Liquid Biopsy Beyond Cancer A miRNA Detection in

RESEARCH ARTICLE
www.advnanobiomedres.com
Liquid Biopsy beyond Cancer: A miRNA Detection in

Serum with Electrochemical Chip for Non-Invasive Coeliac
Disease Diagnosis
Veronica Caratelli, Maria Moccia,* Francesca R. Paggioro, Luca Fiore,
Concetta Avitabile, Michele Saviano, Anna Lisa Imbriani, Principia Dardano, Luca De
Stefano, Danila Moscone, Nicola A. Colabufo, Imane Ghafir El Idrissi, Francesco Russo,
Giuseppe Riezzo, Gianluigi Giannelli, and Fabiana Arduini*
1. Introduction
Coeliac disease is a very common autoimmune disease estimated to affect 1 in
100 people worldwide. It occurs in genetically predisposed people where the Coeliac disease (CD) is one of the most
ingestion of gluten leads to damage in the small intestine, and it is accurately common autoimmune disorders affecting
many people in the Western world. The
diagnosticated through duodenal biopsy, an invasive diagnostic method. The liquid
onset of this enteropathy is triggered by
biopsy, generally used for monitoring cancer, is an appealing alternative even for the common protein fraction of wheat
autoimmune pathology such as coeliac disease, allowing for detecting disease flour, namely gluten, in people with a
progression or resistance to treatment. For this reason, an electrochemical peptide genetic predisposition carrying human leu-
nucleic acid (PNA) device combined with a smartphone-assisted potentiostat for kocyte antigen DQ2 and DQ8 haplotypes.[1]
non-invasive coeliac disease diagnosis is proposed, by measuring the selected The symptoms encompass abdominal
pain, diarrhea, and weight loss[2,3] due to
overexpressed miRNA-486-5p in serum, enlarging the application of liquid biopsy the flattened small intestinal mucosa with
in nontumor pathologies. For highly sensitive detection, the polyester-based a lymphocytic infiltrate, crypt hyperplasia,
printed sensor is nanomodified with gold nanoparticles and a synthetic customized and villous atrophy.[4] The average occur-
PNA probe. The designed sensor can detect the target analyte in the range of rence of this disease is reported in a popu-
10–100 nM with a limit of detection of 0.7 nM by measuring the variation of the lation of between 0.5% and 1%
worldwide,[5,6] with a prevalence of CD in
response of the electrochemical mediator hexaammineruthenium in the presence
patients affected by other conditions,
of PNA–miRNA duplex on the nanostructured working electrode surface. The including insulin-dependent diabetes mel-
analyses testing serum samples are found in agreement with ones obtained by real- litus type 1, autoimmune thyroid disease,
time quantitative polymerase chain reaction (RT-qPCR), demonstrating the reli- inflammatory bowel disease, and irritable
ability of this innovative electrochemical chip developed. bowel syndrome (IBS).[7,8] The CD diagno-
sis starts with the search of immunoglobu-
lin A antibodies, tissue transglutaminase 2,
V. Caratelli, F. R. Paggioro, L. Fiore, D. Moscone, F. Arduini A. L. Imbriani

Department of Chemical Science and Tecnologies Biochemical Systems International S.p.A.
University of Rome “Tor Vergata” Loc, Palazzo del Pero, 23, 52100 Arezzo, Italy
Via della Ricerca Scientifica, 00133 Rome, Italy
E-mail: [email protected] P. Dardano, L. De Stefano
Department of Physical Sciences and Matter Technology
M. Moccia, C. Avitabile, M. Saviano Institute for Applied Science and Intelligent Systems
Institute of Crystallography National Research Council
National Research Council (CNR) Via Pietro Castellino 111, 80131 Napoli, Italy
Via G. Amendola 122/O, 70126 Bari, Italy
E-mail: [email protected] N. A. Colabufo, I. Ghafir El Idrissi
Department of Pharmacy-Pharmaceutical Science
The ORCID identification number(s) for the author(s) of this article University of Bari Aldo Moro
can be found under https://doi.org/10.1002/anbr.202200015. Via Orabona 4, 70125 Bari, Italy
© 2022 The Authors. Advanced NanoBiomed Research published by N. A. Colabufo, I. Ghafir El Idrissi
Wiley-VCH GmbH. This is an open access article under the terms of Biofordrug S.R.L
the Creative Commons Attribution License, which permits use, Spin-off of the University of Bari Aldo Moro
distribution and reproduction in any medium, provided the original Via Dante 99, Triggiano, 70019 Bari, Italy
work is properly cited.
DOI: 10.1002/anbr.202200015
Adv. NanoBiomed Res. 2022, 2, 2200015 2200015 (1 of 14) © 2022 The Authors. Advanced NanoBiomed Research published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advnanobiomedres.com
and the endomysial IgA antibody in the blood, boosting the iden- In the overall scenario of smart analytical methods for miRNA
tification of patients who undergo an endoscopic biopsy. Despite detection, the biosensors have demonstrated several advantages
the high sensitivity of the biochemical analyses, no currently including the cost-effectiveness of the analyses, the miniaturiza-
available analysis in serum provides enough sensitivity and tion of the device, and the capability to work in complex
specificity; therefore, a duodenal biopsy is required.[9] Indeed, matrix.[20,21] Among the different transducers, the electrochemi-
the current recommendations established the duodenal biopsy cal one has revealed its potential to address the challenges related
as the gold-standard method for the accurate diagnosis of this to the detection of circulating nucleic acids such as the low detec-
enteropathy,[10] though it is an invasive diagnostic method. tion limit in clinical samples.[22] For example, Tavallaie et al.
In the case of cancer, even if the tumor tissue remains the proposed a novel approach based on an electric-field-induced
gold-standard source for clinical molecular analyses, the monitor- reconfiguration of labeled DNA-modified gold-coated magnetic
ing of cancer-derived material circulating in the bloodstream, nanoparticles to detect miRNA-21 expression level directly in
called liquid biopsy, has become an appealing alternative. The whole blood, with high sensitivity within the concentration range
high potential of this diagnostic technique has also been recog- of 10 aM–1 nM.[23] Yuan’s group developed a gallium zinc oxide
nized at a worldwide level, being listed in 2017 among the field-effect transistor (IGZO FET) for the detection of miRNA-21
Top 10 Emerging Technologies.[11] The advantage of a liquid in human urine as a trace bladder cancer-associated biomarker.
biopsy is not only related to the lack of invasiveness, but it also Considering the uniform and smooth structure of IGZO with
allows to detect the disease at an early stage. Furthermore, the excellent electrical performance, the single-stranded DNA-
tissue biopsies examine only selected bits of tumors and can miss functionalized IGZO FET exhibited high sensitivity with an
cells that have turned more dangerous than their neighboring ultralow detection limit of 19.8 aM.[24]
cells have. The overall potential of liquid biopsy is its capability Umer et al. designed a novel amplification-free sandwich-type plat-
to detect the full spectrum of mutations, indicating when a more form assay for electrochemical detection of miRNA-DNA chimera by
aggressive treatment is needed, furnishing a fast and easy test for using HRP-labeled p53 able to bind the recognition sequence. The
diagnosis and prognosis of cancer. The application of a liquid amperometric signal was monitored in presence of hydroquinone
biopsy for other autoimmune pathologies is also emerging as and H2O2, obtaining a limit of detection equal to 22 fM.[25]
a potential diagnostic tool for CD. Therefore, considering that Shiddiky’s research group developed different amplification-
the tissue biopsy is the gold-standard method for both cancer free electrochemical approaches for the detection of miRNA
and CD, herein, we investigated the expression of selected using screen-printed gold electrodes. The first device relies on
miRNA in the serum of CD patients. Currently, in the literature, gold-loaded nanoporous superparamagnetic iron oxide nano-
only a few articles reported miRNAs expression in patients with cubes for the detection of miRNA-107 (Au–NPFe2O3NC). The
CD, mainly evaluated in the intestinal mucosa.[12–14] MicroRNAs target miRNA was directly adsorbed onto the gold surfaces of
(miRNAs) have been found circulating in different body fluids, Au–NPFe2O3NC via gold-RNA affinity interaction. The electroca-
such as serum and plasma, protected by a protein binding or talytic activity of Au–NPFe2O3NC was then used for the reduc-
enclosed in vesicles and released in the extracellular space.[15] tion of ruthenium hexaammine(III) chloride bound with target
miRNA. The catalytic signal was further amplified by using the
Circulating miRNAs represent reliable and promising diagnostic,
ferri/ferrocyanide electroactive probe. These multiple signal
prognostic, and therapeutic biomarkers, because miRNAs mod-
enhancement steps enable this device to achieve the detection
ulate >30% of mammalian genes and participate in almost all
limit of 100 aM.[26] The second sensor was developed for the
known cellular processes, including development, DNA repair,
detection of FAM134B mRNA which was selectively isolated
cell cycle regulation, differentiation, apoptosis, as well as malig-
by magnetic separation and adsorbed directly onto an unmodi-
nant cellular transformation and host–pathogen interaction.[16,17]
fied gold-screen-printed electrode (gold-SPE). The surface-
To this regard, Bascunan et al. reported the first description of an
attached miRNA was measured by differential pulse voltammetry
altered pattern of inflammation-related miRNA expression in
in the presence of [Fe(CN)6]4/3 redox system. This method
blood immune cells and plasma in CD adult patients. They eval-
demonstrated good sensitivity, with a limit of detection equal
uated the expression levels of miRNA-146a, miRNA-155, and
to 1.0 pM and reproducibility with relative standard deviation
miRNA-21 in peripheral blood mononuclear cells, miRNA-155
(RSD%) ¼ <5%, for n ¼ 3.[27] Following the improvement of
in monocytes, and miRNA-155, miRNA-21, and miRNA-125b
biosensors using nanomaterials, the new frontiers of e-devices
in plasma of CD patients using TaqMan microRNA Reverse
have been oriented toward miniaturization and portability since
Transcription Kit, suggesting their participation in the immune
the healthcare trend is changing toward becoming more patient
processes underlying this pathology.[18] However, despite the
oriented. To this aim, the development of electrochemical devi-
huge potential for miRNAs in the diagnosis and prognosis of can- ces by combining screen-printed electrodes and smartphone
cer, the current detection approaches are generally confined to technology has allowed obtaining more suitable instruments
classical nucleic acid detection methods such as northern blot- to be applied in healthcare diagnostics, acting as a portable
ting, microarray, quantitative real-time quantitative polymerase and convenient platform for point-of-care testing (POCT).
chain reaction ((q)RT-PCR), and next-generation sequencing.[19] In this regard, Low et al. developed a smartphone-based
F. Russo, G. Riezzo, G. Giannelli F. Arduini

