Search

Expression of the glutathione S-transferase, GSTP1, is associated with phase 1 detoxification of the products of oxidative stress. Recently, GSTP1 expression has been implicated in the regulation of cell proliferation and apoptosis through direct interaction with the c-Jun N-terminal kinase, (JNK). GSTP1 is polymorphic and allelic variants have been associated with disease susceptibility and clinical outcome. However, the influence of GSTP1 alleles on proliferation and apoptosis has not been studied previously. To investigate this, we have examined the effects of inducible expression of wild-type GSTP1*A and mutant GSTP1*C haplotypes on cell proliferation and apoptosis in NIH3T3 fibroblasts. Cells expressing GSTP1*A displayed increased doubling times and a delayed G1-S phase transition compared with cells expressing GSTP1*C. Both GSTP1*A and GSTP1*C haplotypes protected cells from undergoing apoptosis when exposed to oxidative stress. However, analysis of JNK status revealed that only GSTP1*C expression led to a reduction in JNK activity compared with GSTP1*A-expressing cells and non-induced cells. We further examined the effect of GSTP1 alleles on colony-forming efficiency (CFE) in soft agar following exposure to oxidative stress and found that GSTP1*A-expressing clones had increased CFE compared with non-induced and GSTP1*C-expressing clones. Our data suggest that GSTP1 alleles have differential effects on proliferation and apoptosis; GSTP1*A reduces cellular proliferation and protects against apoptosis through a JNK-independent mechanism. In contrast, GSTP1*C does not influence cellular proliferation but protects cells from apoptosis through JNK-mediated mechanisms.

To test the hypothesis of interaction among genetic variants in increasing the individual risk of cancer, we have studied the cumulative effect on lung cancer risk of variants in three metabolic genes, CYP1A1, GSTM1 and GSTT1, which are involved in the metabolism of the tobacco smoke constituents and environmental contaminants, polycyclic aromatic hydrocarbons and of other lung carcinogens. We have selected from the Genetic Susceptibility to Environmental Carcinogens pooled analysis all the studies on lung cancer conducted after 1991 in which all variants were available. The data set includes 611 cases and 870 controls. We found a cumulative effect of the combination of the a priori 'at-risk' alleles for these genes (P for trend 0.004). The risk of lung cancer was increased with the combination of CYP1A1*2B or CYP1A1*4 alleles and the double deletion of both GSTM1 and GSTT1 up to an odds ratio (OR) of 8.25 (95% confidence interval 2.29-29.77) for the combination including CYP1A1*4; among never smokers, the latter combination was associated with an OR of 16.19 (1.90-137). Estimates did not change after adjustment by the number of cigarettes smoked and duration of smoking were consistent across ethnicities and were approximately the same for adenocarcinomas and squamous cell carcinomas. These observations from a large pooled analysis strongly suggest the existence of gene-gene interactions in lung carcinogenesis. People with rare combinations of common gene variants have a high risk of cancer and can be assimilated to subjects with highly penetrant mutations.

Colorectal cancer (CRC) remains a significant cause of mortality accounting for approximately 10% of all deaths from malignancy in the western world. Polymorphism in the glutathione S-transferase GSTT1 gene has been associated with CRC risk in some but not all studies. In this study, we examined associations between GSTT1 genotypes and CRC risk, and prognosis in 361 cases and 881 unrelated controls. GSTT1 null was associated with a small but significant increase in risk (P = 0.0006, odds ratio (OR) = 1.65, 95% confidence interval (CI) = 1.22-2.24). GSTT1 null was also associated with a significantly younger age at diagnosis (mean 65.2 years) compared with GSTT1 A (mean 67.6 years, P = 0.031). There were no significant associations between GSTT1 genotypes and clinical factors (e.g. Dukes stage, differentiation and tumour node metastasis classification) in the total case group. However, following stratification by age (<70 versus > or =70 years at diagnosis), in the patients diagnosed <70 years of age, GSTT1 null was more common in Dukes grade A/B tumours (P = 0.046), stage T1/T2 tumours (P = 0.053) and those with a pushing margin (P = 0.066). We also identified associations between GSTT1 null and increased prevalence of host lymphocyte response, particularly in the younger patients (P = 0.036). Furthermore, GSTT1 null was associated with improved survival in younger patients (P = 0.017, hazards ratio (HR) = 0.52, 95% CI = 0.31-0.89) but poorer survival in older patients (P = 0.017, HR = 1.89, 95% CI = 1.12-3.20). We proposed a model based on the dual functionality of GSTT1 to explain these contrasting results. We suggest that the null genotype is associated with improved immune response in younger patients, but poorer detoxification in older patients. These findings may also provide an explanation for the contrasting finding of other studies on the role of this gene in CRC.

