Search

Solar ultraviolet irradiation is an environmental carcinogen that causes skin cancer. Caspase-7 is reportedly expressed at reduced levels in many cancers. The present study was designed to examine the role of caspase-7 in solar-simulated light (SSL)-induced skin cancer and to elucidate its underlying molecular mechanisms. Our study revealed that mice with genetic deficiency of caspase-7 are highly susceptible to SSL-induced skin carcinogenesis. Epidermal hyperplasia, tumor volume and the average number of tumors were significantly increased in caspase-7 knockout (KO) mice compared with SKH1 wild-type mice irradiated with SSL. The expression of cell proliferation markers, such as survivin and Ki-67, was elevated in SSL-irradiated skin of caspase-7 KO mice compared with those observed in SSL-exposed wild-type SKH1 mouse skin. Moreover, SSL-induced apoptosis was abolished in skin from caspase-7 KO mice. Two-dimensional gel electrophoresis, followed by matrix-assisted laser desorption/ionization-time-of-flight analysis of skin tissue lysates from SSL-irradiated SKH1 wild-type and caspase-7 KO mice revealed an aberrant induction of keratin-17 in caspase-7 KO mice. Immunohistochemical analysis of skin tumors also showed an increase of keratin-17 expression in caspase-7 KO mice compared with SKH1 wild-type mice. The expression of keratin-17 was also elevated in SSL-irradiated caspase-7 KO keratinocytes as well as in human basal cell carcinomas. The in vitro caspase activity assay showed keratin-17 as a substrate of caspase-7, but not caspase-3. Overall, our study demonstrates that genetic loss of caspase-7 promotes SSL-induced skin carcinogenesis by blocking caspase-7-mediated cleavage of keratin-17., (© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com.)

We have previously demonstrated that betel quid containing safrole induced DNA adducts are highly associated with the development of oral squamous cell carcinoma (OSCC) in Taiwan. Sulfotransferase (SULT) is essential for the formation of these adducts. To elucidate the effects of SULT1A1 haplotypes on OSCC susceptibility, 160 male OSCC cases and 218 age- and sex-matched controls were screened for single-nucleotide polymorphisms within the coding region of SULT1A1 by sequencing. We found that 445C>T (His149Tyr) and 507C>T polymorphisms were significantly associated with increased risk of OSCC. Based on the genotype analysis, haplotypes were constructed for 445C>T (His149Tyr), 507C>T, 600G>C and 638G>A (Arg213His) using GENECOUNTING software. After adjustment for age, cigarette smoking and betel quid chewing, we found that haplotype c containing 445C>T (His149Tyr), 507C>T or 600G>C but not 638G>A (Arg213His) variant was significantly associated with increased risk of OSCC (odds ratio, 3.24; 95% confidence interval, 1.57-6.68) when compared with the haplotype a (wild-type). We analyzed the activity in sulfonation of 2-naphthol and 1'-hydroxysafrole of recombinant His149Tyr (445C>T) variant, which led to 51 and 33% reduced activity, respectively; Arg213His (638G>A) variant led to 72 and 54% reduced activity, respectively, when compared with the wild-type. Taken together, haplotype analysis provides a novel evaluation of the SULT1A1 gene as a risk modifier on environmental carcinogen in OSCC and the association of SULT1A1 haplotypes with the risk of OSCC might be modified by betel quid chewing.

