Description
Scaling up the immune resolution of human tumor-draining lymph nodes by single cell analysis
Tosello J1, Ramos R1, Núñez NG1, Richer W1, Dutertre CA2, Bigot J1, Lantz O1, Ginhoux F2, Amigorena S1, Sedlik C1, Helft J1, Piaggio E1
1Paris-Sciences-Lettres, Institut Curie Research Center, INSERM U932 & SiRIC, Translational Immunotherapy Team; 2Singapore Immunology Network, Agency for Science, Technology and Research, Singapore 138648, Singapore
The anti-tumoral immune response is shaped at the tumor bed, but also in tumor-draining lymph nodes (TDLNs) where T cell priming is taking place. Although a strategic target for anti-tumor immunotherapies, little is known about the landscape of TDLNs. In breast tumors, we have observed that: i) TDLNs invasion is associated with increased percentages of CD4+ T regulatory (Tregs) and myeloid cells; and that ii) CD8+ T cells from I-TDLNs are highly functional compared to exhausted CD8+ T cells present at the tumor bed, suggesting that TDLN could benefit from immunotherapies. Therefore, to better understand the phenotypic and functional heterogeneity of immune cells in breast tumors and in their corresponding invaded (I) and non-invaded (NI) TDLNs, and to uncover new therapeutic targets, we have performed high dimensional phenotypic characterization (CyTOF), as well as TCR and single cell RNA sequencing (scRNAseq) of myeloid cells and Tregs present in tumor and I-TDLN of untreated breast cancer patients. We observed that the inflammatory myeloid population enriched in I-TDLNs (CD14+HLA-DR+) expressed CCR2 but did not show suppressive functions. On the other hand, Tregs from I-TDLN expressed CXCR3 and T-bet, yet retained their suppressive function and expressed high levels of targetable immune-checkpoint molecules. Furthermore, the Treg repertoire from TDLN (I and NI) did not overlap with the tumor associated Treg repertoire, and the invasion of TDLN was associated with a more oligoclonal TCR pattern. Finally, scRNAseq analysis revealed an unsuspected heterogeneity of both myeloid and Treg compartments. I-TDLN showed at least 5 sub-populations among CD14+HLA-DR+ myeloid cells and 3 Treg and 3 Tconv subpopulations with differential levels of activation. Our results provide a detailed immune cell atlas of key populations present in I-TDLN and their relationship with those present at the tumor site. Therefore, our work provides a powerful tool for the rational design of immune therapies and to fuel precision medicine targets.