Caucasian renal transplant recipients from Queensland, Australia have the highest non-melanoma skin cancer (NMSC) risk worldwide. Although ultraviolet light (UVR) exposure is critical, genetic factors also appear important. We and others have shown that polymorphism in the glutathione S-transferases (GST) is associated with NMSC in UK recipients. However, the effect of high UVR exposure and differences in immunosuppressive regimen on these associations is unknown. In this study, we examined allelism in GSTM1, GSTM3, GSTT1 and GSTP1 in 361 Queensland renal transplant recipients. Data on squamous (SCC) and basal cell carcinoma (BCC), UVR/tobacco exposure and genotype were obtained. Associations with both NMSC risk and numbers were examined using logistic and negative binomial regression, respectively. In the total group, GSTM1 AB [P = 0.049, rate ratio (RR) = 0.23] and GSTM3 AA (P = 0.015, RR = 0.50) were associated with fewer SCC. Recipients were then stratified by prednisolone dose (< or =7 versus >7 mg/day). In the low-dose group, GSTT1 null (P = 0.006, RR = 0.20) and GSTP1 Val/Val (P = 0.021, RR = 0.20) were associated with SCC numbers. In contrast, in the high-dose group, GSTM1 AB (P = 0.009, RR = 0.05), GSTM3 AB (P = 0.042, RR = 2.29) and BB (P = 0.014, RR = 5.31) and GSTP1 Val/Val (P = 0.036, RR = 2.98) were associated with SCC numbers. GSTM1 AB (P = 0.016) and GSTP1 Val/Val (P = 0.046) were also associated with fewer BCC in this group. GSTP1 associations were strongest in recipients with lower UVR/tobacco exposure. The data confirm our UK findings, suggesting that protection against UVR-induced oxidative stress is important in NMSC development in recipients, but that this effect depends on the immunosuppressant regimen.

Recent studies have proposed that exposure to ultraviolet radiation (UVR) protects against development of some internal cancers including that in prostate. This effect may be mediated by UVR-induced cutaneous synthesis of vitamin D. It is also proposed that ability to pigment in response to UVR will influence susceptibility to prostate cancer through its effects on vitamin D synthesis. We wished to determine first, whether ability to pigment, as assessed by skin type, influences the extent of exposure to UVR, secondly, whether skin type is associated with prostate cancer susceptibility and thirdly, whether such an effect is mediated by the extent of UVR exposure. We studied 453 prostatic adenocarcinoma and 312 benign prostatic hypertrophy (BPH) patients using a validated questionnaire to assess two parameters of exposure; months of cumulative exposure per year and adult sunbathing score. We used analysis of variance to show that cancer cases with sun-sensitive skin (skin type 1) had lower cumulative exposure per year (P = 0.014) and sunbathing scores (P < 0.0001) than those with type 4, possibly because of a tendency to avoid exposure. Further, cumulative exposure per year and sunbathing score were significantly lower in cancer compared with BPH patients (P < 0.001 and P < 0.001, respectively). While the proportion of subjects with skin types 1 and 2 was lower in cancer than BPH patients, these were not significantly different (logistic regression analysis, skin type 1 versus type 4; P = 0.11). We used recursive partitioning to determine if skin type influenced susceptibility to prostate cancer in subgroups stratified by exposure. Analysis of the data showed that in men with low sunbathing scores, skin type 1 conferred protection compared with skin types 2-4 (OR = 4.78, 95% CI 3.01-8.25, P < 0.0009). These findings indicate that susceptibility to prostate cancer is in part determined by extent of exposure to UVR and that ability to pigment mediates this effect.