Cyclooxygenase-2 (COX-2) is well established to play an important role in the tumorigenesis of a variety of human cancers; however, the function of COX-2 in the development of esophageal squamous cell carcinoma (ESCC) remains less clear. Here, we determined, first, the pattern of COX-2 expression in normal esophageal mucosa, dysplasia, carcinoma in situ (CIS) and invasive SCC. Immunohistochemical analysis showed that, while COX-2 was weakly expressed, if at all, in normal squamous epithelium, strong COX-2 expression was detected as early as the stage of dysplasia and frequently in 20 of 26 (77%) CIS and 86 of 111 (77%) invasive SCC. Upregulation of COX-2 in ESCC was found to be significantly associated with tumor progression (R = 0.493, P < 0.01). Further, treatment of human ESCC cell lines (KYSE450 and KYSE510) with NS-398, a COX-2 specific chemical inhibitor, suppressed the production of prostaglandin E2 (PGE2) and induced cell growth inhibition, cell cycle arrest at the G1-S checkpoint, and the expression of cyclin-dependent kinase inhibitors p21waf1/cip1 and p27kip1. Finally, knockdown expression of COX-2 in KYSE450 cells by a specific COX-2 siRNA dramatically inhibited PGE2 production, cell growth and, more importantly, colony formation and tumorigenesis in nude mice. Together, this study suggested that COX-2 may be involved in an early stage of squamous cell carcinogenesis of the esophagus and has a non-redundant role in the regulation of cellular proliferation and tumorigenesis of esophageal epithelial cells.

The p21 (Waf1/Cip1/CDKN1A) protein regulates the transition from the G1 to the S phase and has an important role in modulating cell-cycle control, apoptosis and cell growth. Two polymorphisms of the p21 gene at codon 31 (p21 C98A, dbSNP rs1801270) and at the 3' untranslated region (p21 T70C, dbSNP rs1059234) may have an effect on the protein function and may thus play a role in the development of cancer. We hypothesized that these two p21 polymorphisms are associated with the risk of squamous cell carcinoma of the head and neck (SCCHN). We tested this hypothesis in a hospital-based case-control study of 712 patients newly diagnosed with SCCHN and 1222 cancer-free controls who were frequency-matched by age, sex and ethnicity. All subjects were non-Hispanic whites. Our results showed that the variant alleles and genotypes were more common among cases than among controls (P < 0.001 and P = 0.013 for p21C70T, and P < 0.001 and P = 0.035 for p21C98A, respectively). Compared with the p21 70CC genotype, there was a significantly greater risk of SCCHN associated with the variant p21 70TC [odds ratio (OR) = 1.47, 95% confidence interval (CI) = 1.12-1.93] and combined p21 70TC/TT (OR = 1.49, 95% CI = 1.14-1.95) genotypes. Similarly, compared with the p21 98CC genotype, there was also a significantly greater SCCHN risk associated with the variant p21 98AC (OR = 1.32, 95% CI = 1.00-1.73) and combined p21 98AC/AA (OR = 1.37, 95% CI = 1.05-1.79) genotypes. When these two polymorphisms were evaluated together by the number of risk alleles, there was a significant increase in SCCHN risk that was dependent on the number of risk alleles (P(trend) = 0.001). Our results suggest that the presence of these two p21 polymorphisms may be a marker of genetic susceptibility to SCCHN.

While multiple squamous cell carcinomas are often observed in the esophagus and the head and neck region and confound us about the favorable treatments, the reason why some patients are more likely to develop multiple cancers remains obscure. We statistically analyzed clinical factors in 203 patients with newly diagnosed squamous cell carcinoma, to assess the risk of multiple cancers for the establishment of an effective prevention and screening programs. Widespread epithelial oncogenic alterations were assessed as multiple lugol-voiding lesions (multiple LVL) using lugol chromoendoscopy. Genetic polymorphisms of alcohol dehydrogenase type 3 (ADH3) and aldehyde dehydrogenase type 2 (ALDH2) were identified by PCR-restriction fragment length polymorphism analysis. Forty patients had synchronous multiple cancers and the remaining 163 had solitary cancer. Presence of multiple LVL was the only independent risk factor for multiple cancers [relative risk (RR) = 67; 95%CI, 15-310]. Multiple LVL was observed in only smoking drinkers. Among them, a multivariate analysis demonstrated that the ALDH2 deficiency allele (RR = 5.7; 95%CI, 2.8-11.6) and the slow metabolizing ADH3 allele (RR = 1.9; 95%CI, 1.1-7.9) were the independent risk factors for multiple LVL. Combination of these alleles lead to increase the risk of multiple LVL. In conclusion, an episode of multiple LVL is a remarkable high risk for multiple cancers both at the esophagus and the head and neck region. The interaction between drinking and the ALDH2 deficiency allele increases the risk. In addition, the slow metabolizing ADH3 allele enhances the risk. Prohibiting the use of alcohol and early detection of cancer are strongly recommended for such individuals.

