Rapid Isolation of Anti-Idiotype Aptamers for Quantitation of Human Monoclonal Antibodies Against SARS-CoV-2 Spike Protein

Abstract

Therapeutic antibodies that block viral entry have already proven to be important, first line drugs for treatments of viral infections. In the case of SARS-CoV-2, combinations of multiple therapeutic antibodies may need to be rapidly identified and formulated in a way that blocks each new, predominant variant of the virus. For efficient introduction of any new antibody combination into patients, it is important to be able to monitor patient-specific pharmacokinetics of individual antibodies, which would include the time course of their specific capacity to block the viral spike proteins. Here, we present three examples of microfluidic-based rapid isolation of companion reagents useful for establishing combination antibody therapies. These reagents are specific three-dimensional imprints of variable regions of individual human monoclonal antibodies against the S protein of SARS-CoV-2 virus in the form of oligonucleotide-based ligands (aptamers). We implement these anti-idiotypic aptamers as bioreceptors in graphene-based field-effect transistor sensors to accomplish label free, rapid, and sensitive detection of matching antibodies within minutes. Through this work we have demonstrated the general applicability of anti-idiotype aptamers as capture reagents in quantification of active forms of monoclonal antibodies in complex biological mixtures.

1. Introduction

Therapeutic monoclonal antibodies (mAbs) play a well-established role in treating cancers and autoimmune disorders [1–3]. With global outbreak of coronavirus disease 2019 (COVID-19), caused by the SARS-CoV-2 virus, the significance of combinations of mAbs against its spike (S)-protein was also firmly demonstrated, as a therapeutic option for all individuals at increased risk, those who cannot receive vaccines, or before efficient vaccines become available [4, 5]. The new variants of virus continue to emerge in regular intervals, thus, reformulations of therapeutic combinations from libraries of existing monoclonal antibodies might be necessary to combat variants that evade all previous antibody combinations or existing vaccines. In this context and using our original hybridoma fusion-based approach to generate fully human monoclonal antibodies [6], we have been continuously collecting peripheral blood lymphocytes from patients who survived different variants of SARS-CoV-2, and use them to generate a broad inventory of S-protein specific mAbs. Each new combination of therapeutic antibodies would, however, require rapid, yet full pharmacokinetic characterization of individual antibodies in mixtures, both in healthy volunteers and initial patients. To achieve this, we would require companion reagents that specifically recognize individual human antibodies in these mixtures, against the background of all other immunoglobulins that are present in serum at the same time. With this approach on mind, we describe here a general, highly practical protocol to isolate companion anti-idiotype aptamers [7, 8], which are oligonucleotide-based ligands specific to individual human antibodies that, here, bind to viral S-protein.

Idiotypes refer to molecular structures within variable regions of antibodies and that also comprise their antigen binding sections. Thus, in principle, for each antibody, we could isolate companion idiotype-binding (“anti-idiotype”) reagents, while choosing those that would be specific for only that antibody (anti-idiotype reagents that are shared between non-analogous antibodies and that are true molecular imprints of immunogens are beyond the scope of the current work). The standard approaches to targeting idiotypes rely on either anti-idiotype antibodies or antigen mimotope peptides [9, 10], but these affinity reagents suffer from shortcomings such as lengthy manufacturing time (several months), high manufacturing costs, and batch-to-batch variability. Recently, aptamers have garnered considerable attention due to their unique advantages over protein-based affinity reagents, including ease of synthesis and modification, low cost, thermal stability, reversible binding capabilities, and no batch-to-batch variability [7, 11, 12]. Typically, aptamers are obtained through a process involving recursive affinity separation of binders, followed by their polymerase chain reaction (PCR) amplification from large libraries of random oligonucleotides. This process is commonly referred to as systematic evolution of ligands by exponential enrichment (SELEX) [13, 14]. While there has been an increasing number of aptamers reported for targeting idiotypes or variable regions of antibodies, traditional methods of aptamer isolation remain resource-intensive, time-consuming and laborious, often necessitating approximately 10 or more rounds of SELEX. For instance, Missailidis et al. isolated aptamers against an anti-MUC1 monoclonal antibody using a 10-round benchtop SELEX procedure with both affinity selection and PCR amplification performed in tubes [15]. Hu et al. obtained aptamers for an anti-saxitoxin antibody through 16 SELEX rounds with the antibody coated on the surface of a 96-well plate to which solution-borne library ssDNA molecules were exposed for affinity selection [16]. Similarly, Wang et al. isolated aptamers for anti-zearalenone antibody from 15 SELEX rounds with the antibody coated on 96-well plate surfaces for affinity selection [17]. Saito et al. isolated aptamers for a humanized monoclonal antibody targeting vascular endothelial growth factor A using 14 SELEX rounds, wherein the antibody was immobilized on magnetic beads and affinity selection was performed in tubes [18].

