Materials Today Bio 16 (2022) 100354

Contents lists available at ScienceDirect

Materials Today Bio

journal homepage: www.journals.elsevier.com/materials-today-bio

An ultra-sensitive electrochemical biosensor using the Spike protein for

capturing antibodies against SARS-CoV-2 in point-of-care
Ana R. Cardoso a, b, c, Jo~ao Frederico Alves a, Manuela F. Frasco a, Ana Margarida Piloto b,
nica Serrano a, Daniela Mateus a, d, Ana Isabel Sebasti~
Vero ao d, Ana Miguel Matos e,
An
alia Carmo f, Teresa Cruz d, Elvira Fortunato c, M. Goreti F. Sales a, b, *
a
BioMark@UC/CEB - LABBELS, Faculty of Sciences and Technology, University of Coimbra, Coimbra, Portugal
b
BioMark@ISEP/ CEB - LABBELS, School of Engineering, Polytechnic Institute of Porto, Porto, Portugal
c
CENIMAT|i3N, Department of Materials Science, School of Science and Technology, NOVA University of Lisbon and CEMOP/UNINOVA, Campus de Caparica, 2829-
516, Caparica, Portugal
d
Center for Neuroscience and Cell Biology, Faculty of Pharmacy, University of Coimbra, Coimbra, Portugal
e
Chemical Engineering Processes and Forest Products Research Center, Faculty of Pharmacy, University of Coimbra, Coimbra, Portugal
f
Clinical Pathology Department, Centro Hospitalar e Universit
ario de Coimbra, Coimbra, Portugal

A R T I C L E I N F O A B S T R A C T

Keywords: This work presents an innovative ultra-sensitive biosensor having the Spike protein on carbon-based screen-
Electrochemical biosensor printed electrodes (SPEs), for monitoring in point-of-care antibodies against SARS-CoV-2, a very important tool
Antibodies for SARS-CoV-2 for epidemiological monitoring of COVID-19 infection and establishing vaccination schemes. In an innovative and
SARS-CoV-2
simple approach, a highly conductive support is combined with the direct adsorption of Spike protein to enable an
Spike protein
extensive antibody capture. The high conductivity was ensured by using carboxylated carbon nanotubes on the
Protective immunity
Point-of-care carbon electrode, by means of a simple and quick approach, which also increased the surface area. These were
then modified with EDC/NHS chemistry to produce an amine layer and undergo Spike protein adsorption, to
generate a stable layer capable of capturing the antibodies against SARS-CoV-2 in serum with great sensitivity.
Electrochemical impedance spectroscopy was used to evaluate the analytical performance of this biosensor in
serum. It displayed a linear response between 1.0 pg/mL and 10 ng/mL, with a detection limit of ~0.7 pg/mL.
The analysis of human positive sera containing antibody in a wide range of concentrations yielded accurate data,
correlating well with the reference method. It also offered the unique ability of discriminating antibody con-
centrations in sera below 2.3 μg/mL, the lowest value detected by the commercial method. In addition, a proof-of-
concept study was performed by labelling anti-IgG antibodies with quantum dots to explore a new electro-
chemical readout based on the signal generated upon binding to the anti-S protein antibodies recognised on the
surface of the biosensor. Overall, the alternative serologic assay presented is a promising tool for assessing pro-
tective immunity to SARS-CoV-2 and a potential guide for revaccination.

1. Introduction context, zoonotic viruses are considered the leading potential pathogens,
for which there are neither effective vaccines nor drugs. The most
In just two decades of the 21st century, numerous viral epidemics/ compelling example is the recent outbreak of Severe Acute Respiratory
pandemics have occurred that have been recognised by the World Health Syndrome Coronavirus 2 (SARS-CoV-2), which triggered Coronavirus
Organization. These include Severe Acute Respiratory Syndrome Coro- Disease 2019 (COVID -19) with devastating effects worldwide [3].
navirus (SARS-CoV, 2003), Chikungunya (2005), Influenza A (H1N1) The highly contagious properties led to a rapid increase in the number
(2009), Middle East Respiratory Syndrome Coronavirus (MERS-CoV, of infected individuals. The high prevalence of asymptomatic individuals
2012), Ebola (2014, 2019), and Zika (2015/2016) [1]. Such frequent has also led to a high rate of transmission of the disease. Although there
outbreaks are evidence that future pandemics can be expected [2]. In this are no accurate quantitative data in the literature, reports suggest that

* Corresponding author. BioMark@UC/CEB - LABBELS, Faculty of Sciences and Technology, University of Coimbra, Coimbra, Portugal.
E-mail addresses: [email protected], [email protected] (M.G.F. Sales).

https://doi.org/10.1016/j.mtbio.2022.100354
Received 5 May 2022; Received in revised form 2 July 2022; Accepted 5 July 2022
Available online 9 July 2022
2590-0064/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

