Refining colorectal cancer classification and clinical stratification through a single-cell atlas

Abstract

Background: Colorectal cancer (CRC) consensus molecular subtypes (CMS) have different immunological, stromal cell, and clinicopathological characteristics. Single-cell characterization of CMS subtype tumor microenvironments is required to elucidate mechanisms of tumor and stroma cell contributions to pathogenesis which may advance subtype-specific therapeutic development. We interrogate racially diverse human CRC samples and analyze multiple independent external cohorts for a total of 487,829 single cells enabling high-resolution depiction of the cellular diversity and heterogeneity within the tumor and microenvironmental cells.

Results: Tumor cells recapitulate individual CMS subgroups yet exhibit significant intratumoral CMS heterogeneity. Both CMS1 microsatellite instability (MSI-H) CRCs and microsatellite stable (MSS) CRC demonstrate similar pathway activations at the tumor epithelial level. However, CD8+ cytotoxic T cell phenotype infiltration in MSI-H CRCs may explain why these tumors respond to immune checkpoint inhibitors. Cellular transcriptomic profiles in CRC exist in a tumor immune stromal continuum in contrast to discrete subtypes proposed by studies utilizing bulk transcriptomics. We note a dichotomy in tumor microenvironments across CMS subgroups exists by which patients with high cancer-associated fibroblasts (CAFs) and C1Q+TAM content exhibit poor outcomes, providing a higher level of personalization and precision than would distinct subtypes. Additionally, we discover CAF subtypes known to be associated with immunotherapy resistance.

Conclusions: Distinct CAFs and C1Q+ TAMs are sufficient to explain CMS predictive ability and a simpler signature based on these cellular phenotypes could stratify CRC patient prognosis with greater precision. Therapeutically targeting specific CAF subtypes and C1Q + TAMs may promote immunotherapy responses in CRC patients.

Keywords: CMS classification; Cancer-associated fibroblast; Colorectal cancer; Immunotherapy; Single-cell analysis; Stromal signatures.

Fig. 1

Identification and clustering of single cells. A Workflow of sample collection, sorting, and sequencing (methods contain full description for each step) and t-SNE characterization of the 49,859 cells profiled. B Identification of various cell types based on expression of specified marker genes. C Characterization of the proportion of cell types identified in each sample in tumor vs. normal colon tissue and Consensus Molecular Subtypes (CMS) of bulk RNA-seq data. D Characterization of the proportion of cell types identified in tumor vs. normal colon tissue, sidedness (right vs. left), microsatellite instability (MSI) status, single-cell Consensus Molecular Subtypes (scCMS) classification, Consensus Molecular Subtypes (CMS) of bulk RNA-seq data, and origin of sample. The graph represents total clusters and cell types identified after re-clustering of each cell compartment depicting global heterogeneous landscape of colorectal cancers

Fig. 2

CMS Cell proportions, Gene set enrichment and trajectories of tumor cells. A Proportions of tumor cells classified as various CMS subtypes in various samples. Primary CRC datasets is labeled CAC and Lee et al. 2020 [18] dataset labeled as SMC. Annotation of tumor cells is based on bulk CMS classification. B Gene set variation expression analyses of combined primary CRC data and Lee et al. 2020 [18], in the tumor epithelial cell compartment. C Trajectory analysis of primary CRC dataset colored by bulk CMS status. D Trajectory analysis of Lee et al. 2020 [18] (Korean cohort) CRC dataset colored by bulk CMS status. E Trajectory analysis of Lee et al. 2020 [18] (Belgian cohort) colored by bulk CMS status