This paper presents a method for rapid and streamlined isolation of aptamers targeting idiotypes of mAbs specific to S-protein of SARS-CoV-2 virus, with potential applications in quality control in antibody manufacturing, pharmacokinetic characterization, and antibody monitoring for both healthy volunteers and just-infected patients. We employ microfluidic technology to streamline the process of isolating anti-idiotype aptamers targeting monoclonal antibodies for S protein. Aptamer isolation can be completed in a significantly reduced number of rounds of SELEX. In addition, we overcome the challenge of validating aptamers for specific idiotype-targeting by using transmission electron microscopy (TEM) to directly observe the interaction between aptamers and idiotype regions of monoclonal antibodies. Furthermore, achieving near real-time monitoring of mAb production and capturing rapid change of physiological mAb levels during therapy are in general difficult using conventional methods such as enzyme-linked immunosorbent assay (ELISA), which is often time-consuming (several hours, sometimes overnight) and labor-intensive. We combine the anti-idiotype aptamers and graphene-based affinity nanosensors to enable rapid, label-free, and sensitive quantitation of antibody concentration to inform timely decision-making in both manufacturing and therapeutic contexts. In our experiments, anti-idiotype aptamers with sub-nanomolar affinity toward anti-S protein monoclonal antibodies were isolated within just 5 rounds of SELEX within two days, which represented a significant improvement when compared to conventional methods whose completion requires 10–16 SELEX rounds in up to a month. Employed in graphene-based affinity nanosensors, these isolated aptamers then allowed rapid measurements of antibody concentration in human serum (tested in 4 minutes). With these characteristics, our anti-idiotype aptamers can potentially be used for monitoring of monoclonal antibody drugs for COVID-19.

2. Experimental Section

2.1. Generation of Fully Human Monoclonal Antibodies Against S-Protein of SARS-CoV-2 Virus

Convalescent sera and matching peripheral blood from survivors of COVID-19 were collected under institutional review board (IRB) approved protocol. Peripheral blood mononuclear cells (PMNCs) were isolated and fused with human hybridoma fusion partner cell line MFP-2 as described earlier with some modifications [6]. Briefly, PMNCs were depleted of T-cells using negative magnetic isolation (EasySep^™ Human T Cell Isolation Kit). B-cells were collected and expanded on mouse fibroblast cells L4.5, expressing human CD-4L (kind gift of Dr. Sonia Néron from the Département de Biochimie et Microbiologie, Université Laval, Sainte-Foy, Québec, Canada). Expanded B-cells were fused to MFP-2 and the resulting hybridomas were tested in ELISA using immobilized Spike protein (D164G variant). Selected hybridomas were clones twice, adopted to serum-free media and antibodies were isolated using Protein A/G isolation protocol (BioRad).

2.2. Protocol of Microfluidic Isolation of Anti-Idiotype Aptamers

Anti-idiotype aptamers against anti-S human monoclonal IgG1 antibodies, Am6H2, TNX1, and TNX7, were isolated using microfluidic SELEX. The microfluidic chip was shown in Figure S1. The microfluidic chip consists of six hexagonal chambers (each 7.5 mm long, 2.5 mm wide and 200 μm tall), in which microbeads were trapped via a weir-like structure. The hexagonal shape can minimize the trapping of bubbles. The weir structure made the height (20 μm) of outlet channel smaller than the diameter of microbeads (45~165 μm). Thus, these beads can be uniformly packed in the microscale chamber, allowing for efficient interaction of molecules in a flowing stream with bead-immobilized molecules. The device was fabricated using standard multilayer soft lithography techniques that was described elsewhere [19–21]. Briefly, SU8 and silicon wafers were used to create master molds bearing the device design. PDMS was then poured onto the master molds, cured, peeled, cut, and bonded onto the glass slide.

