while applying the oscillatory magnetic field. While the alloy shows
no sign of destabilization (for virtually indefinite exposure, similar
to regular Au sensor chips), in the case of Au-Co-Au tri-layer there
is a significant exfoliation of the metallic film (Fig. 6a). Increase in
the transparency of the film (observed under transmission micros-
copy Fig. 6a) further amplified with longer exposure confirms the
conclusion of film exfoliation. In the same flow conditions and
buffers, no modifications were observed in the cases of plain gold
and/or Au-Co alloy (Fig. 6b).

3.5 Bioaffinity Assay MOSPR assays of bioaffinity interactions are performed for the
alloy-based MOSPR chip. Because of the corrosive effect of the
saline buffers the Au-Co-Au tri-layers cannot be used even after
surface modification and functionalization.
Affinity tests, using a model biomolecular interaction between
human immuno-globulin G (HIgG) and Anti-HIgG in a direct
assay [1], were performed. IgG is used as model analyte without
restricting the generality of the method.
 Biosensing Based on Magneto-Optical Surface Plasmon Resonance 85

Fig. 6 Optical image of an Au-Co-Au tri-layer chip showing (a) pronounced corrosion in the areas exposed to
saline solution in contrast with (b) non-corroded film

Fig. 7 (a) Relative reflectivity for several specific injections of Anti-HIgG (concentration range 5–80 nM) and
two regeneration steps on the MOSPR Au-Co alloy chip. (b) Signal-to-noise ratio for specific MOSPR (for Au-Co
alloy—black) and SPR (for Au chip—red) responses for several injections of Anti-HIgG

1. The sensor chip surfaces of Au-Co alloy are first functionalized

using a carboxylated thiol and then further modified with
HIgG via an amine coupling reaction.
2. The bioaffinity assay is performed at 100 μl/min, by injecting
successively, for 2 min each, different concentrations of the
analyte (A-HIgG) in the range 5–80 nM in HBS.
3. Each injection is followed by a washing step of 2 min to assess
the binding level.
4. Surface is fully regenerated with two pulses of 10 mM glycine at
pH 2 (2 min).
Several concentrations of Anti-HIgG (ranging from 5 to
80 nM) were injected consecutively over the sensor’s surface and
the response was recorded (Fig. 7a). The signal shows good
86 Sorin David et al.

stability during the experiment: after regeneration (when the bioaf-

finity complexes are broken up and Anti-HIgG washed away) the
signal always returns to initial baseline values, and comparable
signals are obtained for multiple rounds of affinity binding (only
two shown in Fig. 7a) demonstrating the good stability of the alloy
in buffer and regeneration solutions.
The improvement of the SNR for the Au-Co alloy-based
MOSPR assay as compared to the Au-based classic SPR assays
(Fig. 7b) is evaluated by comparing the limit of detection (calcu-
lated as three times the standard deviation of blank) which is
around 0.60 nM for MOSPR and 0.96 nM for the SPR assay, an
improvement of 160% of the sensitivity of MOSPR over SPR assays.

3.6 Concluding The proposed novel solutions, based on innovative chip structure,
Remarks and tailored measurement and improved data analysis methods,
surpass current limitations of MOSPR sensing assays and have
been demonstrated to: (a) enable improved plasmonic and mag-
netic properties yet a structural stability similar to standard Au-SPR
chips, allowing for bioaffinity assays in saline solutions, (b) detec-
tion of minute analyte concentrations (IgG is used as model analyte
without restricting the generality of the method). The custom-
designed MOSPR measurement configuration allows the quasi
simultaneous acquisition of the whole SPR curve at high signal-
to-noise ratios during long time assessment in liquid media, a
significant advancement over existing MOSPR chips, and confirms
the MOSPR increased sensitivity over standard SPR analyses.

4 Notes

1. Functionalization reagents are available from other commercial

sources.
2. Piranha solution reacts violently with many organic materials
and should be handled with extreme care.
3. Aqua regia preparation produces heat and poisonous vapors; it
may react violently with many organic materials and should be
handled with extreme care.
4. A chromium underlayer may also be employed.
5. A 50-nm-thick gold layer is optimal for this wavelength of
incident light (840 nm). Optimal thickness varies with the
wavelength of incident light.
6. Electron beam evaporation may be employed in place of ther-
mal evaporation.
7. The sputtering technique may be used for depositing alloys
[22–24]; however, magnetron sputtering magnetic material
requires expensive magnetron guns with powerful magnets.
 Biosensing Based on Magneto-Optical Surface Plasmon Resonance 87

Moreover, manufacturing custom sputtering targets compris-

ing gold alloys may have prohibitively high costs.
8. E-beam epitaxy may be also used for growing crystalline films
with well-controlled magnetic properties but this is suitable/
limited to substrates with crystalline structure [25].

Acknowledgments

The authors thank the Romanian Executive Unit for Higher Edu-
cation, Research, Development and Innovation Funding for fund-
ing through grants PN II-ID-PCCE-2011-2-0075 and PN-II-RU-
PD-2012-3-0467.

References

1. Schasfoort R, Tudos A (2008) Handbook of 11. Regatos D, Fariña D, Calle A et al (2010) Au/
surface plasmon resonance. The Royal Society Fe/Au multilayer transducers for magneto-
of Chemistry, Cambridge optic surface plasmon resonance sensing. J
2. Raether H (1988) Surface Plasmons on Appl Phys 108:054502. doi:10.1063/1.
smooth and rough surfaces and on gratings. 3475711
Springer, Berlin 12. Manera MG, Montagna G, Ferreiro-Vila E et al
3. Incorporated TI (2003) Spreeta on the sensi- (2011) Enhanced gas sensing performance of
tivity of spreeta. Solutions 2003:1–14 TiO2 functionalized magneto-optical SPR sen-
4. Sepúlveda B, Calle A, Lechuga LM, Armelles G sors. J Mater Chem 21:16049. doi:10.1039/
(2006) Highly sensitive detection of biomole- c1jm11937k
cules with the magneto-optic surface-plasmon- 13. Manera MG, Ferreiro-vila E, Cebollada A et al
resonance sensor. Opt Lett 31:1085–1087 (2012) Ethane-bridged Zn porphyrins dimers
5. Regatos D, Sepúlveda B, Fariña D et al (2011) in langmuir-Scha fer thin films: spectroscopic,
Suitable combination of noble/ferromagnetic morphologic, and magneto-optical surface
metal multilayers for enhanced magneto- plasmon resonance characterization. J Phys
plasmonic biosensing. Opt Express Chem C 116:10734–10742
19:8336–8346 14. David S, Polonschii C, Luculescu C et al
6. Armelles G, González-Dı́az J (2008) Localized (2015) Magneto-plasmonic biosensor with
surface plasmon resonance effects on the enhanced analytical response and stability. Bio-
magneto-optical activity of continuous Au/ sens Bioelectron 63:525–532. doi:10.1016/j.
Co/Au trilayers. Opt Express 16:3112–3114 bios.2014.08.004
7. Temnov VV, Armelles G, Woggon U et al 15. Johnsson B, Löfås S, Lindquist G (1991)
(2010) Active magneto-plasmonics in hybrid Immobilization of proteins to a
metal–ferromagnet structures. Nat Photonics carboxymethyldextran-modified gold surface
4:107–111. doi:10.1038/nphoton.2009.265 for biospecific interaction analysis in surface
plasmon resonance sensors. Anal Biochem
8. Belotelov VI, Kreilkamp LE, Akimov IA et al 198:268–277. doi:10.1016/0003-2697(91)
(2013) Plasmon-mediated magneto-optical 90424-R
transparency. Nat Commun 4:2128. doi:10.
1038/ncomms3128 16. Chinowsky TM, Jung LS, Yee SS (1999) Opti-
mal linear data analysis for surface plasmon
9. Zvezdin A, Kotov V (1997) Modern magne- resonance biosensors. Sensors Actuators B
tooptics and magnetooptical materials. IOP Chem 54:89–97
Publishing Ltd, Bristol
17. Zhan S, Wang X, Liu Y (2011) Fast centroid
10. Armelles G, Cebollada A, Garcı́a-Martı́n A, algorithm for determining the surface plasmon
González MU (2013) Magnetoplasmonics: resonance angle using the fixed-boundary
combining magnetic and plasmonic function- method. Meas Sci Technol 22:025201.
alities. Adv Opt Mater 1:10–35. doi:10.1002/ doi:10.1088/0957-0233/22/2/025201
adom.201200011
88 Sorin David et al.

18. David S, Gheorghiu E (2011) Towards an 22. Kahn D (1992) Magnetic properties and struc-
advanced magneto-plasmonic sensing plat- ture of cobalt hardened gold. SUR/FIN’92
form. Adv Top Electr Eng (ATEE), 2011 7th 1:305–341
International symposium 1:12–14 23. Yang K, Clavero C, Skuza J (2010) Surface
19. Gheorghiu M, David S, Polonschii C et al plasmon resonance and magneto-optical
(2014) Label free sensing platform for amyloid enhancement on Au–Co nanocomposite thin
fibrils effect on living cells. Biosens Bioelectron films. J Appl Phys 107:103924
52:89–97. doi:10.1016/j.bios.2013.08.028 24. Mattox D (2010) Handbook of physical vapor
20. Neuert G, Kufer S, Benoit M, Gaub HE deposition (PVD) processing, 2nd edn. Else-
(2005) Modular multichannel surface plasmon vier, Inc., Berlin
spectrometer. Rev Sci Instrum 76:054303. 25. Ferreiro-Vila E, Iglesias M, Paz E et al (2011)
doi:10.1063/1.1899503 Magneto-optical and magnetoplasmonic prop-
21. Polonschii C, David S, Tombelli S (2010) A erties of epitaxial and polycrystalline Au/Fe/
novel low-cost and easy to develop functiona- Au trilayers. Phys Rev B 83:205120. doi:10.
lization platform. Case study: aptamer-based 1103/PhysRevB.83.205120
detection of thrombin by surface plasmon res-
onance. Talanta 80:2157–2164
 Chapter 6

Nanoplasmonic Biosensor Using Localized Surface Plasmon

Resonance Spectroscopy for Biochemical Detection
Diming Zhang, Qian Zhang, Yanli Lu, Yao Yao, Shuang Li,
and Qingjun Liu

Abstract
Localized surface plasmon resonance (LSPR) associated with metal nanostructures has developed into a
highly useful sensor technique. Optical LSPR spectroscopy of nanostructures often shows sharp absorption
and scattering peaks, which can be used to probe several bio-molecular interactions. Here, we report
nanoplasmonic biosensors using LSPR on nanocup arrays (nanoCA) to recognize bio-molecular binding
for biochemical detection. These sensors can be modified to quantify binding of small molecules to proteins
for odorant and explosive detections. Electrochemical LSPR biosensors can also be designed by coupling
electrochemistry and LSPR spectroscopy measurements. Multiple sensing information can be obtained and
electrochemical LSPR property can be investigated for biosensors. In some applications, the electrochemi-
cal LSPR biosensor can be used to quantify immunoreactions and enzymatic activity. The biosensors exhibit
better performance than those of conventional optical LSPR measurements. With multi-transducers, the
nanoplasmonic biosensor can provide a promising approach for bio-detection in environmental monitor-
ing, healthcare diagnostics, and food quality control.

Key words Localized surface plasmon resonance (LSPR), Nanocup arrays (nanoCA), Electrochemis-
try, Biosensor, BSA (bull serum albumin), Thrombin

1 Introduction

Optical detection is particularly a promising method because it

allows remote transduction without any physical connection
between excitation sources and detecting elements. In recent
years, several nanostructured sensors such as photonic crystal, whis-
pering gallery mode (WGM), and surface plasmon resonance (SPR)
were increasingly being employed by optical spectroscopy for DNA
detection, antigen-antibody recognition, immunoassays probing,
and pathogen identification [1–3]. These sensors can respond to
target molecules at low concentrations by observing shifts in reso-
nance wavelength before and after conjugation of biomolecules to
the sensor surfaces. A common drawback, however, is that optical

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_6,
© Springer Science+Business Media LLC 2017

89
90 Diming Zhang et al.

Fig. 1 Schematic of localized surface plasmon resonance (LSPR), where the free
conduction electrons in the metal sphere are driven into oscillation due to
coupling with incident light

biosensors, such as SPR and WGM, often required a complex

system. They should fabricate devices that ensure precise alignment
of light coupling to bio-detection volume and generate the desired
optical phenomena.
Metal nanostructures were recently reported to modulate loca-
lized surface plasmon resonance (LSPR) spectroscopy for probing
biomolecular interactions [4–6]. LSPR was often induced by inci-
dent light when it interacted with noble metal nanostructures, such
as gold nanoparticles, that had smaller sizes than the wavelength of
the incident light (Fig. 1). The light can be coupled in direct
perpendicular transmission between incident light and spectrome-
ter to obtain LSPR wavelength shifts without complex optical
coupling [7–9]. Thus, it can be measured by a robust optical
sensing platform minimizing the alignment requirement. However,
sizes and positions of single nanostructures such as colloid nano-
particles were always random and difficult to control over a large
area. Thus, the wavelength shift may be different for the same
analyte because nanoparticles of different sizes gave rise to different
optical spectroscopy as biosensors [8]. To generate stable LSPR
spectroscopy, periodic nanostructure arrays such as nanohole
arrays, nanorod arrays, and nanocone arrays can be fabricated by
several nanoelectromechanical systems including electron beam
lithography and focused ion beam milling for biosensing [9, 10].
These nanostructures were often fabricated by scanning beam
lithographies, such as electron beam lithography (EBL) and
focused ion beam (FIB) lithography. Thus, compared to nanopar-
ticles from chemical synthesis, these nanostructure arrays with their
more precise sizes and positions provide a promising approach to
producing repeatable and stable LSPR spectroscopy for biological
 Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 91

and chemical detection. Other detection techniques such as elec-

trochemistry can also be combined with LSPR spectroscopy of
nanostructures while providing unique advantages in sensitivity
and selectivity. For instance, voltammetry can be used to modulate
electrochemical reactions on nanostructures and provide additional
selectivity to LSPR monitoring of α-synuclein–small molecule
interactions [11].
Nanoimprint lithography is a method of fabricating nanometer
scale patterns, and offers advantages including low cost, high
throughput, and high resolution. It often creates nanostructure
patterns by molding and curing of monomer or polymer resist on
nanostructured template, while adhesion between the resist and the
template is controlled to allow proper release. Nanoimprint lithog-
raphy is currently used to fabricate devices for electrical, optical,
photonic, and biological applications [12, 13]. In our work, nano-
cup arrays (nanoCA) were fabricated by nanoimprint and deposited
with nanoparticles to modulate a stable LSPR spectroscopy for
biosensor applications [14–16]. Nanoplasmonic biosensors using
nanoCAs can be employed to monitor small molecule binding to
proteins such as odorant binding proteins (OBPs) and peptides, and
for odorant and explosive detection. The electrochemical detection
technique can also be coupled with LSPR spectroscopy to design an
electrochemical LSPR biosensor. This provides multiple sensing
information for biosensor applications. The electrochemical LSPR
sensor can also be used to quantify immunoreactions and enzymatic
activity with demonstrably higher sensitivity.

2 Materials

2.1 Optical and 1. Halogen cold light source (DT-MINI-2, Ocean Optics Inc.,
Electrochemical Dunedin, Florida, USA).
Measurement System 2. Spectrophotometer (USB2000+, Ocean Optics Inc., Dunedin,
Florida, USA).
3. Three fiber bundles (QP230-0.25-XSR, Ocean Optics Inc.,
Dunedin, Florida, USA).
4. Optical fiber attenuator (FVA-ADP-UV, Ocean Optics Inc.,
Dunedin, Florida, USA).
5. Collimating lens (F230APC-633, Thorlabs Inc., Newton, New
Jersey, USA).
6. Absorb sample pool (SPL-CUV-ABS, Pulei Inc., Hangzhou,
China).
7. Multi-mode microplate reader (SpectraMax Paradigm, Molec-
ular Devices Co., United Sates).
8. Electrochemical workstation (CHI 660E, CH Instruments,
Texas, USA).
92 Diming Zhang et al.

9. Ag/AgCl reference electrode (CHI 111, CH Instruments,

Texas, USA).
10. Pt counter electrode (CHI 102, CH Instruments, Texas, USA).

2.2 Acquisition and 1. Phosphate buffer saline (PBS) (see Note 1).
Immobilization of 2. 2-(morpholino)ethanesulfonic acid (MES) buffer (see Note 2).
Proteins
3. 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) solu-
tion (see Note 3).
4. N-hydroxy-succinimide (NHS) solution (see Note 4).
5. Carboxy-poly(ethylene)-thiol (HOOC-PEG-SH, 2 kDa,
Sigma- Aldrich, USA) (see Note 5).
6. Recombinant expressed plasmid pET-Acer-ASP2 (from Center
of Analysis & Measurement, Zhejiang University) (see Note 6).
7. E. coli BL21 (DE3) (from Center of Analysis & Measurement,
Zhejiang University).
8. BamH I endonuclease (TaKaR Co., Dalian, China).
9. Hind III endonuclease (TaKaR Co., Dalian, China).
10. Isopropyl-β-D-thiogalactopyranoside (Aladdin, Shanghai, China).
11. OBPs, Acer-ASP2.
12. TNT-specific peptides (see Note 7).

2.3 Fabrication of 1. UV light-curing flood lamp system (EC-Series, Dymax, USA).

nanoCA Device 2. E-beam evaporation system (FC/BJD2000, Temescal, USA).
3. Scanning electron microscope (SEM, XL30-ESEM, Philips,
Netherlands).
4. Nanocone quartz template fabricated by e-beam lithography.
5. UV curable polymer (NIL-UV-Si, GuangDuo Nano Ltd., Suz-
hou, China).
6. Dimethyl dichlorosilane (CP, Aladdin, Shanghai, China).
7. Titanium (>99.99%, Sigma-Aldrich, USA).
8. Gold (99.99%, Sigma-Aldrich, USA).
9. (Poly)ethylene terephthalate (PET) substrate.
10. Teflon roller.
11. Transparent epoxy resin adhesive.

3 Methods

3.1 Fabrication of the The nanoCA device can be fabricated for use in LSPR spectroscopy
nanoCA for monitoring binding events occurring on the surface of the
device [14, 17]. The nanostructures were imprinted by nanocone
template made on quartz substrate and deposited with gold nano-
particles by following steps, as shown in Fig. 2.
 Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 93

Fig. 2 Flow chart for nanoCA device fabrication by nanoimprint lithography

1. The template had nanocones in arrays, whose height, upper,

and base diameter were 300, 200, and 500 nm, respectively. It
was passivated with dimethyl dichlorosilane solution for 30 min
and then rinsed three times with ethanol and deionized water.
This promotes formation of a hydrophobic silane layer on the
template, which in turn promotes removal of cured polymer
replica with nanocup structures.
2. A 250 μm flexible (Poly)ethylene terephthalate (PET) sheet
was used as a supporting substrate, and a Teflon roller was
used for evenly distributing the UV curable polymer at the
template/PET interface. The depth and opening diameters of
the nanocup were 500 and 200 nm respectively, while the
diameter of nanoparticles was about 20 nm.
3. To obtain the nanocup array structure, a UV light-curing
flood lamp system was used at an average power density of
105 mW/cm2 to solidify the UV polymer on the template
and PET interface. The polymer was cured by UV light for
60 s at room temperature.
4. To deposit gold nanoparticles, 5-nm-thick titanium was depos-
ited first and used as the adhesive layer. Then, gold nanoparti-
cles were deposited on sidewalls and bottom of nanocups. The
depositions were both performed with a six pocket e-beam
evaporation system. Finally, the nanoCA structure with nano-
particles on sidewalls was successfully fabricated (Fig. 3).
Figure 3c shows a scanning electron microscope image of the
nanoCA device.
94 Diming Zhang et al.

Fig. 3 Fabrication and characterization of nanocup arrays (nanoCA). (a) Photo-

graph of nanoCA device in ~ 2
2 cm2. (b) Structure of periodic nanocups on
nanoCA chip, with Au nanoparticles deposited along the sidewalls. (c) SEM
image of the cup arrays

3.2 Optical Detection Periodic nanostructures have prominent optical features such as
for LSPR Spectroscopy absorption and transmission peaks in the visible spectrum. These
features were elicited by plasmon resonance and can be utilized in
optical sensors. The optical detection can be performed using the
following steps.
1. As shown in Fig. 3a, the nanoCA chips were first made into
small rounds with diameters of 1 cm by hole puncher. Then the
rounds can be immobilized on the bottom of wells by transpar-
ent epoxy resin adhesive, leaving 6 h for the adhesive to solidify.
LSPR spectroscopy was performed using the microplate reader
for multi-mode detection in high throughput.
2. A normal transmission model of the microplate reader was
applied to measure the spectra of nanoCA. Light was emitted
from the light source on the bottom of wells of the plate,
delivered through nanoCA chips and received by the spectrom-
eter. The reader detected LSPR spectroscopy in individual wells
of the plate. The lamp and spectrograph can move from one
well to other wells by a mechanical driver. The scanning range
for spectroscopy was set from 300 to 900 nm and the step was
fixed at 1 nm.
3. During measurement, 10 μl sample solutions were often added
by pipette on the surface of the chip in the wells of the plate.
The small volume of the analyte formed a thin liquid layer and
reduced interference from the solution itself.
 Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 95

Fig. 4 Nanoplasmonic biosensor system for bio- and chemical detection. (a) Schematic diagram of the optical
detection using a 96-well plate instrument in transmission mode. (b) Transmission spectra of nanoCA in the
presence of NaCl at different concentrations (0.1%, 0.2%, 0.3%, 0.4%, and 0.5%). The resonance wavelength
shifts with increasing concentration

4. NaCl at different concentrations was employed as increasing

refractive index solutions, to determine the refractive-index
sensing properties of nanoCA (Fig. 4b). 10 μl NaCl solutions
at concentrations of 0.1%, 0.2%, 0.3%, 0.4%, and 0.5% were
added by pipette on the surface of the chip. Then, optical
detection was performed as described above. The entire trans-
mission spectrum exhibited red shift (shift to right end of the
spectrum) with increasing NaCl concentration, indicating the
refractive index (RI) sensitivity of nanoCA in LSPR detection.
Indeed, transmission peaks around 600 nm were often used as
responses of nanoplasmonic sensors. With synergetic LSPR
effects of periodical array structure and nanoparticles deposited
on the cups, the nanoCA showed a high sensitivity for RI
change with ~104 nm per refractive index unit (RIU), which
was much greater than typical nanoparticle sensors and nano-
hole sensors based on plasmon resonance [18–20]. Thus, a
high-sensitivity LSPR device based on nanoCA is useful as a
high-throughput biosensor for monitoring binding of small
molecules to proteins.

3.3 Acquisition of In sensing property test, the nanoCA device showed a good perfor-
Bio-sensitive Material mance for RIU change on the surface. However, the nanoCA
device was only a physical transducer sensitive to refractive index
changes, rather than a real sensor with high selectivity to special
biochemical analytes. In our work, several olfactory proteins, such
as OBPs and peptides, were used to modify nanoCA for bio- and
chemical detections.
96 Diming Zhang et al.

3.3.1 Expression and Oderant binding proteins (OBPs) are an important class of sensing
Purification of OBPs proteins in biological olfactory systems. OBPs have significant
binding affinities to various biochemical molecules. Distinct from
membrane proteins or receptors, OBPs are soluble, can be
expressed at low cost, purified easily, and maintain bioactivity
in vitro [21, 22]. Thus, OBPs are useful as biosensors and can be
obtained by the following steps:
– The recombinant OBPs of Acer-ASP2 were cloned from full-
length cDNA of adult worker bees, Apis cerana cerana. The
recombinant plasmid pET-Acer-ASP2 was expressed and trans-
formed into E. coli BL21 (DE3) competent cells after 450 bp
fragments were excised with BamH I and Hind III from the
pGEM-Acer-ASP2 plasmid.
– The cells were grown in Luria–Bertani broth (including 30 μg/
ml kanamycin) at 37  C with 1.5 mM isopropyl-β-D-thiogalac-
topyranoside to induce expression of the protein. After 5 h at
28  C, the bacterial cells were harvested and lysed by sonication
and centrifuged into crude cell extracts into pellet and superna-
tant (3 ml, 1740
g, 10 min).
– The OBP was then collected from the cells. The pellets of crude
cell extracts containing recombinant proteins were precipitated
in 1.5 M urea in ddH2O and freeze-dried.
– Finally, the protein was resuspended (500 μg/ml) in phosphate
buffered saline (PBS; pH ¼ 7.4) and stored at 4  C for biosensor
experiments. More details can be found in the previous study
about expression of the OBPs from honeybee [23].

3.3.2 Synthesis of Peptides are short protein fragments composed of a chain of amino
Peptide acids. Peptides can be designed based on known structures of
protein binding sites, synthesized with chemical methods, and
purified to obtain specific sequences. They are useful in designing
artificial receptors to mimic molecular recognition between pro-
teins and analytes. Thus, peptides are ideal candidates for biosen-
sing materials and for biosensor fabrication.
– TNT-specific peptide was chemically synthesized based on the
reported bio-sensitive sequence (WHWQRPLMPVSI) as olfac-
tory protein for TNT [24].
– The peptide (CLVPRGSC) can be cleaved by thrombin, and was
synthesized for use in thrombin detection.
– Chemical synthesis of peptides was performed by standard solid-
phase peptide synthesis (SPPS) using BOC chemistry, with step-
wise addition of protected amino acids to a growing peptide
chain. The synthesis work was performed by Genscript company
(Nanjing, China). Peptides were stored as freeze-dried powders
(see Note 8).
 Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 97

Fig. 5 Self-immobilization of olfactory proteins on sidewalls of nanoCA device

3.4 Self- 1. The self-immobilization of the proteins on nanoCA was per-

Immobilization of formed by linkage of HOOC-PEG-SH (Fig. 5). The immobi-
Proteins lization can be completed by following steps.
2. NanoCA device was first washed with a mixture of 98% H2SO4
and 30% H2O2 (7:3), and DI water respectively, to remove the
organic residues.
3. Subsequently, the nanoCA device was immersed in a petri dish
and reacted with 2 ml HOOC-PEG-SH at 1 mg/ml overnight
(about 18 h ) at room temperature to form Au-S covalent
bonding on the surface of the nanoCA device.
4. 2 ml mixing solution of 8 mg/ml EDC and 12 mg/ml NHS in
0.1 M MES buffer (pH ¼ 5, Sigma) was immersed in petri dish
for 15 min to activate carboxylate groups of HOOC-PEG-SH
after extra HOOC-PEG-SH was washed off with DI water.
5. Finally, the solution was adjusted to pH 8.0 with saturated
NaHCO3, added with 1 ml olfactory proteins at 500 μg/ml,
and incubated at room temperature for 2 h. The nanoCA was
washed with DI water by gentle shaking for 60 s, followed by
drying with nitrogen and stored at 4  C for further experiments
(see Notes 8 and 9).

3.5 Olfactory Protein The interaction between small molecules and olfactory proteins
Modified Sensor Using immobilized on nanoCA surface can affect the refractive index in
LSPR Spectroscopy the metal-dielectric interface of nanocup sidewalls and thus give
rise to resonance wavelength shift in LSPR spectroscopy. The
wavelength shift can then be used to quantify small molecules
involved in binding events, and bio-modified nanoCA can be
used as a nanoplasmonic biosensor for bio- and chemical detections
[15, 16].
98 Diming Zhang et al.

3.5.1 LSPR Spectroscopy The nanoCA can be used as a biosensor by monitoring interactions
Analysis for Monitoring of between small molecules and olfactory proteins. For instance, sev-
Small Molecule Binding to eral chemical molecules can specifically bind into hydrophobic
Protein pockets of OBPs immobilized on the nanoCA surface, which can
modulate protein conformation and change relative dielectric con-
stants of the bio-coating [15] (Fig. 6a).
– The OBPs were immobilized on the surface of the nanoCA
device by self-assembly to obtain OBP-modified nanoCA for
monitoring of molecular binding (see Subheading 3.4). For
odorant measurement one kind of aromatic odorant, β-ionone,
was used as a model molecule to test responses of the bio-
nanoCA with OBPs. Ten microliter β-ionone solutions at differ-
ent concentrations ranging from 10 nM to 10 Mm can be added
in wells of the 96-well plates by pipette (see Note 10).

Fig. 6 Bio-nanoCA with olfactory proteins as nanoplasmonic biosensors for odorant and explosive detection.
(a) Molecular docking of the binding pocket of the OBPs, Acer-ASP2. (b) Transmission spectrum of OBPs-
modified nanoCA for β-ionone. (c) Dose-dependent profiles of bio-nanoCA with OBPs in resonance wavelength
shift of transmission peak for β-ionone and TNT. (d) Dose-dependent profiles of bio-nanoCA with TNT-specific
peptide for β-ionone and TNT. (mean  SD, n ¼ 15)
 Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 99

– Optical detection is performed as described above. The trans-

mission spectra of the bio-nanoCA device can be obtained as
illustrated in Fig. 6b. Intensity and location of the transmission
peak was notably changed by specific binding of β-ionone and
OBPs. Transmission intensity change and LSPR wavelength
shift can both be calculated statistically with the OBPs-modified
devices. As expected, the wavelength shift showed a good linear
dose-dependent response for β-ionone at increasing concentra-
tions (Fig. 6c). These results demonstrate that nanoCA can
monitor specific binding of β-ionone to OBPs by the resonance
wavelength shift in LSPR spectroscopy.

3.5.2 Nanoplasmonic Recent studies reported that trained honeybees are able use their
Biosensor for Explosive sensing proteins to find nitro-explosives, and various OBPs have
Detection been used to modify biosensors for explosive detection [25–27].
We have explored the binding of nitro-explosive molecules to the
OBPs of AcerASP2 by the bio-nanoCA.
– Molecular docking of TNT and the OBP was carried out to
show affinity of these two molecules. Like β-ionone, TNT mole-
cules can enter protein cavities, interact with amino acid resi-
dues, and form a complex with OBPs. This process might elicit
OBP conformation changes like those in β-ionone binding,
thereby modulating the LSPR on the nanoCA in the optical
measurement.
– The OBPs were immobilized on the surface of the nanoCA
device by self-assembly to obtain OBP-modified nanoCA for
monitoring of molecular binding (see Subheading 3.4). Ten
microliter TNT solutions at different concentrations can be
added in wells of the 96-well plates by pipette. Then, optical
detection can be performed as described above (see Subheading
3.2). Figure 6c showed responses of the nanoCA with the OBPs
to TNT at increasing concentrations. The nanoCA can monitor
the binding of TNT molecules to the OBPs. However, the
responses were nonlinear in concentration range from 107 to
105 M, and the nanoCA cannot distinguish TNT and
β-ionone. It indicated that the nanoCA with OBPs was not a
good biosensor strategy for explosive detection of TNT.
– The TNT-specific peptide was immobilized on the surface of the
nanoCA device by self-assembly to obtain peptide-modified
nanoCA for monitoring of molecular binding (see Subheading
3.4). Ten microliter TNT solutions at different concentrations
can be added in wells of the 96-well plates by pipette. Then,
optical detection can be performed as described above (see Sub-
heading 3.2). As shown in Fig. 6d, the nanoCA with the pep-
tides was high selective to TNT, while the nanoCA showed no
response to β-ionone.
100 Diming Zhang et al.

3.6 Electrochemical LSPR is not only influenced by refractive index change on the
LSPR Biosensor for nanostructure surface, but also by potential and current density
Protein Analysis on the surface of a metal film [14]. Thus, LSPR sensors can be
combined with electrochemistry to implement synchronized elec-
trical and optical responses. The combination of electrochemistry
and optical SPR measurement can provide a multitude of informa-
tion from different signal transductions and provide novel
approaches to elucidate detailed processes for electrochemical
reactions.

3.6.1 Electrochemical The apparatus for electrochemical LSPR measurement is depicted

LSPR Measurement in Fig. 7a. Optical measurement was performed in transmission
mode. The light was emitted from the source below the nanoCA
device, delivered through the device, and finally received by a
spectrophotometer. The LSPR signal was measured and analyzed
by transmission spectroscopy.
– For electrochemical measurements the potentiodynamic elec-
trochemical measurement of cyclic voltammetry (CV) was used
to investigate the electrochemistry properties of nanoCA device
and the electrochemical modulation to LSPR. The nanoCA
device was used as the working electrode, while platinum

Fig. 7 Electrochemical LSPR spectroscopy measurement. (a) Schematic of experimental apparatus for
synchronous electrochemical and optical measurement. (b) Transmission LSPR spectroscopy of nanoCA
with different surface currents. (c) Synchronous measurement of LSPR and electrochemistry in CV scanning
Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 101

Fig. 8 Photograph of the electrochemical LSPR measurement apparatus

electrode and Ag/AgCl electrodes were used as counter elec-

trode and reference electrodes, respectively.
– In the synchronous measurement, the range of the transmission
spectrum was from 300 to 1000 nm with a step of 0.5 nm. The
transmission spectrum can be recorded when the surface of
nanoCA had RI change and electrochemical current (Fig. 7b).
The electrochemical reactions were measured using an electro-
chemical workstation in a standard three-electrode system.
5 mM K4[Fe(CN)6]/K3[Fe(CN)6] (1:1) was employed as
redox couple in aqueous solution for electrochemical character-
ization. The scanning rate of CV was fixed at 0.02 V/s and the
transmission spectrum was simultaneously recorded and saved
every 5 s. Thus, the transmission spectrum was obtained to
represent optical LSPR, in 0.1 V steps during voltage scanning
with different electrochemical current (Fig. 7c). Figure 8 shows
a photograph of the electrochemical LSPR measurement appa-
ratus. For optical detection, there were three fiber bundles, two
collimating lenses, an optical fiber attenuator, a light source and
a spectrophotometer. Fiber bundles were used to provide optical
linkage between the lens, the attenuator, the light source, and
the spectrophotometer. The spectrophotometer was also
directly connected to the light source by data cable. Thus, the
spectrophotometer can synchronously capture light emitted by
the light source and delivered through nanoCA devices. For
electrochemical detection, the nanoCA device was used as the
working electrode, while platinum and Ag/AgCl electrodes
were used as counter electrode and reference electrodes,
respectively. These three electrodes were all linked to an
102 Diming Zhang et al.

electrochemical workstation and inserted into a spectrometer

measuring cell filled with redox solution. Careful arrangement
of these three electrodes should maintain successful delivery
through the cell.
– The electrochemical LSPR measurements for nanoCA were fur-
ther carried out and analyzed with transmission spectrum and
synchronous CV scanning. As seen in Fig. 7b, a significant LSPR
dip shift was recorded in the transmission spectrum when elec-
trochemical current on the nanoCA surface was changed from
0 to 0.2 mA. The resonance wavelength was shifted from 558 to
564 nm but there was no significant shift observed for LSPR
peaks.
– Optical measurements within CV scanning were carried out and
the wavelength shifts of LSPR dips were shown in Fig. 7c.
During scanning, the wavelength of dip was shifted from
566 nm to 573 nm with an increase of current from 0.25 to
0.5 mA. When the current returned to 0.25 mA, the wavelength
also decreased with declining current. These results indicate that
the current change on the nanoCA surface can modulate LSPR
of nanoCA and produce a wavelength shift in LSPR. The LSPR
in transmission dip around 550 nm can be enhanced by electro-
chemical current and can be analyzed as an indicator in electro-
chemical LSPR detecting.

3.6.2 Electrochemical Utilizing electrochemical enhancement for LSPR on nanoCA, elec-

LSPR Spectroscopy as trochemical LSPR biosensors can be fabricated with higher sensi-
Biosensor for tivity in the synchronous measurement mode. Many bio-
Immunoreaction interactions such as immunoreaction can be probed by electro-
chemical LSPR spectroscopy on the nanoCA.
– The nanoCA was modified by anti-BSA which can specifically
bind to BSA by self-assembly (for methods see Subheading 3.4).
– One milliliter BSA solutions at different concentrations can be
added in wells of the 96-well plates by pipette.
– The CV scanning can be performed on the nanoCA device (see
Subheading 3.6). The redox current can be generated from the
redox couple and modulated by the binding of BSA to anti-BSA.
The CV scanning can provide redox currents around 0.1 and
0.3 V in the voltammogram with electrochemical reactions.
– The transmission spectrum was recorded when CV scanning
reached 0.3 V (Fig. 9a, see Subheading 3.6). LSPR dip shifts
were elicited by the conjunction of anti-BSA and BSA on the
surface of nanoCA and enhanced by the CV scanning from 9 to
15 nm. Due to the redox current, the negative protein of HSA
also elicited a wavelength shift of about 3 nm. However, this was
much lower than the response to BSA of about 15 nm with
 Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 103

Fig. 9 Electrochemical LSPR detection for specific binding of proteins. (a) The transmission spectrum of
nanoCA with PBS, 100 μg/ml BSA and BSA plus synchronous CV scanning, 100 μg/ml HAS and HSA plus CV
scanning. (b) Statistic for shifts in dip wavelength of HSA, HSA plus CV, BSA and BSA plus CV (mean  SD,
n ¼ 6)

electrochemical enhancement (Fig. 9b). These results suggest

that the electrochemically coupled LSPR measurement had
redox currents that modulate LSPR dip shift, in addition to
that resulting from RI change. In fact, LSPR measurement
with and without CV had detection limits of 13 and 25 μg/
ml, respectively, for BSA detection. Thus, the electrochemically
enhanced biosensor fabricated with electrochemical LSPR had
higher response and sensitivity than that of conventional optical
methods.

3.6.3 Electrochemical The electrochemical LSPR spectroscopy can also be used in throm-
LSPR Spectroscopy as bin detection [28]. The detection can be performed by the follow-
Biosensor for Enzymatic ing steps.
Activity
– As shown in Fig. 10a, polyethylene glycol (PEG), peptide
(CLVPRGSC), and bovine serum albumin (BSA) were immo-
bilized on surface of nanoCA with the self-assembly (methods,
see Subheading 3.4), while peptides were used as probes for
thrombin.
– One milliliter thrombin solutions at different concentrations can
be added in wells by pipette. Thrombin can cleave specific sites
of peptide sequences and remove BSA from the surface of
nanoCA, thereby modulating LSPR transmission spectra.
– Spectrographic measurement was coupled with electrochemical
voltage at 3 V to dissociative BSA from the nanoCA surface and
increase LSPR wavelength shift responses. The measurement
apparatus was the same as described above (see Subheading 3.6).
104 Diming Zhang et al.

Fig. 10 Electrochemical LSPR sensor for thrombin detection. (a) Design of electrochemical LSPR sensor for
thrombin using self-immobilization of protein on nanoCA. (b) LSPR responses of wavelength shifts of the
nanoCA for thrombin detections with (+) and without () electrochemical enhancement, electrochemical
voltage (EV)

The results for thrombin detection are shown in Fig. 10b. The
presence of thrombin can selectively catalyze cleavages at Arg-Gly
bonds of peptides, remove BSA from the nanoCA sidewall, and
generate time-dependent shifts in resonance wavelength of
nanoCA (see Notes 10 and 11).
– The transmission spectrum can be analyzed by LSPR wavelength
shifts with electrochemical enhancement. As expected, as the
electrochemical voltage is increased LSPR wavelength shifts
significantly. There was stronger voltage enhancement in the
presence of thrombin at high concentrations. Indeed, the elec-
trochemically enhanced LSPR biosensor had higher sensitivity
and shorter response time than the sensor utilizing LSPR alone.
This provides a promising approach for designing more sensitive
and rapid LSPR sensors.

4 Notes

1. Phosphate buffer saline (PBS) was used as the solvent for

proteins. The solution was prepared by dissolving 137 mM
NaCl, 2.68 mM KCl, 1.47 mM KH2PO4, and 8.10 mM
Na2HPO4 in pure water and pH was fixed at 7.4.
2. 2-(morpholino)ethanesulfonic acid (MES) buffer was used as
buffer solution for protein self-immobilization. It was prepared
by dissolving 100 mM MES and 500 mM NaCl, and its pH was
fixed at 6.0.
 Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 105

3. 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC)

solution was prepared at 2 mM, while MES buffer was used
as the solvent.
4. N-hydroxy-succinimide (NHS) solution was prepared at
5 mM, while MES buffer was used as the solvent.
5. Carboxy-poly(ethylene)-thiol (HOOC-PEG-SH, 2 kDa) was
used as linkage between proteins and nanoCA device. The
solution was prepared at 1 mg/ml, while ethyl alcohol and
pure water with rate 7:3 were used as solvent of HOOC-
PEG-SH.
6. OBPs, Acer-ASP2, were cloned from the full-length cDNA of
adult worker bees, Apis cerana cerana. The protein was resus-
pended in PBS solution at 500 μg/ml and stored at 4  C for
biosensor experiments (see Note 1).
7. Peptides used in our study were synthesized by the standard
solid phase method. Synthesized peptides were tested by high-
performance liquid chromatography (HPLC) and mass spec-
trometry (MS) to verify amino acid sequence and purity. The
peptides were stored in the form of freeze-dried powders
before experiments and dissolved in PBS at 500 μg/ml.
8. The proteins involved in the study, such as OBPs, peptides, and
BSA, should all be dissolved in PBS solution for use during
experiments. Before experiments, proteins should be stored at
4  C to maintain bio-activity.
9. In designing the electrochemical LSPR biosensor for throm-
bin, the protein immobilization on nanoCA was completed by
two steps of self-assembly of the peptide and BSA. The steps are
both as described for linking proteins to HS–PEG–COOH in
the above Method section.
10. Protein immobilization on nanoCA should be tested to verify
efficient combination of protein and the nanoCA device. Gen-
erally, electrochemical impedance spectroscopy and optical
spectroscopy can be used to verify the combination.
11. In the electrochemical LSPR detection for thrombin, the elec-
trochemical voltage can range from 0.1 to 5 V. The voltage at
3 V used in our study was optimized to obtain better sensitivity
and shorter sensor response times.

Acknowledgments

This work was supported by the National Natural Science Founda-

tion of China (Grant No. 81371643), the Zhejiang Provincial
Natural Science Foundation of China (Grant No. LR13H180002).
106 Diming Zhang et al.

References
1. Homola J (2008) Surface plasmon resonance olfactory proteins from honeybee as biosensor
sensors for detection of chemical and biological for chemical detection. Sensor Actuat B Chem
species. Chem Rev 108:462–493 221:341–349
2. Vollmer F, Arnold S (2008) Whispering-gal- 16. Zhang D, Zhang Q, Lu Y, et al. (2016) Peptide
lery-mode biosensing: label-free detection functionalized nanoplasmonic sensor for
down to single molecules. Nat Methods explosive detection. Nano-Micro Lett 8:36–43
5:591–596 17. Gartia MR, Hsiao A, Pokhriyal A et al (2013)
3. Nair RV, Vijaya R (2010) Photonic crystal sen- Colorimetric plasmon resonance imaging using
sors: an overview. Prog Quant Electron nano lycurgus cup arrays. Adv Opt Mater
34:89–134 1:68–76
4. Saha K, Agasti SS, Kim C et al (2012) Gold 18. Stewart ME, Mack NH, Malyarchuk V et al
nanoparticles in chemical and biological sens- (2006) Quantitative multispectral biosensing
ing. Chem Rev 112:2739–2779 and 1D imaging using quasi-3D plasmonic
5. Willets KA, Van Duyne RP (2007) Localized crystals. Proc Natl Acad Sci U S A
surface plasmon resonance spectroscopy and 103:17143–17148
sensing. Annu Rev Phys Chem 58:267–297 19. Kuznetsov AI, Evlyukhin AB, Goncalves MR
6. Mayer KM, Hafner JH (2011) Localized sur- et al (2011) Laser fabrication of large-scale
face plasmon resonance sensors. Chem Rev nanoparticle arrays for sensing applications.
111:3828–3857 ACS Nano 5:4843–4849
7. Gao HW, Henzie J, Odom TW (2006) Direct 20. Kee JS, Lim SY, Perera AP et al (2013) Plas-
evidence for surface plasmon-mediated monic nanohole arrays for monitoring growth
enhanced light transmission through metallic of bacteria and antibiotic susceptibility test.
nanohole arrays. Nano Lett 6:2104–2108 Sensor Actuat B Chem 182:576–583
8. Anker JN, Hall WP, Lyandres O et al (2008) 21. Lu Y, Li H, Zhuang S et al (2014) Olfactory
Biosensing with plasmonic nanosensors. Nat biosensor using odorant-binding proteins from
Mater 7:442–453 honeybee: Ligands of floral odors and phero-
9. Stewart ME, Anderton CR, Thompson LB mones detection by electrochemical imped-
et al (2008) Nanostructured plasmonic sen- ance. Sensor Actuat B Chem 193:420–427
sors. Chem Rev 108:494–521 22. Lu Y, Yao Y, Zhang Q et al (2015) Olfactory
10. Cao J, Sun T, Grattan KT (2014) Gold biosensor for insect semiochemicals analysis by
nanorod-based localized surface plasmon reso- impedance sensing of odorant-binding pro-
nance biosensors: a review. Sens Actuators B teins on interdigitated electrodes. Biosens
195:332–351 Bioelectron 67:662–669
11. Cheng XR, Wallace GQ, Lagugné-Labarthet F 23. Li H-L, Zhang Y-L, Gao Q-K et al (2008)
et al (2015) Au nanostructured surfaces for Molecular identification of cDNA, immunolo-
electrochemical and localized surface plasmon calization, and expression of a putative
resonance-based monitoring of α-synuclein–s- odorant-binding protein from an Asian honey
mall molecule interactions. ACS Appl Mater bee, Apis cerana cerana. J Chem Ecol
Interfaces 7:4081–4088 34:1593–1601
12. Boltasseva A (2009) Plasmonic components 24. Jaworski JW, Raorane D, Huh JH et al (2008)
fabrication via nanoimprint. J Opt A Pure Evolutionary screening of biomimetic coatings
Appl Opt 11:114001 for selective detection of explosives. Langmuir
24:4938–4943
13. Pimpin A, Srituravanich W (2011) Review on
micro-and nanolithography techniques and 25. Smith RG, D’Souza N, Nicklin S (2008) A
their applications. Eng J 16:37–56 review of biosensors and biologically-inspired
systems for explosives detection. Analyst
14. Zhang D, Lu Y, Jiang J et al (2015) Nanoplas- 133:571–584
monic biosensor: coupling electrochemistry to
localized surface plasmon resonance spectros- 26. Ramoni R, Bellucci S, Grycznyski I et al (2007)
copy on nanocup arrays. Biosens Bioelectron The protein scaffold of the lipocalin odorant-
67:237–242 binding protein is suitable for the design of
new biosensors for the detection of explosive
15. Zhang D, Lu Y, Zhang Q et al (2015) Nano- components. J Phys Condens Matter
plasmonic monitoring of odorants binding to 19:395012
 Nanoplasmonic Biosensor Using Localized Surface Plasmon Resonance. . . 107

27. Kuang Z, Kim SN, Crookes-Goodson WJ et al 28. Zhang D, Zhang Q, Lu Y et al (2015)

(2009) Biomimetic chemosensor: designing Electrophoresis-enhanced nanoplasmonic bio-
peptide recognition elements for surface func- sensor with nanocup arrays for protease detec-
tionalization of carbon nanotube field effect tion in point-of-care diagnostics. China
transistors. ACS Nano 4:452–458 Nanomed 2015:202
 Chapter 7

Plasmonics-Based Detection of Virus Using Sialic Acid

Functionalized Gold Nanoparticles
Changwon Lee, Peng Wang, Marsha A. Gaston, Alison A. Weiss,
and Peng Zhang

Abstract
Biosensor for the detection of virus was developed by utilizing plasmonic peak shift phenomenon of the
gold nanoparticles and viral infection mechanism of hemagglutinin on virus and sialic acid on animal cells.
The plasmonic peak of the colloidal gold nanoparticles changes with the aggregation of the particles due to
the plasmonic interaction between nearby particles and the color of the colloidal nanoparticle solution
changes from wine red to purple. Sialic acid reduced and stabilized colloidal gold nanoparticle aggregation
is induced by the addition of viral particles in the solution due to the hemagglutinin-sialic acid interaction.
In this work, sialic acid reduced and stabilized gold nanoparticles (d ¼ 20.1  1.8 nm) were synthesized by a
simple one-pot, green method without chemically modifying sialic acid. The gold nanoparticles showed
target-specific aggregation with viral particles via hemagglutinin-sialic acid binding. A linear correlation was
observed between the change in optical density and dilution of chemically inactivated influenza B virus
species. The detection limit of the virus dilution (hemagglutinination assay titer, 512) was shown to be
0.156 vol% and the upper limit of the linearity can be extended with the use of more sialic acid-gold
nanoparticles.

Key words Biosensor, Sialic acid, Gold nanoparticle, Viral detection, Colorimetric measurement

1 Introduction

Metallic nanoparticles absorb and scatter with great efficiency when

interacting with light. This strong interaction between metallic
nanoparticles and light occurs because the oscillating electromag-
netic field of light initiates the coherent oscillation of the free
electrons of the metallic nanoparticles. This oscillation is termed
the surface plasmon resonance (SPR). SPR results in dipole oscilla-
tion along the direction of the electric field of light (Fig. 1) . The
amplitude of the oscillation reaches the maximum at certain fre-
quency of light and it is dependent on the particle size, shape, and
refractive index of the solution. For noble metallic nanoparticles,

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_7,
© Springer Science+Business Media LLC 2017

109
110 Changwon Lee et al.

Fig. 1 Schematic illustration of surface plasmon resonance of gold nanoparticles due to collective oscillation
of surface electrons by incident light of appropriate wavelengths

the SPR-induced strong absorption or scattering of light can be

easily measured by UV-Vis absorption spectrometer.
The SPR band position relies strongly on the size and shape of
individual nanoparticles and interparticle distance [1, 2]. Gold
nanoparticles typically display an SPR band at around 520 nm.
The assembly or aggregation of the metallic nanoparticles usually
results in a red shift of the plasmonic band. This SPR band shift can
be easily observed by naked eyes with the color of the colloidal gold
solution changing from red to purple to blue. Kreibig et al. showed
the relationship of SPR band position and interparticle distance
decades ago [3]. The interparticle distance-dependent SPR band
shift is attributed to the electric dipole-dipole interactions and
coupling between plasmons of the neighboring particles in the
aggregates. When the interparticle distance is greater than the
average particle diameter (dispersed state), the SPR band appears
to be red, whereas the color turns to blue when the interparticle
distance decreases to less than the average particle diameter (aggre-
gated state).
This phenomenon has been well adopted in various detection
schemes with specific targeting elements decorated on the nano-
particle surface for the targets of interest, such as polynucleotides
[4, 5], enzymes and proteins [6, 7], cells [8], and heavy metals [9].
In this chapter, we describe a method to develop the gold nano-
particles for the colorimetric detection of influenza virus based on
the interaction between hemagglutinin, a protein expressed on the
viral surface, and sialic acid, utilized as a surface stabilizing ligand
on gold nanoparticles through a simple one-pot synthesis [10].
Sialic acid is a surface ligand presented on the surface of lung
epithelial cells and it is recognized as the primary binding site for
influenza virus. Hemagglutinin is a surface protein on various viral
species that binds to sialic acid on targeted cell surface for the viral
infection process [11]. The binding of viral proteins to sialic acid
has been studied for various pathogenic viruses, such as influenza
virus [11–13], human parainfluenza virus [14], human coronavirus
[15, 16], and a specific serotype of rhinovirus [17]. Based on the
 Plasmonics-Based Detection of Virus Using Sialic Acid Functionalized Gold Nanoparticles 111

Fig. 2 Detection scheme of viral particles based on the aggregation of SA-AuNPs on the virus surface through
sialic acid-hemagglutinin binding

binding capability of sialic acid to certain viruses, it is possible to

design a colorimetric sensor for detecting virus using self-reporting
plasmonic nanoparticles, which display plasmon shift upon binding
to viral particles (Fig. 2). Such a quick and user-friendly detection
method based on the binding between hemagglutinin on the influ-
enza virus surface and sialic acid on the gold nanoparticle surface
provides an effective means to identify individuals or animals
infected with virus, and can help with early therapeutic intervention
and reduce the spread of infection.

2 Materials

2.1 Equipment IKAMAG® Safety Control heating magnetic stirrer (IKA).

Eppendorf bench-top centrifuge (Model 5424).
Nanotrac particle size analyzer (Microtrac).
Ocean Optics USB 4000 UV-Vis spectrometer.
Millipore Milli-DI System.

2.2 Sialic Acid and 1. 1.25 mM N-acetylneuraminic acid (SA) (Catalogue No. A2388,
Gelatin Gold Sigma-Aldrich, St. Louis, MO) was prepared in DI water.
Nanoparticle Synthesis 2. 0.1 wt% gelatin (Catalogue No. 53028, Sigma-Aldrich, St.
Louis, MO) was prepared in DI water.
3. 1 M sodium hydroxide (Catalogue No. S5881, Sigma-Aldrich,
St. Louis, MO) was prepared in DI water.
4. 0.02 M chloroauric acid (HAuCl4) (Catalogue No. 254169,
Sigma-Aldrich, St. Louis, MO) was prepared in DI water.
112 Changwon Lee et al.

2.3 Influenza Virus 1. Dulbecco’s Modified Eagle Media (DMEM) (Catalogue No.
Deactivation 11965, Life Technology, Grand Island, NY) was used for the
culture of virus.
2. Fetal Bovine Serum (FBS) (Catalogue No. 16000-044, Life
Technology, Grand Island, NY) was added to 10% in DMEM.
3. β-propiolactone (Catalogue No. P5648, Sigma-Aldrich, St.
Louis, MO) was added to 0.05% of final concentration in the
viral solution.

3 Methods

3.1 Sialic Acid Gold 1. Ten milliliter of 1.25 mM sialic acid (SA) solution was prepared
Nanoparticle in DI water at room temperature. The SA solution was mixed
(SA-AuNP) Synthesis with 250 μL of 0.02 M HAuCl4 followed by the addition of
50 μL 1 M NaOH in 20 mL glass vial (see Note 1).
2. The mixture was then stirred and heated for 80  C at 20  g
for 15 min on a heating magnetic stirrer. The color of the
solution changed from yellow to a dark red wine.
3. After the solution was cooled to room temperature, the gold
nanoparticles were washed twice by centrifugation at
4500  g) using a bench-top centrifuge (Eppendorf 5424)
for 20 min. After the centrifugation, supernatant was removed
and the dark red pellet was suspended in DI water and stored
until further use.

3.2 Influenza Virus 1. Virus solution was provided in 10% FBS containing DMEM
Inactivation and it was inactivated by the addition of β-propiolactone (final
concentration of 0.05%) for 1 h at 37  C.
2. Inactivated virus was stored at 20  C until further use.

3.3 Particle Size Particle size analysis was performed on a Nanotrac particle size
Analysis analyzer (Microtrac) with a built-in liquid sample holder. The
concentration of the sample was adjusted for the optimal measure-
ment condition. Three 30-s measurements were averaged for parti-
cle size determination, with results shown in Fig. 3c.

3.4 Determination of To determine the concentration of the SA-AuNPs solution, the

SA-AuNP volume-density method was used based on the assumption of
Concentration face-centered cubic (fcc) gold structure and 100% reaction yield.
First, the following equation was used to calculate the number of
gold atoms in a gold nanoparticle.
4πρr 3 N A
N ¼ 
3 M
(N ¼ number of atoms in a gold nanoparticle, ρ ¼ density of
fcc gold (19.3 g/cm3), r ¼ radius of nanoparticle (10.05 nm), NA is
 Plasmonics-Based Detection of Virus Using Sialic Acid Functionalized Gold Nanoparticles 113

A 0.8 B 0.25
0.7
0.2
0.6
0.5 Increase
0.15
over time

OD
0.4
OD

0.3 0.1
0.2
0.05
0.1
0 0
400 500 600 700 800 0 5 10 15
Wavelength (nm) Time (min)
C D100 20
100 50
90 90
80 40 80
70 70

%Passing
%Channel

%Channel
%Passing

60 30 60
50 50 10
40 20 40
30 30
20 10 20
10 10
0 0 0 0
0.1 1 10 100 1,000 10,000 0.000 0.001 0.01 0.1 1 10
Size (Nanometers) Size (Microns)

Fig. 3 Typical results of the UV-Vis spectroscopic results and particle size analysis results of SA-AuNP before
and after virus incubation. UV-Vis absorption spectra measured at 0, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, and 12 min
after the addition of 1.25 vol% influenza B/Victoria. (a) Typical UV-Vis spectral change of SA-AuNP solution after
the addition of the virus, (b) change of OD610 over time, (c) particle size analysis result measured before the virus
addition, and (d) particle size analysis result measured after the addition of 1.25 vol% influenza B/Victoria

Avogadro’s number, and M is gold atom’s atomic weight

(196.97 g/mol).
The calculation result yielded ~250,000 gold atoms in 20.1 nm
spherical-shaped gold nanoparticles. Since the initial concentration
of HAuCl4 is 0.5 mM, the final SA-AuNP concentration is 2 nM
when resuspended in the same volume of DI water.

3.5 UV-Vis Five hundred microliter of the solution containing SA-AuNP with or
Measurements without virus was added into the 1 cm path length quartz cuvette
and placed in the cuvette holder of the UV-Vis spectrometer (Ocean
Optics USB 4000 UV-Vis spectrometer). Then the UV-Vis spec-
trum was measured immediately. Results are shown in Fig. 3a, b.

3.6 Particle Size Particle size of the SA-AuNP with virus was measured in the same
Analysis way as described in Subheading 3.3. In short, the solution contain-
Measurements of SA- ing virus incubated SA-AuNP was placed in the built-in liquid sample
AuNP with Virus holder of the particle size analyzer (Microtrac). The sample concen-
tration was adjusted for the optimum result, as shown in Fig. 3d.

3.7 Colorimetric 1. Five hundred microliter of SA-AuNP solution was mixed with
Detection of Viral different dilutions of influenza virus solution and immediately
Particles added to a 1-cm path length quartz cuvette.
114 Changwon Lee et al.

2. The cuvette was then placed in UV-Vis spectrometer (USB

4000, Ocean Optics) and absorption spectra were measured
every minute until a plateau was reached.
In a typical experiment, it was observed that the absorption
spectrum of SA-AuNPs solution changed immediately after the
addition of influenza B/Victoria solution, with increasing OD
value at 600–610 nm and decreasing OD value at 510 nm over
time. The spectrum becomes stabilized after several minutes, as
shown in Fig. 4. The absorbance of 0.4 nM SA-AuNPs solution
was measured upon the addition of different dilutions of influenza
B/Victoria. We observe that, when the virus concentration was
<1.25 vol%, there was a linear correlation between the change of
OD value (ΔOD) and the amount of added virus solution, with the
detection limit estimated as 0.09 vol%. The final solutions after each
test illustrated the gradual change of color from red to purple as the
amount of virus solution increased (photograph in Fig. 4) .

0.5 0.3 0 min 0.5 0.15

a 0.5 min b 0.10

0.2 1 min

OD610
OD610

2 min
0.4 0.1
3 min
0.4 0.05

4 min 0.00
0.0
0 10 20 30 5 min 0 1 2 3 4 5
0.3 Time (min) 6 min 0.3 Time (min)
OD

7 min
OD

8 min 0 min
0.2 0.2 1 min
2 min
3 min
Increase
0.1 0.1 Increase 4 min
over time 5 min
over time
0.0 0.0
300 400 500 600 700 800 300 400 500 600 700 800
Wavelength (nm) Wavelength (nm)

0.20
c d

0.15
delta (OD610 )

0.10 y=0.104x+0.0003
2
R =0.9915

0.05

0.00
0 2 4 6
Virus (v%)

Fig. 4 (a) Absorption spectra of 0.4 nM SA-AuNPs over time after the addition of 2.5 vol% influenza B/Victoria
solution; inset, OD610 value change. (b) Absorption spectra of 0.4 nM SA-AuNPs over time after the addition of
0.156 vol% virus solution; inset, change in OD610. (c) Change in OD610 value after reaching plateau for
different dilutions of virus. A linear correlation is observed for dilutions <1.25 vol% virus. (d) Photograph of
SA-AuNPs with different dilutions of virus, 5, 2.5, 1.25, 0.625, 0.3125, 0.156, and 0 vol% from left to right,
respectively
 Plasmonics-Based Detection of Virus Using Sialic Acid Functionalized Gold Nanoparticles 115

0.30 0.20
A B
0.25
0.15

delta (OD610)
0.20
delta (OD610)

0.15 0.10
y=0.0957x+0.0003
y=0.0974x+0.0183 2
0.10 2 R =0.9970
R =0.9999
0.05
0.05

0.00 0.00
0 2 4 6 0 2 4 6
Virus (v%) Virus (v%)

Fig. 5 (a) Change in OD610 measured for 0.8 nM SA-AuNP solution after the addition of influenza B/Victoria. A
linear correlation is observed for dilutions <2.5 vol% virus. (b) Change in OD610 measured for 0.4 nM SA-AuNP
solution after the addition of influenza B/Yamagata. A linear correlation is observed for dilutions <1.25 vol%
virus

3.8 Increasing the To increase the linear range of ΔOD vs. concentration of added
Linear Dynamic Range virus solution, we used a higher concentration of SA-AuNP solu-
of Virus Measurements tion (0.8 nM) to conduct the similar tests, with results shown in
Fig. 5a. It was observed that the linear range of ΔOD vs. concen-
tration of added virus solution was extended to 2.5 vol% of virus in
the solution. These results illustrate the importance of using a large
amount of SA-AuNPs in the solution for virus detection when
applying this scheme. We note that this method can also detect
another virus subtype, influenza B/Yamagata. Results obtained
from the influenza B/Yamagata solutions are similar to those of
influenza B/Victoria and shown in Fig. 5b.
In conclusion, we demonstrate a simple method to synthesize
gold nanoparticles using sialic acid as both the reducing agent and
the stabilizing agent. Experimental results support the hypothesis
that sialic acid molecules on SA-AuNP surface interact with viral
envelope protein hemagglutinin, causing a colorimetric change of
the SA-AuNP solution. The resulting SA-AuNPs can readily detect
influenza B virus (HA titer of 512) diluted to 0.156 vol%. The
method is effective for different influenza B lineages (Victoria and
Yamagata) .

4 Notes

1. The amount of 1 M NaOH can be varied in the range of

20–100 μL to achieve the best results, as the local deionized
water may have slightly different pH values.
116 Changwon Lee et al.

Acknowledgment

P.Z. acknowledges support from the National Science Foundation

(CBET-0931677/1065633). A.A.W. acknowledges National
Institute of Health (R01AI089450 from NIAID) for support.

References
1. Zhong Z, Patskovskyy S, Bouvrette P, Luong properties of gold nanoparticles. J Am Chem
JHT, Gedanken A (2004) The surface chemis- Soc 130:8038–8043
try of Au colloids and their interactions with 10. Lee C, Gaston MA, Weiss AA, Zhang P (2013)
functional amino acids. J Phys Chem B Colorimetric viral detection based on sialic acid
108:4046–4052 stabilized gold nanoparticles. Biosens Bioelec-
2. Link S, El-Sayed MA (1999) Size and temper- tron 42:236–241
ature dependence of the plasmon absorption of 11. Sauter NK, Hanson JE, Glick GD, Brown JH,
colloidal gold nanoparticles. J Phys Chem B Crowther RL, Park SJ, Skehel JJ, Wiley DC
103:4212–4217 (1992) Binding of influenza virus hemaggluti-
3. Kreibig U, Genzel L (1985) Optical absorption nin to analogs of its cell-surface receptor, sialic
of small metallic particles. Surf Sci acid: analysis by proton nuclear magnetic reso-
156:678–700 nance spectroscopy and X-ray crystallography.
4. Elghanian R, Storhoff JJ, Mucic RC, Letsinger Biochemistry 31:9609–9621
RL, Mirkin CA (1997) Selective colorimetric 12. Varghese JN, Colman PM, van Donkelaar A,
detection of polynucleotides based on the Blick TJ, Sahasrabudhe A, McKimm-Breschkin
distance-dependent optical properties of gold JL (1997) Structural evidence for a second
nanoparticles. Science 277:1078–1081 sialic acid binding site in avian influenza virus
5. Storhoff JJ, Marla SS, Bao P, Hagenow S, neuraminidases. Proc Natl Acad Sci U S A
Mehta H, Lucas A, Garimella V, Patno T, 94:11808–11812
Buckingham W, Cork W, M€ uller UR (2004) 13. Weis W, Brown JH, Cusack S, Paulson JC,
Gold nanoparticle-based detection of genomic Skehel JJ, Wiley DC (1988) Structure of the
DNA targets on microarrays using a novel opti- influenza virus haemagglutinin complexed
cal detection system. Biosens Bioelectron with its receptor, sialic acid. Nature
19:875–883 333:426–431
6. Thanh NTK, Rosenzweig Z (2002) Develop- 14. Suzuki T, Portner A, Scroggs RA, Uchikawa
ment of an aggregation-based immunoassay for M, Koyama N, Matsuo K, Suzuki Y, Takimoto
anti-protein a using gold nanoparticles. Anal T (2001) Receptor specificities of human
Chem 74:1624–1628 respiroviruses. J Virol 75:4604–4613
7. Pavlov V, Xiao Y, Shlyahovsky B, Willner I 15. Vlasak R, Luytjes W, Spaan W, Palese P (1988)
(2004) Aptamer-functionalized Au nanoparti- Human and bovine coronaviruses recognize
cles for the amplified optical detection of sialic acid-containing receptors similar to
thrombin. J Am Chem Soc 126:11768–11769 those of influenza C viruses. Proc Natl Acad
8. Medley CD, Smith JE, Tang Z, Wu Y, Bam- Sci U S A 85:4526–4529
rungsap S, Tan W (2008) Gold nanoparticle- 16. Kunkel F, Herrler G (1993) Structural and
based colorimetric assay for the direct detec- functional analysis of the surface protein of
tion of cancerous cells. Anal Chem human coronavirus OC43. Virology
80:1067–1072 195:195–202
9. Darbha GK, Singh AK, Rai US, Yu E, Yu H, 17. Uncapher CR, DeWitt CM, Colonno RJ
Chandra Ray P (2008) Selective detection of (1991) The major and minor group receptor
mercury (II) ion using nonlinear optical families contain all but one human rhinovirus
serotype. Virology 180:814–817
 Chapter 8

MicroRNA Biosensing with Two-Dimensional Surface

Plasmon Resonance Imaging
Ho Pui Ho, Fong Chuen Loo, Shu Yuen Wu, Dayong Gu,
Ken-Tye Yong, and Siu Kai Kong

Abstract
Two-dimensional surface plasmon resonance (2D-SPR) imaging, which provides a real-time, sensitive, and
high-throughput analysis of surface events in a two dimensional manner, is a valuable tool for studying
biomolecular interactions and biochemical reactions without using any tag labels. The sensing principle of
2D-SPR includes angular, wavelength, and phase interrogation. In this chapter, the 2D-SPR imaging
technique is applied for sensing a target microRNA by its corresponding oligonucleotide probes, with
sequence complementarity, immobilized on the gold SPR sensing surface. However, the low SPR signal due
to intrinsic properties such as low molecular weight and quantity (pico-nanomolar) of the microRNA in
clinical samples limits the direct detection of microRNA. Therefore, we developed a biosensing technique
known as MARS (MicroRNA-RNase-SPR) assay, which utilizes RNase H to digest the microRNA probes
enzymatically for fast signal amplification, i.e., in order to increase both the SPR signal and readout speed
without the need for pre-amplification of target cDNA by polymerase chain reaction (PCR). Practically, we
targeted microRNA hsa-miR-29a-3p, whose signature correlates to influenza infection, for rapid screening
of influenza A (H1N1) patients from throat swab samples.

Key words Biosensing, Disease screening, Influenza, MicroRNA, MicoRNA-RNase-SPR, Two-

dimensional surface plasmon resonance, Signal amplification

1 Introduction

Surface plasmon resonance (SPR), a physical phenomenon of plas-

monic propagation change near the surface of a nanometer-scale
thin metallic (gold, silver) sensing layer, is a sensitive method to
detect localized refractive index change on the sensing surface [1].
Two-dimensional surface plasmon resonance (2D-SPR) biosen-
sing, sharing the same principle of SPR, has been extensively used
in the last 10 years to study bimolecular interactions [2–5]. The
main reason behind this growth is that 2D-SPR imaging provides
the advantages of real time, label free, and high throughput for
rapid, sensitive detection of target biological binding reactions such

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_8,
© Springer Science+Business Media LLC 2017

117
118 Ho Pui Ho et al.

as antibody–antigen or protein–DNA interactions. There is also a

trend toward detection of low-molecular-weight biomolecules such
as small peptides and oligonucleotides. Researchers have used 2D-
SPR sensing technology for rapid DNA and protein profiling [4, 5].
In this section, we give a brief overview of the principle of SPR and
the development of 2D-SPR for biomedical applications.
The SPR phenomenon is an excited charge density oscillation
induced by optical light propagating along the boundary between a
metal film, such as a gold or silver layer, and a dielectric medium.
The incident optical light is composed of p-polarized (transverse
magnetic polarization) and s-polarized (transverse electric polariza-
tion) light. Only p-polarized light is normal to the boundary and
propagation along the x-axis of a metal film, so it can couple into a
surface plasmon mode for electron oscillation on the film. The
energy of electron oscillation becomes heat and disappears, leading
to the SPR phenomenon. However, the plasmonic propagation
change in SPR occurs on the nanometer scale, so direct observation
of the propagation is impossible with present technologies. There-
fore various techniques, namely angular, wavelength, or phase
interrogation schemes, are used for SPR signal extraction. Angular
interrogation is the measurement of the reflectivity variation with
respect to the incident angle under monochromatic light illumina-
tion. The SPR coupling angle is the reflected light intensity at its
minimum point, or absorption dip (see Fig. 1a, b). It is correlated to
the change in the dielectric constant value of the sample medium,
or the refractive index of the sensing surface. In practice, measuring
angular shift of SPR absorption dip or intensity shift at SPR absorp-
tion dip at a fixed optimal reflected angle can reveal any biomolec-
ular event on the sensing surface. The wavelength interrogation
technique, unlike angular interrogation, uses a fixed illumination
angle and observes the reflectivity variation across the wavelengths
to address the spectral absorption dip, indicating the presence of
SPR, by spectrometer. The spectral location of the adsorption dip,
namely resonant wavelength, is determined by the combined effect
of the spectral dispersion of the prism and sensing layer, as well as
sample medium. In this case, measuring the wavelength shift of
SPR absorption dip or intensity shift at SPR absorption dip at a
fixed optimal wavelength can detect any biomolecular interaction
on the sensing surface. Phase interrogation is based on a resonant
effect in the SPR system in which in or out of resonance signals
cause a massive phase jump of the incident optical wave. The
absorption dip in intensity-angle or wavelength diagram corre-
sponds to the maximum resonance slope in the phase-angle or
wavelength diagram (see Fig. 1c) . The steepness and extent of the
resonance slope, caused by massive phase jump, depends on both
the material and the thickness of the sensing surface. As the phase
jump at the absorption dip leading to a steep phase change across
resonance (which can be readily monitored by an optical
 MicroRNA-RNase-SPR Biosensing 119

A B

Reflected light
Intensity
Sample medium

Sensing surface Absorption dip

Angle /
Prism C Wavelength

Phase
Monochromatic or Resonance slope
broadband light source Reflected light

Angle /
Wavelength

Fig. 1 Angular, wavelength, and phase interrogation schemes are illustrated. SPR sensing surface exhibits the
SPR phenomenon under optical wave using monochromatic or broadband light source. The SPR phenomenon
is observed from the analysis of reflected light (a). In angular interrogation, incident angle of monochromatic
light (fixed wavelength) is scanned to identify the absorption dip of the reflected light. In wavelength
interrogation the spectrum of incident light at a fixed angle results in an absorption dip at a particular
wavelength of the spectral reflected light (b). In phase interrogation, SPR signal extraction of reflected light
search for maximum resonance slope, which corresponds to absorption dip in above two schemes (c). Any
bimolecular event on the sensing surface will lead to a shift in the absorption dip, and analysis of the shift in
absorption dip generates the SPR signal readout

interferometer), the SPR signal has a better sensitivity factor than

angular and wavelength methods using refractive index units (RIU)
to represent SPR sensitivity, i.e., 10
8–10
9 RIU, in which sensi-
tivity factor is 10
6–10
7 RIU. In practice, measuring the phase
change of SPR resonance at a fixed optimal angle and wavelength
can detect any biomolecular reactions on the sensing surface.
Meanwhile, 2D-SPR, also called SPR imaging (SPRI), is the simul-
taneous sensing of multiple regions of a SPR surface. The optical
setups of SPR and SPRI are very similar except for two compo-
nents: illumination and detection. SPRI requires the detection area
to be evenly illuminated so that different regions of sensing probes
on the sensing surface are subjected to the same incident light
intensity. To achieve this it is more common to use LED than
laser illumination. 2D imaging detection is achieved either by 2D
photodiode arrays or CCD camera. The advantage of SPRI over
SPR for biological applications is mainly the ability of detecting
multiple targets, such as cancer biomarkers in one sensing surface.
Thus the speed of overall detection is increased allowing for rapid
diagnosis, especially in clinical use. However, imaging capability
120 Ho Pui Ho et al.

also results in a notable decrease in sensitivity. The sensitivity is

usually one-order magnitude lower in SPRI than in SPR, and a
typical angular interrogation scheme has a sensitivity of 10
5 RIU
in SPRI mode compared to the single-point SPR value of 10
6
RIU.
MicroRNA, the small noncoding RNA encoded from the
human genome to regulate many biological functions, has been
found to play an important roles the immune response during
influenza virus infection [6]. The microRNA expression pattern
thus acts as signature in identifying disease, and rapid detection of
microRNA helps in rapid disease diagnosis. 2D-SPR is used to
detect microRNA by the sequence complementary hybridization
with the oligonucleotide probes immobilized on the SPR sensing
surface. The small size and the concentration of microRNA lead to
small SPR signals, preventing effective detection. Therefore, we
describe the biosensing technique MicroRNA-RNase-SPR
(MARS) using the RNase H enzymatic reaction to improve SPR
signal amplification and readout speed without pre-amplification by
PCR [7]. The workflow for microRNA sensing in the MARS assay
is show in Fig. 2. Designated microRNA probes with spacers at the
30 end and biotin at the 50 end are immobilized on the SPR sensing

5⬘ 3⬘5⬘ cDNA
5⬘ 3⬘ A B
3⬘ C
Mature
Stem-looped
microRNA
primer Intact
hsa-miR-29a-3p D
microRNA
probe
Streptavidin RNase H
addition Digested
MicroRNA microRNA
probes probe

D RNase H
action

Fig. 2 Workflow for microRNA sensing in the MARS assay. First, mature microRNA is converted into cDNA by
stem-loop primers (top, A: Stem-loop primer hybridization; B: Reverse transcription; C: Heat separation/
denature; D: cDNA addition for sensing). On the SPR gold sensing surface (bottom), biotinylated-microRNA
probes with spacer C6 are immobilized through thio-linkage. Unbound region on the sensing surface is
covered with the blocker PEG 1000-SH. Streptavidins are then attached to the immobilized probes through the
biotin-avidin interaction. Target cDNAs with complementary nucleotide sequence are hybridized to the probes
to generate RNA–cDNA hybrids that increase the SPR signal. RNase H is added to digest the RNA probes and to
release the hybridized cDNAs from the RNA–cDNA hybrids, The released cDNAs then binds to the new probes
for further RNase H digestion, leading to the repeated cycle of RNase H reaction for signal amplification as
indicated by forward and backward semicircular arrows. Depletion of the RNA probes on the sensing surface
causes a decrease in SPR signal. Reprinted with permission from ref. 8
 MicroRNA-RNase-SPR Biosensing 121

surface by 30 -thiolated linkage. Since our small microRNA probe

leads to insignificant SPR signal change upon association or disso-
ciation, streptavidin - a 60 kDa protein used to increase SPR signal,
is added to the RNA probes as a cap on top of the probe through
biotin–avidin interaction. This sensing surface is used to capture
target cDNAs. RNase H enzymatic digestion removes a
streptavidin-conjugated microRNA probe from the SPR sensing
surface. Since the cDNA is not degraded, it repeatedly anneals to
another microRNA probe leading to further degradation of probes
as SPR signal amplies isothermally. We demonstrate the detection
of microRNA hsa-miR-29a-3p from throat swab samples for influ-
enza A (H1N1) patient screening. The screening is based on the
analysis of hsa-miR-29a-3p, which is downregulated after influenza
A H1N1 infection. We demonstrate use of an angular interrogation
2D-SPR imager with fixed angle to determine the intensity change
of SPR absorption dip. As mentioned previously, different SPR
interrogation systems, namely wavelength or phase, can be used
without changing the sensing surface MARS assay platform.

2 Materials

2.1 Two- 1. The GWC SPRimager II system (GWC Technologies) as a SPR

Dimensional Surface imager with temperature set at 25  C to preheat for at least 1 h
Plasmon Resonance before sensing application.
System 2. External pump (EP-1 Econo Pump; Bio-Rad) with connection
tubes of inner diameter 1.6 mm for sample loading site.
3. SF10 right-angle prism is used for SPR prism.

2.2 Surface 1. SPRchip™ (GWC Technologies), the uniform gold-coated

Chemistry SF10 glass chip is used as SPR sensing chip, with the ability
for 25 dots of 1 mm-diameter sensing probe to create 5  5
arrays.
2. Probe immobilization buffer is prepared sodium acetate at a
final concentration of 5 mM (optional with 10% glycerol) in
RNase- and DNase-free distilled water and stored at 4  C.
3. Blocking buffer PEG-1000-SH (Takara Biotechnology) is
prepared by dissolving the powder in absolute ethanol to
1 mM and stored at
20  C.
4. Synthesized microRNA probes (microRNA probe hsa-miR-29a-
3p: 50 -(biotin)-uagcaccaucugaaaucgguuauuuuuuuu-(C6)-SH-30 ;
microRNA probe hsa-miR-181-5p: 50 -(biotin)-aacauucaacgcu-
gucggugaguuuuuuuuu-(C6)-SH-30 ) are dissolved individually in
the RNase- and DNase-free distilled water at a final concentration
of 100 μM, aliquoted and stored at
20  C for single use.
122 Ho Pui Ho et al.

5. MicroRNA probes at working concentration are freshly

prepared by dilution to 1 μM with probe immobilization buffer
and stored at 4  C until use.

2.3 Sensing 1. SPR working buffer is prepared at 100 mM KCl, 50 mM Tris,

Solutions 10 mM MgCl2, 10 mM DTT in distilled water, and pH is
adjusted to 7.8 by HCl or KOH. This buffer is stored at
room temperature (see Note 1).
2. Synthesized cDNA (Takara, Inc.) (cDNA of hsa-miR-29a-3p:
TAACCGATTTCAGATGGTGCTA; cDNA of hsa-miR-181-
5p: ACTCACCGACAGCGTTGAATGTT) are dissolved in
distilled water at 100 uM and stored in single-use aliquots at

20  C.
3. Human microRNA sample is collected by throat swabs and
immersed into Hank’s balanced salt solution. Total RNA in
sample solution is extracted by miRNeasy Mini Kit (Qiagen)
according to manufacturing protocol. RNA yield is determined
by NanoDrop ND-1000 spectrophotometer (Thermo Scien-
tific). Target microRNA hsa-miR-29a-3p from 5 ng total RNA
is reversed transcribed into cDNA by TaqMan MicroRNA
Reverse Transcription Kit (Applied Biosystems) with mature-
microRNA-specific stem-loop primer according to the manu-
facturer’s protocol. The cDNA is stored at
20  C until use (see
Note 2).
4. Working solutions of microRNA probes are prepared by dilu-
tion with probe immobilization buffer to a final concentration
of 1 μM.
5. Recombinant streptavidin (Bioss) is stored at
20  C. Working
streptavidin solution is prepared freshly by dilution with the
SPR working buffer to 50 μg/mL.
6. Recombinant RNase H from Escherichia coli (Takara Biotech-
nology) is diluted from stock glycerol solution in SPR working
buffer at 60 U/mL (see Note 3) .

3 Methods

3.1 2D SPR Imager 1. SPRchip (GWC Technologies) from the sealed package is
Setup and Detection freshly opened, rinsed with RNase-free distilled water three
Workflow times, and then air dried before use.
2. MicroRNA probes (final concentration 1 μM, 0.2 μL each) are
blotted on the designated position of the SPR gold sensing
plate (optional with an injector) and incubated at 4  C for 16 h
(see Note 4).
3. Uncoated site of the SPR gold sensing plate is blocked with
PEG 1000-SH (final concentration 1 mM) at 25  C for 2 h.
 MicroRNA-RNase-SPR Biosensing 123

4. The prism is cleaned with absolute ethanol three times and then
air dried. Pump tubing and connection tubing are pre-rinsed
with RNase AWAY (Life Technologies) decontamination solu-
tion and then with RNase- and DNase-free distilled water (Life
Technologies) at 200 μL/min before use.
5. The gold sensing plate is inserted with the prism with refractive
index matching oil and placed into the SPR machine for SPR
signal detection.
6. The external pump is connected to SPR machine, sensing
surface is rinsed with SPR working buffer with a flow rate of
200 μL/min for 10 min (see Note 5).
7. Streptavidin working solution is added at 200 μL/min for
3 min.
8. The sensing surface is rinsed with SPR working buffer at
50 μL/min until a steady baseline with intensity fluctuation
less than 0.01 arbitrary unit is obtained.
9. SPR intensity reading is set to zero.
10. Sample cDNA or artificial ssDNA is added at 50 μL/min for
10 min.
11. Biosensing surface is rinsed with SPR working buffer at 50 μL/
min for 10 min.
12. RNase H working solution is added at 50 μL/min for 15 min.
13. Biosensing surface is rinsed with SPR working buffer at 50 μL/
min until a steady baseline is obtained.

3.2 SPR Sensorgram Specificity is important for this biomedical application since we aim
and Detection to apply it for rapid screening of multiple microRNA signatures.
Specificity This relies on the ability of the recognition element microRNA
probe to specifically capture the target, as well as the specificity of
RNase H to digest the probe for signal output. To evaluate this
specificity, microRNA probes of hsa-miR-29a-3p and hsa-miR-
181-5p, and single-stranded DNA (ssDNA) probe for miR-29a-
3p 3p, are immobilized on different regions of the sensing array, as
illustrated in Fig. 3. cDNAs of hsa-miR-29a-3p and hsa-miR-181-
5p, RNase H are added subsequently at the time points indicated in
the sensorgram of Fig. 3. There is a small increase in signal of
microRNA probe and ssDNA hsa-miR-29a-3p only after the injec-
tion of cDNA of miR-29a-3p, indicating the hybridization between
the microRNA probe and target cDNA occurs with high specificity.
A large decrease in signal of microRNA probe hsa-miR-29a-3p after
RNase H addition indicates that RNase H acts only on the RNA
probes from the RNA–DNA hybrid due to the interaction between
cDNA and microRNA probe, but not DNA–DNA hybrid, due to
the interaction between cDNA and ssDNA probe. Subsequent
injections of cDNA of hsa-miR-181-5p and RNase H cause a
124 Ho Pui Ho et al.

MicroRNA probe ssDNA probe

hsa-miR-29a-3p hsa-miR-29a-3p

MicroRNA probe PEG 1000-SH

has-miR-181-5p Blocker only
Gold sensing surface
cDNA of hsa-miR- cDNA of hsa-miR-
RNase H RNase H Buffer
29a-3p 181a-5p

1
0.5
0
Intensity (Arbitrary Unit)

−0.5

−1
−1.5
−2
−2.5
−3
−3.5
0 1000 2000 3000 4000 5000 6000
Time (second)

Fig. 3 Specificity of microRNA probe on target cDNA detection. The microRNA probes hsa-miR-29a-3p (blue),
hsa-miR-181-5p (green), ssDNA probe miR-29a-3p (red), or PEG 1000-SH blocker only (purple) are immo-
bilized on the surface of a SPR chip as shown in the upper panel. The cDNA of hsa-miR-29a-3p and hsa-miR-
181-5p (100 nM) and RNase H (60 U/mL) are added at the time points indicated in the lower panel. The signal
change is compared with the blank control (purple) to reveal the specificity for probe–target binding and
RNase H action on RNA–cDNA and ssDNA–cDNA hybrids. Reprinted with permission from ref. 8

change in the signal of the microRNA probe hsa-miR-181a-5p,

which like the pattern of hsa-miR-29a-3p, further supports the
specificity of this MARS assay. Saturating the sensor surface with
the blocker PEG 1000-SH causes no change in signal after cDNA
and RNase H addition, indicating the absence of nonspecific bind-
ing of cDNA and RNase H on the sensing surface.

3.3 SPR Detection of We focus on detecting microRNA hsa-miR-29a-3p from throat

microRNA in Throat swab samples to identify influenza A H1N1 infection. The cDNA
Swab Sample of a throat swab sample, instead of artificial cDNA, is added to the
sensing surface for MARS assay. Figure 4 shows the SPR signal of
healthy control and an H1N1-infected sample, with 10 nM artificial
cDNA for easy comparison. The larger decrease in signal in the
healthy control indicates that more hsa-miR-29a-3p is present. Our
published finding showed hsa-miR-29a-3p is correlated to
 MicroRNA-RNase-SPR Biosensing 125

0.1
Buf fer
0
Intensity (Arbitrary Unit)

−0.1

−0.2 H1N1
−0.3 10 nM ssDNA

−0.4

−0.5 Healthy

−0.6
0 100 200 300 400 500 600 700 800 900
Time (second)

Fig. 4 Screening throat swab samples by MARS assay. The cDNA from healthy control and H1N1-infected
samples are added to the SPR sensing surface for hsa-miR-29a-3p detection. Synthesized ssDNA (final
concentration at 10 nM) and buffer are also used in the MRSA assay as the standard and reference. The
concentration of hsa-miR-29a-3p in a patient sample could be determined from the standard curve of known
cDNA concentration of hsa-miR-29a-3p

influenza infection, where low expression level of has-miR-29a-3p

has been found in patients with influenza A H1N1 infection com-
pared with healthy individuals [8]. Using a standard curve having
standard concentrations of artificial cDNA, it is possible to quantify
the microRNA concentration in those samples for more accurate
disease screening [7].

4 Notes

1. Solutions must be prepared with RNase- and DNase-free dis-

tilled water, as minute amounts of RNase in solutions already
lead to SPR signal change. Also, detection of the MARS
response may lead to erroneous results if the pH of the reaction
solutions deviates from the range of pH 7.0–9.0. The reaction
may be halted and bimolecular elements may get denatured
when the pH of the reaction solution is out of this range.
Figure 5 represents the variation of SPR signals of the RNase
H reaction step due to the variation of pH of reaction solutions.
The optimal pH of the RNase H reaction is 7.8.
2. The microRNA extraction from human samples must be con-
ducted in a biosafety level-3 culture hood to prevent human
infection. The human sample immersed in Hank’s balanced salt
solution is stored at
80  C if microRNA extraction is not
performed immediately, in order to prevent nonspecific micro-
RNA digestion by RNase present in the human sample.
3. High concentrations of glycerol (above 10% v/v) inhibit enzy-
matic reaction and causes sudden changes in SPR signal. It is
126 Ho Pui Ho et al.

pH

6 6.5 7 7.5 7.8 8

0

Intensity (Arbitrary unit)

−0.5

−1

−1.5

***

Fig. 5 pH effect on RNase H activity in MARS assay. The RNase H activity under
different conditions is measured in the MARS assay by detecting the SPR
intensity signal change after addition of the cDNA of hsa-miR-29a-3p
(100 nM) on sensing surface with immobilized microRNA probes. SPR signal
change is recorded under different pH values in SPR working buffer. Results are
mean  SEM from three independent experiments and analyzed by Student’s t-
test, where ***p-value <0.001

advisable to dilute the glycerol to a final concentration of 0.1%

or lower.
4. The microRNA probe is carefully dotted on the sensing surface
using a pipette to prevent accidental mixing with the neighbor
probes. After this procedure, the sensing chip can be stored at
4  C for up to 2 weeks without affecting the SPR signal (see
Fig. 6).
5. To prevent false SPR signals, rinse thoroughly to ensure that
bubbles are removed from the sensing surface, otherwise the
bubbles will generate false signals.

Acknowledgment

The research is sponsored by ITF Grant (GHX/002/12SZ), CRF

grant (CUHK1/CRF/12G), AoE (AoE/P-0/12) funding from
the Hong Kong Special Administrative Region as well as the fund
(SGLH20121008144756945) from Shenzhen, China.
 MicroRNA-RNase-SPR Biosensing 127
Intensity (Arbitrary Unit)

Week of
incubation
0, 1, 2

3
4
5

No probe
control

Time (second)
Intensity (Arbitrary Unit)

Week of
incubation 0 1 2 3 4 5 No probe
control

Fig. 6 Stability of the microRNA probes on SPR surface. The microRNA probes immobilized on the SPR gold
sensing surface are stored at 4  C for the times indicated and subjected to SPR detection on target cDNA
hybridization, which increase intensity of the SPR signal. Results in the upper panel are illustrated as bar
charts in the lower panel, with mean  SEM from three independent experiments, where *p<0.05 compared
to SPR intensity signal at 0 weeks of incubation. Reprinted with permission from ref. 8

References
1. Kabashin AV, Patskovsky S, Grigorenko AN plasmon resonance differential phase imaging.
(2009) Phase and amplitude sensitivities in sur- Biosens Bioelectron 24:606–612
face plasmon resonance bio and chemical sens- 6. Tambyah PA, Sepramaniam S, Ali JM et al
ing. Opt Express 17:21191–21204 (2013) MicroRNAs in circulation are altered in
2. Wong CL, Chen GCK, Ng BK et al (2011) response to influenza A virus infection in
Multiplex spectral surface plasmon resonance humans. PLoS One 8:e76811
imaging (SPRI) sensor based on the polarization 7. Loo JFC, Wang SS, Peng F et al (2015) A non-
control scheme. Opt Express 19:18965–18978 PCR SPR platform using RNase H to detect
3. Shao Y, Li Y, Gu D et al (2013) Wavelength- MicroRNA 29a-3p from throat swabs. Analyst
multiplexing phase-sensitive surface plasmon 140:4566–4575
imaging sensor. Opt Lett 38:1370–1372 8. Peng F, He H, Loo JFC et al (2016) Identifica-
4. Wong CL, Ho HP, Suen YK et al (2007) Two tion of microRNAs in throat swab as the bio-
dimensional biosensor arrays based on SPR markers for diagnosis of influenza. Int J Med Sci
phase imaging. Appl Optics 46:2325–2332 13:77–84
5. Wong CL, Ho HP, Suen YK et al (2008) Real-
time protein biosensor arrays based on surface
 Chapter 9

Gold Nanorod Array Biochip for Label-Free, Multiplexed

Biological Detection
Zhong Mei, Yanyan Wang, and Liang Tang

Abstract
Gold nanorod (GNR) based label-free sensing has been attractive due to its unique property of localized
surface plasmon resonance (LSPR). Compared to bulk gold, the SPR of GNRs is more sensitive to the
refractive index change caused by biological binding in the close proximity. Numerous studies have
reported biological detection in solution based GNR probes. However, the biosensing has the intrinsic
problems of fluctuating readings and short storage time due to nanoparticle aggregation. In contrast, a
chip-based nanorod biosensor is a more robust and reliable platform. We have developed a nanoplasmonic
biosensor in a chip format by immobilizing functionalized GNRs on a (3-mercaptopropyl)trimethoxysilane
modified glass substrate. The covalent Au-S bond ensures a strong GNR deposition on the substrate. This
biochip exhibits a high sensitivity and stability when exposed to physiological buffer with high ionic
strength. Another advantage of GNR as optical transducer is its LSPR peak dependence on the aspect
ratio, which provides an ideal multiplexed detection mechanism. GNRs of different sizes that exhibit
distinct SPR peaks are combined and deposited on designated spots of a glass substrate. The spectral shift
of the respective peaks upon the biological binding are monitored for simultaneous detection of specific
analytes. Coupled with a microplate reader, this spatially resolved GNR array biochip results in a high-
throughput assay of samples as well as multiplexed detection in each sample. Since most biological
molecules such as antibodies and DNA can be linked to GNR using previously reported surface chemistry
protocol, the label-free nanosensor demonstrated here is an effective tool for protein/DNA array analysis,
especially for detection of disease biomarkers.

Key words Gold nanorods, Surface plasmon resonance, Biochip, Multiplex

1 Introduction

When the dimension of a matter is reduced to the nanometer scale

(1–100 nm), the material can exhibit distinct properties from its
bulk form, including optical, structural, electrical, and chemical
properties. These changes make nanoparticles promising for a
wide range of biomedical applications such as biosensing, imaging,
and drug delivery [1–3]. Among diverse nanomatrials, plasmonic
nanoparticles (Au, Ag) are attractive because of their unique optical

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_9,
© Springer Science+Business Media LLC 2017

129
130 Zhong Mei et al.

property, surface plasmon resonance (SPR) which arises from free

electron oscillation induced by electromagnetic radiation (i.e.,
light) [4]. The SPR results in a strong absorption of the incident
light at specific wavelength region which can be measured by a
UV–Vis spectrophotometer. The SPR band intensity and wave-
length depends on the metal type, particle size, shape, composition,
and the dielectric constant of the surrounding medium, as theoret-
ically depicted by Mie theory [5]. For a given nanosphere, the SPR
condition is εr ¼ 2εm, where εr is the dielectric function of the
metal, and εm is the dielectric constant of the medium. This indi-
cates an increase in εm requires an increase in εr to satisfy the SPR
condition. Practically, when the refractive index (n) of the sur-
rounding medium increases, the extinction spectrum shifts to lon-
ger wavelengths (known as red-shift). Since most biomolecules
(n  1.4–1.45) have a higher refractive index than water
(n  1.33), a red shift of the SPR peak is observed when the
biomolecules are absorbed to the nanoparitlce surface in aqueous
medium. If the nanoparticle is functionalized with special probe
(i.e., antibody), the shifts can be induced by the specific target (i.e.,
antigen) binding to the probe. Thus, by monitoring the SPR shifts
using UV–Vis spectrophotometer, biological events on the surface
of plasmonic nanoparticles can be quantitatively detected. This is
the mechanism of a label-free SPR biosensor.
Recently, gold nanoparticles (GNPs) have been widely used in
biosensing because of their facile surface modification, chemical
stability and excellent biocompatibility [6, 7]. GNPs with various
shapes including nanosphere, nanarod, nanostar, nanoshell, and
nanocage are reported for different sensing purposes [8]. Both
theoretical study and experimental results have shown that nanor-
ods have a greater performance than nanospheres, which is char-
acterized by the sensitivity (i.e., plasmon peak shift per unit change
in the effective refractive index of the surrounding medium,
RIU1). Given its simple synthesis and large tunability of surface
plasmon, gold nanorods (GNRs) have been a good alternative to
other shapes in SPR-based biosensing [9]. Different from bulk gold
films utilized in conventional SPR techniques, the anisotropic shape
and nanosize effect enable gold nanorods an extremely intensified
local electromagnetic field along its longitudinal direction, which is
highly sensitive to changes in the local refractive index [10]. Once
biological events occur at the surface of GNRs, the change in the
local refractive index will cause a longitudinal peak shift of the
extinction spectra of GNRs, which can be measured by a UV–Vis
spectrophotometer. Compared to the instrumental setup in tradi-
tional SPR techniques [11], this nanosensing method is more
sensitive [12], user-friendly, and cost-effective.
In addition, the longitudinal SPR wavelength of GNR is tun-
able in a broad range from UV–visible to far infra-red region by
varying its aspect ratio (length to width ratio). With advance in the
 Gold Nanorod Array Biochip for Label-Free, Multiplexed Biological Detection 131

Fig. 1 (a) UV–Vis spectra of synthesized GNRs; (b) TEM images of GNRs. The longitudinal plasmon peak is
tunable by varying rod aspect ratio

seed-mediated growth approach [13, 14], high-quality GNR can

be synthesized with precise control of the size and shape. Figure 1
shows the extinction spectra of six GNRs and their respective aspect
ratios. As the nanorod size increases, the longitudinal plasmon peak
132 Zhong Mei et al.

exhibits proportional red shifts. The correlation between aspect

ratio (A) and the longitudinal SPR peak (λLSPR/nm) can be
expressed as follows
λLSPR ¼ 84:4A þ 497:8 ð1Þ

This unique property facilitates a multiplexed biosensing.

GNRs of different aspect ratios can be mixed to fabricate multi-
plexed bioprobes to detect specific analytes simultaneously. For
example, if one sample contains two different antigens (antigen 1
and 2) to be detected/quantified simultaneously, GNR 1
(AR ¼ 2.5, LSPR λ ¼ 682 nm) can be functionalized with antibody
1 for antigen 1 detection while GNR 2 (AR ¼ 5.1, LSPR
λ ¼ 920 nm) is functionalized with antibody 2. These two functio-
nalized GNRs are then mixed and there will distinct plasmon bands
for the specific GNR on the absorption spectrum. The first peak at
~520 nm represents the characteristic gold resonance in transverse
direction which is less sensitive to local refractive index changes.
The second peak at 682 nm and the third peak at 920 nm reflect the
longitudinal peaks from GNR1 and GNR2, respectively. These two
dominate peaks are distinctly separated, thereby enabling a simul-
taneous monitoring of plasmonic shift at both positions indepen-
dently for the dedicated target antigen detection in a single sample.
However, solution-based GNR probe has its intrinsic problems.
First, multiple washes cannot be avoided in the process of fabricat-
ing GNR bioprobes. These cause extinction intensity change and
fluctuate readings. Second, the stability of a GNR suspension
depends on temperature and surfactant concentration, which
results in short storage time of the GNR probes. As such, it is
highly desired to develop a GNR nanosensor in a chip format
which significantly strengthen the utility of this label-free biosen-
sing modality as a powerful bioanalytical device [15].
Given that antibodies have primary amine group, it is feasible to
chemically modify antibody molecules with thiolation reagent to
enable gold–antibody linkage [16]. Thus, these GNR-conjugated
antibodies can work as biological receptors or bioprobes for detect-
ing specific antigens. Since many antibodies or antigens are disease-
representing biomarkers, rapid detection with a low-cost, specific
and sensitive biosensor is of great significance for clinical diagnosis.
Therefore, we develop a chip-based GNR biosensor for antigen
detection and then we extend this biochip into an array format,
which enables high-throughput screening as well as multiplexed
detection. Figure 2 is a schematic of three processes involved in a
label-free, multiplexed LSPR biochip. Human IgG and rabbit IgG
were used as a general analyte for demonstration. First, GNRs of
varying sizes are functionalized with anti-human IgG and anti-
rabbit IgG, respectively. Next, the functional GNRs are immobi-
lized to a thiolated glass substrate to construct chip-based LSPR
 Gold Nanorod Array Biochip for Label-Free, Multiplexed Biological Detection 133

Fig. 2 Schematic of GNR biofunctionalization and design of label-free, multiplexed biosensing by GNRs of
different aspect ratios. (a) Thiolation of antibody, followed by GNR functionalization. (b) Fabrication of
multiplex GNR biochip. (c) Optical detection

sensor. When the antigens (human IgG and rabbit IgG) are present
in the test samples, they will specifically bind to their respective
antibodies (i.e., anti-human IgG and anti-rabbit IgG). Upon this
binding, pronounced red shifts at the respective dominate LSPR
bands are monitored by UV–Vis absorbance measurements. The
magnitude of the red shift in the distinct LSPR peak wavelength is
directly proportional to the amount of antigen binding onto the
GNR probes.