approximately 30% of the population exposed to the virus is asymp- nanomaterials are particularly suitable for carbon electrodes and engi-
tomatic and transmits the disease. In addition, the signs of COVID -19 are neering their atomic properties allow for a variety of carbon allotropes
nonspecific and include respiratory symptoms, fever, cough, dyspnoea, that might answer to a diverse range of applications [18]. Carboxylated
and viral pneumonia that require biochemical testing for diagnosis. single-walled carbon nanotubes (SWCNTs), formed by a nanoscale gra-
Although the inherent infectious properties of the virus cannot be phene sheet folded around itself, are characterized by strong covalent
altered, it is possible to improve the biochemical tests that are currently bonding, high tensile strength, improved electron transfer, high surface
being conducted worldwide to stratify the population, monitor the level area-to-weight ratio and porosity (important for protein immobilization)
of immunity in the community, and conduct viral surveillance [4]. and good thermal stability (which may have a positive impact on the
Separating those infected (who must remain in quarantine) from reproducibility of the sensor) [16,19,20].
those who have an immunological response against the virus (antibodies) Additionally, engineered nanoparticles may also be used as electro-
is essential for controlling pandemics and improving the global economic chemical labels. Cadmium telluride quantum dots (CdTe QDs) are
scenario. This will allow a portion of the population to return to their colloidal semiconductor fluorescent nanoscale crystals that can be easily
“normal” activities without risk. In addition, in order for currently modified with biomolecules such as antibodies, generally improve the
available vaccines to be administered, it must be demonstrated that an- stability of the sensor as well as its photoelectric properties [21–23], and
tibodies against SARS-CoV-2 have formed, and it must be known how present good interaction with carbon nanotubes [24], making these
they develop in individuals over time and how they subside over time. nanoparticles an interesting electrochemical probe for our biosensor.
This is especially important at the community level, but difficult to Thus, this article presents a sensitive electrochemical biosensor that,
implement at the global level [5]. although assembled using an amazingly simple procedure, can detect
Overall, monitoring of antibody levels is necessary to guide antiviral small amounts of antibodies in human serum (and in much smaller
treatment, infection control, epidemiologic measures, and vaccination. amounts than conventional methods). The biosensor uses commercially
Serologic testing is essential for evaluating the qualitative and quanti- available screen-printed electrodes (SPEs) with a carbon working elec-
tative results of immune responses and has proven to be a valuable tool trode containing p-phenylenediamine and protein S. Analytical proper-
since the onset of infection. These tests are important public health tools ties are evaluated, and the devices are tested in human serum and further
to limit the risk of exposure and unintentional spread of the virus. This validated using a commercial benchmark method restricted to laboratory
could be critical to limit the global spread of the disease and allow for the facilities. Electrochemical detection is performed in two ways: with an
establishment of checkpoints for people in transit, including airports. In iron redox probe or with anti-IgG labelled with QDs.
addition, a global antibody test would also allow stratification of pop-
ulations by no, weak, moderate, or strong immune response, which is 2. Experimental section
particularly important for achieving our “normal” life [6,7].
Therefore, tools should be available to detect SARS-CoV-2 antibodies 2.1. Reagents and solutions
in the general population at different time points for each individual.
Current methods include enzyme-linked immunosorbent assay (ELISA) All chemicals were of analytical grade and ultrapure Milli-Q water
tests, which are neither rapid enough to allow immediate action nor laboratory grade (conductivity <0.1 μS/cm) was used. Chemical reagents
sensitive enough to detect early stages of viral infection and disease included potassium hexacyanoferrate III (K3[Fe(CN)6]) from Carlo Erba;
onset. They also require laboratory facilities and specific equipment and potassium hexacyanoferrate II (K4[Fe(CN)6]) trihydrate from Panreac;
reagents. In addition, some of these reagents are expensive and not carboxylated single-walled carbon nanotubes (SWCNT); N-ethyl-N’-(3-
available to meet global needs [8,9]. dimethylaminopropyl) carbodiimide hydrochloride (EDAC), p-phenyl-
Alternative rapid systems using biosensors have also been developed. enediamine, phosphate buffered saline (PBS, 0.01 mol/L, pH 7.4), N,N-
These could be ideal modern tools as they have the potential to develop dimethylformamide (DMF) were purchased from Sigma-Aldrich; N-
simple and inexpensive devices with low limits of detection for anti-SARS hydroxysuccinimide (NHS) was acquired from Merck; recombinant
CoV-2 antibodies. Currently, lateral flow assays are the most popular SARS-CoV-2 Spike protein, S1 subunit, Host Cell Receptor Binding
biosensors for SARS CoV-2 serological assays because they have several Domain (RBD), rabbit anti-SARS-CoV-2 S protein antibody, and rabbit
advantages, such as ease of use and rapid response time. However, they anti-SARS-CoV-2 nucleocapsid (N) protein antibody were obtained from
have limited sensitivity and sensing capability compared to biosensors RayBiotech.
and are also more expensive [8]. In contrast, electrochemical biosensors
can provide efficient, rapid, and accurate analysis, for which many pa- 2.2. Apparatus
pers have been published in the literature using this technology [10].
However, electrochemical biosensors also face some challenges that Electrochemical measurements were performed using a PalmSens4
delay technology transfer to reach commercialization at the point-of-care potentiostat/galvanostat controlled by PSTrace 5.8 software. Commer-
[11]. cial carbon SPEs (C-SPEs) from Metrohm DropSens (DRP-110) were used.
Recent works on antibody detection can use renewable materials as These contain a three-electrode system, including: (a) a carbon counter
substrates [12], optical techniques with special engineered antibodies electrode, (b) a silver reference electrode, and (c) a 4 mm diameter
[13] or plasmonic-based readout devices [14] and specially designed carbon working electrode. The C-SPEs were connected to a PalmSens
nanostructure arrays [15]. Overall, these works provide good analytical switch box that allows connection to the potentiostat.
properties but require several steps for electrode assembly, which can be
very complex and/or require specialised and expensive instrumentation. 2.3. Electrochemical measurements
Therefore, a simple biosensor for serological monitoring of SARS-CoV-2
antibodies is still lacking in the literature, which is particularly impor- Cyclic voltammetry (CV) was performed in the range of
0.3 to þ 0.7
tant to allow scaling up to an industrial production and their global use. V at a scan rate of 50 mV/s. Electrochemical impedance spectroscopy
The performance of an electrochemical biosensor depends on the (EIS) was performed in open circuit with a sinusoidal potential pertur-
characteristics of the electrode's interface and external conditions. bation with an amplitude of 0.01 V and 50 data points logarithmically
Immobilization capacity, sensitivity, conductivity or specificity of the distributed over a frequency range of 0.1–100000 Hz. The SWV data
biosensor may be affected by the electrode's surface area and material require a sampling potential of
0.3 to þ0.7 V, with a frequency of 2 Hz
composition [16,17]. Therefore, nanomaterials emerge as a promising and a step height of up to 2.5 mV.
resource to modify the electrode's interface to enhance its efficiency For the EIS data, Nyquist plots were used to represent the spectra
characteristics, improving the biosensor analytical performance. Carbon obtained, showing the frequency response of the electrolyte system and

2
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

plotting the imaginary component (Z00 ) of the impedance against its real 2.6. Ethics statement
component (Z0 ). The EIS data were fitted to the typical Randles equiva-
lent circuit, typically employed in impedimetric biosensors, and being Human serum was collected at Centro Hospitalar e Universitario de
composed herein of solution resistance (Rs), double layer capacitance Coimbra (CHUC) from patients with or without a clinical history of
(Cdl), and charge transfer resistance (Rct) [25–28]. RS is the resistance COVID -19 and without concomitant comorbidities (cardiovascular,
between the working and reference electrode. RCT is the electron transfer autoimmune, or respiratory disease). Positive samples were collected
resistance across the electrode-electrolyte interface. Cdl is the specific randomly during routine blood analysis by the clinicopathology service
capacitance at the interface of the electrolyte and the electrode and is of the University Hospital of Coimbra (Patients details is shown in
characterized by the non-faradaic charge that arises from the surface. Table S1). The collected human samples were immediately frozen until
Analysing the collected data, circuit elements displayed errors below 5%. use. This study was approved by the CHUC Ethics Committee (Process
In general, the electrical properties at the working electrode surface number OBS.SF.220-2021), always ensuring the protection of personal
were monitored by CV, EIS, and SWV studies using a redox probe solution data through anonymization (General Data Protection Regulation 2016/
of 5.0  10
3 mol/L [Fe(CN)6]3
and 5.0  10
3 mol/L [Fe(CN)6]4
in 679).
PBS buffer pH 7.4.
2.7. Analytical performance of the biosensor
2.4. Preparation of C-SPE biosensor
The analytical performance of the sensor layer was evaluated by in-
The approach used to build the current biosensor is described in Fig. 1 cubation with increasing standard concentrations of antibody solutions
and was inspired by our previous work on malaria and Zika diagnostics (anti-S protein) ranging from 1 pg/mL to 10 ng/mL prepared in PBS
[29,30]. The first step involves the dispersion of carboxylated SWCNT in buffer pH 7.4. Each standard was incubated for 40 min at room tem-
DMF (1.0 mg/mL) during 60 min. Then, the working electrode was perature in a humidified environment. After incubation with each stan-
incubated with 1.16 μL of the SWCNT solution and dried at 72  C for 30 dard concentration, the sensor was washed and its impedimetric
min (Fig. 1A). In the next step, the carboxyl groups on the surface of properties were analysed by EIS using the iron redox probe. The
SWCNT were activated with a solution of EDAC and NHS (0.025 mol/L analytical response of the biosensor was also evaluated in negative
EDAC and 0.0125 mol/L NHS) for 90 min in a humid chamber at room human serum, i.e., serum without anti-S protein antibodies, which served
temperature. The p-phenylenediamine (0.02 mol/L) was then bound to as a control and was spiked with known standard concentrations of anti-S
the SWCNT by incubation for 60 min to obtain an amine layer on the protein antibodies. Standards were prepared in negative human serum
working electrode (Fig. 1B). Finally, S protein (0.019 mg/mL) was bound diluted 500-fold in PBS buffer pH 7.4, and the incubation time was set at
to the amine layer for 40 min at room temperature and in a humid 40 min. All assays were performed in triplicate.
environment (Fig. 1C). All solutions were prepared in PBS buffer pH 7.4, The limit of detection (LOD) corresponded to the x þ 3σ , where x is
unless otherwise indicated. Subsequently, the analytical performance of the average value of the EIS blank signals (obtained in the absence of
the biosensor was determined by detecting antibodies bound to the S anti-S protein) and σ is the known standard deviation of the EIS blank
protein (Fig. 1D). signal in successive measurements [31].
Biosensor selectivity was evaluated by incubation with the anti-N
protein antibody in the same concentration range used to determine
2.5. Characterization by Raman spectroscopy
the calibration curve with the anti-S protein antibody. Similarly, the in-
cubation time was set at 40 min and all electrochemical assays were
Each step of the sensor design was followed by Raman spectroscopy
performed in triplicate.
by direct analysis of the material in a Thermo Scientific DXR Raman
spectroscope equipped with a 532 nm laser. The average signal-to-noise
ratio (peak height/RMS noise) was allowed for 900 s of measurement 2.8. Analysis of human serum samples by VIDAS®
time, after up to 2 min of photobleaching, with up to 3 mW of laser power
and a 50 μm slit aperture. Immunoassay tests were performed using the VIDAS® equipment

