A high-affinity aptamer with base-appended base-modified DNA bound to isolated authentic SARS-CoV-2 strains wild-type and B.1.617.2 (delta variant).

Simple, highly sensitive detection technologies for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are crucial for the effective implementation of public health policies. We used the systematic evolution of ligands by exponential enrichment with a modified DNA library, including a base-appended base (uracil with a guanine base at its fifth position), to create an aptamer with a high affinity for the receptor-binding domain (RBD) of the SARS-CoV-2 spike glycoprotein. The aptamer had a dissociation constant of 1.2 and < 1 nM for the RBD and spike trimer, respectively. Furthermore, enzyme-linked aptamer assays confirmed that the aptamer binds to isolated authentic SARS-CoV-2 wild-type and B.1.617.2 (delta variant). The binding signal was larger that of commercially available anti-SARS-CoV-2 RBD antibody. Thus, this aptamer as a sensing element will enable the highly sensitive detection of SARS-CoV-2.

SARS-CoV-2

SELEX

BAB

RBD

SB

Fw

Rv

dsDNA

ssDNA

SPR

ELAA

FSBC

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the cause of the novel coronavirus disease-2019 (COVID-19) pandemic, triggers acute respiratory diseases, and the COVID-19 pandemic is a global threat to public health [1]. Recently, one of the variants of concern (VoCs), B.1.617.2 (delta variant) has spread rapidly all of the world. SARS-CoV-2 mainly infects human alveolar epithelial cells by binding to angiotensin-converting enzyme 2 via the receptor-binding domain (RBD) of the spike (S) protein, a surface glycoprotein on the viral particle [2,3]. Therefore, the S protein is an important target for detecting SARS-CoV-2 as well as developing antiviral antibodies and compounds [4].

Antibodies, which bind specifically to target proteins, are widely used in research and therapeutics. For example, a lateral flow assay has been developed using antibodies against SARS-CoV-2 [5,6]. Antibodies are the gold standard for recognizing molecules; however, they have drawbacks, including instability and high costs of development and production [7].

Aptamers are single-stranded DNA (ssDNA) or RNA oligonucleotides that bind to specific molecules or cells [8,9]. Therefore, aptamers are able to regulate the functions of the targets similarly as antibodies. Additionally, aptamers have some advantages, such as easiness of both efficient production and chemical modification, and reversible folding without aggregation [[10], [11], [12]].

Aptamers are usually isolated from combinatorial nucleic acid libraries using an iterative selection process called systematic evolution of ligands by exponential enrichment (SELEX) [13]. The SELEX method uses a library comprising primer regions and random regions to select the sequences binding to the target [14,15]. We have developed analogs with modified bases containing other bases, i.e., base-appended bases (BABs). Using these modified bases, we have generated aptamers with extremely high binding affinities for various targets [[22], [23], [24], [25]].

Aptamers for SARS-CoV-2 S protein or RBD have been selected from a DNA library using natural bases [16,17]. Also, an aptamer inhibiting SARS-CoV-2 infection has been identified [[18], [19], [20], [21]]. In this study, a high-affinity artificial nucleic acid aptamer for the SARS-CoV-2 S RBD was obtained using a modified DNA library containing the following base-appended base modifications; analog guanine derivative at the fifth position of uracil (Ugu) [23]. This aptamer was shown to detect authentic SARS-CoV-2 strains belonging to lineages A and B.1.617.2.

Materials and methods

Target protein and virus

The recombinant RBD of the SARS-CoV-2 (2019-nCoV) S protein (YP_009724390.1, Arg319-Phe541) and the His-tagged S1+S2 ECD protein (YP_009724390.1, Val16-Pro1213) (Sino Biological, Beijing, China) were used as the selection targets.

