Association Studies Between Candidate Signaling Pathway Gene Polymorphisms, GWAS Identified Variants And Breast Cancer Susceptibility

With the acceleration of life rhythm and rapid change of life style, as well asenvironmental pollution and occupational exposure, cancer hazards are becoming moreand more serious. Breast cancer is the first pathogenic and lethal cancer worldwide, andthe incidence rate of breast cancer is2times of colorectal cancer and cervical cancer. Itwas estimated that1.38million women was diagnosed as breast cancer (account for23%of the new cancer cases) and0.458million women died of breast cancer every year,while half of the news cases and60%of the deaths happened in developing countries.Recently, the prevalence of breast cancer is growing rapidly in China, and age of onset isbecoming smaller. Thus, it’s essential to identify the cause of breast cancer and set upeffective prevention, diagnosis and treatment methods.Breast cancer is a multifactorial disease caused by complex inherited andenvironmental factors, while27%causes of the disease could be explained by theinherited factors. So far, genetic studies have identified and confirmed four rare high-penetrance genes (BRCA1, BRCA2, TP53, and PTEN), four rare moderate-penetrancegenes (CHEK2, ATM, BRIP1, and PALB2), and52low penetrance SNPs (35loci)which were identified by genome-wide association studies (GWASs). However, thesegenes and loci are estimated to account for only30%of the inherited causes of breastcancer. Considering that the TGF-β signaling pathway is involved in regulating variousessential cellular processes including growth, differentiation, apoptosis, angiogenesis,homeostasis, and both breast and ovarian cancer are hormonally driven, and have similaretiologies, including inherited mutations in BRCA1or BRCA2; Hyperlipidemia isassociated with risk of breast cancer, colorectal cancer, prostate cancer and biliary tractcancers and Cholesterol has been also hypothesized to increase breast cancer risk given that cholesterol is the precursor to steroid hormone synthesis and endogenous sex steroidhormones are directly related to breast cancer risk, here, we comprehensively evaluatethe associations of genetic variants in the TGF-β signaling pathway, GWAS identifiedvariants of ovarian cancer and lipid traits with breast cancer risk.Part I Transforming Growth Factor Beta Signaling Pathway Variants andBreast Cancer Risk: Results From the Asian Consortium StudyObjectives: The TGF-β signaling pathway is composed of several multifunctionalcytokines and receptors that are involved in regulating various essential cellularprocesses including growth, differentiation, apoptosis, angiogenesis, and homeostasis.The objective of this study was to comprehensively evaluate the associations of geneticvariants in the TGF-β signaling pathway with breast cancer risk among Asian women.Subjects and methods: The subjects comes from10Asian consortium studies: theShanghai Breast Cancer Study, the Shanghai Breast Cancer Survival Study, the ShanghaiEndometrial Cancer Study, the Shanghai Women’s Health Study,2,095women fromTaiwan;1,050women from Hong Kong;3,580women from Nanjing,1,657women fromGuangzhou,1,284women from Nagoya, Japan and795women from Nagano, Japan.Appropriate approval was granted from all relevant Institutional Review Boards and allincluded participants provided informed consent. A three-stage case-control study wasdesigned: analysis for stage I was conducted for341SNPs from TGFB1, TGFB2, TGFB3,TGFBR1, TGFBR2, TGFBR3, SMAD2, SMAD3, SMAD4, SMAD7and SARA geneswith a minimum minor allele frequency (MAF) of5%among genotyped controls.Nineteen promising SNPs were selected for additional stage II. One replicated SNP(rs1078985) was further evaluated among participants of SBCGS III and6Asiancollaboration studies (stage III). All included SNPs had call rates and concordance ratesof at least95%among duplicates within each platform, as well as across genotypingplatforms. Laboratory personnel were blinded to the case–control and QC status of allsamples. ORs and corresponding CIs were determined by logistic regression models thatincluded adjustment for age. Additional adjustment for education, age at menarche, ageat menopause, age at first live birth, menopausal status, a first-degree relative with breast cancer, use of hormone replacement therapy, previous history of fibroadenoma, physicalactivity, body mass index, and waist-to-hip ratio were done when appropriate; Additive,dominant, and recessive models of effect were used for all SNPs; stratified analyses weredone by TNM; haplotype analysis were performed by HAPSTAT.Results and conclusions: TGFBR2rs1078985showed consistent association withbreast cancer risk in all three stages (12,004cases and11,789controls). Pooled analysisof all data indicated a highly significant recessive effect (OR,0.76;95%CI,0.65–0.89,P=8.42×10-4). Interaction analysis showed rs1078985was significant associated withfamily history of breast cancer (P=0.0069). This is most significant SNP in recessivemodel which was identified in candidate gene studies. Results from in-silico analysiswere supportive of an significant association between TGFBR2rs1078985and breastcancer risk. In addition, nominally significant associations with breast cancer risk werealso found for TGFB2rs2799086, TGFB2rs17047740, TGFBR1rs2026811, TGFBR1rs10733710, TGFBR2rs304822, TGFBR3rs284185, and SMAD3rs7178117in thecombined analyses, although none