Figure 2.

Immune regulators with the potential to modulate tumor microenvironment revealed by clinical database analysis. (A) Correlations between the expressions of resistance regulators identified in the CRISPR screens and tumor immune infiltration. Patients with cutaneous melanoma whose information was included in The Cancer Genome Atlas (TCGA) were stratified based on lymphocyte infiltration score (L score). A scatterplot was used to demonstrate the differences in mRNA expression levels of these candidate genes in high lymphocyte infiltration (L score >3) and low lymphocyte infiltration (L score ≤3) groups. Candidate genes identified under both screen conditions (effector:target ratio (ET) at 0.3:1 and 1:1) were highlighted with red dots. The dashed line indicates an adjusted p value at 0.05. (B) Boxplots of mRNA expression of PRMT1 and RIPK1 in inflamed (L score >3) and non-inflamed (L score ≤3) groups. The log₂ fold changes (inflamed vs non-inflamed) of TPM of PRMT1/RIPK1 in inflamed and non-inflamed groups were calculated, and their related p values were indicated in the corresponding plot. (C) Comparison of intratumoral cytolytic activity (CYT) in melanomas with different expression levels of PRMT1/RIPK1. The median of mRNA expression (PRMT1 or RIPK1) was used to stratify the clinical samples in the SKCM TCGA dataset. The log₂ fold changes (high vs low) of CYT and their related p values were indicated in the corresponding plot. (D) The profiles of tumor-infiltrating immune cells in melanomas with different expression of PRMT1/RIPK1. CIBERSORT analysis was performed to determine the association between the abundance of immune cells and the levels of PRMT1/RIPK1 expression in SKCM-TCGA database. Immune cell types with significant negative correlation (fraction difference <−0.01 by dashed line on the X axis and p<0.05) were highlighted with blue dots, and immune cell types with significant positive correlation were highlighted with red dots. NK, natural killer.