This study investigates how the conformational state of the SARS‑CoV‑2 spike glycoprotein influences CD4+ T‑cell epitope dominance through proteolytic processing in the MHC class II pathway. Using antigen processing likelihood (APL) modeling across spike conformations and limited proteolysis of stabilized recombinant spike ectodomains (spike‑2P and spike‑6P), the authors mapped protease‑sensitive sites and compared these to reported CD4+ epitope frequencies. They report that protease‑sensitive sites coincide with conformationally unstable domains and that peptides from such unstable, solvent‑exposed regions are more frequently dominant after infection than after vaccination. Differences in spike conformation and exposure circumstances are proposed to influence antigen‑presenting cell behavior, spike fragmentation, peptide loading, and subsequent CD4+ epitope dominance.
CD4+ T cells contribute substantially to adaptive immunity against SARS‑CoV‑2 by supporting antibody responses, promoting CD8+ T‑cell development, recruiting innate effectors, and regulating inflammation. Measuring antigen‑specific CD4+ responses typically employs peptide restimulation readouts such as Elispot or activation‑marker flow cytometry. CD4+ epitope dominance refers to a small subset of peptides that elicit disproportionately strong or frequent responses after priming with intact antigen. Dominant CD4+ epitopes often cluster spatially and are shaped by antigen processing pathways; prior work with HIV Env and influenza hemagglutinin supports a model in which conformationally unstable segments are preferentially cleaved and adjacent sequences are presented by MHCII (antigen processing likelihood, APL).
The SARS‑CoV‑2 spike trimer has multiple domains and conformational states relevant to receptor engagement and fusion. Conformational fluctuations, cleavage at the S1–S2 and S2′ sites, and domain linkers could influence class II pathway proteolysis and hence CD4+ epitope presentation. Circumstances of antigen exposure—infectious virus versus mRNA vaccination—can alter antigen‑presenting cell types, endosomal processing, and spike conformations encountered by processing enzymes, potentially shifting epitope dominance.
The authors applied an APL approach to multiple spike conformations to predict regions of conformational instability and likely proteolytic vulnerability. Structures evaluated included S1 in all‑RBD‑down and 3‑RBD‑up conformations and S2 in pre‑fusion and post‑fusion conformations. Predicted APL profiles were compared to empiric CD4+ epitope mapping data available in the Immune Epitope Database (IEDB). To probe cleavage sensitivity experimentally, limited proteolysis was performed on soluble recombinant spike ectodomains stabilized by proline substitutions (spike‑2P and spike‑6P), using cathepsin S (a class‑II‑pathway protease) and proteinase K (broad specificity). Proteolytic fragments were analyzed by SDS‑PAGE, excised, tryptically digested, and identified by mass spectrometry to map cleavage sites within limits imposed by glycosylation and peptide recovery.
APL predicted subsets of CD4+ epitopes for S1 and S2 conformations but had only modest accuracy for SARS‑CoV‑2 spike overall. For S1, APL from individual structures identified some epitopes but did not reach significant accuracy when assessed against reported epitope frequencies. MHCII binding scores (7‑allele method) and immunopeptidomics assignment similarly lacked significant predictive accuracy for S1. Interestingly, APL performed reasonably well for the pre‑fusion S2 of the common cold coronavirus OC43 and improved modestly using a post‑fusion S2 conformation for SARS‑CoV‑2, suggesting that particular spike conformers better model proteolytic processing for certain regions.
Limited proteolysis with cathepsin S (CS) and proteinase K (PK) generated large fragments corresponding to S1 and S2 domains and revealed multiple cleavage sites. The soluble recombinant spike ectodomain (with proline stabilizing substitutions) migrated broadly at ~200 kDa because of glycosylation. In spike‑2P, major cleavage‑sensitive regions included the NTD‑to‑RBD (N2R) linker and subdomain 1 (SD1) within S1; in S2, sensitive sites were found on the N‑terminal flank of the upstream helix and C‑terminal to the beta hairpin. For spike‑6P, similar cleavage was observed in N2R and SD1 and additional cleavage within an unstable RBD region; in S2, fragments indicated cleavage flanking the upstream helix and fusion peptide and extended to central helix and subdomain 3 in some fragments. Mapping resolution was limited by tryptic peptide boundaries and glycan coverage, but overall cleavage sites localized to conformationally unstable segments. Proline substitutions that alter spike conformation modulated usage of some cleavage sites.
Cleavage‑sensitive sites identified by limited proteolysis coincided with segments that APL flagged as conformationally unstable. Peptides originating from these unstable and solvent‑exposed regions were over‑represented among CD4+ epitopes reported after infection relative to vaccination in IEDB mapping data. Examples cited include peptides near the C‑terminus of the RBD (peptides 64–65) that were more dominant after infection, whereas peptides in the NTD and RBD (peptides 50–51) were more dominant after vaccination. The ratio of CD4+ T‑cell responses to peptide pools derived from conformationally stable versus unstable regions discriminated infection versus vaccination status in two non‑hospitalized cohorts, supporting the notion that exposure circumstances influence epitope dominance via processing differences.
The data support a model in which spike conformation and the cellular context of exposure alter antigen fragmentation and peptide loading into MHCII, thereby shaping CD4+ epitope dominance. Infection and mRNA vaccination differ in the populations of antigen‑presenting cells engaged, their activation state, endosomal pH and protease repertoires, and in the conformational states of spike encountered; these factors together can shift which spike segments are cleaved and which peptides become dominant. The limited proteolysis results validate that domain boundaries and conformationally unstable regions are preferentially protease‑sensitive, and that proline substitutions that stabilize spike conformations can modulate cleavage patterns. The modest accuracy of APL for SARS‑CoV‑2 pre‑fusion S2 emphasizes that static structures may incompletely model the dynamic conformational and processing landscape that determines epitope presentation.
The authors note implications for immune imprinting and for the concept of “hybrid immune damping” that may arise from particular pairings of spike exposures during infection and vaccination. They also highlight how CD4+ epitope dominance patterns could serve as indicators of exposure history and potentially relate to the protective quality of the CD4+ response.
Protease‑sensitive sites in SARS‑CoV‑2 spike align with conformationally unstable domains and domain boundaries. These structural vulnerabilities correlate with CD4+ epitope dominance patterns that differ between infection and vaccination, implicating exposure circumstances and spike conformation in shaping class II antigen processing and CD4+ T‑cell responses.
The source article includes sections for data availability, ethics, author contributions, funding, acknowledgments, conflicts of interest, a generative AI statement, and publisher’s note. Supplementary figures and tables detail proteolysis mapping and APL analyses. Specific data deposition details and ethics statements were included in the source article but are not reproduced here beyond noting their presence in the original publication.