Determination of rSpike Protein by Specific Antibodies with Screen-Printed Carbon Electrode Modified by Electrodeposited Gold Nanostructures

In this research, we assessed the applicability of electrochemical sensing techniques for detecting specific antibodies against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike proteins in the blood serum of patient samples following coronavirus disease 2019 (COVID-19). Herein, screen-printed carbon electrodes (SPCE) with electrodeposited gold nanostructures (AuNS) were modified with L-Cysteine for further covalent immobilization of recombinant SARS-CoV-2 spike proteins (rSpike). The affinity interactions of the rSpike protein with specific antibodies against this protein (anti-rSpike) were assessed using cyclic voltammetry (CV) and differential pulse voltammetry (DPV) methods. It was revealed that the SPCE electroactive surface area increased from 1.49 ± 0.02 cm2 to 1.82 ± 0.01 cm2 when AuNS were electrodeposited, and the value of the heterogeneous electron transfer rate constant (k0) changed from 6.30 × 10−5 to 14.56 × 10−5. The performance of the developed electrochemical immunosensor was evaluated by calculating the limit of detection and limit of quantification, giving values of 0.27 nM and 0.81 nM for CV and 0.14 nM and 0.42 nM for DPV. Furthermore, a specificity test was performed with a solution of antibodies against bovine serum albumin as the control aliquot, which was used to assess nonspecific binding, and this evaluation revealed that the developed rSpike-based sensor exhibits low nonspecific binding towards anti-rSpike antibodies.


Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a highly transmissible and pathogenic coronavirus that first appeared in late 2019 and has since created a pandemic of acute respiratory sickness known as 'coronavirus disease 2019' (COVID- 19), which poses a threat to human health since human-to-human transmission has grown significantly [1]. The progression of the COVID-19 pandemic has shown that there is a crucial need to develop quick and accurate tests to better control the spread of the disease process or analyte-related CV features can be utilized for quantitative findings; however, due to its limitations, EIS is more typically used for exploratory purposes such as assessing the redox process for diverse analytes [28]. In general, pulse techniques such as DPV are more sensitive than linear-sweep-based methods, since CV is the technique most frequently employed for exploratory purposes. Thus, it is rather common in sensor development to employ both these techniques, because CV provides critical information on aspects such as process reversibility and the types of redox processes occurring during the analysis at the interface between the electrode and the solution, whereas potential-pulse-based techniques sometimes enable simplification of the quantification of the analyte [29]. The miniaturization of electrochemical systems enables the determination of protein-based analytes in rather small volumes of aliquots [30].
Therefore, in our present work we compare the applicability of both these two voltametric sensing methods (namely, DPV and CV), taking into account the advantages of their durability and low detection limits in small volumes of aliquots.
The SARS-CoV-2 recombinant spike protein of SARS-CoV-2 (rSpike) was produced by Baltymas (Vilnius, Lithuania) [31]. Serum samples containing antibodies (anti-rSpike) of volunteers vaccinated with a single dose of the Vaxzevria vaccine who had COVID-19 after two weeks were collected [10] according to Lithuanian ethics law. The ethics committee's permission was not required for this project (as confirmed by the Vilnius Regional Biomedical Research Ethics Committee).

Electrochemical Measurements
Electrochemical characterization of the working surface was performed using a potentiostat controlled by the DStat interface software from Wheeler Microfluidics Lab (University of Toronto, Toronto, ON, Canada). DRP-110 screen-printed carbon electrode systems (SPCEs), which are based on a working electrode (geometric area of 0.126 cm 2 ), a carbon counter, and Ag/AgCl reference electrodes, were purchased from Metrohm DropSens (Oviedo, Spain). SPCEs were connected through a specialized 'box-connector' for screen-printed electrodes (DRP-DSC, DropSens, Oviedo, Spain).
All electrochemical measurements were performed in 0.1 M PBS, pH 7.4 solution, adding 2 mM K 3 Fe(CN) 6 /K 4 Fe(CN) 6 ([Fe(CN 6 )] 3−/4− ) solution as a redox probe. Electrochemical characterization of the working electrode at different modification stages was carried out using DPV and CV. DPV experiments were measured in the potential range from −0.4 to +0.6 V vs. Ag/AgCl, with a step size of 0.004 V. CV was registered in the potential window from −0.4 to +0.6 V vs. Ag/AgCl, at a scan rate of 0.05 V/s. All experiments were performed at room temperature (20 • C).
Scanning electron microscope (SEM) images were acquired with a scanning electron microscope (Hitachi-70 S3400 N VP-SEM).

