Functional single nucleotide polymorphisms of the RASSF3 gene and susceptibility to squamous cell carcinoma of the head and neck

Abstract

Background: RASSF3 suppresses tumour formation through uncertain mechanisms, but it is an important gene of p53-dependent apoptosis. RASSF3 depletion impairs DNA repair after DNA damage, leading to polyploidy. The authors hypothesised that potential functional single-nucleotide polymorphisms (SNPs) of RASSF3 are associated with risk of squamous cell carcinoma of the head and neck (SCCHN).

Methods: The authors used a functional SNP approach to evaluate the associations between common (minor allele frequency⩾0.05), putative functional variants in RASSF3 and risk of SCCHN. Four selected such functional SNPs (rs6581580 T>G, rs7313765 G>A, rs12311754 G>C and rs1147098 T>C) in RASSF3 were identified and genotyped in 1087 patients and 1090 cancer-free controls in a non-Hispanic white population.

Results: The authors found that two SNPs were significantly associated with SCCHN risk. Carriers of the variant rs6581580G and rs7313765A alleles were at a reduced SCCHN risk, compared with the corresponding common homozygotes [adjusted odds ratio (OR)=0.75 and 0.73 and 95% confidence interval (CI)=0.62-0.91 and 0.60-0.88, respectively, for dominant models; and Ptrend=0.012 and 0.041, respectively, for additive models], particularly for non-oropharyngeal tumours (adjusted OR=0.68 and 0.60 and 95% CI=0.53-0.86 and 0.47-0.77, respectively, for dominant models). In the genotype-phenotype correlation analysis of peripheral blood mononuclear cells from 102 cancer-free controls, the rs6581580 GG genotype was associated with significantly increased expression levels of RASSF3 mRNA (P=0.038), compared with the TT genotype. Additional functional experiments further showed that variant G allele of rs6581580 had a significantly stronger binding affinity to the nuclear protein extracts than the T allele.

Conclusion: Taken together, these findings indicate that the RASSF3 promoter rs6581580 T>G SNP is potentially functional, modulating susceptibility to SCCHN among non-Hispanic whites. Larger replication studies are needed to confirm our findings.

Keywords: Apoptosis; Biomarker; Genetic susceptibility; Head and neck cancer; Polymorphism; RASSF3.

Fig. 1

(A) Positions of the four potentially functional single-nucleotide polymorphisms (SNPs) in RASSF3 and pairwise linkage disequilibrium (LD) between them. These four SNPs were predicted to affect putative transcription factor-binding activity. (B) Genotype–phenotype correlation for rs6581580 and relative expression levels of RASSF3 mRNA in peripheral blood mononuclear cells (PBMCs) of 102 cancer-free controls. The relative RASSF3 mRNA expression levels were higher for rs6581580 GG genotype (16.62 ± 0.74) than that for the TT genotype (16.11 ± 0.82), and the difference was statistically significant (P < 0.038), the TG genotype (16.38 ± 0.79) and the GG/TG genotypes (16.44 ± 0.78) were not significantly different from that for the TT genotype (P = 0.140, and P = 0.059, respectively). (C) Genotype–phenotype correlation for rs7313765 and relative expression levels of RASSF3 mRNA in PBMCs of 102 cancer-free controls. The relative RASSF3 mRNA expression levels were similar among the three groups with rs7313765 GG, GA and AA genotypes (P = 0.696 for GG versus AA, P = 0.335 for GG versus GA and P = 0.378 for GG versus AA/GA). (D) Electrophoretic mobility shift assay (EMSA) for comparison of DNA–protein binding activities between labelled probes containing different alleles of rs6581580. As it shown in lane 4 and lane 1, probe containing G allele had a stronger binding affinity to the nuclear protein extracts than probe containing T allele. Competition assays further showed that the formation of the DNA–protein complexes was completely eliminated by the 50-fold excess of unlabelled probes containing either T or G alleles (lanes 2, 3 and lanes 5, 6), which indicated that the binding was sequence-specific.