National Institute of Gastroenterology “S. de Bellis” SENSE4MED S.R.L.
Research Hospital Via della ricerca scientifica, 00133 Rome, Italy
Castellana Grotte, 70013 Bari, Italy
biosensing system composed of a disposable screen-printed bio- AGGACCUGCGGGACAAGAUUCUU 3’), miRNA-21-5p (5’
sensor modified with reduced graphene oxide/gold composite, to UAGCUUAUCAGACUGAUGUUGA 3’), miRNA-486-5p (5’
detect salivary circulating miRNA-21. The hybridization between UCCUGUACUGAGCUGCCCCGAG 3’), selecting the miRNA
the miRNA-21 targets and ssDNA probe on the rGO/Au- useful for the diagnosis of CD. Subsequently, the selected
modified electrode resulted in the decrease in peak current with miRNA-486-5p was employed to develop the electrochemical
the increase in miRNA-21 target concentration. The smartphone- chip using a complementary PNA sequence as a molecule probe
based biosensing system showed comparable performance with to realize an innovative and sustainable sensing tool for non-
a commercial electrochemical workstation for miRNA-21 detec- invasive CD diagnosis (Figure 1).
tion in the concentration range of 1 104–1 1012 M.[28]
Herein, we developed a smartphone-assisted electrochemical
chip based on label-free detection for non-invasive CD
2. Results and Discussion
diagnosis, by exploiting the interaction of the positive charge
of the electrochemical probe hexaammineruthenium with the 2.1. Patients Cohort Analysis
well-known different charges of peptide nucleic acid (PNA)
and miRNA.[29,30] First, we investigated the following four Cases for the present study were recruited from among
miRNAs involved in the onset of CD: miRNA-449a (5’ the outpatients of the National Institute of Gastroenterology
UGGCAGUGUAUUGUUAGCUGGU 3’), miRNA-492 (5’ “S. de Bellis” Research Hospital, Castellana G. (Italy). CD
Figure 1. Electrochemical detection of miRNA-486-5p using the smartphone-assisted peptide nucleic acid (PNA)-based printed electrochemical biosen-
sor. Picture of the laboratory setup consisting of a screen-printed electrode a) connected to the miniaturized potentiostat Sensit Smart connected to a
smartphone. b) Measurement of the blank signal. c) Measurement of miRNA-486.
patients were diagnosed based on positive antiendomysium anti- antibody (EMA IgA) and IgA-transglutaminase antibody (tTG
bodies (EMAs) and antitransglutaminase antibodies (TGAs). IgA) were not detectable in IBS patients and control subjects
Biopsies from the proximal part and the distal part of the duo- (Table 1).
denum were taken to confirm the diagnosis according to the For the 12 CD patients, clinical serum data were collected
modified Marsh–Oberhuber criteria (grades 3b–3c).[31] After before and after one year of GFD, and the results are reported
the diagnosis, CD patients followed a specific gluten-free diet in Figure 2. Once CD patients started the GFD, both EMA
(GFD) and received additional counseling at follow-up. IgA and t-Tg IgA dramatically decreased compared to initial val-
Patients with CD were evaluated at the time of diagnosis and ues, thus indicating an optimal response to gluten restriction.[34]
at least 12 months after the start of GFD. The dietary compliance The collection of human serum samples has been carried out
assessment at the 1-year checkup was based on an examination by a medical doctor from a colleague, following ethical guidelines
consisting of a physical exam, the patient’s self-reported compli- provided by occupational health section, University of Oxford.[35]
ance, and the doctor’s opinion. We supplied the Donor Information Sheet and Donor Consent
As for the IBS patient group, the IBS patients with diarrhea form to the donor.
were enrolled according to the description of Longstreth et al.[32]
Only the HLA-DQ2/HLA-DQ8 negative/negative IBS patients 2.2. Analysis of miRNA Expression
were considered for this study, because gluten-sensitive diarrhea
without CD is a clinical entity that has been seen in IBS patients All 40 samples were analyzed, and each subject was tested for
positive for HLA-DQ2 or HLA-DQ8.[33] each of the miRNAs considered: miRNA-39-3p, miRNA-16-5p,
Healthy control (HC) subjects were enrolled among the miRNA-449a, miRNA-492, miRNA- 21-5p, and miRNA-486-5p.
administrative staff of “S. de Bellis” Research Hospital. In addition, samples from CD patients under GFD were also
They reported not having metabolic, endocrine, or immunologi- investigated to detect if any variation in miRNA expression pro-
cal diseases, dyspepsia, or other gastrointestinal (GI) diseases file occurs before and after the elimination of gluten from the
and were not taking any medication. Information on the health diet. Furthermore, the analysis of samples from IBS patients
status of subjects was obtained during an examination consisting can permit the evaluation of differences in terms of miRNAs
of a physical exam and an interview on their current lifestyle, expression in the two diseases. For miRNA-39-3p, a relatively
diet, and medical history. As criteria for admission, EMA and constant expression was shown in all four groups of subjects ana-
TGA had to be negative. Moreover, metabolic parameters (blood lyzed (CD patients, CD patients after 1 year of GFD, IBS patients,
glucose, HbA1c, lipid profile, body weight, and blood pressure) and healthy cases) with an average Ct value of about 27. This data
had to be in the normal range of values. This research was con- confirms that the extraction of miRNAs was efficient and con-
ducted according to the guidelines stipulated in the Declaration stant in the various samples.
of Helsinki and all procedures involving human subjects and miRNA-449a and miRNA-492 were not detected in any of the
patients had been approved by the local Science and Ethics samples analyzed; therefore, from evaluating the first data
Committee. Written informed consent was obtained from all obtained, we can assume that they are not present at detectable
the patients and healthy subjects. This clinical trial had been reg- concentrations in serum.
istered at http://www.clinicaltrials.gov (registration number miRNA-21-5p and miRNA-486-5p were detected in all the
NCT01574209). Overall, the expression patterns of miRNAs were samples analyzed: HCs, CD patients at baseline, CD patients
tested in the serum of a cohort of 12 patients with CD (1 man and under GFD, and patients with IBS. The expression of miRNA-
11 women, mean age 32 11) before and after 1 year of a GFD, 21-5p and miRNA-486-5p are represented both as a function
14 IBS patients (1 man and 13 women, mean ages 39 15), and of Ct and 2ΔΔCT ratio in Figure 3.
14 healthy volunteers (HC) (4 men and 10 women, mean ages As represented in Figure 3a, miRNA-21-5p exhibits a rather
38 14) (number of samples ¼ 52). The quantitative analysis late Ct in all 4 biological groups analyzed (CTRL 32.6 0.1;
of the miRNAs through the RT-PCR technique is described in CD 32 0.04; GFD 32.5 0.2; IBS 33.5 0.2). In CD patients,
the Experimental Section. the Ct value is lower than in the other groups; therefore, the
All the 40 recruited subjects underwent a blood sampling to expression is slightly higher than in HCs, GF diet patients,
evaluate their blood chemistry parameters. All serum data were and IBS patients. In particular, as shown in Figure 3b,
within the normal range. Worthy of note, IgA-endomysial miRNA-21-5p is 1.6 0.1 fold over-expressed in coeliac patients
Table 1. Clinical serum data of recruited cohort.
Cohort Glua) Tagb) Ferrc) RBCd) Leue) Hbf ) Hctg) Plth) Asti) Altj)
CD 87 2.8 95 8 27 8 4.5 0.14 6.6 0.6 13 0.4 39 1 252 21 19 5 25 8
IBS 87 2.6 91 10 48 12 4.6 0.16 9.9 3.5 14 0.2 41 0.8 245 14 15 1.6 23 3
Controls 84 1.6 118 19 53 16 4.7 0.13 5.8 0.3 14 0.3 41 1 257 25 19 2 27 2.7
Glu ¼ glucose (mg dL1); b)Tag ¼ triacylglycerol (mg dL1); c)Ferr ¼ ferritin (μg L1); d)RBC ¼ red blood cells (103 μL1); e)Leu ¼ leucocytes ( 103 μL1);
a)