Fig. 1. Schematic representation of the C-SPEs and the functionalization of the working electrode: (A) addition of carboxylated SWCNT; (B) modification with p-
phenylenediamine through EDAC/NHS coupling reaction; (C) binding of S protein and (D) electrochemical detection of anti-S protein antibody.

3
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

from Biom erieux (Marcy l'Etoile, France). This instrument performs an preparation of anti-human IgG@QDs at 10.10 μg/mL and 101.01 μg/mL
automated qualitative ELISA for the detection of specific anti-SARS-CoV- (Fig. S2).
2 IgG in human serum, which requires the use of the VIDAS SARS -CoV-2
IgG 423834-02® kit (BioMerieux, Marcy l'Etoile, France). This kit con- 2.12. Secondary anti-human IgG@QDs as electrochemical labels
tains the disposable solid phase strips, solid phase vials containing hu-
manized recombinant SARS-CoV-2 antigen, the standard (with First, different concentrations of secondary goat anti-human IgG an-
humanized recombinant anti-SARS-CoV-2 IgG antibody), and the posi- tibodies (10.10, 51.02, and 101.01 μg/mL) labelled with CdTe QDs were
tive (also with humanized recombinant anti-SARS-CoV-2 IgG antibody) each incubated on one of the three sensor-modified surfaces with the
and negative controls (without the humanized recombinant anti-SARS- maximum positive serum concentration detected by our biosensor (752
CoV-2 IgG antibody) (https://www.biomerieux-diagnostics.com/vidas ng/mL) to select the optimal concentration for signal detection (Fig. S2).
-sars-cov-2). In addition, the solutions used to generate the standard The composite working electrode surface of five other biosensors was
curve were obtained by successive dilutions of the purified anti-SARS- incubated with PBS pH 7.4 for 40 min at room temperature until stabi-
CoV-2 S-protein S1 recombinant antibody from BioLegend (San Diego, lized. Then, positive serum solutions diluted in negative serum (107, 150,
CA, USA). 188, 250, and 752 ng/mL) were incubated on each of the five composite
The different solutions of the purified anti-SARS-CoV-2 S-protein S1 sensors. The selected concentration of QDs-labelled IgG antibody con-
recombinant antibody were prepared. Then, 100 μL of these solutions jugates was then added to our device to label the primary antibodies
were added to a designated spot on the kit strip and the automated present in the positive serum by incubating it at room temperature for 90
qualitative ELISA was performed. First, the instrument dilutes the con- min. Instead of a redox probe, a solution containing 0.1 mol/L KCl as
tents of the container. Then, the coated antigen captures the SARS-CoV-2 electrolyte was used to allow a more practical direct reading. The pa-
specific IgG, and a wash step follows to remove the unbound compo- rameters of the SWV method had a potential range of
1.0 V to
0.2 V,
nents. In a second step, the mouse monoclonal antibodies conjugated with a potential step of 2.5 mV with an amplitude of 0.02 V at a frequency
with alkaline phosphatase specifically recognize the IgG present in the of 2 Hz.
contents. Again, the device performs a wash step to remove the unbound
components. Finally, the instrument performs an incubation step by
3. Results and discussion
adding the substrate 4-methyl-umbelliferyl phosphate to the sample/
standard, which is hydrolysed to the fluorescent product 4-methyl-
3.1. Biosensor assembly and characterization
umbelliferone. The fluorescence of this product is measured at 450 nm
and a relative fluorescence value (RFV) is generated (Fig. S1) [32].
The biosensor was constructed in three steps: Carboxylation of the
carbon substrate, subsequent amination of this surface, and adsorption of
2.9. Correlation of VIDAS® signal with standard anti-S solutions S-protein. The preparation of an aminated surface directly on the carbon
substrate could combine the first two steps and further simplify the
Even though VIDAS® gives an RFV value, the information it provides procedure, but this was tried in several other ways that did not show
is only qualitative. It is considered a positive result if the test value sufficiently satisfactory results in terms of electrochemical stability to be
(which is the quotient between the RFV of the sample and the RFV of the a convincing electrochemical biosensor for global application.
instrument standard 1) is equal to or above 1.00, and negative if it is The electrochemical features of the C-SPEs in each of these steps were
below 1.00 [32,33] However, the RFV values are proportional to the followed by CV, EIS and SWV tests, with a standard iron redox probe,
concentration when anti-S standard solutions are analysed by VIDAS®. whose typical data are shown in Fig. 2. Carboxylation of the carbon
This allows the conversion of RFV values to concentration. To this end, working electrode on the C-SPEs was performed by casting carboxylated
anti-S standards of 0; 1.0; 5.0; 10.0; 25.0; 50.0; 100; 250 μg/mL were SWCNTs. As expected, the electrochemical properties showed that the
analysed by VIDAS®. electrical properties of the sensing layer were improved after modifica-
tion with carboxylated carbon nanotubes. The effectiveness of this step
2.10. Analysis of human sera with the electrochemical biosensor was more evident in the EIS and SWV data, with a lower Rct value of
~549 Ω and a higher current of ~48.62 μA, respectively, compared to the
The positive serum samples were tested to evaluate the biosensor original readings. After this modification, amination of the working
response compared to the VIDAS® results. The serum samples only had to electrode with p-phenylenediamine was achieved by an EDAC/NHS
be diluted beforehand (500-fold dilution in PBS, pH 7.4) to fit within the coupling reaction between the carboxyl groups of SWCNT and an amine
linear range of the sensitive biosensor and avoid interference from the group of p-phenylenediamine [34]. Subsequently, the S-protein was
serum matrix. All assays were performed in triplicate. incubated on the amine layer, which resulted in a decrease in current at
both CV (~83.68 μA) and SWV (~33.98 μA). In addition, a significant
2.11. Synthesis of quantum dots and conjugation with anti-human IgG increase in Rct (1038.67 Ω) was observed. These results suggest that the
antibodies S-protein recognition layer was successfully formed on the electrode.
The chemical modification of each step of the biosensor construction
Red emitting cadmium telluride quantum dots capped with 3-mercap- was followed by Raman spectroscopy (Fig. S3). The typical G and D bands
topropionic acid CdTe@MPA QDs were prepared according to a previous of the C-SPE were located at 1583.0 cm
1 and 1350.8 cm
1, respectively,
protocol with minor modifications [23]. Conjugation of antibodies with indicating the presence of a graphitic nanomaterial with sp2 and sp3
red emitting CdTe@MPA QDs was performed as follows. A solution of carbon-hybridization systems. Incubation of the carboxylated SWCNT
QDs (2.5 mg/mL) in PBS 0.01 mol/L pH 7.2 was incubated with goat onto the working electrode significantly changed the Raman spectra
anti-human IgG (51.02 μg/mL), EDAC (1.25 mg/mL), and NHS (1.50 obtained. The G and D bands were now at 1590.62 cm
1 and 1342.21
mg/mL) for 3 h at room temperature in a total volume of 100 μL. Then, cm
1, respectively, with the D band decreasing more than tenfold
30 mL of glycine (10 mmol/L) was added and the reaction mixture was compared to the G band. Overall, this confirmed the presence of the
incubated at room temperature for 30 min. Finally, the resulting solution carbon nanotubes on the C-SPE [35]. Subsequent addition of EDAC/NHS,
was centrifuged at 8500 rpm for 3 min in the Amicon™ Ultra-0.5 cen- p-phenylenediamine, S-protein, and anti-S-protein antibodies did not
trifugal filter device. The filtrate solution was discarded, and the mem- result in significant changes in Raman spectra, also confirmed by calcu-
brane was inverted to collect the anti-human IgG@QDs conjugates at the lating the ratios of the intensities of G and D bands (Table S2). Overall,
final concentration of 51.02 μg/mL. The procedure was repeated for the the significant change in Raman spectra was observed only after the