Five nmol of an initial randomized single-stranded DNA (ssDNA) library that was labelled with FAM (Fluorescein Amidite) was diluted in 200 μL of selection buffer and heated at 95°C for 10 min and immediately cooled at −20 °C for 4 min, followed by incubation at room temperature for 15 min to fold the ssDNA molecules. The prepared library was first incubated with 100 μL of polyclonal IgG-coated beads and 100 μL of protein A/G beads to remove the non-binders. Target IgG-coated beads were injected into the positive selection chamber, and a mixture of polyclonal IgG and protein A/G beads was injected into the counter selection chamber, where the beads were tripped by weir structures. The prepared library was introduced into counter selection chamber with a flow rate of 10 μL/min to remove the background binders (binding to protein A/G beads) and constant region binders in the ssDNA library. After flowing through the counter selection chamber, library was introduced into the positive selection chamber along a tube connecting the two chambers to incubate with the target IgG antibodies. Following this incubation, a 100 μL of washing buffer was introduced into the positive selection chamber at a flow rate of 20 μL/min to remove the weakly binding and nonbinding oligonucleotides. Upon completion of washing, the binding oligonucleotides were released by heating the positive selection chamber to 95 °C on a hotplate for 1 min and a 20 μL of elution was collected for PCR amplification. Asymmetric PCR (with a 20:1 ratio of forward primer to reverse primer) was performed in a 100 μL mixture containing 10 μL of the elution solution as a template, 1 μL of 100 μM FAM-labeled forward primer, 5 μL of 1 μM biotin-labeled reverse primer, 0.5 μL of 5 U/μL Gotaq DNA polymerase, 2 μL of 10 mM dNTPs, 8 μL of 25 mM MgCl₂, 20 μL of Gotaq reaction buffer, and 53.5 μL of nucleic acid free H₂O. Asymmetric PCR amplification with 25 cycles was performed on a thermal cycler using following parameters: an initial denaturation at 95 °C for 3 min followed by amplification cycles at 95 °C for 30 s, annealing at 57 °C for 30 s, extension at 72 °C for 40 s and final extension at 72 °C for 2 min. Then, buffer exchange was performed for the asymmetric PCR amplicons by using a 10 K MWCO centrifugal filter. Finally, a 100 μL of asymmetric amplicons in PBSM buffer was obtained and used for next-round selection by injecting (flow rate: 10 μL/min) into new counter selection chamber and positive selection chamber filled with the protein A/G and poly IgG-coated beads and target IgG antibody-coated beads, respectively. Meanwhile, affinity of enriched pools toward IgG beads was monitored by an inverted fluorescence microscopy (Carl Zeiss Microscopy, LLC) equipped with a X-Cite 120 LED Boost System. Once high-affinity pools were observed, the enriched pool of the last round was sequenced by the next-generation high-throughput sequencing technology (Azenta Life Sciences, New Jersey). The sequencing data was analyzed using the AptaSuite toolkit [22]. The secondary structures of selected aptamer candidates were predicted with the Mfold software (zuker algorithm) [23] at the specific conditions (165 mM Na⁺, 2 mM Mg²⁺, and 25°C). The motif analysis was performed by MEME Suite [24].

2.3. Determination of Aptamer Affinity and Specificity

The equilibrium dissociation constant ($K_{\mathrm{D}}$) for selected aptamer was determined using a bead-based fluorescent binding assay. The FAM-labelled aptamer candidates were diluted with PBSM buffer to several different concentrations (100 μL). These aptamers were heated at 95°C for 10 min, followed by cooling at −20°C for 4 min and then incubated for another 15 min at room temperature (This process was always done before using aptamers for all experiments below). IgG-coated beads (20 μL) were incubated with the folded aptamers. After incubation on a homemade rotator at room temperature for 30 min and washing with 200 μL washing buffer for three times, images were acquired by the inverted fluorescence microscopy. The average fluorescence intensity was quantified with ImageJ software. The $K_{\mathrm{D}}$ values were then calculated by fitting with $Y=B_{\mathrm{m}\mathrm{a}\mathrm{x}}X/\left(K_{\mathrm{D}}+X\right)$, where $Y$ is the fluorescence intensity, $X$ is the aptamer concentration and $B_{\mathrm{m}\mathrm{a}\mathrm{x}}$ is the maximum fluorescence intensity. The specificity of the aptamers against target IgG antibodies was also characterized with the assay by using other protein-coated beads as controls.

2.4. Aptamer-Based Pull-Down Assay

Biotinylated aptamers (30 μL, 100 μM) were incubated with 200 μL of streptavidin beads with gentle mixing end-over-end on a homemade rotator at room temperature for 30 min. After washing with PBSM buffer for six times, the aptamer-beads are suspended in 200 μL PBSM buffer, stored at 4 °C. Dual-aptamer sandwich assay was performed to verify that the target IgG antibody has been successfully captured by aptamer-bead. A 20 μg of IgG antibody spiked in 20 μL PBSM buffer was incubated with the aptamer-beads (5 μL, supernatant removed before incubating with antibodies) on a homemade rotator at room temperature for 30 min. After washing with PBSM buffer for three times, the beads were incubated with a FAM-labelled second aptamer (50 μL, 1 μM) at room temperature for 5 min. After washing with PBSM buffer for three times, the fluorescence images of these beads were carried out by fluorescence microscopy and the fluorescence intensity on the beads was determined by ImageJ. Furthermore, the IgG antibody captured on the beads was released to be characterized by non-reduced SDS PAGE. After washing with PBSM buffer for three times, the beads were heated to 95 °C for 5 min to release the captured IgG. The elution (24 μL) was mixed with 6 μL of 5X loading buffer (10% w/v SDS, 10 mM Dithiothreitol, 20% w/v Glycerol, 0.2 M Tris-HCL with pH 6.8, and 0.05% w/v bromophenolblue) and loaded into the non-reduced 10% SDS-PAGE acrylamide gel. The gel was run in running buffer (25 mM Tris-HCL, 200 mM Glycine, 0.1% w/v SDS) at 120 V until the bromophenolblue in the loading buffer had migrated to end of the gel (typically taking about 1 hour). The gel was then stained with SYPRO RUBY (Invitrogen/ThermoFisher Scientific) and destained overnight prior to photography. In addition, pull-down assay was carried out to prove that the aptamer is binding to fragment antigen-binding (Fab) region of IgG antibody. IgG antibodies were first digested into F(ab’)2 and Fc fragments by using the IdeS enzyme following the instruction of FragIT^™ kit (Cat. No. A0-FR6–010, Genovis). Then, 20 μg of digested IgG fragments were incubated with the aptamer-beads (50 μL, supernatant removed before incubating with antibodies) on a homemade rotator at room temperature for 30 min. After washing with PBSM buffer three times, the captured fragments on beads were released to be characterized by the non-reduced SDS-PAGE gel. The biotinylated reverse primer (/5Biosg/CG TAC AGT GCA CAT GAG GGT) was used as the control aptamer.