Basal cell carcinoma (BCC) patients demonstrate marked variation in tumour numbers and site. Previous studies also show an association between age at first BCC presentation and development of BCC on the trunk. In this study we have investigated the association between age at first presentation and the rate of development of truncal and non-truncal tumours in 747 patients with BCC. We used negative binomial regression analysis to show that increasing age at first presentation was associated with an increased rate of BCC development (rate ratio 1.01/year, 95% CI 1.01-1.02, P < 0.001). In particular, development of tumours was greater in cases aged 60.0-69.9, 70.0-79.9 and 80.0-89.9 years than in those 40.0-49.9 years (P = 0.05, 0.01 and 0.039, respectively). While few cases aged over 70 years of age first present with a truncal BCC, the numbers of BCC/year were greater than in those with a head/neck BCC. The data suggest different genetic factors mediate the appearance of BCC in patients of different ages particularly those aged above and below 60 years.

Susceptibility to lung cancer may in part be attributable to inter-individual variability in metabolic activation or detoxification of tobacco carcinogens. The glutathione S-transferase M1 (GSTM1) genetic polymorphism has been extensively studied in this context; two recent meta-analyses of case-control studies suggested an association between GSTM1 deletion and lung cancer. At least 15 studies have been published after these overviews. We undertook a new meta-analysis to summarize the results of 43 published case-control studies including >18 000 individuals. A slight excess of risk of lung cancer for individuals with the GSTM1 null genotype was found (odds ratio (OR) = 1.17, 95% confidence interval (CI) 1.07-1.27). No evidence of publication bias was found (P = 0.4), however, it is not easy to estimate the extent of such bias and we cannot rule out some degree of publication bias in our results. A pooled analysis of the original data of about 9500 subjects involved in 21 case-control studies from the International Collaborative Study on Genetic Susceptibility to Environmental Carcinogens (GSEC) data set was performed to assess the role of GSTM1 genotype as a modifier of the effect of smoking on lung cancer risk with adequate power. Analyses revealed no evidence of increased risk of lung cancer among carriers of the GSTM1 null genotype (age-, gender- and center-adjusted OR = 1.08, 95% CI 0.98-1.18) and no evidence of interaction between GSTM1 genotype and either smoking status or cumulative tobacco consumption.

Epidemiological studies have suggested that UV exerts a protective effect on prostate cancer. Accordingly, we determined, in 210 prostate cancer cases, whether parameters of exposure, skin type and polymorphism in MC1R, VDR and TYR were associated with the outcome parameters, histological grade, clinical stage and presence of bone metastases. We used logistic regression analysis, with correction for age and metastases, stage and grade in the models, to determine if the frequencies of individual factors were different in the patient groups. The development of metastases was not associated with UV exposure parameters. Paradoxically, patients with skin type 1 were at significantly reduced risk [P = 0.027, odds ratio (OR) 0.17, 95% CI 0.03-0.82] of developing metastases compared with cases with skin type 4. MC1R Val92/Val92 and VDR ff were associated with increased risk of metastases (ORs 4.30 and 4.98, respectively). Further, cumulative exposure (P = 0.005, OR 0.85/year) and increasing proportion of outdoor occupation (P = 0.001, OR 0.84/unit) were associated with reduced risk of advanced stage tumours. Skin types, MC1R or VDR genotypes were not significantly associated with advanced stage. None of the exposure parameters, skin types or genotypes were associated with tumour grade. While MC1R Val92/Val92 and VDR ff were only associated with bone metastases, TYR genotypes were associated with each of the outcome parameters. Thus, in logistic regression models that included age, but not advanced stage and high grade histology, TYR A1A2 was significantly associated with reduced risk of metastases (P = 0.033, OR 0.41). Similarly, in models that included age but not the other outcome parameters, associations between TYR A2A2 and high-grade and advanced stage were significant (P = 0.040, OR 0.41) or approached significance (P = 0.052, OR 0.44), respectively. These data indicate for the first time that pigmentation response to UV is associated with outcome in prostate cancer.