2 Materials

2.1 Chemicals (See 1. Gold(III) chloride trihydrate (HAuCl4: 99%).

Note 1) 2. Sodium borohydride (NaBH4:99%).
3. L-ascorbic acid (AA: 99%).
4. Cetyltrimethylammoniumbromide (CTAB).
5. Sodium oleate (NaOL).
6. Silver nitrate (AgNO3:99%).
7. (3-mercaptopropyl)trimethoxysilane (MPTMS).
134 Zhong Mei et al.

8. Poly (ethylene glycol) methyl ether thiol (SH-PEG,

Mw5000).
9. 2-iminothiolane hydrochloride (Traut’s reagent: 98%).

2.2 Working 1. HAuCl4 is dissolved in MilliQ water at 10 mM and stored at

Solutions 4  C in dark.
2. Bisurfactant solution: a mixture of 7.0 g CTAB and 1.237 g
NaOL is added into 250 mL distilled water and sonicated at
50  C for completely dissolving.
3. NaBH4 is dissolved in ice-cold water at 10 mM before use.
4. Piranha solution: a 3:1 mixture of concentrated sulfuric acid
(98%) and hydrogen peroxide (30%) solution.
5. Phosphate buffered saline (PBS, pH 7.4, 0.01 M) is prepared
by dissolving one pouch of dry powder in 1 L deionized water.
6. Goat anti-human IgG and anti-rabbit IgG (2 mg/mL) are
diluted at 10 μg/mL with PBS and stored at 20  C before
use.
7. Human serum IgG and rabbit serum IgG are dissolved at
various concentrations (10, 20, 30, 40, 50, 60 nM) with PBS.

2.3 Instruments 1. Beckman-Coulter UV-NIR spectrophotometer (DU@720,

Fullerton, CA).
2. BioTek Multi-Detection Microplate Reader (Synergy™ 2,
Winooski, VT).
3. MilliQ@ Direct Water Purification System instruments (EMD
Millipore Corporation, Germany).

3 Methods (All Procedures Are Performed at Room Temperature Unless Notified

Otherwise)

3.1 Chip Based 1. Gold seed solution is prepared by adding 0.025 mL of 10 mM

Nanoplasmonic HAuCl4 into 1 mL of aqueous 0.1 M CTAB, followed by
Biosensing for Single adding 1 mL of 10 mM fresh-made ice-cold NaBH4. The
Antigen resulting solution is incubated at 27  C for 30 min.
3.1.1 Preparation of Gold
2. To prepare gold growth solution, 4 mM AgNO3 is added to
Nanorods
20 mL of bisurfactant solution. The mixture is kept undis-
turbed at 30  C for 15 min. Then 10 mL HAuCl4 is added,
followed by 90 min of stirring. 12.1 M HCl and 100 uL AA
(64 mM) is added.
3. A specific amount of the seed solution is added to the growth
solution. The resulting mixture is incubated at 29  C overnight
for full nanorod growth. To synthesize nanorods of varying
aspect ratios, the amount of AgNO3, HCl and seed solution are
adjusted following the recipe (see Table 1).
 Gold Nanorod Array Biochip for Label-Free, Multiplexed Biological Detection 135

Table 1
Growth conditions for GNRs with respective longitudinal SPR wavelengths

AgNO3 (mL) HCl (μL) Seed (μL) λLSPR (nm)

0.6 120 32 630
1.44 120 32 775
1.92 120 32 840
1.92 140 64 950

4. The suspension is centrifuged twice at 8500 rpm for 30 min.

5 mL MilliQ water is then added to resuspend the solid pellets
and centrifuged again at 13000 rpm for 3 min. The resulting
pellets are then redispersed in 5 mL solution (see Note 2).

3.1.2 Fabrication of 1. Microscopy glass slides (ITO-coated, 7 mm  50 mm  0.7 mm;

Nanoplasmonic Biochip Delta Technologies, Loveland, CO) are cleaned with the pira-
nha solution and heated at 70  C for 40 min (see Note 3).
2. Rinse the glass slides with distilled water, ethanol and dry with
nitrogen gas.
3. Immerse the cleaned slides in a MPTMS (10% v/v in ethanol)
solution.
4. After 2.5 h of incubation, the slides are thoroughly rinsed with
ethanol, water and dried with nitrogen gas.
5. To functionalize GNRs with antibodies, 20 μL of Traut’s
reagents (5 mg/mL) are first added to 0.1 mL of anti-human
IgG for 2 h [16]. Then 200 μL of GNR solution and 100 μL of
SH-PEG (10 mg/mL) are added for incubation overnight (see
Note 4).
6. The anti-IgG conjugated GNRs are separated from the excess
antibody by centrifugation at 8000 rpm for 10 min and finally
redispersed in PBS.
7. Immerse the MPTMS-treated glass in the functionalized GNR
colloids. After 2 h of shaking at 700 rpm, the glass substrates
are rinsed several times with water and dried with nitrogen gas
(see Note 5) .

3.1.3 Antigen Detection 1. Position the GNR-modified glass chip into a UV cuvette
(2.5 mL, 12.5  12.5  45 mm; BrandTech Scientific,
Wertheim, Germany) and insert the cuvette to the cuvette
holder of the Beckman-Coulter UV-NIR spectrophotometer,
as shown in Fig. 3. Measure the UV–Vis absorption spectra of
GNRs on glass as the baseline.
136 Zhong Mei et al.

Fig. 3 Schematic of the measurement of UV–Vis spectrum using Beckman spectrophotometer

Fig. 4 Detection of human IgG targets by anti-human IgG modified GNRs biochip. The black curve is the
absorption spectrum of anti-IgG modified GNRs biochip (baseline). The other six colored curves are absorp-
tions spectra of the GNR/anti-IgG chip incubated in the target human IgG samples of varying concentrations
(a–f: 10–60 nM). The peak wavelength at ~840 nm shows a red shift after IgG binding to the GNR/anti-IgG
chip. A higher concentration of IgG resulted in a larger red shift. Calibration curve of the red-shift versus IgG
concentration shows the longitudinal plasmon shift as a function of human IgG concentration

Dilute human IgG at a series of concentrations from

10 to 60 nM with PBS, then apply these samples on the GNR
biochip surface and incubate for 30 min.
2. After washing the chips with water, the UV–Vis absorption
spectra are measured again.
3. By comparing the spectra before and after human IgG conju-
gation, the red shift of longitudinal plasmon peak is observed in
proportional to the target concentration (Fig. 4, see Note 6).
 Gold Nanorod Array Biochip for Label-Free, Multiplexed Biological Detection 137

3.2 Nanoarray Based The methods for gold nanorod synthesis and functionlization are
Multiplexed similar as described above. To achieve multiplexed biosensing,
Biosensing for Multiple GNRs with varying sizes are mixed and deposited on one substrate.
Antigens Here, we take two-antibody detection for example.
1. GNRs with a LSPR wavelength at 840 nm are functionalized
with anti-human IgG, while GNRs with a LSPR at 630 nm are
functionalized with anti-rabbit IgG.
2. Microscopy glass slides (300  100  1.0 mm; Fisher Scientific,
Pittsburgh, PA) are treated with Piranha and MPTMS solution
as described above.
3. 10 μL of antibody-modified GNRs with mixed sizes are
dropped onto designated spots on the glass slides and incu-
bated for 2 h. The detection spots are arranged to mimic the
layout of a 96 well plate to accommodate an absorption reading
by the BioTek plate reader in a high-throughput fashion
(Fig. 5).
4. Wash the substrate with water three times and dry it with
nitrogen gas.
5. To perform a multiplexed detection, a sample solution (10 μL)
of mixed human IgG and rabbit IgG are dropped onto the
nanosensing spots and incubated for 30 min until equilibrium.
6. The glass substrate is attached to the bottom of a 96-well plate
for absorbance measurement. The UV–Vis absorbance mode is
chosen with a scanning range from 400 to 999 nm. As the light
source scans the spots from well to well, the corresponding
UV–Vis spectra are displayed in the same layout and ready for
analysis. Figure 6 is the result of multiplexed biosensing of
human and rabbit IgG samples at various concentrations
(Note 7).

4 Notes

1. All the materials are available from Sigma-Aldrich (St. Louis,

MO, USA) unless otherwise noted.
2. The washing steps are crucial for GNR functionalization and
immobilization. Three times’ washing is optimal to remove the
unreacted chemicals and most surfactants. Because CTAB
bilayer absorbed on GNR surface could block reactions with
gold, it is necessary to decrease its concentration to the critical
micelle concentration (CMC) at which GNR is stabilized. Fur-
ther washing could break the stability of GNR and causes
aggregation.
3. ITO-coated glass is chosen because of it conductivity, which
makes the GNR on glass observable under scanning electron
138 Zhong Mei et al.

Fig. 5 Representative pictures of (a) antibody functionalized GNR array on glass substrates which is expanded
to a high-throughput biosensing chip. Each dark spot, containing two sizes of functionalized GNRs, works as
an individual sensing platform for detecting two antigens simultaneously in one sample. Thus, this biochip can
detect 12 samples at one time; (b) Reverse side and (c) front side of attaching the chip to the bottom of a 96-
well plate frame for accommodating the reading on microplate reader (d)

microscope (SEM). The ITO coating doesn’t influence

MPTMS treating.
4. As CTAB is replaced by antibodies, SH-PEG is added 10 min
after antibody for stabilizing GNRs.
5. Weakly absorbed (not via covalent bonding) GNR nanoparti-
cles onto the glass can be washed off to ensure a stable and
robust biochip [17].
6. The nanorod biochip shows a linear response to the analyte in the
range from 10 to 60 nM with high sensitivity of 0.27 nm/nM
(R2 ¼ 0.99) for human IgG, which is more than 300% compari-
son with the reported literature with the sensitivity of
0.0607 nm/nM [17].
 Gold Nanorod Array Biochip for Label-Free, Multiplexed Biological Detection 139

Fig. 6 Multiplexed biosensing of human and rabbit IgG samples at various concentrations. (a) UV–Vis spectra
of the GNR biochip before and after detection of six various concentrations of mixed IgG. The shorter
wavelength around 630 nm is dedicated for rabbit IgG assay, while the longer one around 840 nm for
human IgG assay; (b) Plotting curve (black) of wavelength shift around 630 nm versus the concentration of
rabbit IgG; (c) Plotting curve (black) wavelength shift around 840 nm versus the concentration of human IgG.
Both curves suggest that the nanorod chip shows a linear response to the analyte (rabbit or human IgG) in the
range from 10 to 60 nM. For comparison, the calibration curves of single GNR based biochip mentioned in
Subheading 3.1 are also included (red). Interestingly, the multiplexed biosensor has an increased sensitivity
than single biosensor for human IgG detection

7. Nonspecific binding studies indicate minimal cross-reactivity of

the antibody moieties for high specificity in the multiplexed
biosensing. It is observed that the sensitivity is increased as
compared to individual detection, especially for longer GNR
probes.
8. The biochip achieved by this method demonstrated a random
assembly of GNRs on the substrate. Theoretical study has
proved that the local electromagnetic field on the tips of a
140 Zhong Mei et al.

Fig. 7 Local electromagnetic field simulation of a vertical GNR array. (a) model of GNR array; (b) local field of
side view in the cross section along the y–z plane; (c) local field of the top view along the z direction. The
rainbow bar shows the intensity. A redder color indicates a more intensified local field

nanorod is much stronger than that on the side facets. And the
CTAB density of tips is lower than the sides, which ensures a
less steric hindrance for the tip modification [18]. For the
future perspective, controlling the orientation of the GNR
array assembly on the substrate is the current focus. For
instance, if GNRs are orderly organized to form a vertically
standing array, a large amount of homogeneous hotspots will
be created at the GNR surface due to the plasmon coupling
effect. Figure 7 is the local electromagnetic field simulation of a
vertical nanoarray consisting of seven GNRs using COMSOL
5.0. The incident light with an intensity of 1.0  108 V/m
propagates along z-direction, passing through the GNR array.
The simulation results show that the GNR array exhibit highly
uniform and enhanced local field on the tips. This unique
property is desired for biomedical applications such as nanor-
eactor and ultrasensitive SERS sensor [19].
 Gold Nanorod Array Biochip for Label-Free, Multiplexed Biological Detection 141

References
1. Giri S, Trewyn BG, Lin VS (2007) Mesoporous nonspecific tumor marker, based on surface
silica nanomaterial-based biotechnological and plasmon resonance. Biosens Bioelectron 19
biomedical delivery systems. Nanomedicine (9):999–1005
(Lond) 2(1):99–111 12. Sim HR, Wark AW, Lee HJ (2010) Attomolar
2. Lee SH, Sung JH, Park TH (2012) detection of protein biomarkers using biofunc-
Nanomaterial-based biosensor as an emerging tionalized gold nanorods with surface plasmon
tool for biomedical applications. Ann Biomed resonance. Analyst 135(10):2528–2532
Eng 40(6):1384–1397 13. Ye X et al (2013) Using binary surfactant mix-
3. Oyelere AK et al (2007) Peptide-conjugated tures to simultaneously improve the dimen-
gold nanorods for nuclear targeting. Bioconjug sional tunability and monodispersity in the
Chem 18(5):1490–1497 seeded growth of gold nanorods. Nano Lett
4. Chylek P (1986) Absorption and scattering of 13(2):765–771
light by small particles. Appl Opt 25(18):3166 14. Gulati A, Liao H, Hafner JH (2006) Monitor-
5. Bennett HS, Rosasco GJ (1978) Resonances in ing gold nanorod synthesis by localized surface
the efficiency factors for absorption: Mie scat- plasmon resonance. J Phys Chem B 110
tering theory. Appl Optics 17(4):491–493 (45):22323–22327
6. Upadhyayula VK (2012) Functionalized gold 15. Wang Y, Tang L (2015) Multiplexed gold
nanoparticle supported sensory mechanisms nanorod array biochip for multi-sample analy-
applied in detection of chemical and biological sis. Biosens Bioelectron 67:18–24
threat agents: a review. Anal Chim Acta 16. Wang X et al (2015) Gold nanorod biochip
715:1–18 functionalization by antibody thiolation.
7. Yu L, Andriola A (2010) Quantitative gold Talanta 136:1–8
nanoparticle analysis methods: a review. 17. Wang Y, Tang L (2013) Chemisorption assem-
Talanta 82(3):869–875 bly of Au nanorods on mercaptosilanized glass
8. Dreaden EC et al (2012) The golden age: gold substrate for label-free nanoplasmon biochip.
nanoparticles for biomedicine. Chem Soc Rev Anal Chim Acta 796:122–129
41(7):2740–2779 18. Orendorff CJ, Murphy CJ (2006) Quantita-
9. Anker JN et al (2008) Biosensing with plasmo- tion of metal content in the silver-assisted
nic nanosensors. Nat Mater 7(6):442–453 growth of gold nanorods. J Phys Chem B 110
10. Jain PK et al (2006) Calculated absorption and (9):3990–3994
scattering properties of gold nanoparticles of 19. Lim DK et al (2011) Highly uniform and
different size, shape, and composition: applica- reproducible surface-enhanced Raman scatter-
tions in biological imaging and biomedicine. J ing from DNA-tailorable nanoparticles with 1-
Phys Chem B 110(14):7238–7248 nm interior gap. Nat Nanotechnol 6
11. Chou SF et al (2004) Development of an (7):452–460
immunosensor for human ferritin, a
 Chapter 10

Resonant Waveguide Grating Imager for Single Cell

Monitoring of the Invasion of 3D Speheroid Cancer
Cells Through Matrigel
Nicole K. Febles, Siddarth Chandrasekaran, and Ye Fang

Abstract
The invasion of cancer cells through their surrounding extracellular matrices is the first critical step to
metastasis, a devastating event to cancer patients. However, in vitro cancer cell invasion is mostly studied
using two-dimensional (2D) models. Three-dimensional (3D) multicellular spheroids may offer an advan-
tageous cell model for cancer research and oncology drug discovery. This chapter describes a label-free, real-
time, and single-cell approach to quantify the invasion of 3D spheroid colon cancer cells through Matrigel
using a spatially resolved resonant waveguide grating imager.

Key words Adhesion, Colorectal cancer cell, Dynamic mass redistribution, Extracellular matrix,
Invasion, Migration, Multicellular spheroid, PTEN, Resonant waveguide grating

1 Introduction

Tumor metastasis is the most devastating aspect of cancer and has

been an active area for developing targeted therapeutics [1, 2].
Several in vitro assays have been developed to determine the migra-
tory and invasive capacities of tumor and stromal cells, to elucidate
the underlying biochemical and cellular mechanisms of metastasis,
and to screen compounds that can inhibit distinct steps of metasta-
sis [3–7]. These assays differ not only in operational and detection
principles but also in cell models. Cell migration is often character-
ized based on the healing rate of a physically wounded two-dimen-
sional (2D) cell monolayer, the recovery rate of a cell monolayer
having an exclusion zone, or the number of cells adherent on the
underside of a Transwell porous membrane after passing through
the membrane. Cell invasion is often examined based on the num-
ber of individual cells invading through an extracellular matrix
(ECM) (e.g., Matrigel) coated Transwell porous membrane.
Chemotaxis, the movement of cells in response to a chemical

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_10,
© Springer Science+Business Media LLC 2017

143
144 Nicole K. Febles et al.

stimulus, is often studied by introducing a chemoattractant in the

lower chamber of a Transwell device. Common to these assays is
that they only permit narrowly focused snapshots into complex
biological phenomena, and have limited ability to recapitulate in
vivo process as primary tumors often display vast structural hetero-
geneity and microenvironments, and cancer cells can migrate in all
directions [7, 8].
In the past years, there is an increase in using three-dimensional
(3D) cell models including multicellular spheroids for oncology
drug discovery. Compared to 2D cell monolayers, these 3D models
are believed to more closely recapitulate the in vivo biology of solid
cancers, thus enabling more accurate prediction of cancer behaviors
[9–11] and in vivo drug efficacy, potency, and safety [12–14].
Concurrent to the increasing use of 3D models for cancer research
is the development of advanced invasion and migration assays
[15–20]. For instance, spheroid sprouting assay is used to examine
the invasion of cells in a spheroidal structure into surrounding 3D
ECM gel [21, 22]; however, this assay mostly detects late invasion
events. On the other hand, cell invasion into spheroid assay is
useful for studying the invasion of individual cells into a spheroidal
cluster of another cell type [23], or the cell invasion between
spheroids of two different cell types [24, 25]; however, this assay
often uses fluorescently labeled cells, has low temporal resolution,
and is cumbersome in quantification [7].
Recently, we have developed a label-free, real-time, single-cell,
and quantitative assay, termed label-free single cell 3D2 invasion
assay, to track the invasion of 3D spheroid cells through a 3D
matrix [26, 27]. This chapter describes the protocols to determine
the migratory and invasive capacities of three colon cancer cell lines,
HT29, HCT116 wild-type (WT), and HCT16 PTEN
/
using
wound healing, Transwell invasion, and label-free single cell 3D2
invasion assays. PTEN (phosphatase and tensin homolog) is a
tumor suppressor negatively regulating the PI3K signaling path-
way, and PTEN deletion is correlated with colorectal cancer metas-
tasis and poor patient survival [28].

1.1 Sensor For the label-free single cell 3D2 invasion assay, there are four
Configurations and critical components, including resonant waveguide grating
Detection Schemes (RGW) biosensor microplate, Matrigel, spheroids of cancer cells,
and RWG imager (Fig. 1).
The biosensor microplate, commercially known as Epic®
384well cell culture compatible microplate from Corning
Incorporated, is used directly. Within the bottom of each well of
the microplate, there is a RWG biosensor. The biosensor itself
consists of three layers—a glass substrate, a grating structure, and
a waveguide thin film with high refractive index (Fig. 1a). Under
illumination with a light source light at a specific wavelength is
coupled into the waveguide through diffraction, creating a surface
 Label-Free Single Cell 3D2 Invasion 145

Fig. 1 Spatially resolved resonant waveguide grating (RWG) imager for label-free single cell monitoring. (a)
The principle of RWG biosensor for monitoring the invasion and adhesion of cancer cells. (b) The instrument.
(c) Optical setup of the imager, which consists of three components: a tunable light source to sweep the
wavelength range, an array of optical components to guide and expand light path leading to illuminate at a
normal incident angle all biosensors within a plate, and a CMOS camera to record the resonant wavelength
image. (d) A false-colored DMR image of a 3  4 biosensor array. All biosensors were pre-coated with 10 μL
0.1 mg/mL Matrigel. The image was obtained 24 h after spheroids of HT29 cells in the absence and presence
of vandetanib at different doses were placed on the top Matrigel surface. False color bar scale:
500 (blue) to
2000 pm (red). The pattern and area of adherent cells is indicated by the green-to-red spot, while the signal
intensity is indicated by the false color scale—red: high positive signal; green: moderate positive signal; blue:
negative signal. The well corresponding to the second left in the top row did not receive any spheroid as the
negative control. Condition of spheroid formation: 40K per well seeding density, 4 days culture. The adhesion
area and signal was found to be decreased as the dose of vandetanib increases

bound electromagnetic field (also known as evanescent wave) that

has a characteristic penetration depth (~200 nm) extending into the
medium or cell layer. After propagating within the waveguide for a
short distance due to total internal reflection, the light eventually
leaks out the waveguide and reflects back. The optical content
(e.g., wavelength, angle, intensity) of the reflected light is moni-
tored and recorded in real time, depending on the detection sys-
tems. The wavelength of the reflected light, termed as the
146 Nicole K. Febles et al.

resonance wavelength, is identical to that of the coupled light and is

sensitive to local index of reflection, which is proportional to the
density of cellular matters within the penetration depth of the
biosensor [29, 30].
Matrigel is used to create a 3D extracellular matrix on the
biosensor surface by placing equal volume Matrigel solution into
each well of a biosensor microplate at 4  C followed by overnight
gelation at 37  C (Fig. 1a). Matrigel is a gelatinous protein mixture
secreted by Engelbreth-Holm-Swarm (EHS) mouse sarcoma cells,
and is the most widely used basement membrane due to its reliabil-
ity and reproducibility. Matrigel consists of approximately 60%
laminin, 30% collagen IV, and 8% entactin. Entactin is a bridging
molecule that interacts with laminin and collagen IV, and contri-
butes to the structural organization of these ECM molecules.
Matrigel generally contains the transforming growth factor-β
(TGF-β), epidermal growth factor, insulin-like growth factor, fibro-
blast growth factor, tissue plasminogen activator, and other growth
factors which occur naturally in the EHS tumor. There is also
residual matrix metalloproteinases derived from the tumor cells.
However, the Matrigel used here has a reduced level of a variety
of growth factors except for TGF-β. Matrigel is a liquid at 4  C and
starts to form a gel above 10  C [31]. The thickness of Matrigel film
after gelation is determined by the volume of the solution used for
coating. Coating with 10 μL Matrigel solution results in a thickness
of ~630 μm for the 384-well microplate [26]. On the other hand,
the density and porosity of Matrigel is determined by the concen-
tration of total protein concentration used for coating. The higher
concentration the Matrigel solution is the denser 3D matrix is
formed, resulting in slower invasion [27].
Spheroidal clusters of human cancer cells are generated by
seeding equal number of cells into each well, followed by culturing
in Corning ultra-low attachment (ULA) round-bottom 96-well
plates. The ULA plate enables self-assembly of cancer cells into a
single 3D spheroidal cluster in the bottom of each well mediated by
ultralow attachment surface chemistry and gravitational force [32].
A spatially resolved RWG imager (Fig. 1b) is used to achieve
single cell monitoring of cancer cell invasion through the Matrigel.
This imager is based on a swept wavelength interrogation scheme,
wherein a tunable light source combining a broadband light source
with a high precision, narrow-band optical filter is used to illumi-
nate simultaneously a 3  4 biosensor array within a 384-well
microplate (Fig. 1c), and a complementary metal oxide semicon-
ductor (CMOS) digital camera is used to record the resonance
images with a spatial resolution of 12 μm (Fig. 1d) [33, 34]. The
wavelength sweeping from 825 to 840 nm is done stepwise in
100 pm every 20 ms. Owing to the fine difference in resonance
condition among different pixels as well as the use of wavelength
sweeping technique for illumination, the imager permits location-
 Label-Free Single Cell 3D2 Invasion 147

dependent time resolved resonance within a sensor, thus enabling

high resolution imaging, although the coupled light tends to travel
within the waveguide for a short distance [33]. The time series
resonance images are recorded and used to calculate wavelength
shifts (in picometer, pm) at each pixel with a temporal resolution of
3 s, resulting in spatially resolved kinetic responses, one per pixel.
Since the resonance wavelength is sensitive to the change in local
mass density, the kinetic response is often referred to dynamic mass
redistribution (DMR) signal [35].
The label-free single cell 3D2 invasion assay employs the RWG
imager to noninvasively track in real time the adhesion event of cells
after dissociating from a spheroid and invading through the matrix.
The invasion is initiated after a single spheroid of cancer cells is
placed on the top surface of the Matrigel film. Since the thickness of
Matrigel is far greater than the penetration depth of the biosensor, a
cell has to dissociate from the spheroid, invades through the matrix,
and finally adheres onto the sensor surface in order to generate a
detectable signal. Since adhesion is a rapid process at single cell level
[26, 27], the DMR signals obtained can be used to study the
behavior and mechanism of the invasion of cells in spheroid
through the 3D matrix.
Of note, the RWG biosensor has been shown to permit DMR
assay [29, 30, 35], which enables real-time investigation of a wide
range of cell phenotypes [36], receptor signaling [37–39], and
cellular processes [40, 41] (reviewed in [42–44]).

2 Materials

2.1 Tissue Culture 1. Human colorectal adenocarcinoma HT-29 cell line (ATCC®
Medium and Cell Line HTB-38™, American Type Cell Culture, Manassas, VA, USA).
2. The isogenic colorectal carcinoma (CRC) cell lines, HCT116-
WT and HCT116-PTEN
/
cells (Horizon Discovery Ltd.,
Cambridge, UK) (see Note 1).
3. McCoy’s 5A medium (Catalog #16600, Life Technologies,
Carlsbad CA, USA).
4. Complete medium for HT29: McCoy’s 5A modified medium
supplemented with 10% fetal bovine serum, 4.5 g/L glucose,
2 mM glutamine, and 1 Pennstrep.
5. Complete medium for the isogenic cell lines: McCoy’s 5A
medium supplemented with 10% fetal bovine serum, 4.5 g/L
glucose, 1.5 mM glutamine, and 1 Pennstrep.
6. Serum-free medium for the isogenic cell lines: McCoy’s 5A
medium supplemented with 4.5 g/L glucose, 1.5 mM gluta-
mine, and 1 Pennstrep.
7. Pennstrep antibiotic solution 100: 10,000 units penicillin/
mL, 10,000 μg streptomycin/mL.
148 Nicole K. Febles et al.

8. Trypsin–ethylenediaminetetraacetic acid (EDTA) solution

10: 2.5% Trypsin, 0.2% 4Naþ-EDTA.
9. Corning® Matrigel Growth Factor Reduced (GFR) Basement
Membrane Matrix (Catalog #354230).
10. LY294002 and wortmannin (Tocris Chemical Company, St.
Louis, MO, USA).
11. Vandetanib (LC Laboratories, Woburn, MA, USA).
12. Dimethyl sulfoxide (DMSO) .

2.2 Microplates and 1. Epic® BT high-resolution RWG imager (Corning

Instruments Incorporated, Corning, NY, USA).
2. Nikon Eclipse TS100 inverted light microscope equipped with
NIS Elements F3.0 Software (Nikon Instruments Inc., Mel-
ville, NY, USA).
3. Matrix 16-channel electronic pipettor (Thermo Fisher Scien-
tific, Hudson, NH).
4. Corning 75 cm2 tissue-culture treated polystyrene flask with
vented cap (T75).
5. Corning 384-well polypropylene compound storage plate.
6. Corning® Epic® 384-well cell assay microplate (Catalog
#5040).
7. Corning UltraLow Attachment (ULA) 96-well round bot-
tomed plate (Catalog #7007).
8. Corning® HTS Transwell®-24 well permeable supports,
6.5 mm Transwell® with 8.0 μm pore polycarbonate membrane
inserts, sterile (Catalog #3422) .

2.3 Data Analysis 1. Corning Epic® BT Processor software.

Software 2. GraphPad Prism 5 Software (Graph Pad Software Inc., La Jolla,
CA, USA).
3. NIH Image J (http://imagej.nih.gov/ij/).
4. Microsoft Excel (Microsoft Inc., Seattle, WA, USA).

3 Methods

3.1 3D Spheroid Cell This section describes the protocol to form uniform spheroids of
Culture three CRC cell lines by directly culturing an equal number of cells in
each well of a 96-well ULA round bottomed microplate (Fig. 2a, b).
The spheroids formed can be transitional (meaning that it can
grow in size over time) or stable (meaning that it stops growing).
For the 3D2 invasion assay, mechanically stable spheroids are
preferred since it is important to minimize potential interference
with the assay results of mechanically dissociated cells from the
 Label-Free Single Cell 3D2 Invasion 149

Fig. 2 Growth pattern of spheroidal cells of HCT116 and its isogenic PTEN null cell line in 96-well ULA round
bottomed plate. (a) ULA microplate. (b) schematic drawing showing how a spheroid is formed within a ULA
round bottomed well. (c, d) Daily light microscopic images of HCT116 WT cells (c), or HCT116 PTEN
/
cells
(d) at Day 0 to Day 5 (i–v, respectively). Seeding density: 1000 cells per well; scale bar: 500 μm. (e, f) The
diameter of spheroidal clusters of both cell lines as a function of culture duration when seeding density is 1000
(e) or 20,000 (f) cell per well. This figure is adapted from ref. 27 with permission
150 Nicole K. Febles et al.

spheroidal structure. In addition, spheroids that are much smaller

than the dimension of the biosensor (2  2 mm) are preferred for
their effective positioning onto the biosensor surface.
When the seeding density is 1000 cells per well, the two iso-
genic cells become loosely attached right after seeding, started
forming irregular clusters at day 1, and spheroids at day 3, which
continuously grown in size (Fig. 2c–e). This suggests that under
transitional spheroid formation condition PTEN knockout has
little impact on the formation and growth of spheroids. However,
when the seeding density is 20,000 cells per well, the wild-type cells
rapidly formed a spheroid whose size continuously decreased till
day 4, while the PTEN
/
cells formed a spheroid whose size was
barely changed over the 4 day culture (Fig. 2f). This suggests that at
the high seeding density the wild-type cell can self-aggregate and
compact into a stable spheroid due to cell–cell interaction, while the
PTEN null cell can self-aggregate but fail to undergo compaction
due to the loss of epithelial characteristics [45]. For a given cell
type, the spheroids formed can reach a stable dimension likely due
to the limitation of nutrition and oxygen supply and/or accumula-
tion of cellular metabolic waste (e.g., lactate) [46]. HT29 cell can
also form spheroids in the ULA plate [26]. For different cell lines,
the spheroid culture protocol can be optimized by altering
medium, culture duration, and cell seeding numbers.
1. Passage the cells in T-75 flask using 1 trypsin–EDTA when
approaching 90% confluency. Generally, cells are split every
5 days, and a 1:10 split is used to provide maintenance culture.
The cells between 3 and 25 passages are preferred for assays
(see Note 2).
2. Once approaching confluency, harvest cells using 1 trypsi-
n–EDTA solution. Centrifuge the cells down, remove the
supernatant, and resuspend the cell pellet in freshly prepared
complete cell culture medium.
3. Dispense 200 μL cell suspensions at varied numbers of cells per
well into each well of a 96-well ULA round bottomed plate
(see Note 3).
4. Culture in the complete cell culture medium at 37  C with 5%
CO2 for 4–8 days. For HT29, partially exchange the media by
removing 100 μL old media at day 4 or 7 and then replacing
once with 100 μL fresh media. For the two isogenic cell lines,
partially exchange the media at day 3. To minimize disruption
of the spheroids during media changes, hold the pipette tip at
approximately a 45 angle away from the center of the well
bottom.
5. Collect time series light microscopic images for each well at
specific time points.
 Label-Free Single Cell 3D2 Invasion 151

6. Calculate the diameter of each spheroidal structure at a specific

time using NIH ImageJ software.

3.2 Matrigel Coating This section describes the protocol to coat RWG biosensor plates or
Transwell inserts with Matrigel for invasion assays.
1. Precool Matrigel, biosensor plates, Transwell inserts, pipets,
and tips at 4  C before coating. This step is essential to prevent
premature gelation, and to ensure uniform coating.
2. Dilute Matrigel using the complete media to 0.1 mg/mL or
0.2 mg/mL (1  50 or 1  25 dilution, respectively) at 4  C.
3. Add 10 or 20 μL the diluted Matrigel solution to each well of a
384well biosensor microplate, or 100 μL to a Transwell insert
at 4  C.
4. Incubate the microplate or Transwell insert inside an incubator
at 37  C under air/5% CO2 for 24 h to form a 3D gel on the
biosensor surface, or on the Transwell insert.

3.3 Label-Free Single This section describes the protocol to perform label-free single cell
Cell 3D2 Invasion 3D2 invasion assay. This assay starts with placing a single spheroid
Assay onto the top surface of Matrigel, followed by that individual cells
spontaneously transfer from the spheroid to the top Matrigel sur-
face, invade through the matrix, and eventually adhere onto the
biosensor surface (Fig. 3). The RWG imager monitors in real time
the cell adhesion events, which, in turn, are used as an indicator to
determine the invasive potential of a cancer cell line. This assay
effectively mimics the first critical step of cancer metastasis; that is,
the local invasion of cancer cells through surrounding ECM, which
involves adhesion, proteolytic remodeling of the ECM, and migra-
tion [47–50].
This assay enables multiparametric analysis of cell invasion and
adhesion (Fig. 4), The initial adhesion point, designated as the
coordinate of (0, 0), is first identified as the pixelated location at
which the first positive DMR occurs. The lateral distance of any
pixel relative to the initial adhesion point is then calculated. The
maximal DMR at each pixel is extracted from its DMR signal. The
adhesion time at each pixel is also extracted from its DMR as the
time reaches a specific DMR amplitude (e.g., 500 pm for HT29, or
200 pm for the isogenic cell lines). The total adhesion events are
obtained by counting the number of pixels that give rise to a DMR
exceeding a predetermined threshold signal (e.g., 500 pm for
HT29, or 200 pm for the isogenic cell lines). The total adhesion
area is determined using ImageJ as the area that gave rise to a signal
above the background. Correlation analysis can then be performed.
For HCT116, PTEN deletion was found to increase the invasion
distance (Fig. 5a vs. Fig. 5b), and the total adhesion events
(Fig. 5c), suggesting that PTEN deletion increases the invasion
152 Nicole K. Febles et al.

Fig. 3 Label-free single cell 3D2 invasion assay. (a) The assay consists of four critical steps: coating the
biosensor surface with Matrigel; adding medium containing an inhibitor compound to the well; transferring a
spheroid from a ULA round bottomed microplate and placing it onto the top Matrigel surface; and monitoring
the invasion and adhesion in real time. (b–e) The time series DMR images of a biosensor before and after a
single spheroid was placed onto the biosensor surface coated with 10 μL 0.1 mg/mL Matrigel: 0 min (b), 1 h.
(c), 6 h (d), and 24 h (e). Spatial scale bar: 500 μm. Intensity scale bar: false colored (
500 to 2000 pm). (f–i)
The time series light microscopic images of a biosensor before and after a single spheroid was placed onto
Matrigel coated biosensor surface: 0 min (f), 1 h. (g), 6 h (h), and 24 h (i). Scale bar: 500 μm. Condition of
spheroid formation: 40K per well seeding density, 4 days culture. (b–i) is adapted with permission from ref.
26. Copyright (2015) American Chemical Society

speed of the cells in spheroid through 3D matrix. Furthermore, the

invasion and adhesion of both isogenic cell lines was found to be
sensitive to PI3K inhibitors wortmannin and LY294002 (Fig. 5d).
1. Prepare all compound solutions in the complete medium.
2. Add 10 μL the complete medium in the absence and presence
of a compound to the Matrigel coated biosensor plate.
3. Place the Matrigel coated biosensor microplates, RWG imager,
cell media, pipette, pipette tips, and compound solutions all
inside an incubator at least 2 h before the assay. This step is to
ensure that temperature reaches equilibrium, so temperature
mismatch, if occurring, will be minimal.
4. Place the Matrigel coated biosensor microplate onto the
imager and wait for 30 min.
 Label-Free Single Cell 3D2 Invasion 153

Fig. 4 Label-free single cell analysis for the invasion and adhesion of HT29 cell spheroids. (a) A DMR image
taken 24 h after a spheroid was placed on the top Matrigel surface. Scale bar: 500 μm. (b) Real-time
population averaged DMR on the Matrigel coated surface. (c) Pixelated real-time DMR signals for the line
indicated in (a). (d) The correlation between the adhesion time and the lateral distance relative to the first
adhesion point. (e) The correlation between the maximal response and the relative distance. (f) The correlation
between the adhesion time and the maximal response. The biosensor was coated with 10 μL 0.1 mg/mL
Matrigel. Condition of spheroid formation: 40K per well seeding density, 4 days culture. This figure is adapted
(b–j) is adapted with permission from ref. 26. Copyright (2015) American Chemical Society

5. Monitor the baseline until all sensors reached a steady baseline.

This step is to ensure that the baseline drifts <10 pm within
5 min.
6. Establish a baseline by recording resonance images for 5 min.
7. Pause the imager temporally for spheroid transfer. Critical to
this step is to minimize the blackout period (<2 min) between
the baseline and the first recording after spheroid placement.
8. Pick up a 30 μL media solution containing the single spheroid
from a well of a ULA round bottomed microplate using a wide-
bore tip. Critical to this step is to visually inspect the solution
inside the wide-bore tip to ensure the spheroid pickup.
9. Gently place the single spheroid onto the top Matrigel surface
in a well of a 384-well biosensor microplate. Critical to this step
is to ensure that the pipette tip does not damage the coating,
and the spheroid is placed within the biosensor (2  2 mm in
the center of each well). In some cases, two tumor spheroids
can be transferred into a single well to increase the probability
of placing spheroids within the biosensor area.
154 Nicole K. Febles et al.

Fig. 5 PTEN deletion increases the invasion rate of CRC cells in spheroid through 3D Matrigel matrix. (a, b)
Correlation analysis between the adhesion time and lateral distance for the parental (a) and PTEN
/
(b)
cells. (c) The adhesion events versus cell types. (d) The adhesion time to reach 200 pm under different
conditions. Coating: 0.2 mg/mL Matrigel. Data represent mean  s.d. for c, d (n ¼ 3). ***p < 0.001. This
figure is adapted from ref. 27 with permission

10. Restart the imager and continuously record resonance images

for ~24 h.
11. Collect light microscopic images at the top Martigel surface at
the end of the assay (see Note 4).
12. Normalize the starting resonance wavelength at each pixel to
zero, calculate the wavelength shifts, extract kinetic DMR
signals at each pixel of a sensor, and save the data into Excel
using Epic Processor Software.
13. Perform intra-sensor referencing to remove background signal
from the adhesion signals of all pixelated locations (see Note 5).
14. Analyze the background corrected data to determine the initial
adhesion point, the lateral distance relative to the initial adhe-
sion point, the maximal DMR amplitude, the adhesion time,
the total adhesion event, and the adhesion area using Excel (see
Note 6).
15. Perform correlation analysis between two different parameters
using Prism software.
 Label-Free Single Cell 3D2 Invasion 155

3.4 Cell Monolayer This section describes the wound healing assay protocol to detect
Wound Healing Assay the migratory potential of CRC cells. This assay is to first culture
cells to form a monolayer, then physically create a wounded area,
and finally monitor the wound recovery over time using light
microscopy (Fig. 6a). The migratory behavior of cells, one of the
key hallmarks of a metastatic cancer cell, is reflected by the rate of
recovery of the scratch region (Fig. 6b).
1. Harvest cells from T75 flask using 1 trypsin–EDTA.
2. Centrifuge the cells down, remove the supernatant, and resus-
pend the cell pellet in freshly prepared complete cell culture
medium.
3. Add cells to a 6-well plate and grow cells to confluency at 37  C
with 5% CO2.
4. Wash the confluent cells three times with the serum free
medium, and starve the cells overnight.
5. Produce a scratch with a sterile 200 μL pipette tip in a marked
region of the well.
6. Collect immediately light microscopic images of the marked
scratch area.

Fig. 6 2D cell wound recovery assay showing that PTEN deletion increases the
migratory potential of HTC116 cells. (a) Critical steps of the assay. (b) The 2D
migration rate of HCT116 versus HCT116-PTEN
/
cells. Data represents
mean  s.d. (n ¼ 5). ***p < 0.001. This figure is adapted from ref. 27 with
permission
156 Nicole K. Febles et al.

7. Culture the cells in serum free medium at 37  C with 5% CO2

for 48 h.
8. Collect light microscopic images of the marked scratch areas in
the end.
9. Determine the percentage of reduction in wound size (i.e.,
scratch area) using ImageJ.

3.5 Transwell This section describes the Transwell invasion assay protocol to
Invasion Assay of detect the impact of PTEN deletion on the invasive potential of
Individual Cancer Cells HCT116 cells. This assay measures the invasion of individual cells
through Matrigel coated substrates having defined micropores
(Fig. 7a). This assay mimics local invasion of tumor cells through
the ECM or basement membrane. The invasiveness is quantified by
counting the number of invaded cells and normalizing to those in
uncoated Transwell inserts (Fig. 7b, c).
1. Harvest cells from T75 flask using 1 trypsin–EDTA.
2. Centrifuge the cells down, remove the supernatant, and resus-
pend the cell pellet in serum free media.
3. Add 100 μL cell suspension to the top side of the Matrigel
coated Transwell inserts with a cell density of 2.5  104 cells/
mL. Use the uncoated Transwell inserts as the control.

Fig. 7 Transwell invasion assay showing that PTEN deletion increases the invasiveness of the CRC cells. (a)
Critical steps of the assay. (b) Representative light microscopic images of HCT116 versus HCT116-PTEN
/

cells adhered on the underside of Transwell inserts 24 h after invasion. (c) The invasion rate through Matrigel
coated Transwell of HCT116 versus HCT116-PTEN
/
cells. Data represents mean  s.d. (n ¼ 5).
***p < 0.001. This figure is adapted from ref. 27 with permission
 Label-Free Single Cell 3D2 Invasion 157

4. Place the inserts having cells to a 24-well plate, each well being
filled with 750 μL of the cell culture media with 5% FBS as a
chemoattractant, to form invasion chambers.
5. Incubate the invasion chambers for 24 h at 37  C with 5% CO2.
6. Remove the inserts and wash twice with 1 phosphate buff-
ered saline.
7. Discard the non-invading cells using a cotton swab moistened
with culture media.
8. Fix the adhered cells in 4% paraformaldehyde, permeablize in
100% methanol, and stain using crystal violet stain.
9. Manually count the number of invading cells on the underside
of the Transwell using a light microscope.
10. Calculate the percentage invasiveness by normalizing the num-
ber of invading cells in the Matrigel coated inserts to that in the
uncoated inserts.

4 Notes

1. HCT116 is a microsatellite-unstable and metastatic human

CRC cell line and can grow in multilayers but is incapable of
polarization or intestinal epithelial differentiation. Its isogenic
PTEN
/
line is directly derived via homologous knockout of
PTEN in HCT116 by deletion of exon 5 which encodes the
active site of the protein [51]. The wild-type cell line does not
bear any somatic PTEN mutations, but has an H1047R muta-
tion of PI3KCA in exon 20 (kinase domain) [52].
2. Cells can undergo genetic drifting during passaging. Given the
high sensitivity of DMR assays, the DMR of a cell line may be
sensitive to its passages.
3. Optimal seeding density for spheroid formation is cell line
dependent. This can be assessed based on the size and shape
of cell clusters formed, as well as the morphology of spheroids
after manually transferred to the biosensor plate using a wide-
bore pipette tip.
4. Light microscopic images can be acquired at the end of the
assay since RWG DMR imaging is noninvasive. DMR imaging
has a short sensing depth (~200 nm) away from the biosensor
surface, so it only examines cell adhesion on the biosensor
surface (i.e., right below the Matrigel coating) (Fig. 3b–e).
On the other hand, light microscopic imaging has much
thicker imaging depth, so it can examine the spheroid and the
pattern of cells after spontaneously transferring from the spher-
oid to the top Matrigel surface (Fig. 3f–i).
158 Nicole K. Febles et al.

5. The intra-sensor referencing can be used to eliminate any

artifacts associated with mismatch in temperature and bulk
index during the spheroid placement step. RWG biosensor is
known to be sensitive to temperature and bulk index mismatch
[53]. The intra-sensor referencing is possible since the size of a
spheroid is much smaller than the dimension of a biosensor.
Here, a sensor is divided into a background zone (i.e., the cell
free area), and a cell adhesion zone (Fig. 4a). Subtracting the
background signal from the cell adhesion signal led to back-
ground-free responses for the cell adhesion zone (Fig. 4b), or
all pixelated locations (exemplified in Fig. 4c).
6. (a) The initial adhesion point approximates closely to the initial
contact point of a spheroid with the top Matigel surface, at
which the cell has the shortest distance to invade through. The
time for the initial adhesion to occur indicates how fast the first
cell invades through the matrix and reaches the sensor surface.
(b) The lateral distance relative to the initial adhesion point is
not the travel distance of an invading cell. Despite this, given
that all cells need invade the same 3D matrix, the lateral dis-
tance can be used as a useful parameter to investigate the
invasion behavior of spheroidal cells. (c) The maximal DMR
at each pixel is an indicator of cell adhesion degree. (d) The
adhesion time mostly reflects the time required for a cell to
invade through the matrix and arrive at the sensor surface, since
the kinetics of cell adhesion is relatively rapid (Fig. 4c). (e) The
total adhesion events and area are two indicators for total
numbers of cells invading through the Matrigel.

References
1. Steeg PS, Theodorescu D (2008) Metastasis: a 6. Carragher N (2009) Cell migration and inva-
therapeutic target for cancer. Nat Clin Pract sion assays as tools for drug discovery. Clin Exp
Oncol 5:206–219. doi:10.1038/ncponc1066 Metastasis 26:381–397. doi:10.3390/
2. Desgrosellier JS, Cheresh DA (2010) Integrins pharmaceutics3010107
in cancer: biological implications and therapeu- 7. Kramer N, Walzl A, Unger C, Rosner M, Kru-
tic opportunities. Nat Rev Cancer 10:9–22. pitza G, Hengstschl€ager M, Dolznig H (2013)
doi:10.1038/nrc2748 In vitro cell migration and invasion assays.
3. Eccles SA, Box C, Court W (2005) Cell migra- Mutat Res 752:10–24. doi:10.1016/j.mrrev.
tion/invasion assays and their application in 2012.08.001
cancer drug discovery. Biotechnol Annu Rev 8. Quail DF, Maciel TJ, Rogers K, Postovit LM
11:391–421. doi:10.1016/S1387-2656(05) (2012) A unique 3D in vitro cellular invasion
11013-8 assay. J Biomol Screen 17:1088–1095. doi:10.
4. Albini A, Benelli R (2007) The chemoinvasion 1177/1087057112449863
assay: a method to assess tumor and endothelial 9. Chitcholtan K, Asselin E, Parent S, Sykes PH,
cell invasion and its modulation. Nat Protoc Evans JJ (2013) Differences in growth proper-
2:504–511. doi:10.1038/nprot.2006.466 ties of endometrial cancer in three dimensional
5. Kam Y, Guess C, Estrada L, Weidow B, Quar- (3D) culture and 2D cell monolayer. Exp Cell
anta V (2008) A novel circular invasion assay Res 319:75–87. doi:10.1016/j.yexcr.2012.09.
mimics in vivo invasive behavior of cancer cell 012
lines and distinguishes single-cell motility in 10. H€arm€a V, Virtanen J, M€akel€a R, Happonen A,
vitro. BMC Cancer 8:198. doi:10.1186/ Mpindi JP, Knuuttila M, Kohonen P, Lötjönen
1471-2407-8-198 J, Kallioniemi O, Nees M (2010) A
 Label-Free Single Cell 3D2 Invasion 159

comprehensive panel of three-dimensional capillary morphogenesis and maintenance in

models for studies of prostate cancer growth, 3D. Am J Physiol Cell Physiol 297:
invasion and drug responses. PLoS One 5: C179–C187. doi:10.1152/ajpcell.00018.
e10431. doi:10.1371/journal.pone.0010431 2009
11. Tanner, K. and Gottesman, M.M. (2015) 22. Blacher S, Erpicum C, Lenoir B, Paupert J,
Beyond 3D culture models of cancer. Sci Transl Moraes G, Ormenese S, Bullinger E, Noel A
Med 7:283 ps9. doi:10.1126/scitranslmed. (2014) Cell invasion in the spheroid sprouting
3009367. assay: a spatial organization analysis adaptable
12. Justice BA, Badr NA, Felder RA (2009) 3D cell to cell behaviour. PLoS One 9:e97019. doi:10.
culture opens new dimensions in cell-based 1371/journal.pone.0097019
assays. Drug Discov Today 14(1-2):102–107. 23. Ghosh S, Joshi MB, Ivanov D, Feder-Mengus
doi:10.1016/j.drudis.2008.11.006 C, Spagnoli GC, Martin I, Erne P, Resink TJ
13. Pickl M, Ries CH (2009) Comparison of 3D (2007) Use of multicellular tumor spheroids to
and 2D tumor models reveals enhanced HER2 dissect endothelial cell–tumor cell interactions:
activation in 3D associated with an increased a role for T-cadherin in tumor angiogenesis.
response to trastuzumab. Oncogene FEBS Lett 581:4523–4528. doi:10.1016/j.
28:461–468. doi:10.1038/onc.2008.394 febslet.2007.08.038
14. Luca, A.C., Mersch, S., Deenen, R., Schmidt, 24. Oxmann D, Held-Feindt J, Stark AM, Hatter-
S., Messner, I., Sch€afer, K.L., Baldus, S.E., mann K, Yoneda T, Mentlein R (2008) Endo-
Huckenbeck, W., Piekorz, R.P., Knoefel, W. glin expression in metastatic breast cancer cells
T., Krieg, A., and Stoecklein, N.H. (2013) enhances their invasive phenotype. Oncogene
Impact of the 3D microenvironment on phe- 27:3567–3575. doi:10.1038/sj.onc.1211025
notype, gene expression, and EGFR inhibition 25. Sabeh F, Shimizu-Hirota R, Weiss SJ (2009)
of colorectal cancer cell lines. PLoS One 8(3): Protease-dependent versus-independent can-
e59689. doi:10.1371/journal.pone.0059689. cer cell invasion programs: three-dimensional
15. Fisher KE, Pop A, Koh W, Anthis NJ, Saunders amoeboid movement revisited. J Cell Biol
WB, Davis GE (2006) Tumor cell invasion of 185:11–19. doi:10.1083/jcb.200807195
collagen matrices requires coordinate lipid ago- 26. Febles NK, Ferrie AM, Fang Y (2014) Label-
nist-induced G-protein and membrane-type free single cell quantification of the invasion of
matrix metalloproteinase-1-dependent signal- spheroidal colon cancer cells through 3D
ing. Mol Cancer 5:69. doi:10.1186/1476- Matrigel. Anal Chem 86:8842–8849. doi:10.
4598-5-69 1021/ac502269v
16. Brekhman V, Neufeld G (2009) A novel asym- 27. Chandrasekaran S, Deng H, Fang Y (2015)
metric 3D in-vitro assay for the study of tumor PTEN deletion potentiates invasion of colorec-
cell invasion. BMC Cancer 9:415–427. doi:10. tal cancer spheroidal cells through 3D Matri-
1186/1471-2407-9-415 gel. Integr Biol 7:324–334. doi:10.1039/
17. Pathak A, Kumar S (2011) Biophysical regula- c4ib00298a
tion of tumor cell invasion: moving beyond 28. Salmena L, Carracedo A, Pandolfi PP (2008)
matrix stiffness. Integr Biol 3:267–278. Tenets of PTEN tumor suppression. Cell
doi:10.1039/c0ib00095g 133:403–414. doi:10.1016/j.cell.2008.04.
18. Koch TM, M€ unster S, Bonakdar N, Butler JP, 013
Fabry B (2012) 3D traction forces in cancer 29. Fang Y, Ferrie AM, Fontaine NH, Mauro J,
cell invasion. PLoS One 7:e33476. doi:10. Balakrishnan J (2006) Resonant waveguide
1371/journal.pone.0033476 grating biosensor for living cell sensing. Bio-
19. Cheung KJ, Gabrielson E, Werb Z, Ewald AJ phys J 91:1925–1940. doi:10.1529/biophysj.
(2013) Collective invasion in breast cancer 105.077818
requires a conserved basal epithelial program. 30. Fang Y (2007) Non-invasive optical biosensor
Cell 155:1639–1651. doi:10.1016/j.cell. for probing cell signaling. Sensors
2013.11.029 7:2316–2329
20. Liu L, Duclos G, Sun B, Lee J, Wu A, Kam Y, 31. Benton G, Kleinman HK, George J, Arnaou-
Sontag ED, Stone HA, Sturm JC, Gatenby RA, tova I (2011) Multiple uses of basement mem-
Austin RH (2013) Minimization of thermody- brane-like matrix (BME/Matrigel) in vitro and
namic costs in cancer cell invasion. Proc Natl in vivo with cancer cells. Int J Cancer
Acad Sci U S A 110:1686–1691. doi:10.1073/ 128:1751–1757. doi:10.1002/ijc.25781
pnas.1221147110 32. Vinci M, Gowan S, Boxall F, Patterson L, Zim-
21. Kniazeva E, Putnam AJ (2009) Endothelial cell mermann M, Court W, Lomas C, Mendiola M,
traction and ECM density influence both Hardisson D, Eccles SA (2012) Advances in
160 Nicole K. Febles et al.

establishment and analysis of three-dimen- 42. Fang Y (2013) Troubleshooting and deconvo-
sional tumor spheroid-based functional assays luting label-free cell phenotypic assays in drug
for target validation and drug evaluation. BMC discovery. J Pharmacol Toxicol Methods
Biol 10:29. doi:10.1186/1741-7007-10-29 67:69–81. doi:10.1016/j.vascn.2013.01.004
33. Ferrie AM, Deichmann OD, Wu Q, Fang Y 43. Fang Y (2014) Label-free drug discovery.
(2012) High resolution resonant waveguide Front Pharmacol 5:52. doi:10.3389/fphar.
grating imager for cell cluster analysis under 2014.00052
physiological condition. Appl Phys Lett 44. Fang Y (2014) Label-free cell phenotypic drug
100:223701. doi:10.1063/1.4723691 discovery. Comb Chem High Throughput
34. Ferrie AM, Wu Q, Deichmann O, Fang Y Screen 17(7):566–578
(2014) High frequency resonant waveguide 45. Song MS, Salmena L, Pandolfi PP (2012) The
grating imager for assessing drug-induced car- functions and regulation of the PTEN tumour
diotoxicity. Appl Phys Lett 104:183702. suppressor. Nat Rev Mol Cell Biol
doi:10.1063/1.4876095 13:283–296. doi:10.1038/nrm3330
35. Fang Y, Ferrie AM, Fontaine NH, Yuen PK 46. Kasinskas RW, Venkatasubramanian R, Forbes
(2005) Characteristics of dynamic mass NS (2014) Rapid uptake of glucose and lactate,
redistribution of EGF receptor signaling in liv- and not hypoxia, induces apoptosis in three-
ing cells measured with label free optical bio- dimensional tumor tissue culture. Integr Biol
sensors. Anal Chem 77:5720–5725. doi:10. 6:399–410. doi:10.1039/c4ib00001c
1021/ac050887n 47. Hanahan D, Weinberg RA (2011) Hallmarks
36. Fang Y (2011) Label-free biosensors for cell of cancer: the next generation. Cell
biology. Intl J Electrochem 2011:e460850. 144:646–674. doi:10.1016/j.cell.2011.02.
doi:10.4061/2011/460850 013
37. Schröder R, Janssen N, Schmidt J, Kebig A, 48. Valastyan S, Weinberg RA (2011) Tumor
Merten N, Hennen S, M€ uller A, Bl€attermann metastasis: molecular insights and evolving
S, Mohr-Andr€a M, Zahn S, Wenzel J, Smith paradigms. Cell 147:275–292. doi:10.1016/
NJ, Gomeza J, Drewke C, Milligan G, Mohr K, j.cell.2011.09.024
Kostenis E (2010) Deconvolution of complex 49. Stetler-Stevenson WG, Aznavoorian S, Liotta
G protein-coupled receptor signaling in live LA (1993) Tumor cell interactions with the
cells using dynamic mass redistribution mea- extracellular matrix during invasion and metas-
surements. Nat Biotechnol 28:943–949. tasis. Annu Rev Cell Biol 9:541–573
doi:10.1038/nbt.1671
50. Poincloux R, Lizárraga F, Chavrier P (2009)
38. Verrier F, An S, Ferrie AM, Sun H, Kyoung M, Matrix invasion by tumour cells: a focus on
Fang Y, Benkovic SJ (2011) GPCRs regulate MT1-MMP trafficking to invadopodia. J Cell
the assembly of a multienzyme complex for Sci 122:3015–3024. doi:10.1242/jcs.034561
purine biosynthesis. Nat Chem Biol
7:909–915 51. Lee C, Kim JS, Waldman T (2004) PTEN gene
targeting reveals a radiation-induced size
39. Ferrie AM, Wang C, Deng H, Fang Y (2013) checkpoint in human cancer cells. Cancer Res
Label-free optical biosensor with microfluidics 64:6906–6914. doi:10.1158/0008-5472.
identifies an intracellular signalling wave CAN-04-1767
mediated through the β2-adrerengic receptor.
Integr Biol 5:1253–1261 52. Samuels Y, Diaz LA Jr, Schmidt-Kittler O,
Cummins JM, Delong L, Cheong I, Rago C,
40. Li G, Lai F, Fang Y (2012) Modulating cell-cell Huso DL, Lengauer C, Kinzler KW, Vogelstein
communication with a high-throughput label- B, Velculescu VE (2005) Mutant PI3K3CA
free cell assay. J Lab Automation 17:6–15. promotes cell growth and invasion of human
doi:10.1177/2211068211424548 cancer cells. Cancer Cell 7:561–573. doi:10.
41. Pai S, Verrier F, Sun H, Hu H, Ferrie AM, 1016/j.ccr.2005.05.014
Eshraghi A, Fang Y (2012) Dynamic mass 53. Fang Y (2015) Label-free cell phenotypic
redistribution assay decodes differentiation of profiling and screening: techniques, experi-
a neural progenitor stem cell. J Biomol Screen mental design and data assessment. Methods
17:1180–1191. doi:10.1177/ Pharmacol Tox 53:233–252. doi:10.1007/
1087057112455059 978-1-4939-2617-6_2
 Chapter 11

Label-Free Biosensors Based on Bimodal Waveguide

(BiMW) Interferometers
Sonia Herranz*, Adrián Fernández Gavela*, and Laura M. Lechuga

Abstract
The bimodal waveguide (BiMW) sensor is a novel common path interferometric transducer based on the
evanescent field detection principle, which in combination with a bio-recognition element allows the direct
detection of biomolecular interactions in a label-free scheme. Due to its inherent high sensitivity it has great
potential to become a powerful analytical tool for monitoring substances of interest in areas such as
environmental control, medical diagnostics and food safety, among others. The BiMW sensor is fabricated
using standard silicon-based technology allowing cost-effective production, and meeting the requirements
of portability and disposability necessary for implementation in a point-of-care (POC) setting.
In this chapter we describe the design and fabrication of the BiMW transducer, as well as its application
for bio-sensing purposes. We show as an example the biosensor capabilities two different applications: (1)
the immunodetection of Irgarol 1051 biocide useful in the environmental field, and (2) the detection of
human growth hormone as used in clinical diagnostics. The detection is performed in real time by
monitoring changes in the intensity pattern of light exiting the BiMW transducer resulting from
antigen–antibody interactions on the surface of the sensor.

Key words Bimodal waveguide interferometry, Evanescent field biosensor, Integrated optics, Immu-
noassay, Biofunctionalization, Biorecognition

1 Introduction

Most analytical methodologies used in fields such as environmental

monitoring, medical diagnostics and food safety are based on time-
consuming and expensive techniques that must be performed by
specialized personnel in laboratory environments. This usually
involves extensive time and effort due to the need for analysis and
sampling, as well as interpretation of results. These can cause critical
delays in clinical treatment decisions or removal of contaminated
products from the market in order to avoid extended health

*These authors contributed equally to this work.

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_11,
© Springer Science+Business Media LLC 2017

161
162 Sonia Herranz et al.

impacts. Therefore it is imperative to use reliable diagnostic tools

that ensure sensitive, rapid, and simple analysis, and to expedite the
diagnostic process.
In this context, biosensors are beginning to overcome the main
drawbacks of classical analytical techniques. Modern biosensors are
often portable, easy-to-use, and highly sensitive, and can operate in
real-time and in-situ. Most of these photonic biosensors employ
working principles and components (as waveguides) from the field
of Integrated Optics (IO) (for basic information about IO, readers
are referred to references [1, 2]). The IO-based biosensors are
normally fabricated with standard Si-technology which employs
well-known fabrication processes and silicon (Si) or silicon nitride
(Si3N4) as core waveguide materials (see Note 1). Within IO-based
biosensors, photonic interferometric biosensors based on the eva-
nescent wave principle are highly suitable for many applications due
to their high sensitivity and great potential to be integrated in a
POC device [3].

1.1 Interferometric The working principle of interferometric biosensor relies in the

Sensors Based on creation of an interference pattern, generated by the superposition
Integrated Optics of two or more light waves. In a common interferometric device,
the incoming light beam is split in two beams of equal intensity that
travel through different optical paths (arms) and are recombined
before arriving at a detector, which collects the interferometric
signal, as Fig. 1 shows.
For biosensing applications, one of the arms is used as a
reference while the other acts as sensing one. Interferometric bio-
sensors are based on the evanescent wave detection principle, there-
fore any modification in the sensing arm, such as binding events or
concentration variation, induce a change of the effective refractive
index of the guided light wave producing an interference pattern at
the output.
The output intensity, I(t), of an interferometric sensor is related
with the phase difference between the two arms, Δφ(t) as follows:
I ðt Þ / cos ðΔφÞ

2πL
Δφðt Þ ¼ Δneff
λ
where L is the length of the sensing arm, λ the wavelength of the
light, and Δneff the difference between the effective refractive index
of the two propagated modes.
The phase difference, in an experimental interferometric
signal, is determined by the variation between the initial value,
Δφ1, and the final value, Δφ2, given by the number of fringes,
where a complete fringe represents a phase difference of 2π, as
Fig. 1 shows.
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 163

Fig. 1 Scheme of the working principle of an interferometric waveguide biosensor

Our work focuses on the development of IO nano-

immunosensors based on BiMW transducers and their integration
into a complete Lab-On-a-Chip (LOC) platform. BiMW is a novel
interferometric transducer, which has been designed for sensitive,
label-free, and direct detection of biomolecular interactions. In this
IO interferometer, light is coupled in a straight single-mode rib
waveguide, and after passing through a step-junction, two trans-
versal modes with the same polarization are excited (see Fig. 2).
Due to the presence of two propagated modes in a common path
with different evanescent field profiles, any change in the refractive
index causes an interference pattern at the device output. The
intensity distribution at the output depends principally on the
refractive index at the sensor surface, as well as on the dimensions
of the BiMW structure (layers thicknesses and rib dimensions). The
bimodal section of the BiMW includes a sensing window area free
of cladding, in order to have access to the waveguide surface for
sensing purpose (see Fig. 2), following the interferometric working
principles explained previously. Therefore, by measuring the out-
put light intensity the sensor response can be determined in a
quantitative way [4].
164 Sonia Herranz et al.

Fig. 2 (a) Scheme of a BiMW interferometric sensor. (b) Longitudinal side view
of a BiMW showing the distribution of the electromagnetic field in several points

Interferometric sensors, and in particular BiMW interfero-

meters, have proved to be one of the most sensitive devices for
label-free analysis, reaching sensitivity to homogeneous changes in
the refractive index of 107–108 Refractive Index Unit (RIU),
meaning that they have the potential to detect biomolecular inter-
action at pM levels. However, to use this technology for biosensing
applications the transducer must be highly selective for instance,
particularly with applications such as immunoassays (immunosensors).
The limit of detection (LOD) achieved with an immunosensor is
dependent not only on the transducer itself but also on the
biorecognition element and the biofunctionalization protocol
used to combine them.
Here we describe the design and fabrication process of BiMW
interferometers, paying special attention to parameters affecting
their performance (such as waveguide length or rib dimensions) in
order to achieve the highest possible sensitivity. Additionally, we
report the application of BiMW interferometry to biosensor
development. Two applications in the fields of environmental
monitoring and clinical diagnostics are described: the detection
of Irgarol 1051 biocide and that of human growth hormone,
respectively.
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 165

2 Materials

2.1 Equipment 1. Plasma cleaner reactor: Standard plasma system Femto,

version A (40 kHz, 0–100 W) from Diener Electronic GmbH
(Ebhausen, Germany).
2. Ultrasonic bath FB 15054, from Fisher Scientific (Madrid,
Spain).
3. ABBE refractometer (Optic Ivymen System, Spain).
4. Clean Room facilities: the BiMWs are fabricated from a stan-
dard 4-in. (p-type) Si wafer with a thickness of 500 μm
purchased from Siltronic Company. The fabrication of the
BiMW is done at Clean Room facilities (class 100) using stan-
dard microelectronics optical photolithography, reactive ion
etching, wet etching, and various deposition methods.
Over the Si substrate a 2 μm thick layer of thermal oxide
(nbottom,clad ¼ 1.46) is grown. Then, a 340 nm thick core
layer of Si3N4 (ncore ¼ 2.00) is deposited by Low Pressure
Chemical Vapor Deposition (LPCVD). An absorbent layer of
poly-crystalline Si (npoly ¼ 4.06) of 100 nm is deposited by
LPCVD over a silicon dioxide (SiO2) layer of 200 nm previ-
ously deposited. Finally, a SiO2 (ntop,clad ¼ 1.46) layer 1.5 μm
thick is deposited by Plasma Enhanced Chemical Vapor Depo-
sition (PECVD).

2.2 Setup 1. Fluid cell is made of polydimethylsiloxane (PDMS, Sylgard)

Components fabricated using a methacrylate (PMMA) topographic master.
2. Syringe pump (NE300, New Era) for maintaining a constant
flow during all the biochemical detection process.
3. An injection valve (V-451, Idex) allows the injection of
different solutions without changing the flow rate.
4. Laser diode (ML101J27, Mitsubishi) with λ0 ¼ 660 nm and
P ¼ 120 mW, mounted on a compact temperature controlled
laser diode mount (TCLDM9, Thorlabs), is used as light
source. A Temperature controller (TED 200C, Thorlabs) and
current controller (LDC220C, Thorlabs) are employed to sta-
bilize the laser diode.
5. To couple the light into the waveguide a lenses system is used,
composed by: collimated lens (C240TME-D, Thorlabs),
polarization-dependent isolator (IO-3D-660-VLP, Thorlabs)
and coupling objective 40 (Achro, Leica).
6. A four quadrants photodetector (S4349, Hamamatsu) is
employed for collecting the light at the end of the device.
The signals are amplified through standard benchtop instru-
mentation (PDA200C, Thorlabs).
166 Sonia Herranz et al.

7. A digital acquisition card (6251, National instruments) for

reading the photodetector signal in real time. The signal acqui-
sition is controlled by a LabVIEW-based application.
8. 3-axis precision micro-position stage (Nanomax-TS, Thorlabs)
is used for laser-BiMW alignment.
9. A temperature sensor (AD590, Thorlabs) is used to achieve a
temperature feedback circuit together with a thermo-electric
cooler (TEC3-1.5, Thorlabs), operated through a bench top
temperature controller (TED200C, Thorlabs) allowing a
temperature resolution of 0.01  C.

2.3 Reagents 1. Potassium phosphate monobasic (KH2PO4, P0662-500G),

sodium phosphate dibasic (Na2HPO4, S0876-500G), sodium
chloride (NaCl, S9625-1KG-D), potassium chloride (KCl,
60,128-250G-F), 2-(N-Morpholino)ethanesulfonic acid (MES,
M3671-50G), sodium carbonate (Na2CO3, S7795-500G),
sodium hydroxide (NaOH, S8045-500G), sodium dodecyl
sulfate (SDS, 436,143-25G), hexadecyltrimethylammonium
bromide (CTAB, H6269-100G), bovine serum albumin (BSA,
A7906-10G), (3-aminopropyl)triethoxysilane (APTES, A3648-
100ML), N-hydroxysulfosuccinimide sodium salt (sulfo-NHS,
56,485-1G), N-(3-dimethylaminopropyl)-N-ethylcarbodiimide
hydrochloride (EDC, E1769-1G, see Note 2), N,N-dimethyl-
formamide anhydrous (DMF, 227,056-100ML), toluene
(244511-1 L), pyridine (270970-100ML), p-phenylene
diisothiocyanate (PDITC, 258,555-5G), hydrofluoric acid
(HF, 695,068-25ML-D), ammonium fluoride (NH4F,
338,869-25G), and phosphoric acid (85 wt. % in H2O,
W290017-1KG-K) can be purchased from Sigma-Aldrich.
2. Acetone (161007.1211), ethanol absolute (161086.1211),
methanol (161091.1211), nitric acid (65%, 473255.1611),
and hydrochloric acid (HCl 37%, 471020.1611) can be
purchased from Panreac.
3. Recombinant human Growth Hormone (r-hGH, 22 kDa) was
provided by Dr. Parlow (National Hormone and Peptide
Program (NHPP), National Institute of Diabetes and Digestive
and Kidney Diseases (NIDDK), CA, USA). The monoclonal
antibody (anti-hGH) was produced and characterized at the
Department of Immunology and Oncology from the National
Center of Biotechnology (CNB-CSIC, Madrid, Spain) [5].
4. Irgarol 1051, Irgarol 1051 derivative 4e (hapten) [6], and anti-
Irgarol 1051 serum (As87) were provided by Prof. Marco
(Nanobiotechnology for Diagnostics Nb4D group, IQAC-
CSIC, Barcelona, Spain).
5. Deionized water from a Milli-DI® Water Purification System,
Merck Millipore, USA.
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 167

2.4 Buffer 1. Buffered hydrofluoric acid (BHF): mixture of HF and NH4F,

Composition 1:6 (v/v)
2. MES buffer: solution of 0.1 M MES and 0.5 M NaCl in water,
pH adjusted to 5.5. Store at 4  C.
3. Phosphate buffer saline (PBS): solution of 137 mM NaCl,
2.7 mM KCl, 8 mM Na2HPO4, and 2 mM KH2PO4 in
water, pH adjusted to 7.4. Store at 4  C. We recommend
preparing a ten-fold concentration PBS (10 PBS, pH 7–7.5)
as stock solution, which can be stored at room temperature
(RT). Then, prepare a working solution of 1 PBS by dilution
from the stock solution. Alternatively, PBS can be purchased
from Sigma-Aldrich (P7059-1 L).
4. Carbonate buffer (CB): 0.1 M Na2CO3 at pH 9.

3 Methods

One of the first steps to develop a BiMW transducer of high

sensitivity is to simulate a range of different core and cladding
thicknesses, taking into account the rib dimensions and the tole-
rances at the clean room facilities where the fabrication will be
done. The sensitivity of an interferometry-based transducer is
directly related with the dimension of the interaction surface (see
Subheading 1.1), thus higher sensitivity is obtained with a long
sensing area. For that reason a compromise between the total
length of the BiMW chip and the quantity of devices per wafer
needs to be reached, for a cost-effective fabrication.
Once the optimum dimensions have been defined by modelling,
lithographic masks are designed according the chosen dimensions
and then the BiMW devices are fabricated at clean room facilities.
The fabrication processes should be reproducible and must guaran-
tee the same reliable performance for all the BiMW transducers.
To characterize the device, BiMW is placed in a custom-
designed setup. The light from a laser diode is end-fire coupled
using a lens system, as Fig. 3 shows. For light readout, a two-
section photodetector is employed as the interference of the
common path BiMW interferometer generates a variation of the
output intensity distribution, but the total intensity should be
constant. The photodetector must be placed directly at the BiMW
output, and total intensity must be divided in the up and down
quadrant of the photodetector (see Fig. 3). The sensor response, SR,
is determined through the equation:
I up  I down
SR ¼
I up þ I down
168 Sonia Herranz et al.

Fig. 3 Scheme of the experimental setup employed for the evaluation of the BiMW sensors

where Iup and Idown are the intensities acquired for the up and down
quadrants of the photodetector, respectively.
For the application of the BiMW interferometer as a biosensor
the surface must be modified with a recognition element or
biological receptor which, when immobilized onto the surface,
increases the transducer selectivity for the substance to be moni-
tored. The environmental and clinical diagnostic BiMW interfero-
metric biosensor applications will be illustrated.
Irgarol 1051 is a protective algaecide (biocide, s-triazine
compound) used in antifouling paints to prevent attachment and
growth of biofouling organisms in the aquatic environment. It leaks
out from the paint and by its toxicity Irgarol 1051 prevents accu-
mulation of organisms onto the painted surface (boat hulls or
marine installations). Traces of Irgarol 1051 have been detected
in the marine environment [7, 8], and due to its broad toxicity
(photosystem-II inhibitor), Irgarol 1051 is potentially toxic toward
valuable aquatic species (mainly small aquatic plants), and its pres-
ence in the environment can cause significant environmental
problems. In fact, a number of studies have demonstrated its
negative impact on marine organisms such as corals.
Human growth hormone (hGH) is produced by the pituitary
gland and promotes growth, regulates the activity of vital organs
and helps to preserve the health of the whole organism. Its defi-
ciency commonly affects children and has several negative effects
such as hypoglycemia in newborn infants or stunted growth in
children. hGH deficiency is results from several genetic diseases
such as Turner syndrome and Prader-Willi syndrome. Thus the
hGH level in blood or urine can be used for the diagnosis of
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 169

Fig. 4 Immunoassay formats: (a) Competitive immunoassay format selected for Irgarol 1051 detection (b)
Direct immunoassay format selected for hGH detection. For sensor surface details, see Fig. 7

those disorders. Additionally, r-hGH has been used by many sports-

men in an attempt to improve athletic performance, and the control
of hGH level in urine is usually employed as a doping test.
Both biosensor applications are based on antigen–antibody
specific interactions. For the analysis of Irgarol 1051, a low molec-
ular weight pollutant, a competition immunoassay has been
selected (see Fig. 4a). The transducer surface is functionalized
with an Irgarol 1051 derivative (hapten 4e). A fixed amount of an
anti-Irgarol antibody is mixed and allowed to interact with the
sample and then, the mixture is injected to the flow cell where the
sensor is placed. The free antibodies interact with the immobilized
4e receptor and generate a binding response, which is inversely
related to the concentration of Irgarol 1051 in the sample (see
Fig. 5a). The detection of hGH is carried out by using a direct
immunoassay (see Fig. 4b). In this case the transducer surface is
functionalized with an anti-hGH antibody. The hGH present in the
170 Sonia Herranz et al.

a b
Amax Amax
100% 100%
90%
Sensor response

Sensor response
80% 80%

50% 50%

20% 20%
Amin 10% Amin
0% 0%

LOD IC50 [Analyte] LOD EC50 [Analyte]

DR DR

Fig. 5 Standard calibration curves showing the relationship between the analyte concentration ([Analyte],
logarithmic scale) and the sensor response for: (a) competitive and (b) direct immunoassays

Table 1
Layer parameters employed for the evaluation of the single-mode and
bimodal condition

Layer Material Thickness (nm) Refractive index

Top cladding Water/SiO2 1500 1.33/1.46
Core Si3N4 150–340 2.00
Bottom cladding SiO2 2000 1.46

injected sample interacts with the immobilized antibody and

generates a binding response directly related to its concentration
(see Fig. 5b).

3.1 BiMW Simulation (a) To design a BiMW we need to choose optimal materials,
and Fabrication which ensure a cost-effective fabrication, a further integra-
tion onto a complete LOC platform, and a highly sensitive
3.1.1 BiMW Simulation
device. As the transducers are designed for biosensors appli-
cations, SiO2/Si3N4/SiO2 waveguides are selected [9].
Table 1 summarizes the parameters employed in the calcula-
tion, relative to the materials’ refractive indices and their
thicknesses.
(b) Once the material is defined, core dimensions are analyzed to
ensure the best sensitivity of the device. For that, different
considerations have to be taken into account [4] (see Fig. 6):
l Single-mode behavior (mode TE00 or TM00) in section A
in Fig. 6b.
l Bimodal behavior (mode TE00 and TE10 or TM00 and
TM10) in sections B, C, and D in Fig. 6b.
l Performances achievable in the clean room facilities.
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 171

Fig. 6 (a) Front cross section of the employed rib-waveguide, indicating the main parameters described on the
text. (b) Longitudinal side view of the BiMW device. Light is injected in the single-mode section (A). After an
abrupt step junction, two modes propagate till the end of the device. Regions B and D are covered by silicon
dioxide while C is exposed to the external medium (sensing area). (c) Picture of a chip with 20 BiMW sensors

Table 2
Optimal parameters for core dimensions in a high sensitivity BiMW

Section tcore (nm) hrib (nm) wrib (nm)

Single-mode 150 1.5 3000
Bimodal 340 1.5 3000

In bimodal section, bulk and surface sensitivity must be

calculated for different core thicknesses (tcore). Once tcore
has been fixed, the thickness of the single-mode section
is chosen to determine how the energy carried by the funda-
mental mode, propagating in section A in Fig. 6b, is
distributed to the two modes propagating in section B in
Fig. 6b.
Table 2 summarizes the optimal parameters for core
dimensions in a BiMW for a λ ¼ 660 nm and TE polarization
(see Fig. 6a).
(c) Once the core dimensions are defined, the sensor design is
completed by choosing the length of the different sections. It
is required to take into account the following considerations:
l The need to adapt a microfluidic system to the BiMW for
biosensing applications.
l To ensure high sensitivity while keeping a reasonable total
length.
l Distribution of the BiMWs in 4-in. silicon wafer to opti-
mize cost-effective fabrication.
The following values were chosen: L0 ¼ 3 mm, Lin ¼
4.5 mm, Lsa ¼ 15 mm, and Lout ¼ 8.5 mm (see Fig. 6b).
(d) In a 4-in. silicon wafer, 240 BiMW are distributed in 12
different chips (1 cm wide  3.1 cm long) with 20 sensors
per chip (see Fig. 6c).
172 Sonia Herranz et al.

Fig. 7 Scheme of the fabrication process of integrated BiMW sensors. (a) A layer
of SiO2 and Si3N4 is deposited on a Si wafer. (b) Step-junction etched in Si3N4
layer for single-mode section definition. (c) Rib structure is defined. (d) Poly-
crystalline deposition over the Si3N4 and etching process to open desired
window on this layer. (e) SiO2 deposition and sensing area opening

3.1.2 BiMW Fabrication Figure 7 summarizes the fabrication process. Below, each step of
Fig. 7 is explained.
(a) The fabrication processing starts with a 4-in. (p-type) Si
wafer with a thickness of 500 μm, over which a 2 μm thick
layer of thermal oxide is grown. Then a 340 nm thick core
layer of Si3N4 is deposited by LPCVD.
(b) The thickness of the Si3N4 is reduced 190 nm to obtain a
150 nm single-mode section using a wet etching process with
75% phosphoric acid at 160  C. A hard mask constituted by a
layer of SiO2 deposited by PECVD and defined by photoli-
thography is employed in this step to protect unexposed
regions.
(c) Using a wet etching process with BHF and a standard pho-
tolithography process, the rib structure of the waveguide
(3 μm in width and 1.5 nm in height) is defined. The low
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 173

etching rate of the BHF for Si3N4 allows precise control of

the process.
(d) A 100 nm layer of poly-crystalline Si is grown by LPCVD
over 200 nm of SiO2 deposited by PECVD, which is intro-
duced to improve the adhesion to the underlying Si3N4
surface. A standard lithography process is used to define the
structure. Dry etching (RIE) is employed to remove the
poly-crystalline Si layer and 150 nm of silicon dioxide, to
ensure vertical sidewalls. Then, a wet etching process is used
to eliminate the remaining 50 nm of SiO2. Poly-crystalline Si
layer is employed as absorbing material to define lateral bands
along the waveguide path.
(e) A SiO2 layer 1.5 μm thick is deposited by PECVD as top
cladding layer. In order to open the sensing area a standard
lithography process is done, employing a wet etching process
with BFH to remove the desired SiO2.

3.2 Setup (a) The BiMW chip is placed in an aluminum homemade holder
Configuration as shown in Fig. 8. Into the holder a temperature stabiliza-
tion is included, and the entire system is mounted on a 3-axis
stage. In this way, it is possible to focus the light in the
desired BiMW sensor optimally.
(b) As the total output intensity must be divided in the up and
down quadrant of the photodetector, this one is joined in a
2-axis stage assembly to the aluminum chip holder, to ensure
the right position of the photodetector at the BiMW

Fig. 8 Photograph of a BiMW experimental setup

174 Sonia Herranz et al.

output. As photodetector has four quadrants, two BiMW

could be measure simultaneously, using two quadrants per
sensor. This position is essential if a modulated signal is
required [10].
(c) A digital acquisition card connected to a computer is
employed to process the photodetector signal. A homemade
Labview software acquires a continuous photodetector signal
in real time.
(d) A diode laser is placed in a temperature controlled mount. To
guarantee an optimal end-fire incoupling of the laser beam, a
lenses system is employed. A lens is used to collimate the
divergent laser beam of the diode. After this lens a
polarization-dependent isolator is placed to avoid optical
back reflections and to guarantee the required polarization.
Finally, a 40x objective is employed to focus the collimated
beam to the BiMW input for end-fire coupling.
(e) All the biological interactions must be performed in a liquid
medium. For that reason, a PDMS fluid cell must be placed
on top of the sensor chip. This fluid cell has channels of 11 μL
volume, and an inlet and outlet for liquid flow. To avoid
liquid leaks between the PDMS channels and the BiMW
chip, and in order to properly align the fluid cell, four screws
and a methacrylate sheet are employed to press, as Fig. 9
shows.

Fig. 9 Scheme of the fluidic cell employed for BiMW evaluation

 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 175

(f) The flow delivery system is mounted by connecting a syringe

pump with an injection valve and this one to the cell flow
inlet by using Teflon tubes. The length of the tubes should be
reduced as much as possible to minimize diffusion effects.
The loop of the injection valve has a variable length, depend-
ing on the desired volume. Here, 150 μL and 250 μL loops
have been used for Irgarol 1051 and hGH immunoassays,
respectively. An extra tube is used from the cell outlet to the
waste.

3.3 Sensing 1. The determination of the BiMW biosensor bulk sensitivity

Characterization: allows evaluating the response of the transducer, independently
Homogeneous Sensing of the biofunctionalization process. For that, milli-Q water is
supplied to the sensor in a continuous rate (40 μL/min), and
solutions providing small refractive index changes, such as
different concentrations of HCl are injected over the sensor.
2. Previously to the injections, the refractive index of the solutions
is checked with a commercial refractometer, and the difference
of refractive index, Δn, between the continuous buffer running
and the injected solutions is calculated.
3. Taking into account the sinusoidal relationship between the
phase change, Δφ, and the output intensity, I, in an interfero-
metric device [3], changes of the refractive indexes over the
sensor area, induce output signal alterations through the
evanescent field interaction.
4. Phase variation, Δφ, is plotted versus index variation, Δn,
obtaining a calibration curve as Fig. 10 shows. The slope of
the fitting curve, expressed in rad/RIU, represents the bulk
sensitivity of the sensors. For BiMW, the sensitivity (Sbulk) is
around 1700  2π rad/RIU, for TE polarization.
5. The LOD, Δnmin, is calculated through the equation:
Δφmin ΔS R , min π 3  σSR π
Δnmin ¼ ¼ ¼
S bulk S bulk V S bulk V
where SR,min is the minimum sensor response, defined as three
times the system noise, σ S R . V is the visibility of the output
signal, which corresponds to π rad in the phase variation. LOD
for BiMW sensors are in the range of 107–108 RIU.

3.4 Surface The most widely applied approaches to chemical modification of

Biofunctionalization Si3N4 surfaces are based on the molecular grafting of the native
oxide layer present onto this material by forming self-assembled
monolayer (SAM) of alkyl-silanes. Among the wide variety of com-
mercially available silanes, APTES ((3-aminopropyl)triethoxysi-
lane) is one of the most employed for chemical modification of
Si-based photonic devices. Surfaces modified with APTES present
176 Sonia Herranz et al.

Fig. 10 Calibration curve for a BiMW sensor. Insets represent the temporal interferometric signals
corresponding to the entrance of solutions with different refractive indexes

3-aminopropyl moieties, which allow the further covalent immobi-

lization of bioreceptors presenting carboxyl groups in their
structure. Moreover, by using different cross-linker molecules, the
variety of bioreceptors that can be covalently attached is multiplied.
We describe the use of APTES to chemically modify the BiMW
chips. APTES-modified chips are then functionalized by covalently
attaching the hapten 4e, resulting in a biosensor surface which can
detect Irgarol 1051 by a competitive immunoassay (Surface I). In
the second approach, APTES-modified chips are biofunctionalized
by covalently attaching the anti-hGH antibody, using p-phenylene
diisothiocyanate (PDITC) as a cross-linker molecule (Surface II).
PDITC displays two isothiocyanate groups, which can couple the
free amine groups of a bioreceptor. Additionally, PDITC has the
potential to form well-organized assemblies driven by π–π stacking,
resulting in improved antifouling properties [11]. Surface II is used
to detect hGH by a direct immunoassay.
In order to avoid contamination of the surface and to enhance
silanization efficiency and reproducibility, the BiMW chips are
thoroughly cleaned and oxidized forming a fresh prepared oxide
layer with a high density of silanol groups (reactive hydroxyl
groups). This procedure is carried out immediately prior to the
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 177

Fig. 11 Functionalization scheme/protocol: (1) cleaning and oxidation; (2) APTES silanization; (3) covalent
attachment of hapten 4e; (4) activation of APTES-modified surface with PDITC, and (5) covalent attachment of
anti-hGH antibody. Sensor surface I: 4e-functionalized surface, used for Irgarol 1051 immunodetection.
Sensor surface II: anti-hGH antibody-functionalized surface, used for hGH immunodetection

surface biofunctionalization. Avoid storage of cleaned and oxidized

chips, which are susceptible to contamination.
Figure 11 shows a scheme of the different steps of the functio-
nalization protocol. Briefly, the BiMW chips are cleaned, oxidized
(Fig. 11, step 1), and immediately silanized with APTES (Fig. 11,
step 2). Then, sulfo-NHS-ester activated 4e hapten is covalently
attached (Fig. 11, step 3) getting sensor surface I. For preparing
sensor surface II, APTES-modified chips are activated with PDITC
(Fig. 11, step 4) and finally anti-hGH antibody is covalently bound
(Fig. 11, step 5).

3.4.1 Cleaning Procedure (a) Wipe the BiMW chip down using a cleanroom paper soaked
in acetone.
(b) Sequentially rinse the chip with acetone, ethanol, and
deionized water, and dry under nitrogen flow.
(c) Immerse the chip in 1% SDS aqueous solution and sonicate
for 5 min.
(d) Rinse the chip with deionized water and dry again under
nitrogen flow.
178 Sonia Herranz et al.

(e) Immerse the chip in a mixture of methanol and 37% fuming

hydrochloric acid (1:1, v/v) and sonicate for 10 min (see
Note 3).
(f) Finally, generously rinse the chip with deionized water and
dry carefully under nitrogen flow.

3.4.2 Oxidation (a) Treat the cleaned BiMW chip with oxygen plasma (100 W, 45
Procedure sccm) for 5 min (see Note 4).
(b) Immerse the chip in 15% HNO3 at 75  C for 25 min.
(c) Rinse generously the chip with deionized water, dry carefully
under nitrogen flow and immediately immerse it in the
silanization solution.

3.4.3 BiMW (a) Clean and oxidize the BiMW chip according to the previous
Interferometer Silanization sections.
Procedure Using APTES (b) Immediately immerse the BiMW chip in a solution of 1%
APTES and incubate at RT for 1 h. Two different solvents
have been evaluated: absolute ethanol and toluene.
(c) Amply rinse the chip with the silanization solvent and
deionized water, sequentially.
(d) Dry the chip thoroughly under nitrogen flow and thermally
treat it in an oven at 110  C for 1 h.

3.4.4 Covalent (a) At this point, an APTES-modified BiMW (silanization sol-

Immobilization of Irgarol vent: toluene) chip is positioned on the experimental appa-
Derivative “4e” ratus, and a PDMS multichannel gasket (flow-chamber) is
pressed against it, forming the assay flow channels (see
Fig. 9). The immobilization of the Irgarol derivative 4e is
carried out inside the measuring setup, using a stopped-flow
incubation approach, i.e., by filling the flow channel with
the reaction mixture and stopping the flow. After the
appropriate incubation time, the reaction is stopped by
restarting the flow and washing out the surface with
deionized water.
(b) Prepare a solution mix of hapten 4e, EDC, and sulfo-NHS in
MES buffer. Concentrations must be optimized for each
application; however, concentration typically ranges from
10 to 50 μg/mL of hapten, and 0.2–0.4 M and
0.05–0.1 M of EDC and sulfo-NHS, respectively. Pre-
incubate this mixture for 30 min to 4 h at room temperature
(amine-reactive sulfo-NHS ester hapten intermediate
solution).
(c) Initiate a continuous flow (20 μL/min) of deionized water
and wait until stabilization of the interferometric signal.
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 179

(d) Inject into the flow cell a volume of the NHS-activate 4e

solution large enough to fill the whole assay flow channel
(25–50 μL). Once the channel is filled up, stop the flow and
incubate from 4 h to overnight.
(e) Remove the NHS-activate 4e solution by flushing deionized
water through the flow channel for 5 min.
(f) BiMW chip is now functionalized and ready to be used for
Irgarol 1051 immunodetection.

3.4.5 Covalent (a) Prior to the covalent attachment of the anti-hGH antibody,
Immobilization of Anti-hGH the APTES-modified chip (silanization solvent: ethanol)
Antibody must be activated with PDITC, which acts as cross-linker.
After PDITC activation the antibody can be immobilized
through the amine groups present in its structure.
(b) Prepare a 20 mM PDITC solution (5 mL) in anhydrous
DMF containing 1% pyridine.
(c) Immerse the APTES-modified chip in PDITC solution and
incubate in the dark for 2 h.
(d) Wash off unbound reagents by sequentially rinsing the chip
with acetone, ethanol and water, and dry it under nitrogen
stream.
(e) Add 50–100 μL of a solution of anti-hGH antibody in
carbonate buffer onto the PDITC-modified chip, cover it
with a coverslip and incubate overnight, in the dark.
Concentrations must be optimized for each application. For
hGH immunoassay, a 50 μg/mL solution was used.
(f) Rinse the functionalized chip with PBS and carefully dry it
under nitrogen stream.
(g) At this point, the functionalized BiMW chip is positioned on
the experimental apparatus and the PDMS flow-chamber
mounted. Before starting sample evaluation the sensor sur-
face is blocked to avoid nonspecific adsorption and reduce
background signal (see Note 5).
(h) Initiate a continuous flow of PBS and wait until the inter-
ferometric signal stabilizes.
(i) Inject into the flow cell a 2 mg/mL BSA solution in PBS
until the flow channel is full. Then stop the flow and incubate
for 1 h (see Note 6).
(j) The BiMW chip is now functionalized and ready to begin the
sample evaluation. If not used immediately chips must be
dried and stored at 4  C. Before use, the chip is positioned
again on the experimental apparatus and the sensor surface is
conditioning by maintain a continuous flow of PBS for 1 h.
180 Sonia Herranz et al.

3.5 Detection by a (a) Initiate a continuous flow of PBS and wait until stabilization
Competitive of the interferometric signal. This buffer will be used to
Immunoassay perform the Irgarol 1051 immunodetection assay.
3.5.1 Selection of the
(b) Prepare series of solutions (300 μL) with different dilution
Antibody Concentration factors of the antibody (serum As87). For Irgarol 1051
detection, dilutions ranging from 1:10,000 to 1:500 were
used.
(c) Inject into the flow cell the antiserum solutions with a
regeneration step in between, and measure the phase shift
(sensor response). Regeneration consists in disrupting the
receptor–antibody interaction while maintaining the recep-
tor onto the surface with minimal damage, by washing up the
sensor surface with the so-called regeneration solution.
Regeneration solution composition must be empirically opti-
mized (see Note 7). For Irgarol 1051 immunoassay, a 50 mM
NaOH solution was employed.
(d) Construct a plot of sensor response vs. antiserum concentra-
tion. Choose an antibody concentration value (fixed concen-
tration to perform the inhibition assay) below the saturation
limit, able to give high enough response intensity (phase
shift), so that at inhibition, signals would be still clear. Take
into account that by decreasing the antibody concentration, a
better limit of detection can be achieved (see Note 8).

3.5.2 Detection of Irgarol (a) Initiate a continuous flow of PBS (20 μL/min) and wait until
1051 the interferometric signal is stabilized.
(b) Prepare a series of solutions of different concentrations of
Irgarol 1051 in PBS (300 μL, standard solutions). Concen-
trations of the standard solutions to be measure depend on
the limit of detection for each specific application. For Irgarol
1051, the standard solutions were in the ng/L to μg/L
range.
(c) Mix the series of standard solutions with a volume of As87
serum so that the final dilution factor in solution equals that
previously selected. Incubate the mixture for 10 min.
(d) Pump one antibody-Irgarol standard solution into the flow
cell at 20 μL/min. This step produces a shift phase of the
wave (Δφ) due to the interaction of the free antibody with
the immobilized antigen.
(e) Wash off the unbound antibody by pumping PBS for a few
minutes (20 μL/min).
(f) Regenerate the surface by injecting 50 mM NaOH
(20 μL/min, 120 s), and wait until stabilization of the signal
(surface conditioning). At this point, the surface is ready for
the measurement of a second standard solution. In Fig. 12,
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 181

As87 +
As87 Irgarol 1051 NaOH 50 mM As87 NaOH 50 mM
80
60
40
ΔS
ΔS
20
ΔS
S, %

0
–20
–40
–60
NaOH 50 mM
–80
2 12 32 42 72 82
Time, min

Fig. 12 Real monitoring of anti-Irgarol 1051 antibody (As87) detection. The presence of Irgarol 1051 (1 μg/L) in
solution leads to a decrease of the sensor response. The sensor surface was regenerated with 50 mM NaOH

the sensor response for three consecutively samples (in PBS)

is represented, in which the inhibition of the signal by the
presence of Irgarol 1051 in the sample can be noticed.
(g) Construct a plot of sensor response vs. Irgarol 1051 concen-
tration and fit the data to an appropriate curve-fitting model
for obtaining the calibration curve (see Note 9). Limit of
detection (LOD) should be in the desired range for the
application. If not, try to optimize immobilization and
immunoassay conditions. Irgarol 1051 levels typically range
from 0.1–1.7 μg/L in marinas and 1–40 ng/L in coastal and
estuarine waters [12], so a LOD at low ng–pg/L level is
desirable.
(h) Mix the Irgarol samples (unknown concentrations) with an
equal concentration of anti-Irgarol serum used with the stan-
dard solutions. Incubate the mixture for 10 min.
(i) Pump the antibody-sample mixture into the flow cell at
20 μL/min. After monitoring the antigen–antibody interac-
tion, regenerate and condition the sensor surface in order to
prepare it for the next analysis.
(j) The concentration of Irgarol 1051 in the measured samples
can be estimated by interpolating the sensor response in the
fitting curve.

3.6 Detection of hGH 1. Initiate a continuous flow of PBS and wait until the signal is
by a Direct stabilized.
Immunoassay 2. Prepare a series of solutions of different concentrations of hGH
in PBS (400 μL, standard solutions). For hGH, the standard
solutions were in the ng/L to μg/L range.
182 Sonia Herranz et al.

3. Inject into the flow cell the hGH solutions with a regeneration
step in between (regeneration solution: 20 mM HCl), and
measure the phase shift (sensor response).
4. Construct a plot of sensor response vs. hGH concentration and
fit the data to an appropriate curve-fitting model for obtaining
the calibration curve. The hormone hGH usually appears at
ng/L levels in urine, so a LOD at low ng/L is desirable.
5. Inject into the flow cell the sample (unknown hGH concentra-
tion), wait until stabilization of the signal and measure the
sensor response.
6. Estimate the concentration of hGH in the sample by interpo-
lating the measured sensor response in the fitting curve.

3.7 Data Analysis To quantify the concentration of the target analyte in the sample,
the sensor response is interpolated into the calibration curve. In
order to minimize the estimation error and obtain accurate con-
centration values it is extremely important to employ an appropriate
curve-fitting model. The most frequently model used to analyse
immunoassay data is the 4-parameter logistic (4-PL) model, which
fit the data by using the equation:
ðA max  Amin Þ
y ¼ A min þ
b
1 þ xc
where y is the sensor response, x is the concentration of the target
analyte, Amax and Amin are the asymptotic ends, b is the slope at the
inflection point of the curve and c is the concentration at the
inflection point and represents the half maximal inhibitory concen-
tration (IC50) and the half maximal effective concentration (EC50)
in a competitive and a direct assay, respectively [13] (see Note 10).
Typically, the detection limit (LOD) for a competitive and a direct
assay is calculated as the analyte concentration for which the signal
is inhibited or increase by 10% of sensor signal range (Amax – Amin),
respectively, and the dynamic range (DR) of the biosensor is eval-
uated as the analyte concentrations that produced a sensor signal
between 20 and 80% of (Amax – Amin) (see Fig. 4).