Hb ¼ hemoglobin (g dL1); g)Hct ¼ hematocrit (%); h)Plt ¼ number of platelet (103 μL1); i)Ast ¼ aspartate aminotransferase (IU L1); j)Alt ¼ alanine
f)
aminotransferase (IU L1).
Figure 2. Clinical serum results of 12 coeliac patients before and after gluten-free diet.
Figure 3. miRNA-21-5p and miRNA-486-5p expression in four cohorts: control, coeliac disease patients, patients under gluten-free diet, and patients with
irritable bowel syndrome. a) Cycle threshold of miRNA-21-5p in all recruited subjects subdivided into four groups. b) Over-expression of miRNA-21-5p in
all cohorts with respect to controls. c) Cycle threshold of miRNA-486-5p in all recruited subjects subdivided into four groups. d) Over-expression of
miRNA-486-5p in all cohorts with respect to controls. For miRNA analysis in a real-time polymerase chain reaction (PCR), each sample of serum was
tested in triplicate (n ¼ 3) and data expressed as mean standard deviation (SD). Statistical significance using one-way analysis of variance (ANOVA)
with P < 0.05 in comparison with control group and *P < 0.1, **P < 0.01, ***P < 0.001 pathological group, with unpaired t-test.
compared with controls. Even in CD patients after 1 year of GFD, Figure 4 reports the NCM phase and topography of the work-
miRNA-21-5p is over-expressed about 1.1 0.1 fold with respect ing electrode surface during the fabrication process as well as the
to controls, while IBS patients revealed an under-expression of detection step (Figure 4A–L), including a 3D reconstruction of
this miRNA compared to the healthy population with a the working electrode after the gold nanoparticles (AuNPs) depo-
2ΔΔCT ratio of 0.5 0.2. sition and after the last modification step by PNA–miRNA duplex
miRNA-486-5p presented earlier Ct (CTRL 30.5 0.2; CD followed by the measurement with Ru(NH3)63þ (Figure 4M,N).
28 0.6; GFD 28.5 0.8; IBS 29.5 0.3) in the analyzed cohorts The working electrode after AuNPs deposition had an effective
and in particular in the coeliac group. In CD patients, miRNA- surface of about 12 μm2 (increased with respect to the geometri-
486-5p was over-expressed compared to HCs, CD patients after cal scanned surface equal to 9 μm2) and it appeared in the topog-
GFD, and IBS patients, too. In particular, as shown in Figure 3d, raphy image to be fully covered by AuNPs, the distribution of
in CD patients, miRNA-486-5p is 5 1 fold over-expressed with which was not homogeneous at the micrometric level, indeed
respect to controls, in CD patients after GFD, the expression is AuNPs were aggregated in macro-agglomerates (Figure 4A,B).
greater 3.9 0.3 fold compared to HC, and in IBS patients, the After the PNA immobilization (Figure 4C,D), the AuNP aggre-
expression of miRNA-486-5p is 1.7 0.9 fold higher compared to gates were still clearly visible and completely covered the elec-
healthy subjects. trode surface. The NCM phase contrast image did not match
In conclusion, the data obtained suggest that miRNA-21-5p the topography contrast, suggesting an inhomogeneous coverage
and miRNA-486-5p are over-expressed in CD patients compared of the AuNPs by PNA. Moreover, a strong increment (from about
to healthy subjects, and the over-expression is also higher than 12–18 μm2) of the effective surface of the modified electrode was
that in CD patients after GFD and IBS patients. Furthermore, registered, due to a further ultra-structuration of the exposed
miRNA-486-5p exhibits a higher level of expression in blood than AuNPs surfaces.
miRNA-21-5p; therefore, it may be considered a more suitable After 6-mercapto-1-hexanol (MCH) incubation (Figure 4E,F),
marker for CD, helpful in developing a panel to introduce in the effective surface was about the same as the previous fabrica-
the clinical practice. tion step (17 μm2) and a similar NCM phase distribution was still
Noteworthy, our data suggest that following the expression present, even if the topography of the macro-aggregates was
level of both miRNAs, it seems that there is a return to levels slightly changed. After miRNA incubation, a further increment
similar to those of HC. in the NCM phase contrast and effective surface (about 22 μm2)
was registered (Figure 4G,H).
Finally, after the Ru(NH3)63þ incubation, the NCM phase
2.3. Electrochemical and Morphological Platform image and topography were very similar to the sample after
Characterization the first AuNPs deposition (Figure 4I,L). Moreover, the decreas-
ing of the effective surface suggested a full coverage of the
After demonstrating the suitability of miRNA-486-5p as a suitable AuNPs by the PNA–miRNA duplex and Ru(NH3)63þ interaction.
biomarker for CD, we developed the electrochemical PNA-based Successively, EIS characterization was carried out and typical
chip capable of detecting the specific miRNA sequence in serum Nyquist plots were obtained before and after drop-casting 8 μL of
samples. In detail, the sensor is composed of a graphite-based AuNPs onto the graphite working electrode (Figure 5a). AuNPs-
working electrode and counter electrode combined with a silver/ modified SPE (purple plot) showed resistance to charge transfer
silver chloride pseudo-reference electrode. The miRNA-486-5p (Rct) equal to 514 21 kΩ, which is lower than the one obtained
detection was carried out using a signal on a method based on using bare electrode (black plot), that is, 4325 13 kΩ. This
the detection of the electrochemical mediator hexaammineruthe- decrease of Rct demonstrated an enhancement of the conductiv-
nium added to the working solution for monitoring the PNA- ity after modification with AuNPs, thus AuNPs act both as an
miRNA duplex formation.[36] The positively charged Ru(NH3)63þ anchor for probe immobilization and as nanomaterial to increase
is accumulated at the working electrode through electrostatic inter- the electrochemical performances of the sensor. The Rct
action with the negatively charged miRNA, allowing for an increase increased after PNA immobilization (974 13 kΩ, orange plot)
of the signal current due to the formation of PNA–miRNA duplex, and after incubation with MCH (1474 9 kΩ, cyan plot) is
being the PNA oligomer uncharged (Figure 1). ascribed to the formation of PNA and MCH layers, hindering
The biosensor performance depends directly on the success of the reaction at the working electrode surface of the redox couple
device fabrication steps, namely drop-casting of AuNPs on an [Fe(CN)6]3/4. After hybridization with the complementary tar-
electrode surface, PNA immobilization, and PNA-miRNA get sequence (green plot), a significant increase of Rct to
hybridization step. For this reason, atomic force microscopy 3133 13 kΩ was observed due to the negative charge of
(AFM) and electrochemical impedance spectroscopy (EIS) were miRNA-486-5p, resulting in the formation of an electrostatic bar-
employed for the characterization of the biosensor surface at dif- rier for the negative [Fe(CN)6]3/4.
ferent fabrication steps. These EIS results were also confirmed by cyclic voltammetric
The morphological imaging at the nanoscale after any fabrica- measurements reported in Figure 5b, which show a typically
tion steps of the working electrode has been performed by AFM reversible voltammogram of redox couple [Fe(CN)6]3/4.
in noncontact mode (NCM). Because this procedure preserves The voltammogram obtained with bare SPE indicated a pair
samples by damages, NCM AFM characterization is usually of reversible redox peaks with ΔEp of 0.463 V (Figure 5b, black
employed to image and measure functionalized nanostructured curve), while after modification by drop-casting of working elec-
surfaces, ultrasoft materials, and biological samples, including trode surface with AuNPs, the intensity of peak currents signal
PNA and DNA sequences.[37] was considerably increased with a decrease of ΔEp, that is,
0.202 V (Figure 5b, purple curve), confirming that AuNPs pro-

