CHRNA5/CHRNA3 polymorphisms and tobacco smoking risk in a Brazilian population sample

Tobacco smoking is a major risk factor of several diseases such as lung cancer, stroke, chronic obstructive pulmonary disease and increases the susceptibility to infectious diseases. The understanding of smoking addiction requires phenotyping and genotyping studies. Variations in CHRNA5/CHRNA3 can alter receptor responses to nicotine and thus interfere with smoking behavior and risk. Therefore, this study aimed to investigate the association of CHRNA5 rs16969968 and CHRNA3 rs578776 with smoking behavior in a Brazilian population sample, comprising 449 subjects. Smoking data was obtained from a questionnaire. The polymorphisms were genotyped by Polymerase Chain Reaction (PCR). Associations were verified using logistic and linear regression analyses. We found that women with the variant AA genotype for CHRNA5 rs16969968 were at significantly increased risk of smoking, with an OR of 3.09 (95% CI: 1.098.76; p= 0.033). The variant TT genotype of CHRNA3 rs578776 showed protection against smoking and later smoking initiation in the overall population and in women, with an OR of 0.41 (95% CI: 0.19-0.88; p=0.022). In conclusion, CHRNA5 rs16969968 and CHRNA3 rs578776 were associated with increased risk and protective effect against smoking, respectively, in a Brazilian population sample. Gender and recessive homozygosis of the polymorphisms variants resulted in a significant effect regarding the results.


Introduction
Smoking is a worldwide problem Public Health responsible for over 8 million deaths a year. Tobacco use is the leading cause of preventable death and has been associated with an increased risk for several diseases, including cardiovascular diseases, lung disorders, stroke, and several types of cancer [1,2].
Cigarettes are the main source of tobacco consumption, containing over 7000 different chemical substances [2]. Nicotine is the major tobacco substance that establishes and maintains cigarette smoking dependence [3]. Like other psychostimulants, nicotine modulates dopamine activity in the midbrain, specifically the mesocorticolimbic system, which is involved in feeling pleasure when smoking [3,4].
Nicotine is an acetylcholine agonist that binds to nicotinic acetylcholine receptors (nAChRs). These receptors are formed by subunits and are distributed in the central and peripheral nervous system, as well as other tissues such as muscle and the endothelium. The arrangement between subunits results in receptors differentially distributed throughout the body and presenting differential pharmacological properties and responses to nicotine stimulation. Repeated exposure to nicotine can decrease nAchRs responses, leading to their desensitization [5,6]. Consequently, increased nicotine consumption is required to maintain the same response linked to dopamine effects [4]. Some of these receptors are codified by the CHRNA5-CHRNA3-CHRNB4 gene cluster on chromosome 15q25.1, which contains single nucleotide polymorphisms (SNPs) associated with nicotinic dependence [7,8]. These SNPs include CHRNA5 rs16969968 G>A, and CHRNA3 rs578776 C>T, both associated with smoking behavior characteristics, such as number of cigarettes a day (CPD) [9,10,11], age of smoking initiation [12,13], dependence degree as defined by the Fagerström Test for Nicotine Dependence (FTND) [13,14], number of quitting attempts [15,16], smoking cessation [17,18] and heavy smoking [19,20].
Variants in these genes that encode proteins related to nicotine addiction in some mestizo population are associated with cigarette smoking and prevalence of tobacco-related diseases [21]. Apparently, this is the case of Brazil, a country characterized by having a mixed population. However, very few studies have been carried out in the Brazilian population [22,23].
According to Vigitel Brasil 2019, in Brazil, the frequency of adult smokers was 9.8%, being higher in males (12.3%) than in females (7.7%) and a frequency of passive smokers at home was 6.8%, being similar between men and women [24]. Although smoking prevalence has decreased in Brazil due to the actions of the National Tobacco Control Policy, tobacco use is still responsible for 156.216 deaths per year, representing high social and economic costs for the country [24,25]. 74% of deaths caused by chronic obstructive pulmonary disease (COPD), 16% of deaths from heart diseases, 13% from stroke, and 78% from lung cancer are due to the consumption of tobacco. Annual costs with smoking reach up to R$ 56.9 billion in Brazil, of which R$ 39.394.369.233 are spent directly on the treatment of related diseases, and R$ 17.503.786.333 with consequent associated loss of productivity [25]. In this context, this study aimed to investigate potential associations between the CHRNA5 rs16969968 G>A and CHRNA3 rs578776 C>T variants, smoking status risk, and smoking behavior characteristics in a Brazilian population sample.

