Advertisement

Potential Interactions Between Genetic Polymorphisms of the Transforming Growth Factor-β Pathway and Environmental Factors in Abdominal Aortic Aneurysms

  • Author Footnotes
    c These authors contributed equally to the work presented here and should therefore be considered equivalent authors.
    S. Zuo
    Footnotes
    c These authors contributed equally to the work presented here and should therefore be considered equivalent authors.
    Affiliations
    Department of Vascular and Endovascular Surgery, Chinese PLA General Hospital, Beijing, China
    Search for articles by this author
  • Author Footnotes
    c These authors contributed equally to the work presented here and should therefore be considered equivalent authors.
    J. Xiong
    Footnotes
    c These authors contributed equally to the work presented here and should therefore be considered equivalent authors.
    Affiliations
    Department of Vascular and Endovascular Surgery, Chinese PLA General Hospital, Beijing, China
    Search for articles by this author
  • Y. Wei
    Affiliations
    Department of Epidemiology and Biostatistics, School of Public Health, Peking University Health Science Center, Beijing, China
    Search for articles by this author
  • D. Chen
    Affiliations
    Department of Epidemiology and Biostatistics, School of Public Health, Peking University Health Science Center, Beijing, China
    Search for articles by this author
  • F. Chen
    Affiliations
    Department of Vascular and Endovascular Surgery, Chinese PLA General Hospital, Beijing, China
    Search for articles by this author
  • K. Liu
    Affiliations
    Department of Epidemiology and Biostatistics, School of Public Health, Peking University Health Science Center, Beijing, China
    Search for articles by this author
  • T. Wu
    Affiliations
    Department of Epidemiology and Biostatistics, School of Public Health, Peking University Health Science Center, Beijing, China
    Search for articles by this author
  • Y. Hu
    Correspondence
    Corresponding authors.
    Affiliations
    Department of Epidemiology and Biostatistics, School of Public Health, Peking University Health Science Center, Beijing, China
    Search for articles by this author
  • W. Guo
    Correspondence
    Corresponding authors.
    Affiliations
    Department of Vascular and Endovascular Surgery, Chinese PLA General Hospital, Beijing, China
    Search for articles by this author
  • Author Footnotes
    c These authors contributed equally to the work presented here and should therefore be considered equivalent authors.
Open ArchivePublished:May 28, 2015DOI:https://doi.org/10.1016/j.ejvs.2015.04.010

      Objective/Background

      Evidence has accumulated that multiple polymorphisms in the transforming growth factor (TGF)-β pathway and renin–angiotensin system play important roles in determining susceptibility to abdominal aortic aneurysm (AAA). Few studies have considered interactions between these gene polymorphisms and environmental factors. The aim of this study was to evaluate the contribution of single nucleotide polymorphisms (SNPs) and complex gene–environment interactions in AAA.

      Methods

      Six SNPs located in TGFB, TGFBR1, TGFBR2 and AGTR1 were selected. Genotyping of blood samples and collection of lifestyle factors were performed in 155 unrelated participants with AAAs and 310 non-AAA controls. Unconditional logistic regression was performed to assess the effects of SNPs on the risk of AAA. Generalized multifactor dimensionality reduction (GMDR) was used to evaluate gene–gene and gene–environment interactions.

      Results

      Participants carrying TGFB1 rs1800469 TT (odds ratio [OR] 1.83, 95% confidence interval [CI] 1.18–2.85) or AGTR1 rs12695895 TT (OR 4.21, 95% CI 1.41–12.53) genotypes had a higher risk of AAA than those with the common CC genotype. The gene–gene interaction of AGTR1 rs5182, TGFBR1 rs1626340, and TGFB1 rs1800469 was found to be the best model according to the results of the GMDR analysis (cross validation consistency [CVC]) 10/10; p = .010). Smoking, dyslipidemia, and rs1800469 together contributed to the risk of AAA, which demonstrated a potential and complex gene–environment interaction among the three variants that might affect AAA risk (CVC 6/10; p = .001).

      Conclusion

      In this study of the Chinese population, homozygosity of TGFB1 rs1800469-T and AGTR1 rs12695895-T might be associated with increased risk of AAA. The complex gene–gene and gene–environment interactions might contribute to the risk of AAA. As a small study, the preliminary results need extensive validation and replication in larger populations.

      Keywords

      In the present study, two of six genetic variants located in TGFB1, TGFBR1, TGFBR2 and AGTR1 were identified as potential risk factors for the presence of abdominal aortic aneurysms (AAAs) in a Chinese population. Moreover, gene–gene interaction between three polymorphisms and complex gene–environment interactions with smoking, dyslipidemia and rs1800469 may affect AAA risk.

