Screening and affinity optimization of single domain antibody targeting the SARS-CoV-2 nucleocapsid protein

Chemicals and reagents

The HuSdL-2^TM Phage Display Library were provided by Creative Biolabs (Shirley, NY, USA). The HRP-conjugated anti-M13 phage antibody was obtained from Abcam (Cambridge, UK). Bovine serum albumin (BSA), iso-propylthio-β-D-galactoside (IPTG) and triethylamine were purchased from Yeasen (Shanghai, China). The recombinant human SARS-CoV-2-N were purchased from Genscript. 3,3′,5,5′-Tetramethylbenzidine (TMB) was purchased from Solarbio (Beijing, China). All of the other inorganic chemicals and organic solvents were analytical reagent.

Biopanning

Biopanning of anti-SARS-CoV-2-N antibody was performed based on the instruction manual of phage library kit with a few modifications. In the first round of biopanning, 100 µL of SARS-CoV-2-N protein (13.6 µg/mL) dissolved in 0.01 M (pH 7.2) phosphate-buffered saline (PBS) was added into the MaxiSorp^® ELISA plate (Nunc, Rochester, NY, USA) and then incubated at 4 °C for all night. Then, the plate was blocked with PBS containing 5% skim milk. After washing, the phage library (10¹⁰~10¹¹ plaque forming units (PFU)) was incubated at 37 °C for 60 min, then washed. The bound phages were eluted using 100 µL triethylamine (100 mM) for 30 min at 37 °C, and the eluent was neutralized with 200 µL 1M Tris-buffer (pH 7.4). The eluent was amplified in 10 mL of exponentially in logarithmic phase for the next round of selection.

The inputted phages remained unchanged (1.0 × 10¹¹ phage) in the next two rounds of biopanning, but SARS-CoV-2-N protein concentration was reduced to half compared to last round. The input and recovery of each selective step was assessed to evaluate efficiency of each step. The monoclonal phage DNA was extracted, purified and sequenced by Fuzhou Sunya Biotechnology Co., LTD (Fujian, China). Sequence translation software was employed for analysis of amino acid sequence.

Phage ELISA

A total of 96-well plate was coated with the SARS-CoV-2-N (100 μL, 5 μg/ mL) in coating buffer for all night at 4 °C, then blocked with 200 μL 5% BSA for 2 h at 37 °C. 100 μL of phage (cycle 1–4) was added to the blocked well and incubated at 37 °C for 2 h. The plates were washed three times with 200 μL washing buffer. After that, 5,000-fold diluted solution of anti-M13-HRP was added to the wells. The plate was incubated for at 37 °C 1 h followed by washing again. Then, 100 μL of TMB was added to each well. Finally, the plate was quenched with 1.0 M phosphoric acid, and absorbance was tested at 450 nm.

DNA sequence analysis

The obtained positive monoclonal phage single domain antibody genes were identified by PCR and DNA sequencing. The positive clones selected by PCR using the following primers: Forward Primer L1:5′-TGGAATTGTGAGCGGATAACAATT-3′, Reverse Primer S6: 5′-GTAAATGAATTTTCTGTATGAGG-3′. The sequences of the antibody against SARS-CoV-2-N were analyzed by DNA sequencing. Then, the sequences of positive clones were analyzed by Fuzhou Sunya Biotechnology Co., LTD (Fujian, China). Sequence translation software was used for determination of amino acid sequence.

Bioinformatics

Expression of single domain antibody in E. coli and purification

The expression vector was constructed by fusing with the genes fragment of anti-SARS-CoV-2-N and pET-28a(+)/NdeI/XhoI. Then, E. coli strains BL21(DE3) was used as the expression host and were transformed with expression plasmid. Subsequently, the transformed bacteria were cultured grew overnight at 37 °C in LB medium whose Kanamycin concentration was 20 μg/mL. Then, the cultures were diluted at a ratio of 1:100 in fresh liquid LB medium containing 20 μg/mL kanamycin and cultured at 37 °C with shaking. When the absorbance at 600 nm of the cultures reached 0.6, the cultures were induced by 1 mM IPTG and harvested after 5 h by centrifugation at 10,000 g for 10 min.

Subsequently, the collected E. coli cell was resuspended in 100 mL TGE buffer (pH 7.9) which contained 50 mM Tris, 0.5 mM EDTA, 5% and glycine 50 mM NaCl. And then E. coli cell was disrupted by sonication on ice for 30 min, following with centrifugation at 10,000 g at for 10 min 4 °C. After preliminary inclusion body purification, the precipitate was resuspended and purified with HisPur^™ Ni-NTA Magnetic Beads. Then the purified products were pooled and dialyzed in urea with different gradients at 4 °C.

