Moo-Yeal Lee

Microarray
Bioprinting
Technology
Fundamentals and Practices
Microarray Bioprinting Technology
Moo-Yeal Lee

Microarray Bioprinting
Technology
Fundamentals and Practices
Moo-Yeal Lee
Department of Chemical and Biomedical Engineering
Cleveland State University
Washkewicz College of Engineering
Cleveland, OH, USA

ISBN 978-3-319-46803-7    ISBN 978-3-319-46805-1 (eBook)

DOI 10.1007/978-3-319-46805-1

Library of Congress Control Number: 2016961347

© The Author(s) 2016

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.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

In late 2002, I was exposed to microarray bioprinting technology when my postdoc-

toral advisor, Professor Jonathan S. Dordick at Rensselaer Polytechnic Institute,
received a MicroSys microarray spotter manufactured by Cartesian Technologies
(later acquired by Genomic Solutions and DigiLab, Inc.) from VA Medical Center
in Albany, NY. Since then, I was fascinated by the potential of microarray bioprint-
ing technology and have focused my research career on developing surface chemis-
try and methods of printing various biological samples on chip platforms for
miniaturized high-throughput toxicology assays. Through several research pro-
grams at Rensselaer Polytechnic Institute, Solidus Biosciences, Inc., and Cleveland
State University, we have developed highly automatable, high-throughput microar-
ray chip platforms and associated instruments that can be used for miniaturized
biochemical- and cell-based assays to assess human metabolism and toxicology.
As the lead scientist and the principle investigator, I have been working on the
underlying technology for enzyme and cell encapsulation in hydrogel spots and
“microarray three-dimensional (3D) bioprinting” on plastic chip platforms for high-
throughput, high-content imaging assays. Specific areas of current research include
enzyme, virus, and cell printing with inkjet-driven dispensing robots for use in min-
iaturized 3D cell cultures and organotypic tissue constructs created via layer-by-
layer cell printing. In collaboration with engineers from Samsung Electro-Mechanics,
Co., a chip platform (S+ micropillar chip and microwell chip) and associated instru-
ments (S+ MicroArrayer and S+ Scanner) have been commercialized. However,
implementing microarray bioprinting in miniaturized cell-based assays has been
still challenging because of lack of standardized experimental procedures. To this
end, I decided to write a book on microarray bioprinting technology with a number
of fundamental and practical protocols. Although the majority of protocols pro-
vided in this book are based on methodologies we developed and optimized over 10
years, the chapters have all been written with significant contributions from my
graduate students and postdoctoral fellow. Without their assistance, it would be

v
vi Preface

impossible for me to complete this book. I hope that this book will serve as a valu-
able reference manual for graduate students, postdoctoral fellows, scientists, and
researchers working in this area of research. We envision that microarray bioprint-
ing technology offers new opportunities for creating highly organized multicellular
tissue constructs by precisely dispensing multiple human cell types in hydrogel lay-
ers on a chip with printing robots and mimicking the microenvironments of tissues
in vivo, thereby potentially revolutionizing regenerative medicine, oncology, and
drug discovery.

Cleveland, OH Moo-Yeal Lee

Contents

1 Overview of Microarray Bioprinting Technology................................... 1

Moo-Yeal Lee
1.1 Overview............................................................................................. 1
1.2 Microarray Chip Platforms................................................................. 3
1.2.1 Chips and Instrument.............................................................. 3
1.2.2 Cell Microarrays..................................................................... 6
1.2.3 Enzyme Microarrays............................................................... 7
1.2.4 Virus Microarrays................................................................... 11
1.3 Experimental Procedures.................................................................... 12
1.3.1 Cell Spotting........................................................................... 13
1.3.2 Enzyme and Compound Spotting........................................... 13
1.3.3 Chip Incubation, Staining, Scanning, and Data Analysis....... 13
1.4 Advantages and Applications............................................................. 14
References.................................................................................................... 16
2 Microarray Spotter and Printing Technologies....................................... 19
Akshata Datar, Dong Woo Lee, Sang Youl Jeon,
and Moo-Yeal Lee
2.1 Introduction......................................................................................... 20
2.1.1 Contact Printing Techniques................................................... 20
2.1.2 Non-contact Printing Techniques............................................ 21
2.2 Materials............................................................................................. 22
2.3 Components of S+ MicroArrayer....................................................... 22
2.3.1 Main Body Components......................................................... 23
2.3.2 Externally Connected Components......................................... 26
2.4 General Precautions for Operating S+ MicroArrayer......................... 27
2.5 Daily Operations of S+ MicroArrayer................................................ 28
2.5.1 Turning ‘ON’ the System........................................................ 29
2.5.2 Refilling the Pressure Bottles.................................................. 29

vii
viii Contents

2.5.3 Washing Tubes, Solenoid Valves, and Ceramic Tips

with Ethanol and Water........................................................... 30
2.5.4 Dispensing Samples on the Micropillar/Microwell
Chips with a Work File........................................................... 32
2.5.5 Replacing Solenoid Valves and Ceramic Tips........................ 34
2.5.6 Turning ‘OFF’ the System...................................................... 37
2.6 Detailed Programming for Normal Operation.................................... 38
2.6.1 Generating Wash and Dry Sequences..................................... 38
2.6.2 Defining Well Plates............................................................... 38
2.6.3 Registering Chips.................................................................... 41
2.6.4 Registering Spot Layouts........................................................ 44
2.6.5 Optimizing Dispensing Parameters Using
Vision Inspection.................................................................... 47
2.7 Summary............................................................................................. 50
References.................................................................................................... 50
3 Chip Platforms and Chip Surface Treatments........................................ 53
Parnian Bigdelou, Alexander Roth, Yana Sichkar,
and Moo-Yeal Lee
3.1 Introduction......................................................................................... 54
3.1.1 Surface Modification of Micropillar and Microwell Chips.... 54
3.1.2 Surface Modification of Glass Slides...................................... 56
3.2 Materials............................................................................................. 57
3.3 Protocols............................................................................................. 58
3.3.1 Cleaning Microwell Chips with Plasma for Air
Bubble Removal...................................................................... 58
3.3.2 Coating Micropillar Chips with Poly(maleic
anhydride-­alt-­1-octadecene) (PMA-OD)
for Cell Printing...................................................................... 61
3.3.3 Cleaning the Surface of Glass Slides with Acid..................... 61
3.3.4 Coating of Glass Slides with Poly(styrene-co-maleic
anhydride) (PS-MA) for Cell Printing.................................... 63
3.3.5 Coating of Acid-Cleaned Glass Slides
with Methyltrimethoxysilane (MTMOS)
for Enzyme Printing................................................................ 64
3.3.6 Coating of Acid-Cleaned Glass Slides
with (3-Aminopropyl)trimethoxysilane (APTMS)
for Protein Attachment............................................................ 65
3.3.7 Measurement of Silanization on the Surface
of the APTMS Slides with Fluorescein
Isothiocyanate (FITC) Labeling.............................................. 65
3.3.8 Attachment of Glutaraldehyde on the Surface
of the APTMS Slides for Protein Attachment........................ 66
3.3.9 Immobilization of Extracellular Matrix (ECM)
Proteins on Reactive Surfaces................................................. 67
3.4 Summary............................................................................................. 68
References.................................................................................................... 69
Contents ix

4 Biological Sample Printing........................................................................ 71

Parnian Bigdelou, Alexander Roth, Akshata Datar,
and Moo-Yeal Lee
4.1 Introduction......................................................................................... 71
4.2 Materials............................................................................................. 73
4.3 Preparation of Stock Solutions............................................................ 74
4.4 Daily Operation of the S+ MicroArrayer for Dispensing
Biological Samples............................................................................. 76
4.5 Sample Printing Protocols.................................................................. 80
4.5.1 Enzyme Printing for Metabolism-Induced Drug
Toxicity Assays....................................................................... 80
4.5.2 Compound Printing................................................................. 82
4.5.3 Cell Printing............................................................................ 85
4.5.3.1 Preparation of Cell Suspension
in Growth Medium................................................... 85
4.5.3.2 Cell Printing for 2D Monolayer Culture on the
Micropillar Chip...................................................... 86
4.5.3.3 Cell Printing in Alginate for 3D Cultures................ 87
4.5.3.4 Cell Printing in Matrigel for 3D Cultures................ 90
4.5.3.5 Cell Printing in a Mixture of Alginate
and Matrigel for 3D Cultures................................... 93
4.5.3.6 Cell Printing in a Mixture of Alginate
and Fibrinogen for 3D Cultures............................... 95
4.5.3.7 Cell Printing in PuraMatrix for 3D Cultures........... 96
4.5.4 Virus Printing.......................................................................... 98
4.5.4.1 Measurement of Viral Titer in a 96-Well Plate........ 98
4.5.4.2 Adenoviral Transduction on the Micropillar/
Microwell Chip........................................................ 99
4.6 Inspection of Cells Printed on the Micropillar Chips
Using a Bright-Field Microscope........................................................ 101
4.7 Coefficient of Variation (CV) and Z’ Factor for Assay Validation..... 102
4.8 Summary............................................................................................. 103
References.................................................................................................... 103
5 High-Content Cell Staining....................................................................... 105
Kyeong-Nam Yu, Pranav Joshi, and Moo-Yeal Lee
5.1 Introduction......................................................................................... 105
5.1.1 Fluorescent Dyes..................................................................... 106
5.1.2 Immunofluorescence (IF) Assays with Antibodies................. 107
5.1.3 Fluorescent Proteins................................................................ 108
5.2 Materials............................................................................................. 109
5.2.1 Reagents for Fluorescence Staining........................................ 109
5.2.2 Reagents for Immunofluorescence Staining........................... 109
5.2.3 Devices for Cell Staining........................................................ 110
5.2.4 Preparation of Dye Stock Solutions in DMSO....................... 111
x Contents

5.3 Protocols............................................................................................. 112

5.3.1 Staining Cells with Fluorescent Dyes..................................... 112
5.3.1.1 Preparation of a Saline Solution.............................. 112
5.3.1.2 Staining Cells on the Micropillar/Microwell
Chip Platform with Fluorescent Dyes...................... 112
5.3.2 Staining Cells on the Chip Platform
with Fluorophore-­Labeled Antibodies.................................... 116
5.3.2.1 Cell Fixation............................................................ 117
5.3.2.2 Permeabilization of Cell Membranes...................... 117
5.3.2.3 Blocking of Nonspecific Binding and Incubation
with Primary/Secondary Antibodies
for Fluorescence Labeling and Detection................ 118
5.3.3 Measuring the Expression Levels of Drug
Metabolizing Enzymes on a Chip........................................... 119
5.3.3.1 Tyramide Signal Amplification Kit
(Life Technologies).................................................. 121
5.4 Summary............................................................................................. 122
References.................................................................................................... 122
6 3D-Cultured Cell Image Acquisition........................................................ 125
Pranav Joshi, Kyeong-Nam Yu, Emily Serbinowski,
and Moo-Yeal Lee
6.1 Introduction......................................................................................... 125
6.2 Materials............................................................................................. 127
6.3 Protocols............................................................................................. 127
6.3.1 Daily Operational Procedures................................................. 128
6.3.2 Parameter Setting Procedures................................................. 135
6.3.2.1 Setting the Position of Filters and the Distance
of Each Step for Autofocus...................................... 135
6.3.2.2 Setting XYZ Coordinates for Different Chips
and Objective Lenses............................................... 137
6.4 Summary............................................................................................. 139
6.5 Appendix............................................................................................. 139
References.................................................................................................... 141
7 High-Content Image Analysis................................................................... 143
Sean Yu, Pranav Joshi, Dong Woo Lee, and Moo-Yeal Lee
7.1 Introduction......................................................................................... 143
7.2 Materials............................................................................................. 145
7.2.1 ImageJ..................................................................................... 145
7.2.2 S+ Chip Analysis.................................................................... 145
7.3 Protocols............................................................................................. 146
7.3.1 3D Cell Image Analysis with ImageJ..................................... 146
7.3.2 Examples of Image Processing with ImageJ.......................... 149
7.3.2.1 Hue Filter................................................................. 149
7.3.2.2 Background Subtraction, Brightness Filter,
and Region of Interest (ROI)................................... 150
Contents xi

7.3.2.3 Outlier Exclusion..................................................... 151

7.3.2.4 The Performance of the Plugin................................ 151
7.3.3 Image Deconvolution.............................................................. 152
7.3.3.1 Performance of Deconvolution................................ 154
7.3.4 Plotting Dose Response Curves with S+ Chip Analysis......... 154
7.4 Summary............................................................................................. 159
References.................................................................................................... 159
8 Applications of Microarray Bioprinting.................................................. 161
Alexander Roth, Emily Serbinowski, and Moo-Yeal Lee
8.1 Introduction......................................................................................... 161
8.2 Assay Development for Microarray Bioprinting Technologies.......... 162
8.2.1 Hepatotoxicity Assays............................................................ 162
8.2.1.1 Phase I and Phase II Drug Metabolizing
Enzyme Assays........................................................ 164
8.2.1.2 Drug Transporter Assays......................................... 165
8.2.1.3 Oxidative Stress Assays........................................... 166
8.2.2 Neurotoxicity Assays.............................................................. 166
8.2.2.1 Oxidative Stress and Related Assays....................... 167
8.2.2.2 Ion Channel Assays................................................. 167
8.2.2.3 Drug Metabolism Assays......................................... 168
8.3 Simulation of the In Vivo Microenvironment: Liver Applications..... 169
8.4 Other Applications.............................................................................. 171
8.5 Summary............................................................................................. 171
References.................................................................................................... 172
Reviewers

Eben Alsberg Department of Biomedical Engineering and Orthopaedic Surgery,

Case Western Reserve University, Cleveland, OH, USA
Jonathan S. Dordick Department of Chemical and Biological Engineering,
Rensselaer Polytechnic Institute, 4005 Center for Biotechnology and Interdisciplinary
Studies, Troy, NY, USA
Tiago Fernandes Department of Bioengineering, Instituto Superior Técnico
(IST), Lisbon, Portugal
Rayton Gerald 3D MicroArray, Inc., Cleveland, OH, USA
Chandrasekhar Kothapalli Department of Chemical & Biomedical Engineering,
Cleveland State University, Cleveland, OH, USA
Bosung Ku MBD Korea Co., Ltd., Yongin, South Korea
Seok-Joon Kwon Department of Chemical and Biological Engineering, Rensselaer
Polytechnic Institute, 4119 Center for Biotechnology and Interdisciplinary Studies,
Troy, NY, USA
Anand K. Ramasubramanian Department of Biomedical Engineering, The
University of Texas at San Antonio, San Antonio, TX, USA

xiii
Chapter 1
Overview of Microarray Bioprinting
Technology

Moo-Yeal Lee

Contents
1.1 Overview............................................................................................................................ 1
1.2 Microarray Chip Platforms................................................................................................ 3
1.2.1 Chips and Instrument............................................................................................. 3
1.2.2 Cell Microarrays.................................................................................................... 6
1.2.3 Enzyme Microarrays.............................................................................................. 7
1.2.4 Virus Microarrays.................................................................................................. 11
1.3 Experimental Procedures................................................................................................... 12
1.3.1 Cell Spotting.......................................................................................................... 13
1.3.2 Enzyme and Compound Spotting.......................................................................... 13
1.3.3 Chip Incubation, Staining, Scanning, and Data Analysis...................................... 13
1.4 Advantages and Applications............................................................................................ 14
References................................................................................................................................... 16

1.1 Overview

Microarray bioprinting refers to printing extremely small amounts of biological

samples in patterns on a chip platform, such as functionalized silicone wafers, glass
slides, and plastic chips [1, 2]. This technology is mostly based on inkjet printing
technologies that allows nanoliter droplet dispensing in high throughput to deposit
and pattern biological samples on a chip, including genes, carbohydrates, proteins,
viruses, cells, hydrogels, extracellular matrices (ECMs), growth factors, growth

© The Author(s) 2016 1

M.-Y. Lee, Microarray Bioprinting Technology,
DOI 10.1007/978-3-319-46805-1_1
2 1 Overview of Microarray Bioprinting Technology

media, compounds, fluorescent dyes, and other reagents. In general, microarray

chips are classified into six different categories depending on biological samples
printed, fabricated, or immobilized: (1) gene, (2) carbohydrate, (3) protein, (4)
virus, (5) cell, and (6) tissue. These chips have varied applications and methods for
fabrication. For example, GeneChip® from Affymatrix (a well-known chip) con-
tains high-density DNA microarrays on silicone wafers manufactured by photoli-
thography technology. Likewise, carbohydrate microarrays contain immobilized
oligosaccharides on the surface of glass slides to study interactions between oligo-
saccharides and antibodies, or other carbohydrate-recognizing proteins and viruses.
Z Biotech offers several versions of glycan microarrays on glass slides. Antibody
microarrays are a part of protein microarrays, which have a high-density array of
antibodies on functionalized glass slides. RayBiotech is one of pioneers who
invented various protein microarray-based assays. Our research group invented
enzyme microarrays in 2005, which contained 525 spot arrays of drug metabolizing
enzymes (DMEs) encapsulated in silicone precursors to simulate metabolism-­
induced toxicity in human livers [3–7]. The first cell microarrays have been intro-
duced by David M. Sabatini’s group in 2001, which is a variant of DNA microarrays
and based on reverse transfection of genes on cell monolayers [8]. In 2008, our
group demonstrated a method of printing and encapsulating live human cells in
1080 hydrogel spots on a functionalized glass slide for three-dimensional (3D) cell
culture and organ-specific toxicity assessment [9–11]. The cell microarray technol-
ogy has been further improved on a micropillar/microwell chip platform with ade-
novirus microarrays to control expression levels of DMEs in hepatic cell lines and
simulate drug-induced hepatotoxicity [12–17]. Tissue microarrays are somewhat
different from other microarrays in terms of the method of fabrication. Typical tis-
sue microarrays are prepared by removing tissue cores in paraffin-embedded tissue
samples using a hollow needle, inserting them in a recipient paraffin block in an
array pattern, sectioning this block using a microtome, and mounting the slices on
glass slides. Tissue microarrays can be a useful tool for immunohistochemistry and
fluorescent in situ hybridization in cancer research.
Although there have been several books published on microarrays, almost all
books focus on DNA and protein microarrays. Thus, there is a need for a practical
book focusing on broader microarray bioprinting technologies that cover enzyme,
virus, cell, and compound printing. In this book, we provide practical information
and standard operating procedures (SOPs) optimized for printing a myriad of bio-
logical samples for miniaturized 3D cell cultures and mini-tissue constructs that can
be used for biochemical- and cell-based assays to assess human metabolism and
toxicology. We illustrate in-depth knowledge in the area of microarray bioprinting
for assay miniaturization and provide SOPs for constructing 3D cells on the plastic
micropillar/microwell chip via microarray bioprinting. The aims of this practical
book are to introduce key fundamentals of microarray bioprinting, chip platforms
and associated instruments/devices required, and detailed experimental protocols
for chip surface treatment, biological sample printing and incubation, high-content
cell staining for mechanistic toxicology, cell image acquisition, and high-­throughput
data analysis. With several examples of practical applications, we hope that this
1.2 Microarray Chip Platforms 3

book can be particularly beneficial to graduate students and researchers working in

the areas of tissue engineering, regenerative medicine, disease modeling, high-­
throughput compound screening, in vitro toxicology, and cancer research.

1.2 Microarray Chip Platforms

1.2.1 Chips and Instrument

For robust microarray bioprinting, a highly versatile micropillar/microwell chip

system and associated automation instruments have been developed in collabora-
tion with Samsung Electro-Mechanics Co. (SEMCO), South Korea. Therefore,
the majority of SOPs developed and optimized are centered on this chip platform
for rapidly testing drug candidates and environmental toxicants against human
cells isolated from organs. The disposable micropillar/microwell chips with
25 mm × 75 mm in size have been manufactured by plastic injection molding
(Fig. 1.1). The new chip platform is a robust and flexible system for 3D cultures
of human and animal cells, enzymatic reactions, viral infection, and compound
screening. The micropillar chip made of poly(styrene-co-maleic anhydride) (PS-
MA) or polystyrene coated with poly(maleic anhydride-alt-1-octadecene)
(PMA-OD) supports 3D cell cultures and comprises an array of human cells for
gene expression and toxicity screening (Fig. 1.1A). A single micropillar chip
contains 532 micropillars (0.8 mm pillar diameter, 1 mm pillar height, and
1.5 mm pillar-to-pillar distance) onto which an array of 3D human cell cultures
(cells entrapped in 60-nL spots of hydrogel) is dispensed using a microarray
spotter (Fig. 1.2A). The microwell chip made of polystyrene contains a comple-
mentary array of enzymes, recombinant viruses, growth media, reagents, or test
compounds (typically 950-nL solutions in the microwells) (Fig. 1.1B). By sand-
wiching two chips together, various miniaturized biochemical and cell-based

Fig. 1.1 Schematics of (A) a micropillar chip and (B) a microwell chip. (C) Image of micropillar
and microwell chips to scale against a microscopic glass slide.
4 1 Overview of Microarray Bioprinting Technology

Syringe
pumps

Chilling
chip deck

-
Chip-loading deck
Mercury lamp
Objective lens
Color CCD
Solenoid camera
valves Ceramic tips
96-Well plate
4 Channel filter wheel
(A) Water bath Sonicator (B)

Fig. 1.2 Photographs of (A) a microarray spotter (S+ MicroArrayer) with six solenoid valves and
ceramic tips for printing biological samples and (B) a chip-scanning system (S+ Scanner) for cell
image acquisition. The microarray spotter has a printing head with six solenoid valves and a chill-
ing capability that is designed for printing temperature-sensitive hydrogels such as Matrigel.
Several human cell types can be spotted in hydrogels such as alginate, fibrinogen, PuraMatrix, and
Matrigel onto micropillars (typically 60 nL), and various combinations of biological samples as
well as test compounds (typically 950 nL) can be dispensed into microwells. Two chips can be
sandwiched together or separated each other consistently and accurately. The sandwiched chips
can be incubated in a gas-permeable chamber for cell growth.

Water to prevent evaporation

DataChips

Staining dye

(A)
Sandwiched chips (B)

Fig. 1.3 (A) Gas-permeable incubation chamber and (B) Staining plate. The sandwiched chips are
incubated in a chamber that allows good gas exchange. Cell growth occurs during incubation due
to uniform gas exchange over the entire chip. The micropillar chips with cells (DataChip) can be
stained uniformly on a staining plate for various cell-based assays.

assays can be performed on the chip platform (Fig. 1.3). For complete automa-
tion, a microarray spotter has been developed to enable consistent manipulation
of the micropillar/microwell chip, and a chip-scanning system and image analy-
sis software have been developed to analyze data from the scanned images in a
high-throughput fashion (Figs. 1.2 and 1.4).
1.2 Microarray Chip Platforms 5

Fig. 1.4 S+ Chip Analysis from Samsung for cell image analysis. S+ Chip Analysis extracts
green and red fluorescent intensities from living and dead cells in each cell spot on the scanned
chip, plots sigmoidal dose-response curves with the percentage of live cells against the concentra-
tion of the test compound, and then calculates IC50 values for each test condition. It takes about
10 min to process three DataChips, which is equivalent to seventy-two individual dose response
curves and IC50 values.

The micropillar/microwell chip platform represents a promising, high-­throughput

microscale alternative to conventional in vitro multi-well plate platforms and cre-
ates new opportunities for rapid and inexpensive assessment of compound efficacy/
toxicity at very early phases of drug development. We have been making indepen-
dent and important contributions to the many interdisciplinary applications of
protein-, virus-, and cell-based microarray biochips, including the data analysis
toxicology assay chip (DataChip) [9–11, 14–16], metabolizing enzyme toxicology
assay chip (MetaChip) [3, 6, 7], P450 inhibition chip [18], multienzyme lead
­optimization chip (Multizyme Chip) [19], transfected enzyme and metabolism
chip (TeamChip) [12], RNA interference chip (RNAi Chip) [20], and heparin gly-
can chip (HepGly Chip) [21] (Table 1.1). These microarray biochips and instru-
ments will enable low-cost, accurate, high-throughput in vitro analysis of drug
metabolism and toxicology, and serve as a unique platform for high-throughput
toxicology screening in drug discovery.
The chips developed for these microarray technologies can be used indepen-
dently or in combination with other chip types for various applications. For exam-
ple, the DataChip represents cell microarrays on a chip that are developed to assess
acute toxicity of human organs, whereas the MetaChip represents metabolizing enzyme
microarrays on a chip that are specifically designed for analysis of drug metabolism
and enzyme inhibition. With the MetaChip combined with the DataChip, metabolism-
induced compound toxicity has been demonstrated to simulate hepatotoxicity of
6 1 Overview of Microarray Bioprinting Technology

Table 1.1 Examples of microarray chip platforms

Platform Components Output/endpoint References
DataChip Human/rat cells Acute toxicity (metabolism-­ [9–11,
(liver and other organs) induced toxicity with MetaChip) 14–16]
MetaChip Human/rat metabolizing Drug metabolism, Enzyme [3, 6, 7, 18]
enzymes inhibition
Multizyme Chip Biosynthetic enzymes Lead compound optimization [19]
TeamChip Human cells + viruses Metabolism-mediated toxicity [12]
carrying metabolic genes
RNAi Chip Interfering RNA Knock-down of gene expression [20]
HepGly Chip Heparin glycans Anticoagulation [21]

metabolism-sensitive compounds. The Multizyme Chip is a variant of the MetaChip,

which contains biosynthetic enzyme microarrays for optimization of lead com-
pounds. With the Multizyme Chip combined with the DataChip, efficacy of lead
compound derivatives generated on the Multizyme Chip can be tested in high
throughput. The TeamChip represents recombinant virus microarrays that can be
used to express different combinations and levels of DMEs. The RNAi Chip is a
variant of the TeamChip developed for knocking down gene expression using short
hairpin RNA (shRNA) microarrays. HepGly Chip refer to heparin glycan microar-
rays developed to optimize the structure of heparin for enhanced anticoagulation.
These microarray bioprinting technologies are based on the combination of
microscale biocatalysis (for human metabolism of drugs and biosynthesis of lead
compounds), human cell culture (for screening major cell types of the body), dis-
ease modeling (via over-expression and knock-down of specific genes using adeno-
viruses, lentiviruses, and shRNAs), materials science (to all microscale fabrication
and surface treatment), and automation technology to achieve high-throughput
operation with bench top instruments. The most representative cell, enzyme, and
virus microarrays will be discussed in more details in the following sections.

1.2.2 Cell Microarrays

Testing efficacy and toxicity of compounds in more predictive cellular models is

challenging due to a lack of 3D cell systems that can accommodate high-throughput
screening (HTS) assays. The DataChip has the capability to screen drug candidates
on 3D-cultured human cells for organ-specific toxicity at speeds commensurate
with HTS [9–12]. Human cells in hydrogels such as alginate, fibrinogen, PuraMatrix,
and Matrigel are spotted onto the micropillar chip (14 × 38 pillars per chip), and cell
spot volumes are as low as 60 nL (Fig. 1.5). Encapsulated cells are grown on indi-
vidual micropillars of the DataChip when submerged in growth media printed into
the microwells of the complementary chip, leading to 3D morphology, which is
different from morphology observed in 2D cell culture (Fig. 1.6). Typical cell types
accommodated for the DataChip are liver, heart, kidney, bladder, skin, and neural,
among others (Table 1.2).
1.2 Microarray Chip Platforms 7

Hydrogel droplet containing

human cells (60 nL)

Hydrogel
bottom

Fig. 1.5 Schematics of the DataChip for organ-­specific toxicity.

Fig. 1.6 The DataChip—A human 3D cell culture microarray platform: (A) Microscopic pictures
of the sandwiched chips (left) and live Hep3B human hepatoma cell line encapsulated on the
micropillar chip (right), (B) Scanned image of the entire micropillar chip with live Hep3B cells
stained with a green fluorescent dye, (C) Culture of Hep3B cells on the microwell chip over time,
forming 3D spheroids.

1.2.3 Enzyme Microarrays

The MetaChip with active human (or rat) metabolizing enzymes (mixtures or human
liver microsomes) encapsulated in a hydrogel matrix has been developed to investi-
gate drug metabolism and enzyme inhibition by compounds (Fig. 1.7). Typical
DMEs printed in the MetaChip include cytochrome P450 (CYP450), UDP-­
glycosyltransferase (UGT), sulfotransferase (SULT), and glutathione S-transferase
(GST), among others (Table 1.3). The MetaChip allows the production and screen-
ing of human (or animal) metabolites on the chip platform because enzymatic
kinetic parameters are comparable with conventional approaches. Since it relies on
nanoliter-scale enzyme printing, the approach can save the expense of reagents and
test compounds significantly (ca. 100-fold reduction in assay volume compared to
the 96-well plate counterpart).
8 1 Overview of Microarray Bioprinting Technology

Table 1.2 Breadth of cell lines used on the DataChip

Cell type Name
Human primary cells Hepatocytes
Astrocytes
Human transformed cell lines Hep3B, HepG2, THLE-2 (hepatoma)
A293 (adenocarcinoma)
MCF7 (epithelial adenocarcinoma)
BT474 (Her2 overexpressing breast adenocarcinoma)
Mia PaCa-2 (pancreatic adenocarcinoma)
MCF10A (non-tumorigenic epithelial)
Rat primary cells Hepatocytes
Astrocytes
Cardiomyocytes
Rat transformed cell lines H411E (hepatoma)
DITNC1 (astrocytoma)
H9c2(2–1) (cardiomyocyte)
NRK-52E (renal proximal tubular)
Others 46C mouse embryonic stem cell
Embryonic rat neural stem cell
ReNcell VM human neural progenitor cell
E. coli

Compounds
(720 nL)

Metabolizing enzymes
(120 nL)

Fig. 1.7 Schematics of the MetaChip for drug metabolism and inhibition.

By printing test compounds and fluorogenic substrates such as BOMCC and

EOMCC from Life Technologies, it is possible to assess P450 inhibition by com-
pounds on the MetaChip [18]. The microwell chip can accommodate the measure-
ment of fluorescent dyes at different concentrations (Fig. 1.8). To demonstrate P450
inhibition on CYP450 microarrays, CYP2D6, CYP2C9, and CYP3A4 were printed
and immobilized on three different regions of a functionalized glass slide in 21 × 9
arrays. CYP450 isoform-specific inhibitors were printed interspersed with control
reaction spots on each region, and the extent of reaction was assessed [18]. Combined
1.2 Microarray Chip Platforms 9

Table 1.3 Breadth of drug metabolizing enzymes (DMEs) used on the MetaChip
Drug metabolizing enzymes Name
Human Phase I enzymes Cytochrome P450 (CYP1A2, 2C9, 2C19, 2D6, 2E1, 3A4, 3A5)
Flavin monooxygenase (FMO1, FMO3, FMO5)
Monoamine oxidase (MAO-A, MAO-B)
Others (Myeloperoxidase, Esterase, Epoxide hydrolase, DT
diaphorase, NQQ1)
Human Phase II enzymes UDP-glycosyltransferase (UGT1A1, 1A4, 2B4, 2B7)
Sulfotransferase (SULT1A1, 1A3, 1B1)
Glutathione S-transferase (GST)
N-Acetyltransferase (NAT1 and NAT2)
Pooled human liver enzymes Human liver microsomes (HLM)
Human s9 fractions (Hs9)
Rat Phase I enzymes Suite of rat CYP450s
Pooled rat liver enzymes Rat liver microsomes (RLM)
Rat s9 fractions (Rs9)

Fig. 1.8 Scanned image of the microwell chip with varying concentrations of fluorescent dyes
obtained by Cellomics ToxInsight In Vitro Toxicity (IVT) platform.

with the DataChip, the MetaChip has been used for analyzing metabolism-induced
toxicity. For example, the DataChip with Hep3B cells was stamped onto the MetaChip
containing metabolic enzymes (no enzyme, P450 Mix, All Mix, and HLMs) and com-
pounds (six compounds at six different dosages per compound) [3, 7, 9]. No enzyme
contains baculosome® negative control, P450 Mix is a mixture of major P450 iso-
forms, All Mix is a mixture of P450 Mix and Phase II enzymes, and HLMs are pooled
human liver microsomes. This approach allows one to measure not only the toxicity
of parent compounds but also the augmented toxicity by drug metabolites, which are
critically important to predict human drug metabolism in the livers after drug admin-
istration. Thus, it enables early detection of hidden toxicity of metabolites during lead
generation, lead optimization, and preclinical phases of drug development.
The Multizyme Chip, the variant of the MetaChip, encompasses microarrays
of biosynthetic enzymes on a chip to construct artificial synthetic pathways
(Fig. 1.9). Biosynthetically important enzymes are encapsulated into high-density
microarray spots that have the capacity to perform large numbers of enzymatic
reactions on a single chip, ultimately generating a large number of derivatives
starting from a lead compound. Thus, the Multizyme Chip is specifically designed
10 1 Overview of Microarray Bioprinting Technology

Fig. 1.9 Schematics of the

Multizyme Chip for lead
compound optimization.

Lead compound
(100 nL)
Synthetic enzymes
(150 nL)

Table 1.4 Breadth of biosynthetic enzymes used on the Multizyme Chip

Enzyme Name
Peroxidases Lignin peroxidase
Soybean peroxidase
Chloroperoxidase
Bromoperoxidase
Oxidases Tyrosinase
Alcohol oxidase
Aryl alcohol oxidase
Cholesterol oxidase
Bilirubin oxidase
Dehydrogenases Alcohol dehydrogenase
Aldehyde dehydrogenase
Transaminases Transaminase
Amine transaminase
Others Oxynitrilase
Ketoreductase
Type III polyketide synthases (PKSs) Tetrahydroxynaphthalene synthase (THNS)
Chalcone synthase (CHS)
Kinases Tyrosine kinase

for high-throughput derivatization of lead compounds. It is well suited to trans-

form complex lead compounds (e.g., natural products and multifunctional syn-
thetics), as such molecules are difficult to modify selectively and controllably. In
combination with the DataChip, it can be used for cell-based efficacy and toxicity
screens for the rapid identification of biologically active derivatives. This approach
represents a new paradigm for lead optimization by tapping into nature’s vast
biocatalytic repertoire. A list of biosynthetic enzymes tested on the Multizyme
Chip is presented in Table 1.4. The strategy of generating diverse polyketide ana-
logs has been successfully demonstrated using the in vitro metabolic pathway
microarray consisting of type III polyketide synthases (PKSs) and post-PKS tailoring
enzymes [19, 22]. After ­synthesizing diverse polyketide analogs on the enzyme
1.2 Microarray Chip Platforms 11

microarrays, the efficacy of polyketide analog library was assessed by tyrosine

kinase assays in high throughput.

1.2.4 Virus Microarrays

The TeamChip encompasses microarrays of recombinant adenoviruses carrying

genes for DMEs, which is prepared by printing several recombinant adenoviruses in
a combinatorial fashion (Figs. 1.10 and 1.11). By sandwiching the DataChip with
the TeamChip, human cells on the micropillar chip can be infected by adenoviruses
in the microwell chip, leading to creating human cell microarrays expressing different

Fig. 1.10 Schematics of

the TeamChip for viral
gene transduction with the
DataChip.

Recombinant
adenoviruses
(800 nL)

Ad-RFP Mixture of Ad- Ad-GFP

GFP & Ad-RFP

Fig. 1.11 The scanned image of the DataChip containing THLE-2 cells after sandwiching with
the TeamChip containing recombinant adenoviruses carrying GFP/RFP genes (Ad-GFP/RFP).
Human cells (e.g., THLE-2, Hep3B, HepG2, and Beas-2b) in Matrigel spots were 100 % infected
at 60 multiplicity of infection (MOI).
12 1 Overview of Microarray Bioprinting Technology

combinations of metabolizing enzymes on the chip [12]. The TeamChip is being

developed to mimic the first-pass metabolism of the human liver and to predict
enzyme-specific hepatotoxicity, resulting from exposures to various chemicals,
environmental pollutants, biologics, and therapeutic molecules or drugs. Thus, the
reactivity of target compounds with individual human metabolizing enzymes or
combinations of enzymes in the human liver can be assessed and quantified at
speeds commensurate with early-stage predictive human toxicity assessment. The
TeamChip provides metabolizing enzymes to virtually any human cell type, thereby
providing predictions on xenobiotic toxicity in essentially any cell type in the body.
By adjusting the expression levels of the various metabolizing enzymes in human
cells to match the enzyme levels representative of a subgroup unique to an individ-
ual, the TeamChip chip could be tailored for analysis of drug compounds being
developed for specific subgroups of the population and for individual patients.
Similar to the TeamChip, high-throughput shRNA transfection on the DataChip has
been demonstrated using a reverse transfection approach [20]. The DataChip con-
taining CHO and 3T3 cells was transfected with toxic shRNAs in a high-throughput
manner and the percent cell death was measured as a proof of concept.

1.3 Experimental Procedures

The highly versatile microarray biochip platform is based on micropillar/microwell

structures made by plastic injection molding, which is a more robust and flexible
system for mammalian cell culture in 3D, enzymatic reactions, viral infection, and
compound screening compared to its counterpart, microtiter plates. The chip plat-
form doesn’t require constant mixing or multiple wash steps during a cell-based
assay, and solution exchanges can be done by simply transferring the micropillar
chip containing cells from one microwell chip to the other. The experimental proce-
dures are highly specific to each endpoint assay and can be modified to meet require-
ment of the assay [6]. Here we provide a simplified experimental procedure for the
DataChip/MetaChip for metabolism-induced toxicity assays (Fig. 1.12).

Fig. 1.12 Experimental • Chip preparation: cell, enzyme, virus spotting

procedures with the 1
DataChip/MetaChip for • Test compound spotting on the chip
toxicology screening, 2
• Stamping chips to transfer compounds to cells
3
• Chip incubation in a gas-permeable chamber
4
• Cell staining after chip incubation
5
• Chip scanning to obtain scanned images
6
• Data analysis to plot dose-response curves and
7 calculate IC50 values of test compounds
1.3 Experimental Procedures 13

1.3.1 Cell Spotting

To attach cell spots on the micropillar chip, a mixture of poly-L-lysine (PLL) and
barium chloride (BaCl2) is prepared by mixing a 1:2 volume ratio of 0.01 % (wt/vol)
PLL and 24 mM BaCl2. The DataChip is prepared by spotting 60 nL of the PLL/
BaCl2 mixture onto 532 micropillars followed by printing 60 nL of hepatic cell
suspension in alginate on top of the dried PLL/BaCl2 spots. While printing cells, the
micropillar chip is placed on a chilling chip deck at 4 °C to retard evaporation of
water in the spots. A suspension of cells in low-viscosity alginate is prepared by
mixing cell suspension in FBS-supplemented growth medium with alginate solution
in distilled water so that the final concentration of cells and alginate are 2–6 × 106
cells per milliliter and 0.75–1 % (wt/vol), respectively. After nearly instantaneous
gelation, cell spots on the 532 micropillars are immersed in growth medium in the
complementary 532 microwells by stamping. The sandwiched DataChip is incu-
bated in a CO2 incubator at 37 °C for 18 h prior to toxicity tests.

1.3.2 Enzyme and Compound Spotting

The MetaChip is transversely divided into typically four regions (A–D) by printing
120 nL of metabolic enzyme/Matrigel mixtures in the microwell chip laid on a chilling
slide deck at 4 °C. Specifically, regions A to D contain no enzyme as a test compound
only control, a mixture of human cytochrome P450 isoforms (P450 Mix), a mixture of
P450 Mix and human Phase II metabolizing enzymes (All Mix), and human liver
microsomes (HLM). Immediately after spotting, the MetaChip is stored in a −80 °C
freezer until use. For toxicity tests, six different compounds are printed in regions 1–6
of the MetaChip, creating twenty-four distinct regions on the chip, each region contain-
ing a 3 × 6 mini-array. Within each mini-array, six different doses of a compound are
assayed for toxicity. Thus, a single Data chip combined with a single MetaChip has the
capability to generate twenty-four dose response curves for six compounds and their
metabolites generated from three different metabolic conditions on the chip.

1.3.3 Chip Incubation, Staining, Scanning, and Data Analysis

After stamping the DataChip onto the MetaChip and incubating the combined chips
for 24 h in the presence of compounds followed by 48 h post-stamping incubation
in fresh growth medium, the DataChip containing cells is stained with a live/dead
viability/cytotoxicity kit from Life Technologies. Staining dyes such as calcein AM
and ethidium homodimer-1 are used to produce a green fluorescent response from
live cells and a red fluorescent signal from dead cells. The dried DataChip is scanned
with S+ Scanner to obtain fluorescent images of cell spots. Alternatively, a GenePix
Professional 4200 A scanner and Cellomics ToxInsight In Vitro Toxicity (IVT)
14 1 Overview of Microarray Bioprinting Technology

Fig. 1.13 (A) GenePix 4200 A microarray scanner from Molecular Devices and (B) Cellomics
ToxInsight In Vitro Toxicity (IVT) platform from Thermo Fisher are required to obtain scanned
images from the DataChip/MetaChip.

Table 1.5 Comparison of the DataChip/MetaChip platform with the 96-well plate platform
Compound Compound
solution solution/cell Incubation
Platform Spot volume Morphology volume ratio (nL/cell) time (h)
DataChip 60 nL (200 cells/ 3D cells in – – 24–72
(532-micropillar spot) alginate/
chip) Matrigel
MetaChip 120 nL (54 nM Metabolizing 720 nL 3.6 (720 nL/200 24
(532-microwell enzyme) enzymes cells)
chip) in alginate/
Matrigel
Well plate (96-well 100 μL (10,000 2D cell 100 μL 10 (100,000 24–72
plate) cells/well) monolayers nL/10,000 cells)

platform can be used (Fig. 1.13). The dose response curves and IC50 values are
c­ alculated from scanned images using ImageJ and S+ Chip Analysis.
The fundamentally distinctive advantages of the DataChip/MetaChip platform
are the result of product and process characteristics designed and engineered to
replicate the human metabolism process. Consequently, the platform features a 3D
cellular environment that mimics that of human tissues in vivo. It has the ability to
accommodate a diverse array of cells, metabolizing enzymes and enzyme mixtures,
and a broad array of drug candidates. Combined with platform aspects that optimize
cell growth, these system characteristics transform to robust competitive strengths
of high throughput, high predictive reliability, low cost, and consistent reproduc-
ibility of toxicity profiles (Table 1.5).

1.4 Advantages and Applications

The microarray chip technology offers several clear advantages over more conven-
tional in vitro toxicology screening tools. Specifically, it requires extremely small
amounts of enzymes, viruses, cells, compounds, and reagents for analysis. It is well
1.4 Advantages and Applications 15

suited for early stage HTS of compound libraries, which include drug candidates,
chemicals, cosmetic ingredients, and environmental toxicants. The cell encapsula-
tion technology developed is flexible and allows one to culture multiple cell types
from different organs in hydrogel droplets on the chip, thus providing more predic-
tive insight into potential organ-specific toxicity of compounds. 3D-cell cultures on
the chip may provide an environment that simulates the in vivo ECM conditions,
and therefore help to maintain the specific biochemical functions and morphological
features of human cells similar to those found in human organ tissues. Through the
use of recombinant adenoviruses, the chip enables the controlled expression of each
human metabolizing enzyme as well as various combinations of multiple enzymes
in human cells. The gene transduction technology on the chip can be extended to
other cell-based in vitro assays, including gain- and loss-of-function genomic screen-
ing. The chip could be tailored to different subgroups of the population and even to
individual patients by adjusting the expression levels of the various metabolizing
enzymes in human liver cells to match the enzyme levels representative of a subgroup
or unique to an individual. Using this technology, adverse responses of drug candi-
dates and their reactive metabolites by combinations of various DMEs in the human
liver can be assessed and quantified at speeds commensurate with early-stage human
toxicity tests. Ultimately, this approach would provide critical information needed
for the design of patient-specific treatment regimens, as well as for the identification
of pharmacologically safe and effective lead compounds for advancement to clinical
trials. With “high quality” information on chemical toxicity early on, pharmaceutical
companies can make better educated decisions on which compounds to take forward,
accelerate drug development times, and reduce investment in late-stage drug failures.
A list of microarray bioprinting-based assays is provided in Fig. 1.14, and more
information can be found in Chap. 8.

