Lung Cancer 89 (2015) 238–242
Contents lists available at ScienceDirect
Lung Cancer journal homepage: www.elsevier.com/locate/lungcan
Polymorphisms in alternative splicing associated genes are associated with lung cancer risk in a Chinese population Wei Shen a,1 , Rong Yin b,1 , Cheng Wang a , Meng Zhu a , Wen Zhou a , Na Qin a , Jie Sun a , Jia Liu a , Jing Dong a , Guangfu Jin a , Hongxia Ma a , Zhibin Hu a,b , Hongbing Shen a , Lin Xu b,∗∗, Juncheng Dai a,∗ a
Department of Epidemiology and Biostatistics, Jiangsu Key Lab of Cancer Biomarkers, Prevention and Treatment, Collaborative Innovation Center For Cancer Personalized Medicine, School of Public Health, Nanjing Medical University, Nanjing 211166, China b Jiangsu Key Laboratory of Molecular and Translational Cancer Research, Collaborative Innovation Center For Cancer Personalized Medicine, Nanjing Medical University Affiliated Cancer Hospital, Nanjing 210009, China
a r t i c l e
i n f o
Article history: Received 24 December 2014 Received in revised form 19 May 2015 Accepted 14 June 2015 Keywords: Alternative splicing Polymorphism Lung cancer Susceptibility
a b s t r a c t Background: Alternative splicing is an important biological step during mRNA processing. Misregulation of alternative splicing can produce aberrant protein isoforms, thus contributing to cancer. We hypothesized that variants in 5 critical splicing factor-associated genes might play an important role in carcinogenesis of lung cancer. Materials and methods: A case-control study including 1,341 non-small cell lung cancer (NSCLC) cases and 1,982 cancer-free controls were conducted to evaluate the associations of 16 tagging/functional polymorphisms in 5 splicing factor-associated genes with lung cancer risk. Results: We found altogether 8 SNPs were associated with lung cancer risk with adjustment of age, gender, and smoking status after multiple corrections (FDR). Among these, six SNPs were related with SRSF7(rs10197412, OR(95%CI) = 1.23(1.06–1.43), P for FDR = 0.018; rs12621103, OR(95%CI) = 1.25(1.08–1.46), P for FDR = 0.016; rs13024811, OR(95%CI) = 1.25(1.07–1.46), P for FDR = 0.016; rs2037875, OR(95%CI) = 1.23(1.06–1.42), P for FDR = 0.018; rs3134628, OR(95%CI) = 1.25(1.07–1.45), P for FDR = 0.016 and rs6715866, OR(95%CI) = 1.23(1.07–1.43), P for FDR = 0.016); one SNP was near PTBP2 (rs12566237: OR(95%CI) = 1.16(1.05–1.28), P for FDR = 0.016) and one SNP in HNRNPQ (rs16876385: OR(95%CI) = 1.17(1.04–1.32), P for FDR = 0.022). Conclusions: Our findings indicated that genetic variants in these splicing-associated genes might modify individual susceptibility to lung cancer in Chinese population. Further large-scale well-formed population studies and functional researches are warranted to confirm our findings. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Lung cancer is the most commonly diagnosed as well as the leading cause of cancer-related deaths in most countries [1]. In
∗ Corresponding author at: Department of Epidemiology and Biostatistics, Collaborative Innovation Center for Cancer Personalized Medicine, School of Public Health, Nanjing Medical University, Nanjing 211166, China. Tel.: +86 25 8686 8437; Fax: +86 25 8686 8439. ∗∗ Corresponding author at: Department of Thoracic Surgery, Jiangsu Key Laboratory of Molecular and Translational Cancer Research, Collaborative Innovation Center for Cancer Personalized Medicine, Nanjing Medical University Affiliated Cancer Hospital, Nanjing 210009, China. E-mail addresses: [email protected] (J. Dai), [email protected] (L. Xu). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.lungcan.2015.06.010 0169-5002/© 2015 Elsevier Ireland Ltd. All rights reserved.