2.5. Electron Microscopy and Image Processing

Antibody and biotinylated aptamer-streptavidin were incubated at a molar ratio of 1:2 for 1 hour at room temperature, and then diluted to 0.013 mg/mL⁻¹ immediately before negative staining with 2% (w/v) uranyl formate. The negative-stain samples were imaged at room temperature with a Hitachi HT7800 electron microscope. Images were recorded using Leginon on a TVIPS F416 camera using a defocus of ~2 μm and a nominal magnification of 120,000×, resulting in a pixel size of 1.77 Å. A total of 18,406 particles were automatically picked from 114 images and windowed into 256×256-pixel images with program e2boxer_old.py of the EMAN2 software package. After reduction of the images to 64 × 64 pixels, the particles were centered, aligned to each other, and classified with the iterative stable alignment and clustering (ISAC2) procedure implemented in the SPHIRE software package, specifying 100 images per group and a pixel error threshold of 0.7. After 3 generations of ISAC, 100 classes were obtained, accounting for 11,551 particles (62.8% of the entire dataset). Averages of these classes were calculated using the original 256×256-pixel images.

2.6. Competitive Assay with S Protein

FAM-labelled aptamers were heated at 95 °C for 10 min, followed by cooling at −20 °C for 4 min and then incubated for another 15 min at room temperature. The bead-immobilized target IgG antibody (10 μL) was treated with 100 μL of 1 μM aptamer with gentle mixing end-over-end on a homemade rotator at RT for 30 min. After washing with PBSM buffer for three times and removal of supernatant, the beads were incubated with 30 μL of 0–500 nM of S protein with gentle mixing end-over-end on a homemade rotator at RT for 5 min. BSA was used as a control of the S protein. After washing with PBSM buffer for three times, the fluorescence images were obtained by an inverted fluorescence microscopy and the fluorescence intensity on beads were analyzed by ImageJ. To verify that the S protein was binding to the antibody, 50 μL of 1 μM Cy3-labelled anti-His tag aptamer [21] was used to label the S protein by binding to the His tag on the S protein.

2.7. Antibody Detection with Aptameric Graphene Nanosensor

To verify that the aptamer can be applied for antibody drug monitoring, graphene-based field-effect transistor (GFET) nanosensor was combined with the aptamers to detect the antibody in human serum. The nanosensor is a GFET that consists of a graphene-based conducting channel connected to gold drain, source, and gate electrodes. The aptamer specific to the target IgG antibody was immobilized on the graphene with a linker of 1-pyrenebutanoic acid succinimidyl ester (PASE). The fabrication of the graphene nanosensor was described elsewhere [25, 26]. Briefly, 50 nm of Cr/Au electrodes (3 nm/47 nm) were patterned onto the SiO₂ wafer as a substrate via lift-off photolithography processes. A monolayer graphene was transferred onto the drain-source electrode as conducting channel using a polymethyl methacrylate (PMMA) carrier layer. After removing the PMMA with acetone, PASE was immobilized on the graphene through π- π stacking as a linker to link the 5’ amino-conjugated aptamer. Then, the unreacted PASE was passivated by ethanolamine. Finally, an opening 5 mm in diameter was cut through a 2 mm-thick sheet of polydimethylsiloxane (PDMS), which was bonded reversibly to the nanosensor chip. The opening formed a well exposing the graphene to the liquid sample held therein during experiments. The antibodies with varying concentration were spiked in a 0.1× human serum and incubated with the sensor for 4 min before measurements. The electrical properties and transfer curves were characterized by two Keiteley 2400 digital sourcemeters under control of LabVIEW programs.