Microwell chip Micropillar chip

Combined
Chips
Enzyme, virus & Target
compound cells

• Enzyme inhibition • Metabolism-induced • Cellular toxicity

• Metabolic stability toxicity • Enzyme induction
• Cellular defense • Gene transduction
mechanisms • RNA interference
• Lead compound • In-cell immuno
optimization fluorescence assay

Biochemical & cell-based assays

High-throughput predictive human toxicology

Fig. 1.14 Examples of microarray bioprinting-based assays for high-throughput, predictive toxi-
cology screening and drug discovery.
16 1 Overview of Microarray Bioprinting Technology

References

1. Datar, A., Joshi, P., & Lee, M.-Y. (2015). Biocompatible hydrogels for microarray cell printing
and encapsulation. Biosensors, 5(4), 647–663. doi:10.3390/bios5040647.
2. Joshi, P., & Lee, M. Y. (2015). High content imaging (HCI) on miniaturized three-dimensional
(3D) cell cultures. Biosensors, 5(4), 768–790. doi:10.3390/bios5040768.
3. Lee, M.-Y., Park, C. B., Dordick, J. S., & Clark, D. S. (2005). Metabolizing enzyme toxicology
assay chip (MetaChip) for high-throughput microscale toxicity analyses. Proceedings of the
National Academy of Sciences, 102(4), 983–987. doi:10.1073/pnas.0406755102.
4. Lee, M. Y., & Dordick, J. S. (2006). High-throughput human metabolism and toxicity analysis.
Current Opinion in Biotechnology, 17(6), 619–627. doi:10.1016/j.copbio.2006.09.003.
5. Lee, M. Y., Clark, D. S., & Dordick, J. S. (2006). Human P450 microarrays for in vitro toxicity
analysis: Toward complete automation of human toxicology screening. Journal of the
Association for Laboratory Automation, 11(6), 374–380. doi:10.1016/j.jala.2006.08.003.
6. Lee, M.-Y., Dordick, J., & DS, C. (2010). Metabolic enzyme microarray coupled with minia-
turized cell-culture array technology for high-throughput toxicity screening. Methods in
Molecular Biology, 632, 221–237. doi:10.1007/978-1-60761-663-4.
7. Lee, D. W., Lee, M. Y., Ku, B., Yi, S. H., Ryu, J. H., Jeon, R., et al. (2014). Application of the
DataChip/MetaChip technology for the evaluation of ajoene toxicity in vitro. Archives of
Toxicology, 88(2), 283–290. doi:10.1007/s00204-013-1102-9.
8. Ziauddin, J., & Sabatini, D. M. (2001). Microarrays of cells expressing defined cDNAs.
Nature, 411(6833), 107–110. doi:10.1038/35075114.
9. Lee, M.-Y., Kumar, R. A., Sukumaran, S. M., Hogg, M. G., Clark, D. S., & Dordick, J. S. (2008).
Three-dimensional cellular microarray for high-throughput toxicology assays. Proceedings of
the National Academy of Sciences, 105(1), 59–63. doi:10.1073/pnas.0708756105.
10. Fernandes, T. G., Kwon, S. J., Bale, S. S., Lee, M. Y., Diogo, M. M., Clark, D. S., et al. (2010).
Three-dimensional cell culture microarray for high-throughput studies of stem cell fate.
Biotechnology and Bioengineering, 106(1), 106–118. doi:10.1002/bit.22661.
11. Fernandes, T. G., Kwon, S. J., Lee, M. Y., Clark, D. S., Cabral, J. M. S., & Dordick, J. S.
(2008). On-chip, cell-based microarray immunofluorescence assay for high-throughput analy-
sis of target proteins. Analytical Chemistry, 80(17), 6633–6639. doi:10.1021/ac800848j.
12. Kwon, S. J., Lee, D. W., Shah, D. A., Ku, B., Jeon, S. Y., Solanki, K., et al. (2014). High-­
throughput and combinatorial gene expression on a chip for metabolism-induced toxicology
screening. Nature Communications, 5, 3739. doi:10.1038/ncomms4739.
13. Lee, D. W., Yi, S. H., Jeong, S. H., Ku, B., Kim, J., & Lee, M. Y. (2013). Plastic pillar inserts
for three-dimensional (3D) cell cultures in 96-well plates. Sensors and Actuators B: Chemical,
177, 78–85. doi:10.1016/j.snb.2012.10.129.
14. Lee, D. W., Choi, Y. S., Seo, Y. J., Lee, M. Y., Jeon, S. Y., Ku, B., et al. (2014). High-throughput
screening (HTS) of anticancer drug efficacy on a micropillar/microwell chip platform.
Analytical Chemistry, 86(1), 535–542. doi:10.1021/ac402546b.
15. Lee, D. W., Choi, Y.-S., Seo, Y. J., Lee, M. Y., Jeon, S. Y., Ku, B., et al. (2014). High-­
throughput, miniaturized clonogenic analysis of a limiting dilution assay on a micropillar/
microwell chip with brain tumor cells. Small, 10(24), 5098–5105. doi:10.1002/smll.201401074.
16. Lee, D. W., Lee, M.-Y., Ku, B., & Nam, D.-H. (2015). Automatic 3D cell analysis in high-­
throughput microarray using micropillar and microwell chips. Journal of Biomolecular
Screening, 20(9), 1178–1184. doi:10.1177/1087057115597635.
17. Kang, J., Lee, D. W., Hwang, H. J., Yeon, S.-E., Lee, M.-Y., & Kuh, H.-J. (2016). Mini-pillar
array for hydrogel-supported 3D culture and high-content histologic analysis of human tumor
spheroids. Lab on a Chip. doi:10.1039/C6LC00526H.
18. Sukumaran, S. M., Potsaid, B., Lee, M.-Y., Clark, D. S., & Dordick, J. S. (2009). Development
of a fluorescence-based, ultra high-throughput screening platform for nanoliter-scale cyto-
chrome p450 microarrays. Journal of Biomolecular Screening, 14(6), 668–678.
doi:10.1177/1087057109336592.
References 17

19. Kwon, S. J., Lee, M. Y., Ku, B., Sherman, D. H., & Dordick, J. S. (2007). High-throughput,
microarray-based synthesis of natural product analogues via in vitro metabolic pathway con-
struction. ACS Chemical Biology, 2(6), 419–425. doi:10.1021/cb700033s.
20. Zhang, H., Lee, M. Y., Hogg, M. G., Dordick, J. S., & Sharfstein, S. T. (2012). High-throughput
transfection of interfering RNA into a 3D cell-culture chip. Small, 8(13), 2091–2098.
doi:10.1002/smll.201102205.
21. Park, T.-J., Lee, M.-Y., Dordick, J. S., & Linhardt, R. J. (2008). Signal amplification of target
protein on heparin glycan microarray. Analytical Biochemistry, 383(1), 116–121. d­ oi:10.1016/j.
biotechadv.2011s.08.021.Secreted.
22. Kwon, S. J., Mora-Pale, M., Lee, M. Y., & Dordick, J. S. (2012). Expanding nature’s small
molecule diversity via in vitro biosynthetic pathway engineering. Current Opinion in Chemical
Biology, 16(1–2), 186–195. doi:10.1016/j.cbpa.2012.02.001.
Chapter 2
Microarray Spotter and Printing Technologies

Akshata Datar, Dong Woo Lee, Sang Youl Jeon, and Moo-Yeal Lee

Contents
2.1 Introduction........................................................................................................................ 20
2.1.1 Contact Printing Techniques.................................................................................. 20
2.1.2 Non-contact Printing Techniques........................................................................... 21
2.2 Materials............................................................................................................................ 22
2.3 Components of S+ MicroArrayer...................................................................................... 22
2.3.1 Main Body Components........................................................................................ 23
2.3.2 Externally Connected Components........................................................................ 26
2.4 General Precautions for Operating S+ MicroArrayer........................................................ 27
2.5 Daily Operations of S+ MicroArrayer............................................................................... 28
2.5.1 Turning ‘ON’ the System....................................................................................... 29
2.5.2 Refilling the Pressure Bottles................................................................................. 29
2.5.3 Washing Tubes, Solenoid Valves, and Ceramic
Tips with Ethanol and Water.................................................................................. 30
2.5.4 Dispensing Samples on the Micropillar/Microwell Chips
with a Work File........................................................................................... ......... 32
2.5.5 Replacing Solenoid Valves and Ceramic Tips....................................................... 34
2.5.6 Turning ‘OFF’ the System..................................................................................... 37
2.6 Detailed Programming for Normal Operation................................................................... 38
2.6.1 Generating Wash and Dry Sequences.................................................................... 38
2.6.2 Defining Well Plates............................................................................................... 38
2.6.3 Registering Chips................................................................................................... 41
2.6.4 Registering Spot Layouts....................................................................................... 44
2.6.5 Optimizing Dispensing Parameters Using Vision Inspection................................ 47
2.7 Summary............................................................................................................................ 50
References................................................................................................................................... 50

© The Author(s) 2016 19

M.-Y. Lee, Microarray Bioprinting Technology,
DOI 10.1007/978-3-319-46805-1_2
20 2 Microarray Spotter and Printing Technologies

2.1 Introduction

The early stages of drug discovery heavily rely on high-throughput screening (HTS)
of compound libraries to identify effective lead compounds. Biochemical and
cell-­based assays have been commonly performed with the assistance of robotic
liquid dispensing systems to rapidly test and identify the potential efficacy and tox-
icity of drug candidates [1, 2]. In an effort to save costs and reduce expensive
resources such as primary human cells and reagents, pharmaceutical industries have
focused on miniaturizing HTS assays by using higher density well plates including
384- and 1536-well plates, leading to a reduction in reagent volumes and an increase
in the speed of the liquid dispensing [1, 3].
Biological sample printing using microarray spotters is an important advance-
ment in the field of miniaturized assay development and HTS. Compared to tradi-
tional liquid dispensing systems, microarray spotters allow one to dispense
extremely small volumes (typically 200 pL to 950 nL) of biological samples (includ-
ing reagents, growth media, compounds, hydrogels, genes, proteins, viruses, and
cells) in microtiter plates, on glass slides, or on plastic chips. The process of cell
printing in hydrogels on a micropillar/microwell chip platform allows a precise
positioning of human cell spots, leading to the creation of more physiological rele-
vance of human cells grown in three dimensions (3D) [4, 5].
In this chapter, we will briefly introduce the advantages of various printing tech-
nologies and go over general precautions that have to be taken when printing various
biological samples with a microarray spotter. In general, microarray spotting is
divided into direct contact printing using pins and non-contact printing using solenoid
valves, piezoelectric nozzles, etc. [6, 7]. Although we have experience in ­operating
several microarray spotters, including MicroSys and PixSys from DigiLab and Nano-
Plotter from GeSim, we will provide detailed protocols on how to operate S+
MicroArrayer from Samsung Electro-Mechanics, Co. (SEMCO). S+ MicroArrayer
represents the most advanced microarray spotter for cell printing, which is specifically
designed to accommodate the micropillar/microwell chip. The principle of operating
and troubleshooting S+ MicroArrayer can be applicable to any solenoid-driven micro-
array spotters.

2.1.1 Contact Printing Techniques

Contact printing technologies mainly function on the principle of transferring bio-

materials via directly contacting tips to the surface of functionalized glass slides.
Examples of contact printing technology are microarray spotters with an array of
multiple contact pins and atomic force microscopes (AFMs) with scanning tips
(also known as dip-pen nanolithography). The contact microarray spotters have a
single pin or a pin array to transfer liquids from source plates, which is usually a
96- or 384-well plate, to destination glass slides. A droplet forms at the tip when
2.1 Introduction 21

the tip comes in direct contact with a liquid. Several types of tips have been
developed, depending on the material used and the capillary size inside the tip to
manage the volume of the droplet printed [8]. Factors that determine the volume of
the droplet are the capillary size, surface tension of the liquid, affinity of the liquid
with the glass slide, and the surface chemistry of the glass slide. Pin-based printing
is mainly used to print extremely small volumes of proteins and DNA on the sur-
face of glass slides [9, 10]. Although pin-based printing is straightforward and fast,
it is difficult to print colloidal suspensions (such as cells), and the droplet size may
be inaccurate and inconsistent depending on surface and liquid properties [9]. For
dip-pen nanolithography, AFM is used to deposit some biological samples, mostly
small molecules on the surface [7, 11]. Although this technology has capable of
printing small spots in 25–200 nm resolution, it is not widely used in printing
biological samples due to lack of multiplexing capabilities and limited detection
methods [7, 9].

2.1.2 Non-contact Printing Techniques

In non-contact printing, microarray spotters with solenoid valves are the most com-
monly used, which function on the principle of electromagnetic induction. Typical
solenoid valves consist of a metal rod that is surrounded by coiled wires that con-
duct an electric current. When voltage is applied across the coils, a magnetic field is
generated that forces the metal rod to act as a shaft that moves up and down. This
metal rod acts as a gate while printing biological samples. The setup is enclosed in
a nozzle that has constant sample supply and constant pressure maintained using
syringe pumps. When the shaft moves up and down, the outlet for printing biologi-
cal samples opens and closes. This is how the voltage applied to the solenoid valves
controls dispensing of liquid samples [6]. Unlike other dispensing systems, the sole-
noid valves can handle colloidal samples quite well. Typical dispensing volumes are
30–950 nL per droplet. One particular printer, S+ MicroArrayer, is equipped with
six solenoid valves for printing six samples simultaneously.
Piezoelectric nozzles that are used in conventional inkjet printers are also
commonly used in microarray printing with dispensing volume of 200–600 pL per
droplet. Unlike solenoid valves, the dispensing volume is controlled by the inner
structure of piezoelectric nozzles. Microarray spotters with piezoelectric nozzles
use electric pulses that control the expansion and contraction of a piezoelectric
membrane that acts as a pump to push biological samples through the tip [12, 13].
Thus, the dispensing volume can be increased not by changing voltages, but by
depositing multiple spots at the same location. Although piezoelectric nozzles are
capable of printing biological samples, its use is limited due to inconsistent printing
of colloidal samples and difficulty of controlling large dispensing volumes [14].
Piezoelectric nozzles are more commonly used for printing compounds dissolved in
DMSO, proteins, and amino/nucleic acids.
22 2 Microarray Spotter and Printing Technologies

Another popular type of bioprinting technology adapted in the field of tissue

engineering is laser-assisted bioprinting (LAB), also known as laser-induced for-
ward transfer (LIFT) [15]. Typically, LIFT consists of a pulsed laser beam, a glass
slide coated with a biological sample, and a receiving glass slide. The laser beam is
focused onto the absorbing layer, from which the heat is transferred to the biological
sample where a bubble is generated. The bubble expands until its explodes, and the
material in the bubble gets deposited onto the receiving glass slide [16]. However,
this type of printing needs delicate control of the laser spotting, intensity, moisture
and viscosity of the sample, thickness of the absorbing layer, and the printing speed
so that cell viability cannot be compromised [17].
Acoustic printers became popular when single cell-based experiments needed to
be carried out and demand for pL-sized droplets increased [18]. Acoustic printing
requires a source for ultrasound generation, a pool of a biological sample, and a
receiver glass slide. High intensity ultrasound is generated which is focused on air-­
liquid interface. When the energy exceeds the surface tension of the sample, liquid
is ejected from the pool of the sample [19]. The advantage of this technique is its
independence of nozzles and tubing. This technology gains in popularity for print-
ing compounds in DMSO. However, high intensity ultrasound in some cases may
cause an obstruction [20]. For example, printing delicate primary cells could be
problematic because high vibration energy may damage cell membranes and lower
cell viability.

2.2 Materials

• Solenoid valve (Lee Company)

• S+ MicroArrayer with six solenoid valves and ceramic tips (Samsung Electro-­
Mechanics, Co. or SEMCO, Suwon, South Korea)
• Micropillar and microwell chips (SEMCO, Suwon, South Korea)
• Ceramic tips with 150 μm orifice (SEMCO, Suwon, South Korea)
• Reagent alcohol 200 proof ACS grade (VWR)

2.3 Components of S+ MicroArrayer

To better explain the operation of the microarray spotter, we will classify it under
mechanical and software components. Although both components work hand-in-­
hand, their maintenance and parameter setting are different from each other.
Mechanical components are divided into the main body, the utility body, and exter-
nally connected parts that play an important role in chilling, rinsing, and sample
dispensing (Figs. 2.1 and 2.2).
2.3 Components of S+ MicroArrayer 23

Main body

Pressure bottle

Utility body

Waste bottle
Chillers

System water container

for rinsing valves and tips

Fig. 2.1 The picture of S+ MicroArrayer

2.3.1 Main Body Components

The main body components are housed in a chamber where biological samples are
loaded in a 96-well plate and printed on a micropillar/microwell chip at controlled
temperature and humidity. It mainly consists of a robotic arm, a dispensing head with
six solenoid values and ceramic tips, six syringe pumps, a droplet inspection camera,
a chip inspection camera, two chip-loading decks, a well plate deck, a vacuum pump,
a humidifier, a water bath with a sonicator, and a waste drainage basin (Fig. 2.3).
• Dispensing head unit
The dispensing head unit of the microarray spotter is mainly responsible
for moving and printing biological samples in accurate volumes and positions.
It consists of a chip alignment inspection camera, six solenoid valves connected
to syringe pumps through tubing, and six ceramic tips (Fig. 2.4). It is capable of
moving X, Y and Z directions, hence aspirating biological samples from
96-well/384-well plates and dispensing in different patterns on the micropillar/
microwell chip.
–– Solenoid valves: They are the main component behind microarray bioprinting
technology, which function on the principle of electromagnetic induction.
The voltage applied to the valves, creates a magnetic field that forces the gate
to open and close. The syringe pumps maintaining the pressure help the
samples to flow when the gate is open. Each solenoid valve can be controlled
individually.
24 2 Microarray Spotter and Printing Technologies

Controller

Main power switch

Computer

Humidity indicator
Air pressure
gauge
Reset button

Emergency
button

Computer switch Cover lid

Fig. 2.2 The utility body components of S+ MicroArrayer

–– Ceramic tips: They are necessary to dispense small droplets of biological

samples when the gate is open. They are made of an inert material for rinsing
with alcohol and sonication, hence preventing contamination while several
biological samples are printed.
–– Chip alignment inspection camera: It checks the position of micropillar/
microwell chip on the chilled chip-loading deck for accurate sample
dispensing.
• Syringe pumps: The main role of the six syringe pumps is to maintain pres-
sure when the samples need to be aspirated or dispensed. When the gates of the
six solenoid valves open to dispense the samples, the positive pressure main-
2.3 Components of S+ MicroArrayer 25

Syringe pump

Z
axis Sonicator

X axis Vacuum pump

Y axis

Chip deck

Robotic arm Dispensing Droplet inspection

head camera

Fig. 2.3 The main body components of S+ MicroArrayer

Solenoid valve

Sonicator

Chip alignment
inspection camera
Ceramic tips

Plate deck

Fig. 2.4 The dispensing head unit of S+ MicroArrayer

tained by the syringe pumps forces the samples out on the micropillar and
microwell chips.
• Droplet inspection camera: There are two cameras in the main body of the
microarray spotter. The chip alignment inspection camera is installed in the
dispensing head, and the droplet inspection camera is placed beside the chip-­
loading deck, which is used to optimize dispensing parameters such as air
26 2 Microarray Spotter and Printing Technologies

pressure and solenoid valve open time. When sample droplets are dispensed on
a hydrophobic plastic strip, the droplet inspection camera takes pictures of the
droplets and calculate dispensing volumes at the setting by measuring the height
and diameter of the droplet. The dispensing setting and actual dispensing volume
can be compared for parameter optimization.
• Water bath with a sonicator and a vacuum pump: These two components are
sequentially used to clean the tips and tubing that hold the samples between
aspiration and dispensing of samples. The water bath with a sonicator is respon-
sible for rinsing the surface of the tips and breaking down any bigger particles
that tend to settle and clog the tips using ultrasound. The vacuum pump is respon-
sible for drying the tips to prevent sample carry-over.

2.3.2 Externally Connected Components

• Pressure bottles: One pressure bottle contains water for sample printing and the
other bottle holds 70 % ethanol for rinsing and sterilization (Fig. 2.5). The tubing
has to be filled with water to transfer energy from the syringe pumps and push
the samples to flow for printing. The bottles are pressurized with air to supply
water and alcohol.
• Chillers: There are two chillers equipped in the microarray spotter, one for the
dispensing head for printing temperature-sensitive samples and the other for the
chip-loading decks for preventing spot drying, the 96-well plate deck, and the
two pressure bottles (Fig. 2.6).
• Waste bottle: It holds wastewater generated by rinsing and washing the solenoid
valves, the ceramic tips, and the tubing.

Fig. 2.5 The picture of the

two pressure bottles Air inlet tube (Blue)
containing distilled water
and 70 % ethanol

Water outlet

Ethanol outlet
2.4 General Precautions for Operating S+ MicroArrayer 27

Fig. 2.6 The picture of the

two chillers for preventing
water evaporation on the
chip and printing
temperature-sensitive
samples
Main power
Temperature switch
reading

Chip chiller Head chiller

2.4 General Precautions for Operating S+ MicroArrayer

• The S+ MicroArrayer uses surface cooling and condensation so that it requires

low humidity setting. In particular, water condensation on the moving part (not on
the chip-loading deck) or the circuit board would be detrimental. Do not increase
relative humidity higher than 60 % at room temperature. In addition,
maintain the temperature of the chip-loading deck between 4–10 °C to retard
evaporation of water in spots on the chip. It is extremely important to avoid
excess water condensation on the bottom layer (e.g., BaCl2/PLL spots) to
prevent spot detachment.
• Unlike MicroSys and PixSys microarray spotters, there is no vacuum applied to
hold the chips on the S+ MicroArrayer. Before spotting, make sure that all
chips lie flat on the spring-loaded chip deck and no obstacles are on the
work place. It is the main cause of the ceramic tips crashed.
• All ceramic tips should be positioned inside of 96-wells when aspirating
samples. If any one of the tips left outside of the 96-well plates when aspirating
due to wrong programming, then they will be broken. Always use the same kind
of the 96-well plates to avoid tip crashing due to Z height difference in
96-wells.
• Washing solenoid valves and ceramic tips with alcohol is essential to remove air
bubbles prior to sample dispensing. After alcohol washing, thorough rinsing
solenoid valves and ceramic tips with water is necessary to avoid enzyme
deactivation or cell death due to remaining alcohol.
• Do not change Z levels randomly to prevent the ceramic tips crashed into
the chip surface. The typical distance between the chip surface and the tip is
approximately 0.7 mm.
• Always check a streamline of water while rinsing solenoid valves. If the stream-
line is deflected from the vertical or water is beaded up at the ceramic tip end,
clean the ceramic tip with sonication. It is highly important to stop working with
a clogged tip immediately.
28 2 Microarray Spotter and Printing Technologies

• Solenoid valves can be used to print viscous solutions, but may not be suitable
for spotting organic solvents except alcohols because of incompatible plastic
parts in the solenoid valve. Aspirating DMSO with solenoid valves for a short
period of time will be okay. When compounds in DMSO is diluted with solenoid
valves, use a 384-well plate, and do not aspirate more than 20 μL of samples.
• Do not attempt to print spontaneously gelling materials in any case. In case
of printing cells in Matrigel, always maintain the dispensing head at low tem-
perature (4 °C) to prevent the gelation of Matrigel inside of the solenoid valves
while printing. An extensive rinse of tubes and solenoid valves with ice-cold
water is critical before and after printing Matrigel solution.
• Keep in mind that there is no “space” allowed when saving a file name, and do
not use special characters as well.
• If you have a new operational file made, test it without ceramic tips installed first.
Do not install the expensive tips for testing unknown work files.

2.5 Daily Operations of S+ MicroArrayer

Now that we understand the mechanical components of the microarray spotter, it is

important to know how to control these components and make them work as a
whole. ezAOI is a customized software developed to operate S+ MicroArrayer
(Fig. 2.7). This user interface helps us to control the microarray spotter and print

Chip window
Display
window
IO window

Chip alignment
inspection window Droplet inspection
window
Command window

Fig. 2.7 The main screen of ezAOI. The Command window has four groups—Program,
Equipment setup, User operation, and Work
2.5 Daily Operations of S+ MicroArrayer 29

samples on the micropillar and microwell chips in distinctive patterns. The main
screen that pops up displays six windows, including command window, slide
window, IO window, display window, chip alignment inspection window, and
droplet inspection window. The command window in itself has four tabs, includ-
ing program, equipment setup, user operation, and work. Each of these tabs has
different functions, which allow us to program the microarray spotter more
efficiently. For daily operation of the microarray spotter, follow the protocols
provided below.

2.5.1 Turning ‘ON’ the System

1. Open the air cylinder (or in-house air valve) and maintain the pressure of com-
pressed air at 100 psi.
2. Turn on the chiller. Note: For most of applications, there is no need to turn on
the humidifier because of surface cooling. Maintain the temperature of the
chip-loading deck between 4–10 °C to reduce evaporation of water in spots on
the chip. It is extremely important to avoid excess water condensation on the
bottom layer to prevent spot detachment.
3. Turn on the external switch in the utility body.
4. Turn on the computer and the monitor.
5. Prior to running the ‘ezAOI’, push ‘Reset button’ and reset XYZ coordinates.
Note: This step is essential to avoid malfunctioning of the microarray spotter.
Do not skip this step!
6. Run the ‘ezAOI’ software. The software will initialize the system automatically
when running. Note: Make sure that no obstacles are on the work place to
avoid the robotic arm crashing.

2.5.2 Refilling the Pressure Bottles

1. Select the ‘User Operation’ window.

2. Go to ‘Water Alcohol Change’ in the ‘Daily Operation’ box (Fig. 2.8).
3. Click ‘Release’ in the ‘Air pressure’ box to release air pressure in the pressure
bottles.
4. Turn the orange knob connected to the pressure bottles from ‘S’ to ‘O’ position,
which releases the pressure inside.
5. Open the lid of the pressure bottles and fill the pressure bottles with distilled
water and 70 % ethanol. Note: Make sure to have no precipitates at the bottom
of the pressure bottles. Dust and precipitates can be the source of solenoid
valve and tip clogging. Occasionally, the pressure bottles have to be cleaned
and sterilized. To minimize microbial contamination, distilled water with Clear
Bath® or sterilized distilled water can be used.
30 2 Microarray Spotter and Printing Technologies

Fig. 2.8 The screen of

‘User Operation’ > ‘Water
Alcohol Change’

6. Close the lid of the pressure bottles carefully. Note: Make sure that the ‘O’ ring
is properly placed before placing the lid.
7. Turn the orange knob of the pressure bottles from ‘O’ to ‘S’ position, which
close the pressure bottles completely.
8. Click ‘Initialize’ of the ‘Air pressure’ box in the ‘Water & Alcohol Filling Up’
box to apply air pressure in the bottles.

2.5.3  ashing Tubes, Solenoid Valves, and Ceramic Tips

W
with Ethanol and Water

1. Select the ‘User Operation’ window.

2. Select ‘Wash & Dry’ in the ‘Daily Operation’ box (Fig. 2.9).
3. Select the wash and dry sequence you generated.
4. Select the number of solenoid valves you want to wash and dry in the ‘Select
Nozzle’ box.
5. By clicking the ‘Run Wash & Dry’ button, run the wash and dry sequence.
Note: Washing tubes and solenoid valves with alcohol is essential to remove
air bubbles prior to sample dispensing. After alcohol washing, thorough
rinsing tubes and solenoid valves with water is necessary to avoid enzyme
deactivation or cell death due to remaining alcohol. An extensive rinse of tubes
2.5 Daily Operations of S+ MicroArrayer 31

Fig. 2.9 The screen of

‘User Operation’ > ‘Wash
& Dry’

and solenoid valves with ice-cold water is critical before and after printing
Matrigel solutions.
6. Check a streamline of water while washing. Note: If the streamline is deflected
from the vertical or water is beaded up at the tip end, clean the ceramic tip with
sonication.
32 2 Microarray Spotter and Printing Technologies

2.5.4  ispensing Samples on the Micropillar/Microwell Chips

D
with a Work File

1. Select the ‘Work’ window (Fig. 2.10).

2. Select the layout of spots on the chip in ‘Spot Layout’.

Fig. 2.10 The screen of

‘Work’ window
2.5 Daily Operations of S+ MicroArrayer 33

Fig. 2.11 The images of the original micropillar shown as “white spot” (left) and the micropillar
with the BaCl2-PLL bottom layer shown as “black spot” (right)

3. Select the well plate used for sample loading and aspiration in ‘Well Plate ID’.
Note: It is extremely important to use the same kind of the 96-well plates
always to avoid tip crashing due to Z height difference in 96-wells. Enter the
offset values in the X and Y directions to designate the location of samples in
96-wells. The exact location of sample wells will be appeared in the schematic
of the well plate in the ‘Display’ window.
4. Select the type of a chip in ‘Slide/Chip ID’ onto which samples are dispensed.
The schematic of the chip selected will be appeared in the ‘Chip’ window.
5. Select the color of spots on the chip under the camera in ‘Spot Color’. Typically,
it is white when nothing is dispensed on the chip or black when BaCl2/PLL is
dispensed (Fig. 2.11). Note: This color selection is necessary to ­automatically
identify the location of the chip with the chip alignment inspection camera.
6. Enter the air pressure used for sample spotting and the desired droplet volume
for solenoid valves in ‘Spot Volume’. Note: The typical air pressure used is 6
kilopascal (kPa) for 40–950 nL droplets.
7. Enter the air gap between the sample and water. Note: The typical air gap used
is 0 μL for 40–150 nL droplets and 20 μL for 700–950 nL droplets. For accu-
rate sample spotting, small or no air gap is allowed for small dispensing
volumes.
8. Select the wash and dry sequences you want before and after sample dispensing
in ‘Wash Dry’.
9. Select ‘No inspection’ in ‘Camera Inspection ID’. Note: For rapid sample dis-
pensing, we typically skip camera inspection of droplets. Camera inspection
of sample spotting is necessary to optimize dispensing parameters at first.
10. Enter the speed of the axis movement for solenoid valves in ‘Speed of Axis
Movement’. It is typically 10–20 mm/s.
11. Open the optimum dispensing parameters of solenoid valves obtained from the
vision inspection in ‘Optimum Tip Parameter’. Note: Manually change the
open time, if needed. Optimum dispensing parameters will not be changed for
each sample and droplet volume used, unless the solenoid valves are broken
34 2 Microarray Spotter and Printing Technologies

Table 2.1 Typical dispensing parameters for solenoid valves

Open time
Samples printed Spot volume (nL) Air pressure (kPa) Air gap (μL) (μs)
BaCl2/PLL 60 6 0 600
Cells in alginate 60 6 0 600
Enzymes in Matrigel 100 6 0 600
Virus in growth media 320 6 5 4500
Growth media 950 6 20 12,000

or contaminated. Use predetermined parameters without droplet inspection

to reduce the dispensing time. See Table 2.1 for typical dispensing parameters,
including air pressures, air gaps, and solenoid valve open times.
12. Place the chips on the chip-loading deck and select the location of chips in the
‘Display’ window onto which samples are dispensed. Note: It is extremely
important not to leave the chip unattended long time on the chilling chip deck
because excess water condensation on the bottom layer (e.g., BaCl2/PLL spot)
on the chip will facilitate cell spot detachment.
13. Click the ‘File Save’ button and enter a file name to save all information. Note:
There is no space and special character allowed in the file name.
14. Add proper amounts of samples in the designated 96-wells shown in the
‘Display’ window, place the 96-well plate on the deck, and then dispense
samples on the chips by clicking the ‘Run’ button.
15. Click the ‘Pause’ button to pause the work process. Note: In case of emergency
such as robotic arm and tip crashing, push the red ‘Emergency’ button to
stop the process (Fig. 2.2).
16. Click the ‘Reset’ button to reset the work process after clicking the ‘Pause’ button.
17. When dispensing samples with a saved work file, click the ‘File Open’ button,
select the work file, place the chips on the chip-loading deck, select the location
of desired chips to print in the ‘Display’ window, add proper amounts of
samples in the designated 96-wells, place the 96-well plate on the plate deck,
and then click the ‘Run’ button to print the samples on the chips.

2.5.5 Replacing Solenoid Valves and Ceramic Tips

In case of ceramic tips clogging and solenoid valves malfunctioning, ceramic tips
and solenoid valves have to be removed and replaced. Prior to removing tips and
solenoid valves, run ‘Solenoid Fix’, force to dispense large droplets multiple times
(typically 500 times), and rinse solenoid valves with water and sonication.
1. Select the ‘User Operation’ window.
2. Select ‘Wash & Dry’ under Daily Operation options (Fig. 2.12).
3. Select ‘Solenoid Fix’ and select all the nozzles from 1 through 6.
2.5 Daily Operations of S+ MicroArrayer 35

Fig. 2.12 The screen of

‘User Operation’ > ‘Wash
& Dry’
36 2 Microarray Spotter and Printing Technologies

Fig. 2.13 The screen of

‘User_Operation’ > ‘Move
to Tip Change’

4. Execute ‘Run Wash & Dry’.

5. Remove dust by running ‘Priming with water’.
6. In case of compound precipitates clogging solenoid valves, run ‘Priming with
alcohol’ and then execute ‘Run Wash and Dry’.
7. Repeat Steps 3–6 to remove the clogs from ceramic tips and/or solenoid
valves.
8. If the clogging problem persists, select the ‘User Operation’ window and select
‘Water Alcohol Change’
9. Select ‘Move to tip change’ under the ‘For maintenance’ tab (Fig. 2.13). The
robotic arm moves forward so that the dispensing head with ceramic tips and
solenoid valves is accessible for repair.
10. Unscrew the two screws that hold the clogged solenoid valve or the clogged tip
between them (Fig. 2.14).
11. Carefully pull out the malfunctioning solenoid valve with the ceramic tip from
the metal block by holding its connecting wires.
12. Separate the ceramic tip from the solenoid valve by gently pulling the tip.
13. Hold the opening of the ceramic tip across a spray bottle with ethanol and flush
the tip with ethanol to see if ethanol can pass through the orifice without any
obstruction.
14. In case of experiencing a clog, sonicate the ceramic tip until dust is removed. If the
clogging problem persists, replace the ceramic tip.
2.5 Daily Operations of S+ MicroArrayer 37

Fig. 2.14 The pictures of the dispensing head for replacing ceramic tips and solenoid valves

15. Attach a new ceramic tip to the solenoid valve, place them accordingly, and
connect the wire back in.
16. Before putting the screws in, make sure that the ceramic tips are leveled on the
metal block.
17. If the new ceramic tip still doesn’t print samples properly, the solenoid valve is
clogged, thus replacing it with a new one. Note: Run ‘Daily Washing’ after
replacing solenoid valves and before printing samples.

2.5.6 Turning ‘OFF’ the System

1. To avoid potential contamination issues, rinse tubes and solenoid valves with
ethanol and water thoroughly by running ‘Daily Washing’.
2. Close ‘ezAOI’ software.
3. Turn off the computer and the monitor.
4. Turn off the external power switch in the utility body.
5. Turn off the chillers. Note: Make sure to turn off the chillers to avoid excess
water condensation on the chip-loading deck and the dispensing head, causing
a short circuit by water.
6. Close the air cylinder (or in-house air valve). Note: Do not close the air cylinder
while ‘ezAOI’ is on.
38 2 Microarray Spotter and Printing Technologies

2.6 Detailed Programming for Normal Operation

These sections are prepared for advanced users who want to know how to program
the operation of S+ MicroArrayer.

2.6.1 Generating Wash and Dry Sequences

1. Select the ‘Program’ window.

2. Select ‘Wash & Dry’ in the ‘Program’ box (Fig. 2.15).
3. In the second tab select the wash and dry sequence you made in the ‘Daily
Operation’ window.
4. Select the process you want to add in the ‘Step’ box.
5. Click ‘Add Step’ in the ‘Step’ box.
6. If you want to start all over, click ‘New’ in the ‘Step’ box.
7. If you want to delete one of many wash and dry processes you generated, select
the process you want to delete, and then click ‘Remove Selected Step’.
8. Repeat Steps 4 and 5 according to the wash and dry sequences you want to
generate.
9. Change numbers and options in the ‘Value’ column as you need (Fig. 2.15 &
Table 2.2).
10. Click the ‘Save as’ button to generate the wash and dry sequence.
11. Enter a file name. Note: There is no space and special character allowed in
the file name.
12. If you want to modify the wash and dry sequence you generated, select the file
name in the box above the ‘Save as’ button and repeat Steps 8 and 9 as you need.
13. Click the ‘Save’ button to overwrite the changes in the same file.

2.6.2 Defining Well Plates

1. Select the ‘Program’ window.

2. Select ‘Well Plate’ in the ‘Program’ box.
3. Select the type of a well plate, 96 or 384 (Fig. 2.16).
4. Enter the well layout such as the number of wells and the well-to-well distance
(in mm) in X and Y directions. Refer to the 96-well plate figure for correct X
and Y directions (Fig. 2.17).
5. Select the ‘Equipment Setup’ window.
6. Select ‘Axis Position’ in the ‘Equipment Setup’ box (Fig. 2.18).
7. Move X, Y and Z axes to locate the first solenoid valve/ceramic tip at the
bottom of the A1 well (i.e., the tip locating approximately 3 mm above from the
bottom of the well) and then read the current position of X, Y and Z axes.
2.6 Detailed Programming for Normal Operation 39

Fig. 2.15 The screen of

‘Program’ > ‘Wash & Dry’

Note: This will be the position of the first ceramic tip for aspirating samples.
Do not change Z levels randomly to prevent the tips crashed into the well
surface.
8. Select ‘Well Plate’ in the ‘Program’ box again.
9. Enter ‘Well A1 Position’ determined from Step 7.
10. Enter minimum and maximum sample volumes allowed in either 96 or 384 wells.
11. Click ‘Save as’ button to register well plate information.
12. Enter a file name. Note: There is no space and special character allowed in
the file name.
40 2 Microarray Spotter and Printing Technologies

Table 2.2 Processes for the wash and dry sequences

Process Description Property Unit
Sample removal Syringe pumps are moving up to the zero position so – –
that all samples are removed from the solenoid valves
Priming water Tubes and solenoid valves are rinsed with water in Time s
the pressure bottle
Priming alcohol Tubes and solenoid valves are rinsed with 70 % Time s
ethanol in the pressure bottle
Priming system Tubes and solenoid valves are rinsed with system Time s
water water in the big container. This step is applied for
solenoid valves only
Syringe washing Tubes and solenoid valves are rinsed with water by Volume μL
operating syringe pumps
Syringe washing Tubes and solenoid valves are rinsed with water by Volume μL
with sonication operating syringe pumps while sonicating tips Sonication Yes/No
immersed in the water bath
Sonication The outside of ceramic tips are cleaned by sonication Time s
in the water bath Sonication Yes/No
Dry Ceramic tips are dried by vacuum Time s
Solvent wash Tubes and valves are washed with a solvent in the Volume μL
solvent bath. Make sure not to use hydrophobic
solvents. Only alcohols and DMSO are allowed
Pre-dispensing Prior to sample dispensing on the chip, aspirated Droplet Number
samples are pre-dispensed in the waste drainage
basin to equilibrate pressure difference

Fig. 2.16 The screen of

‘Program’ > ‘Well Plate’
2.6 Detailed Programming for Normal Operation 41

Fig. 2.17 The X and Y

directions in the 96-well
plate

13. If you want to modify the well plate you registered, select the file name in the
box above the ‘Save as’ button and repeat Steps 3, 4, 9, and 10.
14. Click the ‘Save’ button to overwrite the changes in the same file.

2.6.3 Registering Chips

1. Select the ‘Program’ window.

2. Select ‘Chip’ in the ‘Program’ box (Fig. 2.19).
3. Enter the information of the chip such as a width (in mm), a height (mm), a
margin (mm) from the top left corner, the number of blocks on the chip, a
block-to-block distance (mm), the number of spots in each block, a spot-to-spot
distance (mm) in X and Y directions. Refer to the slide specification figure for
correct X and Y directions (Fig. 2.20).
4. Enter the number of sacrificial spots. For example, ‘X distance’ 1 means that
there is each one of the sacrificial column inserted on the left and right side of
the chip, whereas ‘Y distance’ 1 indicates that there is each one of the sacrificial
raw inserted on the top and bottom side of the chip.
5. Check the ‘Summary information’ of the chip such as the chip margin and the
number of spots.
6. Click ‘Summary’ button to see the schematics of the chip you generated.
7. Select the ‘Equipment Setup’ window.
8. Select ‘Axis Position’ in the ‘Equipment Setup’ box (Fig. 2.21).
9. Move X, Y and Z axes to locate the ceramic tip approximately 0.7 mm above
the chip surface and then read the current position of the Z axis.
10. Select ‘Chip Inspection Camera’ in the ‘Equipment Setup’ box.
11. Select the box next to ‘Camera Live’ in the ‘Camera Inspection’ box (Fig.
2.21).
12. Move X and Y axes to locate the align inspection camera above the chip sur-
face. Move the Z axis to bring the camera into focus and then read the current
position of the Z axis.
42 2 Microarray Spotter and Printing Technologies

Fig. 2.18 The screen of

‘Equipment Setup’ > ‘Axis
Position’

13. Deselect ‘Camera Live’ in the ‘Camera Inspect’ box.

14. Select ‘Chip’ in the ‘Program’ box again.
15. Enter the Z position of ‘Alignment Camera’ determined from Step 12
(Fig. 2.19).
16. Enter the Z position of ‘Dispensing Height’ determined from Step 9.
2.6 Detailed Programming for Normal Operation 43

Fig. 2.19 The screen of

‘Program’ > ‘Chip’: (A)
top and (B) bottom panel.

17. Click ‘Save as’ button to register the chip.

18. Enter a file name. Note: There is no space and special character allowed in
the file name.
19. If you want to modify the chip information you registered, select the file name
in the box above the ‘Save as’ button and repeat Steps 3, 4, 15, and 16.
20. Click the ‘Save’ button to overwrite the changes in the same file.
44 2 Microarray Spotter and Printing Technologies

Fig. 2.19 (continued)

2.6.4 Registering Spot Layouts

1. Select the ‘Program’ window.

2. Select ‘Spot Layout’ in the ‘Program’ box.
3. Select the name of the chip you registered in the ‘Slide/Chip’ box (Fig. 2.22).
4. Check the chip information such as a margin, spot numbers, and a nozzle-to-­
nozzle distance.
2.6 Detailed Programming for Normal Operation 45

Fig. 2.20 The layout of

the micropillar/microwell
chip. The white circles
indicate either micropillars
or microwells on the chip
platform.

5. Click ‘1st Spotting Layout’ button to generate the first spotting pattern on the
chip with samples.
6. Select ‘Solenoid’ in the ‘Select Nozzle’ box as by default we use it for sample
spotting (Fig. 2.23).
7. Select the number of solenoid valves/ceramic tips and the layout used in the
‘Nozzle Layout’ box.
8. Select the type of the well plate used (96 or 384).
9. Select a 96-well (or a 384-well) where the first solenoid valve/ceramic tip will
be located for sample aspiration by clicking the well. According to the number/
layout of solenoid valves used, other wells will be automatically selected.
46 2 Microarray Spotter and Printing Technologies

Fig. 2.21 The screen of

‘Equipment Setup’ > ‘Chip
Inspection Camera’.

Note: For safety reasons, the entire well will not be selected when any one of
the nozzles is left outside of the well plate for aspiration.
10. Select the region of spots where the sample will be dispensed with the first
solenoid valve/ceramic tip by clicking and dragging on the chip layout.
11. Repeat Steps 9 and 10 according to the spot layout you want to generate.
12. Select the box beside ‘Set Sacrificial Spot’ to designate a sample well for print-
ing sacrificial regions. Note: Always the first solenoid valve/ceramic tip is used
for spotting sacrificial regions.
13. Select the sample well for spotting sacrificial regions.
14. Click ‘Close’ button when done.
15. Click ‘Save as’ button to register the spot layout.
2.6 Detailed Programming for Normal Operation 47

Fig. 2.22 The screen of

‘Program’ > ‘Spot Layout’

16. Enter a file name. Note: There is no space and special character allowed in
the file name.
17. If you want to modify the spot layout you registered, click the ‘Open’ button
and select the file name and repeat Steps 3 through 14.
18. Click the ‘Save’ button to overwrite the changes in the same file.

