SEARCH

SEARCH BY CITATION

Keywords:

  • genetic polymorphism;
  • CHRNA;
  • lung cancer;
  • DNA adducts;
  • TP53

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Genome-wide association studies have demonstrated that genetic polymorphisms influence the risk of developing lung cancer. Nicotinic acetylcholine receptor alpha3, alpha5 and beta4 genes (CHRNA3, CHRNA5 and CHRNB4) cluster at the 15q25.1 lung cancer susceptibility locus. We genotyped 310 patients with non-small cell lung cancer and a control group of 348 cancer-free individuals for seven sequence variants located in CHRNA3 and CHRNA5 genes. Two of the polymorphisms (rs3829787 and rs3841324) statistically influenced the risk of developing lung cancer. We found that four of the variants (rs3829787, rs3841324, rs588765 and rs3743073) were associated with differential levels of genetic alterations measured as the levels of hydrophobic DNA adducts in the adjacent histologically normal tissue of the lung cancer patients and as TP53 mutations in their lung tumors. The seven sequence variants formed three haplotypes with a frequency above 5%. The two most frequent haplotypes were associated with the risk of developing lung cancer and with smoking-related DNA alterations. We also found an association between CHRNA5 mRNA levels and the sequence variants or haplotypes. In conclusion, our results showed that several of the polymorphisms and their haplotypes in CHRNA5/CHRNA3 genes may have functional effects on (i) CHRNA5 mRNA levels, (ii) polycyclic aromatic hydrocarbon–DNA adduct levels, (iii) TP53 mutations and (iv) susceptibility to lung cancer.

Genome-wide association studies have demonstrated that genetic polymorphisms at human chromosome 15q25.1 influence the risk of developing lung cancer.1 Three nicotinic acetylcholine receptor (CHRN) genes encoding nicotinic acetylcholine receptors (nAChRα3, nAChRα5 and nAChRβ4) are clustered at the 15q25 lung cancer susceptibility locus.2, 3 The strong linkage disequilibrium (LD) between the single nucleotide polymorphisms (SNPs) in this region has led to the identification of several SNPs associated with lung cancer risk1–3 or with CHRNA5 mRNA levels.4–6

The nAChRs are prototypic ligand-gated ion channels that are activated by endogenous agonists (acetylcholine and choline) or by xenobiotics like nicotine. The nAChRs are expressed in normal lung tissue, they mediate sensitivity to nicotine and nicotine exposure influences their expression.5, 7 Thus, it is not surprising that genetic polymorphisms in CHRN genes located at the 15q25.1 locus have been associated with lung cancer risk indirectly via tobacco/nicotine addiction or smoking behavior.2 However, the same sequence variants in the CHRN genes have also been reported to be associated with lung cancer risk in never smokers.8 nAChR activity or their subunit composition may regulate critical cellular processes involved in lung carcinogenesis.9 Recent studies have conferred functions to nAChRs in lung cancer independently of smoking quantity or behavior.10–12 Differential expression of CHRNA3 or CHRNA5 genes in human lung cells may regulate the balance between cell survival and apoptosis,11 cell motility and migration12 or wound repair of the respiratory epithelium.10 Also, mRNA levels of CHRNA3 and CHRNA5 in tumor cells are altered compared to normal cells.5, 13 An interesting hypothesis on the link between polymorphisms in CHRN genes, their expression and functions in lung cancer development has been recently proposed.14 According to this hypothesis, some of the mechanisms by which the polymorphisms in CHRN genes may predispose to lung cancer could be that sequence variants may lead to differential expression of CHRN genes which that may affect the receptors activity and the wound repair of smoking-related injuries in the lung epithelium. However, there is lack of functional studies assessing the association between polymorphisms in CHRN genes and smoking-related DNA alterations.