      Introduction

      Abdominal aortic aneurysm (AAA) is characterized by segmental weakening and dilation of the aortic wall. Several environmental factors, such as age, male sex, and smoking, are implicated in the formation of AAA.
      • Lederle F.A.
      In the clinic. Abdominal aortic aneurysm.
      The disease is also recognized to have a strong genetic component. Five chromosomal loci in the human genome have been identified via genome-wide association studies as genetic risk factors for AAA, including CDKN2BAS rs10757278, CNTN3 rs7635818, DAB2IP rs7025486, LRP1 rs1466565, and LDLR rs6511720.
      • Bradley D.T.
      • Hughes A.E.
      • Badger S.A.
      • Jones G.T.
      • Harrison S.C.
      • Wright B.J.
      • et al.
      A variant in LDLR is associated with abdominal aortic aneurysm.
      • Elmore J.R.
      • Obmann M.A.
      • Kuivaniemi H.
      • Tromp G.
      • Gerhard G.S.
      • Franklin D.P.
      • et al.
      Identification of a genetic variant associated with abdominal aortic aneurysms on chromosome 3p12.3 by genome wide association.
      • Gretarsdottir S.
      • Baas A.F.
      • Thorleifsson G.
      • Holm H.
      • den Heijer M.
      • de Vries J.P.
      • et al.
      Genome-wide association study identifies a sequence variant within the DAB2IP gene conferring susceptibility to abdominal aortic aneurysm.
      • Helgadottir A.
      The same sequence variant on 9p21 associates with myocardial infarction, abdominal aortic aneurysm and intracranial aneurysm.
      • Bown M.J.
      • Jones G.T.
      • Harrison S.C.
      • Wright B.J.
      • Bumpstead S.
      • Baas A.F.
      • et al.
      Abdominal aortic aneurysm is associated with a variant in low-density lipoprotein receptor-related protein 1.
      Moreover, AAA susceptibility genes based on candidate gene analysis have been described as MMP3 5A/6A (rs3025058), ACE I/D (rs4646994), AGTR1 1166A/C (rs5186), MTHFR 677C/T (rs1801133), SORT1(rs599839), and IL6R rs7529229.
      • Helgadottir A.
      The same sequence variant on 9p21 associates with myocardial infarction, abdominal aortic aneurysm and intracranial aneurysm.
      • Morris D.R.
      • Biros E.
      • Cronin O.
      • Kuivaniemi H.
      • Golledge J.
      The association of genetic variants of matrix metalloproteinases with abdominal aortic aneurysm: a systematic review and meta-analysis.
      • Takagi H.
      • Manabe H.
      • Kawai N.
      • Goto S.N.
      • Umemoto T.
      Plasma tissue plasminogen activator and abdominal aortic aneurysm presence: a systematic review and meta-analysis.
      • Jones G.T.
      • Thompson A.R.
      • van Bockxmeer F.M.
      • Hafez H.
      • Cooper J.A.
      • Golledge J.
      • et al.
      Angiotensin II type 1 receptor 1166C polymorphism is associated with abdominal aortic aneurysm in three independent cohorts.
      • Harrison S.C.
      • Smith A.J.
      • Jones G.T.
      • Swerdlow D.I.
      • Rampuri R.
      • Bown M.J.
      • et al.
      Interleukin-6 receptor pathways in abdominal aortic aneurysm.
      • Saratzis A.
      • Bown M.J.
      • Wild B.
      • Nightingale P.
      • Smith J.
      • Johnson C.
      • et al.
      Association between seven single nucleotide polymorphisms involved in inflammation and proteolysis and abdominal aortic aneurysm.
      • Saracini C.
      • Bolli P.
      • Sticchi E.
      • Pratesi G.
      • Pulli R.
      • Sofi F.
      • et al.
      Polymorphisms of genes involved in extracellular matrix remodeling and abdominal aortic aneurysm.
      • Korcz A.
      • Mikolajczyk-Stecyna J.
      • Gabriel M.
      • Zowczak-Drabarczyk M.
      • Pawlaczyk K.
      • Kalafirov M.
      • et al.
      Angiotensin-converting enzyme (ACE, I/D) gene polymorphism and susceptibility to abdominal aortic aneurysm or aortoiliac occlusive disease.
      • Giusti B.
      • Saracini C.
      • Bolli P.
      • Magi A.
      • Sestini I.
      • Sticchi E.
      • et al.
      Genetic analysis of 56 polymorphisms in 17 genes involved in methionine metabolism in patients with abdominal aortic aneurysm.
      • Jones G.T.
      • Bown M.J.
      • Gretarsdottir S.
      • Romaine S.P.
      • Helgadottir A.
      • Yu G.
      • et al.
      A sequence variant associated with sortilin-1 (SORT1) on 1p13.3 is independently associated with abdominal aortic aneurysm.
      However, previously identified genetic loci explain only a portion of the contribution of genetic effect. There are also many factors related to heterogeneity, and it is difficult to identify the mechanism of a disease from an isolated study of the association between a single risk factor and the complex disease. How gene–gene and gene–environment interactions lead to AAA is poorly understood and will likely to play a critical role in the future understanding of this disease.
      • Thornton-Wells T.A.
      • Moore J.H.
      • Haines J.L.
      Genetics, statistics and human disease: analytical retooling for complexity.
      The transforming growth factor (TGF)-β family has been shown to regulate alternative pathways that can cause matrix degradation. Biros et al. performed a meta-analysis based on three case control groups,
      • Biros E.
      • Norman P.E.
      • Jones G.T.
      • van Rij A.M.
      • Yu G.
      • Moxon J.V.
      • et al.
      Meta-analysis of the association between single nucleotide polymorphisms in TGF-beta receptor genes and abdominal aortic aneurysm.
      and they concluded that several genetic variations of TGFBR1 and TGFBR2 were associated with AAA. However, Morris et al. failed to detect the relationship between genetic polymorphisms in the main receptors for TGF-β and AAA.
      • Morris D.R.
      • Biros E.
      • Cronin O.
      • Kuivaniemi H.
      • Golledge J.
      The association of genetic variants of matrix metalloproteinases with abdominal aortic aneurysm: a systematic review and meta-analysis.
      Therefore, the role of TGF-β in the development of AAA in genetic epidemiology is controversial.
      Angiotensin II is the final effector of the renin–angiotensin system (RAS), through the activation of AGTR1. Angiotensin II acts to increase TGF-β mRNA levels, which can alter wall stress due to inflammation and proteolysis.
      • Wolf G.
      • Ziyadeh F.N.
      • Zahner G.
      • Stahl R.A.
      Angiotensin II-stimulated expression of transforming growth factor beta in renal proximal tubular cells: attenuation after stable transfection with the c-mas oncogene.
      However, the interaction between AGTR1 and TGF-β pathway-related genetic polymorphisms in AAA is not clear.
      Herein, a preliminary case–control study was performed to explore whether any genetic variants of TGFB1, TGFBR1, TGFBR2, and AGTR1, and their gene–gene or gene–environment interactions contributed to AAA formation in a Chinese Han population.