4
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

Pristine SWCNT S protein

150 600 60
A B C
I(μA)

-Z''(Ω)

I(μA)
400

0 30

200

Potential (V) Z'(Ω) Potential(V)

-150 0 0
-0.3 0.2 0.7 100 700 1300 -0.3 0.2 0.7

150
D E 60 F
Rct(Ω)

I(μA)
I(μA)

1000

75 30
500

S protein
S protein
S protein

Pristine

SWCNT
Pristine

Pristine
SWCNT

SWCNT

0 0 0

Fig. 2. Electrochemical data collected upon the biosensor construction, using CV (A), EIS (B) and SWV (C) measurements. Data includes pristine C-SPEs, after
modified with SWCNT and S protein. Graphics D and F show the current peak's intensity, and graphic E presents the Rct values with the respective error bars,
highlighting the reproducibility of three independent devices.

incubation of the carboxylated SWCNTs, with only minor changes in the Nyquist plots of the EIS measurements are shown in Fig. 3A. They show a
D-band and G-band observed in the subsequent phases of the biosensor positive correlation between the increase in Rct and the successive in-
assembly. crease in antibody concentration. These results indicate that the
impedimetric resistance, i.e., the diameter of the semicircles in the
Nyquist plots, increases with increasing amount of antibody that recog-
3.2. Analytical performance in PBS buffer nizes and binds the S protein. The sensor showed a linear range between
1.0 pg/mL and 10.0 ng/mL, while higher concentrations saturate the
The analytical properties of the biosensor were evaluated by testing sensor layer (Fig. 3B). The reproducibility of the sensor system was
increasing concentrations of the anti-S protein antibody, followed by EIS confirmed by performing three independent calibrations, with new
measurements and the generation of calibration curves. The typical

Fig. 3. EIS measurements of the calibration with anti-S protein antibodies (1.0 pg/mL – 100 ng/mL) in PBS buffer: Nyquist plots (A) and the corresponding calibration
curve (B). Readings were performed with the redox pair 5.0  10
3 mol/L [Fe(CN)6]3
and 5.0  10
3 mol/L [Fe(CN)6]4
prepared in 0.01 mol/L PBS buffer pH 7.4.

5
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

biosensing units. The relative standard deviation of the slope was ~2.6%, Overall, the data suggest that the current biosensor is not affected by
with an average slope of 280.47 Ω/decade. The squared correlation co- the presence of the anti-N protein antibody and that it responds accu-
efficient of the calibrations was >0.99. The reproducibility of the elec- rately to the anti-S protein antibody. These are interesting results, as any
trochemical response was excellent, proving that the system is analytical tool capable of distinguishing spike-from nucleocapsid-specific
reproducible, considering that the RSD values ranged from 0.5% (mini- immunity can help to study the immunological response of post-infected
mum) to 1.11% (maximum) in a linear trend. Overall, these results or vaccinated individuals [10,36].
showed a highly sensitive response to anti-S protein antibodies with a
LOD of 0.77 pg/mL. 3.4. Analytical performance in spiked human serum

