bioRxiv2026
Primary tumors and their sentinel lymph nodes are functionally linked sites of anti-tumor immunity, yet how T cell receptor (TCR) repertoires are organized across these compartments remains incompletely understood. We thus performed TCR{beta} sequencing on paired primary melanoma tumors and sentinel lymph nodes from 24 treatment-naive patients to quantify TCR repertoire diversity and clonal architecture as markers of antigen-driven selection. Primary tumors exhibited markedly reduced TCR diversity and pronounced clonal dominance compared with matched lymph nodes, consistent with selective expansion of tumor-reactive T cells. Tumor-associated clonotypes displayed significantly longer CDR3 sequences driven by increased non-templated nucleotide insertions, a feature associated with neoantigen recognition, and showed biased TRBV and TRBJ gene usage indicative of CD8+ T cell enrichment. Annotation against known melanoma differentiation and cancer testis antigens identified only a small fraction of clonotypes with characterized specificities, and fewer than 10% of these were shared between tumors and lymph nodes, suggesting that most expanded tumor clonotypes recognize patient-specific antigens. To identify shared immune features beyond individual clonotypes, we applied ClustIRR to build a joint graph linking clonotypes across repertoires by CDR3 sequence similarity and detected Communities on the Joint graph (CJs) to uncover recurrent TCR sequence motifs. Differential occupancy analysis identified distinct CJs segregating tumors and lymph nodes, including a limited number of recurrent CJs targeting MART-1 epitopes, while the majority were patient-specific. Together, these data define spatially structured TCR repertoire architectures in human melanoma and establish a scalable framework for interrogating tumor-immune interactions. eLife DigestMelanoma is a skin cancer harboring many UV-associated mutations, making it recognizable to the immune system. T cells attack cancer using specialized molecular structures called T cell receptors (TCRs). They are frequently referred to as specific keys that only fit into the respective locks, i.e. HLA molecules presenting the target antigen. Surgeons often remove both the tumor and the tumor draining lymph node (aka sentinel lymph node). The tumor is where T cells directly fight cancer, while the sentinel lymph node is where new anti-tumor T cells are first activated. How these two sites coordinate immune surveillance is not completely understood. Here, Kitanovski and colleagues sequenced TCRs from paired tumors and sentinel lymph nodes obtained from melanoma patients. These analyses demonstrated that primary tumors have dramatically reduced TCR diversity, dominated by a few massively expanded T cell clones, whereas sentinel lymph nodes contain broad, balanced repertoires. This reveals intense local selection within the tumor-a small set of tumor-specific T cells expand repeatedly while the lymph node preserves diverse immune reserves. Examining TCR structure revealed a second key difference. Tumor-infiltrating T cells possess on average slightly longer CDR3 loops-the region contacting antigens-especially among dominant clones. These longer loops arise from random nucleotide insertions. Prior work shows that self-antigen-specific T cells have shorter CDR3s due to thymic constraints, whereas T cells recognizing tumor-specific mutation-derived neoantigens escape these restrictions and have longer CDR3s. Thus, tumor TCRs appear tailored for recognizing patient-specific neoantigens. The authors could only use TCRs matching known databases for shared melanoma antigens like MART-1 to identify TCR-specific epitopes; most tumor-expanded TCRs could not be annotated because patient-specific neoantigens are not in public databases. This limitation itself is informative- most expanded tumor TCRs likely target private neoantigens. Grouping TCRs by sequence similarity into "communities" revealed striking patterns: sentinel lymph nodes showed conserved community composition across patients, consistent with responses to common viruses, while tumor repertoires were highly individualized, reflecting each tumors unique mutations. These findings establish how anti-melanoma immunity is spatially organized: sentinel lymph nodes contain diverse T cell landscapes, while in primary tumors melanoma-infiltrating T cells appear to have undergone intense selection for neoantigen recognition. This provides a foundation for using TCR features as biomarkers and designing personalized T cell therapies.