Antibody binding affinity determination using ELISA

A total of 96-well Maxisorp plates (Nunc) were incubated with 100 μL of 0.7 μg/mL SARS-CoV-2-N at 4 °C overnight. Subsequently, the plates were washed two times with PBS containing 0.1% Tween-20 and then blocked for 2 h with 300 μL of 5% BSA. After washing, the plate was incubated with 6.8 μg/mL anti- SARS-CoV-2-N antibodies at 37 °C for 1 h. The absorbance of 450 nm was measured using an ELISA plate reader (Bio Tek, Winooski, VT, USA).

Fusion protein

We fused different single domain antibodies with (G₄S)₂ as the linker peptide to construct fusion proteins named as pN3-3 and pN3-1-3, respectively. The method of expression, purification and determination referred to 2.6 and 2.7 with slight modification. After washing, each group was coated with 3.4 μg/mL fusion protein in ELISA experiment.

Statistical analysis

All data were analyzed using GraphPad Prism version 8.0 (GraphPad Software, San Diego, CA, USA) and presented as the mean ± SD. were evaluated by two-tailed test. Multiple group comparisons and two groups were tested using one-way analysis of variance and two-tailed test, respectively. P < 0.05 was determined to be significant.

Screening of specific single domain antibody against SARS-CoV-2-N

HuSdL-2^™ phage library was screened for three rounds to obtain anti-SARS-CoV-2 N protein single domain antibody. The uncoated antigen wells were set as a negative control during each round of screening, and the titer of the eluted phages was tested after each round of screening. The recovery rate is an effective indicator for estimating the degree of screening enrichment. It was shown that with the number of screening rounds increasement, the recovery rates increased, which indicating the effectiveness of the enrichment screening. The affinity increased as the biopanning rounds went on in Fig. 1A, suggesting that the specificity of phages was effectively enriched (Table 1).

DNA was amplified by PCR with primer L1 and primer S6 and the clones containing the complete sequence (approximately 700 bp) were verified by 1% nucleic acid gel (Fig. 1B). Positive clones were selected for DNA sequencing to identify their sequences (Fig. 1C). Finally, eight anti-SARS-CoV-2-N single domain antibody gene sequences were obtained named pN01-pN08, respectively. As is shown in Fig. 1D, the conserved residues at the positions of Val37, Gly44, Leu45 and Trp47 were all conservative disability amino acids in the eight different sequences.

Bioinformatics analysis of antibody against SARS-CoV-2-N

The gene sequence was converted into amino acid sequence by Gene Runner and online software Expasy (https://web.expasy.org/protparam/) was employed to calculate the relevant biochemical information of the sequence. The molecular mass of antibodies is about 13 kDa, the half-life in the mammalian cell is 0.8 h and the instability coefficients of all antibodies are less than 40 that means all single domain antibodies are stable protein. pN01, pN02, pN03 and pN04 are acidic proteins, pN07 is neutral protein, and pN05, pN06 and pN08 are basic proteins (Table 2). pN1 and pN3 were selected to construct fusion proteins named pN3-3 and pN3-1-3, respectively.

Binding pattern of the single domain antibody against SARS-CoV-2-N

Swiss-model software was employed for constructing the 3-D structure of single domain antibody (Fig. 2A). The single-domain antibody 3D model presented a single antigen-binding variable domain (VHH) and classical immunoglobulin fold similar to that of the alpaca antibody, with two layers of antiparallel beta chain folded composition, forming a classic sandwich-like structure. Each nanobody had three highly variable rings called CDRS, which were major regions that undergo genetic changes to present antigen-specific binding surfaces. The single-domain antibody 3D model had the characteristic of “VHH structural antibody”, which does not require complex folding and can be expressed in the prokaryotic system. HawkDock (http://cadd.zju.edu.cn/hawkdock/) was employed to study the molecular docking between single domain antibody against the N-terminal domain protein of 2019-nCoV nucleocapsid (SARS-CoV-2-N-NTD; PDB code: 7CDZ) and the C-terminal domain protein of 2019-nCoV nucleocapsid (SARS-CoV-2-N-CTD; PDB code: 7CE0). The amino acid sequences for SARS-CoV-2-N-NTD and SARS-CoV-2-N-CTD were obtained from the Protein Data Bank (https://www.rcsb.org/). It was showed that pN06 mainly interacted with SARS-CoV-2-N-CTD and the other single domain antibodies mainly interacted with SARS-CoV-2-N-NTD (Fig. 2B).

