Clinical and Translational Discovery (Feb 2023)
Bulk and single‐cell spatial transcriptomic analysis reveals cuproptosis‐related molecular subtypes characterized by distinct tumour microenvironment infiltrates in colorectal cancer
Abstract
Abstract Background The association between tumour microenvironment (TME) in colorectal cancer (CRC) and cell death patterns requires further exploration. Cuproptosis may provide new insight into analyzing CRC's TME. Methods We used cuproptosis‐related genes (CRGs) to stratify the meta‐Gene‐Expression Omnibus cohort by the “non‐negative matrix factorization” consistency matrix algorithm. To clarify the relative abundance of different cells in TME, CIBERSORT and single‐sample gene set enrichment methods were performed. Then, CRGs’ transcription in different cell types and the spatial position of CRGs’ enrichment were demonstrated through single‐cell spatial transcriptomics (STs) analysis. Prediction of clinical outcomes and response rate of immunotherapy was also conducted by constructing CRG_score. Results Three cuproptosis‐related TME phenotypes were figured out in CRC with distinctive clinicopathological and prognostic features. Three phenotypes were identical to immune‐inflamed, immune‐desert and immune‐excluded profiles, respectively. Interestingly, cuproptosis‐related cluster 3 (CPRC3) phenotype‐related gene_score was predominantly upregulated in the fibroblast region, consisting of stromal cells. This result was in line with CPRC3's immune‐excluded profile. Based on the robust cuproptosis‐related risk_score and nomogram, CRC patients’ clinical outcome and response to PD1/PD‐L1 immunotherapy can be well predicted. Conclusions This study characterized three cuproptosis‐related TME patterns with unique immune infiltration and prognosis in both bulk and single‐cell ST profiling. Risk_score was also built to help clinicians decide on personalized anticancer treatment and immunotherapy applications for CRC patients.
Keywords