vide an improvement in the electrochemical performances of
the electrode. The immobilization of PNA probe on AuNPs-mod-
ified electrodes cause an increase ΔEp to 0.410 V (Figure 5b,
orange curve), and the subsequent incubation with MCH causes
a further decrease of the redox peak current with ΔEp ¼ 0.400 V
(Figure 5b, cyan curve), due to the cover of the gold sites neces-
sary for electron transfer.[38] At last, SPE was incubated with
100 nM of miRNA-486-5p for 60 min, and the peak currents
decreased together with the increase of ΔEp ¼ 0.432 V, due to
the electrostatic barrier formation after the hybridization of
PNA with the target analyte.
2.4. Optimization of the Experimental Parameters
Before evaluating the biosensor suitability to detect different con-

centrations of miRNA, we performed the optimization of the
experimental parameters that could influence the analytical per-
formances of this sensing tool.
The first parameter optimized was the probe concentration to
be immobilized onto the working electrode surface. Four differ-
ent concentrations were tested, 25, 50, 100, 200 nM, and the
results are shown in Figure 5c. 100 nM was the concentration
giving the highest signal change and the highest repeatability,
given RSD% equal to 8%. Indeed, a decreasing signal was
obtained at increasing probe concentration. This probe density
dependence of biosensor signaling can result in steric hindrance
that could affect the PNA–miRNA interaction.[39]
For this reason, 100 nM was the probe concentration used for
miRNA target detection. Moreover, we studied the optimal probe
immobilization time for biosensor develpresent in the solutio-
nopment. The times investigated were 30, 60, 120, and
240 min (Figure 5d), and 60 min was selected in terms of repeat-
ability and signal variation obtained (RSD% equal to 8%).
Another critical step was the optimization of the amount of
electrochemical mediator to add to the working solution. After
the PNA–miRNA duplex formation, the hexaammineruthenium
can electrostatically interact with the negatively charged back-
bone of miRNA and the current signal measured is directly pro-
portional to the amount of Ru(NH3)63þ present in the
solution.[34] In this study, Ru(NH3)63þ concentrations examined
were 2.5, 5, and 7.5 mM, 5 mM was chosen, being the lowest
amount of mediator used, giving the maximum signal change
and lowest RSD%, 6% (Figure 5e). Indeed, a higher amount
of redox molecule (10 mM) produced the worst result. Too much
redox probe could affect the biosensor response due to the diffi-
culty in removing its excess during the washing steps, resulting
in increased background signal at low-miRNA concentrations.
The recognition between PNA and its target miRNA
(sequence 5’UCCUGUACUGAGCUGCCCCGAG 3’) and the
consequent duplex formation are crucial factors to establish
Figure 4. Noncontact mode (NCM) phase and topography images for the device sensitivity. Moreover, the binding time directly
each step of fabrication and detection of the working electrode: A,B) impacts the time of the analysis and its suitability to be used
AuNPs deposition; C,D) PNA incubation; E,F) 6-mercapto-1-hexanol
as a non-invasive diagnostic test. Thus, it was necessary to inves-
(MCH) incubation; G,H) miRNA incubation; I–L) PNA–miRNA duplex
and interaction with Ru(NH3)63þ. Topographical 3D reconstructions of
tigate the probe–target interaction, and the binding times tested
a 3 3 μm2 scanned surfaces of the working electrode after AuNPs depo- were 20, 60, 120, and 240 min. Although it is reported in the lit-
sition (M) and after the final modification step by PNA–miRNA duplex and erature that duplex formation of 20 bp requires about
interaction with Ru(NH3)63þ (N). 20 min,[40,41] in this work, 60 min were selected as a compromise
Figure 5. a) Nyquist plots obtained with bare electrode, AuNPs-modified SPE, AuNPs/PNA- (100 nM) modified SPE, AuNPs/PNA (100 nM)/MCH- (2 μM)
modified SPE and after PNA (100 nM)-miRNA-486 (100 nM) hybridization. Electrochemical impedance spectroscopy (EIS) measurements were carried
out using 100 μL of KCl 0.1 M solution containing [Fe(CN)6]3/4 5 mM was used. b) Voltammograms obtained with bare electrode, AuNPs-modified SPE,
AuNPs/PNA- (100 nM) modified SPE, AuNPs/PNA (100 nM)/MCH- (2 μM) modified SPE and after PNA (100 nM)-miRNA-486 (100 nM) hybridization
using 100 μL of KCl 0.1 M solution containing [Fe(CN)6]3/4 5 mM was used (potential range from 0.6 to 0.8, scan rate 0.05 V s1). c) PNA concentra-
tion optimization. Concentrations tested: 25, 50, 100, 200 nM in 50 mM phosphate buffer þ NaCl 150 mM, pH ¼ 7, in RNase-free water. PNA immobili-
zation time ¼ 60 min. [Ru(NH3)63þ] ¼ 5 mM. [miRNA] ¼ 100 nM. Binding time ¼ 20 min. d) PNA immobilization time optimization. Times tested: 30, 60,
120, 240 min. [PNA] ¼ 100 nM. [Ru(NH3)63þ] ¼ 5 mM. [miRNA] ¼ 100 nM. Binding time ¼ 20 min. e) Ru(NH3)63þ concentration optimization.
Concentrations tested: 2.5, 5, 7.5 mM. [PNA] ¼ 100 nM. PNA immobilization time ¼ 60 min. [miRNA] ¼ 100 nM. Binding time ¼ 20 min. f ) Binding time
optimization. Times tested: 20 min, 1, 2, 4 h. [PNA] ¼ 100 nM. PNA immobilization time ¼ 60 min. [Ru(NH3)63þ] ¼ 5 mM. [miRNA] ¼ 100 nM.
Measurements were carried out in triplicates (n ¼ 3) for each concentration or condition tested, and the mean the error of the three measurements
is reported together with relative SD (calculated as SD/mean 100).
between maximum signal variation and highest repeatability spiked with different concentrations of miRNA-486-5p (from
(RSD% equal to 7%); thus, it was chosen as the binding time 0.01 nM to 1 μM), giving the binding curve fitted by using hyper-
for further investigations (Figure 5f ). bolic Langmuir isotherm (R2 ¼ 0.972). Figure 6b shows a linear-
ity from 10 to 100 nM with an affinity constant of 28 7 nM and
the limit of detection equal to 0.7 nM.
2.5. miRNA-486-5p Detection The reproducibility was assessed by measuring each
concentration of the calibration curve in triplicate with RSD%
Once optimized all the experimental parameters, the device per- values lower than 10%, for instance, RSD% ¼ 7% for 10 nM,
formance was evaluated in the presence of different miRNA-486- RSD% ¼ 5% for 50 nM, and RSD% ¼ 4% for 100 nM. These
5p concentrations using the signal-on approach, which allows for results showed the suitability of the printed biosensor to detect
an increase of the measurable signal when the target is present. different concentrations of miRNA target with high sensitivity
The concentrations tested ranged between 0.01 nM and 1 μM and reproducibility, even in a complex matrix such as serum.
(prepared in phosphate buffer solution 50 mM containing Furthermore, this sensor demonstrated valuable analytical fea-
NaCl 150 mM, pH ¼ 7), and the binding curve obtained is shown tures when compared with the other sensing tools reported in
in Figure 6a. The curve was fitted to the hyperbolic Langmuir the literature (Table S1, Supporting Information).
isotherm with R2 ¼ 0.976 and an affinity constant equal to
28 4 nM. Moreover, a linear range between 10 and 100 nM
was obtained with a calculated limit of detection of 0.5 nM, which 2.6. Selectivity and Storage Stability Studies
corresponds to the amount of target able to produce a current
signal equal to 3 times the deviation standard of the blank. The selectivity of the biosensor was investigated by measuring
After evaluating the analytical performance of the biosensor in sequences differing from the target sequence, even for a single
standard solution, the device was tested in serum samples kindly base. The sequences tested were the following: 1 mismatched at
provided by healthy volunteers. In detail, undiluted serum was 3’ end of the oligonucleotide sequence, 1MM 3’ (5’
Figure 6. a) Binding curve in standard solution. Concentrations tested comprised in the range between 0.01 nM and 1 μM in 50 mM phosphate
buffer þ NaCl 150 mM, pH ¼ 7, in RNase-free water. [PNA] ¼ 100 nM. PNA immobilization time ¼ 60 min. [Ru(NH3)63þ] ¼ 5 mM. Binding time ¼ 60 min.
b) Binding curve in undiluted serum sample. Concentrations tested comprised in the range between 0.01 nM and 1 μM in undiluted serum sample.
[PNA] ¼ 100 nM. PNA immobilization time ¼ 60 min. [Ru(NH3)63þ] ¼ 5 mM. Binding time ¼ 60 min. c) Calibration curve of 1MM 3’, d) calibration curve
of 1MM 5’, e) calibration curve of 1MM, and f ) calibration curve of RNA scrambled obtained testing different concentrations in the range comprised
between 0.01 and 1000 nM in serum samples. Experimental condition: [PNA] ¼ 100 nM, PNA immobilization time ¼ 60 min, [Ru(NH3)63þ] ¼ 5 mM and
binding time ¼ 60 min. Measurements were carried out in triplicates (n ¼ 3) for each concentration or condition tested, and the mean the error of the
three measurements is reported together with relative SD (calculated as SD/mean 100).
UCCUGUACUGAGCUGCCCCGAC 3’); 1 mismatched at 5’ 2.7. Detection of miRNA-486–5p Expression on Serum by