The analytical performance of the sensing layer was evaluated using

3.3. Selectivity of the biosensor negative human serum spiked with anti-S protein antibodies to provide
more realistic information in the context of real-world applications. The
The biosensor had a recombinant S protein on the surface of the negative serum was diluted 500-fold to better match the sample to the
working electrode so that it should bind selectively to anti-S protein linear trend of the biosensor. The EIS measurements were used to eval-
antibodies. This selective binding may be confirmed by examining the uate the biosensor response, and the calibration curves were plotted as
response of the biosensor to other competing compounds. Since infected Rct versus the logarithm of concentration (Fig. 5). Rct values increased
humans also have anti-N protein antibodies, it was considered here to with increasing antibody concentration (Fig. 5A), and a linear trend was
evaluate the selectivity by checking the cross-reactivity of the biosensor confirmed in the range of 1.0 pg/mL and 10 ng/mL. The average slope
with these specific antibodies. Distinguishing between antibodies to S was 1825.5 Ω/decade, which was greater than that of the buffer, indi-
and N proteins is also critical for tracking COVID -19. Technically, anti-N cating that sensitivity increased. This likely reflects the higher
proteins are expected to interfere little to not at all with the measurement complexity of the sample matrix (including higher ion and protein con-
of anti-S proteins. tent), as the absolute Rct values were higher than those of the buffer at
In this study, the selectivity of the biosensor was evaluated by comparable concentrations. Overall, the biosensor provided reproducible
examining the response to individual solutions of anti-N and anti-S responses, with relative standard deviations of slope of ~4.0. The
protein antibodies prepared in the same concentration range over the squared correlation coefficient of all calibrations was also >0.99, con-
concentration range of the linear response. The data collected considered firming the quality of the linear trend (Fig. 5B). Overall, these results
the Rct values obtained in each EIS measurement, and the average values showed a sensitive response to detect anti-S antibodies with a very low
were plotted in Fig. 4. In general, in the presence of the anti-N protein LOD. Although tested in a more complex matrix, LOD was not signifi-
antibody, the biosensor showed a uniform response (like the blank) over cantly different from LOD in PBS buffer pH 7.4 and was 0.70 pg/mL. This
the entire concentration range tested. This contrasted with the increasing LOD is in the range or even lower than recent findings reported in the
Rct values of the biosensor in the presence of the increasing concentra- literature for other electrochemical biosensors, mostly also containing
tions of anti-S protein antibody. nanomaterials that improve the sensitivity (Table S3).
Considering the detection of antibodies in infected humans and
because the serum was diluted 500-fold, the biosensor can be used to
3500 Anti- N protein analyse human sera from 0.5 ng/mL to 5000 ng/mL. This is a wide
RCT(Ω)

concentration range that will simplify future application procedures,

Anti-S protein although more concentrated solutions may saturate the biosensor and
require additional dilution. However, further testing was performed with
3000 positive samples, as the level in the serum of a healthy person (negative
serum) is quite different from that of a person with COVID -19.

3.5. Analysis of positive sera samples

2500
The ability of the electrochemical biosensor to determine antibody
levels in different positive human samples was tested by analyzing the
2000 same samples with both the biosensor and VIDAS®. Since VIDAS® only
provides qualitative data (positive or negative) in terms of antibody
levels in sera, we generated a standard curve using different standard
solutions of purified anti-SARS-CoV-2 S1 recombinant protein anti-
1500 bodies. The RFV values obtained for these antibodies (expressed in
arbitrary units) were correlated with the logarithm of the antibody
concentration, enabling the extraction of quantitative data from this
commercial method. The data obtained are shown in Fig. S4 and consider
1000 only the concentration range of the antibodies, which have a linear
behaviour. The anti-S-SARS CoV-2 IgG concentration in the sera could be
now calculated by intersecting this curve, which gives an idea of the
range of anti-S protein IgG antibody concentrations in the COVID-19
500
positive population.
The samples included in this comparative study contained different
ranges of antibody concentrations (n ¼ 5), as expected given the quite
0 variable RFV concentrations in VIDAS®. The concentrations chosen for
1 pg/mL
1 2
100 pg/mL 10 pg/mL
3
1 ng/mL
4 5
10 ng/mL this purpose are well above the capabilities of the biosensor and range
from 5 to 230 μg/mL. Prior to analysis, samples were treated according to
Fig. 4. EIS response of the biosensor in the form of Rct values after the incu- the method used. For VIDAS®, the procedure provided by the commer-
bation of anti-S or anti-N protein antibodies, ranging from 1.0 pg/mL to 10 cial instrument was used, while for the biosensor only an appropriate
ng/mL. dilution was performed.

6
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

Fig. 5. EIS measurements of the calibration with anti-S protein antibodies (1.0 pg/mL – 100 ng/mL) spiked in negative human serum (500-fold diluted in PBS buffer):
Nyquist plots (A) and the corresponding calibration curve (B). Readings were performed with the redox pair 5.0  10
3 mol/L [Fe(CN)6]3
and 5.0  10
3 mol/L
[Fe(CN)6]4
prepared in 0.01 mol/L PBS buffer pH 7.4.

The concentrations of these samples obtained with VIDAS® and the concentration incubated on the biosensor ranged from 70 to 752 ng/mL,
biosensor are shown in Table S4. The concentrations obtained with the with clearly distinguishable values for concentrations of 150, 107, and 83
biosensor were extracted from the correlation of the log concentration of ng/mL. The electrochemical biosensor was thus able to detect differences
antibodies known to be present in the samples with the Rct values of each between samples that had much lower concentrations than those ana-
of these positive samples (Fig. S5). This was done to provide a valuable lysed by VIDAS®, down to 70 ng/mL. Validation of the data obtained by
correlation between the positive sera and the electrochemical method, EIS was not performed because we were not aware of any method suit-
which proved necessary to eliminate the effect of the sample matrix now able for comparison at such low concentrations. This unique feature of
containing additional biomolecules required to combat the disease. At- the proposed devices could be a valuable tool for future epidemiological
tempts were made to compare the readings with the concentrations of studies and for understanding changes in antibody levels associated with
antibodies diluted in negative sera, but the data obtained were not reli- vaccination (as a recommendation for clinicians on when to recommend
able. It was also noted that the Rct values increased well beyond the usual a new vaccine dose).
calibration ranges, which indeed confirmed some influence of the posi-
tive samples, but these had no significant effect upon the concentration of
antibodies yielding a linear trend. 3.7. Alternative electrochemical analysis of human sera
Overall, the relative errors in this analysis ranged from
10.2 to
þ15.3%. Considering that these are human samples, these results can be Considering the good results obtained in testing positive human
considered accurate [37,38]. Furthermore, these are excellent results serum with the electrochemical biosensor, another proof-of-concept was
considering that the samples were highly diluted in the biosensor mea- explored by developing and using nanoparticles as electrochemical
surements and that the two methods have very different operating markers bound to secondary antibodies. These secondary antibodies bind
principles. This also indicates that the biosensor can be considered an the anti-S protein antibodies recognised on the surface of the biosensor
important alternative to current methods, especially when lower and generate an electrochemical signal. For this purpose, the secondary
amounts of antibodies are involved. antibodies were labelled with CdTe QDs, which are nanoscale colloidal
semiconductor crystals that have electrochemical properties (Fig. S2)
[24]. The electrochemical results of the QDs were investigated in terms of
3.6. Comparison of method capabilities to low antibody concentrations metal oxidation by SWV, moving from
1.0 V to positive potentials.
In a proof-of-concept study, different concentrations of anti-human
From the data obtained, it appeared that the biosensor was able to IgG antibodies labelled with CdTe-QDs were tested to evaluate their ef-
detect much lower concentrations of antibodies than VIDAS®. To fects on current measurements at the sensor surface (Fig. 6). The most
confirm this, a positive human sample with a concentration of approxi- diluted and the most concentrated solutions of the IgG antibodies pro-
mately 225.66 μg/mL was diluted down to different concentration levels duced similar current values (1.60 μA and 1.63 μA, respectively).
and evaluated using both methods. Therefore, considering the fluorescence values obtained for the different
VIDAS® was only able to measure concentrations up to 2256 ng/mL, concentrations of QDs-labelled IgG antibodies, we decided to label the
which corresponds to a 100-fold dilution of the original sample. At higher primary antibodies with the lower concentration of IgG antibodies (10.10
serum dilutions, corresponding to 150, 107, and 83 ng/mL, the analytical μg/mL), as this could be a more dispersed solution containing a higher
signal in VIDAS® was similar, indicating that the concentration was ratio of CdTe-QDs nanocrystals for each IgG antibody (thereby ampli-
equal to or lower than 2.3 μg/mL and that further concentration differ- fying the signal).
ences could not be distinguished. Overall, VIDAS® was unable to In this approach, the biosensor was first incubated in a solution
discriminate between antibody concentrations below 2.3 μg/mL. This is a containing diluted positive serum (107, 150, 188, 250, or 752 ng/mL)
fairly high concentration and confirms the limited ability of the and then in a solution containing anti-human IgG@QDs of 10.10 μg/mL.
comparative method to provide accurate data for lower antibody con- The SWV data obtained for these concentrations are shown in Fig. 6D and
centrations (the commercial method works well but only provides reli- reflect the metal oxidation readings. The current signals decreased with
able data for higher concentrations). increasing antibody concentration, yielding a linear trend when the peak
In contrast, serial dilution of this sample resulted in a linear trend height values were plotted against the log concentration of the antibody
against antibody concentration by EIS. This was achieved when the (150 ng/mL to 752 ng/mL). In principle, one would expect that a higher