In our study, we have investigated seven polymorphisms in the CHRNA genes and assessed if they were related with CHRNA mRNA levels or carcinogenic biomarkers such as bulky polycyclic aromatic hydrocarbon (PAH)–DNA adducts and mutations in TP53 in non-small cell lung cancer (NSCLC). We report results on the effects of these polymorphisms and their haplotypes on (i) CHRNA mRNA and hydrophobic PAH–DNA adduct levels in normal lung tissues; TP53 mutations in lung tumors and (ii) risk of NSCLC in a case–control study from Norway.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Cases and controls

The characteristics of lung cancer patients and controls included in our study are summarized in Table 1. Early-stage NSCLC cases (n = 310) and controls (n = 348) were Caucasians of Norwegian origin and were recruited from the same geographical region (West Norway). The cases were recruited consecutively among patients admitted for lung cancer at the Haukeland University Hospital in Bergen. Only patients with histologically confirmed early-stage NSCLC were included in our study. Controls were recruited from the same hospital and frequency-matched with cases on cumulative smoking dose (pack-years). Pack-years smoked [(cigarettes per day/20) × years smoked] were calculated to indicate the cumulative smoking dose. The reason for matching cases and controls on pack-years was that the total amount smoked in a life time greatly influences the risk of developing lung cancer rather than the variable parameter “cigarettes/day.” For mRNA and PAH–DNA adduct studies, samples of adjacent histologically normal lung tissues were collected as far as possible from the tumor (>5 cm) at the time of surgery for lung cancer at Haukeland University Hospital in Bergen between 1986 and 2010. For tumor tissues, tumor histology was reanalyzed by an experienced pathologist, and only tumor tissues containing at least 90% of tumor cells were included in our study. Tumor and normal tissues were snap-frozen after resection in liquid nitrogen and kept at −80°C until further processing. Controls were cancer-free individuals at the time of sampling.15 Cases and controls were interviewed using similar questionnaires and were categorized as never smokers, ex-smokers or current smokers. Never smokers are patients indicating having smoked less than 100 cigarettes in their life time. Ex-smokers were defined as patients having quitted at least 1 year before sampling, and current smokers were those indicating that they were smokers at the time of sampling. All subjects gave written consent, and the study was approved by the regional ethical committee in accordance with the Helsinki Declaration.

Table 1. Characteristics of NSCLC cases and controls
inline image

DNA extraction and genotyping

Genotyping was performed on DNA extracted from frozen normal lung tissue samples for cases and on DNA extracted from blood for controls. DNA was extracted using either DNA isolation kits (Qiagen, Hilden, Germany) or standard proteinase K digestion followed by phenol–chloroform extraction and ethanol precipitation. Seven polymorphisms (rs3829787, rs3841224, rs588765, rs578776, rs6493508, rs3743073 and rs8023462) located at the 15q25.1 locus (Fig. 1a) were genotyped using the core genotyping facilities at the Center for Integrative Genetics (CIGENE) at the Norwegian University of Life Sciences using a Sequenom iPLEX™ MassARRAY® system as instructed by the supplier. The difference in the source of DNA (normal lung tissue or blood) in cases and controls is less likely to cause a bias in genotyping. In our previous studies, we have not experienced genotype differences between the DNA extracted from blood or tissue from same individual.

thumbnail image

Figure 1. Association between genetic polymorphisms or haplotypes with hydrophobic DNA adducts or CHRNA5 mRNA levels. (a) Location of the polymorphisms studied is indicated by arrows, their rs number is given as well as the sequence variations, the major alleles are in bold. The gray boxes represent the introns in the genes. Hydrophobic DNA adduct and polymorphisms (b) or haplotypes (e). CHRNA5 mRNA levels measured in adjacent histological normal tissues and polymorphisms (c) or haplotypes (f). The numbers in boxes and in brackets indicate the number of tissues tested. *p value of Mann–Whitney < 0.05 between median of hydrophobic DNA adduct level (or mRNA levels) measured in adjacent histological normal tissues of homozygous carriers of the common variant versus median measured in tissues from homozygous carriers of the rare variant or heterozygous patients. (d) Linkage disequilibrium (D′ coefficient) in cases (x-axis) and controls (y-axis) between each polymorphism presented as a colored heat map. The values, scaling from 0 to 1, have been assigned various colors as shown in the right panel of the figure. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Construction of SNP haplotypes and LD analysis