2.6.5  ptimizing Dispensing Parameters Using Vision

O
Inspection

1. Select the ‘User Operation’ window.

2. Select ‘Volume Check Vision’ in the ‘Daily Operation’ box (Fig. 2.24).
3. Select the number of nozzles you want to inspect with the camera in the
‘Nozzle’ box.
4. Select the well plate used and indicate the location of a 96 or 384 well contain-
ing a test sample in the ‘Sample Location’ box. The first solenoid nozzle will
be located in the well indicated, and the locations of the other solenoid
48 2 Microarray Spotter and Printing Technologies

Fig. 2.23 The screen of ‘Program’ > ‘Spot Layout’> ‘1st Spotting Layout’

valves/ceramic tips will be determined automatically according to the solenoid

valves selected.
5. Enter the air pressure used for sample spotting and the volume of the sample
loaded for inspection in the ‘Pressure & Sample Aspiration Volume’ box. The
typical air pressure used is 6 kPa for 60–950 nL droplet.
6. Enter aspiration conditions such as air gap and pre-pressurization in ‘Aspiration
Condition Parameter’. Refer to Table 2.1.
2.6 Detailed Programming for Normal Operation 49

Fig. 2.24 The screen of

‘User Operation’ > ‘Vision
Inspection’

7. Select tip rinsing conditions in water before and after sample loading prior to
sample dispensing in the ‘Before’ and ‘After’ boxes.
8. Enter the droplet volume for inspection in ‘Dispensing Volume’, the open time
of the solenoid valves selected in ‘Open Time’, and the number of spots dis-
pensed to test the conditions in ‘Number of Spot’ in the ‘Manual Dispensing
Setup’ box. Note: If the disparity between the droplet volume set up and the
average volume of the measured droplets at a certain condition is less
than the desired CV value (typically 5 %), then the vision inspection will be
successfully finished.
50 2 Microarray Spotter and Printing Technologies

9. After loading test samples, click the ‘Dispense’ button to test the dispensing
parameters manually with the camera.
10. The open time of each solenoid valve will be updated when the vision inspection
is successfully completed; otherwise 0 will be reported in the ‘Open Time’ box.
Repeat Steps 3–9 when failed.
11. After finishing all solenoid valve inspection, click ‘Save as’ button to save the
optimum dispensing parameters.
12. Enter a file name. Note: There is no space and special character allowed in
the file name.
13. If you want to modify the optimum dispensing parameters you saved, click the
‘Open’ button, select the file name, and then repeat Steps 3 through 11.
14. Click the ‘Save’ button to overwrite the changes in the same file.

2.7 Summary

Microarray bioprinting is an important advancement in the field of miniaturized

assay development and HTS. In this chapter, we briefly introduced the advantages
of solenoid-driven bioprinting over other printing technologies in terms of print-
ability of cells, the range of printing volumes, and printing precision. We also
went over general precautions that have to be taken when printing biological sam-
ples with a microarray spotter. Finally, detailed protocols on how to operate the
S+ MicroArrayer for printing various biological samples and how to program
specific functions for advanced users are provided. These protocols will ensure
user safety, proper operation of the equipment, and consistent printing results.
Although we discussed only the S+ MicroArrayer in this chapter, the principle of
operating and troubleshooting S+ MicroArrayer can be applicable to any solenoid-
driven microarray spotters.

References

1. Dunn, D. A., & Feygin, I. (2000). Challenges and solutions to ultra-high-throughput screening
assay miniaturization: Submicroliter fluid handling. Drug Discovery Today, 5(12), 84–91.
doi:10.1016/S1359-6446(00)80089-6.
2. Kapur, R., Giuliano, K. A., Campana, M., Adams, T., Olson, K., Jung., D., et al. (1999).
Streamlining the drug discovery process by integrating miniaturization, high throughput
screening, high content screening, and automation on the CellChip TM system. Biomedical
Microdevices, 2(2), 99–109.
3. Horman, S. R., To, J., Orth, A. P., & Caracino, D. (2013). High-content analysis of three-­
dimensional tumor spheroids: Investigating signaling pathways using small hairpin RNA.
Nature Methods, 10(10), v–vi. doi:10.1038/nmeth.f.370.
4. Justice, B. A., Badr, N. A., & Felder, R. A. (2009). 3D cell culture opens new dimensions in
cell-based assays. Drug Discovery Today, 14(1), 102–107. doi:10.1016/j.drudis.2008.11.006.
5. Comley, J. (2013). Progress made in applying 3D cell culture technologies. Drug Discovery
World., 15, 41–58.
References 51

6. Datar, A., Joshi, P., & Lee, M.-Y. (2015). Biocompatible hydrogels for microarray cell printing
and encapsulation. Biosensors, 5(4), 647–663. doi:10.3390/bios5040647.
7. Salaita, K., Wang, Y., & Mirkin, C. A. (2007). Applications of dip-pen nanolithography.
Nature Nanotechnology, 2(3), 145–155. doi:10.1038/nnano.2007.39.
8. Cho, E. J., & Bright, F. V. (2002). Integrated chemical sensor array platform based on a light
emitting diode, xerogel-derived sensor elements, and high-speed pin printing. Analytica
Chimica Acta, 470(1), 101–110. doi:10.1016/S0003-2670(02)00303-3.
9. Mota, C., & Moroni, L. (2015). Chapter 11—High throughput screening with biofabrication
platforms. In A. Atala & J. J. Yoo (Eds.), Essentials of 3D biofabrication and translation
(pp. 187–213). Amsterdam: Elsevier. doi:10.1016/B978-0-12-800972-7.00011-6.
10. Dusseiller, M. R., Schlaepfer, D., Koch, M., Kroschewski, R., & Textor, M. (2005). An inverted
microcontact printing method on topographically structured polystyrene chips for arrayed
micro-3-D culturing of single cells. Biomaterials, 26(29), 5917–5925. doi:10.1016/j.
biomaterials.2005.02.032.
11. Zhang, L. G., Fisher, J. P., & Leong, K. W. (2015). 3D Bioprinting and Nanotechnology in
Tissue Engineering and Regenerative Medicine. Amsterdam: Elsevier. doi:10.1016/
B978-0-12-800547-7.00004-7.
12. Boland, T., Xu, T., Damon, B., & Cui, X. (2006). Application of inkjet printing to tissue engi-
neering. The Biochemical Journal, 1(9), 910–917. doi:10.1002/biot.200600081.
13. Cui, X., Boland, T., D’Lima, D. D., & Lotz, M. K. (2012). Thermal inkjet printing in tissue
engineering and regenerative medicine. Recent Patents on Drug Delivery & Formulation, 6(2),
1–13. doi:10.2174/187221112800672949.
14. Nishiyama, Y., Nakamura, M., Henmi, C., Yamaguchi, K., Mochizuki, S., Nakagawa, H., et al.
(2009). Development of a three-dimensional bioprinter: construction of cell supporting struc-
tures using hydrogel and state-of-the-art inkjet technology. Journal of Biomechanical
Engineering, 131(3), 1–6. doi:10.1115/1.3002759.
15. PK, W., Ringeisen, B. R., Callahan, J., Brooks, M., Bubb, D. M., Wu, H. D., et al. (2001).
The deposition, structure, pattern deposition, and activity of biomaterial thin-films by matrix-­
assisted pulsed-laser evaporation (MAPLE) and MAPLE direct write. Thin Solid Films, 398–399,
607–614. doi:10.1016/S0040-6090(01)01347-5.
16. Devillard, R., Pagès, E., Correa, M. M., Kériquel, V., Rémy, M., Kalisky, J., et al. (2014). Cell
patterning by laser-assisted bioprinting. Methods in Cell Biology, 119, 159–174. doi:10.1016/
B978-0-12-416742-1.00009-3.
17. Catros, S., Guillotin, B., Bačáková, M., Fricain, J. C., & Guillemot, F. (2011). Effect of laser
energy, substrate film thickness and bioink viscosity on viability of endothelial cells printed by
Laser-Assisted Bioprinting. Applied Surface Science, 257(12), 5142–5147. doi:10.1016/j.
apsusc.2010.11.049.
18. Demirci, U., & Montesano, G. (2007). Single cell epitaxy by acoustic picolitre droplets.
Lab on a Chip, 7(9), 1139–1145. doi:10.1039/b704965j.
19. Fang, Y., Frampton, J. P., Raghavan, S., Sabahi-Kaviani, R., Luker, G., Deng, C. X., et al.
(2012). Rapid generation of multiplexed cell cocultures using acoustic droplet ejection fol-
lowed by aqueous two-phase exclusion patterning. Tissue Engineering. Part C, Methods,
18(9), 647–657. doi:10.1089/ten.tec.2011.0709.
20. Demirci, U., & Montesano, G. (2007). Cell encapsulating droplet vitrification. Lab on a Chip,
7(11), 1428–1433. doi:10.1039/b705809h.
Chapter 3
Chip Platforms and Chip Surface Treatments

Parnian Bigdelou, Alexander Roth, Yana Sichkar, and Moo-Yeal Lee

Contents
3.1 Introduction........................................................................................................................ 54
3.1.1 Surface Modification of Micropillar and Microwell Chips................................... 54
3.1.2 Surface Modification of Glass Slides..................................................................... 56
3.2 Materials............................................................................................................................ 57
3.3 Protocols............................................................................................................................ 58
3.3.1 Cleaning Microwell Chips with Plasma for Air Bubble Removal......................... 58
3.3.2 Coating Micropillar Chips with Poly(maleic anhydride-­alt-­1-octadecene)
(PMA-OD) for Cell Printing.................................................................................. 61
3.3.3 Cleaning the Surface of Glass Slides with Acid.................................................... 61
3.3.4 Coating of Glass Slides with Poly(styrene-co-maleic anhydride)
(PS-MA) for Cell Printing..................................................................................... 63
3.3.5 Coating of Acid-Cleaned Glass Slides
with Methyltrimethoxysilane (MTMOS) for Enzyme Printing............................. 64
3.3.6 Coating of Acid-Cleaned Glass Slides with (3-Aminopropyl)trimethoxysilane
(APTMS) for Protein Attachment......................................................................... 65
3.3.7 Measurement of Silanization on the Surface of the APTMS Slides
with Fluorescein Isothiocyanate (FITC) Labeling................................................. 65
3.3.8 Attachment of Glutaraldehyde on the Surface of the APTMS Slides
for Protein Attachment.......................................................................................... 66
3.3.9 Immobilization of Extracellular Matrix (ECM) Proteins
on Reactive Surfaces.............................................................................................. 67
3.4 Summary............................................................................................................................ 68
References................................................................................................................................... 69

© The Author(s) 2016 53

M.-Y. Lee, Microarray Bioprinting Technology,
DOI 10.1007/978-3-319-46805-1_3
54 3 Chip Platforms and Chip Surface Treatments

3.1 Introduction

There are three significant types of platforms used in microarray printing technol-
ogy: glass slides, micropillar chips, and microwell chips. The surfaces of these plat-
forms are treated for different applications. For example, a glass slide is ideal for
micropatterning and can be used for microscopic image analysis. Micropillar and
microwell chips are ideally suited for applications of high-throughput screening
(HTS) of compounds on 3D-cultured cells. In this chapter, we present various
methods by which these surfaces can be chemically modified for use in bioprinting
applications, with emphasis on the methods used for culture of cells encapsulated
in hydrogels.

3.1.1  urface Modification of Micropillar

S
and Microwell Chips

Since many cells have difficulty in attaching to the surface of glass slides, different
coatings on glass slides have been developed to improve cell attachment and growth,
such as using rat tail collagen. In addition, glass slides need very tedious and precise
cleaning procedures to make sure that any residues on the surface do not affect cel-
lular processes. To alleviate these issues and facilitate robust and reproducible cell
cultures, our research team developed disposable plastic chip platforms such as a
micropillar chip and a microwell chip (Fig. 3.1).
Both micropillar and microwell chips are made of polystyrene, a long carbon
chain polymer with benzene rings attached to every other carbon (Fig. 3.2).
Polystyrene has a very good optical clarity, is easily moldable, and can be sterilized
by UV or plasma irradiation. However, it is hydrophobic in nature without func-
tional groups; thus, cells and hydrogels with cells have difficulty in attaching on its
surface. In order to have proper attachment, the surface of polystyrene is modified

Fig. 3.1 Schematics of (A) a micropillar chip and (B) a microwell chip. (C) Image of micropillar
and microwell chips to scale against a microscopic glass slide
3.1 Introduction 55

PLL

PMA-OD

Polystyrene

Fig. 3.2 Modification of polystyrene surface by coating with poly(maleic anhydride-alt-1-­

octadecene) (PMA-OD) and subsequent covalent attachment of poly-L-­lysine (PLL)

Fig. 3.3 Modification of polystyrene surface via plasma treatment. A variety of different chemical
groups can be introduced by breaking the carbon chain backbone

to allow strong interactions between cell surface/hydrogel and polystyrene surface.

For example, cells can be attached more easily on negatively charged or amine-­
reactive surfaces in general, because extracellular matrices (ECMs) secreted by the
cells as well as serum proteins in culture media can be attached easily on these
surfaces, therefore providing a better surface for cell attachment.
One efficient method for surface modification of polystyrene involves using
amphiphilic co-polymers, utilizing the hydrophobic interactions of the polymer with
the polystyrene, while functional side chains can provide an anchor for hydrogels or
ECMs to attach. One such polymer, poly(maleic anhydride-alt-1-octadecene)
(PMA-OD) has been used for improved hydrogel attachment to micropillar chips.
When the cells are encapsulated in hydrogels, such as alginate, they can attach to
the surface of the polystyrene chip, through the interactions of PMA-OD and the
amine-­reactive residues on the hydrogel. Subsequent improvements to gel adhesion
can be made by using poly-L-lysine (PLL), utilizing the amine groups on PLL as an
anchoring point for hydrogel attachment, as shown in Fig. 3.2. The positively charged
PLL also interacts with negatively charged hydrogels and supports the attachment of
hydrogels on the micropillar chip [1].
The hydrophobic surface of polystyrene can cause an additional problem such as
an entrapment of air bubbles in the microwells. One effective method to convert the
hydrophobic surface of polystyrene to hydrophilic is to irradiate the surface with
56 3 Chip Platforms and Chip Surface Treatments

strong gas plasma under vacuum. This process generates highly energetic oxygen
ions and radicals, which oxidizes the polystyrene chains (Fig. 3.3) so that the
surface becomes hydrophilic and negatively charged [2–5]. This process uses an
ionized gas or plasma, to remove all organic matter from the surface of an object,
which is carried out in a vacuum chamber. During plasma treatment, excited gas
molecules and atoms in the plasma will emit UV light. By introducing different
gases into the plasma chamber such as O2, N2, Ar, He, and CF4, treatment can be
performed. There are no harsh chemicals involved in the process of plasma cleaning;
therefore, it is an environmentally safe process.

3.1.2 Surface Modification of Glass Slides

Glass slides have significant use for biological testing. The distinct advantages of
glass substrates are that they are relatively non-reactive, which is suitable for bio-
compatibility, and they are well suited for imaging purposes. Additionally, glass
slides are thermally stable, which makes them relatively easy to clean and modify
for future uses. The main struggle with glass surfaces is that their non-reactivity,
while potentially useful for culturing cell monolayers, makes them a poor surface
for hydrogel attachment. This hydrogel attachment is critical, as hydrogels and
other structures have the potential for cells to organize themselves into three-­
dimensional (3D) structures, mimicking the cell-ECM interactions found in vivo.
If hydrogels are not properly attached, any treatment towards the glass slide can
result in subsequent loss of cells encapsulated in hydrogels. Thus, it is a goal of
researchers to investigate methods for robustly immobilizing hydrogels onto glass
slides without compromising the integrity of the hydrogel or the normal behavior
of the cells.
Silicone precursor molecules are important to initiate this attachment between
glass slides and biological molecules. The strong covalent bonding between the
glass made of silicone oxide and the silicone precursor allows for the interaction of
biologics (such as hydrogels, proteins, and cell surfaces) with the functionalized
glass slide with the silicone precursor. One potential method that has been used to
improve protein and hydrogel attachment onto glass slides is the use of silanes with
amine groups, which can covalently attach biological compounds to silane surfaces.
The silicon in the silane group anchors well with the silicon glass slide, while func-
tional groups attached to the silane can interact with functional groups on hydrogels
and cell membranes to promote robust attachment. Such functional groups include
thiols, esters, N-hydroxysuccinimide (NHS)-esters, amines, and methacrylates [6].
Seo et al. demonstrated the use of thiol silanes and acryl silanes for the covalent
attachment of polyethylene glycol (PEG) hydrogels and streptavidin onto glass
slides without compromising the hydrogel integrity or the affinity of the streptavidin-­
biotin interaction [6]. While silanes with initial functional groups can be maintained
on a glass surface, H-terminal silanes can also be modified. Carvalho et al. demon-
strated the conjugation of alkenes with H-terminal silanes on glass via ultraviolet
(UV) light [7]. Luo et al. used a similar procedure, but applied acrylamide onto the
3.2 Materials 57

surface for a different surface chemistry [8]. The use of UV light is important for
treating glass surfaces, as it serves the dual function of modification of the surface,
and sterilization.
While UV light has been used as a mechanism for chemical modification of glass
surfaces, the thermal stability of silicon makes glass a suitable surface for thermally
controlled surface modification. Cheawchan et al. has demonstrated this by initially
modifying the glass surface with allyl groups before 70 °C heating was applied, and
the surface was reacted with nitrile n-oxide moieties [9]. The glass slide can then be
further modified for addition of functional groups that can aid in the attachment of
cells and hydrogels to the glass surface.
Another potential method for improving the attachment of biological molecules
onto glass slides involves initial surface treatments with gold. Gold can be easily
coated onto the surface of glass slides using various physical methods, but vapor
deposition is among the most common ways for entirely plating gold onto a glass
slide [10]. This is followed up with subsequent attachment of thiol groups [10].
The covalent attachment of sulfur on gold is easily achievable. This means that cysteine
and potentially methionine residues can be targeted on proteins for immobilization
purposes. Additionally, disulfide linkages can be used for a potential point of interac-
tion between gold and any thiol residues on the surface of the glass slide.

3.2 Materials

• Micropillar/microwell chips (25 × 75 × 2 mm from SEMCO, Suwon, South

Korea)
• Fisherbrand plain microscope slides (25 × 75 × 1 mm from Thermo Fisher
Scientific, Rockford, IL) or custom precision glass slides (25.35 × 75.50 × 1.10 mm
from Coresix Precision Glass, Inc., Williamsburg, VA)
• Wheaton glass 20 slide staining dish with removable rack (Fisher)
• National scientific 20 mL glass sample vials with PTFE-lined solid storage caps
(Fisher)
• Methyltrimethoxysilane (MTMOS from Sigma-Aldrich, St. Louis, MO)
• 3-(Aminopropyl)trimethoxysilane (APTMS from Sigma-Aldrich)
• Poly(maleic anhydride-alt-1-octadecene) (PMA-OD from Sigma Aldrich)
• Poly(styrene-co-maleic anhydride), 14 % (w/w) maleic anhydride content (PS-­MA
from Sigma-Aldrich)
• Glutaraldehyde
• Matrigel™, collagen I, poly-l-lysine, fibronectin, laminin, vitronectin, etc.
• Fluorescein isothiocyanate (FITC)
• Dulbecco’s phosphate-buffered saline (PBS) without Ca2+ and Mg2+
• Potassium phosphate buffer solution (200 mM, pH 8 from Life Technologies,
Carlsbad, CA).
• Nalgen MF75 series sterile disposable tissue culture filter unit (0.2 μm PES filter)
• Spin coater (Model PWM32, Headway Research, Inc., Garland, TX)
• Plasma cleaner (Harrick Plasma, PDC-001-HP)
58 3 Chip Platforms and Chip Surface Treatments

• Vacuum pump (Dekker Vacuum Technologies, RVR003H-01)

• Digital thermocouple gauge control unit (Duniway Stockroom Corp,
DTC531S-115-BX)
• S+ MicroArrayer with six solenoid valves and ceramic tips (SEMCO, Suwon,
South Korea)
• S+ Scanner with four filter sets (SEMCO, Suwon, South Korea)

3.3 Protocols

3.3.1  leaning Microwell Chips with Plasma for Air Bubble

C
Removal

Plasma treatment of microwell chips is carried out to change the surface property of
the microwell chip hydrophilic and remove air bubbles trapped in the microwells
when immersed in growth media or to sterilize the microwell chips. The protocol of
plasma treatment provided here is modified from the user manual from the manu-
facturer [11].
1. Make sure the machine and the thermocouple are plugged into an appropriate
power source, and make sure the vacuum pump is plugged into the plasma
cleaner. Turn on the power switch located on the vacuum pump, but do not touch
the ‘PUMP’ switch on the Plasma Cleaner’s panel just yet.
2. Turn on the Plasma Cleaner’s main power (‘POWER’ switch to ON).
3. Open the door of the Plasma Cleaner. Load microwell chips on the Pyrex glass
tray and put it inside the Plasma Cleaner’s chamber. Maximum ten chips can
fit on the tray (Fig. 3.4). Note: Place the chips facing up in the position for

Fig. 3.4 The position of

the microwell chips on the
Pyrex glass tray
3.3 Protocols 59

plasma treatment. In addition, wipe the back of the microwell chips with
70 % ethanol-­soaked paper towels to prevent microbial contamination.
4. Check that the 3-way valve is closed. It must be leveled in the vertical downward
position (Fig. 3.5).
5. Close and lean the door tightly against the Plasma Cleaner.
6. While holding the door, set the Plasma Cleaner’s ‘PUMP’ switch to ON. It is
located directly on the Plasma Cleaner (Fig. 3.6). The vacuum will maintain the
door tightly closed. Wait for 5 min.
7. Level the 3-way valve into the horizontal position such that it points to the left
(Fig. 3.7).
8. Slightly open the metering valve and wait for 1 min (Fig. 3.7). It will allow the air
to enter the Plasma Cleaner. Note: The metering valve should be operated in the
way the plasma intensity is noticeably increased and the pressure at the thermo-
couple is ranging from 800 to 1000 mTorr. Other pressures may also work
depending on the samples, but 900–1000 mTorr is ideal for microwell chips.
9. Turn on Plasma Cleaner’s RF switch to ‘HI’. Wait for 1 min. A fluorescence should
be observed through the Plasma Cleaner’s window. For a plasma cleaning process
with room air, purple-pink to violet plasma should be observed (Fig. 3.8).

Fig. 3.5 Vertical

downward position of the
3-way valve. The 3-way
valve is closed in this
position

Fig. 3.6 Plasma treatment of microwell chips is performed using (A) a vacuum pump, (B) a digi-
tal thermocouple gauge control unit, and (C) a plasma cleaner unit
60 3 Chip Platforms and Chip Surface Treatments

Fig. 3.7 The 3-way valve

is positioned in the
horizontal position such
that it points to the left

Fig. 3.8 Purple-pink

plasma generated during
the plasma treatment of
microwell chips

Fig. 3.9 The 3-way valve

is positioned in the
horizontal position such
that it points to the right
which allow the system to
vent

10. Process samples for 15 min. Note: The process time has to be optimized for
different applications.
11. At the end of the process, switch the RF level to ‘OFF’.
12. Level the 3-way valve to the vertical downward position (Fig. 3.5)
13. Turn off the vacuum pump by setting the Plasma Cleaner’s ‘PUMP’ switch
to OFF.
14. Slowly level the 3-way valve into the horizontal position such as it points to the
right (Fig. 3.9). The system will vent. Note: The Plasma Cleaner must be vented
3.3 Protocols 61

instantly. Otherwise, oil from the vacuum pump may back stream from the
pump and contaminate the system. Do not open the front door when the chamber
is under vacuum since this will damage the glass sample holder and any samples.
Additionally, a slow vent ensures that the samples and the Pyrex tray do not move
around the chamber when the samples are removed.
15. When the hissing noise from venting the chamber is not observed anymore,
open the front door.
16. Close the 3-way valve. Level it to the vertical downward position (Fig. 3.5).
17. Turn off the Plasma Cleaner’s main power (‘POWER’ switch to OFF).
18. Remove the chip samples.
19. Unplug the thermocouple at the end of the process, and keep it unplugged when
it is not used.

3.3.2  oating Micropillar Chips with Poly(maleic anhydride-

C
­alt-­1-octadecene) (PMA-OD) for Cell Printing

The amphiphilic PMA-OD coating is necessary to create an amine-reactive and

hydrophobic surface of the micropillar chip for cell printing.
1. Prepare a 1 % (weight/volume or w/v) PMA-OD in ethanol solution by dissolv-
ing 200 mg of PMA-OD in 20 mL of 100 % ethanol and letting it mix with a
small stir bar in a sealed glass sample vials with PTFE-lined solid storage caps
for at least 2 days for dissolution.
2. Dilute slightly turbid 1 % PMA-OD with 100 % ethanol at 1:10 ratio and mix
properly to prepare a clear 0.1 % PMA-OD stock solution.
3. Further dilute 0.1 % PMA-OD with 100 % ethanol at 1:10 ratio and mix properly
to prepare a 0.01 % PMA-OD working solution.
4. Add 2 mL of 0.01 % PMA-OD on the shallow staining plate and sandwich the
micropillar chip (Fig. 3.10).
5. Pick up and slant the micropillar chip at an angle of 45° for 1 min to allow excess
PMA-OD on the surface to drain, and wipe off the solution at the bottom of the
chip with Kimwipes.
6. Blow dry the bottom surface of the micropillar chips, then let them dry at
room temperature with the micropillars facing up in a sterile bioassay plate
(Fig. 3.11).

3.3.3 Cleaning the Surface of Glass Slides with Acid

For a very even silanization of a glass surface as well as uniform polymer coating,
removing all dirt and exposing silanol group (─SiOH) on the surface of glass slides
with strong acid are extremely important [12]. Unless pre-cleaned precision glass
62 3 Chip Platforms and Chip Surface Treatments

Fig. 3.10 Coating the surface of a micropillar chip with 0.01 % (w/v) PMA-OD in a shallow
staining plate

Fig. 3.11 Micropillar

chips facing up in a sterile
bioassay plate

slides from Coresix Precision Glass, Inc. are used, the acid-cleaning step is essential
for standard microscope slides.
1. Properly number plain microscope glass slides (25 × 75 × 1 mm) with a
diamond tip.
2. Wipe the glass slides three times with 70 % ethanol-soaked paper towels to
remove oil and solid particles on the slide surface and then clean ethanol with
3.3 Protocols 63

dry paper towels. Note: Do not use Kimwipes because it leaves lots of small
paper particles. After wiping with dry paper towels, the microscope slides
should not be wet with ethanol. It will leave ethanol stains on the slide surface.
3. Place the glass slides in a removable glass rack, immerse the rack in a Wheaton
glass dish filled with concentrated sulfuric acid (96.5 %) for 2 h, and then soni-
cate the Wheaton glass dish with the slides in acid for 10 min. Note: Always
work in a chemical fume hood while handling acid or organic solvent, do not
inhale acidic fume, and do not reuse sulfuric acid more than twice. Dispose of
all chemicals in a labeled waste container properly.
4. Rinse the rack containing the glass slides at least five times with de-ionized dis-
tilled water to remove excess sulfuric acid on the slide surface.
5. Immerse the rack containing the slides twice in a Wheaton glass dish filled with
acetone to remove excess water on the glass slide and dry the acid-cleaned glass
slides thoroughly under N2 gas stream.
6. Bake the acid-cleaned slides in an oven at 120 °C for 10 min to completely
remove water on the slide surface.
7. After drying, carefully inspect the glass slide for cleanliness. If the slide is not
clean enough (i.e., any dirt or particle on the slide surface), return to the Step 2.
Note: Always use tweezers after the Step 2. Do not use fingers to hold the
slides, as this will leave an oily residue on the slide.
8. Store the glass slides in a sterile petri dish (150 mm diameter, maximum 5 slides/
petri dish) or the bioassay plate until use. Never leave the clean slide uncovered
for long time.

3.3.4  oating of Glass Slides with Poly(styrene-co-maleic

C
anhydride) (PS-MA) for Cell Printing

To attach cell spots or prepare cell patterns on the surface of glass slides, it is
required to modify the surface amine-reactive while enhancing hydrophobicity to
prevent spot spreading. The amphiphilic PS-MA coating is suitable to change the
hydrophilic surface of glass slides to hydrophobic while providing amine-reactive
functionality for cell spot attachment.
1. Prepare a fresh 1 % (w/v) solution of PS-MA by dissolving PS-MA in anhydrous
toluene in a scintillation vial with a PTFE-coated lid. Note: Use disposable glass
serological pipets to aspirate toluene precisely.
2. Shake the solution in an incubating shaker at 50 °C, 150 rpm for 40 min until
PS-MA pellets completely dissolved. Note: After 30 min of shaking at 50 °C,
150 rpm, the solution usually become clear. However, leave the solution for
additional 10 min for complete dissolution of PS-MA pellets. The PS-MA solu-
tion should be prepared freshly as reactivity of maleic anhydride groups
decreases over time. Additional note: The PTFE-lined scintillation vial has to
be used for organic solvents. Never leave the vial with toluene in the incubat-
64 3 Chip Platforms and Chip Surface Treatments

ing shaker unattended due to a concern about an explosion at high tempera-

ture. Always keep the maximum shaker temperature below 70 °C for safety
precaution as the boiling point of toluene is 110.6 °C.
3. Prepare a 0.1 % (w/v) PS-MA solution by mixing 18 mL of anhydrous toluene
with 2 mL of the 1 % (w/v) PS-MA solution. Note: Prepare a fresh PS-MA
solution and use it immediately.
4. Prior to spin coating, bake acid-cleaned slides or amine-functionalized slides
in an oven at 120 °C for 10 min to completely remove moisture on the slide
surface.
5. Load 1 mL of the 0.1 % (w/v) PS-MA solution onto the slide and spin coating
the solution at 3000 rpm for 30 s (Model PWM32, Headway Research, Inc.,
Garland, TX).
6. Remove any excess PS-MA remaining on the back of the slide by wiping with
acetone-soaked paper towel.
7. Dry the PS-MA-coated slides overnight at room temperature in a sterile petri
dish and store until use.

3.3.5  oating of Acid-Cleaned Glass Slides

C
with Methyltrimethoxysilane (MTMOS) for Enzyme
Printing

For enzyme printing, amine-reactive surfaces may deactivate the enzymes printed.
Thus, the surface of acid-cleaned glass slides is typically changed to hydrophobic.
A commonly used method is to coat the surface of the glass slides with hydrophobic
silicone precursors such as MTMOS.
1. Prepare a fresh MTMOS-HCl sol solution by mixing 8 mL of MTMOS with
3.2 mL of 5 mM HCl, followed by vortex for 2 min and sonication for 10 min.
Note: The hydrolysis of MTMOS by HCl is an exothermic reaction. So be
careful not to burn skin. The mixture becomes transparent single phase after
vortex mixing (initially two phases). Always prepare fresh MTMOS-HCl sol
solution and use it immediately. Leaving the MTMOS-HCl sol solution longer
than 30 min before spin coating cause uneven coating.
2. Immediately before spin coating, mix 11.2 mL of the MTMOS-HCl sol solution
with 8 mL of potassium phosphate buffer solution (25 mM, pH 8) and use the
mixture within 15 min. Note: Upon addition of the phosphate buffer solution,
gelation of hydrolyzed MTMOS begins. Therefore, use the mixture immedi-
ately. Do not use the mixture when the solution becomes turbid, causing uneven
coating.
3. While spinning the acid-cleaned glass slide at 3000 rpm for 30 s, place 1.5 mL
of the mixture onto the slide.
4. Remove any excess MTMOS on the back of the slide using acetone-soaked
paper towel. Note: Make sure that rainbow-like refraction patterns are seen on
the surface after spin coating. If not, then the MTMOS coating is unsuccessful,
3.3 Protocols 65

presumably because either the acid-cleaned glass slide is too dirty or the
MTMOS-HCl-phosphate buffer mixture is too old.
5. To complete MTMOS gelation, dry the slides overnight at room temperature in
a sterile petri dish and store until use.

3.3.6  oating of Acid-Cleaned Glass Slides

C
with (3-Aminopropyl)trimethoxysilane (APTMS)
for Protein Attachment

Diluted APTMS solutions can be used to achieve amine functionalization on glass

slides, which can be used for protein attachment using a linker.
1. Place acid-cleaned glass slides in a removable glass rack, and then immerse the
rack in a Wheaton glass dish filled with acetone to completely remove moistures
on the surface.
2. After acetone rinsing, dry the slides by blowing nitrogen.
3. Prepare 5 % (v/v) of APTMS in toluene containing 0.5 % (v/v) of methylene
chloride (CH2Cl2). Note: Concentrated APTMS should not be exposed to air
and humidity to prevent oxidation and hydrolysis. Make a fresh APTMS solu-
tion each time for surface functionalization [13]. Toluene is a carcinogenic sub-
stance. Thus, do not inhale the fume, and handle it in a chemical hood.
4. Immerse the glass slides in a fresh APTMS solution and sonicate them for 1 h.
Discard the used APTMS solution after sonication.
5. Rinse the glass slides three times by dipping in toluene for 3 min and then remove
excess toluene by dipping in acetone. Note: Do not dry the slides while rinsing.
Drying can cause uneven silanization. Do not sonicate the slides while rinsing
because sonication doesn’t help to remove multiple layers of APTMS. Sonication
just causes non-uniform amine density on the surface.
6. Dry acetone on the slides by blowing nitrogen
7. Bake the slides at 120 °C for 1 h.
8. After cooling to room temperature, rinse the slides in ethanol three times for
20 min to remove uncoupled APTMS.
9. After drying, store amine functionalized glass slides in a petri dish until use.

3.3.7  easurement of Silanization on the Surface

M
of the APTMS Slides with Fluorescein Isothiocyanate
(FITC) Labeling

1. Allow DMSO (component B) and one vial of reactive FITC dye (component A)
from Life Technologies (FluoReporter® protein labeling kit) to warm to room
temperature immediately before starting the reaction.
66 3 Chip Platforms and Chip Surface Treatments

2. Add 50 μL of DMSO to the vial of reactive FITC dye. Note: Protect the reac-
tive FITC dye from light.
3. Prepare 200 mL of 50 mM potassium phosphate buffer (pH 8 from Life
Technologies) by diluting with de-ionized distilled water. Note: The buffer solu-
tion should not contain any ammonium ions or primary amines because the dye
is reactive with amine groups. Thus, Tris or glycine buffers are unsuitable.
4. Add 50 μL of reactive FITC dye in 200 mL of phosphate buffer and mix well
with a magnetic stir bar. Any remaining reactive dye stock solution should be
discarded.
5. Place amine-functionalized slides and acid-cleaned glass slides (as a control) in
a removable glass rack, and then immerse the rack in a Wheaton glass dish
filled with 100 % acetonitrile (ACN) for 5 min.
6. Rinse the slides in series for 5 min with an equal volume mixture of ACN and
water, and de-ionized distilled water to wet the surface.
7. Immerse the slides in the reactive FITC dye solution and incubate them for 1 h
with magnetic stirring at room temperature. Note: Do not sonicate the slides in
the reactive FITC dye solution because sonication can cause non-specific
binding of the dye on the surface. While incubating the slides, protect the
fluorescent dye solution from light by wrapping it up in aluminum foil.
8. Rinse the slides three times in de-ionized distilled water for 1 h each with magnetic
stirring to remove unbound dye.
9. Rinse the slides in acetone to remove excess water and wrap them up in alu-
minum foil. Note: Be careful not to scratch the surface of the slides with
aluminum foil.
10. Measure fluorescent intensity on the slides at 494 nm (excitation) and 518 nm
(emission).

3.3.8  ttachment of Glutaraldehyde on the Surface

A
of the APTMS Slides for Protein Attachment

1. Rinse the amine-functionalized slides in series for 5 min with 100 % acetonitrile
(ACN), an equal volume mixture of ACN and water, and de-ionized distilled
water to wet the surface. Sterilize de-ionized distilled water with Nalgen MF75
series sterile disposable tissue culture filter unit (0.2 μm PES filter).
2. Prepare 200 mL of 5 % (v/v) of glutaraldehyde in sterile Dulbecco’s phosphate-­
buffered saline (PBS) without CaCl2 & MgCl2.
3. Immerse the amine-functionalized slides in the glutaraldehyde solution and soni-
cate them for 1 h at room temperature. Discard the used glutaraldehyde solution
after incubation.
4. Rinse the slides twice by dipping in de-ionized distilled water for 5 min to
remove unbound glutaraldehyde. Use sterile de-ionized distilled water.
5. Rinse the slides in acetone and dry them by blowing nitrogen. Note: Use the
slides as soon as prepared because of instability of glutaraldehyde.
3.3 Protocols 67

3.3.9 I mmobilization of Extracellular Matrix (ECM) Proteins

on Reactive Surfaces

ECM proteins can be covalently attached on the reactive surfaces of glass slides
such as PS-MA-coated slides and glutaraldehyde-functionalized slides (Figs. 3.12
and 3.13).

HC
C
O O
H O

HC
Si O Si (CH2)3 N C CH C
Si OH O O
HO C CH

CH CH2
Si OH O

CH CH2
Si OH
NH2
OCH3
H3CO Si (CH2)3 NH2 Hydrophobic interaction
OCH3
O

HC
O O O C OH
H
Si O Si (CH2)3 NH2 Si O Si (CH2)3 N C HC
CH C NH Collagen
O O
O
HO C CH

CH CH2
O
CH2 CH CH CH
CH CH2

O C CO y
x O

Fig. 3.12 Immobilization of ECM proteins by APTMS and PS-MA treatment

Si
O
Si
O
Si
O O
(MeOH+HCl) & H2SO4 5% HC (CH2)3 CH

Si OH O O
H
Si O Si (CH2)3 N C (CH2)3 CH
Si OH
O
Si OH
OCH2CH3
10% H3CH2CO Si (CH2)3 NH2 NH2
OCH2CH3

O O H H
Si O Si (CH2)3 NH2 Si O Si (CH2)3 N C (CH2)3 C N
O O

Fig. 3.13 Immobilization of ECM proteins by APTMS and glutaraldehyde treatment

68 3 Chip Platforms and Chip Surface Treatments

Fig. 3.14 Picture of a

humid incubation chamber

1. Prepare ECM protein solutions (40 μg/200 μL) in sterile PBS (1×) and place
them on ice. Typical ECM proteins used are as follows: Matrigel™, collagen I,
fibronectin, laminin, vitronectin, etc.
2. Print 60 nL of ECM protein solutions on the reactive surfaces of glass slides
using S+ MicroArrayer.
3. Incubate the slides for 1 h at room temperature in a humid incubation chamber
to prevent evaporation of water (Fig. 3.14).
4. After drying the slides at room temperature in a biosafety cabinet, keep them in
a sterile petri dish for long-term storage. It may be stored at 2–8 °C for up to
several weeks under sterile conditions.
5. Before adding cell suspensions (e.g., Hep3B, MCF7, and A293 cells), rinse the
slides in sterile PBS twice with shaking for 30 min at 100 rpm to remove unbound
ECM proteins.
6. Culture the cells in proper growth media for 2–3 days in a CO2 incubator at
37 °C to prepare cell patterns or confluent cell monolayers.

3.4 Summary

Glass and polystyrene are both suitable substrates for cellular microarrays, but both
substrates require chemical modifications to ensure proper adhesion of cells or
hydrogels. In the case of plastic chips made of polystyrene, it involves coating with
amphiphilic, water-insoluble polymers. For glass slides, this means chemical
modification with silanes, or spin-coating with amphiphilic polymers. In both cases,
these only serve as preliminary steps to ensure adequate biocompatibility and
cell/hydrogel attachment, as subsequent steps are usually necessary to generate
microarrays. Nevertheless, these steps are necessary to create surfaces that can be
used for biological assay development with living cells.
References 69

References

1. Lee, M.-Y., Kumar, R. A., Sukumaran, S. M., Hogg, M. G., Clark, D. S., & Dordick, J. S. (2008).
Three-dimensional cellular microarray for high-throughput toxicology assays. Proceedings of
the National Academy of Sciences, 105(1), 59–63. doi:10.1073/pnas.0708756105.
2. Ramsey, W., Hertl, W., Nowlan, E., & Binkowski, N. (1984). Surface treatments and cell
attachment. In Vitro, 20(10), 802–808.
3. Curtis, A. S. G., Forrester, J. V., McInnes, C., & Lawrie, F. (1983). Adhesion of cells to poly-
styrene surfaces. The Journal of Cell Biology, 97(5I), 1500–1506. doi:10.1083/jcb.97.5.1500.
4. Amstein, C. F., & Hartman, P. A. (1975). Adaptation of plastic surfaces for tissue-culture by
glow-discharge. Journal of Clinical Microbiology, 2(1), 46–54.
5. Hudis, M. (1974). Plasma treatment of solid materials. In R. H. Hansen & A. T. Bell (Eds.),
Techniques and applications of plasma chemistry (p. 147). New York: John Wiley and Sons.
6. Seo, J. H., Shin, D. S., Mukundan, P., & Revzin, A. (2012). Attachment of hydrogel micro-
structures and proteins to glass via thiol-terminated silanes. Colloids and Surfaces B:
Biointerfaces, 98, 1–6. doi:10.1016/j.colsurfb.2012.03.025.
7. Carvalho, R. R., Pujari, S. P., Lange, S. C., Sen, R., Vrouwe, E. X., & Zuilhof, H. (2016). Local
light-induced modification of the inside of microfluidic glass chips. Langmuir, 32(10), 2389–
2398. doi:10.1021/acs.langmuir.5b04621.
8. Luo, N., Zhong, H., Yang, M., Yuan, X., & Fan, Y. (2016). Modifying glass fiber surface with
grafting acrylamide by UV-grafting copolymerization for preparation of glass fiber reinforced
PVDF composite membrane. Journal of Environmental Sciences, 39, 208–217. doi:10.1016/j.
jes.2015.11.010.
9. Cheawchan S, Uchida S, Sogawa H, Koyama Y, & Takata T. Thermotriggered catalyst-free
modification of a glass surface with an orthogonal agent possessing nitrile N-Oxide and
masked Ketene functions. Langmuir 2015;32(1): 309–315. doi:10.1021/acs.langmuir.5b03881.
(Fig. 1).
10. Tang, J., Peng, R., & Ding, J. (2010). The regulation of stem cell differentiation by cell-cell
contact on micropatterned material surfaces. Biomaterials, 31(9), 2470–2476. doi:10.1016/j.
biomaterials.2009.12.006.
11. User’s manual for PDC-001-HP (115V) or PDC-002-HP (230V) high power expanded plasma
cleaner (and optional PDC-FMG (115V) or PDC-FMG-2 (230V) PlasmaFlo). Ithaca, NY;
2015.
12. Cras, J. J., Rowe-Taitt, C. A., Nivens, D. A., & Ligler, F. S. (1999). Comparison of chemical
cleaning methods of glass in preparation for silanization. Biosensors & Bioelectronics, 14(8–9),
683–688. doi:10.1016/S0956-5663(99)00043-3.
13. No Title. Retrieved from http://www-schreiber.chem.harvard.edu/home/protocols/SMP.htm.
Chapter 4
Biological Sample Printing

Parnian Bigdelou, Alexander Roth, Akshata Datar, and Moo-Yeal Lee

Contents
4.1 Introduction...................................................................................................................... 71
4.2 Materials.......................................................................................................................... 73
4.3 Preparation of Stock Solutions......................................................................................... 74
4.4 Daily Operation of the S+ MicroArrayer for Dispensing Biological Samples................ 76
4.5 Sample Printing Protocols............................................................................................... 80
4.5.1 Enzyme Printing for Metabolism-Induced Drug Toxicity Assays....................... 80
4.5.2 Compound Printing.............................................................................................. 82
4.5.3 Cell Printing......................................................................................................... 85
4.5.3.1 Preparation of Cell Suspension in Growth Medium........................... 85
4.5.3.2 Cell Printing for 2D Monolayer Culture on the Micropillar Chip...... 86
4.5.3.3 Cell Printing in Alginate for 3D Cultures........................................... 87
4.5.3.4 Cell Printing in Matrigel for 3D Cultures........................................... 90
4.5.3.5 Cell Printing in a Mixture of Alginate
and Matrigel for 3D Cultures.............................................................. 93
4.5.3.6 Cell Printing in a Mixture of Alginate and Fibrinogen
for 3D Cultures................................................................................... 95
4.5.3.7 Cell Printing in PuraMatrix for 3D Cultures...................................... 96
4.5.4 Virus Printing....................................................................................................... 98
4.5.4.1 Measurement of Viral Titer in a 96-Well Plate................................... 98
4.5.4.2 Adenoviral Transduction on the Micropillar/Microwell Chip............ 99
4.6 Inspection of Cells Printed on the Micropillar Chips
Using a Bright-Field Microscope..................................................................................... 101
4.7 Coefficient of Variation (CV) and Z’ Factor for Assay Validation.................................. 102
4.8 Summary.......................................................................................................................... 103
References................................................................................................................................. 103

4.1 Introduction

Microarray spotters equipped with solenoid valves are capable of printing a wide
range of biological samples, including reagents, growth media, compounds, hydro-
gels, genes, proteins, viruses, and cells, for biochemical and cell-based assays.
Solenoid valve-driven bioprinting has clear advantages over other printing technologies

© The Author(s) 2016 71

M.-Y. Lee, Microarray Bioprinting Technology,
DOI 10.1007/978-3-319-46805-1_4
72 4 Biological Sample Printing

in terms of controlling sample volume dispensed and flexibility in the type of sam-
ples printed. Unlike robotic liquid dispensers that are more robust due to relatively
large orifice sizes and dispensing volumes, microarray spotters must take precau-
tions to prevent the clogging of solenoid valves and ceramic tips due to the extremely
small sample volume printed. Therefore, it is essential to avoid dust and large pre-
cipitates (e.g., compound precipitates and filamentous microbes), test the viscosity
of the material printed, and understand the mechanism of gelation for hydrogels
used for cell encapsulation. In general, temperature-sensitive hydrogels such as
Matrigel have a high risk of forming a gel spontaneously inside tubes, ceramic tips,
and solenoid valves with a slight change in temperature. Therefore, a lower tem-
perature must be maintained constantly across the chip-loading deck and the dis-
pensing head. The clogging may result in replacement of tubes, solenoid valves, and
ceramic tips, which can be an expensive affair. Gelation of hydrogels used for 3D
cell culture has to be occurred in two steps to avoid clogging. Typically, cross-
linkers/initiators are printed first on the micropillar/microwell chip platform and
then hydrogels containing cells are dispensed on the top to form a gel. Therefore,
ionic crosslinking (e.g., alginate with CaCl2, PuraMatrix with salts), affinity/cova-
lent bonding (e.g., functionalized polymers with streptavidin and biotin), photopo-
lymerization (e.g., methacrylated alginate with photoinitiators), and biocatalysis
(e.g., fibrinogen with thrombin) are favorable mechanisms of gelation. Developing
proper surface chemistry to attach cell spots in hydrogels robustly on the surface of
the micropillar/microwell chip is also critical. Covalent bonding (e.g., poly(maleic
anhydride-alt-1-octadecene) and poly-l-lysine), affinity (e.g., streptavidin and bio-
tin), and ionic interaction (poly-l-lysine and negatively charged alginate) are com-
monly introduced to enhance spot attachment on the chip. Viscosity of the material
printed is another important factor that affects the performance of printing, and
valve/tip clogging and rinsing. In general, biological samples with low viscosity
such as 10–30 centipoise (equivalent to 1 % or lower alginate in distilled water) are
preferred. Samples with high viscosity cannot be printed with solenoid valves and
are difficult to remove from tubes, solenoid valves, and tips by rinsing with water.
Therefore, it has to be diluted properly with either solvents or cell culture media to
ensure reproducible printing. In case of cell printing and encapsulation for 3D cul-
ture, maintaining cell suspension in hydrogels while printing is important to avoid
solenoid valves and tips clogging. In general, over crowded cells (typically seeding
density higher than 10 million cells/mL) can result in clogging issues. In addition,
cells tend to settle down quicker in a low viscosity solution. Finally, mechanical
strength of hydrogels over time needs to be considered to support long-term cell
culture. Peptide-based hydrogels (e.g., fibrinogen, Matrigel, and PuraMatrix) tend
to lose their strength over time due to degradation by matrix metalloproteinases
(MMPs) secreted by many mammalian cells. These MMPs are responsible for
hydrogel degradation and eventual spot detachment. To sustain cell spots for a
longer time and minimize spot detachment due to degradation, peptide-based
hydrogels can be mixed with more stable hydrogels such as alginate. When mixed,
the resulting hydrogel has to be transparent with minimal background fluorescence
4.2 Materials 73

and should not interfere high-content imaging (HCI) assays. The protocols provided
in this chapter will give researchers a guidance towards multiplex, 3D-cell based
assays on the micropillar/microwell chip platform.