Au Deposition on SPCE
The SPCE was covered with 100 µL of the solution containing 0.1 M KNO 3 and 5 mM HAuCl 4 . Electrodeposition was performed at a potential of -0.4 V for 60 s. Then, after AuNS deposition on the SPCE (SPCE/AuNS), the electrode was rinsed with deionized water and dried under a N 2 (%) flow ( Figure 1, step 1). All electrochemical measurements were performed in 0.1 M PBS, pH 7.4 solution, adding 2 mM K3Fe(CN)6/K4Fe(CN)6 ([Fe(CN6)] 3−/4− ) solution as a redox probe. Electrochemical characterization of the working electrode at different modification stages was carried out using DPV and CV. DPV experiments were measured in the potential range from −0.4 to +0.6 V vs. Ag/AgCl, with a step size of 0.004 V. CV was registered in the potential window from −0.4 to +0.6 V vs. Ag/AgCl, at a scan rate of 0.05 V/s. All experiments were performed at room temperature (20 °C).
Scanning electron microscope (SEM) images were acquired with a scanning electron microscope (Hitachi-70 S3400 N VP-SEM).

Au Deposition on SPCE
The SPCE was covered with 100 µL of the solution containing 0.1 M KNO3 and 5 mM HAuCl4. Electrodeposition was performed at a potential of -0.4 V for 60 s. Then, after AuNS deposition on the SPCE (SPCE/AuNS), the electrode was rinsed with deionized water and dried under a N2 (%) flow ( Figure 1, step 1).

Immobilisation of rSpike and Anti-rSpike
The SPCE/AuNS were incubated at 20 °C for 4 h in 5 mM L-Cysteine ethanolic solution to form a self-assembled monolayer (SAM) on the working surface (SPCE/AuNS/SAM) ( Figure 1, step 2). After incubation, the SPCE/AuNS/SAM electrode was rinsed with deionized water and then dried under a N2 flow. SPCE/AuNS/SAM was activated with 10 µL of a mixture of 0.02 M EDC and 0.005 M NHS in PBS, pH 7.4, for 10 min. After the activation, the electrode was incubated with 10 µL of 50 µg/mL rSpike in PBS, pH 7.4, at 20 °C for 20 min. Immobilization of rSpike was performed through covalent coupling of the protein's primary amine functional groups and the activated carboxylic groups of the SAM (SPCE/AuNS/SAM/rSpike) ( Figure 1, step 3). The remaining reactive esters were deactivated by incubating with a 1 mM water solution of ethanolamine for 10 min. Afterwards, SPCE/AuNS/SAM/rSpike was incubated with 10 µL of anti-rSpike

Immobilisation of rSpike and Anti-rSpike
The SPCE/AuNS were incubated at 20 • C for 4 h in 5 mM L-Cysteine ethanolic solution to form a self-assembled monolayer (SAM) on the working surface (SPCE/AuNS/SAM) ( Figure 1, step 2). After incubation, the SPCE/AuNS/SAM electrode was rinsed with deionized water and then dried under a N 2 flow. SPCE/AuNS/SAM was activated with 10 µL of a mixture of 0.02 M EDC and 0.005 M NHS in PBS, pH 7.4, for 10 min. After the activation, the electrode was incubated with 10 µL of 50 µg/mL rSpike in PBS, pH 7.4, at 20 • C for 20 min. Immobilization of rSpike was performed through covalent coupling of the protein's primary amine functional groups and the activated carboxylic groups of the SAM (SPCE/AuNS/SAM/rSpike) ( Figure 1, step 3). The remaining reactive esters were deactivated by incubating with a 1 mM water solution of ethanolamine for 10 min. Afterwards, SPCE/AuNS/SAM/rSpike was incubated with 10 µL of anti-rSpike in PBS, pH 7.4, with a concentration range from 0.5 to 3.5 nM, at 20 • C for 10 min (SPCE/AuNS/SAM/rSpike/ anti-rSpike) ( Figure 1, step 4). After each stage of incubation, the system was rinsed with deionized water and used for further electrochemical measurements.