4 Notes

1. Polymer-based biosensors are increasingly popular, but the

instability of the polymers, especially when they are in contact
with a buffer solution, restricts their use. Progress in polymer
material research could bring about advances in polymer-based
IO biosensors in the near future.
2. EDC is hygroscopic and quickly oxidizes in air. A good way to
handle EDC is to open the container under a dry inert
Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 183

atmosphere such as nitrogen (e.g., inside a glovebox), separate

it into aliquots, using containers which provide a tight seal, and
then store aliquots at 20  C. EDC solution must be prepared
immediately before use.
3. Concentrated hydrochloric acid is corrosive and releases acidic
gas. Therefore, handling of hydrochloric acid under a fume
hood for approved acids and the use of personal protective
equipment is strongly recommended.
4. Oxygen plasma treatment can be replaced by UV/Ozone
plasma treatment (UV/Ozone ProCleaner™, Biorforce
Nanosciences) for 1 h. Both oxidative pretreatments render
highly hydrophilic surfaces (water contact angle lower than
5 ). However, we recommend oxygen plasma rather than
UV/Ozone plasma treatment, since it is less time-consuming
and, in our experience, leads to surfaces with a slightly lower
roughness.
5. Prevention of nonspecific adsorptions of sample matrix com-
ponents is particularly important when working with label-free
optical transducers in which an overvalued sensor response due
to nonspecific adsorption of matrix components is difficult to
discriminate and leads to false positive (direct immunoassay) or
false negative (competitive immunoassay) results.
6. Alternately, adjust the flow rate to a value that keeps the
injected sample volume in contact with the sensor surface
during at least 1 h. Thus flow rates in the range of 2.0–2.5
and 3.5–4 μL/min are recommended for sample volumes of
150 and 250 μL, respectively.
7. The regeneration procedure depends on the antigen/antibody
pair used. The ideal regeneration buffer should effectively
disrupt the antigen-antibody interaction preserving the recep-
tor activity. In practice, all regeneration buffers cause some
damage on the bio-layer immobilized onto the transducer,
limiting the number of cycles that the same sensor surface can
be reused. Usually, either low pH (0.01–1 M HCl, 1–20%
formic acid, 0.01–1 M phosphoric acid) or high pH
(10–100 mM, 0.2 M Na2CO3) solutions, alone or in combina-
tion with chelating and surfactant reagents, are employed. Less
but also employed are solutions of high ionic strength (1 M
NaCl, 2–4 M MgCl2) or chaotropic agents (8 M urea, 6 M
guanidinium chloride), and nonpolar water-diluted solvents
(20% acetonitrile, 10–100% ethanol). For high-affinity interac-
tions, the used of a cocktail of different regeneration solutions
may be necessary [14].
8. Antibody concentration has a strong effect on the sensitivity
of the immunoassay and must be carefully optimized. In
competitive immunoassays it is an advantage to use a low
184 Sonia Herranz et al.

concentration of the antibody since by decreasing antibody

concentration the effect of the competitor (target analyte) is
magnified, usually leading to a lower limit of detection (higher
sensitivity). However, extremely low concentrations of anti-
body give a poor signal-to-noise ratio. Thus, a concentration
leading to a response around 70% of that at the saturation limit
is usually selected.
9. For fitting purposes, several data analysis and graphing software
products such as SigmaPlot, OriginPro, or GraphPad Prism
can be use.
10. 4-PL model gives rise to symmetric curves about the c point. If
the experimental data rise to asymmetric curve, more complex
functions, such as a five parameter logistic model, must be
adopted [15, 16].

Acknowledgments

The nanoB2A is a consolidated research group (Grup de Recerca)

of the Generalitat de Catalunya and has support from the Departa-
ment d’Universitats, Recerca i Societat de la Informació de la
Generalitat de Catalunya (2014 SGR 624). ICN2 is the recipient
of Grant SEV-2013-0295 from the “Severo Ochoa Centers of
Excellence” Program of Spanish MINECO. The authors acknowl-
edge to the European Union (BRAAVOO Grant Agreement No
614010). The authors thank Dr. Parlow (NIDDK, CA, USA), Dr.
Rodrı́guez-Frade and Dr. M. Mellado (CNB-CSIC, Madrid,
Spain), and Prof. Marco (IQAC-CSIC, Barcelona, Spain) for the
supply of the inmunoreagents.

References

1. Hunsperguer RG (2009) Integrated optics: the- monoclonal antibodies specific for the human
ory and technology. Springer, New York, NY growth hormone 22 K and 20 K isoforms. J
2. Nishihara H, Haruna M, Suhara T (1989) Clin Endocrinol Metab 81:1613–1618
Optical integrated circuits, McGraw-Hill Opti- 6. Ballesteros, B., Barceló, D., Sanchez-Baeza, F.,
cal and Electro-optical Engineering. Series, Camps, F. and Marco, M.P (1998) Influence of
London the Hapten Design on the Development of a
3. Kozma P, Kehl F, Ehrentreich-Förster E, Competitive ELISA for the Determination of
Stamm C, Bier FF (2014) Integrated planar the Antifouling Agent Irgarol 1051 at Trace
optical waveguide interferometer biosensors: Levels. Anal Chem 70, 4004–4014.
A comparative review. Biosens Bioelectron 7. Ali, H.R, Arifin, M.M., Sheikh, M.A., Shazili,
58:287–307 N.A.M. and Bachok, Z. (2013) Occurrence
4. Zinoviev KE, González-Guerrero AB, Domı́n- and distribution of antifouling biocide
guez C, Lechuga LM (2011) Integrated Irgarol-1051 in coastal waters of Peninsular
bimodal waveguide interferometric biosensor Malaysia. Mar Pollut Bull 70, 253–257.
for label-free analysis. J Light Technol 8. Martı́nez K, Ferrer I, Hernando MD, Fernán-
29:1926–1930 dez-Alba AR, Marcé RM, Borrull F, Barceló D
5. Mellado M, Rodriguez-Frade JM, Kremer L, (2001) Occurrence of Antifouling Biocides in
Martinez-Alonso C (1996) Characterization of
 Label-Free Biosensors Based on Bimodal Waveguide (BiMW) Interferometers 185

the Spanish Mediterranean Marine Environ- antifouling Irgarol 1051 and the fungicides
ment. Environ Technol 22:543–552 dichlofluanid and 4-chloro-3-methylphenol
9. Prieto F, Sepúlveda B, Calle A, Llobera A, in water samples. J Chromatogr A 839:
Domı́nguez C, Abad A, Montoya A, Lechuga 253–260
LM (2003) An integrated optical interfero- 13. Dunn J, Wild D (2013) Calibration curve
metric nanodevice based on silicon technology fitting. In: Wild D (ed) The immunoassay
for biosensor applications. Nanotechnology handbook, 4th edn. Elsevier, Oxford, pp
14:907–912 323–336
10. Dante S, Duval D, Fariña D, González-Guer- 14. Fisher MJE (2010) Amine coupling through
rero AB, Lechuga LM (2015) Linear readout EDC/NHS: a practical approach. In: Mol NJ,
of integrated interferometric biosensors using a Fischer MJE (eds) Surface Plasmon Resonance
periodic wavelength modulation. Laser Photon Methods and Protocols, 1st edn. Humana
Rev 9:248–255 Press, Springer New York, pp 55–73
11. Gandhiraman RP, Gubala V, Nam LCH, 15. Gottschalk PG, Dunn JR (2005) The five-
Volcke C, Doyle C, James B, Daniels S, parameter logistic: a characterization and com-
Williams DE (2010) Deposition of chemically parison with the four-parameter logistic. Anal
reactive and repellent sites on biosensor chips Bioanal 343:54–65
for reduced non-specific binding. Colloids Sur- 16. Giraldo J, Vivas NM, Vila E, Badia A (2002)
faces B 79:270–275 Assessing the (a)symmetry of concentration-
12. Peñalver A, Pocurull E, Borrull F, Marcé RM effect curves: empirical versus mechanistic
(1999) Solid-phase microextraction of the models. Pharmacol Ther 95:21–45
 Chapter 12

DNA-Directed Antibody Immobilization for Robust Protein

Microarrays: Application to Single Particle Detection
‘DNA-Directed Antibody Immobilization
€ €
Nese Lortlar Unl u*, Fulya Ekiz Kanik*, Elif Seymour, John H. Connor,
€
and M. Selim Unl€ u

Abstract
Protein microarrays are emerging tools which have become very powerful in multiplexed detection
technologies. A variety of proteins can be immobilized on a sensor chip allowing for multiplexed diagnos-
tics. Therefore, various types of analyte in a small volume of sample can be detected simultaneously. Protein
immobilization is a crucial step for creating a robust and sensitive protein microarray-based detection
system. In order to achieve a successful protein immobilization and preserve the activity of the proteins after
immobilization, DNA-directed immobilization is a promising technique. Here, we present the design and
the use of DNA-directed immobilized (DDI) antibodies in fabrication of robust protein microarrays. We
focus on application of protein microarrays for capturing and detecting nanoparticles such as intact viruses.
Experimental results on Single-particle interferometric reflectance imaging sensor (SP-IRIS) are used to
validate the advantages of the DDI method.

Key words Protein microarrays, DNA-directed antibody immobilization, Label free detection, Single
particle detection, SP-IRIS

1 Introduction

Protein microarrays have become indispensable tools with a wide

variety of applications including biomarker detection, protein-
protein interaction analysis, and detection of small molecule targets
as in drug discovery [1, 2]. They are also promising candidates
for multiplexed clinical diagnostics and disease progression moni-
toring [3]. Here, we first describe the challenges of protein micro-
arrays to motivate DNA-directed antibody immobilization (DDI)
for production of robust protein microarrays. For characterization

*
These authors contributed equally to this work.

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_12,
© Springer Science+Business Media LLC 2017

187
188 € u
Nese Lortlar Unl € et al.

Fig. 1 Conceptual representation of IRIS detection platform for digital detection of individual viral particles and
antigens labeled with gold nanoparticles

of the protein microarrays, we use Interferometric Reflectance

Imaging Sensor (IRIS) technology developed in our laboratory.
The IRIS platform has two distinct operation modalities [4]: (1)
high-throughput label-free measurement of biomass accumulation;
and (2) digital detection of single particles with high-magnification,
also known as Single Particle Interferometric Reflectance Imaging
Sensor (SP-IRIS). First modality is utilized for label-free character-
ization of probe immobilization and the latter is used for compari-
son of protein chips for single particle (virus) detection
applications. Figure 1 shows the detection of single virus particles
and single molecule detection of proteins and DNA/RNA, con-
ceptually. With SP-IRIS, particles, which are too small to be
detected by the standard microscopy, such as viruses, can be easily
detected, quantified and differentiated according to their size and
shape. Moreover, it becomes possible to detect individual mole-
cules, such as proteins and DNA, via labeling with small metallic
nanoparticles with high sensitivity [4].
The high-throughput protein microarrays are based on the
technology developed for DNA chips that made profound impact
in genomic analysis. Fabrication of DNA chips benefit from the ease
of uniform immobilization of DNA probes on the sensor surface.
DNA probes of different sequences can be chemically functiona-
lized at a particular end (e.g. amine, thiol, or carboxyl functionali-
zation) or have identical regions in the proximity of the solid sensor
surface to assure uniformity. In contrast, proteins represent a great
diversity of size and conformation resulting in significant variability
of surface immobilization. Therefore, despite many advantages of
protein microarrays in proteomics and multiplexed diagnostics
applications, the technical difficulties associated with fabrication
DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 189

Fig. 2 Variation of antibody immobilization for different types of antibodies. In

this particular example, we compare 13F6 (anti-EBOV glycoprotein antibody),
13C6 (anti-EBOV glycoprotein antibody) and 8G5 (anti-wild type VSV glycoprotein
antibody) were used. Error bars show the variation among 60 different spots in
ten different spotting runs

and reliability of the protein chips limit the potential of this

promising technology. Perhaps the most fundamental challenge
to overcome is related to the immobilization of proteins on the
microarray surface in a repeatable manner. Furthermore, how the
proteins are bound on the surface as well as the orientation of the
protein probes significantly affects their capture efficiency. For
example, on an amine-reactive sensor surface, the protein probes
may have multiple covalent interactions that limit their accessibility
to bind to a target molecule [5]. Moreover, challenges such as
variability of spot properties and protein immobilization within
and across chips influence the accuracy and robustness of protein
microarrays. As illustrated in Fig. 2, when different types of anti-
bodies are spotted, a significant difference in spot heights (as char-
acterized by IRIS), thus a variation in antibody surface densities, is
observed.
In this chapter, we describe DNA-directed antibody immobili-
zation (DDI) for robust protein microarrays specifically for applica-
tions in single particle (or digital) detection - an exciting recent
technological development that provides sensitivity beyond the
reach of ensemble measurements [6–8] but brings along additional
challenges. We have demonstrated an optical imaging technique
termed Single-Particle Interferometric Reflectance Imaging Sensor
(SP-IRIS)—a versatile platform that allows for a large range of
nanoparticle detection including both natural nanoparticles (such
as viruses) and synthetic nanoparticles in a highly-multiplexed
190 € u
Nese Lortlar Unl € et al.

microarray format [9]. The technology is based on interference of

light from an optically transparent multi-layer dielectric structure.
The interference of light reflected from the sensor surface is mod-
ified by the presence of particles producing a distinct signal that
reveals the presence and size of the particle that is not otherwise
visible under a conventional microscope. Size discrimination of the
imaged virions in label-free virus detection allows for rejection of
non-specifically bound particles to achieve a limit-of-detection
competitive with the state-of-the-art laboratory technologies
[10]. SP-IRIS has also shown promising results for detection of
protein [11] and DNA molecules labeled with small gold
nanoparticles—showing attomolar sensitivity and meeting the
requirements for most in vitro tests.
Using our experimental results on digital detection of viruses
via SP-IRIS, we illustrate the additional and more stringent require-
ments for robust protein microarrays. For a typical molecular detec-
tion assay, (for example when protein probes are used to capture
protein biomarker targets), the signal is expected to vary linearly
with the probe density. Therefore, assay variability due to probe
immobilization differences can be accounted for and calibrated as
described in [12]. In detection of intact viruses (or nanoparticles
much larger than individual protein molecules), the influence of
probe density is much more significant as illustrated by the experi-
mental results in Fig. 3. The amount of virus that is captured on the
antibody spots increases rapidly with the surface antibody density.
Also, virus capture nearly vanishes when probe density drops below
5
109 Ab/mm2 corresponding to less than 20% surface coverage.

Fig. 3 The dependence of virus capture on the density of antibody immobilization

on the sensor surface illustrating the strong variation
DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 191

We speculate that multiple antibody-antigen binding interactions

are required for the capture of viruses. A sparse coverage of immo-
bilized antibodies on a solid surface does not allow for multiple
interactions with the proteins on the relatively rigid surface of the
target virus resulting in significant reduction of capture efficiency.
One potential solution is to attach the antibodies to the sensor
surface with flexible tethers thus allowing them to conform to the
surface of the virus to make multiple bonds even for relatively sparse
surface coverage. This is precisely the most significant motivation
for using DNA-directed antibody immobilization in the context of
single particle (such as an intact virus) detection.
Motivated by the limitations of direct immobilization of anti-
bodies, various alternative methods have been explored for protein
immobilization to promote antibody activity and improve assay
sensitivity. DNA-directed immobilization (DDI), one of the recent
alternative methods for protein immobilization, combines both the
robustness of DNA microarrays with the diagnostic utility of pro-
teins through the use of protein-DNA conjugates to functionalize a
DNA surface for subsequent antigen capture [13–15]. Previous
studies demonstrated that DDI enhanced molecular (antigen) cap-
ture efficiency, refined spot homogeneity and improved assay repro-
ducibility contrasted to direct covalent immobilization of
antibodies on the solid surface [16–18]. To validate the improved
capture efficiency in the context of single particle detection, we
have compared DDI to direct antibody immobilization [19]. A
systematic study demonstrated that antibodies attached to the sen-
sor surface with flexible tethers (DNA) provide additional confor-
mational freedom, elevate the capture probes from the solid sensor
surface and increases the target capture efficiency. Therefore, DDI
not only provides a more robust protein microarray fabrication
process directly benefiting from well-established DNA chips but
also enhances sensitivity. Additional advantages of DNA-directed
antibody immobilization include the ability to reprogram the sen-
sor surface by using a different set of antibodies conjugated to the
same DNA sequences and the resilience of DNA microarrays in
elevated temperatures required for assembly of microfluidic car-
tridges [19]. In DDI, selected antibodies are covalently attached
to specific DNA sequences which are complementary to surface
probe ssDNA sequences. The surface probe sequences are immo-
bilized onto the sensor surface and antibody-attached DNA
sequences hybridize to the surface probes on the surface [19] as
represented in Fig. 4.
Below, we describe the detailed method for DNA-directed
immobilization of antibodies and application to single virus detec-
tion. We illustrate the advantages using experimental results on
SP-IRIS platform.
192 € u
Nese Lortlar Unl € et al.

Fig. 4 A schematic representation of SP-IRIS substrate surface with ssDNA spots and the conversion of the
DNA chip into a triplex antibody array through DDI of antibodies. Reprinted with permission from Seymour
et al. Anal. Chem., 2015, 87 (20), pp. 10505–10,512. Copyright 2016 American Chemical Society

2 Materials

2.1 Single-Particle 1. Scientific CCD camera. Retiga 4000R (Qimaging, Corp., Sur-
Interferometric rey, BC, Canada).
Reflectance Imaging 2. Sample illumination source. 530 nm LED source (Thorlabs
Sensor (SP-IRIS) Inc., Newton, NJ, USA).
(Validation Platform: 3. Microscope objectives. 50
0.8 NA Nikon objective for dry
Can Be Replaced by samples or a 40
0.9 NA Nikon objective for samples in liquid
Another Single Particle (Nikon Inc., Melville, NY, USA) (see Note 1).
Detection Method)
4. CRISP (The Continuous Reflection Interface Sampling and
Positioning) autofocus system. MFC-2000 (Applied Scientific
Instrumentation, Eugene, OR, USA).
5. 50:50 Beam splitter (Thorlabs Inc., Newton, NJ, USA).
6. Lenses (Thorlabs Inc., Newton, NJ, USA). Achromatic doub-
lets for the visible spectrum. Catalogue numbers: f1: AC254-
030-A-ML, f2: AC254-050-A-ML and f3: AC254-200-A-ML.
7. Diffuser (Thorlabs Inc., Newton, NJ, USA). N-BK7 Ground
Glass Diffuser, 220 Grit.
8. XY stage (Micronix USA, Santa Ana, CA, USA).
9. Mechanical components for optical setup. Cage system and
lens holders (Thorlabs Inc., Newton, NJ, USA).
10. Computer. Data acquisition and processing.

2.2 Substrate 1. Silicon (Si/SiO2) chips with a patterned thermally grown sili-
Fabrication con dioxide for single particle detection experiments (Silicon
Valley Microelectronics, Santa Clara, CA, USA) (see Note 2).
2. Dimensions of the chips: 10 mm
10 mm
0.5 mm, with a
2.3 mm
2.3 mm active center region for spotting DNA probes.

2.3 Surface 1. Copoly(N,N-dimethylacrylamide (DMA)-acryloyloxysuccini-

Chemistry mide (NAS)-3-(trimethoxysilyl)-propylmethacrylate (MAPS))
(Copoly(DMA-NAS-MAPS)) (MCP-2) (Lucidant Polymers
Inc., Sunnyvale, CA, USA) (see Note 3).
2. Solution A2. Coat-On™ Coating buffer (Lucidant Polymers
Inc., Sunnyvale, CA, USA).
 DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 193

3. Acetone and isopropanol.

4. Plastic petri dish.
5. Plasma asher.
6. Vacuum drying oven.

2.4 Protein 1. HPLC purified 50 -aminated ssDNA molecules (Integrated

Microarray DNA Technologies, Inc., Coralville, IA, USA). A surface
Preparation probe and a linker (for antibody conjugation) sequences.
2. Monoclonal antibody against Ebola virus (EBOV) glyco-
protein as a model antibody (13F6) (Mapp Biopharmaceutical
Inc., San Diego, CA, USA).
3. DNA spotting buffer (2
): 300 mM sodium phosphate buffer,
pH 8.5. 42.6 g of Na2HPO4 and 7.4 g of NaH2PO4 is dis-
solved in 1 L of filtered deionized water (18.2 MΩ resistivity,
0.2 μm-filtered) (Barnstead, NANOpure Diamond water puri-
fication system). The pH is adjusted to 8.5 with HCl. Stored at
room temperature.
4. Tris–HCl buffer (1
): 50 mM Tris–HCl with 150 mM NaCl,
pH 8.5. 6.06 g of Tris is dissolved in 800 mL filtered deionized
water. pH is adjusted to 8.5 with the appropriate volume of
concentrated HCl. Final volume is brought to 1 L with deio-
nized water. Stored at room temperature.
5. Blocking solution: 50 mM ethanolamine in Tris–HCl buffer.
310 μL of ethanolamine is added to 100 mL of Tris–HCl buffer
(see Note 4).
6. Wash buffer (PBST): PBS buffer with 0.1% Tween-20. Stored
at room temperature.
7. Sodium nitrate solution: 0.5 M sodium nitrate. 43 g of NaNO3
is dissolved in 1 L of water.
8. Antibody-DNA conjugation kit (Thunder-Link Oligo Conju-
gation Kit, Innova Bioscience, Cambridge, UK).
9. Piezoelectric microarray spotter. Scienion S3 Fleaxarrayer (Ber-
lin, Germany).
10. 24-well plate.
11. Multipurpose rotating shaker.
12. For comparison purpose only, antibodies are immobilized
directly (see Note 5).

2.5 In-Liquid Virus 1. Microfluidic cartridges. Multilayer polymer laminate devices:

Detection (Selected for Disposable, custom-designed (ALine, Inc., Rancho Domin-
Validation guez, CA, USA). From the bottom pf the costom-designed
Experiments: Other chip, Layer 1: acrylic, Layer 2: Si-PSA, Layer 3: polycarbonate,
Techniques Can Be Layer 4: PET and Layer 5: PET.
Substituted)
194 € u
Nese Lortlar Unl € et al.

Fig. 5 The images of an SP-IRIS chip and microfluidic cartridge for in-liquid SP-IRIS measurements (on the
left). Cross section model of the microfluidic cartridge and objective which indicates the fluidic path (in blue),
the sensor (in gray) and the chip and imaging window (in yellow) not to scale (on the right). Adapted with
permission from Sherr et al. ACS Nano, 2016, 10 (2), pp. 2827–2833. Copyright 2016 American Chemical
Society

2. Syringe pump. PHD 2000 (Harvard Apparatus, Holliston,

MA, USA).
3. Silicon Tubing. FEP Nat 1/16
0.010
20 ft (Upchurch
Scientific, IDEX Health and Science, Middleborough, MA,
USA).
4. Adaptors, syringes, fitting.
5. Figure 5 shows an SP-IRIS chip, a microfluidic cartridge and a
schematic representation of the imaging setup.

2.6 Data Processing 1. MATLAB (MathWorks, Natick, MA, USA).

2. Custom particle detection software developed in MATLAB.

3 Methods

3.1 DNA Sequence 1. DNA sequences are designed to minimize the hairpin, self-
Design dimer and heterodimer structures to increase the hybridization
efficiency and prevent cross hybridization (see Note 6).
2. Twenty base pair of DNA sequence used as the surface probe
(40 mer) is complementary to the antibody conjugated
sequence used in the antibody-conjugation.
3. Amine modification is introduced at the 50 -end of the surface
probe DNA sequence in order to achieve covalent immobiliza-
tion on the copoly(DMA-NAS-MAPS) polymer coated chip
surface. Table 1 shows the DNA sequences used in this work
for antibody immobilization.
4. Lyophilized oligonucleotides are dissolved in ultrapure water
to have a final concentration of 100 μM. 100 μL aliquots
 DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 195

Table 1
Model surface probe DNA sequence and its corresponding sequence used in antibody-conjugation
in this work

DNA sequences conjugated

Antibody to the antibodies DNA sequences immobilized on the chip surface
Anti-EBOV B: 5 AAAAATACAGAGTTA B0 : 50 ATCCGACCT
0

GP (13F6) GTCGCAGTGG30 TGACATCTCTACCACTGCGACTAACTCTGTA30

Complementary sequences are underlined. This particular sequences are selected as an example to be used in DNA-
directed immobilization

are made. The concentration of the DNA solution is deter-

mined spectrophotometrically by measuring the absorbance at
260 nm. The aliquots are stored at 20  C. When needed, a
single aliquot is thawed, diluted to desired concentration and
used immediately.

3.2 Protein 1. The silicon chips are functionalized with the polymer copoly
Microarray Fabrication (DMA-NAS-MAPS). Prior to polymer coating, the chips are
cleaned by sonicating in acetone for 5 min, rinsing in isopro-
panol and deionized water. Then, the chips are dried by blow-
ing nitrogen gas.
2. The chips are cleaned by plasma ashing with oxygen plasma at
300 sccm and 500 W for 10 min.
3. 1% (w/v) copoly(DMA-NAS-MAPS) is dissolved in a mixture
of Solution A2 and water (1:1) by vigorous vortexing.
4. The clean chips are placed in a plastic petri dish and the polymer
solution is poured onto the chips until they are fully covered
with the solution (see Note 7).
5. The chips are incubated in the polymer solution for 30 min at
room temperature while shaking.
6. After functionalization, the chips are rinsed and dipped in
deionized water for 3 min with shaking. This is repeated for
three times. Then, the chips are dried with nitrogen gas.
7. In a vacuum oven, the chips are baked for 15 min at 80  C (see
Note 8).
8. To prepare protein microarrays via DNA-directed immobiliza-
tion technique, surface probe oligonucleotide (Sequence B0 ) is
spotted on the polymer-functionalized chips at an optimum
concentration of 30 μM in sodium phosphate buffer (150 mM,
pH 8.5). In order to determine the effect of probe density, a
varying degree of immobilization for antibody-conjugate was
created and virus capture capacity was determined by spotting
DNA surface probe at different concentrations which are 3, 6,
196 € u
Nese Lortlar Unl € et al.

12, 18, 24 and 30 μM. Each concentration is spotted in six

replicates on the chip (see Note 9).
9. An automated piezoelectric arrayer is used to spot surface
probe DNA. A microarray pattern is designed for six replicate
spots for the probe DNA. The space between the spots is set as
288 μm.
10. The humidity and the temperature inside the spotter chamber
are set to 67% and 20  C, respectively (see Note 10).
11. After spotting, DNA-spotted chips are kept in the spotter
chamber at 67% humidity overnight at room temperature.
12. Following the immobilization, the chips are incubated in
blocking solution for 30 min to quench the excessive NHS
groups left after immobilization.
13. Then, the chips are washed with wash buffer for 30 min while
shaking.
14. Afterwards, the chips are rinsed with phosphate buffer saline
and deionized water, and dried with nitrogen.
15. Antibody-DNA conjugates are synthesized using Thunder-
Link oligo conjugation kit according to manufacturer’s proto-
col. 40 μM 50 -aminated linker DNA (Sequence B) is reacted
with 1 mg/mL antibody (13F6) (see Note 11).
16. Antibody immobilization is achieved by incubating the DNA-
chips with the antibody-DNA conjugate solution (at 5 μg/ml
in PBS with 1% BSA) for 1 h in a 24-well plate on a shaker.
17. After DNA-DNA hybridization and antibody immobilization,
the chips are washed with PBS twice, 0.5 M sodium nitrate
solution twice, and then, dipped in cold 0.1 M sodium nitrate
solution. Finally, the chips are dried with nitrogen gas.
18. For comparison purpose only, antibodies are immobilized
directly (see Notes 12 and 13).

3.3 Quality Control of 1. Biomass quantification is performed for direct antibody-

the Modified Chips immobilized and DNA directed antibody-immobilized chips
using IRIS (see Note 14).
2. To test the activity of DNA conjugate for binding to its target,
biomass measurements of the conjugate hybridization and
virus binding for both directly immobilized and DNA-tethered
antibody are performed. The spots are imaged before
antibody-conjugate hybridization, after hybridization (thus,
antibody immobilization) and after virus (VSV) capture,
shown in Fig. 6. After the immobilization and virus capture,
the spot height change is identified with IRIS measurements
concluding that although DNA-spots have much less probe
density ( 1=3 of antibody-spot density), they capture almost the
 DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 197

Fig. 6 IRIS average spot height measurements for direct-antibody and DNA-directed immobilized-antibody
spots. One nanometer of surface thickness corresponds to 1.2 ng/mm2 of antibody density. (a) Schematic
representation of the IRIS chip spotted with three different probe types: antibody, probe DNA and a negative
DNA sequence (not complementary to target antibody-conjugated DNA). (b) IRIS images of the chip at different
stages of the experiment: before hybridization of DNA B-conjugate, after hybridization of the conjugate and
after incubating with VSV virus. (c) IRIS surface height measurements for three different stages of the
experiment

same amount of virus with the direct-antibody spots. This

shows the significant improvement in the capture efficiency
due to DDI for protein immobilization. The height change in
the antibody spots shows that they are only able to capture a
small amount of virus even though they are spotted at the
highest surface density. This shows that even though the pro-
tein immobilization on the surface is achieved successfully,
functionality of the protein for virus detection after immobili-
zation is questionable. Moreover, it is also seen from the figure
198 € u
Nese Lortlar Unl € et al.

that variability in DNA-tethered protein immobilized spots is

very small. Hence, DDI is found to be an optimum technique
for obtaining robust and reproducible protein microarrays.
These results also suggest that DDI technique can be easily
adapted to immobilization of various proteins without a need
for further optimizations (see Note 15).

3.4 SP-IRIS 1. SP-IRIS optical setup is based on a top-illumination micro-

scope configuration. To create the optical path with the illumi-
nating LED, 2-lens system is used and the LED is imaged onto
the back aperture of the objective for Kohler illumination. LED
light is first collimated with 30 mm-lens in the illumination
path. Then, the collimated light gets focused onto the back
focal aperture of the objective using a 50 mm-lens. Hence, the
light beam is directed onto the sample stage which, then, gets
reflected off the SP-IRIS sensor surface. With a beam splitter,
the light beam is focused; therefore, it is directed down to the
sample stage and the reflected light from the sensor surface is
captured by the camera (see Notes 16 and 17).
2. Via the microscope objective, the light beam is focused.
3. After, the reflected light beam is collected by the objective and
transmitted through the beam splitter.
4. The specular reflected light as well as the scatter light from the
sample are focused via the tube lens onto the CCD camera
where intensity image is recorded. The schematic representa-
tion of the system, light path and imaging principle are shown

Fig. 7 (a) Schematic representation of the SP-IRIS setup. LED is used for illumination and bright field image is
reflected on a CCD camera. (b) With the help of cross term (highlighted) in the equation, visibility of low-index
nanoparticles is enabled with their enhanced intensity by order of magnitude compared to scattered intensity
(middle term). (c) A representation of multiplexed SP-IRIS chip (bottom) and an actual image of detected
particles on the chip after processing (top). Reprinted with permission from Avci et al. Sensors, 2015, 15,
pp 17649–17665. 2016 Creative Commons (http://creativecommons.org/licenses/by/4.0/)
 DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 199

Fig. 8 Representative image of the SP-IRIS setup

in Fig. 7. Figure 8 shows a representative image of the setup

and its configuration.
5. For imaging, the microfluidic device is placed on the sample
holder (stage) and endpoint images are taken.

3.5 Data Processing 1. Each spot is imaged in SP-IRIS.

2. The images for each spot are analyzed for bound virus particles.
3. By subtracting pre-incubation particle counts from post-
incubation counts and dividing the result by the analyzed spot
area, the net particle densities are determined (see Note 18).
4. Six spots for each condition are used to average the particle
densities.
5. The morphological features as well as non-diffraction limited
peaks are removed from the image with appropriate filters.
6. The signal coming from the particle is based on the interfer-
ence between the scattered field from the particle and the
reference field arising from the reflection off the sensor. The
signal from the particle is normalized relative to the nearby
background intensity. Custom particle detection software is
developed in MATLAB to find particle-originated local inten-
sity maxima which fall within the point spread function of the
optical system (see Note 19).
7. The interaction of the light with small particle is considered as
an induced dipole where its strength is proportional with the
200 € u
Nese Lortlar Unl € et al.

polarizability of the particle (α), as given in the following

equation:

ɛp  ɛm
α ¼ 4πɛ0 r 3 ð1Þ
ɛp þ 2ɛm
where r is radius of the particle, ɛp and ɛm are particle and
surrounding medium permittivity, respectively [20]. In interfero-
metric imaging techniques, in the intensity recorded at the detec-
tor, strong reference field and weak scattered fields coming from
the particle are mixed and can be expressed with the following
equation:

I / jE ref j2 þ jE sca j2 þ 2jE ref jjE sca j cos θ ð2Þ

In this equation, for weakly scattering particles, |Esca|2 is negli-

gible. Hence, the signal is dominated by the cross term as seen in
Fig. 7b. This signal appears as a dot in the image as seen in Fig. 7c.
Based upon the contrast of the particle, the size is determined with
the model. Since the virus particles scatter the light upon binding
the surface, enhanced signal coming from the layered chip surface is
detected on the CCD camera.

3.6 Single Particle 1. The flow cell is obtained by assembling the individual layers.
Detection (In Liquid The acrylic layer is placed as the base and then, biosensor chip is
Virus Detection put in the middle of the cartridge. After placing the second
Measurements) layer, the third layer is placed on top providing imaging win-
dow like a polymer cover glass. Then, the fourth and fifth layers
are placed and adaptors are attached to the fifth layer. FEP
tubing is assembled to the adaptors. The syringe pump is
connected via tubing to provide the flow. The flow is directed
in the channels to the chip surface with the precisely designed
microfluidic device as shown in Fig. 5.
2. The flow cell is placed on the sample stage of SP-IRIS under-
neath the objective for image acquisition. DNA-immobilized
SP-IRIS chips are already placed in the microfluidic cartridges.
3. Acquisition software is opened and the camera is turned on.
4. Number of images to be averaged is determined and selected.
Maximum exposure time before saturation of the camera is
selected.
5. First, 13F6-DNA, B-conjugate is flowed at a concentration of
5 μg/mL in PBS with 1% BSA for 1 h to achieve protein
immobilization on the chip.
6. To get pre-incubation particle counts in spots with different
concentration of protein immobilized, SP-IRIS images are
acquired.
 DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 201

Fig. 9 (a) Left IRIS image showing different surface densities of immobilized antibody-DNA conjugate after
hybridization of the B0 sequence with the 13F6-DNA “B” conjugate. Right image shows 13F6-antibody spots at
different concentrations. (b) Effect of surface antibody density (1 nm ¼ 1.2 ng/mm2) on virus capture
efficiency for both direct antibody and DNA-conjugated antibody immobilization. Average capture virus
densities (n ¼ 6 spots) obtained from SP-IRIS images versus average optical thickness of the antibody
spots obtained from IRIS images were plotted. The ellipses indicate the range of surface antibody heights
where the optimal virus capture occurs. Reprinted with permission from Seymour et al. Anal. Chem., 2015, 87
(20), pp. 10505–10512. Copyright 2016 American Chemical Society

7. Then, the chips are incubated with 104 PFU/mL EBOV-VSV

by flowing it over the chip for 30 min.
8. After incubation, 400 μL sodium phosphate buffer is flowed
over the microchannel and SP-IRIS images are acquired (see
Note 20).
202 € u
Nese Lortlar Unl € et al.

Fig. 10 Comparing captured virus densities on direct-spotted and DNA-tethered antibodies from independently
performed experiments. Each letter represents a data point next to it. Experimental conditions corresponding
to A–G experiments are summarized in the table. Error bars show the variation between six spots for a given
experiment. Reprinted with permission from Seymour et al. Anal. Chem., 2015, 87(20), pp 10505–10512.
Copyright 2016 American Chemical Society

9. Finally, end point images of virus-captured chips are acquired.

10. Figure 9 shows the comparison of the virus capture perfor-
mance of both direct antibody-immobilized and DNA-
directed-antibody immobilized assays. IRIS is used to show
the surface morphologies of antibody and DNA-tethered anti-
body spots at different concentrations. According to Fig. 9b,
DNA-conjugated 13F6 performs optimally over a larger range
of surface densities compared to the directly spotted antibody
[19].
11. Figure 10 depicts the performance of the protein immobiliza-
tion technique by varying the capture conditions. In general,
since limited number of target is present at lower concentra-
tions of the virus, a lower capture efficiency is obtained in both
direct and DDI techniques. However, when the available tar-
gets increase in the solution, a drastic increase in DNA-
tethered antibody microarray is observed; whereas, direct-
antibody immobilized microarray does not show the same
sensitivity. DDI makes the capture proteins highly available
and accessible on the sensor surface. Hence, an improved
binding capacity is achieved in shorter time.

4 Notes

1. Adjusting the NA offers a tradeoff between nanomaterial con-

trast and usable field of view (FOV).
DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 203

2. Oxide thicknesses are chosen to maximize the phase angle

term.
3. It can be synthesized as described in the reference [21]. One of
the most affecting components of protein microarrays is micro-
array immobilization surface. For DDI technique, specifically a
reactive-functional surface is needed in order to achieve the
covalent attachment of the surface probe DNA sequences to
the sensor surface. Therefore, MCP-2 is an excellent option for
this purpose. It bears functional NHS groups which take part in
covalent immobilization of the amino-modified oligonucleo-
tides. Moreover, due to its unique 3D structure, it provides
right confirmation for oligonucleotides by enhancing immobi-
lization with its hydrophobic regions. Also, MCP-2 lowers
non-specific bindings of proteins providing a perfect sensor
surface for sensitive and selective detections [22].
4. NHS groups are active for further possible reactions. There-
fore, it is important to block them after the completion of
immobilization. Otherwise, it may cause further crosslinking;
thus, a decrease in the efficiency in protein immobilization due
to the less availability of the probes or proteins.
5. Antibody spots directly immobilized for comparison purpose
only require antibody spotting buffer (1
): 1
PBS with a final
concentration of 25 mM trehalose, pH 7.5. Ready-made 1

PBS (Thermo Fisher, GIBCO PBS 1
). Trehalose stock solu-
tion is prepared at 0.25 M. Threhalose is added to the antibody
solution (which is in PBS) just before spotting. After diluting
antibody solution, trehalose is added to give a final concetra-
tion of 25 mM. 8.6 g of trehalose is dissolved in 10 mL of
filtered deionized water to have 0.25 M trehalose stock
solution.
6. 50 -polyA sequence is added as a spacer to DNA sequence which
is conjugated to the antibody.
7. It is important to use a plastic petri dish in the polymer coating
process of the chips. Since a glass petri dish competes with the
chip in functionalization, for higher coating efficiency, a plastic
petri dish is used.
8. It is critical to store the functionalized chips in a desiccator after
fabrication. The chips can be stored in the desiccator for up to
3 weeks under vacuum. It would be worthwhile to bake the
chips at 80  C for 15 min in a vacuum oven prior to use in order
to have completely dried and NHS-active polymer coating on
the chips for successive spotting.
9. Surface antibody density is an important criterion which dras-
tically affects assay sensitivity. Surface density of the antibodies
should be optimized for a given application to optimize the
target capture.
204 € u
Nese Lortlar Unl € et al.

10. During spotting of DNA, it is very important to keep the

humidity at an optimum level in the spotter chamber. The
optimal level we found is 67% for the humidity. After spotting,
the chips are kept in the spotter chamber at 67% humidity
overnight. However, for antibody spotting, optimum spotting
is found as 57–59%. These changes in the humidity make huge
variations in the spot morphology. If DNA is spotted at the
same humidity with antibody, then, a smaller and non-uniform
spot morphology is observed.
11. DNA concentration is optimized to obtain 1–2 DNA
sequences per antibody. After conjugation, yield of reaction is
determined by Bradford assay. The monoclonal antibody is
conjugated to DNA sequence, B0 , in 1:1.5 ratio. 13F6-DNA,
B, is antibody-conjugates against Ebola-GP pseudotyped VSV.
12. Different antibodies have different probe characteristics.
Therefore, in order to obtain a uniform spot morphology and
surface density, for each type of protein, an optimization study
is needed. This is not time and cost efficient. On the other
hand, oligonucleotides only differ in sequences and length
among each other; therefore, as long as the length of the
sequence is kept constant, DNA immobilization characteristics
generally stay the same. DNA spotting does not depend on the
type or immobilization. Hence, a standard and parameter-
independent protein immobilization can be achieved with
DDI. Thereupon, this technique is preferred in this study.
13. Antibody spots directly immobilized for comparison purpose
only. (a) Antibody solutions are prepared as 3 mg/mL in
antibody spotting buffer. The humidity during spotting is
kept between 57 and 59%. The spotted chips are kept in the
spotter chamber at 67% humidity overnight. (b) To evaluate
the effect of antibody immobilization density on the virus
capture, anti-EBOV GP antibody (13F6) is spotted at varying
concentrations on SP-IRIS substrates (5, 4, 3, 2 ,1 and
0.5 mg/mL).
14. IRIS has a different modality of operation. The important
difference between IRIS and SP-IRIS is the magnification of
the optical system. In IRIS, numerical aperture of the objective
lens is different. Hence, low-magnification can be achieved;
whereas, using SP-IRIS, high-magnification is achieved. This
modality is based on spectral reflectivity. When the biomass
accumulates on the sensor surface, the thickness on the sensor
increases. This increase generates a quantifiable change in the
spectral reflectivity depending on the optical path difference
(OPD) between substrate surface and biomass accumulated
substrate surface. With the help of low-magnification modality,
multiplexed monitoring of biomolecule immobilization or
DNA-Directed Antibody Immobilization for Robust Protein Microarrays. . . 205

capture can be precisely acquired. Therefore, using IRIS, a

larger field of view (FOV), approximately 1 in.2, can be cap-
tured. Hence, IRIS can be used for quality control of the assay
chips [20]. Variations in the probe immobilization can be
detected easily.
15. Spot morphology is another important criterion for single
particle detection. It directly affects probe density, capture
efficiency; thus, the reproducibility in the experiments. Non-
uniform spot morphology causes spot-to-spot and assay-to-
assay changes. Capture efficiency is directly proportional with
the surface probes. A destroyed morphology causes non-
binding of the target particle; therefore, false negative results
may be obtained in quantification of the signal [23]. These
irregularities may be crystallization appearing as bright spots in
the image, coffee rings in the outer region of the spot or
aggregations causing non-uniform regions in the spot. The
detector response should not depend on the parameters. How-
ever, when antibodies are used as the surface probe, a robust
microarray design may not always be achieved due to the
problems in consistency in the spot morphology. On the
other hand, DNA spotting is always uniform. More robust
protein microarrays can be achieved via DDI. DNA spot mor-
phology stays almost identical in repetitive spotting which
provides a uniformity in the sensor response among spot-to-
spot and assay-to-assay measurements.
16. The significant parameters in the optical setup are the illumina-
tion wavelength, uniformity, NA, magnification and camera
pixel size. These parameters affect the resolution and the accu-
racy of sample sizing. In addition, since sample sizing is deter-
mined using the contrast of the central peak to near-neighbor
background pixels; improper and non-uniform illumination
cause sizing obscurity. Kohler illumination is implemented in
the optical system due to its superiority over critical
illumination.
17. To achieve a uniform illumination, a diffuser is placed in front
of the LED light source.
18. In order to minimize signal variations arising from the spot
morphology, it is desirable to take a pre-incubation image to
later subtract from the post-incubation image to calculate the
net number of bound particles.
19. Other image processing software such as ImageJ can be used as
an alternative to MATLAB for post-processing and
quantification.
20. During the microarray preparation processes, especially in
between washings, the chips should always be wetted. Extra
care should be taken.
206 € u
Nese Lortlar Unl € et al.

References
1. MacBeath G, Schreiber SL (2000) Printing pro- reversible, and site-selective surface binding of
teins as microarrays for high-throughput func- proteins by means of covalent DNA-
tion determination. Science 289:1760–1763 streptavidin conjugates. Anal Biochem
2. Sun H, Chen GYJ, Yao SQ (2013) Recent 268:54–63
advances in microarray technologies for prote- 14. Ladd J, Boozer C, Yu Q et al (2004) DNA-
omics. Chem Biol 20:685–699 directed protein immobilization on mixed self-
3. Hall DA, Tacek J, Snyder M (2007) Protein assembled monolayers via a streptavidin bridge.
microarray technology. Mech Ageing Dev 128 Langmuir 20:8090–8095
(1):161–167 15. Schroeder H, Adler M, Gergk K et al (2009)
4. Avci O, Lortlar UN, € Yalcin A et al (2015) User configurable microfluidic device for mul-
Interferometric Reflectance imaging sensor tiplexed immunoassays based on DNA-
(IRIS)-a platform technology for multiplexed directed assembly. Anal Chem 81:1275–1279
diagnostics and digital detection. Sensors 15 16. Washburn AL, Gomez J, Bailey RC (2011)
(7):17649–17665 DNA-encoding to improve performance and
5. Schwenk JM, Lindberg J, Sundberg M et al allow parallel evaluation of the binding charac-
(2007) Determination of binding specificities teristics of multiple antibodies in a surface-
in highly multiplexed bead-based antibody bound immunoassay format. Anal Chem
assays for antibody proteomics. Mol Cell Pro- 83:3572–3580
teomics 6:125–132 17. Wacker R, Niemeyer CM (2004) DNA- μFIA-a
6. Cretich M, Daaboul GG, Sola L et al (2015) readily configurable microarray-fluorescence
Digital detection of biomarkers assisted by immunoassay based on DNA-directed immo-
nanoparticles: application to diagnostics. bilization of proteins. Chem Bio Chem
Trends Biotechnol 33(6):343–351 5:453–459
7. Yurt A, Daaboul GG, Connor JH et al (2012) 18. Wacker R, Schroder H, Niemeyer CM (2004)
Single nanoparticle detectors for biological Performance of antibody microarrays fabri-
applications. Nanoscale 4(3):715–726 cated by either DNA-directed immobilization,
8. Walt D (2013) Optical methods for single mol- direct spotting, or streptavidin-biotin attach-
ecule detection and analysis. Anal Chem 85 ment: a comparative study. Anal Biochem
(3):1258–1263 330:281–287
9. Daaboul GG, Lopez CA, Chinnala J et al 19. Seymour E, Daaboul GG, Zhang X et al (2015)
(2014) Digital sensing and sizing of vesicular DNA-directed antibody immobilization for
stomatitis virus pseudotypes in complex media: enhanced detection of single viral pathogens.
a model for ebola and marburg detection. ACS Anal Chem 87(20):10505–10512
Nano 8(6):6047–6055 20. Avci O, Adato R, Yalcin OA et al (2016) Physi-
10. Monroe MR, Daaboul GG, Tuysuzoglu A et al cal modeling of interference enhanced imaging
(2013) Single nanoparticle detection for multi- and characterization of single nanoparticles.
plexed protein diagnostics with attomolar sen- Opt Express 24(6):6094–6114
sitivity in serum and unprocessed whole blood. 21. Pirri G, Damin F, Chiari M et al (2004) Char-
Anal Chem 85(7):3698–3706 acterization of the polymeric adsorbed coating
11. Sevenler D, Lortlar UN,€ € u MS (2015)
Unl€ for DNA microarray glass slides. Anal Chem 76
Nanoparticle biosensing with interferometric (4):1352–1358
reflectance imaging. In: Vestergaard MC, 22. Yalçin A, Damin F, Özkumur E et al (2009)
Kerman K, Hsing I-M, Tamiya E (eds) Nano- Direct observation of conformation of a poly-
biosensors and nanobioanalyses. Springer, meric coating with implications in microarray
Tokyo, pp 81–95 applications. Anal Chem 81(2):625–630
12. Monroe MR, Reddington A, Collins AD et al 23. Romanov V, Davido SN, Miles AR et al (2014)
(2011) Multiplexed method to calibrate and A critical comparison of protein microarray fab-
quantitate fluorescence signal for allergen- rication technologies. Analyst 139
specific IgE. Anal Chem 83(24):9485–9491 (6):1303–1326
13. Niemeyer CM, Boldt L, Ceyhan B et al (1999)
DNA-directed immobilization: efficient,
 Chapter 13

Reflectometric Interference Spectroscopy

Guenther Proll, Goran Markovic, Peter Fechner, Florian Proell,
and Guenter Gauglitz

Abstract
Reflectometry is classified in comparison to the commercialized refractometric surface plasmon resonance.
The advantages of direct optical detection depend on a sophisticated surface chemistry resulting negligible
nonspecific binding and high loading with recognition sites at the biopolymer sensitive layer of the
transducer. Elaborate details on instrumental realization and surface chemistry are discussed for optimum
application of reflectometric interference spectroscopy (RIfS). A standard protocol for a binding inhibition
assay is given. It overcomes principal problems of any direct optical detection technique.

Key words Label-free optical biosensor, Reflectometric interference spectroscopy (RIfS)

1 Introduction

The methods available for direct monitoring of biomolecular inter-

action can be divided into methods measuring changes in the
refractive index of the interaction layer, and methods measuring
changes in reflectometry at the layer [1]. Regarding the first
method, BiaCore [2] has opened the market by introducing surface
plasmon resonance (SPR) [3] as a very promising tool in biomolec-
ular interaction analysis (BIA). In contrast to refractometry, reflec-
tometry concentrates on the measurement of changes in optical
thickness. Reflectometry has been introduced many decades ago as
ellipsometry using polarized light. Interference at the interfaces of
the layer causes a change in the relative amount of amplitude of the
two polarized radiation beams and in phase. This interferometric
method has been applied to a very simple analytical method, called
reflectometric interference spectroscopy (RIfS) to be used as an
example of a very robust, simple optical detection principle in
chemical and biochemical sensing [4]. This label-free optical detec-
tion method for surface interactions is based on white light inter-
ference at transparent thin layers. At each interface of thin layers of

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_13,
© Springer Science+Business Media LLC 2017

207
208 Guenther Proll et al.

different materials with negligible absorption, radiation is partially

reflected and transmitted. If the optical path length through these
layers is less than the coherence wavelength, the different partial
beams interfere, and form an interference pattern depending on the
wavelength, the optical thickness which is given by the product of
the refractive index of the layer and its physical thickness, the
incident angle, and the refractive index of the surrounding medium
[5, 6]. In case of perpendicular incidence, a non-absorbing layer,
and low reflectances, the reflectance R is given by:
pffiffiffiffiffiffiffiffiffiffiffiffi
R ¼ R1 þ R2 þ 2 R1 R2 cos ð4π nd=λÞ ð1Þ
where R1 and R2 denote the Fresnel reflectance at the two inter-
faces, d is the physical thickness of the film, n its refractive index,
and λ the wavelength of incident light (see Note 1). A typical
interference pattern showing the modulation of reflectance with
cos(1/λ) is given in Fig. 1.
The optical thickness nd can be determined from the position
of an extremum with a given order value m by
mλ
nd ¼ : ð2Þ
2
RIfS uses the change in the optical properties in or at the top
layer of a given layer system as detection principle (Fig. 1). The
binding of an analyte molecule or particle to the sensor surface
causes a shift of the interference pattern in the wavelength domain.
To evaluate the binding signal, the locus of an extremum is tracked
over time; thus, the change of the interference spectrum results in a
time-resolved binding curve representing the binding of the analyte
molecule to the sensor surface.
A major advantage of RIfS is its resistance to changes in tem-
perature [7]. Refractometric methods such as SPR and ellipsome-
try, on the other hand, are very sensitive to temperature variations
due to the high impact of temperature on the refractive index.
Thus, temperature changes during a measurement cause negative
effects with these methods, and quick changes of temperature
between measurements are technically challenging. Since the

Fig. 1 Scheme of the RIfS detection principle. The left part shows the superimposition of the reflected light
beams and the change in optical thickness during a binding event on the sensor surface. The right part shows
the corresponding change of the characteristic interference spectrum and the resulting binding curve
 Reflectometric Interference Spectroscopy 209

refractive index n is dependent on the density given from the

Clausius Mossotti equation, and the density is dependent on the
thermal expansion, the refractive index decreases with increasing
temperature. Due to the thermal expansion of the biopolymer-
layer, the physical thickness d increases with increasing tempera-
ture. These two contrary temperature-dependent effects result in a
rather low influence of temperature on the optical thickness, the
product of the refractive index and the physical thickness.

2 Materials

Common chemicals of analytical grade are purchased from Sigma

or Merck. Milli-Q water is deionized water with a conductivity of
18.2 MΩ cm1.

2.1 Transducer The RIfS standard transducer consists of a glass substrate (D 263
glass, Schott AG, Germany) (~1 mm thick) coated with a 10 nm
layer of a material with high refractive index (usually Ta2O5 or
Nb2O5), and a top layer of SiO2 (330 nm). As a reference use a
transducer without SiO2 layer. Due to the technical specifications
of the components used for setting up a RIfS measurement system
and the investigated sample types (e.g., liquids or gases), other
transducers might be more beneficial. An ideal RIfS transducer is
designed to create the two reflected light beams of interest at
similar intensities. Therefore, it is recommended to use ray tracing
programs (e.g., FilmWizard by SCI Scientific Computing Interna-
tional or WVASE by J. A. Woollam) to simulate the expected
reflectivities for the visible wavelength range for a given RIfS
setup, taking into account the expected optical properties of the
investigated sample types.

2.2 Surface 1. GOPTS (3-glycidyloxypropyltrimethoxysilane) purum: Toxic.

Chemistry Store at room temperature, keep under Argon, sensitive to
humidity (Fluka).
2. AMD (aminodextran): MW 100 kD, level of amination 50%.
Store at 4  C under dry conditions (Innovent e.V. Technolo-
gieentwicklung Jena, Germany).
3. Dicarboxypoly- and diaminopoly(ethylene glycol) (PEG): MW
2000 Da. Store at 20  C under dry conditions (Rapp Poly-
mere, Tuebingen, Germany).
4. NHS (N-hydroxysuccinimide) purum: Store under room tem-
perature (Fluka).
5. DIC (N,N0 -diisopropylcarbodiimide) purum: Toxic. Store at
room temperature, keep under argon, sensitive to humidity
(Fluka).
210 Guenther Proll et al.

6. EMCS (6-maleimidohexanoic acid N-succinimidyl ester)

purum: Store at 20  C under dry conditions (Sigma).
7. TBTU (2-(1H-benzotriazol- 1-yl)-1,1,3,3-tetramethyluro-
nium tetrafluoroborate).
8. HOBT (1-hydroxybenzotriazole) Hydrate purum: Sore at
20  C at dry conditions (Fluka).
9. DIPEA (N,N-diisopropylethylamine): Highly inflammable.
Store at room temperature (Sigma).

2.3 Glass Substrates 1. BK 7 glass, n ¼ 151, Schott, Mainz.

2. WG 345 glass, n ¼ 1699, Schott, Mainz.
3. Interference transducer: D 263, 10 nm Ta2O5, 500 nm SiO2,
Schott, Mainz.
4. Goethe glass: multi-layer system: 1 m D 23, 45 nm, Ta2O5,
20 nm SiO2, Schott, Mainz (Cut in small squares (10 
10 mm)).

2.4 Parallel Setup 1. White light source 100 W/12 V, Osram, Munich.
2. Lenses, mirror, and positioning optics, Spindler&Hoyer,
Göttingen
3. Polymer light guides (PMAA, n ¼ 1490, coupling element
1  2 (50:50), 1 mm, fiber diameter with SMA 905 fiber
connectors, Microparts, Dortmund.
4. Optical 4--1 multiplexer DiCon VX 500-C, Laser Compo-
nents, Olching.
5. MMS diode row spectrometer with Liliput-PC, ZEISS, Jena.
6. 1900 industrial standard housing by RS Components, Walldorf-
Mörfelden.
7. Commonly Microsoft Windows XP, Vista, Windows 7, Win-
dows 8, Windows 8.1, Windows 10 operating systems.

2.5 Single Setup 1. Modified simultaneous spectrometer SPECKOL 1100, Zeiss,

Jena
2. Commonly Microsoft Windows XP, Vista, Windows 7, Win-
dows 8, Windows 8.1, Windows 10 operating systems.

2.6 Advanced Single 1. Light source providing light of a single color (White light
Setup (See Fig. 4) source 100 W/12 V, Osram, Munich with Bandpassfilter
568 nm).
2. Phodiode based detection unit (Lightwave-Multimeter 2832-
C with Si-photodiodes of the series 818 from Newport).

2.7 Liquid Handling 1. Ten position valve VICI, Valco Europa, Schenkon,
Switzerland.
 Reflectometric Interference Spectroscopy 211

2. HPCL 3-way valve VICI, Valco Europa, Schenkon,

Switzerland.
3. Six position valve, Bischoff, Leonberg.
4. Peristaltic pump Reglo-Digital MS2/8-160 ISM 832, Ismatec,
Wertheim.
5. Peristaltic pump MS Fixo, Ismatec, Wertheim.
6. High grade steel capillaries, screws and fittings from Rheodyne,
USA.

2.8 Software 1. MeasureCR for capturing spectras and controlling the systems.
Depending on the selected spectrometer, any other software
control can also be used. As most commercial available spectro-
meters are supplied with ready to use software or plug-ins for
the widely used Lab-View solution (National Instruments Cor-
poration), it is recommended to use one of these solutions.
2. IFZCR for evaluating the interferograms. Any other software
solution capable for performing the same simple mathematical
calculations like Matlab (The MathWorks, Inc.) can be used.
3. MS-Excel (Microsoft) and Origin (MicroCal Inc.) or any
equivalent statistical software packages (open source software
is available) can be used for further processing of measured
data.
4. If an advanced single color setup for reflectometry measure-
ments is used, no spectras have to be analyzed. Therefore,
software only needs to control hardware components. All eval-
uation work can be done with software packages like Matlab or
Origin.

3 Methods

3.1 Setup for RIfS In the standard laboratory setup for RIfS (Fig. 2), the white light is
guided via an optical fiber (1 mm PMMA fiber) to the back of the
transducer mounted in a microfluidic flow cell, which in turn is
attached to a liquid handling system. The reflected light is gathered
in the same waveguide.
The fiber optics used is bifurcated (50:50 ratio), with one tail
leading to the light source and the other to the UV–Vis spectrometer.
Possible light sources are halogen lamps (e.g., 10 W halogen lamp with
fiber in-coupling optics consisting of front surface spherical mirror,
collimating lens, and an infrared absorption filter) or LEDs. For
the spectral detection of the reflected light, diode array spectrometers
are used normally. A gap (approx. 100 μm) between transducer
chip and the fiber output is filled with glycerol (80%) for refractive
index matching. Samples are handled by a flow system (Fig. 3)
(e.g., two peristaltic pumps, inject-load valve, and 6-way valve).
212 Guenther Proll et al.

Fig. 2 Classical RIfS setup

Carrier P1
Flow D Waste

V1 Loading:
Injection:

Waste P2 V2

Sample
Loop

Flow Direction while loading:

while injection:

Fig. 3 Schematic drawing of the FIA system for RIfS

Moreover, this flow system can be equipped with an autosampler. The

various samples are injected by the autosampler into a sample loop.
From there, the sample is driven in continuous flow, passing the
prepared transducer.
The raw spectra are corrected for the dark current of the
spectrometer by subtraction (if necessary), and normalized to the
reflectance spectrum from a glass chip without the SiO2- and bio-
layer. The position of an interference extremum at approx. 550 nm
is determined by a parabolic fit to one halfwave of the interference
spectrum. Optical thickness is calculated according to Eq. 2 (see
Note 2).
 Reflectometric Interference Spectroscopy 213

Fig. 4 Advanced setup for single-color RIfS detection according to [15]

The costs of the setup can be reduced by sequentially illuminat-

ing the transducer with different suitable wavelengths coming from
LEDs or a white light source with appropriate filters, detecting the
intensity without spectral resolution using a photodiode, and
reconstructing part of the interference pattern by fitting a parabola
through the interpolation points. The advanced development is a
single-color (wavelength) RIfS approach, which uses a color filtered
light source (Fig. 4) or even a LED, monitoring the change in
intensity looking at a distinct wavelength rather than the shift of
an extremum (see Fig. 1). This allows a very simple instrumentation
setup and the use of a second wavelength as reference.
This detection principle can be used to realize a parallel screen-
ing system which allows optical online detection of specific biomo-
lecular interaction in 96- or 384-well microplate formats (see Fig. 5)
Reflectometric interference spectroscopy (RIfS):parallel set-up.
Therefore, the whole area of the plate bottom consisting of a RIfS
transducer is illuminated by a halogen light source combined with a
filter wheel, which allows the subsequent passage of monochro-
matic light of seven different wavelengths. This is not only to
reduce the costs, but is necessary because a CCD camera is used
as detector which is able to detect the intensity of the whole area of
interest in a single shot.
Singe-color RIfS also has an advantage in a parallelized setup. A
commercial imaging device was developed by Biametrics GmbH
that can measure up to 20,000 events per square centimeter on the
transducer under flow-through conditions while obtaining infor-
mation on thermodynamics and kinetics in parallel.
214 Guenther Proll et al.

Fig. 5 Parallel RIfS setup

3.2 Surface All surface chemistry steps are applied to the coated side of the RIfS
Chemistry transducer (see Note 3) if not stated otherwise.

3.2.1 Silanization (a) Transducer chips are cleaned with freshly prepared Piranha
solution (mixture of 30% hydrogen peroxide and concen-
trated sulfuric acid at a ratio of 2:3; caution: hot and aggres-
sive!) for 30 min in an ultrasonic bath to clean the chip and to
generate silanol groups on the transducer surface. As an
alternative, this activation step can be carried out using oxy-
gen plasma.
(b) After rinsing with double distilled water and drying in a
nitrogen stream, the surface is immediately activated by
GOPTS at a surface concentration of approx. 10 μl/cm2 for
1 h in dryness (by assembling two slides face-to-face placed in
a tray with ground joint) followed by cleaning with dry
acetone and drying in a nitrogen stream (see Notes 4 and 5).

3.2.2 Biopolymer When investigating interactions of biomolecules on surfaces, non-

specific binding to the sensor must be minimized. Therefore, the
glass slide is coated with a shielding layer which prevents nonspe-
cific binding and additionally provides a large number of functional
groups for the immobilization of the binding partner. The standard
shielding chemicals used are dextrans which form an 3D-hydrogel
loaded with a large number of binding sites, and polyethylene
glycols (PEG) forming a two-dimensional brush-like monolayer
with less binding sites but with a more defined surface. Amino
and carboxy groups are normally functional groups for the immo-
bilization process since the well-established peptide chemistry
methods can be applied. Other shielding chemicals can be used
depending on the application or the surface chemistry needed.
 Reflectometric Interference Spectroscopy 215

(a) Aminodextran.
The coupling of aminodextran as an aqueous solution (1:3)
(approx. 15 μL/cm2) to the silanized surface is carried out by
incubating over night (min 16 h) in a water-saturated atmo-
sphere (see Note 6). After thoroughly rinsing with double
distilled water and drying, the prepared chips are stable for
several months.
(b) Polyethylenglycol.
The coupling of diamino- or dicarboxypoly(ethylene glycol)
(PEG) as a 1 mM solution in dichloromethane (DCM)
(approx. 10 μL/cm2) to the silanized surface is achieved by
incubating overnight at 70  C, followed by thorough rinsing
with Milli-Q water and drying in a nitrogen stream.
(c) Change of functional groups.
For some applications (e.g., immobilization of DNA oligo-
mers with an amino-linker) the follow-up functionalization
via peptide chemistry works better with diaminopoly(ethyl-
ene glycol) treated with 5 M glutaric anhydride (GA) in N,N-
dimethylformamide (DMF) (10 μL/cm2) for 6 h to generate
carboxylic groups on the surface. The same protocol can be
used to modify AMD-surfaces with carboxylic groups.
(d) One step chemistry.
Depending on the expected properties of the biopolymer
layer, complex silanes offering chemically beneficial side
chains like 11-aminoundecyltrimethoxysilane can be used in
a one step procedure combining silanization with biopolymer
coating.

3.2.3 Immobilization of (a) The covalent coupling of amino-terminated molecules/

Amino Terminated Ligands ligands to biopolymers with carboxylic groups is done by
standard peptide chemistry via activated esters (see Note 7).
Therefore, the previously modified transducers are activated
with a solution of NHS and DIC (1 M NHS, 1.2 M DIC in
DMF, 10 μL/cm2) for 2–4 h as sandwich pairs in a DMF
saturated atmosphere (see Note 5). The transducers are
rinsed first with DMF and then with dry Aceton, and dried
in a nitrogen stream (see Notes 4 and 5).
(b) To achieve a high density of covalently bound molecules, it is
necessary to use an excess of reagents. Therefore, the ligand is
applied as a DMF solution (10 μL/cm2) (if your ligand is
insoluble in DMF, use another aprotic solvent of appropriate
polarity) at a concentration of 1–3 μM to the activated sur-
face for 4 h as sandwich pairs in a DMF saturated atmosphere
(see Note 6). Clean the transducers by rinsing them with
DMF and Milli-Q water and drying them in a nitrogen
stream (see Note 4).
216 Guenther Proll et al.

3.2.4 Immobilization of (a) For activation of the carboxy-terminated molecules, the

Carboxy Terminated ligand is dissolved in DMF at a concentration of 1–3 μM
Ligands together with 1.5 M DIC. After quick mixing, this solution is
immediately applied to a transducer modified with a biopoly-
mer providing amino groups at a concentration of 10 μL/
cm2. Place a second transducer on top to form a sandwich
pair (see Note 6).
(b) The reaction is finished after 4 h in a DMF saturated atmo-
sphere (see Note 6). Clean the transducers by rinsing them
with DMF and Milli-Q water and drying them in a nitrogen
stream (see Note 4).

3.2.5 Immobilization of (a) Dissolve 1 mg EMCS per 10 μL DMF and apply this solution
Thiol Terminated Ligands to a transducer modified with a biopolymer offering amino
groups at a surface concentration of 10 μL/cm2 for 6–12 h in
a DMF saturated atmosphere (see Note 6). After rinsing with
DMF (see Note 4), dissolve the ligand at a concentration of
1–3 μM in DMF, and apply this solution to the transducers at
a surface concentration of 10 μL/cm2 for 6–12 h as sandwich
pairs in a DMF saturated atmosphere (see Note 6).
(b) Clean the transducers by rinsing them with DMF and Milli-
Q water and drying them in a nitrogen stream (see Note 4).

3.2.6 Immobilization of (a) Biotin is immobilized by TBTU activation: D-biotin (1 mg,

Biotin 4 mmol), TBTU (1.4 mg, 4.4 mmol), and DIPEA (4 mL,
23.3 mmol) are mixed with DMF (50 mL) until the active
ester is formed and the reaction mixture appears homoge-
neous. This solution is dripped on to a transducer at a surface
concentration of 10 μL cm2 pretreated with a biopolymer
providing amino groups. Two of such transducers are put
together to form a sandwich.
(b) After a reaction time of 4 h in a saturated DMF atmosphere
(see Note 6), the sandwich is separated and both transducers
are rinsed with water and dried in a nitrogen stream.

3.3 Measurement Binding inhibition assay to determine the affinity constant of

Protocol for Standard antigen-antibody interaction: The transducer is modified according
Binding Inhibition to protocols given before either with a carboxy or an amino termi-
Assay for the nated antigen. Then fixed amounts of antibody are pre-incubated
Determination of the with fixed volumes of differently concentrated antigens (see Note 8)
Affinity of an to reach binding equilibrium before injection of the mixture with
Antigen–Antibody the FIA system (see Notes 9–12) on the transducer.
Interaction Antigens with affinity to the antibody in the pre-incubated
solution inhibit the binding of the antibody to the surface immo-
bilized antigen. In the case of diffusion (mass transport) limited
binding, the diffusion rate obeys first Fick’s law. This results in a
 Reflectometric Interference Spectroscopy 217

constant binding rate of the antibody to the surface and a linear

binding curve. The slope of the binding curve is determined by the
concentration of free antibody binding sites in solution. The slope
of the binding curves decreases with increasing concentration of
antigen in solution and reaches zero for high antigen concentra-
tions. Ovalbumin (final concentration of 200 μg/ml) should be
added to all antibody solutions to avoid loss of antibody by non-
specific binding in the fluidic system. The model function describ-
ing the concentration of antibodies with free binding sites is fitted
to the titration curve using a Marquart–Levenberg nonlinear least-
square algorithm (software ORIGIN from Microcal, Northamp-
ton/USA) .

3.4 Data Evaluation Measuring the concentration of active antibody in a sample work-
ing at mass-transport limited conditions is essential. Accordingly,
3.4.1 Determination of
the rate of diffusion to the surface is much slower than the binding
the Concentration of Active
reaction to the surface. The reaction kinetics can therefore be
Antibody
neglected. Mass-transport limited conditions can be achieved by a
high binding capacity of the surface and a low antibody concentra-
tion in solution. Under mass-transport limited conditions accord-
ing to first Fick’s law, the resulting binding curve is linear, and its
slope is proportional to the concentration of functional antibody. If
all samples contain the same amount of protein, determined by UV
spectroscopy or Bradford assay, it is possible to calculate the active
antibody concentration in a.u. by the different slopes of the sam-
ples, where the sample with the highest slope is assigned the a.u. 1
per μg used protein.

3.4.2 Determination of To determine the affinity and the kinetics of an antibody

Affinity and Kinetic (see Note 13), mass-transport to the surface must be much faster
Constants than the rate of binding to the surface. If this is the case, the mass-
transport can be neglected. This can be achieved by a high concen-
tration of bulk antibody and a low surface coverage of hapten
derivative.
The time dependence of the surface coverage can then be
described by a pseudo first-order binding reaction as following:
dΓðt Þ
¼ ka  c Ab  ðΓmax  Γðt ÞÞ  kd Γðt Þ ð3Þ
dt
with cAb: antibody concentration
Γ(t): surface coverage at time t.
Γmax: maximal surface coverage.
ka: association rate constant.
k d: dissociation rate constant.
Solving this differential equation, the surface coverage which is
equal to the resulting signal curve is given by:
218 Guenther Proll et al.

 
Γðt Þ ¼ ΓEq 1  ekobs t ð4Þ
with.
K aff  c Ab
ΓEq ¼ Γmax , ð5Þ
1 þ K aff  c Ab
the equilibrium coverage and.
kobs ¼ ka  c Ab þ kd , ð6Þ
the observed rate constant.
The affinity constant Kaff is given by the fraction of the associa-
tion rate constant, ka, and the dissociation rate constant, kd.
To obtain accurate values for ΓEq and kobs, it is necessary that
only the kinetics-controlled part of the signal curve is approxi-
mated, and that the surface and bulk are nearly in equilibrium at
the end of the measurement. Now it is possible to determine ka as
the slope of the straight line of the plot with kobs as ordinate and
active antibody concentration as abscissa. In order to determine
Kaff Eq. 5 can be rewritten as
1 1 1 1
¼ þ  : ð7Þ
ΓEq Γmax K aff  Γmax c Ab

. it is possible to obtain the value for Γmax with

With this formula,
1
a linear fit of the ΓEq 1
plot. Inserting this value in Eq. 3, a nonlinear
c Ab

least square fit results in a value forKaff. This

 method gives nearly
ΓEq.

the same values as a Scatchard plot cAb .

ΓEq

4 Notes

1. Very thick biolayers (more than approx. 100 nm) lead to mea-
surements out of the linear correlation between change of
optical thickness and shift of the interference spectrum.
2. Correct side of the RIfS transducers: mark the un-coated side
with a diamond pen to avoid mistakes during the surface
chemistry steps. The coated side of the transducers appears
colored because of the light interference.
3. In the case a transducer shows a grey shadow after rinsing and
drying repeat this cleaning procedure. If this does not work,
start the surface chemistry from the beginning (otherwise the
modification will not be homogeneous and will show high
nonspecific binding).
4. All surface chemistry protocols which produce highly reactive
groups are sensitive to humidity (deactivation). This can be
 Reflectometric Interference Spectroscopy 219

avoided by immediately applying the next step and working

under dry ambient conditions.
5. Use a tray with ground joint. Place the transducer as sandwich-
pairs (by assembling two slides face-to-face) in the tray on a
solid support and add the same solvent which is used in the
reaction for a solvent-saturated atmosphere.
6. In the case of limited ligands it is possible to use commercially
available piezo-based microdosing devices for printing a ligand
solution in Milli-Q water.
7. Usually phospahete buffered saline (PBS, 150 mmol NaCl and
10 mmol Di-potassiumhydrogenphosphate in Milli-Q water at
pH 7.4) is used for antigen-antibody interaction analysis. In
general all buffers can be used for RIfS measurements. Only
restriction: do not use buffers with pH >8.5 because of
destruction of the surface modification.
8. Testing of the modified transducer for nonspecific binding: use
ovalbumine (OVA) or bovine serum albumin (BSA) in excess
(e.g., 1 mg/mL) directly before the measurement.
9. Air disturbs the measurements: Use degassed buffers and avoid
negative pressure (sucking) through the flow cell. In addition,
it is helpful to keep the buffer under a slight positive argon
pressure.
10. If possible use the same buffers during a complete measure-
ment to avoid artifacts because of changes in the refractive
index.
11. After the interaction process, the transducer surface can be
regenerated. To remove antibodies from their antigens, a solu-
tion of 0.5% SDS (sodium dodecyl sulfate) at pH 1.9 is applied
via the FIA system. In the case of hybridization experiments, a
solution of either 0.25% SDS at pH 2.5 or a solution contain-
ing 6 M guanidinium hydrochloride and 6 M urea at pH 2 can
be used.
12. RIfS is not restricted for biosensing; this technique can be used
also for chemical sensing. Instead of biopolymers one can
modify the transducer with for example nano-porous polymer
layers, rubber like polymers or molecular imprinted polymers
(MIPs). The determination of kinetics is also possible.
13. Nowadays, even more advanced devices based on the one-color
approach are commercial available. These devices use the
revised reflectometry read out called “1-lambda RIDe” (1-
lambda Reflectometric Interference Detection). Currently, for
example an imaging based version for the label-free readout of
microarrays in the standard microarray format (b-screen®) is
available from Biametrics GmbH. This technology enables the
label-free detection of any microarray experiment like protein
220 Guenther Proll et al.

or peptide microarrays, resulting in multiplexed kinetic experi-

ments of up to 20,000 interactions in a single measurement.
14. Label-free detection like RIfS can be used for a great variety of
applications ranging from standard kinetic analysis [8] to
fragment-based screening [9]. New fields of application
include the detection of small molecules by molecular
imprinted molecules [10] or diagnostic applications [11–14].

References

1. Gauglitz G (2010) Direct optical detection in molecularly imprinted nanospheres: a

bioanalysis: an update. Anal Bioanal Chem 398 promising approach for robust and label free
(6):2363–2372 detection of small molecules. Anal Bioanal
2. www.biacore.com Chem 402(10):3245–3252
3. Homola J, Yee SS, Gauglitz G (1999) Surface 11. Ewald M, Le Blanc AF, Gauglitz G, Proll G
plasmon resonance sensors: review. Sens Actua- (2013) A robust sensor platform for label-free
tors B 54:3–15 detection of anti-Salmonella antibodies using
4. Schmitt HM, Brecht A, Piehler J, Gauglitz G undiluted animal sera. Anal Bioanal Chem
(1997) An integrated system for optical bio- 405(20):6461–6469. doi:10.1007/s00216-
molecular interaction analysis. Biosens Bioelec- 013-7040-9
tron 12:219–233 12. Krieg AK, Gauglitz G (2014) An optical sensor
5. Brecht A, Gauglitz G, Nahm W (1992) Inter- for the detection of human pancreatic lipase.
ferometric measurements used in chemical and Sens Actuators B 203:663–669. doi:10.1016/
biochemical sensors. Analysis 20:135–140 j.snb.2014.07.036
6. Brecht A, Gauglitz G, Kraus G, Nahm W 13. Ewald M, Fechner P, Gauglitz G (2015) A
(1993) Chemical and biochemical sensors multi-analyte biosensor for the simultaneous
based on interferometry at thin layers. Sens label-free detection of pathogens and biomar-
Actuators 11B:21–27 kers in point-of-need animal testing. Anal Bioa-
nal Chem 407(14):4005–4013. doi:10.1007/
7. Proell F, Moehrle B, Kumpf M, Gauglitz G s00216-015-8562-0
(2005) Label-free characterisation of oligonu-
cleotide hybridisation using reflectometric 14. Rau S, Hilbig U, Gauglitz G (2014) Label-free
interference spectroscopy. Anal Bioanal Chem optical biosensor for detection and quantifica-
382(2):1889–1894 tion of the non-steroidal anti-inflammatory
drug diclofenac in milk without any sample
8. Fechner P, Pröll F, Albrecht C, Gauglitz G pretreatment. Anal Bioanal Chem 406
(2011) Kinetic analysis of the estrogen receptor (14):3377–3386. doi:10.1007/s00216-014-
alpha using RIfS. Anal Bioanal Chem 400 7755-2
(3):729–735
15. Frank R, Möhrle B, Fröhlich D, Gauglitz G
9. Pröll F, Fechner P, Proll G (2009) Direct opti- (2005) A lable-free detection method of bio-
cal detection in fragment-based screening. Anal chemical interactions with low-cost plastic and
Bioanal Chem 393(6–7):1557–1562 other transperent transducers. Proc SPIE Int
10. Kolarov F, Niedergall K, Bach M, Tovar GEM, Soc Opt Eng 5826:551–560
Gauglitz G (2012) Optical sensors with
 Chapter 14

Hypermulticolor Detector for Quantum-Antibody Based

Concurrent Detection of Intracellular Markers for HIV
Diagnosis
Annie Agnes Suganya Samson and Joon Myong Song

Abstract
Antiretroviral treatment can reduce the death rate of human immunodeficiency virus (HIV) infection, and
its effectiveness is maximized at the early stage of HIV infection. The present protocol demonstrates an early
stage high-content HIV diagnosis based on multicolor concurrent monitoring of CD4, CD8, and CD3
coreceptors and F-actin cytoskeleton using quantum dot (Qdot)–antibody conjugates at the single cell
level. Artificial HIV infection of peripheral blood mononuclear cells (PBMCs) can be achieved by treating
PBMCs with gp120. Using the present methodology, we can determine the CD4–CD8 ratios of normal
PBMCs and artificial HIV-infected PBMCs. In addition, this protocol enables monitoring of structural
changes of actin filament alignments in PBMCs bound to gp120 proteins using the multicolor single cell
imaging system. Overall, this approach presents a new model for accurate early stage HIV diagnosis.
Simultaneously the approach provides information on actin cytoskeleton and subtypes of PBMCs as well
as their CD4–CD8 ratios.

Key words High-content imaging, Quantum dots, HIV diagnosis, PBMCs and actin filament

1 Introduction

Current HIV diagnosis is performed to detect HIV in blood, saliva,

and urine, and is divided into antibody or antigen tests according to
the target to be detected. The antibody test detects antibodies
arising from HIVs using ELISA or Western blot [1, 2]. However,
after exposure of a human to HIV, HIV antibodies in thebody are
not generally detected with blood tests by the time after infection
and before seroconversion (“window period”) during which mar-
kers of infection (HIV-specific antigen and antibodies) are still
absent or too scarce to be detectable. After the window period
the amount of HIV antibody reaches a high enough level to be
detected. This can often lead to false negative errors. HIV genetic
or p24 antigen tests are capable of complementing errors induced

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_14,
© Springer Science+Business Media LLC 2017

221
222 Annie Agnes Suganya Samson and Joon Myong Song

by an HIV antibody test. An HIV genetic test amplifies DNA or

RNA sequences inherent to HIV using polymerase chain reaction
(PCR), real-time PCR, or reverse transcriptase PCR [3]. Although
HIV genetic tests are very specific and accurate, these tests are
expensive and difficult to administer and interpret compared to
HIV antibody tests [4]. P24 antigen is a protein secreted from
HIV that invades host cells, and its presence verifies that HIV has
invaded host cells. Although the p24 antigen test works before HIV
antibodies are produced in the period immediately after HIV infec-
tion, the detection sensitivity of this test is very low. Gp120 is a
envelope glycoprotein exposed on the surface of HIV. The gp120
of HIV binds selectively to CD4 (cluster of differentiation 4)
expressed on the surface of immune cells such as T helper cells,
monocytes, and macrophages. The binding of HIV gp120 to CD4
causes a change in the conformation of gp120 that permits HIV
to bind to a coreceptor expressed in the host cell, which is
known as CCR5 or CXRX4. A structural change of HIV gp41
(glycoprotein 41) is then promotes fusion between HIV and the
host cell. As a result of this fusion, actin filament rearrangement
occurs within the host cell. At the early stage of HIV infection when
the gp120 of HIV binds to the CD4 coreceptor of a T cell, the
degree of conjugation between the CD4 coreceptor and the CD4
antibody decreases remarkably. This leads to a progressive reduc-
tion in the ratio of CD4–CD8. A ratio less than 1 is suggestive of
advanced immunosuppression. Another characteristic arising from
HIV infection is that morphological change of T cells is caused by
actin filament rearrangement. The present protocol uses quantum
dot (Qdot)–antibody nanoprobes for quantitative concurrent
monitoring of actincytoskeleton, CD3, CD4, and CD8 at the
single peripheral blood mononuclear cell (PBMC) level. The aim
of this procedure is to develop an early stage high-content HIV
diagnosis based on quantitative determination of the CD4–CD8
ratio and subtype of PBMCs, in addition to the monitoring of T cell
morphological change.
Our hypermulticolor single cell imaging system utilizes slightly
defocused PBMC images in the bright field mode. The slightly
defocused cellular images will provide uniform intensity distribu-
tion at the single PBMC level. The PBMC images in the defocused
mode should overlap those obtained by fluorescence intensity at a
particular single PBMC level [5]. A threshold value is fluorescent
intensity at a particular single PBMC region which has the lowest
fluorescence intensity among the PBMCs, but greater intensity
than that of the background signal. The threshold value should
be applied to all overlapped PBMC images to select single PBMCs
larger than the threshold value. This operation enables the quanti-
fication of total cells expressing CD coreceptors or actin filament
structural changes for HIV diagnosis. This quantitative approach
provides a statistical basis for measuring the CD4–CD8 ratio and
determining the subtype of PBMCs.
 Q-Dot Probes for HIV Diagnosis 223

2 Material

1. Human blood (Seoul National University Hospital).

2. HIV envelope glycoprotein gp120 (HIV-1 IIIB gp 120;
Immuno Diagnostics, Inc., Woburn, MA, USA).
3. HISTOPAQUE-1077 (Sigma, Poole, Dorset, UK).
4. Acetone (prechilled).
5. Wash buffer (WB; 1
PBS, 1% FCS (fetal calf serum), 0.01%
sodium azide).
6. Cold 1
PBS (phosphate buffer saline).
7. 3.7% formaldehyde.
8. 1% BSA (bovine serum albumin).
9. Mouse-antihuman CD3-Qdot605,mouse-antihuman CD8-
Qdot565, mouse-antihuman CD4-Qdot655.
10. Alexa Fluor 488 phalloidin ( 200 mM).
11. ITK inhibitor (Interleukin-2-inducible T-cell Kinase) (Inter-
leukin-2-inducible T-cell Kinase) (BMS509744; AdooQ Bio-
science Irvine, CA, USA).
12. Hemocytometer.
13. Centrifuge.
14. Fluorescence microscope.
15. Optical components:
l An Ar ion laser—350 nm ( Melles Griot Laser Group,
Carlsbad, CA, USA).
l An interference filter (U-MGFPHQ, Olympus, Tokyo,
Japan).
l 20
, 40
, and 60
Objective lens (Olympus).
l Acousto-optic tunable filter (AOTF; TEAF10-0.45-0.7-S,
Brimrose).
l Charge-coupled device (CCD) camera.
l A long-pass filter.
l Computer.
l Software for Image analysis—MetaMorph (Version 7.1.3.0,
Molecular Devices, Sunnyvale, CA, USA).

2.1 Special Acousto-optical transmission filter (AOTF) microscope consists of

Equipment different components, which includes a C-mount lens, an acousto-
optic tunable filter (AOTF), a charge coupled device (CCD) cam-
2.1.1 Acousto-Optical
era, an Ar ion laser—350 nm, an interference filter, a fluorescence
Transmission Filter (AOTF)
microscope, and a computer system equipped with image acquisi-
Microscope
tion and analysis software (MetaMorph). In this assembled imaging
224 Annie Agnes Suganya Samson and Joon Myong Song

system, AOTF will function as a tunable emission filter. A frequency

of light gets diffracted at the Bragg angle that is built into the
AOTF. The diffracted light that is the fluorescence emission at a
particular wavelength will be transmitted and collected in a CCD
camera. Fluorescence images at varying wavelength could be
obtained using AOTF. The spectral range varies from 400 to
1000 nm wavelength. The AOTF microscope can select a single
wavelength which will not overlap with other marker emission
wavelengths. A single wavelength can be selected for observation
of HIV markers. Maximum emission wavelengths of the QDs
(quantum dots) used in this protocol occur at 565, 605, and
655 nm. No significant overlap occurs with emission wavelengths
of QDs. Moreover, while using AOTF allowing transmission at a
single wavelength, QD605 is barely detected when detection of
QD655 is executed. Therefore, AOTF will provide much smaller
spectral interference compared to bandpass filters when high-
content HIV diagnosis is performed based on simultaneous moni-
toring of different markers. Simultaneous monitoring of different
intracellular event (e.g., activation or deactivation of protein, recep-
tor activity) in a single cell using fluorescent probes is termed as
high-content based detection. This enhances the accuracy of early
stage HIV diagnosis. As a probe for HIV markers, conventional dye
antibody conjugate can be replaced with QD–antibody conjugate.
Thus QD-antibody conjugate along with AOTF is advantageous
for more accurate early stage HIV diagnosis. QDs having narrow
emission wavelength ranges are shown to be much more suitable to
high-content HIV diagnosis than conventional fluorescent dyes.

3 Methods

3.1 Isolation of Peripheral blood mononuclear cells (PBMCs) include lymphocytes

PBMCs from Human (T cells, B cells, and NK cells), monocytes, and dendritic cells.
Blood PBMCs are considered as the population of immune cells. HIV
infection is associated with hyperactivation of the immune system.
The reproduction of HIV in T cells is closely correlated to the
proliferation of the cells. In general, bulk populations of T cells
(PBMC) are infected with HIV, which is connected with the
increased T cell proliferation leading to greater viral replication. T
cells in PBMC are extremely varied both functionally and pheno-
typically. This segment describes an artificial method to establish
the HIV-infected PBMCs (peripheral blood mononuclear cells).
This segment describes an artificial method to establish the
HIV-infected PBMC (peripheral blood mononuclear cells) model.
1. Add 5 mL of HISTOPAQUE-1077 to a 15 mL conical centri-
fuge tube.
 Q-Dot Probes for HIV Diagnosis 225

Centrifuge
250g, 10 min, 25° C

Whole blood
Plasma layer
PBMC
Histopaque-1077
Histopaque-1077
RBCs PBMC

Fig. 1 Separation of PBMCs from blood sample

2. Add 5 mL human blood (Note: Blood sample should be care-

fully layered over the HISTOPAQUE-1077 to prevent
mixing).
3. Centrifuge at 400
g for 30 min at 25  C.
4. After centrifugation, three separate layers will be observed.
5. First, remove the upper transparent plasma layer of the three
separate layers using a pipette.
6. Next, the opaque interface layer containing PBMCs should be
gently transferred with a pipet into a clean 15 mL centrifuge
tube (Fig. 1).
7. Fill the 15 mL centrifuge tube containing PBMCs with 1
PBS
solution, mix carefully and then centrifuge at 250
g for
10 min at 25  C.
8. After centrifugation, carefully discard the supernatant and
resuspend the white PBMCs with 15 mL of cold 1
PBS
solution. Centrifuge at 250
g for 10 min at 4  C.
9. After centrifugation, discard the supernatant and resuspend the
PBMC pellet in 6.25 mL of 1
PBS solution.
10. Perform cell counting using hemocytometer.

3.2 Treatment of Artificial HIV infection of peripheral blood mononuclear cells

PBMCs with gp120 (PBMCs) is accomplished by treatment of PBMCs with gp120:
1. Centrifuge the PBMCs solution at 250
g for 10 min at 4  C.
2. Discard the supernatant and wash the PBMC pellet twice with
wash buffer.
3. For the treatment of PBMCs with gp120, take PBMCs
(1
106 cells) in 30 μL of WB and add 15 μL gp120
(10 μg/mL).
4. Incubate the mixture for 6 h at 4  C.
5. Finally, after the reaction with gp120 wash the PBMCs thrice
with 5 mL WB (Note: For washing, centrifugation at 250
g
for 10 min at 4  C).
226 Annie Agnes Suganya Samson and Joon Myong Song

3.3 Treatment of Monitoring of the CD4–CD8 ratio can greatly improve HIV diag-
PBMCs with nosis accuracy in early stages of infection, and it is important to
Qdot–Antibody accurately determine this ratio. This segment describes a staining
Conjugates protocol using quantum dot (Qdot)–antibody nanoprobes for
quantitative concurrent monitoring of actincytoskeleton, CD3,
CD4, and CD8 in normal and gp120 treated PBMCs at the level
of single cells.
1. Fix the PBMC or gp120-treated (1
106) cells using 3.7%
formaldehyde for 10 min and then wash twice with 5 mL 1

PBS solution (centrifugation at 250
g for 10 min at 4  C).
2. Place the fixed PBMCs in prechilled acetone for 3 min at
20  C and then wash twice with 1
PBS. (centrifugation at
250
g for 10 min at 4  C).
3. Immerse the PBMCs pellet in 200 μL of 1% BSA aqueous
solution.
4. Add 1 μL of each Qdot–antibody (Mouse-antihuman CD3-
Qdot605, mouse-antihuman CD8-Qdot565, and mouse-
antihuman CD4-Qdot655) stock solutions (1:200 dilutions)
into 200 μL of 1% BSA aqueous solution containing 1
106
normal or gp120-treated PBMCs.
5. Incubate the solution for 1 h at room temperature.
6. After 1 h, wash the Qdot stained cells twice with 1
PBS.
7. Finally, incubate the cells with Alexa Fluor 488 phalloidin in 1

PBS for 20 minto detect F-actin filament in PBMCs.

3.4 Detection of CD3, This section describes a system to simultaneously monitor fluores-
CD4, CD8, and F-Actin cence emission of different wavelengths. PBMCs stained with Alexa
filament in PBMCs Fluor 488 phalloidin, antihuman CD3-Qdot605, antihuman CD8-
Qdot565, and antihuman CD4-Qdot655 can be concurrently
observed using this Hypermulticolor single cell imaging system.
This system provides a scanning rate of 60 wavelengths (nm)/min,
which will enable 60 different cellular images perminute. First,
PBMCs on the sample stage are irradiated by UV using a micro-
scope objective lens to excite all the Qdot–antibody conjugates and
Alexa Fluor 488 phalloidin. Fluorescent emission from the PBMCs
are collected using an identical lens, transmitted through an
acousto-optical transmission filter (AOTF), and images detected
as a function of wavelength using a charge-coupled device (CCD
(Fig. 2).
1. Alexa Fluor 488 phalloidin detects F-actin filament structural
changes of PBMCs at 518 nm. The three quantum dot tagged
antibody conjugates (antihuman CD3-Qdot605, antihuman
CD8-Qdot565, and antihuman CD4-Qdot655) are bound to
CD3, CD4, and CD8 coreceptors expressed in the cellular
membrane of the PBMCs.
a
Blood

Treatment No treatment
HIV-1 gp120

Qdot-anbody
conjugates/Alexa
Fluor®488 phalloidin

Sample

T helper cell CD3 CD8 CD3-Qdot 605 CD8-Qdot565

T cytotoxic cell CD4 Fluor®488 phalloidin CD4-Qdot 655

b
12
Computer
1 Sample

2 Objecve lens

Exciter

3
Laser
Beam splier 4
10
11 5
CCD camera

9 Prism
Long-pass filter Focusing
AOTF
Lens
8 7 6

Fig. 2 (a) Multicolor single cell imaging. (b) Hypermulticolor single cell imaging detector. The components of
the imaging system are as follows: (1) sample stage, (2) objective lens, (3) beam splitter, (4) laser, (5) prism,
(6) focusing lens, (7) AOTF, (8) filter, (9) undiffracted light, (10) diffracted light, (11) CCD camera, and (12)
computer
228 Annie Agnes Suganya Samson and Joon Myong Song

3.5 Treatment of The following steps illustrate an experimental procedure to verify

gp120-Treated PBMC the morphological change in T cells caused by actin filament rear-
with ITK Inhibitor rangement; characteristic changes arise during HIV infection.
1. Preincubate the PBMCs (1
106 cells) with 10 μM ITK
inhibitor for 1 h before the gp120 treatment.
2. After 1 h, for the reaction of PBMCs with gp120; take PBMCs
(1
106 cells) in 30 μL of WB and react with 15 μL gp120
(10 μg/mL).
3. Incubate the mixture for 6 h at 4  C.
4. Wash the PBMCs thrice with 5 mL WB.
5. Fix the PBMCs cells in 3.7% formaldehyde for 10 min and then
wash twice with 1
PBS solution.
6. Place the fixed PBMCs in acetone for 3 min at 20  C and then
wash twice with 1
PBS.
7. Incubate the cells with Alexa Fluor 488 phalloidin in 1
PBS
for 20 min. Detect F-actin filament in PBMCs at 518 nm.

3.6 Hypermulticolor This protocol is based on quantitative multivariate imaging (QMI)

Detector cytometry combined with fluorescent semiconductor nanocrystals,
which can be used for various quantitative cellular image-based
analyses. This standardized protocol can be adapted to various cell
types. Furthermore, the developed assay can overcome the difficul-
ties inherent in flow cytometric assays. It is more miniaturized, cost
effective, and user friendly to operate. A commercial flow cytometer
has 8 detectors; which limits the multicolor/multispectral analy-
sis to detect eight target molecules. But in case of QMI cytometer
provides a wide range of filtering wavelengths at 3.75-nm inter-
vals, making the system more compatible for high-throughput
detection. Hence, this system enables simultaneous monitoring of
several cellular molecules at single cellular level. So, this imaging
system is described as Hypermulticolor detector system. By using a
selective wavelength tuning, Hypermulticolor detector can mini-
mize the interference of unwanted auto fluorescence of cells during
cellular imaging analysis. Therefore, this protocol allows prompt,
quantitative data to visualize and quantify molecular targets, while
simultaneously highlighting there importance in disease diagnosis.
Certainly, this hypermulticolor imaging system can be used for
concurrent monitoring of various intracellular biomolecules by
rapid multicolor/multispectral analysis, which cannot be achieved
by filter-based conventional imaging systems (see Note 1). Subse-
quently, utilizing these concepts may have a significant role in the
early diagnosis of HIV based on targeted molecules by using this
developed protocol.
The Hypermulticolor system is based on uniform threshold
intensity distribution (TID). It consists of an AOTF, C-mount
 Q-Dot Probes for HIV Diagnosis 229

lens, CCD camera, fluorescence microscope, and commercially avail-

able software for image acquisition and analysis. The technique used
for data acquisition and quantitative analysis, involves following
steps: (1) Coordinating the exposure time of CCD camera with
AOTF scanning/filtering wavelength to obtain desired image plane
at desired wavelength parameters. (2) Acquiring a focused image at
an identical X and Y axis position. (3) Applying background correc-
tion and threshold adjustment. Certainly, quantitative analysis is
possible by selecting a region around the threshold objects regardless
of their cellular morphology. Image analysis can be carried out using
MetaMorph software.

3.7 Imaging System As mentioned before, the imaging system consists of various com-
ponents (Fig. 3). A mercury arc lamp acts as an illumination source
for cells (sample) in a fluorescence microscope and the laser beam is
purified by an interference filter, allowing it to pass through the
excitation filter and finally, reflected by a dichroic filter. The filtered
laser beam is focused through a microscope objective lens (20
,
40
, 60
) and then onto the sample platform. The quantum dot
tagged probes in the cells (sample) are induced to emit fluorescence
by the filtered laser beam. The fluorescent emission is collected by
the same objective lens, passes through dichroic filter and comes
out of the side port (ø ¼ 25 mm) of the microscope. In order to
reduce the fluorescence beam diameter, the C-mount lens is
attached to the side port of the fluorescence microscope. Then,
the entire beam passes through the AOTF window (10
10 mm).
The nano crystal in the AOTF splits the fluorescence beam into two
separate beams, the diffracted light (at particular wavelength) and

Fig. 3 Photo of the imaging system

230 Annie Agnes Suganya Samson and Joon Myong Song

undiffracted light. Thus, the fluorescence image of the cells can be

detected at a particular wavelength and the fluorescence images are
captured by a CCD camera. A long-pass filter is placed in front of
the CCD camera to eliminate any laser scattering. The exposure
time of the CCD camera is 1 s and the spectral resolution of the
AOTF is adjustable up to 0.1 nm. Data analyses of fluorescence cell
images are performed automatically using commercially available
software (MetaMorph, Version 7.1.3.0, Molecular Devices).

3.8 Data Analysis Quantitative data analysis can be done using different parameters
like optical density, average gray value, Z position, angle, distance,
area, width, image plane, elapsed time, stage label, wavelength,
region label, intensity S/N, and threshold area. Single cell imaging
cytometry can be used for simultaneous monitoring of cellular
events, that is, expression of different cellular markers. Figure 4a

Fig. 4 (a) Concurrent monitoring of CD3, CD4, CD8 coreceptors and F-actin in PBMCs. The top figures were
obtained from normal PBMCs while the bottom figures were obtained from gp120-treated PBMCs. F-actin
images of normal PBMCs were compared with those of gp120-treated PBMCs. The gp120-treated PBMCs
indicated with arrows showed ununiform distribution of F-actin filament. On the other hand normal
PBMCs showed uniform distribution of F-actin filament. (b) Inhibitory effect of ITK on cytoskeleton structure
of PBMCs treated with gp120 glycoprotein. Top cellular images are PBMCs treated with gp120 glycoprotein
(10 μg/mL) while cellular images at the bottom are PBMCs treated with gp120 glycoprotein (10 μg/mL) and then
ITK inhibitor (10 μM). Gp120 glycoprotein-treated PBMCs show uniform actin cytoskeleton by the ITK inhibitor
treatment
 Q-Dot Probes for HIV Diagnosis 231

represents the actin filament rearrangement of PBMCs by gp120-

induced signaling transduction. The figure data clearly confirms
that normal (control) PBMCs has uniform actin cytoskeleton
than that of gp120 treated PBMCs. This is due to actin rearrange-
ment that occurs when HIV-1 gp120 binds to PBMCs. Hence, a
morphological change occurs in PBMCs. In fact, the rearrange-
ment of actin cytoskeleton is a common happening that occurs in
the HIV infected host cells and it is also considered as a good
indicator to verify HIV infection. Qdot nanoprobes based high-
content HIV diagnosis system make possible and efficient for the
detection of different cellular markers like CD3, CD4, and CD8,
simultaneously in both control and gp-120 treated PBMCs. The
results demonstrate that there is significant reductions in the
expression of CD4 in gp-120 treated PBMCs compared to control
PBMCs. Moreover, ~0.89- and ~0.95-fold difference of CD8 and
CD3 respectively, was observed in control and gp-120 treated
PBMCs. Figure 4b shows inhibitory effect of ITK on the rearrange-
ment of actin cytoskeleton induced by gp120 glycoprotein. ITK is
required for effective HIV transcription, hence leading to increases
in virus-like particle formation. Earlier studies have demonstrated
that ITK affects gp120 protein-induced rearrangement of actin
cytoskeleton that is needed for HIV to enter a host cell. Thus,
ITK inhibitor has been studied to inhibit HIV entry into host
cells and HIV replication. A nonuniform actin cytoskeleton was
not observed in PBMCs treated with ITK inhibitor [6]. These data
verifies that simultaneous monitoring of CD3, CD4, and CD8
coreceptors can clearly prove that actin rearrangement occurs as a
result of HIV infection. To conclude, concurrent quantitative mon-
itoring of actin cytoskeleton, CD3, CD4, and CD8 at the single cell
level enables early stage HIV diagnosis without false negative error.