      Materials and Methods

      Study site and population

      Details of this case control study design, methods, and participants have been described previously.
      • Wei Y.
      • Xiong J.
      • Zuo S.
      • Chen F.
      • Chen D.
      • Wu T.
      • et al.
      Association of polymorphisms on chromosome 9p21.3 region with increased susceptibility of abdominal aortic aneurysm in a Chinese Han population.
      Patients with AAAs (n = 155) who were diagnosed at the Vascular and Endovascular Surgery Department of the Chinese PLA General Hospital, were enrolled. Patients with AAAs selected for the study were diagnosed with infrarenal AAA by computed tomography (CT) and were ready to undergo elective repair surgery. Half (n = 155) of the control participants were from the Vascular and Endovascular Surgery Department of the Chinese PLA General Hospital, and half (n = 155) were from communities in Beijing. Control participants were sex and age matched (within 5 years) with patients with AAAs and were excluded if they had an AAA by abdominal Doppler ultrasound test. Patients with AAAs and controls were excluded if they had a serious inflammatory process that might influence the predictive value of biochemical testing, had any mental disorders, or were pregnant. The study was conducted with the approval of the ethics committee of Chinese PLA General Hospital. Written informed consent was obtained from each participant.

      Data collection procedures

      Using a standardized questionnaire to collect data related to social demographic information, history of chronic diseases, and lifestyle factors, an interview was conducted with each participant. A history of chronic disease among all participants included hypertension, hyperlipidemia, type 2 diabetes, coronary artery disease (CAD), and peripheral arterial disease (PAD). Smoking, alcohol consumption, and leisure activities were recorded as lifestyle factors. Participants were divided into two groups according to smoking status (those who never smoked and smokers), alcohol use (none to almost none and drinkers), and frequency of leisure activity (none to almost none and at least 1 day/week). Participants underwent a standard physical examination including body temperature, blood pressure, and abdominal ultrasound or CT. The maximum diameter of each AAA was measured from a conventional CT scan after vascular surgeons reviewed the entire scan. AAA was defined as a maximal infrarenal aortic diameter ≥30 mm. To exclude AAA, two trained physicians performed abdominal ultrasound examinations for control group participants. After the entire scan from renal artery level to the bifurcation of aorta had been reviewed, the maximum inner to inner diameter of the abdominal aorta was measured in the transverse view between the renal artery level and 1 cm distal to it.
      • Gurtelschmid M.
      • Bjorck M.
      • Wanhainen A.
      Comparison of three ultrasound methods of measuring the diameter of the abdominal aorta.
      Peripheral venous fasting blood samples were obtained in order to evaluate biochemical parameters and to extract DNA from leukocytes. Biochemical parameters were analyzed at the clinical laboratory of PLA General Hospital, Beijing, China.

      Single nucleotide polymorphism selection and genotyping of polymorphisms

      Genomic DNA was extracted from peripheral blood using salt fractionation and stored at −80 °C. Single nucleotide polymorphisms (SNPs) of AGTR1 and TGF-β pathway related genes were chosen based on their functional relevance and importance. These SNPs were in the promoter regions or had been associated with cardiovascular diseases.
      • Lucarini L.
      • Sticchi E.
      • Sofi F.
      • Pratesi G.
      • Pratesi C.
      • Pulli R.
      • et al.
      ACE and TGFBR1 genes interact in influencing the susceptibility to abdominal aortic aneurysm.
      SNPs from four genes were assessed, including rs1800469 of TGFB1, rs1626340 of TGFBR1, rs3773643 and rs4522809 of TGFBR2, and rs5182 and rs12695895 of AGTR1. A polymerase chain reaction was performed using an Applied Biosystems GeneAmp 9700 384 Dual thermal cycler (Applied Biosystems, Foster City, CA, USA). Genotypes were analyzed using Typer 4.0 software (Mass ARRAY Compact System; Sequenom, San Diego, CA, USA).

      Statistical analysis

      Comparisons of continuous variables (age, education years, and abdominal aortic diameter) between patients and controls were performed using the non-parametric Mann–Whitney test. Hardy–Weinberg equilibrium for each SNP in participants in the control group was tested using a chi-square test. The rare allele was defined as the less frequent allele in unaffected participants, also called the minor allele. Using the chi-square test, each minor allele frequency of patients with AAAs was compared with those of the controls. The dominant, recessive, and additive genetic models were used for analysis. Unconditioned logistic regression of genetic models on AAA was performed and odds ratios (ORs) and 95% confidence intervals (CIs) for AAA risk were estimated for the six SNPs. All statistical analyses were performed using SAS 9.13 software (SAS Institute Inc., Cary, NC, USA). Permutation tests (number of permutations = 10,000) were performed as a measure of significance corrected for multiple testing bias via PLINK software (http://pngu.mgh.harvard.edu/∼purcell/plink/) to calculate the corrected probabilities of genetic association between groups.
      • Purcell S.
      • Neale B.
      • Todd-Brown K.
      • Thomas L.
      • Ferreira M.A.
      • Bender D.
      • et al.
      PLINK: a tool set for whole-genome association and population-based linkage analyses.
      • Tian J.
      • Yu W.
      • Qin X.
      • Fang K.
      • Chen Q.
      • Hou J.
      • et al.
      Association of genetic polymorphisms and age-related macular degeneration in Chinese population.
      Statistical tests were considered to be significant at an alpha level of .05 in a two-tailed test. Statistical power was calculated using the PS: Power and Sample Size Calculation v3.0 (http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize).
      Generalized multifactor dimensionality reduction (GMDR; version 0.7, obtained from http://www.medicine.virginia.edu/clinical/departments/psychiatry/sections/neurobiologicalstudies/genomics/gmdr-software-request) was applied to analyze gene–gene and gene–environment interactions. The GMDR method is described in detail elsewhere.
      • Lou X.Y.
      • Chen G.B.
      • Yan L.
      • Ma J.Z.
      • Zhu J.
      • Elston R.C.
      • et al.
      A generalized combinatorial approach for detecting gene-by-gene and gene-by-environment interactions with application to nicotine dependence.
      Briefly, the simplest model with the maximum cross validation consistency (CVC) and the sign test, which is a non-parametric test with a p-value of .05 or lower, was considered as the best interaction model. A total of 1,000 permutations were used to determine the statistical significance of the best model.