Calibration of Anti-rSpike
The initial number of binding antibody units (BAU) per mL against the spike protein of SARS-CoV-2 in the serum sample was 5860 BAU/mL. The concentration was defined by a chemiluminescent microparticle immunoassay performed in the laboratory of Tavo Klinika, Ltd. (Vilnius, Lithuania). The target antibodies in the sample were recalculated from BAU/mL to nM using the ratio 1 BAU/mL: 20 ng/mL (considering the molecular weight of immunoglobulin G as~150 kDa) [32][33][34].
Calibration curves were obtained by the incubation of SPCE/AuNS/SAM/rSpike in serum samples containing 0.5, 1.0, 1.5, 2.5, and 3.5 nM of anti-rSpike, for 10 min for each concentration. DPV and CV data were used to plot the calibration curves. The relative response (RR%) used for the evaluation of the method specificity was calculated using the equation RR% = ((X i − µX blank )/(X blank )) × 100%, where X i is the response for concentration i and X blank is the response for a blank.

Electrochemical Characterisation of SPCE and SPCE/AuNS
In order to improve the surface area for rSpike immobilization and to facilitate better electron transfer kinetics, electrochemical deposition of AuNS was performed on the SPCE working electrode. The CV and DPV results are provided in Figure 2. In addition, the electroactive surface area for SPCE/AuNS was determined using CV in 10 mM H 2 SO 4 ( Figure 3). The characteristic gold reduction and oxidation peaks are present in the potential window from 0 to +1.0 V [35], while the measurements for the unmodified SPCE surface reveal no oxidation or reduction peaks (Figure 3, inset).
With the aim of evaluating the electrochemical performance of the sensor, it is critical to quantify the electrochemically active surface area of the substrate material [36], as well as to define the heterogeneous electron transfer rate constant (k 0 ) [37]. For this purpose, CV at a range of scan rates from 0.01 to 0.15 V/s was performed in PBS, pH 7.4, containing 2 mM [Fe(CN 6 )] 3−/4− for both SPCE and SPCE/AuNS ( Figure 4, Table 1).
Using the Randles-Sevcik equation, the electrochemically active surface areas were calculated as 1.49 ± 0.02 cm 2 for SPCE and 1.82 ± 0.01 cm 2 for SPCE/AuNS ( Figure 5A). The difference between the values can be explained by the increase in the surface roughness ( Figure 6), thus improving the working substrate properties for the subsequent immobilization of the biorecognition element. Furthermore, the data obtained from CV at different scan rates allowed us to assess k 0 by means of the improved Nicholson's approach for the quasi-reversible electrochemical reaction [38,39]. The value for SPCE was 6.30 ± 0.13 × 10 −5 , while that for SPCE/AuNS was 14.56 ± 0.20 × 10 −5 ( Figure 5B), which is more than twice as high. Thus, it can be concluded that the electrodeposition of AuNS contributes not only to an increase in the electrode active area but also to the rate of heterogeneous electron transfer. electron transfer kinetics, electrochemical deposition of AuNS was performed SPCE working electrode. The CV and DPV results are provided in Figure 2. In ad the electroactive surface area for SPCE/AuNS was determined using CV in 10 mM ( Figure 3). The characteristic gold reduction and oxidation peaks are present in the tial window from 0 to +1.0 V [35], while the measurements for the unmodified SPC face reveal no oxidation or reduction peaks (Figure 3, inset).    With the aim of evaluating the electrochemical performance of the sensor, it is critical to quantify the electrochemically active surface area of the substrate material [36], as well as to define the heterogeneous electron transfer rate constant (k 0 ) [37]. For this purpose, CV at a range of scan rates from 0.01 to 0.15 V/s was performed in PBS, pH 7.4, containing 2 mM [Fe(CN6)] 3−/4− for both SPCE and SPCE/AuNS ( Figure 4, Table 1).     scan rates allowed us to assess k by means of the improved Nicholson's approach quasi-reversible electrochemical reaction [38,39]. The value for SPCE was 6.30 ± 0.13 while that for SPCE/AuNS was 14.56 ± 0.20 × 10 −5 ( Figure 5B), which is more than tw high. Thus, it can be concluded that the electrodeposition of AuNS contributes no to an increase in the electrode active area but also to the rate of heterogeneous e transfer. DPV is known to be a potentiostatic method, suggesting some advantages over conventional methods such as CV. In the waveform, DPV is a series of pulses, while for CV the potential is ramped linearly with time. Due to the minimization of the capacitive current, pulse methods, including DPV, are considered to be more sensitive than linear sweep methods. On the other hand, CV is the method most frequently used for research purposes. Hence, it is quite a common practice in sensor development to use both types of electrochemical methods. While CV reveals key electrochemical characteristics such as process reversibility and reflects the redox processes that occur in the system, DPV is employed for quantitative analysis [40].