Haplotype frequency estimations were performed by the use of PHASE version 2.1 as described by Stephens et al.16 To make the genotype data available for PHASE, the input was transformed through a Microsoft Excel macro developed by Cox and Canzian.17 The frequencies were assessed for calculation of D′ and r2 values as well as LD contrast test. D′ and r2 values were calculated from a Microsoft Excel macro kindly provided by Dr. David Cox and which is available upon request.

Aromatic/hydrophobic PAH–DNA adducts and TP53 mutations

Tissues adjacent to tumor samples histologically verified as normal were analyzed for DNA adducts by 32P-postlabeling as described previously.18 PAH–DNA adducts were measured in noncancer tissues as this measurement is considered to provide a direct assessment of their accumulation in normal tissue and of the intrinsic, genetic susceptibility to PAH-induced DNA damage. Indeed, in tumor tissues, levels of PAH–DNA adducts might be significantly altered by cell changes occurring during tumor progression.19 DNA was extracted from the tissue using standard proteinase K digestion followed by phenol–chloroform extraction and ethanol precipitation. Samples were analyzed using the nuclease P1 digestion method of sensitivity enhancement. Results are the means of three to four independent determinations. Benzo[a]pyrene diol-epoxide-modified DNA was used as an external positive control for each assay. Interassay variation was generally <20%. Analysis of TP53 mutations in lung tumor tissues was performed as previously described.20

Quantitative real-time reverse transcription–polymerase chain reaction

Total RNA was extracted from frozen, crushed tissues as previously described.21 Characteristics of the 24 matched pair (adjacent histologically normal versus lung tumor tissues) are given in Supporting Information Table 1. Quantitative analysis of the mRNA levels of CHRNA3 and CHRNA5 was performed by real-time reverse transcription–polymerase chain reaction on an ABI PRISM 7900HT (Applied Biosystems, Foster City, CA) as previously described.21 The expression levels of target genes were normalized to the expression of the ACTB gene (β-actin). The expression of another housekeeping gene encoding the 18S rRNA (RN18S1) was also measured as a second internal control; we found no significant differences in all the results using 18S or β-actin as a housekeeping control gene. The gene-specific sequences of primers are shown in Supporting Information Table 2.

Statistical methods

Statistical analyses were carried out using SPSS software version 17.0. Differences in continuous demographic variables were assessed by nonparametric tests. To test for population stratification, the deviation of the genotype frequencies in the controls from those expected under Hardy–Weinberg equilibrium was assessed by chi-squared test for each polymorphism with p = 0.05 as the threshold. The association between the variant genotypes and risk of lung cancer was estimated by computing odds ratios (ORs) and their 95% confidence intervals (CIs) from unconditional logistic regression analyses adjusted for age, sex and smoking. Homozygous carriers of the common allele among controls were set as the reference group. p < 0.05 was considered statistically significant. The line within each box plot represents the median fold-change value, upper and lower edges of each box, 75th and 25th percentile, respectively. The whiskers represent the lowest datum still within [1.5×(75th − 25th percentile)] of the lower quartile, and the highest datum still within [1.5×(75th − 25th percentile)] of the upper quartile. Two-sided Mann–Whitney test was used to compare medians. The mean level of mRNA levels of two independent reverse transcription and quantitative mRNA level measurements were used for the tissue samples.

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Association between polymorphisms and risk of NSCLC

The relevant characteristics of the subjects studied are shown in Table 1. Genotyping success rate was more than 90% for both cases and controls for all seven polymorphisms (Table 2); hence, there was no evidence of any systematic bias of genotyping between cases and controls. The polymorphisms were in Hardy–Weinberg equilibrium with nonsignificant p-values at p > 0.05.

