CD4 + T-cells are central organisers of anti-viral immune responses11, 41 and dominate memory T-cell responses against the major vaccine target for SARS-CoV-2, S4, 8–10. Studies of SARS-CoV-2-specific CD4 + T-cells have largely examined responses to whole antigens using pools of overlapping peptides. Although S-derived epitopes have been reported12–14, 43, 44, for most, epitope generation from native antigen has not been demonstrated and HLAII restrictions have not been defined. Critically, the current lack of fundamental understanding of SARS-CoV-2-specific CD4 T-cell immunity has prevented important questions being investigated at the epitope level. These include the extent of epitope-specific CD4 + T-cell cross-reactivity with other b-HCoVs and the ability of epitope-specific T-cells induced by ancestral SARS-CoV-2 infection or vaccination to protect against VOCs.
To enable detailed investigation of SARS-CoV-2-specific CD4 + T-cells, we studied HCW infected during the first wave of infection in the UK. In line with previous ex vivo studies assessing whole antigens4, 8–10, we detected robust CD4 + T-cell memory against S, M and N peptide pools in HCW-PI. These varied in magnitude between individuals but were consistently higher in CD8-depleted than matched whole PBMCs, reflecting the numerical dominance of CD4 + T-cell memory4, 8–10. Mapping of responses in HCW-PI following in vitro expansion, showed CD4 + T-cell reactivity to multiple peptides spread across S, M and N in every individual. This sensitive method allows detection of low frequency cells or central memory cells not detectable in ex vivo IFNg assays28, 42. Thus, the total number of SARS-CoV-2 epitope-specific responses induced by infection in each individual may be larger than previously estimated12, 15.
Through isolation and characterisation of over 100 CD4 + T-cell clones we identified 21 HLAII-restricted epitopes; 17 in S, two in M and two in N. This greatly expands the number of known CD4 + T-cell epitopes and provides experimental confirmation and HLAII restriction for some previously reported12–14, 43, 44. Furthermore, all 17 S-derived epitopes we describe were processed and presented from low quantities of protein demonstrating their biological relevance45. Interestingly, the epitopes we characterised were skewed towards HLA-DP and HLA-DR restriction alleles, a pattern also noted for SARS-CoV-2 by others15, 28 and distinct from other human viruses investigated using T-cell clones38, 46, 47. Responses to the same S epitopes were detected in individuals of appropriate HLAII genotype after either SARS-CoV-2 infection or vaccination, indicating that HLAII genotype is a key determinant of the SARS-CoV-2 S-specific CD4 + T-cell response regardless of the route of antigen exposure. We also observed epitopes restricted through prevalent HLAII alleles such as the HLA-DRB1*04 subtypes, DRB1*01:01, DRB1*15:01 and DPB1*04:0148. Thus, it is likely that many of the responses will be widespread in the population.
In addition to SARS-CoV-2, several endemic HCoVs infect humans causing mild disease. If T-cells elicited by prior infection with these viruses cross-reacted with SARS-CoV-2 they could modulate the course of disease. Consistent with several groups, we observed low-frequency CD4 + T-cell responses to SARS-CoV-2 S, M and N in people uninfected by SARS-CoV-2 using high peptide concentrations in ex vivo assays15–18. However, in this study, all the S-specific CD4 + T-cell clones derived from SARS-CoV-2-infected individuals efficiently recognised SARS-CoV-2 S pepmix but never recognised S pepmixes from the endemic OC43 and HKU1 b-HCoV viruses. This suggests strongly that these responses arose from naïve T-cells within the repertoire rather than pre-existing memory. Similar results were recently observed for CD8 + T-cells against the immunodominant HLA-B7:02-restricted N epitope49. Indeed, the only cross-reactivity we observed was within the highly conserved S2 region of SARS (90% aa similarity) to which the UK HCWs studied had never been exposed. Collectively these data argue for limited expansion of pre-existing cross-reactive S-specific CD4 + T-cells during SARS-CoV-2 infection.