Since the obtained cyclic voltammograms were quasi-reversible [41], the character of the correlation between the current peak intensity and the surface modification step was not the same for cathodic and anodic peaks. For instance, in Figure 2, the resolution of the current density signals in the anodic region was higher than in cathodic region. This trend increased with further surface modification, leading to the overlapping of the cathodic peaks ( Figure 7). Hence, to facilitate quantitative data analysis, we used the values of the anodic current density (jpa) as the analytical parameter gained from the CV experiments.
As shown in Figure 2, cyclic and differential pulse voltammograms revealed the same trend of increasing current densities after the working surface modification. Specifically, the values increased from 394.71 ± 0.69 to 536.30 ± 0.42 and from 274.89 ± 0.17 to 632.53 ± 0.83 µA/cm 2 for CV and DPV, respectively. Potential values were also changed, moving left along the axis. This indicates a substrate material change with increasing the conductivity. DPV is known to be a potentiostatic method, suggesting some advantages over conventional methods such as CV. In the waveform, DPV is a series of pulses, while for CV the potential is ramped linearly with time. Due to the minimization of the capacitive current, pulse methods, including DPV, are considered to be more sensitive than linear sweep methods. On the other hand, CV is the method most frequently used for research purposes. Hence, it is quite a common practice in sensor development to use both types of electrochemical methods. While CV reveals key electrochemical characteristics such as process reversibility and reflects the redox processes that occur in the system, DPV is employed for quantitative analysis [40].
Since the obtained cyclic voltammograms were quasi-reversible [41], the character of the correlation between the current peak intensity and the surface modification step was not the same for cathodic and anodic peaks. For instance, in Figure 2, the resolution of the current density signals in the anodic region was higher than in cathodic region. This trend increased with further surface modification, leading to the overlapping of the cathodic peaks ( Figure 7). Hence, to facilitate quantitative data analysis, we used the values of the anodic current density (j pa ) as the analytical parameter gained from the CV experiments.
As shown in Figure 2, cyclic and differential pulse voltammograms revealed the same trend of increasing current densities after the working surface modification. Specifically, the values increased from 394.71 ± 0.69 to 536.30 ± 0.42 and from 274.89 ± 0.17 to 632.53 ± 0.83 µA/cm 2 for CV and DPV, respectively. Potential values were also changed, moving left along the axis. This indicates a substrate material change with increasing the conductivity.  Table 2). The CV oxidation peaks were compared after each of the abovementioned stages of the biosensing element formation.    Table 2). The CV oxidation peaks were compa ter each of the above-mentioned stages of the biosensing element formation.  As considered in the previous section, CV for SPCE/AuNS was characterized by a voltammogram with sharp oxidative/reductive peaks and with a j pa value of 536.30 ± 0.42 µA/cm 2 . After SPCE/AuNS/SAM formation, a decrease in j pa to 436.96 ± 0.18 µA/cm 2 was observed. Then, the activation of the terminal -COOH group of the L-Cysteine took place without accompanying electrochemical measurements, to ensure subsequent effective rSpike immobilization. Afterwards, the remainder of the activated functional groups of the SAM were blocked by 1 mM ethanol amine, to avoid nonspecific interactions during the anti-rSpike coupling stages. CV after antigen immobilization with SPCE/AuNS/SAM/rSpike formation and blocking revealed a further current density decrease to 361.83 ± 0.28 µA/cm 2 .