3. Results and Discussion

3.1. Microfluidic Isolation of Anti-Idiotype Aptamers

Microfluidic SELEX was conducted to isolate anti-idiotype aptamers for monoclonal antibodies Am6H2, TNX1, and TNX7. We adopted a SELEX procedure previously reported by our group [21], with some modifications to enhance the efficiency of anti-idiotype aptamer isolation. Specifically, a mixture of polyclonal IgG beads (NHS-activated beads) and protein A/G beads was used for counter selection (Figures 1a and S1). Here, NHS-activated beads were used to immobilize polyclonal IgG as counter targets because the covalent attachment allows for random immobilization instead of orientated immobilization of antibodies [27, 28] and thus the constant regions of the antibody are maximumly exposed to the ssDNA library, facilitating the removal of non-anti-idiotype aptamers. We monitored the selection process by characterizing the affinity of each enriched pool for IgG-coated beads in the chambers. The initial ssDNA library and forward primer were fluorescently labeled with FAM, allowing us to measure the fluorescence intensity of FAM-labeled ssDNA molecules binding to the antibody-coated beads. This enabled us to track the affinity of the ssDNA pools for these antibody-coated beads throughout each selection round. After three rounds of SELEX, fluorescence intensities on target antibody-coated beads increased significantly, while negligible changes were observed for counter target-coated beads (Figures 1b–d), indicating that enriched pools with high affinity and specificity to target antibodies were obtained. The fifth-round pool was then sequenced using next-generation sequencing technique and analyzed with the AptaSuite toolkit [22]. The sequences were aligned based on the number of copies out of a total of obtained sequences (53,147 for Am6H2, 38,476 for TNX1 and 69,355 for TNX7) after filtering sequences inconsistent with the library in lengths and primer regions. The first 20 most abundant sequences for antibody (Am6H2, TNX1, and TNX7) were listed in Tables S1–S3, respectively.

Figure 1.

Microfluidic isolation of anti-idiotype aptamers toward monoclonal IgG antibodies. (a) Schematic illustration of the microfluidic SELEX process, where counter selection was used to remove background aptamers bound to protein A/G beads and constant region of IgG antibodies. Fluorescence intensity of each enriched pool binding to bead-immobilized antibodies (b) Am6H2, (c) TNX1, and (d) TNX7, as well as counter targets, including polyclonal IgG and protein A/G.

3.2. Aptamer Identification

Aptamers were identified from the 20 most abundant sequences. A motif sequence was identified from these aptamer candidates against Am6H2 and TNX1 (Figures 2a and 2b), resulting in the truncation of nonessential nucleotides to obtain a 17mer aptamer (AH-AptT17) and a 37mer aptamer (TX1-AptT37), respectively (Figures 2d–2e and S2). No motif was obtained for TNX7, so the most abundant 5 sequences were chosen as aptamer candidates and charactered with the bead-based fluorescence assay. Aptamer TX7-Apt3 showed the highest fluorescence intensity (Figures 2c) and was truncated to obtain a 36mer aptamer, TX7-AptT36, based on its secondary structure (Figures 2f and S3). The affinity of these aptamers was evaluated, and the values of equilibrium dissociation constant were determined to be 32 ± 7 nM, 278 ± 39 nM, and 130 ± 25 nM for aptamers AH-AptT17, TX1-AptT37, and TX7-AptT36, respectively (Figures 2g–2i). The nano-molar level of $K_{\mathrm{D}}$ values suggests that these aptamers have strong affinity to their respective target antibodies. The dose of an antibody drug in a treatment varies considerably depending on the specific drug, the intended medical condition, and individual patient characteristics. Typically, a clinical trial involving hospitalized COVID-19 patients used a single intravenous dose ranging from 700 mg to 7000 mg [29]. This administration regimen resulted in mean serum concentrations of the antibody drug over a 29-day period, spanning from 34 μg/mL (227 nM) to 2103 μg/mL (14 μM). The $K_{\mathrm{D}}$ values of these aptamers are either lower or comparable to the clinical concentration range, indicating that the affinities of these aptamers are suitable for monitoring all phases of pharmacokinetics. Furthermore, specificity was characterized using polyclonal IgG, IgE, BSA, and protein A/G as control proteins. High fluorescence intensities were observed for target antibodies with their corresponding aptamers, while negligible fluorescence intensities were detected for control proteins (Figures 2j–2l). These results indicate that the aptamers have high specificity, and there is no cross-reactivity against the three target antibodies that are the same type of antibodies targeting S protein, suggesting that they bind to the idiotype region of antibodies.

Figure 2.

Identification and characterization of aptamers. Motif sequences analyzed from aptamer candidates against (a) Am6H2 and (b) TNX1, respectively. (c) Fluorescence intensity of the most abundant 5 aptamer candidates binding to TNX7. Predicted secondary structures of aptamers (d) AH-AptT17, (e) TX1-AptT37, and (f) TX7-AptT36, respectively. Dissociation constant determination of aptamers (g) AH-AptT17, (h) TX1-AptT37, and (i) TX7-AptT36, respectively. Specificity analysis of aptamers (j) AH-AptT17, (k) TX1-AptT37, and (l) TX7-AptT36, respectively. All measurements were performed in triplicate and the data were shown as the mean ± SD in the graph.