end, 1MM 5’ (5’CCCUGUACUGAGCUGCCCCGAG 3’); 1 mis- Electrochemical Platform
matched at the central zone of the sequence, 1MM
(5’UCCUGUACUGTGCUGCCCCGAG 3’) and a fully random To evaluate if the device was able to detect miRNA-486-5p also in
sequence, scrambled (SCR)–RNA (5’ GCCGCAAGUGG real matrix, the electrochemical platform was used to detect the
CUCCGUACCUU 3’). The calibration curves of these sequences miRNA target in the serum samples of the 12 CD patients before
were compared with the calibration curve of miRNA-486 and after one year of a GFD, the 14 IBS patients and the 14
obtained in the same experimental conditions (serum sample healthy volunteers. In conformity with the results provided by
and concentration tested in the range comprised between 0.01 RT-qPCR, the miRNA-486-5p expression was 2.2 0.4 times
and 1000 nM). The results are shown in Figure 6. All the non- higher in CD patients with respect to healthy cases and
complementary sequences, 1MM 3’ (Figure. 6c), 1MM 5’ 1.7 0.3 and 1.8 0.3 times higher with respect to GFD patients
(Figure 6d), 1MM (Figure 6e), and SCR–RNA (Figure 6f ), caused and IBS patients, respectively (Figure 7). With these results, the
no proper signal variations even at higher concentration with device showed suitability to measure the miRNA target also in
respect to miRNA-486-5p, demonstrating the excellent selectivity serum sample of CD patients before and after GFD giving good
of the biosensor developed toward the target analyte even in the results in monitoring the treatment.
case of single mismatches.
Furthermore, the storage stability of the electrochemical
biosensor was evaluated at RT temperature, and the results 3. Conclusions
are reported in Figure S1, Supporting Information. The
miRNA-486-5p was measured after 1, 2, 3 d and 1, 2, 3, and Herein, we identified a novel biomarker in serum namely
4 weeks after the biosensor fabrication. Figure S2, Supporting miRNA-486-5p for the no-invasive diagnosis of CD, by analyzing
Information, shows that after 1 day we obtained a decrease in the serum of the cohorts of recruited subjects by the qRT-PCR
the signal change equal to 5%, after 2 weeks a decrease of technique. For the design of a smart analytical tool for miRNA-
10% and after the third week, the electrochemical biosensor 486-5p detection, printed electrochemical chip has been fabri-
resulted stable with a decrease of the electrochemical signal cated and modified with AuNPs and PNA probe to deliver an easy
change of the 40%. to use the device, able to work in standard solution and the
phate (HATU) was purchased from IRIS Biotech GMBH, while piperidine,
N,N-diisopropylethylamine (DIPEA), acetic anhydride, and TFA were sup-
plied by Biosolve, the N, methyl-morpholine (NMM) was provided by the
company Sigma-Aldrich. The anhydrous solvents for the synthesis of PNAs
(DCM, DMF) were supplied by the company J.T.Baker. Fmoc-PNA-
cytosine(Bhoc)-OH, Fmoc-PNA-thymine-OH, Fmoc-PNA-guanine(Bhoc)-
OH, and Fmoc-PNA-adenine(Bhoc)-OH were provided by Link
Technologies. Fmoc-Ahx-OH was supplied by Sigma-Aldrich, while
Fmoc-Cys (trt) -OH from Iris Biochem. The analyses and purification of
the products were carried out by high-pressure reverse-phase liquid chro-
matography (RP-HPLC). Analytical analyzes and purification of PNA
oligomers were performed on Agilent 1100 (Hewlett Packard) analytical
HPLC equipped with a UV detector and autosampler. A Phenomenex
Jupiter (5 μm C18 300 Å, 4.6 250 mm) column was used at a flow of
1 mL min1. Liquid chromatography-mass spectrometry (LC-MS) analyses
were conducted using a Phenomenex Jupiter (5 μm C18 300 Å column,
150 460 nm), on an instrument from a Surveyor series HPLC, coupled
Figure 7. miRNA-486-5p expression in four cohorts: control, coeliac dis-
to a Thermo Finnigan mass spectrometer equipped with an electrospray
ease patients, patients under gluten-free diet, and patients with irritable source and a time of flight analyzer.
bowel syndrome detected by electrochemical platform. [PNA] ¼ 100 nM. RNA Extraction: To carry out a gene expression analysis using the
PNA immobilization time ¼ 60 min. [Ru(NH3)63þ] ¼ 5 mM. Binding time RT-PCR technique, it is necessary to isolate the RNA. The validity and pre-
¼ 60 min. Data are shown as multiple unpaired t-test. ##p < 0.01. cision of gene expression estimated by RT-PCR were totally influenced by
the quality, purity, and integrity of the starting genetic material. To be con-
sidered as “high quality” RNA to be used in the retro-transcription, it must
untreated serum sample, with a limit of detection equal to 0.5 be free of RNase, proteins, and genomic DNA.[42,43]
and 0.7 nM, respectively. Due to the greater abundance of RNA in cells than in body fluids, even
The results obtained by testing serum samples from 12 CD small amounts of cell debris may significantly affect RNA profiling of cell-
free fluids such as serum. Therefore, high g-force centrifugation was per-
patients before and after 1 year of a GFD, 14 IBS patients,
formed to isolate circulating nucleic acids from serum samples to remove
and 14 healthy volunteers demonstrated the suitability of the all cell debris potentially present in them. This centrifugation step signifi-
developed miniaturized device for diagnosis and prognosis of cantly reduces the amount of cellular or genomic DNA and cellular RNA in
CD, paving the way for the application of electrochemical mini- the samples.
aturized PNA-based biosensor in autoimmune pathologies. In particular, the serum samples (approximately 1 mL) were centri-
The novelty of this work relies on the identification of a useful fuged in Eppendorf (10 min, 16 000 g) in a refrigerated centrifuge with
biomarker in serum for CD and the development of novel a fixed-angle rotor, and the obtained supernatant was carefully transferred
to a new tube that does not damage the pellet. The samples were subse-
smartphone-assisted electrochemical chip for the selected quently aliquoted and frozen at 80 °C.
biomarker detection in serum, enlarging the use of the electro- Among the specific kits for the extraction of small RNA commercially
chemical printed biosensor beyond the glucose strip and available, two were selected: the MagMAX mirVana (Applied Biosystems)
boosting furtherly the use of point-of-care devices for smart man- and the miRNeasy Serum/Plasma Kit (Qiagen).
agement of diseases. MagMAX mirVana enabled the isolation of total RNA, including
miRNAs, from a wide variety of matrices, including serum and plasma.
The kit used magnetic beads, which ensure a high yield of high-quality
RNA. The extraction protocol included several steps: initial digestion of
4. Experimental Section the serum samples performed with proteinase K, the addition of the lysing
agent 2-mercaptoethanol, followed by the addition of the magnetic beads
Reagents and Equipment: Sodium chloride, sodium dihydrogen
that bind RNA, and by several washes with appropriate solutions and iso-
phosphate hydrate (NaH2PO4 H2O), sodium hydrogen phosphate
(Na2HPO4), MCH, and tris(2-carboxyethyl)phosphine hydrochloride propanol. The samples were treated with DNase, and after proper washing
(TCEP), hexaammineruthenium(III) chloride (Ru(NH3)6Cl3) were pur- and subsequent binding with the beads, we proceed with the final elution
chased from Sigma-Aldrich (St. Louis, MO, USA). AuNPs (5 nm diameter) of the RNA in a specific elution buffer.
were synthesized following the procedure described in our previous The miRNeasy Serum/Plasma Kit (Qiagen) was specifically designed to
work.[36] Polyester-based screen-printed electrodes were supplied from purify total RNA, including miRNAs, from animal and human plasma and
Sense4Med Company (Rome, Italy), composed of graphite counter and serum. It used a phenol/guanidine solution for the lysis of samples, fol-
working electrodes and silver/silver chloride pseudoreference electrodes. lowed by the purification of total RNA through a silica membrane. QIAzol
All the electrochemical measurements were carried out using a miniatur- Lysis Reagent was a monophasic solution of phenol and guanidine thio-
ized potentiostat Sensit Smart (PalmSens, Netherlands) connected to a cyanate that facilitated the lysing process, denatured protein complexes
smartphone. The EIS was performed in [Fe(CN)6]3/4 (5 mM) þ KCl and RNases, and removed most of the DNA with residual proteins from
(0.1 M), and the resistance to charge transfer was calculated using the lysate by organic extraction. First, QIAzol was added to serum samples.
Randles circuit and Z-View software. After the addition of chloroform, the lysate was separated into the aqueous
miRNA-486-5p was purchased from ChemGenes Corp., Wilmington, and organic phases by centrifugation. RNA migrated in the upper aqueous