4 Note

1. Simultaneous monitoring of different intracellular event (e.g.,

activation or deactivation of protein, receptor activity) in a
single cell using fluorescent probes is termed as high-content
based detection. High-content based HIV diagnostic method
is highly sensitive and more compatible for simultaneous mon-
itoring of morphological changes and quantitative detection of
targeted receptors in a single cell. This method is more sensitive
than flow cytometry. Following are the benefits of incorporat-
ing this protocol in early stage HIV detection. It reveals a new
model based on multicolor concurrent monitoring of CD4,
CD8, and CD3 coreceptors and F-actin cytoskeleton using
quantum dot (Qdot)–antibody conjugates at the single cell
level. Due to brightness and photostability of Qdot, they pro-
vide relatively easy operation based on the use of a microscope.
232 Annie Agnes Suganya Samson and Joon Myong Song

This enables simultaneous monitoring of T cellular morpho-

logical changes as well as CD4–CD8 ratio, and greatly contri-
butes to the accuracy improvement of early stage HIV
diagnoses. False negative errors in the early stages of HIV
infection is greatly reduced by the use of hypermulticolor single
cell imaging cytometry. Hence, Hypermulticolor imaging sys-
tem provides new opportunities to efficiently diagnose early
stage HIV infection at single cell cytometric level.

References
1. Chun TW, Engel D, Berrey MM, Shea T, Corey 4. Tcherepanova I, Harris J, Starr A, Cleveland J,
L, Fauci AS (1998) Early establishment of a pool Ketteringham H, Calderhead D, Horvatinovich
of latently infected, resting CD4(+) T cells dur- J, Healey D, Nicolette CA (2008) Multiplex RT-
ing primary HIV-1 infection. Proc Natl Acad Sci PCR amplification of HIV genes to create a
U S A 95(15):8869–8873 completely autologous DC-based immunother-
2. Kahn JO, Walker BD (1998) Acute human apy for the treatment of HIV infection. PLoS
immunodeficiency virus type 1 infection. N One 3(1):e1489. doi:10.1371/journal.pone.
Engl J Med 339(1):33–39. doi:10.1056/ 0001489
NEJM199807023390107 5. Kim MJ, Lee SC, Pal S, Han E, Song JM (2011)
3. Schvachsa N, Turk G, Burgard M, Dilernia D, High-content screening of drug-induced cardi-
Carobene M, Pippo M, Gomez-Carrillo M, otoxicity using quantitative single cell imaging
Rouzioux C, Salomon H (2007) Examination cytometry on microfluidic device. Lab Chip 11
of real-time PCR for HIV-1 RNA and DNA (1):104–114. doi:10.1039/c0lc00110d
quantitation in patients infected with HIV-1 6. Tak YK, Song JM (2013) Early stage high-
BF intersubtype recombinant variants. J Virol content HIV diagnosis based on concurrent
Methods 140(1-2):222–227. doi:10.1016/j. monitoring of actin cytoskeleton, CD3, CD4,
jviromet.2006.11.012 and CD8. Anal Chem 85(9):4273–4278.
doi:10.1021/ac303727e
 Chapter 15

Low-Cost Charged-Coupled Device (CCD) Based Detectors

for Shiga Toxins Activity Analysis
Reuven Rasooly, Ben Prickril, Hugh A. Bruck, and Avraham Rasooly

Abstract
To improve food safety there is a need to develop simple, low-cost sensitive devices for detection of food-
borne pathogens and their toxins. We describe a simple, low-cost webcam-based detector which can be
used for various optical detection modalities, including fluorescence, chemiluminescence, densitometry,
and colorimetric assays. The portable battery-operated CCD-based detection system consists of four
modules: (1) a webcam to measure and record light emission, (2) a sample plate to perform assays, (3) a
light emitting diode (LED) for illumination, and (4) a portable computer to acquire and analyze images. To
demonstrate the technology, we used a cell based assay for fluorescence detection of the activity of the food
borne Shiga toxin type 2 (Stx2), differentiating between biologically active toxin and inactive toxin which is
not a risk. The assay is based on Shiga toxin inhibition of cell protein synthesis measured through inhibition
of the green fluorescent protein (GFP). In this assay, GFP emits light at 509 nm when excited with a blue
LED equipped with a filter at 486 nm. The emitted light is then detected with a green filter at 535 nm.
Toxin activity is measured through a reduction in the 509 nm emission. In this system the level of detection
(LOD) for Stx2 was 0.1 pg/ml, similar to the LOD of commercial fluorometers. These results demonstrate
the utility and potential of low cost detectors for toxin activity. This approach could be readily adapted to
the detection of other food-borne toxins

Key words LED, CCD, Microbial pathogens, Toxin activity, Shiga toxin, GFP, Webcam, Fluores-
cence, Fluorometer

1 Introduction

In recent years, charged-coupled devices (CCDs) [1–9] or comple-

mentary metal–oxide–semiconductors (CMOSs) [10–13] are
increasingly utilized as optical detectors because of their low cost,
small size, sensitivity, and low power consumption. Their ability to
image large surfaces makes them ideal for sample detection because
many samples can be imaged and analyzed simultaneously. The cost
of technical and scientific-grade CCD and CMOS imagers for
biodetection is typically high. While webcams, digital cameras,
digital imagers for amateur astronomy, and other consumer

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_15,
© Springer Science+Business Media LLC 2017

233
234 Reuven Rasooly et al.

electronics incorporating imaging components are more afford-

able, there is a need for low-cost alternatives providing sufficient
sensitivity and imaging quality for biodetection at significantly
lower cost. This is especially important for point-of-care (POC)
diagnostics for global health where it is desirable to enhance the
quality of health care delivery for underserved populations at very
low cost.

1.1 Considerations In both CCD and CMOS devices, photons striking the semicon-
for Optical Detector ducting material generate electron–hole pairs, and the resulting
Selection electric charge is processed into electronic signals. However, the
reading of these charges in each device is different. In CCDs, the
charge from each pixel is sequentially transported across the chip to
an analog-to-digital (A/D) converter, which converts the charge
into a digital value. The use of a single converter for all pixels
increases the uniformity of the output—an important factor in
image quality. However, blooming is a problem in CCD sensors
when a part of the image is over-exposed, causing leakage of elec-
trical charge to adjacent pixels whose signals become corrupted.
The low noise associated with a CCD sensor enables longer expo-
sure times for low-level light detection, provided the CCD has a
low thermal noise level relative to the incoming photonic signal.
This typically necessitates the use of a cooling system in order to
minimize temperature and the subsequent thermal noise for CCD
or CMOS sensors.
In most CMOS devices each pixel has its own charge-to-volt-
age conversion and the sensor may include transistors at each pixel,
noise correction, and digitization circuits to convert the output
signal to digital so that the chip outputs digital bits. When light
hits the sensor, noise may result from some photons hitting the
transistor instead of the pixel. By placing the transistors under the
pixels (i.e., back-side illuminated CMOS sensor) this type of noise
can be reduced. This configuration increases the chip complexity,
reducing the area available for light capture. Moreover, because
each pixel is doing its own conversion, uniformity is reduced,
especially at higher temperatures. Because the output of a CMOS
is massively parallel compared to a CCD, it is possible to achieve
higher image acquisition speeds and more efficient use of energy.
CMOS sensors are also immune to the blooming effect.
CCDs utilize global shutters in which the entire imager is reset
before integration in order to remove any residual signal in the
sensor pixels; the light collection starts and ends at exactly the same
time for all pixels with no motion blur. CMOS devices utilize a
rolling shutter in which light is not collected simultaneously by all
of the pixels. Thus the time that light collection starts and ends is
slightly different for each row; the top row of the imager is the first
one to start and finish collecting light. The row-by-row process of
 Low-Cost Charged-Coupled Device (CCD) Based Detectors. . . 235

light collection in a rolling shutter camera can result in motion blur

and switch noise effect, particularly at high gain settings.
To summarize, CMOS sensors have cheaper manufacturing
costs, greater energy efficiency, no blooming and faster data-
throughput speeds, while CCDs have lower dark currents, larger
areas available for light capture in low light level applications, and
more pixel uniformity. Given the rapid improvement in sensor
technologies such as complex pixel with correlated double sam-
pling (CDS), many CMOS devices overcome some of the technol-
ogy limitations and gradually blur the boundary between the
capabilities of the two competing imaging technologies. Therefore,
both CCD and CMOS technologies are now being broadly used for
imaging in a variety of fields, with the exception that low-cost
devices still primarily use CMOS sensors due to their substantially
lower manufacturing costs.

1.2 Optics for Digital The optics for digital detectors are relatively simple, consisting of a
Detectors lens and filters. A good lens for biodetection has a focal length of
4–12 mm at f1.2. Such a wide aperture, wide-angle lens is capable
of imaging a wide field-of-view (FOV) from a short distance,
providing a more compact device configuration. However, to
achieve this wide FOV at a short distance, the portion of the
image captured along the edges of the lens ends up being curved,
resulting in barrel distortion.
In addition to fixed focal length lenses, zoom lenses have
variable focal lengths enabling them to image both wide and nar-
row FOVs. Their main disadvantage is they are slower, depending
on the focal length used (usually operated at f/2.8–f/5.6), than
comparable fixed focal length lenses (e.g., f/1.2), resulting in
longer exposure times. Extension tubes are needed for greater
magnification to achieve macro (i.e., close-up) imaging. The exten-
sion tube contains no optical elements and is placed between the
sensor and the lens. It moves the lens farther from the sensor image
plane and closer to the imaged surface. Alternatively, reversing the
lens so small objects close to the image sensor can be magnified and
projected onto the image sensor [14] can achieve a macro effect
similar to extension tubes for imaging a very small detection surface
at high magnification. There is a trade-off in using extension tubes:
the farther the lens is from the sensor, the greater the loss of light.
Furthermore, extension tubes may not be the only optical element
affecting light intensity. For fluorometry, emission and excitation
filters are also needed that further reduce the optical signal. There-
fore, the choice of lens, extension tube, and filters must all be
compatible in order to image with enough sensitivity over smaller
surface areas on a detection plate.

1.3 Light Sources for Light sources are critical for optical detection, especially for
Optical Detection fluorescence-based detection. Traditional optical detectors utilize
236 Reuven Rasooly et al.

high intensity, high power and bulky bench-top excitation sources

including tungsten, mercury, or xenon lamps. These light sources
are usually expensive and nondurable. They also require high volt-
age, which further limits the portability of the device. Lasers [1, 7,
15, 16] have been used for illumination sources since they are low
cost, highly efficient, small, simple to operate, durable and consume
relatively little power. These characteristics make lasers ideal excita-
tion light sources for portable detectors. However, lasers are inher-
ently spot light sources that can only be used effectively with
photodiodes or photomultiplier detectors. When using lasers as
spot light sources to illuminate larger surface areas, mechanical
translation stages can be used for raster scanning, which compli-
cates the operation of the imaging system and slows down the
imaging. Alternatively, a line generator and evanescence light can
be used with a laser [1, 7, 15, 16] to expand the light source and
cover larger surface areas compatible with the spatial imaging of a
CCD or CMOS sensor. However, this may introduce light distri-
bution and uniformity problems, and still complicates the design.

1.3.1 Electrolumine- Electroluminescence (EL) illuminators are illuminated surfaces

scence (EL) Illumination providing spatial illumination, which is inherently more compatible
with CCD and CMOS optical detectors [17, 18]. EL illuminators
are semiconductor surfaces that emit light in response to an alter-
nating electric current. In EL, electrons that are excited into a
higher state leave “holes” and when dropping back to lower-energy
ground state, the excited electrons release their energy as photons
upon recombining with their “holes”. The spectrum of the light
emitted by EL depends on the electroluminescent materials.
Although EL material is widely used for signs and instrument dial
illumination (indigo watches, clock radio, personal organizers,
cockpit instrumentation, etc.), they are not commonly used for
fluorescence excitation and biodetection. Nevertheless, they have
great potential for biodetection applications given that the features
of EL include spatial illumination (a great advantage for lab-on-a-
chip (LOC) applications), a range of wavelengths, low cost, high
efficiency, small footprint, simple operation, durability, and low
power consumption. However, EL generally emits fewer photons
(i.e., less light) than other light sources used for fluorescence
excitation such as tungsten, mercury, xenon lamps, LEDs, or lasers.
Therefore they require longer exposure times. Another limitation is
that they can require high voltage (e.g., 100 V) to enhance photon
emission.

1.3.2 Light-Emitting Light-emitting diodes (LEDs) have been used as illumination

Diode (LED) Illumination sources in various optical detectors such as portable real-time
PCR detection [19] and fluorometers [18, 20, 21]. LEDs are
semiconductor devices that emit light when an electric current
passes through them. Similar to ELs, LEDs operate by
 Low-Cost Charged-Coupled Device (CCD) Based Detectors. . . 237

electroluminescence, in which the emission of photons is caused by

electronic excitation of a material. In the LED, the two semicon-
ductor materials are of different composition, one negative, or n-
type, and the other positive, or p-type, forming a p–n junction.
Under an electric field, current flows across the p–n junction. The
free electrons moving across a diode can fall into empty holes from
the p-type layer. This involves a drop from the conduction band to a
lower orbital, so the electrons release energy in the form of
photons. The LED light is visible when the diodes are characterized
by a wider gap between the conduction band and the lower electron
orbit (i.e., wide band-gap). The size of this gap (depending on the
semiconductor materials) determines the frequency of the photon
(i.e., the wavelength the light). Unlike EL, LED intensity is high
and suitable for fluorophore excitation in fluorescence detection.
However, several LEDs are required to illuminate a surface, and
unlike EL, which provides uniform illumination, the illumination
of a panel of LEDs is not uniform. However, LEDs have great
potential use for optical sensor illumination [20] because they are
available in a wide variety of wavelengths. Furthermore, some
LEDs can emit light over a range of wavelengths, enabling multi-
wavelength analysis. They are low cost, highly efficient, small,
simple to operate, durable and consume relatively little power.
Their main advantage over ELs is they emit far more light than
ELs and require low voltage DC for operation. Their main disad-
vantage is, unlike ELs, they are spot illuminators like lasers. There-
fore, surface illumination with LEDs requires optical systems
(mirrors, diffusers, lenses, etc.) that may complicate the design
and introduce light distribution and uniformity problems similar
to lasers.

1.3.3 Screens for The screen of a portable device can also be used for effective
Illumination of Portable illumination. The screen enables high intensity uniform illumina-
Devices tion in red, blue and green. Screens for a portable devices such as a
smartphones can be organic LED (OLED) or liquid crystal display
(LCD). The major difference between the two is the use of a
backlight for an LCD display, where the pixels contain liquid crys-
tals that regulate the wavelengths that pass through the pixel result-
ing in the desired color. One advantage of OLEDs over LCDs is the
individual electrical control of each pixel in an OLED reduces
power consumption, since even a “black” screen for an LCD
requires electrical power to control the orientation of the liquid
crystals to prevent any light from passing through, while the OLED
is naturally black without any electrical power. However, the use of
a single backlight for all pixels in an LCD results in greater
uniformity.
238 Reuven Rasooly et al.

1.4 Shiga Toxin In order to demonstrate the use of CCD-based biodetection we

Detection will discuss a specific system we have previously developed for the
detection of Shiga toxin activity [22]. Sensitive methods for detec-
tion of Shiga toxin are an important component of the effort to
prevent food poisoning. However, current detection technologies
cannot distinguish between biologically active and inactive toxins.
Current detection approaches rely mainly on immunological assays
using expensive microplate readers or spectrophotometers. How-
ever, these approaches cannot differentiate between a biologically
active toxin and an inactive toxin not posing a risk. In our
previous work we developed a cell based assay [22] for fluorescence
detection of the activity of the food borne Shiga toxin type 2 (Stx2)
which can differentiate between biologically active and inactive
toxins. The assay is based on Shiga toxin inhibition of protein
synthesis measured by green fluorescent protein (GFP), so toxin
activity can be measured as a function of GFP light emission. When
excited by a blue LED equipped with a filter at 486 nm, a light
emission peak occurs at 509 nm which can be detected with a green
filter at 535 nm.
For this assay, mammalian Vero cells were transduced with an
adenovirus coding GFP (Fig. 2a). The cells were treated with
various amounts of Shiga toxin, and after incubation the cell fluo-
rescence was detected with the CCD detector. The plate containing
the cells was illuminated by the LED illuminator using the blue
LED with a blue excitation filter. The detection plate was imaged
with a green emission filter and the signals of the wells were quan-
tified by the CCD camera operating in a video mode. The image
was analyzed and quantified by the open-source imaging software
ImageJ. By using this open source software it is possible to calculate
the average value of each pixel for a series of frames.

2 Materials

2.1 CCD-Based 1. Point Grey Research Chameleon monochrome camera

Detector equipped with a C-mount CCTV lens was used as the photo-
detector used to measure GFP (emission peak at 509 nm).
2. Pentax 12 mm f1.2 lens. Item # C61215KP (Spytown, Utopia,
NY).
3. Alternative lens: Tamron manual zoom CCTV 4–12 mm, f1.2
lens. Item # 12VM412ASIR (Spytown, Utopia, NY).
4. Green emission filter HQ535/50 m (Chroma Technology
Corp Rockingham, VT).
5. Blue excitation filter D486/20x (Chroma Technology Corp,
Rockingham, VT).
6. LED illumination box containing red, green, blue, and white
LEDs custom built by Luminousfilm (Shreveport, Louisiana,
www.luminousfilm.com/led.htm).
 Low-Cost Charged-Coupled Device (CCD) Based Detectors. . . 239

2.2 Computer 1. PC (laptop or desktop) with USB port.

Control and Data 2. Image analysis software ImageJ open source NIH software
Analysis (http://rsb.info.nih.gov/ij/download.html).
3. Excel (Microsoft, Redmond, WA) and SigmaPlot (SigmaPlot,
Ashburn, VA) data analysis software.

2.3 Biological and 1. Stx2 was obtained from Toxin Technology (Sarasota, FL).
Cell Culture Regents 2. Vero cells: African Green Monkey adult kidney cells (ATCC
CCL-81) were obtained from American Type Culture Collec-
tion (Manassas, VA).
3. HEK293 cells (ATCC CRL-1573) were obtained from Ameri-
can Type Culture Collection (Manassas, VA).
4. Dulbecco’s Modified Eagle’s Medium (DMEM) from Gibco
(Carlsbad, CA).
5. Fetal bovine serum (FBS) (Hyclone, Waltham, MA).
6. Fetal calf serum (FCS) (Hyclone, Waltham, MA).
7. Penicillin and streptomycin from Gibco (Carlsbad, CA).
8. Trypsin (Gibco).
9. 96-well plates (Greiner 655,090 obtained from sigma).

2.4 Plate Assay 1. Black 3.2 mm acrylic (Piedmont Plastics, Beltsville, MD).
Material 2. Clear 0.5 mm polycarbonate film (Piedmont Plastics, Beltsville,
MD).
3. 3 M 9770 Adhesive transfer Tape (Piedmont Plastics, Beltsville,
MD).
4. Epilog Legend CO2 65 W cutter (Epilog, Golden, CO).

3 Methods

3.1 Cell Culture Vero cells were maintained at 37  C in a CO2 incubator in Dulbec-
co’s Modified Eagle’s Medium (DMEM) containing 10% fetal
bovine serum and 100 units/ml of both penicillin and streptomy-
cin. Cells were trypsinized when ready to harvest.

3.2 Preparation of Since most of the virus remains associated with infected cells until
High-Titer Viral Stocks very late in the infection process, high-titer viral stocks were
prepared by concentrating the infected cells. Tissue culture flasks
(175 mm) were seeded with HEK293 cells in DMEM containing
10% FCS and 100 units/ml penicillin/streptomycin. When cells
reached 90% confluency, they were infected with the virus at a
multiplicity of infection (MOI) of 10. When the cytopathic effect
was nearly completed (after 48–72 h) and most of the cells were
rounded but not yet detached, the cells were harvested (they were
240 Reuven Rasooly et al.

easily dislodged by tapping). Cells were pelleted by centrifugation

at 800  g for 5 min at 4  C. Both cell pellet and supernatant were
collected. Since most progeny viruses remain cell-associated,
infected cells were disrupted by freeze–thaw cycles followed by
lysis as described below. To every cell pellet collected from 18 flasks
(175 mm), 18 ml of PBS/1 mM MgCl2/0.1% NP-40/1 mM
CaCl2, were added for hypotonic lysis followed by three rounds
of freeze–thawing. Crude lysates were then pelleted to remove
cellular debris by centrifugation at 9500  g for 20 min at 4  C.
The supernatants that contained the crude viral lysates were care-
fully removed. The viral supernatants were loaded onto CsCl step
gradients made by layering three densities of CsCl (1.25, 1.33, and
1.45 g/ml) and centrifuged at 50,000  g for 2 h in a Beckman
SW41 rotor at 14  C. The lower band containing the intact pack-
aged virus was removed, and a second step density gradient of
1.33 g/ml CsCl was centrifuged for 16 h at 48,000  g at 14  C.
The lower band containing the packaged virus was collected and
dialyzed against 10% glycerol in saline solution. The stock viral titer,
determined by plaque assays, was 1011 plaque-forming units/ml.

3.3 Plaque Assays Plaque assays depend on the ability of the adenovirus to propagate
for Purification and in HEK293 cells. Six 35 mm tissue culture plates were seeded with
Titration of the HEK293 cells. The cells were incubated at 37  C in a CO2 incuba-
Adenovirus tor until the cells were 90% confluent. Serial dilutions (10 8–10 13
of the adenovirus stock) were made in DMEM supplemented with
2% FBS. The diluted virus was added to the cells. After 2 h, the
medium was removed and replaced with 1 Modified Eagle’s
Medium and 1% sea-plaque agarose. The agar overlay was added
to keep the virus localized after the cells had lysed. After 5 days,
plaques were visible, and counted for titer determination after
7 days.

3.4 Cell Transduction The cells were transduced with adenovirus coding for GFP as
described in previous work [23] to produce label free cells that
express the GFP gene used for the assay. Vero cells were plated on
black wells 96-well plates at 1  104 cells in 100 μl of medium per
well. Cells were incubated overnight to allow time for cells to attach
to the plate and then the cells were transduced with Ad-GFP
multiplicity of infection (MOI) of 100.

3.5 Shiga Toxin Transduced cells were treated in triplicate with various amounts of
Activity Assay Shiga toxin for 48 h. The toxin activity is measured through a
decrease in fluorescence because the biologically active toxin inhi-
bits protein synthesis. The controls in these experiments are cells
without toxin. After incubation with toxin, the cell fluorescence
was imaged with a green emission filter, and the signals of the wells
were quantified by the CCD camera operating in a still single frame
mode. Various concentrations of the toxins were employed and the
effect of toxin on the cellular fluorescence was measured.
 Low-Cost Charged-Coupled Device (CCD) Based Detectors. . . 241

A B
1 400
w
2
300

Magnitude
3 w
200
4
5
100
B G R
0
6 350 400 450 500 550 600 650 700 750
Wave length (nm)

Fig. 1 Fluorescence detector. (A) The main system elements: CCD detector, LED multi-wavelength illuminator
and cell toxicity assay. The CCD detector components: (1) a camera, (2 ) lens, (3 ) emission filter mounted on
the end of the lens, (4 ) assay plate, (5 ) excitation filter, and (6 ) multi-wavelength LED. (B) The spectra
(measured by a spectrometer) for the (W ) white, (B ) blue, (g) green, and (R ) red LEDs

3.6 CCD Imager The basic configuration of the CCD detector platform is shown
schematically in Fig. 1a along with the spectra of the LED illumi-
nator. The system consists of three main elements (a) the CCD
detector module, (b) the assay chip module, and (c) the illumina-
tion/excitation module (see Note 1). All of these elements are
contained within a black acrylic box designed for portability, recon-
figurability, prevention of light loss and internal light reflection, and
to minimize extraneous signals from ambient light.

3.7 Illumination The multi-wavelength spatial LED illuminator module [20] shown
Modules in Fig. 1a6 comprises a custom built multi-wavelength LED illumi-
nation box with two different types of LED: (1) white, which
generates across the full optical spectrum, and (2) RGB, which
generates at the red, green, and blue wavelengths. The LED illumi-
nator contains four RGB LED strips and four white LED strips,
both 18 LEDs per foot. Each LED emission color is controlled
individually with a switch in the front panel. The top of the box
consists of a diffusion panel (milky white plastic panel), which
assures uniformity of the light and hence even illumination of the
sample chip. The dimensions of the diffusion panel are
8.5  5.5 cm. The LEDs are powered by a 3 V DC battery. It is
important to measure the illumination uniformity across the
measured surface (see Note 5).

3.8 Optical System The optical system includes a Pentax 12 mm f1.2 CCD lens. The
fixed focal length enables the use of a wider aperture. The C mount
fixed focal length lens was connected to the imager. The f1.2 lens
permits operation at very low light levels. To minimize green light
emission, which increases the noise and limits the detection
242 Reuven Rasooly et al.

sensitivity, a green emission filter HQ535/50 m was used (see Note

2). Similarly, a blue excitation filter (D486/20) was used to block
the blue light emission to the CCD and to enable the measurement
of the 523 nm light from the excited GFP (see Note 6).

3.9 Assembling the The assembly of the CCD-based fluorometer based on LED illu-
CCD Detector mination was described in previous work [17, 18, 20]. The CCD
imager (Fig. 1a1) is placed inside the light-sealed black acrylic
enclosure (see Note 4). The camera is mounted in the enclosure
with holes drilled in the top of the enclosure to enable air circula-
tion for cooling the CCD. The lens is mounted onto the camera
body (Fig. 1a2) and a green filter (EmF) is mounted on the lens
(Fig. 1a3). The LOC or detection plate (Fig. 1a4), excitation filter
(ExF) (Fig. 1a5), and the LED (Fig. 1a6) are interchangeable (see
Notes 1–7)

3.10 Light Unlike the monochromatic light emitted from lasers, the multi-
Characteristics of the wavelength LED illumination box provides illumination in the
Multi-Wavelength LED blue, green, red, and white ranges (Fig. 1b), covering a spectrum
of 450–650 nm (red 610–650 nm, green 492–550 nm, and blue
450–495 nm) as described in our previous work [20]. The only
region of the spectrum not represented by the RGB LEDs strip is
the range between 550 and 610 nm. The white LED provide broad
illumination spectra (440–680 nm with lower intensity in the range
472–510 nm). The multi-wavelength LED enables illumination in
narrower spectra by using only one of the R, G, B, or white LEDs,
or combinations of them. Within each LED spectrum, interchange-
able filters can be used to narrow the excitation band (see Note 6).

3.11 Image Images were captured using a camera incorporating a CCD sensor
Capturing with high quantum efficiency and response linearity as well as the
ability to generate high-quality, low noise images. The camera used
in this case employed a monochrome sensor, and fluorescent emis-
sion was detected through a green filter. The images were analyzed
and the data quantified with ImageJ [24, 25]. In PBS we observed
no photobleaching when the cells were exposed for less than 5 min.
The longest exposure time employed in our experiments was 20 s.
The average sensor background signal was subtracted from sample
images. This background signal was recorded by capturing images
with exposure and gain settings identical to those used to image
samples, but with a lens cap in place.

3.12 Image and Data The exposure time for the signal will depend on the type of illumi-
Analysis nation and the light emitted, which is determined empirically (see
Note 8). The spot intensity can be analyzed using any standard
image processing software. For our system, we chose the freeware
ImageJ developed at NIH (see Note 8), which enables 2D image
analysis of the intensity of each spot in a row or column (Fig. 2b), as
 Low-Cost Charged-Coupled Device (CCD) Based Detectors. . . 243

B
A
A
I + GFP II

III + IV B

V C

1 2 3
C D 40000
Average Pixel Value (ADU)

35000

30000
R² = 0.8596
25000

20000

A 15000

B 3 10000
C 1 2 0.000001 0.001 1
Shiga Toxin Concentration (ng/mL)

Fig. 2 Fluorescence detection of cell toxicity assay: (a) cell toxicity assay (I ) Vero cells transduced with
adenovirus with GFP (II ) the resultant cells that express the GFP (III ) emit light at 509 nm. A toxin effect on
Vero cells (IV ) will reduce light emission (V ) after the addition of the toxin (VI ) measured by CCD detector
shown in (b). (b) Vero-GFP cells response to low concentrations of Shiga toxin. Vero-GFP cells were treated
with various amounts of Shiga toxin. The signals were detected by the CCD operating in a still single frame
mode, the amount of toxin used 3A control no toxin, 3B 0.01 pg/mL, 3C 0.1 pg/mL, 2A 1 pg/mL, 2B 10 pg/mL,
2C 100 pg/mL, 1A 1 ng/mL, 1B 10 ng/mL, 1C 100 ng/mL. The corresponding ImageJ 3D image is shown in (c).
Average signal brightness (ADU) was plotted against the various Shiga toxin concentrations (d)

well as 3D (Fig. 2c) spatial analysis (2D of the spatial position of the
spots and the third dimension is the intensity of the spots), to
provide visual representation of the image as shown in Fig. 2c.
For spot analysis, the intensity value of every spot was exported to
244 Reuven Rasooly et al.

an Excel spreadsheet and to the scientific data analysis and graphing

software Sigmaplot, which was used to plot the data. Several analy-
sis methods were performed, including subtracting baseline noise
level and calculating the signal-to-oise ratio (S/N).

3.13 Assay Plate A 9-well plate (Fig. 1a6) was fabricated as described in our previous
Fabrication work [21, 26, 27], using black acrylic with one side coated by 3 M
9770 black adhesive transfer tape laser machined to have the
needed array of wells (see Note 10 and 11). A layer of thin polycar-
bonate sheet was attached to the adhesive transfer tape to form the
bottom of the sample wells. To minimize light reflection, the
polycarbonate material used for fabrication did not show detectable
autofluorescence, and all the materials used to fabricate the device
did not appear to inhibit enzymatic reactions. The assay plate
fabrication includes:
1. Micromachining: The fluidics was made of 3.2 mm black
acrylic, micromachined with a computer controlled 65 W Epi-
log Legend CO2 laser system (see Notes 9 and 10). The well
holes are 2 mm, which allows analysis of 20 μl samples. Smaller
holes make the loading of the sample less reproducible (see
Notes 10 and 11).
2. Bonding: The polycarbonate bottom was bonded to the acrylic
with double sided pressure-sensitive adhesive transfer tape (see
Notes 12 and 13) which was bonded to the micromachined
layer so there is no contact between the adhesive and the fluids.
Assembled devices were bonded together and processed with a
laminating machine to eliminate air bubbles and improve uni-
formity of the bonding.

3.14 Detection of Transduced cells were treated with various amounts of Shiga toxin
Active SHIGA Toxin in three replicates. As previously mentioned, the toxin activity is
Using CCD Based measured through decreases in fluorescence using controls without
Optical Detector toxin. The cells were cultured for 48 h and after incubation with
toxin the cell fluorescence was detected with the CCD detector.
The plate containing the cells was illuminated by the LED illumi-
nator using the blue LED with a blue excitation filter. The detec-
tion plate was imaged with a green emission filter, and the signals of
the wells were quantified by the CCD camera operating in a still
single frame mode (Fig. 2b). Various concentrations of the toxins
were employed and the effect of toxin on the cellular fluorescence
was measured as shown in Fig. 2b. The amount of toxin used was
well 3A: control no toxin, 3B: 0.01 pg/mL, 3C: 0.1 pg/mL, 2A:
1 pg/mL, 2B: 10 pg/mL, 2C: 100 pg/mL, 1A: 1 ng/mL, 1B:
10 ng/mL, 1C: 100 ng/mL. The corresponding ImageJ 3D image
is shown in Fig. 2c. Average signal brightness measured by the A/D
unit (ADU) was plotted against the various Shiga toxin concentra-
tions (Fig. 2d). The reduction in cell fluorescence signal is
 Low-Cost Charged-Coupled Device (CCD) Based Detectors. . . 245

proportional to the amount of toxin due to toxin inhibition of

protein synthesis, with the lowest signal at 100 ng/mL as shown
in Fig. 2b well 1C. In other wells with lower levels of toxin (1A:
1 ng/mL, 1B: 10 ng/mL), the level of fluorescence increases
slightly.
The image was analyzed and quantified using ImageJ software.
By this method it is possible to calculate the average value of each
pixel for a series of frames. Figure 2c illustrates the decreasing
fluorescence signal with increasing amount of the toxin. The aver-
age signal intensities (brightness) displayed as Average Pixel Value
(ADU count) of three replicates was plotted against the various
Shiga toxin concentrations (Fig. 2c) which shows decreased fluo-
rescence as the amount of toxin increased, suggesting that toxin
level can be measured as a function of decreased fluorescence as
shown in Fig. 2c. Another toxin (Clostridium difficile toxin B) was
used with this assay with no effect on fluorescence, which demon-
strates the specificity of the assay. A plot of the ADU versus Shiga
toxin concentration can be seen in Fig. 2d, where there is a nearly
linear variation of the fluorescence signal with toxin concentration.

3.15 Detection Limit To measure lower concentrations of toxin, three replicates of the
of Biologically Active transduced Vero-GFP cells were cultured for 24 h and treated with
Stx2 Using CCD various amounts of Shiga toxin (100 ng to 0.01 pg/mL). The
Measurements signals of the wells were detected by the CCD camera. The image
was quantified with ImageJ. The limit of detection (LOD) was
calculated as the mean pixel value in three control samples (cells
with no toxin) minus three times the standard deviation of these
samples. The plotted LOD is indicated as a horizontal dashed line
(Fig. 3). Concentrations of toxin which yielded a mean signal out of
three replicates that was under this limit were considered to be
detected. Unlike the usual LOD calculation, the standard deviation
was subtracted and not added because the toxin reduces signal.
Thus, the LOD was determined to be 0.1 pg/mL, which is similar
to a commercial fluorimeter.

3.16 Factors There are several factors contributing to improve sensitivity of

Contributing to System mobile low cost optical detection:
Sensitivity
1. Detector: While camera phone lenses are not interchangeable,
many webcams permit lenses to be changed (we used an f/1.2
lens to maximize the amount of light transmitted to the sensor.
Using cooled CCS/CMOS devices to reduce thermal noise
and improve SNR. In general, monochrome sensors are capa-
ble of higher detail and sensitivity than similar color sensors.
Monochrome sensors capture all incoming light at each pixel—
regardless of color. Each pixel therefore receives up to 3 more
light.
246 Reuven Rasooly et al.

40000

35000

30000
Average Pixel Value (ADU)

25000

20000

15000

10000

5000

0
0.00000010.000001 0.00001 0.0001 0.001 0.01 0.1 1 10 100
Shiga Toxin Concentration (ng/mL)

Fig. 3 Limit of Shiga toxin detection (LOD). Vero-GFP cells were treated with various amounts of Shiga toxin
(100 ng to 0.01 pg/mL). The signals of the wells were detected by the CCD camera. The image was
quantitated with ImageJ The level of detection (LOD) was calculated as the mean pixel value in three control
samples (cells with no toxin) minus three times the standard deviation of these samples. The LOD is indicated
as a horizontal dashed line. Concentrations of toxin which yielded a mean signal out of three replicates that
was under this limit were considered to be detected

2. Illumination source: Increasing the power of the excitation

source in fluorescent detection increases florescence emission.
Increasing the intensity of the LED illumination (i.e., the use of
more LEDs or using more powerful LEDs) increases the fluo-
rescent signal. Using color LEDs enables illumination in the
range of florescence excitation which may reduce noise and
improve SNR (Fig. 1). The use of low cost lasers equipped
with line generator or removing the laser lens may increase
light intensity and provide narrow wavelength illumination.
3. Exposure time: For single frame imaging, cameras and some
webcams allow for long exposure times (>1 s). Longer expo-
sures can be used to detect low signal intensities; however,
longer exposure will also increase image noise in high noise
detectors.
4. Video imaging: The use of video imaging mode combined with
the image stacking [24] computational approach improves
SNR.
5. Filters: The quality of filters is critical. Using high quality
narrow band filters at the emission/excitation wavelengths
reduces noise and improves detection.
6. Assays: For immunoassays it is possible to increase primary
antibody immobilization by increasing surface area for anti-
body binding using high surface area nanoparticles for
 Low-Cost Charged-Coupled Device (CCD) Based Detectors. . . 247

antibody binding such as gold nanoparticles [28] or carbon

nanotubes [29, 30].
7. Plate material: For reducing light noise, the use of a light
absorbing polycarbonate material combined with the compact
design of the device enclosure minimizes light losses and cross
talk.

4 Notes

1. Make sure the lens, filters, and sample plate are vertically cen-
tered and aligned.
2. Make sure the arrows on the coated filters are facing the CCD.
3. The filters should be leveled
4. The filters should be always in the same position.
5. The imager enclosure must be light sealed without any light
leaks and use light reflecting material. A long exposure (e.g.,
3 min) can be used to detect light leaks.
6. To measure light uniformity, it is important to measure the
light with a long enough exposure time (e.g., 60 s) to measure
the CCD uniformity when all of the wells are loaded with the
same GFP fluorescence sample. If the lighting is not uniform,
correction values for each well can be calculated and use for
measurement composition.
7. To measure the effectiveness of the filters, it is recommended to
perform two long exposures (e.g., 3 min) without the assay
plate, one with the LED on and one with the LED off. Ideally,
the two measurements should be very similar (the blue filters
pass only blue light which is blocked by the green filters). The
difference between measurements may suggest that the blue
filters do not block all green light and/or the green filters do
not block all blue light.
8. For data analysis, use the ImageJ high contrast visualization,
which will not affect the values but will enable easy selection of
the spots
9. A fully open aperture (e.g., f1.2) will enable shorter exposure
time, but because of the limited depth of field, focusing will be
more limited and the image less sharp. The laser power and
speed for cutting polymers has to be determining empirically. It
is recommended to use the minimum laser power to reduce
overheating or burning the material.
10. The “holes” for the wells on the assay plate are 2 mm in
diameter, which allows analysis of 20 μL samples. Smaller
holes make the loading of the sample less reproducible because
248 Reuven Rasooly et al.

the fluid meniscus within the wells causes light diffraction,

which complicates quantification.
11. For strong bonding, remove 1 cm of the adhesive tape cover,
align the tape with acrylic surface, and attached the exposed
tape to the acrylic surface. With a ruler, press the tape to the
surface and slowly move the ruler across the tape with little
pressure in order to prevent air bubbles. When bonding the
tape, it should be aligned with the acrylic. For assembling of
the assay plate, remove the protective cover from the other side
of the double side adhesive (taped to the acrylic) and remove
the protective cover from the polycarbonate. Then, align the
two pieces, apply pressure, and run through the assay plate and
attach the polycarbonate to the acrylic.
12. Controlling optical noise: Fluorescence emission and scattered
excitation light can propagate through the chip, causing cross
talk between adjacent channels. This can become a major
source of optical noise in the system [18, 31, 32] by increasing
background noise, thus reducing the sensitivity of the measure-
ments. To limit the effect of fluorescence background, PC, and
not Mylar, which is the commonly used material for lamination
based fabrication, was used as the main fabrication material due
to its lower fluorescence background [18]. Using black mate-
rial decreases the noise.

References

1. Ligler FS, Taitt CR, Shriver-Lake LC, Sapsford monitoring ions and metabolites 2: color
KE, Shubin Y, Golden JP (2003) Array biosen- responses to pH, K+ and glucose. Anal Sci
sor for detection of toxins. Anal Bioanal Chem 22:937–941
377:469–477 7. Feldstein MJ, Golden JP, Rowe CA, Maccraith
2. Roda A, Manetta AC, Portanti O et al (2003) A BD, Ligler FS (1999) Array biosensor: optical
rapid and sensitive 384-well microtitre format and fluidics systems. Biomed Microdevices
chemiluminescent enzyme immunoassay for 1:139–153
19-nortestosterone. Luminescence 18:72–78 8. Sohn YS, Goodey A, Anslyn EV, McDevitt JT,
3. Svitel J, Surugiu I, Dzgoev A, Ramanathan K, Shear JB, Neikirk DP (2005) A microbead
Danielsson B (2001) Functionalized surfaces array chemical sensor using capillary-based
for optical biosensors: applications to in vitro sample introduction: toward the development
pesticide residual analysis. J Mater Sci Mater of an “electronic tongue”. Biosens Bioelectron
Med 12:1075–1078 21:303–312
4. Liu Y, Danielsson B (2007) Rapid high 9. Knecht BG, Strasser A, Dietrich R, Martlbauer
throughput assay for fluorimetric detection of E, Niessner R, Weller MG (2004) Automated
doxorubicin—application of nucleic acid-dye microarray system for the simultaneous detec-
bioprobe. Anal Chim Acta 587:47–51 tion of antibiotics in milk. Anal Chem
5. Burkert K, Neumann T, Wang J, Jonas U, 76:646–654
Knoll W, Ottleben H (2007) Automated prep- 10. Balsam J, Bruck HA, Rasooly A (2015)
aration method for colloidal crystal arrays of Smartphone-based fluorescence detector for
monodisperse and binary colloid mixtures by mHealth. Methods Mol Biol 1256:231–245
contact printing with a pintool plotter. Lang- 11. Balsam J, Rasooly R, Bruck HA, Rasooly A
muir 23:3478–3484 (2014) Thousand-fold fluorescent signal
6. Tohda K, Gratzl M (2006) Micro-miniature amplification for mHealth diagnostics. Biosens
autonomous optical sensor array for Bioelectron 51:1–7
 Low-Cost Charged-Coupled Device (CCD) Based Detectors. . . 249

12. Balsam J, Bruck HA, Rasooly A (2014) 23. Rasooly R, Do PM (2010) Shiga toxin Stx2 is
Webcam-based flow cytometer using wide- heat-stable and not inactivated by pasteuriza-
field imaging for low cell number detection at tion. Int J Food Microbiol 136:290–294
high throughput. Analyst 139:4322–4329 24. Balsam J, Bruck HA, Kostov Y, Rasooly A
13. Balsam J, Bruck HA, Rasooly A (2013) Ortho- (2012) Image stacking approach to increase
graphic projection capillary array fluorescent sensitivity of fluorescence detection using a
sensor for mHealth. Methods 63:276–281 low cost complementary metal-oxide-semicon-
14. Balsam J, Ossandon M, Bruck HA, Lubensky I, ductor (CMOS) webcam. Sens Actuators B
Rasooly A (2013) Low-cost technologies for 171–172:141–147
medical diagnostics in low-resource settings. 25. Balsam J, Ossandon M, Bruck HA, Rasooly A
Expert Opin Med Diagn 7:243–255 (2012) Modeling and design of microma-
15. Sapsford KE, Taitt CR, Loo N, Ligler FS chined optical Soller collimators for lensless
(2005) Biosensor detection of botulinum tox- CCD-based fluorometry. Analyst
oid A and staphylococcal enterotoxin B in 137:5011–5017
food. Appl Environ Microbiol 71:5590–5592 26. Sapsford KE, Francis J, Sun S, Kostov Y,
16. Golden JP, Floyd-Smith TM, Mott DR, Ligler Rasooly A (2009) Miniaturized 96-well
FS (2007) Target delivery in a microfluidic ELISA chips for staphylococcal enterotoxin B
immunosensor. Biosens Bioelectron detection using portable colorimetric detector.
22:2763–2767 Anal Bioanal Chem 394:499–505
17. Kostov Y, Sergeev N, Wilson S, Herold KE, 27. Sun S, Yang M, Kostov Y, Rasooly A (2010)
Rasooly A (2009) A simple portable electrolu- ELISA-LOC: lab-on-a-chip for enzyme-linked
minescence illumination-based CCD detector. immunodetection. Lab Chip 10:2093–2100
Methods Mol Biol 503:259–272 28. Yang M, Kostov Y, Bruck HA, Rasooly A
18. Sapsford KE, Sun S, Francis J, Sharma S, Kostov (2009) Gold nanoparticle-based enhanced
Y, Rasooly A (2008) A fluorescence detection chemiluminescence immunosensor for detec-
platform using spatial electroluminescent exci- tion of Staphylococcal enterotoxin B (SEB) in
tation for measuring botulinum neurotoxin A food. Int J Food Microbiol 133:265–271
activity. Biosens Bioelectron 24:618–625 29. Yang M, Kostov Y, Bruck HA, Rasooly A
19. Higgins JA, Nasarabadi S, Karns JS et al (2003) (2008) Carbon nanotubes with enhanced
A handheld real time thermal cycler for bacte- chemiluminescence immunoassay for CCD-
rial pathogen detection. Biosens Bioelectron based detection of Staphylococcal enterotoxin
18:1115–1123 B in food. Anal Chem 80:8532–8537
20. Sun S, Francis J, Sapsford KE, Kostov Y, 30. Yang M, Kostov Y, Rasooly A (2008) Carbon
Rasooly A (2010) Multi-wavelength Spatial nanotubes based optical immunodetection of
LED illumination based detector for in vitro Staphylococcal enterotoxin B (SEB) in food.
detection of botulinum neurotoxin A activity. Int J Food Microbiol 127:78–83
Sens Actuators B 146:297–306 31. Irawan R, Tjin SC, Yager P, Zhang D (2005)
21. Sun S, Ossandon M, Kostov Y, Rasooly A Cross-talk problem on a fluorescence multi-
(2009) Lab-on-a-chip for botulinum neuro- channel microfluidic chip system. Biomed
toxin a (BoNT-A) activity analysis. Lab Chip Microdevices 7:205–211
9:3275–3281 32. Hawkins KR, Yager P (2003) Nonlinear
22. Rasooly R, Balsam J, Hernlem BJ, Rasooly A decrease of background fluorescence in poly-
(2015) Sensitive detection of active Shiga toxin mer thin-films – a survey of materials and how
using low cost CCD based optical detector. they can complicate fluorescence detection in
Biosens Bioelectron 68:705–711 microTAS. Lab Chip 3:248–252
 Chapter 16

Smartphone-Enabled Detection Strategies for Portable

PCR–Based Diagnostics
Aashish Priye and Victor M. Ugaz

Abstract
Incredible progress continues to be made toward development of low-cost nucleic acid-based diagnostic
solutions suitable for deployment in resource-limited settings. Detection components play a vitally impor-
tant role in these systems, but have proven challenging to adapt for operation in a portable format. Here we
describe efforts aimed at leveraging the capabilities of consumer-class smartphones as a convenient platform
to enable detection of nucleic acid products associated with DNA amplification via the polymerase chain
reaction (PCR). First, we show how fluorescence-based detection can be incorporated into a portable
convective thermocycling system controlled by a smartphone app. Raw images captured by the phone’s
camera are processed to yield real-time amplification data comparable to benchtop instruments. Next, we
leverage smartphone imaging to achieve label-free detection of PCR products by monitoring changes in
electrochemical reactivity of embedded metal electrodes as the target DNA concentration increases during
replication. These advancements make it possible to construct rugged inexpensive nucleic acid detection
components that can be readily embedded in a variety of portable bioanalysis instruments.

Key words PCR, Smartphone, Point of care, Label free detection, Mobile health care, Image analysis

1 Introduction

The emerging demand for sophisticated medical diagnostic tools

that can be readily deployed in resource limited settings has focused
renewed attention on the need for advanced bio-analytical detec-
tion techniques. Next-generation instruments based on these tech-
nologies promise to catalyze a revolutionary shift away from
dedicated laboratory-based approaches toward field-deployable
systems that are compact, portable, inexpensive, and easy to use.
But existing methods employed to perform post-analytical evalua-
tion predominantly involve optical analysis (e.g., microscopy, fluo-
rometry, colorimetry, cytometry) whose cost and complexity
severely constrain the development of portable systems. Smart-
phones are poised to provide a versatile platform that can enable
many of these limitations to be overcome. In addition to their

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_16,
© Springer Science+Business Media LLC 2017

251
252 Aashish Priye and Victor M. Ugaz

ubiquity for voice communication, today’s mobile devices embed a

robust array of sensors (microphones, accelerometers), wireless
data transmission protocols (Wi-Fi, Bluetooth), navigation and
positioning tools (global positioning system (GPS)), and perhaps
most significantly high-resolution CMOS (complementary metal–-
oxide semiconductor) optical sensors associated with their built-in
cameras. The high sensitivity of these CMOS imaging systems,
coupled with advanced processing capabilities, makes smartphones
an attractive standalone platform to support development of porta-
ble analytical systems.
Incredible progress has already been made toward leveraging
these advantages to position smartphones as an alternative to
benchtop microscopes in biosensing applications [1–5]. Many of
these efforts have focused on harnessing the small footprint and
versatility of smartphones to produce custom devices tailored
toward microscopy [6, 7], spectroscopy [8], single molecule analy-
sis [9], colorimetry [10–13], paper-based microfluidics [14, 15],
and label-free detection [16, 17]. In this chapter we describe recent
efforts by our research group to harness the CMOS imaging cap-
abilities embedded in ordinary consumer-class smartphones for
development of real-time fluorescence- and electrochemical-based
detectors for bioanalytical applications involving DNA replication
via the polymerase chain reaction.
The polymerase chain reaction (PCR) is considered a gold-
standard nucleic acid-based detection assay owing to its favorable
combination of specificity and sensitivity. The assay protocol
involves cyclically heating and cooling reagents through distinct
temperature zones (denaturing (95
C), annealing (55–65
C),
and extension (72
C)) to produce exponential amplification of a
specific target sequence within a template nucleic acid strand
(Fig. 1a). Such thermal cycling protocols are conventionally carried
out in “thermocycler” instruments that are quite slow (timescales
of ~1 h or more) and power intensive. Alternatively, thermal
convection-driven techniques can overcome these limitations by
applying a temperature gradient to drive microliter-scale convective
flow that transports reagents through the denaturing, annealing
and extension temperature zones passively [18–24] (Fig. 1a). Thus,
convective PCR offers pseudo isothermal operation by housing the
reagents in a confined cylindrical enclosure subjected to a tempera-
ture gradient (Fig. 1b).
Despite the inherent simplicity of convective PCR format, the
resulting micro-scale flows are quite intricate [19, 20, 25]. A spec-
trum of flow states (static fluid, periodic, quasi-periodic, chaotic,
and transient flow) can be accessed by tuning the cylinder geometry
(aspect ratio; h/d; d is the diameter of the cylinder) and the thermal
driving force, i.e., Rayleigh number (Ra ¼ [gβ(T2  T1)h3]/να,
where β is the fluid’s thermal expansion coefficient, g is the gravita-
tional acceleration, T1 and T2 are the temperatures of the top (cold)
 Smartphone-Enabled Detection Strategies for Portable PCR–Based Diagnostics 253

Fig. 1 Microscale convective flow states enable convective PCR. (a) Thermocycling is performed in a
cylindrical reactor whose top and bottom surfaces are maintained at different fixed temperatures, establishing
a flow field that circulates PCR reagents sequentially through denaturing, annealing, and extension conditions.
(b) Reagents are loaded in cylindrical wells machined into polycarbonate rods and sealed, after which the
reactor is subjected to a vertical temperature gradient. (c) A 3D computational fluid dynamics model coupled
with PCR kinetics yields predictions of DNA replication time scales expressed in terms of a characteristic
generation rate (i.e., DNA doubling events per hour) in a Ra-h/d parametric space. The parametric map reveals
a zone in the chaotic flow regime where accelerated DNA replication is stably achievable over a span of two
orders of magnitude in Ra. Insets originating from the map depict that the convective flow field has a strong
chaotic component in lower aspect ratio geometries as compared to higher aspect ratio geometries (quasi-
periodic flow trajectories). Trajectories were obtained from 3D CFD simulations performed over 5 min of
convective flow with a vertically imposed temperature gradient of 40
C

and bottom (hot) surfaces, respectively, h is the height of the

cylindrical reactor, α is the thermal diffusivity, and ν is the kinematic
viscosity). Furthermore, PCR kinetics are strongly coupled with the
underlying convective flow state [19]. For example, PCR is exe-
cuted most robustly in a broad chaotic flow regime (spanning two
orders of Ra) in Ra–h/d parametric map where the fluid elements
exhibit 3D chaotic trajectories as opposed to at higher aspect ratio
geometries where the fluid elements exhibit more 2D periodic
trajectories (Fig. 1c). These results provide crucial insight into
design principles for convective PCR reactors as they can function
universally in the chaotic flow regime encompassing all possible
combinations of temperature and reactor volume associated with
realistic PCR conditions.
254 Aashish Priye and Victor M. Ugaz

2 Materials

2.1 Convective PCR 1. Convective PCR reactors/tubes: holes are machined by drilling
Experiments into polycarbonate rods (Amazon Supply; Brand name: Small
Parts; Part #: PlasticRod137).
2. Bovine serum albumin (cat. no. A2153; Sigma-Aldrich).
3. Rain-X Anti-Fog (SOPUS Products).
4. SYBR Green PCR Master Mix (2; Cat. no. 4309155, Life
Technologies).
5. AmpliTaq Gold DNA polymerase (included in the SYBR Green
PCR Master Mix).
6. Forward and Reverse primers (10 μM; 50 -
0 0
CTGAGGCCGGGTTATTCTTG-3 and 5 -CGACTGGC-
CAAGATTAGAGA-30 Integrated DNA Technologies).
7. λ-phage template DNA (1 μg/mL; Cat. no. N3011S, New
England Biolabs).
8. PCR grade water.
9. Aluminum tape (cat. no. PCR-AS-200; Axygen, Inc.) and clear
tape.
10. A smartphone with PCR to go app.
11. Clip on mini-microscope attachment for the phone (60 mini
pocket microscope for iPhone 4; supply).
12. Portable battery pack (Vinsic Tulip battery charger).
13. Optical filter (THORLABS; Ø25 mm OG515 Colored Glass
Filter; Part # FGL515).
14. Microcontroller (Atmega328).
15. Ceramic resistor (Wire wound 10 Ω; Mouser electronics).
16. Temperature sensor (TMP35; Mouser electronics).
17. 5 mm round RGB LED, 5 mW.
18. RN41 Bluetooth module (Microchip).
19. Resistors and capacitors.
20. Polydimethylsiloxane (PDMS).

2.2 Electrode 1. Glass microscope slides (cat. no. 12-550-A3, Fisher).

Dissolution 2. Dry transfer film (Press-n-Peel Blue, Techniks, Inc.).
Experiments
3. Metal etchant (Gold Etchant TFA, Transene).
4. Acetone.
5. Cyanoacrylate adhesive.
6. Polycarbonate cylindrical cells (item 1 in Subheading 2.1).
7. Conductive tape (xyz-axis Electrically Conductive Tape, 3 M).
 Smartphone-Enabled Detection Strategies for Portable PCR–Based Diagnostics 255

8. DC power supply (E3612A, HP/Agilent).

9. Aluminum tape.

3 Methods

3.1 Smartphone- PCR amplification is typically detected by a fluorescent signal pro-

Enabled Analysissee duced by an intercalating dye or fluorophore-labeled probe whose
Smartphones, PCR of signal can be used to monitor the target sequence concentration
Real-Time PCR over time. Adaptation of PCR-based analysis to a portable format
Fluorescence: Design has proven challenging due to the electrical power requirements
Considerations associated with temperature cycling, and a number of innovations
have been made to overcome these issues [26–29]. But while these
approaches simplify the thermal cycling apparatus, quantitative
analysis of the amplified products still relies heavily on expensive
optical detection components. A typical PCR detection unit incor-
porates an excitation source, optical focusing lenses, and a detector
(e.g., photo multiplier tube (PMT), charged-coupled device
(CCD) camera, photo-diode) to enable fluorescence quantifica-
tion. Fluorescence data recorded from the PCR sample as a func-
tion of time are then analyzed and processed to yield a real-time
PCR amplification curve.
To explore feasibility of performing portable real-time PCR
fluorescence analysis, we developed a smartphone application
(app) that integrates with a portable thermocycling instrument
developed in our laboratory [30]. Instrument portability is
enhanced by adopting a convective PCR format whereby a temper-
ate gradient established between opposing surfaces establishes a
microscale convective flow field that can be harnessed to drive
DNA replication [18, 19, 31]. In this way, PCR can be actuated
isothermally by maintaining a single heater at a constant tempera-
ture, drastically reducing electrical consumption to a level that can
be supplied by standard 5 V USB sources that power consumer
mobile devices. A smartphone is able to wirelessly integrate with
the heating and detection module via Bluetooth, creating a seam-
less interface between the peripheral device components (heater,
LED illumination source) and a smartphone app (Fig. 2a). Tem-
perature settings can be controlled through the app, with thermal
management achieved by using off the shelf ceramic resistors that
convert electrical current to heat via joule heating (essentially a
simple mini-hot plate). A temperature sensor regulates the applied
current via a feedback loop programmed to maintain the tempera-
ture at 95
C (the PCR denaturing setpoint) using an Arduino
microcontroller (Fig. 2b). A polydimethylsiloxane (PDMS) casing
around the resistor provides thermal insulation and directs heat
transfer to the reactor.
256 Aashish Priye and Victor M. Ugaz

Fig. 2 A portable convective PCR setup. A smartphone enabled portable convective PCR device. (a) Convective
PCR reactor is placed on an isothermal mini hotplate and is illuminated with an LED excitation source to emit
fluorescence signal from the reactor which is then captured by the phone camera for further analysis. (b) The
phone interfaces with the device’s peripheral components (isothermal heater and LED source) wirelessly via
Bluetooth module coupled with an Arduino microcontroller via the RX and TX lines (Universal asynchronous
receiver/transmitter) (c) Device in operation

1. A simple convective thermocycling device consists of inter-

changeable polycarbonate reaction chambers (Fig. 1b) and a
ceramic mini hot plate that functions as an isothermal heater
(Fig. 2a).
2. Cylindrical reactor wells are embedded by machining holes in
polycarbonate rods, with different combinations of hole diam-
eter and plastic sheet thickness employed to achieve the desired
aspect ratios (height/diameter ¼ h/d).
3. The height of the reactor is chosen such that at steady state, the
temperature of the top surface is maintained at the annealing
temperature (55–65
C) and the diameter is chosen such that
geometry lies in the chaotic flow regime to enable robust
convective PCR. Based on these design considerations the
final reactor geometry in the results presented here are:
h/d ¼ 5, h ¼ 10 mm and d ¼ 2 mm.
4. The bottom surface of the reactor is heated using a mini hot
plate constructed from two wire wound ceramic resistors
(10 Ω) connected in series. The heaters were encapsulated
using poly (dimethyl siloxane) (PDMS) to provide insulation.
Smartphone-Enabled Detection Strategies for Portable PCR–Based Diagnostics 257

A temperature probe (tmp35) senses the temperature of the

heater, and the signal is sent to one of the analog pins in the
microcontroller (Arduino UNO R3). This analog reading is
converted into temperature which was continuously monitored
to operate the heaters isothermally at 95
C by actuating the
pulse width modulation (PWM) pin of the ATMEGA micro-
processor. The PWM pin outputs 100% of its signal until the
temperature of the heater reaches the setpoint, and reduces to
85% thereafter. If the temperature exceeds 97
C, the PWM
signal reduces to zero. In our testing, this simple scheme was
sufficient to maintain the heater at 95  2
C.
5. The reactor is placed on the isothermal mini hot plate surface
and a 5 mW blue LED excitation source connected to low
dropout (LDO) regulator illuminates the reactor from the
side (Fig. 2a).
6. The isothermal heater and LED excitation source can be oper-
ated wirelessly through the smartphone which can communi-
cate to a Bluetooth module linked to the Arduino
microcontroller. The incorporated Bluetooth module reports
the isothermal heater temperature and LED state. The smart-
phone app can establish a RX/TX connection (universal asyn-
chronous receiver/transmitter at a baud rate of 9600 bits/s with
no parity setting) via the inbuilt iOS core Bluetooth framework
and send commands to the services advertised by the peripheral
device to control the heater and the LED source.
7. The final circuit and all the components are soldered on a
printed circuit board which is housed in a plastic case designed
using FreeCAD and printed using a MakerGear M2 3D printer.
8. Reactors are first rinsed with a 10 mg/mL aqueous solution of
bovine serum albumin followed by Rain-X Anti-Fog, and
dried (see Note 1).
9. Convective PCR experiments are performed to replicate a 237
base pair target from a λ-phage DNA template using the fol-
lowing protocol.

Reagent Volume (μL)

SYBR Green PCR Master Mix (2) 50
Forward primer (10 μM) 10
Reverse prime (10 μM) 10
λ-phage template DNA (1 μg/mL) 1
Water, molecular biology grade 29
258 Aashish Priye and Victor M. Ugaz

10. The lower surface of the polycarbonate reactor is sealed

using aluminum tape, after which PCR reagent mixture
(from step 9) is pipetted inside and the top surface sealed
with another layer of clear tape (see Note 2).
11. After loading the reagents, the convective PCR reactor is
placed on the preheated ceramic heater based mini hot plate
maintained at 95
C for 20 min.
12. The smartphone app is programmed to actuate an LED exci-
tation source that illuminates the sample, and the top of the
PCR reactor is imaged using the smartphone camera to quan-
tify fluorescence at regular time intervals.

3.2 PCR to Go: A The PCR to Go application has been developed for both iOS (Xcode
Fluorescence Analysis 5.0) and Android (Android studio) platforms. It provides a user
Smartphone App friendly GUI that sends commands to the detector/heater system
and seamlessly switches between various in-app functions.
1. Once the app is launched it automatically pairs with the periph-
eral Bluetooth module of the detector system. The user can
then set the heater and LED operation parameters.
2. The analysis can either be carried out in end point or real time
detection modes. In each case, the camera exposure time and
focal plane settings are manually fixed with the sample placed
~5 cm away from the camera lens (see Note 3).
3. A clip on mini microscope attachment (TOMTOP Mini Pocket
Microscope Magnifier Jewelers Loupe for iPhone 4 4S with
LED Light (Amazon.com)), supplemented with the ability to
digitally zoom images via the app’s affine transformation frame-
work, enables a total of 60 magnification to resolve an area of
~1 mm2 without pixel degradation (Fig. 3).
4. In real-time detection mode, the user inputs additional para-
meters such as the frequency of image acquisition and the
number of images to be captured. The excitation LED is illu-
minated for 5 s during which the app synchronizes image
capture.
5. Raw image data from the CMOS sensor are stored as bitmap
data files containing 4 bits per pixel (one each for red, blue,
green, and alpha values) corresponding to standard RGB color
space (sRGB).
6. The user can then define a region of interest within the
acquired image using touch gestures.
7. When the analyze button is tapped, the app extracts the
corresponding sRGB byte matrix from the region of interest.
The alpha channel, which measures transparency, is discarded
as it is constant in all images.
 Smartphone-Enabled Detection Strategies for Portable PCR–Based Diagnostics 259

Fig. 3 Smartphone-based fluorescence analysis workflow. RGB bitmap data from the cropped raw images first
undergoes a gamma transformation. The green channel intensity of the pixel matrix is then extracted to
analyze the average fluorescence intensities. In end point detection mode pre- and post-PCR images are
compared, while in real time detection mode averaged intensities from a series of acquired images are fit
using a nonlinear sigmoid function to obtain a qPCR curve

8. A gamma transformation is applied to the RGB matrix with a

gamma value of 2.2 to account for the nonlinear intensity scale
of the CMOS sensor [32].
9. The red and blue channels’ 8 bit intensity values are discarded,
yielding a green channel pixel matrix that best represents the
fluorescence signal.
10. The built-in nonlinear solver is invoked to fit the averaged
green channel intensity over time using a four parameter sig-
moid function via the Marquardt Levenberg algorithm
[33, 34].

I max
I ¼  þ Io
ðtt 1=2 Þ=k
1þe

where Imax and Io represent the maximum and background

fluorescence intensities, respectively. The quantities t1/2 and k
260 Aashish Priye and Victor M. Ugaz

Fig. 4 Smartphone-based quantitative fluorescence detection (a) The PCR to Go app presents a simple GUI to
control image acquisition, processing, and data analysis (https://itunes.apple.com/us/app/pcr-to-go/
id909227041?mt¼8). (b) Smartphone-based quantification of a 237 bp target sequence from a λ-phage
DNA template is demonstrated for three initial copy numbers ([DNA]0 ¼ 105, 104, and 103 copies/μL). (c)
Sigmoidal fits are applied to the smartphone acquired real time data, and a standard curve is constructed
using reaction times when fluorescence exceeds a threshold value of 20 units (inset, CT ¼ 9.4, 11.8, and
13.3 min for starting DNA concentrations of [DNA]0 ¼ 105, 104, and 103 copies/μL, respectively). (d) A
benchtop real time PCR instrument (LightCycler 96, Roche) generates comparable results with a nearly
identical standard curve

express the time required for intensity to reach half of the

maximum value and slope of the curve, respectively, and are
used as fitting parameters.
11. The fitting algorithm also allows background fluorescence
subtraction yielding a constant baseline. The initial guess for
the curve fitting iterations is determined from the parameters
of the first image. Critical threshold reaction times are deter-
mined from the point where the sigmoidal fit exceeds a user-
defined threshold value.
12. Finally, an in-app plotting feature enables results to be viewed
graphically or exported as a .csv file for further analysis, along
with a time stamp and GPS location of the run for archival.
To demonstrate the utility of our smartphone app, we applied it
to analyze fluorescence during a PCR reaction run using a conven-
tional SYBR Green reagent chemistry (Fig. 4a). Results were
obtained by replicating a 237 bp target sequence from a λ-phage
DNA template over a series of initial copy numbers ranging from
103 to 105 (Fig. 4b). An in-app algorithm fits the recorded fluores-
cence data into a four parameter sigmoidal curve and extracts a
critical reaction time to exceed a threshold value of 20 fluorescence
 Smartphone-Enabled Detection Strategies for Portable PCR–Based Diagnostics 261

units—a time scale analogous to the critical cycle number obtained

when real-time PCR is performed in a conventional thermocycler
(Fig. 4c). These data can be used to construct a standard curve
whose slope yields an effective doubling time of 35.8 s. For com-
parison, we evaluated the same series of reactions in a benchtop
real-time PCR instrument (LightCycler 96, Roche), (Fig. 4d) and
converted the convective data in Fig. 4c (expressed in units of
reaction time) to equivalent cycle numbers by assuming that one
conventional replication cycle occurs during the 35.8 s doubling
time.
Both approaches agree remarkably well, yielding virtually iden-
tical standard curves (insets Fig. 4c, d). Our adaptation of
smartphone-based fluorescence detection therefore enables quan-
titative analysis at template concentrations down to 1000 copies/μ
L, a level sufficient for many diagnostic scenarios (i.e., depending
on the severity of an infection), and particularly impressive consid-
ering the simplified instrument format. It is also envisioned that, in
practice, each PCR reactor will include a calibration target on the
surface and multiple reactor cells that enable parallel control reac-
tions to be simultaneously performed.

3.3 Detection Via An alternative approach to leverage smartphone imaging, poten-

Electrochemical tially useful for detection of unlabeled biomolecules, involves mon-
Dissolution itoring electrochemical reactivity of metal electrodes and its
dependence on the concentration of DNA in solution. We previ-
ously demonstrated this approach in a device that exploited the
electrochemical reactivity of chromium to enable visual detection of
PCR products [17]. This design consisted of a microfabricated
electrode array patterned on a silicon substrate and bonded to an
etched glass microchannel, yielding a cross section with electrodes
spanning the floor of an enclosed chamber (Fig. 5a). Electrode
dissolution is monitored by recording the change in reflectivity
with time under oblique illumination with ordinary white light,
from which the average intensity within a region of interest cen-
tered on the electrode can be evaluated. In this way, changes in
dissolution rate in response to the bulk solution’s chemical compo-
sition can be readily distinguished (Fig. 5b). We demonstrated
feasibility of performing this detection using a smartphone camera
outfitted with an inexpensive mini-microscope attachment
(Fig. 5c). Fabrication of these devices is briefly summarized below:
1. Silicon wafers (P(100), 500 μm thick, 15 cm diameter, 5000 Å
oxide layer; University Wafer) are cleaned in a reactive ion
etcher.
2. Wafers are spin coated with hexamethyldisilazane (J.T. Baker)
followed by a positive photoresist (Shipley 1827; Rohm &
Haas).
262 Aashish Priye and Victor M. Ugaz

Fig. 5 Label-free detection via electrode dissolution. (a) Microdevice design incorporating a glass micro-
channel (275  45 μm cross section) bonded to a Si substrate patterned with a Cr microelectrode array. When
a 2.5 V potential is applied, the Cr electrodes electrochemically dissolve at a rate dependent on the chemical
composition of the bulk solution (arrows). (b) The evolution of reflected light intensity from the active electrode
displays composition-dependent behavior. The upper panel compares dissolution kinetics of histidine buffer
(50 mM, pH 7.6), a 100 bp dsDNA ladder (20 μg/mL in 50 mM histidine), a reducing agent (0.06% w/v Na2S2O4
in 50 mM histidine), and at low pH (50 mM histidine, pH 4.4). Pre- and post-reaction PCR reagent mixtures are
compared in the lower panel. (c) Electrodes are large enough to enable dissolution to be directly imaged using
an ordinary smartphone (Apple iPhone 4S). Scale: all electrodes are 50 μm wide horizontally

3. Wafers are patterned, and developed (MF-319 developer;

Rohm & Haas).
4. Electrodes are fabricated by depositing a 600 Å chromium layer
by thermal evaporation.
5. Glass microchannels (275  45 μm cross section) are etched on
glass wafers (borofloat, 500 μm thick, 15 cm diameter; Preci-
sion Glass and Optics) using hydrofluoric acid.
6. Assembled devices are wire bonded to a printed circuit board so
that electrodes can be individually addressed.
Building on this proof of concept, we developed a greatly
simplified design based on electrochemical dissolution of copper
in a format that can be directly integrated with convective PCR
reactors. Here, copper electrodes experience electrochemical cor-
rosion at non-neutral pH (Fig. 6a), with degradation rates strongly
dependent on bulk solution composition. To demonstrate this
approach in the context of the above mentioned convective PCR
platform, we patterned 100 nm-thick addressable copper electrodes
on the top surface of the PCR reactor. This arrangement enables
degradation kinetics to be straightforwardly monitored because the
electrode surface (500 μm-wide, 1 mm inter electrode spacing) is
large enough to easily view using the smartphone camera (Fig. 6a).
This can be seen by comparing degradation kinetics before and after
replication of a 237 bp target sequence from a λ-phage DNA
 Smartphone-Enabled Detection Strategies for Portable PCR–Based Diagnostics 263

Fig. 6 Integrated label-free convection PCR detection by imaging local electrode dissolution. (a) Addressable
Cu electrodes are patterned on a glass substrate affixed to the top surface of a cylindrical PCR reactor,
enabling electrochemical dissolution to be viewed from above (drawing not to scale). Anodic dissolution
occurs slowly at neutral pH, but is accelerated under alkaline conditions (images of the anode taken after a 3 V
potential was applied for 1 min). (b) In the pre-PCR solution (above), the anode surface is not visibly changed
when a 3 V potential is applied. Dissolution becomes favored (below) as more negatively charged DNA
molecules are generated during PCR, forming an electrophoretically compacted film at the anode. (c) Images
of the electrodes can be analyzed to quantify anodic dissolution before and after PCR, and compared against
appropriate controls. Scale: all electrodes are 500 μm wideSee Polymerase chain reaction (PCR)

template. Electrode dimensions were selected such that small

potentials (3 V, much lower values can be applied by reducing the
inter-electrode spacing) generate electric fields high enough to
electrophoretically trap negatively charged DNA at the anode
[35, 36]. The resulting DNA film imposes a membrane-like barrier
against transport of electrochemically generated OH– ions into the
surroundings, leading to a local increase in pH that acts to acceler-
ate the rate of copper dissolution (Fig. 6b). DNA replication can
then be monitored in situ by analyzing video recordings of the
electrodes to quantify the dissolved mass of Cu before and after
PCR, revealing a faster dissolution rate owing to increase in target
DNA concentration (Fig. 6c).
1. Copper film coatings (100 nm thick) were sputter deposited on
glass microscope slides.
264 Aashish Priye and Victor M. Ugaz

2. Electrodes (500 μm wide, 1 mm inter-electrode spacing) were

patterned on the copper coated slides using dry transfer film.
3. The patterned electrode on the glass slide was immersed in
metal etchant (Gold Etchant TFA, Transene) for ~1 min after
which the remaining dry transfer film was stripped using
acetone.
4. Patterned glass slides were affixed to the upper surface of the
cylindrical acrylic cells using cyanoacrylate adhesive, and elec-
trical connections were made using conductive tape.
5. Wires are soldered directly to pads on the patterned copper,
enabling the individual electrode pair to be addressed.
6. Potentials were applied using a DC power supply. The bottom
surfaces of the cells were sealed with aluminum tape and the
assembly was heated from below using the same apparatus
described above.
7. A smartphone was used to continuous monitor the electrode
dissolution process using video recordings.
8. The acquired video was converted to image sequences, and
analyzed using ImageJ software.
9. Dissolution was quantified within a region of interest overlay-
ing the anode surface by converting the images to 8-bit gray-
scale and applying a black/white threshold of 160 out of 255.
10. The total white pixel count was then calculated to determine
the mass of copper dissolved before and after convective PCR
(Fig. 5c).

4 Notes

1. For the convective PCR runs we found that additional rinsing

steps (step 8, Subheading 3.1) helped minimize adsorption at
the sidewalls that may otherwise inhibit the reaction while also
enhancing surface wettability so that reagents could be loaded
without trapping air pockets inside the reactor. It is also essen-
tial to ensure that air bubbles are not introduced during sample
loading as they can not only hinder the optical detection from
the top but larger bubbles can also disrupt the convective flow
resulting in failed PCR. Most of the time they are introduced at
the time of sealing the reactor so proper care should be taken at
this step. Sealing the reactors under pressure may also help to
suppress any initial bubble formation. Additionally, injection
molding can be applied to construct reactors, as opposed to the
current machining process. Molding processes are envisioned
to be used for mass production, and are likely to significantly
reduce sidewall roughness that generates bubble nucleation
 Smartphone-Enabled Detection Strategies for Portable PCR–Based Diagnostics 265

sites. These modifications suggest strong potential to signifi-

cantly reduce bubble formation while convective PCR and thus
improve operational reliability.
2. Prior to beginning step 10 in Subheading 3.1, it is important
that all the other pre-PCR steps be performed. For example,
the enzyme we used in the SYBR master mix requires a hot start
step. We ensured that enzyme activation was performed by
heating the sample for 10 min at 95
C in a conventional
thermocycler (T-Gradient, Biometra).
3. A key challenge in smartphone-based optical detection is
accounting for variability of ambient lighting. Inherent to all
smartphones is a dynamic color balancing functionality opti-
mized for photography that introduces variability in sensor
parameters applied to adjust to the dynamic ambient condi-
tions. This mode of operation is not ideal for fluorescence
detection, and previous studies have explored approaches
including the use of color calibration targets and reference
samples to reduce the resulting image-to-image variability
[37, 38]. Our system employs a 3D printed “dark box” to
block ambient light and provides consistent illumination form
the built-in LEDs. Additionally, our analysis app overrides the
preset camera parameters so that fixed focal length and expo-
sure times can be employed. Together, these approaches elimi-
nate the most significant sources of fluorescence intensity
variations.

References
1. Ozcan A (2014) Mobile phones democratize 7. Tseng D et al (2010) Lensfree microscopy on a
and cultivate next-generation imaging, diag- cellphone. Lab Chip 10:1787–1792
nostics and measurement tools. Lab Chip 8. Yu H, Tan Y, Cunningham BT (2014) Smart-
14:3187–3194 phone fluorescence spectroscopy. Anal Chem
2. Laksanasopin T et al (2015) A smartphone 86:8805–8813
dongle for diagnosis of infectious diseases at 9. Wei Q et al (2014) Imaging and sizing of single
the point of care. Sci Transl Med 7:273re1 DNA molecules on a mobile phone. ACS Nano
3. Zhu H, Isikman SO, Mudanyali O, Green- 8:12725–12733. doi:10.1021/nn505821y
baum A, Ozcan A (2013) Optical imaging 10. Lee S, Oncescu V, Mancuso M, Mehta S, Erick-
techniques for point-of-care diagnostics. Lab son D (2014) A smartphone platform for the
Chip 13:51–67 quantification of vitamin D levels. Lab Chip
4. Erickson D et al (2014) Smartphone technol- 14:1437–1442
ogy can be transformative to the deployment of 11. Oncescu V, Mancuso M, Erickson D (2014)
lab-on-chip diagnostics. Lab Chip Cholesterol testing on a smartphone. Lab
14:3159–3164 Chip 14:759–763
5. Xu X et al (2015) Advances in smartphone- 12. Oncescu V, O’Dell D, Erickson D (2013)
based point-of-care diagnostics. Proc IEEE Smartphone based health accessory for colori-
103:236–247 metric detection of biomarkers in sweat and
6. Breslauer DN, Maamari RN, Switz NA, Lam saliva. Lab Chip 13:3232–3238
WA, Fletcher DA (2009) Mobile phone based 13. Shen L, Hagen JA, Papautsky I (2012) Point-
clinical microscopy for global health applica- of-care colorimetric detection with a smart-
tions. PLoS One 4:e6320 phone. Lab Chip 12:4240–4243
266 Aashish Priye and Victor M. Ugaz

14. San Park T, Li W, McCracken KE, Yoon J-Y 28. Kopp MU, De Mello AJ, Manz A (1998)
(2013) Smartphone quantifies Salmonella from Chemical amplification: continuous-flow PCR
paper microfluidics. Lab Chip 13:4832–4840 on a chip. Science 280:1046–1048
15. Fronczek CF, San Park T, Harshman DK, Nico- 29. West J et al (2002) Application of magnetohy-
lini AM, Yoon J-Y (2014) Paper microfluidic drodynamic actuation to continuous flow
extraction and direct smartphone-based identi- chemistry. Lab Chip 2:224–230
fication of pathogenic nucleic acids from field 30. Priye A et al (2016) Lab-on-a-drone: toward
and clinical samples. RSC Adv 4:11103–11110 pinpoint deployment of smartphone-enabled
16. Gallegos D et al (2013) Label-free biodetection nucleic acid-based diagnostics for mobile
using a smartphone. Lab Chip 13:2124–2132 health care. Anal Chem 88:4651–4560
17. Huang Y-W, Ugaz VM (2013) Smartphone- 31. Ugaz VM, Krishnan M (2004) Novel convec-
based detection of unlabeled DNA via electro- tive flow based approaches for high-throughput
chemical dissolution. Analyst 138:2522–2526 PCR thermocycling. JALA 9:318–323
18. Krishnan M, Ugaz VM, Burns MA (2002) 32. Skandarajah A, Reber CD, Switz NA, Fletcher
PCR in a Rayleigh-Benard convection cell. Sci- DA (2014) Quantitative imaging with a mobile
ence 298:793–793 phone microscope. PLoS One 9:e96906
19. Priye A, Hassan YA, Ugaz VM (2013) Micro- 33. Liu W, Saint DA (2002) Validation of a quanti-
scale chaotic advection enables robust convective tative method for real time PCR kinetics. Bio-
DNA replication. Anal Chem 85:10536–10541 chem Biophys Res Commun 294:347–353
20. Muddu R, Hassan YA, Ugaz VM (2011) Cha- 34. Rutledge R (2004) Sigmoidal curve-fitting
otically accelerated polymerase chain reaction redefines quantitative real-time PCR with the
by Microscale Rayleigh–Bénard convection. prospective of developing automated high-
Angew Chem Int Ed 50:3048–3052 throughput applications. Nucleic Acids Res
21. Yao D-J, Chen J-R, Ju W-T (2007) Micro–- 32:e178
Rayleigh-Bénard convection polymerase chain 35. Huang Y-W, Shaikh F, Ugaz VM (2011) Tun-
reaction system. JM3 6:043007–043009 able synthesis of encapsulated microbubbles by
22. Braun D, Goddard NL, Libchaber A (2003) coupled electrophoretic stabilization and elec-
Exponential DNA replication by laminar con- trochemical inflation. Angew Chem Int Ed
vection. Phys Rev Lett 91:158103 50:3739–3743
23. Hennig M, Braun D (2005) Convective poly- 36. Shaikh F, Ugaz VM (2006) Collection, focus-
merase chain reaction around micro immersion ing, and metering of DNA in microchannels
heater. Appl Phys Lett 87:183901 using addressable electrode arrays for portable
24. Braun D (2004) PCR by thermal convection. low-power bioanalysis. Proc Natl Acad Sci U S
Mod Phys Lett B 18:775–784 A 103:4825–4830
25. Priye A, Muddu R, Hassan YA, Ugaz VM 37. Sari YA, Ginardi RH, Sarno R (2013) Assess-
(2011) Royal Society of Chemistry, Cambridge ment of color levels in leaf color chart using
smartphone camera with relative calibration.
26. H€ uhmer A, Landers J (2000) Noncontact Inform Syst 2:4
infrared-mediated thermocycling for effective
polymerase chain reaction amplification of DNA 38. Yetisen AK, Martinez-Hurtado J, Garcia-
in nanoliter volumes. Anal Chem 72:5507–5512 Melendrez A, da Cruz Vasconcellos F, Lowe
CR (2014) A smartphone algorithm with inter-
27. Pal R et al (2005) An integrated microfluidic phone repeatability for the analysis of colori-
device for influenza and other genetic analyses. metric tests. Sens Actuators B 196:156–160
Lab Chip 5:1024–1032
 Chapter 17

Streak Imaging Flow Cytometer for Rare Cell Analysis

Joshua Balsam, Hugh Alan Bruck, Miguel Ossandon, Ben Prickril,
and Avraham Rasooly

Abstract
There is a need for simple and affordable techniques for cytology for clinical applications, especially for
point-of-care (POC) medical diagnostics in resource-poor settings. However, this often requires adapting
expensive and complex laboratory-based techniques that often require significant power and are too massive
to transport easily. One such technique is flow cytometry, which has great potential for modification due to
the simplicity of the principle of optical tracking of cells. However, it is limited in that regard due to the flow
focusing technique used to isolate cells for optical detection. This technique inherently reduces the flow rate
and is therefore unsuitable for rapid detection of rare cells which require large volume for analysis.
To address these limitations, we developed a low-cost, mobile flow cytometer based on streak imaging. In
our new configuration we utilize a simple webcam for optical detection over a large area associated with a
wide-field flow cell. The new flow cell is capable of larger volume and higher throughput fluorescence
detection of rare cells than the flow cells with hydrodynamic focusing used in conventional flow cytometry.
The webcam is an inexpensive, commercially available system, and for fluorescence analysis we use a 1 W
450 nm blue laser to excite Syto-9 stained cells with emission at 535 nm. We were able to detect low
concentrations of stained cells at high flow rates of 10 mL/min, which is suitable for rapidly analyzing larger
specimen volumes to detect rare cells at appropriate concentration levels. The new rapid detection cap-
abilities, combined with the simplicity and low cost of this device, suggest a potential for clinical POC flow
cytometry in resource-poor settings associated with global health.

Key words POC, Webcam, Flow cytometry, Rare cells, Fluidics, Fluorescence detection

1 Introduction

In many clinical applications, environmental analysis, and basic

research, flow cytometers have become important diagnostic
tools. One current limitation for flow cytometry is that the asso-
ciated low flow rates makes it practical for analyzing small specimen
volumes. This is not appropriate for detecting rare cells, including
circulating tumor cells (CTCs), and therefore prevents flow cyt-
ometers from being used for rare cells analysis. Thus there is a need
for new flow cytometry techniques for rare cell detection that are

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_17,
© Springer Science+Business Media LLC 2017

267
268 Joshua Balsam et al.

capable of handling larger specimen volumes. If these techniques

can be made more affordable and portable it would also be possible
to successfully introduce them in resource-poor settings. To this
end, a new imaging point-of-care (POC)see Point-of-care (POC)
flow cytometer for detection of rare cells [1–3] based on streak
imaging is described in this protocol.

1.1 Point-of-Care The development of POC analytical tools has largely focused on
Cytometry lab-on-a-chip (LOC) microfluidic technologies due to their small
size, simplicity, and low manufacturing cost [4, 5]. In addition,
smartphone cameras are being used effectively as optical detection
systems for mobile POC devices based on LOC platforms [6–14].
For example, an optofluidic fluorescent imaging cytometer using a
smartphone camera with a spatial resolution of 2 μm has been
recently described [9, 13]. However, the flow rate for this system
is 1 μL/min, limiting analyses to small volumes. In addition to
smartphones, standalone inexpensive webcam imaging systems
may also overcome limitations of smartphone cameras. For exam-
ple, smartphone cameras often have limited video capabilities (e.g.,
many phones are limited to 30 fps) and have less versatile optical
systems (e.g., inability to change lenses or directly set imaging
parameters). To overcome these limitations we developed a small,
mobile, and low-cost flow cytometer based on webcam imaging
capable of large volume analysis and rare cell detection for a variety
of POC applications.
The main components of current flow cytometers consist of:
[1] a fluidic system for carrying fluorescently labeled biological
material (e.g., cells), and [2] an optical system for fluorescence
detection. Most current flow cytometers utilize microfluidic
sheathing to focus the cells or particles to a channel whose width
permits only a single cell to pass, enabling narrow-field detection
using photomultipliers or other narrow-field photodetectors. This
technique is known as “hydrodynamic focusing.” In one recent
device a passive hydrodynamic flow focusing was achieved
using chevron grooves imbedded on the walls of the microchannel
[15, 16]. This enabled the sheath fluid to completely surround
the sample stream so that cells could be individually interrogated
by the narrow-field fluorescent detector. However, one limitation
of the focused stream is a low flow rate due to the high hydrody-
namic resistance and pressure constraints of the cell, ultimately
limiting the device to small volumes or long analysis times. In
addition, the commonly used photomultipliers used for detection
are delicate, expensive, and difficult to transport.
In general, current flow cytometers using hydrodynamic focus-
ing flow cells are not suitable for POC applications or for rare cell
detection due to their size, expense, lack of durability, small sample
capacity, and high power requirement. An alternative approach to
the focused microfluidic sheathing used in conventional flow
 Streak Imaging Flow Cytometer for Rare Cell Analysis 269

cytometers has been recently developed using a wide-field flow cell

[1–3] for large volume analysis. By expanding the cross section of
the flow field by orders of magnitude over conventional hydrody-
namic focusing flow cells, low flow rates can be used to interrogate
larger specimen volumes over the same period of time while
enabling easier detection by reducing the linear velocity of target
cells and increasing the detection area.
With lower flow velocities, wide-field flow cell cytometry can
use less-sensitive alternatives to photomultipliers for optical detec-
tion such as imaging devices based on complementary metal–ox-
ide–semiconductor (CMOS) or charge-coupled device (CCD)
detectors. Because of the larger flow field it is necessary to use
these imaging devices to interrogate large numbers of cells in
parallel. However, due to their inherent portability CMOS or
CCD cameras have already been employed as relatively simple,
low cost devices for optical detection over large areas in several
array assays [17–20]. Their main advantage is that they have suffi-
cient sensitivity to analyze light from an area large enough to cover
the entire surface of a LOC or array [21–23]. This has made CMOS
or CCD-based detectors an ideal choice for POC detectors.

1.2 Streak Imaging Using CCD or CMOS imaging sensors for flow cytometry opens
Flow Cytometry up new possibilities for image analysis. One approach is to image
moving cells at long exposure times such that at a given flow rate a
cell moves a distance X in time t to produce a line of pixels described
as a “streak image.” The dimensions of the streak are directly
related to the size of the cell and its velocity. Unlike conventional
imaging aimed of capturing precise sharp image of the cell using
low flow rate and short exposure, streak imaging enables the use of
higher flow rates and long exposure to screen larger volumes of
specimens.
Figure 1a illustrates a cell image captured as a line generated
using a flow rate of 20 mL/min with a frame rate of 20 FPS. Unlike
a spot that would be generated during regular imaging, the line
pattern associated with exposing many pixels during streak imaging
enables easy identification, especially for low quality detectors with
high background noise. The minimum streak length required is
such that at least one pixel measures the maximum possible signal in
order to maximize signal. This streak length is equal to twice the
length of the cell image in pixels before streaking (L) plus one pixel
(i.e., for a cell image of 3
3 pixels, a streak length of 7 is required).
This effect is shown graphically in Fig. 1b, where a cell of length 3
pixels is seen to move a distance of 4 pixels, thereby producing an
image streak with length 7 pixels with a single bright pixel in its
center. An actual cell streak image is shown in Fig. 1c for a similar
cell image size and similar flow conditions, along with a plot of
actual pixel brightness along the center of the cell streak in Fig. 1d.
Cell streaking such as this results in at least one pixel measuring the
maximum possible signal from the cell.
270 Joshua Balsam et al.

d 20

15
Gray Value

10

0 2 4 6 8
Distance (pixels)

Fig. 1 Streak imaging of Cells—(a) An image of a cell captured at a flow rate of

20 mL/min with a frame rate of 20 FPS (the arrow shows the flow direction). (b)
Schematic of a cell traversing a number of pixels showing a maximum
brightness achieved in the pixel at the image center. (c) An actual cell image
is shown under these approximate flow conditions (125 μL/min, 20 FPS), along
with (d) a plot of its brightness along the center line of pixels
 Streak Imaging Flow Cytometer for Rare Cell Analysis 271

Recently, a new computational enumeration approach for accu-

rate counting of cells in streak mode using low resolution cameras
has been developed (Miguel Ossandon unpublished results). The
image analysis algorithm identifies potential cells based on the
streak intensity, length and relative location of the streaks in con-
secutive frames. The algorithm is based on: (1) finding streaks, (2)
identifying candidate cells, and (3) filtering out spurious cells to
identify true cells. This approach may enable automation of streak
imaging cytometry.

2 Materials

2.1 Cell Culture and 1. THP-1 human monocytes cells, ATCC #TIB-202 (ATCC
Fluorescence Staining Manassas, VA).
2. For cell culturing, RPMI-1640 medium supplemented with
10% (v/v) heat-inactivated FBS, 1% antibiotic–antimycotic
solution, and 2 mM glutamine is used.
3. SYTO-9 dye (Molecular Probes Eugene, OR).
4. The cells are cultured in CO2 incubator.

2.2 Optical 1. As a detector, Sony PlayStation® Eye webcam is used as an

Components for inexpensive color CMOS photodetector (Amazon, Seattle
Imaging Flow WA).
Cytometer 2. An alternative camera with improved sensitivity is a C mount
1.3 Mp gray scale CCD camera CMLN 13S2M-CS (Point
Grey Research, Richmond BC Canada).
3. A C-mount CCTV lens (Pentax 12 mm f/1.2) is used to
replace the webcam lens to improve imaging (Spytown, Utopia
NY).
4. Green emission filter HQ535/50M (Chroma Technology
Corp Rockingham VT).
5. 1 W 450 nm blue laser is used for fluorescent excitation illumi-
nation (Hangzhou BrandNew Technology Co., Zhejiang,
China).

2.3 Flow Cell Fluid 1. 3 mm clear acrylic sheet (Piedmont Plastics, Beltsville MD).
Delivery, and Imaging 2. Quartz glass microscope slide.
3. 3 M 9770 Adhesive transfer Tape (Piedmont Plastics, Beltsville
MD).
4. Nylon rod (McMaster-Carr, Robbinsville, NJ).
5. Fusion100 syringe pump is used for fluid delivery (Chemyx,
Stafford TX).
6. Epilog Legend CO2 65 W laser cutter (Epilog, Golden CO)
used for flow cell fabrication.
272 Joshua Balsam et al.

2.4 Computer 1. The webcam sensor is connected to a 32-bit Windows-based

Control and Data laptop computer via a USB2 port.
Analysis 2. Drivers and software allowing the webcam to be controlled on
a personal computer are developed and freely distributed by
Code Laboratories, Inc. (Henderson NV).
3. The camera control software (CL-Eye Test) is used to set
camera parameters (exposure time, frame rate, and gain) and
to capture and save video in uncompressed AVI format.
4. Video files are analyzed using ImageJ software (freely
distributed by NIH, http://rsb.info.nih.gov/ij/).
5. Data analysis and plotting is carried out in Microsoft Excel
(Redmond, WA).

3 Methods

3.1 Cell Culture and Fluorescently stained THP-1 human monocytes are used as a
Labeling model to simulate rare CTCs. Though monocytes themselves are
not rare, they are diluted to levels similar to those associated with
rare cells.
1. Cells are cultured with RPMI-1640 medium supplemented
with 10% (v/v) heat-inactivated FBS, 1% antibiotic–antimyco-
tic solution, and 2 mM glutamine (see Note 1) in a 5% CO2
environment at 37  C.
2. Cells are removed from an active culture, pelleted by centrifu-
gation and resuspended in deionized water at an approximate
concentration of 106 cells/mL (staining protocol for Syto-9
dye advises against the use of phosphate buffer solutions).
3. To stain cells, 10 μL of Syto-9 dye (3.34 mM stock concentra-
tion) is added to 1 mL of suspended cells and allowed to rest at
room temperature in the dark for 20 min. Washing of cells to
remove excess dye is not necessary due to the low intrinsic
quantum yield when not bound to nucleic acids (<0.01), and
because of the later dilution of this stock solution by factors of
104–106. One known factor which contributed substantially to
the decreasing counting efficiency at lower concentrations is
the cell staining protocol. The protocol recommended by the
dye manufacturer is followed. However, dye which has entered
the cell but not yet bound to DNA is noted to diffuse out of
cells over time, leading to diminishing cell brightness. Because
all dilutions are prepared simultaneously and then analyzed
sequentially from highest to lowest concentration, lower dilu-
tions exhibited cells with lower fluorescent intensity. Allowing
the cells to equilibrate at a low concentration for several hours
would allow this diffusion process time to complete.
 Streak Imaging Flow Cytometer for Rare Cell Analysis 273

4. After staining, cells are diluted to approximately 1–10 cell/μL

(measured by microscopy) to allow for accurate manual count-
ing. Cell concentration is measured by placing 3 μL sample
droplets on a microscope slide and counting cells in the droplet
in real time under laser excitation using the same imaging
platform employed to image the flow cells in these experi-
ments. This is repeated many (N > 20) times. An average cell
concentration of 3.73 cells/μL with a standard error of 0.3 is a
typical population estimate resulting from these measurements.
From this relatively high concentration, lower concentration
samples of 100, 10, and 1 cell/mL are generated by single-step
dilution. For each dilution, 268 μL of stock solution (with
measured concentration of 3.73 cells/μL) is diluted into a
volume of deionized water to yield the final target concentra-
tion (10, 100, and 1000 mL of deionized water for concentra-
tions of 100, 10, and 1 cell/mL, respectively). Based on a
normal sampling distribution with standard deviation of
0.3 cells/μL, the 95% confidence range for each concentration
is 84–116, 8.4–11.6, and 0.84–1.16 cells/mL for concentra-
tions of 100, 10, and 1 cell/mL, respectively. Pipetting volume
error is measured to be less than 1%.

3.2 Wide-Field Flow A wide-field flow cytometer is developed that is simple, lightweight,
Cytometer for POC and inexpensive. The components are available commercially, and
Applications the key to the design is the integration of the elements in a compact
configuration that is isolated from ambient light to enhance imag-
ing of the fluorescence signal, and which allows for easy placement
and removal of the wide-field flow cell.
1. The flow cytometer for fluorescence detection (Fig. 2) consists
of four main elements: (1) a webcam used as an optical detec-
tor, (2) a laser excitation source for illumination, (3) a wide-
field flow cell, and (4) an optical stage to hold each module in
alignment for stable and clear imaging.
2. For fluorescence detection, a green emission filter with a center
wavelength of 535 nm and bandwidth of 50 nm is used with a
1 W 450 nm blue laser (see Note 1). The optical elements must
be vertically centered (see Note 2).
3. The optical detector consists of the electronics of a webcam
which is disassembled in order to remove the original lens, and
a 12 mm f/1.2 CCTV lens for improved light collection and
image magnification (see Note 3).
4. The fluidics module consists of a flow cell and a programmable
syringe pump. The flow cytometer platform (Fig. 2) consists of
a stationary platform and a moveable stage for focusing using a
screw mechanism.
274 Joshua Balsam et al.

Flow Cytometer
Flow Cell
In from syringe Out to Waste
pump

Movable stage

12mm
CCTV
Lens
Emission
Filter
Computer
Webcam circuit board and
sensor with USB output

Fig. 2 Schematic of mobile wide-field flow cytometer for POC—The flow

cytometer consists of four modules: (1) a webcam as an optical detector, (2) a
blue 450 nm excitation source for illumination, (3) a wide-field flow cell, and (4) a
stage to focus the image and to hold each module in alignment. The sensing
element consists of the electronic elements of a webcam, a 12 mm f/1.2 CCTV
lens, a green emission filter. It is connected to a computer to collect and analyze
data. The excitation source is a 450 nm 1 W blue laser. The sample handling
module consists of the wide-field flow cell, a programmable syringe pump, and a
waste container

5. The flow cytometer platform is constructed using 0.5 in. thick

clear acrylic sheet and nylon rod which is used as a rail for
focusing. The webcam electronic circuitry and optics are
attached to the stationary platform while the flow cell is
attached to the translating stage. It is important to enclose
the device to reduce background light (see Note 4).

3.3 Wide-Field Flow To maximize residence time of cells in the interrogation window
Cell Fabrication and maximize the number of fluorescent photons captured (i.e., to
increase sensitivity), a new high throughput flow cell is designed
using a wide flow channel that increases volumetric flow rates at
lower flow velocities through an orders of magnitude higher cross-
sectional area than used for hydrodynamic focusing.
1. The flow cell (Fig. 3) consists of a 4–20 mm wide flow channel
that maximizes the volume of fluid in the interrogation win-
dow. Fluorescent labeled samples are injected via an inlet
through the channel, excited by a laser in the channel, and
 Streak Imaging Flow Cytometer for Rare Cell Analysis 275

Inlet Inlet

A B D

Field of Field of
View View

Laser spot Laser Line

Outlet Outlet
Fig. 3 Wide-field flow cell—(a) A schematic of the wide-field flow cell with key elements illuminated with spot
laser. (b) An image from the flow cell illuminated with spot laser. (c) A schematic of the wide-field flow cell
with key elements illuminated with line laser illumination. (d) An image from the flow cell illuminated with line
laser illumination

then imaged by the webcam before being collected at the

outlet.
2. The wide-field flow cell is fabricated using an Epilog Legend
CO2 65 W laser cutter (see Note 5) using techniques similar to
those described in our previous work [5, 24–28] (see Note 6).
3. The flow cell consisted of three functional layers: (1) a plain
glass or quartz microscope slide lower layer, (2) a middle layer
laser machined from 3 M 9770 double-sided adhesive transfer
tape to define the geometry of the fluid channel, and (3) a glass
or quartz microscope slide on the top layer that has two holes
drilled for the inlet and outlet ports aligned with the ends of
the fluid channel layer (see Note 6).
4. Fluid interfacing is accomplished by bonding 1 cm2 acrylic
plates over each port hole. The acrylic plates have through
holes laser machined, into which 18 G needles are bonded.
Tygon 1.3 mm ID tubing is used to connect directly from the
needles to either a waste container or an input syringe. The
experiments consist of injecting fluorescently labeled samples
into the wide channel via an inlet, excite them with the laser,
and then image them with the webcam before they are col-
lected at the outlet (see Note 6).
276 Joshua Balsam et al.