All the experiments using SARS-CoV-2 were performed under the guidelines of the Biosafety Management Committee on Pathogens and Other Hazardous Agents of Hokkaido University and the International Institute for Zoonosis Control. SARS-CoV-2 strains belonging to lineages A [26], B.1.1.7 (alpha variant), B.1.351 (beta variant), P.1 (gamma variant), B.1.617.2 (delta variant), and human coronavirus OC43 (HCoV-OC43) were prepared separately. Those SARS-CoV-2 strains were isolated and provided by Drs. Saijyo and Shimojima at the National Institute of Infectious Diseases, Tokyo, Japan. First, each strains of SARS-CoV-2 was inoculated into Vero-TMPRSS2 cells [27] cultured in Dulbecco's modified Eagle's medium containing 2% fetal bovine serum, and the supernatants were collected upon observation of cytopathic effects. Meanwhile, HCoV-OC43 was inoculated into MRC-5 cells, and the supernatants were collected in the same manner. Then, the viruses were pelleted by ultracentrifugation at 110,000×g at 4 °C for 2 h with a 20% sucrose cushion. Afterward, the pellets were resuspended in phosphate-buffered saline (PBS) and stored at −80 °C. Viral titers were measured by plaque assays [27].

SELEX

SELEX was conducted as previously reported [23]. Dynabeads MyOne Carboxylic Acid and Dynabeads MyOne SA C1 magnetic beads (Invitrogen, Waltham, MA) were used for target solidification and biotinylated DNA retrieval, respectively. The target beads were prepared by binding MyOne Carboxylic Acid to recombinant RBD according to the manufacturer's instructions and washed with the selection buffer [SB; 40 mM HEPES (pH 7.5), 125 mM NaCl, 5 mM KCl, 1 mM MgCl2, and 0.01% Tween 20]. A double-stranded DNA (dsDNA) with inserted Ugu was prepared using a 5′-biotin-modified complementary strand (5′-GAATACAAGACACCTCGGCTTTGC-N30-GATTTCAGTGGCGGAGACATACC-3′), a forward (Fw) primer (5′-GGTATGTCTCCGCCACTGAAATC-3′), and KOD Dash (Toyobo, Osaka, Japan). After the dsDNA was bound to MyOne SA C1 magnetic beads, ssDNA was eluted with 0.02 M NaOH and neutralized with 0.08 M HCl to prepare the Ugu ssDNA library. The primers, random pool, and aptamer clone templates were purchased from Integrated DNA Technologies (Tokyo, Japan).

The 130 pmol library was mixed for 15 min with 250 μg of target beads at 25 °C. Then, the beads were washed with SB, and the bound ssDNA was eluted with 7 M urea and amplified by polymerase chain reaction (PCR) using the Fw primer and a reverse (Rv) primer (5′-GAATACAAGACACCTCGGCTTTGC-3′) with modification by the 5′-biotin. Next, the amplified dsDNA was bound to MyOne SA C1 magnetic beads, and the Fw chain was eluted with 0.02 M NaOH. The ssDNA produced by this method, using the Rv chain, Fw primer, and Ugu-immobilized in magnetic beads, were used in the next round. After eight rounds of selection, PCR with the Fw and Rv primers and subsequent sequencing were conducted using a MiniSeq System (Illumina, San Diego, CA). The sequence data obtained were clustered by fast string-based clustering (FSBC), and 1 sequence each was selected from the top and another higher-ranked clusters [28].

Surface plasmon resonance analysis

All the surface plasmon resonance (SPR) measurements were performed at 25 °C using the ProteON XPR360 (Bio-Rad, Hercules, CA) [22]. For the aptamer clones, the ligand was set by appending poly-A20 at the 3′-end and hybridizing the 5′-end to an NLC sensor chip (Bio-Rad) with biotin-modified oligo (dT20) [23]. The analytes included the RBD, S1+S2, recombinant 2019-nCoV S1 (Elabscience, Houston, TX), and recombinant 2019-nCoV S trimers (Elabscience). Bovine serum albumin (Sigma-Aldrich, St. Louis, MO) with SB was used as the running buffer. The dissociation constants between the aptamer and recombinant RBD or S protein were calculated using a simple 1:1 biomolecular interaction model, the most common kinetic fit model for SPR data analysis, according to the manufacturer's instructions. The dissociation constant of S protein was estimated as the apparent Kd value due to its trimeric formation.