4.2 Materials

• Poly maleic anhydride-alt-1-octadecene (PMA-OD) (Sigma-Aldrich)

• Alginate (Alginic acid sodium salt; Sigma-Aldrich)
• Barium chloride, dehydrate (BaCl2; Sigma-Aldrich)
• Poly-l-lysine, 0.01 % (PLL; Sigma-Aldrich)
• PureCol® bovine collagen solution, Type I, 3 mg/mL (Advanced Biomatrix)
• Matrigel (Corning)
• Fibrinogen from bovine plasma (Sigma-Aldrich)
• Thrombin from bovine plasma (Sigma-Aldrich)
• Dulbecco’s phosphate-buffered saline (DPBS) without Ca2+ and Mg2+
• 0.05 % Trypsin-EDTA (1×), phenol red (Life Technologies)
• 0.25 % Trypsin-EDTA (1×), phenol red (Life Technologies)
• Penicillin-Streptomycin, liquid (Life Technologies)
• Control Baculosomes® plus reagents (Life Technologies)
• P450-3A4 Baculosomes® plus reagents, rHuman (Life Technologies)
• P450-2D6 Baculosomes® plus reagents, rHuman (Life Technologies)
• P450-2C9 Baculosomes® plus reagents, rHuman (Life Technologies)
• P450-2E1 Baculosomes® plus reagents, rHuman (Life Technologies)
• P450-1A2 Baculosomes® plus reagents, rHuman (Life Technologies)
• P450-3A5 Baculosomes® plus reagents, rHuman (Life Technologies)
• P450-2C8 Baculosomes® plus reagents, rHuman (Life Technologies)
• P450-2C19 Baculosomes® plus reagents, rHuman (Life Technologies)
• P450-2B6 Baculosomes® plus reagents, rHuman (Life Technologies)
• Vivid® regeneration system, 100X (Life Technologies)
• UGT1A1, human, 5 mg/mL (Corning)
• UGT1A3, human, 5 mg/mL (Corning)
• UGT1A4, human, 5 mg/mL (Corning)
• UGT1A9, human, 5 mg/mL (Corning)
• UGT2B4, human, 5 mg/mL (Corning)
• UGT2B7, human, 5 mg/mL (Corning)
• UGT reaction mix A (containing 25 mM UDP-GA) (Corning)
• UGT reaction mix B (containing alamethicin) (Corning)
• Sulfotransferase isozyme 1A1 human (SULT1A1, Cypex)
• Sulfotransferase isozyme 1A3 human (SULT1A3, Cypex)
• Sulfotransferase isozyme 1B1 human (SULT1B1, Cypex)
• Glutathione S-transferase (GST) A1 human (Cypex)
• Glutathione S-transferase (GST) M1 human (Cypex)
74 4 Biological Sample Printing

• β-Nicotinamide adenine dinucleotide phosphate sodium salt (NADP;

Sigma-Aldrich)
• Adenosine 3′-phosphate 5′-phosphosulfate lithium salt hydrate, >60 % (PAPS;
Sigma-Aldrich)
• Glutathione reduced (GSH; Sigma-Aldrich)
• Micropillar/microwell chips (25 × 75 × 2 mm from SEMCO, Suwon, South
Korea)
• Staining plates for micropillar/microwell chips (SEMCO, Suwon, South Korea)
• Round bottom 96-well plate (Corning)
• S+ MicroArrayer with six solenoid valves and ceramic tips (SEMCO, Suwon,
South Korea)
• S+ Scanner with four filter sets (SEMCO, Suwon, South Korea)
Materials for Hepatocytes
• Human cryopreserved hepatocytes (ThermoFisher Scientific) Note: Cryopreserved
hepatocytes must be stored in the vapor phase of a liquid nitrogen freezer at less
than −150 °C to assure the long term viability of these cells. When stored under
the recommended conditions, the cryopreserved hepatocytes may be stable for up
to 5 years.
• CHRM™ cryopreserved hepatocyte recovery medium (ThermoFisher
Scientific)
• Thawing/plating supplements (ThermoFisher Scientific)
• Cell maintenance supplements (ThermoFisher Scientific)
• 1× DMEM with 4.5 g/L glucose and sodium pyruvate without L-glutamine and
phenol red (Corning)
• Corning® 500 mL vacuum filter (Corning)
• Trypan blue solution, 0.4 % liquid, sterile-filtered, cell culture tested
(Sigma-Aldrich)
• Matrigel (Corning). Note: Store Matrigel solution in aliquots at −20 °C and
keep frozen until used. Do not store in a frost-free freezer. Avoid multiple
freeze-thaws.
• Dulbecco’s phosphate-buffered saline (DPBS) without CaCl2 & MgCl2, 500 mL
(ThermoFisher Scientific)

4.3 Preparation of Stock Solutions

Preparation of CaCl2 Stock Solution

1. Prepare 200 mM CaCl2 stock solution in sterile distilled water by dissolving
1.5 g of CaCl2⋅H2O powder in 50 mL of deionized distilled water and passing it
through a 0.2 μm syringe filter.
2. Store the stock solution in a 4 °C refrigerator until use.
4.3 Preparation of Stock Solutions 75

Preparation of Matrigel® Stock Solution

1. Thaw Matrigel from Corning by storing the bottle in a 4 °C refrigerator
overnight.
2. While cold, quickly vortex the bottle to ensure that Matrigel is evenly
dispersed.
3. Place Matrigel on ice and take aliquots of 500 μL Matrigel with pre-cooled
pipette tip. Note: Do not change the pipette tip as Matrigel is viscous and stick-
ing to the side of the tip. It is important to avoid multiple freezing and thawing
of the Matrigel stock solution.
4. Immediately after taking aliquots in Eppendorf tubes, store them in a −20 °C
freezer until use. Note: Aliquots of Matrigel should be prepared quickly and
stored to prevent gelation. Gelled Matrigel may be re-liquefied if placed at 4 °C
on ice for 24–48 h.
Preparation of Thrombin Stock Solution
1. Add 20.7 mg of thrombin in 20.3 mL of DPBS to prepare a stock solution of 100
NIH units/mL. Thrombin comes in 20.7 mg of lyophilized powder which is
equivalent to 98 NIH units/mg of protein (i.e., 2028.6 NIH units of thrombin in
total). Note: Thrombin can be dissolved in distilled water, but a buffer solution
is preferred.
2. Prepare aliquots of 500 μL of thrombin in DPBS to avoid multiple freezing and
thawing of the enzyme stock solution.
3. Store the thrombin stock solution at −20 °C. Note: Aliquots of thrombin should
be prepared quickly and stored to prevent autolytic digestion.
Preparation of Fibrinogen Stock Solution
1. Prepare 0.9 % NaCl solution in distilled water and sterile filter it.
2. Warm the 0.9 % NaCl solution to 37 °C in a water bath.
3. Add 100 mg of fibrinogen in 1 mL of 0.9 % NaCl to prepare a stock concentra-
tion of 100 mg/mL.
4. Agitate the solution slightly by shaking to dissolve. If necessary, further dilute
the stock concentration to obtain a clear solution. Note: Fibrinogen solution
should not be vortexed. Fibrinogen solution cannot be stored, and should be
prepared fresh every time.
Preparation of BaCl2 Stock Solution
1. Prepare a 0.1 M barium chloride (BaCl2) solution by adding barium chloride
dehydrate in sterile de-ionized distilled water and vortex well for complete
dissolution.
2. After sterile filtering the solution using a 0.2 μm syringe filter, store the 0.1 M
BaCl2 solution in aliquots in a 4 °C refrigerator until use.
76 4 Biological Sample Printing

Preparation of Alginate Stock Solution

1. Prepare a 3 % (w/v) alginic acid solution by dissolving alginic acid sodium salt
in sterile de-ionized distilled water in a sterile glass sample vial.
2. Place a small stir bar in the vial and let it mix for at least 3 days for complete
dissolution.
3. Store the 3 % alginate solution in aliquots in a 4 °C refrigerator until use.

4.4  aily Operation of the S+ MicroArrayer for Dispensing

D
Biological Samples

1. Open the air cylinder (or in-house air valve) and maintain the pressure of com-
pressed air at 100 psi.
2. Turn on the chiller, the external switch in the utility body, the computer, and the
monitor.
3. Prior to running the ‘ezAOI’, push ‘Reset button’ and reset XYZ coordinates.
Note: This step is essential to avoid malfunctioning of the microarray spotter.
Do not skip this step!
4. Run the ‘ezAOI’ software (Fig. 4.1). Note: Make sure that no obstacles are on
the work place to avoid the robotic arm crashing.
5. Go to ‘Water Alcohol Change’ in the ‘User Operation’ box (Fig. 4.2).
6. Click ‘Release’ in the ‘Air pressure’ box to release air pressure in the pressure
bottles.

Fig. 4.1 The main screen of ezAOI. The Command window has four groups—Equipment setup,
Program, User operation, and Work
4.4 Daily Operation of the S+ MicroArrayer for Dispensing Biological Samples 77

Fig. 4.2 Screen of ‘User

Operation’
window > ‘Water Alcohol
Change’

7. Turn the orange knob connected to the pressure bottles from ‘S’ to ‘O’ position,
which releases the pressure inside.
8. Open the lid of the pressure bottles and fill the pressure bottles with distilled
water and 70 % ethanol.
9. Close the lid of the pressure bottles carefully. Note: Make sure that the ‘O’
ring is properly placed before placing the lid.
10. Turn the orange knob of the pressure bottles from ‘O’ to ‘S’ position, which
close the pressure bottles completely.
11. Click ‘Initialize’ of the ‘Air pressure’ box in the ‘Water Alcohol Change’ box
to apply air pressure in the bottles.
12. Click the ‘Start’ button in the ‘Daily Washing’ box. Note: Washing tubes and
solenoid valves with alcohol is essential to remove air bubbles prior to sample
dispensing. After alcohol washing, thorough rinsing of tubes and solenoid
valves with water is necessary to avoid enzyme deactivation or cell death due
to remaining alcohol. An extensive rinse of tubes and solenoid valves with
ice-cold water is critical before and after printing Matrigel solutions.
13. Check a streamline of water while washing. Note: If the streamline is deflected
from the vertical or water is beaded up at the tip end, clean the ceramic tip
with sonication.
14. Select the ‘Work’ window (Fig. 4.3).
78 4 Biological Sample Printing

Fig. 4.3 Screen of ‘Work’

window
4.4 Daily Operation of the S+ MicroArrayer for Dispensing Biological Samples 79

Fig. 4.4 Screen of ‘Display’ window

15. Click the ‘File Open’ button, and select the desired work file.
16. Enter the offset values in the X and Y directions to designate the location of
samples in 96-wells. The exact location of sample wells will be appeared in the
schematic of the well plate in the ‘Display’ window (Fig. 4.4).
17. Click the ‘Chip Loading’ button, and place the chips on the chilled chip deck
shortly before sample printing and select the location of chips in the ‘Display’
window onto which samples are dispensed (Fig. 4.4).
18. Add proper amounts of samples based on aspiration and sacrificial volumes
shown in the ‘Display’ window, in the designated wells, and place the well plate
on the deck. Note: Always add 20 μL more to the aspiration and sacrificial
volumes in the 96-well plate in case of evaporation.
19. Click the ‘Run’ button.
20. To avoid potential contamination issues after sample printing, rinse tubes and
solenoid valves with ethanol and water thoroughly by running ‘Daily Washing’.
21. Close ‘ezAOI’ software.
80 4 Biological Sample Printing

22. Turn off the computer, the monitor, the external power switch in the utility
body, and the chillers. Note: Make sure to turn off the chillers to avoid excess
water condensation on the chip-loading deck and the dispensing head, caus-
ing a short circuit by water.
23. Close the air cylinder (or in-house air valve). Note: Do not close the air c­ ylinder
while ‘ezAOI’ is on.

4.5 Sample Printing Protocols

4.5.1  nzyme Printing for Metabolism-Induced Drug Toxicity

E
Assays

A miniaturized enzyme array (Metabolizing Enzyme Toxicology Assay Chip or

MetaChip) has been prepared on the microwell chip to emulate the metabolizing
reactions in the human liver [1–3]. The MetaChip is an array of drug metabolizing
enzymes encapsulated in hydrogels, which is complementary to the micropillar chip
containing hepatic cell lines cultured in 3D (Data Analysis Toxicology Assay Chip
or DataChip) [1–3]. The MetaChip sandwiched with the DataChip has been used to
identify metabolism-induced toxicity of compounds. Briefly, six different com-
pounds can be printed in sections 1–6 of the MetaChip, each region containing a
3 × 6 mini-array (Fig. 4.5). Within each mini-array, six different doses of a com-
pound can be assayed for toxicity. For on-chip drug metabolism, human metaboliz-
ing enzymes can be transversely printed into four regions (A–D), each containing no
enzyme and different mixtures of enzymes encapsulated in Matrigel, thereby creating
24 distinct regions on the MetaChip (Table 4.1). Other human and rat metabolizing
enzymes can be spotted similarly on the microwell chip for drug metabolism.
1. Thaw a Matrigel aliquot overnight in a refrigerator at 4 °C. Note: Matrigel may
gel at slightly elevated temperatures in a refrigerator.
2. Keep metabolizing enzymes and Matrigel on ice before use, and use pre-cooled
pipettes, tips, and tubes when preparing a mixture of Matrigel and enzyme for
spotting.
3. Turn on the chiller connected to the dispensing head, the chip-loading deck and
the 96-well plate deck, set the temperature at 4–12 °C, and circulate cold water.
Note: The chilling temperature should be low enough to avoid water evapora-
tion in the spots. The chilling temperature may vary depending on the tem-
perature and humidity of the room. If there is severe water condensation
observed on the dispensing head or the chip-loading deck, then the chilling
temperature is too low. Make sure to turn the chiller connected to the dispens-
ing head off after Matrigel spotting to avoid extensive water condensation on
the dispensing head, potentially causing a short circuit.
4. Operate the daily washing program with ice-cold water and make sure that all
tubing, solenoid valves, and ceramic tips are cold prior to Matrigel spotting.
4.5 Sample Printing Protocols 81

Fig. 4.5 (a) A layout of the MetaChip (432 spots/chip) for in-situ drug metabolism. Specifically,
region A contains no enzyme as a test compound only control, regain B contains a mixture of
human Cytochrome P450 isoforms, region C contains a mixture of human phase II drug-­
metabolizing enzymes and P450 enzymes (i.e., all enzyme mixture), and region D contains human
liver microsome. (b) A layout of compounds printed in the MetaChip (432 spots/chip). Each dif-
ferent compound is printed in sections 1–6 of the MetaChip, and the concentrations (C1–C6) vary
within each mini-array. The grey dots represent the microwells

At least 2 cycles of running the wash program with cold water is necessary to
cool it down.
5. Prepare cold metabolizing enzyme solutions in Matrigel in a 96-well plate on ice
(see Table 4.1 for details). Note: Always use the same kind of the well plates to
avoid tip crashing due to Z height difference in wells. Always prepare fresh
enzyme-Matrigel mixtures on ice and use them immediately.
82 4 Biological Sample Printing

Table 4.1 Typical preparation of enzyme solutions for the MetaChip

Region Enzymes Co-substrates Matrix
A 120 μL Insect cell microsomes controla 120 μL RPMI 80 μL Matrigel
B 120 μL P450 mixtureb 120 μL Phase I cofactore 80 μL Matrigel
C 120 μL All Mixc 120 μL All cofactorf 80 μL Matrigel
D 120 μL Human liver microsomed 120 μL All cofactor 80 μL Matrigel
a
Insect cell microsomes (P450 negative control, 5 mg/mL) is purchased from BD biosciences and
used without dilution
b
P450 mixture contains 52 % 3A4 (1 μM), 20 % 2D6 (1 μM), 8 % 2C9 (1 μM), 5 % 2E1 (1 μM),
4 % 1A2 (1 μM), 4 % 3A5 (1 μM), 3 % 2C8 (1 μM), 3 % 2C19 (1 μM), and 1 % 2B6 (1 μM), all
from Life Technologies
c
All Mix contains 50 % P450 mixture and 50 % Phase II enzyme mixture
Phase II mixture contains 12 % UGT1A1 (5 mg/mL), 12 % UGT1A3 (5 mg/mL), 12 % UGT1A4
(5 mg/mL), 12 % UGT1A9 (5 mg/mL), 12 % UGT2B4 (5 mg/mL), 12 % UGT2B7 (5 mg/mL)
from BD biosciences, 4 % SULT1A1 (250 μg/50 μL), 4 % SULT1A3 (250 μg/50 μL), 4 %
SULT1B1 (250 μg/50 μL) from Cypex, 6 % GST (5 mg/200 μL) from Sigma, 5 % NAT1 (2.5 mg/mL),
and 5 % NAT2 (2.5 mg/mL) from BD biosciences
d
Human liver microsome purchased from BD biosciences and used without dilution
e
Phase I cofactor solution contains 50 % NADP (10 mM) and 50 % regeneration system from Life
Technologies kits.
f
All cofactor solution contains 50 % Phase I cofactor and 50 % Phase II cofactor
Phase II cofactor solution contains 50 % UDP-GA (2 mL BD solution A + 3 mL BD solution
B = 10 mM in 50 mM Tris-HCl buffer, pH 7.5), 20 % GSH (100 mM in 20 mM PBS buffer, pH 8),
20 % PAPS (25 mM in 20 mM PBS buffer, pH 8), and 10 % acetyl CoA (25 mM in 10 mM PBS
buffer, pH 8)

6. Print metabolizing enzyme solutions in Matrigel into microwells (typically

432 spots/chip, 100 nL per spot, 1.5 mm well-to-well distance) placed on the
cold chip deck. Maintain humidity under 50 % with the cooling chip deck to
retard evaporation of water in the spots.
7. Immediately after enzyme printing, place the microwell chip with metabolizing
enzymes in a petri dish (4 MetaChips in 150 mm-diameter petri dish) and store
in a - 80 °C freezer until use. Note: Do not dry the enzyme spots, causing enzyme
deactivation.
8. Operate the daily washing program at least twice with ice-cold water to make
sure that no Matrigel remains in tubing, solenoid valve, and ceramic tip.

4.5.2 Compound Printing

Test compounds can be printed in the MetaChip with drug metabolizing enzymes
for assessing metabolism-induced toxicity or in the microwell chip (with no
enzymes) for measuring parent compound toxicity. Typical printing volume of com-
pounds in growth media on the chip is 800–950 nL. It is critical to minimize the use
of organic solvents in the compound solutions. Typically, less than 0.5 % of DMSO
4.5 Sample Printing Protocols 83

in the final compound solutions is allowed. Thus, at least 200-fold dilution of

compounds in growth media is necessary. In addition, compounds have to be either
completely dissolved in growth media or form stable colloidal suspension. Big
precipitates of compounds in growth media will clog solenoid valves and ceramic
tips, resulting in unpredictable printing outcomes.
1. Dissolve a powder form of a compound in DMSO to prepare a compound stock
solution at 200-fold higher concentrations than the desired final concentration.
A typical concentration of the compound in DMSO is 50–300 mM. Note:
Nonvolatile and water-miscible DMSO is the most commonly used as a sol-
vent unless the compound is insoluble in DMSO. Check the Sigma-Aldrich
website for solubility data of compounds in organic solvents.
2. Add 20–40 mg of the compound in a glass scintillation vial (6 mL) and dis-
pense a predetermined volume of DMSO. Note: The volume of DMSO added
can be varied depending on the molecular weight and the toxicity of the com-
pound. Typically, 200–2000 μL of DMSO is added to prepare a compound stock
solution.
3. Sonicate the compound in DMSO to completely dissolve it. The solution can be
warmed up slightly to facilitate the dissolution process. Note: Do not vortex the
compound in DMSO as the undissolved compound will be attached on the side
wall of the glass vial, making it difficult to dissolve. Prior to placing the vial
with the compound in a warm water bath, check the Sigma-Aldrich website for
temperature stability of the compound. Typical temperature used is 40–70 °C.
4. If the compound is insoluble in DMSO, further dilute the compound stock by
adding more DMSO, or try different organic solvents. Ethanol or methanol is
the most commonly used as an alternative of DMSO.
5. Test solubility of the compound in growth media by adding 2 μL of the com-
pound in DMSO to 398 μL of growth media in an Eppendorf tube (1 mL) and
mixing well with sonication for 1–2 min. Note: Do not vortex it when dissolv-
ing in growth media. Evaluate solubility at 200-fold dilution as the maximum
allowable DMSO content is 0.5 %. The compound in growth media has to be
completely dissolved or form stable colloidal suspension.
6. Leave the Eppendorf tube with the compound in growth media for at least
30 min to see if big precipitates are formed. Note: The formation of compound
precipitates must be avoided to prevent solenoid valve and ceramic tip clog-
ging. In case of big precipitates formed, further dilute the compound stock
with DMSO.
7. After testing solubility in growth media, take aliquots of the compound stock in
DMSO (typically 40 μL aliquot) and store them in a −20 °C or −80 °C freezer
until use. Note: Do not freeze and thaw the compound stock in DMSO mul-
tiple times. The aliquot of the compound stock is for a single use for generat-
ing reproducible data. Compounds dissolved in volatile organic solvents
cannot be aliquoted and stored due to evaporation issues.
8. Prepare a compound-in-DMSO plate by thawing the frozen compound stock at
room temperature and adding 40 μL of the compound stocks in a round-bottom
96-well plate (e.g., the 96-wells A6, B6, C6, D6, E6, and F6 in Fig. 4.6A).
84 4 Biological Sample Printing

4-Fold serial dilution in DMSO

A1 A2 A3 A4 A5 A6
DMSO Compound
alone in DMSO

A1 A2 A3 A4 A5 A6

B1 B2 B3 B4 B5 B6

C1 C2 C3 C4 C5 C6

D1 D2 D3 D4 D5 D6

E1 E2 E3 E4 E5 E6

F1 F2 F3 F4 F5 F6

(A)

A1 A2 A3 A4 A5 A6

B1 B2 B3 B4 B5 B6

C1 C2 C3 C4 C5 C6

D1 D2 D3 D4 D5 D6

E1 E2 E3 E4 E5 E6

F1 F2 F3 F4 F5 F6

(B)

Fig. 4.6 (a) Preparation of a compound-in-DMSO plate in a 96-well plate by fourfold serial dilu-
tion. The 96-wells A6, B6, C6, D6, E6, and F6 contain 40 μL of compound stocks in DMSO and
the remaining 96-wells containing 30 μL of DMSO. For fourfold serial dilution, 10 μL of com-
pounds in DMSO are transferred to adjacent 96-wells and mixed well with a multi-channel pipette
sequentially until the last 96-wells with DMSO alone remains. Thus, the 96-wells A1, B1, C1, D1,
E1, and F1 containing DMSO alone are used as controls. (b) Preparation of a compound-in-growth
medium plate in a 96-well plate. For 200-fold dilution of compounds in growth media, dispense
298.5 μL of growth media in the 96-wells A1–F6, transfer 1.5 μL of serially diluted compounds in
the compound-in-DMSO plate to the 96-well plate containing growth media with a multi-channel
pipette, and mix the solutions well by aspirating and dispensing at least 20 times. Samples in the
96-wells A1, B1, C1, D1, E1, and F1 are transferred first, and then sequentially from low to high
concentrations to avoid compound carry-over
4.5 Sample Printing Protocols 85

Note: Make sure of the complete dissolution of the compound in DMSO

before serial dilution. In case of precipitates observed, sonicate or warm up
the compound stock in DMSO.
9. For fourfold serial dilutions of compound stock solutions in the 96-well plate,
add 30 μL of DMSO in other 96-wells, take 10 μL of the compound stock and
transfer it to the adjacent 96-wells, and then mix the solution well by aspirat-
ing and dispensing the mixture with a multi-channel pipette at least 10 times
(Fig. 4.6A).
10. This step is repeated sequentially until the last 96-wells with DMSO alone
remains. Solvent alone (without compound) is used as a control (e.g., the
96-wells A1, B1, C1, D1, E1, and F1 in Fig. 4.6A). Typically, 6 compounds
(each compound with 5 dosages and 1 solvent-alone control) are serially diluted
from high concentrations to low concentrations in the 96-well plate.
11. Prepare a compound-in-growth medium plate by dispensing 298.5 μL of growth
media in a round-bottom 96-well plate, adding 1.5 μL of serially diluted com-
pounds in DMSO with a multi-channel pipette, and mixing the solution well by
aspirating and dispensing (Fig. 4.6B). Note: To maintain final 0.5 % of DMSO
in growth media, 200-fold compound dilution is used. Make sure to use the
same kind of the 96-well plates to avoid tip crashing due to Z height differ-
ence in 96-wells.
12. This step is repeated from low to high concentrations of compounds by trans-
ferring 1.5 μL of the serially diluted compounds in DMSO to the adjacent
96-wells containing 298.5 μL of growth media with a multi-channel pipette and
then mixing the solution well by aspirating and dispensing at least 20 times.
Note: Due to potential compound carry-over, always add the diluted com-
pounds in DMSO from low to high concentrations in the 96-wells.
13. Place the compound-in-growth medium plate and the microwell chips in S+
MicroArrayer.
14. Print 950 nL of compound solutions (typically 6 compounds, each with 5 dosages
and 1 no compound control per chip) into the microwell chips using a work file
(drug_6tip_36block) (Fig. 4.5B).

4.5.3 Cell Printing

4.5.3.1 Preparation of Cell Suspension in Growth Medium

It is very important to prepare well suspended cells in growth media prior to mixing
with hydrogels for cell printing. Clumpy cells can clog solenoid valves and ceramic
tips, resulting in unreliable printing problems. Always make sure to resuspend cells
in the 96-well plate manually with a pipette before sample aspiration for uniform
printing.
86 4 Biological Sample Printing

1. Carefully remove the medium from a tissue culture flask with approximately
80 % confluent monolayer of mammalian cells.
2. Rinse the flask once with DPBS. Add DPBS slowly from the side to avoid
detaching the cells. Use 12 mL of DPBS for a T75 flask.
3. Aspirate out DPBS.
4. Add 1–2 mL of 0.05 % trypsin in the T75 flask, uniformly coat the cell mono-
layer with the trypsin solution, and incubate for 1–2 min at 37 °C.
5. Inspect the flask and ensure complete detachment of the cells by gently tapping
the side of the flask with the palm of your hand.
6. Apply 5 mL of growth medium pre-warmed at 37 °C to the flask.
7. Aspirate and dispense the growth medium containing cells rigorously at least
20 times to break apart big cell clumps. Note: The trypsin step and this step are
critical to prepare good cell suspension as clumpy cells can clog solenoid
valves and ceramic tips.
8. Gently rotate the flask to suspend the cells in growth medium.
9. Transfer the detached and suspended cells to a 15 mL conical tube.
10. Centrifuge the tube at 1200 rpm (typically 200 g) for 4 min to pellet the cells.
11. Aspirate out the entire supernatant gently without disturbing the cell pellet.
12. Add 2 mL of a complete growth medium supplemented with 10 % FBS and 1 %
Pen/Strep and resuspend the cells thoroughly with a 1 mL pipette.
13. While cells are in good suspension, take 5 μL of the cell suspension out and mix
it with 495 μL of growth media in an Eppendorf tube (1.5 mL) for cell counting
at 100-fold dilution.
14. After well mixing with a pipette, take 75 μL of cell suspension out from the
Eppendorf tube and load it in a Moxi Z cassette.
15. Insert the cassette in a Moxi Z mini automated cell counter to measure cell
density in the number of cells per milliliter. Actual cell density is 100-fold
higher than this density because of the dilution factor. Note: Do not vortex the
cell suspension for mixing as mammalian cells are fragile.
16. Dilute the cell suspension with growth media at a desired seeding density (typi-
cally, 6–10 million cells/mL) and then mix it with a hydrogel solution. The final
cell seeding density is typically in the range of 1–4 million cells/mL, varied
depending on printing volume on the chip and doubling time of the cells printed.
The required volume of cell suspension can be calculated from the program
(ezAOI) in S+ MicroArrayer by selecting an appropriate work file and the num-
ber of chips necessary.

4.5.3.2 Cell Printing for 2D Monolayer Culture on the Micropillar Chip

A variety of cell-based assays have been carried out on two-dimensional (2D)

monolayers of mammalian cells because of straightforward method of preparation,
low cost, and ease of use, compared to 3D cell culture [4]. Thus, 2D cell monolayer
culture in 96-well plates has been widely used as a standard tool for high-­throughput
screening of drug candidates. Similar 2D cell culture has been demonstrated on the
micropillar chip, and the protocol of cell printing is summarized below.
4.5 Sample Printing Protocols 87

1. Coat the micropillar chip with 2 mL of 0.01 % (w/v) PMA-OD in ethanol, and
let it dry at room temperature at least for 2–3 h (see the Sect. 3.3.2 in Chap. 3).
2. Prepare 0.0033 % poly-l-lysine (PLL) by mixing 0.01 % PLL stock (Sigma-­
Aldrich) with sterile deionized water at 1:2 ratio to obtain a final concentration
of 0.0033 % PLL. Note: In order for robust cell attachment onto the micropil-
lar chip, in addition to PLL, we may need collagen or laminin coating for
some cells.
3. Print 60 nL of 0.0033 % poly-l-lysine (PLL) on the micropillar chip using a
work file and dry it for 2–3 h. Note: Coating the micropillar chip with biocom-
patible polymer solutions is critical for robust attachment of cells on the
surface.
4. Prepare cell suspension in the complete growth medium at a desired seeding
density (1–6 million cells/mL).
5. Print 950 nL of the cell suspension in the microwell chips. Note: It is important
to resuspend the cells in the growth medium immediately before printing.
Printing cell suspension at too high seeding density may cause clogging of
solenoid valves and ceramic tips.
6. Sandwich the micropillar chip with polymer coating with the microwell chip
containing cell suspension.
7. Place the sandwiched chips with the microwell chip on the top in a humidified
petri dish with 10 mL of distilled water so that the cells can be settled on the
surface of the micropillars.
8. To robustly attach the cells on the surface of the micropillars, incubate the sand-
wiched chips in the humidified petri dish in a 5 % CO2 incubator at 37 °C for
6 h. Note: It is extremely important to prevent water evaporation during cell
incubation as drying can result in cell death.
9. After 6 h incubation, flip the sandwiched chips around to place the micropillar
chip on the top, which may prevent leaching the growth medium out.
10. Incubate the sandwiched chips overnight for cell growth.
11. After robust cell attachment on the micropillar chip, change the growth medium
every 1–2 days by printing fresh growth medium in the microwell chip, discard-
ing the microwell chip containing old growth medium, and sandwiching the
freshly prepared microwell chip. Note: The cells are cultured on the micropillar
chip until 80–90 % confluency.

4.5.3.3 Cell Printing in Alginate for 3D Cultures

Since 2D-cultured cells rapidly lose phenotypic properties and functions of tissues
and may not accurately emulate the microenvironment and cellular architecture
found in vivo, 3D cell culture strategies have been developed in efforts to foster
retention of phenotypic tissue functions. We have successfully demonstrated human
cell cultures in 3D on the micropillar/microwell chip platform using various hydro-
gels, including alginate, PuraMatrix™, fibrinogen, Matrigel, Geltrex, and photo-
crosslinkable, methacrylated alginate. In particular, alginate is an ideal printable
scaffold for tissue engineering because it is a naturally-derived hydrogel that is non-
toxic, non-immunogenic, and non-adhesive in nature [5]. Alginate is also able to
88 4 Biological Sample Printing

deliver soluble signaling molecules to cells, act as a support structure for cell growth
and function, and provide space filling for future tissue ingrowth after hydrogel
degradation [5]. Cells can easily be encapsulated and remain viable in alginate
hydrogels, and it has been widely used in a variety of tissue engineering applica-
tions [5]. Alginate is a negatively charged biopolymer which forms a gel when
interacting with a divalent cation such as Ca2+, Ba2+, Mg2+, etc. Some divalent cat-
ions such as Ba2+ can be relatively toxic to cells, thus requiring additional rinsing of
cells to remove excess Ba2+ ions. Alginate can be functionalized with methacrylate
groups for photopolymerization [5, 6]. Photocrosslinkable alginate has been used as
a biodegradable hydrogel system with independently adjustable physical, cell adhe-
siveness, and bioactive factor binding properties [7, 8]. Optimization of photoinitia-
tors and their concentrations are necessary to minimize basal toxicity of cells
printed. The protocol of cell printing in alginate using S+ MicroArrayer is summa-
rized below.
1. Coat the micropillar chip with 2 mL of 0.01 % (w/v) PMA-OD in ethanol and
dry it at room temperature at least for 2–3 h.
2. Prepare a mixture of 0.0033 % PLL and 25 mM CaCl2 by mixing 0.01 % PLL
with 37.5 mM CaCl2 in distilled water at 1:2 ratio. Note: For human cell lines,
a mixture of 0.0033 % PLL and 16 mM BaCl2 can be used to improve spot
attachment. BaCl2 is much stronger crosslinker, but more toxic against some
primary human cells and delicate cells.
3. Print 60 nL of the PLL-CaCl2 mixture on the micropillar chip using a work file
(60 nL PLL_CaCl2) and dry it for at least 2–3 h.
4. Print 950 nL of growth medium in the microwell chip placed on the chilling chip
deck at 4–12 °C using a work file (medium_6tip_6block_950nL). Note: Turn on
the chiller at least 1 h before sample printing.
5. Immediately after printing, incubate the microwell chips with growth medium in
a humid chamber with water to avoid water evaporation (Fig. 4.7).
6. Prepare 1–4 million cells/mL in 0.75 % alginate by mixing 1 mL of 2–8 million
cells/mL, 0.5 mL of 3 % alginate in distilled water, and 0.5 mL of growth medium
using a 1 mL pipette. Note: Do not vortex the mixture as mammalian cells are

Fig. 4.7 Picture of a

humid chamber with water.
The lid will be put in place,
and the chamber will be
completely sealed to
prevent water evaporation
from growth medium in
the microwell chip
4.5 Sample Printing Protocols 89

fragile. Mix well by aspirating and dispensing the solution repeatedly

(20 times) with a 1 mL pipette.
7. Place the micropillar chips with dried PLL-CaCl2 spots on the chilling chip
deck at 4–12 °C. Note: This is one of the most important steps in cell printing.
Do not leave the micropillar chips too long on the chilling chip deck unat-
tended as there might be severe water condensation observed on the chip
surface, which will affect spot attachment. The chilling temperature should
be adjusted to prevent spot drying as well as to avoid severe water evaporation
on the surface.
8. Add 100–200 μL/96-well of 1–4 million cells/mL in 0.75 % alginate in a
96-well plate. Note: Always use the same kind of the 96-well plates to avoid
tip crashing due to Z height difference in 96-wells.
9. Immediately before cell printing, resuspend the cells in 0.75 % alginate with a
pipette. Note: This is one of the most important steps in cell printing. For uni-
form cell printing and preventing solenoid valve and tip clogging, it is
extremely important to resuspend the cells immediately before printing using
a 100 μL pipette. There might be severe clogging observed if cells are clumpy
in alginate or the cell seeding density is too high (above 8 million cells/mL).
10. Print 60 nL of the cell suspension in 0.75 % alginate on the micropillar chips
placed on the chilling chip deck at 4–12 °C using a work file (Cell_Alginate_6
tip_6 block_60nL).
11. After 1–2 min gelation on the chilling chip deck, sandwich the micropillar chip
with encapsulated cells with the microwell chip containing 950 nL of growth
medium.
12. Place the sandwiched chips with the micropillar chip on the top in a humid
cell incubation chamber or a petri dish with 10 mL water (Fig. 4.8B) for
30 min at 37 °C to remove excess CaCl2. Note: In case of BaCl2 used for algi-
nate gelation, double rinsing with fresh growth medium might be necessary
every 30 min.
13. Discard the microwell chip and sandwich the micropillar chip with encapsulated
cells with the microwell chip containing 950 nL of fresh growth medium.

Fig. 4.8 Culture of cells on sandwiched chips in (a) a humid incubation chamber with 25 mL
water and (b) a petri dish with 10 mL water
90 4 Biological Sample Printing

14. Inspect the sandwiched chips under a bright field microscope with 4× or 10×
objective lens for spot attachment and uniform cell printing. Note: Make sure
to avoid water evaporation during chip inspection. Do not leave out the
sandwiched chips longer than 2 min for chip inspection. Water evaporation
will change the concentration of salts in growth medium, thus influencing
osmotic pressure inside cells and reducing cell viability and growth.
15. Incubate the sandwiched chips in the humidified petri dish placed in a 5 % CO2
incubator at 37 °C for 3D cell culture.
16. Change growth medium every 1–2 days by printing fresh growth medium in the
microwell chip, discarding the microwell chip containing old growth medium,
and sandwiching the freshly prepared microwell chip.

4.5.3.4 Cell Printing in Matrigel for 3D Cultures

Matrigel consists of laminin, collagen IV, entactin, heparin, and entactin which is pro-
duced by Engelbreth-Holm-Swarm (EHS) mouse sarcoma cells. It has been widely
used in cell growth, differentiation, angiogenesis, and tissue vascularization, since it
resembles cellular microenvironment in various tissues [9, 10]. Although Matrigel can
be used for cell printing and encapsulation on the chip platform, extra caution has to be
exercised as Matrigel is a temperature-sensitive hydrogel that forms a gel at a tempera-
ture ranging from 24 to 37 °C [11]. Therefore, it is necessary to maintain a low tem-
perature on the chip-loading deck and in the dispensing head while printing Matrigel to
avoid premature gelation and clogging in solenoid valves and ceramic tips. Typically,
cells are mixed with cold Matrigel and printed immediately onto the micropillar chip.
The printed cells in Matrigel can be gelled at 37 °C within 30 min [12]. Specific
example of primary hepatocyte printing in Matrigel is presented below.

Preparation of Growth Media

1. Freshly prepare a thawing/plating medium by mixing 100 mL of 1× DMEM,

5 mL of FBS, 10 μL of dexamethasone, and 3.6 mL of cocktail A.
2. Freshly prepare a maintenance medium by mixing 100 mL of 1× DMEM, 1 μL
of dexamethasone, and 4 mL of cocktail B.
3. Sterile filter the thawing/plating and maintenance media using the 0.2 μm filter
kit from VWR.

