Analysis of T cell receptor clonotypes in tumor microenvironment identifies shared cancer-type-specific signatures

Despite the conventional view that a truly random V(D)J recombination process should generate a highly diverse immune repertoire, emerging reports suggest that there is a certain bias toward the generation of shared/public immune receptor chains. These studies were performed in viral diseases where public T cell receptors (TCR) appear to confer better protective responses. Selective pressures generating common TCR clonotypes are currently not well understood, but it is believed that they confer a growth advantage. As very little is known about public TCR clonotypes in cancer, here we set out to determine the extent of shared TCR clonotypes in the intra-tumor microenvironments of virus- and non-virus-driven head and neck cancers using TCR sequencing. We report that tumor-infiltrating T cell clonotypes were indeed shared across individuals with the same cancer type, where the majority of shared sequences were specific to the cancer type (i.e., viral versus non-viral). These shared clonotypes were not particularly enriched in EBV-associated nasopharynx cancer but, in both cancers, exhibited distinct characteristics, namely shorter CDR3 lengths, restricted V- and J-gene usages, and also demonstrated convergent V(D)J recombination. Many of these shared TCRs were expressed in patients with a shared HLA background. Pattern recognition of CDR3 amino acid sequences revealed strong convergence to specific pattern motifs, and these motifs were uniquely found to each cancer type. This suggests that they may be enriched for specificity to common antigens found in the tumor microenvironment of different cancers. The identification of shared TCRs in infiltrating tumor T cells not only adds to our understanding of the tumor-adaptive immune recognition but could also serve as disease-specific biomarkers and guide the development of future immunotherapies.

Predicting the spectrum of TCR repertoire sharing with a data-driven model of recombination

Despite the extreme diversity of T-cell repertoires, many identical T-cell receptor (TCR) sequences are found in a large number of individual mice and humans. These widely shared sequences, often referred to as "public," have been suggested to be over-represented due to their potential immune functionality or their ease of generation by V(D)J recombination. Here, we show that even for large cohorts, the observed degree of sharing of TCR sequences between individuals is well predicted by a model accounting for the known quantitative statistical biases in the generation process, together with a simple model of thymic selection. Whether a sequence is shared by many individuals is predicted to depend on the number of queried individuals and the sampling depth, as well as on the sequence itself, in agreement with the data. We introduce the degree of publicness conditional on the queried cohort size and the size of the sampled repertoires. Based on these observations, we propose a public/private sequence classifier, "PUBLIC" (Public Universal Binary Likelihood Inference Classifier), based on the generation probability, which performs very well even for small cohort sizes.