3.3. Aptamer-Based Pull-Down

Pull-down assay was performed to confirm the binding of aptamer to target antibody. Aptamer AH-AptT17 was chosen as a demonstration. Streptavidin beads were used to immobilize the aptamer, and a fluorescent sandwich assay was performed using a second FAM-labelled aptamer as a reporter to detect the captured antibody (Figure 3a). A higher fluorescence intensity was observed on AH-AptT17 beads that were used to pull down Am6H2, while negligible fluorescence intensity was detected on aptamer beads for pulling down polyclonal IgG and control aptamer beads for pulling down Am6H2 (Figure 3b). IgG proteins captured by the aptamer beads were eluted and analyzed by non-reduced SDS-PAGE, and a brighter band was observed at the IgG antibody size of 150 kDa for the AH-AptT17 beads, indicating successful capture of the target antibody (Figure 3c). Lanes 1 and 2 showed the IgG that was subjected to capture by the control aptamer-coated beads from polyclonal IgG and Am6H2 solution, respectively. In both cases, negligible bands were observed, indicating that neither polyclonal IgG nor Am6H2 was pulled down by the control aptamer-coated beads. Lane 3 showed the polyclonal IgG that was captured by the aptamer AH-AptT17-coated beads from polyclonal IgG solution, and again, negligible bands were observed in this lane. This indicated that the aptamer was unable to pull down polyclonal IgG. Therefore, these results confirmed that aptamer AH-AptT17 was specifically binding to the target antibody Am6H2.

Figure 3.

Aptamer-based pull-down assay. (a) Schematic illustration of the dual aptamer sandwich assay and (b) pull-down verification by measuring the fluorescence intensity on beads. All measurements were performed in triplicate and the data were shown as the mean ± SD. (c) SDS-PAGE analysis of pull-down of antibody Am6H2. (d) SDS-PAGE analysis of pull-down of digested antibody Am6H2. Lanes 1 and 2 showed the fragments pulled by the control aptamer-beads from digested polyclonal IgG and Am6H2 solution, respectively. Lanes 3 and 4 showed the fragments pulled by the aptamer-beads from digested polyclonal IgG and Am6H2 solution, respectively. Lanes 5 and 6 showed the digested polyclonal IgG and Am6H2, respectively.

To further confirm that the aptamer binds to the fragment antigen-binding (Fab) region, the antibody Am6H2 was digested into F(ab’)₂ and Fc fragments using IdeS protease, a highly specific IgG-degrading enzyme. The aptamer-coated beads were incubated with the digested IgG solution to capture the F(ab’)2 fragments, which were then released and characterized by non-reduced SDS-PAGE (Figure 3d). Two bands at 100 kDa and 25 kDa were observed in lanes 5 and 6, which indicate that the IgG has been successfully digested into F(ab’)2 and Fc fragments. A significantly brighter band at 100 kDa, corresponding to the size of F(ab’)2, was only observed for the digested Am6H2 pulled down by aptamer AH-AptT17 beads (lane 4), indicating that only the F(ab’)2 fragment has been pulled down by aptamer AH-AptT17 beads and the Fc fragment has been removed. This suggested that the aptamer AH-AptT17-coated beads were binding to F(ab’)₂ fragment rather than the Fc fragment of Am6H2. Furthermore, no bands were observed in lanes 1 and 2, indicating that the control aptamer-coated beads did not pull down any fragments from the digested polyclonal IgG and digested Am6H2 solution. Moreover, no bands were observed in lane 3, indicating that the aptamer AH-AptT17-coated beads did not capture any fragments from the digested polyclonal IgG solution. This provided further evidence that the aptamer AH-AptT17-coated beads did not bind to polyclonal IgG. These consistent results confirmed the specific binding of aptamer AH-AptT17 to the Fab region of the target antibody Am6H2.

3.4. Negative Stain Transmission Electron Microscopy (TEM) Imaging

Negative stain TEM imaging was employed to directly visualize the binding of the aptamers to their respect target antibodies. Since the size of aptamer is too small to be visualized by TEM imaging, the aptamers were linked to streptavidin protein via biotin-streptavidin interaction. We used single-biotin-labeled aptamer AH-AptT17 linked with streptavidin to study its binding to antibody Am6H2. We observed the attachment of the aptamer-linked streptavidin to the Fab region of the antibody, providing evidence of the binding of the aptamer to the Fab region (Figures 4a and S4). However, the yield of streptavidin-aptamer-antibody complexes was low, possibly due to the binding region of antibodies being occupied by some free aptamers that were not able to be linked with streptavidin. As the research progressed, our group identified two additional antibodies, TNX1 and TNX7. Since TNX7 exhibited the best affinity to the S protein, our research further prioritized TNX7, and we then utilized TEM to comprehensively characterize the aptamer binding to TNX7. To improve the yield, we used dual-biotin-labelled aptamer, which can increase the efficiency of streptavidin binding. We used this approach to visualize the binding between antibody TNX7 and aptamer TX7-AptT36, resulting in an improved yield of streptavidin-aptamer-antibody complexes compared to that of the single-biotin-labeled aptamer (Figures 4b and S5). We also observed two distinct binding conformations, one in which the aptamer-linked streptavidin binds to one of the Fab and the other in which the streptavidin is located between two Fabs. As the second conformation is unlikely a different view of the first conformation in the TEM experiment, this is possibly due to the different number of dual-biotin-conjugated aptamers linked to streptavidin (Figure 4c). Overall, our findings directly confirmed the specific binding of aptamers to the Fab region of antibodies.

Figure 4.