Arachidonic acid (AA) is the major precursor of several classes of signal molecules and the alteration of its metabolism is involved in human carcinogenesis. For instance, 5-lipoxygenase (5-LOX) converts AA to hydroxyeicosatetraenoic acids or leukotrienes (LTs), which are able to enhance proliferation, increase survival and suppress the apoptosis of human cells. To determine the potential use of 5-LOX inhibitors in the prevention of esophageal cancer, we first analyzed the 5-LOX expression in esophageal tissue samples using immunohistochemistry and then examined the effect of the 5-LOX inhibitors AA861 and REV5901 on cell viability and apoptosis in esophageal cancer cell lines. 5-LOX expression was present in 79% (127/161) of esophageal cancer but in only 13% (4/32) of normal esophageal mucosa. 5-LOX was also expressed in all the eight esophageal cancer cell lines. Moreover, 5-LOX inhibitors caused a dose- and time-dependent reduction of cell viability, which was due to the induction of apoptosis and associated with LTB4 suppression. Our data also showed that both LTB4, a product of 5-LOX and LTB4 receptor antagonist U-75302 were able to prevent AA861 and REV5901 on induction of apoptosis. The present study demonstrated that 5-LOX protein expression is increased in esophageal cancer and that 5-LOX inhibitors can induce esophageal cancer cells to undergo apoptosis, suggesting that 5-LOX may be an effective target in the prevention of esophageal cancer.

Polymorphisms in GSTM1, GSTT1 and GSTP1 genes in humans are associated with the reduction of enzymatic activity toward several substrates, including those in tobacco smoke. To investigate the potential role these polymorphisms have, as modulators of early-onset lung cancer risk, a population-based case-control study involving early-onset lung cancer cases was performed. Biological samples were available for 350 individuals diagnosed <50 years of age identified from the metropolitan Detroit Surveillance, Epidemiology and End Results (SEER) program and 410 cases of age, race and sex-matched controls ascertained through random digit dialing. African Americans carrying at least one G allele at the GSTP1 locus were 2.9-fold more likely to have lung cancer compared with African Americans without a G allele after adjustment for age, sex, pack years of smoking and history of lung cancer in a first-degree relative (95% CI 1.29-6.20). African Americans with either one or two risk genotypes at the GSTM1 and GSTP1 loci were at increased risk of having lung cancer compared with those having fully functional GSTM1 and GSTP1 genes (OR = 2.8, 95% CI 1.1-7.2 and OR = 4.0, 95% CI 1.3-12.2, respectively). No significant single gene associations between GSTM1, GSTT1 or GSTP1 and early-onset lung cancer were identified in Caucasians, after adjusting for age, sex, pack years and family history of lung cancer. However, our results suggest that specific combinations of glutathione S-transferase polymorphisms increase the risk of early-onset of lung cancer. Joint analysis of these genotypes may identify individuals who are at a higher risk of developing early-onset lung cancer with a greater certainty than single gene studies.