Enzyme-linked aptamer assays

Enzyme-linked aptamer assays (ELAAs) were conducted as previously described [29]. Briefly, recombinant RBD protein and purified SARS-CoV-2 were diluted with 50 mM carbonate buffer (pH 9.6), added to MaxiSorp plates (Thermo Scientific, Waltham, MA), at 1 μg/well and 1.0 × 106 PFU/well, respectively, and solidified at 4 °C overnight. Recombinant His-tagged hemagglutinin (HA) protein of influenza A/California/04/2009H1N1 (Sino Biological, Beijing, China) was used as a negative control. Then, the wells were washed once with SB and blocked with Tris-buffered saline Blocking Buffer (Pierce Biotechnology, Rockford, IN) at 25 °C for 1 h. Then, a 5′ biotinylated aptamer at 1 μM was prepared in SB, denatured by heating at 95 °C for 5 min, and cooled. Next, the aptamer was added and incubated at room temperature for 1 h. After three washes with SB, streptavidin–horseradish peroxidase (HRP) (1:1,000 diluted; Citiva, New York, NY) was added, and the plates were incubated for 30 min at room temperature to detect the bound biotinylated aptamers. After three washes with SB, a 3,3′,5,5′-Tetramethylbenzidine (TMB) solution (Thermo Fisher Scientific) for RBD detection and 1-step™ Ultra TMB-ELISA substrate solution (Thermo Fisher Scientific) for SARS-CoV-2 detection was added, and the plates were incubated at room temperature for 10 min. The reaction was stopped by adding 0.5 N sulfuric acid, and the absorbance was measured at 450, 490 and 620 nm.

Statistical analysis

All statistical analyses were performed using R version 4.0.4. For analyses between two groups, a one-tailed Student's unpaired T-test was applied. For comparisons among more than two groups, one-way ANOVA with Tukey's multiple comparisons was used. The methods of statistical analysis are described in the figure legends for each experiment.

Results and discussion

For the detection of SARS-CoV-2, the S protein-binding aptamers were screened using a Ugu-modified nucleic acid library. Four S protein-binding aptamers candidates [28], two for the RBD and two for the S1+S2 protein (Table S1), were selected from sequence analysis. Four analytes, S1 and S trimers in addition to RBD and S1+S2, were used for SPR analysis. The two RBD-binding aptamers, RBD-Ugu1 and RBD-Ugu2, bound all four analytes. However, the two S1+S2-binding aptamers, S1S2-Ugu3 and S1S2-Ugu4, failed to bind the RBD (Fig. S1).

The RBD-binding ability of a previously reported aptamer, CoV2-RBD-1 [16], was compared with that of RBD-Ugu1 and RBD-Ugu2 using SPR analysis. The dissociation constants of RBD-Ugu1 and RBD-Ugu2 were 1.2 and 1.7 nM, respectively, demonstrating more strong binding affinities than CoV2-RBD-1 (Fig. 1 a). In a solution containing dextran sulfate, a polyanion suppressing nonspecific binding to nucleic acids [30,31], CoV2-RBD-1 significantly lost its binding activity. Conversely, the dissociation constants for RBD-Ugu1 and RBD-Ugu2 were 3.2 and 3.5 nM, respectively, indicating that their binding activities remained largely unaffected (Fig. 1b). This observation suggests that charge has little influence on the ability of RBD-Ugu1 and RBD-Ugu2 to bind RBD, even though the binding of aptamers to the target is usually influencing by charge. These data indicate that RBD-Ugu1 and RBD-Ugu2 exhibit little nonspecific binding. Additionally, the evaluation of the binding of RBD-Ugu1 and RBD-Ugu2 to the S trimer using SPR revealed a dissociation constant of less than 1 nM for both aptamers, indicating exceptionally strong binding (Fig. S2). The binding sites on RBD for RBD-Ugu1 and RBD-Ugu2 are unclear. Since BAB clones have different structures from natural-base clones with corresponding sequences (in which dUx is dT) [22], the more complex structures of RBD-Ugu1 and RBD-Ugu2 may result in stronger RBD binding.