Massachusetts, USA while the RNA sequences for the selectivity study phase, DNA at the interphase, and proteins in the lower organic phase and
were purchased from Biomers.net the polymer factory Ulm (Germany). interphase. Then, the upper aqueous phase was separated, and ethanol
For the synthesis of PNA, the Fmoc-PAL-polyethylene glycol (PEG)-poly- was added to provide appropriate binding conditions for all RNA mole-
styrene (PS) resin was used, supplied by Applied Biosystem. The hydrox- cules, approximately 18 nucleotides long. The sample was then placed into
ybenzotriazole (HOBT) activators, hexafluorophosphate benzotriazole the RNeasy MinElute spin column, where the total RNA would bind to the
tetramethyl uronium (HBTU) were purchased from Novabiochem, 2- membrane while the phenol and other contaminants were washed away.
(1H-7-Azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronoium hexafluorophos The RNA was then eluted in a small volume of RNase-free water.
Both techniques were tried, and the miRNeasy Serum/Plasma Kit was fluorescent signal. The emission of the fluorescent signal depends on
selected for the RNA extractions due to its easier applicability. the activity of the Taq DNA polymerase.
To identify the expression profile of miRNAs in serum or plasma sam- During the PCR cycles, the collection of the fluorescent signal took
ples, there was currently no shared opinion within the scientific commu- place, and the intensity of the fluorescence of the reaction mixture in each
nity regarding the use of normalizers. The most common trend is to insert well was measured and plotted with respect to the reaction cycle. If a target
a synthetic miRNA (Spike-In) into samples to monitor the quality of RNA is present, its amplification can be monitored in real time.
extraction and cDNA synthesis. The Spike-In was inserted into the samples For the expression analysis of the miRNAs of our interest, the TaqMan
after adding the denaturant (e.g., QIAzol), followed by adding the chloro- Advanced miRNA cDNA Synthesis Kit and the TaqMan Advanced miRNA
form and subsequent separation phase. Assays (Applied Biosystems) were used.
The use of the Spike-In Control was strongly recommended to evaluate The TaqMan Advanced miRNA cDNA Synthesis Kit allowed the prepa-
the purification and amplification of miRNAs. Indeed, the yield of RNA ration of cDNA templates from the entire pool of mature microRNAs pres-
extracted from serum or plasma was highly variable from sample to sam- ent in the sample.
ple coming from different individuals and was generally too low for quan- miRNAs were modified by ligation of an adapter at the 5’ UTR and
tification by measuring its optical density (e.g., Nanodrop). The use of a extension of the 3’ UTR of the mature transcript by adding a poly(A) tail.
normalizer allowed overcoming this limit since, after RT-PCR, the obtained Modified miRNAs underwent a reverse transcription reaction followed by a
cycle threshold (CT) value relative to the synthetic miRNA allowed the nor- miRNA cDNA amplification reaction (miR-Amp).
malization between the different samples. Following assays (TaqMan Advanced miRNA Assays) were character-
Cel-miR-39-3p (Thermo Fisher Scientific) extracted from Caenorhabditis ized by specific probes and primers for a specific target and could detect
elegans has been selected among the recommended Spike-In, whose and quantify the expression levels of the miRNAs of interest through qPCR
nucleotide sequence is shown as follows: analysis. The assays can detect mature forms of miRNAs starting from
cel-miRNA-39-3p: 5’ UCACCGGGUGUAAAUCAGCUUG 3’ 2 μL of total RNA from serum or plasma.
The lyophilized preparation of cell-miRNA-39-3p was first resuspended When working out with small RNA, any slight variations in the quantity
in RNase-free water to obtain a solution with a concentration of 100 μM; of starting material, the modalities of sample collection, the purification
subsequently, dilutions were carried out until a concentration of 10 μM was and quality of the RNA, and the efficiency of the reverse transcription can
reached. Aliquots of the obtained solution were stored at 80 °C. determine errors in quantifying the expression level.
Extraction protocol: 1) Serum samples stored at 80 °C were thawed Normalization with an endogenous constitutively expressed control
and brought to room temperature (15–25 °C); 2) 100 μL of serum was col- gene is currently the most accurate method for correcting potential bias.
lected for each sample to which 5 volumes (500 μL) of QIAzol Lysis An ideal endogenous control shows a relatively constant and moderately
Reagent were added; 3) tubes containing lysate were vortexed and left abundant expression level in various tissues and cell types and various
at room temperature for 5 min; 4) to each sample, Spike-In Control types of treatment.
(3.5 μL, 10 pM) (cel-miR-39-3p) was added and the solutions thoroughly In our experiments, miRNA-16-5p was selected because it was reported
mixed; 5) chloroform (100 μL) was added to the matrix and vortexed (15 s) in the literature and known to be expressed at relatively constant levels in
until a completely homogeneous solution was obtained; 6) after incuba- plasma and serum:
tion (2–3 min) at room temperature, the tubes were centrifuged (12 000 g, miRNA-16-5p 5’ UAGCAGCACGUAAAUAUUGGCG 3’
for 15 min, 4 °C). Separation of three phases was obtained: an upper aque- Nevertheless, it is always suggested to validate the chosen endogenous
ous phase, colorless, containing RNA; white interphase containing DNA; control, or a set of controls, since no control can act as a universal endog-
and a red lower organic phase containing proteins and lipids; 7) the upper enous control under all experimental conditions.[44]
phase (approximately 300 μL) was transferred to a new tube, and 1.5 vol- Protocol: 1) Polyadenylation: 2 μL of RNA from each sample were
umes (450 μL) of 100% ethanol were added by mixing thoroughly; treated with a mix of Poly(A) Buffer, ATP, Poly(A) Enzyme, and RNase-free
8) 700 μL of each sample were pipetted into a RNeasy MinElute spin col- water and loaded into the thermal cycler for the polyadenylation reaction at
umn and centrifuged (8000 g, 15 s, 15–25 °C); the contents of the collec- the 3’ UTR (37 °C 45 min, 65 °C 10 min, 4 °C storage); 2) Ligation: after
tion tube were discarded; 9) the membranes were washed with suitable having prepared a reaction mix with appropriate volumes of 50% polyeth-
buffers by centrifuging (15 s, 8000 g), and a final wash was carried out with ylene glycol (PEG) 8000, Ligation Adapter, DNA Ligase Buffer, RNA Ligase,
80% ethanol (8000 g for 2 min); 10) the RNeasy MinElute spin columns, and RNase-free water, 10 μL of this mix were added to 5 μL of the product
with an open lid, were centrifuged at maximum speed for 5 min to dry the of previously obtained polyadenylation and the ligation reaction was car-
membranes well. This step is crucial to eliminate any ethanol residues that ried out in a thermal cycler (16 °C, 60 min, 4 °C storage);
could interfere with subsequent reactions; 11) to each column, RNase-free 3) Retrotranscription: 15 μL of the product obtained from the ligation reac-
water (14 μL) was added directly to the center of the membrane, and cen- tion were treated with 15 μL of a mix composed of dNTP, Universal RT
trifugation (1 min) was carried out at maximum speed to elute RNA; and primer, RT Enzyme Mix, RT Buffer, and RNase water followed by the
12) The RNA obtained was stored at 80 °C. reverse transcription reaction (42 °C, 15 min, 85 °C 5 min, 4 °C storage);
Analysis of MicroRNA Expression by using RT-PCR: RT-PCR, or qPCR 4) miR-Amplification: a mix containing miR-Amp Master Mix, miR-Amp
(quantitative PCR), allowed to quantify the PCR product’s synthesis at each Primer Mix, and RNase-free water were prepared, and 45 μL of this mix
amplification cycle in real-time. This technique allowed a quantitative anal- was added to 5 μL of the previously obtained product and was performed
ysis of the amount of initial DNA template (or cDNA), and that is why RT- at thermal cycler the specific amplification reaction for miRNAs:
PCR is often used, in combination with the retro-transcription reaction a) activation of the enzyme: at 95 °C 5 min; b) denaturation: at 95 °C 3
(RT), to quantify the levels of expression of specific genes. The signal s for 14 cycles; c) annealing and extension: at 60 °C, 30 s for 14 cycles;
was quantified by the fluorescence emitted by fluorophores (fluorescent d) stop reaction: at 99 °C, 10 min; and e) storage: at 4 °C; 5) RT-PCR: a