China, the estimated incidence in 2012 was 36.1 per 100,000 and estimated mortality was 32.5 per 100,000, being the top one of all the cancers [2]. Although tobacco smoking remains the most important aetiological risk factor for lung cancer, genetic factors, such as Single Nucleotide Polymorphisms (SNPs), have proved of importance in the development of lung cancer [3]. During the past decades, genome-wide association studies (GWAS), a powerful approach in exploring the genetic determinants of complex diseases, have identified a number of SNPs associated with the risk of lung cancer [4]. However, the results of heritability analysis showed that these loci can only interpret a small fraction of lung cancer risk [5]. Additional efforts based on important candidate genes which were related with the pathogenesis may help us detect the missing heritability of lung cancer in the post-GWAS era.
W. Shen et al. / Lung Cancer 89 (2015) 238–242
Splicing is a modification of pre-messenger RNA transcript and alternative splicing (AS) may lead to multiple mRNAs from a single gene. Recent study in genome-wide scale has found more than 90% of human genes undergo AS [6]. Misregulation of AS can produce aberrant protein isoforms, which may contribute to cancer [7]. Well studied splicing factors with antagonistic functions mainly include two families: serine/arginine-rich (SR) proteins and heterogeneous nuclear ribonucleoproteins (HNRNPs) [8]. Generally, SR proteins recruit spliceosomal components and promote exon recognition [9], while HNRNP factors inhibit splicing directly or promote exon skipping [10]. Although some related factors have been clearly illustrated for lung cancer risk previously, such as ASF/SF2 (SRSF1) [7], SRSF2 (SC35) [11], SRSF5 (SRp40) [12], SRSF6 (SRp55) [12], HNRNPA1 [13], HNRNP A2/B1 [14], HNRNP C1/C2, HNRNP K [13], there are still some critical splicing factors involved in tumorigenesis and tumor progression (SRSF7, PTBP2, HNRNPI, HNRNPQ, HNRNPM, etc.) [15–17], and few studies of them were available for lung cancer. Therefore, we hypothesized that variants in these critical genes may associate with lung cancer risk in Chinese population, and a case-control study including 1,341 non-small cell lung cancer (NSCLC) cases and 1,982 controls was conducted to evaluate the effects of 16 polymorphisms located at 5 splicing factor-associated genes. 2. Materials and methods 2.1. Study subjects A total of 1,341 lung cancer cases and 1,982 cancer-free controls were included. These cases were incident lung cancer patients and consecutively recruited between 2003 and 2009 from the First Affiliated Hospital of Nanjing Medical University and the Cancer Hospital of Jiangsu Province (Nanjing, China). All of them were histopathologically or cytologically confirmed by at least two pathologists. Those with a history of other cancers and ever received radiotherapy or chemotherapy were excluded. The control subjects were randomly selected from individuals receiving routine physical examinations in local hospitals or participating in a community-based screening program for non-infectious diseases conducted in Jiangsu province during the same period when the cases were recruited. All the control subjects were frequencymatched to the cases by age (±5years), gender and residential area (urban or countryside). Participants were unrelated Han Chinese and were face-to-face interviewed by trained interviewers to collect individual information on demographic data, life styles such as histories of smoking were obtained from each subject after the informed consent. Those who smoked less than one cigarette a day and with a cumulative time of less than one year were defined as non-smokers, and smokers on the contrary, with those quit smoking for more than one year as former smokers. Five-ml of venous blood was collected from each subject after the interview. All study subjects and experiments used in the study were approved by the Institutional Review Board of Nanjing Medical University. 