The matrix metalloproteinases (MMPs), a family of proteolytic enzymes that degrade different components of the extracellular matrix, play important roles in tumor development and invasion. A single adenine insertion/deletion polymorphism (6A/5A) in the MMP3 promoter region causes transcriptional elevation. The aim of this study was to assess the effects of this single nucleotide polymorphism (SNP) on the development and clinical staging of esophageal squamous cell carcinoma (ESCC) and gastric cardiac adenocarcinoma (GCA). The MMP3 SNP was genotyped by polymerase chain reaction-restriction fragment length polymorphism analysis in 417 cancer patients (234 ESCC and 183 GCA) and 350 controls in north China. The overall distribution of the MMP3 SNP in ESCC and GCA patients was not significantly different from that in healthy controls. However, smoking individuals with the 5A/5A or 5A/6A genotype were significantly more common in ESCC patients than in controls (37.5 versus 23.3%, xi(2) = 5.13, P = 0.02). Thus, smokers with at least one 5A allele had a significantly increased risk of ESCC, compared with 6A homozygotes (age and sex adjusted OR = 1.95, 95% CI = 1.08-3.53). The significant difference in the SNP distribution between ESCC patients, GCA patients and controls was not observed when stratified by family history of upper gastrointestinal cancer. In addition, the frequency of the 5A/5A + 5A/6A genotypes in ESCC patients with and without lymphatic metastasis was significantly different (45.8 versus 27.8%, xi(2) = 4.56, P = 0.03). Therefore, patients with at least one 5A allele were significantly more prone to lymphatic metastasis of ESCC. In contrast, no significant difference in the SNP distribution between patients with and without lymphatic metastasis was observed in GCA. The present study suggests that the MMP3 promoter SNP might be associated with a risk of development and lymphatic metastasis in ESCC but not in GCA.

Polymorphisms in the untranslated regions (UTRs) of the thymidylate synthase (TS) gene, which may modulate TS transcription and expression, have been associated with susceptibility and prognosis of several tumors. However, their effects on the development and clinical staging of esophageal squamous cell carcinoma (ESCC) and gastric cardiac adenocarcinoma (GCA) have not been assessed so far. In this study, the 28-bp tandem repeat and the G/C single nucleotide polymorphism in the TS 5'UTR, the 6-bp deletion (6 bp-) polymorphism in the TS 3'UTR, were genotyped in 465 cancer patients (232 ESCC, 233 GCA) and 348 control subjects in North China. The genotype and allelotype distribution of the TS variants in ESCC, GCA patients and controls did not show significant difference. However, the frequency of the 6 bp-/2R haplotype in ESCC and GCA patients was marginally or significantly lower than that in controls (P = 0.05 and 0.006, respectively). Thus, the 6 bp-/2R significantly reduced the risk to ESCC and GCA, compared with the 6 bp-/3G haplotype [odds ratio (OR) = 0.61 and 0.48, 95% confidence interval (CI) = 0.37-1.00 and 0.28-0.81, respectively]. In addition, the 6 bp+/3G haplotype in ESCC patients was also significantly less common than in controls (P = 0.002). Compared with the 6 bp-/3G haplotype, the 6 bp+/3G significantly reduced the risk to ESCC (OR = 0.30, 95% CI = 0.14-0.67). Moreover, the TS 2R/3G genotype frequency in ESCC patients with and without lymphatic metastasis was significantly different (27.1 versus 4.9%, P < 0.001). Therefore, the 2R/3G genotype had an approximately 11-fold increase in the risk of lymphatic metastasis of ESCC, compared with the 3G/3G genotype (95% CI = 2.67-49.74). The results suggested that the TS polymorphisms might be associated with the susceptibility to ESCC and GCA, and the 2R/3G genotype might be a candidate marker to predict the potential of lymphatic metastasis in ESCC.