7
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

Fig. 6. (A and B) Fluorescence signals of red emitting QDs in the presence of increasing concentrations of goat anti-human IgG, in PBS 0.01 mol/L pH 7.2. (C) UV–vis
absorption spectra of goat anti-human IgG in PBS 0.01 mol/L pH 7.2 (a) and the UV/vis spectra of red emitting QDs 2.5 mg/mL in PBS 0.01 mol/L pH 7.2, (b) along
with the correspondent fluorescent spectra (c). (D) SWV voltammograms of different concentrations of positive serum containing primary antibodies conjugated with a
10.10 ng/mL concentration of IgG secondary antibodies labelled with CdTe QDs.

amount of anti-S antibody would lead to a higher number of IgG@QDs This novel electrochemical biosensor can be used post-infection or post-
conjugates, which in turn should lead to a higher current value due to the vaccination when circulating antibody concentrations are lower and
higher amount of QDs. However, the system may undergo opposite ef- cannot be detected by conventional ELISA methods. Its detection capa-
fects during the incubation of the IgG@QDs conjugates. While the anti- bilities significantly outperform commercial methods available on the
bodies block the electrical surface and decrease the current signals, the market and are comparable or better than other electrochemical methods
current value of the observed peak increases with increasing number of described in the literature.
QDs. Since this leads to opposite effects and the linear trend has a It is important to note that a cross-response of the biosensor against
negative slope, it was obvious that the anti-IgG dominated the signal other SARS virus was not evaluated. However, a cross-reactivity with
obtained. other antibodies against these means that one's serum has antibodies that
In any case, this proof-of-concept approach is an alternative way to also recognize S protein, meaning that there is an expected degree of
read the electrochemical response if one wishes to obtain a more selec- protective immunity conferred by these. Thus, as the main target of the
tive response. Alternatively, direct electrochemical measurements using biosensor is to monitor the protective immunity by means of antibodies,
the iron redox probe already outperform conventional methods such as an eventual cross-reactivity would not hamper the result of a test pro-
VIDAS® in terms of LOD and sensitivity. vided by this biosensor in point-of-care.
Overall, the biosensor provides very low LOD values and can reach
4. Conclusions very low concentrations of positive sera when testing human sera.
Therefore, the biosensor described here has the potential to become a
A rapid serological test to diagnose and/or monitor immunity after valuable tool for early diagnosis and monitoring of antibody levels during
SARS-CoV-2 infection or vaccination remains a challenge given current vaccination follow-up. Future work will therefore focus on miniaturiza-
technologies. ELISA-based methods are not very sensitive and reliable tion and the use of different nanomaterials as well as an environmentally
quantitative information is scarce. Here we demonstrate the successful friendly substrate, as paper. The development of multiplex biosensors
development of a simple, innovative, highly sensitive and selective covering a variety of different viruses is also foreseeable.
biosensor for the detection of antibodies to the S protein in human serum.

