SCORe: SARS-CoV-2 Omicron variant RBD-binding DNA aptamer for multiplexed rapid detection and pseudovirus neutralization

Abstract

During the COVID-19 (coronavirus disease 2019) pandemic, several SARS-CoV-2 variants of concern emerged, including the Omicron variant, which has enhanced infectivity and immune invasion. Many antibodies and aptamers that bind the spike (S) of previous strains of SARS-CoV-2 either do not bind or bind with low affinity to Omicron S. In this study, we report a high affinity SARS-CoV-2 Omicron RBD-binding aptamer (SCORe) that binds Omicron BA.1 and BA.2 RBD with nanomolar K_D1. We employ aptamers SCORe.50 and SNAP4.74 in a multiplexed lateral flow assay (LFA) to distinguish between Omicron and wild-type S at concentrations as low as 100 pM. Finally, we show that SCORe.50 and its dimerized form SCOReD can neutralize Omicron S-pseudotyped virus infection of ACE2-overexpressing cells by >70%. SCORe therefore has potential applications in COVID-19 rapid diagnostics as well as in viral neutralization.

Graphical Abstract

First identified to infect humans in 1967, the coronavirus subfamily has low pathogenic strains, such as HCoV-OC43, and severe disease-causing strains, such as SARS-CoV and SARS-CoV-2.¹ The evolution rate of coronavirus is impacted by its large RNA-based genome, a low fidelity of viral polymerase, and recombination with other coronavirus variants within the host.¹ SARS-CoV-2, the pathogen responsible for the on-going COVID-19 pandemic, has already evolved since its initial characterization in 2019. The World Health Organization has five SARS-CoV-2 variants as “variants of concern” (VOC) due to their increased global health significance: Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2) and Omicron (B.1.1529 AKA BA.1, BA.2) strains.² First detected in November 2021, Omicron variant SARS-CoV-2 has both improved immune evasion and more efficient spread compared to the ancestral virus.^{3, 4} Plasma from both convalescent and vaccinated individuals have less neutralizing activity against Omicron pseudovirus versus earlier strains. In addition, the Omicron strain binds with 2.4-fold higher affinity to the angiotensin converting enzyme-2 (ACE2) receptor compared with wild-type Wuhan-Hu-1 virus.³

The SARS-CoV-2 enveloped virus has a lipid shell containing spike protein (S protein or S), a trimeric glycosylated protein that mediates virus binding to angiotensin converting enzyme 2 (ACE2) on host cells. Each S monomer is comprised of the exposed subunit 1 (S1) connected to the membrane-anchored subunit 2 (S2) by a furin cleavage sequence.⁵ S1 contains both the critical receptor-binding domain (RBD) that contacts the ACE2 receptor through the receptor-binding motif (RBM, residues 438–506) as well as the amino-terminal domain (NTD). S1 is the primary target for anti-SARS-CoV-2 neutralizing antibodies, which have been reported to bind both the RBD and NTD domains of SARS-CoV-2 S protein.⁶

Molecular recognition agents that bind new variants are critical for rapid detection of SARS-CoV-2 in patient samples as well as in emerging technologies to sense intact virus in the environment. The two major types of molecular recognition agents are antibodies and aptamers, which are nucleic acid sequences that fold into secondary structures. Aptamers have target binding affinities on par with antibodies while offering several advantages such as small size and extended shelf life.⁷ As chemically synthesized agents, aptamers are also produced more affordably and reproducibly than antibodies,⁸ and are therefore well-suited for rapid viral detection assays. Numerous platforms for aptamer-based SARS-CoV-2 detection have been reported, including electrochemical sensors^{9, 10}, ion transport-sensing nanopores¹¹, metal nanostructure-based optical sensing¹², and lateral flow assays (LFAs), which provide rapid, point-of-care results from patient samples^{13, 14}.

