Influence of single nucleotide polymorphisms among cigarette smoking and non-smoking patients with coronary artery disease, urinary bladder cancer and lung cancer

Introduction Cigarette smoke is suggested to be a risk factor for coronary artery disease (CAD), urinary bladder cancer (UBCa) or lung cancer (LCa). However, not all heavy smokers develop these diseases and elevated cancer risk among first-degree relatives suggests an important role of genetic factor. Methods Three hundred and ten healthy blood donors (controls), 98 CAD, 74 UBCa and 38 LCa patients were included in this pilot study. The influence of 92 single nucleotide polymorphisms (SNPs) and impact of cigarette smoking were analysed. Results Out of 92 SNPs tested, differences in distribution of 14 SNPs were detected between controls and patient groups. Only CTLA4 rs3087243 showed difference in both CAD and UBCa patient group compared to control group. Stratified by smoking status, the impact of smoking was associated to frequencies of 8, 3 and 4 SNPs in CAD, UBCa, LCa patients, respectively. None of these 92 SNPs showed a statistically significant difference to more than one type of disease among smoking patients. In non-smoking patients, 7, 3 and 6 SNPs were associated to CAD, UBCa, LCa, respectively. Out of these 92 SNPs, CTLA4 rs3087243 was associated to both non-smoking CAD and UBCa. The XRCC1 rs25487 was associated to both non-smoking UBCa and LCa. Conclusion SNPs might be important risk factors for CAD, UBCa and LCa. Distribution of the SNPs was specific for each patient group, not a random event. Impact of cigarette smoking on the disease was associated to the specific SNP sequences. Thus, smoking individuals with SNPs associated to risk of these serious diseases is an important target group for smoking cessation programs.

Introduction Thirty ml peripheral blood was collected from the controls and patients using EDTA containing tubes. All blood samples were stored at -80˚C in the biobank, at Ryhov hospital, Sweden until analysed.

Ethics statement
This pilot study was conducted in accordance with the Declaration of Helsinki and Regional Research Ethics Review Board of Linköping, Sweden approved our study (Dnr 2011/271-31 and 2015/178-32). Written informed consent was obtained from all participants.

SNP analysis
High molecular weight DNA was extracted from blood samples using QIAGEN Bio Robot M48 with MagAttract DNA Blood M48 kits EZ1 (www.qiagen.com). The quantity and quality of DNA was determined by NanoDrop. According to genetics home reference-NIH (https://ghr.nlm.nih.gov/gene) and previous investigations, 92 SNPs located in genes associated to cell cycle controls, cell death, immune response, DNA repair, inflammation, micro-RNAs, oncogenes and tumour suppressor genes ( Table 2) were selected for this study [4, 12-14, 17, 18].
All 92 SNP sequences were tested and passed two-hits in the dbSNP database and were Hap Map-validated with Illumina design ability score [19]. Genotyping of SNP was done at the SNP & SEQ Technology Platform, Uppsala University, Sweden (www.genotyping.se). Highthroughput genotyping with the Illumina Golden gate assay was used according to the manufacturer's protocol (www.illumina.com).

Statistical analysis
Results are presented as numbers, means and standard deviations. Fishers exact test were used to evaluate possible differences in the frequency distribution of SNPs between the patient and control group. To examine the impact of cigarette smoking and SNPs, comparisons between SNP frequencies among the cigarette smoking (smoking) patients in each group to their corresponding smoking controls were analyzed. For the impact of unknown co-factors, SNP frequencies in patients who never had used any type of tobacco products (non-smoking) in each patient group were compared to non-smoking controls.
In the univariable logistic regression, the results were presented as Wald Chi-square test pvalues, odds ratios (OR) and corresponding 95% confidence intervals (95% CI). The most frequent SNP among controls were used as reference level for OR in regression models. Stratified by disease and smoking status, no other interaction effects were analyzed. Statistical analyses were done using SAS 9.4 software (SAS Institute, Cary, NC). All comparisons were two-sided and a Fishers exact test p value �0.05 was considered statistically significant.