Preparation of Primary Hepatocyte Suspension

1. Warm up the thawing/plating and CHRM™ media for at least 30 min in a 37 °C

water bath before use.
2. Take out cryopreserved human hepatocytes in cryo-vial from a liquid nitrogen
storage tank, release high nitrogen pressure in the cryo-vial by slightly unscrew-
4.5 Sample Printing Protocols 91

ing the cap and then retightening it, and remove the paper label so that the ice
crystal melting can be easily observed during the thawing process.
3. Immerse the cryo-vial containing hepatocytes in a 37 °C water bath and gently
shake the cryo-vial in water until the last ice crystals floated to the top of the
cryo-vial. Note: Do not take the vial out of water during the thawing process
to check ice crystal melting. Always check it in water. The whole thawing
process typically takes about 3 min.
4. Immediately pour the melted hepatocytes from the cryo-vial into 50 mL of the
CHRM™ medium.
5. Rinse the cryo-vial three times with 1 mL of the CHRM™ medium to recover
hepatocytes attached on the wall.
6. Centrifuge the hepatocyte suspension in the CHRM™ medium at 100 g (RCF)
and 20 °C for 10 min with the slowest brake and acceleration settings available on
the centrifuge. Note: This is a critical step as primary hepatocytes are delicate.
7. Aspirate out the supernatant into a waste bottle, leaving the hepatocyte pellet at
the bottom.
8. Pour 3 mL of the thawing/plating medium onto the hepatocyte pellet and resus-
pend the pellet by gently swirling the tube or by tapping on the hood surface or
by dragging the bottom of the tube over corrugated surface. Note: Do not use a
vortex or a pipette to make well suspension of hepatocytes as they are very
fragile. Do not shake the hepatocyte suspension.
9. Split the thawing/plating medium containing suspended hepatocytes into two in
Eppendorf tubes using a 1 mL pipette.
10. Centrifuge the hepatocyte suspension in thawing/plating medium again under
the conditions stated above.
11. Aspirate out the supernatant into a waste bottle, leaving the hepatocyte pellet at
the bottom.
12. Add 1.5 mL of the thawing/plating medium onto the hepatocyte pellet and
resuspend the resulting hepatocyte pellet by gentle swirling. Note: Primary
hepatocytes are extremely fragile. Do not vortex, shake, or aspirate/dispense
hepatocytes to make the suspension.
13. Combine two hepatocyte suspensions into one and centrifuge the combined
hepatocyte suspension in thawing/plating medium under the conditions stated
above.
14. Aspirate out the supernatant into a waste bottle, leaving the hepatocyte pellet at
the bottom.
15. Add 300 μL of the thawing/plating medium onto the hepatocyte pellet, resus-
pend the resulting hepatocyte pellet by gentle swirling, and then store the
hepatocyte suspension (8 million cells/mL) on ice until use.
16. To count the total number of hepatocytes, remove 10 μL of the hepatocyte
suspension, add 490 μL of the thawing/plating medium, and measure the cell
number in suspension using an automated cell counter.
17. To measure the viability of hepatocytes, remove 10 μL of the hepatocyte sus-
pension, add 40 μL of the thawing/plating medium, mix the diluted hepatocyte
suspension with 50 μL of trypan blue, and place 10 μL of the hepatocyte-trypan
blue mixture on a hemocytometer.
92 4 Biological Sample Printing

Preparation of the Micropillar Chip with Primary Hepatocytes

1. Warm up the thawing/plating and maintenance media for at least 30 min in a

37 °C water bath before use.
2. Prior to use, thaw Matrigel overnight in a 4 °C refrigerator. Matrigel may gel at
slightly elevated temperatures in the refrigerator.
3. Keep DPBS, Matrigel, and hepatocytes suspension on ice before use, and use
pre-cooled pipettes, tips, and tubes when preparing a mixture of Matrigel and
DPBS/cell suspension for spotting.
4. Turn on both chillers connected to the chilling chip deck, the dispensing head,
the 96-well plate deck, and the water/alcohol containers at least 1 h before sam-
ple printing, set the temperature at 4 °C, and then circulate cold water to the
dispensing head. Note: It is particularly important to maintain low tempera-
ture when printing Matrigel. Wait until the temperatures reach 4 °C. Make
sure to turn off the chiller connected to the dispensing head after Matrigel
spotting due to severe water condensation.
5. Operate the daily washing program with ice-cold water and make sure that all
tubes, solenoid valves, and ceramic tips are cold prior to Matrigel spotting.
Note: At least running 2 cycles of the wash program with cold water are nec-
essary to cool it down.
6. Prepare a thawing/plating microwell chip by printing the thawing/plating
medium into the microwell chip (typically 532 spots/chip, 950 nL per spot into
microwell, 1.5 mm well-to-well distance).
7. Add 20 mL of sterile distilled water to the surrounding trench of the gas-­
permeable incubation chamber to retard water evaporation during incubation
(Fig. 4.8A).
8. Place the thawing/plating microwell chip with 950 nL of the medium in the
gas-permeable chamber and incubate it in a 5 % CO2 incubator at 37 °C prior
to stamping.
9. Prepare a Matrigel-DPBS mixture on ice by mixing 200 μL of cold Matrigel
with 200 μL of cold Dulbecco’s phosphate-buffered saline (DPBS).
10. Add approximately 500 μL of the cold Matrigel-DPBS mixture in a shallow-­
well plate (Fig. 4.9) and spread the solution uniformly on the surface of the
shallow well using a pipette tip.

Fig. 4.9 Micropillar chip

stamped on the shallow-­
well plate containing
500 μL of cold Matrigel-­
DPBS mixture
4.5 Sample Printing Protocols 93

11. Sandwich the micropillar chip (typically 532 spots/chip, 1.5 mm pillar-to-pillar
distance) with the shallow-well plate containing the cold Matrigel-DPBS
mixture and then dry Matrigel-DPBS spots on the micropillar chip for 1 h in a
sterile petri dish with a lid slightly open to yield flat Matrigel layers.
12. Prepare a suspension of hepatocytes in Matrigel on ice by mixing 300 μL of the
hepatocyte suspension with 300 μL of Matrigel so that the final concentration
of cells will be 4 × 106 cells/mL. Note: To prepare well-suspended hepatocytes
in Matrigel, gently mix them on ice by aspirating and dispensing for 10 times.
Don’t mix cell suspension with Matrigel by vortex as some cell membranes
are very fragile.
13. Print the cold Matrigel solution containing hepatocytes (532 spots/chip, 50 nL
per spot) atop each Matrigel-DPBS spot at 4 °C. Note: Put the 96-well plate
with the cold Matrigel-hepatocyte mixture back on ice immediately after aspi-
ration, or chill the 96-well plate deck by turning on the chiller and setting the
temperature at 4 °C.
14. Incubate the micropillar chip with hepatocytes in the black humid chamber
with 700 μL water in each well at 37 °C for 20 min, resulting in gelation of the
Matrigel matrix (Fig. 4.7). Note: Ensure the Matrigel gelling condition in
your lab by adjusting the incubation time at 37 °C.
15. If necessary, add freshly Matrigel solution containing suspended hepatocytes in
the round-bottom 96-well plate to make sure well suspension of hepatocytes
while spotting. After printing every chips (i.e., after every 5–10 min of spot-
ting), add freshly Matrigel solution containing suspended hepatocytes in the
well repeatedly. Note: Remind that hepatocytes are gradually settling down
while spotting. For uniform cell spotting, keep re-suspending hepatocytes.
After cell spotting, operate the daily washing program at least twice with ice-­
cold water to make sure that no Matrigel remains in tubes, solenoid valves,
and ceramic tips.
16. After incubation of the micropillar chip for Matrigel gelation, stamp the micropillar
chip with hepatocytes onto the microwell chip with 950 nL of the thawing/plating
medium, place the stamped chips in the gas-permeable chamber, and then incubate
it in a 5 % CO2 incubator at 37 °C for 2 h prior to compound toxicity tests.
17. Prepare compound solutions in the maintenance medium, print the medium in
the microwell chip, stamp the micropillar chip with hepatocytes onto the
microwell chip with compound solutions, and then incubate the stamped chips
in a 5 % CO2 incubator for 24 h for acute toxicity assays.
18. Stain hepatocytes on the micropillar chip with a Live⁄Dead® viability/cytotoxicity
kit from Life Technologies.

4.5.3.5 C
 ell Printing in a Mixture of Alginate and Matrigel for 3D
Cultures

Matrigel can be mixed with alginate to enhance cell viability and growth in alginate,
while avoiding temperature-induced gelation. Thus, the concentration of Matrigel
in alginate is substantially lower than its counterpart, Matrigel alone. The general
protocol of cell printing in a mixture of alginate and Matrigel is presented below.
94 4 Biological Sample Printing

1. Coat the micropillar chip with 2 mL of 0.01 % (w/v) PMA-OD in ethanol and
dry it at room temperature at least for 2–3 h.
2. Prepare a mixture of 0.0033 % PLL and 25 mM CaCl2 by mixing 0.01 % PLL
with 37.5 mM CaCl2 in distilled water at 1:2 ratio.
3. Print 60 nL of the PLL-CaCl2 mixture on the micropillar chip using a work file
(60 nL PLL_CaCl2) and dry it for at least 2–3 h.
4. Print 950 nL of growth medium in the microwell chip placed on the chilling
chip deck at 4–12 °C using a work file (medium_6 tip_6 block). Note: It is
particularly important to maintain low temperature when printing Matrigel.
Turn on both chillers connected to the chilling chip deck, the dispensing
head, the 96-well plate deck, and the water/alcohol containers at least 1 h
before sample printing. Wait until the temperatures reach 4–12 °C. Matrigel
is gelled at 37 °C.
5. Immediately after printing, incubate the microwell chips with growth medium
in a humid chamber with water to avoid water evaporation (Fig. 4.7).
6. Take out one aliquot of 500 μL Matrigel from a −20 °C freezer and thaw it
overnight in a 4 °C refrigerator.
7. Place Matrigel on ice to avoid premature gelation and dilute the Matrigel stock
solution (typically 9 mg/mL) with cold growth medium to obtain a Matrigel
working solution (2 mg/mL).
8. Prepare 1–4 million cells/mL in 0.75 % alginate containing 0.5 mg/mL Matrigel
by mixing 1 mL of 2–8 million cells/mL, 0.5 mL of 3 % alginate in distilled
water, and 0.5 mL of Matrigel working solution (2 mg/mL) using a 1 mL
pipette. Note: Do not vortex the mixture as mammalian cells are fragile. Mix
well by aspirating and dispensing the solution repeatedly (20 times) with a
1 mL pipette.
9. Place the micropillar chips with dried PLL-CaCl2 spots on the chilling chip
deck at 4–12 °C. Note: Do not leave the micropillar chips too long on the
chilling chip deck unattended as there might be severe water condensation
observed on the chip surface, which will affect spot attachment.
10. Add 100–200 μL/96-well of 1–4 million cells/mL in 0.75 % alginate containing
0.5 mg/mL Matrigel in a 96-well plate. Note: Always use the same kind of the
96-well plates to avoid tip crashing due to Z height difference in 96-wells.
11. Immediately before cell printing, resuspend the cells in 0.75 % alginate con-
taining 0.5 mg/mL Matrigel with a pipette. Note: For uniform cell printing
and preventing solenoid valve and tip clogging, it is extremely important to
resuspend the cells immediately before printing using a 100 μL pipette.
12. Print 60 nL of the cell suspension in 0.75 % alginate containing 0.5 mg/mL
Matrigel on the micropillar chips placed on the chilling chip deck at 4–12 °C
using a work file (Cell_Alginate_6 tip_6 block_60nL).
13. After 1–2 min gelation on the chilling chip deck, sandwich the micropillar chip
with encapsulated cells with the microwell chip containing 950 nL of growth
medium.
14. Rinse tubing, solenoid valves, and ceramic tips with cold distilled water by run-
ning the daily washing program twice. Note: Although Matrigel cannot be
4.5 Sample Printing Protocols 95

gelled in alginate at this low concentration, rinsing with cold water can
ensure to remove residual Matrigel.
15. Place the sandwiched chips with the micropillar chip on the top in a humid cell
incubation chamber or a petri dish with 10 mL water (Fig. 4.8B) and incubate
the chips in a 5 % CO2 incubator at 37 °C for 3D cell culture. Note: Unless
cytotoxicity noticed due to CaCl2, there will be no need to rinse the micropil-
lar chip with growth medium.
16. If necessary, inspect the sandwiched chips under a bright field microscope with
4× or 10× objective lens for spot attachment and uniform cell printing. Note:
Make sure to avoid water evaporation during chip inspection.
17. Change growth medium every 1–2 days by printing fresh growth medium in the
microwell chip, discarding the microwell chip containing old growth medium,
and sandwiching the freshly prepared microwell chip.

4.5.3.6  ell Printing in a Mixture of Alginate and Fibrinogen

C
for 3D Cultures

Fibrinogen is a glycoprotein, which is converted into fibrin gel via polymerization

in the presence of thrombin, and can be used for cell encapsulation. Fibrin gel is
widely used in gene delivery, cell growth and differentiation, and tissue engineering
applications [13–16]. Since it is a soft gel and susceptible to hydrolytic degradation
by proteases, fibrinogen alone may not be suitable for microarray 3D bioprinting
applications. However, fibrinogen can be mixed with alginate to improve physical
strength and used to facilitate 3D cell cultures. In this chapter, we provide an experi-
mental procedure for cell encapsulation and culture in a mixture of alginate, fibrino-
gen, and cells on the micropillar chip in the presence of thrombin [12].
1. Coat the micropillar chip with 2 mL of 0.01 % (w/v) PMA-OD in ethanol and
dry it at room temperature at least for 2–3 h.
2. Prepare a mixture of 0.0033 % PLL and 25 mM CaCl2 by mixing 0.01 % PLL
with 37.5 mM CaCl2 in distilled water at 1:2 ratio.
3. Print 60 nL of the PLL-CaCl2 mixture on the micropillar chip using a work file
(60 nL PLL_CaCl2) and dry it for at least 2–3 h.
4. Take out one aliquot of 100 U/mL thrombin from the −20 °C freezer and mix it
with growth medium at 1:10 ratio to achieve a final concentration of 10 U/mL
thrombin. Note: Do not vortex it. Just mix it with a 1 mL pipette.
5. Print 950 nL of growth medium containing 10 U/mL thrombin in the microwell
chip placed on the chilling chip deck at 4–12 °C using a work file (medium_6
tip_6 block).
6. Immediately after printing, incubate the microwell chips with growth medium
in a humid chamber with water to avoid water evaporation (Fig. 4.7).
7. Prepare 100 mg/mL fibrinogen freshly by adding 100 mg of fibrinogen in 1 mL
of 0.9 % NaCl and shaking the solution slightly to dissolve. Note: Fibrinogen
solution should not be vortexed. Fibrinogen solution cannot be stored, and
should be prepared fresh every time.
96 4 Biological Sample Printing

8. Prepare 1–4 million cells/mL in 0.75 % alginate containing 25 mg/mL fibrino-

gen by mixing 1 mL of 2–8 million cells/mL, 0.5 mL of 3 % alginate in distilled
water, and 0.5 mL of 100 mg/mL fibrinogen using a 1 mL pipette. Note: Do not
vortex the mixture as mammalian cells are fragile. Mix well by aspirating
and dispensing the solution repeatedly (20 times) with a 1 mL pipette.
9. Place the micropillar chips with dried PLL-CaCl2 spots on the chilling chip
deck at 4–12 °C. Note: Do not leave the micropillar chips too long on the
chilling chip deck unattended as there might be severe water condensation
observed on the chip surface, which will affect spot attachment.
10. Add 100–200 μL/96-well of 1–4 million cells/mL in 0.75 % alginate contain-
ing 25 mg/mL fibrinogen in a 96-well plate. Note: Always use the same kind
of the 96-well plates to avoid tip crashing due to Z height difference in
96-wells.
11. Immediately before cell printing, resuspend the cells in 0.75 % alginate con-
taining 25 mg/mL fibrinogen with a pipette. Note: For uniform cell printing
and preventing solenoid valve and tip clogging, it is extremely important to
resuspend the cells immediately before printing using a 100 μL pipette.
12. Print 60 nL of the cell suspension in 0.75 % alginate containing 25 mg/mL
fibrinogen on the micropillar chips placed on the chilling chip deck at 4–12 °C
using a work file (Cell_Alginate_6 tip_6 block_60nL).
13. After 1–2 min gelation on the chilling chip deck, sandwich the micropillar chip
with encapsulated cells with the microwell chip containing 950 nL of growth
medium supplemented with 10 U/mL thrombin.
14. Place the sandwiched chips with the micropillar chip on the top in a humid cell
incubation chamber or a petri dish with 10 mL water (Fig. 4.8B) and incubate
the chips overnight in a 5 % CO2 incubator at 37 °C for complete gelation of
fibrinogen. Note: Unless cytotoxicity noticed due to thrombin, there will be no
need to rinse the micropillar chip with growth medium.
15. If necessary, inspect the sandwiched chips under a bright field microscope with
4× or 10× objective lens for spot attachment and uniform cell printing. Note:
Make sure to avoid water evaporation during chip inspection.
16. Change growth medium every 1–2 days by printing fresh growth medium
(without thrombin) in the microwell chip, discarding the microwell chip con-
taining old growth medium, and sandwiching the freshly prepared microwell
chip.

4.5.3.7 Cell Printing in PuraMatrix for 3D Cultures

PuraMatrix from BD Biosciences is a synthetic peptide-based hydrogel commonly

used for 3D cell cultures. The peptide component of PuraMatrix forms 3D fibrous
structures with salts via self-assembly. The average pore size of the PuraMatrix
hydrogel is 50–200 nm. One of the biggest problems of using PuraMatrix for 3D
cell culture is its strong acidity, which can be toxic to many cells. Therefore, rinsing
cells with growth media immediately after cell printing is of utmost importance.
4.5 Sample Printing Protocols 97

1. Coat the micropillar chip with 2 mL of 0.01 % (w/v) PMA-OD in ethanol and
dry it at room temperature at least for 2–3 h.
2. Sonicate 500 μL of 1 % PuraMatrix (from BD Biosciences) in a 1.5 mL
Eppendorf tube for 30 min every time to decrease the viscosity. Note: If air
bubbles are present, centrifuge the PuraMatrix stock solution.
3. Prepare 0.25 % PuraMatrix by mixing 1 % PuraMatrix with sterile deionized
water in a 1:3 (v/v) ratio.
4. Print 60 nL of 0.25 % PuraMatrix on the micropillar chip using a work file
(60 nL PLL_CaCl2) and dry it for at least 4 h. Note: If the micropillar chips
with dried PuraMatrix are not being used within 24 h of the print, store over-
night at 4 °C.
5. For Hep3B cells, prepare RPMI medium containing 10 % FBS, 1 % Pen/Strep
(100 x dilution), and 50 μg/mL gentamycin (1000 × dilution).
6. Print 950 nL of growth medium onto the microwell chip using a work file
(medium_6 tip_6 block) while maintaining the chilling chip deck at 4–12 °C
and incubate it in a humid chamber at 37 °C until use to avoid water evaporation.
Note: Three microwell chips with growth medium are necessary for every
micropillar chip with cells encapsulated in PuraMatrix.
7. Mix PuraMatrix stock with 20 % sucrose in sterile deionized water in a 1:1
(v/v) ratio to generate a 0.5 % PuraMatrix solution in 10 % sucrose.
8. Trypsinize Hep3B cells, mix them with 7 mL growth media, and spin down cell
suspension at 1200 rpm (200 g) for 4 min.
9. Remove the supernatant and resuspend cell pellet with 3 mL of growth media.
10. Take a small aliquot and measure cell density using a cell counter. Note:
Sucrose might be toxic to cells, and interferes cell counting. Thus, do not use
solutions containing sucrose yet.
11. Seed cells in a T75 flask for near future experiments, if necessary.
12. Recentrifuge the cell suspension for the same speed and duration as described
in step 8, completely remove growth media from cell pellet to eliminate salts,
and resuspend the cells with 7 mL of 10 % sucrose in sterile deionized water to
maintain osmolarity. Note: Resuspension of the cells must be done very gently
with the 10 % sucrose solution as sucrose may lyse the cells.
13. Centrifuge the cells at 200 g for 7 min, remove the supernatant, and then resus-
pend the cells in 10 % sucrose at 2× the final desired cell concentration (12 mil-
lion cells/mL). Note: The gelation of PuraMatrix is initiated by salt
concentrations greater than 1 mM in buffers or growth media. Therefore, do
not combine PuraMatrix with salt-containing buffers or growth media until
gelation is desired.
14. Mix the 0.5 % PuraMatrix solution in 10 % sucrose with the 2× cell suspension
in 10 % sucrose in a 1:1 (v/v) ratio very gently for a final concentration of
6 million cells/mL in 0.25 % PuraMatrix and 10 % sucrose. Note: Vigorous
mixing can lyse the cells and release intracellular salts, resulting in premature
gelation of PuraMatrix.
15. Immediately after preparing the mixture freshly, print 60 nL of the cell suspension
in 0.25 % PuraMatrix and 10 % sucrose on the micropillar chip with dried
98 4 Biological Sample Printing

PuraMatrix using a work file (60 nL cell_alginate_printing) while maintaining

the chip deck at 4–12 °C. Note: This is one of the most important steps. Since
PuraMatrix exhibits a pH of 2–3, printing cells within 5–10 min of mixing
the cells and the PuraMatrix minimizes cell lysis and the chance for prema-
ture gelation. Freshly mix the 0.5 % PuraMatrix solution in 10 % sucrose
with the 2× cell suspension in 10 % sucrose right before each cell printing.
16. Wait two min for the gels to set.
17. Sandwich the micropillar chip with cells with the microwell chip containing
950 nL of pre-warmed growth medium for 30 min for gelation. Afterwards,
discard the microwell chip. Note: Fresh growth medium without sucrose is
used to remove 10 % sucrose and equilibrate the pH.
18. Repeat the previous step once. Note: Since PuraMatrix is very acidic and toxic
to cells, double rinsing with fresh growth medium is necessary every 30 min.
19. Sandwich the micropillar chip again with another microwell chip containing
pre-warmed media. Incubate the sandwiched chips in a humid incubation
chamber placed in a 5 % CO2 incubator at 37 °C for 3D cell cultures.

4.5.4 Virus Printing

Recombinant viruses have been used to transfer genes to human cells encapsulated
in hydrogels on the micropillar chip. For example, recombinant lentiviruses with
genes for neural stem cell (NSC)-specific biomarkers, including Synapsin1 for neu-
ron differentiation, glial fibrillary acidic protein (GFAP) for astrocyte differentia-
tion, myelin basic protein (MBP) for oligodendrocyte differentiation, and SOX2 for
self-renewal, have been used on the chip platform to monitor real-time NSC differ-
entiation. Recombinant adenoviruses carrying genes for cytochrome P450
(CYP450), including CYP1A2, CYP2C9, CYP2D6, CYP2E1, and CYP3A4, have
been constructed and used on the chip platform for transiently controlled expression
of CYP450 isoforms for metabolism-induced toxicity assays. Protocols for printing
viruses are simple and straightforward, compared to printing cells and enzymes.

4.5.4.1 Measurement of Viral Titer in a 96-Well Plate

Prior to virus printing, it is important to know the number of virus particles in the
solution. The multiplicity of infection (MOI) which is the number of virus particles
per cell has to be measured to precisely control the expression levels of recombinant
viruses used and minimize the basal toxicity due to excessive use of the viruses.
1. Seed HEK293 cells at a density of 10,000 cells in 100 μL Eagle’s minimum
essential medium (EMEM) per well in a 96-well plate.
2. Incubate the 96-well plate in a 5 % CO2 incubator at 37 °C for 24 h. Check to see
if cells are attached to the surface of 96 wells and spread out.
4.5 Sample Printing Protocols 99

Controls
1 2 3 4 5 6 7 8 9 10 11 12

A (10-12)

B (10-11)

C (10-10)

D (10-9)

E (10-8)

F (10-7)

G (10-6)

H (10-5)

Fig. 4.10 Layout of a 96-well plate with different virus dilution in growth medium. The dilution
factors used in the rows are recommended ranges. Sometimes there is need to increase/decrease the
dilution factors accordingly. Growth medium alone (without viruses) is used as a control

3. Dilute viruses with EMEM at 10−5–10−12 fold dilution (Fig. 4.10) and incubate
the 96-well plate in the 5 % CO2 incubator at 37 °C for 7–10 additional days.
4. After 7–10 days of incubation, check each well for cytopathic effect (CPE). This
assay is only valid if the following are true.
1) The control wells (columns 11 and 12) are completely CPE negative.
2) The highest viral concentration (row H1–10) is completely CPE positive.
3) The lowest viral concentration (row A1–10) is completely CPE negative.
5. Sum the fraction of wells in individual row that exhibit CPE.
6. Calculate viral titer by using the following equation:
H
0.8 + n + å fi .
æ pfu ö
Titer ç ÷ = 10 i= A

è mL ø

Where A, B, C, …, and H are rows inspected

f = fraction of wells in row that are CPE positive
n = −1−log (highest concentration virus dilution)

4.5.4.2 Adenoviral Transduction on the Micropillar/Microwell Chip

1. Print 950 nL of recombinant adenoviruses diluted with growth medium at different

MOIs into the microwell chip using a work file (medium_6tip_6block_950nL).
The layout of the microwell chip containing viruses is shown in Fig. 4.11.
2. Sandwich the micropillar chip with 3D-cultured cells with the microwell chip
containing adenoviruses and incubate the sandwiched chips in the humidified
100 4 Biological Sample Printing

Fig. 4.11 Layout of the microwell chip containing different MOI of adenoviruses for gene expres-
sion. For example, C1 is no virus condition, C2 is 5 MOI of adenovirus carrying gene for GFP
(Ad-GFP), C3 is 20 MOI of Ad-GFP, C4 is 10 MOI of Ad-GFP plus 10 MOI of Ad-RFP, C5 is 5
MOI of Ad-RFP, and C6 is 20 MOI of Ad-RFP

petri dish with 10 mL of distilled water placed in a 5 % CO2 incubator at 37 °C

for 24 h. Note: For 2D-grown cells, 4–6 h of virus exposure are commonly
used. For 3D-cultured cells, it requires longer incubation time with viruses
due to virus diffusion through the hydrogel matrix. Alginate is the worst
hydrogel for virus infection because of strong binding affinity of virus particles
to alginate. For higher efficiency of virus infection, 3D cells may be encapsu-
lated in Matrigel, PuraMatrix, or alginate mixture with Matrigel and
fibrinogen.
3. During cell exposure to viruses, print 950 nL of growth medium (with no virus)
onto the microwell chip using a work file (medium_6 tip_6 block) while main-
taining the chilling chip deck at 4–12 °C and incubate it in a humid chamber at
37 °C until use to avoid water evaporation. Note: Two microwell chips with
growth medium are necessary for every micropillar chip with infected cells
for removing excess viruses.
4. Discard the microwell chip containing viruses, and sandwich the micropillar
chip with infected cells with the microwell chip containing 950 nL of fresh
growth medium for 1 h for rinsing and removing excess viruses.
5. Discard the microwell chip again, and sandwich the micropillar chip with
infected cells with the microwell chip containing fresh growth medium for gene
expression.
4.6 Inspection of Cells Printed on the Micropillar Chips Using a Bright-Field Microscope 101

6. After 48 h incubation, inspect the sandwiched chips under a fluorescent micro-

scope with 4× or 10× objective lens for gene expression and spot attachment.
Note: Make sure to avoid water evaporation during chip inspection.
7. Dry the micropillar chip and observe fluorescence using the S+ scanner under
green, red, and multiband filter settings, or measure activity of various drug
metabolizing enzymes using appropriate kits.

4.6 I nspection of Cells Printed on the Micropillar Chips

Using a Bright-Field Microscope

1. To inspect cells printed on the micropillar chip or in the microwell chip under a
bright-field microscope (EVOS XL Core Imaging System, ThermoFisher
Scientific), turn on the microscope by pressing the on/off switch on the right side
of the microscope and wait until the starting messages disappear and the screen
is ready for inspection (Fig. 4.12).
2. Select an objective lens for chip inspection. There are three different magnifica-
tions of objective lenses (including 4×, 10×, and 20×) installed in the bright-
field microscope. Note: Typically, 4× objective lens is used for rapid inspection
of spot detachment, and 20× objective lens is used for inspecting cell
morphology.
3. Place the chip with cells on the chip deck and adjust focus positions by rotating
the coarse and fine focus adjustment knob for visualizing cells (Fig. 4.13a).
Note: Make sure to avoid water evaporation during chip inspection.
4. Move around the chip with cells to inspect spot detachment and uniform cells
printing in each block. Note: Make sure that the overall number of cells in

Fig. 4.12 Picture of the microscope with the on/off switch, objective lenses, and the focus adjust-
ment knob: (Left) front view and (Right) side view
102 4 Biological Sample Printing

Fig. 4.13 (a) Adjusting focus positions, (b) freezing the image, and (c) saving the cell image on
the chip.

spots printed by different solenoid valves are similar. The block-to-block varia-
tion (i.e., solenoid valve-to-solenoid valve variation) should be less than
20 %. In addition, spot detachment should be below 1 %.
5. Press the “Freeze” button to freeze the cell image before taking a picture
(Fig. 4.13b).
6. Press the “Save” button to store the picture in the USB connected to the micro-
scope (Fig. 4.13c). Note: Do not move the chip until the save message disap-
pears from the screen.

4.7 Coefficient of Variation (CV) and Z’ Factor for Assay

Validation

For robust assay development on the chip platform, it is important to measure the
range of errors and chip-to-chip and day-to-day reproducibility. The coefficient of
variation (CV) and the Z’ factor are commonly measured to evaluate error ranges
and robustness of an assay. The CV is defined as the ratio of the standard deviation
(σ) to the average (μ).
s
CV =
m

It is the extent of variability in relation to the average of the signal, thus the
inverse of the signal-to-noise ratio. The range of acceptable CV is typically
below 20 %.
References 103

Instead of using signal to noise (S/N) or signal to background (S/B), the Z’ factor
is the most widely used to measure the robustness of a new assay. The Z’ factor is
defined as the following equation where the averages of maximum/minimum sig-
nals from positive/negative controls are μ+/μ− and the standard deviations of posi-
tive/negative controls are σ+/σ−.

3 (s + + s - )
Z¢ = 1 -
( m+ - m- )
Assays with a Z’ factor between 0.5 and 1.0 are considered excellent. In addition,
assays with a Z’ factor between 0 and 0.5 are considered marginal. The Z’ factor is
used to identify the quality of a new assay prior to testing a large numbers of com-
pounds using high-­throughput screening (HTS).

4.8 Summary

In this chapter, we have reviewed the essential methods for printing biological sam-
ples. These methods were optimized for each individual sample, as a suspension of
cells in media will have a different viscosity than cells suspended in each of the vari-
ous hydrogels. Additionally, we have optimized preparation and printing protocols
for hydrogels with various gelation mechanisms, including ion-responsive, pH-­
responsive, enzymatically-activated, and thermos-responsive hydrogels. Optimizing
these protocols are essential for ensuring cell viability and reproducing mechanical
and physical interactions observed between cells and ECM in vivo. Additionally,
with successful enzyme, virus, and compound printing protocols established, it is
possible to transiently control gene expression, and predict drug efficacy and toxic-
ity for current and future therapeutics.

References

1. Lee, D. W., Lee, M. Y., Ku, B., Yi, S. H., Ryu, J. H., Jeon, R., et al. (2014). Application of the
DataChip/MetaChip technology for the evaluation of ajoene toxicity in vitro. Archives of
Toxicology, 88(2), 283–290. doi:10.1007/s00204-013-1102-9.
2. Lee, M.-Y., Kumar, R. A., Sukumaran, S. M., Hogg, M. G., Clark, D. S., & Dordick, J. S.
(2008). Three-dimensional cellular microarray for high-throughput toxicology assays.
Proceedings of the National Academy of Sciences, 105(1), 59–63. doi:10.1073/
pnas.0708756105.
3. Lee, M.-Y., Park, C. B., Dordick, J. S., & Clark, D. S. (2005). Metabolizing enzyme toxicology
assay chip (MetaChip) for high-throughput microscale toxicity analyses. Proceedings of the
National Academy of Sciences, 102(4), 983–987. doi:10.1073/pnas.0406755102.
104 4 Biological Sample Printing

4. Kadletz, L., Heiduschka, G., Domayer, J., Schmid, R., Enzenhofer, E., & Thurnher, D. (2015).
Evaluation of spheroid head and neck squamous cell carcinoma cell models in comparison to
monolayer cultures. Oncology Letters, 1281–1286. doi:10.3892/ol.2015.3487.
5. Pawar, S. N., & Edgar, K. J. (2012). Alginate derivatization: A review of chemistry, properties
and applications. Biomaterials, 33(11), 3279–3305. doi:10.1016/j.biomaterials.2012.01.007.
6. Jeon, O., Bouhadir, K. H., Mansour, J. M., & Alsberg, E. (2009). Photocrosslinked alginate
hydrogels with tunable biodegradation rates and mechanical properties. Biomaterials, 30(14),
2724–2734. doi:10.1016/j.biomaterials.2009.01.034.
7. Jeon, O., Powell, C., Ahmed, S. M., & Alsberg, E. (2010). Biodegradable, photocrosslinked
alginate hydrogels with independently tailorable physical properties and cell adhesivity. Tissue
Engineering. Part A, 16(9), 2915–2925. doi:10.1089/ten.tea.2010.0096.
8. Hsiong, S. X., Boontheekul, T., Huebsch, N., & Mooney, D. J. (2009). Cyclic arginine-­glycine-­
aspartate peptides enhance three-dimensional stem cell osteogenic differentiation. Tissue
Engineering. Part A, 15(2), 263–272. doi:10.1089/ten.tea.2007.0411.
9. Morritt, A. N., Bortolotto, S. K., Dilley, R. J., Han, X., Kompa, A. R., McCombe, D., et al.
(2007). Cardiac tissue engineering in an in vivo vascularized chamber. Circulation, 115(3),
353–360. doi:10.1161/CIRCULATIONAHA.106.657379.
10. Ponce, M. L. (2009). Tube formation: An in vitro matrigel angiogenesis assay. Methods in
Molecular Biology, 467, 183–188.
11. Thiele, J., Ma, Y., Bruekers, S. M. C., Ma, S., & Huck, W. T. S. (2014). 25th anniversary arti-
cle: Designer hydrogels for cell cultures: A materials selection guide. Advanced Materials,
26(1), 125–148. doi:10.1002/adma.201302958.
12. Datar, A., Joshi, P., & Lee, M. Y. (2015). Biocompatible hydrogels for microarray cell printing
and encapsulation. Biosensors, 5(4), 647–663. doi:10.3390/bios5040647.
13. Zhang, Z., He, Q., Deng, W., Chen, Q., Hu, X., Gong, A., et al. (2015). Nasal ectomesenchymal
stem cells: Multi-lineage differentiation and transformation effects on fibrin gels. Biomaterials,
49, 57–67. doi:10.1016/j.biomaterials.2015.01.057.
14. Eyrich, D., Brandl, F., Appel, B., Wiese, H., Maier, G., Wenzel, M., et al. (2007). Long-term
stable fibrin gels for cartilage engineering. Biomaterials, 28(1), 55–65. doi:10.1016/j.
biomaterials.2006.08.027.
15. Luyckx, V., Dolmans, M.-M., Vanacker, J., Scalercio, S. R., Donnez, J., & Amorim, C. A.
(2013). First step in developing a 3D biodegradable fibrin scaffold for an artificial ovary.
Journal of Ovarian Research., 6(1). doi:10.1186/1757-2215-6-83.
16. Huang, Y.-C. Y., Dennis, R. R. G. R. R. G., Larkin, L., & Baar, K. (2005). Rapid formation of
functional muscle in vitro using fibrin gels. Journal of Applied Physiology, 98(2), 706–713.
doi:10.1152/japplphysiol.00273.2004.
Chapter 5
High-Content Cell Staining

Kyeong-Nam Yu, Pranav Joshi, and Moo-Yeal Lee

Contents
5.1 Introduction...................................................................................................................... 105
5.1.1 Fluorescent Dyes.................................................................................................. 106
5.1.2 Immunofluorescence (IF) Assays with Antibodies.............................................. 107
5.1.3 Fluorescent Proteins............................................................................................. 108
5.2 Materials.......................................................................................................................... 109
5.2.1 Reagents for Fluorescence Staining..................................................................... 109
5.2.2 Reagents for Immunofluorescence Staining........................................................ 109
5.2.3 Devices for Cell Staining..................................................................................... 110
5.2.4 Preparation of Dye Stock Solutions in DMSO.................................................... 111
5.3 Protocols.......................................................................................................................... 112
5.3.1 Staining Cells with Fluorescent Dyes.................................................................. 112
5.3.1.1 Preparation of a Saline Solution.......................................................... 112
5.3.1.2 Staining Cells on the Micropillar/Microwell Chip Platform
with Fluorescent Dyes......................................................................... 112
5.3.2 Staining Cells on the Chip Platform with Fluorophore-­Labeled Antibodies....... 116
5.3.2.1 Cell Fixation........................................................................................ 117
5.3.2.2 Permeabilization of Cell Membranes.................................................. 117
5.3.2.3 Blocking of Nonspecific Binding and Incubation with Primary/
Secondary Antibodies for Fluorescence Labeling and Detection....... 118
5.3.3 Measuring the Expression Levels of Drug Metabolizing Enzymes on a Chip.... 119
5.3.3.1 Tyramide Signal Amplification Kit (Life Technologies).................... 121
5.4 Summary.......................................................................................................................... 122
References................................................................................................................................. 122

5.1 Introduction

Fluorescence-based cell imaging is an important technology for analyzing various

biological processes at cellular and molecular levels [1]. Morphological and func-
tional features in a cell can be labeled with multiple fluorescent probes/reagents,
imaged with automated fluorescence microscopes, and quantified with image

© The Author(s) 2016 105

M.-Y. Lee, Microarray Bioprinting Technology,
DOI 10.1007/978-3-319-46805-1_5
106 5 High-Content Cell Staining

analysis algorithms [2]. In particular, high-content imaging (HCI) has gained popu-
larity for systematic and accurate evaluation of drug candidates [3, 4] because of its
capability to assess specific signals, including changes in the nucleus, organelle
structure, protein translocation, oxidative stress, apoptosis/necrosis, mitochondrial
impairment, calcium homeostasis, morphology, and phenotype profiling as readouts
[2, 4]. Three different types of cell labeling are typically used for HCI assays, which
include fluorescent dyes for direct cell staining, antibodies for immunofluorescent
labeling, and genetically expressed fluorescent proteins such as green fluorescent
protein (GFP) [1, 5]. This chapter summarizes basic cell staining protocols with
various fluorescent dyes and antibodies for HCI assays on the micropillar/microwell
chip platform.

5.1.1 Fluorescent Dyes

Fluorescent dyes (also known as fluorophores) are generally polyaromatic hydro-

carbons or heterocycles which function at distinct excitation and emission wave-
lengths to generate fluorescence [6, 7]. Fluorescent dyes are taken up by cells and
concentrated in different organelles based on charge or molecular affinity. Many
fluorescent dyes are available commercially for various cell-based assays and
applications. For example, Hoechst 33342, 4′,6-diamidino-2-phenylindole (DAPI),
and Draq5 are commonly used for the assessment of nucleus morphology and cell
counting. Tetramethyl rhodamine methyl ester (TMRM) and MitoTracker are used
for measuring mitochondrial impairment. Propidium iodide, calcein AM, and
ethidium homodimer-1 are commonly used for measuring cell viability. TO-PRO-3
and BOBO-1 are used for assessing permeability of cell membrane, whereas
YO-PRO-1 has been used for apoptotic cell death. H2DCFDA and BODIPY
665/676 are commonly used for measuring the generation of reactive oxygen (O2)
species (ROS) [8–15].
Compared to immunofluorescence assays and fluorescent protein assays, fluo-
rescent dye-based cell staining offers several clear advantages, such as ease of
use, high sensitivity for fluorescence detection, and a wide range of selection for
specific cellular mechanisms [7]. Among a myriad of commercially available
fluorescent dyes, proper evaluation of several factors are required prior to cell
staining. For example, it is necessary to avoid excitation/emission spectrum over-
lapping among different fluorescent probes to obtain clear separation of fluores-
cence signals when multiple dyes are used simultaneously. Another problem is
that fluorescent dyes generally have limited photostability. Thus, stained cells
have to be always stored in the dark as most fluorescent molecules may be photo-
bleached, as repeated illumination on stained cells can cause continuous decrease
in fluorescent signals. Fluorescent dye solutions have to be prepared freshly
because exposing the solution to ambient air for a long time can reduce the fluo-
rescence intensity as dissolved oxygen (O2) may interact with fluorescent dyes.
5.1 Introduction 107

As some fluorescent dyes are pH sensitive, it is required to use a pH neutral buffer

solution when preparing the working solution. Precaution has to be taken for
background fluorescence from samples and sample holders when acquiring cell
images. Finally, fluorescence intensity is only linearly proportional to the concen-
tration of fluorophore in diluted solutions [1, 16].

5.1.2 Immunofluorescence (IF) Assays with Antibodies

As very few fluorescent dyes can stain proteins expressed within cells or secreted by
cells specifically, immunofluorescent labeling with antibodies has been developed
to resolve this issue and visualize the distribution of target proteins in biological
samples. Target proteins can interact with antibodies, which belong to a family of
globular proteins called immunoglobulins (Igs) (Fig. 5.1). Ig has four polypeptide
chains divided into two heavy (~50 kD, H) chains and two light (~23 kD, L) chains,
one of which has a specific binding site for the epitope of an antigen (i.e., target
protein). These specific binding of an antibody to an antigen is the major principle
of IF assays [17].
Two types of antibodies such as monoclonal (mAbs) and polyclonal antibodies
(pAbs) are commonly used for IF assays. The major difference is that a mAb
­recognizes one epitope on an antigen whereas a pAb recognizes multiple epitopes on
any one antigen. Thus, a mAb is used to detect a specific target protein which leads
to reproducible cell staining with significantly lower background fluorescence. On
the other hand, a pAb is used to detect a target protein with low expression levels and
amplify fluorescent signals due to multiple binding sites. The difference in their
properties are summarized in Table 5.1 [17].
The major advantages of IF staining are their wide applications with sectioned
tissues, primary cells, and cultured cell lines as well as their capability to detect
proteins, glycans, and small biological molecules. The IF-labeled samples can be

Fig. 5.1 Schematic diagram depicting the structure of an antibody

108 5 High-Content Cell Staining

Table 5.1 Properties of monoclonal antibody (mAb) and polyclonal antibody (pAb)
mAb pAb
• High specificity, detects only one • High affinity, binds multiple epitopes on any one
epitope on an antigen antigen and amplifying signals for robust detection
for immunoprecipitation (IP) and chromatin
immunoprecipitation (ChIP)
• Reduces background staining and • High background and potentially generates
highly reproducible between inaccurate results
experiments
• Used as primary antibody • Used as secondary antibody
• Expensive and takes a long time • Inexpensive and relatively easy to produce
for hybridoma production
• Too susceptible to the loss of • High batch-to-batch variability
epitope after chemical fixation

analyzed quantitatively and qualitatively by monitoring the fluorescent intensity.

Diverse primary or secondary antibodies with fluorescent tags or enzymes have
been developed and used for HCI assays in the fields of modern medicine, molecu-
lar biology, and toxicology. However, a variety of factors can interrupt the quality of
IF assays, including nonspecific antibody binding, excitation/emission spectrum
overlapping, autofluorescence, and photobleaching. For example, some mammalian
cells contain flavin coenzymes such as flavin adenine dinucleotide (FAD) and flavin
mononucleotide (FMN) (absorption: 450 nm and emission: 515 nm) and reduced
pyridine nucleotide such as nicotinamide adenine dinucleotide (NADH) (absorp-
tion: 340 nm and emission: 460 nm) that may have spectra overlapping with fluoro-
phores tagged on antibodies [18]. In addition, fluorophores bound to antibodies can
react with ROS generated by biological reactions, resulting in destruction of the
fluorophore [19]. Furthermore, aldehyde used for cell fixation can react with pro-
teins and amines, generating unwanted fluorescence [20].

5.1.3 Fluorescent Proteins

Cells expressing fluorescent proteins are used in HCI assays to investigate various
cellular processes, such as protein trafficking, gene activation, protein-protein
interactions, organelle condition, and cellular development/differentiation [5].
Fluorescent proteins in mammalian cells are expressed by transferring genes encod-
ing fluorescent proteins into the cells using plasmids, viral vectors, or other means.
The processes known as transfection or transduction can be either transient or
permanent. In transient transfection/transduction, the gene introduced into the cell
does not integrate into the chromosomes, therefore expressing the protein for a short
period of time. On the other hand, the gene can be permanently incorporated into
the genome to create transformed cell lines. Selection of fluorescent proteins for
HCI generally depends on several properties, including efficiency of protein expres-
sion, toxicity of the protein in the cells, brightness, photostability, and insensitivity
5.2 Materials 109

to environmental effects [21]. For example, enhanced green fluorescent protein

(EGFP) offers higher photostability than emerald in green spectrum ranges, and
mOrange offers high brightness but low photostability. In addition, mCherry and
mPlum offer high brightness and photostability in red and far-red spectrum ranges.
Moreover, CyPet offers better photostability than mCFP, but its expression effi-
ciency is relatively low at 37 °C [21]. The major advantage of fluorescent proteins
is the possibility of fluorescence labeling to almost any protein of interest [5].
However, fluorescent proteins are relatively large in size, which can affect the func-
tion and localization of endogenous proteins of interest [5].

