Cavalluzzo, B. | bioRxiv Science

bioRxiv2026

Differential serum binding patterns predicting healthy subjects and cancer patients

A viral exposure signature (VES) has been previously described predicting the development of Hepatocellular carcinoma (HCC) in at-risk patients. This has been achieved by a serological profiling of the viral infection history using a synthetic human virome including >100k epitopes (VirScan). In the present study we applied the same VirScan strategy to identify a differential serum binding pattern (DSBP) for classifying patients of different cancer types from healthy individuals. In particular, the healthy group included both age-matched (ADULTS) as well as elderly (ELDERS) individuals, the latter counting also nonagenarians and centenarians. The class comparison performed with serological data show DSBPs supporting class predictions, as confirmed by the receiver operating characteristic (ROC) curve analysis. Antibody responses supporting the class predictions are specific to peptides from persistent herpesviruses, acute-infecting viruses and, consistently in all comparisons, human respiratory syncytial virus (HRSV). Strikingly, the DSB of the ELDERS vs. CANCER comparison is characterized by higher titers in the healthy subjects; on the contrary, the DSB of the ADULTS vs. CANCER comparison is characterized by lower titers in the healthy subjects. Overall, the results show a differential serological binding pattern predicting healthy individuals (ADULTS or ELDERS) from patients with different types of cancer. Such results provide the first evidence suggesting a close link between anti-microbial immunity and cancer development. They may be of the highest relevance in terms of predictive, diagnostic and/or prognostic impact in oncology.

immunology↗

bioRxiv2026

Cross-reactivity of SARS-CoV-2-specific T cells against tumor-associated antigens via molecular mimicry

BACKGROUNDWe have recently described SARS-COV-2 antigens showing sequence and conformational homology to tumor associated antigens (TAAs). Moreover, cross-reactive T cells have been identified in individuals either infected by the SARS-CoV-2 virus or vaccinated with the BNT162b2 preventive vaccine. In the present study, we analyzed the specific cross-binding TCRs by single cell RNA TCR sequencing. METHODS AND RESULTSThe paired SARS-CoV-2 epitope LLLDDFVEI (VIR) and the PRDX5 tumor associated antigen LLLDDLLVS (TAA) were selected to elicit cross-reacting T cells ex vivo. PBMCs from 5 healthy individuals were cultured for 10 days with 10 ug every 3 days of one of the two peptides and cells were selected for single cell RNA TCR sequencing. Results in CD8+ T Effector cells (TTE) showed the amplification or the de novo identification of a handful number of TRAV/TRBV genes and of CDR3{beta} motifs upon treatment ex vivo with both epitopes, which are specific for each subject in the analysis. The very same clonotypes were identified also in the CD8+ T proliferating subset, confirming that both epitopes induced a highly activated and plastic state. Conformational prediction analyses of pMHC-TCR complexes showed perfect structural overlap, supporting the functional cross-reaction of CD8+ T cells with both the viral and the tumor antigens. CONCLUSIONSOur results describe for the first time the TCR CDR3{beta} motifs amplified or de novo expanded by induction with a viral antigen showing a molecular mimicry with a tumor antigen. They are strictly individual and do not match with any motif in the publicly available TCR repository. However, considering the significant degeneracy in the TCR binding to the same epitope, the finding of identical TCR CDR3{beta} motifs elicited by two homologous epitopes is of the highest functional relevance. Such results provide a clear experimental validation proof that microbial epitopes mimicking TAAs can be used to develop off-the-shelf preventive/therapeutic vaccine formulations. Indeed, such non-self antigens are much stronger immunogens and may elicit a potent cross-reacting anti-cancer T cell response.

cancer biology↗