5. Channel depth is set to keep the flow field within the depth of
field of the lens being used. A Pentax CCTV 12 mm f/1.2 is
operated at approximately f/2.4 to reduce field curvature and
improve depth of field.
6. A Fusion100 syringe pump connected to the flow cell is used
for flow rate control. The maximum flow rate achievable
through this flow cell is 10 mL/min. The adhesive may fail at
higher flow rates due to the dynamic pressures required. The
flow rate in experiments presented in this chapter is limited to
500 μL/min due to the maximum frame rate of the webcam
employed for sensing. For very faint cells the flow rate should
be decreased in order to improve sensitivity. Sensitivity also
increases if exposure time is increased to allow cell images to
form short streaks, with length greater than or equal to
the length of the cell image plus one pixel (detailed in Sub-
heading 3.7).

3.4 Optical System The optical system consists of: (1) the blue laser for fluorophore
for Webcam-Based excitation, (2) a webcam detector, and (3) an excitation filter.
Cytometer Platform 1. Laser Excitation: For fluorescent imaging of a wide-field, a
1 W laser is used for illumination in two modes of illumination,
spot illumination and line illumination:
(a) In spot illumination the laser illuminates the flow cell at an
angle of incidence of approximately 45 (see Note 2) at
this angle the laser illuminate a bright elliptical spot which
covered the width of the flow cell (Fig. 3b). The main
advantage of spot illumination is that it is high intensity,
which is suitable for higher flow rates.
(b) In line illumination, the laser illuminate the edge of the
slide an angle of incidence of approximately 45 and the
light illuminate the flow cell through the slide (Fig. 3d)
(c) The 1 W consumer grade laser is fairly expensive (~$300)
for a device designed for use in a low-resource setting. To
reduce the cost of the laser, a lower power laser with line
generator optics could be used to further focus the laser
spot. This could allow the critical parameter of photon flux
to remain unchanged while reducing overall power; how-
ever, this approach may reduce the area analyzed. Alterna-
tively, high power LEDs could be utilized. This would
require the addition of an excitation filter as well as colli-
mating optics, leading to a more complex and potentially
more expensive configuration.
2. Webcam detector and emission filter: A Sony PlayStation®
Eye webcam is used as the photodetector in this platform. The
Sony PlayStation Eye device was disassembled and the main
circuit board (with attached image sensor and USB cable) is
 Streak Imaging Flow Cytometer for Rare Cell Analysis 277

removed. A C-mount CCTV lens (Pentax 12 mm f/1.2) is

used to replace the stock camera lens (see Note 2). For fluores-
cence detection, a green emission filter with center wavelength
535 and 50 nm bandwidth is used for detecting fluorescent
emission. To improve sensitivity, other video devices which
employ CCDs with electronics for high frame rate video imag-
ing can be used, such as the Point Grey Research CMLN
13S2M-CS.

3.5 Configuration of The new imaging flow cytometer, the webcam integrated the sen-
Imaging Flow sor into the bottom of the platform below the flow cell, as shown in
Cytometer Fig. 2. It is connected to a 32-bit Windows-based laptop computer
via a USB2 port. The drivers and software controlling the webcam
are developed and freely distributed by Code Laboratories. The
camera control software (CL-Eye Test) enables setting of several
camera parameters (exposure time, frame rate, and gain), and the
capture of video in uncompressed AVI format. The imaging para-
meters for the camera (exposure time and frame rate) depend on
the brightness of the fluorescently labeled cells and the flow rate (see
Note 7). An example of fluorescence labeled cells is the single video
frame of THP-1 stained with SYTO-9 dye in a wide-field flow cell at
a flow rate of 500 μL/min, as shown in Fig. 4.

3.6 Background Frames may contain a signal of interest (e.g., image of a cell),
Signal and Noise interfering background signal (e.g., autofluorescence), and random
Reduction noise. To enhance detection, the constant component of the inter-
fering signal can be subtracted. For background subtraction, the
median value and the maximum value of each pixel in a video file of
the relevant color channel can be calculated. Figure 5 illustrates the
green channel video frames of a video clip containing 2000 frames
(10.7 s of video) showing a single cell passing through the flow cell
shown in different positions along its path. The median image of all
the frames, shown in Fig. 5a, represents the average background
signal with no information from the cells. The maximum image,
shown in Fig. 5b, indicates the highest signal recorded at each pixel
during the video with the cell position in each frame marked. In
order to improve cell detection, especially for lower SNR images
than shown in this example, the median image is subtracted from
the maximum image (Fig. 5c), which allowed for improved visuali-
zation of cell movement by removing background autofluorescence
from the flow cell.

3.7 Noise Reduction Figure 6a shows an actual cell streak image (circled) with flow
in Streak Mode Images direction indicated. Figure 6b shows a close-up of the cell streak
image containing background noise and showing individual pixels
and. In order to reduce noise each column of pixels is averaged
along the streak length n to produce a single averaged row of pixels,
labeled avg(n) in Fig. 6. Figures 6d–i and ii shows a plot of pixel
278 Joshua Balsam et al.

Fig. 4 Cell imaged using wide-field flow cytometer. (a) Single video frame of
fluorescent labeled THP-1 human monocytes in wide-field flow cell at
500 μL/min, and (b) 3D visualization of A

values before and after averaging, respectively, indicating a 3

improvement in SNR. SNR is calculated as follows:
μsignal
SNR ¼
σ noise
Where μsignal is the peak value in the plot, and σ noise is the
standard deviation of the pixel values on either side of the peak.
When calculating σ noise, a five pixel exclusion zone on either side of
a peak is used to prevent the inclusion of any signal components in
the noise measurement. Twenty-five pixel windows on either side
of this exclusion zone are used to measure local noise levels (i.e., a
total of fifty pixels around each peak).
Fig. 5 Background subtraction to improve imaging. (a) Median pixel value from
2000 frames showing average background autofluorescence from the flow cell.
(b) Maximum pixel value from 2000 video frames showing a single cell moving
through the laser spot. (c) Result of subtracting image B from image C, allowing
for improved visualization of cell movement and faint cell images
280 Joshua Balsam et al.

Fig. 6 Noise reduction in streak mode images. (a) A cell streak image (circled) with flow direction indicated. (b)
Close-up of cell streak image showing individual pixels and background noise. In order to reduce noise, each
column of pixels is averaged over the streak length n to produce a single averaged row of pixels, labeled avg
(n). (c) A plot of pixel values before (i) and after (ii) averaging, showing a 3
improvement in SNR. The plot in (i)
is for the row with the brightest pixel value in (d).
 Streak Imaging Flow Cytometer for Rare Cell Analysis 281

3.8 Color Channel When using a color camera, spectral filtering can be employed to
Extraction improve SNR by reducing non-specific background light (e.g.,
excitation light). The webcam CMOS sensor employs a typical
Bayer color filter array in which one half of all pixels have a green
filter, 25% have a red filter, and 25% have a blue filter. The emission
profile of Syto-9 dye used in this work has a large overlap in the
green spectrum resulting in this color channel having the highest
SNR. The blue and red channels typically showed SNRs of one half
and one tenth those of the green channel, respectively. For this
reason the red and blue channels are discarded and only the green
channel is analyzed. The “Split Channels” function in ImageJ is
used to divide a color image into its constituent channels.

3.9 Rare Cell Counting of rare cell events is simulated using dilutions of fluores-
Counting Using cently labeled THP-1 monocytes. Dilutions are prepared as
Streaking Mode described in Subheading 1, and then measured using the wide-
Imaging field flow cytometer in Fig. 2. Figure 7a shows results using the
flow cell depicted in Fig. 3a. Dilutions of 100, 10, and 1 cell per mL
respectively yielded average concentrations of 84, 7.9, and
0.56 cells/mL, with 95% confidence bounds of 61–107, 6.9–9,
and 0.16–0.96 cells/mL. Error bars in Fig. 7 represent a 95%
confidence interval.
Following this set of experiments, flow cell geometry and sys-
tem operating parameters (flow rate, frame rate) are optimized

1000

A 1000
B
Measured Cell Count (cells/mL)

100
Measured Cell Count (cells/mL)

100

10
10

1 1

0.1 0.1

0.01 0.01
100 10 1 100 10 1 0.1
Target Cell Count (cells/mL) Target Cell Count (cells/mL)

Fig. 7 Rare cell counting using streaking mode—(a) Results of counting dilutions of rare cells using line laser
illumination. The large error bars evident at the level of 1 cell/mL prompted further optimization of the flow cell
geometry and operating parameters. (b) Results of counting dilutions of rare cells using an updated flow cell
design to improve image SNR by reducing cell velocity. The field-of-view (FOV) was also improved. Following
this optimization, performance of the wide-field flow cytometer improved substantially at the level of 1 cell/mL,
allowing measurement expansion to as low as 0.1 cell/mL. Measurements made via microscopy
(sparsely filled columns) vs. wide-field flow cytometry (dense fill) were not significantly different at the 95%
confidence level
282 Joshua Balsam et al.

based on the methods described in Subheading 3.6. This resulted in

the flow cell pictured in Fig. 7b with increased channel width and
an increased field-of-view. The excitation laser spot is modified to
form a line covering the width of the channel. The modifications to
the flow cell resulted in cells moving with lower velocity, which
translates to higher image SNR. This allowed for substantially
improved results at low cell dilutions as seen in Fig. 7c. Dilutions
of 1 and 0.1 cell/mL are prepared and measured via microscopy
(95% confidence bounds: 0.84–1.16 and 0.094–0.106, respec-
tively). These dilutions are then measured in cell streaking mode
with the optimized wide-field cytometer, yielding average counts of
0.91 and 0.083 cells/mL, each with a 95% confidence bound of
0.85–0.97 and 0.065–0.102 cells/mL, respectively. In both cases
the two means are not significantly different at the 95% confidence
level.
Our technology has many promising clinical applications,
including detection of CTCs for POC. This may enable better
disease prognosis, earlier detection of metastasis-capable malig-
nancy, guidance in selection of treatments, and evaluation of treat-
ment efficacy to help prevent patients from being exposed to
ineffective treatments. The simplicity and low cost of the
webcam-based wide-field flow cytometer suggests that this config-
uration may have potential for developing clinically relevant POC
flow cytometry for rare cell detection in resource-poor settings.

3.10 Factors for In order to optimize the performance of a wide-field flow cyt-
Optimization of Streak ometer, the following method for setting various device parameters
Imaging Flow can be used. This will result in maximized cell image SNR, max-
Cytometer imized sample throughput, and accurate flow sampling.
1. The distance between the imaging lens and the flow cell should
be set such that images of cells are projected onto at most a
3
3 array of pixels. Distance should not be so great that a cell
image is less than one pixel in size. At those distances, photon
flux per pixel begins to diminish and SNR drops quickly. When
cells are imaged on more than one pixel, photon flux per pixel is
a constant and independent of the distance between flow cell
and lens. The largest possible distance should be chosen in
order to maximize the field-of-view (FOV). Set the flow cell
width so that it covers the FOV of the image sensor at this lens-
to-flow cell distance. This allows the largest possible flow cell
width to be used, thereby minimizing flow velocity and max-
imizing SNR at a given flow rate.
2. The relationship between average cell velocity and flow rate
should be determined empirically. The maximum velocity
achievable will be determined later and is a function of cell
image signal-to-noise ratio (SNR) (e.g., higher SNR allows
higher velocity). This in turn depends on the cell staining
 Streak Imaging Flow Cytometer for Rare Cell Analysis 283

characteristics (e.g., quantum yield of the dye, number of

fluorophores bound, and excitation field intensity), the camera
used (e.g., camera sensitivity, characteristic noise levels, and
maximum frame rate), and the lens used (e.g., focal ratio).
3. Exposure time should be set such that cell image brightness is
maximized. For a given average flow velocity and average cell
brightness (i.e., photon emission rate) there will be a maximum
cell image brightness that can be produced. This is based on the
number of photons than can strike a pixel as the cell image
passes over it.
4. Volumetric flow rate should be set such that the entire distri-
bution of cells is significantly above the noise background of
the system (this typically corresponds to an SNR > 5). At
higher flow rates image SNR will decrease. Above a certain
flow rate, cells at the faint end of the distribution will be
indistinguishable from background noise and the average cell
count will decrease. In order to set flow rate, a stock solution of
fluorescent cells should be prepared and counted beginning at a
high flow rate and decreasing until the average cell count per
sample reaches a constant value. As previously mentioned, this
maximum flow rate will depend on several factors including the
inherent brightness of cell staining and the excitation photon
flux. The parameters of exposure time and frame rate depend
directly on flow rate, so these parameters will need to be
updated as flow rate is varied in order to maintain similar levels
of device sensitivity across various tests.

4 Notes

1. Make sure the lens, filters, and flow cell are vertically centered
and aligned and that the laser illuminates the image area of the
flow cell.
Cell cultures with very high density may result in cell anu-
cleation (ghost cells), cells with low fluorescence signal when
labeled with nucleic acid stain.
2. Make sure the arrows on the coated filters are facing the cam-
era, and that if the filters being used have angular dependence
(e.g., interference filters) the excitation source beam is con-
fined to the correct range of angles.
3. A fast focal ratio (e.g., f/1.2) will enable shorter exposure time,
but focusing will be more difficult as the depth of field will be
reduced. Extension tubes are added between the lens and
sensor to allow for close focusing and enlargement of cell
images. This can result in noticeable image field curvature
(i.e., image corners out of focus while image center is in
focus), so spacing and imaging distance must be optimized.
Focal ratio can be reduced to reduce field curvature.
284 Joshua Balsam et al.

4. The imager enclosure must be sealed to block external light

sources. A long exposure can be used to detect light leaks.
5. For flow cell fabrication, the laser power and speed for cutting
polymers has to be determined empirically. The minimum laser
power is recommended in order to reduce overheating or
excessive burning of the material.
6. For strong bonding, all air bubbles between the surfaces and
the adhesive tape should be pressed out manually. The adhesive
layer should be attached to the acrylic side first to ensure that
proper alignment exists between the inlet and outlet ports.
Bubbles between these layers can be pressed out before the
protective paper layer on the back of the adhesive tape is
removed. After this, with the protective paper removed, the
glass layer can be bonded. Bubbles between the glass and
adhesive layer can be removed by carefully applied pressure,
or by use of a heated press. Care must be taken to ensure even
pressure because of the risk of glass breakage. To avoid this, the
glass layer can be replaced by another layer of acrylic. However,
the high autofluorescent emission rate of acrylic will result in
reduced SNR and diminished dynamic range of the sensor. For
improved sensitivity, two quartz microscope slides can be used
in place of the acrylic layer and the glass slide. Inlet and outlet
holes must be cut using a glass cutting drill bit or other appro-
priate method. High pressure ports can be constructed by
pressing an 18G needle into a laser cut hole in a small square
of acrylic sheet, sealing with cyanoacrylate glue, and bonding
over the inlet and outlet holes in the quartz slide using double
sided adhesive transfer tape with a center hole cut out.
7. For image quantification, the webcam performance will vary
depending both on the device and the application used to
collect the images from the camera. Prior to preparing a sample
for measurement it is important to understand the perfor-
mance of the camera and application being used. First, block
any light from reaching the camera sensor by taping a layer of
aluminum foil over the lens aperture. Take two images and
open them in ImageJ or a comparable software package. If a
color sensor is used, use only the green channel for analysis.
Convert the images to 32-bit float format, and subtract one
image from the other. Finally, use the software package to
produce a histogram of all pixel values. This histogram should
have an approximately Gaussian profile. If the original images
have pixel values that are predominantly zero valued, you must
change the settings of the camera capture software so that the
black end of the sensor histogram is not being artificially
clipped. This clipping behavior is typically set by an adjustment
labeled brightness. If the images captured by the sensor with
 Streak Imaging Flow Cytometer for Rare Cell Analysis 285

no light arriving cannot be made to show values above zero

with an approximately Gaussian profile, it is likely that the
device is not suitable for performing sensitive fluorescence
measurements of faint objects.

References
1. Balsam J, Bruck HA, Rasooly A (2015) Cell 14. Zhu H, Yaglidere O, Su TW, Tseng D, Ozcan
streak imaging cytometry for rare cell detec- A (2011) Cost-effective and compact wide-
tion. Biosens Bioelectron 64:154–160 field fluorescent imaging on a cell-phone. Lab
2. Balsam J, Bruck HA, Rasooly A (2015) Mobile Chip 11(2):315–322
flow cytometer for mHealth. Methods Mol 15. Golden JP, Kim JS, Erickson JS, Hilliard LR,
Biol 1256:139–153 Howell PB, Anderson GP et al (2009) Multi-
3. Balsam J, Bruck HA, Rasooly A (2014) wavelength microflow cytometer using groove-
Webcam-based flow cytometer using wide- generated sheath flow. Lab Chip 9
field imaging for low cell number detection at (13):1942–1950
high throughput. Analyst 139(17):4322–4329 16. Howell PB Jr, Golden JP, Hilliard LR, Erickson
4. Sun S, Ossandon M, Kostov Y, Rasooly A JS, Mott DR, Ligler FS (2008) Two simple and
(2009) Lab-on-a-chip for botulinum neuro- rugged designs for creating microfluidic sheath
toxin a (BoNT-A) activity analysis. Lab Chip 9 flow. Lab Chip 8(7):1097–1103
(22):3275–3281 17. Taitt CR, Anderson GP, Ligler FS (2005) Eva-
5. Sun S, Yang M, Kostov Y, Rasooly A (2010) nescent wave fluorescence biosensors. Biosens
ELISA-LOC: lab-on-a-chip for enzyme-linked Bioelectron 20(12):2470–2487
immunodetection. Lab Chip 10(16):2093–2100 18. Ngundi MM, Qadri SA, Wallace EV, Moore
6. Wei Q, Qi H, Luo W, Tseng D, Ki SJ, Wan Z MH, Lassman ME, Shriver-Lake LC et al
et al (2013) Fluorescent imaging of single (2006) Detection of deoxynivalenol in foods
nanoparticles and viruses on a smart phone. and indoor air using an array biosensor. Envi-
ACS Nano 7(10):9147–9155 ron Sci Technol 40(7):2352–2356
7. Coskun AF, Nagi R, Sadeghi K, Phillips S, 19. Moreno-Bondi MC, Taitt CR, Shriver-Lake
Ozcan A (2013) Albumin testing in urine LC, Ligler FS (2006) Multiplexed measure-
using a smart-phone. Lab Chip 13 ment of serum antibodies using an array biosen-
(21):4231–4238 sor. Biosens Bioelectron 21(10):1880–1886
8. Navruz I, Coskun AF, Wong J, Mohammad S, 20. Ligler FS, Sapsford KE, Golden JP, Shriver-
Tseng D, Nagi R et al (2013) Smart-phone Lake LC, Taitt CR, Dyer MA et al (2007)
based computational microscopy using multi- The array biosensor: portable, automated sys-
frame contact imaging on a fiber-optic array. tems. Anal Sci 23(1):5–10
Lab Chip 13(20):4015–4023 21. Kostov Y, Sergeev N, Wilson S, Herold KE,
9. Zhu H, Ozcan A (2013) Wide-field fluorescent Rasooly A (2009) A simple portable electrolu-
microscopy and fluorescent imaging flow cyto- minescence illumination-based CCD detector.
metry on a cell-phone. JoVE 74:e50451. Methods Mol Biol 503:259–272
10. Zhu H, Sencan I, Wong J, Dimitrov S, Tseng 22. Sapsford KE, Sun S, Francis J, Sharma S, Kos-
D, Nagashima K et al (2013) Cost-effective tov Y, Rasooly A (2008) A fluorescence detec-
and rapid blood analysis on a cell-phone. Lab tion platform using spatial electroluminescent
Chip 13(7):1282–1288 excitation for measuring botulinum neuro-
11. Zhu H, Sikora U, Ozcan A (2012) Quantum toxin A activity. Biosens Bioelectron 24
dot enabled detection of Escherichia coli using (4):618–625
a cell-phone. Analyst 137(11):2541–2544 23. Sun S, Francis J, Sapsford KE, Kostov Y,
12. Zhu H, Yaglidere O, Su TW, Tseng D, Ozcan Rasooly A (2010) Multi-wavelength Spatial
A (2011) Wide-field fluorescent microscopy on LED illumination based detector for in vitro
a cell-phone. Conf Proc IEEE Eng Med Biol detection of Botulinum Neurotoxin A Activity.
Soc 2011:6801–6804 Sens Actuators B 146(1-8):297–306
13. Zhu H, Mavandadi S, Coskun AF, Yaglidere O, 24. Rasooly A, Bruck HA, Kostov Y (2013) An
Ozcan A (2011) Optofluidic fluorescent imag- ELISA Lab-on-a-Chip (ELISA-LOC). Meth-
ing cytometry on a cell phone. Anal Chem 83 ods Mol Biol 949:451–471
(17):6641–6647
286 Joshua Balsam et al.

25. Rasooly A, Kostov Y, Bruck HA (2013) 27. Balsam J, Rasooly R, Bruck HA, Rasooly A
Charged-coupled device (CCD) detectors for (2014) Thousand-fold fluorescent signal
Lab-on-a Chip (LOC) optical analysis. Meth- amplification for mHealth diagnostics. Biosens
ods Mol Biol 949:365–385 Bioelectron 51:1–7
26. Balsam J, Bruck HA, Rasooly A (2013) Capil- 28. Balsam J, Ossandon M, Bruck HA, Lubensky I,
lary array waveguide amplified fluorescence Rasooly A (2013) Low-cost technologies for
detector for mHealth. Sens Actuators B medical diagnostics in low-resource settings.
186:711–717 Expert Opin Med Diagn 7(3):243–255
 Chapter 18

Rapid Detection of Microbial Contamination Using

a Microfluidic Device
Mustafa Al-Adhami, Dagmawi Tilahun, Govind Rao, Chandrasekhar
Gurramkonda, and Yordan Kostov

Abstract
A portable kinetics fluorometer is developed to detect viable cells which may be contaminating various
samples. The portable device acts as a single-excitation, single-emission photometer that continuously
measures fluorescence intensity of an indicator dye and plots it. The slope of the plot depends on the
number of colony forming units per milliliter. The device uses resazurin as the indicator dye. Viable cells
reduce resazurin to resorufin, which is more fluorescent. Photodiode is used to detect fluorescence change.
The photodiode generated current proportional to the intensity of the light that reached it, and an op-amp
is used in a transimpedance differential configuration to ensure amplification of the photodiode’s signal. A
microfluidic chip is designed specifically for the device. It acts as a fully enclosed cuvette, which enhances the
resazurin reduction rate. In tests, the E. coli-containing media are injected into the microfluidic chip and the
device is able to detect the presence of E. coli in LB media based on the fluorescence change that occurred in
the indicator dye. The device provides fast, accurate, and inexpensive means to optical detection of the
presence of viable cells and could be used in the field in place of more complex methods, i.e., loop-
meditated isothermal amplification of DNA (LAMP) to detect bacteria in pharmaceutical samples (Jimenez
et al., J Microbiol Methods 41(3):259–265, 2000) or measuring the intrinsic fluorescence of the bacterial
or yeast chromophores (Estes et al., Biosens Bioelectron 18(5):511–519, 2003).

Key words Contamination detection device, Resazurin, Resorufin, E. coli detection, Microfluidic
device, Thermal bonding of PMMA

1 Introduction

Each year thousands of people die and millions are infected due to
food, water, or medicine contamination [1]. In order to prevent
most of these poisonings a rapid, precise, and sensitive microbial
detection method is highly desirable [2, 3]. There are many com-
mon methods to detect pathogens and other biologics [4]. Most of
these methods require well-equipped and environmentally stable
laboratories as well as highly trained staff to handle devices and
reagents or antibodies [2]. These methods are hard to apply in the

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_18,
© Springer Science+Business Media LLC 2017

287
288 Mustafa Al-Adhami et al.

field and especially in the biotechnology industry where the final

product has to always be sterile to prevent infection of the patients
[1, 5]. The recent complex biologics are hard and sometimes
impossible to sterilize with the traditional methods of sterilization
[6]. Therefore, it has been of increasing importance to develop a
mechanically robust and easily handled device that can detect con-
tamination rapidly. Catching the contamination at early stage can
save both time and labor for the manufacturer and, more impor-
tantly, increases the safety of the final product. For example, having
near-real-time feedback in bioreactors can be very critical since each
batch takes many days to grow before harvest; it would be of great
interest to abort the process early in case of contamination [7].
Another important application of rapid microbial detection is
in the quality control processes for biopharmaceuticals. Since bio-
logics are highly susceptible for contamination by adventitious
agents such as mycoplasma, there is a need for risk mitigation
procedure like testing to confirm the absence of any unwanted
contaminants [8]. In this way, contaminated products can be
caught early, which minimizes the risk of having them produced
and then sold in pharmacies. Furthermore, if the device is low-cost
and low-effort, it could be used during the biological process to test
manufactured drugs for contamination [9].
Methods of contamination detection considered to be quick
usually refer to a time frame of one day. For example [9], Jimenez
et al. discuss the use of PCR analysis for detecting low levels of
bacteria and mold contamination in pharmaceutical samples; in this
case, the protocol states that the method can detect presence of
viable cells only in their exponential phase of growth, which might
take up to 24 h. The method mentioned using PCR analysis for
contamination detection above is very labor-intensive and takes
27 h to achieve results. It can detect contamination of colony
forming units as low as 10 CFU. It is not clear though whether
the detected cells are dead or alive at the time of detection. There
are other reported methods that are faster, for example measuring
the intrinsic fluorescence of the bacterial or yeast chromophores
[12]. This approach is fast and sensitive but it is limited to a slightly
higher number of colony forming units and it is very specific which
requires a high level of knowledge about the environment in which
the contamination is detected as to compensate for background
signal [13].
Cell viability assays are widely used in drug discovery applica-
tions to determine the ability of organs, cells, or tissues to live on
their own and develop. These assays are very useful for the rapid
detection and quantification of microorganisms [3, 10–13]. For
example, there are many assays like tetrazolium reduction, and
protease activity assays that measure different aspects of the viable
cell activity. They require an incubated population of viable cells to
convert a substrate to a colored product. On the other hand, ATP
 Rapid Detection of Microbial Contamination Using a Microfluidic Device 289

–
a O b
N
+

HO O O
Resazurin

Viable Cells

Reduction
Reaction

HO O O
Resorufin

Fig. 1 (a) The reduction of resazurin to resorufin chemical formula. (b) The color transformation when resazurin
is converted to resorufin

assays use a different approach in which adding a reagent ruptures

the cells and therefore it does not need any incubation time.
More recently, resazurin has been found to be an inexpensive
alternative that can provide a fluorescent readout at good sensitivity
(see Note 1) [14, 15].
Resazurin is a blue dye which itself is weakly fluorescent [14].
However, viable cells retain the ability to reduce resazurin into
resorufin, which is highly fluorescent [15]. Nonviable cells rapidly
lose metabolic capacity which is why they do not reduce this indi-
cator dye. Resazurin-based assays are used for viability detection of
cells other than bacterial cells, e.g., human cells for clinical trans-
plantation [16], stem cells [17], CD4 T cells [18], and malarial
gametocytocidal assay [19]. Also, resazurin-based assay can be
used for the screening of bacteria for the radiation sensitivity
(see Note 2) [20].
As shown in Fig. 1, the conversion of resazurin to resorufin
changes the color of the dye from blue to red, which is accompa-
nied with significant increase of the absorption green–yellow–green
region of visible light (λmax ¼ 570 nm). The fluorescence emission
in the orange–red region (λmax ¼ 590 nm) is enhanced as well.
These wavelengths are quite far from the absorption and emission
of the fluorophores in most media, which allows the assay to be
used directly in the cell culture media under test. The conversion
290 Mustafa Al-Adhami et al.

process gives a linear curve over a wide range of cell concentration

[21]. The resazurin assay is as sensitive as thymidine assay [22] for
detecting cell proliferation [23].
In this study, we present a highly sensitive, accurate, fast, and a
USB-powered portable device that can detect the presence of viable
cells in a given sample. We demonstrate that the device can detect
colony forming units as low as 10 CFU/mL in a given E. coli
sample in 30 min or less (see Note 3). The device results are then
validated using a standard plating method. This low-cost, tablet
powered device can be used for the field measurements of viable-
cell-induced fluorescent assay.
The main purpose of the device is to detect viable cells in the
tested samples. The device does not identify the species or the
growth phase of the cells that are detected. The main application
of the device is to do initial screening on sample which would be
followed by more traditional plate-based tests where the culture can
be grown and identified. However, having the ability to rapidly
screen a variety or samples irrespectively of the contaminating
species is of great value in the pharmaceutical and food industries.
The device provides means for reliable detection of contamination
in less than 30 min. Each reading is compared with a standard plate
count method. The device has proven to be accurate and fast, which
makes it suitable for rapid detection of contamination applications
(see Note 4). When used quantitatively, it is necessary to develop of
custom calibration curve such as that in Fig. 9. It will depend on the
type of species to be detected.

2 Materials

1. Green Light Emitting Diode (NSPG-500, Nichia).

2. Resistors RA—1 kΩ, RA—1 kΩ, RC—10 Ω, capacitor C—
330 pF.
3. Operational amplifier OPA 354 (Texas Instruments) for the
LED driver.
4. Excitation filter 532.0-35-75 (Intor, Soccorro, NM).
5. Emission filter 590.0-40-75 (Intor, Soccorro, NM).
6. Photodiode S12232-01 (Hamamatsu).
7. Two operational amplifiers OPA2301 (Texas Instruments) for
the photodetector amplifier.
8. Resistors RM—100 kΩ, RT—3 MΩ, CL—220 nF, RK—619 Ω,
RP—12 kΩ, CH-100 pF.
9. Analog-to-digital converter (16 bit ADC, ADS8318, Texas
Instruments).
 Rapid Detection of Microbial Contamination Using a Microfluidic Device 291

10. Digital-to-analog converter (12 bit DAC, DAC6571, Texas

Instruments).
11. Microcontroller (MSP430F2272 Texas Instruments).
12. E. coli NM303.
13. Luria–Bertani (LB)-agar plate to verify the viability of the
Escherichia coli NM303. Luria–Bertani broth is purchased in
the form of powder from MP Biomedicals (Santa Ana, CA).
The agar is purchased from Fisher Scientific.
14. Resazurin sodium salt powder is from Acros Organics (Fair
Lawn, NJ).
15. Plates are purchased from BD Biosciences (San Jose, CA).
16. Other reagents like NaCl, Na2HPO4, and NaH2PO4.
17. PBSr from Gibco.
18. Syringe Filters.
19. PMMA sheets purchased from Astra products (Thickness ¼ 0.2
and 1.5 mm).
20. Ethyl alcohol (90% concentration).

3 Methods

3.1 Portable Kinetics 1. The kinetics fluorometer is a single-excitation, single emission

Fluorometer photometer that can detect fluorescence and then plots it. The
block schematics of the device is presented in Fig. 2. The green
denotes the excitation light path (light source, filter, aperture,
direction of the excitation light), while the red denotes the
emission light path). Note that excitation and emission light
paths are oriented at 90 at each other to each other [26].

Fig. 2 Block schematics of the fluorometer. VCCS voltage controlled current

source, LED light emitting diode, uFc microfluidic cassette, ExF, EmF Excitation
and emission filters, PD photodiode, A amplifier, ADC analog to digital converter,
uP microprocessor, USB universal serial bus
292 Mustafa Al-Adhami et al.

Fig. 3 Schematics of the voltage-controlled current source for LEDs. The

extraction module consists of an LED with emission maximum 525 nm and
voltage controlled current source which sets a constants current through the LED

2. The excitation module consists of an LED with emission maxi-

mum 525 nm and voltage controlled current source which sets
a constant current through the LED. The schematics of the
module is shown in Fig. 3.
3. The current through the LED is controlled via the voltage
applied at Uin. With the suggested values, an input voltage in
the range between 0 and 1 V will result in current through the
LED between 0 and 50 mA. The capacitor decreases the spuri-
ous transients on the LED which result in unwanted electro-
magnetic high-frequency noise. The driver is good for
modulating the current through the LED for frequencies up
to 3 MHz.
4. An excitation filter must be mounted in front of the LED in
order to filter all the light that enters the cassette. Ideally, the
filter and the LED should be mounted in a holder that keeps
them mechanically aligned. The holder must be made of non-
fluorescent materials
5. Photodiode is employed to detect the fluorescence intensity by
generating current proportional to the intensity of the light
reaching them. The photodiode should be mounted at an angle
of 90 degrees to the excitation beam, with the emission filter
mounted in front of it.
6. The overall positioning of the LED, photodiode, the filters and
the cassette is presented in Fig. 4.
7. The first stage of the amplifier converts the current in voltage
with 107 transimpedance gain and second stage provides addi-
tional amplification of 20 (Fig. 5).
8. The second stage is separated from the first using a blocking
capacitor, which prevents the dc offsets from the output of the
first stage being amplified (Fig. 5). The high-pass filter formed
between CL and RK also significantly attenuates the ever-
present 60 Hz. The maximum amplification is in the band
between 5 and 20 kHz.
Rapid Detection of Microbial Contamination Using a Microfluidic Device 293

Fig. 4 Spatial positioning of the optics and optoelectronics. ExF Excitation Filter,
EmF Emission Filter, PD Photo diode, LED Light emitting diode

Fig. 5 Schematics of the differential photoreceiver Photodiodes do not exhibit

internal amplification as do photomultipliers or avalanche photodiodes,
therefore the photocurrent is converted to voltage using an op amp in
transimpedence differential configuration and additional amplification is
required

9. The output of the second stage is digitized using a fast analog-

to-digital converter(ADS8318, TI) with a maximum through-
put of 500 kilosamples per second. This is a 16-bit ADC with
SPI output. It is connected to the SPI port of microcontroller
(MSP430F2272, TI).
10. The ADC conversion is initiated by the microcontroller by
toggling the CONVST (conversion start) pin of the microcon-
troller. The microcontroller enters conversion phase, which
lasts for 1.4 μs. After the conversion end, ADC enters acquisi-
tion phase, during which the result is read by the microcon-
troller via the SPI protocol. It also performs the LED control
and times the acquisition of the intensities [22]. 997 readings
are initiated and collected and averaged by the microcontroller
294 Mustafa Al-Adhami et al.

as a single data point. The averaged value is transferred to the

computer via serial-to-USB bridge (FT232R, FTDI chip).
11. The computer control program is written in Labview. It allows
to set the number of the averaged readings (997 in this exam-
ple) and controls the brightness of light source (LED) via DAC
(DAC7571, TI). The computer initiates the measurements at a
prescribed interval (i.e., once a second) and stores the received
data points as an array. A second array with the time of mea-
surements is also stored. Every time a new data point is
obtained, it is added to the data array and the time at which it
has been acquired is added to the time array. These values are
approximated with a straight line. It is fitted to the data using
the least squares method. The slope of this line is the resazurin
reduction rate.

3.2 Microfluidic 1. A PMMA sheet of 1.5 mm in thickness is first cut using any
Cassettes available equipment (CNC mill, laser cutter). In order to
remove the internal stresses from the plastic sheet, it is annealed
using a PID controlled oven [25].
2. After the sheet is annealed, its surface is flattened with sand
paper [26], treated with 90% ethanol and immediately covered
with a 0.2 mm thick PMMA sheet acting as the cover sheet
[27].
3. The ethanol treated sheets are then clamped with a flat surface
clamp and placed in the oven at 55  C for 5 min (Fig. 6). After
the oven, the assembly is taken out to cool down and the
clamps are removed [28].
4. After the microfluidic chip is fabricated, holes are drilled on the
side of the device. These holes are then filled with silicone and
left to cure to act like a septum for the microfluidic chip.
Figure 1b shows one of the chip designs (see Note 4) [11].

a b 23.5mm
PMMA Assembly
PMMA

Ethanol

Vent
Removal of Clamps

The bonded sheets

Inlet
The Sheets are clamped and put in 6mm
the oven

Fig. 6 (a) Step sequence to bond two PMMA sheets. (b) Schematics of the microfluidics cassette
 Rapid Detection of Microbial Contamination Using a Microfluidic Device 295

3.3 Preparation and 1. Initially, 100 mL primary culture is prepared using 10 μL of E.

Plating of E. coli Cells coli NM303 cells, which is grown at 37  C in a shaker at
150 rpm (Labline Instruments, Melrose park, IL) for 8 h
until the optical density of primary culture is measured to be
2.5 at 600 nm (model 8453, Agilent Technologies, Santa
Clara, CA).
2. The primary seed culture (1%) is used to inoculate into 200 mL
secondary culture are grown at 37  C in a shaker at 150 rpm to
reach an optical density of 0.4 at 600 nm [3].
3. This sample is used for making serial dilutions (in 50 mL tubes)
from 1:10 to 1:105. The 1:105 diluted sample is used to make
further dilutions to make a final concentration of viable cells
which is calibrated against a standard plate count. An appropri-
ate volume is calculated (from the 1:105 aliquot) for the enu-
meration of 1000 CFU/mL to 10 CFU/mL and plated on
LB-agar plate.
4. It is tested for viable cell number using both standard plate
count method and the resazurin reduction test. The block-
schematics of the protocol are presented in Fig. 7.
5. One milliliter aliquots (in triplicates) of sample containing cells
from the serially diluted tubes are plated on LB-agar plate. The
plates are incubated at 37  C for 24 h. CFU are counted from
the each plate.

3.4 Preparation of 1. 2.3 mg of resazurin is mixed in a 1 mL of filtered water to

Resazurin Dye create a master mix. 100 μL of master mix is added to a 1.5 mL
centrifuge tube and diluted with 900 μL of filtered PBSr.
2. 980 μL PBS is then mixed with 20 μL of resazurin stock
solution to create reaction mix. The master mix is then
returned to the freezer to be used later.
3. 1 mL of the cell culture with the respective number of CFU is
deposited in an Eppendorf tube. Then, 3.3 μL of the freshly
made reaction mix is added to it and. The final mix is
vortexed and 300 μL are injected into the microfluidic cassette
(see Note 6). Two needles are used on the both sides of
the cassette. The first needle is used to inject the mix, while
the second needle serves as a vent for the air. After withdrawing
both needles, the cassette is inserted into the device cassette
holder. The leftover reaction mix is stored in a refrigerator to be
used within 12 h, otherwise it is discarded.

3.5 Measurement of 1. Once the cassette is in the device, the control program is
the Resazurin started. It continuously measures and displays the fluorescence
Reduction intensity. The control program also calculates the running
value of the slope of the fluorescence intensity change. If the
slope is close to zero (Fig. 8, squares), there is no
296 Mustafa Al-Adhami et al.

Fig. 7 Experimental protocolfor low CFU detection. The dashed box represents the actual measurement
(reprinted from the PDA Journal of Pharmaceutical Science and technology)

contamination. If the slope is significant (Fig. 8, triangles), the

media is contaminated.
2. To build a calibration curve, several dilutions of bacteria are
plated on LB-agar plates and grown for 24 h at 37  C. The
plates are analyzed for colony-forming units (CFU) with
respect to each dilution.
3. In parallel the bacterial dilutions are analyzed for resazurin
reduction over a time period of 3 min.

4 Notes

1. An even lower limit of detection can be achieved if the sample is

pre-concentrated. This can be achieved by aspirating the
 Rapid Detection of Microbial Contamination Using a Microfluidic Device 297

Fig. 8 Fluorescene intensity changes in contaminated vs. non-contaminated media. Triangles: Fluorescene
intensity increase in media contaminated with 10,000 CFU. Squares: fluroscene intensity stays practically
constant in sterile LB media. Both measurements performed at room temperature (25  C). (a.u. refers to
arbitrary unit)

sample into the syringe through a 0.2 μm filter. As a result, the

cells will be trapped on the filter. For example, a 100 mL sample
can be filtered fairly quickly. Then, a small percentage (i.e., 10%
or even 1%) of the filtered media is used to wash the cells from
the filter and into the Eppendorf tube. If now the same proto-
col is followed, ideally 10 or 100 times lesser CFU could be
detected. However, the approach requires additional validation
due to the possibility of for some of the cells to get stuck on the
filter or to be lysed during filtration.
2. The indicator (resazurin) is a blue dye which itself is weakly
fluorescent. Viable cells reduce resazurin into resorufin, which
is highly fluorescent. Nonviable cells do not have metabolic
capacity and do not reduce indicator dye. By examining the
increase of fluorescence intensity with time, we are able to
correspond the slope with the number of viable cells. It is
worth mentioning that the magnitude of fluorescence change
is small. Therefore, the device is working at relatively high
amplifications of the fluorescence signal (see Note 6). Strong
ambient light may interfere with device operation, even though
it is shielded (the cassette holder is made of black plastic). Care
should be taken to avoid placing it close to room lights.
3. To confirm the normal operation of the device, different
CFU/mL values are monitored and the slopes are determined
as in Fig. 8. The reduction rate slope of resazurin is correlated
to the number of bacterial cells in a given sample. The bacterial
cells (E. coli) concentrations ranged from 10 CFU/mL to
10,000 CFU/mL. All sample readings are analyzed in
298 Mustafa Al-Adhami et al.

Fig. 9 Correlation between the CFU number and the slope of fluorescence
increase. Samples below 10,000 CFU/mL were verified using agar plates while
the other numbers were estimated

triplicates (some of the standard deviations as shown in Fig. 9

are too small to be visible). As seen in the figure, there is a good
correlation between the resazurin reduction over the specified
range and the log number of viable cells.
4. The system can be used in the field for contamination detection
like the contamination of media. The portable fluorometer is
powered from a USB port of a computer and has been shown
to work with laptop or tablet featuring full USB port. It is
designed to house a specifically designed microfluidic cassette
(Fig. 1).
5. The cassettes are fully enclosed and have long input channels to
provide mixing. The design helps to decrease possible ingress
of air in the solution, which may interfere with the measure-
ment. The actual input and output shape of the chamber are
designed to feature gradual increase and decrease of the chan-
nel width. This prevented the formation of bubbles. The light
enters the cassette sideways from the edge, and the wide-area
photodiode picks up the fluorescence from the sample.
6. A control and visualization program written in Labview pro-
vides the timing for the fluorescence acquisition. It also can
calculate the slope value with every incoming reading and saves
the data for further analysis on the computer’s internal drive.
7. The protocol for field use of the device is developed with a goal
of simplicity. Only a syringe and an Eppendorf tube in addition
to the cassette and the portable fluorometer are used. The use
of a vortex mixer is optional, as the sample and the indicator
dye could be mixed by hand.
 Rapid Detection of Microbial Contamination Using a Microfluidic Device 299

References
1. Pharmtech.com (2015) An overview of rapid of human cells for clinical transplantation. Cell
microbial-detection methods|Pharmaceutical Transplant 15(2):169–174
technology. N.p. 2015. Web. 3 June 2015 17. Nagaoka M, Hagiwara Y, Takemura K, Mura-
2. Hoehl MM et al (2012) Rapid and robust kami Y, Li J, Duncan SA, Akaike T (2008)
detection methods for poison and microbial Design of the artificial acellular feeder layer
contamination. J Agric Food Chem 60 for the efficient propagation of mouse embry-
(25):6349–6358 onic stem cells. J Biol Chem 283
3. Hobson NS, Tothill I, Turner AP (1996) (39):26468–26476
Microbial detection. Biosens Bioelectron 11 18. Longhi MP, Wright K, Lauder SN, Nowell
(5):455–477 MA, Jones GW, Godkin AJ, Jones SA, Galli-
4. Fda.gov (2015) Archived BAM method: rapid more AM (2008) Interleukin-6 is crucial for
methods for detecting foodborne pathogens. recall of influenza-specific memory CD4 T
N.p. Web. 23 July 2015 cells. PLoS Pathog 4(2):e1000006
5. Vogel SJ, Tank M, Goodyear N (2013) Varia- 19. Tanaka TQ, Williamson KC (2011) A malaria
tion in detection limits between bacterial gametocytocidal assay using oxidoreduction
growth phases and precision of an ATP biolu- indicator, alamarBlue. Mol Biochem Parasitol
minescence system. Lett Appl Microbiol 58 177(2):160–163
(4):370–375 Web 20. Hudman DA, Sargentini NJ (2013) Resazurin-
6. Folsome CE (1964) Functional transformation based assay for screening bacteria for radiation
in mammalian cell culture systems. Nature 202 sensitivity. Springerplus 2(1):55
(4936):1023–1024 Web 21. Fields RD, Lancaster MV (1993) Dual-
7. Celsis.com (2010) Quality control—microbial attribute continuous monitoring of cell prolif-
testing: rapid microbiological methods in lean eration/cytotoxicity. Am Biotechnol Lab 11
manufacturing. N.p. Web. 23 July 2015 (4):48–50
8. Pettit AC, Kropski JA, Castilho JL, Schmitz JE, 22. Ahmed SA, Gogal RM Jr, Walsh JE (1994) A
Rauch CA, Mobley BC, Wang XJ, Spires SS, new rapid and simple non-radioactive assay to
Pugh ME (2012) The index case for the fungal monitor and determine the proliferation of
meningitis outbreak in the United States. N lymphocytes: an alternative to [3H]thymidine
Engl J Med 367(22):2119–2125 incorporation assay. J Immunol Methods 170
9. Gurramkonda C et al (2014) Fluorescence- (2):211–224
based method and a device for rapid detection 23. Al-Nasiry S, Geusens N, Hanssens M, Luyten
of microbial contamination. PDA J Pharm Sci C, Pijnenborg R (2007) The use of Alamar
Technol 68(2):164–171 Web Blue assay for quantitative analysis of viability,
10. Jimenez L, Smalls S, Ignar R (2000) Use of migration and invasion of choriocarcinoma
PCR analysis for detecting low levels of bacteria cells. Hum Reprod 22(5):1304–1309
and mold contamination in pharmaceutical 24. Kostov Y et al (2014) Portable system for the
samples. J Microbiol Methods 41(3):259–265 detection of micromolar concentrations of glu-
11. Ncbi.nlm.nih.gov 2015 Polymerase chain reac- cose. Measurement Science and Technology 25
tion (PCR). N.p. Web. 23 July 2015 (2):025701 Web
12. Estes C, Duncan A, Wade B, Lloyd C, Ellis W 25. Henderson RM, Selock N, Rao G 2012 Robust
Jr, Powers L (2003) Reagentless detection of and easy microfluidic connections in acrylic.
microorganisms by intrinsic fluorescence. Bio- Chips and Tips. http://blogs.rsc.org/
sens Bioelectron 18(5):511–519 chipsandtips/2012/04/23/robust-and-easy-
macrofluidic-connections-in-acrylic. Accessed
13. Brown NA (1990) Cell based assays. Develop- 16 Sept 2015
mental toxicity assays in vitro. Anal Proc 27
(9):246 Web. 26. Zhu X, Liu G, Guo Y, Tian Y (2007) Study of
PMMA thermal bonding. Microsystem Tech-
14. Bionity.com (2015) Alamarblue® assay for nology 13:403–407
assessment of cell proliferation using the Fluos-
tar OPTIMA. N.p. Web. 23 July 2015 27. Tran H, Wu W, Lee N (2013) Ethanol and UV-
assisted instantaneous bonding of PMMA
15. Boyce ST, Anderson BA, Rodriguez-Rilo HL assemblies and tuning in bonding reversibility.
(2006) Quantitative assay for quality assurance Sens Actuators B 181:955–962
of human cells for clinical transplantation. Cell
Transplant 15(2):169–174 28. Ng S, Tjeung R, Wang Z, Lu A, Rodriguez I,
Rooij N (2007) Thermally activated solvent
16. Boyce ST, Anderson BA, Rodriguez-Rilo HL bonding of polymers. Microsystem Technol-
(2006) Quantitative assay for quality assurance ogy 14:753–759
 Chapter 19

Resonance Energy Transfer-Based Nucleic Acid

Hybridization Assays on Paper-Based Platforms Using
Emissive Nanoparticles as Donors
Samer Doughan*, M. Omair Noor*, Yi Han, and Ulrich J. Krull

Abstract
Quantum dots (QDs) and upconverting nanoparticles (UCNPs) are luminescent nanoparticles (NPs)
commonly used in bioassays and biosensors as resonance energy transfer (RET) donors. The narrow and
tunable emissions of both QDs and UCNPs make them versatile RET donors that can be paired with a wide
range of acceptors. Ratiometric signal processing that compares donor and acceptor emission in RET-based
transduction offers improved precision, as it accounts for fluctuations in the absolute photoluminescence
(PL) intensities of the donor and acceptor that can result from experimental and instrumental variations.
Immobilizing NPs on a solid support avoids problems such as those that can arise with their aggregation in
solution, and allows for facile layer-by-layer assembly of the interfacial chemistry. Paper is an attractive solid
support for the development of point-of-care diagnostic assays given its ubiquity, low-cost, and intrinsic
fluid transport by capillary action. Integration of nanomaterials with paper-based analytical devices (PADs)
provides avenues to augment the analytical performance of PADs, given the unique optoelectronic proper-
ties of nanomaterials. Herein, we describe methodology for the development of PADs using QDs and
UCNPs as RET donors for optical transduction of nucleic acid hybridization. Immobilization of green-
emitting QDs (gQDs) on imidazole functionalized cellulose paper is described for use as RET donors with
Cy3 molecular dye as acceptors for the detection of SMN1 gene fragment. We also describe the covalent
immobilization of blue-emitting UCNPs on aldehyde modified cellulose paper for use as RET donors with
orange-emitting QDs (oQDs) as acceptors for the detection of HPRT1 gene fragment. The data described
herein is acquired using an epifluorescence microscope, and can also be collected using technology such as a
typical electronic camera.

Key words Quantum dots, Upconverting nanoparticles, Resonance energy transfer, Paper-based
bioassays, Nucleic acid hybridization

*
These authors contributed equally to this work.

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_19,
© Springer Science+Business Media LLC 2017

301
302 Samer Doughan et al.

1 Introduction

1.1 Quantum Dots in Quantum dots (QDs) are colloidal semiconductor nanocrystals
Resonance Energy with diameter in the range of 2–10 nm. [1] The two commonly
Transfer reported structural types of QDs are core QDs (e.g., CdSe, CdTe,
and CdS) and core/shell QDs (e.g., CdSe/ZnS, CdSe/CdS, and
CdSexS1
x/ZnS) [1]. The core/shell structural type is prevalent in
bioassay development as the shell passivates the optical properties
of QDs that originate from its core. QDs exhibit robust and unique
optical properties that originate from quantum confinement
effects. These properties include narrow (full-width-at-half-maxi-
mum in the range of 25–40 nm), symmetric and size tunable
emission spectra, greater photostability and brightness than organic
fluorophores, high quantum yields and a large extinction coefficient
over a broad range of wavelengths that stretches from the UV
region to their first exciton peak in the visible region [1]. These
properties are well suited for the use of QDs in optical multiplexing
for the development of bioassays. QDs are commercially available
with numerous surface chemistries allowing for easy conjugation of
biorecognition elements for use in bioassays [1].
QDs are popular RET donors due to their high quantum yield
and tunable photoluminescence (PL) spectra. The surface area of
QDs allows for immobilization of multiple recognition elements,
where a single QD can participate in multiple RET events [1]. Our
group has paired green-emitting QDs (gQDs) and red-emitting
QDs (rQDs) with various molecular dyes as RET acceptors in
nucleic acid hybridization assays [2, 3]. While QDs are popular
RET donors, their use as RET acceptors has been limited to time
gated measurements [4], chemiluminescence resonance energy
transfer (CRET) [5] and bioluminescence resonance energy trans-
fer (BRET) [6]. The broad absorbance band of QDs that stretches
into the UV region of the spectrum makes them susceptible to
direct excitation in the process of exciting the RET donor. We
have recently paired QDs as RET acceptors with UCNP donors
[7, 8]. UCNPs are excited in the near-IR and IR region and emit in
the UV to near-IR region of the spectrum. This permits the excita-
tion of the RET donor without the direct excitation of the QDs.
We have previously shown the utility of UCNP/QD RET pair for
bioassay development using analytes such as proteins [7] and
nucleic acids [8].

1.2 Upconverting UCNPs are luminescent nanocrystals that are typically tens of
Nanoparticles as nanometers in size. They exhibit anti-Stokes emission based on
Donors in Resonance the process of upconversion. Low-energy pump photons are accu-
Energy Transfer mulated in multiple long-lived excited-states of lanthanide ions that
are supported in an inert crystal lattice. The subsequent relaxation
to the ground electronic state results in the emission of radiation of
 Paper-Based Nucleic Acid Assay 303

higher energy than the excitation light [9]. A common class of

UCNPs is lanthanide doped inorganic crystals, such as NaYF4 and
Y2O3. The unique 4fn 5d0–1 electronic orbitals of Ln3+ ions, which
are shielded by the 5s2 and 5p6 sub-shell electrons, present energy
states that can be long lived (up to 0.1 s) and are thus ideal for UC
processes. There are five major upconversion processes observed in
lanthanide doped UCNPs: excited state absorption (ESA), energy
transfer upconversion (ETU), cooperative upconversion (CUC),
photon avalanche (PA), and energy migration upconversion
(EMU) [9]. Of these processes, ETU is the most efficient and is
based on co-doping the inorganic lattice with two lanthanide spe-
cies, a sensitizer (donor) and an activator (acceptor). For selection
criteria of dopants, refer to Ref. 9. The sensitizer ions are capable of
absorbing photons in the NIR or IR region of the spectrum and
effectively transfer the energy non-radiatively to an activator ion for
UC luminescence. The sequential excitation of one activator and its
subsequent relaxation to the ground electronic state results in an
anti-Stokes emission [9]. Herein, we use Yb3+ and Tm3+ as sensi-
tizer and activator, respectively, to obtain emission peaks in the UV
and blue regions of the spectrum. By changing the identity and
concentration of the dopant ions, different emission profiles can be
obtained. Lanthanide doped UCNPs are protected with an inert
shell of the same material as the host lattice to prevent non-radiative
relaxation pathways resulting from interactions with the high-
energy vibrational modes of surface ligands and collisions with
solvent molecules [9].
Our group has used UCNPs on paper as direct labels [10] and
as RET donors with molecular dyes [11] and QDs [8] as acceptors
in nucleic acid hybridization assays. Blue emitting NaYF4: 0.5%
Tm3+, 30% Yb3+/NaYF4 core/shell UCNPs have been paired
with orange emitting QD (oQD) acceptors [8]; green and red
emitting NaYF4: 2%Er3+, 18% Yb3+/NaYF4 core/shell UCNPs
have been paired with Cy3 and Cy5.5 dyes, respectively [11]. The
use of a near-IR excitation source minimizes background signals
due to the suppression of light scatter and autofluorescence [9].

1.3 Solid-Phase Solid-phase resonance energy transfer (RET)-based bioassays are

Assays characterized by immobilization of a donor nanoparticle (e.g., QDs
or UCNPs) on the surface of a solid support that is modified with a
suitable surface chemistry to facilitate its immobilization [1]. The
immobilized donor nanoparticle surface is then conjugated to a
biorecognition element to allow selective interaction with a
solution-phase analyte. The selective binding event modulates the
luminescence of donor nanoparticle, which is achieved by pairing
an appropriate acceptor (a fluorescent dye or another nanoparticle)
with the donor nanoparticle. The selective binding interaction
modulates the efficiency of energy transfer between the donor and
304 Samer Doughan et al.

the acceptor, where the resulting emission from an acceptor serves

as an analytical signal [1].
Solid-phase RET-based bioassays that make use of nanoparti-
cles as an active component of a transduction interface offer various
advantages in terms of assay development and analytical perfor-
mance when compared with the corresponding solution-phase
assays. Immobilization of nanoparticles on a solid support elimi-
nates the need to maintain their colloidal suspension, which allows
implementation of solution conditions (e.g., solvent composition
and reaction conditions) that may be beneficial to the analytical
performance of the assay [1, 12]. On the other hand, exposure of
such solution-phase conditions to a colloidal suspension of nano-
particles can potentially compromise their colloidal stability, the
latter being an integral requirement for the performance of
solution-phase nanoparticle assays. Immobilization of a biorecog-
nition element is greatly simplified in the case of solid-phase assays,
where the immobilized nanoparticles can be exposed to a high
concentration of reagents to facilitate bioconjugation. Excess
reagents can be rinsed from the surface without the need for
time-consuming purification steps that are typically encountered
in a multistep synthetic protocol. As a result, bioconjugation can be
readily achieved [1, 12]. A solid support that is modified with a
selective chemistry can be used to capture targets of interest from a
complex matrix. The selectivity of biomolecular interaction can be
improved by exposing the substrate to washes in order to suppress
the contribution of interferents prior to measurement [1, 12].
Solid-phase assays are also useful for the development of reusable
assays. The biomolecule–target complex can be dissociated by
changing the environment (e.g., by means of manipulation of
temperature, ionic strength or introduction of denaturing agents),
while retaining the immobilized selective chemistry on the surface
of a solid support for subsequent interrogation of another sample
solution for the analyte of interest [1, 12]. It should be noted that
reusability is an important feature of biosensors. Solid-phase nano-
particle assays are also amenable to integration with a number of
near-field optical transduction techniques, which include evanes-
cent wave excitation [13, 14] and plasmonics [15]. These techni-
ques not only improve the versatility of nanoparticle-based solid-
phase assays, but they can also have a pronounced effect on the
analytical performance of the assays.
Another significant advantage of solid-phase assays employing
nanoparticle mediated RET for signal transduction is the improve-
ment in assay sensitivity that arises from enhancement of energy
transfer efficiency at an interface. When donors and acceptors are
immobilized at sufficiently high density at an interface there are
additional energy transfer pathways between donors and acceptors,
where multiple donors can interact with a single acceptor in addi-
tion to the interaction of one donor with multiple acceptors [1]. At
 Paper-Based Nucleic Acid Assay 305

an interface, there are no discrete donor–acceptor pairs. Instead,

there is a two-dimensional plane of donors and acceptors. As a
result, the probability of energy transfer to a given acceptor
increases for solid-phase RET-based assays. This enhancement is
sufficient to compensate for the loss of available surface area of a
nanoparticle for selective interaction, which is experienced when
one face of a nanoparticle is blocked by immobilization to an
interface [1].

1.4 Paper as a There is a growing interest in the development of decentralized

Support for diagnostic assays that can be applied at the point-of-care and point-
Development of of-need settings in order to improve patient care and to make
Diagnostic Assays health care more accessible. In this regard, paper-based assays
have attracted considerable attention in recent years given the
advantageous attributes of paper substrates. These attributes
include: (1) low-cost and widespread commercial availability of
paper substrates with different pore sizes and flow rates [16]; (2)
cost-effective, high-throughput and ease of patterning of paper-
based analytical devices (PADs) using methods such as wax print-
ing, stamping, and drawing [17, 18]; (3) PADs can be operated
independent of any supporting equipment given the autonomous
fluid flow offered by the hydrophilic cellulosic fibers of paper sub-
strates; (4) well-established methods for surface modification of
cellulose for biomolecule immobilization; (5) an elegant approach
to eradicate biohazard by means of incineration of paper; (6)
requirement of a small reagent/sample volume; and (7) compati-
bility with biological samples [19]. A number of these attributes are
in congruence with the ASSURED (affordable, sensitive, specific,
user-friendly, rapid and robust, equipment-free, and deliverable to
end-users) guidelines set by the World Health Organization
in order to develop diagnostics for the developing world [19].
Examples of PADs include dipsticks assays, lateral flow devices,
paper-based 96-zone and 384-zone plates, microfluidic paper-
based analytical devices (μPAD), three-dimensional PADs and
origami PADs [20].

1.5 Paper-Based In this chapter, we describe methods for solid-phase nucleic acid
Solid-Phase Nucleic hybridization assays on a paper-based platform using QDs and
Acid Hybridization UCNPs as donors in RET-based transduction scheme. The paper
Assays Using the gQD/ substrates are patterned using a wax printing method and subse-
Cy3 and UCNP/QD RET quently chemically derivatized with functional groups that are suit-
Pairs able for the immobilization of the donor nanoparticles. In case of
the gQD/Cy3 (donor/acceptor) RET pair (Fig. 1a), the paper
substrates are chemically modified with imidazole groups to allow
immobilization of gQDs that are pre-modified with oligonucleo-
tide probes. Subsequent hybridization of the target strand and the
Cy3-labeled reporter strand brought the Cy3 acceptor dye in close
proximity to the surface of immobilized gQDs to allow RET-
306 Samer Doughan et al.

Fig. 1 Illustration of interfacial chemistry for paper-based solid-phase transduction of nucleic acid hybridiza-
tion using the (a) gQD/Cy3 RET pair and the (b) UCNP/QD RET pair. The paper substrates were patterned using
wax printing and the circular hydrophilic paper zones were derivatized with (a) imidazole groups for the
immobilization of QD–probe oligonucleotide conjugates and (b) aldehyde moieties for immobilization of
UCNPs. Sequential addition and hybridization of target and the (a) Cy3-labeled or the (b) QD-labeled reporter
oligonucleotides provided the necessary proximity for RET sensitized emission from the acceptor (a) Cy3 dye
or (b) oQD upon excitation of donor (a) gQDs (emission maximum at 525 nm) or (b) UCNP with (a) a UV or (b) IR
excitation source. Upon a selective hybridization event, the emission PL shifts from the donor to the acceptor
and the modulation of the donor and the acceptor PL serve as an analytical signal
 Paper-Based Nucleic Acid Assay 307

sensitized emission from the Cy3 acceptor dye, which served as an

analytical signal upon excitation of gQDs. In case of the UCNP/
oQD RET pair (Fig. 1b), the paper substrates are modified with
aldehyde groups to allow covalent immobilization of blue-emitting
UCNPs. The immobilized UCNPs are subsequently bioconjugated
to an oligonucleotide probe. The hybridization of the target strand
and the reporter strand that is bioconjugated to an oQD resulted in
a close proximity of the UCNP and the oQD, where the resulting
RET-sensitized emission from the oQD served as an analytical
signal upon excitation of UCNPs. This chapter also entails experi-
mental and data analysis methods for a ratiometric transduction of
nucleic acid hybridization using the aforementioned RET pairs.
Readout of the donor and acceptor emissions is described using
both an epifluorescence microscope and a low-cost electronic cam-
era employing colored digital imaging. Given the advantageous
attributes of paper substrates for the development of low-cost
diagnostic assays, the nucleic acid hybridization assays described
herein can potentially find applications in field-portable and remote
diagnostic applications.

2 Materials

2.1 Reagents 1. (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) (HEPES)

buffer: 100 mM HEPES (pH 7.2).
2. 1-(3-aminopropyl)imidazole (API, 97%).
3. 1.7 mL microcentrifuge tubes.
4. 2.5 M NaCl solution prepared using cartridge-purified water
(the Milli-Q water).
5. Anhydrous ethanol.
6. Avidin from egg white.
7. Borate buffer (BB1): 50 mM borate (pH 9.2).
8. Borate buffered saline (BB2): 50 mM borate (pH 9.2, 100 mM
NaCl).
9. Chloroform, reagent grade.
10. Dithiothreitol (>99%).
11. Ethyl acetate, reagent grade.
12. Hydrophobic green-emitting QDs (gQDs) and orange-emit-
ting QDs (oQDs). We use commercially available alkyl-ligand
coated ternary alloyed CdSexS1
x/ZnS QDs (gQDs with peak
PL at 525 nm and oQDs with peak PL at 575 nm) from
Cytodiagnostics Inc. (Burlington, ON, Canada).
13. Hexanes, reagent grade.
14. L-glutathione, reduced (GSH).
308 Samer Doughan et al.

15. Lithium chloride (LiCl, anhydrous, ACS reagent, 99%).

16. Maleimidohexanoic acid—G(Aib)GHHHHHH from Can-
Peptide (Pointe-Claire, QC).
17. Methanol, reagent grade.
18. NHS-PEG4-biotin from Thermo Scientific (Rockford, IL,
USA).
19. Octadecene, reagent grade.
20. Oleic acid, reagent grade, 90%.
21. Phosphate-buffered saline (PBS).
22. Phosphorylethanolamine (PEA).
23. Sodium (meta)periodate (NaIO4, 99%).
24. Sodium cyanoborohydride (NaCNBH3, reagent grade, 95%).
25. Sodium hydroxide.
26. Sterile ultrapure Milli-Q water (specific resistance
18 MΩ cm).
27. Tetramethylammonium hydroxide (TMAH): 25% w/w solu-
tion in methanol.
28. Thulium (III) acetate hydrate (99.9%).
29. Tris(2-carboxyethyl)phosphine hydrochloride (TCEP) powder
(98%).
30. Whatman® cellulose chromatography papers (Grade 1,
20 cm  20 cm).
31. Ytterbium(III) acetate hydrate (99.9%).
32. Yttrium(III) acetate hydrate (99.9%).
33. Oligonucleotide sequences listed in Table 1 (see Note 1). The
sequences are from Integrated DNA Technologies (Coralville,
IA, USA).

2.2 Oligonucleotide See Table 1.

Sequences for the
Hybridization Assays

2.3 Instrumentation 1. Nikon Eclipse L150 epifluorescence microscope (Nikon, Mis-

and Equipment sissauga, ON, Canada) custom fitted with the following com-
ponents: 4 Nikon WD Plan Fluor objective lens (NA ¼ 0.13);
a filter cube comprising ZET405/20 as an excitation filter,
Z405rdc as a dichroic mirror and a HQ430lp as an emission
filter (Chroma Technologies Corp., Bellows Falls, VT, USA); a
25 mW diode laser excitation source with an output of 402 nm
(Radius 402, Coherent Inc. Santa Clara, CA, USA) and a diode
array spectrometer (QE65000, Ocean Optics Inc., Dunedin,
FL, USA) as a detector.
 Paper-Based Nucleic Acid Assay 309

Table 1
Probe, target and reporter oligonucleotide sequences used in the hybridization assays with the
gQD/Cy3 and UCNP/QD RET pairs

Name Sequence (50 to 30 direction)

gQD/Cy3 RET
SMN1 probe DTPA-50 -ATT TTG TCT GAA ACC CTG T-30
SMN1 FC target 50 -TCC TTT ATT TTC CTT ACA GGG TTT CAG ACA AAA T-30
SMN1 Cy3 reporter Cy3-50 -AAGGAAAATAAAGGA-30
UCNP/QD RET pair
HPRT1 probe 50 -Biotin-CAA AAT AAA TCA AGG TCA-3’
HPRT1 FC target 50 -GAT GAT GAA CCA GGT TAT GAC CTT GAT TTA TTT TG-3’
HPRT1 reporter 50 -TAACCTGGTTCATCATC-Thiol-3’
Abbreviations: FC fully complementary, DTPA dithiol phosphoramidite, Cy3 cyanine 3

2. Nikon Eclipse L150 epifluorescence microscope (Nikon) cus-

tom fitted with the following components: 40 Nikon WD
Plan Fluor objective lens (NA ¼ 0.60); a filter cube comprising
a ZET964/86 bp excitation filter (Chroma), a ZT1064rdc-sp
dichroic mirror (Chroma) and a 455/35 (Nikon) or 570/20
(Chroma) emission filters; a 2.5 W tunable 980 nm collimated
diode laser (Laserglow Technologies, Toronto, ON, CAN) and
a H5784-20 photomultiplier tube (Hamamatsu, Bridgewater,
NJ, USA).
3. Handheld ultraviolet (UV) lamp (UVGL-58, LW/SW, 6W,
The Science Company®, Denver, CO, USA).
4. HP8452A Diode-Array Spectrophotometer (HP Corporation,
Palo Alto, CA, USA).
5. iPad mini (Apple, Cupertino, CA, USA).
6. Low-binding polypropylene microcentrifuge tubes, 1.7 mL
capacity.
7. Micropipettes and tips.
8. Neutral density (ND) filters (ND 4, ND 8 and ND 16).
9. Orbital shaker.
10. 0.2 μm Polyethersulfone (PES) syringe filters.
11. Xerox ColorQube 8570DN solid ink wax printer (Xerox
Canada, Toronto, ON, Canada).
12. Amicon Ultra-0.5 mL 100 kDa centrifugal filters (Millipore
Corporation, Billerica, MA, USA).
310 Samer Doughan et al.

3 Methods

3.1 UCNP Synthesis 1. Add 0.4562, 0.2534 and 0.0042 g of Y(CH3CO2)3xH2O, Yb

(CH3CO2)34H2O, Tm(CH3CO2)3xH2O and a stir bar into a
3.1.1 Synthesis of Core
100 mL three-neck round bottom flask.
NaYF4: 0.5% Tm3+, 30%
Yb3+ UCNPs 2. Add 30 mL of octadecene and 12 mL of oleic acid into the
round bottom flask.
3. Place the round bottom flask on a heating mantle controlled by
a temperature precision controller.
4. Insert the precision controller thermometer in the central neck
of the three-neck round bottom flask through a septum. Attach
a two-neck collection round bottom flask via a bent adaptor to
one of the side necks of the reaction flask. Seal the remaining
necks of the reaction and the collection flasks with septa (see
Note 2).
5. Stir the mixture gently under vacuum at 115  C for 30 min.
The solution should turn clear and colourless at this point.
6. Cool the mixture to 50  C under a gentle stream of argon.
Insert the in-line via a needle in the unused neck of the three-
neck round bottom flask. Insert the out-line via a needle in the
unused neck of the collection round bottom flask. The argon
flow should be maintained for the remainder of the synthesis.
7. Prepare a 20 mL methanol solution containing 0.20 g NaOH
and 0.30 g NH4F via sonication. Shake the solution occasion-
ally to help dissolve the solids (see Note 3). Add the methanol
solution to the reaction mixture. The solution will turn cloudy.
8. Allow the reaction to stir for 30 min at 50  C before heating it
to 75  C to evaporate the methanol. The solution should
become clear at this point (see Note 4).
9. Increase the temperature to 300  C rapidly and maintain it for
1 h. Wrap the reaction flask with glass fibre to help keep the
solution warm (see Note 5).
10. Allow the reaction mixture to cool to room temperature under
argon.
11. Add an equal volume of absolute ethanol to the reaction mix-
ture and centrifuge the solution at 4500 rpm to collect the
UCNPs.
12. Resuspend the UCNPs in hexanes and repeat step 11. Repeat
twice.
13. Store the washed core UCNPs in hexanes in a glass vial at 4  C
for subsequent growth of shell.
 Paper-Based Nucleic Acid Assay 311

3.1.2 Synthesis of NaYF4: 1. Add 0.5738 g of Y(CH3CO2)3xH2O into a 100 mL three-

0.5% Tm3+, 30% Yb3+/ neck round bottom flask.
NaYF4 Core/Shell UCNPs 2. Follow steps 2–5 in Subheading 3.1.1.
3. Cool the mixture to 80  C under argon and add the core
UCNPs from step 13.
4. Cool the reaction temperature to 50  C after all the hexane has
evaporated.
5. Prepare a 20 mL methanol solution containing 0.14 g NaOH
and 0.26 g NH4F via sonication. Add the mixture to the
reaction vessel. The reaction mixture will turn cloudy (see
Note 3).
6. Allow the reaction to stir for 30 min before heating the mixture
to 75  C to evaporate the methanol. The solution should be
clear at this point (see Note 4).
7. Increase the temperature to 300  C rapidly and maintain it for
1 h (see Note 5).
8. Allow the reaction mixture to cool to room temperature under
argon.
9. Add an equal volume of absolute ethanol and centrifuge at
2173 g to collect the UCNPs.
10. Resuspend the UCNPs in hexanes and repeat step 9. Repeat
twice.
11. Store the oleic acid capped core/shell UCNPs in hexanes or
toluene in a glass vial at 4  C for subsequent modification.

3.2 Preparation of 1. Mix 2 mL of hexanes containing 100 mg of oleic acid capped

Water Soluble QDs and core/shell UCNPs from step 11 in Subheading 3.1.2, 400 mg
UCNPs of phosphorylethanolamine (PEA) and 1 mL of tetramethy-
lammonium hydroxide (TMAH) in 10 mL of absolute ethanol
3.2.1 Preparation of in a capped glass vial.
Water Soluble UCNPs
2. Stir the reaction vigorously overnight at 70  C in an oil bath.
3. Allow the reaction to cool to room temperature and recover
PEA capped UCNPs by centrifugation at 1315 g.
4. Resuspend the UCNPs in ethanol via sonication and mix the
solution with an equal volume of hexanes before centrifugation
at 4500 rpm.
5. Repeat step 4 two times.
6. Resuspend the washed PEA capped UCNPs in 10 mL of water
and pass the solution through a 0.2 μm polyether sulfone (PES)
syringe filter to remove aggregates. If desired, evaporate water
using a rotary evaporator to concentrate UCNPs.
7. Store the PEA capped UCNPs in excess PEA at 4  C.
312 Samer Doughan et al.

3.2.2 Preparation of 1. Add 75 μL of 10 μM alkyl QDs into 2 mL of chloroform in a

Water Soluble QDs glass vial.
2. Dissolve 0.2 g of reduced L-Glutathione (GSH) in 600 μL of
TMAH.
3. Add the solution of QDs drop-wise to the GSH solution while
swirling.
4. Allow the cloudy mixture to sit overnight in the dark.
5. Extract water soluble GSH coated QDs using BB2 in 100 μL
fractions. Place the collected fractions in a 1.7 mL microcen-
trifuge tube.
6. Add an equal volume of ethanol and collect the QDs by centri-
fugation at 6869 g for 5 min.
7. Resuspend the QDs in 200 μL of BB2.
8. Repeat steps 6 and 7 two more times.
9. Resuspend the GSH-QDs in BB1.
10. Measure the absorbance spectrum of the QDs and use their
first exciton peak and known molar extinction coefficient to
determine the NP concentration.
11. Store the GSH-QD at 4  C for subsequent use.

3.3 Bioconjugation 1. Incubate thiol-terminated oligonucleotides (27 nmol in

of Oligonucleotides to 100 μL) with 500 equivalents of dithiothreitol (DTT) (27 μL
QDs of 0.5 M DDT solution in 1 PBS) and 100 μL of 1 PBS
buffer for 1 h at room temperature to reduce disulfides moi-
3.3.1 Bioconjugation of eties into sulfhydryl groups.
Hexahistidine-Terminated
Oligonucleotides to GSH- 2. Extract excess DTT with 600 μL of anhydrous ethyl acetate
QDs four times.
3. Add 6-maleimidohexanoic acid—G(Aib)GHHHHHH
(0.7 mg in 30 μL of DMSO) to the oligonucleotide solution
in 20 times molar excess and allow the reaction mixture to
shake for 12 h at room temperature (see Note 6).
4. Purify the modified oligonucleotides using a NAP-5 desalting
column as per the manufacturer’s instructions.
5. Quantify the purified hexahistidine-modified oligonucleotides
by UV-vis spectroscopy (λmax ¼ 260 nm).
6. Store the hexahistidine-modified oligonucleotides at
20  C.
7. Incubate hexahistidine-modified oligonucleotides (HPRT1
reporter) with GSH-QDs at the desired QD:DNA ratio for
1 h in BB2 in a 1.7 mL microcentrifuge tube on an orbital
shaker (see Note 7).
8. Purify the hexahistidine-modified oligonucleotides by centrifu-
gation three times using an Amicon Ultra-0.5 mL 100 kDa
 Paper-Based Nucleic Acid Assay 313

centrifugal filter as per the manufacturer’s instructions to

remove excess nucleic acids.
9. Store the washed nucleic acid conjugated QDs in BB2 at 4  C.

3.3.2 Bio-Conjugation of 1. Prepare 50 mM solution of TCEP by dissolving 7 mg of TCEP

DTPA-Terminated in 500 μL of borate buffer (BB1).
Oligonucleotides to GSH- 2. To a 1.7 mL microcentrifuge tube, pipette 169 μL of borate
QDs buffered saline (BBS).
3. To the same tube, pipette 82 μL of 50 mM TCEP solution that
has been prepared in step 1.
4. Add 8.2 nmol of DTPA terminated oligonucleotide probe
(SMN1 probe) to a solution prepared in step 3 and place the
tube on an orbital shaker for 15 min (see Note 8).
5. Pipette 200 pmol of GSH-gQDs to the solution in step 4 (see
Note 9).
6. Pipette additional volume of borate buffered saline (BB2) to
the solution in step 5 such that the final solution volume is
493 μL. For the example of QD–probe conjugates preparation
provided in this section, an additional 169 μL of borate buff-
ered saline (BB2) will need to be added.
7. Agitate the contents of the tube overnight on an orbital shaker.
8. Prepare a fresh 50 mM TCEP solution using step 1.
9. Pipette a 40 μL aliquot of 50 mM TCEP solution from step
8 into the contents of the tube in step 7.
10. Incrementally pipette 100 μL of 2.5 M NaCl solution to the
tube in step 9. We use an interval time of 10 min for each
sequential addition of 10 μL of 2.5 M NaCl (see Note 10).
11. Place the tube in step 10 on an orbital shaker for an overnight
incubation and subsequently store the contents of the tube
(QD–probe conjugates) inside a fridge at 4  C until further
use. The concentration of QD–probe conjugates in this solu-
tion is ca. 312 nM (total solution volume is 633 μL).

3.4 Wax Printing and 1. Using an appropriate software, draw a pattern of the paper
Chemical Modification device that is to be printed on the Whatman® cellulose chro-
of Paper matography paper substrates. We use AutoCAD 2012 software
(see Note 11). For the purpose of the work described in this
3.4.1 Wax Printing of chapter, the dimensions of each paper device is 25 mm  60 mm
Paper Substrates and a single sheet of 20 cm  20 cm Whatman® chromatogra-
phy paper substrate comprised 18 paper devices (arranged in a
6  3 array format). We use paper devices that contain 32
circular zones with a diameter of 5 mm (see Note 12), arranged
in a 4  8 array format. We print on bare paper, and the zones
are surrounded by a filling with black ink.
314 Samer Doughan et al.

2. Print the paper substrates using a Xerox ColorQube 8570DN

solid ink wax printer (see Note 13).
3. Preheat an oven or a hot plate to 120  C.
4. Place the paper sheet (printed side up) inside the oven or on
top of the hotplate for 2.5 min to confine the hydrophilic paper
zones across the thickness of the paper substrate (see Note 14).
5. Allow the paper sheet to cool to room temperature.
6. Using scissors, trim the patterned paper sheet to individual
paper devices.

3.4.2 Chemical 1. Attach a binder clip to the end of a wax-printed paper substrate
Derivatization of Paper in order to suspend the paper device in air. We do so by
Zones with Aldehyde inserting a micropipette tip in the arms of the binder clip
Functionality such that the bottom of the micropipette tip rests inside one
of the holders of the microcentrifuge tray rack (see Note 15).
2. Prepare 1.4 M solution of LiCl by dissolving 30 mg of LiCl in
500 μL of Milli-Q water.
3. Prepare 94 mM solution of NaIO4 by dissolving 10 mg of
NaIO4 in 500 μL of Milli-Q water.
4. Mix the two solutions prepared in steps 2 and 3 in equal
volumes.
5. Spot a 5 μL aliquot of the solution prepared in step 4 onto each
paper zone and incubate the spotted paper devices inside an
oven set at 50  C for 30 min.
6. Repeat step 5.
7. Wash the paper devices three times with Milli-Q water (see
Note 16).
8. Place the washed paper devices on any form of absorbent
blotting paper in order to wick off the excess water.
9. Dry the paper devices inside a vacuum desiccator.

3.4.3 Chemical 1. Suspend aldehyde modified paper substrates in air using a

Derivatization of Aldehyde binder clip as mentioned previously.
Modified Paper Zones with 2. Prepare 320 mM solution of NaCNBH3 by dissolving 10 mg of
Imidazole Functionality NaCNBH3 in 500 μL of 100 mM HEPES buffer (pH 8.0).
3. To 400 μL of 320 mM solution of NaCNBH3 prepared in step
2, pipette 12 μL of API. The concentration of API in the
resulting solution is 0.24 M.
4. Pipette a 5 μL aliquot of the solution prepared in step 3 onto
each paper zone that has been modified with the aldehyde
functionality.
5. Incubate the spotted paper substrates under ambient condi-
tions for 1 h.
 Paper-Based Nucleic Acid Assay 315

6. Submerge each paper device inside a 50 mL conical centrifuge

tube filled with BB1.
7. Place the conical centrifuge tubes on an orbital shaker for
15 min in order to wash the paper devices.
8. Remove the imidazole modified paper substrates from the
conical tubes and place them on any form of an absorbent
blotting paper in order to wick off the excess solution.
9. Dry the paper substrates inside a vacuum desiccator.

3.5 Immobilization of 1. Suspend imidazole modified paper substrates in air using a

QDs and UCNPs on binder clip as mentioned previously.
Paper 2. Spot a 3 μL aliquot of 312 nM solution of QD–probe con-
3.5.1 Immobilization of jugates solution onto each paper zone that has been modified
QD–Probe Oligonucleotide
with imidazole functionality and incubate for 1 h at room
Conjugates on Imidazole temperature in dark (see Note 17).
Modified Paper Substrates 3. Wash the paper devices for 15 min with BB1 using an orbital
shaker by submerging each paper device inside a 50 mL conical
centrifuge tube containing BB1.
4. Wick off the excess solution from the paper substrates using any
form of absorbent blotting paper and dry the QD–probe con-
jugates modified paper substrates inside a vacuum desiccator.

3.5.2 Immobilization of 1. Mix 200 μL of PEA-UCNPs (3 mg/mL) with 200 μL of a

UCNPs on Aldehyde 0.1 mM sodium cyanoborahydride solution in HEPES buffer
Modified Paper Substrates (100 mM, pH 7.2).
2. Suspend aldehyde modified paper substrates in air using a
binder clip as mentioned previously.
3. Pipet 5 μL of the solution in step 1 onto each aldehyde mod-
ified paper zone.
4. After incubating for 10 min., wash the paper substrate in 0.1%
v/v aqueous Tween® 20 solution for 5 min in a conical centri-
fuge tube.
5. Wash the paper substrate with purified water for 2 min.
6. Place the washed paper substrate on any form of absorbent
blotting paper to wick off the excess water.
7. Dry the paper substrate inside a vacuum desiccator.

3.5.3 Layer-By-Layer 1. Suspend paper substrates with immobilized UCNPs in air

Assembly of the Assay using a binder clip as mentioned previously.
Using Immobilized UCNPs 2. Pipet 3 μL of a freshly prepared 2 mM aqueous NHS-PEG-
on Paper Substrates biotin solution onto each spot of the paper substrate contain-
ing immobilized UCNPs and allow to air dry.
3. Wash the paper for 2 min with purified water in a conical
centrifuge tube.
316 Samer Doughan et al.

4. Place the washed paper substrate on an absorbent blotting

paper to wick off the excess water before placing it inside a
vacuum desiccator to dry.
5. Pipet 5 μL of a 20 μM avidin solution in HEPES buffer
(100 mM, pH 7.2) onto each paper zone. Allow the zones to
air dry.
6. Wash the paper for 2 min with BB1 in a conical centrifuge tube.
7. Place the washed paper substrate on an absorbent blotting
paper to wick off the excess water before placing it inside a
vacuum desiccator to dry.
8. Pipet 5 μL of 10 μM biotinylated oligonucleotide probe
(HPRT1 probe) solution in BB1 onto each paper zone. Allow
the paper to air dry.
9. Repeat steps 6 and 7.

3.6 Calibration Curve 1. Prepare dilutions of the SMN1 FC or HPRT1 FC target in the
of Target DNA concentration range of 20 nM to 15 μM using 50 mM borate
buffered saline (pH 9.2, 100 mM NaCl) (see Note 18).
2. Prepare a dilution of SMN1 Cy3 reporter at 10–15 μM con-
centration or HPRT1 oQD reporter at 2 μM using borate
buffered saline (BB2).
3. Spot 3 μL of borate buffered saline (BB2) onto the paper zones
that belong to the top row of the paper device (see Note 19).
4. Spot a 3 μL aliquot of SMN1 FC or HPRT1 FC target at
various concentrations onto each of the different rows of the
paper device (excluding the first row, see point 3). In order to
collect replicate measurements, the same concentration of the
target can be spotted onto the paper zones that belong to the
same row of the paper device (see Note 20).
5. Allow the targets/sample solutions to incubate on the paper
device for 1 h (see Note 21).
6. Spot a 3 μL aliquot of 10–15 μM solution of SMN1 Cy3
reporter or 2 μM HPRT1 oQD reporter onto each of the
paper zones that have been subjected to the target/sample
hybridization and allow the reporter solution to incubate on
the paper device for 30 min (see Note 22).
7. Wash the paper substrates for 5 min with BB2 using an orbital
shaker by submerging each paper device inside a 50 mL conical
centrifuge tube containing BB2. Add 1% Tween® 20 by volume
to the wash solution when QDs are used as acceptors.
8. Wick off the excess solution from the paper substrates using an
absorbent blotting paper and dry the paper substrates inside a
vacuum desiccator prior to the data collection (see Note 23).
 Paper-Based Nucleic Acid Assay 317

3.7 Data Acquisition 1. gQD/Cy3 RET Pair.

(a) Using the epifluorescence microscope platform (402 nm
laser excitation source and diode array spectrometer as a
detector), measure the PL spectrum from each paper zone
that has been modified with the selective chemistry
(immobilized QD–probe conjugates) (see Note 24).
Also, collect a background spectrum from imidazole
modified paper zone (without immobilized QD–probe
conjugates). Proceed to subheading 3.8, step 1 for data
analysis.
(b) Representative data acquired using the epifluorescence
microscope platform is shown in Fig. 2b and c.
(c) For the hybridization assays using the gQD/Cy3 RET
pair, the red-green-blue (RGB) color palette of a digital
camera can also be used to acquire quantitative informa-
tion. Using the long wavelength (365 nm) setting of a
handheld UV lamp (UVGL-58, LW/SW, 6W, The Sci-
ence Company®), illuminate the paper devices for the
excitation of immobilized QDs on paper substrates and
collect a colored digital image of the paper device under
the default settings of the built-in Camera application of
an iPad mini in a dark environment (see Note 25). Pro-
ceed to Subheading 3.8, step 1 for data analysis.
(d) Representative data from the iPad detection platform is
shown in Fig. 2a and c.
2. UCNP/QD RET Pair.
(a) Using the epifluorescence microscope platform with a
980 nm laser excitation source and the PMT detector,
acquire the PL image from each paper zone. Appropriate
optical emission filters are used to collect the UCNP and
QD signals independently. Herein, we use 455/35 nm
and 570/20 nm band-pass filters to collect UCNP and
QD luminescence, respectively. The paper is scanned on
the microscope stage and an image is obtained using Lab-
View (see Note 26). Proceed to subheading 3.8, step 1 for
data analysis.
(b) Representative data from the epifluorescence detection
platform is shown in Fig. 3.

3.8 Data Analysis 1. Analysis of PL spectra for gQD/Cy3 RET Pair.

(a) Subtract background spectrum from each PL spectrum
that has been collected from the paper zones with immo-
bilized QD–probe conjugates.
(b) Normalize each background corrected PL spectrum to
the maximum PL intensity of gQDs, i.e., divide all the
PL intensities by the maximum PL intensity of gQDs (the
318 Samer Doughan et al.

Fig. 2 Solid-phase QD-RET transduction of nucleic acid hybridization on paper substrates using the gQD/Cy3
RET pair. (a) Colored digital image and the corresponding pseudo-colored PL images of gQDs (G channel) and
Cy3 (R channel) after R-G-B splitting of the colored digital image with increasing amount of SMN1 FC TGT. The
amounts of SMN1 FC TGT in (i) to (viii) were 0, 0.94, 1.9, 3.8, 7.5, 15, 30, and 45 pmol, respectively. (b)
Normalized PL spectra acquired using the epifluorescence microscope platform showing the assay response
with increasing amount of SMN1 FC TGT. The amounts of SMN1 FC TGT in (i) to (xii) were 0, 0.057, 0.12, 0.23,
0.47, 0.94, 1.9, 3.8, 7.5, 15, 30, and 45 pmol, respectively. (c) Calibration curves showing the RET ratio
response (red) and the R/G ratio response (black) with increasing amount of SMN1 FC TGT. Figure adapted
with permission from Ref. 21. Copyright 2014 American Chemical Society

λmax for gQDs is in the range of 520–530 nm). The

resulting spectra will depict profiles as shown in Fig. 2b.
(c) Calculate RET Ratio for each background subtracted and
normalized PL spectrum using Eq. 1. In Eq. 1, the wave-
length range of 560–590 nm in the numerator of each
term is used to integrate the Cy3 (acceptor) PL intensity,
while the wavelength range of 510–540 nm in the
denominator of each term is used to integrate the gQD
(donor) PL intensity. The subscripts DA and D denote
 Paper-Based Nucleic Acid Assay 319

Fig. 3 Solid-phase UCNP/QD (donor/acceptor) RET transduction of nucleic acid hybridization on paper
substrates. (a) Pseudo-colored microscope images of UCNPs (blue channel), oQDs (orange channel) and an
overlay of both images with increasing amount of HPRT1 FC target. The amounts of HPRT1 FC target ranged
from 15 fmol to 6 pmol. (b) Calibration curve showing the RET ratio with increasing amount of HPRT1 FC target

measurements made in the presence and absence of the

acceptor, respectively (see Note 27).
0λ¼590 1 0λ¼590 1
P P
B PLðλÞC B PLðλÞC
Bλ¼560 C Bλ¼560 C
FRET ratio ¼ Bλ¼540 C
Bλ¼540 C ð1Þ
@ P A @ P A
PLðλÞ PLðλÞ
λ¼510 DA λ¼510 D

(d) Plot the calculated RET ratios in step 1c as a function of

the amount of target DNA that has been spotted onto
each paper zone. An example of the resulting data set is
shown in Fig. 2c.
2. Analysis of Digital Images.
(a) Open the acquired colored digital images using the Ima-
geJ software (National Institute of Health, Bethesda, MB,
USA).
(b) Click on the image tab and sequentially select the color
and then split channels option from the dropdown menu
in order to split each colored digital image into the
corresponding red, green and blue color channels.
320 Samer Doughan et al.

The channels of interest are the red and green color

channels, which respectively interrogate the acceptor
(Cy3) and the donor (gQDs) PL intensities as shown in
Fig. 2a (see Note 28).
(c) Draw a circle around a paper zone in the split red and
green color channels images by selecting the oval drawing
tool (see Note 29).
(d) Click on the Analyze tab and select the measure option
from the dropdown menu. This will open a new window
displaying the area, mean, min and max values
corresponding to the paper zone that has been selected.
The value of interest is the mean value.
(e) Determine the mean value from each of the paper zones
by dragging the same drawn circle from one paper zone to
another and subsequently repeating step 2d for different
paper zones.
(f) Calculate the R/G ratio corresponding to each of the
paper zones using Eq. 2. In Eq. 2, IR corresponds to the
mean intensity of the paper zone in the red channel, while
IG corresponds to the mean intensity of the same paper
zone in the green channel. The subscripts DA and D
denote measurements made in the presence and absence
of the acceptor, respectively (see Note 30).
   
IR IR
R=G ratio ¼
ð2Þ
I G DA IG D

(g) Plot R/G ratios as a function of the amount of target

DNA that has been spotted onto each paper zone. An
example of the resulting data set is shown in Fig. 2c.
3. Analysis of Microscope Images for the UCNP/QD RET Pair.
(a) Using the ImageJ software (National Institute of Health,
Bethesda, MB, USA), open the images that have been
acquired by scanning the paper using an epifluorescence
microscope.
(b) Click on Plugins, Macros, Record (see Note 31).
(c) Draw circles around each paper zone for the resultant
image acquired using 455/35 nm optical filter using the
oval drawing tool. The circle drawn for one zone can be
dragged over to other zones to ensure that the mean
intensity determined from each zone represents the
same area.
(d) After each circle is drawn, press the “M” key on the
keyboard to measure an integrated intensity in the
paper zone. This will display the area, and the mean,
 Paper-Based Nucleic Acid Assay 321

minimum, and maximum intensity values of the paper

zone that has been selected. Use the mean value as the
average signal from each zone.
(e) In the Macros window, save the resultant code.
(f) Open the image that has been scanned using the
560/20 nm filter.
(g) Go to Plugins, Macro, Run and select the code that has
been saved. This will automatically give results from each
zone on this image in the same order as the zones have
been created in the previous image.
(h) Calculate a RET Ratio (Eq. 3) for each reaction zone by
taking the ratio of the signal in the QD (acceptor) chan-
nel and UCNP (donor) channel.
   
Io Io
RET Ratio ¼
ð3Þ
I B DA IB D

(i) Plot the calculated RET ratios as a function of amount of

DNA target that has been spotted onto each paper zone.
Take an average of the three paper zones that have been
subjected to the same target concentration. An example of
the resulting data set is shown in Fig. 3.

4 Notes

1. SMN1 sequence is a genetic marker that is diagnostic of the

neuromuscular disorder known as spinal muscular atrophy
[22]. We have used the SMN1 sequences as model sequences
to demonstrate the principles of solid-phase QD-RET nucleic
acid hybridization assays. HPRT1 is a housekeeping gene
found in mammalian cells. It is selected as a generic target to
demonstrate the principles of solid-phase RET between
UCNPs and QDs. Housekeeping genes are routinely used for
control and calibration in biotechnological applications and
genomic studies [8].
2. A small incision is made in the septum for the thermometer.
3. This solution can be prepared after step 5 has been set up and
will be ready to use in this step.
4. Mark the solvent level on the reaction flask before the addition
of the methanol and wait for the solvent level to return to the
same level. The methanol should evaporate and condense in
the collection flask. Check the side flask to determine the end of
condensation.
322 Samer Doughan et al.

5. Be careful not to get any material trapped between the heating

mantel and the reaction flask as it is a fire hazard.
6. Ensure that the 6-maleimidohexanoic acid solution is prepared
before extracting the excess DTT. Add the 6-maleimi-
dohexanoic acid solution to the reduced thiol DNA immedi-
ately after extraction to ensure that the sulfhydryl functionality
is fully available.
7. A maximum of approximately 20 oligonucleotide strands can
be loaded on one QD using this method.
8. As an example, if the concentration of SMN1 probe is
239.8 μM, then 34 μL of the probe solution will need to be
pipetted. TCEP is used as a reducing agent to reduce the
disulfide (DTPA) moiety associated with each oligonucleotide
probe to a dithiol [23]. TCEP is added at 500 times molar
excess of the DTPA probe concentration. Therefore, if a differ-
ent amount of DTPA probe is used, then the volume of 50 mM
TCEP solution should be adjusted accordingly.
9. As an example, if the concentration of GSH-gQDs is 5.1 μM,
then ca. 39 μL of the QD solution will need to be pipetted. The
amount of QDs added in this step is such that the stoichiomet-
ric ratio between the QDs and DTPA probe is 1 to 40, respec-
tively. The bioconjugation of DTPA probe to CdSeS/ZnS
(core/shell) QDs is driven by self-assembly via the coordina-
tion linkage of dithiol moiety to the Zn2+ atoms on ZnS shell of
the QDs [24–26].
10. This step is referred to as the salt aging of QD–probe conju-
gates. Salt aging is used to increase the loading capacity of
oligonucleotide probes to the surface of QDs by screening
the electrostatic repulsion associated with the negative charge
of the sugar-phosphate backbone of oligonucleotides [21]. We
have previously reported that without the salt-aging step, an
average of 17 oligonucleotide probes can be self-assembled to
the surface of similar GSH-QDs when incubated with 40 times
molar excess of the oligonucleotide probes [21]. However,
with the salt-aging step, an average of 35–40 oligonucleotide
probes can be self-assembled to the surface of the GSH-QDs
for the aforementioned preparation [21].
11. Ensure that the pattern of the paper device fits within the
20 cm  20 cm dimensionality of the Whatman® cellulose
chromatography paper sheet.
12. After the wax melting step, the actual diameter of each paper
zone will be smaller than what has been originally designed,
i.e., less than 5 mm. An estimate of the size of the features that
are wax printed on the paper substrates after the melting step
can be made as described by Carrilho et al. [17].
 Paper-Based Nucleic Acid Assay 323

13. A custom setting of the printer may be required in order to

print the paper substrate.
14. This step melts and spreads the wax laterally and vertically,
resulting in a formation of hydrophobic wax barriers across
the thickness of the paper. Owing to the lateral spreading of
the wax, the resulting diameter of each hydrophilic paper zone
on the paper device will be approximately 3 mm [17].
15. This ensures that any solution that is spotted onto the paper
zones does not come in contact with any other surface, which
can otherwise result in reagent loss.
16. We wash paper devices by submerging each paper device inside
a 50 mL conical centrifuge tube that is filled with Milli-Q
water.
17. Dilutions of the stock QD–probe conjugates solution can also
be prepared and spotted onto the imidazole modified paper
zones. Quantitative information can also be acquired using
various concentrations of the QD–probe conjugates solution
(e.g., 8.7, 52, or 167 nM) provided that calibration data exist
for these concentrations of immobilized QD–probe
conjugates.
18. It is preferred that a serial dilution method is used in order to
prepare various dilutions of SMN1 FC TGT. Also, ensure that
the NaCl concentration in each of the target solutions is
adjusted to 100 mM. We use 2.5 M NaCl solution in order
to adjust NaCl concentration to 100 mM.
19. This allows for the collection of PL spectra and images from
paper zones that have been modified with immobilized
QD–probe conjugates but not exposed to the target sample.
20. There are four paper zones along each row of the paper device.
This allows four replicate measurements to be collected by
spotting the same target concentration along a particular row
of the paper device. To evaluate nonspecific adsorption, a row
of paper zones can also be spotted with SMN1 NC target.
21. During this incubation step, the target/sample solutions may
dry on the paper device. However, this does not affect the
experimental outcome.
22. We typically apply reporter that is in two- to threefold molar
excess of the target concentration. For experiments involving
single nucleotide polymorphism discrimination, the paper sub-
strates are first washed with 15% (v/v) formamide solution in
BB1 for 15 min and then with BB2 for 1 min prior to the
reporter hybridization step.
23. We have previously reported that data collection from dry
paper substrates offers at least an order of magnitude higher
assay sensitivity and at least an order of magnitude lower limit
324 Samer Doughan et al.

of detection as compared to the corresponding hydrated paper

substrates [21].
24. We use a 4 objective lens (NA ¼ 0.13) for the collection of
PL spectra from paper substrates. Depending on the concen-
tration of QD–probe conjugates that has been spotted onto
the paper zones and the power of the laser excitation source, a
neutral density (ND) filter may be required in order to attenu-
ate the laser excitation intensity to prevent detector saturation.
We use a ND 8 filter for experiments that rely on excitation
with the 25 mW diode laser and QD–probe conjugates spotted
onto the paper substrates at 312 nM (936 fmol) concentration.
Due to ratiometric signal processing, fluctuations in the abso-
lute PL intensities of donor and acceptor that are caused by
factors such as variations in detector sensitivity, excitation
source intensity and sample dilutions do not significantly affect
the experimental outcome [27]. While QDs exhibit a strong
and broad absorption band [1], which is a useful attribute to
efficiently excite QDs with a broad range of excitation wave-
lengths, the selection of excitation wavelength close to 400 nm
is done in order to minimize direct excitation of the Cy3
acceptor [1].
25. For the collection of PL images, place the paper substrates and
the UV lamp on a horizontal surface and hold the iPad mini
orthogonal to the horizontal surface. Depending on the con-
centration of QD–probe conjugates that has been spotted onto
the paper zones, a placement of ND filter in front the iPad mini
camera may be required to prevent pixel saturation. For
QD–probe conjugates spotted onto the paper devices at
312 nM (936 fmol) concentration, we place paper devices
10 cm away from the UV lamp and use ND 16 filter. The
selection of ND filter (ND 4, ND 8, or ND 16) will also be
dependent on the separation distance between the lamp and
the paper device.
26. We use a 40 objective lens (NA ¼ 0.60) to collect the photo-
luminescence from the paper. The tunable power of the IR
laser is set between 0.4 and 1 W depending on the concentra-
tion and brightness of immobilized UCNPs.
27. The second term in Eq. 1 is a correction factor and accounts for
the crosstalk of the gQDs donor PL with the Cy3 acceptor PL.
28. After splitting the color channels, the images will be in the grey
color scheme. Pseudo-coloring of the images can be done by
clicking on the image tab and then selecting the Lookup
Tables option from the dropdown menu, followed by selecting
the desired color scheme. We use red color for displaying the
red color channel and green color for displaying the green
color channel.
 Paper-Based Nucleic Acid Assay 325

29. Make sure the drawn circle fits within the dimensions of the
paper zone. The same drawn circle can be dragged from one
paper zone to another in order to analyze the mean PL inten-
sity from each paper zone. This also ensures that the mean
intensity determined from each of the paper zones represents
the same unit area.
30. The second term in Eq. 2 is a correction factor and accounts for
the crosstalk of the gQDs donor intensity in the red channel.
31. This allows the software to track the locations on the paper
from where the signals are collected as an integrated area for
the image collected with the 455/35 nm filter. This can then
be applied to the image collected with the 560/20 nm filter.

