Integrated multi-omics characterization of KRAS mutant colorectal cancer

Abstract

KRAS mutation is the most frequent oncogenic aberration in colorectal cancer (CRC). The molecular mechanism and clinical implications of KRAS mutation in CRC remain unclear and show high heterogeneity within these tumors. Methods: We harnessed the multi-omics data (genomic, transcriptomic, proteomic, and phosphoproteomic etc.) of KRAS-mutant CRC tumors and performed unsupervised clustering to identify proteomics-based subgroups and molecular characterization. Results: In-depth analysis of the tumor microenvironment by single-cell transcriptomic revealed the cellular landscape of KRAS-mutant CRC tumors and identified the specific cell subsets with KRAS mutation. Integrated multi-omics analyses separated the KRAS-mutant tumors into two distinct molecular subtypes, termed KRAS-M1 (KM1) and KRAS-M2 (KM2). The two subtypes had a similar distribution of mutated residues in KRAS (G12D/V/C etc.) but were characterized by distinct features in terms of prognosis, genetic alterations, microenvironment dysregulation, biological phenotype, and potential therapeutic approaches. Proteogenomic analyses revealed that the EMT, TGF-β and angiogenesis pathways were enriched in the KM2 subtype and that the KM2 subtype was associated with the mesenchymal phenotype-related CMS4 subtype, which indicated stromal invasion and worse prognosis. The KM1 subtype was characterized predominantly by activation of the cell cycle, E2F and RNA transcription and was associated with the chromosomal instability (CIN)-related ProS-E proteomic subtype, which suggested cyclin-dependent features and better survival outcomes. Moreover, drug sensitivity analyses based on three compound databases revealed subgroup-specific agents for KM1 and KM2 tumors. Conclusions: This study clarifies the molecular heterogeneity of KRAS-mutant CRC and reveals new biological subtypes and therapeutic possibilities for these tumors.

Phenotypic Heterogeneity Analysis of APC-Mutant Colon Cancer by Proteomics and Phosphoproteomics Identifies RAI14 as a Key Prognostic Determinant in East Asians and Westerners

“…These mutations may activate other signaling pathways and increase tumor malignancy ( 13 ). Molecular subtyping based on proteomic and phosphoproteomic data ignores the influence of genetic background, simplifies tumor heterogeneity characterization, and provides potential biomarkers and therapeutic targets for each subtype with APC mutations ( 16 , 17 , 18 , 19 ). Molecular subtyping of APC -MUT colon cancer identifies a combination of molecular markers predicting prognosis in APC -MUT tumors, in which high expression of RAI14 in tumors is associated with poor prognosis in both WCH and CPTAC cohorts.…”

“…The combination of omics technologies and molecular classification methods has been widely used to help elucidate the heterogeneity of cancer ( 16 , 17 , 18 , 19 ). Inspired by these achievements, in this study, we carried out genomics of 76 pairs of colon cancer tissues and proteomics and phosphoproteomics analysis of 69 pairs of colon cancer tissues, of which 35 patients identified mutated APC s. Molecular classification based on the proteome and phosphoproteome data demonstrated the heterogeneity in APC -MUT tumors and identified RAI14 as a key prognostic marker.…”

Dementia with Lewy Bodies: Genomics, Transcriptomics, and Its Future with Data Science

“…Further analyses have identified novel associations within pathways of amyloid pathology, neuronal injury, and tau hyperphosphorylation from the CSF of AD cases [147]. Similar analyses within oncology have identified molecular features associated with tumour invasiveness [148], immunotherapy susceptibility [149], and the subtypes of oncogene mutations [150]. Emerging methods are also supplementing unsupervised models with labelled datasets through semi-supervised learning to improve the identification of novel signatures.…”

Impact of KRAS mutation on the tumor microenvironment in colorectal cancer

“…Herein, we summarize several features identified by multiomics studies. Based on integrated multiomics data, Chong et al reported that the KRAS-M1 (KM1) subtype was related to the CMS4 subtype, which, as opposed to the KRAS-M2 (KM2) subtype, is associated with a poor prognosis and increased angiogenesis, EMT, and TGF-β pathways 23 . Liu et al integrated drug sensitivity information and the phosphoproteome and reported that the combination of SHP2 and DOT1L inhibitors is an effective treatment for patients with subset 2 KRAS-mutant cancers 24 .…”