      Results

      Characteristics of participants

      The study included a total of 465 participants. The demographic and clinical characteristics of the participants are given in Table 1. A total of 155 patients (138 men; 89.0%) with AAAs were enrolled and the mean abdominal aortic diameter was 54.2 ± 16.1 mm. The mean age of patients with AAAs was 69.2 ± 9.9 years. Statistically significant differences were found for traditional cardiovascular risk factors, including smoking habits, drinking habits, hypertension, dyslipidemia, and PAD. The presence of type 2 diabetes and CAD were not significantly different between the groups (p > .05).
      Table 1Demographic and clinical characteristics of patients with abdominal aortic aneurysms and control participants.
      CharacteristicCase group (n = 155)Control group (n = 310)Chi-square (F)p
      Age, yr (mean ± SD)69.2 ± 9.969.5 ± 9.90.12.73
      Sex (male)138 (89.0)276 (89.0)0.001.00
      Married (yes)138 (89.0)265 (85.5)0.55.45
      Years of education (mean ± SD)10.1 ± 5.17.0 ± 4.837.81<.001
      Smoking habits (yes)132 (85.2)167 (53.9)39.58<.001
      Drinking habits (yes)79 (51.0)125 (40.3)4.73.03
      Physical activity habits (yes)94 (60.6)195 (62.9)0.10.64
      Hypertension (yes)108 (69.7)143 (46.1)22.39<.001
      Dyslipidemia (yes)118 (76.1)138 (44.5)25.42<.001
      Type-2 diabetes (yes)18 (11.6)46 (14.8)0.91.34
      CAD (yes)53 (34.2)80 (25.8)3.52.06
      PAD (yes)51 (32.9)51 (16.5)15.80<.001
      Abdominal aorta diameter, mm (mean ± SD)54.2 ± 16.117.0 ± 3.217.0<.001
      Note. Data are given as n (%) unless otherwise indicated. CAD = coronary artery disease; PAD = peripheral arterial disease.

      SNP analysis

      The available sample sizes of 155 AAA cases and 310 controls had >60% and 85% power to detect ORs of ∼1.5 and ∼1.8, respectively, for the association of risk alleles with AAA (Supplementary Appendix I). None of the SNPs showed significant deviation from Hardy–Weinberg equilibrium. The genotype and minor allele frequencies of TGFB1 rs1800469; TGFBR1 rs1626340; TGFBR2 rs3773643, rs4522809; and AGTR1 rs5182, rs12695895 and their association with AAA risk are shown in Table 2. After permutation tests, the allele frequencies did not differ between the case and control groups (pcorr > .05). Individuals homozygous for the TGFB1 rs1800469-T had a significantly higher likelihood of AAA than the CC genotype in the recessive model (OR 1.83, 95% CI 1.18–2.85) (Table 3). This suggested that carriers of the rs1800469 TT genotype were more likely to develop an AAA than those with the CC genotype. Homozygosity of AGTR1 rs12695895-T, which was the rare allele, also increased the likelihood of developing an AAA (OR 4.21, 95% CI 1.41–12.53) in the recessive genetic model (Table 3). Other SNP genotypes did not increase AAA risk in any of the three genetic models tested.
      Table 2Gene variants observed in patients with abdominal aortic aneurysms (AAAs) and controls.
      GenedbSNPAAA group (n = 155)Control group (n = 310)p
      Allele distributions were compared between AAA and control groups by chi-square analysis, and p-values reported.
      pcorr
      pcorr is the corrected probabilities of allelic association estimated by permutation test with PLINK (number of permutations = 10,000).
      Genotype frequency (%)MAF
      MAF in case and control groups.
      Genotype frequency (%)MAF
      TGFB1rs1800469TT (31.2)CT (45.5)CC (23.4)T = 54.0TT (19.7)CT (52.9)CC (27.4)T = 46.2.01.16
      TGFBR1rs1626340AA (24.2)GA (51.6)GG (24.2)A = 50.0AA (23.6)GA (45.5)GG (31.0)A = 46.4.29.88
      TGFBR2rs3773643GG (5.9)GA (30.9)AA (63.2)G = 21.4GG (6.5)GA (33.3)AA (60.2)G = 23.2.55.99
      TGFBR2rs4522809CC (8.1)CT (43.9)TT (48.0)C = 30.1CC (7.2)CT (44.3)TT (48.5)C = 29.4.821.00
      AGTR1rs5182CC (9.0)CT (46.5)TT (44.5)C = 32.3CC (7.7)CT (40.0)TT (52.3)C = 27.7.15.65
      AGTR1rs12695895TT (6.5)TC (16.8)CC (76.8)T = 14.9TT (1.6)TC (18.1)CC (80.3)T = 10.7.07.36
      Note. dbSNP = Single Nucleotide Polymorphism database; MAF = minor allele frequency.
      a MAF in case and control groups.
      b Allele distributions were compared between AAA and control groups by chi-square analysis, and p-values reported.
      c pcorr is the corrected probabilities of allelic association estimated by permutation test with PLINK (number of permutations = 10,000).
      Table 3Association between abdominal aortic aneurysms and polymorphisms in dominant, recessive, and additive models.
      GenedbSNPMinor allele
      The minor allele was defined as the less frequent allele in unaffected participants. Groups of major allele homozygous participants were defined as referent groups in genetic models.
      Dominant model
      The dominant model compares individuals with one or two minor alleles with the group of homozygous participants. The recessive model compares individuals with two minor alleles with the combined group of heterozygous and major allele homozygous participants. The additive model assumes that there is a linear gradient in risk between the three genotypes.
      Recessive modelAdditive model
      OR (95% CI)
      Logistic regressions of genotypes were performed and ORs, 95% CIs, and p-values estimated.
      ppcorr
      pcorr is the corrected probabilities of genotype association estimated by permutation test with PLINK (number of permutations = 10,000).
      OR (95% CI)ppcorrOR (95% CI)ppcorr
      TGFB1rs1800469T1.21 (0.77–1.88)0.410.911.83 (1.18–2.85).007.031.35 (1.03–1.78).03.11
      TGFBR1rs1626340A1.33 (0.86–2.06)0.190.521.02 (0.65–1.60).941.001.12 (0.86–1.47).39.84
      TGFBR2rs3773643G0.86 (0.58–1.28)0.460.980.89 (0.40–2.01).791.000.89 (0.65–1.23).481.00
      TGFBR2rs4522809C0.95 (0.65–1.40)0.791.001.10 (0.53–2.28).801.000.98 (0.72–1.34).921.00
      AGTR1rs5182C1.36 (0.93–2.01)0.120.511.18 (0.59–2.36).631.001.25 (0.92–1.68).15.84
      AGTR1rs12695895T1.24 (0.78–1.97)0.380.934.21 (1.41–12.53).01.021.39 (0.95–2.01).09.10
      Note. dbSNP = Single Nucleotide Polymorphism database; OR = odds ratio; CI = confidence interval.
      a The minor allele was defined as the less frequent allele in unaffected participants. Groups of major allele homozygous participants were defined as referent groups in genetic models.
      b The dominant model compares individuals with one or two minor alleles with the group of homozygous participants. The recessive model compares individuals with two minor alleles with the combined group of heterozygous and major allele homozygous participants. The additive model assumes that there is a linear gradient in risk between the three genotypes.
      c Logistic regressions of genotypes were performed and ORs, 95% CIs, and p-values estimated.
      d pcorr is the corrected probabilities of genotype association estimated by permutation test with PLINK (number of permutations = 10,000).