Table 2. Association between polymorphisms and risk of NSCLC
inline image

A statistically significant reduced risk of developing NSCLC was found (Table 2) for homozygous carriers of the rare variants of rs3829787 (OR = 0.515, 95% CI: 0.28–0.93; p = 0.028) and rs3841324 (OR = 0.60, 95% CI: 0.36–0.99; p = 0.047). Both polymorphisms are situated in the promoter of the CHRNA5 gene (Fig. 1a). For the five other SNPs, we did not find any significant association with lung cancer risk (Table 2). However, a meta-analysis has shown that rs588765 and rs578776 could be associated with risk of lung cancer.22 Also, rs3743073 has been shown to be significantly associated with lung cancer in a Chinese Han population.23 Thus, the sample size, limited statistical power or the ethnic origin of the cases and controls are parameters to be considered when comparing our results to those of previous studies.

Polymorphisms influence the PAH–DNA adduct levels in normal lung tissue

Epidemiological studies have shown the relevant role of PAHs in lung carcinogenesis.24, 25 The PAH-related bulky DNA adduct levels may be considered as a molecular marker of the biologically effective carcinogen dose reaching the target DNA in the lung.26 The levels of aromatic/hydrophobic DNA adducts measured in the histologically normal lung tissues of 199 cases were analyzed in relation to the seven polymorphisms. The NSCLC patients homozygous for the rare allele of rs3829787 (A/A), rs384132 (del/del), rs588765 (T/T) or rs3743073 (C/C) had significantly lower levels of PAH–DNA adducts than individuals homozygous for the respective common alleles (Fig. 1b).

Polymorphisms influence the TP53 mutational status in lung tumors

The hydrophobic PAH–DNA adducts may lead to mutations in key genes such as TP53.27 The TP53 mutational status was assessed in 265 tumor tissues of the cancer patients recruited in our study; 148 of the 265 patients had at least one TP53 mutation in their lung tumors (for detailed mutation types, see Ref.20). Interestingly, we found that patients homozygous for the rare allele of five of seven polymorphisms (rs3829787, rs3841324, rs588765, rs3743073 and rs8023462) were less likely to harbor a mutated TP53 in their lung tumors; the homozygous patients for the rare alleles had significantly lower frequencies of TP53 mutations than expected (Table 3). The risk of NSCLC was also assessed in the group of patients with a nonmutated TP53 in their tumor and the group with a mutated TP53. The polymorphisms were not associated with lung cancer risk in the group of patients with wild-type TP53 when compared to the control group (Table 3). In the group of patients with at least one TP53 mutation in their tumor, several polymorphisms were found to affect the risk (Table 3). The results indicate that being homozygous for the rare allele was associated with lower PAH–DNA adducts, protected against TP53 mutations during multistage carcinogenesis and thus protected against TP53-driven lung cancer.

Table 3. Association between polymorphisms and risk of NSCLC and TP53 mutational status
inline image

Polymorphisms influence mRNA levels of the CHRNA5 gene

Quantitative analysis of CHRNA5 mRNA levels in 24 histologically normal lung tissues (characteristics of subjects are shown in Supporting Information Table 1) revealed an association between three of the polymorphisms (rs3829787, rs3841324 and rs588765) situated in CHRNA5 gene and mRNA levels. Notably, increasing CHRNA5 mRNA levels were observed in subjects harboring at least one rare allele for each polymorphism (Fig. 1c). We did not find any association between genetic variants in CHRNA3 gene and mRNA levels (Supporting Information Fig. 1).

A variant haplotype affects PAH–DNA adduct, CHRNA5 mRNA levels, is associated with TP53 mutations and risk of lung cancer

Haplotype and LD analysis of polymorphisms showed partial linkage between the polymorphisms. LD in cases and controls calculated as D′ and r2 are illustrated in Figure 1d and Supporting Information Figure 2, respectively. We identified nine haplotypes with a frequency above 1% and only three haplotypes with a frequency above 5%. For assessment of the association between the haplotypes and lung cancer risk, we used the most common haplotype as reference, which is also the haplotype that contains all common alleles of the seven polymorphisms (Table 4). Interestingly, the second most frequent haplotype was (i) protective against lung cancer (Table 4), (ii) associated with lower hydrophobic PAH–DNA adduct levels (Fig. 1e), (iii) protective against TP53 mutations (Table 5) and (iv) was associated with higher CHRNA5 mRNA levels (Fig. 1f).