The protein-ligand complex was further analyzed and visualized through Pymol software and Ligplus. As shown in Fig. 2C and Table 3, five hydrogen bond interactions were observed between SARS-CoV-2-N-NTD and pN01, with Asp31 (2.00 Å), Glu32 (2.34 Å), Asp53 (2.90 Å), Arg98 (2.83 Å) and Tyr109 (3.29 Å), respectively. Moreover, pN01 formed hydrophobic interactions with SARS-CoV-2-N-NTD. Focusing on pN07 and pN08, the amino acids and hydrogen bonds involved in the binding interface of the two proteins are the same. According to previous studies, SARS-CoV-2-N had high binding affinity to 11 kinds of human chemokines (Alturaiki et al., 2023). Increase in chemokines can attract immune cells to the infected site, leading to inflammation and tissue damage (Catalán, Gómez-Ambrosi & Rodríguez, 2013). On the basis of the molecular docking map, the combination situations of CCL5, CCL21 and CXCL4 with SARS-CoV-2-N were similar to that of pN01-pN08. As shown in Fig. 2D, CCL5, CCL21 and CXCL4 mainly interacted with SARS-CoV-2-N-NTD. The binding affinity between protein complexes were calculated using HawkDock tool. The binding affinity of CCL5, CCL21 and CXCL4 were more than or equal to −9.0. The binding affinity of all the single domain antibodies were less than −9.0 except the pN05 (Table 4). It was suggested that there was a competitive relationship between pN01-pN08 and chemokines to bind SARS CoV-2-N, and pN01-pN08 might inhibit the combination between SARS-CoV-2-N and chemokines to reduce inflammatory response.

Construction, expression, and purification of the single domain antibody against SARS-CoV-2-N

The vector was conducted with a C-terminal 6× His tag for purification (Fig. 3A). The recombinant plasmid was verified by double enzymatic digestion and digested to two bands, one of which was approximately 700 bp and the other band was approximately 4,500 bp (Fig. 3B). This indicated that the insertion of the gene was correct. DNA sequencing analysis result also suggested that plasmid was successfully constructed (Fig. 3C). Then, the ligated product was transformed into E. coli BL21 (DE3) cells. Colony PCR result showed a band of approximately 700 bp (Fig. 3D). The subsequent expression of the single domain antibody was induced with IPTG and the molecular weight of single domain antibody approximately was 13 kDa. The purification of single domain antibody was performed with HisPur^™ Ni-NTA Magnetic Beads and the purity of target protein was more than 95% (Fig. 3E). The purified product was identified by western blot and the result showed that the purified product was target protein (Fig. 3F).

Binding activity evaluation of the single domain antibody against SARS-CoV-2-N

The binding activity of the antibodies to SARS-CoV-2-N were determined by ELISA. As shown in Fig. 3G, all antibodies could specifically bind to SARS-CoV-2-N. Furthermore, pN03 had best affinity with SARS-CoV-2-N in all single domain antibodies (Fig. 3G).

Binding pattern of SARS-CoV-2-N to fusion proteins against SARS-CoV-2-N

Swiss-model software was used to construct the 3D structures of fusion proteins (Fig. 4A). The molecular docking between fusion proteins and SARS-CoV-2-N-CTD or SARS-CoV-2-N-NTD were analyzed by HawkDock software and Ligplus. Figs. 4B and 4C showed the interaction between fusion proteins and SARS-CoV-2-N-CTD. As shown in Table 5, pN3-3 formed five hydrogen bond interactions and multiple hydrophobic interactions with SARS-CoV-2-N-CTD. The hydrogen bond interactions were observed at Thr9 (3.23 and 2.89 Å), Gln18 (2.47 Å), Lys113 (3.12 Å) and Asp199 (2.38 Å) of pN3-3. pN3-1-3 showed two hydrogen bond interactions with SARS-CoV-2-N-CTD at same amino acid of pN3-3 at Thr9 and Lys113. The binding affinity of pN3-3 and pN3-1-3 was −10.8 and −12.7, respectively (Table 6).

Construction, expression, and purification of the fusion proteins against SARS-CoV-2-N

The construction process of the fusion proteins was similar to that of single domain antibody. In brief, the constructed product was transformed into competent cells. Colony PCR verified the inserted gene was correct. The PCR products of pN3-3 was approximately 800 bp and the PCR products of pN3-1-3 was approximately 1,500 bp (Fig. 5A). To confirm the cloning, plasmid extracted from transformed cells was digested by XhoI and MluI. Results showed pN3-1-3 was about 6,000 bp, and pN3-3 was about 4,500 bp (Fig. 5C). DNA sequencing analysis was consistent with the above results (Fig. 5B). SDS-PAGE results confirmed that the protein was greatly induced after the addition of IPTG. The molecular weight of pN3-3 was approximately 27 kDa and the molecular weight pN3-1-3 was approximately 40 kDa (Table 6). The recombinant fusion protein was purified by HisPur^™ Ni-NTA Magnetic Beads, the purity of target protein was more than 95% (Fig. 5D). ELISA was performed to compare the binding activity of the two fusion proteins. As is shown in Fig. 5E, all fusion proteins could specifically bind to SARS-CoV-2-N and the affinity of fusion proteins is stronger than that of single domain antibody.