rs66651343 and rs12909095 confer lung cancer risk by regulating CCNDBP1 expression

Abstract

Lung cancer is a malignant tumor with high rates of mortality and shows significant hereditary predisposition. Previous genome-wide association studies suggest that rs748404, located at promoter of TGM5 (transglutaminase 5), is associated with lung carcinoma. By analysis of 1000 genomes project data for three representative populations in the world, another five SNPs are identified to be in strong linkage disequilibrium with rs748404, thus suggesting that they may also be associated with lung carcinoma risk. However, it is ambiguous about the actually causal SNP(s) and the mechanism for the association. Dual-luciferase assay indicates that the functional SNPs are not rs748404, rs12911132 or rs35535629 but another three SNPs (rs66651343, rs12909095 and rs17779494) in lung cell. By chromosome conformation capture, it is disclosed that the enhancer encompassing the two SNPs, rs66651343 and rs12909095, can interact with the promoter of CCNDBP1 (cyclin D1 binding protein 1). RNA-seq data analysis indicates that CCNDBP1 expression is dependent on the genotype of these two SNPs. Chromatin immunoprecipitation assay suggests that the fragments spanning rs66651343 and rs12909095 can bind with the transcription factors, cut like homeobox 1 and SRY-box transcription factor 9, respectively. Our results establish the connection between genetic variations at this locus and lung cancer susceptibility.

Fig 1. Relative luciferase activity for different alleles of the six SNPs in Beas-2B (A) or A549 (B) cell.

For each part, the above chart is for promoter activity while the below one for enhancer. Each bar indicates one plasmid. For the two above charts, the above plasmid is the original construct, while the below two are from mutagenesis. Mg represents mutagenesis. For each SNP in below charts, the above allele is from haplotype 1 while the below one from haplotype 2. The x axis represents relative luciferase activity. All data is presented as mean ± standard deviation (SD) and normalized by the reading of empty vector. * represents P<0.01.

Fig 2. Interaction efficiency in Beas-2B cell between the enhancer containing rs66651343 and rs12909095 (A) or rs17779494 (B) and surrounding genome regions in 15q15.2.

The x axis shows the start and end of the restrictive segments in chr15 (relative to human genome build 37) while the y axis indicates the relative interaction efficiency. The above arrow shows the schematic position and transcript direction of the gene in this region. All data is shown as mean ± SD.

Fig 3. Gene expression analysis (RT-PCR) for EPB42 (left), TGM5 (middle) and TGM7 (right) in Beas-2B cell.

For each gene, the left lane is the result from positive control cell line while the right one from Beas-2B. The cell lines K562, HepG2 and SKOV-3 are used as a positive control for EPB42, TGM5 and TGM7, respectively. L represents DNA ladder, which includes DNA fragments with known size (from top to bottom, 2000, 1000, 750, 500, 200 and 100 bp). GAPDH (glyceraldehyde-3-phosphate dehydrogenase) is utilized as internal control in PCR. The arrows point out the position of PCR product for different genes.

Fig 4. Interaction efficiency in Beas-2B (A) or A549 (B) cell between the enhancer containing rs66651343 and rs12909095 and surrounding genome regions after removing three non-expressed genes.

The x axis indicates the restrictive segments location in chr15 (relative to human genome build 37) while the y axis shows the relative interaction efficiency. The above arrow shows the schematic position and transcript direction of the gene in this region. All data is shown as mean ± SD.

Fig 5. Relationship between rs12909095 genotype and CCNDBP1 expression in LCL at CEU population from literature [12] (A) and GTEx database (B).

The x-axis indicates genotype, while y-axis shows gene expression. For part A, the expression is displayed by FPKM and log transformed. For part B, the sample size for each group is shown in bracket.

Fig 6. Chromatin enrichment of the region containing position rs66651343 and rs12909095 in Beas-2B (A) or A549 (B) cell.

For each part, the left chart is for rs66651343 while the right one for rs12909095. The y axis represents relative enrichment. The result is normalized by input, and the data is presented as mean ± SD. * represents P<0.05.

Fig 7. Binding affinity difference of rs66651343 and rs12909095 alleles in Beas-2B (A) or A549 (B) cell.

For each part, the left chart is for rs66651343 while the right one for rs12909095. The top line indicates different alleles. NE denotes nuclear extracts, and the arrow points out the position of protein-probe complex.