Negative stain TEM 2D class averages showing the antibody-aptamer-streptavidin complexes. (a) 2D class averages of the complex between antibody Am6H2 and aptamer AH-AptT17-linked streptavidin. The green boxes indicate those antibodies that are bound to streptavidin. (b) 2D class averages of the complex between antibody TNX7 and aptamer TX7-AptT36-linked streptavidin. (c) 2D class averages showing two distinct binding conformations between the aptamer and antibody in the complex between antibody TNX7 and aptamer TX7-AptT36-linked streptavidin.

3.5. Competitive Assay with S Protein

To determine the exact binding site of the aptamer in the idiotype region of antibody, direct competition with S protein was performed using a bead-based fluorescence microscope assay. The assay involved saturating bead-immobilized antibodies with FAM-labelled aptamers, followed by treatment with varying concentrations of S protein. If the fluorescence signal on the beads decreases, it indicates that S protein can compete out aptamer. This means paratope-specific binding of the aptamer, whereas no significant change indicates non-paratope binding (Figure 5a). The coexistence of the aptamer and S protein on the antibody (Figure 5b) and no significant change in fluorescence intensity when S protein was added suggested non-paratope binding of the aptamers to Am6H2 and TNX1 (Figures 5c and 5d). A significant decrease in fluorescence signal upon addition of S protein was observed for the binding between aptamer TX7-AptT36 and TNX7 (Figure 5e), indicating that the S protein competes with the aptamer, and the aptamer binds to the paratope region of TNX7. As each mAb has two independent antigen-binding sites, if both sites bind to S proteins, it indicates a fully bound form. If only one site binds to S protein, the mAb can potentially still bind to another S protein, indicating a partially bound (or partially free) form. In human body’s systemic circulation, various forms of mAb can coexist, including the free (unbound), partially bound, and fully bound variants. As the free and partially bound mAb forms are bioactive, assessment of their concentration offers a more accurate estimate of efficacious concentration and safety margin [30]. Given the competition between the S protein and aptamer TX7-AptT36, this specific aptamer can, in principle, serve as a tool for measuring both free and partially S-bound antibody TNX7.

Figure 5.

Aptamer-based competitive assay with S protein. (a) Schematic illustration of the experimental design. (b) Fluorescence images of Am6H2-coated beads binding to FAM-labelled aptamer AH-AptT17 in the absence of S protein and in the presence of 500 nM S protein. The S protein was labeled with a Cy3-conjugated anti-His tag aptamer. Scale bar is 200 μm. Normalized fluorescence intensity of aptamers (c) AH-AptT17, (d) TX1-AptT37, and (e) TX7-Apt36 binding to antibodies Am6H2, TNX1, and TNX7, respectively, after treated with S protein as competitor with varying concentrations (0–500 nM). BSA was used as a control competitor. All measurements were performed in triplicate and the data are shown as the mean ± SD in the graph.

3.6. Antibody Detection by Aptameric Graphene Nanosensor

To accurately evaluate the efficacy of antibody therapy in human serum, a sensing platform that is highly sensitive, specific and capable of rapid detection is required. These aptamers were combined with graphene field-effect transistor nanosensors, which were developed by our group previously [25, 26], to enable the specific detection of antibodies in complex human samples. Compared to antibodies, aptamers have a smaller size, which can improve the detection sensitivity of electronic-based biosensors, especially in nanomaterial-based field-effect transistor biosensors, by alleviating the shielding effect created by the electrical double layer [31, 32]. The GFET device consists of a graphene-based conducting channel that connects to gold drain source and gate electrodes (Figure S6). During operation, the electric double layer at the interface of graphene and electrolyte medium serves as the gate dielectric. The specific binding between the aptamer and the target antibody results in a change in the drain-source current in the graphene, which can be measured to determine the concentration of antibody in the sample (Figure 6a).

Figure 6.

Measurement of IgG spiked in human serum by aptameric graphene-based affinity nanosensor. (a) Schematic of graphene nanosensor with modification of aptamers. Dirac point shift $\left(\mathrm{\Delta }{V}_{\text{Dirac}}\right)$ of the device modified with aptamers AH-Apt17, TX1-Apt37, and TX7-Apt36 when binding to target antibodies (d) Am6H2, (e) TNX1, and (f) TNX7 at varying concentrations. Error bars were determined by the standard deviation of three device measurements.