Acknowledgments

The authors gratefully acknowledge the Natural Sciences and Engi-

neering Research Council of Canada (NSERC) for financial sup-
port of their research. S.D. acknowledges NSERC for the provision
of a graduate scholarship. M.O.N. is grateful to the Ontario Cen-
tres of Excellence (OCE) for provision of a Talent Edge postdoc-
toral fellowship. Y.H. acknowledges the Ministry of Training,
Colleges and Universities for provision of an Ontario Graduate
Scholarship (OGS).

References
1. Noor MO et al (2014) Building from the 5. Freeman R, Liu XQ, Willner I (2011) Chemi-
“Ground” up: developing interfacial chemistry luminescent and Chemiluminescence Reso-
for solid-phase nucleic acid hybridization assays nance Energy Transfer (CRET) detection of
based on quantum dots and fluorescence reso- DNA, metal ions, and aptamer-substrate com-
nance energy transfer. Coord Chem Rev plexes using hemin/G-quadruplexes and
263:25–52 CdSe/ZnS quantum dots. J Am Chem Soc
2. Noor MO, Krull UJ (2013) Paper-based solid- 133(30):11597–11604
phase multiplexed nucleic acid hybridization 6. Kumar M et al (2011) A rapid, sensitive, and
assay with tunable dynamic range using immo- selective bioluminescence resonance energy
bilized quantum dots as donors in fluorescence transfer (BRET)-based nucleic acid sensing sys-
resonance energy transfer. Anal Chem 85 tem. Biosens Bioelectron 30(1):133–139
(15):7502–7511 7. Doughan S et al (2014) Solid-phase covalent
3. Noor MO, Tavares AJ, Krull UJ (2013) On- immobilization of upconverting nanoparticles
chip multiplexed solid-phase nucleic acid for biosensing by luminescence resonance
hybridization assay using spatial profiles of energy transfer. ACS Appl Mater Interfaces 6
immobilized quantum dots and fluorescence (16):14061–14068
resonance energy transfer. Anal Chim Acta 8. Doughan S, Uddayasankar U, Krull UJ (2015)
788:148–157 A paper-based resonance energy transfer
4. Algar WR et al (2012) Quantum dots as simul- nucleic acid hybridization assay using upcon-
taneous acceptors and donors in time-gated version nanoparticles as donors and quantum
forster resonance energy transfer relays: charac- dots as acceptors. Anal Chim Acta 878:1–8
terization and biosensing. J Am Chem Soc 134 9. DaCosta MV et al (2014) Lanthanide upcon-
(3):1876–1891 version nanoparticles and applications in
326 Samer Doughan et al.

bioassays and bioimaging: a review. Anal Chim 19. Martinez AW et al (2010) Diagnostics for the
Acta 832:1–33 developing world: microfluidic paper-based
10. Ju Q, Uddayasankar U, Krull U (2014) Paper- analytical devices. Anal Chem 82(1):3–10
based DNA detection using lanthanide-doped 20. Yetisen AK, Akram MS, Lowe CR (2013)
LiYF4 upconversion nanocrystals as bioprobe. Paper-based microfluidic point-of-care diag-
Small 10(19):3912–3917 nostic devices. Lab Chip 13(12):2210–2251
11. Zhou F, Krull UJ (2014) Spectrally matched 21. Noor MO, Krull UJ (2014) Camera-based
duplexed nucleic acid bioassay using two-colors ratiometric fluorescence transduction of
from a single form of upconversion nanoparti- nucleic acid hybridization with reagentless sig-
cle. Anal Chem 86(21):10932–10939 nal amplification on a paper-based platform
12. Algar WR, Tavares AJ, Krull UJ (2010) Beyond using immobilized quantum dots as donors.
labels: a review of the application of quantum Anal Chem 86(20):10331–10339
dots as integrated components of assays, biop- 22. Watterson JH et al (2004) Rapid detection of
robes, and biosensors utilizing optical trans- single nucleotide polymorphisms associated
duction. Anal Chim Acta 673(1):1–25 with spinal muscular atrophy by use of a reus-
13. Algar WR, Krull UJ (2009) Interfacial trans- able fibre-optic biosensor. Nucleic Acids Res
duction of nucleic acid hybridization using 32(2):1–9
immobilized quantum dots as donors in fluo- 23. Noor MO, Shahmuradyan A, Krull UJ (2013)
rescence resonance energy transfer. Langmuir Paper-based solid-phase nucleic acid hybridiza-
25(1):633–638 tion assay using immobilized quantum dots as
14. Algar WR, Krull UJ (2009) Toward a multi- donors in fluorescence resonance energy trans-
plexed solid-phase nucleic acid hybridization fer. Anal Chem 85(3):1860–1867
assay using quantum dots as donors in fluores- 24. Pong BK, Trout BL, Lee JY (2008) Modified
cence resonance energy transfer. Anal Chem 81 ligand-exchange for efficient solubilization of
(10):4113–4120 CdSe/ZnS quantum dots in water: a procedure
15. Robelek R et al (2004) Multiplexed hybridiza- guided by computational studies. Langmuir 24
tion detection of quantum dot-conjugated (10):5270–5276
DNA sequences using surface plasmon 25. Algar WR et al (2011) The controlled display
enhanced fluorescence microscopy and spec- of biomolecules on nanoparticles: a challenge
trometry. Anal Chem 76(20):6160–6165 suited to bioorthogonal chemistry. Bioconjug
16. Martinez AW et al (2007) Patterned paper as a Chem 22(5):825–858
platform for inexpensive, low-volume, portable 26. Sapsford KE et al (2013) Functionalizing
bioassays. Angew Chem Int Ed 46 nanoparticles with biological molecules: devel-
(8):1318–1320 oping chemistries that facilitate nanotechnol-
17. Carrilho E, Martinez AW, Whitesides GM ogy. Chem Rev 113(3):1904–2074
(2009) Understanding wax printing: a simple 27. Algar WR, Krull UJ (2008) Quantum dots as
micropatterning process for paper-based donors in fluorescence resonance energy trans-
microfluidics. Anal Chem 81(16):7091–7095 fer for the bioanalysis of nucleic acids, proteins,
18. Martinez AW et al (2008) FLASH: a rapid and other biological molecules. Anal Bioanal
method for prototyping paper-based microflui- Chem 391(5):1609–1618
dic devices. Lab Chip 8(12):2146–2150
 Chapter 20

Enhanced Performance of Colorimetric Biosensing

on Paper Microfluidic Platforms Through Chemical
Modification and Incorporation of Nanoparticles
Ellen Flávia Moreira Gabriel, Paulo T. Garcia, Elizabeth Evans, Thiago
M.G. Cardoso, Carlos D. Garcia, and Wendell K.T. Coltro

Abstract
This chapter describes two different methodologies used to improve the analytical performance of colori-
metric paper-based biosensors. Microfluidic paper-based analytical devices (μPADs) have been produced by
a stamping process and CO2 laser ablation and modified, respectively, through an oxidation step and
incorporation of silica nanoparticles on the paper structure. Both methods are employed in order to
overcome the largest problem associated with colorimetric detection, the heterogeneity of the color
distribution in the detection zones. The modification steps are necessary to improve the interaction
between the paper surface and the selected enzymes. The enhanced performance has ensured reliability
for quantitative analysis of clinically relevant compounds.

Key words Oxidation process, Nanoparticle, Colorimetric detection, Clinical diagnostic, Urinalysis
test

1 Introduction

According to the definition, a biosensor is an analytical device that

combines two specific characteristics: an element able to interact
specifically with a target and a transducer able to convert the
recognition process into a measurable signal [1, 2]. The extensive
use of biosensors can be understood since these devices are able to
provide reliable, specific, and rapid responses. There are numerous
examples in literature reporting the construction of biosensors
dedicated to different applications such as clinical and biological
monitoring [3–5], environmental pollutants [6], and food and
water contamination [7, 8]. In a particular case of clinical and
diagnostic monitoring, the development of biosensors has
increased significantly, since it has been demonstrated that early
diagnosis can dramatically change the clinical treatment type [1].

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_20,
© Springer Science+Business Media LLC 2017

327
328 Ellen Flávia Moreira Gabriel et al.

In order to understand the biological systems and mainly the

specific biomolecular interaction involved in the process, different
techniques have been proposed. The most commonly explored
technologies are the surface plasmon resonance (SPR) [9] and
electrochemical impedance spectroscopy (EIS) [10]. Both techni-
ques offer high sensitivity and are commercially established. How-
ever, the instrumentation required in both techniques is expensive
and it demands trained personal to operate the instrument accord-
ingly. These disadvantages hinder promptly the implementation
of both SPR and EIS in research groups with limited financial
support [11]. Alternatively, different techniques or simpler tools
are emerging as promising alternatives to monitor binding events.
Some examples involve the use of chemosensors [12], supercapaci-
tive admittance tomoscopy technique [13], and capacitively
coupled contactless conductivity detection [11, 14].
In general, the fabrication of biosensors involves the thin-film
deposition followed by surface functionalization. These procedures
are laborious and raise the final cost of the biosensor. Therefore, as
an alternative that could maintain the main characteristics of bio-
sensors (specific interaction with target analyte and measurement of
signal response), microfluidic paper-based analytical devices
(μPADs) have emerged as a new generation of powerful platforms
serving as a low-cost substrate for biosensing applications. Since the
pioneering report published by Whitesides group in 2007 [15],
many research groups have demonstrated that μPADs could offer
important advantages in the design of novel biosensing systems,
especially for clinical diagnostics. The main advantages of μPADs
are the low-cost, low sample consumption, minimal waste disposal,
ease-of-use, global affordability, disposability, biocompatibility, and
capability to be used in remote areas [3, 15–17]. All of these
features make μPADs attractive to be implemented in low-income
communities, principal targets to improve public health care [1].
Different techniques have been employed to produce paper-
based biosensors including photolithography [15, 18, 19], inkjet
etching [20, 21], wax printing [22], laser cutting [23, 24], and
stamping process [4, 25]. All fabrication methods allow the pro-
duction of microfluidic channels on paper surface, which due to its
porous structure facilitates solution transport through capillary
action. In addition to the fluidic channels, specific regions known
as detection zones are designed and integrated with channels to
proceed the chemical reaction responsible for the specific recogni-
tion process. In order to provide the response, suitable transduc-
tion methods are coupled with paper-based biosensors including
colorimetric [3–5, 24], electrochemical [26–28], fluorescence
[29], chemiluminescence [30, 31], and mass spectrometry [32,
33] detectors. Among the different detection systems, colorimetric
detection represents one of the most popular tools used as trans-
duction mode on μPADs. Color is usually developed through
 Colorimetric Biosensing on Paper Microfluidic Platforms 329

enzymatic or complexometric reactions between analyte and chro-

mogenic agent. The presence of target analyte in sample may
further be qualitatively identified by the naked eye inside the detec-
tion zones. Moreover, the concentration level can be quantified
based on the color intensity, which is measured by the pixel inten-
sity recorded in a digital image captured with scanner [3–5], digital
camera [34], or cell phone cameras [35–37].
Despite the wide applicability of colorimetric biosensors on
μPADs, many reports have presented drawbacks that still need to
be addressed to ensure reliability for quantitative measurements.
The most remarkable is related the lack of color homogeneity or
uniformity generated in the detection zones during the reaction
process between the analyte and indicator, and typically catalyzed
by a specific enzyme. The heterogeneous color is due to the poor
interaction between the support material (paper) and the enzyme
or chromogenic reagents.
To overcome this problem, this chapter describes two different
methods to enhance the analytical performance of colorimetric
biosensors on μPADs. The improvement is achieved through the
(1) oxidation process and (2) incorporation of silica nanoparticles
onto paper structure. Both modification processes are adopted to
provide better support for enzyme adsorption and therefore
improve the color homogeneity and uniformity inside the detection
zones. For each procedure, the optimal conditions of the modifica-
tion steps are investigated. The μPADs modified by the oxidation
process are evaluated for the detection of glucose and uric acid. On
the other hand, the μPADs incorporated with silica nanoparticles
are explored on glucose, lactate, and glutamate assays. In both
instances, a flatbed scanner is used to quantify each analyte using
the histogram tools in the graphic software. The feasibility of both
modification methodologies are evaluated by urinalysis tests using
an artificial urine samples.

2 Materials

2.1 Paper-Based 1. Computer equipped with graphic software Corel Photo-

Analytical Device Paint™ and Adobe Photoshop®.
2. Paraffin.
3. Metal Stamp designed in stainless steel by a local shop (MS
Máquinas, Goiânia, GO, Brazil).
4. CO2 Laser ablation system model Mini 24, 30 W obtained
from Epilog Laser Systems (Golden, CO, USA).

2.2 Modification 1. 0.5 M sodium periodate solution prepared in water.

Process (Oxidation 2. Silicon dioxide nanopowder (SiO2—15 nm) obtained from
and Incorporation Sigma-Aldrich (St. Louis, MO, USA).
with SiO2)
330 Ellen Flávia Moreira Gabriel et al.

3. 5% (v/v) solution of 3-aminopropyltrielthoxysilane (APTES)

(Sigma-Aldrich; St. Louis, MO, USA) in ethanol.
4. Nanoparticle suspension: 3.3 mg of SiO2 nanoparticles sus-
pended in 10 mL of 5% APTES solution.

2.3 Colorimetric 1. Flatbed Scanner.

Detection 2. Histogram tools from graphic software.

2.4 Bioassay 1. Solution 1: potassium iodide (0.6 M) and trehalose (0.3 M)

Solutions dissolved in 100 mM phosphate buffer at pH 6.0.
2. Solution 2: 4-aminoantipyrine (4-AAP) and sodium 3,5-
dichloro-2-hydroxybenzenesulfonic acid (DHBS), both in
6.6 mM, dissolved in 100 mM phosphate buffer at pH 6.0.
3. Solution 3: 15 mM 3,30 ,5,50 -tetramethylbenzidine (TMB) dis-
solved in ethanol and wrapped with aluminum foil to protect
from light.
4. Solution 4: A 5:1 mixture of glucose oxidase (120 U/mL) and
horseradish peroxidase (30 U/mL) is prepared in 100 mM
phosphate buffer at pH 6.0.
5. Solution 5: A 1:1 mixture of lactate oxidase (100 U/mL) and
horseradish peroxidase (339 U/mL) dissolved in 100 mM
phosphate buffer at pH 6.0.
6. Solution 6: A 1:1 mixture of L-glutamate oxidase (4.16 U/mL)
and horseradish peroxidase (339 U/mL) prepared in 100 mM
phosphate buffer at pH 7.4.
7. Solution 7: A 1:1 mixture of uricase oxidase (80 U/mL) and
horseradish peroxidase (339 U/mL) dissolved in phosphate
buffer solution pH 6.0.
8. All reagents employed on colorimetric assays are acquired from
Sigma-Aldrich (St. Louis, MO, USA).

2.5 Artificial Urine 1. Artificial urine is prepared at pH 6.0 and consists of 2 mM citric
Sample acid (Sigma-Aldrich; St. Louis, MO, USA), 25 mM sodium
bicarbonate (Sigma-Aldrich; St. Louis, MO, USA), 170 mM
urea (Sigma-Aldrich; St. Louis, MO, USA), 2.5 mM calcium
chloride (E.M. Science; Gibbstown, NJ, USA), 90 mM sodium
chloride (Mallinckrodt Baker; Center Valley, PA, USA), 2 mM
magnesium sulfate (Mallinckrodt Baker; Center Valley, PA,
USA), 10 mM sodium sulfate (Sigma-Aldrich; St. Louis, MO,
USA), 7 mM sodium phosphate monobasic anhydrous (Fisher
Scientific; Waltham, MA, USA), 7 mM sodium phosphate
dibasic anhydrous (Fisher Scientific; Waltham, MA, USA),
and 25 mM ammonium chloride (Sigma-Aldrich; St. Louis,
MO, USA). This solution is stored in the refrigerator (4  C)
until use.
 Colorimetric Biosensing on Paper Microfluidic Platforms 331

3 Methods

3.1 Methodologies 1. Stamped μPADs are fabricated according to a procedure previ-

to Produce μPADs ously developed by our group [38] and schematically described
in Fig. 1.
3.1.1 Stamping Process
2. Initially, a sheet of filter paper is immersed into liquid paraffin
(at 90  C) for 60 s.
3. Then, the paper is removed from the paraffin solution, allowed
to solidify at room temperature and placed above sheet of paper
without paraffin, thus forming a sandwiched paper piece.
4. A metal stamp is preheated at 150  C on a hot plate for 2 min.
5. Then, the metal stamp is put in contact with sandwiched piece
and ca. 0.1 MPa of pressure is applied to the metal stamp for
2 s. This process is necessary to enable the thermal transfer of
paraffin from top paper to bottom paper, thus forming hydro-
phobic barriers.
6. The layout of the proposed μPADs (45 mm  45 mm) con-
sisted of eight circular detection zones for bioassays
interconnected by microfluidic channels and one central zone
as sample inlet.
7. All channels are nominally fabricated with a 10 mm length and
3 mm width. The diameter values for detection and central
zones are 5 and 10 mm, respectively.

Fig. 1 Scheme showing the fabrication process of μPADs based on stamping process. In (a), a paraffinized
paper is placed over the native paper surface; in (b), a metal stamp preheated at 150  C is brought into contact
with the layered paper pieces; in (c), it is presented a typical μPAD fabricated by the stamping method. The
optical micrograph in (d) depicts a real image showing the stamped μPAD. Reproduced from ref. 4 with
permission
332 Ellen Flávia Moreira Gabriel et al.

3.1.2 CO2 Laser System 1. Laser cutting is also selected out of the many techniques used
to fabricate μPADs due to its simplicity of being a straightfor-
ward single step and quick process.
2. A schematic representation of the laser cutting step can be seen
in Fig. 2.
3. The μPAD is first designed using CorelDraw™ X6 Software
with dimensions of a 1.25 mm main channel that branches into
three channels of the same width leading to three 5 mm detec-
tion circles on each end.
4. The design is sent to the laser engraver and cut at 600-dpi
resolution using 30% power and 30% speed. A steady flow of
N2 gas impinged at the engraving point to minimize the risk of
ignition.
5. The laser cuts the paper and define the geometry on the surface
by the ablation process.
6. The laser cuts at a rate of approximately 5 s per μPAD.

3.2 Modification 1. Paper substrates are chemically modified by an oxidation pro-

Steps cess using NaIO4 to allow the covalent coupling of enzymes on
the cellulose surface.
3.2.1 Oxidation Process
with NaIO4 2. Prior to stamping, the paper sheets are immersed in a petri dish
containing 10 mL of 0.5 M NaIO4 solution and allowed to
react at room temperature in the dark for 30 min.
3. After oxidation, the paper sheets are washed three times with
ultrapure water.
4. Then, μPADs are stamped according to procedure described on
Subheading 3.1.2.
5. All detection zones are spotted with 0.75 μL of a mixed solu-
tion of 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide
(EDC, 0.1 M) (Sigma-Aldrich; St. Louis, MO, USA) and
N-Hydroxysuccinimide (NHS, 0.1 M) (Sigma-Aldrich; St.
Louis, MO, USA) prepared in ultrapure water. Then, each
detection zone is allowed to dry at room temperature for

Fig. 2 Procedure for fabricatin μPADs by CO2 laser process. (a) CO2 laser machine. (b) Scheme showing the
laser cut of a paper sheet to produce paper biosensor devices. All μPADs were cut using the vector mode, at
30% speed (of a maximum linear speed of 1.65 cm/s) and 30% power (of a maximum intensity of 30 W)
 Colorimetric Biosensing on Paper Microfluidic Platforms 333

OH
OH

NaIO4 O
O O
O O
O O
HO O
OH

EDC-NHS

OH OH

O Enz-NH2 O O
O
O O
O O
O O
HN O N O
Enz

Fig. 3 Reactional scheme showing the modification process on paper surface with NaIO4 to obtain a paper
biosensor. Adapted from ref. 17 with permission

20 min to convert the aldehyde groups of the cellulose to

amine reactive esters.
6. Afterwards, enzymes are covalently conjugated on the modi-
fied paper surface by adding 0.75 μL of each enzyme solution
to the detection zones.
7. All reactions involved in the modification process of the cellu-
lose surface to obtain a biosensor in paper platform are shown
in Fig. 3.

3.2.2 Incorporation of Silica nanoparticles (SiO2NP) are also employed to modify the
Silica Nanoparticle surface of the paper in order to accomplish a uniform distribution
(SiO2NP) of the enzyme, therefore increasing the color homogeneity in the
detection zones. The SiO2NP incorporated on detection zone of
μPAD can been seen in Fig. 4.
1. First, a silane-coupling reaction is employed to introduce
amine groups onto the surface of the nanoparticles. 3.3 mg of
15 nm SiO2NP are added to a 5% (v/v) solution of APTES in
ethanol.
2. The length of time the nanoparticles remained in contact with
the APTES solution is investigated from 1 to 24 h due to the
variables that can affect the density of the amine groups, and
ultimately the adsorption of the enzymes to the modified
μPAD.
3. The incorporation of SiO2NP on μPADs is realized by two
different procedures: dispensing the nanoparticle suspension
334 Ellen Flávia Moreira Gabriel et al.

Fig. 4 Scheme of modification process with SiO2NP. Scanning electron micrograph images show the silica
incorporated on cellulose surface. The micrograph images was reproduced from ref. 5 with permission

using an automatic pipette and full immersion of the paper in

the nanoparticle suspension.
4. The μPADs are carefully rinsed with 0.1 M PBS pH 6 to
remove any loosely bound particles and finally allowed to dry
at room temperature before use.

3.3 Enzymatic The enzymatic assays are performed by adding specific color
Procedures reagents for each bioassay on the detection zones. The processes
are described below.
3.3.1 μPADs—Oxidation
Process 1. For the glucose assay, each detection zone is spotted with
0.75 μL of solution 1 and dried at room temperature for
10 min. Then, 0.75 μL of solution 4 is spotted in each detec-
tion zone and allowed to dry at room temperature for 10 min.
2. For the uric acid assay, each detection zone is spotted with
0.75 μL of solution 2 and dried at room temperature for
10 min. Afterwards, 0.75 μL of solution 7 is added in each
detection zone and allowed to dry at room temperature for the
same time used previously.

3.3.2 μPADs—SiO2NP 1. For the lactate assay, each detection zone is spotted with 1 μL
Incorporation of solution 2 and dried at room temperature for 15 min. After-
wards, the zone is spotted with 1 μL of solution 5 and dried for
the same duration as before.
2. For the glucose assay, the detection zone is spotted with 1 μL of
solution 1 and dried at room temperature for 15 min. After-
wards, the zone is spotted with 1 μL of solution 4 and dried for
additional 15 min.
3. For the glutamate assay, the detection zone is spotted with 1 μL
of solution 3 and dried at room temperature for 15 min. Then
 Colorimetric Biosensing on Paper Microfluidic Platforms 335

the zone is spotted with 1 μL of solution 6 and dried for

10 min.
4. In order to perform simultaneously all three bioassays on
μPADs, a 10 μL solution containing the analytes is introduced
at the bottom stem of the μPAD and allowed to wick up the
three channels by capillarity to the detection zones.

3.4 Colorimetric The scanner mode of a DeskJet multifunctional printer (Hewlett-

Detection System Packard, model F4280) is used to capture the colorimetric infor-
mation related to the semi-quantitation of glucose, lactate, gluta-
mate, and uric acid.
1. Following the colorimetric enzymatic procedures described in
Subheading 3.3, images of the μPADs are taken with a scanner
using 600-dpi resolution.
2. The recorded images are first converted to a 24-bit color scale
(RGB or CMYK dimension) and then analyzed in either Corel
Photo-Paint™ or Adobe Photoshop software. Some free soft-
wares, as for example ImageJ, can also be used to obtained the
pixel information.
3. The arithmetic mean of the pixel intensity within each test zone
is obtained with the histogram tool of graphic software and
used to obtain the color intensity.
4. The color intensity achieved is used to achieve the analyte
concentration and thus to perform quantitative measurements.

4 Paper-Based Biosensor: Characterization Step

The poor color uniformity and homogeneity associated with colori-

metric measurements represent some the most important short-
comings on μPADs (see Notes 1 and 2). Aiming to solve this
problem, the oxidation of cellulose and incorporation of SiO2
nanoparticles are used as analytical strategies to support the enzymes
required for the selected bioassays (see Note 3). The improvements
achieved with both methodologies are described below.

4.1 Oxidation with 1. The oxidation process is used for the conversion of the
NaIO4 hydroxyl groups on paper surface to aldehyde groups. Then,
these groups are chemically activated with an EDC/NHS solu-
tion enabling the covalent coupling with different enzymes.
2. The conversion process is monitored by infrared spectroscopy.
The technique confirmed the presence of aldehyde groups on
paper surface after the oxidation step. As it can be observed in
Fig. 5, the specific band at 1727 cm1, characteristic of alde-
hyde groups, is found only on the oxidized paper.
336 Ellen Flávia Moreira Gabriel et al.

60 (A) Native Paper 60 (B) Native Paper

Oxidized Paper Oxidized Paper
50 50

% Reflectance
% Reflectance

40 40

30 30

20 20

10 10

0 0

4000 3500 3000 2500 2000 1500 1000 500 2500 2250 2000 1750 1500 1250 1000
-1 -1
Wavenumbers (cm ) Wavenumbers (cm )

Fig. 5 Characterization of native and oxidized paper by Fourier transform infrared spectroscopy. Reproduced
from ref. 4 with permission

3. Besides the characterization process, the feasibility of the μPADs

is tested by performing the glucose and uric acid assays. The
procedure is performed by spotting the specific color reagents
on each detection zone and ca. 40 μL of standard solution on the
central zone. After 15 min, the images are obtained by colori-
metric detection as previously described in Subheading 3.4.
4. The chemical modification has improved the color uniformity
for both assays, as observed in Fig. 6. The values found of color
gradient is lower than 10%. The color gradient or uniformity
values are obtained from the standard deviation provided by
the histogram tools on the graphic software. The better gradi-
ent is expected once the oxidized process ensures the chemical
bond between activated groups on the paper surface and amino
group from each enzyme.

4.2 Incorporation of 1. The performance of amino group incorporation on silica nano-

SiO2NP particle is evaluated on color intensity and gradient response.
As it can be observed in Fig. 7, the modification time has
significant effects on the color intensity but little effect on the
color gradient. These behaviors can be interpreted by relation
between the amount of protein adsorbed on paper modified
with SiO2NP and the strength of the interaction. In short time
of reaction just few amino groups are present on silica surface
to support the enzyme dispensed, and in long reaction times
the modification can make the surface too hydrophobic. Con-
sidering these results, 3 h is selected as the optimum time for
the silica nanoparticles to be modified with the amino groups.
2. The best method to incorporate the SiO2NP on paper surface
is also investigated. As described on Subheading 3.2, basically
two different (dispense by pipette and full immersion) proce-
dures are used. The better homogeneity distribution of
 Colorimetric Biosensing on Paper Microfluidic Platforms 337

Fig. 6 Optical micrographs showing the improvements on the color uniformity for glucose (left images) and
uric acid assays (right images) before (a, b) and after (c, d) chemical oxidation on paper surface. Reproduced
from ref. 4 with permission

30
195
Color Intensity (AU)

Color Gradient (AU)

25
180

165 20

150 15
0 5 10 15 20 25
Time (hours)

Fig. 7 Effect of modification time of silica nanoparticles with APTES on the color intensity and gradient.
Reproduced from ref. 5 with permission
338 Ellen Flávia Moreira Gabriel et al.

Fig. 8 Optical images showing the colorimetric assays for glucose, lactate and glutamate assay with and
without nanoparticle. Reproduced from ref. 5 with permission

nanoparticle on paper surface is achieved when the paper is full

immersed in the suspension solution (see Fig. 4 on SEM
image). This system is used for the remaining experiments.
3. Once the optimized conditions for the incorporation process
have been established, the feasibility of these devices are shown
for lactate, glucose, and glutamate bioassays. In this case the
μPADs are prepared as described in Subheading 3.3.2.
4. All three selected reactions are catalyzed by enzymes with well-
known properties, and use a simple, oxygen-dependent reac-
tion that produces H2O2. The H2O2 is then utilized to oxidize
a chromogenic agent in a secondary reaction catalyzed by HRP.
5. The enzymatic tests are carried out on the paper devices with
nanoparticles and the results are compared with those
conducted on native μPADs. As can be observed on Fig. 8
significant improvements in the signal intensity and color
homogeneity are obtained when silica nanoparticles are
incorporated on paper device. In the glucose assay, for example,
the increase in color intensity is 46% while the color gradient
decreased by 75%.

4.3 Clinical Assay 1. μPADs modified with SiO2NP have been explored for the
with Artificial Urine detection of glucose, lactate, and glutamate in artificial urine
Sample samples.
2. Once demonstrated individual assays on modified μPADs, all
three assays (glutamate, lactate, and glucose) are simultaneously
 Colorimetric Biosensing on Paper Microfluidic Platforms 339

Fig. 9 Optical image showing the analysis of an artificial urine sample spiked
with lactate, glucose, and glutamate on the proposed paper based biosensor.
Reproduced from ref. 5 with permission

performed on the same devices using artificial urine as biological

sample (Fig. 9). For these experiments, urine (prepared as
described in the Subheading 2.5) is spiked with glucose
(7.33 mM), lactate (2.14 mM), and glutamate (4.21 mM).
These levels are chosen once they can be related to different
biological disorders such as diabetes, liver disturbers, and risk of
migraines [39–41].
3. The pixel intensity information is correlated with concentra-
tion using a calibration curve for each analyte. The linear
equations for glucose, lactate, and glutamate are y ¼ 2 + 8.7
[Glucose]; y ¼ 22 + 14[Lactate] and y ¼ 16.5 + 7.5[Gluta-
mate], respectively. The concentration values obtained from
glucose, lactate, and glutamate are 7.9  0.6 mM,
2.2  0.2 mM, and 5.5  1.0 mM, respectively.
4. The modified μPAD has exhibited good analytical performance
for simultaneous assays. The relative standard deviation (RSD)
values achieved for colorimetric measurements ranged from
0.9% to 8%. The estimated error between found concentrations
and known concentrations is ca. 20%.
5. Based on the results, μPADs have offered good accuracy and
precision when used as colorimetric biosensors. They also have
great potential for quantitative determination of clinically rele-
vant analytes. In addition, these biosensors are fabricated using
a low cost substrate allowing their use in places with limited
resources.

5 Notes

1. The poor color uniformity and homogeneity associated with

colorimetric measurements represent one the most important
shortcomings on μPADs.
340 Ellen Flávia Moreira Gabriel et al.

2. These both parameters have affected the reliability of low-cost

platform, compromising its use in clinical assays.
3. In our experiment, we used two different strategies associated
with cellulose oxidation and incorporation of SiO2 nanoparti-
cles aiming the chemical modification of paper surface to pro-
vide a better support for enzyme adsorption.

References
1. Parolo C, Merkoci A (2013) Paper-based 11. Coltro WKT, de Santis Neves R, de Jesus
nanobiosensors for diagnostics. Chem Soc Motheo A, Da Silva JAF, Carrilho E (2014)
Rev 42:450–457 Microfluidic devices with integrated dual-
2. Nery EW, Kubota LT (2013) Sensing capacitively coupled contactless conductivity
approaches on paper-based devices: a review. detection to monitor binding events in real
Anal Bioanal Chem 405:7573–7595 time. Sens Actuators B 192:239–246
3. Martinez AW, Phillips ST, Carrilho E, Thomas 12. Delaney TL, Zimin D, Rahm M, Weiss D,
SW III, Sindi H, Whitesides GM (2008) Sim- Wolfbeis OS, Mirsky VM (2007) Capacitive
ple telemedicine for developing regions: cam- detection in ultrathin chemosensors prepared
era phones and paper-based microfluidic by molecularly imprinted grafting photopoly-
devices for real-time, off-site diagnosis. Anal merization. Anal Chem 79:3220–3225
Chem 80:3699–3707 13. Gamby J, Abid J-P, Abid M, Ansermet J-P,
4. Garcia PT, Cardoso TMG, Garcia CD, Car- Girault H (2006) Nanowires network for bio-
rilho E, Coltro WKT (2014) A handheld molecular detection using contactless imped-
stamping process to fabricate microfluidic ance tomoscopy technique. Anal Chem
paper-based analytical devices with chemically 78:5289–5295
modified surface for clinical assays. RSC Adv 14. Lima RS, Piazzetta MH, Gobbi AL,
4:37637–37644 Rodrigues-Filho UP, Nascente PA, Coltro
5. Evans E, Gabriel EFM, Benavidez TE, Coltro WK, Carrilho E (2012) Contactless conductiv-
WKT, Garcia CD (2014) Modification of ity biosensor in microchip containing folic acid
microfluidic paper-based devices with silica as bioreceptor. Lab Chip 12:1963–1966
nanoparticles. Analyst 139:5560–5567 15. Martinez AW, Phillips ST, Butte MJ, White-
6. Gomes HI, Sales MGF (2015) Development sides GM (2007) Patterned paper as a platform
of paper-based color test-strip for drug for inexpensive, low-volume, portable bioas-
detection in aquatic environment: application says. Angew Chem Int Ed 46:1318–1320
to oxytetracycline. Biosens Bioelectron 16. Martinez AW, Phillips ST, Whitesides GM,
65:54–61 Carrilho E (2009) Diagnostics for the develop-
7. Sicard C, Glen C, Aubie B, Wallace D, ing world: microfluidic paper-based analytical
Jahanshahi-Anbuhi S, Pennings K, Daigger devices. Anal Chem 82:3–10
GT, Pelton R, Brennan JD, Filipe CD (2015) 17. Pelton R (2009) Bioactive paper provides a
Tools for water quality monitoring and low-cost platform for diagnostics. Trends Anal
mapping using paper-based sensors and cell Chem 28:925–942
phones. Water Res 70:360–369 18. Martinez AW, Phillips ST, Wiley BJ, Gupta M,
8. Zhang Y, Zuo P, Ye B-C (2015) A low-cost and Whitesides GM (2008) FLASH: a rapid
simple paper-based microfluidic device for method for prototyping paper-based microflui-
simultaneous multiplex determination of dif- dic devices. Lab Chip 8:2146–2150
ferent types of chemical contaminants in food. 19. Klasner S, Price A, Hoeman K, Wilson R, Bell
Biosens Bioelectron 68:14–19 K, Culbertson C (2010) Paper-based micro-
9. Chung J, Kim S, Bernhardt R, Pyun J (2005) fluidic devices for analysis of clinically relevant
Application of SPR biosensor for medical diag- analytes present in urine and saliva. Anal Bioa-
nostics of human hepatitis B virus (hHBV). nal Chem 397:1821–1829
Sens Actuators B 111:416–422 20. Abe K, Suzuki K, Citterio D (2008) Inkjet-
10. Lin Z, Chen L, Zhang G, Liu Q, Qiu B, Cai Z, printed microfluidic multianalyte chemical
Chen G (2012) Label-free aptamer-based elec- sensing paper. Anal Chem 80:6928–6934
trochemical impedance biosensor for 17β- 21. Wang J, Monton MRN, Zhang X, Filipe CDM,
estradiol. Analyst 137:819–822 Pelton R, Brennan JD (2014) Hydrophobic
 Colorimetric Biosensing on Paper Microfluidic Platforms 341

sol-gel channel patterning strategies for paper- biosensor for simultaneous determination of
based microfluidics. Lab Chip 14:691–695 glucose and uric acid. Lab Chip 11:1286–1291
22. Carrilho E, Martinez AW, Whitesides GM 32. Wleklinski M, Li Y, Bag S, Sarkar D, Narayanan
(2009) Understanding wax printing: a simple R, Pradeep T, Cooks RG (2015) Zero volt
micropatterning process for paper-based paper spray ionization and its mechanism.
microfluidics. Anal Chem 81:7091–7095 Anal Chem 87:6786–6793
23. Nie J, Liang Y, Zhang Y, Le S, Li D, Zhang S 33. Bag S, Hendricks PI, Reynolds JC, Cooks RG
(2013) One-step patterning of hollow micro- (2015) Biogenic aldehyde determination by
structures in paper by laser cutting to create reactive paper spray ionization mass spectrom-
microfluidic analytical devices. Analyst etry. Anal Chim Acta 860:37–42
138:671–676 34. Apilux A, Siangproh W, Praphairaksit N, Chai-
24. Evans E, Gabriel EFM, Coltro WKT, Garcia lapakul O (2012) Simple and rapid colorimetric
CD (2014) Rational selection of substrates to detection of Hg(II) by a paper-based device
improve color intensity and uniformity on using silver nanoplates. Talanta 97:388–394
microfluidic paper-based analytical devices. 35. Wang H, Li Y-j, Wei J-f, Xu J-r, Wang Y-h,
Analyst 139:2127–2132 Zheng G-x (2014) Paper-based three-
25. Zhang Y, Zhou C, Nie J, Le S, Qin Q, Liu F, Li dimensional microfluidic device for monitoring
Y, Li J (2014) Equipment-free quantitative of heavy metals with a camera cell phone. Anal
measurement for microfluidic paper-based ana- Bioanal Chem 406:2799–2807
lytical devices fabricated using the principles of 36. Salles M, Meloni G, de Araujo W, Paixão T
movable-type printing. Anal Chem (2014) Explosive colorimetric discrimination
86:2005–2012 using a smartphone, paper device and chemo-
26. Dungchai W, Chailapakul O, Henry CS (2009) metrical approach. Anal Methods
Electrochemical detection for paper-based 6:2047–2052
microfluidics. Anal Chem 81:5821–5826 37. Souza FR, Duarte-Junior GF, Garcia PT, Col-
27. Noiphung J, Songjaroen T, Dungchai W, tro WKT (2014) Avaliação de dispositivos de
Henry CS, Chailapakul O, Laiwattanapaisal W captura de imagens digitais para detecção col-
(2013) Electrochemical detection of glucose orimétrica em microzonas impressas. Quim
from whole blood using paper-based microflui- Nova 37:1171–1176
dic devices. Anal Chim Acta 788:39–45 38. Gabriel EFM, Coltro WKT, Garcia CD (2014)
28. Santhiago M, Wydallis JB, Kubota LT, Henry Fast and versatile fabrication of PMMA micro-
CS (2013) Construction and electrochemical chip electrophoretic devices by laser engraving.
characterization of microelectrodes for Electrophoresis 35:2325–2332
improved sensitivity in paper-based analytical 39. Dungchai W, Chailapakul O, Henry CS (2010)
devices. Anal Chem 85:5233–5239 Use of multiple colorimetric indicators for
29. Yamada K, Takaki S, Komuro N, Suzuki K, paper-based microfluidic devices. Anal Chim
Citterio D (2014) An antibody-free microflui- Acta 674:227–233
dic paper-based analytical device for the deter- 40. Lin C-C, Tseng C-C, Chuang T-K, Lee D-S,
mination of tear fluid lactoferrin by Lee G-B (2011) Urine analysis in microfluidic
fluorescence sensitization of Tb3+. Analyst devices. Analyst 136:2669–2688
139:1637–1643 41. Ragginer C, Lechner A, Bernecker C, Horejsi
30. Delaney JL, Hogan CF, Tian J, Shen W (2011) R, Moller R, Wallner-Blazek M, Weiss S, Faze-
Electrogenerated chemiluminescence detec- kas F, Schmidt R, Truschnig-Wilders M, Gru-
tion in paper-based microfluidic sensors. Anal ber HJ (2012) Reduced urinary glutamate
Chem 83:1300–1306 levels are associated with the frequency of
31. Yu J, Ge L, Huang J, Wang S, Ge S (2011) migraine attacks in females. Eur J Neurol
Microfluidic paper-based chemiluminescence 19:1146–1150
 Chapter 21

A Smartphone-Based Colorimetric Reader for Human

C-Reactive Protein Immunoassay
A.G. Venkatesh*, Thomas van Oordt, E. Marion Schneider, Roland
Zengerle, Felix von Stetten, John H.T. Luong, and Sandeep Kumar Vashist*

Abstract
A smartphone-based colorimetric reader (SBCR), comprising a Samsung Galaxy SIII mini, a gadget (iPAD
mini, iPAD4, or iPhone 5s) and a custom-made dark hood and base holder assembly, is used for human C-
reactive protein (CRP) immunoassay. A 96-well microtiter plate (MTP) is positioned on the gadget’s
screensaver to provide white light-based bottom illumination only in the specific regions corresponding to
the well’s bottom. The images captured by the smartphone’s back camera are analyzed by a novel image
processing algorithm. Based on one-step kinetics-based human C-reactive protein immunoassay (IA), SBCR
is evaluated and compared with a commercial MTP reader (MTPR). For analysis of CRP spiked in diluted
human whole blood and plasma as well as CRP in clinical plasma samples, SBCR exhibits the same precision,
dynamic range, detection limit, and sensitivity as MTPR for the developed IA (DIA). Considering its
compactness, low cost, advanced features and a remarkable computing power, SBCR is an ideal point-of-
care (POC) colorimetric detection device for the next-generation of cost-effective POC testing (POCT).

Key words Smartphone-based colorimetric reader, Human C-reactive protein, Graphene, One-step
kinetics, Sandwich immunoassay, Monitoring inflammation

1 Introduction

Cost-effective point-of-care (POC) devices are of importance for

the readout of in vitro diagnostic (IVD) assays in decentralized,
remote, and personalized settings. Current smartphones with
advanced features provide a promising digital platform for POC
diagnostics and mobile healthcare [1]. Smartphones are equipped
with a high resolution camera, a powerful processor with high
storage capacity, wireless connectivity, real-time geo-tagging,
secure data management, and cloud computing. The potential of

*
These authors are contributed equally to this work.

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_21,
© Springer Science+Business Media LLC 2017

343
344 A.G. Venkatesh et al.

smartphones in such applications is phenomenal considering above

seven billion cellphone users with more than 70% subscription in
developing countries [2]. Smartphones are constantly upgraded
with more advanced features at low cost. Thus, many smartphone
devices and smart applications have been commercialized recently
for health and fitness, diabetes and weight management, and mea-
surement of routine healthcare characteristics [3]. In addition,
many specialized smartphone-based bioanalytical applications
have been demonstrated in the last few years [1]. Consequently,
smartphones will have a significant impact on healthcare monitor-
ing and management to revolutionize the fields of IVD and mobile
healthcare by enabling real-time on-site analysis and telemedicine
opportunities in remote settings [4].
The microtiter plate (MTP)-based enzyme linked immunosor-
bent assay (ELISA), a subject of over 350,000 publications, is the
gold standard of IVD in clinical and bioanalytical settings. There-
fore, there is an immense need for a portable and cost-effective
SBCR for the readout of IVD assays at decentralized and remote
settings. Considering the extensive outreach of cellphone that cov-
ers more than 95% of the world population, the developed SBCR
will be of immense utility for IVD and mobile healthcare.
This chapter describes an extremely low-cost SBCR (Figs. 1
and 2) that employs a commercial gadget (iPAD mini, iPAD4, or
iPhones 5s), an inexpensive polyamide dark hood and a polyamide
base holder [5]. It serves as the colorimetric readout of graphene-
based CRP IA [6] (Fig. 3) and compared with commercial MTPR
for its analytical performance. As an important biomarker for infec-
tion and inflammation [7], CRP analysis is needed for a wide range
of diseases, disorders, and pathophysiological conditions, such as
neonatal sepsis [8–12], inflammasome related diseases [13],
depressive and posttraumatic stress [14–16], diabetes [17–19],
and cardiovascular diseases [20–24]. It is a signal-enhanced IA
format, where graphene nanoplatelets (GNPs) provide an increased
surface area for the higher binding of capture CRP antibody (Ab).
The Ab is covalently bound to MTP in a leach-proof technique by
admixing EDC-activated Ab with GNPs in APTES inside the MTP
wells (Fig. 3). The DIA detectes CRP with linearity of 0.3–81 ng/
mL and a limit of detection (LOD), a limit of quantification (LOQ)
of 0.07 and 0.9 ng/mL, respectively (Fig. 4). The DIA enables a
precise and specific determination of CRP in diluted human whole
blood and plasma (Fig. 4a), as it is unaffected by immunological
reagents and nonspecific control proteins (Fig. 4b). There is the
leach-proof binding of capture Ab as the Ab-prebound MTPs
retained their activity at 4
C in 0.1 M PBS, pH 7.4 for up to
6 weeks (Fig. 4c). Moreover, there is no batch-to-batch variability
for various preparations of GNP-anti-CRP Ab (Fig. 4d). The bioa-
nalytical performance of the DIA employing SBCR is comparable
to that of the conventional sandwich ELISA using MTPR and the
 A Smartphone-Based Colorimetric Reader for Human C-Reactive Protein Immunoassay 345

Fig. 1 (a) Schematics of the smartphone-based colorimetric readers (SBCRs) developing using the gadgets’
(iPAD4, iPAD mini, or iPhone 5s) screen-based bottom illumination, Samsung Galaxy SIII mini’s back camera
(5 MP) based imaging, and a custom-made polyamide dark hood and polyamide base holder assembly. (b)
Screensavers used for the screen-based bottom illumination of the 96-well microtiter plate (MTP) in gadgets.
(c) Image processing algorithm employed for the analysis of smartphone-captured colorimetric images

clinically accredited analyzer-based IA. Similarly, the SBCR readout

based DIA is identical in sensitivity to MTPR readout based DIA, as
confirmed by the overlay plot of the normalized signals (Fig. 5).
346 A.G. Venkatesh et al.

Fig. 2 Typical SBCR assemblies, demonstrating the steps involved for SBCR-based colorimetric readout of
immunoassays, for (a) iPAD4, (b) iPAD mini, and (c) iPhone 5s. The process involves the placement of the
gadget, with an illuminated screensaver, within the engraved cavity of the base holder. The MTP with
colorimetric products is then placed on top of the screensaver after aligning with the respective illumination
spots. Subsequently, the dark hood is placed on top of the base holder and the smartphone imaging is
performed after placing the smartphone in the containment provided at the top of the dark hood and aligning
the smartphone’s back camera with the hole

2 Materials

2.1 Preparation 1. Nunc microwell 96-well polystyrene plates, flat bottom (non-
of Capture Ab-Bound treated), sterile (Catalog No. P7491, Sigma-Aldrich).
MTP 2. Eppendorf microtubes (1.5 mL; Catalog No. Z 606340,
Sigma-Aldrich).
3. Deionized water (18.2 MΩ, DIW). (Direct-Q®3 Water Purifi-
cation System, Millipore).
4. 70
C freezer (operating range 60 to 86
C) (New
Brunswick).
5. 2–8
C refrigerator (Future, UK).
6. Direct-Q®3 water purification system (Millipore, USA).
7. Mini incubator (Labnet Inc., UK).
8. PVC fume cupboard Chemflow range (CSC Ltd.).
9. KOH pellets (99.99%), semiconductor grade (Catalog No.
306568, Sigma-Aldrich).
 A Smartphone-Based Colorimetric Reader for Human C-Reactive Protein Immunoassay 347

Fig. 3 Schematic of the developed bioanalytical procedure for human CRP IA

10. 3-aminopropyltriethoxysilane (3-APTES) (Catalog No.

A3684, Sigma-Aldrich).
11. Graphene nanoplatelets (diameter 5 μm) (Catalog No. Grade
2 Graphene nanoplatelets, Cheap Tubes Inc): GNPs (1 mg) are
mixed with 1 mL of 0.25% APTES and sonicated for 1 h.
12. 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochlo-
ride (EDC) (Catalog No. 22981, Thermo Scientific).
348 A.G. Venkatesh et al.

Fig. 4 The bioanalytical performance of the developed SBCR-based human CRP IA. (a) Detection of CRP in PBS
(10 mM, pH 7.4), diluted human plasma and diluted human whole blood. (b) Experimental process controls.
Anti-CRP1 and anti-CRP2 are capture and detection antibodies, respectively. The control proteins employed
are human serum albumin (HSA), human lipocalin-2 (LCN2), human fetuin A (HFA), interleukin (IL)-1 beta, IL-6,
IL-8, and tumor necrosis factor (TNF)-alpha. (c) Storage stability of anti-human CRP capture Ab-bound
graphene-functionalized MTPs stored at 4
C in PBS for 6 weeks. (d) Determination of batch-to-batch
variability for various preparations of GNP-anti-CRP Ab, as demonstrated for the detection of 9 ng/mL CRP
by the developed graphene-based IA. All experiments are performed in triplicate with the error bars
representing the standard deviation. Reproduced with permission from Elsevier Inc. [6]

13. BupH phosphate buffered saline packs (0.1 M sodium phos-

phate, 0.15 M sodium chloride, pH 7.2) (Catalog No. 18372,
Thermo Scientific).
14. BupH MES buffered saline packs (0.1 M MES [2-(N-morpho-
lino)ethane sulfonic acid], 0.9% (w/v) sodium chloride,
pH 4.7) (Catalog No. 28390, Thermo Scientific).
15. Human CRP Ab (Catalog No. MAB17071, R&D Systems).
 A Smartphone-Based Colorimetric Reader for Human C-Reactive Protein Immunoassay 349

Fig. 5 An overlay plot of the developed human CRP IAs based on the use of SBCR and MTPR, and employing
the normalized signals (the signal response obtained for a particular human CRP concentration divided by the
signal response obtained for the highest human CRP concentration). Reproduced with permission from
Elsevier Inc. [6]

16. Blocker BSA in PBS (10), pH 7.4, 10% w/v (Catalog No.
37525, Thermo Scientific).
17. 0.1% BSA in PBS, pH 7.4 is used as the binding buffer.
18. 0.05% Tween® 20 in PBS, pH 7.4 (PBST) is used as the
washing buffer.

2.2 Human CRP IA 1. Recombinant human CRP (Catalog No. 1707-CR, R&D
Systems).
2. Biotinylated mouse anti-human CRP detection Ab (Catalog
No. BAM17072, R&D Systems).
3. Streptavidin-conjugated horseradish peroxidase (SA-HRP,
Catalog No. , R&D Systems).
4. Biotinylated anti-CRP detection Ab conjugated to SA-HRP is
prepared by adding 1 μL of biotinylated anti-CRP detection Ab
(0.5 mg/mL) to 1 μL of SA-HRP to 2998 μL of the binding
buffer followed by 20 min of incubation at RT. The resulting
concentration of biotinylated anti-CRP detection Ab is
0.17 μg/mL, while SA-HRP dilution is 1:3000.
5. 3,30 ,5,50 -Tetramethylbenzidine (TMB) substrate kit (Catalog
No. 34021, Thermo Scientific).
(a) TMB solution (0.4 g/L).
(b) Hydrogen peroxide solution (containing 0.02% v/v
H2O2 in citrate buffer).
350 A.G. Venkatesh et al.

6. Human whole blood (HQ-Chex level 2) (Catalog No.

232754, Streck).
7. Human serum (CRP free) (Catalog No. 8CFS, HyTest Ltd.).
8. Recombinant human serum albumin (HSA, Catalog No.,
R&D Systems), human fetuin A (HFA, Catalog No., R&D
Systems), human lipocalin 2 (LCN2, Catalog No., R&D Sys-
tems), interleukin (IL)-1β (Catalog No., R&D Systems), IL-6
(Catalog No., R&D Systems), IL-8 (Catalog No., R&D Sys-
tems) and tumor necrosis factor (TNF)-α (Catalog No., R&D
Systems).
9. The CRP spiked diluted human whole blood or serum samples
(0.3–81 ng/mL) are prepared by mixing the desired CRP
concentrations in 1:100 diluted human whole blood/serum.
10. Anonymized EDTA plasma samples from patients: The leftover
anonymized EDTA plasma samples of patients are provided by
Dr. Eberhard Barth, University Hospital Ulm, Germany. They
are diluted 1:1000 and 1:4000 in the binding buffer so that the
CRP concentration in these samples falls within the linear
range of the DIA, which enables the detection of an entire
pathophysiological concentration range of human CRP.
11. Tecan Infinite M200 Pro microplate reader (Tecan, Austria
GmbH).
12. Roche COBAS® 8000 modular analyzer.

2.3 SBCR Analysis 1. Samsung Galaxy SIII mini.

2. iPAD mini, iPAD4, iPhone 5s.
3. Custom-made polyamide dark hood and polyamide base
holder assembly (Figs. 1 and 2).
4. Image J 1.48v (freely available to download and use at http://
imagej.nih.gov/ij).

3 Methods

3.1 Preparation 1. PBS: A BupH PBS pack is dissolved in 100 mL of autoclaved

of Capture Ab-Bound DIW. The resulting volume is adjusted to 500 mL using auto-
MTP claved DIW. Each BupH PBS pack makes 500 mL of PBS at
pH 7.2 (see Notes 1 and 2).
2. MES: A BupH MES pack is dissolved in autoclaved DIW with a
final concentration of 500 mL. Each BupH MES pack makes
500 mL of MES at pH 4.7 (see Notes 1 and 3).
3. APTES: The commercial APTES solution (99% purity) is
reconstituted in autoclaved DIW to make an effective 0.25%
(v/v) solution (see Note 4).
 A Smartphone-Based Colorimetric Reader for Human C-Reactive Protein Immunoassay 351

4. EDC: An effective concentration of 8 mg/mL is made by

reconstituting EDC in 0.1 M MES buffer, pH 4.7 (see Notes
5 and 6).
5. Binding buffer: 0.1% BSA in PBS, pH 7.4.
6. Washing buffer: 0.05% Tween® 20 in PBS, pH 7.4 (PBST).
7. GNPs: GNPs (1 mg) is sonicated in 1 mL of 0.25% APTES for
1 h.
8. EDC-activated anti-human CRP Ab: EDC (0.4 mg or 2 mM)
is dissolved in 100 μL of 0.1 M MES, pH 4.7. Thereafter,
990 μL of the anti-human CRP Ab (4 μg/mL) is incubated
with 10 μL of EDC (4 mg/mL) for 15 min at RT (see Note 7).
9. Each MTP well is incubated with 100 μL of 1% (w/v) KOH in
DIW for 10 min and then washed five times with 300 μL DIW
(see Note 8).
10. EDC-activated Ab is mixed with GNPs (1 mg/mL) in 0.25%
APTES in the ratio of 1:1 (v/v). The resulting anti-human
CRP Ab solution (100 μL, 2 μg/mL, 0.5 mg/mL GNPs and
0.125% APTES) is then added to the MTP wells and incubated
for 30 min at RT. The Ab-bound and GNPs-functionalized
MTP wells are washed five times with 300 μL of PBST.
11. The MTP wells are blocked by incubating with 300 μL of 5%
(w/v) BSA for 30 min at 37
C and washing subsequently with
300 μL of wash buffer five times. Washing can also be per-
formed with an automatic plate washer. This step is essential to
prevent nonspecific binding to the unbound sites available on
the GNPs and the MTP (see Notes 9 and 10).

3.2 Human CRP IA 1. CRP spiked diluted human whole blood or serum: The CRP
spiked samples (0.3–81 ng/mL) are prepared by mixing the
desired CRP concentrations in 1:100 diluted human whole
blood/serum.
2. EDTA plasma samples from anonymized patients. The EDTA
plasma samples from anonymized patients are diluted 1:1000
and 1:4000 in the binding buffer, which makes the resulting
CRP concentration in these samples to fall within the DIA
linear range, thereby enabling the detection of an entire patho-
physiological concentration range of human CRP.
3. Biotinylated anti-CRP detection Ab (100 μL, 0.17 μg/mL)
pre-conjugated to SA-HRP and 100 μL of CRP (varying con-
centrations; 0.3–81 ng/mL) are dispensed sequentially to the
Ab-bound and BSA-blocked MTP wells, and incubated for
15 min at 37
C (see Note 11).
4. The sandwich immune complex-bound MTP is washed with
300 μL of PBST five times to remove nonspecifically bound
substances and excess IA reagents.
352 A.G. Venkatesh et al.

5. A TMB-H2O2 mixture (100 μL) is added to each MTP well

and incubated for 4 min at RT to allow the enzymatic reaction
by horseradish peroxidase to develop color. TMB acts as a
hydrogen donor for the reduction of hydrogen peroxide to
water by the enzyme. The resulting diimine causes the solution
blue and its absorbance can be read at 650 nm.
6. The enzymatic reaction is stopped by adding 50 μL of 2 N
H2SO4 to each MTP well (see Note 12).
7. If there are problems such as no development of color or a high
background, Notes 13 and 14 should be followed.

3.3 SBCR Analysis 1. The colorimetric readout of the MTP is performed by smart-
phone imaging using our developed SBCR set up (Figs. 1 and 2).
The gadget (iPAD 4, iPAD mini or iPhone 5s) is placed in the
designated groove of the base holder followed by placing the
96-well MTP on the gadget’s screensaver that provides white
light-based bottom illumination only in the specific regions
corresponding to the MTP well’s bottom. The dark box is then
placed on top of the base holder in the alignment groove.
Subsequently, the Samsung SIII mini’s back camera-based imag-
ing of the colorimetric reaction in the MTP is performed by
placing the smartphone inside the designated groove on top of
the dark hood (see Note 15).
2. The desired pixel intensity (PI) of the captured smartphone
image is then determined by our novel image analysis algo-
rithm using Image J 1.48v (http://imagej.nih.gov/ij)
(Fig. 1c). The color channel of the image is split and the pixel
coordinate of an individual MTP well’s center is identified
using the red channel image. A mean of neighboring pixels
(varies based on the resolution of the camera) from the center is
calculated individually for all the channels. The composite
mean PI (CMPI) is derived from the following color-weighted
formula [CMPI ¼ 0.7 MPIBlue + 0.2 MPIGreen + 0.1 MPIRed].
The desired PI is determined as 255  CMPITest Conc  CMPI-
Blank. The resulting CMPI is plotted against their respective—
CRP concentrations using the four-parameter logistic based
standard curve analysis in SigmaPlot version 11.2 software.
The determination of CMPI and PI is done in Microsoft
Excel after taking the PI values of red, blue and green channels
of the smartphone-captured image using image J. The concen-
tration of unknown CRP samples is determined on the basis of
their CMPI from the resulting calibration plot. As described in
Subheading 3.2, a standard curve is plotted using various
known concentrations of CRP (0.3–81 ng/mL). A schematic
of the image analysis algorithm, comprising various steps, is
provided in Fig. 6.
 A Smartphone-Based Colorimetric Reader for Human C-Reactive Protein Immunoassay 353

Fig. 6 Schematic of the image analysis procedure for the SBCR based colorimetric readout

4 Notes

1. Inhalation of BupH PBS and MES buffer packs must be

avoided and the solutions should be prepared in autoclaved
DIW (18.2 MΩ).
2. The reconstituted BupH PBS is good for a week at RT and up
to 4 weeks at 4
C.
3. The reconstituted BupH MES can be stored at RT for up to
2 weeks.
4. The APTES solution should be prepared fresh just prior to use.
354 A.G. Venkatesh et al.

5. The recommended concentration of EDC should be used for

the optimal cross-linking of capture Ab.
6. The aliquots of EDC can be stored at 20
C for up to 6
months.
7. The reconstituted Ab and antigen, stored at 2–8
C, are good
within a month. The aliquots stored at 20 to 70
C are
good for up to 6 months.
8. KOH treatment over 10 min may cause strong surface aberra-
tion, which in turn modifies its properties.
9. The blocker BSA is filtered with 0.2 μm pore size filter paper
prior to use to avoid contamination.
10. The human CRP concentrations must be prepared in BSA-
preblocked sample vials to minimize the analyte loss due to
its nonspecific surface binding.
11. SA-HRP should not be frozen and simply stored in the dark at
4
C as streptavidin is light-sensitive. In the case of its exposure
to light, a freshly prepared SA-HRP solution should be used.
12. Personal protective equipment (PPE), such as chemical safety
glasses, chemical-resistant shoes and lab coats, must be used for
handling sulfuric acid, H2O2, KOH, APTES, and EDC, in a
fume cabinet. Skin contact must be avoided with sulfuric acid, a
strong corrosive agent and an irritant. In the case of skin
contact, the exposed area should be washed immediately with
acid neutralizers and medical advice should be sought urgently.
KOH can cause severe burns. Its concentration must be 1%
(w/v) in autoclaved DIW as higher concentrations may affect
the surface properties, which leads to a decreased binding of
capture Ab. APTES is a skin and eye irritant, and highly toxic to
the kidney. Being hygroscopic, EDC absorbs moisture that
leads to its activity loss, so it should be equilibrated to RT
before opening the sample vial.
13. If there is no development of color in DIA, the possible reasons
could be the degradation of KOH; loss of activity of APTES
due to hydrolysis; loss of activity of EDC due to hydrolysis;
inactivity of capture Ab, detection Ab or CRP; and/or degra-
dation of TMB or H2O2. This problem could be solved by
employing a new batch or stock of these chemicals or immu-
nological reagents.
14. If there is a high background signal, this could be due to the
loss of activity or contamination of BSA. Therefore, a fresh
stock of BSA should be used to obviate this problem.
15. The illumination of the gadget’s screen is set to maximal set-
tings before its use in the developed SBCR.
 A Smartphone-Based Colorimetric Reader for Human C-Reactive Protein Immunoassay 355

Acknowledgments

We thank PD Dr. Eberhard Barth for providing the leftover anon-

ymized EDTA plasma samples of patients treated by intensive care
at University Hospital Ulm, Germany for validating the DIA.

References
1. Vashist SK, Mudanyali O, Schneider EM et al 12. Tappero E, Johnson P (2010) Laboratory eval-
(2014) Cellphone-based devices for bioanaly- uation of neonatal sepsis. Newborn Infant
tical sciences. Anal Bioanal Chem 406: Nurs Rev 10:209–217
3263–3277 13. http://www.nomidalliance.org/downloads/
2. http://www.itu.int/en/ITU-D/Statistics/ comparative_chart_front.pdf. Accessed 8 June
Documents/publications/mis2013/MIS2013_ 2015
without_Annex_4.pdf. Accessed 8 June 2015 14. Danner M, Kasl SV, Abramson JL et al (2003)
3. Vashist SK, Schneider EM, Luong JHT (2014) Association between depression and elevated
Commercial smartphone-based devices and C-reactive protein. Psychosom Med
smart applications for personalized healthcare 65:347–356
monitoring and management. Diagnostics 15. Raison CL, Capuron L, Miller AH (2006)
4:104–128 Cytokines sing the blues: inflammation and
4. Ozcan A (2014) Mobile phones democratize the pathogenesis of depression. Trends Immu-
and cultivate next-generation imaging, diag- nol 27:24–31
nostics and measurement tools. Lab Chip 16. Spitzer C, Barnow S, Volzke H et al (2010)
14:3187–3194 Association of posttraumatic stress disorder
5. Vashist SK, van Oordt T, Schneider EM et al with low-grade elevation of C-reactive protein:
(2015) A smartphone-based colorimetric evidence from the general population. J Psy-
reader for bioanalytical applications using the chiatr Res 44:15–21
screen-based bottom illumination provided by 17. Lin MS, Shih SR, Li HY et al (2007) Serum
gadgets. Biosens Bioelectron 67:248–255 C-reactive protein levels correlates better to
6. Vashist SK, Schneider EM, Zengerle R et al metabolic syndrome defined by Interna-
(2015) Graphene-based rapid and highly- tional Diabetes Federation than by NCEP
sensitive immunoassay for C-reactive protein ATP III in men. Diabetes Res Clin Pract
using a smartphone-based colorimetric reader. 77:286–292
Biosens Bioelectron 66:169–176 18. Morimoto H, Sakata K, Oishi M et al (2013)
7. Marnell L, Mold C, Du Clos TW (2005) C- Effect of high-sensitivity C-reactive protein on
reactive protein: ligands, receptors and role in the development of diabetes as demonstrated
inflammation. Clin Immunol 117:104–111 by pooled logistic-regression analysis of annual
8. Chiesa C, Natale F, Pascone R et al (2011) C health-screening information from male Japa-
reactive protein and procalcitonin: reference nese workers. Diabetes Metab 39:27–33
intervals for preterm and term newborns dur- 19. Testa R, Bonfigli AR, Sirolla C et al (2008) C-
ing the early neonatal period. Clin Chim Acta reactive protein is directly related to plasmino-
412:1053–1059 gen activator inhibitor type 1 (PAI-1) levels in
9. Chiesa C, Panero A, Osborn JF et al (2004) diabetic subjects with the 4G allele at position
Diagnosis of neonatal sepsis: a clinical and lab- 675 of the PAI-1 gene. Nutr Metab Cardio-
oratory challenge. Clin Chem 50:279–287 vasc Dis 18:220–226
10. Dollner H, Vatten L, Austgulen R (2001) Early 20. Kuo HK, Yen CJ, Chen JH et al (2007) Asso-
diagnostic markers for neonatal sepsis: compar- ciation of cardiorespiratory fitness and levels of
ing C-reactive protein, interleukin-6, soluble C-reactive protein: data from the National
tumour necrosis factor receptors and soluble Health and Nutrition Examination Survey
adhesion molecules. J Clin Epidemiol 1999–2002. Int J Cardiol 114:28–33
54:1251–1257 21. Park HE, Cho GY, Chun EJ et al (2012) Can
11. Nguyen-Vermillion A, Juul SE, McPherson RJ C-reactive protein predict cardiovascular events
et al (2011) Time course of C-reactive protein in asymptomatic patients? Analysis based on
and inflammatory mediators after neonatal sur- plaque characterization. Atherosclerosis
gery. J Pediatr 159:121–126 224:201–207
356 A.G. Venkatesh et al.

22. Ridker PM (2003) High-sensitivity C-reactive with a schizophrenia spectrum disorder treated
protein and cardiovascular risk: rationale for in routine medical practice. Eur Psychiatry
screening and primary prevention. Am J Car- 28:161–167
diol 92:17K–22K 24. Wilson AM, Ryan MC, Boyle AJ (2006) The
23. Sicras-Mainar A, Rejas-Gutierrez J, Navarro- novel role of C-reactive protein in cardiovascu-
Artieda R et al (2013) C-reactive protein as a lar disease: risk marker or pathogen. Int J Car-
marker of cardiovascular disease in patients diol 106:291–297
 Chapter 22

A Novel Colorimetric PCR-Based Biosensor for Detection

and Quantification of Hepatitis B Virus
Li Yang, Mei Li, Feng Du, Gangyi Chen, Afshan Yasmeen, and Zhuo Tang

Abstract
Hepatitis B virus (HBV) can cause viral infection that attacks the liver and it is a major global health problem
that puts people at a high risk of death from cirrhosis of the liver and liver cancer. HBV has infected one-
third of the worldwide population, and 350 million people suffer from chronic HBV infection. For these
reasons, development of an accurate, sensitive, and expedient detection method for diagnosing, monitor-
ing, and assessing therapeutic response of HBV is very necessary and urgent for public health and disease
control. Here we report a new strategy for detection of viral load quantitation of HBV based on colorimet-
ric polymerase chain reaction (PCR) with DNAzyme-containing probe. The special DNAzyme adopting a
G-quadruplex structure exhibited peroxidase-like activity in the presence of hemin to report colorimetric
signal. This method has shown a broad range of linearity and high sensitivity. This study builds an important
foundation to achieve the specific and accurate detection level of HBV DNA with a low-cost and effective
method in helping diagnosing, preventing and protecting human health form HBV all over the world, and
especially in developing countries.

Key words Hepatitis B virus, Molecular diagnosis, Biosensor, G-quadruplex, DNAzyme, Peroxidase-
like activity, Colorimetric PCR

1 Introduction

Hepatitis B virus (HBV) can cause a potential life-threatening liver

infection and pathology ranging from self-limited illness to chronic
hepatitis, cirrhosis, and hepatocellular carcinoma [1]. Approxi-
mately 400 million people are infected globally by HBV and
about one-third of the world population has been infected once
in their lives [2]. It is an important health hazard for the worldwide
population and the most serious type of viral hepatitis.
Presently, serologic immunity and nucleic acid testing are the
most commonly used methods for HBV diagnosis, prevention, and
treatment in clinical medicine [3]. The HBV serological tests such
as ELISA assays are used to distinguish acute, self-limited infections
from chronic HBV infections and monitor vaccine-induced

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_22,
© Springer Science+Business Media LLC 2017

357
358 Li Yang et al.

immunity by testing for a series of serological markers of HBV. But

this method can only be used to qualitative diagnosis with low
sensitivity and cause false-positive frequently. Besides these, nucleic
acid detection of HBV DNA based on the PCR techniques is also
used to quantify HBV viral load and track the effectiveness of
therapeutic drugs [4]. With PCR amplification, specific segment
of viral DNA can be magnified and a large number of copies of the
target DNA sequence are produced across several orders of magni-
tude [5]. Compared to the traditional immunological assays, the
nucleic acid assays show higher sensitivity, accuracy, and specificity,
which get rid of the evident limit for the accurate detection of HBV
DNA in the progression of disease as well as the therapeutic effect.
Because the nucleic acid assays can solve these problems which are
brought about by immunological test and offer several obvious
advantages, it is the most reliable diagnosis technique for detecting
and monitoring HBV infection as well as assessing therapeutic
response.
TaqMan technique in nucleic acid assays is one of the most
popular methods used for detection of HBV DNA. The said tech-
nique uses a TaqMan probe containing a fluorophore and a
quencher at both the ends respectively. In the single-stranded
form, florescence is not detected because of fluorescence resonance
energy transfer (FRET). When the PCR extension progresses, a
DNA polymerase with the activity of 50 –30 exonuclease leads to
degrading the fluorophore-modified DNA probe that will anneal
to the target strand [6]. But the main limitation of the TaqMan
method is the high-cost modified flurogenic oligonucleotide
probes and sophisticated equipment, which are restricted to well-
equipped laboratories and less accessible to many ordinary users.
G-quadruplexes are higher-order DNA and RNA structures
formed from G-rich sequences that are built around tetrads of
hydrogen-bonded guanine bases. The criteria for a potential quad-
ruplex sequence is restricted to: G3–5NL1G3–5NL2G3–5NL3G3–5‚
where NL1–3 are loops of unknown length, within the limits
1 < NL1–3 < 7 nt. DNAzymes are catalytic single-stranded deox-
yribonucleic acids that are obtained through in vitro selection. The
DNAzyme we use has been found to exhibit peroxidase-like activity
by forming G-quadruplex structure and binding with hemin. In
recent years, this DNAzyme has been used to design various color-
imetric or chemiluminescent assays. Herein, we report a novel
approach for both qualitative and quantitation of HBV DNA
based on our previous work [7]. This method takes the key advan-
tage of the TaqMan technology (i.e., the elegant use of the
50 -exonuclease activity of Taq polymerase) [8], and applies a low-
cost catalytic DNA molecular beacon as probe [9]. In the process of
PCR amplification, Taq DNA polymerases cleave the probe and
release a DNAzyme sequence [10–12], which has been embedded
in the probe. After PCR amplification, the DNAzyme can form
 A Novel Colorimetric PCR-Based Biosensor for Detection and Quantification. . . 359

G-quadruplex and bind with hemin, possessing a peroxidase-like

activity which catalyzes the oxidation of different substrates by
H2O2 to generate either colorimetric or fluorometric signals
[13–15].

2 Materials

2.1 Preparation 1. T4 polynucleotide kinase (10 U/μL; TransGen Biotech,

of 32P Labeled Probe Beijing, China).
2. [γ-32P] ATP (Furui Biological Engineering, Beijing, China).
3. Probe-HBV-2 (Sangon Biotech, Shanghai, China) is dissolved
in sterile water.

2.2 Clinical Serum 1. QIAamp DNA Blood Mini Kit (Qiagen).

Sample Extraction 2. HBV infection clinical serum samples (from HBV patients).
3. Normal clinical serum samples (from normal population).

2.3 Amplification 1. HBV DNA template (the DNA is extracted from the HBV
infection clinical serum samples) is dissolved in sterile water,
stored at 20  C.
2. Primers (forward primer: 1.5 μM, reverse primer: 1 μM; San-
gon Biotech, Shanghai, China) are dissolved in sterile water,
stored at 20  C.
3. Probe (32P labeled, 0.6 μM; Sangon Biotech, Shanghai, China)
is dissolved in sterile water, stored at 20  C.
4. 1.2 mM dNTPs (600 μM dUTP, 200 μM dATP, 200 μM
dCTP, 200 μM dGTP; TransGen Biotech, Beijing, China).
5. FastStart Taq DNA polymerase (5 U; Roche).
6. Taq DNA polymerase (5 U; TransGen Biotech, Beijing,
China).
7. UNG (1 U; Takara).
8. MgCl2 (0.5 mM).
9. 1 Taq polymerase buffer (15 mM Tris–HCl (pH 8.2), 30 mM
KCl, 5 mM (NH4)2SO4, 2.5 mM MgCl2, 0.002% BSA; Trans-
Gen Biotech, Beijing, China).
10. C1000 thermal cycler (Bio-Rad).

2.4 Colorimetric 1. 400 mM NaCl solution is prepared by dissolving 0.2340 g

Detection NaCl in 10 ml of ultrapure water.
2. Hemin (Alfa Aesar) is dissolved in DMSO.
3. ABTS (2,20 -azinobis-(3-ethylbenzthiazoline-6-sulphonate))
(Wolsen, Xi’an, China) is dissolved in ultrapure water.
360 Li Yang et al.

4. H2O2 (Bodi Chemical Holding Co., Ltd., Tianjin, China) is

diluted in ultrapure water.
5. Varioskan Flash (λ ¼ 414 nm, 40 readings with a 30 s interval
are recorded; Thermo Scientific).

2.5 TaqMan Assay 1. Hepatitis B Viral DNA Quantitative Fluorescence Diagnostic

Kit (48 Tests; Sansure Biotech, Changsha, China).
2. Real-Time PCR System (Thermo Scientific).

3 Methods

3.1 The Principle of 1. The principle of our detection strategy is depicted in Fig. 1.
the Colorimetric The DNA probe containing three parts of sequences (A, B, C)
Detection is designed to form a hairpin structure at room temperature.
After denaturation step of PCR, the concentration of probes is
much higher than that of HBV dsDNA templates, so the
chance for probe to hybridize with the targeted HBV single
strand DNA is much higher than that of its complementary
single strand DNA. Then, in the annealing step of PCR, the
loop domain (part C) of the probe hybridizes to the conserved
region of HBV genome, which will be amplified by two pri-
mers. The stem part of the probe contains blocking sequence A
(black) and a DNAzyme sequence B (blue) that could report
the detection result through oxidation reaction [16–20]. The
blocking sequence A is used to prohibit DNAzyme sequence B
fold into catalytic G-quadruplex structure at room temperature
to decrease the background of colorimetric assay [7]. As the

Fig. 1 Strategy of the colorimetric PCR-based detection of HBV. A part of the probe is the blocking sequence, B
represents the sequence for DNAzyme forming G-quadruplex, loop C is the complementary sequence of the
conserved region of HBV genome. (1) Melting of double-stranded DNA and hairpin loop during PCR denatur-
ation. (2) Hybridization of probe and primers to targeted ssDNA respectively during PCR annealing step. (3)
Cleaving of the hybridized probe by DNA polymerase during primers extension. (4) Release of G-quadruplex
sequence after cleavage by DNA polymerase. (5) Horseradish peroxidase-mimicking DNAzymes formation
with the addition of hemin, reacting with ABTS and H2O2. (6) Colorimetric test result
 A Novel Colorimetric PCR-Based Biosensor for Detection and Quantification. . . 361

PCR proceeding, the probe formed a stable duplex with target

sequence of HBV genome and Taq DNA polymerase with 50 –30
exonuclease activity can cleave the obstruction of part A, con-
sequently released the part B. After PCR amplification, the
cleaved fragment of the probe could form a stable
G-quadruplex structure and bind with hemin to exhibits the
peroxidase-like activity to oxidize ABTS in the presence of
H2O2, leading to produce a colorimetric green product. With
the exception of the HBV sequence, the blocking sequence and
the sequence for DNAzyme forming G-quadruplex can be used
for detection of other targets.

3.2 DNA Extraction 1. Extraction of DNA from the HBV infection and normal clinical
from Clinical Serum serum samples from HBV patients or normal population is
Sample carried out using QIAamp DNA Blood Mini Kit as described
by the manufacturer (Qiagen).

3.3 The Serum 1. 108 copies/mL clinical serum samples are chosen to dilute with
Calibration Curve negative serum by decuple to ten copies successively. These
Preparation samples are extracted with QIAamp DNA Blood Mini Kit
(Qiagen) to build the calibration curve.

3.4 Manual Design of 1. According to our previous procedure, this method requires
Probe and Primers two consecutive PCR steps to achieve the successful detection
of a target DNA, which include the amplification of target
sequence and subsequently the cleavage of probe [7]. In our
experiment, three different hairpin probes (probe-HBV-1,
probe-HBV-2, and probe-HBV-3) and corresponding primers
are designed against the different conserved regions of the
HBV genome including overlapping genes encoding
X-protein, S gene, and X gene region.
2. Under the optimized conditions, Fig. 2 shows that probe-
HBV-2 for S gene region of HBV genome exhibits the best
colorimetric result comparative to that of other probes. Fur-
thermore, this region of HBV genome is rather conserved in
different genotypes of HBV strains and can be extensively and
accurately used to detect all HBV genotypes. Therefore, the
probe-HBV-2 has been proved to be the best fitted probe for
HBV detection based on our method. The selected sequences
are:
Primer-HBV-Forward-2:CCTGGTTATCGCTGGATGTGT,
Primer-HBV-Reverse-2:GGACAAACGGGCAACATACCTT,
Probe-HBV-2: CCCTACCCA TTCATCCTGCTGCTATG
CCTCATCTTCTT TGGGTAGGG CGGGTTGGGAAA -
NH2 (see Note 1).
The sequence marked with box represents the stem part of
Probe-HBV-2. Sequence in the second box would be closed
362 Li Yang et al.

Fig. 2 Colorimetric PCR reaction with three different probes. Tube 1: probe-HBV-
1 with HBV DNA targets; tube 2: probe-HBV-1 without targets; tube 3: probe-
HBV-2 with targets; tube 4: probe-HBV-2 without targets; tube 5: probe-HBV-3
with targets; tube 6: probe-HBV-3 without targets

at room temperature. The underlined sequence could anneal to

S gene region of HBV genome and is the loop domain of the
probe. Sequence for DNAzyme forming G-quadruplex is
shown in boldface.

3.5 One-Step PCR 1. Thaw 10 Taq polymerase buffer, dNTPs, and primers.
2. The PCR is performed in 50 μL volume containing 2 μL HBV
DNA template, 1.5 μM Primer-HBV-Forward-2, 1 μM
Primer-HBV-Reverse-2, 0.6 μM Probe-HBV-2, 1.2 mM
dNTPs (600 μM dUTP, 200 μM dATP, 200 μM dCTP,
200 μM dGTP), 5 U of FastStart Taq DNA polymerase, 1 U
of UNG (see Note 2), 0.5 mM MgCl2, 1 Taq polymerase
buffer (15 mM Tris–HCl (pH 8.2), 30 mM KCl, 5 mM
(NH4)2SO4, 2.5 mM MgCl2, 0.002% BSA).
3. The negative control is prepared without HBV DNA.
4. The amplification conditions are as follows: 20  C for 10 min;
95  C for 2 min followed by 40 cycles of 94  C for 30 s and
60  C for 90 s.

3.6 32P Labeled 1. To verify the probe had been cleaved by DNA polymerase
Selected Probe PCR successfully, Probe-HBV-2 and mark are 50 end-labeled with
Reaction [γ-32P] ATP. The reaction mixture containing oligonucleotides
with 10 μCi [γ-32P] and 10 U of T4 polynucleotide kinase is
incubated for 1 h at 37  C for DNA phosphorylation.
2. The labeled product is purified by 10% denaturing PAGE.
3. The 50 32P labeled probe-HBV-2 is added into the PCR reaction.
4. The PCR product is used for PAGE analysis. As shown in
Fig. 3, in the positive control containing HBV genome, a
A Novel Colorimetric PCR-Based Biosensor for Detection and Quantification. . . 363

Fig. 3 PAGE analysis of the cleavage of isotope-labeled probe-HBV-2 in PCR

amplification. Lane 1, from top to bottom 17-nt, 13-nt, 9-nt DNA marker; lane 2,
PCR reaction containing isotope-labeled probe-HBV-2 and HBV DNA targets; lane
3, negative control: PCR reaction without HBV DNA targets; lane 4, 32P-labeled
probe-HBV-2 as marker

13-nt fragment is released from probe-HBV-2 during the PCR

reaction (lane 2, Fig. 3). Comparatively, in the negative con-
trol, without HBV target, the probe displayed no cleavage (lane
3, Fig. 3). When Taq DNA polymerase encounters the hybri-
dized complementary duplex between the target site of HBV
genome and the probe-HBV-2, it cleaves the probe at the
second nucleotide of stable double-stranded domain. Then, a
13-nt 32P-labeled fragment from probe-HBV-2 is released.
The result is consistent with our previous research about 50 –30
exonuclease activity of DNA polymerase. So these results prove
that the final colorimetric result of positive control is caused by
the cleavage of probe to release enzymes as reporter in the
process of PCR amplification.
364 Li Yang et al.

3.7 Colorimetric 1. NaCl at the concentration of 400 mM (see Note 3) is added to

Detection the PCR products.
2. The reaction mixture is heated at 94  C for 5 min.
3. The reaction mixture is incubated at room temperature for
30 min.
4. Hemin (2 μM), ABTS (2.4 mM), and H2O2 (2 mM) (see
Note 4) are added successively.
5. Colorimetric signals of different concentrations of HBV DNA
are recorded (Fig. 4a). Samples containing different concentra-
tions of HBV DNA (from 10 to 107 copies) has been detected,
the colorimetric results reveal a gradual increase of color inten-
sity from light to dark green, while the negative control con-
taining no HBV target remained colorless. As few as ten copies
of HBV DNA could be qualitatively detected with naked eyes
in 50 μL1 reaction mixture. The method represents a simply
visual way for HBV DNA detection, which refrains from the
advanced equipment and harsh detection conditions.

Fig. 4 Colorimetric detection of HBV DNA. (a) Photograph of the colorimetric detection of different concentra-
tion of HBV DNA: tube 1, negative control (without HBV DNA); tubes 2–8 containing 10, 102, 103, 104, 105, 106,
107 copies of HBV DNA, successively. (b) Time-dependent colorimetric detection of different concentrations of
HBV DNA. The inset is a calibrated curve of the average absorbance (414 nm) at 3 min plotted against the
number of HBV DNA. The solid line indicates linear least squares fitting between 10 and 107 copies of HBV
DNA, and their formulation is FI ¼ 0.0061 lg(CHBVDNA) + 0.0678 (R2 ¼ 0.9788). The correlation coefficient is
0.9788. The error bars were determined by standard deviation (SD) of the triplicate data
 A Novel Colorimetric PCR-Based Biosensor for Detection and Quantification. . . 365

6. Absorbance detection of ABTS is performed at 414 nm.

About 40 readings with a 30 s interval are recorded. With the
aid of simple UV-spectrometer of microplate-reader, more
accurate quantitative detection can be realized. A time-
dependent optical absorption change (414 nm) is recorded
(experiments are conducted in triplicate), and the relationship
between different concentrations of HBV DNA and absor-
bance is studied (Fig. 4b). The optical density is proportional
to the concentration of HBV DNA over the range of 10–107
copies. The inset is a calibration curve with colorimetric inten-
sity data at 414 nm obtained from the spectrum, corresponding
to different concentrations of HBV DNA (the average values of
triplicate data are used for plotting at 3 min). The calibration
curve reveals that there is a good linear relationship between
absorbance and the concentration of HBV DNA with a corre-
lation coefficient of 0.9788, which shows linear dynamic ranges
from 10 to 107 copies. We test the detection limit from 107 to
10 copies by serial dilutions, and the ten copies is the lowest
concentration in samples that we can dilute and detect repro-
ducibly. So a minimum detection for ten copies of HBV DNA
is achieved based on our method.

3.8 Comparation 1. TaqMan assay is the most advanced technique for HBV detec-
with TaqMan Assay tion, which is widely used in clinical diagnosis, so we compare
our colorimetric assay with the quantitative TaqMan PCR
assay. Seven different concentrations of clinical HBV serum
samples and seven negative clinical serum samples are measured
with the TaqMan method.
2. Prepared reagents according to the instruction of Hepatitis B
Viral DNA Quantitative Fluorescence Diagnostic Kit.
3. 200 μL of negative control, positive control, quantitative refer-
ence A–D, and the sample under test is respectively added to
seven clean 1.5 ml microcentrifuge tubes, each containing
300 μL of DNA extraction solution. Mix completely by vibrat-
ing for about 10 min.
4. Add 100 μL of DNA extraction solution 2-mix to each tube.
Mix for 10 s, then quiescence at room temperature for 10 min.
5. Place the microcentrifuge tubes into the magnetic bead separa-
tor for 3 min after a short centrifugation, then sucked out the
solution slowly.
6. 600 μL DNA extraction solution 3 and 200 μL DNA extraction
solution 4 are added to each tube. Mix for 5 s.
7. Place the microcentrifuge tubes into the magnetic bead separa-
tor again after a short centrifugation. 3 min later, there are two
layers of the solution. Discard the lower layer and transfer all of
the microcentrifuge tubes to tube rack.
366 Li Yang et al.

8. Add 50 μL of PCR-mix to each tube. Mix completely. The

mixed solution is transferred to clean 0.2 ml PCR tube
respectively.
9. Apply the mixed solution for Real-Time PCR under the
instruction of Hepatitis B Viral DNA Quantitative Fluores-
cence Diagnostic Kit.
10. Apply colorimetric assay to detect the same clinical sample in
triplicate; meanwhile consecutive serum samples ranging from
101 to 108 copies/mL are determined in triplicate to record
the time-dependent optical absorption changes (Fig. 5a) and

a
1.2

NC
1
101 Copies

0.8 102 Copies

Absorbance

103 Copies

0.6 104 Copies

105 Copies

0.4 106 Copies

107 Copies
0.2 108 Copies

0
1 2 3 4 5 6 7 8 9 10 11 12 13
T (min)
b
0.8
y = 0.0731x + 0.0319
0.7 R2 = 0.9822
0.6
Absorbance

0.5
0.4
0.3
0.2
0.1
0
101 102 103 104 105 106 107 108
Concentration (copy)

Fig. 5 Calibration curve establishment of HBV serum samples. (a) Time-

dependent colorimetric detection of different concentrations of HBV DNA
(414 nm). (b) Calibrated curve of the average absorbance (414 nm) at 3 min
plotted against the number of HBV DNA. The solid line indicates linear least
squares fitting between 10 and 108 copies of HBV DNA, and their formulation is
FI ¼ 0.0731 lg(CHBVDNA) + 0.0319 (R2 ¼ 0.9822). The correlation coefficient is
0.9822. The error bars were determined by standard deviation (SD) of the
triplicate data
 A Novel Colorimetric PCR-Based Biosensor for Detection and Quantification. . . 367

the calibration curve which used the average values of triplicate

data is obtained plotting at 3 min (Fig. 5b). It shows a good
linear relationship between absorbance and the concentration
of HBV DNA. The correlation coefficient reached 0.9822 and
their formulation is FI ¼ 0.0731 lg(CHBVDNA) + 0.0319. The
results reveal the excellent proximity of estimated values in
curve showed the correlation between the two assays is very
high, whose correlation coefficient attained 0.913 (Fig. 6).
Besides, the other negative clinical samples determined by
TaqMan assay are also negative measured in triplicate by our
colorimetric method, which failed to produce a detectable
PCR product.

10
Colormetric method
9
Taqman method
8
7
Concentration (1g)

6
9
TaqMan assay (lg of concentration)

y = 0.9091x + 0.8262
5 8 2
R = 0.913
7

4 6
5
4
3
3
2
2 1
0
1 0 1 2 3 4 5 6 7 8 9
Colormetric assay (lg of concentration)

0
1 2 3 4 5 6 7
Sample number
Fig. 6 Comparison of colorimetric and TaqMan assay. Serum samples of consecutive concentration were
measured by colorimetric and TaqMan assay. Red points represent the values of TaqMan assay and blue
points represent the values of colorimetric assay. The inset is the comparison of the values by colorimetric and
TaqMan assay. The concentration of target used for experiment was copies/ml. “Ig” represents the denary
logarithm value of the concentration of target. X-axis represents the lg value of concentration by colorimetric
assay and Y-axis represents the lg value of concentration by TaqMan assay. The correlation coefficient is
0.913
368 Li Yang et al.

4 Notes

1. The catalytic beacons probe could be extended from the 30 end

by DNA polymerase unexpectedly, which could interfere the
forming of right structure of DNAzyme and the following
colorimetric reaction as well. The probe with amino modifier
at the 30 end has been introduced into our PCR amplification
because the modification on the probe could prohibit the
undesired extension caused by the DNA polymerase.
2. Uracil N-glycosylase (UNG) treatment system is most com-
monly used method to prevent the contamination caused by
PCR amplicon. To overcome the shortage of low-efficiency
PCR amplification caused by dUTP, we optimized the PCR
conditions, containing the PCR temperature, cycles, times,
buffer constitution, the quantity of enzyme, concentration of
sensor and primer, and concentration of ion.
3. NaCl solution can help the original probe form hairpin struc-
ture, which keep solution colorless. So the concentration of the
NaCl solution should be tested to hold the background of
original probe solution.
4. To ensure the result of colorimetric reaction, Hemin, ABTS,
and H2O2 are freshly prepared.

Acknowledgments

This study was supported by the financial supports from the CAS
(Hundreds of Talents Program), National Science Foundation of
China (Grant No. 21172215 and No. 21102140), Innovation
Program of the CAS (Grant No. KSCX2-EW-J-22).

References
1. Mahoney FJ (1999) Update on diagnosis, Infect Dis Med Microbiol J Can Mal Infect
management, and prevention of hepatitis B Microbiol Med/AMMI Can. 16:65–72
virus infection. Clin Microbiol Rev 12: 5. Mullis KB, Faloona FA (1987) Specific synthe-
$351–366 sis of DNA in vitro via a polymerase-catalyzed
2. Loeb KR, Jerome KR, Goddard J et al (2000) chain reaction. Methods Enzymol
High-throughput quantitative analysis of hep- 155:335–350
atitis B virus DNA in serum using the TaqMan 6. Lee LG, Connell CR, Bloch W (1993) Allelic
fluorogenic detection system. Hepatology discrimination by nick-translation PCR with
32:626–629 fluorogenic probes. Nucleic Acids Res
3. Gish RG, Locarnini SA (2006) Chronic hepa- 21:3761–3766
titis B: current testing strategies, clinical gas- 7. Du F, Tang Z (2011) Colorimetric detec-
troenterology and hepatology. Off Clin Pract J tion of PCR product with DNAzymes
Am Gastroenterol Assoc 4:666–676 induced by 50 -nuclease activity of DNA
4. Krajden M, McNabb G, Petric M (2005) The polymerases. Chembiochem Eur J Chem
laboratory diagnosis of hepatitis B virus. Can J Biol 12:43–46
 A Novel Colorimetric PCR-Based Biosensor for Detection and Quantification. . . 369

8. Holland PM, Abramson RD, Watson R et al 14. Majhi PR, Shafer RH (2006) Characterization
(1991) Detection of specific polymerase chain of an unusual folding pattern in a catalytically
reaction product by utilizing the 50 –30 exonucle- active guanine quadruplex structure. Biopoly-
ase activity of Thermus aquaticus DNA polymer- mers 82:558–569
ase. Proc Natl Acad Sci U S A 88:7276–7280 15. Kong DM, Wu J, Wang N et al (2009) Peroxi-
9. Xiao Y, Pavlov V, Niazov T et al (2004) Catalytic dase activity-structure relationship of the inter-
beacons for the detection of DNA and telome- molecular four-stranded G-quadruplex-hemin
rase activity. J Am Chem Soc 126:7430–7431 complexes and their application in Hg2+ ion
10. Li Y, Sen D (1996) A catalytic DNA for por- detection. Talanta 80:459–465
phyrin metallation. Nat Struct Biol 3:743–747 16. Emilsson GM, Breaker RR (2002) Deoxyribo-
11. Travascio P, Li Y, Sen D (1998) DNA- zymes: new activities and new applications. Cell
enhanced peroxidase activity of a DNA-apta- Mol Life Sci 59:596–607
mer-hemin complex. Chem Biol 5:505–517 17. Baum DA, Silverman SK (2008) Deoxyribo-
12. Travascio P, Witting PK, Mauk AG et al (2001) zymes: useful DNA catalysts in vitro and
The peroxidase activity of a hemin–DNA oligo- in vivo. Cell Mol Life Sci 65:2156–2174
nucleotide complex: free radical damage to spe- 18. Navani NK, Li Y (2006) Nucleic acid aptamers
cific guanine bases of the DNA. J Am Chem and enzymes as sensors. Curr Opin Chem Biol
Soc 123:1337–1348 10:272–281
13. Deng M, Zhang D, Zhou Y et al (2008) Highly 19. Breaker RR (1997) DNA enzymes. Nat Bio-
effective colorimetric and visual detection of technol 15:427–431
nucleic acids using an asymmetrically split per- 20. Liu J, Cao Z, Lu Y (2009) Functional nucleic
oxidase DNAzyme. J Am Chem Soc acid sensors. Chem Rev 109:1948–1998
130:13095–13102
 Chapter 23

CCD Camera Detection of HIV Infection

John R. Day

Abstract
Rapid and precise quantification of the infectivity of HIV is important for molecular virologic studies,
as well as for measuring the activities of antiviral drugs and neutralizing antibodies. An indicator cell line,
a CCD camera, and image-analysis software are used to quantify HIV infectivity. The cells of the P4R5 line,
which express the receptors for HIV infection as well as β-galactosidase under the control of the HIV-1 long
terminal repeat, are infected with HIV and then incubated 2 days later with X-gal to stain the infected cells
blue. Digital images of monolayers of the infected cells are captured using a high resolution CCD video
camera and a macro video zoom lens. A software program is developed to process the images and to count
the blue-stained foci of infection. The described method allows for the rapid quantification of the infected
cells over a wide range of viral inocula with reproducibility, accuracy and at relatively low cost.

Key words Charge-coupled device (CCD), Imaging, HIV, Infectivity, Quantify, Rapid

1 Introduction

The quantification of a biological process can sometimes be cum-

bersome or labor-intensive. In the field of virology, several assays
have been developed to measure the infectivity of the human
immunodeficiency virus (HIV) during a single round of replication:
plaque assays [1], syncytium formation assays [2, 3], focal immu-
noassays [4], assays using indicator cell lines expressing β-galactosi-
dase or luciferase [5–12], and assays using fluorescently labeled
reporter viruses [13, 14]. The quantification of infection in several
of these assays requires counting the number of plaques, syncytia,
foci, or cells within an infected population of cells. High through-
put assays may utilize plate readers or flow cytometers. The use of

Portions reprinted from the Journal of Virological Methods, Vol. 137, J.R. Day, L.E. Martı́nez, R. Šášik,
D.L. Hitchin, M.E. Dueck, D.D. Richman and J.C. Guatelli, A computer-based, image-analysis method to
quantify HIV-1 infection in a single-cycle infectious center assay, pp. 125–133, Copyright 2006, with permission
from Elsevier.

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_23,
© Springer Science+Business Media LLC 2017

371
372 John R. Day

indicator cell lines, such as those containing an integrated β-galac-

tosidase gene under the control of the HIV long terminal repeat
(LTR), has enabled the straightforward determination of viral
infectivity by visualizing infected cells that are stained blue with
5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal). The
Magi-CCR5 [6] and P4R5 [7] indicator cell lines are frequently
used as targets of infection. These are HeLa-based adherent cells
that have been engineered to express CD4, the HIV receptor, as
well as CCR5, one of the two predominantly used coreceptors;
HeLa cells naturally express CXCR4, the other major coreceptor
used by HIV.
To quantify the number of infected cells by using these β-
galactosidase-based indicator cell lines, one approach is to manually
count blue cells by eye through a microscope. This laborious
method is subject to observer error and is difficult when the density
of infected cells is high, potentially limiting the dynamic range of
the assay. To facilitate the counting of infected cells with speed,
accuracy, and optimal dynamic range, a charge-coupled device
(CCD) camera is used and software is developed to analyze digital
images of infected cell monolayers [15]. Accurate and reproducible
cell counts over a wide range of viral inocula are obtained with this
method. CCD camera technology has been available for years, but
the explosion in the use of digital cameras for personal use has
accelerated technology development and facilitated the incorpora-
tion of affordable CCD detection in the laboratory.

2 Materials

2.1 Imaging 1. Five megapixel CCD color camera with real-time viewing,
Apparatus C-mount optical lens interface, and FireWire IEEE 1394 digi-
tal interface: MicroPublisher 5.0 RTV (Model # MP5.0-RTV-
CLR-10, QImaging, Burnaby, BC, Canada) (see Note 1).
2. Macro video zoom lens: Optem 18–108 mm, f/2.5, C-mount
(Qioptiq Imaging Solutions, Rochester, NY, formerly Thales
Optem, Inc.) (see Note 2).
3. Copy stand with a 1/400 -20 threaded mounting screw (model #
CS-3, Testrite Instrument Company, Inc.). Something similar
would suffice.
4. Fluorescent light box (typically used for viewing X-ray films) (see
Note 3).
5. Camera mount adapter, custom made by our university
machine shop, consisting of a tripod mounting screw (1/400 -
20 thread) and a rectangular piece of aluminum
(2.25  1  0.25 in.) with two drilled holes (see Note 4).
 CCD Camera Detection of HIV Infection 373

Use of the adapter requires that the copy stand have a mount-
ing rod that can extend away from the stand.
6. Personal computer for capturing and analyzing images: Win-
dows operating system, 1 available FireWire port (IEEE 1394)
or a FireWire expansion card.