5.2 Materials

5.2.1 Reagents for Fluorescence Staining

• Hoechst 33342 (Thermo Fisher Scientific)

• Tetramethyl rhodamine methyl ester (TMRM; Thermo Fisher Scientific)
• Fluo-4 acetoxy methyl ester (Fluo-4 AM; Thermo Fisher Scientific)
• Monochlorobimane (mBCl; Thermo Fisher Scientific)
• Live/dead viability/cytotoxicity kit for mammalian cells (Thermo Fisher
Scientific)
• YO-PRO-1 (Thermo Fisher Scientific)
• Propidium iodide (Thermo Fisher Scientific)

5.2.2 Reagents for Immunofluorescence Staining

• 1× Phosphate-buffered saline (PBS) containing 137 mM NaCl, 2.7 mM KCl,

10 mM Na2HPO4, and 2 mM KH2PO4. Adjust pH to 7.4 with 1 M HCl.
• 1× Tris-buffered saline (TBS) containing 50 mM Tris-HCl and 150 mM NaCl.
Adjust pH to 7.4 with 1 M HCl. Note: If cells to be stained are encapsulated in
alginate, TBS should only be used in the staining process, instead of PBS. PBS
is incompatible with alginate samples. TBS is used to prevent degradation of
alginate by chelating agents such as phosphate ions and EDTA.
• Paraformaldehyde (Sigma-Aldrich)
• Triton X-100 (Sigma-Aldrich)
• Bovine serum albumin (BSA) (Sigma-Aldrich)
• Antibodies
○○ Primary antibodies should be selected according to target proteins. Note: The
primary antibody should be considered based on the species of the target
protein stained, which avoids cross-reactivity of the secondary antibody
with endogenous antibodies in the sample. For example, if a mouse protein
is studied, select a primary antibody obtained from different species (e.g.,
human, rabbit, or goat).
110 5 High-Content Cell Staining

Table 5.2 Considerations to select secondary antibodies [22]

Considerations Description
Reactivity Anti-mouse, anti-rat, anti-chicken, anti-rabbit, anti-goat, anti-human, etc.
Non fluorescent Horseradish peroxidase (HRP)-labeled, alkaline phosphatase (AP)-labeled,
conjugates biotin, unconjugated, etc.
Fluorescent Alexa fluor™ dyes, fluorescein isothiocyanate (FITC), tetramethylrhodamine
conjugates (TRITC), DyLight™ dyes, rhodamine, Texas Red™ & Texas Red™-X,
R-phycoerythrin (R-PE), allophycocyanin (APC), Qdot™ probes, Pacific dyes,
etc.
Target Ig class IgG, IgM, IgA, IgG2a, IgG1, IgD, Kappa or lambda type light chain, etc.

○○ Secondary antibodies should be selected based on reactivity against primary

antibodies, fluorescent conjugates, and target immunoglobulin (Ig) class (rec-
ognition of whole IgG or any fragments) (Table. 5.2).
–– Host species: If the primary antibody is from a mouse, the secondary anti-
body used should be anti-mouse.
–– Experimental methods: For immuno-staining assays (e.g., IF, IHC, etc.), it
requires to select a secondary antibody conjugated with fluorescent dyes or
enzymes.
–– Class/subclass of antibody: In general, polyclonal primary antibodies are
produced from rabbit, goat, or sheep, and the majority of them has diverse
Ig isotypes. Therefore, secondary antibodies used should be anti-Ig
antibodies.
–– Ig class/subclass and type/subtype: Ig has diverse isoforms, and each Ig
isoform has a unique sequence for antigen binding. Before selecting a sec-
ondary antibody, the Ig class has to be checked.

Anti-mouse IgG (H+L), highly cross-adsorbed, Alexa Fluore® 488 antibody produced in donkey

Name of Description of antibody Antibody can be Description of the Description of

species specificity (Isotype of adsorbed to the fluorescent dye host species
antibody antibody, Heavy chain serum of diverse labeled for
with light chain) species detection

• 4′,6-Diamidino-2-phenylindole (DAPI; Sigma-Aldrich)

• Mounting solution with DAPI staining (Sigma-Aldrich)
• Mounting solution without DAPI staining (Fisher Scientific)

5.2.3 Devices for Cell Staining

• Deep-well staining plate (Samsung Electro-Mechanics, Co. or SEMCO, Suwon,

South Korea)
• Shallow-well staining plate (SEMCO, Suwon, South Korea)
5.2 Materials 111

5.2.4 Preparation of Dye Stock Solutions in DMSO

• Hoechst 33342
–– Dissolve 20 mg of Hoechst 33342 in 3.25 mL DMSO to prepare a stock solution
of 10 mM Hoechst 33342.
–– Store aliquots of 10 mM Hoechst 33342 in a −20 °C freezer protected from
light until use.
–– To prepare a working solution of 25 μM Hoechst 33342, add 20 μL of the stock
solution in 8 mL of Dulbecco’s phosphate buffered saline (DPBS). Note: If
possible, do not reuse the stock solution of the dye for reproducible results.
• Tetramethyl rhodamine methyl ester (TMRM)
–– Dissolve 25 mg TMRM in 1 mL DMSO and vortex it for 1 min to prepare a
stock solution of 50 mM TMRM.
–– Store aliquots of 50 mM TMRM in a −20 °C freezer protected from light
until use.
–– Dilute the stock solution 100-fold to get a working stock concentration of
0.5 mM. Store aliquots of 0.5 mM TMRM in a −20 °C freezer, and do not
reuse this stock solution for reproducible results.
–– To prepare a working solution of 0.5 μM TMRM, add 8 μL stock in 8 mL
DPBS.
• Fluo-4 acetoxy methyl ester (Fluo-4 AM)
–– Dissolve 50 μg of Fluo-4 AM in 23 μL DMSO to prepare a stock solution of
2 mM Fluo-4 AM. Note: Fluo-4 AM may require addition of 20 % (w/v)
Pluronic F-127 to enhance solubility.
–– Store aliquots of 2 mM Fluo-4 AM in a −20 °C freezer protected from light
until use.
–– To prepare a working solution of 5 μM Fluo-4 AM, add 10 μL stock in 2 mL
DPBS.
• Monochlorobimane (mBCl)
–– Dissolve 25 mg of mBCl in 550 μL DMSO to prepare a stock solution of
200 mM mBCl.
–– Store aliquots of 200 mM mBCl in a −20 °C freezer protected from light until use.
–– To prepare a working solution of 100 μM mBCl, add 4 μL stock in 8 mL
DPBS.
• Calcein AM
–– Calcein AM comes in a liquid form in the live/dead viability/cytotoxicity kit
at a concentration of 4 mM without the need for adding solvent.
–– Store aliquots of calcein AM in a −20 °C freezer protected from light until use.
–– To prepare a working solution of 1 μM calcein AM, add 2 μL stock in 8 mL
DPBS.
112 5 High-Content Cell Staining

• Ethidium homodimer-1
–– Ethidium homodimer-1 comes in a liquid form in the live/dead viability/cyto-
toxicity kit at a concentration of 2 mM without the need for adding solvent.
–– Store aliquots of ethidium homodimer-1 in a −20 °C freezer protected from
light until use.
–– To prepare a working solution of 1 μM ethidium homodimer-1, add 4 μL
stock in 8 mL DPBS.
• Propidium iodide
–– Propidium iodide comes in a liquid form at a concentration of 1.5 mM with-
out the need for adding solvent.
–– Store aliquots of 1.5 mM propidium iodide in a 4 °C refrigerator protected
from light until use. Note: Propidium iodide is stable for at least 6 months
when stored at 4 °C.
–– To prepare a working solution of 5 μM propidium iodide, add 27 μL stock in
8 mL DPBS.
• YO-PRO-1
–– YO-PRO-1 comes in liquid form at a concentration of 1 mM without the need
for adding solvent.
–– Store aliquots of 1 mM YO-PRO-1 in a −20 °C freezer protected from light
until use.
–– To prepare a working solution of 10 μM YO-PRO-1, add 80 μL stock in 8 mL
DPBS.

5.3 Protocols

5.3.1 Staining Cells with Fluorescent Dyes

5.3.1.1 Preparation of a Saline Solution

1. Dissolve 8.1 g of NaCl and 2.9 g of CaCl2⋅H2O in 1 L of sterile deionized water

to prepare a saline solution containing 140 mM NaCl and 20 mM CaCl2. Note:
The saline solution with 20 mM CaCl2 is used only if the viability of the cells
encapsulated in alginate on the chip is not affected by the salts. Alginate is
sensitive to chelating agents such as phosphate ions in some buffer solutions.
DPBS is recommended when working with hydrogels other than alginate.

5.3.1.2 S
 taining Cells on the Micropillar/Microwell Chip Platform
with Fluorescent Dyes

1. Prepare working solutions of desired fluorescent dyes in the saline solution prior
to initial washing of the chip, and keep it protected from light (Table 5.3).
 5.3 Protocols

Table 5.3 The working conditions of commonly used fluorescent dyes

Solvent Working Working
Quantity Stock conc. added Solvents used to stock conc. Working Working Dye volume volume
Fluorescent dye Mol. wt. (mg) (mM) (μL) dissolve dyes (mM) conc. (μM) solvent (μL) (mL)
Hoechst 33342 615.99 20 10 3250 DMSO 10 25 DPBS 20 8
Tetramethyl rhodamine 500.93 25 50 998 DMSO 0.5 0.5 DPBS 8 8
methyl ester (TMRM)
Fluo-4 acetoxy methyl 1096.95 0.05 2 23 DMSO 1 5 DPBS 10 2
ester (Fluo-4 AM)
YO-PRO-1 629.32 1 mL 1 NA DMSO 1 10 DPBS 80 8
Propidium iodide (PI) 668.4 10 mL 1.5 NA Water 1.5 5 DPBS 27 8
Monochlorobimane 226.66 25 200 551 DMSO 200 100 DPBS 4 8
(mBCl)
BODIPY 665/676 448.32 5 10 1115 DMSO 10 2.5 DPBS 2 8
Calcein AM 994.87 40 μL 4 NA DMSO 4 1 DPBS 2 8
Ethidium homodimer-1 856.77 200 μL 2 NA DMSO 2 1 DPBS 4 8
113
114 5 High-Content Cell Staining

Fig. 5.2 Pictures of (a) an empty deep-well plate, (b) a deep-well plate with a saline solution and
(c) a deep-well plate with the micropillar chips placed on the saline solution for cell rinsing.

2. Using 10 mL serological pipette, aspirate the saline solution and dispense 5.5 mL
in each well of the deep-well staining plate (Fig. 5.2B). Note: Make sure to avoid
dust particles while washing and staining. To remove dust and tiny particles
from deep-well and shallow-well plates, blow air on the wells of both deep-well
and shallow-well staining plates before starting the cell staining process.
3. Place the micropillar chip facing down on top of the deep well with 5.5 mL of
the saline solution so that cells on the micropillars can be immersed in the solu-
tion, and rinse the chip twice for 5 min each (Fig. 5.2C). Note: The cells on the
chip are immersed in the saline solution to remove growth media and fetal
bovine serum (FBS), which may interfere cell staining. Make sure that there
are no bubbles trapped underneath the chip.
4. Add 2 mL of a desired fluorescent dye solution in each well of a shallow-well
plate (Fig. 5.4B). Note: Spread the dye solution with a pipette tip evenly on the
surface of the shallow well. As the surface of the shallow well is relatively
hydrophobic, it may not be spread well enough for uniform cell staining.
5. Once the chip is rinsed on the deep-well plate, drain an excess saline solution
from the chip by tilting the chip at an angle of 45° and remove the remaining
saline solution from the side of the chip with a paper towel (Fig. 5.3). Note:
Do not dry the cell spots on the micropillar chip as spot drying will cause
cell death.
6. Immediately after draining the excess saline solution, place the micropillar chip
facing down on top of the shallow well with 2 mL of the desired fluorescent dye
solution and incubate for 60 min for cell staining (Fig. 5.4C). Note: The staining
time can be varied depending on the fluorescent dye used and the morphology
of cells stained.
5.3 Protocols 115

Fig. 5.3 Removing the excess saline solution from the side of the micropillar chip with a paper
towel by tilting the chip at a 45° angle

Fig. 5.4 Pictures of (a) an empty shallow-well plate, (b) a shallow-well plate with a fluorescent
dye solution and (c) a shallow-well plate with the micropillar chips placed on the fluorescent dye
solution for cell staining

7. During the period of cell staining, cover the entire shallow-well plate with the
chips with aluminum foil. Note: The majority of fluorescent dyes are very sensi-
tive to light illumination. To avoid photobleaching of fluorescent dyes during
staining, make sure to prevent the chips from light exposure.
8. After cell staining, remove an excess dye solution by rinsing the micropillar
chip with stained cells twice for 10 min each in the deep-well plate with 5.5 mL
of the saline solution. The micropillar chips should be protected from light dur-
ing the rinsing steps too.
9. After rinsing, drain and remove the excess saline solution from the chip as
described in Step 5.
116 5 High-Content Cell Staining

Fig. 5.5 Representative images of Hep3B cells cultured in 3D (60 nL) on the micropillar chip and
stained with (b) calcein AM for cell viability, (b) TMRM for mitochondrial membrane potential,
(c) Hoechst 33342 for nucleus morphology and cell count, (d) mBCl for glutathione level, (e)
propidium iodide for cell viability and necrosis, and (f) Fluo-4 AM for intracellular calcium level

10. Dry the stained chips completely in the dark for at least 2 h. Once the micropillar
chips are completely dried, they are ready for scanning. As an example, images
of 3D-cultured cells stained with various fluorescent dyes are presented in Fig.
5.5. Note: Do not scan a wet micropillar chip as dripping water with salts can
damage objective lenses and filter sets in the S+ Scanner. Always completely
dry the micropillar chip or seal the microwell chip with wet samples with a gas-
permeable sealing membrane (Sigma-Aldrich) before scanning.

5.3.2  taining Cells on the Chip Platform with Fluorophore-­

S
Labeled Antibodies

In general, immunofluorescence (IF) assays on the chip consist of cell fixation,

permeabilization of cell membranes, blocking of nonspecific binding, and incuba-
tion with primary/secondary antibodies for fluorescence labeling and detection.
Carrying out IF assays on the micropillar/microwell chip require careful optimiza-
tion of experimental conditions due to several reagents that can facilitate spot
detachment. For example, some detergents used can detach cell spots from the
micropillar chip and phosphate ions in PBS buffers can degrade alginate spots.
5.3 Protocols 117

5.3.2.1 Cell Fixation

1. Rinse the cells by immersing the micropillar chip containing cells twice in
5.5 mL of 1× sterilized PBS (or 1× TBS) in the deep-well plate for 5 min each.
2. Incubate the chip in the shallow-well plate with 2 mL of a fixation reagent for
10–15 min. Note: Refer Table 5.4 for various fixation reagents and their
methods.
3. Rinse the cells three times in 5.5 mL of 1× sterilized PBS (or 1× TBS) in the
deep-well plate for 3 min each.

5.3.2.2 Permeabilization of Cell Membranes

4. Incubate cells with permeabilization reagents for 10 min. Note: Refer Table 5.5
for various permeabilization reagents and their methods.
5. After permeabilization, rinse cells with 1× PBS or 1× TBS.

Table 5.4 Fixation reagents commonly used for target antigens

Fixation reagents Methods of use Target antigens
Ice-cold acetone (100 %) • Fix cells in −20 °C acetone for 5–10 min Large proteins
• The chip made of polystyrene may not be (e.g., immunoglobulin)
compatible with [23, 24]
• 100 % acetone
• No permeabilization step needed after
acetone fixation
Ice-cold methanol • Fix cells in −20 °C methanol for
(100 %) 5–10 min
• No permeabilization step needed after
methanol fixation
4 % Paraformaldehyde • Fix cells in 4 % paraformaldehyde with Most proteins, peptides,
with 1 % glutaraldehyde 1 % glutaraldehyde for 10–20 min and enzymes with low
• Rinse cells briefly with 1× PBS (or 1× molecular weight [25]
TBS)
4 % Paraformaldehyde • Fix cells in 4 % paraformaldehyde for Most proteins, peptides,
10–20 min and enzymes with low
• Rinse cells briefly with 1× PBS (or 1× molecular weight [26]
TBS)
10 % Neutral-buffered • Fix cell in 10 % neutral buffered formalin Most proteins, peptides,
formalin (NBF) for 10–20 min and enzymes with low
• Rinse cells briefly with 1× PBS (or 1× molecular weight [27]
TBS)
Carnoy’s solution • Carnoy’s solution contains 60 % ethanol, Nucleic acids [28]
30 % chloroform, and 10 % acetic acid
• Fix cells in Carnoy’s solution for
10–20 min and rinse the cells with 1×
PBS (or 1× TBS)
118 5 High-Content Cell Staining

Table 5.5 Various permeabilization reagents and their methods of use

Permeabilization reagents Method of use References
Solvents Acetone and methanol • Acetone and methanol can be used to fix [29]
and permeabilize cells simultaneously
• It can be used with crosslinking agents
such as formaldehyde
Detergents Triton X-100 and • Treat cells with 0.1–0.2 % of the [30]
NP-40 detergents in PBS for 10 min
• It can partially dissolve the nuclear
membrane, thus suitable for nuclear
antigen staining
Tween 20, saponin, • Treat cells with detergents for 10–30 min [30–32]
digitonon (0.01 %), • These detergents are mild membrane
and leucoperm solubilizers and make large pores on
cellular membrane without dissolving
membrane
• It is suitable for cytoplasma membrane
or soluble nuclear antigen staining

Table 5.6 Blocking reagents, their methods of use, and limitations

Blocking reagents Methods of use Characteristics References
Normal serum • Use it before • Common blocking reagent [33, 34]
introduction of
primary antibody
• This can be used • Make sure to use proper serum
for the dilution of of species. It may interact with
antibodies detection antibody
Protein solution • Use it before • Can be made easily in labs [35, 36]
(0.1–5 % BSA, introduction of
gelatin, skim milk) primary antibody
• This can be used • Must be made freshly before use
for the dilution of • Competitively binding to
antibodies analogous ligands
Pre-formulated • Use it before • Contain a refined concentration [37]
commercial buffers introduction of of individual proteins
primary antibody
• This can be used • Using to diverse application
for the dilution of and improved shelf life
antibodies

5.3.2.3  locking of Nonspecific Binding and Incubation with Primary/

B
Secondary Antibodies for Fluorescence Labeling and Detection

6. After cell fixation and permeabilization, incubate the cells in 1× PBS (or 1×
TBS) containing 3 % BSA (or other blocking reagents) for 1 h to prevent non-
specific binding of antibodies. Note: Blocking reagents may vary depending on
the application. Refer to Table 5.6 for various blocking reagents and their
methods of use.
5.3 Protocols 119

Fig. 5.6 Picture of a

humidified chamber for
cell staining with primary/
secondary antibodies

7. Dilute a primary antibody in 1× PBS (1× TBS) containing a blocking reagent.

The dilution factor of the primary antibody can be varied and determined
experimentally. Check the manufacturer protocol or references. Note: A typical
dilution factor ranges from 1:50 to 1:500.
8. Incubate the cells overnight with the primary antibody at 4 °C in a humidified
chamber (Fig. 5.6).
9. Decant the primary antibody and rinse the cells three times in 1× PBS containing
Triton X-100 (1× PBS-T) or 1× TBS containing Triton X-100 (1× TBS-T) for
10 min each on a plate shaker.
10. Incubate the cells with a secondary antibody in 1× PBS (or 1× TBS) containing
3 % BSA for 1–3 h at room temperature in the humidified chamber. Note:
For proper multicolor staining, ensure that primary antibodies and their
corresponding secondary antibodies are from different species.
11. Decant the secondary antibody and rinse three times in 1× PBS-T (or 1× TBS-­
T) for 10 min each in the dark on the plate shaker.
12. Stain the cells with DAPI at a final concentration of 500 nM for 10 min.
13. Rinse the cells twice with 1× sterile PBS-T (or 1× TBS-T) for 3 min each.
14. Apply a mounting solution for image acquisition.

5.3.3  easuring the Expression Levels of Drug Metabolizing

M
Enzymes on a Chip

1. Prepare 1× Tris-buffered saline (TBS) containing 0.1 % Tween 20 (1× TBS-T)

2. Rinse the cells by immersing the micropillar chip containing cells three times in
25 mL of 1× TBS in a petri dish (10 cm in diameter) for 5 min each. Note: Place
the micropillar chip face down in the petri dish containing 1× TBS to allow the
chip floating on the surface of the solution. Make sure not to damage cell spots
120 5 High-Content Cell Staining

Table 5.7 Examples of primary antibodies for detecting cytochrome P450 isoforms
Name Target protein Origin Dilution factor Company
Rabbit polyclonal anti-CYP1A1 CYP1A1 Rabbit 1:1000 Abcam
Rabbit polyclonal anti-CYP3A4 CYP3A4 Rabbit 1:100 Thermo Fisher
Scientific
Rabbit polyclonal anti-CYP2C9 CYP2C9 Rabbit 1:100 Thermo Fisher
Scientific

by touching the bottom of the petri dish when lifting the floating chip up with
a pair of tweezers. Do not shake in order to avoid spot detachment.
3. Fix and permeate the cells by immersing the chip in 25 mL of 1× TBS containing
3.7 % formaldehyde for 20 min at room temperature and then transferring the
chip in 25 mL of 1× TBS containing 0.15 % Triton X-100 for 10 min. In case the
chip made of glass is used, fix and permeate the cells by immersing the glass chip
in 25 mL of methanol and acetone (1:1, v/v) for 20 min at −20 °C.
4. Rinse the cells by immersing the chip twice in 25 mL of 1× TBS containing
0.15 TBS for 5 min each.
5. Dry the chip under weak nitrogen gas (or air) stream.
6. Incubate the chip overnight in 25 mL of a blocking buffer in TBS (SuperBlock
from Fisher Scientific) at 4 °C.
7. Wash the chip three times in 25 mL of TBS-T for 5 min each. Note: Tween 20 in
TBS-T is necessary to reduce hydrophobic interactions between protein and
surface.
8. Label the cells with a primary antibody (e.g., mouse anti-human IgG from Life
Technologies) by diluting the primary antibody (typically 1:100–1:500, v/v, see
primary antibody dilutions used in Table 5.7 as a reference) in TBS-T containing
1 % (w/v) bovine serum albumin (BSA), applying 2 mL of the primary antibody
solution to the shallow-well plate, and then incubating the chip stamped onto the
staining plate overnight at 4 °C.

9. Rinse the cells by immersing the chip three times in 25 mL of TBS-T for 15 min
each with gentle shaking.
10. Label the cells with a secondary antibody (e.g., HRP-conjugated goat anti-­
mouse IgG from Life Technologies T20912) by diluting the secondary anti-
body (typically 1:500–1:1000, v/v) in TBS-T containing 1 % BSA, applying
2 mL of the secondary antibody solution to the shallow-well plate, and then
incubating the chip stamped onto the staining plate for 3 h at room
temperature.
11. Rinse the cells by immersing the chip three times in 25 mL of TBS-T for 15 min
each with gentle shaking.
12. Label the cells with Alexa Fluor 488 tyramide (Tyramide signal amplification
kit, Life Technologies T20912) by diluting the tyramide stock (1:200, v/v) in
amplification buffer containing 0.0015 % H2O2 just prior to labeling, applying
5.3 Protocols 121

2 mL of the tyramide working solution to the shallow-well plate, and incubating

the chip stamped onto the staining plate for 30 min at room temperature.
To freshly prepare the tyramide working solution, add 5 μL of 30 % H2O2 to
995 μL of amplification buffer and then 20 μL of this intermediate solution
(0.15 % H2O2) to further a 1980 μL of amplification buffer.
13. Rinse the cells by immersing the chip three times in 25 mL of 1× TBS for
15 min each. Note: Tween 20 is unnecessary for the final rinsing.
14. Dry the chip under weak air stream.
15. Scan the chip for green fluorescence using S+ Scanner. Note: The fluorescence
signal of β-actin (determined by the same protocol using triplicate chips for
each condition) can be used as an internal control.

5.3.3.1 Tyramide Signal Amplification Kit (Life Technologies)

Contents
• Labeled tyramide (Component A), one vial
• Dimethylsulfoxide (DMSO; Component B), 200 μL
• HRP-conjugated secondary antibody (Component C), 100 μg
• Blocking reagent (Component D), 3 g
• Amplification buffer (Component E), 25 mL (containing 0.02 % thimerosal)
• Hydrogen peroxide (H2O2; Component F), 200 μL of a 30 % stabilized solution
Upon receipt and prior to use, the kit should be stored at −20 °C, desiccated and
protected from light.
Preparation of Solutions
1. Prepare tyramide stock solution by dissolving the solid material provided
(Component A) in 150 μL of DMSO (Component B). Invert the vial several
times to dissolve any tyramide coating the sides of the vial. Store unused por-
tions of this stock solution in small aliquots at −20 °C, desiccated and protected
from light.
2. Prepare the HRP-conjugated antibody stock solution by reconstituting the mate-
rial provided (Component C) in 200 μL of PBS. This solution may be stored at
2–6 °C for up to 3 months if required. Optionally, add 0.02 % thimerosal as a
preservative. Note: Sodium azide must not be used for this purpose.
3. Prepare a 1 % (10 mg/mL) solution of blocking reagent in PBS. Note: We recom-
mend preparing only as much as is needed for immediate use. However, unused
solution can be stored frozen at −20 °C for 1 month if necessary.
4. Prepare amplification buffer/0.0015 % H2O2 by adding 30 % hydrogen peroxide
(Component F) to amplification buffer (Component E) to obtain a final concen-
tration of 0.0015 % H2O2. For example, add 5 μL of 30 % H2O2 to 995 μL of
amplification buffer and then add 20 μL of this intermediate dilution (0.15 %
H2O2) to a further 1980 μL of amplification buffer.
122 5 High-Content Cell Staining

5.4 Summary

In this chapter, we introduced principal methods of cell staining on the micropillar/

microwell chip platform, which include 3D cell-based HCI assays with fluorescent
dyes as well as IF assays with primary/secondary antibodies. To demonstrate and
retrofit traditional HCI assays and IF assays on the micropillar/microwell chip, the
capability to optimize experimental parameters is of importance. These protocols
can be easily modified to other cell staining and extended to high-throughput screen-
ing of compounds for predictive toxicology.

References

1. Charvin, G., Oikonomou, C., & Cross, F. R. (2010). Labels and probes for live cell imaging:
Overview and selection guide. Methods in Molecular Biology, 591(2), 229–242.
doi:10.1007/978-1-60761-404-3_14.
2. van Vliet, E., Danesian, M., Beilmann, M., Davies, A., Fava, E., Fleck, R., et al. (2014).
Current approaches and future role of high content imaging in safety sciences and drug discovery.
ALTEX, 31(4), 479–493. doi:10.14573/altex.1405271.
3. Buchser, W., Collins, M., Garyantes, T., Guha, R., Haney, S., Lemmon, V., et al. (2012). Assay
development guidelines for image-based high content screening, high content analysis and
high content imaging. In G. S. Sittampalam, N. P. Coussens, H. Nelson, M. Arkin, D. Auld,
C. Austin, et al. (Eds.), Assay guidance manual (pp. 1–69). Bethesda, MD: Eli Lilly &
Company and the National Center for Advancing Translational Sciences.
4. Zanella, F., Lorens, J. B., & Link, W. (2010). High content screening: Seeing is believing.
Trends in Biotechnology, 28(5), 237–245. doi:10.1016/j.tibtech.2010.02.005.
5. Valeur, B., & Berberan-Santos, M. N. (2012). Autofluorescence and fluorescence labeling in
biology and medicine. Molecular Fluorescence: Principles and Applications, 479–505.
6. Demchenko, A. P. (2008). Introduction to fluorescence sensing. Berlin: Springer.
doi:10.1007/978-1-4020-9003-5.
7. Virus, E. D., Sobolevsky, T. G., & Rodchenkov, G. M. (2008). Introduction to fluorescent
techniques. Journal of Mass Spectrometry, 43(7), 949–957. doi:10.1002/jms.1447.
8. Towne, D. L., Nicholl, E. E., Comess, K. M., Galasinski, S. C., Hajduk, P. J., & Abraham, V. C.
(2012). Development of a high-content screening assay panel to accelerate mechanism of
action studies for oncology research. Journal of Biomolecular Screening, 17(8), 1005–1017.
doi:10.1177/1087057112450050.
9. Martin, H. L., Adams, M., Higgins, J., Bond, J., Morrison, E. E., Bell, S. M., et al. (2014).
High-content, high-throughput screening for the identification of cytotoxic compounds based
on cell morphology and cell proliferation markers. PLoS ONE, 9(2), 1–8. doi:10.1371/journal.
pone.0088338.
10. Alonso-Padilla, J., Cotillo, I., Presa, J. L., Cantizani, J., Peña, I., Bardera, A. I., et al. (2015).
Automated high-content assay for compounds selectively toxic to trypanosoma cruzi in a myo-
blastic cell line. PLoS Neglected Tropical Diseases, 9, 1–17. doi:10.1371/journal.pntd.0003493.
11. Sirenko, O., Hesley, J., Rusyn, I., & Cromwell, E. F. (2014). High-content assays for hepato-
toxicity using induced pluripotent stem cell-derived cells. Assay and Drug Development
Technologies, 12(1), 43–54. doi:10.1089/adt.2013.520.
12. Håkanson, M., Cukierman, E., & Charnley, M. (2014). Miniaturized pre-clinical cancer mod-
els as research and diagnostic tools. Advanced Drug Delivery Reviews, 69–70, 52–66.
doi:10.1016/j.addr.2013.11.010.
References 123

13. Mioulane, M., Foldes, G., Ali, N. N., Schneider, M. D., & Harding, S. E. (2012). Development
of high content imaging methods for cell death detection in human pluripotent stem cell-­
derived cardiomyocytes. Journal of Cardiovascular Translational Research, 5, 593–604.
doi:10.1007/s12265-012-9396-1.
14. Fujisawa, S., Romin, Y., Barlas, A., Petrovic, L. M., Turkekul, M., Fan, N., et al. (2014).
Evaluation of YO-PRO-1 as an early marker of apoptosis following radiofrequency ablation of
colon cancer liver metastases. Cytotechnology, 66(2), 259–273. doi:10.1007/
s10616-013-9565-3.
15. Donato, M. T., Tolosa, L., Jiménez, N., Castell, J. V., & Gómez-Lechón, M. J. (2012). High-­
content imaging technology for the evaluation of drug-induced steatosis using a multiparamet-
ric cell-based assay. Journal of Biomolecular Screening, 17(3), 394–400.
doi:10.1177/1087057111427586.
16. Lakowicz, J. R. (2009). Fluorophores. Princ Fluoresc Spectrosc, 954. doi:10.1002/
smll.201090041.
17. Lipman, N. S., Jackson, L. R., Weis-Garcia, F., & Trudel, L. J. (2005). Monoclonal versus
polyclonal antibodies: Distinguishing characteristics, applications, and information resources.
ILAR Journal, 46(3), 258–268. doi:10.1093/ilar.46.3.258.
18. Aubin, J. E. (1979). Autofluorescence of viable cultured mammalian cells. The Journal of
Histochemistry and Cytochemistry, 27(1), 36–43.
19. Ntziachristos, V. (2010). Going deeper than microscopy: the optical imaging frontier in biol-
ogy. Nature Methods, 7(8), 603–614. doi:10.1038/nmeth.1483.
20. Lee, K., Choi, S., Yang, C., Wu, H.-C., & Yu, J. 2013. Autofluorescence generation and elimi-
nation: A lesson from glutaraldehyde. Chemical Communications (Cambridge, England),
49(29):3028–3030. doi:10.1039/c3cc40799c.
21. Shaner, N. C., Steinbach, P. A., & Tsien, R. Y. (2005). Aguide to choosing fluorescent proteins.
Nature Methods, 2(12), 905–909. doi:10.1038/nmeth819.
22. Manning, C. F., Bundros, A. M., & Trimmer, J. S. (2012). Benefits and pitfalls of secondary
antibodies: Why choosing the right secondary is of primary importance. PLoS ONE, 7(6).
doi:10.1371/journal.pone.0038313.
23. Kaku, T., Ekem, J. K., Lindayen, C., Bailey, D. J., Van Nostrand, A. W., & Farber, E. (1983).
Comparison of formalin- and acetone-fixation for immunohistochemical detection of
­carcinoembryonic antigen (CEA) and keratin. American Journal of Clinical Pathology, 80(6),
806–815.
24. Levitt, D., & King, M. (1987). Methanol fixation permits flow cytometric analysis of immuno-
fluorescent stained intracellular antigens. Journal of Immunological Methods, 96(2), 233–237.
doi:10.1016/0022-1759(87)90319-X.
25. Kosaka, T., Nagatsu, I., JY, W., & Hama, K. (1986). Use of high concentrations of glutaralde-
hyde for immunocytochemistry of transmitter-synthesizing enzymes in the central nervous
system. Neuroscience, 18(4), 975–990. doi:10.1016/0306-4522(86)90112-0.
26. Pollice, A. A., McCoy, J. P., Shackney, S. E., Smith, C. A., Agarwal, J., Burholt, D. R., et al.
(1992). Sequential paraformaldehyde and methanol fixation for simultaneous flow cytometric
analysis of DNA, cell surface proteins, and intracellular proteins. Cytometry, 13(4), 432–444.
doi:10.1002/cyto.990130414.
27. Grizzle, W. E. (2009). Models of fixation and tissue processing. Biotechnic & Histochemistry,
84(5), 185–193. doi:10.3109/10520290903039052.Models.
28. Miething, F., Hering, S., Hanschke, B., & Dressler, J. (2006). Effect of fixation to the degrada-
tion of nuclear and mitochondrial DNA in different tissues. The Journal of Histochemistry and
Cytochemistry, 54(3), 371–374. doi:10.1369/jhc.5B6726.2005.
29. Jamur MC, Oliver C. (2010). Permeabilization of cell membranes. In Oliver C, & Jamur MC
(Eds.), Methods and Protocols (Vol. 588, pp. 63–66). New York: Humana Press.
doi:10.1007/978-1-59745-324-0.
30. Amidzadeh, Z., Behzad Behbahani, A., Erfani, N., Sharifzadeh, S., Ranjbaran, R., Moezi, L.,
et al. (2014). Assessment of different permeabilization methods of minimizing damage to the
124 5 High-Content Cell Staining

adherent cells for detection of intracellular RNA by flow cytometry. Avicenna Journal of
Medical Biotechnology, 6(1), 38–46.
31. Ohsaki, Y., Maeda, T., & Fujimoto, T. (2005). Fixation and permeabilization protocol is criti-
cal for the immunolabeling of lipid droplet proteins. Histochemistry and Cell Biology, 124(5),
445–452. doi:10.1007/s00418-005-0061-5.
32. Misra, D. P., Chaurasia, S., & Misra, R. (2016). Increased circulating Th17 Cells, Serum
IL-17A, and IL-23 in Takayasu Arteritis. Autoimmune Diseases, 2016, 7841718.
doi:10.1155/2016/7841718.
33. Kopen, G. C., Prockop, D. J., & Phinney, D. G. (1999). Marrow stromal cells migrate through-
out forebrain and cerebellum, and they differentiate into astrocytes after injection into neonatal
mouse brains. Proceedings of the National Academy of Sciences of the United States of
America, 96(19), 10711–10716. doi:10.1073/pnas.96.19.10711.
34. Italiano, J. E., Richardson, J. L., Patel-Hett, S., Battinelli, E., Zaslavsky, A., Short, S., et al.
(2008). Angiogenesis is regulated by a novel mechanism: Pro- and antiangiogenic proteins are
organized into separate platelet α granules and differentially released. Blood, 111(3), 1227–
1233. doi:10.1182/blood-2007-09-113837.
35. Kwong, K. F., Schuessler, R. B., Green, K. G., Laing, J. G., Beyer, E. C., Bioneau, J. P., et al.
(1998). Differential expression of gap junction proteins in the canine sinus node. Circulation
Research, 82, 604–612.
36. Bianchi, L., Shen, Z., Dennis, A. T., Priori, S. G., Napolitano, C., Ronchetti, E., et al. (1999).
Cellular dysfunction of LQT5-minK mutants: Abnormalities of I(Ks), I(Kr) and trafficking in
long QT syndrome. Human Molecular Genetics, 8(8), 1499–1507. doi:10.1093/hmg/8.8.1499.
37. Thompson, K., Trowern, A., & Fowell, A. (1998). Primary rat and mouse hepatic stellate cells
express the macrophage inhibitor cytokine interleukin-10 during the course of activation
in vitro. Hepatology, 28(6), 1518–1524. doi:10.1002/hep.510280611.
Chapter 6
3D-Cultured Cell Image Acquisition

Pranav Joshi, Kyeong-Nam Yu, Emily Serbinowski, and Moo-Yeal Lee

Contents
6.1 Introduction...................................................................................................................... 125
6.2 Materials.......................................................................................................................... 127
6.3 Protocols.......................................................................................................................... 127
6.3.1 Daily Operational Procedures.............................................................................. 128
6.3.2 Parameter Setting Procedures.............................................................................. 135
6.3.2.1 Setting the Position of Filters and the Distance of Each Step
for Autofocus...................................................................................... 135
6.3.2.2 Setting XYZ Coordinates for Different Chips
and Objective Lenses.......................................................................... 137
6.4 Summary.......................................................................................................................... 139
6.5 Appendix.......................................................................................................................... 139
References................................................................................................................................. 141

6.1 Introduction

Acquiring high-content images of 3D-cultured cells for analyzing multiple cellular

events is a daunting task, requiring an automated fluorescent microscope and high-­
throughput image analysis software. High throughput is an important feature for
high-content imaging (HCI) devices to enable rapid image acquisition. Various fac-
tors come into play when dealing with the speed of image acquisition. For example,
capturing large number of cells with lower magnification or reducing sample vol-
ume/size can increase data acquisition speed. In addition, reducing exposure time
with the use of high intensity light sources, optimizing fluorescence staining proto-
cols for brighter colors, and using an objective lens with relatively high numerical
aperture also significantly increases the image acquisition speed [1].

© The Author(s) 2016 125

M.-Y. Lee, Microarray Bioprinting Technology,
DOI 10.1007/978-3-319-46805-1_6
126 6 3D-Cultured Cell Image Acquisition

The basic components of any imaging device consist of light sources, detectors,
objective lenses, and optical filters among which light sources and detectors play
critical role in determining fate of any HCI system. Light sources can comprise of
lamps, lasers, and light-emitting diodes (LEDs) with each of them having their own
benefits and limitations. For example, lamps provide a broad excitation source from
UV to IR, but should be replaced frequently due to shorter lifetime and may need
realignment for optimal excitation. Lasers, on the other hand, have longer lifetime,
but may not offer optimal excitation spectra for certain target fluorophores such as
blue fluorescent dyes. LEDs offer both long lifetime and optimal excitation, but are
expensive. Most of currently available image acquisition systems use xenon or mer-
cury lamps as a light source due to economic feasibility and a broad range of spec-
tra. Detectors are another important component of HCI devices with two types of
detectors commonly used. These detectors are charge-coupled devices (CCDs) and
photomultiplier tubes (PMTs). CCD-based detectors are widely used in HCI devices
as it offers excellent acquisition speed (100 frames per second), large dynamic
ranges (>20,000:1), broad spectral sensitivity (400–900 nm and higher), and high
resolution (>2000 × 2000 pixels), but with reduced sensitivity to low intensity light.
On the other hand, PMT-based detectors offer high sensitivity towards low intensity
light, but with compromise in throughput. Multiple PMTs are used in parallel to
acquire images through various fluorescent channels [1–3].
Conventional HCI devices can be divided broadly into two categories; wide-field
imaging devices and confocal imaging devices. Wide-field imaging devices are
basically inverted microscopes which offer an excellent image acquisition speed.
Some examples of commercially available CCD-based wide-field imaging devices
are ArrayScan XTI HCS Reader manufactured by Thermo Fisher, IN Cell Analyzer
2000 by GE Healthcare, ImageXpress Micro HCS system by Molecular Devices,
Operetta by Perkin Elmer, WiSCAN by IDEA Bio-Medical, and S+ Scanner by
SEMCO. These commercially available wide-field imaging devices commonly use
lamps and LEDs as their light source. For example, ArrayScan XTI HCS Reader
uses a metal halide lamp and LED, and IN Cell Analyzer 2000 uses a metal halide
lamp. Similarly, ImageXpress Micro HCS system uses a xenon lamp, and Operetta
uses a xenon lamp and LED. In addition, S+ scanner uses a mercury lamp, and
WiSCAN uses a mercury lamp and LED. Although different types of light sources
are used by various wide-field imaging devices, the difference is very little in terms
of optical performance. The difference between the various wide-field imagers
available arise generally from the numerical aperture of the objectives and the qual-
ity of light path [1, 4, 5].
Confocal imaging devices use light barrier with a fixed or adjustable pinhole to
eliminate light in front or behind the focus plane of an objective. Confocal devices
offer high resolution and improved contrast compared to wide-field devices, but it
also causes reduction in light signal and only scans single points of a specimen at
any given moment. Although these issues have been overcome by using technolo-
gies like laser scanning confocal microscopy (LSCM) and Nipkow spinning disk,
the image acquisition speed is compromised, and there is always a chance of photo-
6.3 Protocols 127

bleaching [1, 4, 6, 7]. Confocal imaging devices are expensive compared to wide-­
field imaging devices and most of its applications are towards imaging small
intracellular structures, small cells, complex 3D structures, etc. IN Cell Analyzer
6000 by GE Healthcare is a commonly used confocal microscope, which uses
LSCM with varying apertures. It consists of four laser lines and a LED for transmit-
ted light and a scientific complementary metal oxide semiconductor (sCMOS)-
based detector. ImageXpress ULTRA developed by Molecular devices is another
confocal imager consisting of four lasers and four PMTs that can be operated seri-
ally or in parallel [1, 8, 9]. Extensive information on various HCI devices and their
components can be found in Assay Guidance Manual published by Eli Lilly &
Company and the National Center for Advancing Translational Sciences (http://
www.ncbi.nlm.nih.gov/books/NBK53196/) [1].
While other HCI devices are highly sophisticated and require trained personnel
to operate the machine, the S+ Scanner offers researchers/users the ability to acquire
high quality fluorescent images in a high-throughput manner without the need of
complex training. This chapter aims to provide simple procedures for operating S+
Scanner to rapidly acquire 3D-cultured cell images from the micropillar/microwell
chip platform.

6.2 Materials

• Micropillar/microwell chips (25 × 75 × 2 mm from Samsung Electro-Mechanics,

Co. or SEMCO, Suwon, South Korea)
• S+ MicroArrayer with six solenoid valves and ceramic tips (SEMCO, Suwon,
South Korea)
• S+ Scanner with four filter sets (SEMCO, Suwon, South Korea)
• Light source (Olympus, Model: U-HGLGPS)
• Micropillar chip zig (SEMCO, Suwon, South Korea)
• Microwell chip zig part I and II (SEMCO, Suwon, South Korea)
• Metal frame for the micropillar chip (SEMCO, Suwon, South Korea)

6.3 Protocols

S+ Scanner, is a wide-field imager consisting of a mercury lamp as a light source

and a CCD-based detector. It is an automated fluorescence microscope with four
filter channels for detecting multicolor, blue, green, and red fluorescent dyes, indi-
vidually or simultaneously. The basic components of S+ Scanner are provided in
Fig. 6.1. Certain precautions are required before starting the image acquisition from
the micropillar/microwell chips. For example, micropillar chips stained with fluo-
rescent dyes should be dried properly before scanning to avoid damage to the
128 6 3D-Cultured Cell Image Acquisition

Chip-loading deck

Light source

Detector
Objective lens
Filter set

Fig. 6.1 Mechanical components of S+ Scanner

objective lens and stored in the dark to prevent photobleaching. In addition, when
acquiring the image from microwell chips containing liquid samples it should be
covered with a gas-permeable sealing membrane to prevent spilling of the samples
and to minimize immediate drying of the spots.