Glutathione S-transferase GSTM1 B and GSTT1 null, and cytochrome P450 CYP2D6 EM have been associated with cutaneous basal cell carcinoma (BCC) numbers, although their quantitative effects show that predisposition to many BCC is determined by an unknown number of further loci. We speculate that other loci that determine response to oxidative stress, such as NAD(H):quinone oxidoreductase (NQO1) are candidates. Accordingly, we assessed the association between NQO1 null and BCC numbers primarily to rank NQO1 null in a model that included genotypes already associated with BCC numbers. We found that only 14 out of 457 cases (3.1%) were NQO1 null. This frequency did not increase in cases with characteristics linked with BCC numbers including gender, skin type, a truncal lesion or more than one new BCC at any presentation (MPP). However, the mean number of BCC in NQO1*0 homozygotes was greater than in wild-type allele homozygotes and heterozygotes, although the difference was not quite significant (P = 0.06). These data reflect the link between NQO1 null and BCC numbers in the 42 MPP cases rather than the whole case group. We identified an interaction between NQO1 null and GSTT1 null that was associated with more BCC (P = 0.04), although only four cases had this combination. The relative influence of NQO1 null was studied in a multivariate model that included: (i) 241 patients in whom GSTM1 B, GSTT1 null and CYP2D6 EM genotype data were available, and (ii) 101 patients in whom these genotypes, as well as data on GSTM3, CYP1A1 and melanocyte-stimulating hormone receptor (MC1R) genotypes were available. NQO1 null (P = 0.001) and MC1R asp294/asp294 (P = 0.03) were linked with BCC numbers, and the association with CYP2D6 EM approached significance (P = 0.08). In a stepwise regression model only these genotypes were significantly associated with BCC numbers with NQO1 null being the most powerful predictor.

Basal cell carcinoma (BCC) is the commonest cancer in Caucasians. Its incidence is rising and many patients develop multiple primary tumours at separate sites. Factors determining time between first primary tumour presentation and the next new primary lesion are unclear. We used Cox's proportional hazards model to study, in 856 Caucasians, the influence of tumour site, individual characteristics and polymorphism in glutathione S-transferase (GSTM1, GSTT1) and cytochrome P450 (CYP2D6, CYP1A1) loci on time to next primary tumour presentation. More than one tumour at first presentation (P <0.0001, hazard ratio 2.72) and GSTT1 null (P = 0.028, hazard ratio 1.74) were associated with decreased time to next primary tumour presentation. Significant two-factor interactions, corrected for number of tumours at presentation, were identified between a truncal tumour at first presentation and each of male gender, GSTM1 null and CYP2D6 EM (P <0.003, hazard ratios 3.09-3.82). In each of these cases, all patients with the risk combination demonstrated further separate tumours within 5 years of first presentation. Thus, patients with a truncal tumour at first presentation, especially males and those presenting with more than one lesion have a significantly decreased time to presentation of further tumours and should receive more meticulous follow-up. Polymorphism in GSTM1 and CYP2D6 also influences the rate of new primary tumour accrual giving insights into the link between ultraviolet exposure and multiple tumour development.

The genetic factors that mediate the pathogenesis of multiple primary cutaneous basal cell carcinomas (BCC) are largely unclear. Thus, some patients suffer many BCC (>30) and/or rapid accrual (number of tumours/year from first presentation) of further lesions. We have studied, in 827 English Caucasians, the influence of polymorphism in carcinogen-metabolizing enzymes on susceptibility to this cancer. Accordingly, we describe, first, a cross-sectional analysis of the influence of GSTM1, GSTT1, CYP2D6 and CYP1A1 genotypes on tumour numbers, and secondly, a longitudinal analysis, in 169 of these cases, of the effect of these genes on tumour accrual. We have confirmed the expected importance of age and number of lesions at presentation, and male gender and skin type as risk factors. Furthermore, the cross-sectional analysis showed CYP1A1 m1m1 (P = 0.004; rate ratio 1.242) and CYP2D6 EM (P < 0.001, rate ratio 1.266) are associated with increased numbers of BCC. The longitudinal study showed, after adjustment for age and tumour number at presentation, that GSTT1 null (P < 0.001, rate ratio 2.677) and CYP2D6 EM (P < 0.001, rate ratio 2.154) were significant determinants of accrual while CYP1A1 Ile/Ile was associated with slower accrual than the Ile/Val and Val/Val genotypes (P = 0.008, rate ratio 0.690). We believe these are the first genetic factors to be associated with tumour accrual. No significant interactions between genotypes were identified, though the combinations GSTM1 null/skin type 1 (P < 0.001, rate ratio 2.702), CYP2D6 EM/male gender (P = 0.049, rate ratio 1.279) and CYP2D6 EM/blue+green eyes (P = 0.046, rate ratio 1.388) influenced tumour numbers. Previous studies indicate the importance of effective repair of UV-damaged DNA in the pathogenesis of multiple BCC; indeed the influence of GSTM1 may result from its ability to utilize 5'-hydroxymethyluracil. However, the finding that CYP2D6 and CYP1A1 influence tumour numbers and accrual indicates detoxification of unknown molecules is important and supports the view that factors other than UV are important in the pathogenesis BCC.