Table 4. Association between haplotypes and risk of NSCLC
inline image
Table 5. Association between haplotypes, risk of NSCLC and TP53 mutational status
inline image

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

It has been suggested that the association between genetic polymorphisms in the CHRN genes at the 15q25 locus and lung cancer risk may be indirect, deriving from association with smoking addiction.28 However, recent studies have revealed a role for CHRN genes in lung cancer development and in lung cells and tissues homeostasis.9–12 A recent hypothesis has proposed that genetic variants in the CHRN genes may be associated with interindividual differences in nAChR receptor functions, which may modulate cell migration and wound repair of bronchial mucosa injured by inhaled toxic substances and thus participate in lung cancer development independently of tobacco addiction.14 This hypothesis suggested that polymorphisms associated with low cell motility might lead to more persistent mucosal damage and inflammation, which could render mucosal cells more sensitive to mutagens and would promote the formation of precursor lesions.14 It has been shown that the presence of the receptor heteropentamere, nAChR α5(α3β2)2, was essential at the edge of wounds of human lung tissue for the cell migration and repairing to be efficient.10 In addition, a recent study identified CHRNA5 as an important determinant of human lung cell motility.12 Thus, interindividual differences in CHRNA5 expression in normal lung tissue might influence the healing of tissue damage. Some of the polymorphisms at the 15q25 locus have been associated with differential mRNA levels of CHRNA5.4, 5 Similar to our study, an inverse relationship was observed between the risk allele dosage and mRNA levels.5 We found here that several of the polymorphisms and their haplotypes in the 15q25 locus were associated with (i) CHRNA5 mRNA levels, (ii) PAH–DNA adduct levels, (iii) TP53 mutations in tumors and (iv) susceptibility to lung cancer.

It has been previously observed an association between polymorphisms in CHRNA3 (rs1051730) or CHRNA5 (rs16969968) and the amount of tobacco-specific carcinogenic nitrosamines found in the urine of smokers. The carriers of the risk variants had more carcinogenic nitrosamines in their urine and thus, may be exposed to higher levels of tobacco-specific nitrosamines.29 These compounds form highly mutagenic DNA adducts such as O6-methylguanine and can also bind to nAChR and affect their function.30 However, a recent study using the A/J mouse model of lung tumorigenesis did not find any relationship between chronic consumption of nicotine and formation of O6-methylguanine adduct in the lung.31 It still could be interesting to investigate the levels of O6-methylguanine in relation to the polymorphisms of the CHRN genes studied here and also to compare the formation of this adduct with PAH adducts in the lung. We have previously studied rs1051730 or rs16969968 in relation with PAH–DNA adduct levels, which were measured in adjacent histologically normal tissues from lung tumors.32 We found that these two SNPs were associated with lung cancer risk, but there was no link between polymorphisms and the levels of PAH–DNA adduct.32 PAHs are present in tobacco smoke, and they comprise one major class of carcinogenic compounds. There is lack of studies investigating how CHRN polymorphisms may affect the smoking-related DNA alterations such as bulky PAH–DNA adducts and mutations in cancer-related genes such as TP53. Mutations in the TP53 gene in tumor may reflect some of the genetic alterations occurring in lung tissue due to formation of various DNA adducts.19 In our study, we found that homozygous carriers of the rare variants of several polymorphisms or carriers of a specific haplotype were at reduced risk of developing lung cancer. The lung cancer patients with these variants had also (i) higher CHRNA5 mRNA levels in their histologically normal lung tissues, (ii) lower levels of bulky aromatic/hydrophobic DNA adducts in their histologically normal lung tissues and (iii) lower frequency of TP53 mutations in their tumor tissue. Our results may thus support a functional role for the studied polymorphisms that could be related to differences in cellular responses of lung cells in sensitivity to mutagens. In fact, we demonstrate here that low mRNA levels of CHRNA5 associated with certain genetic polymorphisms may be permissive for the accumulation of DNA damage as measured by the hydrophobic DNA adduct levels and mutations in the tumor suppressor gene TP53. Interestingly, five of the seven variants studied strongly influenced the risk of developing lung cancer in the subgroup of lung cancer patients harboring TP53 mutations in their tumor tissue. Mutations in the TP53 gene occur in about 50% of NSCLC and show characteristics such as an excess in G:C to T:A transversions predominantly caused by PAH–DNA adducts and they often occur at hot spots.33 TP53 is important in determining the fate of damaged cells, and its activation may initiate DNA repair, cell cycle arrest, senescence or cell death.34 Besides its role in the regulation of cell survival and death, recent studies demonstrate the importance of TP53 in carcinogenesis.35

