S-Space College of Medicine/School of Medicine (의과대학/대학원) Dept. of Biomedical Sciences (대학원 의과학과) Theses (Ph.D. / Sc.D._의과학과)
Identification of single nucleotide polymorphisms and copy number variations associated with colorectal cancer susceptibility in Korean population
- 의과대학 의과학과
- Issue Date
- 서울대학교 대학원
- 학위논문 (박사)-- 서울대학교 대학원 : 의과학과 의과학전공, 2015. 8. 김종일.
- Introduction: Colorectal cancer (CRC) is one of the most common cancer worldwide. Many genome-wide association studies (GWAS) have been performed in the past few years in order to identify genetic variants contributing to CRC risk. Recent GWAS studies have identified several common single nucleotide polymorphisms (SNPs) associated with CRC risk in populations of European descent. Considering a significant racial and ethnic diversity in the human genome, it is doubtful whether the GWAS identified CRC susceptibility SNPs discovered in European populations would also be relevant in the Korean population. In addition, the incidence of CRC has lately increased significantly in Asian population, especially in Korean population. However, CRC-susceptibility genetic variants in Korean population are not well known. In this study, we aimed to identify genetic variants associated with colorectal cancer risk in Korean population. In the first part of this study, we evaluated associations between CRC risk and five SNPs, which have previously been reported to be European CRC-susceptibility SNPs, in Korean population. Furthermore, we performed genome-wide screening to identify novel SNPs associated with CRC in Korean population. In the second part of this study, we examined the presence of copy number variations (CNVs). Several GWAS studies have identified more than 20 CRC susceptibility loci
yet, these studies have focused only on the analysis of SNPs, and these variants account for only a small fraction of heritability in CRC. Recent studies have reported that copy number variations (CNVs) are considered as important human genomic variants related to cancer predisposition. However, the contribution of CNVs to the CRC development and progression remains unclear. Therefore, we performed array comparative genomic hybridization to identify CNVs associated with CRC.
Methods: To investigate the racial and ethnic diversity of CRC-susceptibility genetic variants, we genotyped 5 established European CRC-susceptibility SNPs (rs3802842, rs4779584, rs4939827, rs6983267 and rs10795668) in 198 CRC cases and 329 controls in Korean population. Genotyping was conducted using TaqMan SNP Genotyping Assays. To find novel genetic variants using genome-wide screening in Korea, Illumina HumanHap 370K and 610K BeadChips were performed on 45 and 60 CRC patients, respectively. The frequencies of genotypes in CRC patients were compared with those of Korean HapMap data. 8 candidate CRC-susceptibility SNPs were selected (rs12266240, rs10491619, rs10941887, rs1859915, rs727235, rs17051076, rs1599695 and rs902960). Subsequently, genotyping for replication was done in 189 CRC cases and 190 controls using TaqMan SNP Genotyping Assays. In the second part of this study, we performed array comparative genomic hybridization using the Agilent 180K microarray to detect CNVs associated with CRC in 36 patient and 47 control specimens. Using breakpoint PCR, we identified precise breakpoint location of PKD1L2 copy number deletion region. We validated the association between PKD1L2 copy variation and CRC risk in 1,874 cases and 2,088 controls using break point PCR.
Results: Of the five SNPs that were associated with CRC susceptibility in the European population, rs4939827 in SMAD7 was significantly associated with a decreased risk of Korean CRC [age/gender-adjusted OR (95%CI), P]: additive model, 0.67 (0.47-0.95), P=.024
dominant model, 0.59 (0.39-0.91), P=.016]. Rs10795668 [additive model, 0.61 (0.39-0.95), P=.029] and rs4779584 [dominant model, 2.50 (1.16-5.39), P=.020] were associated with CRC risk in males and females, respectively. Interestingly, in subgroup analysis with the clinical data, we found that rs4939827 had a significant association in advanced CRC patients with stage III/IV status [additive model, 0.63 (0.43-0.93), P=.020], lymph node metastasis [additive model, 0.50 (0.61-0.82), P=.006], and distant metastasis [additive model, 0.22 (0.08-0.66), P=.007]. Among 8 candidate CRC susceptibility SNPs selected from genome-wide screening, rs17051076 was found to be significantly associated with an increased risk of microsatellite instability-high (MSI-H) CRC [age/gender-adjusted OR (95%CI): additive model, 4.25 (1.51-11.98, P=.006)
dominant model, 3.52 (1.13-10.94) , P=.030] in the replication study. In the second part of the study, among the candidate CNVs identified in this experiment, only one variation, which involved PKD1L2, was located in a protein coding region and was therefore chosen as a final candidate CRC-associated CNV. Using breakpoint PCR, we confirmed the true breakpoint of the PKD1L2 deletion region, and validated the association between PKD1L2 variation and CRC risk in 1,874 cases and 2,088 controls (OR=1.44, 95% CI=1.04-1.98, P=.028). In addition, CN loss of PKD1L2 is associated with increased CRC risk in patients with age younger than 50 years (OR=2.14, 95% CI 1.39-3.30, P=5.8x10-4). In subgroup analysis according to BMI (body mass index), we found that the CN loss of PKD1L2 with BMI above or equal to 25 exhibited a significant increase in CRC risk (OR=2.29, 95% CI 1.29-4.05, P=.005). Furthermore, PKD1L2 CN loss with BMI above or equal to 25 and age below 50 is associated with a remarkably increased risk of colorectal cancer (OR=5.24, 95% CI 2.36-11.64, P= 4.8x10-5). In addition, we found that PKD1L2 variation in obese patients (BMI>=25) was associated with poor survival rate (P=.026).
Conclusions: Among the 5 European CRC-susceptibility SNPs, rs4939827, rs10795668 and rs4779584 may contribute to the risk of CRC in Korean population as well as in European populations. The new susceptibility SNP rs17051076 could be associated with MSI-H CRC in Koreans. Also, the common copy number variation in PKD1L2 is associated with cancer predisposition in CRC patients, and CN loss of PKD1L2 with BMI below or equal to 25 and/or age below 50 exhibited a significant increased risk of CRC. In addition, in obese patients, PKD1L2 variation was associated with poor survival rate.