Fig. 1

SPR response curves of the interaction between the SARS-CoV-2 S RBD and the aptamer candidates, RBD-Ugu1, RBD-Ugu2, and CoV2-RBD-1

(a–b) Different concentrations of RBD (25–400 nM) were injected over the respective aptamer-immobilized sensor chips for 120 s at a flow rate of 50 μL/min, and the measurements were performed using multicycle kinetics. Black dot line and red line represent the measured and fitting curves, respectively. The average of the squared differences between the data points and the corresponding fitted values is represented as χ2. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Generally, a longer sequence results in the instability of an aptamer's conformation at the interface of the target [32]. Hence, the length of aptamers should be minimized to enhance the aptamers' binding ability and reduce their cost of production [33]. We designed a truncated aptamer sequence (RBD-Ugu1-1) with the region in RBD-Ugu1 most strongly related to binding (5′-GGAATTCATG-3′), which was estimated with FSBC, considering the maintenance of the secondary structure of the binding region. RBD-Ugu1-1 was a 42-mer with the sequence of 5′-CCACTGAAATCCGtGCCtAAtCtCACCCCACGGAAttCAtGG-3′, with t indicating Ugu. The secondary structure of RBD-Ugu1-1 was predicted with RNAfold of ViennaRNA package [34] (Fig. 2 a). The dissociation constant of RBD-Ugu1-1 for RBD was 2 nM in SB (Fig. 2b) and 5.3 nM in dextran sulfate (Fig. 2c), exhibiting a binding ability comparable with that of the original sequence. Next, a direct ELAA was conducted using the 5′-terminal biotinylated aptamers to the recombinant RBD proteins and various SARS-CoV-2 strains immobilized on 96-well plates. RBD-Ugu1 and RBD-Ugu1-1 bound equally well to the RBD. However, S1S2-Ugu3 did not bind to the RBD (Fig. 3 ). These results were consistent with those of the SPR binding assays. No binding signal of aptamers to influenza virus HA protein was detected, suggesting that aptamers specifically recognize spike proteins of SARS-CoV-2.

Fig. 2

Properties of the truncated RBD-Ugu1-1 aptamer

(a) The predicted secondary structure of RBD-Ugu1-1 which was estimated with RNAfold of ViennaRNA package [34] was displayed using Forna [37].

(b–c) The SPR response curves of the interaction between SARS-CoV-2 S RBD protein and RBD-Ugu1-1. Black dot line and red line represent the measured and fitting curves, respectively. The average of the squared differences between the measured data points and the corresponding fitted values is represented as χ2. In (b), SB was used as a running buffer; in (c), SB with 0.1% dextran sulfate was used. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Binding capacities of RBD-Ugu1, RBD-Ugu1-1, and S1S2-Ugu−3 to the RBD of the spike protein and HA protein derived from influenza H1N1 virus as measured by direct ELAA

The amount of the 5′ biotinylated aptamer captured by immobilized RBD and HA was determined the absorbance at 490 nm. The experiment was performed in triplicates. Error bars represent standard deviation for 6 data points. Buffer was used Selection Buffer [40 mM HEPES (pH 7.5), 125 mM NaCl, 5 mM KCl, 1 mM MgCl2, and 0.01% Tween 20].One-tailed unpaired T-test was applied to compare the mean of ELAA signals of the RBD with that of HA protein or buffer for each aptamer. The symbol of two asterisks stands for P value less than 0.01.

