Abstract
Background: Hereditary nonpolyposis colorectal cancer (HNPCC) is associated with germ-line mutations in DNA mismatch repair genes. There is considerable variation in disease expression that cannot be explained by genotype/phenotype correlation, which is likely to be the result of polymorphic modifier genes. One candidate group of modifiers is the xenobiotic clearance enzyme genes that encode CYP1A1, GSTM1, GSTT1, GSTP1, and NAT2. Alterations in these xenobiotic clearance genes can potentially influence the host response to carcinogen exposure and thereby alter cancer risk. We have investigated eight polymorphisms in xenobiotic clearance genes to assess the effect on the risk of disease in mutation positive HNPCC patients.
Methods: DNA samples from 220 mutation-positive HNPCC participants (86 Australian and 134 Polish) were genotyped for single nucleotide polymorphisms (SNP) in CYP1A1, GSTM1, GSTT1, GSTP1, and NAT2. The association between the SNPs and disease characteristics, disease expression and age of diagnosis of colorectal cancer (CRC), was tested with Pearson's χ2 and Kaplan-Meier survival analysis.
Results: The HNPCC population displays a significant difference in the genotype frequency distribution between CRC patients and unaffected mismatch repair gene mutation carriers for the CYP1A1 SNP where the CRC patients harbor more of the mutant genotype.
Conclusions: Evidence from this study is not conclusive, but our data suggest that the CYP1A1 influences disease expression in individuals with HNPCC. (Cancer Epidemiol Biomarkers Prev 2006;15(11):2307–10)
Introduction
Hereditary nonpolyposis colorectal cancer (HNPCC) is an autosomal dominant inherited condition caused by germ-line mutations in DNA mismatch repair (MMR) genes, mainly affecting hMLH1 and hMSH2 (1, 2). MMR genes are important in the DNA repair mechanism, repairing mistakes caused by DNA polymerase slippage during DNA replication (3). Individuals with HNPCC have an 80% risk of developing colorectal cancer (CRC) during their lifetime, at a mean age lower than that of the general population and they are at risk of developing other epithelial malignancies, most notably cancer of the endometrium (4).
There is evidence to suggest that most cancers arise, at least in part, as a consequence of exposure to environmental mutagenic agents. Because of the significance of xenobiotics in the environment, perturbations in the ability to remove them are likely to alter disease risk in HNPCC. The detoxification and elimination of foreign chemicals are controlled by complex mechanisms involving phase I and phase II enzymes (5). A person's susceptibility to environmental and occupational carcinogens and predisposition to cancer can therefore be influenced by metabolism of xenobiotics (6).
The single nucleotide polymorphisms (SNP) chosen for this study have been associated with a variety of cancers but the roles that the different SNPs have on cancer risk are controversial (7-13). The aim of this study was to investigate SNPs in several candidate xenobiotic clearance genes that may confer an altered susceptibility to disease expression in HNPCC patients. We have investigated CYP1A1 T3801C (MspI), GSTM1 and GSTT1 deletions, GSTP1 I105V and NAT2 T341C, C481T, G590A, and G857A.
Materials and Methods
Subjects
Local ethics committees in Australia and Poland approved the study. The HNPCC patients in this study have been described previously (14); they all harbored a mutation in either hMLH1 or hMSH2 and consisted of 86 samples collected in the state of New South Wales, Australia from 1998 to 2003, whereas 134 samples were collected in Poland from 1997 to 2002. Age of diagnosis of CRC was defined as the patient's age at diagnosis, whereas age for the unaffected patients was determined by using their date of birth and current year. Age was unknown for six Australian and five Polish patients, and disease status was unknown for two Australian patients. Each study subject had previously contributed blood from which DNA was extracted, by the salt precipitation method (15).
Multiplex PCR—GSTM1/GSTT1 (Gene Deletion)
Genotyping for the GSTM1 and the GSTT1 polymorphisms was determined simultaneously in a single assay using a multiplex PCR approach. The samples were genotyped with a method described previously by Abdel-Rahman et al. (16).
