Mature and migratory dendritic cells promote immune infiltration and response to anti-PD-1 checkpoint blockade in metastatic melanoma
Jiekun Yang†‡, Cassia Wang†, Doris Fu†, Li-Lun Ho, Kyriakitsa Galani, Lee Chen, Dennie T Frederick, Liang He, Amy Y. Huang, Emily J. Robitschek, Sandeep K. Yadav, Wentao Deng, Rahul Tacke, Kelly P. Burke, Tatyana Sharova, Kunal Rai, David Liu, Genevieve M. Boland‡, Manolis Kellis‡
† These authors contributed equally to this work
‡ Corresponding author
Immune checkpoint inhibitors (ICIs) have heralded a remarkable shift in cancer care, significantly extending survival for advanced cancer patients. However, despite their remarkable clinical successes, a substantial majority of patients fail to achieve a lasting response to ICIs. To address this challenge and gain insights into the complex landscape of the tumor microenvironment (TME), we conducted an extensive analysis using single-cell RNA sequencing (scRNA; ~216K cells across 39 samples) and single-nucleus transposase-accessible chromatin sequencing (snATAC; ~43K cells from 15 samples) in a metastatic melanoma cohort. This systematic approach delineates 14 distinct cell types and 55 cell subtypes, including the identification of 15 transcriptional hallmarks of malignant cells. Through correlation analysis of cell subtype proportions, we unveiled six distinct clusters associated with varying tumor responses. Particularly intriguing was the identification of the mature DC enriched in immunoregulatory molecules (mregDC) subtype exhibiting correlations with naive T and B cells, forming an anti-tumor program that underscores the importance of multiple immune cell types in mediating anti-tumor immunity. Notably, we found that mregDC abundance represents a good prognostic predictor of progression-free survival (PFS) in the context of ICI treatment, and when combined with the TCF7+/- CD8 T cell ratio, it reliably predicts patient PFS across treatments beyond ICI. We validated our findings using an independent cohort of 274 ICI-treated melanoma samples analyzed using tissue-level expression. We next compared mregDCs and conventional dendritic cell types 1 and 2 (cDC1 and cDC2) using transcriptome signature, differentiation trajectory, interactome, cytokine milieu, and epigenome landscape analyses. This comparative analysis shed light on the unique attributes of mregDCs within the TME. Finally, we investigated cell type/subtype-specific genes, pathways, immune response enrichment, and ligand-receptor interactions closely associated with the proportions of mregDCs within the TME. These molecular and cellular insights, with their critical roles in enhancing the immune response against cancer, offer valuable prospects for predicting the efficacy of ICI regimens, and potentially guiding the selection of rational combinatorial therapies.
This repository contains the scripts to replicate the findings displayed in the manuscript. It is organized into two folders Figures and Methods. The Figures folder has the scripts to replicate the figures in the manuscript and the files are named accordingly. The methods folder contains the different methods used and developed for the manuscript.
The Figures files require specific outputs generated with the Methods scripts.