      Gene–gene interactions

      GMDR was performed to explore the gene–gene interactions of six genetic polymorphisms. Moreover, to decrease confounding from other factors, demographic factors, such as age, sex, marital status, and years of education, were adjusted. Table 4 shows the results for two factor to six factor models adjusted by covariates, as mentioned above. The three factor interaction model of AGTR1 rs5182, TGFBR1 rs1626340, and TGFB1 rs1800469 was the best gene–gene interaction model identified. It had a maximum CVC of 10/10, the maximum testing balance accuracy (0.61), and a sign test p-value of .01. These findings suggested the interaction of rs1800469, rs1626340, and rs5182 could contribute to AAA risk.
      Table 4Generalized multifactor dimensionality reduction models of gene–gene interactions and aortic abdominal aneurysm risk.
      Factor numbersBest model
      Adjusted by covariates, including age, sex, marital status, and years of education.
      Training balance accuracyTesting balance accuracyCVCSign test (p-value)Permutation test (p-value)
      2rs1800469, rs16263400.590.488/103 (.95).29
      3rs1800469, rs1626340, rs51820.650.6110/109 (.01).001
      4rs1800469, rs1626340, rs5182, rs45228090.710.599/109 (.01).002
      5rs1800469, rs1626340, rs5182, rs4522809, rs37736430.770.5910/106 (.38).011
      6rs1800469, rs1626340, rs5182, rs4522809, rs3773643, rs126958950.810.5810/108 (.05).031
      Note. CVC = cross-validation consistency.
      a Adjusted by covariates, including age, sex, marital status, and years of education.

      Gene–environment interactions

      GMDR was used to assess the contribution of combinations of environmental factors and genetic polymorphisms to the risk of AAA. Possible environmental risk factors, including smoking, drinking, physical exercise habits, hypertension, dyslipidemia, type 2 diabetes, CAD, PAD, and the six SNPs, were brought into the interaction model. The results of gene–environment interactions on AAA risk for two factor to six factor models adjusted by covariates are presented in Table 5. The potential interaction of smoking and dyslipidemia on AAA risk was discovered (CVC = 10/10; p < .001) and the best gene–environment interaction was the model of smoking, dyslipidemia, and TGFB1 rs1800469 (CVC = 6/10; sign test p-value = .001). This indicated a potential gene–environment interaction among the three variants affecting AAA risk.
      Table 5Generalized multifactor dimensionality reduction models of gene–environment interactions and aortic abdominal aneurysm risk.
      Factor numbersBest model
      Adjusted by covariates including age, sex, marital status, and years of education.
      Training balance accuracyTesting balance accuracyCVCSign test (p-value)Permutatio test (p-value)
      2Smoking, dyslipidemia0.730.7210/1010 (.001)<.001
      3Smoking, dyslipidemia, rs18004690.730.706/1010 (.001)<.001
      4Smoking, dyslipidemia, physical exercise, rs37736430.750.663/109 (.01).001
      5Smoking, rs1800469, rs1626340,rs5182, rs45228090.790.624/1010 (.001)<.001
      6Dyslipidemia, hypertension, rs1800469, rs1626340, rs5182, rs45228090.850.603/108 (.05).031
      Note. CVC = cross-validation consistency.
      a Adjusted by covariates including age, sex, marital status, and years of education.