6.3.1 Daily Operational Procedures

1. Turn on the power source (by using the blue switch on the step-up transformer).
2. Turn on the S+ Scanner using the green power switch and then press the ‘Reset’
button (Fig. 6.2). Note: Pushing the ‘Reset’ button resets XYZ coordinates of
the chip-loading deck. This step is essential to avoid malfunctioning of the S+
Scanner. Do not skip this step!
3. Turn on the light source and the computer. Use the black switch on the back to
turn on the light source and the ‘ON/OFF’ button in the front to turn on the lamp
(Fig. 6.3). Note: After the lamp is turned ON, do not turn OFF for 2 min to
avoid damage to the lamp. In addition, do not turn ON the lamp within 10 min
of turning OFF. The operation of switches is disabled for 5 min after the lamp
is turned OFF. The lamp is connected to the light source in S+ Scanner via a
fiber optic cable, which needs to be placed properly to avoid damage to the fiber
optic cable.
4. Open the scanner program by double clicking the shortcut of S+ Scanner located
on the desktop (Fig. 6.4).
5. The user interface of the scanner program will be opened (Fig. 6.5).
6. Make sure that the sliding door on top of the scanner is closed (Fig. 6.2).
6.3 Protocols 129

Sliding door/
Chip-loadingdeck
Reset button (Blue)
Emergency button (Red)

Power switch
Fig. 6.2 The picture of S+ Scanner showing the power switch on the back, reset and emergency
buttons on the side, and sliding door/chip-loading deck on the top

Fig. 6.3 Pictures of the light source: (a) The front side with ON/OFF and Reset buttons and a LED
indicator displaying the burner status and (b) The back side with the main switch and the fiber
optic cable

7. In the ‘Home’ window, click the ‘Home’ icon in the ‘XYZ Move’ tab to activate
the ‘Chip loading’ and ‘Live’ icon (Fig. 6.6).
8. Click ‘Chip loading’ icon in the ‘XYZ Move’ tab to bring the chip-loading deck
in position for loading the micropillar chip zig (Fig. 6.6).
Fig. 6.4 The shortcut of S+ Scanner program

Fig. 6.5 The user interface of S+ Scanner program showing ‘Home’ window highlighted in a red
box

Fig. 6.6 ‘Home’ window in the user interface of the scanner program

Fig. 6.7 Preparation of micropillar chips before loading on the chip-loading deck for image acqui-
sition. (a) Placing a metal frame on the micropillar chip at an angle of 45°, (b) Inserting all the
micropillars into the holes on the metal frame, (c) Placing the micropillar chip with the metal frame
on the magnetic micropillar chip zig. The circle highlighted in yellow indicates the bottom of the
zig aligned with the bottom of the micropillar chip.
6.3 Protocols 131

9. In order to place micropillar chips on the micropillar chip zig and load it into the
chip-loading deck, follow the steps below (Fig. 6.7).
(a) Align a metal frame on a micropillar chip by holding the metal frame at an
angle of 45° against the edge of the micropillar chip (Fig. 6.7A). The dents
at the bottom of the metal frame should be perfectly aligned with the circles
at the bottom of the micropillar chip.
(b) Slowly drop the metal frame so that all the micropillars go into the holes on
the metal frame (Fig. 6.7B).
(c) Gently insert the micropillar chip with the metal frame into the mag-
netic micropillar chip zig by holding the side of the chip with two fin-
gers (Fig. 6.7B).
(d) The micropillar chip will be laid flat and held firmly while scanning due to
strong magnets in the micropillar chip zig (Fig. 6.7C). Note: The strong
magnetic attraction between the zig and the metal frame may impose
severe force on the micropillar chip. Therefore, make sure that all the
micropillars go into the holes on the metal frame properly before placing
it on the zig.
10. Open the sliding door and load the micropillar zig with the chips into the chip-
loading deck by inverting the zig (Fig. 6.8). Note: The symbol (4) on the zig
faces the same direction as the symbol on the chip-loading deck.
11. In case of scanning microwell chips, the chip should be sealed with a gas-­
permeable membrane to prevent water evaporation in the microwells and spill-
ing of the sample over the objective lens.
12. A microwell chip zig consists of two parts—part I is an open metal frame for
placing microwell chips and part II is a magnetic frame that can hold the
microwell chips firmly while scanning due to strong magnets (Fig. 6.9).

Fig. 6.8 The micropillar

chip zig with micropillar
chips loaded into the
chip-loading deck. The
symbol (4) on the
micropillar chip zig should
be aligned in the same
direction with the symbol
(4) on the chip-loading
deck.
132 6 3D-Cultured Cell Image Acquisition

(A) (B)

Symbol (4) on the microwell chip zig

Bottom of the microwell chip

(C) (D)

Fig. 6.9 (a) A microwell chip inserted in the part I microwell chip zig, (b) The front side of the
part II microwell chip zig, (c) The part I zig combined with the part II zig, and (d) The back side
of the part II zig combined with the part I zig. The symbol (4) on the back of the part II zig high-
lighted in red should be aligned with the bottom of the microwell chips inserted.

13. In order to place microwell chips on the microwell chip zig and load it into the
chip-loading deck, follow the steps below (Fig. 6.9).
(a) Place and insert microwell chips in the part I microwell chip zig (Fig. 6.9A)
so that the back side of the microwell chip is exposed for scanning while
the front side of the chip is sealed with the gas-permeable membrane, fac-
ing the part II microwell chip zig (Fig. 6.9B).
(b) Take the part II zig and place it on the part I zig so that the symbol [4] on
the part II zig is aligned with the bottom of the microwell chips (Fig. 6.9C, D).
Note: The microwell chip will be laid flat and held firmly while scanning
due to strong magnets in the microwell chip zig.
14. Click ‘Open’ in the ‘Chip layout’ tool bar (Fig. 6.10). A window displaying a
list of chip files will appear (Fig. 6.11). Note: These chip files contain XYZ
coordinates for the micropillar chip and the microwell chip at different objec-
tive lenses.
6.3 Protocols 133

Fig. 6.10 The display of the ‘Chip layout’ tool bar

Fig. 6.11 Window displaying a list of chip files

15. Select a desired chip file from the list by clicking ‘Open’ in the ‘Chip layout’
window (Fig. 6.11). Note: Refer to parameter setting procedures to set and
change XYZ coordinates in a given chip file.
16. In the ‘Home’ window, click the ‘Spot position’ icon in the ‘XYZ Move’ tab to
enable accurate selection of individual spots (i.e., micropillars or microwells)
for live view (Fig. 6.12).
17. Select the number of chips being scanned from the ‘Chip selection’ tab in the
‘Home’ window (Fig. 6.12).
18. In the ‘Live view’ tab, click the ‘Live’ icon to observe a real-time cell image and
check an optimum exposure time by dragging the cursor left to right in the
‘Exposure time’ section or manually entering the value in the ‘Exposure time’
section (Fig. 6.12).
134 6 3D-Cultured Cell Image Acquisition

Fig. 6.12 ‘Home’ window of S+ Scanner program

Fig. 6.13 ‘Histogram’ window displaying the histogram of the red-fluorescent image from a red
filter. The red intensity is higher than 255, meaning that the exposure time is a bit too high

19. Click the ‘Histogram’ icon in the ‘Home’ window to enable the display of fluo-
rescence histogram (Fig. 6.12).
20. To find a proper exposure time, use the ‘Histogram’ tool bar to analyze the
wavelength and also visually inspect the real-time cell image. An exposure time
which gives a histogram between 80 and 90 % of maximum pixel range
(i.e., 255) for the given channel is considered as an optimum exposure
time (Fig. 6.13). Note: Do not select too high exposure time as photobleach-
ing of fluorescence might occur at higher exposure time.
21. In the ‘Region of interest’ tab, click ‘ROI’ icon to display the region of interest
(ROI) in the image window and use the ‘Radius’ and ‘Margin’ cursor or enter
the value to increase and decrease the size of the ROI (Fig. 6.12).
22. In the ‘Chip scan’ tool bar, click ‘Set’ to activate the setting window for each
chip. Assign a label for each chip in the ‘Folder’ section, set the exposure time
in ‘Exposure’ section, and select the desired filter from the drop-down list in the
‘Filter’ section. (Fig. 6.14). Note: A single chip can be scanned using four dif-
ferent filters and exposure setting depending on the user requirement. Make
sure that the rest of the drop-down list in ‘Filter’ section is set to ‘UNUSED’
when scanning the chip with less than four filters setting. Select ‘Multiband’
filter for acquiring images from cells stained with multiple fluorescent dyes.
The drop-down list in ‘Auto-Exp’ field is usually set to ‘UNUSED’ when
scanning the chip with a user-defined exposure setting.
6.3 Protocols 135

Fig. 6.14 The display of the ‘Chip scan’ tool bar

23. Select the magnification of an objective lens used for image acquisition from the
drop down list in the ‘Lens’ field and select the number of steps for autofocus
from the drop down list in the ‘Steps’ field in the ‘Chip scan’ tab (Fig. 6.12). Note:
Select a larger number of steps when the chip is bent or not placed flat. It will
cover a larger distance in the Z direction to obtain an optimum image in focus.
24. Click the ‘Scan’ icon to begin chip scanning.
25. The images obtained from individual filters are saved in multiple folders with
the name of the specific filter and the exposure time used for that filter in the
format of ‘filter_exposure time’ (e.g., Multiband_150). These folders are stored
inside another folder with the name of the chip provided by users (Refer to Step
22). The user-named folders for multiple chips are stored in a big folder which
is labeled in the format of ‘year-month-day’ (e.g., 2016-06-01).
26. After the scanning is done, remove the chips from the scanner and store the
scanned chips in the dark until disposed.
27. Close the scanner software and turn off the computer, the S+ Scanner, and the
light source (Fig. 6.3). Note: To switch off the light source, first press the ‘ON/
OFF’ button in the front until the blue LED turns off and the countdown
begins from 300 on the digital display. Wait until the countdown ends and
then turn off the main switch of the light source located on the back.

6.3.2 Parameter Setting Procedures

6.3.2.1  etting the Position of Filters and the Distance of Each Step
S
for Autofocus

This tool bar displays all the list of filters available for ‘Live’ view of cells and chip
scanning, allows to set and change the position of the filters, and also allows to
modify the distance of each step for autofocus (Refer to Step 22). Note: The num-
ber of steps and the distance of each step are important to obtain an optimum
image in focus using the autofocus function. The S+ Scanner takes multiple
136 6 3D-Cultured Cell Image Acquisition

Fig. 6.15 The display of the ‘Filter’ tool bar

pictures in the Z direction but save only one best-looking image. Ideally, the dis-
tance of each step should be equal to the height of cell spheroids.
1. Click ‘Move’ button to select the desired filter, which has predetermined excita-
tion and emission spectra for image acquisition (Fig. 6.15 and Appendix). For
example, clicking the ‘Move’ button for the ‘Orange’ filter enables ‘Live’ view
of cells/spheroids stained with orange/red fluorescent dyes in the excitation/
emission range of the ‘Orange’ filter.
2. Filters are set to the position displayed on the ‘Position’ column by default.
Change or reset the position of individual filters only when the filter position is
misplaced due to machine malfunctioning or user errors.
3. To change or reset the filter position, click the ‘Set’ button to activate the ‘Filter
setting’ window and enter the position of a respective filter so that ‘Live’ cell
image is visible uniformly without distortion (Fig. 6.15). Note: The filter posi-
tion has to be increased or decreased by few units only if needed. For example,
the ‘Multiband’ filter position has to be increased or decreased to 23 or 21,
respectively.
4. The ‘Excitation’ and ‘Emission’ columns are used to calculate and set the dis-
tance of each step for autofocus (Fig. 6.15). Note: It doesn’t indicate actual
excitation and emission wavelengths. When we select the number of ‘Steps’ in
the ‘Chip scan’ tab for autofocus (Fig. 6.12), it uses the distance of each step
(i.e., the thickness of one step) calculated from the following formula.

Distance of each step ( m m = é Emission wavelength / ( Lens NA ) ´ 10000) ù ´ ( 2 / 3 )

2

ë û

where NA is numerical aperture.

For example, when 4× and 20× lenses are used at the emission wavelength of
500 nm, the distance of each step is calculated as follows:

Step distance for 4 ´ lens = éë500 / ( 0.13 ´ 0.13 ´ 1000 ) ùû ´ ( 2 / 3 ) = 19.7 m m

6.3 Protocols 137

Step distance for 20 ´ lens = éë500 / ( 0.5 ´ 0.5 ´ 1000 ) ùû ´ ( 2 / 3 ) = 1.3 m m

where the numerical apertures for the 4× lens and the 20× lens are 0.13 and 0.5,
respectively.
Note: The recommended distance of each step is 10–20 μm (e.g., 400 nm emission
wavelength with 4× lens) for 3D-cultured cells and 2–5 μm (e.g., 120 nm emission
wavelength with 4× lens) for 2D cell monolayers. For a small step ­distance, it will
require a large number of ‘Steps’ for better autofocus. The emission wavelength
setting can be changed for each filter used and the morphology of cells.

6.3.2.2  etting XYZ Coordinates for Different Chips

S
and Objective Lenses

The ‘Chip layout’ tool bar allows to design the scanning layout (i.e., the number of
micropillars or microwells per chip) and determine specific XYZ coordinates of left
top (LT), right top (RT), and left bottom (LB) corners of the chip for accurate image
acquisition.
1. Click ‘Open’ in the ‘Chip layout’ tool bar to select a desired chip file from the
list (Fig. 6.16).
2. Select a desired chip file and click ‘Open’ in the window. For example, the layout
of the micropillar chip at 4× magnification lens is selected (Fig. 6.17).
3. To change the X, Y, and Z coordinates of the chips in the desired chip file, click
‘Set/Save’ button among four chips loaded (i.e., Chip 1, Chip 2, Chip 3, and
Chip 4) (Fig. 6.16). Note: This will allow one to change the X, Y, and Z coordi-
nates of LT, RT, and LB corners of the chip selected.
4. To read accurate XYZ coordinates, click the ‘Axis’ icon in the ‘XYZ Move’ tab
in ‘Home’ window to open the ‘Axis’ window (Fig. 6.18).
5. In the ‘Live view’ tab, click the ‘Live’ icon to observe a real-time cell image.
Note: We need to place three micropillars or microwells at the LT, RT, and LB
corners in the center of the ROI.

Fig. 6.16 The display of the ‘Chip layout’ tool bar

138 6 3D-Cultured Cell Image Acquisition

Fig. 6.17 Window displaying the list of chip files

Fig. 6.18 ‘Home’ window showing the ‘Axis’ icon highlighted in red in the ‘XYZ Move’ tab

Fig. 6.19 The display of

‘Axis’ window

6. In the ‘Axis’ window (Fig. 6.19), increase/decrease X, Y, and Z coordinates by

clicking the x+/x−, y+/y−, z+/z− buttons respectively to identify three ­micropillars
or microwells at the LT, RT and LB corners of the chip. Note: The changes in the
XYZ coordinates will be reflected immediately in the live image of cells.
7. Select the optimum XYZ coordinates at which the image is properly focused,
and enter those coordinates in the ‘Chip layout’ tool bar (Fig. 6.16).
6.5 Appendix 139

6.4 Summary

Imaging technology is the key determinant for a successful HCI assay. An ideal HCI
system enables the user to acquire high-resolution images of fluorescently labeled
cells at a high speed in real time. In this chapter, we introduced some commonly
available HCI devices and provided simple procedures for acquiring high-content
images in a high-throughput fashion from the micropillar/microwell chip platform
using the S+ Scanner. These procedures can be further modified to acquire high-­
content images from traditional HTS platforms such as 96- and 384-well plates.

6.5 Appendix

In the S+ Scanner, the following four filter sets are installed.

1. Multiband filter

Model (Semrock) DA/FI/TR/Cy5-A-000 (Multi-band)

http://www.semrock.com/SetDetails.aspx?id=2777
140 6 3D-Cultured Cell Image Acquisition

2. Orange filter

Model (Semrock) TxRed-4040C-000 (RED)

http://www.semrock.com/SetDetails.aspx?id=2799
3. Blue filter

Model (Semrock) DAPI-5060C-000 (Blue)

http://www.semrock.com/SetDetails.aspx?id=2779
References 141

4. Green filter

Model (Omega) XF404 (Green)

http://www.omegafilters.com/Products/Curvomatic

References

1. Buchser, W., Collins, M., Garyantes, T., Guha, R., Haney, S., Lemmon, V., et al. (2012). Assay
development guidelines for image-based high content screening, high content analysis and
high content imaging. In S. GS, C. NP, H. Nelson, M. Arkin, D. Auld, C. Austin, et al. (Eds.),
Assay guidance manual. (pp. 1–69). Bethesda, MD: Eli Lilly & Company and the National
Center for Advancing Translational Sciences.
2. Talbot, C. B., McGinty, J., Grant, D. M., McGhee, E. J., Owen, D. M., Zhang, W., et al. (2008).
High speed unsupervised fluorescence lifetime imaging confocal multiwell plate reader for
high content analysis. Journal of Biophotonics, 1(6), 514–521. doi:10.1002/jbio.200810054.
3. Celli, J. P., Rizvi, I., Blanden, A. R., Massodi, I., Gidden, M. D., Pogue, B. W., et al. (2014).
An imaging-based platform for high-content, quantitative evaluation of therapeutic response in
3D tumour models. Scientific Reports. doi:10.1038/srep03751.
4. Romano, S. N., & Gorelick, D. A. (2014). Semi-automated imaging of tissue-specific fluores-
cence in Zebrafish Embryos. Journal of Visualized Experiments, 87, e51533. doi:10.3791/51533.
5. Bickle, M. (2010). The beautiful cell: High-content screening in drug discovery. Analytical and
Bioanalytical Chemistry, 398, 219–226. doi:10.1007/s00216-010-3788-3.
142 6 3D-Cultured Cell Image Acquisition

6. Jahr, W., Schmid, B., Schmied, C., Fahrbach, F. O., & Huisken, J. (2015). Hyperspectral light
sheet microscopy. Nature Communications, 6, 1–7. doi:10.1038/ncomms8990.
7. Reynaud, E. G., Peychl, J., Huisken, J., & Tomancak, P. (2015). Guide to light-sheet micros-
copy for adventurous biologists. Nature Methods, 12(1), 30–34. doi:10.1038/nmeth.3222.
8. Rae Chi, K. (2012). High on high content: A guide to some new and improved high-content
screening systems. The Scientist Magazine, 1–7.
9. Starkuviene, V., & Pepperkok, R. (2007). The potential of high-content high-throughput
microscopy in drug discovery. British Journal of Pharmacology, 152(1), 62–71. d­ oi:10.1038/
sj.bjp.0707346.
Chapter 7
High-Content Image Analysis

Sean Yu, Pranav Joshi, Dong Woo Lee, and Moo-Yeal Lee

Contents
7.1 Introduction...................................................................................................................... 143
7.2 Materials.......................................................................................................................... 145
7.2.1 ImageJ.................................................................................................................. 145
7.2.2 S+ Chip Analysis................................................................................................. 145
7.3 Protocols.......................................................................................................................... 146
7.3.1 3D Cell Image Analysis with ImageJ.................................................................. 146
7.3.2 Examples of Image Processing with ImageJ....................................................... 149
7.3.2.1 Hue Filter............................................................................................ 149
7.3.2.2 Background Subtraction, Brightness Filter, and Region
of Interest (ROI................................................................................... 150
7.3.2.3 Outlier Exclusion................................................................................ 151
7.3.2.4 The Performance of the Plugin........................................................... 151
7.3.3 Image Deconvolution........................................................................................... 152
7.3.3.1 Performance of Deconvolution........................................................... 154
7.3.4 Plotting Dose Response Curves with S+ Chip Analysis...................................... 154
7.4 Summary.......................................................................................................................... 159
References................................................................................................................................. 159

7.1 Introduction

High-content imaging (HCI) and image processing of cells grown in 3D pose a

significant challenge because 3D cells are not grown in a single focal plane, and the
cell culture systems are often incompatible with traditional microscopes. Confocal
microscopy is an important tool for imaging 3D cells due to its ability to acquire
high definition images at various optical sections. However, the low scanning rate
and depth induce low throughput in image acquisition and also incur additional
problems such as photobleaching or phototoxicity [1–3]. To alleviate these issues,
miniaturized 3D cell cultures on a micropillar/microwell chip platform have been

© The Author(s) 2016 143

M.-Y. Lee, Microarray Bioprinting Technology,
DOI 10.1007/978-3-319-46805-1_7
144 7 High-Content Image Analysis

demonstrated, which facilitate high-throughput spheroid cultures in hydrogels

while offering better imaging capabilities. The whole sample depth can fit within
the focus depth of a normal objective lens due to the small dimensions.
For more accurate image acquisition, one needs to take precautions of carrying out
experiments, which will minimize defects, accentuate features, and help in extracting
meaningful information from the cell images acquired. For example, it is absolutely
important to avoid dust particles while rinsing and staining cells, which can be a
source of artifacts. In addition, it is necessary to avoid excitation/emission spectrum
overlapping among different fluorescent probes to obtain clear separation of fluores-
cence signals when multiple dyes are used simultaneously. When acquiring images,
various parameters of the imaging system such as exposure time and focus positions
should be optimized to avoid photo saturation and blurry image. After acquiring
images, the images can be corrected to account for various imperfections, including
high background noise, dust, imaging artifacts, uneven illumination, and blurring.
This is followed by segmentation to extract a particular feature in the image and quan-
tify the information. For example, to calculate the average size of cells, we need to
identify what constitutes a cell in the images—this may involve understanding the
morphology of a cell, the stained color of the cell, the size range of cells, etc. Once
these features have been properly identified and segmented, we can quantify the infor-
mation, transport and warehouse the data, and then manipulate and process the data.
Various biological image analysis tools such as CellProfiler, Fiji, BioImageXD,
EBImage, CellCognition, and other commercial software sold with HCI instruments
are available for high-content image analysis. These tools measure the features and
extract the quantitative values for analysis. If the experiment requires techniques
that are highly instrumentation-specific, the complementary software can serve as
an excellent starting point. Features measured typically consist of cell count, size
and shape, intensity, texture, location, and clustering. CellProfiler for instance was
developed for measuring cellular features such as the number of cells, size, and
other morphological features, protein level per cell, and cell cycle distribution.
Major steps of image processing and data analysis include image processing, object
segmentation, feature extraction and selection, and statistical analysis.
An example of image acquisition and analysis tools for 3D cells is the processing
and analysis software that comes with the S+ Scanner, including S+ Chip Analysis.
Other tools which can be highly subject-specific or research group-specific can be
found in research articles. There are groups and organizations that maintain and
curate these tools within their specialties, and the listings can often be found online.
For example, the Neuroimaging Tools and Resource Clearinghouse hosts and
categorizes various tools pertaining to neuroimaging at nitrc.org. General image
analysis tools can also be modified and extended to address a broader host of prob-
lems. Most notably, ImageJ is one of such tools that has been widely adopted due to
open source and extensibility. There are countless plugins developed for ImageJ. In
addition, there are novel image analysis platforms that base their architecture on
ImageJ such as Icy and Vaa3D. Due to its incredible versatility and widespread use,
ImageJ with an in-house plugin will be discussed extensively in the following section.
The landscape of imaging processing and analysis is broad and complicated; please
refer to Biological Imaging Software Tools by Kevin W Eliceiri et al. for an excel-
7.2 Materials 145

lent overview [4]. In this chapter, we summarize protocols for analyzing images of
3D-cultured cells on the micropillar/microwell chip obtained by staining with
multiple fluorescent dyes.

7.2 Materials

ImageJ is used to extract fluorescent intensity from 3D-cultured cell images, and S+
Chip Analysis is used to plot dose response curves from the extracted fluorescence
data in high throughput.

7.2.1 ImageJ

Among several image analysis software available, including Image-Pro, CellProfiler,

and the Image Processing Toolbox for MATLAB, we determined to use ImageJ
with an in-house BioPrinting plugin. ImageJ is a Java-based image analysis tool that
is designed for biological research and is free, open-source, lightweight, and exten-
sible. Its extensibility is an enormous advantage as it allows researchers to develop
macros or plugins specific to their research applications. We downloaded version
1.49v ImageJ (https://imagej.nih.gov/ij/download.html) and generated an in-house
plugin for high-throughput 3D cell image analysis by modifying existing plugins
and ImageJ-native functionality (http://rsb.info.nih.gov/ij/plugins/). The plugin can
apply a hue filter, a brightness filter, subtract background, select a region of interest,
ignore problematic outliers, save the processed images, and save the quantified fluo-
rescence data. After processing cell images obtained from the micropillar/microw-
ell chip using the BioPrinting plugin, the resulting fluorescence data are stored in an
analysis-friendly format, which are further analyzed with S+ Chip Analysis for
plotting dose response curves and calculating IC50 values, as well as Minitab or R
for statistical analysis. The IC50 value represents the concentration of a compound
at which cell growth is inhibited by 50 %.

7.2.2 S+ Chip Analysis

S+ Chip Analysis is an image processing and data analysis software particularly

developed for visualizing and analyzing images from microarray chip platforms.
This software is capable of automatic detection of spots on the microarray chip, and
removal of outlier spots. It can analyze multiple images in a single batch and gener-
ate dose response curves in multiple formats along with curve fitting. In addition,
this software exports the results in MS-Word format for easy access and storage of
data. All the above features make it an ideal software for high-throughput com-
pound toxicity analysis.
146 7 High-Content Image Analysis

7.3 Protocols

7.3.1 3D Cell Image Analysis with ImageJ

1. Download ImageJ from https://imagej.nih.gov/ij/download.html and install it in

your computer.
2. Place the BioPrinting plugin in the plugin folder within the ImageJ installation
folder to prevent the user from having to relocate the plugin every time they want
to run it.
3. Run ImageJ, locate the ‘BioPrinting’ tab under the ‘Plugins’ menu, and select
the ‘BioPrinting’ plugin (Fig. 7.1).
4. The Plugin will ask the user to select a source directory—select the directory
where the scanned images are stored (Fig. 7.2). Note: Multiple images in the
folder are processed simultaneously in a single batch.
5. In the ‘BioPrinting Image Analysis’ window, configure the processing variables
(Fig. 7.3).
6. Provide a name for a set of raw images in the ‘Label’ field. This name will be
used for the data files containing the fluorescence information as well as the
folder containing that data files.
7. If cell images contain multiple distinct dyes, the plugin can separate them using
a hue filter - specify the hue range by designating a minimum and maximum hue
value for one of the dyes and then re-run the plugin with a different hue range for

Fig. 7.1. Selection of the BioPrinting plugin from the Plugins menu
7.3 Protocols 147

Fig. 7.2. Selecting a directory with raw images to process

Fig. 7.3. BioPrinting plugin interface

the other dye to get separate quantitative data for two separate dyes in the same
image set. As for the hue range, calcein AM, for example, has a working range
of 20–130 (Table 7.1).
148 7 High-Content Image Analysis

Table 7.1. Hue ranges of fluorescent dyes with green, red, and blue colors.
Dyes Fluorescent color Minimum hue Maximum hue
Calcein AM Green 20 130
Ethidium homodimer-1 Red 0 50
Hoechst 33342 Blue 120 200
Tetramethyl rhodamine methylester Red 0 50
(TMRM)

8. Reduce the background by applying a uniform brightness filter in a 0–50 scale.

Note: If the images exhibit very low or no background fluorescence, then no
brightness filter is required (brightness filter value 0). If there is significant
background fluorescence, a brightness filter value of up to 50 is recommended.
Be wary that fluorescence from cells will be lost if the brightness filter value is
too high compared to the fluorescence signal of the stained cells.
9. If the background fluorescence is significant even after the application of the
brightness filter, and especially if the images are beset with high frequency
background noise, check the ‘Background Subtraction’ checkbox to apply the
rolling ball background subtraction method [5]. Note: This background
subtraction method is slow, so use it only when needed.
10. Save the processed images using the ‘Save Image’ checkbox, or forgo that
option to reduce plugin runtime.
11. Check the ‘Measure only the pillar area (ROI)’ checkbox to analyze a particular
region of interest. Note: As with most culture platforms, if the region of inter-
est is circular, this option can allow you to only measure fluorescence from
the circular central area.
12. Select ‘Exclude Outliers’ to ignore the data points where the fluorescence value
is two or more standard deviations away from the mean (Fig. 7.9). Note: Due
to dust or optical aberrations, there may be huge spikes of fluorescence. In
addition, certain procedural failures can lead to spot detachment from micro-
pillars thereby resulting in lack of cells.
13. Save quantified fluorescence data. Two files are generated (Figs. 7.4). The first
one includes fluorescence (i.e., raw integrated fluorescent intensity from image
processing) in a 14 by 38 matrix to match the layout of the micropillar/microw-
ell chip. Although this format is intuitive because it matches the layout of the
chip, it requires manipulation before it can be used in other statistical packages.
For that purpose, another file stores the data in a 2 column format, where
the first column indicates the block and the second stores the fluorescence.
Note: The chip is divided into 6 blocks, each block with a 14 by 6 matrix,
which is a common experimental layout on the chip for testing compound
toxicity (e.g., 6 dosages with 14 replicates).
14. The time log option is available for those who want to tinker with the plugin
and improve its performance. It will log how long it takes to perform each
processing step and how long it takes to process each image.
7.3 Protocols 149

14
Block Fluorescence

Block 1 1

Block 2 2

Block 3 3

532
38 Block 4 4

Block 5 5

Block 6 6
(A) (B)

Fig. 7.4. (a) The *.scn data output file that matches the layout of the micropillar/microwell chip
(14 × 38 data array) and (b) The *.txt output is in 6 block format (1 × 532 data array) for ease of
use elsewhere such as statistical software.

7.3.2 Examples of Image Processing with ImageJ

Note: The following images were selected for their undesirable features to demon-
strate the performance of the BioPrinting plugin and are not representative of images
acquired from the micropillar/microwell chip platform using the S+ Chip Scanner.

7.3.2.1 Hue Filter

As in live/dead assays with calcein AM and ethidium homodimer-1, fluorescent

images obtained from a multiband filter can have multiple peaks of fluorescence, but
one may want to acquire information about one type of fluorescence at a time. In
addition, optical aberrations and debris are often present in hues distinct from the
emission spectra of the stain. In these cases, we can apply a hue-based band-filter
through the ImageJ BioPrinting plugin to isolate the desired fluorescence (Fig. 7.5).

Fig. 7.5. The performance of hue filtering: (a) Original image with cells (in green), dust (in red)
and artifacts (in blue) and (b) Hue filtered image with cells alone. Green dots represent live cells
on the micropillar chip
150 7 High-Content Image Analysis

7.3.2.2  ackground Subtraction, Brightness Filter, and Region of Interest

B
(ROI)

When the hue of the stained cells is similar to background fluorescence, dust, and artifacts,
we can use several methods to eliminate unwanted noise fluorescence. Brightness filter
(Fig. 7.3) simply reduces pixel brightness values below a certain threshold to zero
(Fig. 7.6). This method is particularly useful for images with significant base fluores-
cence. Background subtraction (Figs. 7.3 and 7.7) can be used as an alternative of, or
in conjunction with, the brightness filter in reducing background fluorescence via the
rolling ball algorithm [5]. Often times, the undesirable fluorescence signals can be

Fig. 7.6. The effect of brightness filter: (a) Original image with no background filter applied, (b)
Brightness filter 10 applied, (c) Brightness filter 30 applied, and (d) Brightness filter 50 applied.
Blue dots represent nucleus stained with Hoechst 33342

Fig. 7.7. The effect of background subtraction: (a) Original image with no rolling background
subtraction applied and (b) Rolling background subtraction applied
7.3 Protocols 151

Fig. 7.8. Typical image processing: (a) Original image with edge artifacts, (b) Brightness filtered
image, (c) Background subtracted image, and (d) Region of interest (ROI) for fluorescence extrac-
tion. Blue dots represent nucleus stained with Hoechst 33342

found outside the area of the micropillar/microwell where cells are seeded and grown.
In that case, we can specify a region of interest (ROI) and have the software only
extract fluorescence data from those regions (Fig. 7.8).

7.3.2.3 Outlier Exclusion

When attempting to derive meaningful information from fluorescence data

(e.g., dose response curves), a few images can highly skew the results. For example,
spot detachment will lead to no fluorescence on certain images whereas smearing or
dust can lead to extremely high fluorescence on others. The burden, of course, lies
on the scientist to modify or fix the experimental procedures, leading to those
problems to ensure the reliability of the fluorescence data. In the meanwhile, there
is an option to exclude erroneous fluorescence data beyond two standard deviations
of the mean to discount invalid data points (Fig. 7.9). If the variability is expected
or part of the experimental design, then this option is not recommended.

7.3.2.4 The Performance of the Plugin

Without any processing, the scanned images often have high background fluores-
cence and high variability. By applying the various aforementioned features of the
plugin, we can significantly reduce the unwanted background fluorescence and
152 7 High-Content Image Analysis

Fig. 7.9. Excluded data points are indicated in red

diminish the inflated variability as shown in the following images. Briefly, the
micropillar chips were seeded with varying cell concentrations and stained with 2
different dyes—fluo-4 acetoxymethyl ester (AM), and monochlorobimane
(mBCl)—to determine the relationship between fluorescence and cell seeding den-
sity. Unprocessed raw fluorescence data acquired from the scanned micropillar
chips were compared with the processed data (Fig. 7.10). Both Fluo-4 AM and
mBCl had high background fluorescence, which was significantly reduced by the
plugin. Variability caused by spot detachment, dust, or artifacts was also signifi-
cantly reduced—for example, see mBCl at 2 million cells/mL.

7.3.3 Image Deconvolution

Image processing is a well-developed field, and there are vast and increasing num-
ber of solutions that could be borrowed and implemented to enhance signals while
eliminating noises. One such method is deconvolution. When cells are encapsulated
in a 3D gel, some of them may not be in focus at the same time. As a result, the out-­
of-­focus cell images appear somewhat poor and blurred, which could impact the
resulting fluorescence data. Mathematically, blurring can be considered as a convo-
lution between the in-focus (theoretical) image and a blurring factor. The blurring
factor could be removed using deconvolution so that an un-blurred image can be
remained. Therefore, we are focusing on image deconvolution as a potential solu-
tion to the blurring problem. Various image processing toolboxes or suites exist that
can perform deconvolution such as the image processing toolbox for MATLAB or
the several deconvolution plugins for ImageJ. We decided to use Image Restoration
software developed by Advanced Technology Inc. (ATI) for its decoupling and cus-
tomizability of kernel estimation methods and deconvolution methods. The Kernel
estimation method refers to the process of identifying the blurring factor, and the
deconvolution algorithm uses the estimated kernel to deconvolve the images. The
blind form of Richardson and Lucy (RL) algorithm combined with Inter-Level
Intra-Level Deconvolution (ILILD) was determined to be ideal for images of cells
7.3 Protocols 153

3.5 1.4

Fluorescence (Million) 3.0 1.2

Fluorescence (Million)
2.5 1.0

2.0 0.8

1.5 0.6

1.0 0.4

0.5 0.2

0.0 0.0
0 2 4 6 8 10 0 2 4 6 8 10

(A) Seeding density (Million cells/mL) (B) Seeding density (Million cells/mL)

10 3.0

2.5
Fluorescence (Million)

Fluorescence (Million)
8

2.0
6
1.5
4
1.0

2
0.5

0 0.0
0 2 4 6 8 10 0 2 4 6 8 10

(C) Seeding density (Million cells/mL) (D) Seeding density (Million cells/mL)

Fig. 7.10. The effect of the BioPrinting plugin (hue filter, background subtraction, brightness fil-
ter, ROI, and outlier exclusion) on cell images from the microarray chip: (a) Unfiltered data of cells
stained with fluo-4 AM, (b) Filtered fluo-4 AM data, (c) Unfiltered data from cells stained with
mBCl, and (d) Filtered mBCl data

obtained from the microarray chips. RL is a traditional deconvolution algorithm

developed independently by William Richardson and Leon Lucy in 1972 and 1974,
respectively. The algorithm was derived from Bayes’ theorem, and is an iterative
process, the blind form of which alternates between estimating the point spread
function (kernel) and the object (deblurred image), and converges on the maximum
likelihood solution for the kernel [6]. ILILD is a pyramid structured (multi-layered)
algorithm where inter-level deconvolution is performed at multiple levels of resolution,
from coarse to fine. In addition, at each level, residual deconvolution (intra-­level
deconvolution) is performed to recover edge features and details [7].
1. Install and open the Image Restoration software from ATI (Fig. 7.11).
2. Select the ‘Kernel estimation’ method from the drop-down list. Note: RL is
recommended.
154 7 High-Content Image Analysis

Fig. 7.11. ATI’s Image Restoration software for deconvolution. Blue dots represent nucleus
stained with Hoechst 33342

3. Select the ‘Deconvolution’ method from the drop-down list. Note: ILILD is
recommended.
4. Click on the ‘Batch test’ button and select the directory where the images are stored.
5. Click ‘Execute’ to run the software for deconvolution.

7.3.3.1 Performance of Deconvolution

3D cell cultures, especially multi-layered 3D cell cultures in the microwell chip, can
be challenging to image due to the cells not being in the same focal plane due to
relatively large spotting volume (typically 350–1000 nL). To test the efficacy of
deconvolution on multi-layered 3D cell cultures, microwell chips were seeded with
a double layer of hoechst 33342-stained cells in varying concentrations, imaged,
and processed using the ImageJ Bioprinting plugin as well as the deconvolution
software. Deconvolution was found to increase the sensitivity and the slope of the
fluorescence calibration as well as the wellness of fit (Fig. 7.12).

7.3.4 Plotting Dose Response Curves with S+ Chip Analysis

1. Open S+ Chip Analysis program by double clicking the S+ Chip Analysis icon
on the desktop (Fig. 7.13).
2. Click the ‘Experiment’ button on the ‘Menu’ bar to open the ‘Experiment
condition’ window (Fig. 7.14).
7.3 Protocols 155

Original
5

Fluorescence (Million)
RL - ILILD
4

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

Seeding density (Million cells/mL)

Fig. 7.12. The effect of deconvolution on fluorescence calibration curves. The blind Richardson
and Lucy (RL) kernel estimation method and the Inter-Level Intra-Level Deconvolution (ILILD)
algorithm were applied to the images of double-layered Hoechst 33342-stained cells

Fig. 7.13. S+ Chip Analysis software icon

Fig. 7.14. S+ Chip Analysis window with the ‘Menu’ bar and the ‘Experiment condition’
window

3. Click the ‘Browse’ button in the ‘Experiment condition’ window and select the
experiment folder where image data are saved (Fig. 7.15).
4. Click the ‘Show’ button to see the data. The data will be displayed on a popup
window (Fig. 7.15).
5. Select the ‘Chip Platform’ for the layout of spots (i.e., experimental conditions)
on the chip from the drop down list (Fig. 7.16).
6. Enter the dilution factor in the ‘Compound Dilution’ field (Fig. 7.16).
7. Enter the name of the cells used in the ‘Cell Type’ field, select the unit of com-
pound measurement (e.g., μM) from drop down list in the ‘Unit’ field, and then
click the ‘Apply’ button (Fig. 7.16).
156 7 High-Content Image Analysis

Fig. 7.15. The ‘Experiment condition’ window with browse button for selecting the desired folder
and the ‘Show’ button for displaying the data, both highlighted in red

Fig. 7.16. The display of the ‘Experiment condition’ window

Fig. 7.17. The ‘Status’ window displaying the status of data analysis

8. The cell type and the unit of test compounds will be applied to all the data and
will appear in the ‘Cell Type’ and ‘Unit’ grid.
9. Enter the compound names and the highest concentrations tested against the
cell type in the first and second row, respectively (Fig. 7.16). Note: The name
of six compounds and the respective concentrations should be entered
individually in column numbered from 1 to 6, respectively.
10. Enter the name of enzymes or fluorescent dyes or any reagents applied to the
chip in the ‘Enzyme’ grid (Fig. 7.16). Note: While analyzing multiple chips, if
the compound names, concentration values, and enzymes/fluorescent dye
names are identical, the user can copy and paste parameters from one chip to
other chips.
11. If outliers are to be detected and removed automatically, check the ‘Detect
Outliers’ option (Fig. 7.16).
12. After entering the parameters press the ‘Save’ button to save the parameters
(Fig. 7.16).
7.3 Protocols 157

Fig. 7.18. A window

displaying ‘Graph plotting
completed’ message

Fig. 7.19. The display of the ‘Experiment result’ window

13. Press ‘Run’ button to start analyzing the data (Fig. 7.16). Status of analysis will
be displayed in ‘Status’ window (Fig. 7.17).
14. After the data is analyzed, ‘Graph plotting completed’ message will appear.
Press OK to complete the analysis (Fig. 7.18).
15. The ‘Experiment Result’ window will show up after the completion of data
analysis. Inspect the individual dose response curve and check ‘Select All’ to
select all the graphs obtained or select individual graphs from the chip prior to
combining the replicates from different chips (Fig. 7.19).
16. Select ‘Export’ in the menu bar to export the result/graph into an MS-Word file
(Fig. 7.19). Note: This process allows one to combine dose response curves
from multiple replicate chips and draw one dose response curve.
17. Once the exporting process is complete, a ‘Save As’ window will appear
with ‘Report.doc’ as its default file name. Press save if you want to save it
with the default file name or else enter a desired file name and press ‘Save’
button (Fig. 7.20).
18. A ‘Select Plot Types’ window will show up, check ‘% Viability vs. Log concen-
tration—summary plot’ and click ‘Export’ button (Fig. 7.21).
19. A pop up window with a message ‘Document file created’ will appear. Press
OK to finish exporting the file (Fig. 7.22).
20. A summary table with test conditions and IC50 values as well as corresponding
dose response curves will be saved in the same folder containing data file
(Fig. 7.23). Note: To produce a conventional sigmoidal dose-response curve,
the fluorescence intensities of all cell spots are normalized to the fluorescence
intensity of 100 % live cell spots (e.g., cell spots contacted with no compound)
158 7 High-Content Image Analysis

Fig. 7.20. The ‘Save As’ window appears to save the result in a desired filename inside a desired
folder

Fig. 7.21. ‘Select Plot Types’ window

Fig. 7.22. A pop up window displaying ‘Document file created’ message

References 159

Fig. 7.23. Examples of dose-response curves with IC50 values obtained from data analysis in S+
Chip Analysis software. Dose-response curves of lovastatin for (a) nucleus morphology, (b) mito-
chondrial membrane potential, and (c) glutathione level are presented

and then plotted against the logarithm of test compound concentrations. The
sigmoidal dose-response curves (variable slope) and IC50 values for each test
condition are obtained using the following equation:

(
Y = Bottom + é( Top - Bottom ) / 1 + 10(
ë
LogIC50 - X )* H
)ùû
where IC50 is the midpoint of the curve, H is the hill slope, X is the logarithm
of test concentration, and Y is the response (% live cells), starting at Bottom and
going to Top with a sigmoid shape.

7.4 Summary

In this chapter, we introduced protocols for processing the images of 3D-cultured cells
on the micropillar/microwell chip in a high-throughput manner using ImageJ. We also
demonstrated a deconvolution algorithm for deblurring the out of focus images using
Image Restoration software developed by Advanced Technology Inc. (ATI), South
Korea. Finally, we explained step-by-step protocols for plotting dose-response curves
and calculating IC50 values using S+ Chip Analysis software, specifically designed for
visualizing and analyzing images from microarray chip platforms.

References

1. Lang, P., Yeow, K., Nichols, A., & Scheer, A. (2006). Cellular imaging in drug discovery.
Nature Reviews. Drug Discovery, 5(4), 343–356. doi:10.1053/j.gastro.2006.06.028.
2. Jahr, W., Schmid, B., Schmied, C., Fahrbach, F. O., & Huisken, J. (2015). Hyperspectral light
sheet microscopy. Nature Communications, 6, 1–7. doi:10.1038/ncomms8990.
3. Scherf, N., & Huisken, J. (2015). The smart and gentle microscope. Nature Biotechnology,
33(8), 815–818. doi:10.1038/nbt.3310.
160 7 High-Content Image Analysis

4. Eliceiri, K. W., Berthold, M. R., Goldberg, I. G., Ibáñez, L., Manjunath, B. S., Martone, M. E.,
et al. (2012). Biological imaging software tools. Nature Methods, 9(7), 697–710. doi:10.1038/
nmeth.2084.
5. Sternberg, S. (1983). Biomedical image processing. IEEE Computer, 16(1), 22–34.
6. Fish, D. A., Brinicombe, A. M., Pike, E. R., & Walker, J. G. (1995). Blind deconvolution by
means of the Richardson–Lucy algorithm. Journal of the Optical Society of America. A, 12(1),
58–65.
7. Y. Ding, I. Park, X. Cui, Van Huan. Nguyen, Hakil Kim, Trung Dung Do et al. (2015). Inter-­
level and intra-level deconvolution based image deblurring algorithm for wide field micros-
copy. In 2015 8th International Conference on Biomedical Engineering and Informatics
(BMEI). Shenyang: IEEE (pp. 90–95). doi:10.1109/BMEI.2015.7401479.
Chapter 8
Applications of Microarray Bioprinting

Alexander Roth, Emily Serbinowski, and Moo-Yeal Lee

Contents
8.1 Introduction...................................................................................................................... 161
8.2 Assay Development for Microarray Bioprinting Technologies....................................... 162
8.2.1 Hepatotoxicity Assays......................................................................................... 162
8.2.1.1 Phase I and Phase II Drug Metabolizing Enzyme Assays.................. 164
8.2.1.2 Drug Transporter Assays.................................................................... 165
8.2.1.3 Oxidative Stress Assays...................................................................... 166
8.2.2 Neurotoxicity Assays........................................................................................... 166
8.2.2.1 Oxidative Stress and Related Assays.................................................. 167
8.2.2.2 Ion Channel Assays............................................................................ 167
8.2.2.3 Drug Metabolism Assays.................................................................... 168
8.3 Simulation of the In Vivo Microenvironment: Liver Applications.................................. 169
8.4 Other Applications........................................................................................................... 171
8.5 Summary.......................................................................................................................... 171
References................................................................................................................................. 172

8.1 Introduction

Applications for microarray bioprinting include compound toxicity and efficacy

assessments, disease diagnostics, and simulation of tissue microenvironments on a
small scale. The low volume, high-throughput nature of this system makes it ideal for
rapid screening and assay development, particularly in regards towards assays that
can be used on various tissue engineering platforms. In this section, we highlight the
development of assays for high-throughput screening (HTS) using microarray bio-
printing technologies, with a specific focus on toxicity assays that affect the liver and
the central nervous system (CNS). Next, we will highlight the various organ systems
that can be recapitulated using 3D bioprinting, and the advantages and disadvantages
of tissue miniaturization for microarray technologies. Finally, we’ll discuss other
applications, including tissue regeneration studies, and drug patterning.