DNA aptamers against SARS-CoV-2 S protein have been identified using a library selection process called systematic evolution of ligands by exponential enrichment (SELEX).¹⁵ Most of the currently-reported aptamers were selected against wild-type or Alpha variant S protein.^{13, 16, 17, 18, 19} To our knowledge, only the Li group has reported a DNA aptamer that binds to BA.1 S protein²⁰ that doesn’t also bind non-specifically to His-tag-containing proteins (Fig. S1A).¹⁸ We tested S-binding aptamers (SP6.51,¹⁸ SNAP1,¹³ SNAP4.74,¹⁴ CoV2-4C¹⁹, CoV2-6C3²¹), including two reported by our group, but none of these aptamers show specific binding to BA.1 NTD or S1 protein (Fig. S1) in our testing conditions.^{13, 14, 18} While the CoV2-4C¹⁹ and CoV2-6C3²¹ aptamers showed binding to BA.1 NTD, S1 and RBD proteins, they also showed binding to a control, His-tagged protein (Fig. S1A).

In this work, we report a SARS-CoV-2 Omicron variant RBD-binding aptamer (SCORe) identified in protein SELEX using BA.1 S1 protein. Biolayer interferometry (BLI) measurements reveal that SCORe and its truncation SCORe.50 bind BA.1 RBD with a K_D1 of 1.73 nM and 2.80 nM, respectively, and that SCORe.50 binds BA.2 RBD with 2.86 nM affinity. We demonstrate a proof-of-concept rapid lateral flow assay (LFA) that uses SCORe.50 and SNAP4.74 to differentiate between BA.1 and wild-type SARS-CoV-2 S protein. Lastly, we show that monomeric SCORe.50 as well as SCORe.50 dimers (SCOReD) can reduce BA.1 S pseudotyped virus infection of ACE2-expressing human cells by over 70%.

Experimental Section

Oligonucleotides.

All oligonucleotides were synthesized by Integrated DNA Technologies. The ssDNA library used in the protein SELEX process was 5’-TCGCTCTTTC CGCTTCTTCGCGG-N40-CCGCGTAAGTCCGTGTGTGCGAA-3’. The forward primer for amplification was 5’-FAM-TCGCTCTTTCCGCTTCTTCG-3’, and the reverse primer for amplification was 5’-Biotin-TTCGCACACACGGACTTACG -3’. Oligonucleotide sequences are listed in Table S2. Modifications to oligonucleotides included: FITC (5’- FITC – sequence – 3’), biotin (5’- biotin – iSp18 spacer – sequence – 3’), and Cy5 (5’ – Cy5 – sequence – 3’). All aptamer pools and aptamers were annealed before use by diluting to 1 – 10 μM in SELEX wash buffer (SELEX WB), heating at 95°C for 5 min, and snap-cooling on ice for 5 min. Dimers (ex. SCOReD) were annealed by combining T linker aptamer (ex. SCORe.50–30T) and A linker aptamer (ex. SCORe.50–30A) sequences at 25 μM in SELEX WB, heating at 95°C for 5 min, and snap-cooling on ice for 5 min.

Buffers.

SELEX wash buffer (SELEX WB) was prepared as previously described.¹³ Binding buffer contained 0.1 mg/mL yeast tRNA (Invitrogen) (tRNA), 0.1 mg/mL salmon sperm DNA (Invitrogen) (SS DNA), and 2% MACS BSA solution (Miltenyi Biotec).

Recombinant proteins.