dyes) to bind the DNA molecules produced at each amplification cycle. 1:10 dilution of products obtained from the miR-Amplification with TE
The fluorophores could intercalate with DNA in a nonspecific way or buffer was carried out, the Fast Advanced Master mix (15 μL) was added
act as markers of oligonucleotide probes complementary to specific to this diluted cDNA (5 μL), and TaqMan Advanced miRNA Assays were
sequences. In our experiments, the TaqMan probe, a hydrolytic probe with specific to the targets of interest. The 52 samples to be analyzed were ali-
a high-energy fluorophore (reporter) at the 5’ untranslated region (UTR), quoted for the different assays and loaded on the 96-well reaction plates,
and a fluorophore inhibitor (quencher) at the 3’ UTR, was used. When the already containing the targets of interest (miRNA-449, miRNA-492,
probe was paired with the target sequence, the inhibitor was close enough miRNA-21-5p, and miRNA-486-5p), the endogenous control (miRNA-
to the fluorophore to block the emission of the fluorescent signal. During 16-5p), and the exogenous control (cel-miR-39-3p). A healthy subject
the elongation, in each amplification cycle, the polymerase hydrolyzed the was chosen to be used as a calibrator, loaded on each plate, and used
probe. In this way, the fluorophore was released into the reaction mixture as a reference for comparative analysis. A technical replicate was per-
and moved away from the inhibitor, resulting in the emission of the formed for each sample to improve accuracy and exclude operator errors
in the dispensing phase of samples. Negative controls necessary to iden- PNA Bioprobe Synthesis: PNA bioprobe was synthesized employing a
tify the presence of nonspecific products or sample contamination were synthetic protocol still reported in literature which consisted of repetitive
included in each assay. The negative control consists of all components of cycles of deprotection, coupling and acetylation, at room temperature.[46]
the reaction mixture except the genetic material to be amplified. To improve the efficiency of the coupling reaction, double couplings were
The qPCR analysis was performed on the StepOnePlus Real-Time PCR performed on G and A monomers. The Cys conjugate was obtained adding
System instrument (Thermo Fisher Scientific) in compliance with the fol- Fmoc-Ahx-OH and Fmoc-Cys (trt)-OH at the N terminus of PNA oligomer
lowing protocol: 1) activation of the enzyme: at 95 °C, 20 s; 2) denaturation: still immobilized on resin and employing the standard synthetic procedure
at 95 °C, 3 s for 40 cycles; and 3) annealing and extension: at 60 °, 30 s for for peptide synthesis. The purification of PNA was carried out through RP-
40 cycles. HPLC on a semi-preparative column, using a gradient of acetonitrile (0.1%
Data Analysis: RT-PCR signal analysis was performed with the software TFA) in H2O (0.1% TFA) from 5% to 70% in 30 min. The purified product
StepOne and StepOnePlus Software v2.3. For the analysis of miRNA was characterized by electrospray ionization time-of-flight (ESI-TOF) mass
expression levels, relative quantification of the fluorescence signal was per- spectrometry.
formed. The Cycle threshold (Ct) of each sample is inversely proportional Cys-Ahx-AntimiR-486-5p: Cys-Ahx-ctcggggcagctcagtacagga (Da): found:
to the initial quantity; this means that the smallest is the quantity of start- 6243.6600; [M þ 3H]3þ: 2082.1577; [M þ 4H]4þ: 1561.9401; [M þ 5H]5þ:
ing material, more cycles will be required to reach the threshold level. With 1249.8037; calculated: 6243.1016; [M þ 3H]3þ: 2082.0339; [M þ 4H]4þ:
a relative quantification, it is possible to determine the changes in the 1561.7754; [M þ 5H]5þ: 1249.6203;
expression levels of a target miRNA in different samples compared to Electrochemical PNA-Based Chip Preparation: First, for the immobiliza-
internal control, which can be co-amplified together with the target of inter- tion of PNA oligomer, AuNPs (8 μL) were previously drop-cast onto the
est. The endogenous control is a reference miRNA whose expression level working electrode surface to anchor the PNA sequence thanks to the thiol
remains constant in all tested samples and is not affected by experimental group introduced at the N terminus of the oligomer. The modification with
treatments. A reference control is advantageous, especially when quanti- AuNPs and relative working area were evaluated using cyclic voltammetry
fying the starting material is impossible or when a small amount of initial as technique in 0.1 M H2SO4.[47] Then, PNA (100 μM) prepared in phos-
template is available. phate buffer (50 mM) containing NaCl (150 mM) (pH ¼ 7, in RNase-free
For each experiment, the following agents are needed: 1) target: the water) was reduced for 1 h in the presence of TCEP (10 mM), prepared in
cDNA sequence to be analyzed; 2) calibrator: the sample to be used as phosphate buffer (50 mM) containing NaCl (150 mM) (pH ¼ 7, in
a reference for comparative analysis; and 3) endogenous control: an RNase-free water). After the reduction, the PNA (20 μL of 100 nM) was
miRNA constitutively expressed in all samples, necessary to normalize immobilized onto the SPE surface for 1 h at room temperature in a humid
the data with respect to the quantity of loaded cDNA and any variations chamber. Successively, the screen-printed electrodes were rinsed with
in the efficiency of the reaction. RNase-free water to remove the excess PNA, and they were incubated with
The calibrator was used when comparing multiple samples; the expres- 6-mercapto-1-hexanol (20 μL of 2 mM) prepared in phosphate buffer
sion of the target miRNAs in all samples was represented by a relative (50 mM) þ NaCl (150 mM) (pH ¼ 7, in RNase-free water) to passivate
increase or decrease compared to the calibrator. The advantage of the the electrode surface to help PNA-miRNA duplex formation. Then,
method based on the normalizer with reference genes or miRNAs was SPEs were rinsed again with RNase-free water to wash away the unbound
represented by chance to avoid an absolute quantification and the need MCH, and they were stored with a buffer solution at 4 °C in a humid
to have an exact quantification of the template loaded for the reaction. chamber.
The relative quantification of miRNA expression was performed with Electrochemical Analysis: In detail, all the electrochemical measure-
the 2(ΔΔCt) method. It assumed the hypothesis that both target and ref- ments were performed by differential pulse voltammetry technique
erence genes were amplified with an efficiency of approximately 100%, (DPV) scanning the potential from 0.1 to 0.5 V using as a working solu-
with an acceptable deviation of 5%. Therefore, it is necessary to verify tion 50 mM phosphate buffer containing NaCl 150 mM prepared in RNase-
the validity of this hypothesis, checking the amplification efficiencies of free water at pH 7. The detection of miRNA-486-5p was carried out by
both the target genes and the reference genes. Once established that comparing the signal obtained in the absence of the target (blank signal)
the initial hypothesis is correct, it was possible to proceed with the nor- with the signal recorded in the presence of the miRNA-486-5p. For the
malization of target gene (or miRNA) Ct with respect to the reference gene blank signal, 5 μL of the redox mediator (final concentration 5 mM) were
(or miRNA, ref ), both for treated samples (test) and for controls (cal). added to the working solution on the PNA-modified SPE. After 20 min, the
SPE was rinsed with RNase-free water to wash away the mediator in
ΔCt ðtestÞ ¼ Ct ðtarget; testÞ Ct ðref ; testÞ (1) excess, and the voltammetric measurement was carried out in a buffer
solution. Although the PNA was uncharged, a slight peak was observed
ΔCt ðcalÞ ¼ Ct ðtarget; calÞ Ct ðref ; calÞ (2) due to nonspecific interactions between probe and mediator
(Figure 1b). The PNA-modified SPE was then incubated with the desired
ΔCT of the test must be normalized with the ΔCt of the calibrator
concentration of miRNA-486-5p for 60 min. After the duplex formation,
5 μL of Ru(NH3)63þ (final concentration 5 mM) were added to the solution,
ΔΔCt ¼ ΔCt ðtestÞ–ΔCt ðcalÞ (3)
and after 20 min, the SPE was rinsed with RNase-free water to remove all the
Ratio ¼ 2ΔΔCt (4) unbound molecules. Finally, 100 μL of buffer solution was added to the SPE
surface, and the square-wave voltammetry (SWV) scanning was performed.
In our case, the reference miRNA (ref ) was represented by the miRNA- In this case, an increase of the signal is obtained thanks to the electrostatic
16-5p, while the control sample (cal) was the healthy (non-coeliac) subject interactions between the positively charged Ru(NH3)63þ and the negative