2.2. Selection and genotyping assays of polymorphisms Several important genes were involved in the mechanistic regulation of AS, five of them (SRSF7, PTBP2, HNRNPI, HNRNPQ, HNRNPM) were selected according to systematically review on cancer related studies [15–18]. The detailed selection strategies were described as followings: Firstly, according to HapMap database (phase II + III Feb 09, CHB + JPT, on NCBI B36 assembly), a series of criteria (P for Hardy–Weinberg equilibrium (HWE) ≥ 0.05; genotyping rate ≥ 90%; Mendel errors ≤ 1; Minor allele frequency
239
(MAF) ≥ 0.05) were checked for the regions of the 5 genes (plus 10-kb up-stream region), 87 SNPs were selected for further evaluation; Secondly, according to a web-based tool (http://snpinfo.niehs. nih.gov/snpinfo/snpfunc.htm), potentially functional SNPs were marked; Lastly, Haploview 4.2 was used to find tagging SNPs with a r2 threshold of 0.8. All the potentially functional SNPs were compulsively included. Finally, 20 SNPs were eventually selected and 4 of them were excluded for failed probe design (Supplementary Table 2). The methods of isolating genomic DNA and genotyping were similar with the previous study [19]. Briefly, the genotyping was performed without knowing the subjects’ case or control status by Illumina Infinium® Human Exome+ BeadChip (Illumina Inc.), a customized version of Human Exome BeadChip. Genotype calling was performed using the GenTrain version 1.0 clustering algorithm in GenomeStudio V2011.1 (Illumina). 2.3. Statistical analysis Fisher’s Exact Test was used to evaluate the difference of distributions among cases and controls for categorical variables. Similarly, Student’s t-test was used for continuous variables and Welch’s t-test was applied for the situations of unequalled variances. Odds ratios (ORs) and their 95% confidence intervals (CIs) were calculated by using logistic regression model to evaluate the associations between genotypes and lung cancer risk with adjustment for age, gender and pack years of smoking. HWE was evaluated using goodness-of-fit 2 test among the controls. The heterogeneity between subgroups was assessed using 2 -based Cochrane’s Q test. False Discovery Rate (FDR) (Benjamini and Hochberg, 1995) was used for multiple correction [20]. Haplo.stats (an R package) was used to estimate the haplotype for each individual based on the observed genotypes. All the statistical analyses were two-sided with 0.05 as significant level and performed with R software (Version 3.1.0, 2014-04-10; R Foundation for Statistical Computing), unless indicated elsewhere. 3. Results 3.1. General description of subjects The distribution of demographic and smoking status between cases and controls was summarized in Supplementary Table 1. Briefly, age and gender status between cases and controls showed no significance (P > 0.05). The mean age was 61.06 ± 10.15 for the cases and 61.32 ± 11.07 for the controls. For smoking status, the cases were more likely to be current or former smokers (61.08% in cases vs. 48.54% in controls, P < 0.001). Among 1,341 NSCLC cases, 481 (35.87%) were squamous cell carcinoma and 860 (64.13%) were adenocarcinoma. 