Modified eQTL and Somatic DNA Segment Alterations in Esophageal Squamous Cell Carcinoma for Genes Related to Immunity, DNA Repair, and Inflammation

Abstract

We integrated ESCC expression and GWAS genotyping, to investigate eQTL and somatic DNA segment alterations, including somatic copy number alteration, allelic imbalance (AI), and loss of heterozygosity (LOH) in ESCC. First, in eQTL analysis, we used a classical approach based on genotype data from GWAS and expression signals in normal tissue samples, and then used a modified approach based on fold change in the tumor vs. normal samples. We focused on the genes in three pathways: inflammation, DNA repair, and immunity. Among the significant (p < 0.05) SNP-probe pairs from classical and modified eQTL analyses, 24 genes were shared by the two approaches, including 18 genes that showed the same numbers of SNPs and probes and 6 genes that had the different numbers of SNPs and probes. For these 18 genes, we found 28 SNP−probe pairs were correlated in opposite directions in the two approaches, indicating an intriguing difference between the classical and modified eQTL approaches. Second, we analyzed the somatic DNA segment alterations. Across the 24 genes, abnormal gene expression on mRNA arrays was seen in 19−95% of cases and 26−78% showed somatic DNA segment alterations on Affymetrix GeneChip Human Mapping Arrays. The results suggested that this strategy could identify gene expression and somatic DNA segment alterations for biological markers (genes) by combining classical and modified eQTLs and somatic DNA evaluation on SNP arrays. Thus, this study approach may allow us to understand functionality indicative of potentially relevant biomarkers in ESCC.

Keywords: DNA segment alterations; ESCC; SNP; eQTL.

Figure 1

Flow chart for integrated analysis for classical and modified eQTL and somatic DNA segment alterations in immunity, DNA repair and inflammation in ESCC.

Figure 2

(a) Boxplot of the signal in normal and the Spearman correlation between signal and genotype with p = 0.003 and a negative rho = −0.29 for the gene–SNP pair DAPK1 (211214_s_at) and rs1964911. (b) Distribution of the tumor vs. normal fold change and the correlation with p = 0.03 and a positive rho = 0.22 for the same gene–SNP pair.

Figure 3

(a) Distribution of signal in normal and the correlation with p = 0.02 and a positive rho = 0.26 for the gene–SNP pair ST6GAL1 (214970_s_at) rs4686849. (b) Distribution of the tumor vs. normal fold change and the correlation with p = 0.03 and a negative rho = −0.25 for the same gene–SNP pair.

Figure 4

Somatic alteration on ST6GAL1 and RARB in 76 ESCC were analyzed. The copy number change images for two genes, (a) ST6GAL1 and (b) RARB were shown for 20 ESCC samples as assay examples. On each image, the top portion showed the distribution of somatic alteration and the bottom portion showed individual samples in multiple rows. The colors blue, red, and purple indicated copy number gain, copy number loss, and allelic imbalance, respectively.