Then, the ability of RBD-Ugu1-1 to detect authentic SARS-CoV-2 was evaluated with SARS-CoV-2 lineages A, B.1.617.2, B.1.1.7, B.1.351 and P.1 using direct ELAA (Fig. 4 ). As a control, a commercially available anti-SARS-CoV-2-RBD antibody conjugated with HRP (Abcam, Cambridge, UK) was used with 1:1,000 dilution. The aptamer was bound to authentic SARS-CoV-2 wild-type and B.1.617.2 immobilized on a plate (Fig. 4a). The binding signals of the aptamer to SARS-CoV-2 wild-type and B.1.617.2 were higher than that of the antibody, indicating that this aptamer has a potential as a better alternative for SARS-CoV-2 detection. The aptamer binding signals to B.1.1.7, B.1.351, and B.1.617.2 are reduced compared with that of wild-type at a significance level of 5%, suggesting that some mutations of these variant affect the aptamer binding. The binding signal to B.1.617.2 is the lowest between them. The mutations introduced in RBD are only two residues, L452R and T478K (https://www.ecdc.europa.eu/en/covid-19/variants-concern). Residue 452 is located in a “hull” portion, while residue 478 is located in the outer loop region (PDB entry number: 6m0j, [35]). Based on the modeling results of previous report [36], the replacement of residue 452 may cause to reduce the aptamer binding by steric effect or re-arrangement of the small beta-sheet. Even though the aptamer binding signal to B.1.617.2 is decreased slightly compared to wild-type, it was able to bind all examined variants of SARS-CoV-2 and not to HCoV-OC43 at all that has distinct RBD (Fig. 4b). Hence, this aptamer has a potential to bind to RBD of SARS-CoV-2 strains with high specificity.

Fig. 4

Binding capacities of RBD-Ugu1-1 aptamer to SARS-CoV-2 strains as measured by direct ELAA

(a) The amount of the 5′-biotinylated aptamer and an anti-SARS-CoV-2 RBD antibody captured by immobilized SARS-CoV-2 strains lineage A (wild-type) and B.1.617.2 (delta variant) was determined using SA-HRP and TMB substrate. The absorbance was determined by subtracting the absorbance at 620 nm from that at 450 nm. The experiment was repeated twice with each data point measured in triplicates. Error bars represent standard error. Selection Buffer [40 mM HEPES (pH 7.5), 125 mM NaCl, 5 mM KCl, 1 mM MgCl2, and 0.01% Tween 20] was used as a negative control.

(b) The amount of the 5′-biotinylated aptamer captured by immobilized SARS-CoV-2 strains (1.0 × 106 PFU), including lineage A (wild-type), B.1.1.7 (alpha variant), P.1 (gamma variant), B.1.617.2 (delta variant) of SARS-CoV2 and HCoV-OC43 was determined in the SB containing 0.1% dextran sulfate using SA-HRP and TMB substrate. The absorbance was measured at 450–620 nm. The experiment was repeated twice with each data point measured in triplicates. Error bars represent standard error. One-tailed unpaired T-test was applied to compare the mean of ELAA signals between each strain and buffer [SB; 40 mM HEPES (pH 7.5), 125 mM NaCl, 5 mM KCl, 1 mM MgCl2, and 0.01% Tween 20]. Double asterisks (∗∗) indicate p < 0.01; single asterisk (∗) indicates p < 0.05. For multiple comparisons between SARS-CoV-2 strains, one-way ANOVA with Tukey's multiple comparisons was used.

In conclusion, we have demonstrated that synthesized aptamers by BAB modification, RBD-Ugu1 and RBD-Ugu1-1, efficiently bind with recombinant SARS-CoV-2 S protein and authentic SARS-CoV-2 strains, including wild-type and B.1.617.2, suggesting that these aptamers may be a useful tool for detecting SARS-CoV-2 in the environment.

Declaration of competing interest

This work was partly supported by grants for the Japan Program for Infectious Diseases Research and Infrastructure (JP21wm0225003 and JP21fk0108104j) from and fund from NEC Solution Innovators, Ltd.