Metastatic colorectal cancer (CRC) is a major cause of cancer-related death and incidence is rising in the younger population (<50 years).  Current chemotherapies can achieve response rates above 50%, but immunotherapies have limited value for patients with microsatellite-stable (MSS) cancers.  The present study investigates the impact of chemotherapy on the tumor immune microenvironment.  We treat human liver metastases slices with 5-Fluorouracil (5FU) plus either irinotecan or oxaliplatin, then perform single-cell transcriptome analyses.  Results from eight cases reveal two cellular subtypes with divergent responses to chemotherapy. Susceptible tumors are characterized by a stemness signature, an activated interferon pathway, and suppression of PD-1 ligands in response to 5FU+irinotecan.  Conversely, immune checkpoint TIM-3 ligands are maintained or up-regulated by chemotherapy in CRC with an enterocyte-like signature, and combining chemotherapy with TIM-3 blockade leads to synergistic tumor killing.  Together, our analyses highlight chemo-modulation of the immune microenvironment and provide a framework for combined chemo-immunotherapies. 

Following surgical resection of CRLM specimens greater than 2 cm in diameter, sterile 6 mm tumor tissue cores were punch biopsied and immediately placed in BELZER-UW solution on ice. Within hours, cores were cut into 250 μm thick slices by vibratome and placed with media onto Millicell Cell Culture Inserts in a 24-well cell culture plate. Tumor slices were treated with 1 µg/ml 5-Fluorouracil in combination with 1 µg/ml oxaliplatin in the FOLFOX group or 1 µg/ml 5-Fluorouracil in combination with 2 µg/ml irinotecan in the FOLFIRI group. Control group consisted of slices treated with 0.2% DMSO in medium.

Tumor slices were dissociated using the MACS Tumor Dissociation Kit according to the Miltenyi Biotec “dissociation of soft tumors” protocol. Cells were processed by 10xGenomics Chromium using the single-cell 3' RNA version 2 protocol.  RNAseq libraries were sequenced on the NextSeq500 instrument for 150 cycles (26 bp for Read 1 and 124 bp for Read 2). Reads were aligned to the human genome (GRCh38) and quantified using the Cell Ranger version 2.0 with default settings.  All gene expression data was saved to a single Cell Ranger-like Market Exchange Format (MEX) sparse data matrix. 

Python and R libraries exists for reading MEX data matrices.  In the style of 10x Cell Ranger, we provide here the MEX file (filtered_matrix.mtx) and TSV files (filtered_genes.tsv, filtered_barcodes.tsv) with feature and barcode sequences corresponding to row and column indices, respectively.  One such library is ScanPy (https://scanpy.readthedocs.io) with its "read_mtx" function. We also provide a CSV file (sample_identification.csv) with sample and treatment information.