Evolution of T cells in the cancer-resistant naked mole-rat

The study investigates how T-cell subsets have evolved and are organized in the cancer-resistant, long-lived naked mole-rat (NMR). Lymphocytes mediate innate and adaptive immunity, with CD8 and CD4 αβ T cells acting against intracellular and extracellular infections via MHC-I and MHC-II, respectively, and NK cells detecting altered MHC-I and stress ligands. TCR loci undergo somatic recombination to generate diverse clonotypes; a subset differentiates into γδ T cells that recognize stress ligands and typically reside in peripheral tissues. Prior work showed NMRs lack NK cells, possess a unique cervical lymph node (ectopic thymus) contributing to T-cell maturation, display atypical thymic involution trajectories, and show spleen size variations with social rank. Given NMR cancer resistance and the immune system’s role in immunosurveillance, this study aims to characterize NMR T-cell subsets (αβ and γδ), their clonotypic diversity, thymic development, and the evolution of TCR and MHC gene families, testing the hypothesis that relaxed intracellular pathogen pressure shaped distinctive NMR T-cell features favoring cancer resistance and longevity.

Background literature outlines roles of αβ T cells and NK cells in pathogen and tumor surveillance, and γδ T cells as bridging innate and adaptive immunity with tissue-resident functions, including anti-tumor and pro-tumor roles. Prior NMR studies reported absence of NK cells, an ectopic cervical thymus, atypical thymic aging, and social rank-associated spleen size differences. Evolutionary pressures on immune multigene families (TCR and MHC) via birth-and-death processes are well established. These works contextualize the present investigation into NMR T-cell composition, development, and gene family evolution.

- Comparative single-cell RNA sequencing (scRNA-seq): Profiled spleen, bone marrow, and thymus from NMRs and mice. Initial splenic datasets (previously generated) were reanalyzed; new T cell–enriched spleen datasets were generated using 10x Genomics 3' v3 chemistry (NMR adults 2 years, n=3 males; NMR old 26–28 years, n=4 males; mouse adults 2 months, n=4 males; mouse old 2 years, n=4 males). Bone marrow datasets: adult NMRs and mice (n=2 females and n=2 males per species). Thymus scRNA-seq: NMR thoracic and cervical thymuses (n=7 adults; 4F/3M) and mouse thoracic thymus (n=4 adults; 2F/2M). Standard 10x workflows (Cell Ranger), rigorous barcode/gene filtering, variable-gene selection, PCA, SNN graph, Louvain clustering, and marker-based annotation were used. - Single-cell TCR sequencing via hybridization-capture: Developed a capture approach for full-length TCR transcripts (α, β, γ, δ) from 3' barcoded scRNA-seq cDNA using biotinylated lockdown probes against constant regions; PacBio Sequel II HiFi sequencing; processed with SMRT Link, lima, Iso-Seq3 (UMI/barcode tagging, refine, dedup), alignment (minimap2), and IgBLAST for V(D)J and CDR3 calls. Clonotypes defined by C,V,D,J gene IDs and both-chain CDR3 amino acid sequences. Diversity metrics computed (Hill number, Shannon entropy, Gini-Simpson) and compared with bootstrap-based tests (simboot mcpHill). - Comparative genomics (phyletic patterns): Constructed a mammalian phyletic pattern for TCR constant/variable regions and MHC-I/II gene families across 67 genomes by merging Ensembl and RefSeq assemblies/annotations, identifying annotated genes and putative pseudogenes via reciprocal BLAST against conserved domains (CDD), orthogroup filtering (OrthoFinder), and distance-based clustering of conserved domain hits into genes. Phylogenetic least squares modeled γ/δ variable-region counts as a function of α/β counts to derive residual diversity measures. - Molecular evolution analysis: Assessed relaxation of purifying selection on Cd8b in hystricomorphs (NMR, Damaraland mole-rat) relative to murines using codon models (HyPhy RELAX) on one-to-one orthologs (MAFFT alignments), comparing Cd8b to Cd8a, Cd4, and CD3 subunit genes. - Thymus age-trajectory and histology: Measured thymic and body weights across physiologically equivalent ages in mice (n=55; 1–24 months) and NMRs (n=28; 1–120 months; additional 312 months), with histopathological assessment (H&E) at set ages to evaluate involution (corticomedullary distinction, cortical reduction, adipocyte infiltration, cysts). Polynomial regression modeled weight trajectories; comparative histology by a veterinary pathologist. - Statistics: Mixed-effects multinomial logit models for cell-type proportion changes with age; multiple testing controlled by FDR.