BA.1 SARS-CoV-2 S protein S1 domain (ACROBiosystems S1N-C52Ha), BA.1 SARS-CoV-2 S protein trimer (ACROBiosystems SPN-C52Hz), BA.1 SARS-CoV-2 S protein NTD (ACROBiosystems SPD-C522df), BA.1 SARS-CoV-2 S protein RBD (ACROBiosystems SPD-C522e), BA.2 SARS-CoV-2 S protein RBD (ACROBiosystems SPD-C522g), SARS-CoV S1 protein (ACROBiosystems S1N-S52H5), human Axl (tyrosine-protein kinase receptor UFO) (ACROBiosystems AXL-H5226), and recombinant human CD8A protein (Sino Biological 10980-H08H) were purchased in lyophilized form. All proteins contained Histag. Omicron variant (BA.1) S1, S trimer, NTD, and RBD contained the following mutations: A67V, HV69-70del, T95I, G142D, VYY143-145del, N211del, L212I, ins214EPE, G339D, S371L, S373P, S375F, K417N, N440K, G446S, S477N, T478K, E484A, Q493R, G496S, Q498R, N501Y, Y505H, T547K, D614G, H655Y, N679K, P681H, N764K, D796Y, N856K, Q954H, N969K, L981F. BA.2 RBD contained the following mutations: G339D, S371F, S373P, S375F, T376A, D405N, R408S, K417N, N440K, S477N, T478K, E484A, Q493R, Q498R, N501Y, Y505H. Anti-SARS-CoV-2 Spike RBD antibody, chimeric mAb, Human IgG1 (ACROBiosystems S1N-M130) and anti-SARS-CoV-2 Spike RBD broadly neutralizing antibody, Human IgG1 (ACROBiosystems SPD-M265) were purchased in lyophilized form. All proteins were reconstituted, aliquoted, and stored according to manufacturer’s recommendations. SARS-CoV-2 wild-type S protein trimer was kindly provided by the Institute of Protein Design (IPD) at the University of Washington.

Viruses.

Inactivated (UV or heat-inactivated) SARS-CoV-2 (wild-type, Alpha, Beta, Delta, BA.1) and HCoV-OC43 virus samples at known concentrations were kindly provided by the NIH RADx-Radical Data Coordination Center (DCC) at the University of California San Diego and BEI Resources.

Protein-SELEX.

The protocol was based on the guidelines outlined by Wang et al.²² and is outlined in Fig. 1. Experimental conditions for 7 rounds of SELEX are summarized in Table S1. Aptamers were partitioned by Dynabeads His-tag isolation & Pulldown (Novex by Life Technologies) on a rack magnet. Aptamers were amplified with Phusion^® High-Fidelity DNA Polymerase (NEB). Annealing, strand separation, and composition of buffers were described previously².

Figure 1.

Schematic of SELEX process. A single-stranded DNA (ssDNA) aptamer library is subjected to SARS-CoV-2 Omicron (BA.1) S1 protein for seven rounds of positive selection with increasing stringency, and SARS-CoV-2 Omicron (BA.1) NTD negative selection in round 6 and 7 only. Binding of aptamer round pool is evaluated by protein binding assay, and aptamer round pools with positive binding are sequenced and analyzed.

Next-generation sequencing (NGS) and data analysis.

The DNA pools from SELEX rounds were PCR amplified with barcoded primers described in Table S3 using the MiSeq Reagent Kit v2 (300-cycles) (Illumina) and MiSeq System (Illumina) per manufacturer’s protocol. Sequence reads were analyzed as previously described.¹³ NUPACK was used to predict secondary structure of candidate aptamers.²³

Biolayer interferometry (BLI).

Studies were performed with an Octet Red96 machine (Sartorius) at 25°C and 1,000 rpm sample agitation. Pre-soaked Ni-NTA (NTA) sensors were rinsed in 1% BSA, 0.1 mg/mL tRNA, 0.1 mg/mL SS DNA, 0.01% Tween-20 SELEX WB (“diluent”) for 100s. For experiments containing nasal swab in buffer, nasal swab samples were collected as previously described and spiked into diluent.¹⁴ Next, tips were loaded with 50 nM His-tagged protein until signal reached a 3.5 nm signal threshold. Subsequently, tips were cross-linked in 0.1 M EDC 0.025 M NHS in H₂O for 60 s, quenched in 1 M ethanolamine PBS (pH 8.5) for 60 s, and rinsed in diluent for 100 s. Then biosensors were baselined in diluent for another 100 s. After association with protein of interest diluted to the desired concentration, sensor tips were returned to the baseline diluent well for dissociation. Data were analyzed with Octet Data Analysis 9.0 (Sartorius). Kinetic values were determined from a global fit of several curves generated from serial dilutions of the protein with a 1:2 binding model (heterogenous ligand). The K_D1 describes the rate of the initial binding and the K_D2 describes the rate of a second binding event after the initial binding.