We reported previously that subjects homozygous for the cytochrome P450 2A6 (CYP2A6) (*)4 have a lower risk of lung cancer. The purpose of this study was to clarify whether or not the alterations of smoking behavior and risk for lung cancer could be found in subjects possessing novel CYP2A6 variants discovered recently. An epidemiological study was performed with 1094 cases and 611 controls in male Japanese smokers. It was found that the amounts of daily cigarette consumption in subjects who harbored CYP2A6(*)4/(*)7, (*)4/(*)10, (*)7/(*)7, (*)7/(*)9 and (*)4/(*)4 genotypes were significantly less than those in subjects carrying the (*)1/(*)1 genotype (P < 0.01). Even after adjustment with cigarette consumption, the adjusted odds ratios (ORs) for lung cancer were significantly lower in subjects who harbored CYP2A6(*)1/(*)4, (*)1/(*)7, (*)1/(*)9, (*)1/(*)10, (*)4/(*)4, (*)4/(*)7, (*)4/(*)9, (*)7/(*)7 and (*)7/(*)9 genotypes than those who possessed the (*)1/(*)1 genotype (P < 0.05). When participants were classified into four groups according to the CYP2A6 genotypes, group 1 ((*)1/(*)1), group 2 (heterozygotes for the (*)1 and a variant allele), group 3 (heterozygotes and homozygotes for variant alleles except for (*)4/(*)4) and group 4 ((*)4/(*)4), lung cancer risk was found to be less in subjects with the variant of CYP2A6 alleles [group 2, OR of 0.59 [95% confidence interval (CI), 0.44-0.79]; group 3, OR of 0.52 (95% CI, 0.37-0.72); group 4, OR of 0.30 (95% CI, 0.16-0.57)]. The reduced risk for lung cancer was seen more clearly in heavy smokers than in light smokers. Additional stratification analysis showed that the ORs for squamous cell carcinoma (OR of 0.07) and small cell carcinoma (OR of 0.10) were lower than that of adenocarcinoma (OR of 0.39) in group 4. These results suggest that the CYP2A6 is one of the principal determinants affecting not only smoking behavior but also susceptibility to tobacco-related lung cancer.

A decreased incidence of squamous cell carcinoma of the head and neck (SCCHN) associated with fruit and vegetable intake may act through chemopreventive compounds, which may be more available to persons homozygous for the deletion genotypes of the glutathione S-transferase (GST). We evaluated interactions between fruits and vegetables and GSTM1 and GSTT1 on incidence of SCCHN using data from a case-control study of 149 cases and 180 age- and gender-matched controls. After adjustment for age, gender, race, tobacco and alcohol use, weekly consumption of four or more servings of raw vegetables was inversely associated with SCCHN [odds ratio (OR) = 0.66, 95% confidence intervals (CI) 0.30-1.3]. Contrary to expectation, relatively high intake of cooked vegetables (14 or more weekly servings) and legumes (two or more weekly servings) were associated with increased incidence (OR = 2.5, 95% CI 1.1-6.0; OR = 2.5, 95% CI 1.2-5.2, respectively). In general, our results did not suggest a clear or consistent pattern of modification by GST genotypes of the association between foods and SCCHN. For example, eating cruciferous vegetables, foods of a priori interest, and having the GSTM1-deletion genotype was not associated with the expected reduction in incidence compared with abstaining from cruciferous vegetable intake and having the GSTM1-present genotype. Among non-consumers of cruciferous vegetables, the GSTM1-deletion genotype was inversely associated with SCCHN (OR = 0.55, 95% CI 0.07-4.2). Raw vegetables were associated with a reduction in incidence only among persons with the GSTM1-deletion genotype (OR = 0.69, 95% CI 0.29-1.6), whereas either factor alone had a null association. Future research of GST-diet interactions and SCCHN would benefit from larger, population-based studies.

Sulfotransferase 1A1, an important member of sulfotransferase superfamily, is involved in the biotransformation of many compounds including tobacco carcinogens. A single nucleotide polymorphism (G638A) in the sulfotransferase 1A1 (SULT1A1) gene causes Arg213His amino acid change and consequently results in significantly reduced enzyme activity and thermostability. We thus hypothesized that the variant SULT1A1 allele may protect against the risk of lung cancer related to tobacco smoking. To examine this hypothesis, we analyzed 805 patients with lung cancer and 809 controls for this polymorphism in a hospital-based, case-control study. We observed that, compared with the GG genotype, the variant SULT1A1 genotype (638GA or AA) was associated with a significantly increased risk for overall lung cancer [odds ratio (OR) 1.85; 95% confidence interval (CI) 1.44-2.37]. Stratification analysis showed that the increased risk of lung cancer related to the variant SULT1A1 genotypes was more pronounced in younger subjects and limited to smokers but not non-smokers [OR 2.28 (95% CI 1.66-3.13) versus OR 1.35 (95% CI 0.91-1.99); P for homogeneity = 0.000]. Furthermore, the risk of lung cancer for the variant genotypes was increased consistently with cumulative smoking dose, with the ORs being 1.66 (95% CI, 0.75-3.68), 2.28 (95% CI, 1.47-3.54) and 3.35 (95% CI, 1.71-6.57) for those who smoked <15 pack-years, 15-36 pack-years and >36 pack-years, respectively (P for trend = 0.000). When analysis was stratified by histological subtypes of lung cancer, consistent results were observed for all three major types of the cancer, i.e. squamous cell carcinoma, adenocarcinoma and other types. Our results, which are against the original hypothesis, demonstrate that the variant SULT1A1 638A allele is associated with susceptibility to lung cancer in relation to tobacco smoking.