Restriction Fragment Length Polymorphism—CYP1A1
The CYP1A1 T3801C (MspI, rs4646903) SNP was genotyped using primers reported by Sugawara et al. (17) in a 50 μL PCR containing 2.5 mmol/L MgCl2, 10 pmol of each primer (Invitrogen, Frederick, MD), 5 μL of 10× PCR buffer, 0.2 μmol/L deoxynucleotide triphosphates (Promega Corp., Sydney, NSW), and 2 units Platinum Taq DNA polymerase (Invitrogen). PCR was done at 95°C for 5 minutes for the initial denaturing followed by 35 cycles of 95°C for 45 seconds, 60°C for 45 seconds, and 72°C for 1 minute and 15 seconds before final extension at 72°C for 10 minutes.
SNP Genotyping
NAT2 and GSTP1 genotyping was done using the Taqman 5′-nuclease assay (Applied Biosystems, Foster City, CA). Assay-by-Design was used to design primers and probes for NAT2 T341C (rs1801280; 5′-ACTGGCATGGTTCACCTTCTC-3′ (forward), 5′-CCCAGCATCGACAATGTAATTCCT-3′ (reverse), 5′-VIC-CCGTCAATGGTCACC-3′ [wild-type (WT) probe], and 5′-FAM-CGTCAGTGGTCACC-3′ (mutant probe)], whereas Assay-by-Demand was used for GSTP1 (rs947894, assay ID: C_3237198_1), NAT2 C481T (rs1799929, assay ID: C_1204092_10), NAT2 G590A (rs1799930, assay ID: C_1204091_10), and NAT2 G857A (rs1799931, assay ID: C_572770_10).
NAT2 Classification
A review from 1999 provided a table of human NAT2 allele designations (18) that was adopted when classifying the NAT2 SNPs used in this study. Human NAT2*4 allele is defined as the reference human NAT2 allele because it is associated with high activity (WT allele). The presumed phenotypes are classified fast (homozygous and heterozygous WT allele carriers) and slow (homozygous mutant allele carriers) acetylators.
Statistical Analysis
Pearson's χ2 statistics was used to evaluate the deviation from the expected Hardy-Weinberg genotypic proportions in the subject group. Statistical analysis was undertaken to assess whether the SNPs segregate with disease expression or age of diagnosis of CRC in HNPCC patients. All statistical tests were done using the statistical software package Intercooled Stata 8.0 (Stata Corp., College Station, Texas) and GraphPad Instat version 3.06 (GraphPad Software, San Diego, CA). The significance levels for all tests were set at P < 0.05. Pearson's χ2 test was used for comparison of the distributions of the SNPs, whereas odds ratio (OR) and 95% confidence intervals (95% CI) were calculated from a 2 × 2 table. Kaplan-Meier survival analysis was used to compare genotype and age of diagnosis of CRC. The log-rank test, Wilcoxon's test, and Tarone-Ware test were used to examine the homogeneity of the Kaplan-Meier survival curves. Comparison of the distribution of slow and rapid acetylators in the different groups were analyzed using Pearson's χ2 test. Bonferroni correction was applied if the results were initially significant to take into account the problem of multiple testing.
Results
The differences/similarities observed in disease expression between Australian and Polish HNPCC participants have been reported previously (14).
Genotype Frequency Distribution
Two of the SNPs in the study deviated from Hardy-Weinberg equilibrium, CYP1A1 (χ2 = 4.58 at 1 degree of freedom; P = 0.03) and NAT2 G857A (χ2 = 16.37 at 1 degree of freedom; P = 0.0001). The distribution of the genotype frequencies of the SNPs was examined; significant findings were then explored to examine evidence potentially exaggerated by geographic differences. Two samples consistently failed to amplify for NAT2 C481T and NAT2 G857A, whereas one sample consistently failed to amplify for NAT2 T341C and NAT2 G590A; these samples were left out of the study for the SNP in question. The genotype and allele frequency distribution for each SNP is presented in Table 1.