Aptamer round pool binding assay.

The protocol was conducted using an ELISA-like plate binding assay as previously described.¹³ In brief, 96 well plates were coated overnight with protein of interest. After washing and blocking wells with 5% BSA, 0.1 mg/mL tRNA, 0.1 mg/mL SS DNA and 0.01% Tween-20 SELEX WB, fluorescein-labeled aptamer pools were incubated in the wells (30 min, RT) before staining with HRP anti-fluorescein antibody for detection with TMB substrate. Aptamer binding was quantified by UV absorbance with measurements at 450 nm.

Protein binding plate assay.

Nunc MaxiSorp^™ flat-bottom 96-well plates (Thermo Fisher Scientific) were coated with 50 nM protein (except Omicron S trimer at 25 nM) at 4°C overnight. Then wells were washed four times with wash buffer (0.5% BSA, 0.01% Tween-20 SELEX WB.) Next, wells were incubated with blocking buffer (5% BSA, 0.1 mg/mL tRNA, 0.1 mg/mL SS DNA, 0.01% Tween-20 SELEX WB) for 1.5 hr at room temperature. After blocking buffer removal via flick and tap, 100 nM biotin-labeled aptamer pools or control biotin-labeled aptamer in binding buffer were incubated for 30 min at room temperature. Then the plate was washed four times with wash buffer and stained with Pierce^™ High Sensitivity streptavidin HRP (1:16,000 1 mg/mL, Thermo Scientific 21130) for 1 h at room temperature. Lastly, the plate was washed four times with wash buffer. Pierce^™ TMB substrate kit (Thermo Scientific 1854050) was mixed at 1:1 according to manufacturer’s protocol, then added to wells and incubated at room temperature until desired blue color was developed (5–30 min.) 2M sulfuric acid was added to each well to terminate the HRP reaction. The absorbance of the solution (yellow) was measured using the Infinite 200 PRO plate reader (Tecan) at 450 nm with 550 nm as reference. The absorbance values of the groups were deducted from wells without aptamer and plotted.

Gold nanoparticle synthesis.

Gold nanoparticles were prepared by the citrate synthesis method that involves sodium citrate reduction of tetrachloroauric acid as previously described.¹⁴

Gold nanoparticle and antibody conjugation.

Antibody-conjugate gold nanoparticles were prepared by adapting a protocol from Wang et al.²⁴ Anti-RBD chimeric mAb human IgG1 antibody (ACROBiosystems S1N-M130) was activated with 10-fold excess Traut’s reagent in 2mM EDTA PBS (pH 8.0) for 1 hr, then desalted by Zeba^™ Spin Desalting Columns, 7K MWCO (Thermo Scientific).²⁴ 10 μg/mL of activated antibody was added to 1 mL gold nanoparticles, and incubated for 5 min before addition of 150 μM thiol-PEG12-acid (Broadpharm BP-21916). After 2 hr incubation, 0.25% BSA was added to the solution and further incubated for 5 min before use.

LFA dipstick manufacturing.

The protocol was conducted as previously described,¹⁴ with adjustments to the striped test lines: Test line 1: 2 mg/mL polystreptavidin R (pSA) (Eagle Biosciences, # 10 120 030) in DPBS. Test line 2: 0.5 mg/mL anti-FITC antibody (BioLegend, #408302). Control line: 0.5 mg/mL anti-human IgG antibody (BioLegend #410701).

Multiplex LFA.

Binding buffer (1% BSA, 0.1 mg/mL tRNA, 0.1 mg/mL SS DNA, 0.01% Tween-20), 25 nM biotinylated aptamer, 25 nM FITC-labeled aptamer, and 50 μL antibody-functionalized gold nanoparticles were prepared, then analyte was added to a total volume of 100 μL. The solution was incubated at RT for 20 min with rocking. Then, a lateral flow device was dipped into the solution for 15–20 min (entire volume absorbed). Lastly, the lateral flow devices are dipped in 25 μL SELEX WB for 5 min (partial volume absorbed. For each strip, the average pixel values in the control band, test bands, and background (blank area between bands) were calculated. The band values were background-subtracted, then the percent maximum signal across all strips for the same band was calculated. The LOD was determined from the lowest concentration of sample that was significantly higher than the negative control.