Controls and patients
A total of 512 participants were enrolled from one community based population in Jönköping region, Sweden during 2016 to 2019 (Table 1). There were 302 controls (155 males and 147 females). The median age of females was 55 years (range 20-89) and of males was 56 years (range 19-79). Among controls, 150 had a history of cigarette smoking (smoking) and 160 had never use any type of tobacco products (non-smoking).
A total 210 patients (154 males and 56 females) were included. The median age of females was 69 years (range 52-84) and of males was 71 years (range 39-91). The 98 CAD patients (83 males and 15 females) were 53 smoking and 45 non-smoking. The 74 UBCa patients (59 males

SNP distribution among CAD, UBCa and LCa patient groups and control group
Out of the 92 SNPs tested (Table 2), differences in distribution of 14 SNPs were detected between controls and the CAD, UBCa or LCa patients (Table 3). Differences between controls and CAD patients were detected in six SNPs located in ATM, CTLA4, BRCA1, CCND3, HRas and IL2. Differences between controls and UBCa patients were detected in six SNPs located in ABCC5, CDH13, CRP, CTLA4, p2 and TNFSF1 Differences between controls and LCa patients were detected in three SNPs located in DNMT3B, MTHFR, and Serpin1.
Of the 92 SNPs (Table 2), only CTLA4 rs3087243 showed a statistically significant difference in two groups of the patients, CAD and UBCa, compared to the controls (Table 3).

SNP distribution among smoking CAD, UBCa or LCa patients and smoking controls
Out of these 92 SNPs, differences in the distribution of 15 SNPs were detected between smoking controls and smoking patients (Table 4). Differences in seven SNPs located in CCND3, CTLA4, KDM4C, PFA1, PPAP2B, PRF1 and Rad52, were found to be specific for smoking CAD patients. Differences in three SNPs located in ABCA1, CCND1 and MiR206 were specific for smoking UBCa patients. Differences in four SNPs located in CDH13, HTR3B1, CRP and ZMF830 were specific for smoking LCa patients.
None of these 92 SNPs showed a statistically significant difference in distribution in more than one disease among smoking patients compared to their corresponding smoking controls (Table 4).

SNPs distribution among non-smoking CAD, UBCa, LCa patients and non-smoking controls
Out of these 92 SNPs tested, differences in distribution of 14 SNPs were detected between the non-smoking controls and non-smoking patients ( Table 5). They were seven SNPs located in ATM, BRCA1, CTLA4, CYP19A1, FGFR4, MiR34A and PRKDC in non-smoking CAD patients. Three SNPs located in MTHFR, CTLA4, and XRCC1, in non-smoking UBCa patients and four SNPs located in HTRB1, Lig4, Serpin1, TGFb, and XRCC1 in non-smoking LCa patients ( Table 5).
Two of these 92 SNPs, CTLA4 rs3087243 and XRCC1 rs25487 showed different distribution in more than one disease among non-smoking patient groups compared to their non-smoking controls ( Table 5). The SNP in CTLA4 rs3087243 showed a difference in distribution among non-smoking controls and non-smoking CAD or UBCa patients. The SNP in XRCC1 rs25487 showed a difference in distribution among non-smoking controls and non-smoking UBCa or LCa patients.

Discussion
The genetic and the environment influence human risk of various diseases and clinical outcome. We found that specific genetic variations influenced the risk of diseases that are suggested to associate to cigarette smoking such as CVD, UBCa and LCa. These diseases are heterogeneous groups regarding site location and their pathological parameters.