As the devices were exposed to 0.1× diluted human serum solutions spiked with increasing concentrations of target antibodies, the Dirac point $\left.(V_{\text{dirac}}\right)$ decreased, indicating that the binding between the aptamer and antibody induced n-type doping to the graphene. In contrast, there was negligible shift in the Dirac point when the nanosensor was exposed to polyclonal IgG (Figure S7), indicating high specificity and the ability to distinguish specific target IgG from polyclonal IgG in a complex human matrix. The Dirac point shift, $\mathrm{\Delta }{V}_{\text{dirac}}\left(=V_{\text{dirac}}-V_{\text{dirac,}0}\right)$, was plotted against varying concentrations of IgG, where $V_{\text{dirac,}0}$ is the Dirac point when the nanosensor was exposed to 0.1× human serum solution without target IgG antibody. Concentration-dependent increase in $\mathrm{\Delta }{V}_{\text{dirac}}$ was observed for target antibodies Am6H2, TNX1, and TNX7, when binding to their respective aptamers AH-Apt17, TX1-Apt37, and TX7-Apt36 (Figures 6b–d). The data was well fitted by the Hill-Langmuir binding equation, $\mathrm{\Delta }{V}_{\text{dirac}}=A{c}^n/\left(K^n+c^n\right)$, where $A$ is the maximum response with all binding site occupied, $c$ is the concentration of the applied IgG, $K$ is the effective dissociation constant, and $n$ is the Hill coefficient [33, 34]. The $K_{\mathrm{D}}$ values obtained from the binding curves of the aptamers on the GFET platform, namely 13 nM, 243 nM, and 123 nM for aptamers AH-Apt17, TX1-Apt37, and TX7-Apt36, respectively, are indeed in close alignment with the $K_{\mathrm{D}}$ values shown in Figure 2, which are 32 nM, 278 nM, and 130 nM, respectively. The slight discrepancies between these values are reasonable and can be attributed to the distinct methodologies employed. For example, the immobilization of aptamers on the GFET sensor surface, as opposed to having them free in solution in the fluorescence assay in Figure 2, may contribute to the variation in $K_{\mathrm{D}}$ values. The values of limit of detection (LOD), obtained by limit of blank plus 1.645 standard deviations of the lowest measured concentration [35], were 1.1 pM (0.17 ng/mL), 12.5 pM (1.9 ng/mL), and 55 pM (8.3 ng/mL), respectively. These values are better or comparable to the LOD of commercial ELISA kits (10~11 ng/mL) [36, 37] and significantly lower than that of peptide-based ELISA (5000 ng/mL) [38]. However, ELISA is time-consuming (several hours), labor-intensive, and expensive (anti-idiotype antibodies are costly). The aptamer-based GFET sensor can achieve simple and rapid measurement (tested within 4 minutes). Typically, the administration regimen in clinical trials could result in mean serum concentrations of the antibody drug ranging from the nanomolar level to the micromolar level [29]. The clinical concentration range significantly exceeds the limits of detection values (1~55 pM) of our GFET sensors. This suggests that aptamer-based GFETs have the capability to detect clinically relevant concentrations, even with potential dilution factors ranging from 10 to 1000 times.

We used human serum without the target antibody as baseline for our sensors. In this serum, there is already a whole population of IgG. Our sensors can distinguish human serum samples spiked with the target antibody at a low concentration of 0.01 nM from those without the target antibody. This demonstrates our sensors’ capability to detect a single target antibody within the context of the entire IgG population in the human body. In real-world scenarios, the immunogenicity of antibody drugs may prompt the human body to generate anti-drug polyclonal antibodies. However, in the case of our fully humanized antibody drugs, the formation of anti-drug polyclonal antibodies may be prevented. We acknowledge the need for further investigation and validation in a clinical context to fully understand these antibodies’ limitations and capabilities. To account for the possibility of polyclonal anti-drug antibody generation, we have determined the limit of detection values for this situation, which are 13 pM, 197 pM, and 54 pM for target antibodies, Am6H2, TNX1, and TNX7, respectively. In addition, there is a possibility that the aptamers may bind to antibodies generated by the immune system of patients infected with SARS-CoV-2, potentially sharing the same idiotypes as the therapeutic antibodies. To address this potential cross-reactivity, a pre-administration assessment can be performed in clinical case scenarios using the aptamer-based sensors. By establishing a baseline measurement, the interference of antibodies generated by the immune system on the sensing device can be eliminated, allowing for the subsequent monitoring of therapeutic antibody drug levels following injection. In addition, it is worth mentioning that these selected aptamers are versatile and can be applied not only on the GFET platform but also adapted to alternative sensor platforms, such as lateral flow assays and electrochemical platforms.

4. Conclusion

Monitoring patient-specific pharmacokinetics of individual antibodies is crucial for effective administration of new antibody combinations in patients. This includes tracking the time course of their ability to block viral S-proteins, optimizing dosage regimens and evaluating treatment efficacy. We used microfluidic technology to enable rapid (within 5 rounds) and efficient isolation of anti-idiotype aptamers as the affinity reagents for three examples of monoclonal IgG antibodies targeting S protein of SARS-CoV-2 virus. These nanomolar-level aptamers were validated by bead-based fluorescence microscopy assay, pull-down assay and native stain transmission electron microscopy, and integrated with graphene field effect transistor biosensors to enable rapid (tested within 4 minutes), sensitive (with LOD of 1~55 pM), and specific detection of the monoclonal antibody in human serum. These efforts highlight that anti-idiotype aptamers, with many advantages over anti-idiotype antibodies, can be excellent molecular recognition elements for recognizing monoclonal antibody drugs. With the increasing generation of monoclonal antibodies, corresponding aptamers will be needed for each individual antibody. The microfluidic-enabled fast acquisition of anti-idiotype aptamers will allow high throughput aptamer-based bioanalyses to be conducted, facilitating therapeutic program optimization, simple production quantification, and quality control during the antibody manufacturing processes.