m6A-related lncRNA-based immune infiltration characteristic analysis and prognostic model for colonic adenocarcinoma.

Background: Colonic adenocarcinoma (COAD) is a common gastrointestinal tract tumor, and its occurrence and progression are typically associated with genomic instability, tumor-suppressor gene and oncogene mutations, and tumor mutational load. N6-methyladenosine (mA) modification of RNAs and long non-coding RNA (lncRNA) expression are important in tumorigenesis and progression. However, the regulatory roles of mA-associated lncRNAs in the tumor microenvironment, stratification of prognosis, and immunotherapy are unclear.

Methods: We screened 43 prognostic lncRNAs linked to mA and performed consistent molecular typing of COAD using consensus clustering. The single-sample Gene Set Enrichment Analysis and ESTIMATE algorithms were used to assess the immune characteristics of different subgroups. Covariation between methylation-related prognostic lncRNAs was eliminated by least absolute shrinkage and selection operator Cox regression. A nomogram was created and evaluated by combining the methylation-related prognostic lncRNA model with other clinical factors. The relationship between the prognostic model grouping and microsatellite instability, immunophenotype score, and tumor mutation burden was validated using R scripts. Finally, we used a linkage map to filter sensitive medicines to suppress the expression of high-risk genes. Three mA-associated lncRNA modes were identified in 446 COAD specimens with different clinical endpoints and biological statuses. Risk scores were constructed based on the mA-associated lncRNA signature genes. Patients with lower risk scores showed superior immunotherapy responses and clinical benefits compared to those with higher risk scores. Lower risk scores were also correlated with higher immunophenotype scores, tumor mutation burden, and mutation rates in significantly mutated genes (e.g., FAT4 and MUC16). Piperidolate, quinostatin, and mecamylamin were screened for their abilities to suppress the expression of high-risk genes in the model.

Conclusions: Quantitative assessment of mA-associated lncRNAs in single tumors can enhance the understanding of tumor microenvironment profiles. The prognostic model constructed using mA-associated lncRNAs may facilitate prognosis and immunotherapy stratification of patients with COAD; finally, three drugs with potential therapeutic value were screened based on the model.