Expansion of the HSV-2-specific T cell repertoire in skin after immunotherapeutic HSV-2 vaccine

Abstract

The skin at the site of HSV-2 reactivation is enriched for HSV-2-specific T cells. To evaluate whether an immunotherapeutic vaccine could elicit skin-based memory T cells, we studied skin biopsies and HSV-2-reactive CD4⁺ T cells from peripheral blood mononuclear cells (PBMCs) by T cell receptor β (TRB) sequencing before and after vaccination with a replication-incompetent whole virus HSV-2 vaccine candidate (HSV529). The representation of HSV-2-reactive CD4⁺ TRB sequences from PBMCs in the skin TRB repertoire increased after the first vaccine dose. We found sustained expansion after vaccination of unique, skin-based T-cell clonotypes that were not detected in HSV-2-reactive CD4⁺ T cells isolated from PBMCs. In one participant a switch in immunodominance occurred with the emergence of a T cell receptor (TCR) αβ pair after vaccination that was not detected in blood. This TCRαβ was shown to be HSV-2-reactive by expression of a synthetic TCR in a Jurkat-based NR4A1 reporter system. The skin in areas of HSV-2 reactivation possesses an oligoclonal TRB repertoire that is distinct from the circulation. Defining the influence of therapeutic vaccination on the HSV-2-specific TRB repertoire requires tissue-based evaluation.

Fig. 1.. Number, fold change, and clonality of TCR clonotypes in HSV lesion site and arm biopsies before and after vaccination.

(a) Schematic of vaccine study timeline and procedures. CD4⁺ (b) and CD8⁺ (c) T-cell densities from biopsies from control skin at the time of enrollment and from the site of a symptomatic lesion (N=4) are shown in comparison to HSV lesion site and control skin biopsies over the course of a 3-dose vaccine trial in nine vaccine recipients. Each dot represents the mean of three counted sections in a single participant. Median and interquartile range are shown in grey. (d) Representative 10x micrographs of CD4⁺ (green) and CD8⁺ (red) T-cell density by IFA in HSV lesion site biopsies at specified time points before and after vaccination. CD4⁺ and CD8⁺ T-cell IFA from (e) control skin and (f) lesion site during a symptomatic HSV-2 outbreak. All images are from P4. White bars are 100μm. (g) Total and (h) number of TCRβ clonotypes detected at >4 copies are shown from the HSV lesion site and arm biopsies. (h) Clonality calculated from Shannon entropy of the TCR repertoire from each sample. Each dot represents a single participant. Median and interquartile range are shown in grey. For comparisons, *p <0.05, **p <0.01 by paired Wilcoxon.

Fig. 2.. Overlap of HSV-2-reactive CD4⁺ T cells from PBMCs in skin biopsies by clonal tracking before and after initiation of vaccination series.

(a) Unique TRB sequences detected only in blood at day 0 (‘blood only’), clonotypes from blood at day 0 also detected at any time in skin (‘both’) and unique clonotypes in skin at day 0 (‘skin only’) and (b) the longitudinal detection of each of the overlapping (‘both’) clonotypes in lesion-area skin over time, by whether they were observed before or after vaccination. Each line represents a unique clonotype. “ND” is not detected at that time point. (c) Unique TRB sequences detected in blood and skin at day 10 by detection in one or both sites and (d) the longitudinal detection of each of the overlapping clonotypes in lesion-area skin over time. Each line represents a unique clonotype. “ND” is not detected at that time point. Shading of both histograms and longitudinal graphing represents whether clonotypes were observed before (purple) or only after (red) vaccination. (e) All HSV-2-reactive CD4⁺ T cells from PBMCs identified in P4 at day 10 at >10 copies and their representation in serial tissue biopsies after immunotherapeutic vaccination. Empty boxes indicate no detection. Each horizontal row is a distinct nucleotide sequence. Quantitation of PBMC copy number at day 10 (but not day 0) in this participant is influenced by T cell ex vivo expansion prior to stimulation and sequencing (Fig. S1, method 2).

Fig. 3.. Prevalent and elicited clonotypes in HSV-2-enriched CD4+ T cells from PBMCs (“PBMC”), healed lesion site (“HSV lesion-area skin”), and arm biopsy expanding or detected at high copy number.

(a) Prevalent clonotypes detected in P4, by fold increase over dose 1. Each row represents a single clonotype (by nucleic acid sequence). Co-detection of clonotypes in HSV-2-enriched CD4+ T cells from PBMCs, active lesion, quiescent lesion-area skin and arm skin is compared at each of the time points. Number of copies detected in PBMCs from P4 at day 10 reflect ex vivo expansion prior to sequencing and are denoted by # (see Methods). Blank boxes indicate lack of detection. (b) Prevalent and (c) elicited clonotypes either expanding ≥6 fold after dose 1 or present after dose 1 (day 10) at ≥6 fold above a single copy and their longevity in lesion area tissue over the course of the vaccine trial (left), with the corresponding abundance in the arm control (right). ND = not detected. Highly prevalent clonotypes that expanded after the first vaccine dose are seen to persist in tissue throughout the vaccine trial. Color indicates whether the nucleotide sequences were detected in blood (black indicates a tissue-only sequence, red was also seen in blood). Prevalent expanded clonotypes are of greater abundance than elicited clonotypes. Most in-tissue expansion was not from clonotypes observed in blood. (d) Number of clonotypes expanding by ≥6 fold after each dose in the lesion-area skin by detection prior to initiation of vaccine series (prevalent) vs clonotypes only seen after initiation of vaccination (elicited).

Fig. 4.. Evaluation of TRA/TRB V and J gene usage.

(a) V gene usage in participants 1-6 across all sites, blood, arm, and genital skin, over time. Pie demonstrates proportion of the total repertoire represented by each gene. Skin site clonotypes were limited to those that were present at greater than 4 copies; all HSV-reactive CD4⁺ T cells from blood are shown. (b) Combination V and J gene usage in P4 showing the shift in immunodominance from two years prior to vaccination through ten days after the first vaccine dose. Z axis represents numbers of copies in each combination. X and y axes represent V and J genes listed sequentially. The V and J genes are labeled for the most abundant combinations. (c) HSV-2-specificity of a synthetic TCR composed of the most abundant TRA and TRB sequences from the day 10 sample in P4, using a Jurkat Nur77-mNeonGreen reporter cell system (right). Negative control (‘no TCR’, left) and positive control (PMA + Ionomycin, middle) for TCR stimulation and negative treatment (‘no HSV’, top) or experimental treatment (UV-irradiated HSV, bottom) to assess TCR specificity.

Fig. 5.. Representative data from fine specificity determination of blood CD4 T-cell clones overlapping with TCRβ CDR3 sequences detected in HSV lesion site biopsies.

(a) Clonotype tracking with assigned specificity of 13 of the 16 clones queried in P4 by abundance and fold change in skin over dose one (day 0 to 10). (b) Example of fine-specificity mapping of a single clone confirmed to be specific to UL11. Both specimens were obtained from day 10 after HSV529 vaccination. At left is T-cell proliferation in response to matrix pools of HSV-2 antigens with positive and negative controls. Pools containing US11 (Pool 1, Pool 19) are positive. At right is confirmatory assay with recombinant UL11 and controls. Blue and orange bars represent replicate assays. (c) Logoplot representation of 3 sequence-similar (NN-distance <13) clonotype clusters in HSV-2 reactive CD4+ T cells from blood and clonotypes expanding after dose 1. (d) Number of shared HLA-II alleles (among DR, DQ, DP) between participants.