Electrochemical Characterisation of the Biosensing Element
DPV measurements for the above-mentioned stages of biosensing element formation showed the same tendency toward a stepwise decrease in the current density to 632.53 ± 0.83, 363.52 ± 0.28, and 185.26 ± 1.17 µA/cm 2 for SPCE/AuNS, SPCE/AuNS/SAM, and SPCE/AuNS/SAM/rSpike. These results are summarized in Table 1.
The decrease in current density according to both CV and DPV methods can be explained by the increasing layer thickness on the working electrode surface, thus hampering electron transfer. The stepwise broadening of the DPV peaks could be related to a reduced electron exchange rate.
For CV measurements, the potential values for j pa moved within the 0.1-0.2 V window. Again, this could be related to alterations in the electron transfer process and/or to changes in the reference Ag/AgCl electrode, which is quite sensitive to experimental conditions such as the presence of Cl − in PBS, pH 7.4, during AuNS electrodeposition. At the same time, the DPV is characterized by rather stable potential value, changing only slightly in the range of 0.0 to 0.1 V, which is observed due to different nature of the electrochemical signal recording/assessment principles in the CV and DPV techniques.

Electrochemical Characterisation of the Anti-rSpike Detection
The next step of the experiment was to test the ability of the biosensor to detect anti-rSpike. For this purpose, SPCE/AuNS/SAM/rSpike was sequentially incubated with 10 µL of anti-rSpike in a concentration range from 0.5 to 3.5 nM. Each subsequent incubation was accompanied by CV and DPV measurements ( DPV experiments revealed the same effect, with a sequential decrease in j p , i.e., 185.26 ± 1.17, 148.86 ± 1.02, 124.25 ± 0.32, 105.86 ± 0.32, 82.23 ± 0.59, and 66.93 ± 0.2 µA/cm 2 for 0, 0.5, 1.0, 1.5, 2.5, and 3.5 nM, respectively. In contrast to CV-based experiments, the peaks of the differential pulse voltammograms for solutions with different concentrations of anti-Spike antibodies are characterized by higher resolution and more stable potential values, which correspond to particular concentrations of anti-Spike antibodies.

Limit of Detection and Limit of Quantification
Data gained from the performed electrochemical measurements were used to evaluate the limit of detection (LOD) and limit of quantification (LOQ) for the developed immunosensor, using both the CV and DPV methods. The j pa and j p values were used as analytical signals for CV and DPV, respectively. Figure 9 shows the calibration curves.
Biosensors 2022, 12, x FOR PEER REVIEW 0.5, 1.0, 1.5, 2.5, and 3.5 nM, respectively. In contrast to CV-based experiments, th of the differential pulse voltammograms for solutions with different concentra anti-Spike antibodies are characterized by higher resolution and more stable poten ues, which correspond to particular concentrations of anti-Spike antibodies.

Limit of Detection and Limit of Quantification
Data gained from the performed electrochemical measurements were used t ate the limit of detection (LOD) and limit of quantification (LOQ) for the develo munosensor, using both the CV and DPV methods. The jpa and jp values were analytical signals for CV and DPV, respectively. Figure 9 shows the calibration cu The LOD was calculated as 3.33σ/s and LOQ was calculated as 10σ/s, where standard deviation for the blank response and s is the slope of the calibration cu It was revealed that the LOD and LOQ values for the CV-based method were 0.27 0.81 nM, respectively, while the values calculated from DPV data were 0.14 nM a nM, respectively.