Cell culture.

The HEK293 ACE2 cell line stably expressing ACE2, generously provided by Dr. Jesse Bloom (Fred Hutchinson Cancer Research Center), was cultured in DMEM (Gibco) 10% FBS at 5% CO₂ 37°C.²⁵

HEK293 ACE2 competition.

25 nM of His-tagged Omicron (BA.1) S trimer was incubated with 250 nM or 2.5 μM of SCOReD, 250 nM or 2.5 μM of SNAP1D, or 500 nM or 5 μM of SCORe.50 in binding buffer for 20 min at RT, then added to 200,000 HEK293T ACE2 cells for 20 min on ice. Cells were stained with anti-His-tag antibody-FITC (Invitrogen MA1-81891) and fluorescence read by flow cytometer (Attune NxT, Invitrogen). The median mean fluorescence intensity (MFI) of experiment groups were background subtracted from that of untreated cells. The percent inhibition was calculated by subtracting experiment value from S only average value and dividing by S only average value.

Virus neutralization studies.

10,000 HEK293 ACE2 cells were plated in a 96-well plate and cultured to 75% confluence. Aptamers were annealed at 25 μM in DMEM 3 mM MgCl₂. 12.5 μL (1:500 dilution in DMEM 3mM MgCl₂ 2% FBS) of Omicron variant (BA.1) SARS-CoV-2 S protein pseudotyped virus (ACROBiosystems PSSO-HLGB003) and indicated concentrations of aptamer or neutralizing antibody (ACROBiosystems SPD-M265) were incubated together in DMEM 3 mM MgCl₂ for 1 hr at RT. Incubation media was removed and virus added for 2 hr at 37°C. Next, media was replaced with fresh complete media, and cells were incubated for 48 hr. Cells were rinsed with 200 μL 1X PBS, then 40 μL Reporter Lysis Buffer (Promega) was added. Plate was frozen then thawed before 20 μL cell lysate was mixed with 100 μL Luciferase Assay Reagent. Light was measured using the Infinite 200 PRO plate reader (Tecan) at 1000 nm integration time. The relative light units (RLUs) of experiment groups were background subtracted from that of untreated cells. The percent infectivity was calculated by normalizing to virus-only treatment group values.

Results & Discussion

Aptamer Selection

We designed a protein SELEX strategy to identify aptamers that bind to BA.1 S protein RBD. Our previous attempts using the RBD domain alone for SELEX were not successful, so we used SARS-CoV-2 BA.1 S1 protein for positive selection and SARS-CoV-2 BA.1 NTD for negative selection in later rounds. We conducted seven rounds of SELEX with increasing stringency, accomplished by increasing the amount of competitors (BSA, tRNA, SS DNA) and decreasing the amount concentration of aptamer pool and positive selection target Omicron (BA.1) S1 protein (Fig. 1). For the last three rounds, we evaluated binding of the fluorescein-labeled aptamer libraries from each selection round and observed binding to BA.1 S1 protein, but not to NTD or Axl (tyrosine-protein kinase receptor UFO) (negative control protein), starting in round 5 (R5) and increasing significantly in round 7 (R7) (Fig. S2). We analyzed libraries from rounds 3 through 7 by next generation sequencing (NGS) and identified four aptamers (named A1 to A4) of interest. A1 and A2 first appeared as early as round 3, while A3 appeared in round 4 and A4 in round 5. Between round 5 and 6, all motifs sharply increased in prevalence (Fig. S3). We tested A1, A2, A3, and A4 and observed that A3 bound BA.1 S1 but not BA.1 NTD (Fig. S1). Aptamers A1, A2 and A4 did not bind either protein and therefore were not further characterized.