Study demographics of the eight SNPs studied in HNPCC patients according to disease expression: affected with CRC (CRC+) and unaffected MMR gene mutation carriers (CRC−)
CYP1A1 T3810C . | TT (%) . | TC (%) . | CC (%) . | Any C (%) . | P* . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Subject group (n = 220) | 187 (85) | 29 (13) | 4 (2) | |||||||
Allele frequency | 0.916 | 0.084 | ||||||||
CRC+ (n = 118) | 94 (80) | 20 (17) | 4 (3) | 24 (20) | 0.03 | |||||
CRC− (n = 100) | 91 (91) | 9 (9) | 0 (0) | 9 (9) | ||||||
OR, 0.39;† 95% CI, 0.17-0.88; P = 0.03 | ||||||||||
GSTM1 del | WT (%) | Deletion (%) | Any A (%) | P* | ||||||
Subject group (n = 220) | 99 (45) | N/A | 121 (55) | |||||||
Allele frequency | 0.450 | 0.550 | ||||||||
CRC+ (n = 118) | 53 (45) | N/A | 65 (55) | 65 (55) | 0.99 | |||||
CRC− (n = 100) | 45 (45) | N/A | 55 (55) | 55 (55) | ||||||
OR, 0.99;† 95% CI, 0.58-1.70; P = 0.99 | ||||||||||
GSTT1 del | WT (%) | Deletion (%) | Any A (%) | P* | ||||||
Subject group (n = 220) | 175 (80) | N/A | 45 (20) | |||||||
Allele frequency | 0.795 | 0.205 | ||||||||
CRC+ (n = 118) | 93 (79) | N/A | 25 (21) | 25 (21) | 0.83 | |||||
CRC− (n = 100) | 80 (80) | N/A | 20 (20) | 20 (20) | ||||||
OR, 0.93;† 95% CI, 0.48-1.80; P = 0.96 | ||||||||||
GSTP1 I105V | AA (%) | AG (%) | GG (%) | Any G (%) | P* | |||||
Subject group (n = 220) | 109 (50) | 87 (40) | 24 (11) | |||||||
Allele frequency | 0.693 | 0.307 | ||||||||
CRC+ (n = 118) | 56 (47) | 49 (42) | 13 (11) | 62 (53) | 0.80 | |||||
CRC− (n = 100) | 52 (52) | 38 (38) | 10 (10) | 48 (48) | ||||||
OR, 0.83;† 95% CI, 0.49-01.42; P = 0.59 | ||||||||||
NAT2 T341C | TT (%) | TC (%) | CC (%) | Any C (%) | P* | |||||
Subject group (n = 219) | 71 (32) | 105 (48) | 43 (20) | |||||||
Allele frequency | 0.564 | 0.436 | ||||||||
CRC+ (n = 117) | 33 (28) | 61 (52) | 23 (20) | 84 (72) | 0.33 | |||||
CRC− (n = 100) | 37 (37) | 43 (43) | 20 (20) | 63 (63) | ||||||
OR, 0.67;† 95% CI, 0.38-1.19; P = 0.22 | ||||||||||
NAT2 C481T | CC (%) | CT (%) | TT (%) | Any T (%) | P* | |||||
Subject group (n = 218) | 76 (35) | 103 (47) | 39 (18) | |||||||
Allele frequency | 0.585 | 0.415 | ||||||||
CRC+ (n = 117) | 36 (31) | 59 (50) | 22 (19) | 81 (69) | 0.41 | |||||
CRC− (n = 99) | 39 (39) | 43 (43) | 17 (17) | 60 (60) | ||||||
OR, 0.68;† 95% CI, 0.39-1.20; P = 0.24 | ||||||||||
NAT2 G590A | GG (%) | GA (%) | AA (%) | Any A (%) | P* | |||||
Subject group (n = 219) | 126 (58) | 77 (35) | 16 (7) | |||||||
Allele frequency | 0.751 | 0.249 | ||||||||
CRC+ (n = 118) | 72 (61) | 38 (32) | 8 (7) | 46 (39) | 0.63 | |||||
CRC− (n = 99) | 54 (55) | 37 (37) | 8 (8) | 45 (45) | ||||||
OR, 1.30;† 95% CI, 0.76-2.24; P = 0.41 | ||||||||||
NAT2 G857A | GG (%) | GA (%) | AA (%) | Any A (%) | P* | |||||