of these SNPs had significant associations in stage II.The results above supported the essential role of TGF-β signaling pathway in thecarcinogenesis of breast.Part II Selected GWAS Identified Variants and Breast Cancer riskChapter I GWAS Identified Ovarian Cancer Susceptibility Variants andBreast Cancer riskObjectives: Both breast and ovarian cancer are hormonally driven, and have similaretiologies, including inherited mutations in BRCA1or BRCA2. The objective of thisstudy was to comprehensively evaluate the associations of GWAS identified variants ofovarian cancer with breast cancer risk among Asian women.Subjects and methods: This analysis included a total of2,918breast cancer casesfrom the Shanghai Breast Cancer Study (SBCS), Shanghai Breast Cancer Survival Study(SBCSS), and the Shanghai Women’s Health Study (SWHS), and2,324controls fromSBCS and SWHS. Study protocols were approved by the relevant review boards of allinstitutions, and written informed consent was obtained from all participants. The103 variants from5loci were included in this analysis. Genotyping data was from theAffymetrix Genome Wide Array (6.0) or from MACH1.0imputation. All included SNPshad call rates and concordance rates of at least95%among duplicates within eachplatform, as well as across genotyping platforms; r2>0.3were included for imputationdata. Laboratory personnel were blinded to the case–control and QC status of all samples.Multivariate logistic regression was used to estimate odds ratios (ORs) and95%confidence intervals (CIs) for associations between breast cancer risk and geneticvariants using additive effect models that included adjustment for age and education.Additional adjustment for education, age at menarche, age at menopause, age at first livebirth, menopausal status, a first-degree relative with breast cancer, use of hormonereplacement therapy, previous history of fibroadenoma, physical activity, body massindex, and waist-to-hip ratio were done when appropriate; stratified analyses were doneby menopause status, ER, PR, TNM, BMI, WHR and age at menarche.Results and conclusions: No associations with breast cancer risk in additive effectmodels adjusted for age and education were identified for the103variants. Stratifiedanalyses by age at menarche and WHR identified that5SNPs in loci2(rs717852，rs2857540，rs1318778，rs2072590，rs711830) statistically increased breast cancer riskamong women with age at menarche smaller than14.7,while one SNP in loci7(rs7206951)statistically increased breast cancer risk among women with WHR smallerthan0.82. The results were still significant when corrected by Bonferroni method.Among them, the5SNPs in loci2have significant interaction with age at menarche. Theresults above showed the shared disease cause between ovarian cancer and breast cancer.Chapter II GWAS Identified Lipid Trait Susceptibility Variants andBreast Cancer riskObjectives: Hyperlipidemia, generally characterized by increased levels of totalcholesterol (TC), triglycerides (TG), LDL cholesterol (LDL-C) and decreased level ofHDL cholesterol (HDL-C), has been associated with risk of breast cancer, colorectalcancer, prostate cancer and biliary tract cancers. Cholesterol has been also hypothesizedto increase breast cancer risk given that cholesterol is the precursor to steroid hormone synthesis and endogenous sex steroid hormones are directly related to breast cancer risk.The objective of this study was to comprehensively evaluate the associations of GWASidentified variants of lipid traits with breast cancer risk.Subjects and methods: The NBHS is a population-based case-control study ofincident breast cancer conducted in the Nashville, TN metropolitan area. Eligible caseswere women diagnosed with invasive breast cancer or ductal carcinoma in situ, who werebetween the ages of25and75, had no prior history of cancer, had a resident telephones,spoke English, and who were able to provide consent to the study. Eligibility controlscould not have a prior cancer diagnosis and were frequency matched to cases on5-yearage group, race, and county of residence. In the current analysis, a total of1,514casesand1,285controls from NBHS were included and all the subjects were Caucasians.Study protocols were approved by the relevant review boards of all institutions, andwritten informed consent was obtained from all participants. These SNPs were genotypedusing an exome chip (Illumina Inc, San Diego CA, USA); Odds ratios and correspondingconfidence intervals (OR,95%CI) of the163lipid-trait GWAS identified variants (104loci) with breast cancer risk were determined by logistic regression model that includedadjustment for age and education levels; To measure the cumulative effect of multiplegenetic risk variants, we calculated GRSs by counting the number of risk alleles at eachlocus (0,1, or2) together; Interactions were evaluated using likelihood ratio tests andcase-only study.Results and conclusions:8SNPs (rs2652834， rs6450176， rs2142672，rs10128711，rs3177928，rs6756629，rs2068888and rs10195252) showed consistentrisk alleles with our hypothesis and significantly association with breast cancer risk.Among them rs10128711and rs2142672showed a marginal interaction (P=0.056). Noneof the GRSs were significantly associated with breast cancer risk. We found a statisticallysignificant interaction between GRSHDL-Cand family history of breast cancer (P=0.023).The conclusion need to be confirmed by Mendelian randomization among largerpopulations.