Characterization of SCORe (A3)

To determine the binding location of A3 on S1 protein, we evaluated A3 (Fig. 2A) binding to Omicron S protein, S1, RBD and NTD via an ELISA-like plate binding assay, and found that A3 bound BA.1 RBD, BA.2 RBD, BA.1 S1, and BA.1 S protein (Fig. 2B) but not BA.1 NTD. Given the binding location, we renamed A3 to SARS-CoV-2 Omicron RBD-binding Aptamer (SCORe). The selection of an RBD-binding aptamer demonstrates that our SELEX strategy using negative selection against NTD was effective.

Figure 2.

A) Secondary structure prediction with the following conditions: 22°C, 137 mM Na⁺, 5.5 mM Mg⁺⁺. Arrows represent the 3’ end of the aptamer. B, D) Proteins (B) or inactivated virus (D) adsorbed onto surface are incubated with 100 nM biotinylated aptamer and stained with streptavidin-HRP with detection of the HRP substrate TMB by UV absorbance. Bars indicate mean, brackets indicate S.D., n = 3. C) Biolayer interferometry assessment of binding. His-tagged RBD was loaded on Ni-NTA biosensors and associated with SCORe-FITC or SCORe.50-biotin. The gray line indicates the switch from analyte association to dissociation after 600 s. K_D values (mean ± S.D., n = 3–4) were determined from a global fit (dotted line) of the kinetic data at various concentrations of proteins for a 2:1 binding model.

We compared the sequence of SCORe to other aptamers reported to bind SARS-CoV-2 S protein at the RBD (CoV2-RBD-1C, CoV2-RBD-4C, CoV2-6C3, nCoV-S1-Apt1, MSA5, and MSA52-T5)^{17, 19, 20} and found no similarity between SCORe and the other sequences when compared with local alignment (Fig. S4).²⁶

We next truncated the aptamer sequence to reduce synthesis cost and potentially improve binding affinity.⁹ The NUPACK secondary structure prediction of SCORe shows a stem formed by the constant sequences used for PCR amplification. We removed this stem in a 50-nt truncation called SCORe.50 (Fig. 2A).

We used biolayer interferometry (BLI) to measure the binding affinity of SCORe and SCORe.50 for immobilized BA.1 RBD, revealing a K_D1 of 1.73 nM and 2.80 nM, respectively (Fig. 2C). In addition, we show that SCORe.50 binds BA.2 RBD with a K_D1 of 2.86 nM. We did not observe binding of RBD to biotinylated SCORe, SCORe.50, or SCORe.50-30T (SCORe.50 with a 30-nt thymidine linker) bound to streptavidin sensor surfaces. We hypothesize that immobilization of the 5’ end of the aptamer may prevent conformational mobility required for binding. Lastly, we showed that SCORe and SCORe.50 bound to UV-inactivated BA.1 virus with 18.8-fold and 11.2-fold higher signal, respectively, than heat-inactivated BA.1 virus (Fig. 2D).

We also evaluated aptamer binding to spike proteins of other variants of SARS-CoV-2 and other coronaviruses. We observed binding of SCORe and SCORe.50 to wild-type RBD and wild-type S protein in plate binding assay (Fig. S5A) but not by BLI (Fig. S5B). We observed binding of SCORe and SCORe.50 to Delta variant S1 protein, but not SARS-CoV S1 protein by plate binding assay (Fig S5C). We also observed binding of both SCORe and SCORe.50 to UV-inactivated SARS-CoV-2 virus of different variants, including Omicron (BA.1), Delta, and wild-type (Fig. 2D). SCORe and SCORe.50 did not bind as highly to heat-inactivated SARS-CoV-2 virus presumably due to denaturation of S protein. Based on these results, we hypothesize that while SCORe can bind to RBD of several variants, such as wild-type and Delta variant, it binds with the highest affinity to Omicron variant RBD. Due to similarity in binding patterns and affinities on the same order of magnitude (K_D1 of 1.73 nM for SCORe and K_D1 of 2.80 nM for SCORe.50), we used SCORe.50 in subsequent studies.

