Dynamic heterogeneity of colorectal cancer during progression revealed clinical risk-associated cell types and regulations in single-cell resolution and spatial context

Haoxian Ke; Zhihao Li; Peisi Li; Shubiao Ye; Junfeng Huang; Tuo Hu; Chi Zhang; Ming Yuan; Yuan Chen; Xianrui Wu; Ping Lan

doi:10.1093/gastro/goad034

Dynamic heterogeneity of colorectal cancer during progression revealed clinical risk-associated cell types and regulations in single-cell resolution and spatial context

Gastroenterol Rep (Oxf). 2023 Jun 24:11:goad034. doi: 10.1093/gastro/goad034. eCollection 2023.

Authors

Haoxian Ke^{1

2}, Zhihao Li^{1

2}, Peisi Li^{2

3}, Shubiao Ye², Junfeng Huang^{1

2}, Tuo Hu^{1

2}, Chi Zhang^{1

2}, Ming Yuan^{1

2}, Yuan Chen³, Xianrui Wu^{1

2}, Ping Lan^{1

2}

Affiliations

¹ Department of General Surgery (Colorectal Surgery), The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong, P. R. China.
² Guangdong Provincial Key Laboratory of Colorectal and Pelvic Floor Diseases, Guangdong Institute of Gastroenterology ,The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong, P. R. China.
³ School of Medicine, Sun Yat-sen University, Shenzhen, Guangdong, P. R. China.

Abstract

Background: Tumor heterogeneity is contributed by tumor cells and the microenvironment. Dynamics of tumor heterogeneity during colorectal cancer (CRC) progression have not been elucidated.

Methods: Eight single-cell RNA sequencing (scRNA-seq) data sets of CRC were included. Milo was utilized to reveal the differential abundance of cell clusters during progression. The differentiation trajectory was imputed by using the Palantir algorithm and metabolic states were assessed by using scMetabolism. Three spatial transcription sequencing (ST-seq) data sets of CRC were used to validate cell-type abundances and colocalization. Cancer-associated regulatory hubs were defined as communication networks affecting tumor biological behaviors. Finally, quantitative reverse transcription polymerase chain reaction and immunohistochemistry staining were performed for validation.

Results: TM4SF1⁺, SOX4⁺, and MKI67⁺ tumor cells; CXCL12⁺ cancer-associated fibroblasts; CD4⁺ resident memory T cells; Treg; IgA⁺ plasma cells; and several myeloid subsets were enriched in stage IV CRC, most of which were associated with overall survival of patients. Trajectory analysis indicated that tumor cells from patients with advanced-stage CRC were less differentiated, when metabolic heterogeneity showed a highest metabolic signature in terminal states of stromal cells, T cells, and myeloid cells. Moreover, ST-seq validated cell-type abundance in a spatial context and also revealed the correlation of immune infiltration between tertiary lymphoid structures and tumors followed by validation in our cohort. Importantly, analysis of cancer-associated regulatory hubs revealed a cascade of activated pathways including leukocyte apoptotic process, MAPK pathway, myeloid leukocyte differentiation, and angiogenesis during CRC progression.

Conclusions: Tumor heterogeneity was dynamic during progression, with the enrichment of immunosuppressive Treg, myeloid cells, and fibrotic cells. The differential state of tumor cells was associated with cancer staging. Assessment of cancer-associated regulatory hubs suggested impaired antitumor immunity and increased metastatic ability during CRC progression.

Keywords: colorectal cancer; single-cell RNA sequencing; spatial transcription sequencing; tumor heterogeneity; tumor progression.