      Discussion

      AAA is a multifactorial disease influenced by complex genetic and environmental factors. In this study, the association of AGTR1 and TGF-β-related genetic polymorphisms, and their gene–gene and gene–environment interactions in a Chinese Han population, were evaluated.
      Six common variants of TGFB1, TGFBR1, TGFBR2, and AGTR1 were investigated. Allele frequencies of these SNPs were similar to other results obtained in Chinese populations but there were obvious differences compared with other ethnic groups (Supplementary Appendix II). These results demonstrate that, when evaluating a disease related gene, owing to genetic heterogeneity, it is necessary to study the genetic associations in different ethnic groups.
      TGFB1 rs1800469 is in the promoter region and appears to influence the regulation of TGF-β1 plasma levels. rs1800469 has been reported to be associated with cardiovascular diseases.
      • Lu Y.
      • Boer J.M.
      • Barsova R.M.
      • Favorova O.
      • Goel A.
      • Muller M.
      • et al.
      TGFB1 genetic polymorphisms and coronary heart disease risk: a meta-analysis.
      However, genetic associations between rs1800469 and AAA are still confounding. Only Thompson et al. demonstrated that a rare rs1800469 homozygote was associated with the presence of AAA in a UK cohort.
      • Thompson A.R.
      • Cooper J.A.
      • Jones G.T.
      • Drenos F.
      • van Bockxmeer F.M.
      • Biros E.
      • et al.
      Assessment of the association between genetic polymorphisms in transforming growth factor beta, and its binding protein (LTBP), and the presence, and expansion, of abdominal aortic aneurysm.
      However, this association was lost after adjusting for confounders.
      AGTR1 rs12695895 is located in an intronic region. Few studies have evaluated the role of rs12695895 in cardiovascular diseases. Only Nie et al. have reported rs12695895 to be associated with essential hypertension in a Han Chinese population.
      • Nie S.J.
      • Wen-ru T.
      • Bi-feng C.
      • Jin L.
      • Wen Z.
      • Sheng-jun L.
      • et al.
      Haplotype-based case-control study of the human AGTR1 gene and essential hypertension in Han Chinese subjects.
      The present study found that homozygosity of rs12695895-T could increase the likelihood of developing AAA. However, there is a need for further research to better characterize their exact association. AGTR1 rs5182 is located in an exonic region and its variation might influence the normal progress of microRNA transcription, leading to altered AGTR1 expression.
      • Xu M.
      • Sham P.
      • Ye Z.
      • Lindpaintner K.
      • He L.
      A1166C genetic variation of the angiotensin II type I receptor gene and susceptibility to coronary heart disease: collaborative of 53 studies with 20,435 cases and 23,674 controls.
      The present study failed to find an association between rs5182 and AAA. In the participants in the present study, the frequency of rs5182-C was exactly the same as that of the general Chinese population (27.9% vs. 27.7%).
      • Jia E.Z.
      • Chen Z.H.
      • An F.H.
      • Li L.H.
      • Li L.
      • Guo C.Y.
      • et al.
      Relationship of renin-angiotensin-aldosterone system polymorphisms and phenotypes to mortality in Chinese coronary atherosclerosis patients.
      In contrast, rs5182-C was present in 44% of white people.
      • Fatini C.
      • Sticchi E.
      • Sofi F.
      • Said A.A.
      • Pratesi G.
      • Pulli R.
      • et al.
      Multilocus analysis in candidate genes ACE, AGT, and AGTR1 and predisposition to peripheral arterial disease: role of ACE D/-240T haplotype.
      Therefore, one possible explanation for the differences in the results might be ethnic variations.
      TGF-β and RAS play an essential role in matrix deposition and extracellular cell matrix (ECM) degradation. An interaction between TGF-β and RAS related genes had been found that they could influence the development of AAA.
      • Lucarini L.
      • Sticchi E.
      • Sofi F.
      • Pratesi G.
      • Pratesi C.
      • Pulli R.
      • et al.
      ACE and TGFBR1 genes interact in influencing the susceptibility to abdominal aortic aneurysm.
      In the present study, the effect of the interaction of rs5182, rs1626340, and rs1800469 on AAA were also estimated, suggesting close and complicated correlations exist between the TGF-β family and AGTR1. Angiotensin II could increase TGF-β mRNA expression and regulate TGF-β serum levels. In addition, polymorphisms in TGFB1 could correspond to reduced levels of TGF-β and angiotensin II.
      • Jimenez-Sousa M.A.
      • Fernandez-Rodriguez A.
      • Heredia M.
      • Tamayo E.
      • Guzman-Fulgencio M.
      • Lajo C.
      • et al.
      Genetic polymorphisms located in TGFB1, AGTR1, and VEGFA genes are associated to chronic renal allograft dysfunction.
      Therefore, the interaction between TGF-β and the RAS pathway is close and complicated. The correlation between genes and biological mechanisms of AAA require further study.
      The GMDR analysis demonstrated that TGFB1 rs1800469, smoking, and dyslipidemia potentially worked together in affecting AAA risk. Smoking and dyslipidemia are strong risk factors for AAA and are thought to contribute to the initiation and progression of vascular fibrosis. TGF-β plays a critical role in ECM accumulation and vascular remodeling by upregulating fibroblast growth factor expression and participating in the fibrotic process.
      • Lan T.H.
      • Huang X.Q.
      • Tan H.M.
      Vascular fibrosis in atherosclerosis.
      These findings suggested that smoking, dyslipidemia, and the TGF-β pathway might influence AAA formation via the vascular fibrotic process, but the specific process is unknown.
      There were several limitations to this study that warrant consideration. First, the sample size was small. It was found that the available sample sizes of 155 AAA cases and 310 controls had >60% and 85% statistical power to detect ORs of ∼1.5 and ∼1.8, respectively, for the association of the risk alleles with AAA. Therefore, the results should be considered preliminary and requiring verification. Second, the number of SNPs genotyped was limited. Only six polymorphisms were chosen and one gene–environment interaction affecting AAA risk discovered. However, several factors that influenced the presence of AAA, such as type 2 diabetes, hypertension, and atherosclerotic diseases, were not discovered to have potential interactions with AAA risk, warranting a more comprehensive study of complex interactions in AAA.