Subject group (n = 218) | 212 (97) | 5 (2) | 1 (0) | |||||||
Allele frequency | 0.984 | 0.016 | ||||||||
CRC+ (n = 118) | 115 (97) | 2 (2) | 1 (1) | 3 (3) | 0.53 | |||||
CRC− (n = 98) | 95 (97) | 3 (3) | 0 (0) | 3 (3) | ||||||
OR, 1.30;† 95% CI, 0.24-6.14; P = 0.82 |
CYP1A1 T3810C . | TT (%) . | TC (%) . | CC (%) . | Any C (%) . | P* . | |||||
---|---|---|---|---|---|---|---|---|---|---|
Subject group (n = 220) | 187 (85) | 29 (13) | 4 (2) | |||||||
Allele frequency | 0.916 | 0.084 | ||||||||
CRC+ (n = 118) | 94 (80) | 20 (17) | 4 (3) | 24 (20) | 0.03 | |||||
CRC− (n = 100) | 91 (91) | 9 (9) | 0 (0) | 9 (9) | ||||||
OR, 0.39;† 95% CI, 0.17-0.88; P = 0.03 | ||||||||||
GSTM1 del | WT (%) | Deletion (%) | Any A (%) | P* | ||||||
Subject group (n = 220) | 99 (45) | N/A | 121 (55) | |||||||
Allele frequency | 0.450 | 0.550 | ||||||||
CRC+ (n = 118) | 53 (45) | N/A | 65 (55) | 65 (55) | 0.99 | |||||
CRC− (n = 100) | 45 (45) | N/A | 55 (55) | 55 (55) | ||||||
OR, 0.99;† 95% CI, 0.58-1.70; P = 0.99 | ||||||||||
GSTT1 del | WT (%) | Deletion (%) | Any A (%) | P* | ||||||
Subject group (n = 220) | 175 (80) | N/A | 45 (20) | |||||||
Allele frequency | 0.795 | 0.205 | ||||||||
CRC+ (n = 118) | 93 (79) | N/A | 25 (21) | 25 (21) | 0.83 | |||||
CRC− (n = 100) | 80 (80) | N/A | 20 (20) | 20 (20) | ||||||
OR, 0.93;† 95% CI, 0.48-1.80; P = 0.96 | ||||||||||
GSTP1 I105V | AA (%) | AG (%) | GG (%) | Any G (%) | P* | |||||
Subject group (n = 220) | 109 (50) | 87 (40) | 24 (11) | |||||||
Allele frequency | 0.693 | 0.307 | ||||||||
CRC+ (n = 118) | 56 (47) | 49 (42) | 13 (11) | 62 (53) | 0.80 | |||||
CRC− (n = 100) | 52 (52) | 38 (38) | 10 (10) | 48 (48) | ||||||
OR, 0.83;† 95% CI, 0.49-01.42; P = 0.59 | ||||||||||
NAT2 T341C | TT (%) | TC (%) | CC (%) | Any C (%) | P* | |||||
Subject group (n = 219) | 71 (32) | 105 (48) | 43 (20) | |||||||
Allele frequency | 0.564 | 0.436 | ||||||||
CRC+ (n = 117) | 33 (28) | 61 (52) | 23 (20) | 84 (72) | 0.33 | |||||
CRC− (n = 100) | 37 (37) | 43 (43) | 20 (20) | 63 (63) | ||||||
OR, 0.67;† 95% CI, 0.38-1.19; P = 0.22 | ||||||||||
NAT2 C481T | CC (%) | CT (%) | TT (%) | Any T (%) | P* | |||||
Subject group (n = 218) | 76 (35) | 103 (47) | 39 (18) | |||||||
Allele frequency | 0.585 | 0.415 | ||||||||
CRC+ (n = 117) | 36 (31) | 59 (50) | 22 (19) | 81 (69) | 0.41 | |||||
CRC− (n = 99) | 39 (39) | 43 (43) | 17 (17) | 60 (60) | ||||||
OR, 0.68;† 95% CI, 0.39-1.20; P = 0.24 | ||||||||||
NAT2 G590A | GG (%) | GA (%) | AA (%) | Any A (%) | P* | |||||
Subject group (n = 219) | 126 (58) | 77 (35) | 16 (7) | |||||||
Allele frequency | 0.751 | 0.249 | ||||||||
CRC+ (n = 118) | 72 (61) | 38 (32) | 8 (7) | 46 (39) | 0.63 | |||||
CRC− (n = 99) | 54 (55) | 37 (37) | 8 (8) | 45 (45) | ||||||