8
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

Author statement E.M. Gillis-buck, Z.B. Steinhart, Y. Lee, R. Apathy, M.J. Lipke, J.A. Smith, T. Zheng,
I.C. Boothby, E. Isaza, J. Chan, D.D.A. Ii, J. Lee, T.A. Macrae, T.S. Kyaw, D. Wu,
D.L. Ng, W. Gu, V.A. York, H.A. Eskandarian, L.A. Farrington, R.P. Loudermilk,
Ana R. Cardoso. Investigation. Methodology. Formal analysis. Vali- K. Koshal, K.C. Zorn, W.F. Garcia-beltran, D. Yang, M.G. Astudillo, B.E. Bernstein,
dation. Visualization. Data Curation. Writing - original draft. Jo~ao Fred- J.A. Gelfand, E.T. Ryan, R.C. Charles, A.J. Iafrate, J.K. Lennerz, S. Miller, C.Y. Chiu,
erico Alves. Investigation. Formal analysis. Validation. Writing - original S.L. Stramer, M.R. Wilson, J.G. Cyster, J.D. Ernst, A.H.B. Wu, K.L. Lynch, C. Bern,
P.D. Hsu, A. Marson, Evaluation of SARS-CoV-2 serology assays reveals a range of
draft. Manuela F. Frasco. Investigation. Methodology. Formal analysis. test performance, Nat. Biotechnol. 38 (2020) 1174–1183, https://doi.org/10.1038/
Validation. Writing - review & editing. Ana Margarida Piloto. Investiga- s41587-020-0659-0.
tion. Formal analysis. Validation. Writing - original draft. Veronica [9] M.A. Macmullan, A. Ibrayeva, K. Trettner, L. Deming, S. Das, F. Tran, J.R. Moreno,
J.G. Casian, P. Chellamuthu, J. Kraft, K. Kozak, F.E. Turner, V.I. Slepnev, L.M. Le
Serrano. Investigation. Formal analysis. Daniela Mateus. Investigation. Page, ELISA detection of SARS - CoV - 2 antibodies in saliva, Sci. Rep. 10 (2020)
Methodology. Formal analysis. Ana Isabel Sebasti~ ao. Investigation. 1–8, https://doi.org/10.1038/s41598-020-77555-4.
Methodology. Formal analysis. Ana Miguel Matos. Validation. Resources. [10] N. Koerber, A. Priller, S. Yazici, T. Bauer, C. Cheng, H. Mijo, H. Wintersteller,
S. Jeske, E. Vogel, M. Feuerherd, K. Tinnefeld, C. Winter, J. Ruland, M. Gerhard,
Formal analysis. Supervision. Analia do Carmo. Validation. Data curation. B. Haller, C. Christa, O. Zelger, H. Roggendorf, M. Halle, J. Erber, P. Lingor,
Writing - review & editing. Teresa Rosete. Methodology. Validation. Re- Dynamics of spike-and nucleocapsid speci fi c immunity during long-term follow-up
sources. Project administration. Supervision. Elvira Fortunato. Resources. and vaccination of SARS-CoV-2 convalescents, Nat. Commun. 13 (2022) 1–14,
https://doi.org/10.1038/s41467-021-27649-y.
Writing - review & editing. Supervision. M. Goreti F. Sales. Conceptuali- [11] N. Kumar, N.P. Shetti, S. Jagannath, T.M. Aminabhavi, Electrochemical sensors for
zation. Methodology. Funding acquisition. Project administration. Su- the detection of SARS-CoV-2 virus, Chem. Eng. J. 430 (2022) 1–15, https://doi.org/
pervision. Visualization. Writing - review & editing. 10.1016/j.cej.2021.132966.
[12] A. Yakoh, U. Pimpitak, S. Rengpipat, N. Hirankarn, O. Chailapakul, S. Chaiyo,
Paper-based electrochemical biosensor for diagnosing COVID-19 : detection of
Declaration of competing interest SARS-CoV-2 antibodies and antigen, Biosens. Bioelectron 176 (2021) 1–8, https://
doi.org/10.1016/j.bios.2020.112912.
[13] S.K. Elledge, X.X. Zhou, J.R. Byrnes, A.J. Martinko, I. Lui, K. Pance, S.A. Lim,
The authors declare that they have no known competing financial J.E. Glasgow, A.A. Glasgow, K. Turcios, N.S. Iyer, L. Torres, M.J. Peluso,
interests or personal relationships that could have appeared to influence T.J. Henrich, T.T. Wang, C.M. Tato, K.K. Leung, B. Greenhouse, J.A. Wells,
Engineering luminescent biosensors for point-of-care SARS-CoV-2 antibody
the work reported in this paper. detection, Nat. Biotechnol. 39 (2021) 928–935, https://doi.org/10.1038/s41587-
021-00878-8.
Acknowledgements [14] O. Calvo-lozano, M. Sierra, M. Soler, M. Estevez, L. Chiscano-Camon, A. Ruiz-
Sanmartin, J.C. Ruiz-Rodriguez, R. Ferrer, J.J. Gonzalez-L opez, J. Esperalba,
C. Fernandez-Naval, L. Bueno, R. L opez-Aladid, A. Torres, L. Fernandez-Barat,
The authors acknowledge funding through project TecniCov (POCI- S. Attoumani, R. Charrel, B. Coutard, L.M. Lechuga, Label-free plasmonic biosensor
01-02B7-FEDER-069745), co-funded by FEDER through COMPETE2020 for rapid , quantitative , and highly sensitive COVID-19 serology : implementation
and clinical validation, Anal. Chem. 94 (2022) 975–984, https://doi.org/10.1021/
and Lisboa2020 and CY-Sensors (POCI-01-0145-FEDER-032359) acs.analchem.1c03850. In press.
through Fundaç~ ao para a Ci^encia e a Tecnologia (FCT), Portugal. ARC [15] A. Ali, C. Hu, S. Jahan, B. Yuan, M.S. Saleh, E. Ju, S. Gao, R. Panat, Sensing of
acknowledge funding to National Foundation for Science and Technol- COVID-19 antibodies in seconds via aerosol jet nanoprinted reduced-graphene-
oxide-coated 3D electrodes, Adv. Mater. 33 (2021) 1–15, https://doi.org/10.1002/
ogy, I.P., Portugal (FCT) through the PhD. Grant, reference SFRH/BD/
adma.202006647.
130107/2017. [16] X. Bin, E.H. Sargent, S.O. Kelley, Nanostructuring of sensors determines the
efficiency of biomolecular capture, Anal. Chem. 82 (2010) 5928–5931, https://
doi.org/10.1021/ac101164n.
Appendix A. Supplementary data [17] L. Soleymani, Z. Fang, E.H. Sargent, S.O. Kelley, Programming the detection limits
of biosensors through controlled nanostructuring, Nat. Nanotechnol. 4 (2009)
Supplementary data to this article can be found online at https:// 844–848, https://doi.org/10.1038/nnano.2009.276.
[18] Z. Wang, Z. Dai, Carbon nanomaterial-based electrochemical biosensors: an
doi.org/10.1016/j.mtbio.2022.100354.
overview, Nanoscale 7 (2015) 6420–6431, https://doi.org/10.1039/c5nr00585j.
[19] M. Pagan, D. Suazo, N. del Toro, K. Griebenow, A comparative study of different
References protein immobilization methods for the construction of an efficient nano-structured
lactate oxidase-SWCNT-biosensor, Biosens. Bioelectron. 64 (2015) 138–146,
https://doi.org/10.1016/j.bios.2014.08.072.
[1] WHO, Managing epidemics, Thomson Pr, Switzerland, 2018.
[20] N. Etminan, M. Yoosefian, H. Raissi, M. Hakimi, Solvent effects on the stability and
[2] A.G.P. Ross, S.M. Crow, M.W. Tyndal, Planning for the next global pandemic, Int. J.
the electronic properties of histidine/Pd-doped single-walled carbon nanotube
Infect. Dis. 38 (2015) 89–94, https://doi.org/10.1016/j.ijid.2015.07.016.
biosensor, J. Mol. Liq. 214 (2016) 313–318, https://doi.org/10.1016/
[3] P. Zhou, X. Lou Yang, X.G. Wang, B. Hu, L. Zhang, W. Zhang, H.R. Si, Y. Zhu, B. Li,
j.molliq.2015.12.009.
C.L. Huang, H.D. Chen, J. Chen, Y. Luo, H. Guo, R. Di Jiang, M.Q. Liu, Y. Chen,
[21] H. Ashiba, Y. Sugiyama, X. Wang, H. Shirato, K. Higo-Moriguchi, K. Taniguchi,
X.R. Shen, X. Wang, X.S. Zheng, K. Zhao, Q.J. Chen, F. Deng, L.L. Liu, B. Yan,
Y. Ohki, M. Fujimaki, Detection of norovirus virus-like particles using a surface
F.X. Zhan, Y.Y. Wang, G.F. Xiao, Z.L. Shi, A pneumonia outbreak associated with a
plasmon resonance-assisted fluoroimmunosensor optimized for quantum dot
new coronavirus of probable bat origin, Nature 579 (2020) 270–273, https://
fluorescent labels, Biosens. Bioelectron. 93 (2017) 260–266, https://doi.org/
doi.org/10.1038/s41586-020-2012-7.
10.1016/j.bios.2016.08.099.
[4] D.K.W. Chu, Y. Pan, S.M.S. Cheng, K.P.Y. Hui, P. Krishnan, Y. Liu, D.Y.M. Ng,
[22] Z. Deng, Y. Zhang, J. Yue, F. Tang, Q. Wei, Green and orange CdTe quantum dots as
C.K.C. Wan, P. Yang, Q. Wang, M. Peiris, L.L.M. Poon, Molecular diagnosis of a
effective pH-sensitive fluorescent probes for dual simultaneous and independent
novel coronavirus ( 2019-nCoV) causing an outbreak of pneumonia, Clin. Chem. 66
detection of viruses, J. Phys. Chem. B 111 (2007) 12024–12031, https://doi.org/
(2020) 549–555, https://doi.org/10.1093/clinchem/hvaa029.
10.1021/jp074609z.
[5] H. Nishiura, T. Kobayashi, T. Miyama, A. Suzuki, S. Jung, K. Hayashi, R. Kinoshita,
[23] A.M. Piloto, D.S.M. Ribeiro, S.S.M. Rodrigues, C. Santos, J.L.M. Santos,
Y. Yang, B. Yuan, A.R. Akhmetzhanov, N.M. Linton, Estimation of the asymptomatic
M.G.F. Sales, Plastic antibodies tailored on quantum dots for an optical detection of
ratio of novel coronavirus infections (COVID-19), Int. J. Infect. Dis. 94 (2020)
myoglobin down to the femtomolar range, Sci. Rep. 8 (2018) 1–11, https://doi.org/
154–155, https://doi.org/10.1016/j.ijid.2020.03.020.
10.1038/s41598-018-23271-z.
[6] B. Sun, Y. Feng, X. Mo, P. Zheng, Q. Wang, P. Peng, X. Liu, Z. Chen, H. Huang,
[24] Q. Liu, X. Lu, J. Li, X. Yao, J. Li, Direct electrochemistry of glucose oxidase and
F. Zhang, W. Luo, X. Niu, P. Hu, L. Wang, H. Peng, L. Feng, F. Li, F. Zhang, F. Li,
electrochemical biosensing of glucose on quantum dots/carbon nanotubes
N. Zhong, L. Chen, Kinetics of SARS-CoV-2 specific IgM and IgG responses in
electrodes, Biosens. Bioelectron. 22 (2007) 3203–3209, https://doi.org/10.1016/
COVID-19 patients, Emerg. Microb. Infect. 9 (2020) 940–948, https://doi.org/
j.bios.2007.02.013.
10.1080/22221751.2020.1762515.
[25] M.A. Macdonald, H.A. Andreas, Method for equivalent circuit determination for
[7] K.K.-W. To, O.T.-Y. Tsang, W.-S. Leung, A.R. Tam, T.-C. Wu, D.C. Lung, C.C.-Y. Yip,
electrochemical impedance spectroscopy data of protein adsorption on solid
J.-P. Cai, J.M.-C. Chan, T.S.-H. Chik, D.P.-L. Lau, C.Y.-C. Cho, L.-L. Chen, W.-
surfaces, Electrochim. Acta 129 (2014) 290–299, https://doi.org/10.1016/
M. Chan, K.-H. Chan, J.D. Ip, A.C.-K. Ng, R.W.-S. Poon, C.-T. Luo, V.C.-C. Cheng,
j.electacta.2014.02.046.
J.F.-W. Chan, I.F.-N. Hung, Z. Chen, H. Chen, K.-Y. Yuen, Temporal profiles of viral
[26] E.B. Bahadir, M.K. Sezgintürk, A review on impedimetric biosensors, Artif. Cells
load in posterior oropharyngeal saliva samples and serum antibody responses
Nanomed. Biotechnol. 44 (2016) 248–262, https://doi.org/10.3109/
during infection by SARS-CoV-2 : an observational cohort study, Lancet Infect. Dis.
21691401.2014.942456.
20 (2020) 565–574, https://doi.org/10.1016/S1473-3099(20)30196-1.
[27] F. Lisdat, D. Sch€afer, The use of electrochemical impedance spectroscopy for
[8] J.D. Whitman, J. Hiatt, C.T. Mowery, B.R. Shy, R. Yu, T.N. Yamamoto, U. Rathore,
biosensing, Anal. Bioanal. Chem. 391 (2008) 1555–1567, https://doi.org/10.1007/
G.M. Goldgof, C. Whitty, J.M. Woo, A.E. Gallman, T.E. Miller, A.G. Levine,
s00216-008-1970-7.
D.N. Nguyen, S.P. Bapat, J. Balcerek, S.A. Bylsma, A.M. Lyons, S. Li, A.W. Wong,