© The Author(s) 2016 161

M.-Y. Lee, Microarray Bioprinting Technology,
DOI 10.1007/978-3-319-46805-1_8
162 8 Applications of Microarray Bioprinting

8.2  ssay Development for Microarray Bioprinting

A
Technologies

Adverse drug reactions (ADRs) are among the five leading causes of death in the
United States [1]. Among ADRs, hepatotoxicity and cardiotoxicity are the most
common events observed. Other organ-specific ADRs, including neurotoxicity and
renal toxicity, are observed in the presence of certain drugs. Additionally, the cost
for successful drug development can range from $160 million to $2.5 billion, and it
takes ten to 15 years from lead compound discovery to clinical evaluation [2]. About
60 % of this cost comes from clinical trials [3]. With roughly a post-marketing fail-
ure rate of one drug per year, the human, financial, and capital tolls are further
increased.
One of the major issues with drug development is the reliance on animal models
to predict ADRs before clinical trials are conducted. Animal models have been
shown to be poor predictors for ADRs, as other mammals do not adequately mimic
all the human tissues and organ functions [4–7]. Additionally, animal models are
generally significantly more expensive than preliminary in vitro toxicity studies.
Thus, it is a goal of scientists to develop a suite of in vitro toxicity assays that can
be used to accurately predict human in vivo toxicity, and potentially reduce the costs
associated with drug development.
In this section, we will focus on assays that can be implemented for in vitro tox-
icity studies on the microarray printing platform. Specific focus will be placed on
hepatotoxicity assays, focusing on phase I and phase II drug metabolizing enzymes
(DMEs), transporter assays, and oxidative stress detection. Additionally, neurotox-
icity will be explored in regards to assays that detect specific mechanisms for neu-
rotoxicity, and how this toxicity is differentiated between adult neuronal cells and
neural stem cells (NSCs).

8.2.1 Hepatotoxicity Assays

The liver is arguably the most important organ tied to toxicity assays as the liver is
responsible for clearance of toxic substances from the body, and hepatotoxicity is
the most common form of ADR that is associated with market withdrawals. Toxicity
in the liver can stem from general elevated exposures to certain compounds, known
as intrinsic drug-induced liver injury (DILI), or it can stem from individual varia-
tions in the tissue microenvironment and the genetic of the individual, known as
idiosyncratic drug-induced liver injury (IDILI). Here, we will focus on the assays
associated with detection of DILI.
Drug metabolism in the liver is illustrated in Fig. 8.1. An external, apical membrane
transporter can allow drugs to enter the cell, though it is also possible for the
structure of the molecule to pass through the membrane via diffusion. These com-
pounds are then covalently modified in 2–3 steps to be transported into the bile via
8.2 Assay Development for Microarray Bioprinting Technologies 163

Drug
Passive
diffusion
Influx
transporter
(NTCP, OCT,
OATP) Necrosis/
Protein Apoptosis Nucleus
Lipid
Parent peroxidation damage
drug ER stress
DNA
Phase I metabolism damage
(CYP, FMO, MAO) ROS/RNS
generation NFkB, Nrf2 Cell
Reactive Antioxidants survival
metabolite (GSH, SOD,
catalase, GPx)
Phase II metabolism Protein
(UGT, SULT, GST, NAT) adduct
ROS
MPT,
Conjugated Necrosis
mtDNA
metabolite (RIPK, PGAM) ATP/glucose damage
depletion
Apoptosis ROS
Efflux (JNK, caspase, Bcl2) Mitochondria
transporter
(BSEP, BCRP,
MDR, MRP)

Hypersensitive immune response

Toxicity (APC, KC, T cell, cytokine)
Excretion

Fig. 8.1. Simplified mechanisms of metabolism-mediated drug toxicity in liver cells. Abbreviations
are used as follows: NTCP Na+-taurocholate cotransporting polypeptid, OCT organic cation trans-
porter, OATP organic anion transporting polypeptide, BSEP bile salt export pump, BCRP breast
cancer resistance protein, MDR multidrug resistance protein, MRP multidrug resistance-associated
protein, CYP cytochrome P450, FMO flavin-containing monooxygenase, MAO monoamine oxi-
dase, UGT UDP-glucuronosyltransferases, SULT sulfotransferase, GST glutathione S-transferase,
NAT N-acetyl transferase, ROS reactive oxygen species, RNS reactive nitrogen species, NFκB
nuclear factor kappa-light-chain-enhancer of activated B cells, Nrf2 nuclear factor erythroid
2-related factor 2, GSH glutathione, SOD superoxide dismutase, GPx glutathione peroxidase, MPT
mitochondrial pore transition, mtDNA mitochondrial DNA, RIPK receptor-interacting serine/
threonine protein kinase, PGAM phosphoglycerate mutase, JNK c-Jun N-terminal kinase, Bcl2
B-cell lymphoma 2, APC antigen-presenting cell, KC Kupffer cell

basolateral membrane proteins for eventual elimination. The modification of the

structure of the parent drug involves the formation of reactive functional groups
such as alcohols and aldehydes (phase I drug metabolism), followed by subsequent
reaction with bulkier, less reactive functional groups, such as acetyl, sulfate, and
glucuronyl functional groups (phase II drug metabolism). The parent drug, reactive
intermediates, and final products can all be transported out of the cell. ADRs in the
liver can result from the dysfunction of any DMEs or specific drug transporters,
which can have effects ranging from localized inflammation to complete liver
failure.
In addition, all forms of the drug may produce reactive byproducts in the form of
reactive oxygen species (ROS) and reactive nitrogen species (RNS), which can have
164 8 Applications of Microarray Bioprinting

the effects of damaging DNA, proteins, and lipids. Mechanisms that clear ROS
and RNS from the body can contribute to limiting the effects of ADRs, but if the
activities of the enzymes that combat oxidative stress are compromised, this can
also have a deleterious effect on hepatocytes. Finally, molecules damaged by any
forms of the drug or ROS/RNS can be presented on the surface of the hepatocyte,
potentially triggering an acute inflammatory response. While most responses are
immediate and short term, sometimes the inflammatory response is sustained and
this can lead to further damage, or potentially autoimmunity against the liver, which
is potentially fatal.

8.2.1.1 Phase I and Phase II Drug Metabolizing Enzyme Assays

Assays that focus on detection of hepatotoxicity rely on metabolic profiling associ-

ated with various DMEs and potential interactions with drugs. The enzyme that is
responsible for the majority of the phase I drug metabolism in the liver is cyto-
chrome P450 oxidase (CYP450). CYP450 reacts O2 with functional groups on par-
ent drugs to produce alcohols, using NADPH as a cofactor. CYP450 has many well
studied isoforms that are associated with the metabolism of compounds with signifi-
cantly varied structures. While other enzymes involved in phase I drug metabolism
exist within hepatocytes, polymorphisms in various CYP450s have resulted in the
ADRs being observed in the human liver.
CYP450 activity for most clinically relevant isoforms has been determined for
both ex vivo and in vitro studies using coumarin-based metabolism. Each of the
CYP450 isoforms can metabolize variants of the fluorogenic substrate of coumarin.
For the CYP2B6, CYP2E1, and CYP3A4 isoforms, this results in the formation of
7-hydroxy-4-trifluoromethyl coumarin (HFC) from different substrates [8, 9].
Previous results have shown that HFC production from these substrates is linear
with time, meaning there is good correlation with the enzymatic activity. All of the
reactions utilize hydroxylation of the ether oxygen that is conjugated to the func-
tional group. Zarowna-Dabrowska et al. have alternatively exploited the fluorescent
properties of resorufin to detect CYP1A1/2 activity via a 7-ethoresorufin-O-­
deethylase (EROD) assay under confocal microscopy [10]. CYP450 activity detec-
tion isn’t strictly limited to using direct fluorescent and fluorogenic substrates. Meli
et al. have demonstrated the use of immunofluorescent microarrays for the detection
of CYP2C9 and CYP3A4 activity [11].
The advantage of fluorogenic substrates is that they can be both printed in micro-
array chip platforms, and detected under fluorescence microscopy. Generally, fluo-
rogenic substrates are also more sensitive over a broader concentration range, and
can hence detect both low and high activities of CYP450s quite easily. Promega
provides several commercially available assays that can be utilized within the
microarray chip platforms for detection of phase I metabolic activity, in fluorescent,
colorimetric, and luminescent forms, through fluorescence is most compatible with
cellular microarrays. The drawback to these substrates is that if the cell line is capa-
ble of significant phase II metabolism, inhibitors need to be added to the solution to
8.2 Assay Development for Microarray Bioprinting Technologies 165

properly assess CYP450 activity. Additionally, there are substrates that exist which
can be metabolized by isolated CYP450s, but may not be adequate to permeate the
cell membrane. Human liver microsomes (HLMs) may be used to convey or alter
CYP450 activity to cells lacking CYP450 activity, or can be used as a suitable con-
trol for assaying activity of certain compounds with various CYP450s [9]. HLM
activity can be assayed using microarray bioprinting technology, which offers the
alternative of conventional 96-well plate assays at a lower cost.
Phase II drug metabolism is the replacement of reactive functional groups added
by phase I drug metabolism with bulkier, easily cleared functional groups. Phase II
enzymes include uridine 5′-diphosphoglucuronosyltransferases (UGTs), sulfotrans-
ferases (SULTs), glutathione S-transferases (GSTs), and N-acetyltransferases
(NATs). For all of these enzymes, there exist clinically relevant polymorphisms that
make certain individuals more susceptible to ADRs. Specific protein assays that are
relevant to ADRs have been developed for measuring UGT and GST activity. UGT
activity can be measured by using 4-methylumbelliferone (4-MU), a fluorescent sub-
strate of the coumarin family that undergoes glucuronidation in the presence of UGT
[12, 13]. Additionally, owing to the role of UGT in clearance of steroids, β-estradiol
can be used to assay UGT1A1 activity via detection of estradiol 3-­glucuronide [14].
GST can be measured and distinguished using 1-chloro-2,4-­dinitrobenzene (CDNB)
as a substrate [15]. This reaction involves the removal of the chlorine via reaction
with the sulfur on glutathione (GSH). While CDNB can be used as a substrate for
total GST activity, it is measured via absorbance, and less ideal for microarray bio-
printing applications. Monochlorobimane (MCB) is recommended as the reaction
with GSH produces a fluorescent compound (excitation 380 nm/emission 461 nm),
which has been used previously to assess GSH levels in cells.

8.2.1.2 Drug Transporter Assays

Transporters represent a very diverse class of enzymes. Influx transporters and

efflux transporters are associated with the transport of various compounds from the
lumen (influx) to the bile (efflux). Generally, toxicity can result from dysfunction in
efflux transporter, as reactive metabolites can accumulate in the cytoplasm if they
cannot escape the cell. Included in the list of efflux transporters are breast cancer
resistance protein (BCRP), bile salt export pump (BSEP), multidrug resistance pro-
tein (MDR), and multidrug resistance-associated protein (MRP). BCRP, BSEP, and
certain MDRs and MRPs have all been linked with particular ADRs if their function
is compromised.
BSEP is the most common target for transporter activity measurement owing to
its altered function during cholestasis. Assays used for measuring BSEP activity
include the vesicular transport assay (VTA) and the ATPase assay [16]. Both assays
utilize ATP as a co-transporter molecular necessary to drive metabolites out of the
cell against concentration gradients. While this is specialized for the purpose of
measuring BSEP activity, this method can be used for measuring ABC transporter
activity. In addition to these assays, drug uptake assays are available for various
166 8 Applications of Microarray Bioprinting

metabolic compounds [17]. MDR1 has a significant amount of substrates available

that have similar substrate specificity with CYP3A4, owing to the co-regulation of
both proteins by the pregnane X receptor (PXR) transcription factor [18–20].
Induced MDR1 activity can be measured by rifampicin, while inhibition can be
measured verapamil [20]. Additionally, MRP2 can be inhibited by 5(6)-carboxy-­
2′,7′-dichlorofluorescein (CDCF), which is a commercially available fluorogenic
substrate [14, 21, 22]. While transporter detection is possible with fluorescent and
fluorogenic substrates in the standard well-plate format, only genetic expression of
transporters has been achieved in cellular microarrays.

8.2.1.3 Oxidative Stress Assays

While oxidative stress can affect numerous organelles, antioxidant enzymes local-
ized to the mitochondria are the antioxidants most associated with individualized
ADRs. This makes the mitochondrion the primary indicator for liver functionality
tests related to oxidative stress. For general mitochondrial toxicity assays, tetra-
methyl rhodamine (TMRM) can be used to stain for measuring mitochondrial mem-
brane potential, as depolarization of the membrane is a symptom of apoptosis [23].
Cytochrome c release and oxygen consumption rate (OCR) may also be used to
determine effects on energetic metabolism [24]. OCR determination utilizes the
metabolic turnover of O2 in energetic metabolism [25, 26]. OCR can be measured
fluorescently using a variety of methods, though fluorescent probes are most com-
mon. MitoXpress from Luxcel (ex/em 380/650) can be used for direct fluorescence
measurement of OCR by quantifying oxidative phosphorylation, though time-­
resolved fluorescence (TR-F) is recommended instead of direct fluorescence.
Similar assays are available for OCR measurements from other companies.
If antioxidant levels are innately low, GSH may also be used to determine
effects of drug on mitochondria. MCB may be used as a fluorogenic substrate to
quantify GSH levels [27], but the activity of GSH as a function of antioxidant
stress must be distinguished from that of GSH used in phase II drug metabolism.
While GSH acts as a general indicator for oxidative stress, this can be sufficient
enough to indicate dysfunction with antioxidant enzymes, or even the general
presence of an ADR.

8.2.2 Neurotoxicity Assays

Aside from hepatotoxicity, ADRs are also found in neural cell lineages, particularly
neural stem cells (NSCs) [28]. NSCs are the cells that eventually differentiate into
the various lineages associated with the central nervous system, including neurons,
oligodendrocytes, and astrocytes. Much of the toxicity that drives ADRs in NSCs
relates to damaging or blocking ion channels on the cell membrane surface, induc-
ing oxidative stress, denaturing of nucleic acids and proteins from drug metabolites,
8.2 Assay Development for Microarray Bioprinting Technologies 167

membrane compromise, and both caspase-dependent and -independent apoptosis.

This section will deal with assays that detect these various mechanisms, focusing on
assays that can be used in microarray bioprinting applications.

8.2.2.1 Oxidative Stress and Related Assays

Like in hepatocytes, NSCs may undergo apoptosis in the presence of certain

compounds. Apoptosis may be derived from either direct interaction of certain
compounds with molecules within the pathways necessary for apoptosis, or it can
derive from oxidative stress related to reactive metabolites. In order to detect this
behavior, assays have focused on observing the behavior of specific molecules
within these pathways. Some of these methods, such as simple live-dead stains dis-
cussed in Chap. 5 can elucidate the basic mechanisms of toxicity, pointing towards
mitochondrial dysfunction, glutathione depletion, or caspase-mediated apoptosis.
In addition, oxidative stress often couples with reduced membrane integrity, and the
potential for nucleic acid and peptide conjugation from reactive metabolites.
While these cellular symptoms aren’t necessarily indicative of individualized
ADRs, they are all indicative of generalized ADRs.
Fluorescent methods used to quantify these mechanisms are more advantageous
for microarray bioprinting technologies owing to the small volume dispensed in the
microarray, and thus the increased sensitivity necessary to detect the potential for
ADRs. For protein and nucleic acid modification, benzo(a)pyrene diol epoxide
(BPDE) is used in conjunction with a colorimetric enzyme-linked immunosorbent
assay (ELISA) to detect such molecules that may have been conjugated due to drug
byproducts. The commercially available ELISA quantifies adduction via horse-­
radish peroxidase (HRP) conjugated to secondary antibodies. However, this second-
ary antibody could potentially be replaced by a fluorescent one to make it suitable for
microarray platforms [29, 30]. Abcam provides a kit (ab112158, ex/em: 490/520 nm)
that can quantify adduction and GSH levels via measuring thiol groups with their
reagent as it conjugates with free thiols on proteins and GSH.
For cell membrane compromise, lactate dehydrogenase (LDH) can be used to
quantify the viability of cells in 96-well plates and in microwell chips [31–33].
When cells lyse or the cellular membrane is compromised, LDH may leak into the
surrounding environment. CytoTox One has manufactured an assay that utilizes the
conversion of lactate to pyruvate via NAD+ as a cofactor for NADH. It then quanti-
fies the level of NADH via NADH reductase, which converts resazurin to resorufin,
a fluorescent compound.

8.2.2.2 Ion Channel Assays

Another way ADRs are triggered in NSCs involve the blocking of ion channels. Ion
channels are associated with maintaining plasma membrane potential, but they can
also be found on the membranes of various organelles. Within axonal differentiated
168 8 Applications of Microarray Bioprinting

Fig. 8.2. The conversion of CoroNa green via intracellular esterases to a fluorescent indicator for
intracellular Na+ ions

neural cell lineages and NSCs, ion channels also play a role in synaptic transmission
and the generation of action potentials, which is a common behavior associated with
cells of the CNS and all versions of muscle cells. Disruption in ion channels may
lead to disruption in the transmittance of action potentials, and the subsequent loss
of signals. K+, Na+, Ca2+, and Cl− may all be blocked within neural cells and NSCs.
The disruption in ion flow could result from competitive ionic interactions of ions
with similar valence charges (i.e., transition elements Ni2+ and Cd2+ with Ca2+),
covalent modification of proteins, or competitive binding with necessary ligands
that may allow channels to open and close. Common compounds used as ion chan-
nel blocker controls include 4-aminopyridine, disopyramide, amlodipidine besyl-
ate, and 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS).
There are commercially available assays used to detect ion channel activity and ion
channel blocking. FluxOR potassium ion channel assay from Life Technologies uses
thallium ions to trigger the opening of K+ channels, activating the fluorescent dye inside
the cell, while also using probenecid to prevent the exit of the fluorescent dye. The
CoroNa green sodium indicator from Life Technologies uses a fluorescein based mol-
ecule with a binding pocket for Na+, which is activated inside cells despite Na+ ion
concentrations generally being greater extracellularly (Fig. 8.2). Fluo-4AM reagent
from Life Technologies is believed to detect Ca2+ molecules intracellularly via cleavage
of the acetoxymethyl ester groups. This may be coupled with probenecid to maintain
the fluorescent moieties inside the cell. Premo halide sensor from Life Technologies
uses I− to mimic the influx of Cl− ions. I− will quench a yellow fluorescent agent inside
the cell. It is known so far that Fluo-4AM is compatible with a high-throughput bio-
printing system, and that the rest of these reagents are cell compatible.

8.2.2.3 Drug Metabolism Assays

Another such mechanism for detecting NSC toxicity is through drug metabolism
itself. Like hepatocytes, adult neural cells and NSCs metabolize drugs and can be
adversely affected by mutations in DMEs [34]. Drug metabolism in NSCs utilizes a
8.3 Simulation of the In Vivo Microenvironment: Liver Applications 169

similar process of phase I and phase II metabolism, and there are separate classes of
transporters for compound influx and efflux. CYP450s are the DMEs primarily
responsible for phase I drug metabolism in NSCs, and there is some overlap with
the individual CYP450 isoforms that are responsible for drug metabolism.
Additionally, phase II DME of the GST family have been tied to individual ADRs,
and dysfunction in transporter MDR1 may also increase susceptibility to ADRs.
NSC assays which deal with drug metabolism focus on similar assays that are
involved in assessing DME-related toxicity in hepatocytes. While the protocols are
identical for detection of activity with these enzymes, many of the CYP450 iso-
forms unique to NSCs are involved in the metabolism of fatty acids and steroids [35,
36]. Thus, substrates for metabolism and assaying activity of neuronal CYP450
enzymes can be such biological compounds, or their analogues. Likewise, many of
the GST and MDR1 assays used on hepatocytes are also compatible with NSCs,
including CDNB and MCB for GST activity, and verapamil for MDR1. These
assays are compatible with cells and are fluorescent, making them suitable assays to
be used for determining toxicity on a microarray chip platform.

8.3  imulation of the In Vivo Microenvironment: Liver

S
Applications

Automatic liquid dispensation for microarray technologies has until recently seen
limited use outside the organism. While research has focused on development of
whole 3D tissue constructs, technology has not reached the point where whole
organs can be printed for transplantation purposes [37]. The creation of physio-
logical networks is quite complex, and maintaining and mimicking all of the nec-
essary functions of a 3D-bioprinted tissue is difficult. However, microarray 3D
bioprinting can be used to mimic small scale in vivo behaviors. While this does
not make it suitable for large scale cultures, miniaturization of the bioprinting
process can still give insight into certain in vivo behaviors in a high-throughput
manner. In this section, we will focus on the use of bioprinting 3D cell cultures for
high-throughput applications, with a focus on co-culture systems related to mimicking
the liver (Fig. 8.3).
In order to understand the nature of co-culture systems, we must also understand
the cells and vasculature associated with the tissue that is to be mimicked. In the
case of livers, there are four types of cells that make up over 95 % of the cell volume
and count: hepatocytes, Kupffer cells (KCs), sinusoidal endothelial cells, and
hepatic stellate cells [38, 39]. Hepatocytes make up the majority of the cell volume
and count (roughly 60 % of the cell count and 80 % of the cell mass) [38]. These
cells serve a myriad of functions, including drug metabolism, bile flow, cholesterol
synthesis, glycolysis, gluconeogenesis, and amino acid catabolism [39]. The roles
the hepatocytes take on relate closely towards their proximity towards the portal
vein, the main supplier of oxygen to the liver [39]. The other cell types are localized
to other parts of the liver. The KCs are the resident macrophages of the liver that
initiate inflammation in response to toxic injury [38, 40, 41]. Sinusoidal endothelial
170 8 Applications of Microarray Bioprinting

Bioprinting with S+
microarray spotter
Combinatorial cell printing Chip
into the microwell chip In vitro In vivo scanning
study study

Chip incubation
in growth media Chip
implanting
Cell staining
Cells mixed with Chip
“bioink” Photocrosslinking recovery
in a 96-well plate

Image
analysis

High-content cell imaging

Creating human tissues
for toxicology assessment

Fig. 8.3. Microarray three-dimensional (3D) bioprinting technology for creating human tissues
by dispensing multiple human cells in biomimetic hydrogels (‘bioinks’) layer-by-layer with preci-
sion printing robots, thereby potentially revolutionizing tissue engineering and disease modeling
for screening therapeutic drugs and studying toxicology. Miniaturized liver tissue blocks (as small
as 1 mm3) are generated by printing several layers of human hepatic cells in photocrosslinkable
hydrogels with extracellular matrices (ECMs) and growth factors (GFs) into a microwell chip
using S+ MicroArrayer. After gelation, the microwell chip containing hundreds of biomimetic
conditions is incubated in a petri dish with growth media or implanted in immunocompromised
animals for rapidly testing optimum microenvironments to create biomimetic human tissues.
Created human liver tissue constructs are stained and scanned with S+ Scanner for high-content
imagining (HCI) of hepatic functions as well as predictive assessment of drug toxicity

cells border the blood vessels, and are important as they have interactions with any
blood cells (including KCs) and can present antigens to CD4+ (mature) helper T
cells [42, 43]. They are also important on limiting the effects of shear on hepato-
cytes [38, 39, 44]. Finally, hepatic stellate cells exist in the space of disse, a region
between the endothelial cells, and the hepatocytes. The stellate cells store fatty acids
and some vitamins, but they may behave like fibroblasts and synthesize collagen if
liver restructuring is necessary [38, 39, 45].
To mimic the behavior of the liver, focus has shifted towards 3D co-culture sys-
tems, utilizing both hepatocytes and non-parenchymal cells. While KCs are most
frequently incorporated into these systems owing to their role in response towards
ADRs, the other cell types have been used too [46, 47]. Model systems that have
been used to create this environment include large scale hepatic bioreactor [48, 49],
microfluidic technologies [50–52], layered-sheets of fibroblasts to initiate growth
conditions [53], and 3D hydrogel technologies for microarrays [54]. For the drug
discovery process, including testing for ADRs, 3D hydrogel microarrays are ideal
because of their high throughput nature. Hepatic bioreactors cannot be scaled down
at low cost, microfluidics can provide some replicates, but not with the same effi-
ciency as a microarray can, and fibroblast systems could potentially be more com-
plicated than the hydrogels, and many really do not adequately mimic the 3D liver
microenvironment. Therefore, it is more desirable to use a microarray system for
drug development testing.
8.5 Summary 171

Currently, several systems are in place to print co-cultures of 3D systems, reca-

pitulate the in vivo-like behavior of various liver cell lines on both 2D and 3D
­platforms, or at the very least co-culture various cell types in a hydrogel platform.
Kwon et al. have demonstrated the feasibility of using a micropillar chip as a plat-
form for culture of printed cells, and have modified the expression of various DMEs
via adenoviral transduction systems to examine individual toxicities [30]. Kang et
al. established the viability of primary rat hepatocytes (PRHs) co-cultured with
sinusoidal endothelial cells in a transwell membrane plate and established the main-
tenance in the level of expression of CYP450 activity. Bhise et al. printed co-cul-
tures of HepG2 and C3A into spheroids for the sake of bioreactor functions on a
liver-on-­a-chip platform, and retained transporter function. All of these results show
promise for the ability of microarray bioprinting towards the recapitulation of the
3D in vivo-like behavior.

8.4 Other Applications

The major goal of most moderate to large scale 3D bioprinting is to recapitulate in

vivo physiology in such a way that the system can be used as a model for various
tissues, or even be used to replace tissue. While the microarray system is not suit-
able for direct implantation, small scale mimics of various tissues can be used to
model regeneration and cell growth. Though the early part of this chapter has
detailed the printing of cells and compounds for applications in hepatic tissues and
neural tissues, microarray bioprinting is not limited to just these organs. Ma et al.
have printed cell-laden gelatin methacrylate (GelMA) hydrogels to test for regen-
eration of periodontal tissue in periodontitis [55]. The platform for printing was not
a micropillar/microwell chip as discussed in much of the book; rather, it was a flat
surface pre-treated with other chemicals to improve gel attachment [55]. Other tis-
sues have also been printed at microarray levels. However, the printing technique
can best be described as micropatterning for the development of larger 3D scaffolds.
In addition to cell and tissue engineering, scientists have printed hydrogels contain-
ing test compounds on micropatterned surfaces to test drug toxicity [56]. While
compound-cell interactions have been discussed earlier in this chapter, this feature
can be used to examine hydrogels as drug delivery systems on a microarray plat-
form, rather than just for cell and tissue engineering constructs.

8.5 Summary

Microarray bioprinting is a useful tool for miniaturized assessments of in vivo-like

behavior. It offers a small-scale, high-throughput approach in tissue engineering,
disease modeling, and drug development research. Assays for detecting cellular
functions and toxicity can be performed to minimize the use of otherwise expensive
172 8 Applications of Microarray Bioprinting

reagents. While macroscale 3D bioprinting has yet to effectively recapitulate whole

organ systems, the 3D printing of hydrogels at the microarray level still gives insight
into particular details of organ functions, and cells may still cluster and reorganize
themselves in ways which are comparable to in vivo structures. It is this miniaturiza-
tion that allows a high-throughput testing of conditions to optimize treatments for
disease targets, determine effective nontoxic drugs, and hopefully lead towards
greater discoveries in tissue engineering and tissue regeneration research.

References

1. Reuben, A., Koch, D. G., & Lee, W. M. (2010). Drug-induced acute liver failure: Results of a
U.S. multicenter, prospective study. Hepatology, 52(6), 2065–2076. doi:10.1002/hep.23937.
2. Morgan, S., Grootendorst, P., Lexchin, J., Cunningham, C., & Greyson, D. (2011). The cost of
drug development: A systematic review. Health Policy (Amsterdam, Netherlands), 100(1),
4–17. doi:10.1016/j.healthpol.2010.12.002.
3. Preventable adverse drug reactions: A focus on drug interactions. 2016. Retrieved from http://
www.fda.gov/Drugs/DevelopmentApprovalProces.
4. Elliott, N. T., & Yuan, F. A. N. (2011). A review of three-dimensional in vitro tissue models for
drug discovery and transport Studies. Journal of Pharmaceutical Sciences, 100(1), 59–74.
doi:10.1002/jps.22257.
5. Godoy, P., Hewitt, N. J., Albrecht, U., Andersen, M. E., Ansari, N., Bhattacharya, S., et al.
(2013). Recent advances in 2D and 3D in vitro systems using primary hepatocytes, alternative
hepatocyte sources and non-parenchymal liver cells and their use in investigating mechanisms
of hepatotoxicity, cell signaling and ADME. Archives of Toxicology, 87(8), 1315–1530.
doi:10.1007/s00204-013-1078-5.
6. Kimlin, L. C., Casagrande, G., & Virador, V. M. (2013). In vitro three-dimensional (3D)
models in cancer research: An update. Molecular Carcinogenesis, 52(3), 167–182. doi:10.1002/
mc.21844.
7. Keogh, J. P. (2012). Membrane transporters in drug development. Advances in Pharmacology,
63, 1–42. doi:10.1016/B978-0-12-398339-8.00001-X.
8. Lin, J., Schyschka, L., Mühl-Benninghaus, R., Neumann, J., Hao, L., Nussler, N., et al. (2012).
Comparative analysis of phase I and II enzyme activities in 5 hepatic cell lines identifies Huh-7
and HCC-T cells with the highest potential to study drug metabolism. Archives of Toxicology,
86(1), 87–95. doi:10.1007/s00204-011-0733-y.
9. Walsky, R. L., & Obach, R. S. (2004). Validated assays for human cytochrome P450 activities.
Drug Metabolism and Disposition, 32(6), 647–660. doi:10.1124/dmd.32.6.647.
10. Zarowna-Dabrowska, A., McKenna, E. O., Schutte, M. E., Glidle, A., Chen, L., Cuestas-­Ayllon,
C., et al. (2012). Generation of primary hepatocyte microarrays by piezoelectric printing.
Colloids and Surfaces B: Biointerfaces, 89(1), 126–132. doi:10.1016/j.colsurfb.2011.09.016.
11. Meli, L., Jordan, E. T., Clark, D. S., Linhardt, R. J., & Dordick, J. S. (2012). Influence of a
three-dimensional, microarray environment on human cell culture in drug screening systems.
Biomaterials, 33(35), 9087–9096. doi:10.1016/j.biomaterials.2012.08.065.
12. Leite, S. B., Iwona, W.-Z., Zaldivar, J. M., Airola, E., Reis-Fernandes, M. A., Mennecozzi, M.,
et al. (2012). 3D HepaRG model as an attractive tool for toxicity testing. Toxicological
Sciences, 130(1), 106–116.
13. Leite, S. B., Teixeira, A. P., Miranda, J. P., Tostões, R. M., Clemente, J. J., Sousa, M. F., et al.
(2011). Merging bioreactor technology with 3D hepatocyte-fibroblast culturing approaches:
Improved in vitro models for toxicological applications. Toxicology in Vitro, 25(4), 825–832.
doi:10.1016/j.tiv.2011.02.002.
References 173

14. Chang, J. H., Plise, E., Cheong, J., Ho, Q., & Lin, M. (2013). Evaluating the in vitro inhibition
of UGT1A1, OATP1B1, OATP1B3, MRP2, and BSEP in predicting drug-induced hyperbiliru-
binemia. Molecular Pharmaceutics, 10(8), 3067–3075.
15. Liang, Q., Sheng, Y., Jiang, P., Ji, L., Xia, Y., Min, Y., et al. (2011). The gender-dependent
difference of liver GSH antioxidant system in mice and its influence on isoline-induced liver
injury. Toxicology, 280(1–2), 61–69. doi:10.1016/j.tox.2010.11.010.
16. Kis, E., Ioja, E., Rajnai, Z., Jani, M., Méhn, D., Herédi-Szabó, K., et al. (2012). BSEP inhibi-
tion: In vitro screens to assess cholestatic potential of drugs. Toxicology In Vitro, 26(8), 1294–
1299. doi:10.1016/j.tiv.2011.11.002.
17. Lee, J. K., Marion, T. L., Abe, K., Lim, C., Pollock, G. M., & Brouwer, K. L. R. (2010).
Hepatobiliary disposition of troglitazone and metabolites in rat and human sandwich-cultured
hepatocytes: Use of Monte Carlo simulations to assess the impact of changes in biliary excre-
tion on troglitazone sulfate accumulation. The Journal of Pharmacology and Experimental
Therapeutics, 332(1), 26–34. doi:10.1124/jpet.109.156653.tion.
18. Cascorbi, I., & Haenisch, S. (2010). Pharmacogenetics of ATP-binding cassette transporters
and clinical implications. Methods in Molecular Biology, 596, 95–121.
doi:10.1007/978-1-60761-416-6.
19. Fromm, M. F. (2004). Importance of P-glycoprotein at blood-tissue barriers. Trends in
Pharmacological Sciences, 25(8), 423–429. doi:10.1016/j.tips.2004.06.002.
20. Stieger, B. (2011). The role of the Sodium-Taurocholate Cotransporting Polypeptide (NTCP)
and of the Bile Salt Export Pump (BSEP) in physiology and pathophysiology of bile forma-
tion. In M. F. Fromm & R. B. Kim (Eds.), Drug transporters. Berlin: Springer.
doi:10.1007/978-3-642-14541-4.
21. Goral, V. N., Hsieh, Y.-C., Petzold, O. N., Clark, J. S., Yuen, P. K., & Faris, R. A. (2010).
Perfusion-based microfluidic device for three-dimensional dynamic primary human hepato-
cyte cell culture in the absence of biological or synthetic matrices or coagulants. Lab on a
Chip, 10(24), 3380–3386. doi:10.1039/c0lc00135j.
22. Thompson, R. A., Isin, E. M., Y, L., Weidolf, L., Page, K., Wilson, I., et al. (2012). In vitro
approach to assess the potential for risk of idiosyncratic adverse reactions caused by candidate
drugs. Chemical Research in Toxicology, 25(8), 1616–1632. doi:10.1021/tx300091x.
23. Xu, J. J., Diaz, D., & O’Brien, P. J. (2004). Applications of cytotoxicity assays and pre-lethal
mechanistic assays for assessment of human hepatotoxicity potential. Chemico-Biological
Interactions, 150(1), 115–128. doi:10.1016/j.cbi.2004.09.011.
24. Porceddu, M., Buron, N., Roussel, C., Labbe, G., Fromenty, B., & Borgne-Sanchez, A. (2012).
Prediction of liver injury induced by chemicals in human with a multiparametric assay on
isolated mouse liver mitochondria. Toxicologic Pathology, 129(2), 332–345.
25. Hynes, J., Swiss, R. L., & Will, Y. (2012). High-throughput analysis of mitochondrial oxygen
consumption. In C. M. Palmeira & A. J. Moreno (Eds.), Mitochondrial bioenergetics: Methods
and protocols (Vol. 810, pp. 103–117). New York: Humana Press.
doi:10.1007/978-1-61779-382-0.
26. Aleo, M. D., Luo, Y., Swiss, R., Bonin, P. D., Potter, D. M., & Will, Y. (2014). Human drug-­
induced liver injury severity is highly associated with dual inhibition of liver mitochondrial
function and bile salt export pump. Hepatology, 60(3), 1015–1022. doi:10.1002/hep.27206.
27. Ong, M. M. K., Latchoumycandane, C., & Boelsterli, U. A. (2007). Troglitazone-induced
hepatic necrosis in an animal model of silent genetic mitochondrial abnormalities. Toxicological
Sciences, 97(1), 205–213. doi:10.1093/toxsci/kfl180.
28. Fernandes, T. G., Kwon, S. J., Bale, S. S., Lee, M. Y., Diogo, M. M., Clark, D. S., et al. (2010).
Three-dimensional cell culture microarray for high-throughput studies of stem cell fate.
Biotechnology and Bioengineering, 106(1), 106–118. doi:10.1002/bit.22661.
29. Fernandes, T. G., Kwon, S. J., Lee, M. Y., Clark, D. S., Cabral, J. M. S., & Dordick, J. S.
(2008). On-chip, cell-based microarray immunofluorescence assay for high-throughput analy-
sis of target proteins. Analytical Chemistry, 80(17), 6633–6639. doi:10.1021/ac800848j.
30. Kwon, S. J., Lee, D. W., Shah, D. A., Ku, B., Jeon, S. Y., Solanki, K., et al. (2014). High-­
throughput and combinatorial gene expression on a chip for metabolism-induced toxicology
screening. Nature Communications, 5, 3739. doi:10.1038/ncomms4739.
174 8 Applications of Microarray Bioprinting

31. Ulasov, I., Nandi, S., Dey, M., Sonabend, A. M., & Lesniak, M. S. (2011). Inhibition of Sonic
hedgehog and Notch pathways enhances sensitivity of CD133+ glioma stem cells to temozolo-
mide therapy. Molecular Medicine, 17(1–2), 103–112. doi:10.2119/molmed.2010.00062.
32. Tegenge, M. A., Rockel, T. D., Fritsche, E., & Bicker, G. (2011). Nitric oxide stimulates
human neural progenitor cell migration via cGMP-mediated signal transduction. Cellular and
Molecular Life Sciences, 68(12), 2089–2099. doi:10.1007/s00018-010-0554-9.
33. Blurton-jones, M., Spencer, B., Michael, S., Castello, N. A., Agazaryan, A. A., Davis, J. L.,
et al. (2014). Neural stem cells genetically-modified to express neprilysin reduce pathology in
Alzheimer transgenic models Neural stem cells genetically-modified to express neprilysin
reduce pathology in Alzheimer transgenic models. Stem Cell Research & Therapy, 5(2), 46.
34. Farrel, K., Joshi, P., Roth, A., Kothapalli, C. R., & Lee, M.-Y. (2016). High-throughput screen-
ing of toxic chemicals against neural stem cells. In J. L. Sherley (Ed.), Human stem cell toxi-
cology (pp. 31–63). Cambridge: The Royal Society of Chemistry.
35. Ferguson, C. S., & Tyndale, R. F. (2011). Cytochrome P450 enzymes in the brain: Emerging
evidence of biological significance. Trends in Pharmacological Sciences, 32(12), 708–714.
doi:10.1016/j.tips.2011.08.005.
36. Miksys, S., & Tyndale, R. F. (2013). Cytochrome P450-mediated drug metabolism in the
brain. Journal of Psychiatry & Neuroscience, 38(3), 152–163. doi:10.1503/jpn.120133.
37. Murphy, S. V., & Atala, A. (2014). 3D bioprinting of tissues and organs. Nature Biotechnology,
32(8), 773–785. doi:10.1038/nbt.2958.
38. Arias, I., Wolkoff, A., Boyer, J., Shafritz, D., Fausto, N., Alter, H., & Cohen, D. (2011).
The liver: Biology and pathobiology. Chichester: Wiley.
39. Boron, W. F., & Boulpaep, E. L. (2009). Medical physiology (2nd ed.). Philadelphia, PA:
Saunders Elsevier.
40. Jaeschke, H. (2011). Reactive oxygen and mechanisms of inflammatory liver injury: Present
concepts. Journal of Gastroenterology and Hepatology, 26(Suppl 1), 173–179.
doi:10.1111/j.1440-1746.2010.06592.x.
41. Zimmermann, H. W., Trautwein, C., & Tacke, F. (2012). Functional role of monocytes and
macrophages for the inflammatory response in acute liver injury. Frontiers in Physiology, 3,
1–18. doi:10.3389/fphys.2012.00056.
42. Kim, Y., & Rajagopalan, P. (2010). 3D hepatic cultures simultaneously maintain primary hepa-
tocyte and liver sinusoidal endothelial cell phenotypes. PloS One, 5(11), 1–10. doi:10.1371/
journal.pone.0015456.
43. Sato, Y., Tsukada, K., & Hatakeyama, K. (1999). Role of shear stress and immune responses
in liver regeneration after a partial hepatectomy. Surgery Today, 29(1), 1–9.
44. Tuleuova, N., Lee, J. Y., Lee, J., Ramanculov, E., Zern, M. A., & Revzin, A. (2010). Using
growth factor arrays and micropatterned co-cultures to induce hepatic differentiation of
embryonic stem cells. Biomaterials, 31(35), 9221–9231. doi:10.1016/j.
biomaterials.2010.08.050.
45. Ueno, T., Sata, M., Sakata, R., Torimura, T., Sakamoto, M., Sugawara, H., et al. (1997).
Hepatic stellate cells and intralobular innervation in human liver cirrhosis. Human Pathology,
28(8), 953–959.
46. Messner, S., Agarkova, I., Moritz, W., & Kelm, J. M. (2013). Multi-cell type human liver
microtissues for hepatotoxicity testing. Archives of Toxicology, 87(1), 209–213. doi:10.1007/
s00204-012-0968-2.
47. Kostadinova, R., Boess, F., Applegate, D., Suter, L., Weiser, T., Singer, T., et al. (2013). A
long-term three dimensional liver co-culture system for improved prediction of clinically rel-
evant drug-induced hepatotoxicity. Toxicology and Applied Pharmacology, 268(1), 1–16.
doi:10.1016/j.taap.2013.01.012.
48. Fiegel, H. C., Kneser, U., Kluth, D., & Rolle, U. (2010). Hepatic tissue engineering.
Handchirurgie, Mikrochirurgie, Plast Chir Organ der Deutschsprachigen Arbeitsgemeinschaft
für Handchirurgie Organ der Deutschsprachigen Arbeitsgemeinschaft für Mikrochirurgie der
Peripher Nerven und Gefässe Organ der Vereinigung der Deut, 42(6), 337–41. doi:10.105
5/s-0030-1252045.
References 175

49. Hoekstra, R., Nibourg, G. A., van der Hoeven, T. V., Plomer, G., Seppen, J., Ackermans, M. T.,
et al. (2013). Phase 1 and phase 2 drug metabolism and bile acid production of HepaRG cells
in a bioartificial liver in absence of dimethyl sulfoxide. Drug Metabolism and Disposition,
41(3), 562–567. doi:10.1124/dmd.112.049098.
50. Bhatia, S. N., & Ingber, D. E. (2014). Microfluidic organs-on-chips. Nature Biotechnology,
32(8), 760–772. doi:10.1038/nbt.2989.
51. Toh, Y.-C., Lim, T. C., Tai, D., Xiao, G., van Noort, D., & Yu, H. (2009). A microfluidic 3D
hepatocyte chip for drug toxicity testing. Lab on a Chip, 9(14), 2026–2035. doi:10.1039/
b900912d.
52. Bale, S. S., Vernetti, L., Senutovitch, N., Jindal, R., Hegde, M., Gough, A., et al. (2014). In
vitro platforms for evaluating liver toxicity. Experimental Biology and Medicine (Maywood,
N.J.), 239(9), 1180–1191. doi:10.1177/1535370214531872.
53. Nakazawa, K., & Shinmura, Y. (2011). Effects of culture conditions on a micropatterned co-­
culture of rat hepatocytes with 3T3 cells. Journal of Bioprocessing & Biotechniques, 01(3).
doi:10.4172/2155-9821.S3-002.
54. Li, C. Y., Stevens, K. R., Schwartz, R. E., Alejandro, B. S., Huang, J. H., & Bhatia, S. N.
(2014). Micropatterned cell-cell interactions enable functional encapsulation of primary hepa-
tocytes in hydrogel microtissues. Tissue Engineering. Part A, 20(617), 2200–2212. doi:10.1089/
ten.TEA.2013.0667.
55. Ma, Y., Ji, Y., Huang, G., Ling, K., Zhang, X., & Bioprinting, X. F. (2015). 3D cell-laden
hydrogel microarray for screening human periodontal ligament stem cell response to extracel-
lular matrix. Biofabrication, 7(4), 044105. doi:10.1088/1758-5090/7/4/044105.
56. Bailey, S. N., Sabatini, D. M., & Stockwell, B. R. (2004). Microarrays of small molecules
embedded in biodegradable polymers for use in mammalian cell-based screens. Proceedings
of the National Academy of Sciences of the United States of America, 101(46), 16144–16149.