Multiplexed LFA with SCORe.50 and SNAP4.74

We next applied SCORe.50 for multiplexed SARS-CoV-2 S protein detection in a lateral flow assay (LFA). Since SCORe.50 binds to Omicron S protein with higher affinity than wild-type S protein and our previously-reported SNAP4.74 binds to wild-type S protein,¹⁴ but not Omicron S protein, we hypothesized that we could use the two aptamers in a rapid detection format to differentiate between Omicron and wild-type S protein. We used SNAP4.74-fluorescein for wild-type S protein capture using anti-FITC antibody and SCORe.50-biotin for Omicron (BA.1) S protein capture using streptavidin (Fig 3A). To test our custom LFA, we incubated S protein with a solution containing biotinylated SCORe.50, FITC-labeled SNAP4.74, and anti-RBD IgG gold nanoparticles (Ab-AuNP). We then dipped into the solution a lateral flow strip with three immobilized bands: (1) poly-streptavidin to capture SCORe.50-biotin, (2) anti-FITC antibody to capture SNAP4.74-FITC, and (3) anti-IgG antibody control to capture remaining Ab-AuNPs. Ab-AuNPs bind to captured S proteins, allowing users to read results with the naked eye (Fig. 3A). The lateral flow device differentiates between Omicron (BA.1) and wild-type S protein, with detection limit as low as 50 pM of Omicron (BA.1) S protein and 100 pM of wild-type S protein (Fig. 3B). Although SCORe.50 can bind wild-type S protein by plate binding studies, we did not observe binding of SCORe.50-biotin with 1 nM wild-type S protein in lateral flow, even in the absence of SNAP4.74-FITC. We also show that SCORe.50-biotin can bind BA.1 RBD in buffer containing nasal swab by BLI, demonstrating potential for patient nasal swab sample LFA (Fig. S6). Our system is a proof-of-concept for a modular aptamer-based LFA system for differentiation between SARS-CoV-2 variants or diseases with similar symptoms, such as influenza. Because disease severity differs between variants,²⁷ an LFA system that can distinguish between SARS-CoV-2 variants may provide information useful for patient care and disease epidemiology.

Figure 3.

A) LFA schematic. B) Detection and differentiation between Omicron (BA.1) and wild-type SARS-CoV-2 S protein. “None” and “0 pM” indicate negative controls, which were treated identically as other samples except with no S protein added. Bars indicate mean. Brackets indicate S.D. Statistical significance determined by one-way ANOVA with Dunnett correction; ** indicates p < 0.01, **** indicates p < 0.0001 when compared to no protein; n = 3 per experimental group. Representative images are shown.

Pseudovirus neutralization by SCORe.50 and SCOReD

Multivalent aptamers have previously been shown to have tighter affinity for their target protein¹⁰. A trimeric RNA aptamer was shown to neutralize wild-type SARS-CoV-2 entry to cells. ²⁸ Therefore, designed SCORe.50 dimers that connect two SCORe.50 with a 12-, 20-, or 30-nt stretch of A-T complementary binding in order to augment binding affinity for Omicron RBD (Fig. S7A). We first show co-staining between fluorescently-labeled SCORe.50 and BA.1 RBD binding to ACE2-overexpressing cells (Fig. S7B). We observe no correlation with negative control aptamer (SNAP1.50), which binds wild-type NTD. Next, we screened the three dimer types by flow cytometry for competition with BA.1 S protein to ACE2-overexpressing HEK293 cells. We found that 20 A/T and 30 A/T linker length dimers were slightly more effective in reducing binding of S protein than dimers with 12 A/T (Fig. S7C). We also tested binding of dimers to BA.1 RBD by BLI and observed similar binding patterns and affinities for each (Fig. S7D). We selected the 30 A/T linker length SCORe.50 dimer (SCOReD) to focus on in following experiments (Fig. 4A).

Figure 4.