9
A.R. Cardoso et al. Materials Today Bio 16 (2022) 100354

[28] N.O. Laschuk, E.B. Easton, O.V. Zenkina, Reducing the resistance for the use of antibodies, J. Virol. Methods 293 (2021) 1–7, https://doi.org/10.1016/
electrochemical impedance spectroscopy analysis in materials chemistry, RSC Adv. j.jviromet.2021.114173.
11 (2021) 27925–27936, https://doi.org/10.1039/d1ra03785d. [34] K. Jiang, L.S. Schadler, R.W. Siegel, X. Zhang, M. Terrones, Protein immobilization
[29] G. Cabral-miranda, A.R. Cardoso, L.C.S. Ferreira, M.G.F. Sales, M.F. Bachmann, on carbon nanotubes via a two-step process of diimide-activated amidation,
Biosensor-based selective detection of Zika virus speci fi c antibodies in infected J. Mater. Chem. 14 (2004) 37–39.
individuals, Biosens. Bioelectron 113 (2018) 101–107, https://doi.org/10.1016/ [35] A.I. Lopez-Lorente, B.M. Simonet, M. Valcarcel, Raman spectroscopic
j.bios.2018.04.058. characterization of single walled carbon nanotubes : in fl uence of the sample
[30] A.R. Cardoso, G. Cabral-miranda, A. Reyes-sandoval, M.F. Bachmann, M.G.F. Sales, aggregation state, Analyst 139 (2014) 290–298, https://doi.org/10.1039/
Detecting circulating antibodies by controlled surface modi fi cation with speci fi c c3an00642e.
target proteins : application to malaria, Biosens. Bioelectron. 91 (2017) 833–841, [36] J.S. Terry, L. Br, M.S. Scherman, C.E. Mcalister, R. Perera, T. Schountz, B.J. Geiss,
https://doi.org/10.1016/j.bios.2017.01.031. Development of a SARS-CoV-2 nucleocapsid specific monoclonal antibody, Virology
[31] D. Harvey, in: K. Kane (Ed.), Modern Analytic Chemistry, first ed.Mod. Anal. Chem., 558 (2021) 28–37, https://doi.org/10.1016/j.virol.2021.01.003.
McGraw-Hill Higher Education, 2000, p. 797. [37] D. Brinati, A. Campagner, D. Ferrari, M. Locatelli, G. Banfi, F. Cabitza, Detection of
[32] N. Renard, S. Daniel, N. Cayet, M. Pecquet, F. Raymond, S. Pons, J. Lupo, COVID-19 infection from routine blood exams with machine learning: a feasibility
C. Tourneur, C. Pretis, G. Gerez, P. Blasco, M. Combe, I. Canova, M. Lesenechal, study, J. Med. Syst. 44 (2020) 1–12, https://doi.org/10.1007/s10916-020-01597-
F. Berthier, Performance characteristics of the vidas SARS-CoV-2 IgM and IgG 4.
serological assays, J. Clin. Microbiol. 59 (2021) 1–14. [38] M. Plebani, Errors in clinical laboratories or errors in laboratory medicine? Clin.

[33] M. Bubonja-Sonje, L. Baticic, M. Abram, Đ.C. Grbesa, Diagnostic accuracy of three Chem. Lab. Med. 44 (2006) 750–759, https://doi.org/10.1515/CCLM.2006.123.
SARS-CoV2 antibody detection assays , neutralizing effect and longevity of serum

10