      Conclusion

      This study is the first to report that TGFB1 rs1800469 and AGTR1 rs12695895 genetic polymorphisms might be associated with AAA risk in a Chinese Han population. Furthermore, complex interactions between environmental factors and polymorphisms might contribute to the risk of AAA. Owing to small sample size, the results should be considered preliminary and requiring extensive validation and replication in larger populations.

      Acknowledgments

      We thank all participants for their enrollment in this study.

      Conflict of Interest

      None.

      Funding

      The study was supported by the National Natural Science Foundation of China (81230066) and The Eleventh Five-year Plan in Health Care Foundation of PLA (09BJZ04).

      Appendix A. Supplementary data

      The following is the supplementary data related to this article:

      References

        • Lederle F.A.
        In the clinic. Abdominal aortic aneurysm.
        Ann Intern Med. 2009; 150 (ITC5-1-15)
        • Bradley D.T.
        • Hughes A.E.
        • Badger S.A.
        • Jones G.T.
        • Harrison S.C.
        • Wright B.J.
        • et al.
        A variant in LDLR is associated with abdominal aortic aneurysm.
        Circ Cardiovasc Genet. 2013; 6: 498-504
        • Elmore J.R.
        • Obmann M.A.
        • Kuivaniemi H.
        • Tromp G.
        • Gerhard G.S.
        • Franklin D.P.
        • et al.
        Identification of a genetic variant associated with abdominal aortic aneurysms on chromosome 3p12.3 by genome wide association.
        J Vasc Surg. 2009; 49: 1525-1531
        • Gretarsdottir S.
        • Baas A.F.
        • Thorleifsson G.
        • Holm H.
        • den Heijer M.
        • de Vries J.P.
        • et al.
        Genome-wide association study identifies a sequence variant within the DAB2IP gene conferring susceptibility to abdominal aortic aneurysm.
        Nat Genet. 2010; 42: 692-697
        • Helgadottir A.
        The same sequence variant on 9p21 associates with myocardial infarction, abdominal aortic aneurysm and intracranial aneurysm.
        Nat Genet. 2008; 10: 217-224
        • Bown M.J.
        • Jones G.T.
        • Harrison S.C.
        • Wright B.J.
        • Bumpstead S.
        • Baas A.F.
        • et al.
        Abdominal aortic aneurysm is associated with a variant in low-density lipoprotein receptor-related protein 1.
        Am J Hum Genet. 2011; 89: 619-627
        • Morris D.R.
        • Biros E.
        • Cronin O.
        • Kuivaniemi H.
        • Golledge J.
        The association of genetic variants of matrix metalloproteinases with abdominal aortic aneurysm: a systematic review and meta-analysis.
        Heart. 2014; 100: 295-302
        • Takagi H.
        • Manabe H.
        • Kawai N.
        • Goto S.N.
        • Umemoto T.
        Plasma tissue plasminogen activator and abdominal aortic aneurysm presence: a systematic review and meta-analysis.
        Ann Vasc Surg. 2010; 24: 686-689
        • Jones G.T.
        • Thompson A.R.
        • van Bockxmeer F.M.
        • Hafez H.
        • Cooper J.A.
        • Golledge J.
        • et al.
        Angiotensin II type 1 receptor 1166C polymorphism is associated with abdominal aortic aneurysm in three independent cohorts.
        Arterioscler Thromb Vasc Biol. 2008; 28: 764-770
        • Harrison S.C.
        • Smith A.J.
        • Jones G.T.
        • Swerdlow D.I.
        • Rampuri R.
        • Bown M.J.
        • et al.
        Interleukin-6 receptor pathways in abdominal aortic aneurysm.
        Eur Heart J. 2013; 34: 3707-3716
        • Saratzis A.
        • Bown M.J.
        • Wild B.
        • Nightingale P.
        • Smith J.
        • Johnson C.
        • et al.
        Association between seven single nucleotide polymorphisms involved in inflammation and proteolysis and abdominal aortic aneurysm.
        J Vasc Surg. 2014 Mar 7; ([Epub ahead of print])
        • Saracini C.
        • Bolli P.
        • Sticchi E.
        • Pratesi G.
        • Pulli R.
        • Sofi F.
        • et al.
        Polymorphisms of genes involved in extracellular matrix remodeling and abdominal aortic aneurysm.
        J Vasc Surg. 2012; 55 (171–179 e2)
        • Korcz A.
        • Mikolajczyk-Stecyna J.
        • Gabriel M.
        • Zowczak-Drabarczyk M.
        • Pawlaczyk K.
        • Kalafirov M.
        • et al.
        Angiotensin-converting enzyme (ACE, I/D) gene polymorphism and susceptibility to abdominal aortic aneurysm or aortoiliac occlusive disease.
        J Surg Res. 2009; 153: 76-82
        • Giusti B.
        • Saracini C.
        • Bolli P.
        • Magi A.
        • Sestini I.
        • Sticchi E.
        • et al.
        Genetic analysis of 56 polymorphisms in 17 genes involved in methionine metabolism in patients with abdominal aortic aneurysm.
        J Med Genet. 2008; 45: 721-730
        • Jones G.T.
        • Bown M.J.
        • Gretarsdottir S.
        • Romaine S.P.
        • Helgadottir A.
        • Yu G.
        • et al.
        A sequence variant associated with sortilin-1 (SORT1) on 1p13.3 is independently associated with abdominal aortic aneurysm.
        Hum Mol Genet. 2013; 22: 2941-2947
        • Thornton-Wells T.A.
        • Moore J.H.
        • Haines J.L.