A) Secondary structure prediction as described in Fig 2. B) BLI characterization of SCOReD binding to immobilized RBD with experimental details and K_D determination as described in Figure 2. C) ACE2-overexpressing HEK293 cells were incubated with 25 nM Omicron S protein and 250 nM or 2.5 μM of indicated dimer or 500 nM or 5 μM of monomeric SCORe.50-30T. S binding was measured by flow cytometry. Ratios (20:1, 200:1) indicate molar ratio of monometic aptamer units to S protein. Three technical replicates shown. Bars represent mean, brackets represent S.D. **** denotes p < 0.0001 as determined by 2-way ANOVA with Tukey’s multiple comparison test. D) Omicron S pseudotyped virus infection of ACE2-overexpressing HEK293 cells in the presence of various aptamer constructs. Three biological replicates performed with four technical replicates each shown. Bars represent mean, brackets represent S.D. * denotes p < 0.05 and ** denotes p < 0.01 as determined by 2-way ANOVA with Dunnett’s multiple comparison test.

We confirmed that SCOReD binds to BA.1 RBD with a K_D1 of 1.04 nM, similar to that of SCORe.50 (Fig. 4B). SCOReD inhibited BA.1 S protein from binding to ACE2-overexpressing HEK293 cells, knocking down binding by 38.0% at 10-fold molar excess of SCOReD to S and 69.0% at 100-fold molar excess of SCOReD to S (Fig. 4C). In comparison, monomeric SCORe.50 reduces binding by 38.5% at 200-fold molar excess (equivalent to 100-fold molar excess of SCOReD). Analogous negative control dimer with wild-type-NTD-binding SNAP1.50 (SNAP1D) shows no inhibition. Lastly, we show that SCORe.50 and SCOReD can block Omicron (BA.1) S pseudotyped virus entry (OPV) to ACE2-overexpressing HEK293 cells (Fig. 4D). Cells were transduced with OPV virus carrying the luciferase reporter gene in the presence of SNAP1D (control dimer), monomeric SCORe.50-30T and SCOReD. Both SCORe.50-30T and SCOReD significantly reduced pseudovirus infection compared to control dimer. Both SCORe.50 and SCOReD significantly reduced infectivity in comparison to SNAP1D at 5 μM (Fig. 4D) and 10 μM (Fig. S8) of dimer or monomer molar equivalent. In particular, SCOReD reduced infection by 42.2±4.4% at 5 μM (Fig. 4D) and 71.1±9.9% at 10 μM (Fig. S8). To probe whether SCOReD dimer provided an advantage over SCORe.50, we compared treatment with heterodimer composed of SCORe.50 and SNAP1.50 to treatment with SNAP1D and SCOReD. There was a significant difference between SNAP1D and SCOReD, but there was no significant difference between heterodimer and SNAP1D or heterodimer and SCOReD. This indicates that dimerization may not significantly augment the neutralizing capabilities of SCORe.50 (Fig. S9). Treatment with 10 μM SCOReD is less effective than with 10 nM or 100 nM of neutralizing antibody, which both achieved virtually complete neutralization (Fig. S10).

Conclusion

In this report, we discover and characterize SCORe, a high affinity aptamer for the Omicron BA.1 and BA.2 S protein RBD domain. We apply a SCORe truncation, SCORe.50, together with our previously reported wild-type S protein-binding aptamer, SNAP4.74, in a multiplexed lateral flow assay and demonstrate sub-nanomolar detection of and differentiation between Omicron (BA.1) and wild-type S. We demonstrated that SCORe.50 and its dimer SCOReD can compete with ACE2 for binding to Omicron (BA.1) S and inhibit pseudovirus infection. We use SCORe in rapid lateral flow assays and pseudovirus neutralization, demonstrating its versatile uses and potential for diagnostic and therapeutic applications.

ABBREVIATIONS

SARS-CoV-2

severe acute respiratory syndrome coronavirus 2

S

spike protein

ACE2

Angiotensin-converting enzyme 2

RBD

receptor binding domain

NTD

N-terminal domain

LFA

lateral flow assay

SCORe

SARS-CoV-2 Omicron variant RBD-binding aptamer