Efficient Identification of Monoclonal Antibodies Against Rift Valley Fever Virus Using High-Throughput Single Lymphocyte Transcriptomics of Immunized Mice

Abstract

Background: Rift Valley fever virus (RVFV) is a zoonotic virus that poses a significant threat to both livestock and human health and has caused outbreaks in endemic regions. In humans, most patients experience a febrile illness; however, in some patients, RVF disease may result in hemorrhagic fever, retinitis, or encephalitis. While several veterinary vaccines are being utilized in endemic countries, currently, there are no licensed RVF vaccines or therapeutics for human use. Neutralizing antibodies specifically targeting vulnerable pathogen epitopes are promising candidates for prophylactic and therapeutic interventions. In the case of RVFV, the surface glycoproteins Gc and Gn, which harbor neutralizing epitopes, represent the primary targets for vaccine and neutralizing antibody development. Methods: We report the implementation of advanced 10x Genomics technology, enabling high-throughput single-cell analysis for the identification of rare and potent antibodies against RVFV. Following the immunization of mice with live attenuated rMP-12-GFP virus and successive Gc/Gn boosts, memory B cell populations (both general and antigen-specific) were sorted from splenocytes by flow cytometry. Deep sequencing of the antibody repertoire at a single-cell resolution, together with bioinformatic analyses, was applied for BCR pair selection based on their abundance and specificity. Results: Twenty-three recombinant monoclonal antibodies (mAbs) were selected and expressed, and their antigen-binding capacities were characterized. About half of them demonstrated specific binding to their cognate antigen with relatively high binding affinities. Conclusions: These antibodies could be used for the future development of efficacious therapeutics, as well as for studying virus-neutralizing mechanisms. The current study, in which the single-cell sequencing approach was implemented for the development of antibodies targeting the RVFV surface proteins Gc and Gn, demonstrates the effective applicability of this technique for antibody discovery purposes.

Keywords: 10x Genomics; RVFV Gc glycoprotein; RVFV Gn glycoprotein; Rift Valley fever virus (RVFV); antibody discovery; antibody repertoire profiling; recombinant antibody; single-cell analysis.

Figure 1

Single-cell BCR repertoire-profiling experimental approach implemented in the current study for identification of RVFV-specific antibody sequences. (a) The mouse immunization regimen consisted of priming and three boosts with live attenuated RVFV rMP-12, followed by recombinant antigen administration. (b) Eleven days after the last boost, splenocytes were isolated, stained, and subjected to sorting of IgG-presenting B cells and antigen-specific sub-populations. (c) Total transcriptome, full V(D)J, and capture barcode sequences (antigen-specific barcode acquired by the dCODE Klickmer^®-PE reagent) were determined at the single-cell level using 10x Genomics technology and high-throughput sequencing (HTS). (d) The data analysis enabled the determination of phenotypic subsets of splenocytes and, most importantly, the selection of BCR pairs for recombinant expression. (e) The binding capacities of the expressed mAbs were characterized. The figure was created with BioRender.com.

Figure 2

Specific humoral responses of RVFV-immunized mice. ELISA assessments of polyclonal reactivity of sera collected from mice prior to spleen harvesting (11 days after last boost). (a,b) The binding capacities against rGc (a) and whole live virus (b) of the serum samples collected from m1–m4. (c,d) The binding capacities against rGn (c) and whole live virus (d) of the serum samples collected from m5. The data represent average of duplicates ±SD. (e) The calculated Dil50 values (the dilution factor at 50% of maximal binding) of each tested serum sample.

Figure 3

Single-cell phenotypic profiling. Gene expression profiling (scRNA-seq) using 10x Genomics platform and unsupervised cell clustering was performed according to differentially expressed markers (using Seurat). (a,b) Heatmaps of the top 10 differentially expressed genes in each cluster of the Gc-sample and Gn-sample, respectively. (c,d) UMAP visualization of individual cells (each point represents a single barcode, which corresponds to a single cell) in the Gc-sample and Gn-sample, respectively. The plots indicate the cluster numbers as well as the number of cells in each cluster (in brackets). Each cell is colored according to its annotation: B.GC, germinal center B cell; B.Fo, follicular B cells; B1a cells; MΦ, macrophages; and T cells. (e,f) UMAPs indicating cells expressing immunoglobulin heavy chain isotypes (IGHG1, IGHG2, and IGHM) in the Gc-sample and Gn-sample, respectively. NA indicates cells with no BCR information or isotypes represented by less than 1% of the cells.

Figure 4

Characterization of Gc-sample and Gn-sample recombinant mAbs. (a) The specificity of the recombinant mAbs selected from the Gc-sample (Gc_mAbs) or Gn-sample (Gn_mAbs) was assessed by ELISA against rGc, rGn, and BSA. The chimeric antibody HN_2 was used as an isotype control. The data represent an average of technical triplicates ±SD and are representative of at least two independent experiments. (b,d) The reactivity profiles of the indicated mAbs, as determined by ELISA, against Gc ((b) Gc_mAbs) or Gn ((c,d) Gn_mAbs). The values represent the average of technical triplicates ±SD. (e). The kinetic binding parameters of the indicated mAbs were determined using biolayer interferometry. Increasing concentrations of each mAb were incubated with immobilized biotinylated rGc or rGn. The binding sensograms were fitted using the 1:1 binding model.