PLOS ONE
Impact of SNPs and cigarette smoking on coronary artery disease, urinary bladder cancer and lung cancer

PLOS ONE
Impact of SNPs and cigarette smoking on coronary artery disease, urinary bladder cancer and lung cancer  Endothelial malignancies are very rare in the western population and consists mainly of sarcomas. The endothelial cells seems not to be transformed into malignancies from smoking [20]. However, the blockage of the CAD patient artery caused by smoking is a complex mechanism involving inflammation and benign smooth muscle proliferation [21,22].
Out of these 92 SNPs, only distribution of CTLA4 rs3087243 differed between controls and two patient groups, CAD and UBCa. CTLA4 located on chromosome 2q33.2, encodes for CTLA4 protein that functions as an immune checkpoint and downregulates immune responses [23,24]. CTLA4 is constitutively expressed in regulatory T cells. Upon activation, conventional T cells in cancer patients could up-regulated CTLA4 [25]. Alteration in CLTA4 expression cells were detected in the inflammatory heart disease [26]. CTLA4 polymorphism and risk of cancer were reported [27,28]. The possible benefits of immunotherapy with focus on CTLA4 protein as target therapy of CAD and UBCa need further investigation [25,29,30].
The impact of cigarette smoking on CAD risk was more pronounced among individuals with specific SNPs in genes involving the immune response (CTLA4, PFA and PRF), cell cycle control (CCND3 and p21), DNA repair (BRCA1, ATM) and oncogene (HRas). The impact of cigarette smoking on UBCa risk seem to be accumulated among individuals with SNPs in genes involved in cell cycle control (ABCC1, CDH13 and p21,) and immune response (CRP, CTLA4 and TNFSF1). The impact of cigarette smoking on LCa risk might be increased among individuals with SNPs in genes involving cell cycle control (CDH13 and HTR3B1) and immune response (CRP). The variation in SNPs among cigarette smoking CAD, UBCa and LCa patients indicates an impact of general toxic agents from cigarette smoke and specific DNA sequences, not random event on the risk for these diseases.
There are subgroups of the CAD, UBCa and LCa patients that never have used any type of tobacco products, non-smoking patient group. Despite similar type of disease, SNPs that associated to the non-smoking UBCa patients differed from those of the smoking UBCa patients. Unknown risk factors, apart from cigarette smoking, in combination with the specific SNPs could also increase the risk of UBCa in non-smoking individuals.
SNP sequence variation in CTLA4 rs3087243 was associated to risk of CAD in both smoking and non-smoking patients. Thus, the influence of immune response on smooth muscle cell proliferation of CAD patients [26] was independent from cigarette smoking. SNPs in the CTLA4 rs3087243 also associate to risk of non-smoking UBCa patients, not in smoking UBCa patients. This suggests an influence of CTLA4 rs3087243 on risk of CAD and UBCa might be independent from the cigarette smoke.
The differences in XRCC1 rs25487 of non-smoking UBCa and non-smoking LCa patients compared to non-smoking controls, confirms that DNA repair play an important role on UBCa and LCa [34,35]. If patients with AA or AG sequences in XRCC1 rs25487 are prone to DNA damage from other agents/microenvironments more than cigarette smoke needs further investigation [37][38][39].
Our pilot study has several limitations. Firstly, this is a single center, non-randomized retrospective study with a relatively low number of included patients [40]. In addition, the lower age among controls compared to patients was based on the blood bank regulation, needs to be considered. Secondly, cigarette smoking history of the patients and controls was self-reported and this could be influenced by recall bias [41]. Thirdly, passive smoking conditions that might have impact on risk of these diseases [8,42], were not recorded.
In summary, our results indicate an important role of specific genetic variations on risk of CAD, UBCa or LCa. Impact of cigarette smoking was also found in a proportion of these patients in association with individual specific SNPs. The SNPs are lifelong stable genetic variation that could predict the risk to develop specific diseases.
Unable to quit smoking during or after treatment could increase rate of diseases recurrence, progression, development of a second primary tumour and disease-specific mortality. The identification of SNPs that associated to risk or disease progression could increase the cessation rate in smoking patients.
Healthy individual with SNPs associated to a smoking related risk to develop CAD, UBCa or LCa will also get higher motivation for smoking cessation. Unknown environmental factors associated to risk of these diseases in non-smoking group [43] and the possibility to use SNPs as prognostic biomarkers for treatment selection and prediction of clinical outcome needs future investigation.