VNAR development through antigen immunization of Japanese topeshark (Hemitriakis japanica)

Abstract

The VNAR (Variable New Antigen Receptor) is the smallest single-domain antibody derived from the variable domain of IgNAR of cartilaginous fishes. Despite its biomedical and diagnostic potential, research on VNAR has been limited due to the difficulties in obtaining and maintaining immune animals and the lack of research tools. In this study, we investigated the Japanese topeshark as a promising immune animal for the development of VNAR. This shark is an underutilized fishery resource readily available in East Asia coastal waters and can be safely handled without sharp teeth or venomous stingers. The administration of Venus fluorescent protein to Japanese topesharks markedly increased antigen-specific IgM and IgNAR antibodies in the blood. Both the phage-display library and the yeast-display library were constructed using RNA from immunized shark splenocytes. Each library was enriched by biopanning, and multiple antigen-specific VNARs were acquired. The obtained antibodies had affinities of 1 × 10^-8 M order and showed high plasticity, retaining their binding activity even after high-temperature or reducing-agent treatment. The dissociation rate of a low-affinity VNAR was significantly improved via dimerization. These results demonstrate the potential utility of the Japanese topeshark for the development of VNAR. Furthermore, we conducted deep sequencing analysis to reveal the quantitative changes in the CDR3-coding sequences, revealing distinct enrichment bias between libraries. VNARs that were primarily enriched in the phage display had CDR3 coding sequences with fewer E. coli rare codons, suggesting translation machinery on the selection and enrichment process during biopanning.

Keywords: Japanese topeshark; VNAR; biopanning; deep sequencing; phage display; yeast display.

FIGURE 1

Induction of antigen-specific antibodies in the plasma of Japanese topesharks immunized with Venus protein. (A) Japanese topeshark. (B,C) IgM and IgNAR antibody titers in immunized Japanese topesharks. Two Japanese topesharks were immunized with Venus once every 2 weeks. Antibody titers in the plasma were quantified via direct ELISA. (B) IgM. (C) IgNAR. The mean and SD are shown (n = 4). Arrowheads indicate immunizations. (D) IgNAR antibody titer in long-term immunization. Four Japanese topesharks were immunized with Venus once every 2 weeks. The mean and SD are shown (n = 4). Arrowheads indicate immunizations.

FIGURE 2

Construction and enrichment of VNAR-displaying libraries from Japanese topeshark. (A) Illustration of VNAR-displaying M13 phage. (B) The process of biopanning using an immuno tube to enrich for specific VNAR sequences from the phage library. (C) Phage ELISA. The wells were coated with Venus or BSA, after which VNAR-, displaying phage at a concentration of 1 × 10¹¹ pfu/mL, was applied. (D) VNAR displayed on a yeast via Aga1 and Aga2. (E) Enrichment of yeast libraries using MACS. (F) Flow cytometry. Yeast library was mixed with Venus protein and APC-conjugated anti-AGIA antibody and then applied to a cell analyzer.

FIGURE 3

Deep sequencing analysis of VNAR display libraries. (A) An overview of the deep sequencing analysis. The total counts of CDR3 reads and unique CDR3 sequences analyzed during each biopanning round are presented. (B) The diversity of CDR3 sequences throughout the biopanning process. An amino acid sequence homology analysis was conducted on the top 100 most abundant CDR3 sequences from each biopanning round. Sequence clustering was performed using the Clustal Omega website, and phylogenetic trees were visualized using the FigTree application. Each sequence is represented using a circle, with variations in the size and color of the symbol denoting differences in abundance (the ratio of the number of sequences to the total CDR3 sequences in the round). The top 10 CDR3s in terms of abundance before (input) and after (output, biopanning round 4) biopanning are displayed as follows: Y, yeast; P, phage; R0, the input library; R4, the library enriched via 4-round biopanning; the last number indicates the ranking of abundance. The detailed amino acid sequence and abundance of each CDR3 are available in Supplementary Table S1. Major clusters in round 4 are highlighted with blue circles. (C,D) The lengths of CDR3 sequences in both VNAR display libraries. Histograms depict the number of CDR3 sequences for each length. (C) Input (phage, 7,259 unique sequences; yeast, 6645); (D) round 4 (phage, 2469; yeast, 2879). (E,F) A comparison of abundance in both VNAR display libraries for each CDR3 sequence. The abundance of each CDR3 sequence is plotted on the horizontal axis for the phage display and on the vertical axis for the yeast display, respectively. Only CDR3 sequences detected in both libraries are shown, while sequences detected in just one of the libraries are excluded from the analysis. (E) Input; (F) round 4.

FIGURE 4

CDR3 sequences enriched by biopanning. (A) The sequence and phylogenetic tree of the enriched CDR3 sequences. Sequence alignment was conducted using the MAFFT website, and the phylogenetic tree was visualized using the FigTree application. (B,C) Changes in the abundance of CDR3 sequences throughout the biopanning process. Detailed abundance data are available in Supplementary Table S2. (B) CDR3s that increased after biopanning in both phage and yeast display libraries; (C) CDR3s that increased in either library.

FIGURE 5

Codon usage analysis. All DNA sequences encoding a CDR3 were extracted from the deep sequencing results. Codon codes that appeared were classified into 3 groups, focusing on the frequency of codon use in either bacteria or yeast³⁵. Bacterial codon usage was applied to the analysis of phage library sequences, and yeast codon usage was applied to yeast library sequences. The figures in the bars indicate the number of codons. (A) Phage display libraries; (B) yeast display libraries.

FIGURE 6

AlphaScreen binding assay of VNARs with enriched CDR3s. VNAR clones with CDR3 sequences enriched using biopanning were, respectively, selected (Figure 3B; Supplementary Table S2) and were fused to human IgG Fc to produce recombinant VNAR-Fc antibodies. The binding of each recombinant VNAR-Fc to biotinylated Venus proteins was detected using AlphaScreen (orange circles and lines). Biotinylated DHFR was used as a negative control (gray circles and lines). The mean and SD are shown (n = 4). Graphs were arranged in order of strongest binding response.

FIGURE 7

Functional analysis of three recombinant VNAR clones. (A) Binding assay to denatured antigen. Membranes on which denatured Venus proteins were blotted were treated with the respective antibodies. (B–D) Kinetics assay with surface plasmon resonance. Venus-His protein was immobilized as a ligand on a CM5 sensor chip, and purified VNAR-AGIA-His was applied as an analyte. (E,F) Binding assay of bivalent VNAR with CDR3 P-R4-01 to Venus. (E) VNAR-rabbit Fc dimer; (F) VNAR-GS linker-VNAR-His tandem VNAR dimer. (G) Tolerance test for high temperature. VNAR-AGIA-His antibodies were incubated at temperatures ranging from 25°C to 90°C for 1 h. After the samples were allowed to incubate at room temperature for 30 min, binding to Venus was confirmed using AlphaScreen. An anti-GFP mouse monoclonal antibody (clone mFX73, Fuji Film Wako Pure Chemical) was used as a control. The mean and standard deviation are shown (n = 4). (H) Tolerance test for reducing agents. VNARs were mixed with 2–8 mM DTT and incubated for 30 min. They were then diluted 100-fold in 5% skim milk-PBS and reacted with Venus-coated ELISA plates. The mean and standard deviation are shown (n = 4).