One could argue that genetic variations or haplotypes of our study at 15q25 associated with different CHRNA5 mRNA levels may also modulate the tobacco addiction of the subjects, because studies using CHRNA5 knockout mice have reported that α5 subunit was associated with substance dependence and nicotine addiction.36 However, when we analyzed the self-reported smoking habits (cigarette smoked per day and years smoked) of the lung cancer patients (cases) and the control group in relation to the polymorphisms, we did not find any association between the genetic variants and smoking (data not shown). Even if self-reported smoking may not be the best indicator of the real smoking habits or of the quantity reaching the lung,37 our results indicate that the genetic variants studied here are linked with risk of developing lung cancer independently of smoking proportion.

Quantitative analysis of mRNA levels of 24 matched pair normal–tumor tissues revealed significantly higher CHRNA5 mRNA levels in lung tumor tissues compared to the corresponding adjacent histologically normal tissues (Supporting Information Fig. 3). As CHRNA5 expression may modulate lung cells motility, it is tempting to speculate that increased CHRNA5 mRNA levels in tumors may enhance motility of cancerous cells allowing tumor to metastasize.12 It should be noted that CHRNA5 mRNA levels in the tumor tissues were not related to the genotypes of the patients (data not shown), indicating that other factors than polymorphisms may be related to increased CHRNA5 mRNA levels in tumor tissues. To our knowledge, when sequence variants in CHRNA5 have been associated with the mRNA levels, it was in the adjacent normal human lung tissues4, 5 or in normal brain38 but not in tumor tissues.

The polymorphisms we studied at the 15q25 were in strong LD. We found that the seven variants formed only three haplotypes with a frequency above 5%. The rs3841324 polymorphism has been previously studied by Falvella et al.,4 and the authors found that this polymorphism was in strong LD with three other SNPs (rs503464, rs55853698 and rs55781567) and that the haplotypes formed were related to CHRNA5 mRNA levels. Our results are in concordance with these findings, because in both studies, the genotyping of rs3841324 revealed that the minor allele (deletion) was related to higher CHRNA5 mRNA levels in “normal” lung tissue.

In summary, our study showed that several of the polymorphisms and their haplotypes in the locus 15q25 may be associated with (i) CHRNA5 mRNA levels, (ii) PAH–DNA adduct levels, (iii) TP53 mutations in tumors and (iv) susceptibility to lung cancer. Particularly, these results indicate a functional role for the CHRNA5 polymorphisms in lung cancer susceptibility related to the sensitivity to mutagens.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors gratefully acknowledge collaboration of Dr. Lodve Stangeland, Haukeland, for recruiting the lung cancer patients. They are grateful to Mrs. Tove Andreassen, Kristine Haugen Anmarkrud and Elin Einarsdottir Thorner for excellent technical assistance.

References

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
IJC_27870_sm_SuppFig1.ppt354KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.