Early esophageal squamous cell carcinoma detected by esophageal iodine staining can be easily treated by endoscopic mucosectomy, and identifying its predictors is important in better selecting candidates to screen for this high-mortality cancer. The common etiologies of elevated mean corpuscular volume (MCV) and esophageal cancer, including folate deficiency, smoking, drinking and high acetaldehyde exposure, suggest testing MCV as such a predictor. Japanese alcoholic men with (n = 65) and without (n = 206) esophageal squamous cell carcinomas, excluding those with liver cirrhosis, were assessed for MCV within 7 days of their last drink, alone or in combination with findings from either the alcohol flushing questionnaire or genotyping to identify inactive aldehyde dehydrogenase-2 (ALDH2*1/2*2) and the less-active form of alcohol dehydrogenase-2 (ADH2*1/2*1), which pose risks for esophageal squamous cell carcinoma. MCV was higher in cancer patients than in the control group. MCV was higher in both groups in those who were heavier smokers, had lower body mass index (BMI), experienced alcohol flushing, and had ALDH2*1/2*2. After adjusting for age, drinking and smoking habits, BMI and ALDH2/ADH2 genotypes, macrocytosis of MCV > or =106 fl was associated with increased risk for esophageal cancer (OR = 2.75). Men with both MCV > or =106 fl and alcohol flushing had an even higher cancer risk (OR = 5.51). The combinations of MCV > or =106 fl with ALDH2*1/2*2 or ADH2*1/2*1 alone, and both ALDH2*1/2*2 and ADH2*1/2*1 (ORs = 11.44, 21.22 and 319.7, respectively) showed consistently higher risk than the corresponding group with MCV <106 fl (ORs = 7.24, 4.71 and 27.01, respectively). In conclusion, MCV measurement, alone or in combination with the markers of alcohol sensitivity, provides a new means of predicting risk for esophageal squamous cell carcinoma in Japanese alcoholic men.

NAD(P)H: quinone oxidoreductase 1 (NQO1) is an antioxidant enzyme, important in the detoxification of environmental carcinogens. A single base substitution (C --> T) polymorphism at nucleotide 609 (null-allele) of NQO1 gene impairs stability and function of the NQO1 protein. To investigate the association of this NQO1 polymorphism with susceptibility to esophageal squamous cell carcinoma (ESCC), the NQO1 C609T genotypes were determined by PCR-RFLP analysis in 450 patients with ESCC (257 German Caucasians and 193 northern Chinese) and 393 unrelated healthy controls (252 German Caucasians and 141 northern Chinese). Additionally, NQO1 protein expression was determined by immunohistochemistry in a subset of 74 ESCC (50 German, 24 Chinese). A significant difference in NQO1 C609T genotype distribution was observed between Caucasian healthy controls (C/C, 73.4%; C/T, 25.0%; T/T, 1.6%) and Chinese healthy controls (C/C, 34.0%; C/T, 49.7%; T/T, 16.3%) (chi(2) = 68.40, P < 0.001). The NQO1 T/T genotype significantly increased the risk for developing ESCC in both Caucasian subjects (OR = 4.62, 95% CI = 1.54-13.86) and Chinese subjects (OR = 1.81, 95% CI = 1.04-3.15), compared with the combined C/C and C/T genotypes. In Chinese subjects, this increased susceptibility was pronounced in patients with family history of upper gastrointestinal cancers (OR = 2.18, 95% CI = 1.14-4.17). Immunohistochemical analysis showed NQO1 protein expression in 53 carcinomas, whereas 21 carcinomas were negative. Negativity for NQO1 expression correlated strongly with the NQO1 genotype, being present in 8.6% of cases with C/C, 22.2% of cases with C/T and 100% of cases with T/T genotype (chi(2) = 16.60, P < 0.001). In summary, the association of the NQO1 C609T polymorphism with ESCC in genetically distinct populations makes a strong argument for its importance in carcinogenesis of ESCC in the German Caucasian and the northern Chinese population.