Study design and population
A total of 449 subjects from Duque de Caxias city in Rio de Janeiro, Brazil, were enrolled in a cross-sectional study with between 2016 and 2019 by a convenience sampling. The study was open to participation by both male and female individuals over 18 years old. The study was approved by the Oswaldo Cruz Foundation Sergio Arouca National School of Public Health (ENSP) ethical committee, number CAAE 40514415.0.0000.5240. Informed consent was obtained from all individual participants included in the study. The participants were interviewed by trained personnel, and a questionnaire was applied to obtain general information, including demographic characteristics.

Smoking status
Smoking status was defined based on available self-report questionnaire data. According to the smoking status, study participants were classified into never, current and former smokers. Current smokers had smoked at least 100 cigarettes during their lifetime, never smokers had smoked less than 100 cigarettes or had never smoked regularly, and former smokers had quit smoking for ≥ 1 year at the time of interview. The FTND was used to assess the nicotine dependence degree in Current smokers [26]. The risk of smoking was evaluated between Never smokers and Ever smokers (Current and Former smokers). Age of initiation and CPD were also evaluated as smoking behavior among Ever smokers.

Genotyping
Approximately 5 mL of blood samples were collected from each study participant, and stored under refrigeration until DNA extraction, carried out by the salting-out method from whole blood. After extraction, the genomic DNA amount was assessed at 260 nm and DNA quality was evaluated through the 260/280 nm ratio using a NanoDrop 2000 spectrophotometer (Thermo Fisher ScientificTM). SNPs were determined using TaqMan single nucleotide polymorphism Genotyping Assays (assay IDs: rs16969968: C_26000428_20; rs578776: C_721253_10), whose amplifications were conducted according to the manufacturer's instructions and performed on a 7500 real-time PCR (Polymerase Chain Reaction) platform (Applied Biosystems). In order to verify genotyping efficiency, random samples were selected and re-genotyped for each SNP (10% of all samples).

Statistical analysis
Statistical analysis was carried out using the SPSS statistical software package 20.0 (Statistical Package for Social Sciences Inc., Chicago, IL, USA). Data normality distribution was assessed by the Kolmogorov-Smirnov test. Continuous variables were presented as means and standard deviation (SD), while categorical data were expressed as frequencies.
The Mann-Whitney U-test and χ2-test were used to evaluated differences between means and frequencies.
Possible associations between CHRNA3 rs578776 and CHRNA5 rs16969968 SNPs with risk of smoking was assessed by applying a logistic regression, considering smoking status as the dependent variable and age, gender, skin color (white/non-white; auto-declared), years of education and current alcohol consumption (yes/no) as covariates in the modeling. Odds ratios (ORs) with 95% confidence intervals were calculated and a linear regression test was used to access the association between polymorphisms and FTND, age of initiation, and CPD (95% CI). Alcohol consumption and gender were used as covariates for FTND and age of initiation, and age was included for CPD. The significance level for all tests was set at p ≤ 0.05. Table 1 presents the sociodemographic characteristics of the study population according to smoking status. From a total of 449 individuals, 33.6% were ever smokers, comprising current (n=55) and former smokers (n=96). Significant differences were detected for gender (p =0.006), age (0.003), years of education (0.002) and alcohol use (0.015). Age ranged from 18 to 87 years old. Most of the study population (74%) auto-declared themselves as non-white, including black, medium brown, Asian, and indigenous.    . Individuals homozygous for the T allele began smoking about two and three years later than allele C homozygous subjects in the overall study population and in women subjects, respectively ( Figure 1). No significant difference was noted for men. Regarding FTND and CPD, the linear regression analysis results indicated no association with CHRNA3 polymorphisms. For the CHRNA5 (rs16969968) SNP, a lower mean age of initiation for the AA genotype compared to GG and AG genotypes was observed, albeit non-statistically significant, as displayed in Table 3. Likewise, FTND score and CPD were not associated to SNPs (data not shown).

Figure 1
Age of smoking initiation among genotypes for CHRNA3 rs578776.
Mean age of smoking initiation is presented at a 95% confidence interval (CI), using CC as reference in the regression analysis. *p value was ≤0.05.