3.2. General information of selected SNPs The aliases, locus, position and corresponding tagging and functional SNPs of selected 5 genes were listed in Supplementary Table 2. The general information and associations of the 16 SNPs in 5 genes with lung cancer risk using additive model were exhibited in Supplementary Table 3. The observed genotype frequencies for these SNPs were all in agreement with HWE among controls. All the individuals were successfully genotyped for each SNP with a call rate above 99.00%. 3.3. Association analysis of SNPs and lung cancer We observed eight SNPs from three genes with significant effects on lung cancer in both additive and dominant models with
240
W. Shen et al. / Lung Cancer 89 (2015) 238–242
Table 1 Association results between eight SNPs and lung cancer risk. Genotype
Case
Control
Crude OR (95%CI)
Adjusted OR (95%CI)a
Pa
P for FDR
SRSF7 rs10197412(A > G) AA AG GG AG/GG Additive model
1001 315 25 340
1546 406 30 436
1.00 1.20(1.01–1.42) 1.29(0.75–2.20) 1.20(1.02–1.42) 1.18(1.02–1.37)
1.00 1.25(1.05–1.49) 1.35(0.78–2.34) 1.26(1.07–1.49) 1.23(1.06–1.43)
0.010 0.290 0.007 0.008
0.018
SRSF7 rs12621103(G > A) GG GA AA GA/AA Additive model
1015 302 24 326
1.00 1.23(1.04–1.46) 1.33(0.77–2.3) 1.24(1.05–1.46) 1.21(1.04–1.4)
1.00 1.28(1.08–1.53) 1.39(0.79–2.45) 1.29(1.09–1.53) 1.25(1.08–1.46)
0.006 0.256 0.004 0.004
0.016
SRSF7 rs13024811(G > A) GG GA AA GA/AA Additive model
1008 309 24 333
1.00 1.24(1.05–1.47) 1.24(0.72–2.14) 1.24(1.05–1.46) 1.2(1.04–1.4)
1.00 1.3(1.09–1.54) 1.29(0.74–2.26) 1.3(1.09–1.53) 1.25(1.07–1.46)
0.004 0.365 0.003 0.004
0.016
1.00 1.19(1.01–1.40) 1.35(0.80–2.27) 1.20(1.02–1.41) 1.18(1.02–1.36)
1.00 1.24(1.04–1.47) 1.41(0.83–2.42) 1.25(1.06–1.48) 1.23(1.06–1.42)
0.014 0.205 0.008 0.007
0.018
1.00 1.24(1.04–1.47) 1.24(0.72–2.13) 1.24(1.05–1.46) 1.2(1.04–1.4)
1.00 1.29(1.09–1.54) 1.29(0.74–2.25) 1.29(1.09–1.53) 1.25(1.07–1.45)
0.004 0.369 0.003 0.004
0.016
1.00 1.21(1.03–1.43) 1.24(0.76–2.04) 1.21(1.03–1.43) 1.18(1.03–1.36)
1.00 1.28(1.08–1.52) 1.28(0.77–2.14) 1.28(1.08–1.51) 1.23(1.07–1.43)
0.005 0.339 0.004 0.005
0.016
1.00 1.13(0.97–1.33) 1.36(1.11–1.66) 1.19(1.03–1.38) 1.16(1.05–1.28)
1.00 1.12(0.95–1.31) 1.35(1.10–1.66) 1.18(1.01–1.37) 1.16(1.05–1.28)
0.183 0.004 0.035 0.005
0.016
1.00 1.18(1.02–1.37) 1.32(0.96–1.82) 1.20(1.04–1.38) 1.17(1.04–1.31)
1.00 1.17(1.01–1.36) 1.36(0.98–1.89) 1.19(1.03–1.38) 1.17(1.04–1.32)
0.041 0.065 0.017 0.011
0.022
1.00 1.15(0.96–1.37) 1.45(1.16–1.82) 1.31(1.06–1.61) 1.66(1.13–2.43) 1.00 1.24(1.06–1.45) 1.37(1.13–1.65)
1.00 1.13(0.94–1.35) 1.48(1.17–1.86) 1.36(1.10–1.69) 1.74(1.18–2.58) 1.00 1.23(1.05–1.45) 1.43(1.17–1.74)
SRSF7 rs2037875(A > G) AA AG GG AG/GG Additive model
994 320 27 347
SRSF7 rs3134628(G > A) GG GA AA GA/AA Additive model
1010 307 24 331
SRSF7 rs6715866(A > G) AA AG GG AG/GG Additive model
982 327 29 356
PTBP2 rs12566237(A > G) AA AG GG AG/GG Additive model
404 657 280 937
HNRNPQ rs16876385(G > C) GG GC CC GC/CC Additive model
768 499 74 573
1573 381 28 409
1564 387 30 417
1535 416 31 447
1564 384 30 414
1514 416 36 452
673 966 343 1309
1221 672 89 761
b
Combined Analysis 0–1 2 3–6 7–8 9–15 0–1 2–6 ≥7 Trend P
546 332 184 219 57 546 516 276
922 488 214 283 58 922 702 341
0.189
Contents lists available at ScienceDirect
Lung Cancer journal homepage: www.elsevier.com/locate/lungcan