We describe studies to assess the influence of polymorphism in the human glutathione S-transferase GSTM3 gene on susceptibility to high grade astrocytoma. Immunohistochemical studies using a GSTM3-specific antiserum identified expression of the GSTM3 subunit in astrocytes. The relative levels of expression of GSTM1 and GSTM3 in brain cytosols were determined after resolution of these enzymes using chromatofocusing. We found no differences in the level of GSTM3 activity in individuals with GSTM1 null and those with GSTM1-positive genotypes (GSTM1 A, GSTM1 B and GSTM1 A/B). A case-control study was performed to determine if GSTM3 alone or in combination with GSTM1 or GSTT1 influenced susceptibility to high grade astrocytoma. After correction for differences in age and gender, GSTM3 AA was not significantly different in cases compared with controls. No significant interactions between GSTM3 AA and GSTM1 null were identified. The significant interaction between GSTM3 AA and GSTT1 null appeared to result from the strength of the main effect (GSTT1 null). The data show that while GSTM3 is expressed in astrocytes and contributes significantly to total GST activity in human brain, it does not appear to influence susceptibility to high grade astrocytoma. Further, unlike lung, there appears to be no relationship between the level of GSTM3 activity in brain and GSTM1 genotype.

Allelism in glutathione S-transferase GSTM1 and GSTT1 has been suggested as a risk factor in various cancers. Accordingly, we describe a group of case-control studies carried out to identify associations between GSTT1 genotypes and susceptibility to lung, oral, gastric and colorectal cancers. The frequencies of the putatively high risk GSTT1 null genotype were not increased in the lung, oral or gastric cancer cases compared with controls but the frequency of this genotype was significantly increased (P = 0.0011, odds ratio = 1.88) in the colorectal cancer cases. No significant interactions between the GSTT1 and GSTM1 null genotypes types were identified in the cancer groups studied. Indeed, no significant associations between GSTM1 genotypes and susceptibility were identified though further evidence was obtained that the protective effect of GSTM1*A and GSTM1*B is not equal. The data complement studies showing that GSTT1 null is associated with an increased susceptibility to total ulcerative colitis and suggests that this enzyme is important in the detoxification of unidentified xenobiotics in the large intestine.

The frequency of the GSTM1 0 polymorphism at the glutathione S-transferase M1 locus has been determined in controls and patients with pituitary adenomas by using the polymerase chain reaction to amplify genomic DNA in the exon 4-5 region of the gene. The frequency of the genotype in patients with prolactinomas, non-functional adenomas, corticotrophinomas and somatotrophinomas varied between 52-67% compared with 44% in the controls. In the patients with prolactinomas the frequency of the genotype (67%) was significantly greater than in controls with odds ratio analysis indicating that GSTM1 0 individuals have a 2.56-fold greater risk of developing this adenoma.

The suggestion that individuals with the GST1 0 phenotype have a greater susceptibility to carcinogens than those with other GST1 phenotypes has been examined by using a starch gel zymogram approach to compare the frequency of this phenotype in control subjects and a group of patients with adenocarcinoma of stomach and colon. A significantly greater proportion of the patients with adenocarcinoma demonstrated the null phenotype, odds ratio analysis indicating that individuals with this polymorphic variant have an approximately 3-fold greater risk of developing these cancers.