Discussion
Possible associations between CHRNA5 rs16969968 and CHRNA3 rs578776 SNPs and the risk of smoking and smoking behavior in a Brazilian population were evaluated, since these SNPs have been linked to interindividual variability concerning nicotine responses. CHRNA5 rs16969968 and CHRNA3 rs578776, have been investigated as candidates of genetic factors for risk of smoking susceptibility and protection, respectively.
The distribution of variant SNP alleles varies according to population ethnicity. For CHRNA5 rs16969968, the frequencies of allele A are higher in Europeans/Caucasians, at approximately 42%, and lower in Asians and Africans descendants, ranging from 0 to 3% (12,27,28,29). The inverse is noted for CHRNA3 rs578776. The presence of allele T is lower in Caucasians (24%), and higher in African descendants (80%) and Asians (65%) [29]. The Brazilian population is known to be highly mixed, due to the presence of diverse ethnicities in its establishment [30]. In the present study, most of the participants reported themselves as black, medium brown, indigenous or Asians (74%). Herein, variant allele frequencies were 24% and 45% for CHRNA5 rs16969968 and CHRNA3 rs578776, respectively. This distribution appears to be intermediary compared to other populations, and similar to previous studies performed on Brazilian populations [22,23].
The CHRNA5 rs16969968 and CHRNA3 rs578776 SNPs were shown to influence smoking risks, with their recessive genotypes indicating a higher risk of smoking for CHRNA5 rs16969968, and protective effect for CHRNA3 rs578776. An association between the TT genotype of CHRNA3 rs578776 and later initiation age in the overall population was also verified, particularly in women, but not in men. These results are in accordance with previous studies carried out in Caucasian populations [14,28,31,32,33]. No investigations regarding increased risk or protection against smoking in Brazilian population have been conducted for CHRNA5 rs16969968 and CHRNA3 rs578776. In addition, few molecular mechanism propositions for these effects are available in literature. For the CHRNA5 rs16969968 variant allele A, a possible explanation is that an amino acid replacement in the receptor caused by the G>A variation in the gene appears to decrease receptor responses to nicotine, leading to an increased need for consumption [28]. With regard to CHRNA3 SNP, however, no plausible explanation for its protective effect has been postulated.
In the present study, gender was proven to be a relevant factor, which should be taken into account in other investigations on genetic variability associated to smoking. The CHRNA3 rs578776 variant was associated to protection against smoking in the overall population and in women, but not in men. The same was noted for the CHRNA5 rs16969968 variant, only associated with increased risk of smoking in women. Some authors have suggested that women tend to evolve from experimentation to addiction more quickly and have greater difficulty in quitting smoking [34,35]. Additionally, they are more susceptible due to hormonal differences [36]. Several studies providing evidence of gender influence on the genetic-smoking relationship are available [9,11,35]. Interestingly, as association between the assessed SNPs and smoking was only observed for the homozygous variant genotypes, in accordance to other reports for European populations [28,32,37].
FTND, CPD, and initiation were evaluated as smoking behavior characteristics associated to the CHRNA5 rs16969968 and CHRNA3 rs578776 polymorphisms among ever smokers. A significant association was observed for initiation, but not with FTND and CPD, probably due to the reduced sample size of ever smokers. Subjects homozygote for the C variant of CHRNA3 rs578776 began smoking 2 years earlier than those presenting the TT genotype. This indicates that the CHRNA3 rs578776 variant has a protective effect on age of smoking initiation in the overall population, which is stronger in women. Even with no statistically significant difference, subjects with CHRNA5 rs16969968 variant AA genotype started smoking about 2 years earlier than those presenting other genotypes. Although few studies focusing on the interference of genetic variations in smoking initiation of smoking are available, the findings reported herein were in accordance with other assessments [12][13][14]37,38], who reported that people who began smoking earlier exhibited a greater risk of smoking dependence. Weiss et al. only observed an association between CHRNA5 rs16969868 AA and CHRNA3 rs578776 TT, established risk and protection genotypes, respectively, and smoking dependence in adults who began smoking up to 16 years old [14].
The major limitation of this study comprises the inability to assess a possible combined effect of the risk CHRNA5 rs16969968 SNP and protective CHRNA3 rs578776 SNP on smoking, due to an insufficient sample size. This could indicate whether the homozygous variant genotype TT CHRNA3 rs578776 may minimize increased smoking risks associated to the CHRNA5 rs16969968 variant, and may have been the reason for the lack of association between the evaluated SNPs and FTDN and CPD.

Conclusion
In conclusion, the CHRNA5 (rs16969968) variant was only associated with a higher risk of smoking in women, while the CHRNA3 (rs578776) variant was associated with a protective effect regarding risk of smoking and with age of smoking initiation in both the overall study population and in women. For both SNPs, the association was only statistically significant in subjects with variant homozygote genotypes, i.e., the presence of one copy of the allele variant did not affect smoking. Gender influenced the effects of the CHRNA5 rs16969968 and CHRNA3 rs578776 polymorphisms on smoking. This study contributes to elucidate the influence of the CHRNA5 rs16969968 and CHRNA3 rs578776 variants on smoking, especially in the Brazilian population, and may aid in advancing smoking cessation treatments.

Disclosure of conflict of interest
There is no conflict of interest in publishing the present data of the study.

Statement of informed consent
Informed consent was obtained from all individual participants included in the study.