Polymorphisms in alternative splicing associated genes are associated with lung cancer risk in a Chinese population Wei Shen a,1 , Rong Yin b,1 , Cheng Wang a , Meng Zhu a , Wen Zhou a , Na Qin a , Jie Sun a , Jia Liu a , Jing Dong a , Guangfu Jin a , Hongxia Ma a , Zhibin Hu a,b , Hongbing Shen a , Lin Xu b,∗∗, Juncheng Dai a,∗ a
Department of Epidemiology and Biostatistics, Jiangsu Key Lab of Cancer Biomarkers, Prevention and Treatment, Collaborative Innovation Center For Cancer Personalized Medicine, School of Public Health, Nanjing Medical University, Nanjing 211166, China b Jiangsu Key Laboratory of Molecular and Translational Cancer Research, Collaborative Innovation Center For Cancer Personalized Medicine, Nanjing Medical University Affiliated Cancer Hospital, Nanjing 210009, China
a r t i c l e
i n f o
Article history: Received 24 December 2014 Received in revised form 19 May 2015 Accepted 14 June 2015 Keywords: Alternative splicing Polymorphism Lung cancer Susceptibility
a b s t r a c t Background: Alternative splicing is an important biological step during mRNA processing. Misregulation of alternative splicing can produce aberrant protein isoforms, thus contributing to cancer. We hypothesized that variants in 5 critical splicing factor-associated genes might play an important role in carcinogenesis of lung cancer. Materials and methods: A case-control study including 1,341 non-small cell lung cancer (NSCLC) cases and 1,982 cancer-free controls were conducted to evaluate the associations of 16 tagging/functional polymorphisms in 5 splicing factor-associated genes with lung cancer risk. Results: We found altogether 8 SNPs were associated with lung cancer risk with adjustment of age, gender, and smoking status after multiple corrections (FDR). Among these, six SNPs were related with SRSF7(rs10197412, OR(95%CI) = 1.23(1.06–1.43), P for FDR = 0.018; rs12621103, OR(95%CI) = 1.25(1.08–1.46), P for FDR = 0.016; rs13024811, OR(95%CI) = 1.25(1.07–1.46), P for FDR = 0.016; rs2037875, OR(95%CI) = 1.23(1.06–1.42), P for FDR = 0.018; rs3134628, OR(95%CI) = 1.25(1.07–1.45), P for FDR = 0.016 and rs6715866, OR(95%CI) = 1.23(1.07–1.43), P for FDR = 0.016); one SNP was near PTBP2 (rs12566237: OR(95%CI) = 1.16(1.05–1.28), P for FDR = 0.016) and one SNP in HNRNPQ (rs16876385: OR(95%CI) = 1.17(1.04–1.32), P for FDR = 0.022). Conclusions: Our findings indicated that genetic variants in these splicing-associated genes might modify individual susceptibility to lung cancer in Chinese population. Further large-scale well-formed population studies and functional researches are warranted to confirm our findings. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Lung cancer is the most commonly diagnosed as well as the leading cause of cancer-related deaths in most countries [1]. In
∗ Corresponding author at: Department of Epidemiology and Biostatistics, Collaborative Innovation Center for Cancer Personalized Medicine, School of Public Health, Nanjing Medical University, Nanjing 211166, China. Tel.: +86 25 8686 8437; Fax: +86 25 8686 8439. ∗∗ Corresponding author at: Department of Thoracic Surgery, Jiangsu Key Laboratory of Molecular and Translational Cancer Research, Collaborative Innovation Center for Cancer Personalized Medicine, Nanjing Medical University Affiliated Cancer Hospital, Nanjing 210009, China. E-mail addresses: [email protected] (J. Dai), [email protected] (L. Xu). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.lungcan.2015.06.010 0169-5002/© 2015 Elsevier Ireland Ltd. All rights reserved.