Fig. 3

Fibroblast clusters in colon and colorectal tumors. A t-SNE of fibroblasts colored by Normal, CAF-S1 and CAF-S4 subtypes. B Dot plots showing the variable expression of fibroblast specific marker genes across CAF-S1 and CAF-S4. C Integration analysis of five CAFs subtypes from a breast cancer (BC) cohort to validate the existence of specific CAF subtypes in the CRC samples. D IHC representative images of CAF-S1 (FAP+, PDGFR-ß+) and CAF-S4 (RGS5+, MCAM+) in CRC sections from five independent patients. Asterisk (*) indicates tumor cells and arrows (>) indicate CAFs. Annotated by a board-certified GI pathologist. All images are × 20 magnification. High-resolution images are available on GitHub as source data [43]. E Boxplots show the distribution of cell types in two CRC bulk expression datasets, within tumors based on CMS status. The whiskers depict the 1.5 x IQR. The P-values for pairwise t-tests comparisons (with Benjamini-Hochberg correction) of cell abundance across CMS are shown in the figure. Note “NS”: P > 0.05, *P </=0.05, **P </=0.01, ***P </=0.001, ****P </=0.0001

Fig. 4

Myeloid cell clusters in colon and colorectal tumors. A t-SNE of myeloid cells colored. by distinct cell types. B Heatmap showing the variable expression of myeloid specific marker genes across various myeloid cell types. C Identification of various myeloid cell subtypes based on expression of specified marker genes. D Boxplots show the distribution of cell types in two CRC bulk expression datasets, within tumors with varying CMS status [45, 48]. The whiskers depict the 1.5 x IQR. The P-values for pairwise t-tests comparisons (with Benjamini-Hochberg correction) of cell abundance across CMS are shown in the figure. E Box plot show TAMs infiltration and influence on CD8+ T cells in relation to CMS1 and CMS4. Note in GSE17536 Ythdf2 deficiency trended in CMS4 but did not reach the statistical significance. Note “NS”: P > 0.05, *P </=0.05, **P </=0.01, ***P </=0.001, ****P </=0.0001

Fig. 5

T cell clusters in colon and colorectal tumors. A t-SNE of 22525 T cells colored by distinct clusters. B t-SNE plot showing the variable expression of T cell specific marker genes across various clusters. C Boxplots show the distribution of cell types in two CRC balk expression datasets, within tumors with varying CMS status. The whiskers depict the 1.5 x IQR. The P-values for pairwise t-tests comparisons (with Benjamini-Hochberg correction) of cell abundance across CMS are shown in the figure. Note “NS”: P > 0.05, *P </=0.05, **P </=0.01, ***P </=0.001, ****P </=0.0001

Fig. 6

Continuous subtype scoring across cell type (GSE39582 [45], GSE1736 [48]). A, B Continuous scores reported by CMS classification across cell types show minimal separation in the top 2 principal components in GSE39582 and GSE17536 datasets respectively. All cell types are represented in CMS1-4 using PCSS1 and PCSS2 scores. Note that the cell types largely form a continuum along CMS status and are not clustered in discrete subtypes separate from one another. Cells and markers are colored by bulk CMS status accordingly to the tumor sample of origin. (PCSS1 = PC Cluster Subtype Scores1, PCSS2 = PC Cluster Subtype Scores1)

Fig. 7

Survival analysis of two independent bulk dataset. The relationship between relative cell abundance and disease-free survival (DFS) in the GSE39582 [45] (A) and GSE17536 [48] (B) (COX regression analysis). Kaplan Meier curves depicting DFS in GSE39582 [45] (A) and GSE17536 [48] (B). Note, in addition to CMS4, CMS1-3 subgroups (good prognosis subtypes) with high CAF and C1Q+ TAMs signatures were associated with poor DFS. CAF's and C1Q+ TAMs stratified all CMS subgroups into high and low-poor survival subgroups beyond CMS categorization. HR and P values are indicated

Fig. 8

CAF, TAM and CD8+ T-cell interactions in the CRC microenviroment. A Dotplot showing cell-cell interaction between CD8+ T cells and TAMs. B Circle plot directed ligand receptor interactions in CD8+ T cells and TAMs for better visualization. C Dotplot showing cell-cell interaction between CD8+ T cells and CAFs. D Circle plot directed ligand receptor interactions in CD8+ T cells and CAFs for better visualization. E Dotplot demonstrating cell-cell interaction between TAMs and CAFs. F Circle plot directed ligand receptor interactions in TAMs and CAFs for better visualization