OR, 1.30;† 95% CI, 0.76-2.24; P = 0.41 | ||||||||||
NAT2 G857A | GG (%) | GA (%) | AA (%) | Any A (%) | P* | |||||
Subject group (n = 218) | 212 (97) | 5 (2) | 1 (0) | |||||||
Allele frequency | 0.984 | 0.016 | ||||||||
CRC+ (n = 118) | 115 (97) | 2 (2) | 1 (1) | 3 (3) | 0.53 | |||||
CRC− (n = 98) | 95 (97) | 3 (3) | 0 (0) | 3 (3) | ||||||
OR, 1.30;† 95% CI, 0.24-6.14; P = 0.82 |
NOTE: CRC+ pertains to the CRC patients and CRC− pertains to the unaffected MMR gene mutation carriers.
Comparison of genotype frequencies using Pearson's χ2.
OR is the relative risk for patients with “any variant allele” genotype relative to those being homozygote WT.
There was initially a significant difference between CRC patients (CRC+) and unaffected MMR gene mutation carriers (CRC−) for the CYP1A1 MspI SNP [P = 0.03 (genotype frequency); OR, 0.39; 95% CI, 0.17-0.88; P = 0.03), the individuals being CRC− harboring more of the WT genotype (see Table 1), but after Bonferroni correction the significance disappeared. When splitting the samples by geographic difference, a similar trend can be seen in the Australian population (P = 0.08) but not in the Polish population (P = 0.24).
Kaplan-Meier Analysis
No significant difference between age of diagnosis of CRC and genotype was observed for any of the SNPs in this study. The median age of diagnosis of CRC or the age at which 50% of the population is cancer-free differed by up to 8 years in three SNPs but because of a small number of cases with the mutant genotype this was not statistically significant. The reminder of the SNPs differed in age by between 1 and 4 years.
Combined NAT2 Analysis
NAT2 genotyping allowed the distinction between the NAT2*4 allele and the alleles NAT2*5A (C481T and T341C), NAT2*6B (G590A), and NAT2*7A (G857A). Homozygous or heterozygous carriers of NAT2*4 allele (WT alleles) were phenotypically classified as fast acetylators. Possession of two mutant alleles is equivalent to a slow acetylation phenotype. A patient was considered as a slow acetylator if one of the three alleles (*5A, *6B, or *7A) was present as a slow acetylator phenotype. After classifying individuals into fast or slow acetylators, frequency of these variables was assessed. This revealed no significant difference in the distribution between fast and slow acetylator phenotypes. NAT2 allele frequencies for CRC patients (CRC+) and unaffected MMR gene mutation carriers (CRC−) can be seen in Table 2 and NAT2 acetylator phenotype in Table 3. Kaplan-Meier analysis did not reveal any significant association when assessing age of diagnosis of CRC and acetylator phenotype (P = 0.86); median age of diagnosis of CRC was 47 years (fast acetylators) and 46 years (slow acetylators).