2.2 Image 1. Image acquisition software, QCapture Pro, v.5.0.0.16, with

Acquisition and USB key (dongle) and drivers for PC (QImaging, Burnaby,
Analysis Software BC, Canada). The Micropublisher 5.0 camera is bundled with
QCapture Suite software, however, the upgrade to QCapture
Pro is purchased to increase control over the image capture
process.
2. Image analysis software: application (5MGL.exe) and parame-
ter file (5MGL.par), collectively termed the “Romanizer.” The
software is developed in-house in the Fortran 95 programming
language for the specific purpose of counting HIV-infected
cells. Use of the software for other purposes may not be suit-
able. Alternative image-analysis applications exist, both free-
ware and for purchase (see Note 5).

2.3 HIV Infectivity 1. Indicator cells: P4R5 HeLa cells (P4.R5 MAGI cells, AIDS
Assay Research and Reference Reagent Program, Division of AIDS,
NIAID, NIH contributed by Dr. Nathaniel Landau). The
P4R5 cell line contains the β-galactosidase gene under the
control of the HIV-1 promoter region (LTR). The derivation
of the parental P4 line has been reported [7]. P4R5 cells
express on their surface the HIV entry receptors, CD4,
CXCR4, and CCR5.
2. Complete media: Dulbecco’s Modified Eagle Medium
(DMEM; Gibco, Grand Island, NY) supplemented with 10%
fetal bovine serum (FBS; GemCell, Woodland, CA),
100 U/mL penicillin, 100 μg/mL streptomycin (pen/strep;
Gibco, Grand Island, NY), 2 mM L-glutamine (Gibco, Grand
Island, NY) and 1 μg/mL puromycin. Store at 4  C.
3. 48-well tissue culture plate (Cat. # 353078, BD Falcon, San
Jose, CA).
4. Phosphate buffered saline, 1 (PBS; Invitrogen, Carlsbad, CA).
5. Fix solution: 1% formaldehyde, 0.2% glutaraldehyde in PBS.
Can be made in advance and stored in the dark at 4  C for
1–2 months.
6. X-gal stain solution made fresh: For each 1.0 mL of solution,
combine 949 μL PBS, 20 μL 0.2 M potassium ferrocyanide,
20 μL 0.2 M potassium ferricyanide, 1.0 μL 2.0 M Mg2CL, and
10 μL 40 mg/mL X-gal (5-bromo-4-chloro-3-indolyl-β-D-
galactopyranoside). Stock solutions of potassium ferrocyanide,
potassium ferricyanide, Mg2Cl, and X-gal can be prepared and
374 John R. Day

stored in aliquots at 20  C. X-gal is dissolved in dimethyl

sulfoxide (DMSO) and should be stored in the dark. The X-gal
solution will turn yellow over time, but this does not affect the
activity. Discard the stock if it becomes green/brown.
7. The HIV virus is either isolated and cultured from blood, or
produced in vitro by transient transfection of proviral plasmid
DNA into cultured cells. Viruses do not need a complete HIV
genome, but they must be competent for entry into P4R5 cells
and they must be capable of expressing the HIV Tat protein.
Tat transactivates the HIV LTR promoter region and will
enable expression of β-galactosidase in the P4R5 indicator cells.

3 Methods

In order to facilitate the counting of blue-stained HIV-infected

cells, a CCD camera, macro lens, and a computer program are set
up to capture and process images of plated cells. The setup does not
require the exact equipment listed in Subheading 2. A camera, lens,
light source, computer, and a method to mount the camera are the
basic required elements. The ideas described here can be adapted to
different configurations to suit the needs of the application. The
method for analyzing the images will vary depending on the final
goal of the analysis. Although the software we developed to count
HIV-infected cells may not be suitable for other applications, alter-
native software programs may provide the necessary tools for your
particular application.

3.1 Setup of Imaging 1. Place the copy stand on a table and the light box (or light
Apparatus source of choice) on the base of the stand.
2. Assemble the camera by screwing the lens onto the camera.
3. Mount the camera to the stand using the 1/400 -20 threaded
mount. The most straightforward way to mount the camera is
illustrated in Fig. 1a. An alternative method to mount the
camera requires a custom-built adapter (see Note 6). The alter-
native method is illustrated in Fig. 1b and a photograph of
the setup is shown in Fig. 2. Screw the mount adapter onto the
copy stand, then attach the camera to the adapter using the
tripod mount screw (Fig. 3).
4. The height of the camera above the light box is adjustable and
depends on the zoom setting of the lens. In our setup, the
distance between the surface of the light box and the mounting
screw hole of the camera is 14.25 in. (36.2 cm). The tip of the
lens is 6.25 in. (15.9 cm) above the light box. Level the light
box and camera using a standard bubble level.
 CCD Camera Detection of HIV Infection 375

Fig. 1 Schematic diagrams of the CCD camera setup. (a) The simplest setup using direct attachment of the
camera to the copy stand. The FireWire ports on the camera are on the backside in the perspective shown. (b)
Use of an aluminum mount adapter and tripod screw to mount the camera rotated at 180 (see Notes 4 and 6)

Fig. 2. Photograph of the actual camera setup showing the light box, copy stand,
CCD camera, macro lens, and computer
376 John R. Day

Fig. 3. (a) Photograph of the camera mounted to the copy stand. (b) Close-up of the mounted camera with
mount adapter

5. Attach a FireWire cable to either one of the ports on the

camera. Do not connect the cable to the computer until after
loading the software and drivers.

3.2 Capture and 1. These instructions assume the use of the Micropublisher 5.0
Analysis of Images RTV camera, Optem 18–108 mm macro video zoom lens, and
QCapture Pro software. Other cameras, capture software, and
3.2.1 Capturing Images
analysis software may be used. Please follow the directions
included with your specific components.
2. Follow the setup instructions from QImaging to load the
camera drivers and QCapture Pro software. You will connect
the FireWire cable to the computer during this process.
3. To define the capture settings, launch QCapture Pro and click
the camera icon on the toolbar (or select Acquire/Video-Digital
from the menu) to open the acquisition dialog box. Use the
Basic Dialog (rather than the Advanced Dialog) by clicking the
“Basic Dialog” button. If the button in the lower left says
“Advanced” then you are already looking at the Basic Dialog.
4. Check the settings by clicking the “More >>” button at the
bottom of the window. This expands the window. Using ver-
sion 5.0.0.16 of QCapture Pro and with the specific light
source we used, the settings are as follows (Fig. 4) (see Note
7 for possible variations):
l Exp Acq: 00.300.00, Adjust Exp for Binning (checked).
l Binning: Pvw (2  2 for focusing, 4  4 for moving the
plate); Acq (1  1).
l Capture area dimensions: Left (320), Top (0), Right (2239),
Bottom (1919) [final resolution, 1920  1920 pixels].
 CCD Camera Detection of HIV Infection 377

Fig. 4. Screen shot of the capture settings within QCapture Pro capture software

l Capture depth: 24-bit color.

l Gain: 1100; Gamma: 1.90; Offset: 1120.
l White balance: R (2.327), G (1.0), B (1.375).
5. Click “Less <<” to close the settings side of the window, then
click “Preview” to see a live image.
6. Turn on the light source and place a 48-well plate (or object to
be imaged) under the camera.
7. Adjust the aperture (top ring on lens) to change the amount of
light passing through the lens. The amount of this adjustment
will depend on the desired result. We adjusted the aperture so
that a slightly gray image appeared in the preview window (see
Note 8).
8. Adjust zoom (middle ring) and focus (bottom collar) to fill the
preview window with a single well of the 48-well plate. An
378 John R. Day

initial height adjustment of the camera above the light box may
be necessary as well, to achieve the desired result. Use 2  2
Preview (Pvw) binning to focus, then switch to 4  4 for
moving the plate. As long as Acq binning remains at 1  1
the full resolution image will be acquired. The binning status of
Pvw does not affect the acquired image, but only the image
seen in the preview window. Binning at 4  4 allows for faster
on-screen response of plate movement (see Note 9).
9. Capture images by clicking the “Snap” button. Move the 48-
well plate to view the next well, click “Snap,” and repeat for the
entire plate.
10. After acquiring all the images, close the live preview window by
clicking the “Stop” button, then the “X” in the upper right
corner of the acquisition window.
11. Save the images to a new folder. A quick way to save all the
images is to select the menu Window/Close All, then click “Yes
All.” When prompted, type a name for each file. Save the files as
TIFF for analysis by the Romanizer software.

3.2.2 Analyzing Images 1. The Romanizer program will automatically count blue-stained
HIV-infected cells in 1920  1920 pixel, 24-bit color images.
Note that in addition to counting blue-stained cells, the soft-
ware may also count overly clumped cells, dust, or fibers that
appear dark in the image (see Note 10).
2. Open the file folder where the Romanizer program is stored.
Copy 5MGL.exe and 5MGL.PAR into the folder where you
saved your images. The program must reside in the same folder
as your TIFF images, and it will analyze all images within that
folder.
3. Double-click 5MGL.exe to launch the counting program. You
should see one of your images with a red circle superimposed
upon it (Fig. 5). The circle defines the region to be analyzed.
Using the computer mouse, click and drag the circle to the
center of your well to check its size. The circle should be a little
smaller than the size of the well to prevent analyzing the edge
of the well. If you need to adjust the size follow these steps:
(a) Close the program by clicking Exit on the menu.
(b) Open the file 5MGL.PAR file (see Note 11) using a simple
text editor such as the Notepad.
(c) Increase or decrease the number next to “circle radius in
tiff pixels” depending on the desired change in the circle
radius.
(d) Save your changes and close the file.
 CCD Camera Detection of HIV Infection 379

Fig. 5. Screen shot of the Romanizer program. The red circle indicates the area
that will be analyzed by the program. The position of the analysis circle was
defined for each image because of potential variations in the placement of the
well within the digital image

(e) Launch 5MGL.exe again to see the result of the size

change.
(f) Repeat steps a–e if needed to find the appropriate size.
4. Position the circle in the center of the well, then click “Next.”
The next image in the folder will appear. Repeat the procedure
for the remaining images. Due to variation in the placement of
the well by hand within the frame of the digital photograph, the
location of the analysis circle needs to be defined separately for
each image analyzed.
5. When all circles have been defined, the program will process the
images and start counting. Press “OK” when it has finished.
Figure 6a shows a digital photograph taken with the Micro-
publisher 5.0. The resulting image after being processed by the
software is shown in Fig. 6b.
6. After background subtraction and processing, the images are
placed in the file folder with an “f” preceding each filename.
The results of the analysis are exported to a Microsoft Excel file
called “results.xls.” “Simple count” is a raw count of spots
found. “Smart count” takes into consideration that some
380 John R. Day

Fig. 6. Image-analysis of HIV-infected P4R5 cells. (a) Digital photograph of a monolayer of P4R5 cells infected
with HIV and stained with X-gal within one well of a 48-well tissue culture plate using a MicroPublisher 5.0
camera and macro video zoom lens. (b) The same well after processing by the analysis software. The boxed
region is magnified in Fig. 7

spots may in fact be several adjacent cells, as it counts the

adjacent cells separately. In most cases, “Smart count” is the
preferred result to use (see Note 12).

3.3 HIV Infectivity 1. These instructions describe an assay to quantify the infectivity
Assay of live HIV virus. Sterile cell culture technique and safe labora-
tory practices for handling live HIV are beyond the scope of
this description. We performed these assays in a Biosafety Level
3 (BL3) facility approved by the Environmental Health and
Safety department of our institution.
2. The P4R5 cell line is an adherent cell line derived from HeLa
cells. The cells are engineered to express β-galactosidase after
infection with HIV. Maintain them according to standard cell
culture procedures for adherent cell lines.
3. Plate cells at a density of 2  104 cells per well in a 48-well
tissue culture plate and place them in a humidified, 37  C, 5%
CO2 incubator overnight.
4. The next day, carefully aspirate the media, avoiding contact
with the cell monolayer. Infect the cells in duplicate wells
with 100 μL of serial dilutions of virus in complete media.
After a 2 h incubation at 37  C, add 400 μL of complete
media. Incubate for 2 days at 37  C.
5. Carefully aspirate the media from the wells and add 1 mL of fix
solution to each well. Incubate for 5 min at room temperature.
The fix solution renders the virus non-infectious.
6. Aspirate the fix solution and wash the wells twice with PBS.
 CCD Camera Detection of HIV Infection 381

7. Add 250 μL of X-gal stain solution and incubate overnight at

37  C in a non-CO2 incubator.
8. Aspirate the stain solution and wash once with PBS. Wash again
with deionized water. Aspirate the water from the well.
After wiping down the outside of the plate with 70% ethanol,
the plate can be removed from the BL3 and handled at the lab
bench.
9. Invert the plate over a paper towel and allow the wells to dry.
10. The plate is now ready for imaging and analysis. Figure 7
illustrates the final result in a magnified region of Fig. 6. Syn-
cytia (fused neighboring HIV-infected cells) of various sizes
and single cells are accurately distinguished within the mono-
layer of P4R5 cells. The magnified image illustrates the amount
of detail seen in the original digital photograph (Fig. 7a), and
for comparison, the same region is photographed using the
20 objective of a Leitz Labovert microscope (Fig. 7b). In the
processed image, every object that is counted is marked with a

Fig. 7 Magnified view of HIV-infected P4R5 cells. Magnified images of the boxed region of Fig. 6 depicting (a)
the original digital photograph, (b) the same region viewed through the 20 objective of a microscope, (c) the
processed image including small red dots as indicators of the objects counted and (d) outlines of infected cells
surrounding red dot markers. The arrows indicate cell groups counted singly using Simple count and as two
cells using Smart count
382 John R. Day

small red dot to allow the user to examine more closely how the
software is performing and exactly what objects in the image
are counted (Fig. 7c). To illustrate this more clearly, the out-
lines of infected foci are shown encircling each red dot
(Fig. 7d). Two or more red dots within a group of cells that
come in contact in the image are counted as one infection,
using “Simple” count and as multiple foci using “Smart” count
(see Note 12).

4 Notes

1. With today’s advances in digital camera technology, there are

many types of affordable options. The first parameter to consider
when choosing a camera is the spatial resolution that the camera
provides. The resolution corresponds to the ability to discrimi-
nate between points that are closely spaced in the image [16].
We initially used an analog CCD camera with a standard VGA
resolution of 640  480 pixels (approximately 0.3 megapixels).
The camera is able to resolve the larger blue-stained foci, but
smaller ones remained undetected. Furthermore, the analysis
software is not able to distinguish between small foci in close
proximity. The resolution of the camera chosen depends largely
on the object to be imaged. Smaller objects require larger arrays
of pixels (higher resolution) so that they can be discriminated
from one another. We upgraded to a 5 megapixel digital CCD
camera (2560  1920 pixels) when they became more readily
available and reasonably priced. In addition to resolution, a
second consideration in choosing a camera is the sensitivity
requirement. The sensitivity of a camera refers to the lowest
signal that can be detected [17]. If the object to be imaged is
fluorescent or chemiluminescent, a camera with increased sen-
sitivity may be required. For example, QImaging makes a
camera that will detect a single photon of light (Rolera-MGi);
however the resolution of the camera is only 512  512. An
increase in camera sensitivity is often accomplished at the
expense of resolution. Higher sensitivity applications can also
be subjected to an increase in noise; therefore, the signal-to-
noise ratio becomes important, as well. Because heat generated
in a camera can cause an increase in noise (dark noise), cooled
cameras, such as the cooled version of the Micropublisher
(MP5.0-RTV-CLR-10-C), can maintain a lower temperature
and reduce the amount of noise in the system [18]. Choosing
the right camera for your application requires careful consider-
ation of these factors. Consult camera manufacturers for more
information and advice.
 CCD Camera Detection of HIV Infection 383

2. Like the choice of camera, the choice of lens depends on the

specific object to be imaged. Macro lenses are designed for
close-up photography and are capable of focusing on objects
within a short distance. Using a lens with zoom capability
enables greater control of the size of the object in the frame.
The Optem lens we used is recommended to us and it meets
our needs well.
3. The choice of a light source also depends on what is to be
imaged. Illumination from above or from below depends on
the transparency of the subject. For our purposes, a simple
fluorescent light box provided ample light to illuminate the
wells of a 48-well plate from below. The light source does not
necessarily need to emit a specific intensity of light because
there are camera (aperture) and software (exposure) settings
that can be adjusted to change the brightness of the resulting
image. The color spectra emitted by the light source is also not
a large factor because white balance settings can be adjusted
within the capture software. A commonly used light box for
viewing X-ray films is more than adequate.
4. A tripod mounting screw and a rectangular piece of aluminum
are the materials used for the mount adapter (Fig. 8). The
dimensions of the aluminum rectangle are 2.25  1  0.25 in.
(5.7  2.5  0.6 cm). Two holes are made in the aluminum,
one large enough for the tripod mounting screw to pass
through easily, and the other threaded so that it would
screw onto the copy stand mount (1/400 -20 thread) (Fig. 8).
The holes are 1.25 in. (3.2 cm) apart. The adapter works only
if the copy stand mounting rod can extend far enough and
away from the stand to leave at least 3.2 in. (8.1 cm) of space

Fig. 8 (a) Photograph of the mount adapter with tripod mounting screw. (b) Photograph of mount adapter
holes. One hole was drilled wide enough for the tripod mounting screw to pass through easily. The second hole
was drilled with a 1/4”-20 thread for mounting onto the copy stand
384 John R. Day

for the camera body (Fig. 3). This custom-built adapter

allowed the camera to be rotated 180 and mounted on the
copy stand.
5. The Romanizer program is freely available for use in the type of
analysis described here. Its use for counting other kinds of
objects of different sizes may not produce acceptable results.
Processed images should be scrutinized carefully to determine
the program’s performance. You may download a copy of the
software and setup instructions here: http://cfar.ucsd.edu/
links/downloads/software. QCapture Pro performs some
basic analyses and may be one alternative. Another freely avail-
able program that performs image analysis is the NIH Image/
ImageJ program developed by the National Institutes of
Health [19].
6. When the camera is mounted directly on the camera stand, the
image on-screen is inverted, i.e., moving a microplate to the
left caused the on-screen live preview image to move to the
right, and moving the plate up caused the preview to move
down. One can become accustomed to this type of motion, but
it is not ideal. The capture software did not have the option to
reverse the orientation of the image, but future versions may.
Our solution to the problem is to rotate the camera 180 .
Unfortunately there is no threaded mounting hole on the
opposite side of the camera. Instead, we designed a mount
adapter and asked our university machine shop to construct it
(Fig. 8). This custom-built adapter allowed the camera to be
rotated and mounted on the copy stand.
7. For consistency between experiments, the same settings are used
each time. Some changes have been made in newer versions of
QCapture Pro such that the parameters may not be exactly the
same as shown here. We used the default settings for gain,
gamma and offset, and since version 5.0.0.16 they have changed
the value ranges for gain and offset. The default settings are
recommended; alternately, consult QImaging for advice. Differ-
ent light boxes will most likely require different settings for red,
green, and blue than those shown here. White balance should be
set using the Auto White Balance feature of the software. The
exposure of 00.300.000 (0.3 s) can be adjusted to change the
brightness of the image. Differences in light box intensity may
require a change to the exposure time and/or the aperture (see
Note 8). Finally, setting the capture dimensions to a square area
(1920  1920 pixels) rather than using the full rectangular
capture area (2560  1920) reduced TIFF file size and elimi-
nated unnecessary image data on either side of the circular well.
8. One way to adjust the brightness of the image is to adjust the
aperture. Turning the aperture ring on the lens changes the
 CCD Camera Detection of HIV Infection 385

diameter of a circular opening. A larger aperture will allow

more light to reach the CCD array. For consistency in plate
analysis, the same settings should be used for all experiments,
including lens aperture. In our experiments, the aperture of the
lens is empirically adjusted so that the images are not too bright
to obscure faint blue cells, and not too dark to increase back-
ground counts of uninfected wells. Once an aperture setting is
selected, the aperture ring is marked so that the same setting is
used consistently.
9. Binning is a function in which the charges from adjacent pixels
on the CCD array are combined. This is accomplished at the
expense of spatial resolution [17]. For example, 2  2 binning
combines the charge from a 2  2 grid of pixels (4 pixels total)
into one signal, effectively reducing the resolution two times in
both the x and y axes. Similarly, 4  4 binning combines the
charge from 16 adjacent pixels into one signal. The greatest
benefit to binning is an increase in signal output and dynamic
range [17]. The increase in signal-to-noise ratio allows for
detection of lower light levels. In QCapture Pro we checked
the box “Adjust Exp for Binning.” When binning is increased
from 1  1 (no binning) to 2  2 or 4  4, the exposure time is
reduced to compensate for the increase in signal. Another
benefit of binning is that the resolution of the image is reduced
and it is more easily processed by the computer, effectively
speeding up the on-screen response time when moving the
object. We used binning solely for that purpose. Increased
signal is not necessary because there is ample light emitted
from the fluorescent light box.
10. An evenly plated monolayer of target cells is critical to prevent-
ing cells from overgrowing and forming dark clumps that may
be detected by the software. Using clean reagents that will not
leave dust or fibers in the wells is also important.
11. The parameter file, 5MGL.PAR, consists of the following
default settings:
l Circle radius in tiff pixels: 860.
l Window size in screen pixels: 640.
l Noise cutoff: 45.
l Spot area cutoff in pixels: 10.
l Minimum distance of spot maxima: 3.
l Size of the structuring element: 10.
Other than the circle radius, these settings should not be
changed. The window size may need adjustment depending on
the screen resolution of the computer. If you feel you need to adjust
some of the other parameters, do so in a careful manner, comparing
the results before and after changes to assess the modified perfor-
mance of the program.
386 John R. Day

12. To count HIV infection, blue-stained cells are identified as

dark spots in all three channels of the RGB spectrum even
though they appeared predominantly blue to the naked eye.
5MGL.exe therefore analyzed a gray-scale image created by
adding all three channels of the original 24-bit color TIFF
image. To remove the impulse noise from the image, gray
scale thinning and thickening morphological filters are
applied with a point structuring element [20]. Because the
illumination of the image is not perfectly uniform, straight-
forward thresholding of the image cannnot be used to count
the spots. Instead, using a square structuring element whose
size is larger than the typical spot size, the software calculated
the morphological opening of the image and then subtracted
it from the image. The resulting image has a uniform, fluctu-
ating background. Signal and noise are then separated by
thresholding. The remaining islands (defined as unions of all
side-to-side adjacent pixels with non-zero intensity) repre-
sented the actual blue-stained cells, plus a few small ghost
spots that emerge by random agglomeration of noise after
thresholding. The latter are typically much smaller than the
islands representing actual blue cells and are removed by
applying an area opening to the image (ibid). The remaining
islands are considered as signals and are counted; this analysis
is termed “Simple count.” However, sometimes a single
island represented several adjacent cells that should have
been counted separately; this is especially the case under con-
ditions of high inoculum. Because a stained cell is darker on
the inside than near the edge, adjacent cells have separate dark
intensity maxima, which could be counted individually. This
analysis is termed “Smart count.” Using Smart count, in
terms of the intensity landscape, the hills on the islands rather
than the islands themselves are counted. Both Simple and
Smart counts are reported in the exported results file. For
user verification of software performance, a processed version
of each image containing a small red dot on every object
counted is saved into the file folder containing the original
images.

Acknowledgments

We thank the coauthors of the original publication, Laura Martı́-

nez, Roman Šášik, Douglas Hitchin, Megan Dueck, Douglas Rich-
man, and John Guatelli for their contributions to this work. The
system we have developed was conceived with ideas from Chris
Aiken. We are also grateful to Prentice Higley, Sherry Rostami
and Nanette Van Damme for assistance during setup and evalua-
tion. This work was supported by grants AI27670, AI043638,
 CCD Camera Detection of HIV Infection 387

AI038201, the UCSD Center for AIDS Research (AI 36214),

AI29164, AI047745, from the National Institutes of Health and
the Research Center for AIDS and HIV Infection of the San Diego
Veterans Affairs Healthcare System.

References
1. Harada S, Koyanagi Y, Yamamoto N (1985) of primary strains of human and simian immu-
Infection of HTLV-III/LAV in HTLV-I-car- nodeficiency viruses. Virology 233:193–198
rying cells MT-2 and MT-4 and application in a 11. Wei X, Decker JM, Wang S, Hui H, Kappes JC,
plaque assay. Science 229:563–566 Wu X, Salazar-Gonzalez JF, Salazar MG, Kilby
2. McKeating JA, McKnight A, McIntosh K, JM, Saag MS, Komarova NL, Nowak MA,
Clapham PR, Mulder C, Weiss RA (1989) Hahn BH, Kwong PD, Shaw GM (2003) Anti-
Evaluation of human and simian immunodefi- body neutralization and escape by HIV-1.
ciency virus plaque and neutralization assays. J Nature 422:307–312
Gen Virol 70:3327–3333 12. Montefiori DC (2009) Measuring HIV neu-
3. Nara PL, Fischinger PJ (1988) Quantitative tralization in a luciferase reporter gene assay.
infectivity assay for HIV-1 and -2. Nature Methods Mol Biol 485:395–405
332:469–470 13. Tilton CA, Tabler CO, Lucera MB, Marek SL,
4. Chesebro B, Wehrly K (1988) Development of Haqqani AA, Tilton JC (2014) A combination
a sensitive quantitative focal assay for human HIV reporter virus system for measuring post-
immunodeficiency virus infectivity. J Virol entry event efficiency and viral outcome in pri-
62:3779–3788 mary CD4+ T cell subsets. J Virol Methods
5. Akrigg A, Wilkinson GW, Angliss S, Greenaway 195:164–169
PJ (1991) HIV-1 indicator cell lines. AIDS 14. Xu R, El-Hage N, Dever SM (2015)
5:153–158 Fluorescently-labeled RNA packaging into
6. Chackerian B, Long EM, Luciw PA, Over- HIV-1 particles: direct examination of infectiv-
baugh J (1997) Human immunodeficiency ity across central nervous system cell types.
virus type 1 coreceptors participate in posten- J Virol Methods 224:20–29
try stages in the virus replication cycle and 15. Day JR, Martı́nez LE, Šášik R, Hitchin DL,
function in simian immunodeficiency virus Dueck ME, Richman DD, Guatelli JC (2006)
infection. J Virol 71:3932–3939 A computer-based, image-analysis method to
7. Charneau P, Mirambeau G, Roux P, Paulous S, quantify HIV-1 infection in a single-cycle infec-
Buc H, Clavel F (1994) HIV-1 reverse tran- tious center assay. J Virol Methods 137:125–133
scription. A termination step at the center of 16. Beyon JDE, Lamb DR (eds) (1980) Charge-
the genome. J Mol Biol 241:651–662 coupled devices and their applications.
8. Deng H, Liu R, Ellmeier W, Choe S, Unutmaz McGraw-Hill, UK
D, Burkhart M, Di Marzio P, Marmon S, Sut- 17. Holst GC (1998) CCD arrays, cameras, and
ton RE, Hill CM, Davis CB, Peiper SC, Schall displays, 2nd edn. JCD Publishing/SPIE Opti-
TJ, Littman DR, Landau NR (1996) Identifi- cal Engineering Press, Winter Park, FL/Bel-
cation of a major co-receptor for primary iso- lingham, WA
lates of HIV-1. Nature 381:661–666 18. Howes MJ, Morgan DV (eds) (1979) Charge-
9. Kimpton J, Emerman M (1992) Detection of couple devices and systems. John Wiley &
replication-competent and pseudo-typed Sons, New York, NY
human immunodeficiency virus with a sensitive 19. NIH. (2005) NIH Image/ImageJ, a public
cell line on the basis of activation of an domain program developed at the US.
integrated beta-galactosidase gene. J Virol National Institutes of Health. Available at
66:2232–2239 http://rsb.info.nih.gov/nih-image/
10. Vodicka MA, Goh WC, Wu LI, Rogel ME, 20. Soille P (2003) Morphological image analysis:
Bartz SR, Schweickart VL, Raport CJ, Emer- principles and applications. Springer-Verlag,
man M (1997) Indicator cell lines for detection New York, NY
 Chapter 24

“Dipstick” Colorimetric Detection of Metal Ions Based

on Immobilization of DNAzyme and Gold Nanoparticles
onto a Lateral Flow Device
Debapriya Mazumdar, Tian Lan, and Yi Lu

Abstract
Real-time, on-site detection and quantification of different trace analytes is a challenge that requires both
searching a general class of molecules to recognize a broad range of contaminants and translating this
recognition to easily detectable signals. Functional nucleic acids, which include DNAzymes (DNA with
catalytic activity) and aptamers (nucleic acids that bind an analyte), are ideal candidates for the target
recognition. These nucleic acids can be selected by a combinatorial biology method called in vitro selection
to interact with a particular analyte with high specificity and sensitivity. Furthermore, they can be
incorporated into sensors by attaching signaling molecules. Due to the high extinction coefficients and
distance-dependent optical properties, metallic nanoparticles such as the commonly used gold nanoparti-
cles have been shown to be very attractive in converting analyte-specific functional DNA into colorimetric
sensors. DNAzyme directed assembly of gold nanoparticles has been used to make colorimetric sensors for
metal ions such as lead, uranium, and copper. To make the operation even easier and less vulnerable to
operator’s errors, dipstick tests have been constructed. Here, we describe protocols for the preparation of
DNAzyme-linked gold nanoparticles (AuNP) that are then immobilized on to lateral flow devices to make
easy-to-use dipstick tests for metal ions.

Key words Dipstick, Nanoparticle, Sensor, Colorimetric, Lateral flow, DNAzyme, Metal ions

1 Introduction

1.1 DNAzymes or Nucleic acids have recently emerged as an important platform for
Deoxyribozymes as selective molecular recognition, one major requirement for sensors.
Sensors for Metal Ions Long considered as passive molecules for the storage of genetic
information, RNA and DNA molecules with catalytic function
similar to protein enzymes were discovered in the early 1980s and
1990s, respectively [1–3]. These enzymes are called ribozymes
(catalytic RNA) and deoxyribozymes or DNAzymes (catalytic
DNA). The nucleic acid enzymes usually require a metal ion co-
factor to perform their catalytic function and can be tailored to be
specific for a particular metal ion. In addition, nucleic acids that

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_24,
© Springer Science+Business Media LLC 2017

389
390 Debapriya Mazumdar et al.

bind to a target material with high specificity and affinity (analo-

gous to protein antibodies) have also been obtained, and these are
called aptamers [4–7]. Nucleic acid enzymes and aptamers have also
been fused to form a new class of allosteric enzymes called apta-
zymes [8, 9]. Collectively, the nucleic acid enzymes, aptamers and
aptazymes are termed functional nucleic acids. In this chapter, our
focus is on DNAzymes that catalyze cleavage reactions in the pres-
ence of a metal ion.
Although a number of naturally occurring ribozymes have been
discovered in nature [1, 2], DNAzymes have not been found in
nature. DNAzymes and aptamers are obtained by a combinatorial
biology technique called in vitro selection (also called systematic
evolution of ligands by exponential enrichment (SELEX for apta-
mer selections)) in test tubes [4–6, 10]. Figure 1a is a schematic
representation of the selection process. A large pool of nucleic acid
sequences represented by the colored objects (
1014–1016 differ-
ent sequences) is incubated with a target of interest in each round
of selection. The ‘winner sequences’ which bind to the target
analyte (in the case of aptamer selection) or catalyze a reaction in
the presence of target (in the case of nucleic acid enzyme selection)
are separated by various techniques such as gel electrophoresis,
column separation and capillary electrophoresis. These “winners”
are amplified using polymerase chain reaction to be used for the
next round of selection. During each round of selection, the strin-
gency can be increased by decreasing the interaction time between
the target and the nucleic acid or by decreasing the concentration of
the target. Iterative rounds of selection are continued until the pool
is sufficiently enriched with sequences of desired sensitivity and
specificity (represented by blue cubes in Fig. 1a). A review of
in vitro selection of DNAzymes, including a detailed protocol has
been published in a previous Methods in Molecular Biology series
book [11]. This technique can be used to obtain nucleic acid
sequences that recognize a target contaminant with sensitivity and
specificity. In vitro selection has been used to obtain a number of
metal specific-DNAzymes dependent on Pb2+ [12, 13], Zn2+ [14],
Co2+ [15], Cu2+ [16], UO22+ [17], Hg2+ [18], and As5+ [18],
some of which have been converted into fluorescent and colorimet-
ric sensors as described in the following sections. A number of these
metal ions, notably Pb2+, Hg2+, and As5+, are heavy metals that are
particularly toxic; UO22+ is a radionuclide. The maximum level of
these metal ions in drinking water is strictly regulated by the US
Environmental Protection Agency (EPA). Only a few sensors avail-
able today can detect metal ions below EPA regulatory levels, and
therefore the selectivity of sensors should be improved in order to
be practically applicable. For this purpose, the utility of DNAzymes
as toxic metal ion sensors is of great importance. The predicted
secondary structures of a few DNAzymes are shown in Fig. 1b. The
strands in green represent the enzyme and the strands in black are
 Dipstick Metal ion Detection 391

Fig. 1 In vitro selection of functional nucleic acids. (a) Schematic depiction of

general in vitro selection scheme for obtaining functional nucleic acid that
interacts with a specific target (contaminant). (b) Predicted secondary structures
of metal specific RNA cleaving DNAzymes obtained by in vitro selection. The
black strands represent the substrate and the green strands are the enzyme.
The cleavage site is depicted by the black arrow and there is a single RNA base
at the cleavage site of the Pb2+, Hg2+ and UO22+ DNAzyme denoted by rA.
(c) General schematic representation of the catalytic activity of metal ion
dependent RNA cleaving DNAzyme. Strand in black represents the substrate
strand with an embedded riboadenosine base (rA in red); strand in green
represents the enzyme strand which is capable of recognizing a specific metal
ion and catalyze the cleavage of a phosphodiester bond at the rA site
392 Debapriya Mazumdar et al.

the nucleic acid substrate. All these DNAzymes (except the Cu2+
dependent DNAzyme) are RNA cleaving enzymes that catalyze the
cleavage of a single phosphodiester bond within the rA (RNA)
embedded in the DNA substrate. They share a similar secondary
structure containing a substrate strand and an enzyme strand hybri-
dizing to each other through the two arms on each side of the
cleavage site (Fig. 1c). It has been determined that, unlike the DNA
sequences in the single stranded loop, the sequence identity of two
arms do not contribute significantly to either metal binding or
activity, as long as they can form double stranded DNA to provide
stability [12]. Therefore the sequence length and GC content can
be designed so that, in the absence of the target metal ion, the
enzyme strand and substrate strand have a melting temperature
above the ambient temperature, making them hybridize to each
other. In the presence of target metal ion, however, the cleavage of
the substrate strand changed the melting temperature between the
two halves of the cleaved products and enzyme strand to below the
ambient temperature so that they can be released from the enzyme
strand. Some of the fastest DNAzymes have a catalytic efficiency
(kcat/Km) of 109 M1 min1 [19], rivaling that of protein enzymes
and thus they are ideal for fast sensing.
As a major component for sensors, nucleic acids possess many
advantages. First, DNA/RNA targeting essentially any molecule of
choice can be obtained through combinatorial selections [4–6, 10],
which provide a unique opportunity to construct a general sensing
platform for a broad range of analytes. Second, nucleic acids, par-
ticularly DNA, are very stable and can be denatured and renatured
many times without losing their binding abilities, allowing a long
shelf life. Third, nucleic acids have predictable base pairing interac-
tions, which have been proven to be very useful for rational sensor
design that are difficult to attain when developing protein or
organic molecule-based sensors. Finally, a broad range of DNA
modifications can be chemically synthesized with relatively low
cost, allowing convenient and precise conjugation. These proper-
ties make DNA and RNA ideal candidates to create sensors.
Since natural nucleic acids do not possess functional groups
that can generate absorption in the visible region, external signaling
labels need to be applied to convert them into sensors. To achieve
this goal, many inorganic nanoparticles have been employed. The
next few sections describe the properties of these nanoparticles and
preparation of a DNAzyme based lateral flow device using gold
nanoparticles.

1.2 Physical Depending on the composition, size, and shape of inorganic nano-
Properties of particles, a wide range of properties can be obtained. Inorganic
Nanoparticles metallic [20, 21], semiconductor [22], and magnetic nanoparticles
[23] can all be used to make functional sensors with different
detection modes. In this protocol, metallic nanoparticles are used.
 Dipstick Metal ion Detection 393

Dispersed AuNPs (diameter from several nanometers to about

100 nm) display red colors resulting from their surface plasmons.
In addition to such distance-dependent optical properties, AuNPs
also possess very high extinction coefficients, which are usually 3–5
orders of magnitude higher than the brightest organic chromo-
phores. Thiol-modified DNA can be attached to the surface of
AuNPs and these functionalized nanoparticles can be crosslinked
by complementary DNA to form blue-colored aggregates [24].
This process has been applied by Mirkin and coworkers to design
highly sensitive and selective colorimetric sensors for DNA detec-
tion [25]. By using functional DNA (aptamers [20, 26], catalytic
DNA or DNAzymes [27, 28], and aptazymes [29]) that can recog-
nize a diverse range of analytes, AuNP-based colorimetric detection
method can be applied to detect many analytes beyond DNA.

1.3 Nanoparticle Many metal specific DNAzymes have been successfully converted
Conjugated DNAzymes into fluorescent sensors using a catalytic beacon technology [30]
for Sensing where the DNA enzyme and substrate are tagged with a combina-
Application tion of organic fluorophore and quencher. Sensors for metal ions
such as Pb2+ [13], UO22+ [17], Hg2+ [31], Cu2+ [32] are reported
in literature and these have been commercialized by ANDalyze Inc.
using a platform that consists of consumable fluorescent sensor
cartridges and a handheld fluorimeter.
Although fluorescent sensors are very sensitive and provide a
method of precise quantitative detection, colorimetric sensors have
been developed to make the detection possible by the naked eye.
In 1996, Mirkin and coworkers utilized the DNA induced assembly
of AuNPs to make a colorimetric sensor for nucleic acids [25].
Lu and coworkers expanded the scope of sensing to analytes
beyond nucleic acids, by combining AuNPs with DNAzymes [27,
28, 33, 34]. The sensing method is depicted using the Pb2+ depen-
dent 17E DNAzyme as a representative example (Fig. 2). AuNPs
functionalized with short thiol modified DNA are assembled on the
arms of the extended substrate which is in turn hybridized to the
enzyme. Since each AuNP is functionalized with many DNA
strands, blue aggregates are formed. In the presence of Pb2+, the
enzyme catalyzed cleavage of the substrate will disassemble the
aggregate producing red color. The color can be spotted on a
TLC plate and one such representative test is shown in the inset
of Fig. 2. Red color of increasing intensity is seen with Pb2+,
whereas the sensors containing other metals are blue. The reaction
is fast with color change occurring within 10 min under optimized
conditions and the detection limit is
100 nM.
Although not described in detail in this chapter, AuNP-based
aptamer sensors for various analytes, including several small organic
molecules are reported in literature [20, 33, 35–37]. The AuNPs
can also be replaced by other metallic nanoparticles, so that differ-
ent color changes can be achieved in the presence of different
394 Debapriya Mazumdar et al.

Fig. 2 DNAzyme based colorimetric sensor for Pb2+ detection. In the absence of Pb2+, the oligonucleotide
functionalized AuNPs are assembled on the substrate to form blue aggregates. When Pb2+ is present, the
substrate is cleaved and the aggregate is disassembled to yield red colored dispersed AuNPs. Inset: Picture of
sensor incubated with increasing concentrations of Pb2+ and with other metal ions (that are all 5 μM in
concentration)

analytes. In addition to metallic nanoparticles, other nanomaterials

including magnetic nanoparticles, quantum dots, nanotubes, or
polymer beads can also be used to suit various applications.

1.4 Dipstick Tests While these colorimetric biosensors are an important step toward
Based on real-time sensing as the signal is detectable by the naked eye,
Immobilization of without the need for expensive instrumentation, they still require
DNAzyme-Gold laboratory type operations, such as precise transfer and mixing of
Nanoparticles onto a multiple solutions. In addition, although the sensitivity is high
Lateral Flow Device when the absorbance is recorded using a UV–Vis spectrophotome-
ter, it is often difficult to distinguish the red color of dispersed
nanoparticles against a blue background from the aggregates, par-
ticularly at low metal-ion concentrations. Furthermore, AuNPs are
not very stable in solution state and are vulnerable to aggregation
under a variety of conditions thereby making it difficult to store the
sensors for a long period of time.
Lateral flow devices are an ideal platform for making dipstick
type tests to further improve the performance of DNAzyme-AuNP
colorimetric sensors. In addition to eliminating precise solution
transfer and allowing separation of AuNPs to make it easier to
distinguish colors, the reagents can be prepared in a dry or nearly
dry state, making the device stable at ambient conditions for a long
period of time. The home pregnancy test is the most commonly
used lateral flow devices and several other antibody-based tests are
also well-known; however, DNA-based dipstick tests are not com-
mon. Glynou et al. reported a lateral flow device for the detection
of DNA [38]. To expand on the range of analytes detected, we
previously reported dipstick tests for the detection of adenosine and
cocaine using AuNPs conjugated to aptamers [39]. The protcol to
 Dipstick Metal ion Detection 395

make these aptamer-AuNP based dipsticks has been reported pre-

viously [40]. A paper-based bioassay using aptamers and the pro-
tein enzyme DNase I, which also involved the disassembly of
nanoparticle aggregates that were dried onto paper substrates was
also reported by Yingfu Li and coworkers [41]. Even though our
previously reported methodology can be applied to almost any
target for which aptamers can be obtained, because aptamers with
high affinity for metal ions have been difficult to select, this meth-
odology has not been applied to dipstick tests for metal ions. To
extend the applicability of this test for metal ions, metal-specific
DNAzymes are an excellent choice. However, it is not trivial to
adopt the aptamer-based dipstick methodology by simply replacing
aptamers with DNAzymes because DNAzymes undergo not only
binding as aptamers do but also catalytic activity and product
release, making the design more complicated. The Lu group has
previously reported a method to convert DNAzyme-AuNPs into
dipstick tests for metal ions, specifically for Pb2+ [42]. We showed
that the cross-link based method used in previous aptamer-AuNP
systems cannot be used for DNAzyme-AuNP system. Instead, we
succeeded in developing a non-cross-linked DNAzyme-AuNP sys-
tem for detection of Pb2+. These dipstick tests are ideal for detec-
tion of lead in household paints, in accordance with the EPA
defined threshold of 1 mg/cm2 Pb2+ for paint to be classified as
lead-based.
The 8–17 DNAzyme is used to construct the dipstick tests for
Pb2+ because of its very high activity as shown by a fast cleavage rate
(estimated kobs
50 min1 at pH 7.0) [12]. Figure 3a shows its
reaction scheme. In the presence of Pb2+, the enzyme strand (called
17E) catalyzes the cleavage of the chimeric substrate (called 17S) at
the single ribo-adenosine (rA) base. Unlike aptamer-based colori-
metric tests, the presence of target does not cause immediate disas-
sembly of the aggregates in the colorimetric tests. Disassembly in
the case of DNAzymes has been shown to require heating [27] or
the use of invasive DNA strands [33] to release the cleaved product
trapped in the nanoparticle aggregates. Either of these methods
leads to added complexity and are not highly feasible for lateral flow
devices. Therefore, we decided to use an alternate approach that
does not involve formation of nanoparticle aggregates and thus the
detection is not based on a change in the optical properties of
AuNPs.
In our scheme, the 8–17 DNAzyme is modified to form the
construct shown in Fig. 3b. The 17S substrate is modified on the 30
end to have a biotin moiety. On the 50 side, 18 additional bases
(AAG)6 are added to act as a free site for DNA hybridization
required for the capture of cleaved product. In addition, the 50
end is functionalized with a thiol group in order to conjugate the
substrate to 13 nm AuNPs. When the original 8–17 DNAzyme
construct with 9 symmetric base pairs on either substrate binding
396 Debapriya Mazumdar et al.

Fig. 3 (a) 8–17 DNAzyme reaction: In the presence of Pb2+, the 17E enzyme catalyzes the cleavage of the
chimeric substrate, 17S at the single ribo-linkage (rA). (b) Modified 8–17 construct conjugated to AuNPs
(called Enz-SubAuNP) used for the dipstick tests. B inside the purple box denotes biotin. (c) Assembled lateral
flow device. (d) Negative control: In the absence of Pb2+, AuNP-uncleaved substrate is captured at the control
zone via streptavidin-biotin interaction, producing a single red line. (e) Positive test: Substrate is cleaved in the
presence of Pb2+ and the AuNP-cleaved product migrates beyond the control zone to be captured at the test
zone by hybridization to complementary DNA. Two lines are produced

arm is tested, this construct failed to give a positive signal as

designed. To facilitate the release of the cleavage product without
compromising the binding of the substrate, three bases are deleted
from the 50 end of the enzyme and two bases are added to the 30 end
of the 17E enzyme. The shortened arm facilitates the release of the
cleaved product on the 50 end after cleavage reaction, while the
overall stability of the construct before cleavage is maintained due
to the extra base-pairing on the opposite arms. The enzyme–sub-
strate complex is prepared by hybridizing the enzyme to the sub-
strate conjugated to AuNPs and this was referred to as Enz-
SubAuNP.
 Dipstick Metal ion Detection 397

The lateral flow device is constructed using a Millipore Assem-

bly kit by placing four overlapping pads on a backing. Streptavidin
and capture DNA are applied on the capture zone and test zone of
the membrane, respectively and the complex, Enz-SubAuNP is
spotted on the conjugation pad and allowed to dry (Fig. 3c). We
hypothized that if the dipstick is dipped in a flow buffer the Enz-
SubAuNP complex will be rehydrated. In the absence of Pb2+, the
substrate would remain uncleaved and Enz-SubAuNP would
migrate on the membrane till it reached the control zone, where
the biotin-containing Enz-SubAuNP could be captured by strepta-
vidin, thus producing a single red line at the control zone (Fig. 3d).
In the presence of Pb2+, the substrate would be cleaved, and the
cleaved product would migrate past the control zone to be captured
at the test zone by a 27 base long DNA sequence complementary to
the cleaved substrate piece (called capture DNA), producing a red
line at the test zone. Since the cleavage reaction may not be 100%
complete, the positive tests would normally result in two red lines,
one being the cleaved product and the other being the uncleaved
enzyme–substrate (Fig. 3e).

2 Materials

1. Oligonucleotides: All oligonucleotides are purchased from a

commercial source (e.g., Integrated DNA Technologies Inc.,
(Coralville, IA)). The oligonucleotides are purified by HPLC
or polyacrylamide gel electrophoresis (PAGE) to ensure high
purity.
2. Other chemicals: Hydrogen tetrachloroaurate (III) (HAuCl4),
trisodium citrate dihydrate, tris-(2-carboxyethyl)phosphine
hydrochloride (TCEP), tris-(hydroxymethyl) aminomethane
(Tris), lead chloride, sodium hydroxide, concentrated hydro-
chloric acid (HCl), and concentrated nitric acid (HNO3) are
purchased from Sigma-Aldrich (St. Louis, MO). Streptavidin is
purchased from Promega (Madison, WI).
3. Buffers: Tris–HCl buffers are used in the experiments. 500 mM
of Tris–HCl buffer stock at pH 8.0 is prepared by adding
hydrochloric to 500 mM Tris solution until the desired pH
value is achieved. The buffer stock solutions are incubated with
metal chelating resin (iminodiacetic acid, sodium form,
Aldrich) overnight to eliminate trace divalent metal ions.
Finally, the buffer stock solutions are filtered through 0.2 μm
syringe filters (Nalgene, Rochester, NY) and stored in a 20  C
freezer.
4. Equipment: A two-neck flask (100 ml), a condenser, and a
stopper; hot plate with magnetic stirring and a stir bar; dispos-
able scintillation vials (20 ml), polypropylene microcentrifuge
398 Debapriya Mazumdar et al.

tubes (1.7 ml; catalog no. MCT-175-C; Axygen Scientific),

temperature-controlled UV–visible spectrophotometer
(Hewlett-Packard 8453), quartz UV–visible cell (Hellma),
0.2-μm syringe filter (Nalgene), and Sep-Pak desalting column
(Waters).
5. Lateral flow devices: Lateral flow devices are constructed from
membranes and pads obtained from Millipore Corporation
(Bedford, MA). Hi-Flow Plus Cellulose Ester Membrane with
a nominal capillary flow time of 240 s/4 cm and a nominal
membrane thickness of 135 μm direct cast onto 2 mil polyester
backing is used. Cellulose fiber sample pads and glass fiber
conjugate pads are also purchased from Millipore. Avery Self-
Adhesive Laminating Sheets, 9  12 in., are obtained from
Amazon.com.

3 Methods

3.1 Preparation Preparation of high quality AuNPs ensures the success of

of AuNPs subsequent steps of the experiment. For current applications, we
chose to synthesize
13 nm diameter AuNPs for the following
reasons. First, the protocol for such synthesis is well-established
and requires only a simple mixing step [43]. Second, the resulting
AuNPs can be readily used for conjugation of thiol-modified DNA
[25], and the conjugates are usually highly stable against aggrega-
tion. In our experience, if nanoparticles of 40 nm diameter are
used, it is more difficult to obtain DNA conjugates with compara-
ble stability.
1. Prepare 500 ml of aqua regia by mixing 3:1 concentrated HCl:
HNO3 in a large beaker in a fume hood. The color of the
mixture changes to deep orange/red in several minutes. Be
extremely careful when preparing and working with aqua
regia. Wear goggles and gloves, and perform the experiment
in a fume hood.
2. Soak a two-neck flask, magnetic stir bar, stopper, and condenser
in the aqua regia solution for at least 15 min (see Note 1).
The volume of the flask can vary depending on the scale of
synthesis, and usually 100–500 ml is used. Rinse the glassware
with copious amount of deionized water and then Millipore
water.
3. Prepare 50 mM HAuCl4 solution by dissolving the solid in
Millipore water. Do not use metal spatula while weighing out
the HAuCl4. Filter the solution with a 0.2 μm pore size syringe
filter. Prepare 38.8 mM trisodium citrate solution by dissolving
the salt in Millipore water and filter the solution.
 Dipstick Metal ion Detection 399

4. To prepare about 100 ml of AuNPs, add 98 ml of Millipore

water into the two-neck flask. Add 2 ml of 50 mM HAuCl4
solution so that the final HAuCl4 concentration is 1 mM.
Connect the water condenser to one neck of the flask, and
place the stopper in the other neck. Put the flask on a hot
plate to reflux while stirring.
5. When the solution begins to reflux, remove the stopper.
Quickly add 10 ml of 38.8 mM sodium citrate, and replace
the stopper. The color should change from pale yellow to
grayish blue to deep red in 1 min. Allow the system to reflux
for another 20 min.
6. Turn off the heating and allow the system to cool to room
temperature (23–25  C) under stirring. The diameter of such
prepared AuNPs is
13 nm. The extinction value of the
520 nm plasmon peak is
2.4, and the nanoparticle concentra-
tion is
13 nM. The color of the solution should be burgundy
red, and the AuNP shape should be spherical under transmis-
sion electron microscopy (TEM). The prepared nanoparticles
are stable for months when stored in a clean container (glass or
plastic) at room temperature. Do not freeze the nanoparticles.

3.2 Functionalization 1. Soak disposable scintillation vials (20 ml volume) in 12 M

of AuNPs with NaOH for 1 h at room temperature (see Note 2). Rinse the
Thiol-Modified DNA vials with copious amounts of deionized water and then Milli-
pore water. Be extremely careful when preparing and working
with concentrated NaOH. Wear goggles and gloves. When
preparing 12 M NaOH solution, the temperature of the system
increases significantly. Occasional stirring is needed to avoid the
condensation of solid NaOH on the bottom of the container.
The concentrated NaOH solution can be reused many times
for soaking glass vials.
2. Prepare 10 mM fresh TCEP solution by dissolving a tiny crystal
of TCEP in Millipore water.
3. Pipette 9 μl of 1 mM thiol modified substrate DNA into a
microcentrifuge tube.
4. Add 1 μl of 500 mM acetate buffer (pH 5.2) and 1.5 μl of
10 mM TCEP to the tube to activate the thiol-modified DNA.
Incubate the sample at room temperature for 1 h. This activa-
tion step is necessary because the thiol-modified DNA from
IDT is shipped in the oxidized form with a disulfide bond.
5. Transfer 3 ml of the already prepared AuNPs into the NaOH-
treated glass vials, and then add the TCEP-treated thiol DNA
with gentle shaking by hand.
6. Cap the vial and store in a drawer at room temperature for at
least 16 h. Although all the operations described in this
400 Debapriya Mazumdar et al.

protocol can be carried out under light, it is advised to keep

nanoparticles in the dark for long-term storage.
7. After the initial incubation, add 30 μl of 500 mM Tris–HCl
(pH 8.0) buffer dropwise to the vial with gentle hand shaking.
The final Tris acetate concentration is 5 mM.
8. Add 300 μl of 1 M NaCl dropwise to the vial with gentle hand
shaking. Cap the vial tightly and store them in a drawer for at
least another day before use. When sealed tightly, the functio-
nalized AuNPs can be stored at room temperature for several
months. However, slow degradation of the DNA on AuNPs
may happen to change the properties of the AuNPs. See Note 3
for discussion on the storage of DNA-functionalized AuNPs.

3.3 Preparation 1. Transfer 500 μl of functionalized particles into two 1.7-ml

of DNAzyme-Linked microcentrifuge tubes.
AuNP Aggregates 2. Centrifuge the two tubes at 16,110  g at room temperature
(23–25  C) on a benchtop centrifuge for 15 min.
3. Gently remove the two tubes from the centrifuge. The super-
natant should be clear, and the AuNPs should be at the bottom
of the tubes. If a red color can still be observed in the superna-
tant, centrifuge for another 5 min.
4. Gently pipette off as much supernatant as possible to remove
free DNA. Again, disperse the AuNPs in 200 μl of buffer
containing 100 mM NaCl, 25 mM Tris–HCl, pH 8.0.
5. Centrifuge again for 10 min at 16,110  g at room
temperature.
6. Remove the supernatant. Again, disperse the nanoparticles in
200 μl of buffer containing 100 mM NaCl, 25 mM Tris–HCl,
pH 8.0. We found that most of the free DNA can be removed
by two centrifugations. If desired, repeat steps 2–5 to remove
more of the free DNA.
7. To each tube containing the substrate DNA functionalized
AuNP, add 25 μl buffer containing 25 mM Tris–HCl pH 8.0,
100 mM NaCl, 8% sucrose and mix the contents of two tubes
together to a total volume of
50 μl. This constitutes the
AuNP functionalized with the thiol modified substrate called
SubAuNP. (See Note 4 about optimizing sucrose concentra-
tion and salt concentration.)
8. To prepare the hybridized construct of substrate functionalized
with gold nanoparticles and enzyme (called Enz-SubAuNP),
add 5 μl of 1 mM enzyme to the 50 μl of SubAuNP.
9. Anneal by heating at 70  C for 2 min and slowly cooling to
room temperature over
1 h, to give the hybridized construct
of substrate functionalized with gold nanoparticles and enzyme
 Dipstick Metal ion Detection 401

(called Enz-SubAuNP). See Note 5 for discussion on the

design of DNAzymes for dipstick application.
3.4 Dipstick Tests 1. Preparation of Lateral Flow Devices
Lateral flow devices are constructed from membranes and pads
obtained from Millipore Corporation. Hi-Flow Plus Cellulose
Ester Membrane with a nominal capillary flow time of 240 s/
4 cm and a nominal membrane thickness of 135 μm direct cast
onto 2 mil polyester backing is used. The flow time of 240 s/
4 cm gives the highest sensitivity. Using a membrane which has
faster flow times (such as 90 s/4 cm) will increase the speed,
but reduce the sensitivity. The absorption pad and wicking pad
are cut from Millipore cellulose fiber sample pads, and the
conjugation pad is cut from the Millipore glass fiber conjugate
pad.
– Cut out the absorption pad (15 mm  300 mm) and
the wicking pad (15 mm  300 mm) from Millipore cellu-
lose fiber sample pads, and the conjugation pad
(13 mm  300 mm) from Millipore glass fiber conjugate pad.
– Cut the membrane (50 mm  300 mm) from the mem-
brane sheet obtained from Millipore. Touch the membranes
from the edges only.
– Attach the absorption pad, wicking pad, and conjugation
pad to a plastic backing which is at least 300 mm wide. The
overlap for each pad should be
2 mm. The assembled
components are cut using a paper cutter into individual
lateral flow devices with a width
8 mm. Discard the lateral
flow devices that are cut from the edges as those membranes
may have been damaged during handling.

3.5 Application 1. Apply 5 μl of the construct, Enz-SubAuNP on each conjuga-

of Reagents on Lateral tion pad. Apply 1.5 μl of 10 mg/ml streptavidin on the control
Flow Device and zone and 1.5 μl of 1 mM capture DNA on the test zone of the
Detection membrane using a 2 μl pipet (as shown in Fig. 3c). Allow the
devices to dry for 6–8 h.
2. In order to test the dipsticks, prepare solutions of various
concentrations of Pb2+ dissolved in a flow buffer containing
25 mM Tris (pH 8.0), 30 mM NaCl, 1% glycerol. The glycerol
is used to slow down the flow.
3. Dip the wicking pad into the buffer solution containing lead
(Pb2+) for 10 min or until the liquid has migrated to the
adsorption pad.
4. Remove the device from the buffer and lay it down on a flat
plastic surface for flow to continue for about 5 min. Scan the
dipsticks when they are dry.
5. If Pb2+ is not present, a single red line is observed at the control
zone (Fig. 4a). In the presence of Pb2+, a second red line is
402 Debapriya Mazumdar et al.

Fig. 4 Results of dipstick test for lead: (a) performance when the Enz-SubAuNP was pre-immobilized on the
conjugate pad and the Pb2+ reaction occurred on the surface of the device; (b) performance when Enz-
SubAuNP was allowed to react with Pb2+ in solution and then placed on the conjugate pad

observed at the test zone and its intensity increases with

increasing Pb2+ concentration (Fig. 4a). This test can be quali-
tative or semiquantitative because a color chart can be used to
estimate the Pb2+ concentration, like a pH paper. The detec-
tion limit for this test is determined to be
5 μM. This
detection limit is higher than
0.1 μM reported for the
DNAzyme-AuNP sensor in solution [2]. This is because
the reaction is slowed down by diffusion limitation of Pb2+
to the DNAzyme active site on the surface, resulting in long
reaction time (up to 1 h) [10]. Since the reaction on the
lateral flow device takes place within 10 min, it is difficult to
complete such a slow reaction on surface.
6. When the Pb2+- induced cleavage reaction is done in solution and
the lateral flow device as a medium to visualize the results, the
sensitivity is improved. For the tests in which the Pb2+ reaction is
performed in solution; add 1 μl of 6X concentration Pb2+ solu-
tion to 5 μl of Enz-SubAuNP construct and let the reaction
proceed for 15 min. Apply this reaction mixture on the conjuga-
tion pad and then dip into the flow buffer for 10 min. Remove
the device and lay it flat for 5 min and scan when dry (Fig. 4b).
7. A clear red line at the test zone can be seen in Fig. 4b at 0.5 μM
Pb2+, and thus the sensitivity is
10 times better than the
system shown in Fig. 4a. The sensitivity of this lateral flow
device shown in Fig. 4b is also good in comparison with our
previously reported colorimetric lead sensor that is based on
the assembly of AuNPs and produces a blue to red color change
with Pb2+ [2]. Although the detection limit for the solution-
based method is 0.1 μM when the absorbance is measured
using a UV–Vis instrument, it is difficult to visualize the red
color below 5 μM Pb2+, because of the large blue background
from the aggregates. Thus, this lateral flow device provides a

10-fold improvement in sensitivity for visual detection over
the previously reported colorimetric sensor [2]. Further,
performing the reaction in solution only requires one addi-
tional step of placing the reacted DNAzyme construct on the
conjugation pad, which can be carried out using a dropper.
 Dipstick Metal ion Detection 403

4 Notes

1. Care should be taken to make sure that no contamination is

introduced during AuNP synthesis. Low quality AuNPs can
result in poor DNA conjugation, which will induce nanoparti-
cle precipitation during later processing steps such as centrifu-
gation. Obtaining high-quality nanoparticles is the first
important step toward the success of the experiment.
2. If the glass vials are not treated with concentrated NaOH,
nanoparticles tend to stick to the surface of the vials, especially
after addition of NaCl to the particles. If this problem occurs,
the effective concentration of nanoparticles decreases.
3. The functionalized nanoparticles can still be used to form
aggregates even after storing at room temperature for months,
although their properties may change slightly with the passage
of time as a result of events such as degradation of DNA. These
changes could affect the properties of the prepared aggregates.
To be sure that the results are consistent, use freshly functio-
nalized AuNPs.
4. For lateral flow devices it is important to optimize the sucrose
and NaCl concentration for drying the constructs on the con-
jugation pad. Sucrose is important for preserving the DNA-
linked aggregates and aiding their rehydration. If sucrose is not
added, then the aggregates do not disassemble. We tested
sucrose concentration in the range of 0–30% and chose 8%
sucrose as the optimum condition for drying our constructs.
Optimizing the NaCl concentration (typically 50–500 mM) is
important to obtain a balance between stabilizing the DNA
base pairing interactions and achieving fast disassembly in the
presence of analyte. The optimum concentration will depend
upon the DNAzyme construct that is chosen for making dip-
sticks and therefore we recommend testing at least 3–5 separate
NaCl concentrations when optimizing the tests.
5. It is important to optimize the sequence of the DNAzyme
construct used for the dipstick tests, so that the cleavage effi-
ciency is maintained but the release of the cleaved substrate is
quick. For the system described here one arm is elongated to
11 base pairs, instead of the 9 base pairs in the original 17E
DNAzyme (these bases have not been shown in the figure for
clarity). The number of base pairs on the other arm (which in
linked to the AuNP)is varied in order to facilitate release. For
the construct shown in Fig. 5a, there are 9 base pairs, and the
dipstick stick test with Pb2+ does not show any red line at the
test zone. For the construct shown in Fig. 5b, there are 6 base
pairs, and the dipstick stick test with Pb2+ shows a faint red line
404 Debapriya Mazumdar et al.

Fig. 5 Constructs used for the optimization of dipstick test. Construct (c) had the best performance in the
presence of Pb2+ and was used for all other tests reported in this chapter. Construct (a) and (b) have different
arm lengths and worse performance comparing to construct (c)

at the test zone. This is because the release of the cleaved

substrate piece after Pb2+ reaction is easier as only 6 base pairs
are holding it. For the construct shown in Fig. 5c, there are 6
base pairs; however, in this case only the enzyme arm is reduced
in length to 6 bases whereas the substrate arm has 9 bases. The
dipstick stick test with Pb2+ shows a dark red line at the test
zone. By keeping the substrate intact, it is possible that the
activity of the DNAzyme is higher as the structural perturba-
tion is less. In addition, the capture of the cleaved piece at the
test zone is better due to 3 additional base pairs with the
capture DNA. Construct (c) is used in all other tests reported
in this method, because its performance has been the best
among the constructs tested.

Acknowledgment

This material is based on work from the following funding agencies—

National Institute of Health (Grant no. ES016865), Department of
Energy (DE- FG02-08ER64568), the National Science Foundation
(Grant no. CTS-0120978 and DMI-0328162), and Department of
 Dipstick Metal ion Detection 405

House and Urban Development (ILLHT0112-06). Yi Lu is a

cofounder of both ANDalyze and GlucoSentient, Inc.

References
1. Kruger K, Grabowski PJ, Zaug AJ, Sands J, selection of catalytic DNA. Comb Chem High
Gottschling DE, Cech TR (1982) Self-splicing Throughput Screen 5:327–335
RNA: autoexcision and autocyclization of the 16. Carmi N, Shultz LA, Breaker RR (1996) In
ribosomal RNA intervening sequence of Tetra- vitro selection of self-cleaving DNAs. Chem
hymena. Cell 31:147–157 Biol 3:1039–1046
2. Guerrier-Takada C, Gardiner K, Marsh T, Pace 17. Liu J, Brown AK, Meng X, Cropek DM, Istok
N, Altman S (1983) The RNA moiety of ribo- JD, Watson DB, Lu Y (2007) A catalytic bea-
nuclease P is the catalytic subunit of the con sensor for uranium with parts-per-trillion
enzyme. Cell 35:849–857 sensitivity and millionfold selectivity. Proc Natl
3. Breaker RR, Joyce GF (1994) A DNA enzyme Acad Sci U S A 104:2056
that cleaves RNA. Chem Biol 1:223–229 18. Vannela R, Adriaens P (2007) In vitro selection
4. Tuerk C, Gold L (1990) Systematic evolution of Hg (II) and As (V)-dependent RNA-cleaving
of ligands by exponential enrichment: RNA DNAzymes. Environ Eng Sci 24:73–84
ligands to bacteriophage T4 DNA polymerase. 19. Santoro SW, Joyce GF (1997) A general pur-
Science 249:505–510 pose RNA-cleaving DNA enzyme. Proc Natl
5. Ellington AD, Szostak JW (1990) In vitro Acad Sci U S A 94:4262–4266
selection of RNA molecules that bind specific 20. Liu J, Lu Y (2006) Fast colorimetric sensing of
ligands. Nature 346:818–822 adenosine and cocaine based on a general sen-
6. Wilson DS, Szostak JW (1999) In vitro selec- sor design involving aptamers and nanoparti-
tion of functional nucleic acids. Annu Rev Bio- cles. Angew Chem Int Ed Engl 45:90–94
chem 68:611–647 21. Liu J, Lu Y (2006) Preparation of aptamer-
7. Gold L, Polisky B, Uhlenbeck O, Yarus M linked gold nanoparticle purple aggregates for
(1995) Diversity of oligonucleotide functions. colorimetric sensing of analytes. Nat Protoc
Annu Rev Biochem 64:763–797 1:246–252
8. Breaker RR (2002) Engineered allosteric ribo- 22. Liu J, Lee JH, Lu Y (2007) Quantum dot
zymes as biosensor components. Curr Opin encoding of aptamer-linked nanostructures
Biotechnol 13:31–39 for one pot simultaneous detection of multiple
9. Hesselberth J, Robertson MP, Jhaveri S, analytes. Anal Chem 79:4120–4125
Ellington AD (2000) In vitro selection of 23. Yigit MV, Mazumdar D, Kim H-K, Lee JH,
nucleic acids for diagnostic applications. Rev Odintsov B, Lu Y (2007) Smart “turn-on”
Mol Biotechnol 74:15–25 magnetic resonance contrast agents based on
10. Famulok M (1999) Oligonucleotide aptamers aptamer-functionalized superparamagnetic
that recognize small molecules. Curr Opin iron oxide nanoparticles. Chembiochem
Struct Biol 9:324–329 8:1675–1678
11. Ihms HE, Lu Y (2012) In vitro selection of 24. Mirkin CA, Letsinger RL, Mucic RC, Storhoff
metal ion-selective DNAzymes. Methods Mol JJ (1996) A DNA-based method for rationally
Biol 848:297–316 assembling nanoparticles into macroscopic
12. Brown AK, Li J, Pavot CMB, Lu Y (2003) A materials. Nature 382:607–609
lead-dependent DNAzyme with a two-step 25. Storhoff JJ, Elghanian R, Mucic RC, Mirkin
mechanism. Biochemistry 42:7152–7161 CA, Letsinger RL (1998) One-pot colorimet-
13. Li J, Lu Y (2000) A highly sensitive and selec- ric differentiation of polynucleotides with sin-
tive catalytic DNA biosensor for lead ions. J gle base imperfections using gold nanoparticle
Am Chem Soc 122:10466–10467 probes. J Am Chem Soc 120:1959–1964
14. Li J, Zheng W, Kwon AH, Lu Y (2000) In vitro 26. Liu J, Lu Y (2006) Smart nanomaterials
selection and characterization of a highly effi- responsive to multiple chemical stimuli with
cient Zn(II)-dependent RNA-cleaving deoxyr- controllable cooperativity. Adv Mater
ibozyme. Nucleic Acids Res 28:481–488 18:1667–1671
15. Bruesehoff PJ, Li J, Augustine AJ, Lu Y (2002) 27. Liu J, Lu Y (2003) A colorimetric lead biosen-
Improving metal ion specificity during in vitro sor using DNAzyme-directed assembly of gold
406 Debapriya Mazumdar et al.

nanoparticles. J Am Chem Soc DNA nanotechnology. Springer, Berlin, Hei-

125:6642–6643 delberg, pp 277–305
28. Liu J, Lu Y (2004) Accelerated color change of 37. Zhang J, Liu B, Liu H, Zhang X, Tan W (2013)
gold nanoparticles assembled by DNAzymes Aptamer-conjugated gold nanoparticles for
for simple and fast colorimetric Pb2+ detec- bioanalysis. Nanomedicine 8:983–993
tion. J Am Chem Soc 126:12298–12305 38. Glynou K, Ioannou PC, Christopoulos TK,
29. Liu J, Lu Y (2004) Adenosine-dependent Syriopoulou V (2003) Oligonucleotide-
assembly of aptazyme-functionalized gold functionalized gold nanoparticles as probes
nanoparticles and its application as a colorimet- in a dry-reagent strip biosensor for DNA anal-
ric biosensor. Anal Chem 76:1627–1632 ysis by hybridization. Anal Chem 75:
30. Liu J, Lu Y (2006) Fluorescent DNAzyme bio- 4155–4160
sensors for metal ions based on catalytic molec- 39. Liu J, Mazumdar D, Lu Y (2006) A simple and
ular beacons. Methods Mol Biol 335:275–288 sensitive “dipstick” test in serum based on lat-
31. Liu J, Lu Y (2007) Rational design of “Turn- eral flow separation of aptamer-linked nanos-
On” allosteric DNAzyme catalytic beacons for tructures. Angew Chem Int Ed Engl
aqueous mercury ions with ultrahigh sensitivity 45:7955–7959
and selectivity. Angew Chem Int Ed Engl 40. Lu Y, Liu J, Mazumdar D (2009) Nanoparti-
46:7587–7590 cles/dip stick. In: Mayer G (ed) Nucleic acid
32. Liu J, Lu Y (2007) A DNAzyme catalytic bea- and peptide aptamers, vol 535. Humana Press,
con sensor for paramagnetic Cu2+ ions in New York, pp 223–239
aqueous solution with high sensitivity and 41. Zhao W, Ali MM, Aguirre SD, Brook MA, Li Y
selectivity. J Am Chem Soc 129:9838–9839 (2008) Paper-based bioassays using gold nano-
33. Liu J, Lu Y (2005) Stimuli-responsive disas- particle colorimetric probes. Anal Chem
sembly of nanoparticle aggregates for light-up 80:8431–8437
colorimetric sensing. J Am Chem Soc 42. Mazumdar D, Liu J, Lu G, Zhou J, Lu Y
127:12677–12683 (2010) Easy-to-use dipstick tests for detection
34. Liu J, Lu Y (2006) Design of asymmetric of lead in paints using non-cross-linked gold
DNAzymes for dynamic control of nanoparti- nanoparticle-DNAzyme conjugates. Chem
cle aggregation states in response to chemical Commun 46:1416–1418
stimuli. Org Biomol Chem 4:3435–3441 43. Handley DA (1989) Methods for synthesis of
35. Torabi SF, Lu Y (2014) Functional DNA nano- colloidal gold. In: Hayat MA (ed) Colloidal
materials for sensing and imaging in living cells. gold principles, methods, and applications, vol
Curr Opin Biotechnol 28:88–95 1, 1st edn. Academic Press, San Diego, pp
36. Li L, Lu Y (2013) Functional DNA-integrated 13–32
nanomaterials for biosensing. In: Fan C (ed)
 Chapter 25

Liposome-Enhanced Lateral-Flow Assays for Clinical

Analyses
Katie A. Edwards, Ricki Korff, and Antje J. Baeumner

Abstract
Clinical and environmental analyses frequently necessitate rapid, simple, and inexpensive point-of-care or
field tests. These semiquantitative tests may be later followed up by confirmatory laboratory-based assays,
but provide an initial scenario assessment important for resource mobilization and threat confinement.
Lateral-flow assays (LFAs) and dip-stick assays, which are typically antibody-based and yield a visually
detectable signal, provide an assay format suiting these applications extremely well. Signal generation is
commonly obtained through the use of colloidal gold or latex beads, which yield a colored band either
directly proportional or inversely proportional to the concentration of the analyte of interest. Here, dye-
encapsulating liposomes as a highly visible alternative are discussed. The semiquantitative LFA biosensor
described in this chapter relies on a sandwich immunoassay for the detection of myoglobin in whole blood.
After an acute myocardial infarction (AMI) event, several cardiac markers are released into the blood, the
most common of which are troponin, creatine kinase MB, C-reactive protein, and myoglobin. Due to its
early release, myoglobin has value as an indicator of a recent heart attack amongst conditions which present
with similar symptoms and its lack of elevation can effectively rule out a heart attack (Brogan et al., Ann
Emerg Med 24:665–671, 1994). The assay described within relies on sandwich complex formation
between a membrane immobilized capture monoclonal antibody against myoglobin, a detector biotiny-
lated monoclonal antibody against a different epitope on myoglobin, and streptavidin-conjugated visible
dye (sulforhodamine B)-encapsulating liposomes to allow for signal generation.

Key words Lateral-flow assay, Liposome, Whole blood, Myoglobin

1 Introduction

Clinical and environmental analyses frequently necessitate rapid,

simple, durable, and inexpensive point-of-care or field tests.
Lateral-flow assays (LFAs), such as those used in pregnancy
tests, often fulfill these requirements. These assays rely on the
appearance of a colored band, the intensity of which corresponds
to the concentration of the target analyte in the sample. The goal
for such assays is typically a qualitative “yes” or “no” answer,
though some assays can provide semiquantitative results. LFAs

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_25,
© Springer Science+Business Media LLC 2017

407
408 Katie A. Edwards et al.

are commercially available for the detection of a wide variety of

analytes. These targets include pathogenic organisms such as
Legionella pneumophilia [1, 2], Leptospira [3], Cryptosporidium
parvum, and Giardia lamblia [4]; viruses such as HIV [5], influ-
enza [6], and respiratory syncytial virus (RSV) [7, 8]; food aller-
gens [9] and drugs of abuse [10]; health states such as pregnancy,
ovulation, or menopause [11, 12]; diseases such as prostate cancer
[13, 14]; or biological toxins such as botulinum toxin [15, 16],
and verotoxins [17, 18]. Some lateral flow assays have proven to
be equivalent or better than enzyme-linked immunosorbent
assays (ELISAs), direct fluorescent antibody testing, and viral
cultures [5, 6, 8, 19]. Lateral-flow assays used in a clinical setting
can obviate the need for patients to return to the clinician’s office
for results [20] and can provide guidance for the administration of
chemoprophylaxis after potential occupational exposure to viruses
[21]. In environmental or occupational settings, LFAs provide an
initial assessment for mobilization of resources and containment
of potential threats. LFAs for research and more fundamental
applications such as probe selection, PCR-product identification,
genomic sequence search, and identification of single-nucleotide
polymorphisms have been suggested [22–24].
Here we discuss in detail the use and preparation of LFAs for
analytes present in whole blood, using the 16.7 kDa cardiac marker
myoglobin as an example analyte. After an acute myocardial infarc-
tion (AMI) event, several cardiac markers are released into the
blood, the most common of which are cardiac troponin I (cTnI),
cardiac troponin T (cTnT), creatine kinase MB subform (CK-MB),
C-reactive protein (CRP), and myoglobin. cTnI, cTnT, and CK-
MB are initially elevated 4–6 h after an AMI event; C-reactive
protein is initially elevated up to 6 h after the event, and myoglobin
is initially elevated after 1–3 h [25, 26]. The relative increase in
these markers over time following an incident of chest pain is
shown in Fig. 1.
Due to the rapid and reliable release of myoglobin from dam-
aged myocardium, tests that use serum myoglobin levels to deter-
mine the risk of a recent myocardial infarction have high
sensitivities and high negative predictive values within 90 min of
symptom onset [27]. Especially when combined with detection of
other cardiac markers like troponin I, the absence of elevated serum
myoglobin is useful in quickly ruling out acute myocardial infarc-
tion in patients with symptoms such as chest discomfort, dyspnea,
and syncope [28, 29]. Since approximately 80% of patients present-
ing to emergency rooms with AMI-associated symptoms do not
develop heart attacks [30], a test that can rapidly rule out AMI may
allow hospitals to prevent unnecessary admission of patients to
coronary care units. While a serum myoglobin test cannot confirm
AMI since myoglobin release is not specific to cardiac muscle
damage, detecting elevated levels of myoglobin in the blood
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 409

Hours After Chest Pain

0 4 6 8 16 24 36 48 72
1000
Myo

Relative Marker Increase (log scale)

CK-MB
cTnl
cTnT
100
cTnT
Myoglobin
10
CK-MB cTnl

0.1

Fig. 1 Temporal release of myoglobin, CK-MB, and cTnT and cTnI. Reprinted with
permission from Christenson RH, Azzazy HME. Biomarkers of Myocardial Necro-
sis: Past, Present and Future. In Morrow DA, ed. Cardiovascular Biomarkers:
Pathophysiology and Clinical Management. Totowa, NJ: Humana Press, 2006

regardless of its source is important in preventing kidney damage.

In rhabdomyolysis (severe muscle breakdown), serum myoglobin
levels may rise to levels that lead to acute renal failure [31].
The approach described within allows for the analysis of myoglobin
levels in whole blood using plasma separators incorporated into a
LFA device and the use of dye-encapsulating liposomes as a highly
visible signaling species in contrast to commonly used latex particles
and colloidal gold.

1.1 Sandwich While simple for the end-user to operate, from an engineering
Immunoassay Format standpoint, the sandwich lateral flow immunoassay format is a
relatively complicated multicomponent design encompassing a
sample pad, conjugate pad, analytical membrane, absorbent pad,
signaling species, blocking constituents, and multiple antibodies.
When these components are assembled appropriately, the end-user
simply needs to add a liquid sample and wait typically 5–15 min
before a visible signal is assessed. The flow of the fluid in such an
assay is detailed in Fig. 2.
A liquid sample ideally with minimal preparation requirements
is applied to the sample pad via an opening in the housing. Depend-
ing on the sample matrix, the sample pad can serve as a particulate
filter, site to retain interfering substances, and provide pH buffering
capacity with the overall aim to prepare the fluid for subsequent
analysis and dispense it evenly. Once the fluid passes through the
sample pad, it then passes through a conjugate pad where it rehy-
drates a previously dehydrated signaling species, which is typically
410 Katie A. Edwards et al.

Fig. 2 Lateral flow assay fluid path. (Top) A liquid sample is applied to a sample pad through a port in the
plastic housing (not shown). The liquid sample then passes through the conjugate pad where the signaling
species (most commonly colloidal gold) with an attached detection antibody becomes rehydrated. The solution
then passes onto the analytical membrane were capture and control antibodies are immobilized, then lastly is
wicked by the absorbent pad at the opposite end of the assembly. (Middle) If the analyte is present, it can form
sandwich complexes with the detection antibody on the colloidal gold and the capture antibody immobilized
on the analytical membrane. A visible signal where the capture antibody is immobilized is observed as well as
a visible signal where the control antibody is immobilized to indicate successful fluid flow and conjugate
release. (Bottom) If the analyte is not present, a signal at only the control line is observed

colloidal gold or visibly dyed latex beads. The surface of the signal-
ing species is functionalized with “detection” antibodies against the
analyte of interest and can form antibody-antigen complexes with
analyte in the sample solution if it is present. The fluid then passes
onto the analytical membrane where a second antibody is immobi-
lized at the test line. This “capture” antibody, most commonly a
monoclonal antibody (mAb), can recognize a different epitope on
the target analyte than that recognized by the detection antibody
and serves to retain the passing analyte-signaling particle complexes.
The sandwich complex formed between the immobilized capture
antibody, analyte, and detection antibody-tagged particles allows a
visible signal to be observed in the presence of the target analyte due
to the attached signaling species. When the analyte is not present, no
sandwich complexes are formed and thus no signal is generated. The
intensity of the signal is proportional to the analyte concentration,
which can be used for semiquantitative determinations. At a location
downstream of the test line is a control line. This control line
typically consists of a secondary antibody, which can bind to the
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 411

primary antibody immobilized on the surface of the signaling species

(e.g., a goat anti-mouse secondary antibody could be used as a
control line if the primary antibody on the colloidal gold particles
was a mouse anti-myoglobin antibody). This control line captures
passing signaling species regardless of the presence or absence of the
target analyte and allows the end user to confirm that sufficient fluid
was applied to yield a valid assay and that the conjugate successfully
released. At the distal end of the membrane lies an absorbent pad,
which serves to draw fluid through the assay and retain excess fluid
for the duration of the assay.
The inherent format of the sandwich immunoassay places
certain constraints on the types of analytes that can be detected.
In order to accommodate the binding of two separate antibodies
which are
150 kDa each, the analyte must be relatively large.
As such, it is predominantly used for the quantification of large
proteins or whole cells. While this format precludes the detection of
small molecules, it has been used for quantification of oligopeptides
including, for example, insulin C-peptide (31 amino acids (AA),
3 kDa) [32] provided an appropriate pair of antibodies is available.
To satisfy the criteria for appropriate antibodies, either two anti-
bodies that can bind to distinct, nonoverlapping epitopes on the
target molecule are required, or if a single antibody type is used,
the target molecule must have multiple copies of a single epitope.
In practice, the method developer typically needs to determine
which antibodies can successfully form a sandwich complex with a
given analyte through in-house experimentation. With the price of
commercially available primary antibodies typically ranging from

$350 to $700/mg, this can be a costly trial and error process.
Thus, it is a great benefit when “paired” antibodies that are known
to work together in sandwich assays are available from antibody
vendors. Other considerations to the choice of antibody include its
level of purification and concentration. Antibodies used as capture
antibodies in LFAs should not contain any stabilizing proteins or
other species that may compete for adsorption to the nitrocellulose.
Similarly, if the detection antibody is covalently attached to the
signaling species, it should not contain stabilizing proteins or buffer
constituents that will interfere with the conjugation chemistry.
Unless the capture antibody has a very strong affinity, a minimum
concentration of 1 mg/mL is recommended. The sensitivity of
lateral flow assays stems in part from the basic design where fluid
must pass the test line where it has the opportunity to contact the
immobilized capture antibody with negligible mass transport lim-
itations. This is in contrast to enzyme-linked immunosorbent assays
(ELISAs) where diffusion from the bulk sample volume must occur
to the plate surface where capture antibodies are immobilized [33].
While the latter can often achieve low limits of detection using
much lower concentrations of antibodies (typically 5–10 μg/mL),
extended incubation times (typically 1–2 h) are required. Given that
412 Katie A. Edwards et al.

there is a limited amount of time in which an analyte is in contact

with antibody immobilized on the test line which is typically
0.5–1.0 mm wide, binding kinetics are key in LFAs. Aside from
high affinity, the association rate is a critical parameter [33, 34].

1.1.1 Membrane Types Most commonly, nitrocellulose membranes are used as analytical
membranes for LFAs due to their high binding capacity, low cost,
and wide availability [35]. The mechanism behind protein binding
to nitrocellulose is not fully understood, but is believed to be non-
covalent through hydrophobic, hydrogen bonding, and electro-
static interactions [36, 37]. The physical adsorption of nucleic
acid probes through drying to nitrocellulose is believed to result
in attachment through hydrophobic interactions, an effect which
may be enhanced through modification with a poly-T tail [38].
We have also had success using polyethersulfone membranes for
nucleic acid sandwich hybridization assays [22, 23, 39].
More recently, membranes based on electrospun nanofibers
have been applied for LFAs [40]. Electrospun nanofibers are
prepared by using a simple set-up consisting of a power voltage
supply, a syringe with metal needle, a syringe pump and a ground
plate. A polymer solution (dope) is filled into the syringe and
pumped out at a steady flow rate. A high voltage (7–15 kV) is
applied between the tip of the needle and the ground plate. A taylor
cone is formed and a nanofiber in the size range of a few tens of
nanometers to a few micrometers are collected on the grounded
plate, depending on the spinning dope and electrospinning condi-
tions chosen. Nitrocellulose and nylon via electrospinning have been
shown for protein slot blot and Western blot applications [41].
By doping the composite with poly-L-lysine (PLL) or polylactic
acid (PLA), specific functional groups including amino and carbox-
ylic acid groups, respectively, may be incorporated [42]. Additionally,
through control over polymer dope composition, voltage, distance
between needle tip and ground plate control over the mat thickness
and porosity can be achieved and offers significant surface area for
immobilization.

1.1.2 Membrane The ideal properties of the membrane include particle and pore size
Properties consistency, hydrophilicity, high protein binding, and durability.
As nitrocellulose membranes are a fibrous polymeric matrix rather
than uniform particles, the membranes are characterized by their
capillary rise rate (or capillary flow time), rather than by a specific
particle or pore size. This rate is the amount of time required for fluid
to flow 4 cm, and based on a survey of current commercial offerings
from major manufacturers, is generally between 60 and 280 s. Since
fluid flow speed in lateral flow membranes decreases with the
distance traveled, the capillary rise rate is a more consistent measure
of the speed of flow [43]. This parameter is important to consider
during assay development since a fast wicking rate allows for more
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 413

rapid results and lower background, but may adversely affect assay
sensitivity since less time for interaction between the target and the
antibodies immobilized at the test line is available [44].
Commercially available nitrocellulose membranes vary in phys-
ical properties based on the effective “pore size,” the amount of
trapped air, and the thickness of their nitrocellulose layer. These
properties affect the surface area available for immobilization of
capture reagents. The binding capacity of nitrocellulose itself
is inherently high, with estimates ranging from 50 to 200 μg
IgG/cm2 [45]. The thickness of commercially available membranes
typically ranges from 100 to 150 μm. Thicker membranes can allow
more sample fluid to be accommodated by the membrane during the
assay and more capture reagent to be applied. However, the latter
does not necessarily yield greater sensitivity in lateral flow assays with
optical detection since due to the opacity of nitrocellulose, only the
top
10 μm of membrane is visible to the user [45]. One advantage
to thicker membranes is improved tensile strength, however, for ease
of handling, durability, and avoidance of adhesive interactions, a
supported membrane should be chosen. These membranes are man-
ufactured by directly casting the nitrocellulose membrane material
onto a plastic (polyester or cellulose acetate) backing. The backing
thickness typically ranges from 50 to 225 μm [43]. Alternatively,
non-supported membranes can be laminated manually prior to use.
LFAs may be further made more durable and user-friendly by encas-
ing them in an injected molded plastic housing such as that shown in
Fig. 3.

Fig. 3 Lateral flow assay membranes assembled in injection molded housings. In

the presence of the target analyte, a sandwich complex forms with dye-
encapsulating liposomes resulting in a magenta-colored band at the capture
zone that is proportional to target concentration (top). In the absence of target
analyte, no band is visible as no sandwich complex has formed (bottom). Though
not necessary, these LFA membranes were housed in more user friendly
packaging using a plastic cassette made using injection molding and designed
for addition of liposome/target mixture in hole #1 and addition of running buffer
in hole #2. Readings can be taken in the “R” hole
414 Katie A. Edwards et al.

In addition, a sample pad, conjugate release pad, and wicking

material are often utilized for LFAs to allow for sample pretreat-
ment such as filtration or pH adjustment; storage of dehydrated
signaling species; and accommodation of larger fluid volumes,
respectively. These are critical components for any assay developed
in complicated sample matrices and also those intended for con-
sumer usage where the number of steps needs to be minimized. For
the assays described in this chapter, the conjugate pad is omitted
since the liposomes used to provide the signal are maintained in the
solution phase. The sample pad is selected based on the type of
sample matrix that is employed. For the assays described within, a
sample pad that is capable of retaining white and red blood cells
from whole blood while allowing the plasma to pass is required.
The ideal sample pad minimizes hemolysis from red blood cells,
which can contribute to an undesired colored background signal on
the nitrocellulose membranes (Fig. 4).
The capture antibody can be applied to the membrane either
manually using a pipettor, with a thin-layer chromatography (TLC)
applicator, or similar approaches. Different binding characteristics,
buffers (pH, salt), application rate, and biorecognition element
concentration lead to differences in thickness in the capture zone
line. Once the biorecognition element is applied, the membrane
material is typically immersed in a blocking agent, which is used to

Fig. 4 Whole blood applied to lateral flow assay membranes assembled in

injection molded housings to test sample pad types. No test or control lines
were immobilized. Significant hemolysis was observed in the left assembly
which yielded a pink colored background which could potentially obscure
weak signals, whereas a clear background was available in the right assembly
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 415

reduce nonspecific binding of the assay components. Common

agents include proteins such as bovine serum albumin (BSA),
casein, and gelatin; non-ionic surfactants such as Tween 20, Triton
X-100, or Brij; or synthetic polymers such as polyvinylpyrrolidone
(PVP), polyvinyl alcohol (PVA), and polyethylene glycol (PEG).
These components compete for protein binding sites, interfere
with hydrophobic interactions, and protein binding, respectively
[46, 47]. In addition to serving as a blocking agent, Tween 20 has
also been reported to have a renaturing effect on immobilized anti-
gens, which can improve assay sensitivity through enhanced antige-
n–antibody binding [48, 49]

1.2 Detection LFAs typically utilize antibody-labeled colloidal gold [5, 19, 50] as
Mechanisms a visualization method. Colloidal gold ranges in size 2–250 nm,
though 30–80 nm particles are preferred for LFAs [24]. A diameter
of 40 nm has been reported to be optimal, allowing for clear
visualization, dense packing of these small particles at the capture
zone, and reduced steric hindrance for protein binding [51]. These
particles have a characteristic red color, allowing for visual detection
or semiquantitative measurements with portable reflectometers or
scanners [52–54]. Proteins are associated with gold particles ioni-
cally through the particle’s negative charge with positively charged
amino acids such as lysine; through hydrophobic interactions of
amino acids such as tyrosine and tryptophan; and through sharing
of electrons between gold and sulfur atoms of cysteine [51]. When
labeled with anti-biotin or streptavidin, such species allow for rec-
ognition of biotinylated biorecognition elements, such as DNA
oligonucleotides or antibodies. The sensitivity of LFAs using such
particles can be increased with silver enhancement [55]. Alterna-
tively, antibody-tagged latex particles [56, 57], up-converting
phosphors [58], superparamagnetic particles [59], or visible dye-
encapsulating liposomes [60] have been used as a signal enhance-
ment means. The latter species are the focus of this chapter.
Liposomes are vesicles formed through the association of phos-
pholipid molecules in an aqueous environment, yielding a structure
with the hydrocarbon tails forming a lipid bilayer and hydrophilic
headgroups directed at both the aqueous core and external aqueous
medium. One of the common methods for liposome formation is
known as the reverse-phase evaporation method [61, 62]. Here,
phospholipids are dissolved in organic solvent; the mixture is intro-
duced to an aqueous medium containing high concentrations of
visible dye; the solvent is removed under vacuum; the resulting
liposomes are passed through defined pore-size membranes to
improve homogeneity and reduce the number of lipid bilayers
(lamellarity); then the unencapsulated material is removed through
size-exclusion chromatography and dialysis. We commonly use
sulforhodamine B (SRB) dye, which is relatively inexpensive, highly
416 Katie A. Edwards et al.

soluble in water, has a high molar extinction coefficient, and has

good photostability. However, the most appropriate dye color can
be chosen to ensure good visibility against any background signal in
the LFA that is not removed by the sample pad. For example, for
whole blood, blue liposomes made using Patent Blue dye can
improve optical detection against a potentially red or brownish
background.
The advantages of dye-encapsulating liposomes as a label for
immunoassays include long-term stability of the encapsulated sig-
naling molecules; the substantial signal enhancement afforded by
the encapsulation of hundreds of thousands of dye molecules
within the large interior liposomal volume; the instantaneous signal
provided through either visual detection of intact liposomes or the
release of hydrophilic dye molecules from their aqueous cores upon
surfactant-induced liposome lysis; and the ease of labeling through
the covalent modification of lipid headgroups or the direct incor-
poration of hydrophobic biorecognition elements into their lipid
bilayers (Fig. 5) [60, 63–65].
The stability of the liposomes is a function of their lipid com-
position, buffer composition, and storage conditions. In the for-
mulation that we most commonly employ (described within), we
have found that streptavidin tagged sulforhodamine B-
encapsulating liposomes have retained their functionality in LFAs
for at least 400 days when stored at either 4 or 21  C, while loss of
encapsulated dye and signal occurs at elevated temperatures [66].

Fig. 5 Liposome structure. (a) Biorecognition elements can be covalently conjugated to or inserted into lipid
bilayers through hydrophobic interactions (not to scale). (b) The large internal volume of unilamellar vesicles
can encapsulate hundreds of thousands of hydrophilic signaling molecules and provide for their stability. (c)
Surfactant introduction can provide for instantaneous signal enhancement through release of encapsulants.
Fluorophores encapsulated within liposomes at high concentrations undergo self-quenching with is overcome
upon release into the surrounding medium
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 417

1.3 Universal Assays Both immunoassays and nucleic acid based lateral flow assays can
also be made in a universal format, obviating the need for the
generation of specifically labeled membranes and detection
elements. This can be done through membrane immobilization of
streptavidin [23] or anti-fluorescein antibodies [22] and
subsequent addition of biotinylated or fluorescein-labeled capture
antibodies, respectively, with the remaining assay components.
Protein A/G [67], streptavidin/avidin [22, 68], or generic oligo-
nucleotide [23] labels can be conjugated to the liposomes to form
generic species capable of facile recognition of the Fc0 region of
antibodies, biotinylated biorecognition elements, or complemen-
tary generic oligonucleotides, respectively. Anti-fluorescein, anti-
biotin, or anti-digoxigenin are also options for common universal
liposome tags [69]. These assays are of particular interest in
research laboratories where a variety of antibodies or probes need
to be screened or where different analytes need to be tested. From a
commercial and manufacturing standpoint, the universal format is
of interest since only one type of membrane and liposome need to
be prepared which simplifies packaging of tests for any analyte.

2 Materials

Materials listed here are those that have been used successfully in
our laboratory, though substitutions may be made. Unless other-
wise specified, reagents are molecular biology grade and are pur-
chased from VWR (Bridgeport, NJ). Table 1 lists the biomolecules
used for the detection of myoglobin as an example.

2.1 Liposome 1. Bath sonicator (Aquasonic Model 150D, VWR, Bridgeport, NJ).
Preparation 2. Rotary evaporator (Model R-114, Buchi, New Castle, DE).
3. 50-mL round bottom flask (Catalog # 80068-756, VWR).
4. Mini-extruder (Catalog # 610000, Avanti Polar Lipids, Alabas-
ter, AL), including two 1 mL syringes, Teflon supports and o-
rings, and extruder holder/heating block. When purchasing
initially, the membranes and supports listed in #5 of this section
are included with this catalog number as of the date that this
chapter was prepared.
5. Extrusion membranes (0.4 and 1.0 μm polycarbonate mem-
branes*, 19 mm) and filter supports (Catalog # 610007,
610010, and 610014, respectively, Avanti Polar Lipids).
6. 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), choles-
terol, 1,2-dipalmitoyl-sn-glycero-3-[phospho-rac-(1-glycerol)],
sodium salt (DPPG), 1,2-dipalmitoyl-sn-glycero-3-phos-
phoethanolamine-N-glutaryl, sodium salt (N-glutaryl DPPE)
(Catalog #: 850355, 700000, 840455, and 870245, respec-
tively, Avanti Polar Lipids).
418 Katie A. Edwards et al.

Table 1
Reagents needed for the development of a LFA for human myoglobin

Function Information Vendor Catalog number

Capture antibody Anti-myoglobin (mouse, Meridian Life Science, Inc. H86703M
mAb, IgG1)
Detection antibody Biotinylated anti-myoglobin Meridian Life Science, Inc. H86142B
(mouse, mAb, IgG1)
Control antibody Anti-streptavidin Vector Labs SP-4000
(goat, pAb)
Liposome tag Streptavidin Thermo Fisher Scientific S888
Test matrix Myoglobin-free serum Meridian Life Science, Inc. N86580H
Test analyte Recombinant full-length Meridian Life Science, Inc. A01428H
myoglobin
Function Information Vendor Catalog Number
Capture antibody Anti-myoglobin (mouse, Meridian Life Science, Inc. H86703M
mAb, IgG1)
Detection antibody Biotinylated anti-myoglobin Meridian Life Science, Inc. H86142B
(mouse, mAb, IgG1)
Control antibody Anti-streptavidin (goat, Vector Labs SP-4000
pAb)
Liposome tag Streptavidin Thermo Fisher Scientific S888
Test matrix Myoglobin-free serum Meridian Life Science, Inc. N86580H
Test analyte Recombinant full-length Meridian Life Science, Inc. A01428H
myoglobin

7. Chloroform, methanol, and isopropyl ether (HPLC grade).

8. 10HEPES-saline buffer is composed of 0.1 M HEPES, 2.0 M
sodium chloride, and 0.1% (w/v) sodium azide, adjusted to
pH 7.5 with NaOH.
9. 1HEPES–saline–sucrose buffer (1HSS) is prepared by dis-
solving 205.4 g sucrose (0.2 M) in 300 mL 10 HEPES-saline
and 900 mL water, then bringing the final volume to 3 L with
deionized water. However, we usually prepare a 2 M stock
solution of sucrose instead of weighing for this solution.
10. 0.419 g sulforhodamine B (SRB, Catalog # S1307, Molecular
Probes, Inc., Eugene, OR) is added to 0.5 mL 0.2 M HEPES,
pH 7.5 and diluted to a final volume of 5 mL with deionized
water to yield a 150 mM solution. For blue liposomes, Patent
Blue dye should be substituted.
11. Sephadex G-50 (Catalog # G-50-150, Sigma, St. Louis, MO)
and a glass chromatography column (such as VWR catalog #
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 419

KT420400-1530). The column is packed by first swelling 1 g

of Sephadex in 120 mL deionized water in 3  50 mL centri-
fuge tubes overnight, and decanting off water either after
centrifugation or settling by gravity then pouring. The volume
is then replaced by 1HSS and decanting procedure repeated.
Overall, three 1HSS buffer exchanges are typically sufficient.
The Sephadex mixture is then poured into the chromatography
column to a height of approximately 20 cm and allowed to
settle while maintaining a flow of 1HSS through the column
for at least 30 min. The top of the column should be level with
no gaps or bubbles throughout the remaining height.
12. Dialysis membranes, Specta/Por 2 (Catalog # 132678, Spec-
trum Laboratories, Inc., Rancho Dominguez, CA).
13. 15- and 50-mL centrifuge tubes (Catalog #: 21008-216 and
21008-242, respectively, VWR).