chosen[45] as a reference and reported in each assay. The result obtained charges of miRNA-486-5p (Figure 1c). The presence of the target was quan-
quantified the increase (or decrease) of the target gene in the test sample tified in comparison with the signal recorded in the absence of the target
relative to the calibrator sample and normalized with respect to the expres- and the signal change obtained was plotted against the concentration of
sion of the reference gene.[44] miRNA analyzed. For the detection of miRNA-486-5p, a miniaturized elec-
Assuming a value of ΔΔCT > 0, which corresponded to the condition in trochemical chip connected to a smartphone-assisted portable potentiostat
which ΔCT (test) > ΔCT (cal), and substituting its value in the expression was used.
of the ratio, we obtained a ratio < 1, index of under-expression of the gene Statics: For microRNA analysis in RT-PCR, each sample of serum was
of interest in the sample treated (test), compared to the control sample tested in triplicate (n ¼ 3) and data expressed as mean SD from at least
(cal). Conversely, a value of ΔΔCT > 0, that is, ΔCT (test) < ΔCT (cal), 3 independent experiments (n ≥ 3). Analysis was performed using
gave a ratio > 1, which corresponded to an increase in gene expression one-way analysis of variance (one-way ANOVA) with unpaired t-test.
in the treated sample compared to the control. p < 0.05 was considered to be statistically significant, p < 0.05 in
comparison with control group, and *p < 0.1, **p < 0.01, ***p < 0.001 [13] V. Vaira, L. Roncoroni, D. Barisani, G. Gaudioso, S. Bosari,
pathological group. The software used was GraphPad 3.0. G. Bulfamante, L. Doneda, D. Conte, C. Tomba, M. T. Bardella,
S. Ferrero, M. Locatelli, L. Elli, Clin. Sci. 2014, 126, 417.
[14] S. Magni, G. B. Comani, L. Elli, S. Vanessi, E. Ballarini, G. Nicolini,
M. Rusconi, M. Castoldi, R. Meneveri, M. U. Muckenthaler,
Supporting Information
M. T. Bardella, D. Barisani, Am. J. Gastroenterol. 2014, 109, 1662.
Supporting Information is available from the Wiley Online Library or from [15] A. Baldassarre, C. Felli, G. Prantera, A. Masotti, Genes 2017, 8,
the author. 234.
[16] Y. Cai, X. Yu, S. Hu, J. Yu, Genomics, Proteomics Bioinf. 2009, 7,
102.
Acknowledgements [17] W. Filipowicz, S. N. Bhattacharyya, N. Sonenberg, Nat. Rev. Genet.
2008, 9, 102.
V.C. and M.M. contributed equally to this work. M.M., M.S., and N.A.C. [18] K. A. Bascuñán, F. Pérez-Bravo, G. Gaudioso, V. Vaira, L. Roncoroni,
acknowledged H2020-PON-MISE 2014/2020—Project. n. F/050013/03/ L. Elli, E. Monguzzi, M. Araya, Dig. Dis. Sci. 2020, 65, 1982.
X32. D.M. acknowledges PRIN 2017 prot. 2017Y2PAB8_004 project [19] M. K. Masud, M. Umer, M. S. A. Hossain, Y. Yamauchi, N. T. Nguyen,
“Cutting Edge Analytical Chemistry Methodologies and Bio-Tools to M. J. Shiddiky, Trends Biochem. Sci. 2019, 44, 433.
Boost Precision Medicine in Hormone-Related Diseases.” [20] Y. Wang, Z. Li, Q. Lin, Y. Wei, J. Wang, Y. Li, R. Yang, Q. Yuan, ACS
Sensors 2019, 4, 2124.
[21] J. K. Cheong, Y. C. Tang, L. Zhou, H. Cheng, H. P. Too, Curr. Opin.
Conflict of Interest Biotechnol. 2022, 74, 2124.
[22] M. N. Islam, M. K. Masud, M. H. Haque, M. S. A. Hossain,
The authors declare no conflict of interest. Y. Yamauchi, N. T. Nguyen, M. J. Shiddiky, Small Methods 2017,
1, 1700131.
[23] R. Tavallaie, J. McCarroll, M. Le Grand, N. Ariotti, W. Schuhmann,
Data Availability Statement E. Bakker, R. D. Tilley, D. B. Hibbert, M. Kavallaris, J. J. Gooding,
Nat. Nanotech. 2018, 13, 1066.
The data that support the findings of this study are available from the [24] J. Guo, R. Shen, X. Shen, B. Zeng, N. Yang, H. Liang, Y. Yang,
corresponding author upon reasonable request. Q. Yuan, Chin. Chem. Lett. 2022, 33, 979.
[25] M. Umer, N. B. Aziz, R. G. Mahmudunnabi, Y. B. Shim, C. Salomon,
M. J. Shiddiky, Analyst 2021, 146, 5496.
[26] M. N. Islam, M. K. Masud, N. T. Nguyen, V. Gopalan, H. R. Alamri,
Keywords Z. Alothman, M. S. Al Hossain, Y. Yamauchi, A. K. Lamd,
PNA, point-of-care devices, screen-printed electrodes, smartphone- M. J. Shiddiky, Biosens. Bioelectron. 2018, 101, 275.
assisted potentiostat [27] M. N. Islam, V. Gopalan, M. H. Haque, M. K. Masud, M. S. Al
Hossain, Y. Yamauchi, N. Nguyen, A. K. Lam, M. J. Shiddiky,
Received: February 21, 2022 Biosens. Bioelectron. 2017, 98, 227.
Revised: May 13, 2022 [28] S. Shin Low, Y. Pan, D. Ji, Y. Li, Y. Lu, Y. He, Q. Chen, Q. Liu, Sens.
Published online: July 3, 2022 Actuators, B 2020, 308, 127718.
[29] G. Buoli Comani, R. Panceri, M. Dinelli, A. Biondi, C. Mancuso,
R. Meneveri, D. Barisani, Genes Nutr. 2015, 10, 32.
[1] F. Biagi, G. R. Corazza, Nat. Rev. Gastroenterol. Hepatol. 2010, 7, 158. [30] M. Moccia, A. Antonacci, M. Saviano, V. Caratelli, F. Arduini,
[2] P. H. R. Green, C. Cellier, N. Engl. J. Med. 2007, 357, 1731. V. Scognamiglio, Trends Anal. Chem. 2020, 132, 116062.
[3] N. Chand, A. A. Mihas, J. Clin. Gastroenterol. 2006, 40, 3. [31] A. Raiteri, A. Granito, A. Giamperoli, T. Catenaro, G. Negrini,
[4] P. H. Green, B. Jabri, Lancet 2003, 362, 383. F. Tovoli, World J. Gastroenterol. 2022, 28, 154.
[5] A. Fasano, C. Catassi, N. Engl. J. Med. 2012, 367, 2419. [32] G. F. Longstreth, W. G. Thompson, W. D. Chey, L. A. Houghton,
[6] P. Singh, S. Arora, A. Singh, T. A. Strand, G. K. Makharia, F. Mearin, R. C. Spiller, Gastroenterology 2006, 130, 1480.
J. Gastroenterol. Hepatol. 2016, 31, 1095. [33] E. F. Verdu, D. Armstrong, J. A. Murray, Am. J. Gastroenterol. 2009,
[7] S. Ashtari, M. A. Pourhoseingholi, K. Rostami, M. Rostami-Nejad, 104, 1587.
L. Busani, M. R. Tavirani, M. R. Zali, JGLD 2019, 28, 95. [34] F. Ricci, R. Y. Lai, A. J. Heeger, K. W. Plaxco, J. J. Sumner, Langmuir
[8] R. Costa Gomes, J. Cerqueira Maia, R. Fernando Arrais, C. André 2007, 23, 6827.
Nunes Jatobá, M. Auxiliadora Carvalho Rocha, M. Edinilma Felinto [35] http://www.admin.ox.ac.uk/uohs/policies-guidance/blood/.
Brito, A. Laissa Oliveira Nazion, C. Marques Maranhão, (accessed: December 2021).
H. De Sousa Maranhão, Scand. J. Gastroenterol. 2016, 51, 178. [36] M. Moccia, V. Caratelli, S. Cinti, B. Pede, C. Avitabile, M. Saviano,
[9] U. Volta, A. Fabbri, C. Parisi, M. Piscaglia, G. Caio, F. Tovoli, E. Fiorini, A. L. Imbriani, D. Moscone, F. Arduini, Biosens. Bioelectron. 2020,
Expert Rev. Gastroenterol. Hepatol. 2010, 4, 31. 165, 112371.
[10] W. P. Pais, D. R. Duerksen, N. M. Pettigrew, C. N. Bernstein, [37] G. Oliviero, S. D’Errico, B. Pinto, F. Nici, P. Dardano, I. Rea, L. De
Gastrointest. Endosc. 2008, 67, 1082. Stefano, L. Mayol, G. Piccialli, N. Borbone, ChemistryOpen 2017, 6,
[11] World Economic Forum, Top 10 Emerging Technologies 599.
2017, http://www3.weforum.org/docs/WEF_Top_10_Emerging_ [38] S. Cinti, E. Proietti, F. Casotto, D. Moscone, F. Arduini, Anal. Chem.
Technologies_report_2017.pdf (accessed: September 2021). 2018, 90, 13680.
[12] M. Capuano, L. Iaffaldano, N. Tinto, D. Montanaro, V. Capobianco, [39] D. A. Antonioli, Mod. Pathol. 2003, 16, 342.
V. Izzo, F. Tucci, G. Troncone, L. Greco, L. Sacchetti, PLoS One 2011, [40] S. D. Keighley, P. Estrela, P. Li, P. Migliorato, Biosens. Bioelectron.
6, e29094. 2008, 24, 906.
[41] S. Hassani, M. R. Akmal, A. Salek-Maghsoudi, S. Rahmani, [44] M. Filion, Quantitative Real-Time PCR In Applied Microbiology, Caister
M. R. Ganjali, P. Norouzi, M. Abdollahi, Biosens. Bioelectron. 2018, Academic Press, Norfolk, UK 2012.
120, 122. [45] K. J. Livak, T. D. Schmittgen, Methods 2001, 25, 402.
[42] S. Kennedy, N. Oswald, PCR Troubleshooting and Optimization: The [46] C. Avitabile, L. Moggio, L. D. D’Andrea, C. Pedone, A. Romanelli,
Essential Guide, Caister Academic Press, Norfolk, UK 2011. Tetrahedron Lett. 2010, 51, 3716.
[43] J. Kuang, X. Yan, A. J. Genders, C. Granata, D. J. Bishop, PLoS One [47] S. Cinti, S. Politi, D. Moscone, G. Palleschi, F. Arduini, Electroanalysis
2018, 13, e0196438. 2014, 26, 931.

Liquid Biopsy Beyond Cancer A miRNA Detection in

Uploaded by

Copyright:

Available Formats

Liquid Biopsy Beyond Cancer A miRNA Detection in

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Liquid Biopsy Beyond Cancer A miRNA Detection in

Uploaded by

Copyright:

Available Formats

RESEARCH ARTICLE

Liquid Biopsy beyond Cancer: A miRNA Detection in

V. Caratelli, F. R. Paggioro, L. Fiore, D. Moscone, F. Arduini A. L. Imbriani

F. Russo, G. Riezzo, G. Giannelli F. Arduini

Table 1. Clinical serum data of recruited cohort.

aminotransferase (IU L1).

0.202 V (Figure 5b, purple curve), conﬁrming that AuNPs pro-

2.4. Optimization of the Experimental Parameters

Before evaluating the biosensor suitability to detect different con-

UCCUGUACUGAGCUGCCCCGAC 3’); 1 mismatched at 5’ 2.7. Detection of miRNA-486–5p Expression on Serum by

You might also like