Specificity Test
The experiment for nonspecific binding on SPCE/AuNS/SAM/rSpike was pe by comparison of the relative electrochemical signal responses (initial values fro 2) after incubation of the electrode in 10 mM PBS, pH 7.4, with solutions of 1.5 n rSpike and 15 nM anti-BSA, (Figure 10). The comparison of the relative responses r that for CV, the RR(%) values were 17.80 ± 0.07% and 3.24 ± 0.46% for anti-rSpike a BSA, respectively. Similarly, the RR(%) values for the DPV method were 42.86 ± 0 anti-rSpike and 7.57 ± 0.09% for anti-BSA. The LOD was calculated as 3.33σ/s and LOQ was calculated as 10σ/s, where σ is the standard deviation for the blank response and s is the slope of the calibration curve [42]. It was revealed that the LOD and LOQ values for the CV-based method were 0.27 nM and 0.81 nM, respectively, while the values calculated from DPV data were 0.14 nM and 0.42 nM, respectively.

Specificity Test
The experiment for nonspecific binding on SPCE/AuNS/SAM/rSpike was performed by comparison of the relative electrochemical signal responses (initial values from Table 2) after incubation of the electrode in 10 mM PBS, pH 7.4, with solutions of 1.5 nM anti-rSpike and 15 nM anti-BSA, (Figure 10). The comparison of the relative responses revealed that for CV, the RR(%) values were 17.80 ± 0.07% and 3.24 ± 0.46% for anti-rSpike and anti-BSA, respectively. Similarly, the RR(%) values for the DPV method were 42.86 ± 0.32% for anti-rSpike and 7.57 ± 0.09% for anti-BSA.

Conclusions
In this work, electrochemical characterization of SPCE/rSpike and SPCE/AuNS/SAM/rSpike was performed. The electroactive surface area and the heterogeneous electron transfer rate constants were determined and were 22% and 131% higher for SPCE with electrodeposited AuNS, making the SPCE/AuNS surface more suitable for electrochemical measurements. The formation of the SPCE/AuNS/SAM/rSpike biosensing element, as well as the interaction between immobilized rSpike and anti-rSpike, were accompanied by CV and DPV measurements after key stages. For both detection methods, a stepwise decrease in current density was measured after each modification stage, including that applied for the detection of anti-rSpike occurring due to increasingly prohibited access of the [Fe(CN)6] 3−/4− redox probe to the working electrode. The DPV method was more reliable and more sensitive compared to CV, resulting in 48% lower LOD and LOQ values, making the DPV method more suitable for quantitative analysis. Specificity tests with anti-BSA showed low nonspecific binding for this antibody type. In conclusion, it is expected that the electrochemical immunosensor designed in this research will prove suitable for the diagnosis of the immunological response generated during the course of COVID-19 disease or after vaccination.

Conclusions
In this work, electrochemical characterization of SPCE/rSpike and SPCE/AuNS/SAM/ rSpike was performed. The electroactive surface area and the heterogeneous electron transfer rate constants were determined and were 22% and 131% higher for SPCE with electrodeposited AuNS, making the SPCE/AuNS surface more suitable for electrochemical measurements. The formation of the SPCE/AuNS/SAM/rSpike biosensing element, as well as the interaction between immobilized rSpike and anti-rSpike, were accompanied by CV and DPV measurements after key stages. For both detection methods, a stepwise decrease in current density was measured after each modification stage, including that applied for the detection of anti-rSpike occurring due to increasingly prohibited access of the [Fe(CN) 6 ] 3−/4− redox probe to the working electrode. The DPV method was more reliable and more sensitive compared to CV, resulting in 48% lower LOD and LOQ values, making the DPV method more suitable for quantitative analysis. Specificity tests with anti-BSA showed low nonspecific binding for this antibody type. In conclusion, it is expected that the electrochemical immunosensor designed in this research will prove suitable for the diagnosis of the immunological response generated during the course of COVID-19 disease or after vaccination.