        Genetics, statistics and human disease: analytical retooling for complexity.
        Trends Genet. 2004; 20: 640-647
        • Biros E.
        • Norman P.E.
        • Jones G.T.
        • van Rij A.M.
        • Yu G.
        • Moxon J.V.
        • et al.
        Meta-analysis of the association between single nucleotide polymorphisms in TGF-beta receptor genes and abdominal aortic aneurysm.
        Atherosclerosis. 2011; 219: 218-223
        • Wolf G.
        • Ziyadeh F.N.
        • Zahner G.
        • Stahl R.A.
        Angiotensin II-stimulated expression of transforming growth factor beta in renal proximal tubular cells: attenuation after stable transfection with the c-mas oncogene.
        Kidney Int. 1995; 48: 1818-1827
        • Wei Y.
        • Xiong J.
        • Zuo S.
        • Chen F.
        • Chen D.
        • Wu T.
        • et al.
        Association of polymorphisms on chromosome 9p21.3 region with increased susceptibility of abdominal aortic aneurysm in a Chinese Han population.
        J Vasc Surg. 2013; 59: 879-885
        • Gurtelschmid M.
        • Bjorck M.
        • Wanhainen A.
        Comparison of three ultrasound methods of measuring the diameter of the abdominal aorta.
        Br J Surg. 2014; 101: 633-636
        • Lucarini L.
        • Sticchi E.
        • Sofi F.
        • Pratesi G.
        • Pratesi C.
        • Pulli R.
        • et al.
        ACE and TGFBR1 genes interact in influencing the susceptibility to abdominal aortic aneurysm.
        Atherosclerosis. 2009; 202: 205-210
        • Purcell S.
        • Neale B.
        • Todd-Brown K.
        • Thomas L.
        • Ferreira M.A.
        • Bender D.
        • et al.
        PLINK: a tool set for whole-genome association and population-based linkage analyses.
        Am J Hum Genet. 2007; 81: 559-575
        • Tian J.
        • Yu W.
        • Qin X.
        • Fang K.
        • Chen Q.
        • Hou J.
        • et al.
        Association of genetic polymorphisms and age-related macular degeneration in Chinese population.
        Invest Ophthalmol Vis Sci. 2012; 53: 4262-4269
        • Lou X.Y.
        • Chen G.B.
        • Yan L.
        • Ma J.Z.
        • Zhu J.
        • Elston R.C.
        • et al.
        A generalized combinatorial approach for detecting gene-by-gene and gene-by-environment interactions with application to nicotine dependence.
        Am J Hum Genet. 2007; 80: 1125-1137
        • Lu Y.
        • Boer J.M.
        • Barsova R.M.
        • Favorova O.
        • Goel A.
        • Muller M.
        • et al.
        TGFB1 genetic polymorphisms and coronary heart disease risk: a meta-analysis.
        BMC Med Genet. 2012; 13: 39
        • Thompson A.R.
        • Cooper J.A.
        • Jones G.T.
        • Drenos F.
        • van Bockxmeer F.M.
        • Biros E.
        • et al.
        Assessment of the association between genetic polymorphisms in transforming growth factor beta, and its binding protein (LTBP), and the presence, and expansion, of abdominal aortic aneurysm.
        Atherosclerosis. 2010; 209: 367-373
        • Nie S.J.
        • Wen-ru T.
        • Bi-feng C.
        • Jin L.
        • Wen Z.
        • Sheng-jun L.
        • et al.
        Haplotype-based case-control study of the human AGTR1 gene and essential hypertension in Han Chinese subjects.
        Clin Biochem. 2010; 43: 253-258
        • Xu M.
        • Sham P.
        • Ye Z.
        • Lindpaintner K.
        • He L.
        A1166C genetic variation of the angiotensin II type I receptor gene and susceptibility to coronary heart disease: collaborative of 53 studies with 20,435 cases and 23,674 controls.
        Atherosclerosis. 2010; 213: 191-199
        • Jia E.Z.
        • Chen Z.H.
        • An F.H.
        • Li L.H.
        • Li L.
        • Guo C.Y.
        • et al.
        Relationship of renin-angiotensin-aldosterone system polymorphisms and phenotypes to mortality in Chinese coronary atherosclerosis patients.
        Sci Rep. 2014; 4: 4600
        • Fatini C.
        • Sticchi E.
        • Sofi F.
        • Said A.A.
        • Pratesi G.
        • Pulli R.
        • et al.
        Multilocus analysis in candidate genes ACE, AGT, and AGTR1 and predisposition to peripheral arterial disease: role of ACE D/-240T haplotype.
        J Vasc Surg. 2009; 50: 1399-1404
        • Jimenez-Sousa M.A.
        • Fernandez-Rodriguez A.
        • Heredia M.
        • Tamayo E.
        • Guzman-Fulgencio M.
        • Lajo C.
        • et al.
        Genetic polymorphisms located in TGFB1, AGTR1, and VEGFA genes are associated to chronic renal allograft dysfunction.
        Cytokine. 2012; 58: 321-326
        • Lan T.H.
        • Huang X.Q.
        • Tan H.M.
        Vascular fibrosis in atherosclerosis.
        Cardiovasc Pathol. 2013; 22: 401-407

      Comments

      Commenting Guidelines

      To submit a comment for a journal article, please use the space above and note the following:

      • We will review submitted comments as soon as possible, striving for within two business days.
      • This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
      • We require that commenters identify themselves with names and affiliations.
      • Comments must be in compliance with our Terms & Conditions.
      • Comments are not peer-reviewed.