- Large circulating γδ T-cell population in NMR spleen: Reanalysis and new scRNA-seq revealed two γδ subsets—cytotoxic (high TRG/TRD constant regions, Gzma, Nkg7, Xcl1, Il2rb; inhibitory Klra1/Klrd1; Cd8a without Cd8b) and non-cytotoxic (TRG/TRD high, lacking cytotoxic markers and Cd8a/b). In mice, splenic γδT cells are a minor subset, with NK cells instead expressing cytotoxic markers. - Tissue distribution and development: γδT cells present in NMR bone marrow (0.35% ±0.01 of total marrow cells) versus mouse marrow where NK cells are 0.65% ±0.01; pre-T and immature T cell proportions were similar across species (NMR: pre-T 2.7% ±0.2, immature 0.5% ±0.05; mouse: pre-T 2.2% ±0.05, immature 0.3% ±0.005). In thymus, γδT cells comprise <1% at the terminus of maturation in both species. Mouse thymic γδT cells expressed Itgae (CD103) highly, whereas NMR thymic γδT cells did not, consistent with a circulating γδ pool in NMRs. - Cd8b evolution and expression: Outside thymic double-positive cells, NMR Cd8b expression was negligible despite Cd8a expression, contrasting mice where Cd8b matched Cd8a broadly. HyPhy RELAX showed NMR and Damaraland mole-rat Cd8b evolved under significantly relaxed purifying selection relative to murines, unlike Cd8a and CD3 genes, suggesting functional divergence/loss for Cd8b. - TCR locus diversity (phylogenomics): NMR α and β variable region repertoires are reduced (~1.75-fold and ~1.5-fold smaller than rodent and mammal means, respectively), whereas γ and δ variable regions are comparatively expanded (residuals from phylogenetic least squares indicate ~18-fold more γ/δ diversity than expected from α/β). This hystricomorph pattern holds except in guinea pig. Murine genomes generally showed negative γ/δ residuals (except blind mole-rat). - Dominant γδ clonotype in NMR cytotoxic subset: Single-cell TCR-seq identified a highly public, dominant cytotoxic γδ clonotype present in all NMR samples (15.6%–65.1% of cytotoxic γδ cells; mean 34% ±6.7), with closely related clonotypes cumulatively reaching 26.8%–81.4% (mean 34% ±7.8). Exemplary chains: Vγ4-2/Jγ5-3 with CDR3 TYWDSNYAKK; Vδ1-4/Dδ3/Jδ2 with CDR3 ALWELRTGGITAQLV. Non-cytotoxic γδ clonotypes were more diverse and less public (publicity ~0.4%) than cytotoxic ones (publicity 5.32%). Mouse γδ clonotypes also included a dominant public clonotype (e.g., Vγ6/Jγ1 CDR3 ACWDSSGFHKV; Vδ4/Dδ2/Jδ2 CDR3 GSDIGGSSWDTRQMF), but sample sizes differed (~47 mouse γδ vs ~4500 NMR γδ clonotyped cells). - αβ clonotypic diversity and CD8/CD4 balance: In NMR spleen, CD8 αβ T cells had consistently lower clonotypic diversity (richness, evenness, rareness) than CD4 αβ T cells in old samples. NMR total splenic T-cell proportion approximated mouse (NMR 26%, mouse 27.2%), but NMR showed a strong CD4 bias: adult NMR CD8 7% vs CD4 17%; old NMR CD8 2.3% vs CD4 11%. Mice showed more balanced naive/memory CD8 and CD4 proportions across ages. Cross-species, NMR CD8 αβ clonotypes were less diverse than mouse naive CD8 clonotypes (especially in adults), whereas old NMR CD4 clonotypes tended to be more diverse than old mouse CD4. - MHC gene-family sizes align with selective-pressure hypothesis: NMR MHC-I gene family is markedly reduced (~11-fold vs rodent mean; ~10-fold vs mammal mean), while MHC-II is modestly reduced (~1.6-fold vs rodent; ~2-fold vs mammal). Consequently, the MHC-I/MHC-II ratio is ~7.8-fold (rodents) and ~5.3-fold (mammals) lower in NMR, consistent across hystricomorphs (except guinea pig). - Thymus development and involution: NMR thymus peaks at ~2 months with roughly half the maximal mouse thymus weight (absolute and relative to body weight), declines between 2–12 months but less steeply than mouse (which declines notably between 1–2 months). Histology showed similar microscopic involution signs in both species at physiologically equivalent ages (reduced corticomedullary distinction, cortical reduction, adipocyte infiltration, cysts), present in both NMR thoracic and cervical thymi. - Age-related changes: In NMR spleen, ILCs appeared predominantly in old samples; cytotoxic γδT cells were enriched with age, CD8 αβ underrepresented with age. NMR cytotoxic γδ clonotypes increased in evenness/rareness with age; mouse naive CD8/CD4 clonotypes declined in diversity with age, while Tregs increased.