China, the estimated incidence in 2012 was 36.1 per 100,000 and estimated mortality was 32.5 per 100,000, being the top one of all the cancers [2]. Although tobacco smoking remains the most important aetiological risk factor for lung cancer, genetic factors, such as Single Nucleotide Polymorphisms (SNPs), have proved of importance in the development of lung cancer [3]. During the past decades, genome-wide association studies (GWAS), a powerful approach in exploring the genetic determinants of complex diseases, have identified a number of SNPs associated with the risk of lung cancer [4]. However, the results of heritability analysis showed that these loci can only interpret a small fraction of lung cancer risk [5]. Additional efforts based on important candidate genes which were related with the pathogenesis may help us detect the missing heritability of lung cancer in the post-GWAS era.
W. Shen et al. / Lung Cancer 89 (2015) 238–242
Splicing is a modification of pre-messenger RNA transcript and alternative splicing (AS) may lead to multiple mRNAs from a single gene. Recent study in genome-wide scale has found more than 90% of human genes undergo AS [6]. Misregulation of AS can produce aberrant protein isoforms, which may contribute to cancer [7]. Well studied splicing factors with antagonistic functions mainly include two families: serine/arginine-rich (SR) proteins and heterogeneous nuclear ribonucleoproteins (HNRNPs) [8]. Generally, SR proteins recruit spliceosomal components and promote exon recognition [9], while HNRNP factors inhibit splicing directly or promote exon skipping [10]. Although some related factors have been clearly illustrated for lung cancer risk previously, such as ASF/SF2 (SRSF1) [7], SRSF2 (SC35) [11], SRSF5 (SRp40) [12], SRSF6 (SRp55) [12], HNRNPA1 [13], HNRNP A2/B1 [14], HNRNP C1/C2, HNRNP K [13], there are still some critical splicing factors involved in tumorigenesis and tumor progression (SRSF7, PTBP2, HNRNPI, HNRNPQ, HNRNPM, etc.) [15–17], and few studies of them were available for lung cancer. Therefore, we hypothesized that variants in these critical genes may associate with lung cancer risk in Chinese population, and a case-control study including 1,341 non-small cell lung cancer (NSCLC) cases and 1,982 controls was conducted to evaluate the effects of 16 polymorphisms located at 5 splicing factor-associated genes. 2. Materials and methods 2.1. Study subjects A total of 1,341 lung cancer cases and 1,982 cancer-free controls were included. These cases were incident lung cancer patients and consecutively recruited between 2003 and 2009 from the First Affiliated Hospital of Nanjing Medical University and the Cancer Hospital of Jiangsu Province (Nanjing, China). All of them were histopathologically or cytologically confirmed by at least two pathologists. Those with a history of other cancers and ever received radiotherapy or chemotherapy were excluded. The control subjects were randomly selected from individuals receiving routine physical examinations in local hospitals or participating in a community-based screening program for non-infectious diseases conducted in Jiangsu province during the same period when the cases were recruited. All the control subjects were frequencymatched to the cases by age (±5years), gender and residential area (urban or countryside). Participants were unrelated Han Chinese and were face-to-face interviewed by trained interviewers to collect individual information