NAT2 allele frequencies in colorectal cancer patients (CRC+) and unaffected MMR gene mutation carriers (CRC−)
Cancer status . | n . | NAT2*4 . | NAT2*5A . | NAT2*6B . | NAT2*7A . |
---|---|---|---|---|---|
CRC+ | 117 | 0.72 | 0.20 | 0.07 | 0.01 |
CRC− | 94 | 0.73 | 0.21 | 0.06 | 0 |
Cancer status . | n . | NAT2*4 . | NAT2*5A . | NAT2*6B . | NAT2*7A . |
---|---|---|---|---|---|
CRC+ | 117 | 0.72 | 0.20 | 0.07 | 0.01 |
CRC− | 94 | 0.73 | 0.21 | 0.06 | 0 |
NAT2 acetylator status in cancer patients (CRC+) and unaffected MMR gene mutation carriers (CRC−)
NAT2 phenotype . | CRC+ . | CRC− . | |
---|---|---|---|
. | n (%) . | n (%) . | |
Slow acetylator | 31 (26.5) | 26 (27.7) | |
Fast acetylator | 86 (73.5) | 68 (72.3) | |
P = 0.85 |
NAT2 phenotype . | CRC+ . | CRC− . | |
---|---|---|---|
. | n (%) . | n (%) . | |
Slow acetylator | 31 (26.5) | 26 (27.7) | |
Fast acetylator | 86 (73.5) | 68 (72.3) | |
P = 0.85 |
Discussion
Polymorphisms of xenobiotic clearance genes have the potential to affect individual response to carcinogens, thereby influencing cancer risk. Environmental carcinogens directly or indirectly damage DNA and there are complex processes, which include bioactivation, detoxification, and chemical modification, that are involved in their removal (19).
The results of this study indicate that there is a trend toward a decreased risk of CRC in HNPCC participants harboring the WT CYP1A1 MspI allele. These results are different to a pharmacogenetic study of CRC where no change in disease risk was observed in patients with the MspI polymorphism but there was a difference with two other CYP1A1 SNPs (T461N and −1738A>C) both associated with a reduced risk of cancer (20). The CYP1A1 polymorphism used in this study increases the enzyme activity resulting in the accumulation of more aromatic DNA adducts than phase II enzymes can handle, contributing to disease risk.
The CYP1A1 MspI SNP was not in Hardy-Weinberg equilibrium and a significant difference between the observed and expected genotype frequencies was observed. Deviation from Hardy-Weinberg equilibrium may be indicative of a genotyping error or as a result of there being several relatives in the sample population. However, if it is the population (and not controls) that fails to fulfil Hardy-Weinberg equilibrium, it can be taken as supporting evidence for a correlation between the observed genotype and disease (21). Although this study is thus far one of the largest of its kind, more HNPCC populations need to be studied to confirm these results. Population stratification could also be a confounder in this study; however, this is unlikely to be a problem because we are searching for modifying genes affecting disease expression in HNPCC patients. Nevertheless, there remain environmental factors that are possibly different in the two countries, which could potentially affect on the results. We, however, believe this not to be the case as it has been shown that for most of the common disease-associated polymorphisms, ethnicity is likely to be a poor predictor of an individual's genotype (22).
NAT2 is expressed in the digestive system and ∼50% of individuals in Caucasian population have a slow acetylator phenotype (23). It has been reported that HNPCC and familial adenomatous polyposis patients have somewhere between 40% to 50% slow acetylators (11, 24). In our study, we have fewer individuals with the slow acetylator phenotype (27%). If being a slow acetylator affected disease expression in HNPCC patients, more individuals harboring this phenotype would be expected in the CRC patients, which is not the case in this study. Several studies have investigated the effect of NAT2 polymorphisms in patients with CRC, familial adenomatous polyposis, and HNPCC. In a HuGE review (18), it was concluded that in 10 of 11 studies of invasive CRC and NAT2 acetylator genotype, no association was observed. The different conclusions about the association of slow acetylation phenotype in different populations might be due to different genetic and environmental exposures.
The failure to identify any difference between the age of disease onset and the harboring of a particular genotype or combination of genotypes was tested using Kaplan-Meier analysis. The lack of any obvious difference may be a result of the number of participants included in this study but this is unlikely, as previous work by others using smaller study populations have been able to show an association between genotype and the age of onset of CRC (25).
Finally, the evidence from this study is not conclusive, but the data suggest that the CYP1A1 MspI polymorphism influences disease expression in HNPCC patients by increasing the chance of developing CRC.
Grant support: Hunter Medical Research Institute and Pomeranian Medical University.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Acknowledgments
We thank Dr. Kim Colyvas for advice on how to do the statistical analysis and Hunter Medical Research Institute and the Cancer Institute of New South Wales for supporting this project as well as the Pomeranian Medical University.