The findings support a model in which NMRs evolved under relaxed intracellular pathogen pressure, reflected in loss of NK cells, reduced MHC-I and α/β TCR variable-region diversity, a reduced and CD4-biased αβ T-cell compartment, and relaxed selection on Cd8b. Concurrently, γδT cells appear repurposed as a circulating effector arm with an expanded genomic γ/δ variable-region repertoire and a cytotoxic subset harboring a dominant public clonotype, potentially recognizing common stress ligands pertinent to cancer surveillance or tissue homeostasis. The lack of Itgae expression in NMR thymic γδT cells aligns with their circulating rather than epithelial-homing fate, explaining their high splenic abundance. Thymic size and involution trajectories are consistent with a smaller early-life thymic output in NMRs due to lower αβ repertoire demands, yet both species show involution at equivalent physiological ages, supporting energy-allocation theories of thymopoiesis shutdown after repertoire establishment. Overall, the immune architecture in NMRs appears shifted from intracellular pathogen defense toward tumor surveillance and extracellular pathogen responses, potentially contributing to their longevity and cancer resistance.

This work provides an integrated cellular, clonotypic, genomic, and developmental portrait of NMR T cells. Key contributions include discovery of abundant circulating γδT cells with a dominant public cytotoxic clonotype, reduced α/β and expanded γ/δ TCR genomic diversity, evidence of relaxed selection on Cd8b, a reduced MHC-I relative to MHC-II gene-family size, and a small early-life thymus with involution comparable to mice. The hybridization-capture single-cell TCR approach introduced here generalizes clonotype profiling to non-model species and all TCR loci using existing 3' scRNA-seq libraries. Future directions include functional characterization of NMR γδT-cell cytotoxicity, ligand identification for dominant clonotypes, and spatial single-cell/TCR profiling in NMR tumor microenvironments to link γδ repertoires with anti-tumor activity.

- Functional roles of NMR γδT cells were inferred from transcriptomic profiles; direct effector assays against pathogens or tumors were not performed. - Evidence for relaxed intracellular pathogen pressure is indirect; experimental infection data in NMRs are sparse and difficult to obtain. - Phyletic pattern inferences depend on genome assembly/annotation integrity; gene-family size estimates may be approximate despite benchmarking. - Single-cell TCR capture yielded paired chains for ~1/3 of profiled T cells, potentially affected by probe efficiency and sequencing depth; this may underrepresent true clonotype diversity. - Cross-species comparisons involve differing sample sizes for some subsets (e.g., mouse γδ clonotyped cells were few), which can affect diversity and publicity estimates.