on demographic data, life styles such as histories of smoking were obtained from each subject after the informed consent. Those who smoked less than one cigarette a day and with a cumulative time of less than one year were defined as non-smokers, and smokers on the contrary, with those quit smoking for more than one year as former smokers. Five-ml of venous blood was collected from each subject after the interview. All study subjects and experiments used in the study were approved by the Institutional Review Board of Nanjing Medical University. 2.2. Selection and genotyping assays of polymorphisms Several important genes were involved in the mechanistic regulation of AS, five of them (SRSF7, PTBP2, HNRNPI, HNRNPQ, HNRNPM) were selected according to systematically review on cancer related studies [15–18]. The detailed selection strategies were described as followings: Firstly, according to HapMap database (phase II + III Feb 09, CHB + JPT, on NCBI B36 assembly), a series of criteria (P for Hardy–Weinberg equilibrium (HWE) ≥ 0.05; genotyping rate ≥ 90%; Mendel errors ≤ 1; Minor allele frequency
239
(MAF) ≥ 0.05) were checked for the regions of the 5 genes (plus 10-kb up-stream region), 87 SNPs were selected for further evaluation; Secondly, according to a web-based tool (http://snpinfo.niehs. nih.gov/snpinfo/snpfunc.htm), potentially functional SNPs were marked; Lastly, Haploview 4.2 was used to find tagging SNPs with a r2 threshold of 0.8. All the potentially functional SNPs were compulsively included. Finally, 20 SNPs were eventually selected and 4 of them were excluded for failed probe design (Supplementary Table 2). The methods of isolating genomic DNA and genotyping were similar with the previous study [19]. Briefly, the genotyping was performed without knowing the subjects’ case or control status by Illumina Infinium® Human Exome+ BeadChip (Illumina Inc.), a customized version of Human Exome BeadChip. Genotype calling was performed using the GenTrain version 1.0 clustering algorithm in GenomeStudio V2011.1 (Illumina). 2.3. Statistical analysis Fisher’s Exact Test was used to evaluate the difference of distributions among cases and controls for categorical variables. Similarly, Student’s t-test was used for continuous variables and Welch’s t-test was applied for the situations of unequalled variances. Odds ratios (ORs) and their 95% confidence intervals (CIs) were calculated by using logistic regression model to evaluate the associations between genotypes and lung cancer risk with adjustment for age, gender and pack years of smoking. HWE was evaluated using goodness-of-fit 2 test among the controls. The heterogeneity between subgroups was assessed using 2 -based Cochrane’s Q test. False Discovery Rate (FDR) (Benjamini and Hochberg, 1995) was used for multiple correction [20]. Haplo.stats (an R package) was used to estimate the haplotype for each individual based on the observed genotypes. All the statistical analyses were two-sided with 0.05 as significant level and performed with R software (Version 3.1.0, 2014-04-10; R Foundation for Statistical Computing), unless indicated elsewhere. 3. Results 3.1. General description of subjects The distribution of demographic and smoking status between cases and controls was summarized in Supplementary Table 1. Briefly, age and gender status between cases and controls showed no significance (P > 0.05). The mean age was 61.06 ± 10.15 for the cases and 61.32 ± 11.07 for the controls. For smoking status, the cases were more likely to be current or former smokers (61.08% in cases vs. 48.54% in controls, P < 0.001). Among 1,341 NSCLC cases, 481 (35.87%) were squamous cell carcinoma and 860 (64.13%) were adenocarcinoma. 