2.2 Liposome 1. Streptavidin (Catalog # S888, Molecular Probes, Inc., Eugene,

Conjugation OR) is diluted to 200 μM with 50 mM potassium phosphate,
pH 7.0 and aliquotted into 50 μL portions prior to storage at
20  C.
2. 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochlo-
ride (EDC) (Catalog #22980, Thermo Fisher Scientific).
3. 0.1 M 2-(N-morpholino)ethanesulfonic acid (MES) buffer,
pH 4.65.
4. Sepharose CL-4B (Sigma-Aldrich #CL4B200) packed as the
Sephadex G50 column is above in Subheading 2.1. Sepharose
CL-4B is provided as a slurry by the manufacturer, which
reduces the preparation time to that required for the HSS
buffer exchanges.

2.3 Membrane 1. TLC applicator (Linomat IV, CAMAG Scientific, Wilmington,

Preparation NC).
2. Vacuum sealer (Foodsaver, San Francisco, CA).
3. Vacuum oven, capable of 23 and 50  C.
4. Paper cutter.
5. Fine-tip tweezers (Catalog #25729-081, VWR).
6. Kimwipes, 1500  1700 (Catalog # 21905-049, VWR).
7. Nitrocellulose membrane cards (HF090 membrane cards, Cat-
alog # HF09004XSS, EMD Millipore, Cheshire, CT) cut into
20 cm sections.
8. Anti-myoglobin capture antibodies diluted to 1.0 mg/mL
with a solution of 0.4 M NaHCO3/Na2CO3, pH 9.0 contain-
ing 5% (v/v) methanol. Anti-streptavidin control antibodies
diluted to 1.5 mg/mL with the same buffer.
420 Katie A. Edwards et al.

9. The blocking reagent is prepared by diluting 100 mL 1% (w/v)

casein sodium salt and 3.85 mL 2 M sucrose to a final volume
of 1000 mL with HPLC grade water.
10. Sample pads (Unbound glass, Catalog #VFE, Whatman Inter-
national, Ltd., Florham Park, NJ) cut into 1.9  20 cm sections.
11. Absorbent pads (Cellulose fiber sample pads, Catalog
#CFSP223000, EMD Millipore, Billerica, MA) cut into
1.8  20 cm sections.

2.4 Assay 1. Flat bottom, non-binding microtiter plates (Corning, #3610,

Optimization Corning, NY).
Using Recombinant 2. Prior to running the assay, recombinant full length myoglobin
Myoglobin and Serum is diluted in TBS with 0.05% (w/v) Tween 20 and 0.1% (w/v)
BSA to concentrations of 100 μg/mL, 10 μg/mL, 1 μg/mL,
100 ng/mL, 10 ng/mL, 1 ng/mL, 100 pg/mL, and 0 pg/
mL. 30 μL portions are appropriate.
3. Biotinylated anti-myoglobin antibody is diluted to 10 μg/mL
with 0.05% (w/v) Tween 20 and 0.1% (w/v) BSA.
4. Streptavidin-tagged liposomes of known phospholipid
concentration.
5. Membranes with capture and control antibodies immobilized
cut into 4.5 mm wide strips.
6. Reflectometer (λ ¼ 560 nm, ESECO Speedmaster, Cushing,
OK) or scanner with image analysis software (Optional if more
than qualitative results by eye are desired).

3 Methods

3.1 Liposome A flowchart of this process is shown in Scheme 1. Note that the
Preparation times listed in this flowchart, and others in this chapter, reflect
overall times for each step, including setup and incubations. For
specific times for each step, please refer to the text.

Scheme 1 Flow chart for the preparation of dye-encapsulating liposomes

 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 421

1. Set rotary evaporator bath and sonicator bath to 45  C. Fill the

condenser on the rotary evaporator with ice and seal condenser
top rim to rotary evaporator glass housing with Parafilm.
2. Prepare dye solution as described in materials section using a
15-mL centrifuge tube, cover tube with aluminum foil, and
store in the 45  C sonicator bath.
3. DPPC, DPPG, N-glutaryl DPPE, and cholesterol
(40.9:20.1:7.3:51.7 μmol, respectively) are added to a 50-mL
round bottom flask (see Note 1).
4. A solvent mixture containing 3 mL chloroform and 0.5 mL
methanol is added and the mixture sonicated for 1 min in a
bath sonicator. We use sonication level 6 with the Aquasonic
Model 150D bath sonicator.
5. 2 mL of the 45  C sulforhodamine B dye solution is added to
the lipid mixture during the first 30 s while sonicating for a
total of 4 min. Swirl the flask manually during sonication (see
Note 2).
6. The mixture is then placed onto a rotary evaporator at the
highest rotation speed with the bath at 45  C. The vacuum
should be adjusted such that bubbling or foaming of the con-
tents does not occur. Using the Buchi R-114 rotary evaporator,
a vacuum of 500 mBar for 20 min, followed by 400 mbar for
20 min is recommended. During evaporation, return unused
portion of the SRB solution to the 45  C sonicator bath.
7. The mixture is then transiently vortexed preceding and follow-
ing an additional introduction of 2 mL 45  C 150 mM SRB.
Using a VWR mini-vortexer, we have found that holding the
flask at a 45 angle and vortexing at level 4–5 is appropriate,
however, be sure to use a stopper on the flask.
8. The mixture is returned to the rotary evaporator for an addi-
tional 30 min under slightly higher vacuum than in step 6 (350
mBar, in our case.) Generally, the evaporation is considered
complete 15 min after no further solvent is observed from the
condenser. Removal of nearly all organic solvent is required for
successful formation of liposomes.
9. While the liposome mixture is on the rotary evaporator, set the
extruder support block of the mini-extruder on a hot plate and
adjust temperature such that the temperature of the block does
not exceed 65  C.
10. Set-up the mini-extruder as outlined in the manufacturer’s
instructions. Label 2–50 mL centrifuge tubes for each extru-
sion size and clamp them in a 45  C water bath during the
extrusion process and before application to the size-exclusion
column.
422 Katie A. Edwards et al.

11. The liposomes are then extruded at 60–65  C 19 times

through 1.0 μm pore membranes, followed by 19 times
through 0.4 μm pore membranes. The liposomes must be
maintained in the 45  C water bath during and after extrusion.
12. The level of the 1HSS buffer volume on the Sephadex G-50
column prepared in step 12 of Subheading 2.1 is reduced to
just below the level of the Sephadex. Do not allow the Sepha-
dex to dry out.
13. The liposomes are pipetted onto the top of the column without
disturbing the top of the Sephadex. This addition should be
done as carefully, but rapidly, as possible using a glass Pasteur
pipet in a circular path along the inside of the glass column just
above the Sephadex. Important: the tubes containing the lipo-
somes need to be kept in 45  C at all times and the transfer to
the column needs to be efficient. If the liposomes cool to room
temperature, they will form clumps with external dye on the
column and will not separate.
14. Once all of the liposome volume has entered the column, two

1 mL aliquots of 1HSS should be pipetted onto the top of
the column allowing each one to enter the column before the
next addition. This is done to create a separation between the
liposomes/free dye and the remaining 1HSS used for lipo-
some elution.
15. After carefully pipetting another 1–2 mL of 1HSS on top of
the column, allow 1HSS to flow freely (typically
4 mL/min)
and collect the eluting liposomes in test tubes. The liposomes
will elute from the column first, forming a dark magenta band,
followed by the free dye yielding a darker and larger elution
volume. If the column is run properly, some separation of
liposomes and free dye can usually be seen.
16. The liposome containing fractions are then combined based
either visually or on their measured optical density at 532 nm.
Optical density measurements are made at 532 nm by diluting
5 μL liposomes with 995 μL 1HSS in a 1.5 mL spectrometer
cell, or, more conveniently, 1 μL liposomes in 199 μL 1HSS in a
clear microtiter plate (Corning #3795). Be selective about which
liposomes are pooled together: only the most highly concentrated
fractions should be together, the intermediates together, and if
needed, the low concentration fractions. This is very important if
the liposomes are to be subsequently coupled to proteins where a
high liposome concentration is necessary (see Note 3).
17. The combined fractions are placed into dialysis bags and dialyzed
overnight against the sucrose-HEPES-saline buffer. Be sure to
transfer liposomes from the test tubes to the dialysis bags over a
beaker to reclaim any potential spills. Replace buffer and con-
tinue dialysis until external dialysate is no longer notably pink.
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 423

Scheme 2 Flow chart for the preparation of lateral flow membranes with immobilized capture antibodies

18. Clean everything thoroughly, including removing the o-rings

from the Teflon holders and the needles and plungers from the
syringes of the extruder.
19. Store dialyzed liposomes in 15-mL centrifuge tubes at 4  C.

3.2 Membrane A flowchart of this process is shown in Scheme 2.

Preparation
1. Prepare blocking reagent (0.1% (w/v) casein, sodium salt,
0.25% (w/v) sucrose) or remove previously made blocking
reagent from the refrigerator and allow to reach ambient
temperature.
2. Cut the HF090 membrane cards into 20 cm sections (see Note 4).
3. Label each membrane card section with a line drawn
0.5 cm
from its top on the back side using a permanent marker. The
top of the HF090 membrane cards has a single covered adhe-
sive section, whereas the bottom has an adhesive section cov-
ered by two separate peel strips.
4. For each 20 cm membrane card section, prepare 50 μL of
capture antibodies at 1.0 mg/mL in 0.4 M NaHCO3/
Na2CO3, pH 9.0 containing 5% (v/v) methanol. Prepare also
50 μL of control antibodies at 1.5 mg/mL in the same solution.

3.2.1 Automated The following instructions are for operation of the CAMAG Lino-
Application of Capture mat IV, but should be adaptable for similar applicators using the
Antibodies Using a Linomat respective manufacturer’s directions.
IV or Similar TLC Plate
5. Set the Limonat IV parameters as follows:
Applicator

Plate width: 200 Space: 0

Start position: 0 Rate: 12 s/μL
Band width: 190 Volume: 38

6. Fill the syringe slowly with the capture antibody solution mix-
ture to avoid air bubbles, then insert syringe fully into syringe
holder.
424 Katie A. Edwards et al.

7. Gradually lower the plunger mechanism by pressing in the

front and rear buttons until the silver lever is just above the
syringe plunger.
8. Press the GAS button on the front of the Linomat, then lower
the silver lever onto the syringe plunger by holding down the #
key until you see liquid consistently bubble from the end of
your syringe. You may simultaneously hold down the + and #
buttons (faster) if there is a fair distance between the silver lever
and syringe plunger.
9. Place one of the 20 cm membrane sections so that is lined up
with the bottom and side markers on the black surface. Use the
supplied flat magnets to hold down the membrane at its top
and right sides. Slide the tower to 3.0 cm from the base of the
membrane card using the ruled marks.
10. Press Calc, then Run. Repeat steps for remaining membrane
sections.
11. Thoroughly wash the syringe with HPLC grade water, then
repeat steps with the anti-streptavidin solution to form a con-
trol zone 3.5 cm from the base of the membrane card following
the same procedure.

3.2.2 Manual Application If a TLC or similar applicator is not available, the capture antibody
of Capture Antibodies Using solution may be applied manually to the lateral flow membranes.
a Pipettor Here, a round spot for the capture and control zones would result,
versus the line seen in Fig. 3 generated using the Linomat for
antibody application. This procedure is useful for optimization
purposes, but is more laborious than the TLC applicator procedure
due to the need to apply antibodies and tape individual membranes.
1. Cut the membrane sections into 4.5 mm pieces. We’ve found it
beneficial to tape a laminated piece of paper to the paper cutter
with lines every 4.5 mm for a guide. Then, align one end of one
section with this paper and advance the section, making per-
pendicular cuts every 4.5 mm.
2. Using tape with the sticky side exposed placed onto a piece of
cardboard, lightly tape the top of each membrane to a piece of
clean cardboard using tweezers.
3. Apply 1 μL of the capture antibody solution to each membrane
3.0 cm from the bottom of the strip using a microvolume
pipettor. The more consistent the antibody application is, the
more consistent the LFA results will be. The antibody solution
will rapidly seep into the membrane.
4. Deposit the control zone 3.5 cm from the base of the mem-
brane card following the same procedure.
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 425

3.2.3 Membrane Drying 5. Place the membranes or membrane cards in the vacuum oven at
and Blocking Steps 40  C and set vacuum to 1500 Hg for 1.5 h. Alternatively,
membranes or membrane cards may be stored overnight in a
tightly sealed Tupperware container with the bottom filled to
0.500 with Drierite and covered with layers of large Kimwipes.
6. Remove membranes from vacuum oven or desiccated storage
container.
7. Pour blocking reagent into a plastic Tupperware container so
that it is <50% full.
8. Using tweezers, gently remove each membrane or membrane
card and place in the blocking reagent laminated side down. Be
sure to allow each membrane to gradually sink into your block-
ing reagent, otherwise blocking will be uneven. You can gently
help them along by pressing the ends under the solution.
9. Place container on shaker using slow agitation for 30 min. The
speed should be set such that the membranes are being covered
by solution and moving freely, but not so much that they
become bunched up and solution comes over the sides.
10. Using tweezers, remove each membrane or membrane section
and lie flat on three layers of large Kimwipes. Place 2–3 layers of
Kimwipes on top of membranes, then add a heavy flat object on
top of the membranes, such as a heavy lab supplies catalog.
11. Retape membranes back onto the cardboard.
12. Return the taped membranes to the vacuum oven at 25–30  C
and set vacuum to 1500 Hg for 3–4 h.
13. Remove membrane strips or sheets from cardboard and store in
vacuum sealed bags at 4  C or as above in step 17 at in sealed
Tupperware container at ambient temperature.
LFA assembly steps
14. Cut the absorbent pad into 1.8  20.0 cm sections
corresponding to the number of membrane cards prepared in
step 2.
15. Gently peel away the film covering the top section of the
HF090 card.
16. Apply the absorbent pad such that it overlaps the nitrocellulose
by 2 mm (Fig. 6).

3.0 cm 0.5 cm

4.5 mm
1.9 cm 2 mm 2 mm 1.8 cm

Fig. 6 (Left) Layout of assembled LFA membrane showing overlaps and positions of sample and absorbent
pads and antibody deposition zones. (Right) Image of assembled HF090 membrane with VFE unbound glass
sample pad and cellulose fiber absorbent pad before insertion into cassette
426 Katie A. Edwards et al.

17. Apply gentle, but even, pressure along the length of the absor-
bent pad to secure it to the underlying adhesive.
18. If the membranes are to be used for assay optimization pur-
poses only in the absence of whole blood, the bottoms may be
cut off. To cut off the bottoms, align the paper cutter such that
the lowest 0.5 mm of nitrocellulose will be removed when the
cut is made (see Note 5).
19. The bottom of the HF090 membrane cards can accommodate
both a conjugate and a sample pad. For the assays described
within, only a sample pad is used. Prior to assembly, this sample
pad is cut into 1.9  20 cm sections and blocked with 0.05%
(w/v) casein, 0.25% (w/v) sucrose for 30 min and dried in a
vacuum oven at 21  C and 1500 Hg overnight.
20. Gently peel away the film covering the bottom sections of the
HF090 card.
21. Apply the sample pad such that it overlaps the nitrocellulose by
2 mm.
22. Apply gentle, but even, pressure along the length of the sample
pad to secure it to the underlying adhesive. Excessive pressure
may crush the delicate sample pads and impede flow of the
samples.
23. The assembled membrane sheets may be cut into strips of
desired widths prior to storage as described in step 12. Our
membranes are typically cut into 4.5 mm widths.

3.3 Analysis of If available, liposome size measurements should be made using

Liposomes dynamic light scattering or alternative sub-micron particle size
measurement technique. We make liposome size distribution mea-
surements using a DynaPro LSR (Proterion Corporation, Piscat-
away, NJ) using the Dynamics (version 6.3.01) software program
and the Cumulants method of analysis [70, 71]. The Bartlett assays
are necessary for the preparation of protein or antibody-conjugated
liposomes (see Notes 6 and 7).

3.3.1 Bartlett Assays for 1. Liposome samples (20 μL) in triplicate are dehydrated at
Phospholipid Content 180  C for 10 min, then mixed and heated with 1.5 mL of
3.33 N H2SO4 for 2 h at the same temperature. Standards
prepared from potassium phosphate dibasic in deionized
water (16, 32, 64, 128, and 256 nmol phosphate per tube in
triplicate) are subjected concurrently to the same procedure.
Note that inorganic phosphates will interfere with the Bartlett
assay and need to be considered if these are used in alternate
buffers or diluents used for liposome preparation.
2. 100 μL of 30% hydrogen peroxide is added and the mixture is
returned to the oven for 1.5 h. The tubes are permitted to cool
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 427

to ambient temperature prior to, and vortexed vigorously fol-

lowing, each addition.
3. Lastly, 4.6 mL of 0.22% ammonium molybdate and 0.2 mL of
the Fiske–Subbarow reagent [72] are added.
4. The Fiske-Subbarow reagent is prepared by mixing 40 mL 15%
(w/v) sodium bisulfite, 0.2 g sodium sulfite, and 0.1 g 1-
amino-4-naptholsulfonic acid at ambient temperature for
30 min then filtering out undissolved solids. This reagent
needs to be prepared fresh at least weekly.
5. The tubes are then heated in a boiling water bath for 7 min,
then quickly cooled in an ice water bath. The color of the
highest standard should be moderate to dark blue after heating.
If it is not sufficiently dark, continue heating for several minutes
until it becomes this shade of blue.
6. The absorbance at 830 nm is recorded using a Powerwave XL
microtiter plate reader (Bio-Tek Instruments, Winooski, VT).
7. The phospholipid content of the liposomes is determined from
a calibration curve prepared from the phosphate standards
analyzed in each run. The total lipid concentration is calculated
by multiplying the phospholipid concentration by the initial
ratio of total lipid to phospholipid.

3.4 Liposome The liposome conjugation procedure assumes that the phospho-
Conjugation lipid concentration, determined as described in Subheading 3.3 or
by other means, is known. The procedure described within is has
been optimized for the conjugation of streptavidin to liposomes
and may be modified to accommodate other proteins by varying the
EDC and protein concentrations. It utilizes EDC to form a zero-
length linkage between carboxylic acid groups on the liposomes
and protein amine groups. Before beginning this procedure, ensure
that the protein to be conjugated is purified and does not contain
any stabilizing proteins or buffer components with amine groups.
1. Remove EDC from 20  C storage and allow to warm to
ambient temperature while proceeding with the following
steps.
2. Add desired volume of liposomes prepared in Subheading 3.1
to a 1.5-mL microcentrifuge tube.
3. Using the known phospholipid concentration, calculate the
nmol of total lipid present. Be sure to account for non-
phospholipid constituents such as cholesterol in your
calculation.
4. Determine the amount of protein needed for conjugation.
Typically, 0.05 mol% of the total lipid content is a reasonable
amount.
428 Katie A. Edwards et al.

5. Determine the amount of EDC needed for conjugation. 0.3

molar equivalents EDC based on the total lipid content is
optimal. Calculate how much volume of a 100 mg/mL solu-
tion of EDC is needed.
6. Add the amount of protein calculated in step 23 to the micro-
centrifuge tube containing the liposomes and gently vortex
until homogeneous.
7. Weigh the requisite amount of EDC into a clean, dry micro-
centrifuge tube.
8. Dilute EDC to a concentration of 100 mg/mL with 0.1 M
MES buffer, pH 4.65 immediately before proceeding with
step 9.
9. Add desired volume of EDC and immediately vortex the solu-
tion for approximately 30 s.
10. Incubate the mixture at room temperature in the dark for
30 min.
11. Purify the conjugated liposomes through a SEC column
packed with Sepharose CL-4B using 1 HSS as an elution
buffer [73].
12. All fractions may be combined together for subsequent phos-
pholipid determination measurements as described in Sub-
heading 3.3.

3.5 Assay The steps within this section are helpful for determining the best
Optimization Using conditions for antibody/antigen interaction and signaling. Here,
Recombinant the buffer composition, biotinylated antibody, and liposome con-
Myoglobin centrations may be varied to attain the desired level of detection.
This information can be utilized to streamline the development of
the assay in whole blood. If needed, the capture antibody concen-
tration can also be optimized during membrane preparation which
can be done easily if the manual application approach is used.
1. Add 10 μL myoglobin solutions to the bottom of wells of a flat-
bottom, non-binding microtiter plate (see Note 7).
2. Insert a 4.5 mm membrane strip with immobilized capture and
control antibodies with the base cut off and allow volume to
wick up membrane.
3. Once the solution has completely wicked, the membrane strips
are transferred to wells containing 10 μL of MESS with 0.1%
(w/v) BSA.
4. Once this solution has wicked, membranes are transferred to
wells containing a solution with 6 μg/mL biotinylated anti-
myoglobin antibodies and liposomes diluted to 250 μM
phospholipid.
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 429

5. Once this solution has wicked, membranes are then transferred

to wells containing 30 μL of MESS buffer.
6. Remove membrane once all solution has wicked and place on a
sheet of paper to dry. A dark colored sheet allows for more
contrast when results are scanned. Drying takes
10 min at
room temperature.

3.6 Assay Format In our studies, 16 mL of human venous blood was drawn into EDTA
for the Analysis vacutainer tubes by venipuncture at the Gannett Health Center at
of Myoglobin in Whole Cornell University. Participants provided informed consent for this
Blood study, which was approved by the institutional review board at Cornell
University. No participant identifier was associated with each sample.
Important: Human blood should be treated as it is potentially
infectious and proper precautions for handling and disposal should
be followed, in accordance with your organization’s protocols.
Obtaining human blood samples also requires informed consent
of your subjects, in accordance with your organization’s institu-
tional review board.
1. Fully assembled membrane strips with blocked sample pads,
immobilized capture and control antibodies, and absorbent
pads are inserted into a commercially available cassette. The
dimensions of the absorbent pad may be adjusted to ensure
that the pressure points of the cassette line up with the sample
and absorbent pad overlaps; the sample hole lines up with the
sample pad; and the test and control lines are visible.
2. Whole blood is spiked with 0–350 ng/mL myoglobin.
3. 50 μL of the spiked blood is added to 50 μL of an antibody/
liposome solution and is gently vortexed for 5 min.
4. 100 μL of the blood mixture is transferred to the sample pad.
5. Allow assay to run for 15 min. A signal proportional to the
myoglobin concentration should result (Fig. 7).

3.7 Analysis The membranes may be read by eye for qualitative results, however,
of Results semiquantitative results may be obtained if a scanner, digital cam-
era, or reflectometer is available. After the membranes have fully
dried (typically 20–30 min), scan an image of the membranes using
a computer and flatbed scanner, then use any available densitome-
try software to quantify band intensity. Alternatively, if a reflectom-
eter is available, the readings should be taken at this time following
the instrument manufacturer’s instructions. For optimal reproduc-
ibility and sensitivity, the intensity of the bands should be read as
soon as the membranes dry as it tends to decrease over time (days to
weeks) and exposure to light. While membranes may also be
scanned or read using a reflectometer when wet, care must be
taken to avoid leaving impressions in the nitrocellulose. Whether
read wet or dry, consistency in the timing of measurement is key for
optimal reproducibility.
430 Katie A. Edwards et al.

Fig. 7 Whole-blood lateral flow assay at spiked myoglobin concentrations ranging from 350 ng/mL (left) to
0 ng/mL. 50 μL blood samples were spiked with various myoglobin concentrations and added to 50 μL of an
antibody/liposome solution. This mixture was gently vortexed, and after 5 min, all 100 μL were added to the
sample pad. Results were observed after 15 min. A visible signal at the test line could be seen at a spiked
myoglobin concentration of 75 ng/mL (see Notes 8 and 9)

4 Notes

1. For the preparation of liposomes to be tagged to antibodies or

other proteins, the N-glutaryl DPPE phospholipid introduces a
carboxylic acid group to the lipid bilayer allowing for facile
conjugation to primary amine groups using 1-ethyl-3-[3-
dimethylaminopropyl]carbodiimide hydrochloride (EDC)
[73].
2. Be sure that the solution in the flask is being sonicated moder-
ately vigorously and that the flask is not located in a “dead
spot” of the bath. We use sonication level 6 with the Aquasonic
Model 150D bath sonicator.
3. If liposomes are instead formed with a cholesterol-modified
DNA probe for nucleic acid sequence detection [74], the
focus on liposome concentration is less important since no
post-formation conjugation needs to take place.
4. Here, HF090 membranes worked out well for this assay as they
provided the requisite capillary flow, performance for our ana-
lyte, and ability to secure the sample and absorbent pads.
However, blank backing cards (#MIBA-010, DCN, Carlsbad,
CA) with adhesive for the center portion may be purchased to
allow the user to apply their preferred nitrocellulose or other
analytical membrane material.
5. By cutting off a small portion of the nitrocellulose, good
contact of the nitrocellulose with the fluid during the assay is
ensured.
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 431

6. For antibody conjugation, the liposomal phospholipid concen-

tration needs to be determined using Bartlett assays [75] (see
Subheading 3.3). Both antibodies of the matched set should be
conjugated to liposomes to determine which serves as the
better detection antibody. Note that EDC is very
hydroscopic—use care to protect the reagent from moisture.
7. Alternatively, the phospholipid concentration of liposomes can
be measured by inductively coupled plasma—atomic emission
spectroscopy (ICP-AES). A volume of 20 μL of liposomes is
diluted in nitric acid (c ¼ 0.5 mol/L) to a total volume of 3 mL.
The mixture is vortexed and sonicated thoroughly. Phosphorus
standard solutions are prepared with concentrations of 1, 5, 10,
25, 50, and 100 μmol/L phosphate in 0.5 mol/L HNO3 for
calibration. The phosphorus content is then measured via ICP-
AES (at the phosphorus specific wavelength of 178.29 nm
[76].
8. No signal was observed in this assay with normal levels of
circulating myoglobin, but a visible signal was observed if a
normal blood sample was spiked with 75 ng/mL myoglobin,
which mimics an elevation outside of the normal range. Nor-
mal levels of myoglobin in human blood range from 0 to
85 ng/mL, whereas the median level of myoglobin 90 min
after an AMI has been reported to be 353 ng/mL [77]. The
level at which a signal is observed in a LFA can be altered by
optimizing the assay as noted in Subheading 3.5.
9. In any sandwich immunoassay run without a wash step, the risk
of a high-dose Hook effect needs to be checked for to avoid
underestimating the concentration of the analyte in the sample.
This can be done by making dilutions of the sample and repeat-
ing the assay to ensure that the signal decreases as expected,
rather than increases as would be the case with the Hook
Effect.

Acknowledgments

The authors gratefully acknowledge Vincent Sy and Peter Bent of

GE Healthcare for their sage advice in blood separation for LFA
applications. We also are indebted to Tara Holter, Jocelyn Jing Tan,
Kit Meyers, and Barb Leonard for their preliminary work in the
development of this assay. The authors acknowledge partial funding
through the Multistate Federal Formula Grant “Development of a
novel rapid-on-site biosensor for food safety” Project #2012-13-
132; the CD4 Initiative, Imperial College, London; and a Cornell
Engineering Learning Initiatives Undergraduate Research Award
funded by James Moore, ‘62 to Ricki Korff.
432 Katie A. Edwards et al.

References
1. Helbig JH, Luck PC, Kunz B, Bubert A (2006) membrane strip tests compared. Urology 53
Evaluation of the Duopath Legionella lateral (1):155–160
flow assay for identification of Legionella pneu- 14. http://www.brittneylimited.com/PDF%20Files/
mophila and Legionella species culture isolates. PSA-Prostrate-strip.pdf
Appl Environ Microbiol 72(6):4489–4491 15. http://www.nhdiag.com/botulism.shtml
2. Koide M, Haranaga S, Higa F, Tateyama M, 16. Sharma SK, Eblen BS, Bull RL, Burr DH,
Yamane N, Fujita J (2007) Comparative evalu- Whiting RC (2005) Evaluation of lateral-flow
ation of Duopath Legionella lateral flow assay clostridium botulinum neurotoxin detection
against the conventional culture method using kits for food analysis. Appl Environ Microbiol
Legionella pneumophila and Legionella anisa 71(7):3935–3941
strains. Jpn J Infect Dis 60(4):214–216
17. Park CH, Kim HJ, Hixon DL, Bubert A
3. http://www.kit.nl/biomedical_research/assets/ (2003) Evaluation of the duopath verotoxin
images/Lepto_lateral_flow_protocol.doc test for detection of shiga toxins in cultures of
4. Johnston SP, Ballard MM, Beach MJ, Causer human stools. J Clin Microbiol 41
L, Wilkins PP (2003) Evaluation of three com- (6):2650–2653
mercial assays for detection of Giardia and 18. Capps KL, McLaughlin EM, Murray AWA et al
Cryptosporidium organisms in fecal specimens. (2004) Validation of three rapid screening
J Clin Microbiol 41(2):623–626 methods for detection of verotoxin-producing
5. Ketema F, Zeh C, Edelman DC, Saville R, Escherichia coli in foods: interlaboratory study.
Constantine NT (2001) Assessment of the per- J AOAC Int 87(1):68–77
formance of a rapid, lateral flow assay for the 19. Smits HL, Eapen CK, Sugathan S et al (2001)
detection of antibodies to HIV. J Acquir Lateral-flow assay for rapid serodiagnosis of
Immune Defic Syndr 27(1):63–70 human leptospirosis. Clin Diagn Lab Immunol
6. Cazacu AC, Demmler GJ, Neuman MA et al 8(1):166–169
(2004) Comparison of a new lateral-flow chro- 20. Kassler WJ, Dillon BA, Haley C, Jones WK,
matographic membrane immunoassay to viral Goldman A (1997) On-site, rapid HIV testing
culture for rapid detection and differentiation with same-day results and counseling. AIDS 11
of influenza A and B viruses in respiratory spe- (8):1045–1051
cimens. J Clin Microbiol 42(8):3661–3664
21. Prevention USCfDCa (1996) Provisional pub-
7. Mokkapati VK, Sam Niedbala R, Kardos K et al lic health service recommendations for chemo-
(2007) Evaluation of UPlink-RSV: prototype prophylaxis after occupational exposure to
rapid antigen test for detection of respiratory HIV. MMWR Morb Mortal Wkly Rep
syncytial virus infection. Ann N Y Acad Sci 45:468–480
1098:476–485
22. Baeumner AJ, Jones C, Wong CY, Price A
8. Slinger R, Milk R, Gaboury I, Diaz-Mitoma F (2004) A generic sandwich-type biosensor
(2004) Evaluation of the QuickLab RSV test, a with nanomolar detection limits. Anal Bioanal
new rapid lateral-flow immunoassay for detec- Chem 378(6):1587–1593
tion of respiratory syncytial virus antigen. J
Clin Microbiol 42(8):3731–3733 23. Baeumner AJ, Pretz J, Fang S (2004) A univer-
sal nucleic acid sequence biosensor with nano-
9. van Hengel AJ, Capelletti C, Brohee M, molar detection limits. Anal Chem 76
Anklam E (2006) Validation of two commer- (4):888–894
cial lateral flow devices for the detection of
peanut proteins in cookies: interlaboratory 24. Seal J, Braven H, Wallace P (2006) Point-of-
study. J AOAC Int 89(2):462–468 care nucleic acid lateral-flow tests. IVD Tech-
nol 8:41–54
10. http://www.craigmedical.com/salivascreen6_
drug_tests.htm 25. Tomoda H, Aoki N (2000) Prognostic value of
C-reactive protein levels within six hours after
11. http://www.brittneylimited.com/PDF%20 the onset of acute myocardial infarction. Am
Files/FDA_Files/Pregnancy-strip.pdf Heart J 140(2):324–328
12. Tiplady S (2013) Lateral flow and consumer 26. Qureshi A, Gurbuz Y, Niazi JH (2012) Bio-
diagnostics. In: Wild D (ed) The immunoassay sensors for cardiac biomarkers detection: a
handbook, 4th edn. Elsevier, Oxford, U.K. review. Sens Actuators B Chem
13. Jung K, Zachow J, Lein M et al (1999) Rapid 171–172:62–76
detection of elevated prostate-specific antigen 27. Anderson JL, Adams CD, Antman EM et al
levels in blood: performance of various (2011) 2011 ACCF/AHA Focused Update
 Liposome-Enhanced Lateral-Flow Assays for Clinical Analyses 433

Incorporated Into the ACC/AHA 2007 40. Reinholt SJ, Sonnenfeldt A, Naik A, Frey MW,
Guidelines for the Management of Patients Baeumner AJ (2014) Developing new materi-
With Unstable Angina/Non-ST-Elevation als for paper-based diagnostics using electro-
Myocardial Infarction: A Report of the Ameri- spun nanofibers. Anal Bioanal Chem 406
can College of Cardiology Foundation/Amer- (14):3297–3304
ican Heart Association Task Force on Practice 41. Manis AE, Bowman JR, Bowlin GL, Simpson
Guidelines. Circulation 123(18):e426–e579 DG (2007) Electrospun nitrocellulose and
28. Brogan GX Jr, Friedman S, McCuskey C et al nylon: design and fabrication of novel high
(1994) Evaluation of a new rapid quantitative performance platforms for protein blotting
immunoassay for serum myoglobin versus CK- applications. J Biol Eng 1:2–2
MB for ruling out acute myocardial infarction 42. Matlock-Colangelo L, Baeumner AJ (2012)
in the emergency department. Ann Emerg Recent progress in the design of nanofiber-
Med 24(4):665–671 based biosensing devices. Lab Chip 12
29. McCord J, Nowak RM, McCullough PA et al (15):2612–2620
(2001) Ninety-minute exclusion of acute myo- 43. Mansfield M (2009) Nitrocellulose membranes
cardial infarction by use of quantitative point- for lateral flow immunoassays: a technical trea-
of-care testing of myoglobin and troponin I. tise. In: Wong RC, Tse HY (eds) Lateral flow
Circulation 104(13):1483–1488 immunoassay. Humana Press, New York, pp
30. Alexander LL (2016) Early prediction of acute 95–113
myocardial infarction from clinical history, 44. Bangs LI (1999) Lateral flow tests. Vol Tech
examination, and electrocardiogram in the Note 303:1–6
emergency room. Ann Emerg Med 20 45. Millipore. Rapid lateral flow test strips—
(11):1270 considerations for product development.
31. Schiff HB, MacSearraigh ET, Kallmeyer JC http://www.emdmillipore.com/US/en/
(1978) Myoglobinuria, rhabdomyolysis and products/ivd-oem-materials-reagents/lateral-
marathon running. Q J Med 47(188):463–472 flow-membranes/n6mb.qB.L0YAAAE_gut3.
32. C-Peptide ELISA Kit ab178641. www.abcam. Lxi,nav?bd¼1. Accessed 18 Mar 2016
com/C-Peptide-ELISA-Kit-ab178641.pdf. 46. Batteiger B, Newhall WJ, Jones RB (1982) The
Accessed 23 Mar 2016 use of Tween 20 as a blocking agent in the
33. Brown M (2008) Antibodies: Key to a robust immunological detection of proteins trans-
lateral flow immunoassay. In: Wong RC, Tse ferred to nitrocellulose membranes. J Immunol
HY (eds) Lateral flow immunoassay. Humana Methods 55(3):297–307
Press, New York 47. Jones KD (1999) Troubleshooting protein
34. O’Farrell B (2013) Lateral flow immunoasay binding in nitrocellulose membranes. Part I:
systems: Evolution from the current state of Principles. IVD Technol 5:32–41
the art to the next generation of highly sensi- 48. Van Dam AP, Van den Brink HG, Smeenk RJ
tive, quantitative rapid assays. In: Wild D (ed) (1990) Technical problems concerning the use
The immunoassay handbook, 4th edn. Else- of immunoblots for the detection of antinu-
vier, Oxford clear antibodies. J Immunol Methods 129
35. O’Farrell B, Bauer J (2006) Developing highly (1):63–70
sensitive, more-reproducible lateral-flow 49. Zampieri S, Ghirardello A, Doria A et al (2000)
assays. Part 1: New approaches to old pro- The use of Tween 20 in immunoblotting assays
blems. IVD Technol 10(6):11–15 for the detection of autoantibodies in connec-
36. Hans H, Beer EJ, Pflanz K, Klewitz TM (2002) tive tissue diseases. J Immunol Methods 239
Qualification of cellulose nitrate membranes (1–2):1–11
for lateral-flow assays. IVD Technol 8:35–42 50. Smits HL, Chee HD, Eapen CK et al (2001)
37. Harlow E, Lane D (1988) Antibodies: a labo- Latex based, rapid and easy assay for human
ratory manual. Cold Spring Harbor, Cold leptospirosis in a single test format. Trop Med
Spring Harbor Laboratories Int Health 6(2):114–118
38. Jones K (2001) Membrane immobilization of 51. Carney J, Braven H, Seal J, Whitworth E
nucleic acids. Part 2: Probe attachment techni- (2006) Present and future applications of gold
ques. IVD Technol pp 1–7 in rapid assays. IVD Technol 12:41–49
39. Edwards KA, Baeumner AJ (2006) Optimiza- 52. Chan CP, Sum KW, Cheung KY et al (2003)
tion of DNA-tagged dye-encapsulating lipo- Development of a quantitative lateral-flow
somes for lateral-flow assays based on assay for rapid detection of fatty acid-binding
sandwich hybridization. Anal Bioanal Chem protein. J Immunol Methods 279
386:1335–1343 (1–2):91–100
434 Katie A. Edwards et al.

53. Wang S, Zhang C, Wang J, Zhang Y (2005) sandwich-hybridization assays. Anal Biochem
Development of colloidal gold-based flow- 244(2):260–269
through and lateral-flow immunoassays for 65. Rongen HAH, Bult A, vanBennekom WP
the rapid detection of the insecticide carbaryl. (1997) Liposomes and immunoassays. J
Anal Chim Acta 546(2):161–166 Immunol Methods 204(2):105–133
54. Klewitz T, Gessler F, Beer H, Pflanz K, Scheper 66. Edwards KA, Baeumner AJ (2013) Periplasmic
T (2006) Immunochromatographic assay for binding protein-based detection of maltose
determination of botulinum neurotoxin type using liposomes: a new class of biorecognition
D. Sens Actuat Chem 113(2):582–589 elements in competitive assays. Anal Chem 85
55. Shyu RH, Shyu HF, Liu HW, Tang SS (2002) (5):2770–2778
Colloidal gold-based immunochromato- 67. Chen CS, Baeumner AJ, Durst RA (2005) Pro-
graphic assay for detection of ricin. Toxicon tein G-liposomal nanovesicles as universal
40(3):255–258 reagents for immunoassays. Talanta 67
56. Greenwald R, Esfandiari J, Lesellier S et al (1):205–211
(2003) Improved serodetection of Mycobacte- 68. Plant AL, Brizgys MV, Locasio-Brown L,
rium bovis infection in badgers (Meles meles) Durst RA (1989) Generic liposome reagent
using multiantigen test formats. Diagn Micro- for immunoassays. Anal Biochem 176
biol Infect Dis 46(3):197–203 (2):420–426
57. Birnbaum S, Uden C, Magnusson CG, Nilsson 69. Edwards KA, Baeumner AJ (2007) DNA-
S (1992) Latex-based thin-layer immunoaffi- oligonucleotide encapsulating liposomes as a
nity chromatography for quantitation of pro- secondary signal amplification means. Anal
tein analytes. Anal Biochem 206(1):168–171 Chem 79(5):1806–1815
58. Corstjens P, Zuiderwijk M, Brink A et al 70. Koppel DE (1972) Analysis of macromolecular
(2001) Use of up-converting phosphor repor- polydispersity in intensity correlation
ters in lateral-flow assays to detect specific spectroscopy—method of Cumulants. J Chem
nucleic acid sequences: a rapid, sensitive DNA Phys 57(11):4814–4820
test to identify human papillomavirus type 16 71. Frisken BJ (2001) Revisiting the method of
infection. Clin Chem 47(10):1885–1893 cumulants for the analysis of dynamic light-
59. LaBorde R, O’Farrell B (2002) Paramagnetic- scattering data. Appl Opt 40(24):4087–4091
particle detection in lateral-flow assays. IVD 72. Fiske CH, Subbarow Y (1925) The colorimet-
Technol 8(3):36–41 ric determination of phosphorus. J Biol Chem
60. Rule GS, Montagna RA, Durst RA (1996) 66(2):375–400
Rapid method for visual identification of spe- 73. Edwards KA, Curtis KL, Sailor J, Baeumner A
cific DNA sequences based on DNA-tagged (2008) Universal liposomes: preparation and
liposomes. Clin Chem 42(8):1206–1209 usage for the detection of mRNA. Anal Bioanal
61. Szoka F, Olson F, Heath T, Vail W, Mayhew E, Chem 391(5):1689–1702
Papahadjopoulos D (1980) Preparation of uni- 74. Edwards KA, Baeumner AJ (2009) Liposome-
lamellar liposomes of intermediate size enhanced lateral-flow assays for the sandwich-
(0.1–0.2 mumol) by a combination of reverse hybridization detection of RNA. Methods Mol
phase evaporation and extrusion through poly- Biol 504:185–215
carbonate membranes. Biochim Biophys Acta
601(3):559–571 75. Bartlett GR (1959) Phosphorus assay in col-
umn chromatography. J Biol Chem 234
62. Szoka F Jr, Papahadjopoulos D (1978) Proce- (3):466–468
dure for preparation of liposomes with large
internal aqueous space and high capture by 76. Fenzl C, Hirsch T, Baeumner AJ Liposomes
reverse-phase evaporation. Proc Natl Acad Sci with high refractive index encapsulants as tun-
U S A 75(9):4194–4198 able signal amplification tools in surface Plas-
mon resonance spectroscopy. Anal Chem 87
63. Edwards KA, Baeumner AJ (2006) Sequential (21):11157–11163
injection analysis system for the sandwich
hybridization-based detection of nucleic acids. 77. Sallach SM, Nowak R, Hudson MP et al
Anal Chem 78(6):1958–1966 (2004) A change in serum myoglobin to detect
acute myocardial infarction in patients with
64. Rule GS, Montagna RA, Durst RA (1997) normal troponin I levels. Am J Cardiol 94
Characteristics of DNA-tagged liposomes (7):864–867
allowing their use in capillary-migration,
 Chapter 26

Development of Dual Quantitative Lateral Flow

Immunoassay for the Detection of Mycotoxins
Yuan-Kai Wang, Ya-Xian Yan, and Jian-He Sun

Abstract
Lateral flow immunoassays have been widely used in recent years for detection of toxins, heavy metals, and
biomarkers. To improve the efficiency of individual lateral flow immunoassays, multiplex analytical strips
play an important role in the detection of several important analytes. In this chapter, development of a dual
lateral flow immunoassay is presented for detection of a variety of low molecular weight molecules. Various
buffers, additives, and materials are introduced and evaluated. Depending on the analyte to be tested, the
technique allows for selection of optimum buffers, additives, and other materials.

Key words Lateral flow, Gold nanoparticles, Fumonisin B1, Deoxynivalenol, Aflatoxin B1

1 Introduction

With the advantages of speed, low cost, and ease of use, lateral flow
immunoassays can be developed to detect toxins, biomarkers,
drugs, etc., and are especially useful for point-of-care applications
[1]. Like other immunoassays, the principle of lateral flow immu-
noassays includes antigen–antibody and antibody–antibody (such
as mouse antibody and goat anti-mouse antibody) binding. More-
over, capillary action is employed to induce fluid flow through the
device without need of further mechanical or human intervention.
Signal detection is provided via gold nanoparticles or quantum dot
fluorescence. For most lateral flow immunoassays minimal training
is necessary and results can be observed in 20 min. These assays will
help address future needs in area as diverse as environmental moni-
toring, biomarker assays, and identification of toxins and drugs.
However, qualitative lateral flow immunoassays still lack ade-
quate sensitivity for many applications. Current detection limits
range from 1–10 ng/mL. Moreover, this immunoassay provides
only “yes or no” qualitative results. The future challenge of

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0_26,
© Springer Science+Business Media LLC 2017

435
436 Yuan-Kai Wang et al.

Table 1
Comparison between competitive and sandwich lateral flow immunoassay

Indirect competitive lateral flow

Differences of strip immunoassay Sandwich lateral flow immunoassay
Conjugation pad Colloidal gold–monoclonal antibody Colloidal gold–monoclonal antibody 1#
conjugate conjugate
Test line Analyte–protein conjugate Monoclonal antibody 2#
Detect analytes Low-weight molecules Proteins, bacteria, etc.
Result judgment of Red line indicates negative Red line indicates positive
test line

analytical methods such as this will be to provide more sensitive,

quantitative, and accurate results.
Based on the measurement of strip reader device for the test
line and control line, quantitative lateral flow immunoassay can be
developed to determine certain concentrations of analytes [2].
Also, different analytes can be detected simultaneously in multiplex
quantitative lateral flow immunoassay for the high through-put
screening, which can save time, cost, and human power.
According to the different analytes, lateral flow immunoassay
can be divided into competitive analysis [3] and sandwich analysis
[4] (see Table 1). Competitive lateral flow immunoassays are
employed to detect the low-weight molecules (toxins, heavy
metals, drugs, etc.). In most competitive lateral flow assays, the
competition is between the migrating target analyte in the sample
and the immobilized analyte on the strip (capture antigen) for the
binding to the migrating labeled antibody. The degree of inhibition
of migrating labeled antibody binding to the capture antigen is
proportional to the level of the migrating target analyte in the
sample.
To form a strip for lateral flow immunoassay, several essential
parts including sample pad, conjugation pad, absorption pad, nitro-
cellulose (NC) membrane, and a sticky PVC baseplate are needed
(see Fig. 1). A sticky baseplate is used as the framework, and other
pads can be readily pasted onto it. Because the mycotoxins are low
molecular weight molecules, competitive lateral flow immunoassay
is selected in this study. To assemble the competitive test strip, the
antigen conjugates/secondary antibody-coated NC membrane is
first pasted onto the sticky side of PVC baseplate. Next, the conju-
gation pad (0.5
0.5 cm) sample pad (2
0.5 cm) and absorption
pad (3
0.5 cm) are pasted to the lower and upper portions of the
baseplate, with 1–2 mm overlap to the NC membrane on both
sides. For the reactant in each part of the strip, colloidal gold–anti-
body (such as mouse monoclonal antibody from against seleted
 Development of Dual Quantitative Lateral Flow Immunoassay. . . 437

Fig. 1 Schematic diagram of the quantitative lateral flow dual immunoassay

analytes) is added on the conjugation pad and pretreated for dehy-

dration. Analyte–protein conjugate is immobilized in a line onto
the NC membrane as the test line. A secondary antibody (such as
goat anti-mouse antibody) is immobilized in a line onto the NC
membrane as the control line, which can react with the antibody in
the colloidal gold solution. When the sample solution is added on
the sample pad, the solution will flow on the strip in the direction of
arrow under capillary action, which passes through the conjugation
pad, NC membrane, and absorption pad in sequence. Normally the
colloidal gold and colloidal gold–antibody solutions are red, thus if
reaction occurred on the test line or control line, a red color can be
observed. To judge the positive or negative samples, if the extract
contains no analytes (negative sample), the dehydrated colloidal
gold–monoclonal antibody will be resolved and flow with the
extract under capillary action. Then the mixture will react with
the analyte–protein conjugate to form a complex. Thus the red
color can be observed in the test line. The fluid moves forward
continuously and reacts with the secondary antibody to form a new
complex, and the clear red line can be observed in the control line
(see Fig. 2). When the extract contains analyte (positive sample), the
colloidal gold–monoclonal antibody will first bind with the analyte.
Then the complex of the colloidal gold–monoclonal antibody will
flow over the strip. Because the antibody has already bound with
the analyte, the colloidal gold–monoclonal antibody cannot bind
with the analyte–protein conjugate in the test line, and no color will
be observed. But the antibody in colloidal gold solution can still
438 Yuan-Kai Wang et al.

Fig. 2 Results judgment and reactions in test line of indirect competitive and sandwich lateral flow
immunoassay

bind with secondary antibody in the control line, and red color will
be observed on the control line (see Fig. 2).
The sandwich lateral flow immunoassays are mainly used to
detect large molecule analytes, such as proteins, bacteria, and bio-
markers. For each analyte, the monoclonal antibodies 1# and 2# are
the antibodies against the different epitopes of the analyte. Nor-
mally, the monoclonal antibody conjugated with colloidal gold
nanoparticles named 1#, and the monoclonal antibody is immobi-
lized in a line on the NC membrane as the test line named 2#. The
anaylte in the extract of samples will be reacted with colloidal
gold–monoclonal antibody 1# conjugate firstly, and continuously
reacted with the monoclonal antibody 2# on the test line. Thus a
clear red line can be observed, which indicate the positive result (see
Fig. 2).
To our knowledge, the uses of inappropriate buffers for lateral
flow immunoassay will induce instability of the gold nanoparticle-
labeled antibodies and affect sensitivity (see Note 1). The intensities
of the test lines, stability and release rate of the mixtures (gold
nanoparticle-labeled antibodies and sample extraction), and detec-
tion sensitivity are markedly affected by the types of buffer and
concentrations of their components. Moreover, in different lateral
flow immunoassays (different analytes, antibodies, or materials),
the optimum buffers may be different. To optimize the immunoas-
say, different buffers, additives, surfactants (see Note 2) would be
evaluated. The optimum buffers and concentrations of components
are needed to determine for lower detection limits, balance of
 Development of Dual Quantitative Lateral Flow Immunoassay. . . 439

intensities of the test and control lines, and stability of the gold
nanoparticle-labeled antibodies. The buffers in lateral flow immu-
noassay are mainly PBS and borate buffer (from 2 mM to 100 mM),
and the pH values are ranges from 6.5 to 8.2. The surfactant is also
an essential component for the conjugation/sample pad pretreat-
ing buffer, and it will affect the color of lines and the sensitivity.
Different types of surfactants (Tween 20, Triton X-100, Tetronic
1307, etc.) can be evaluated, and selected in the certain lateral flow
immunoassay.

2 Materials

1. Tetrachloroauric acid (Sigma Chemical, St. Louis, MO, USA).

2. The monoclonal antibodies against zearalenone (2C9, mAb-
ZEN) and fumonisin B1 (6H3, mAb-FB1) are prepared in our
laboratory based on the hybridoma approach [5].
3. Bovine serum albumin (BSA, Sangon Biotech, Shanghai,
China).
4. Ovalbumin (OVA, Sangon Biotech, Shanghai, China).
5. ZEN-BSA and FB1-OVA conjugates are prepared in our
laboratory.
6. Goat anti-mouse antibody (ICL, Portland, OR, USA).
7. Tetronic 1307 (Pragmatic co. ltd, Elkhart, IN, USA).
8. NC membrane (Millipore 180, from Millipore, Bedford, MA,
USA).
9. Glass fiber (8964, from Ahlstrom, Helsinki, Finland).
10. Glass fiber (SB08, from Jiening Biotech, Shanghai, China).
11. Strip reagent dispenser (XYZ-3060, from Bio-Dot, Irvine, CA,
USA).
12. Strip reader (CHR-110R, from Kaiwood, Taiwan, China).
13. MX2 shaker (Finepcr, Gyeonggi-do, Korea).
14. Incubator (DHP-9082, from Yiheng, Shanghai, China).
15. Centrifuge (5418, from Eppendorf, Hamburg, Germany).
16. Spectrophotometer (UV-1800, from Shimadzu, Tokyo,
Japan).
17. Transmission electron microscope system (JEM 2100, from
JEOL, Tokyo, Japan).
Other reagents such as sodium citrate, potassium carbonate,
sodium chloride, sodium phosphate, sodium dihydrogen phos-
phate, trehalose, sucrose, sodium azide, potassium chloride, mono-
potassium phosphate, Tween 20, methanol are purchased from
Sinopharm, Shanghai, China.
440 Yuan-Kai Wang et al.

3 Methods

3.1 Preparation The flow chart of methods in this study is shown on Fig. 3. Firstly,
of Colloidal Gold 100 mL tetrachloroauric acid solution (0.01%, w/v) is stirred and
Nanoparticles (25 nm) heated until boiling (see Note 3) in conical flask. A volume of
0.7 mL 1% sodium citrate (w/v) is then added to the solution
quickly [6] (see Note 4). In the next 2 min, the color of mixture
is changed from transparent to dark, then clear wine red. Decreas-
ing the power to gentle heating for 5 min, then the mixture is
cooled down to room temperature and stored at 4  C. The diame-
ter of nanoparticles is determined by the wavelength scanning and
transmission electron microscope (see Fig. 4). The maximum
absorbing wavelength is 525 nm for the 25 nm colloidal gold
nanoparticles, based on the equation relating the gold nanoparticle
size X (nm) and the maximum absorption wavelength Y (nm):
Y ¼ 0.4271X + 514.56 [7]. In transmission electron microscopy
the mean diameter of gold nanoparticles is 24.8  3.2 nm as
measured by averaging a random selection of 100 particles.

Preparation of colloidal gold nanoparticles

Monoclonal antibodies against analytes

Concentration of antibodies
Preparation of colloidal gold-labeled monoclonal antibodies
Optimization pH of colloidal gold solution

Analytes-protein conjugates Concentration of colloidal

gold-labeled monoclonal antibodies

Preparation of lateral flow immunoassay Concentration of analytes-protein

Optimization conjugates

Types, pH and components of

Standard solution of analytes buffers

Preparation of standard curves

Sample preparation

Determine the content of analytes

Fig. 3 Flow chart of quantitative lateral flow immunoassay in this study. The optimization steps are needed for
each lateral flow immunoassay
 Development of Dual Quantitative Lateral Flow Immunoassay. . . 441

a b
0.5

0.4
Absorbance

0.3

0.2

0.1

0
400 450 500 550 600
100 nm
Wavelength (nm)

Fig. 4 Measurement of 25 nm colloidal gold nanoparticles by wavelength scanning (a) and transmission
electron microscope (b)

3.2 Preparation The monoclonal antibody mAb-ZEN is evaluated and showed no

of Colloidal Gold- cross-reactivity with FB1, deoxynivalenol, or aflatoxin B1, which
Labeled Monoclonal usually coexist in samples. The monoclonal antibody mAb-FB1 is
Antibodies also evaluated and showed no cross-reactivity with ZEN, deoxyni-
valenol, and aflatoxin B1. To optimize the binding between colloi-
dal gold particles and antibodies, colloidal gold solutions at
different pH and antibody concentrations are evaluated as follows
(see Note 5). The pH values of colloidal gold solutions (1 mL) are
adjusted by 0.2 M potassium carbonate to 6.0, 6.5, 7.0, 7.5, 8.0,
8.5, and 9.0. 0.1 mg/mL of antibody in 100 μL 2 mM borate
buffer (pH 7.4) is added to each colloidal gold solution. After
vortexing for 10 min, let the mixture sit for 10 min and room
temperature. Then 100 μL of 10% sodium chloride are added to
each tube. After vortexing for 10 min, the mixtures are incubated at
room temperature for 30 min. After adding sodium chloride, the
coagulation of colloidal gold–monoclonal antibody will be
observed in the improper pH values. Finally, the solutions are
determined by wavelength scanning respectively, and the highest
optical density of certain pH value solution in 525 nm is selected as
the optimum pH value.
Different concentrations of antibodies are also optimized. The
optimum pH which evaluated in the previous method is employed.
Different concentrations of antibodies mAb-ZEN (0.5, 1, 2, 3, 4, 5,
and 6 μg/mL) or mAb-FB1 (3, 4, 5, 6, 7, 8, and 9 μg/mL) in
100 μL 2 mM borate buffer (pH 7.4) are prepared and added to each
colloidal gold solution. After vortexing for 10 min, let the mixture sit
for 10 min and room temperature. Then 100 μL of 10% sodium
chloride are added to each tube. After vortexing for 10 min, the
mixtures are incubated at room temperature for 30 min. At last, the
solutions are determined by wavelength scanning respectively and
the highest optical density in 525 nm, and relative lower concentra-
tion (to improve the sensitivity) of certain antibody solution is
selected as the optimum antibody concentration.
442 Yuan-Kai Wang et al.

To prepare the colloidal gold-labeled monoclonal antibodies,

1.0 mL of 2 mM borate buffer (pH 7.4) containing mAb-ZEN
(3 μg/mL) or mAb-FB1 (7 μg/mL) is added slowly with 1-min
duration to 10 mL colloidal gold solution (pH 6.5 for mAb-ZEN
or pH 7.0 for mAb-FB1). After gentle stirring for 30 min, 1.0 mL
10% BSA (w/v) is added to the mixture, which is again stirred for
another 30 min. The mixture is then centrifuged at 1500
g for
20 min. The supernatant sample is collected and centrifuged at
8000
g for 30 min. Finally, the colloidal liquid is washed three
times by adding 10 mL of 2 mM borate buffer (pH 7.4) and
centrifuging at 6000 g for 30 min. The resulting colloidal gold-
labeled antibodies are then stored in 2 mM borate buffer (pH 7.4)
containing 6% trehalose (w/v), 4% sucrose (w/v), 1% BSA (w/v),
and 0.05% sodium azide (w/v) at 4  C. To evaluate the quality of
colloidal gold-labeled monoclonal antibodies, 2 μL 0.1 mg/mL
goat anti-mouse antibody is added on the NC membrane firstly.
After dried at room temperature, 2 μL of colloidal gold-labeled
monoclonal antibodies are added on the NC membrane which
located lower than the goat anti-mouse antibody respectively.
At last, 50 μL ddH2O is added on the low side of the NC mem-
brane. Red dot is observed at the goat anti-mouse antibody point,
which indicated the well quality of colloidal gold-labeled monoclo-
nal antibodies.

3.3 Preparation of Mycotoxin-protein conjugates (ZEN-BSA and FB1-OVA) are

Lateral Flow prepared firstly. To improve the sensitivity of detection, the con-
Immunoassay centrations of coated antigens (0.05 mg/mL for ZEN-BSA, and
0.2 mg/mL for FB1-OVA) are optimized by checkerboard titra-
tion test which the details are showed as follows (see Fig. 5).
Different concentrations of ZEN-BSA (0.025, 0.05, 0.1,
0.2 mg/mL) or FB1-OVA (0.05, 0.1, 0.2, 0.4 mg/mL) in
50 mM phosphate buffer saline (PBS, pH 7.4) with 7% methanol
(w/w) are prepared. Then the dispensed volume is set to
1.0 μL/cm, which means that 1.0 μL of solution is dispensed
onto each centimeter of NC membrane. Each of the above solu-
tions is dispensed onto the NC membrane to form the test lines in
each strip. The conjugation pad made of glass fiber (8964) is dipped
in and pretreated by 10 mM PBS (pH 7.4) containing 6% trehalose
(w/v), 1% BSA (w/v), 0.5% Tetronic 1307 (w/v), and 0.05%
sodium azide (w/v), and stored at 4  C after freeze dehydration.
The pad is cut into 0.5
0.5 cm sections, and 10 μL different
dilution ratio of colloid gold labeled antibody (diluted by 20 mM
borate buffer, pH 8.2, containing 6% trehalose, 1% BSA, and 0.05%
sodium azide) is added and dehydrated under vacuum (see Note 6).
The sample pad made of glass fiber (SB08) is dipped in and
pretreated by 50 mM borate buffer (pH 7.4) containing 5% treha-
lose, 1% BSA, 0.1% Tetronic 1307, and stored at 4  C after freeze
dehydration. Finally, 150 μL ddH2O is added on the sample pad.
 Development of Dual Quantitative Lateral Flow Immunoassay. . . 443

Fig. 5 Different concentrations of coating ZEN-BSA and FB1-OVA conjugates and colloidal gold labeled
antibody in lateral flow immunoassay for ZEN (a) and FB1 (b)

The fluid will pass through the sample pad, conjugation pad, NC
membrane, and absorption pad under capillary action. Clear red
lines (test lines) can be observed if the concentration of coated
antigen and colloidal gold-labeled monoclonal antibody are
enough, and selected as the optimum coating concentration.
Based on the optimization of individual test strip, dual lateral
flow strip is developed subsequently. ZEN-BSA (0.05 mg/mL) and
FB1-OVA (0.2 mg/mL) are dispensed onto the NC membrane as
two test lines (see Fig. 1). The goat anti-mouse antibody (0.07 mg/
mL) is also dispensed onto the NC membrane (1.0 μL/cm) to form
the control line, positioned at 0.5 cm above the Test 2 line. Mix-
tures of the two colloidal gold-labeled antibodies (mAb-ZEN and
mAb-FB1) are prepared at different ratios (1:9, 1.5:8.5, 2:8,
2.5:8.5, and 3:7, v/v) to acquire similar product intensities in the
Test 2 and Test 1 lines. A low concentration of colloidal gold-
labeled antibodies results in higher sensitivity, but narrows the
detection range. Thus, the use of an optimum dilution ratio
could balance the sensitivity and detection range in the lateral
flow immunoassay. The optimum ratio of the gold nanoparticles-
mAb-ZEN and gold nanoparticles-mAb-FB1 is 1.5:8.5 (v/v). Fur-
thermore, different dilution ratios (1:1.5, 1:2, 1:2.5, 1:3, 1:3.5,
and 1:4, v/v) of these mixtures are assessed using the 20 mM
borate buffer (pH 8.2) containing 6% trehalose (w/v), 1% BSA
(w/v), and 0.05% sodium azide (w/v). The optimum dilution ratio
of the colloidal gold-labeled antibodies mixture is 1:2.5 (v/v) in
the colloidal gold–monoclonal antibody dilution buffer. Then
10 μL mixtue is added on the pretreated conjugation pad and
444 Yuan-Kai Wang et al.

dehydrated under vacuum. The assembly of dual lateral flow immu-

noassay is similar with the individual method which introduced
previously.

3.4 Test Procedure Twofold dilution series of ZEN (from 25 ng/mL) or FB1 (from
500 ng/mL) are prepared in 50 mM PBST. 250 μL standard
solutions are tested to determine the detection limit and ranges of
the strip. During reaction in the test strip, ZEN or FB1 in the
mixture would compete for binding with the specific antibodies
gold nanoparticles–monoclonal antibody for ZEN or gold nano-
particles–monoclonal antibody for FB1 against the coating antigen
conjugates, thus altering the reaction intensity on the test lines
(Test 1 or Test 2 lines, respectively, see Fig. 6).
The intensity of the test line and control lines are captured with
the strip reader after 20 min of reaction. Concentrations of ZEN or
FB1 in the samples are quantified from the dose–response curves
(band intensity vs concentrations of ZEN or FB1 in the standard
solutions) which are run simultaneously in triplicate. The strip
reader results from different concentrations of ZEN and FB1 are
used to determine the calibration curves of ZEN and FB1 (see
Fig. 7, by Origin 6.0, from OriginLab, Northampton, MA,
USA). Relative standard deviations for different samples are deter-
mined and normally need to be lower than 15%. After extraction

Fig. 6 Lateral flow dual immunoassay strips for fumonisin B1 (top) and zearalenone (bottom). The concentra-
tions of FB1 from 1 to 9 (top): 500, 250, 125, 62.5, 31.3, 15.6, 7.81, 3.91 and 0 ng/mL, and those of ZEN from
1 to 9 (bottom): 25, 12.5, 6.25, 3.13, 1.56, 0.781, 0.391, 0.195 and 0 ng/mL
 Development of Dual Quantitative Lateral Flow Immunoassay. . . 445

Fig. 7 Calibration curves of zearalenone (panel a) and fumonisin B1 (panel b) in

the lateral flow dual immunoassay. The X-axes are the log concentrations of ZEN
or FB1. The Y-axes are the ratio of the relative optical density of the test line to
the control line, a ratio that represents the degree of competitive inhibition. The
detection range is 0.94–7.52 ng/mL for ZEN and 9.34–100.45 ng/mL for FB1.
The error bars indicate the standard deviation

from the spiked and natural samples, the dilution ratios (1:2.5, 1:5,
1:10, 1:15, and 1:20, v/v) of the sample extracts are also optimized
in order to decrease the methanol content in the extraction buffer
and the influence of proteins or other components in the sample
matrixes. The spiked and natural samples are determined for ZEN
or FB1 by quantitative lateral flow immunoassay and LC-MS/MS.
The recovery rates of the spiked samples are compared and the
446 Yuan-Kai Wang et al.

correlation between lateral flow immunoassay and LC-MS/MS

results is investigated using linear regression (SPSS software,
IBM, New York, USA).

4 Notes

1. PBS could result in the appearance of deposits in gold nano-

particle–antibody conjugation. Borate buffer could reduce line
intensities when used for coating of the antigens or antibody.
Borate buffer is more suitable than PBS for use in conjugation,
dilution and storage of gold nanoparticle-labeled antibodies.
Trehalose at 6% (w/v) is better than sucrose for stability of the
gold nanoparticle-labeled antibodies during storage. BSA is
essential for gold nanoparticle-labeled antibody storage and
dilution.
2. To optimize lateral flow immunoassay performance, PBS and
borate buffers with four different molar concentrations (2, 10,
20, and 50 mM) and two different pH values (7.4 and 8.2) are
assessed. The buffers for antigen or antibody coating and sam-
ple pad pretreating as well as for conjugation, storage, and
dilution of gold nanoparticle–antibodies are investigated.
Types and concentrations of components in the buffers, i.e.,
BSA (0, 0.5%, and 1%, w/v), sucrose (0, 2%, 4%, 6%, 8%, and
10%, w/v), and trehalose (0, 2%, 4%, 6%, 8%, and 10%, w/v),
are also investigated by assessing the detection limits of the test
strip and the stability of gold nanoparticles-labeled antibodies.
3. The quality of solvent (ddH2O) is essential for the preparation
of gold nanoparticles.
4. The volume of 1% sodium citrate can be adjusted to prepare
different diameters of gold nanoparticles. Normally, the more
the volume of 1% sodium citrate, the smaller the nanoparticles
prepared.
5. For the different diameters of gold nanoparticles and antibo-
dies, the optimum pH of gold nanoparticles and concentra-
tions of antibodies are different, but normally from 6.0 to 9.0.
Thus a series of dilutions from 6.0–9.0 (such as 6.0, 6.5, 7.0,
7.5, 8.0, 8.5, 9.0) can be used for other different diameters and
antibody optimizations.
6. Not every colloidal gold-labeled monoclonal antibody works
well after the dehydration (freeze or evaporation). The
researchers can add the liquid colloidal gold-labeled monoclo-
nal antibody solution on the sample pad directly. But the
sample pad is needed to be pretreated by 10 mM PBS
(pH 7.4) containing 6% trehalose (w/v), 1% BSA (w/v), 0.5%
Tetronic 1307 (w/v), and 0.05% sodium azide (w/v), and
stored at 4  C after freeze dehydration.
 Development of Dual Quantitative Lateral Flow Immunoassay. . . 447

References
1. Quesada-González D, Merkoçi A (2015) paper microfluidics. Proc Natl Acad Sci U S A
Nanoparticle-based lateral flow biosensors. Bio- 111(10):3671–3676
sens Bioelectron 73:47–63 5. Wang YK, Wang J, Wang YC et al (2011) Prepa-
2. Feng S, Caire R, Cortazar B et al (2014) Immu- ration of anti-zearalenone monoclonal antibo-
nochromatographic diagnostic test analysis dies and development of an indirect
using Google Glass. ACS Nano 8(3): competitive ELISA for zearalenone. Microbiol-
3069–3079 ogy China 38(12):1793–1800
3. Kolosova AY, Sibanda L, Dumoulin F et al 6. Frens G (1973) Preparation of gold dispersions
(2008) Lateral-flow colloidal gold-based immu- of varying particle size: controlled nucleation for
noassay for the rapid detection of deoxynivalenol the regulation of the particle size in monodis-
with two indicator ranges. Anal Chim Acta 616 perse gold suspensions. Nat Phys Sci 241:20–22
(2):235–244 7. Chen X, Liu S (2004) Colloidal gold labeling
4. Warren AD, Kwong GA, Wood DK et al (2014) immunoassay and its application in rapid detec-
Point-of-care diagnostics for noncommunicable tion of small molecule. Pharm Biotechnol
diseases using synthetic urinary biomarkers and 11:278–280
 INDEX

A UV/Ozone plasma treatment ................................ 183

Biorecognition............................................................... 164
Acousto-optical transmission filter (AOTF) Biosensors
microscope ................................................... 223 characteristics........................................................... 328
Acute myocardial infarction (AMI).............................. 408
HBV. see Hepatitis B virus (HBV)
Affinity biosensor ............................................... 78, 84–86 fabrication................................................................ 328
Aflatoxin B1................................................................... 441 Bovine serum albumin (BSA).............101, 103–105, 415
Angle-resolved surface plasmon resonance .................. 74,
79, 80, 82 C
Anti-biotin ..................................................................... 413
Aptazymes............................................................. 390, 393 Carboxylic acid .............................................................. 430
Artificial urine................................................................ 330 CCD camera detection, HIV ....................................... 230
Au nanodisk array .............................................. 4, 7–9, 13 camera setup ............................................................ 375
HIV Infectivity Assay ..................................... 379, 381
B image capture ................................................. 375, 377
imaging apparatus ................................................... 374
Bimodal waveguide (BiMW) interferometry P4R5 cells ................................................................ 381
antibody concentration........................................... 183
QCapture Pro capture software ............................. 377
antigen/antibody interaction ................................. 183 quantify .................................................................... 372
biofunctionalization ....................................... 174–179 Romanizer program .............................. 377, 379, 380
buffer composition......................................... 166–167
CCDs. See Charged-coupled devices (CCDs)
calibration curve ............................................. 174, 176 CD4–CD8 ratio ................................................... 222, 226
data analysis .................................................... 182, 184 CD8-Qdot565 .............................................................. 226
environmental monitoring, medical diagnostics,
Cell monolayer wound healing assay ..........................152,
and food safety ............................................ 161 155, 156
equipment................................................................ 165 Charged-coupled devices (CCDs) ...................... 226, 236
fabrication....................................................... 171–173 assay plate fabrication.............................................. 243
hGH ....................................................... 168, 169, 181
biological and cell culture regents.......................... 239
homogeneous sensing............................................. 174 cell culture ............................................................... 239
hydrochloric acid..................................................... 183 cell transduction ...................................................... 240
immunoassay
and CMOS ..................................................... 234, 235
antibody concentration.............................179–180 computer control and data analysis........................ 239
hGH ................................................................... 182 configuration ........................................................... 241
Irgarol 1051 ............................................. 180, 181
digital detectors....................................................... 235
integrated optics (IO) .................................... 162–164 EL illumination ....................................................... 236
Irgarol 1051 ................................................... 168, 169 fluorescence detector ..................................... 241, 243
light, laser diode...................................................... 167
fluorescence emission.............................................. 248
optimum dimensions .............................................. 167 fluorometer.............................................................. 242
photodetector................................................. 167, 168 high-titer viral stocks ..................................... 239, 240
photonic biosensors ................................................ 162 image and data analysis ........................................... 242
polymer-based biosensors ....................................... 182
image capturing....................................................... 242
reagents.................................................................... 166 LEDs. see Light-emitting diodes (LEDs)
sensitivity ................................................................. 167 light sources.................................................... 235–237
setu-up components ...................................... 165, 166
optical system .......................................................... 241
setu-up configuration..................................... 173, 175 plaque assays and adenovirus.................................. 240
simulation ....................................................... 170, 171

Avraham Rasooly and Ben Prickril (eds.), Biosensors and Biodetection: Methods and Protocols Volume 1:
Optical-Based Detectors, Methods in Molecular Biology, vol. 1571, DOI 10.1007/978-1-4939-6848-0,
© Springer Science+Business Media LLC 2017

449
450 B IOSENSORS
Index
AND BIODETECTION

Charged-coupled devices (CCDs) (cont.) thermal bonding of PMMA .......................... 292, 294
plate assay material .................................................. 239 Conversion process ....................................................... 335
screens, portable devices ......................................... 237 CorelDraw™ ................................................................. 332
Shiga toxin activity ........................................ 237–238, Cryptosporidium parvum .............................................. 408
240, 243, 245
system sensitivity, factors ............................... 245, 247 D
technical cost and scientific-grade.......................... 233 Deoxynivalenol.............................................................. 441
Chemoprophylaxis ........................................................ 408 Deoxyribozymes.......................................... 389, 390, 392
CMOSs. See Complementary
DeskJet multifunctional printer ................................... 335
metal–oxide–semiconductors (CMOSs) 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine
CO2 laser system ........................................................... 332 (DPPC) ........................................................ 416
Colocalization, 3D plasmonic nanogap arrays
Dipstick tests ................................................................. 401
biomolecular interaction........................................... 16 DMR. See Dynamic mass redistribution (DMR)
circular and triangular patterns ................................ 23 DNA hybridization ....................................................... 395
colocalized biomolecular interaction ....................... 17
DNAzymes ................................................. 358, 360, 362,
DNA hybridization 368, 389, 390, 392
molecular binding capacity .................... 25, 26, 28 3D spheroid cell culture ...................................... 148–151
preparation .................................................... 23–24
2D-SPRi. See Two-dimensional surface plasmon
sensitivity enhancement ......................... 24, 25, 27 resonance imaging (2D-SPRi)
DNA molecules ......................................................... 22 Dulbecco’s Modified Eagle Medium
2D nanogap production ........................................... 21 (DMEM).................................... 35, 37, 38, 44
DNA preparation ...................................................... 19
Dynamic mass redistribution (DMR) .........................145,
electromagnetic field distribution ............................ 22 147, 150, 152–154, 157, 158
evaporation ..........................................................16, 28
fluorescence-based sensing ....................................... 15 E
nanogap fabrication ..................................... 17, 19, 20
nanogap reducuion ................................................... 22 ECM. See Extracellular matrix (ECM)
near-field intensity distribution ................................ 22 EGFR. See Epithelial growth factor receptor (EGFR)
optical set-up ..........................................17–18, 20, 21 Electroluminescence (EL) illumination .............. 236, 237
surface plasmon (SP) ................................................ 15 Electrospun nanofibers ................................................. 410
Colorectal cancer cell .................................................... 145 Enz-SubAuNP...................................................... 396, 401
Colorimetric biosensors ................................................ 329 Enzymatic tests.............................................................. 338
Colorimetric detection system ............................ 330, 335 Enzyme-linked immunosorbent assays
Colorimetric PCR ................................................ 360, 362 (ELISAs) ............................................. 408, 411
Colorimetric sensors ................................... 393, 394, 402 Enzyme–substrate complex .......................................... 396
Complementary metal–oxide–semiconductors Epithelial growth factor receptor
(CMOSs)................... 233–236, 245, 269, 281 (EGFR) ............................................. 64, 67, 68
Contamination detection device Evanescent field biosensor ................................... 163, 174
biopharmaceuticals.................................................. 288 Extracellular matrix
bioreactors ............................................................... 288 (ECM).......................143, 144, 146, 150, 156
cell viability assays ................................................... 288
F
colony-forming units (CFU)......................... 296–298
control and visualization......................................... 298 Fixed-angle surface plasmon resonance ......................... 76
E. coli .............................................................. 290, 295 Fluorescent sensors ....................................................... 393
fluorescene intensity................................................ 297 Fluorometer.......................................................... 236, 242
growth phase, cells .................................................. 290 Free prostate-specific antigen
materials.......................................................... 290–291 (f-PSA) ................................................ 4, 12, 13
methods ................................................................... 288 Fumonisin B1 ....................................................... 439, 444
microfluidic cassettes...................................... 292, 298
optics and optoelectronics ...................................... 293 G
pathogens and biologics ......................................... 287
β-Galactosidase.............................................................. 372
photodiodes............................................................. 293
GFP. See Green fluorescent protein (GFP)
portable kinetics fluorometer ........................ 291–294
Giardia lamblia ............................................................ 408
resazurin .................................................289, 295–297
Glucose assay ................................................................. 334
 BIOSENSORS AND BIODETECTION
Index 451
Glutamate assay ............................................................. 334 electromagnetic radiation ....................................... 130
GNRs. See Gold nanorods (GNRs) functionalization and immobilization.................... 137
Gold-coated tilted fiber Bragg gratings (TFBGs) growth conditions ................................................... 135
architectures .............................................................. 49 instruments.............................................................. 134
biochemical reactions................................................ 49 ITO-coated glass ..................................................... 137
bioreceptors ............................................................... 51 local electromagnetic field
Bragg wavelength...................................................... 53 simulation ........................................... 139, 140
configurations............................................................ 49 multiplexed bioprobes ............................................ 132
cytokeratins 7 (CK7) .......................................... 63–65 multiplexed biosensing, human and rabbit IgG
definition, biosensors ................................................ 47 samples ................................................ 137, 139
differential diagnosis ................................................. 51 nanoplasmonic biosensing
EGFR ......................................................................... 64 antigen detection ..................................... 135, 136
glass optical fiber ....................................................... 52 Beckman-Coulter UV-NIR
gold deposition, optical fiber surface .................51, 56 spectrophotometer ...................................... 136
high-resolution sensing............................................. 61 fabrication, nanoplasmonic biochip ................. 135
Kretschmann-Raether approach .........................48, 62 preparation ............................................... 134, 135
light coupling mechanism ........................................ 53 shapes ....................................................................... 130
optical fiber devices ................................................... 62 Sigma-Aldrich.......................................................... 137
optical system, production ....................................... 51 SPR band intensity and wavelength....................... 130
phase mask technique .........................................53, 56 UV–V spectra ................................................. 130, 131
photo-inscription ................................................53, 56 working solutions.................................................... 134
plasmonic generation ................................................ 50 Gp120............................................................................ 222
polyclonal antibodies ................................................ 67 G-quadruplex .............................................. 358, 360, 362
proteins and cells.......................................... 58–63, 67 Graphene nanoplatelets (GNPs) ......................... 344, 347
sensitivities ................................................................. 50 Graphic software ........................................................... 329
single-mode optical fiber .......................................... 52 Green fluorescent protein (GFP) ....................... 238, 240,
spectrum of................................................................ 51 242, 243, 246
SPR optical fiber sensors........................................... 50
SPR signature, EH and TM modes ...................58, 59 H
SRI sensitivity ......................................................60, 61 Hanks’ balanced salts (HBBS) .................................36–38
surface chemistry....................................................... 52 Hepatitis B virus (HBV)
surface functionalization.....................................56, 57
amplification ............................................................ 359
temperature changes ................................................. 55 colorimetric detection................... 359–361, 363–365
transducers................................................................. 49 colorimetric PCR .................................................... 360
transmembrane receptors
DNA extraction....................................................... 361
biosensing platform.......................................68, 69 DNAzyme................................................................ 358
development ........................................................ 69 G-quadruplexes ....................................................... 358
EGFR ...................................................... 64, 67, 68
nucleic acid testing .................................................. 358
optical fiber-based biochemical sensors ............. 68 One-Step PCR ........................................................ 362
surface functionalization..................................... 64 32
P labeled probe .................................. 359, 362, 363
transmission spectrum .............................................. 54
probe and primers .......................................... 361, 362
transmitted amplitude spectrum .......................50, 57, serologic immunity ................................................. 357
58, 60 serum calibration curve........................................... 361
wavelength tracking .................................................. 55 serum sample extraction ......................................... 359
Gold nanoparticles (AuNPs) ............................... 393–395
serum samples ......................................................... 367
lateral flow device and detection ................... 401, 402 TaqMan assay ................................358, 360, 365, 366
preparation ..................................................... 398, 400 Uracil N-glycosylase (UNG) .................................. 368
thiol-modified DNA ...................................... 399, 400
HIV genetic tests .......................................................... 222
Gold nanorods (GNRs) HIV infectivity assay ..................................................... 373
antibodies ....................................................... 132, 133 HIV long terminal repeat (LTR) ................................. 372
biofunctionalization ................................................ 133
Human growth hormone (hGH) ...................... 168, 169,
biosensing, imaging and drug delivery .................. 129 175–177, 181, 182
chemicals.................................................................. 133 Hydrocarbon ................................................................. 413
452 B IOSENSORS
Index
AND BIODETECTION

Hydrophobic interactions............................................. 410 fabrication................................................................ 6–8

Hypermulticolor detector............................................. 228 f-PSA ....................................................................12, 13
material ........................................................... 223–224 functionalization ....................................................... 12
PBMCs..................................................................... 222 materials and equipment......................................... 5–6
optical measurement ............................................... 7, 9
I sensing principle and configuration ........................... 4
Imaging sensitivity ............................................................. 10–12
and analysis software ............................................... 373 LOD. See Limit of detection (LOD)
LSPR-FO nanoprobe. See Localized surface plasmon
apparatus......................................................... 372, 373
Immunosensing, gold-coated TFBGs. See Gold-coated resonance coupled fiber-optic (LSPR-FO)
tilted fiber Bragg gratings (TFBGs) nanoprobe
Lysine............................................................................. 413
Inductively coupled plasma—atomic emission
spectroscopy (ICP-AES) ............................. 431
M
Influenza A H1N1 infection ............................... 121, 124
Integrated optics (IO) ......................................... 162, 164 Magnetic alloys..........................................................78, 79
In vitro selection ........................................................... 390 Magneto optical surface plasmon resonance (MOSPR)
angle-resolved............................................................ 76
L Au-Co alloy ............................................................... 75
Label-free single cell 3D2 invasion assay ....................150, chips functionalization and bioaffinity
152–155 assays .............................................................. 78
description .................................................... 75, 85–87
Lateral-flow assays (LFAs)
bartlett assays.................................................. 426, 427 drawbacks .................................................................. 75
capture antibodies .......................................... 421–424 electron beam evaporation ....................................... 85
fabrication............................................................77, 78
detection mechanisms.................................... 413, 416
liposome conjugation .................................... 418, 427 fluidics assembly ........................................................ 77
liposome preparation .................................... 416, 418, measurement setup ................................................... 76
measurements and data processing
420–422
membrane drying and blocking Au-Co-Au tri-layer chip................................84, 85
steps..................................................... 424–426 bioaffinity assay.............................................. 84–86
membrane preparation...........................418, 421–426 calibration ............................................................ 80
electromagnet...................................................... 79
myoglobin ...................................................... 428, 429
myoglobin and serum ............................................. 420 fitting function ....................................... 79, 82, 83
pipettor .................................................................... 424 Kretschmann configuration ................................ 80
real-time............................................................... 78
Universal Assays ...................................................... 416
Lateral flow devices .............................................. 394, 398 reflectivity curves.................................... 78, 79, 82
Legionella pneumophilia................................................ 408 refractive index variations vs. time...................... 80
sensitivity, SPR sensors ................................. 82–84
Leptospira ....................................................................... 408
Light-emitting diodes (LEDs) SPREETA sensor model TSPR1K23................. 79
blue excitation filter ................................................ 243 voltage.................................................................. 80
preparation ................................................................ 76
ELs ........................................................................... 237
intensity .......................................................... 237, 246 sputtering technique ................................................. 85
luminous film .......................................................... 238 surface chemistry....................................................... 76
multi-wavelength, light characteristics.......... 241, 242 MARS. See MicroRNA-RNase-SPR (MARS)
Mercury arc lamp .......................................................... 229
organic LED (OLED) ............................................ 237
semiconductor devices ............................................ 236 Metal ions ............................................389, 390, 392, 393
types ......................................................................... 241 Microfluidic paper-based analytical devices (μPADs)
oxidation process..................................................... 334
Limit of detection (LOD) ................................... 245, 246
Localized surface plasmon resonance coupled SiO2NP incorporation ............................................ 334
fiber-optic (LSPR-FO) nanoprobe MicroRNA biosensing, 2D-SPRi
advantages................................................................ 117
advantages.................................................................... 3
angular/wavelength interrogation ............................. 2 angular, wavelength and phase
biomarker detection.................................................... 1 interrogation....................................... 118, 119
experimental setup and procedure ............................. 3 description ............................................................... 121
disease screening ..................................................... 125
 BIOSENSORS AND BIODETECTION
Index 453
DNA and protein profiling..................................... 118 sensing property test ................................................. 95
MARS .................................................... 120, 121, 126 small molecules and olfactory protein ..................... 98
optical light.............................................................. 118 transmission spectrum ............................................ 103
optical setups ........................................................... 119 Nanofabrication............................................................. 6–8
plasmonic propagation............................................ 118 Nanoparticles........................................................ 390, 393
RNase H activity, pH effect........................... 125, 126 Nanoplasmonic biosensor, LSPR
sensing solutions ..................................................... 122 acquisition and immobilization, proteins ................ 92
setup and detection workflow ....................... 122, 123 electrochemistry
specificity, cDNA detection ........................... 123, 124 immunoreaction ....................................... 101, 103
stability, microRNA probes ........................... 126, 127 measurement .............................................100–102
surface chemistry............................................ 121, 122 thrombin detection .................................. 103, 104
throat swab samples ....................................... 124, 125 metal nanostructures...........................................90, 91
MicroRNA-RNase-SPR (MARS) ....................... 120, 121, nanoCA. see Nanocup arrays (nanoCA)
125, 126 nanoimprint lithography.....................................91, 93
Mobile health care ........................................................ 265 OBPs. see Oderant binding proteins (OBPs)
MOSPR. See Magneto optical surface plasmon resonance olfactory proteins
(MOSPR) explosive detection.............................................. 99
Muscarine .............................................. 36, 38, 39, 42–45 small molecules.............................................. 97–99
Mycotoxins optical and electrochemical
colloidal gold nanoparticles.................................... 440 measurement............................................91, 92
gold-labeled monoclonal antibodies ..................... 437, optical detection........................................................ 89
441, 442 peptides............................................................. 96, 105
immunoassays .......................................................... 435 self-immobilization, proteins.................................... 97
lateral flow immunoassay .............................. 436, 439, spectroscopy, optical detection...........................93, 95
442, 443 Nerve growth factor (NGF) .................37, 38, 42, 43, 45
materials................................................................... 439 Nitrocellulose ................................................................ 410
mouse monoclonal antibody .................................. 436 NPGS. See Nano pattern generation system (NPGS)
NC membrane......................................................... 437 Nucleic acid hybridization. See Resonance energy
PBS........................................................................... 445 transfer (RET)
signal detection ....................................................... 435 Nucleic acids......................................................... 389, 392
strip reader device ................................................... 436
test procedure.......................................................... 443 O
Myoglobin ............................................................ 428, 429 Oderant binding proteins (OBPs)
AcerASP2............................................. 92, 98, 99, 105
N
expression and purification ....................................... 96
Nano pattern generation system (NPGS) ................... 6, 8 nanoCA device .......................................................... 98
Nanoapertures .................................................... 21, 23, 28 and peptides............................................................... 91
Nanocup arrays (nanoCA) Oligonucleotides ........................................................... 395
BSA ................................................................. 103, 104 Oxidation process ................................................ 332, 335
characterization ......................................................... 94
electrochemistry properties .................................... 100 P
fabrication............................................................ 92–94 Paper-based analytical devices (PADs) ................ 304, 329
immunoreaction ...................................................... 101 Paper-based biosensor ......................................... 335–339
microplate reader ...................................................... 93
PCR. See Polymerase chain reaction (PCR)
NaCl........................................................................... 95 Peripheral blood mononuclear cells (PBMCs)...........222,
nanoimprint lithography........................................... 93 224, 225
nanoparticles.............................................................. 91
Peroxidase-like activity......................................... 358, 361
OBP-modified ........................................................... 99 Personal protective equipment (PPE) ......................... 354
olfactory proteins ...................................................... 98 Phosphatase and tensin homolog (PTEN).................144,
petri dish .................................................................... 97
145, 149, 150, 154–157
platinum and Ag/AgCl electrodes......................... 101 Phospholipids ................................................................ 413
protein immobilization ........................................... 105 Pipettor .......................................................................... 424
self-immobilization, proteins.................................... 97 PKC. See Protein kinase C (PKC)
454 B IOSENSORS
Index
AND BIODETECTION

Point-of-Care (POC) R
CMOS...................................................................... 269
development ............................................................ 268 Reflectometric interference spectroscopy (RIfS)
fluidics ...................................................................... 268 advanced set-up .............................................. 210, 213
fluorescence ............................................................. 268 advantage ................................................................. 208
hydrodynamic focusing.................................. 268, 269 affinity and kinetic constants ......................... 217, 218
smartphone cameras................................................ 268 antibody concentration........................................... 217
wide-field flow cytometer .............................. 273, 274 antigen-antibody interaction ......................... 216, 217
Polyacrylamide gel electrophoresis (PAGE) ................ 395 BiaCore .................................................................... 207
Polyethylene glycol (PEG) ........................................... 415 biomolecular interaction......................................... 207
Polylactic acid (PLA) .................................................... 410 biopolymer ..................................................... 212, 215
Poly-L-lysine (PLL) ....................................................... 410 biosensing ................................................................ 219
Polymerase chain reaction (PCR) cleaning procedure.................................................. 218
convective format ..........................252, 254, 256, 262 detection principle .................................................. 208
electrode dissolution experiments.......................... 254 glass substrates ........................................................ 210
experiments ............................................................. 254 immobilization
gold-standard nucleic acid-based detection amino-terminated ligands................................. 215
assay.............................................................. 252 biotin.................................................................. 216
kinetics ..................................................................... 253 carboxy terminated ligands............................... 216
medical diagnostic tools ......................................... 251 thiol terminated ligands.................................... 216
microscale convective flow states .................. 252, 253 1-lambda RIDe ....................................................... 219
optical analysis ......................................................... 251 label-free optical biosensor ............................ 207, 220
smartphones. see Smartphones, PCR liquid handling ........................................................ 210
thermal cycling protocols ....................................... 252 optical thickness ...................................................... 208
voice communication .............................................. 252 parallel setup................................................... 210, 214
Polyvinyl alcohol (PVA)................................................ 415 phospahete buffered saline ..................................... 219
Polyvinylpyrrolidone (PVP) ......................................... 415 set-up .............................................................. 211–213
PPE. See Personal protective equipment (PPE) silanization............................................................... 212
Prostate-specific antigen (f-PSA).................................... 13 single setup .............................................................. 210
Protein biomarker biosensors...................................12, 13 software.................................................................... 211
Protein kinase C (PKC) .......................................... 35, 39, surface chemistry............................................ 209, 210
40, 42, 44, 45 temperature ............................................................. 208
PTEN. See Phosphatase and tensin homolog (PTEN) transducers...................................................... 209, 218
Resonance energy transfer (RET)
Q attributes of paper substrates.................................. 304
calibration curve, target DNA................................ 316
QCapture Pro capture software ................................... 377 data acquisition .............................................. 316–317
Quantitative data analysis ............................................. 230 data analysis
Quantitative multivariate imaging (QMI) ................... 228 digital images............................................ 319, 320
Quantum dots (QDs) .......................................... 222, 226 gQD/Cy3 RET Pair ................................ 317, 319
bioconjugation microscope images, UCNP/QD RET
DTPA-terminated oligonucleotides................. 313 Pair ...................................................... 320, 321
hexahistidine-terminated decentralized diagnostic assays ............................... 304
oligonucleotides ................................. 312, 313 instrumentation and equipment.................... 308, 309
gQD/Cy3....................................................... 304, 307 oligonucleotide sequences,
imidazole modified paper substrates ...................... 315 hybridization....................................... 308, 309
neutral density (ND) filter...................................... 324 paper-based solid-phase nucleic acid
photoluminescence ................................................. 324 hybridization.............................. 304, 306, 307
properties................................................................. 302 pseudo-coloring of images ..................................... 324
salt aging.................................................................. 322 QDs. see Quantum dots (QDs)
structural types ........................................................ 302 reagents........................................................... 306, 308
surface area .............................................................. 302 SMN1 sequence ............................................. 318, 322
water soluble................................................... 311–312 solid-phase assays .......................... 303–305, 318, 319
 BIOSENSORS AND BIODETECTION
Index 455
UCNPs. see Upconverting nanoparticles (UCNPs) BupH PBS ............................................................... 353
wax printing, papers CRP analysis ............................................................ 344
aldehyde and imidazole functionality ............. 314, DIA .......................................................................... 345
315 human CRP IAs ............................ 347–349, 351, 352
aldehyde functionality ....................................... 314 iPAD4 ...................................................................... 346
substrates .................................................. 313, 314 iPAD mini ................................................................ 346
Resonant waveguide grating (RGW) iPhone 5s ................................................................. 346
3D cell models ........................................................ 144 IVD .......................................................................... 344
3D spheroid cell culture ................................ 148–151 MTP......................................................................... 352
adhesion.......................................................... 157, 158 pixel intensity (PI)................................................... 352
biosensor microplate ...................................... 144, 146 POC ......................................................................... 343
cell invasion ............................................................. 143 PPE .......................................................................... 354
cell migration........................................................... 143 Smartphones, PCR
chemotaxis ............................................................... 143 advantages................................................................ 252
data analysis software .............................................. 148 amplification ............................................................ 255
extracellular matrix (ECM) .................................... 143 DNA replication ...................................................... 263
intra-sensor referencing .......................................... 158 electrochemical dissolution ...................259, 262–264
label-free single cell monitoring.................... 144, 145 enzyme..................................................................... 265
Matrigel ................................................. 146, 147, 150 fabrication................................................................ 259
microplates and instruments .................................. 148 fluorescence analysis....................................... 258–261
multicellular spheroid ............................................. 144 instrument portability ............................................. 255
optimal seeding density, spheroids......................... 157 label-free detection......................................... 262, 263
single cell monitoring ............................................. 146 molding processes ................................................... 262
spheroids, cancer cells ............................................. 146 photography ............................................................ 265
tissue culture medium and cell line............... 145, 148 temperature settings....................................... 255–258
transwell invasion assay .................................. 156, 157 voice communication .............................................. 252
tumor metastasis ..................................................... 143 SPPs. See Surface plasmon polaritons (SPPs)
wavelength sweeping technique ............................. 146 SPR. See Surface plasmon resonance (SPR)
wound healing assay.............................. 152, 155, 156 Streak imaging flow cytometry
Respiratory syncytial virus (RSV)................................. 408 adhesive layer........................................................... 284
RGW. See Resonant waveguide grating (RGW) background signal and noise
Ribozymes ..................................................................... 389 reduction............................................. 277–280
cell culture and fluorescence staining .................... 271
S cell culture and labeling ................................. 272, 273
cell image ........................................................ 269–271
SA-AuNP. See Sialic acid gold nanoparticle (SA-AuNP)
Sandwich immunoassay ................................................ 351 CMOS imaging sensors .......................................... 269
“capture” antibody ................................................. 410 color channel extraction ................................ 278–281
components ............................................................. 271
“detection” antibodies............................................ 410
flow of fluid ............................................................. 409 computer control and data analysis........................ 272
membrane properties ..................................... 410, 414 configuration ........................................................... 277
development ............................................................ 268
membrane types ...................................................... 410
quantification........................................................... 411 description ............................................................... 267
Sialic acid gold nanoparticle (SA-AuNP) flow cell fluid delivery and imaging ....................... 271
absorption spectra ................................................... 114 focal ratio ................................................................. 283
optimization ................................................... 282, 283
concentration ................................................. 112, 113
description ............................................................... 112 POC. see Point-of-care (POC)
hemagglutinin ......................................................... 114 rare cell counting ........................................... 281, 282
webcam performance ............................ 275, 277, 284
linear correlation ..................................................... 115
particle size .............................................................. 113 wide-field flow cell fabrication....................... 274–276
Signal amplification ....................................................... 120 Streptavidin.................................................................... 413
Sulforhodamine B (SRB) dye ....................................... 413
Signal transduction ........................................................... 5
Silica nanoparticles (SiO2NP) ............................. 333, 338 Surface biofunctionalization, BiMW
Smartphone-based colorimetric reader (SBCR) (3-aminopropyl)triethoxysilane
Ab-bound MTP .................................... 346, 350, 351 (APTES)..................................... 174, 177, 178
456 B IOSENSORS
Index
AND BIODETECTION

Surface biofunctionalization, BiMW (cont.) living cells .................................................................. 32

cleaning procedure ......................................... 176, 178 mammalian cells ........................................................ 35
immobilization materials............................................................... 35–37
anti-hGH antibody ........................................... 179 PC12 cells
Irgarol derivative 4e.................................. 178, 179 KCl ....................................................................... 40
oxidation procedure ................................................ 178 muscarine, NGF ............................................42, 43
p-phenylene diisothiocyanate (PDITC) ................. 176 PKC agonist and antagonist .........................40, 42
silanization procedure ............................................. 178 real-time cell-based sensor system............................ 35
Surface plasmon polaritons (SPPs) ..........................48, 60 reflection intensity...............................................34, 44
Surface plasmon resonance (SPR)...................... 3, 16, 49, single-cell activity ...................................................... 31
130, 328 staurosporine ............................................................. 44
band position and interparticle distance ................ 110
biosensing assays ....................................................... 75 U
colocalization, 3D plasmonic nanogap arrays. see Upconverting nanoparticles (UCNPs)
Colocalization, 3D plasmonic nanogap arrays
aldehyde modified paper substrates ....................... 315
definition ..................................................................... 2 donors ............................................................. 302, 303
GNRs. see Gold nanorods (GNRs) gQD/Cy3....................................................... 304, 307
gold-coated TFBGs. see Gold-coated tilted fiber Bragg
layer-by-layer assembly, paper
gratings (TFBGs) substrates ............................................ 315, 316
Kretschmann configuration ..................................2, 74 microscope images ......................................... 320, 321
LSPR-FO nanoprobe. see Localized surface plasmon NaYF4 ............................................................. 310, 311
resonance coupled fiber-optic (LSPR-FO)
water soluble............................................................ 311
nanoprobe Uracil N-glycosylase (UNG) ........................................ 368
optical sensing ........................................................... 16 US Environmental Protection Agency
oscillation, gold nanoparticles ....................... 109, 110
(EPA)............................................................ 390
TMOKE..................................................................... 75 UV–Vis spectrophotometer.......................................... 394
total internal reflection (TIR) conditions ................ 73
variables ....................................................................... 2 V
T Viral detection, gold nanoparticles
colorimetric measurement ............................. 113, 114
TaqMan assay .............................................. 358, 365, 366
equipment................................................................ 111
Thin-layer chromatography (TLC).............................. 414 hemagglutinin ......................................................... 110
Transverse magneto-optic Kerr effect influenza virus deactivation .................................... 112
(TMOKE) ...................................................... 75
influenza virus inactivation ..................................... 112
Transwell invasion assay ....................................... 156, 157 linear dynamic range ...................................... 114, 115
Tris–HCl buffers ........................................................... 395 metallic nanoparticles.............................................. 109
Two-dimensional surface plasmon resonance imaging
particle size analysis................................................. 112
(2D-SPRi) pathogenic viruses, sialic acid ........................ 110, 113
angle shift measurements.......................................... 34 polynucleotides, enzymes, proteins, cells and heavy
biosensing, microRNA. see MicroRNA biosensing,
metals ........................................................... 110
2D-SPRi SA-AuNP. see Sialic acid gold nanoparticle
biosensorsbiosensors ................................................. 32 (SA-AuNP)
cell culture and sensor surface ............................37, 38 sialic acid and gelatin .............................................. 111
cell-stimulation ....................................................32, 34
UV-V measurements............................................... 113
cell suspension ........................................................... 44
data analysis .................................................. 34, 39, 40 W
experiments .........................................................38, 39
fluorescent labeling ................................................... 31 Webcams ..................................................... 233, 245, 246,
intracellular signal analysis ........................................ 34 275, 277
Kretchmann configuration .................................32, 33 Whole blood.................................................................. 429
light emitting diode .................................................. 32 Winner sequences.......................................................... 390