This view is consistent with prior reports of S cross-reactivity that suggest cross-reactive CD4 + T-cells are rare components within a limited number of S epitope-specific responses15, 16, 21. At the epitope level, spike-specific CD4 + T cell cross-reactivity has mainly been observed within the S2 809–830 linker region (encompassing the HLA-DRB1*08:01 SFIE epitope studied here); in polyclonal T-cell lines16, 19, T-cell clones15, 21 and using HLA-DP*04-restricted MHCII tetramers14. Compared to SARS-CoV-2 primed T-cell responses, the HCoV primed CD4 + T-cells were present at low frequency in the blood14, had 5-fold lower functional avidity against SARS-CoV-2 peptides21 and exhibited less expansion following SARS-CoV-2 infection14.
In our study we only investigated cross-reactivity of CD4 + T cells specific for S, a protein highly targeted by mutation. In comparison, other viral proteins are more conserved across HCoVs and T-cell cross-reactivity is therefore more likely. Accordingly, in SARS-CoV-2 unexposed individuals, pre-existing T cell immunity has been reported against the more highly conserved N and ORF1ab-encoded NSP proteins16, 17, 50. HCoV cross-reactive T cells in SARS-CoV-2-seronegative HCWs were frequently directed against epitopes within the early expressed replication transcription complex (encompassing Nsp7, Nsp12 and NSp13)50. Interestingly, these expanded following putative abortive SARS-CoV-2 infection in exposed HCW who remained seronegative, consistent with a role in protection from overt infection. Ultimately, the extent to which pre-existing cross-reactive T-cell immunity contributes to controlling SARS-CoV-2 infection in an individual will be determined by a complex combination of factors including the conservation of the epitopes presented by their HLA genotype as well as their TCR repertoire and previous history of HCoV exposure14.
The SARS-CoV-2 pandemic has been exacerbated by emergence of VOCs with mutations in S that reduce the neutralisation capacity of antibodies induced against ancestral S sequences24–26. Because spike is the only viral antigen in the most widely used vaccines, understanding how sequence mutations affect T-cell recognition is vital for vaccine development efforts and understanding the potential impact of emerging VOCs for vaccinated individuals. Studies of total S-specific T-cells have shown minimal reductions in the overall frequency of response to Alpha, Beta, Gamma, Delta VOCs26, 31, 51 and Omicron33, 34, 52 in previously infected or vaccinated individuals. However, the use of high concentrations of peptide pools in these studies masks biologically important differences in epitope-recognition efficiency35.
We identified aa substitutions or deletions in 10 of 17 S CD4 + T-cell epitopes; some were common to multiple VOCs while others were unique to particular viral isolates. Combining data from T-cell clones and experiments using ex vivo PBMCs we found variable effects of intra-epitope mutations on CD4 + T-cell recognition. As expected, triple aa deletion and multiple aa substitutions within individual epitopes had the greatest impact on CD4 + T-cell recognition53. Of note, the GGNY and FNCY epitopes within the frequently mutated RBD region were eliminated in Omicron, despite point mutations in earlier VOCs being recognised. Overall, three of the four epitopes mutated in Omicron were no longer recognised by epitope-specific CD4 + T cells. However, the effect of point mutation is complex. A single mutation dramatically decreased T-cell recognition (e.g. SGTN D80Y, P2; TLVK N969K, Omicron) or had no effect (LSET T307A, P3; STEC N764K, Omicron; TYVT A1022S, Beta). The epitope mapping we present therefore provides a rational basis for VOC risk stratification. It highlights the need for further careful experimental definition of immunodominant CD4 + T-cell epitopes, identification of the essential aa required for HLAII binding and TCR engagement, and the need to consider each epitope individually.
In conclusion, our study demonstrates the fine sensitivity of SARS-CoV-2 S-specific CD4 + T-cells to aa differences in epitope sequence and the potential for SARS-CoV-2 evasion of ancestral virus or vaccine-elicited CD4 + T-cells through viral evolution. The breadth of SARS-CoV-2 S epitopes targeted in every individual shows current VOC mutations are likely to have only limited impact on overall CD4 + T-cell surveillance. However, continued mutation of SARS-CoV-2 could lead to further epitope loss highlighting the importance of continued virus sequence monitoring of emerging VOCs.