3.2. General information of selected SNPs The aliases, locus, position and corresponding tagging and functional SNPs of selected 5 genes were listed in Supplementary Table 2. The general information and associations of the 16 SNPs in 5 genes with lung cancer risk using additive model were exhibited in Supplementary Table 3. The observed genotype frequencies for these SNPs were all in agreement with HWE among controls. All the individuals were successfully genotyped for each SNP with a call rate above 99.00%. 3.3. Association analysis of SNPs and lung cancer We observed eight SNPs from three genes with significant effects on lung cancer in both additive and dominant models with
240
W. Shen et al. / Lung Cancer 89 (2015) 238–242
Table 1 Association results between eight SNPs and lung cancer risk. Genotype
Case
Control
Crude OR (95%CI)
Adjusted OR (95%CI)a
Pa
P for FDR
SRSF7 rs10197412(A > G) AA AG GG AG/GG Additive model
1001 315 25 340
1546 406 30 436
1.00 1.20(1.01–1.42) 1.29(0.75–2.20) 1.20(1.02–1.42) 1.18(1.02–1.37)
1.00 1.25(1.05–1.49) 1.35(0.78–2.34) 1.26(1.07–1.49) 1.23(1.06–1.43)
0.010 0.290 0.007 0.008
0.018
SRSF7 rs12621103(G > A) GG GA AA GA/AA Additive model
1015 302 24 326
1.00 1.23(1.04–1.46) 1.33(0.77–2.3) 1.24(1.05–1.46) 1.21(1.04–1.4)
1.00 1.28(1.08–1.53) 1.39(0.79–2.45) 1.29(1.09–1.53) 1.25(1.08–1.46)
0.006 0.256 0.004 0.004
0.016
SRSF7 rs13024811(G > A) GG GA AA GA/AA Additive model
1008 309 24 333
1.00 1.24(1.05–1.47) 1.24(0.72–2.14) 1.24(1.05–1.46) 1.2(1.04–1.4)
1.00 1.3(1.09–1.54) 1.29(0.74–2.26) 1.3(1.09–1.53) 1.25(1.07–1.46)
0.004 0.365 0.003 0.004
0.016
1.00 1.19(1.01–1.40) 1.35(0.80–2.27) 1.20(1.02–1.41) 1.18(1.02–1.36)
1.00 1.24(1.04–1.47) 1.41(0.83–2.42) 1.25(1.06–1.48) 1.23(1.06–1.42)
0.014 0.205 0.008 0.007
0.018
1.00 1.24(1.04–1.47) 1.24(0.72–2.13) 1.24(1.05–1.46) 1.2(1.04–1.4)
1.00 1.29(1.09–1.54) 1.29(0.74–2.25) 1.29(1.09–1.53) 1.25(1.07–1.45)
0.004 0.369 0.003 0.004
0.016
1.00 1.21(1.03–1.43) 1.24(0.76–2.04) 1.21(1.03–1.43) 1.18(1.03–1.36)
1.00 1.28(1.08–1.52) 1.28(0.77–2.14) 1.28(1.08–1.51) 1.23(1.07–1.43)
0.005 0.339 0.004 0.005
0.016
1.00 1.13(0.97–1.33) 1.36(1.11–1.66) 1.19(1.03–1.38) 1.16(1.05–1.28)
1.00 1.12(0.95–1.31) 1.35(1.10–1.66) 1.18(1.01–1.37) 1.16(1.05–1.28)
0.183 0.004 0.035 0.005
0.016
1.00 1.18(1.02–1.37) 1.32(0.96–1.82) 1.20(1.04–1.38) 1.17(1.04–1.31)
1.00 1.17(1.01–1.36) 1.36(0.98–1.89) 1.19(1.03–1.38) 1.17(1.04–1.32)
0.041 0.065 0.017 0.011
0.022
1.00 1.15(0.96–1.37) 1.45(1.16–1.82) 1.31(1.06–1.61) 1.66(1.13–2.43) 1.00 1.24(1.06–1.45) 1.37(1.13–1.65)
1.00 1.13(0.94–1.35) 1.48(1.17–1.86) 1.36(1.10–1.69) 1.74(1.18–2.58) 1.00 1.23(1.05–1.45) 1.43(1.17–1.74)
SRSF7 rs2037875(A > G) AA AG GG AG/GG Additive model
994 320 27 347
SRSF7 rs3134628(G > A) GG GA AA GA/AA Additive model
1010 307 24 331
SRSF7 rs6715866(A > G) AA AG GG AG/GG Additive model
982 327 29 356
PTBP2 rs12566237(A > G) AA AG GG AG/GG Additive model
404 657 280 937
HNRNPQ rs16876385(G > C) GG GC CC GC/CC Additive model
768 499 74 573
1573 381 28 409
1564 387 30 417
1535 416 31 447
1564 384 30 414
1514 416 36 452
673 966 343 1309
1221 672 89 761
b
Combined Analysis 0–1 2 3–6 7–8 9–15 0–1 2–6 ≥7 Trend P
546 332 184 219 57 546 516 276
922 488 214 283 58 922 702 341
0.189