The genetic polymorphisms of aldehyde dehydrogenase-2 (ALDH2), alcohol dehydrogenase-2 (ADH2), ADH3, and glutathione S-transferase M1 (GSTM1) influence the metabolism of alcohol and other carcinogens. The ALDH2*1/2*2 genotype, which encodes inactive ALDH2, and ADH2*1/2*1, which encodes the low-activity form of ADH2, enhance the risk for esophageal cancer in East Asian alcoholics. This case-control study of whether the enzyme-related vulnerability for esophageal cancer can be extended to a general population involved 234 Japanese men with esophageal squamous cell carcinoma and 634 cancer-free Japanese men who received annual health checkups. The GSTM1 genotype was not associated with the risk for this cancer. Light drinkers (1-8.9 units/week) with ALDH2*1/2*2 had an esophageal cancer risk 5.82 times that of light drinkers with ALDH2*1/2*1 (reference category), and their risk was similar to that of moderate drinkers (9-17.9 units/week) with ALDH2*1/2*1 (odds ratio = 5.58). The risk for moderate drinkers with ALDH2*1/2*2 (OR = 55.84) exceeded that for heavy drinkers (18+ units/week) with ALDH2*1/2*1 (OR = 10.38). Similar increased risks were observed for those with ADH2*1/2*1. A multiple logistic model including ALDH2, ADH2, and ADH3 genotypes showed that the ADH3 genotype does not significantly affect the risk for esophageal cancer. For individuals with both ALDH2*1/2*2 and ADH2*1/2*1, the risk of esophageal cancer was enhanced in a multiplicative fashion (OR = 30.12), whereas for those with either ALDH2*1/2*2 or ADH2*1/2*1 alone the ORs were 7.36 and 4.11. In comparison with the estimated population-attributable risks for preference for strong alcoholic beverages (30.7%), smoking (53.6%) and for lower intake of green and yellow vegetables (25.7%) and fruit (37.6%), an extraordinarily high proportion of the excessive risk for esophageal cancer in the Japanese males can be attributed to drinking (90.9%), particularly drinking by persons with inactive heterozygous ALDH2 (68.5%). Education regarding these risky conditions in connection with ALDH2 and ADH2 is vitally important in a new strategic approach aimed at preventing esophageal cancer in East Asians.

Patients with squamous-cell carcinoma in the head and neck (HNSCC) often develop second primary esophageal squamous-cell carcinomas (ESCC). In addition, widespread epithelial oncogenic alterations are also frequently observed in the esophagus and can be made visible as multiple Lugol-voiding lesions (multiple LVL) by Lugol chromoendoscopy. Multiple occurrences of neoplastic change in the upper aerodigestive tract have been explained by the concept of 'field cancerization', usually associated with repeated exposure to carcinogens such as alcohol and cigarette smoke. However, the etiology of second ESCC in HNSCC patients remains unclear and acetaldehyde, the first metabolite of ethanol, has been implicated as the ultimate carcinogen in alcohol-related carcinogenesis. We first investigated the relation between second ESCC and multiple LVL in 78 HNSCC patients. Multiple LVL and second ESCC were observed in 29 (37%) and 21 (27%) patients, respectively. All of the second ESCC were accompanied by multiple LVL. This may indicate that episodes of multiple LVL are precursors for second ESCC. We then examined the association of multiple LVL with the patients' characteristics, including genetic polymorphisms of the alcohol metabolizing enzymes, alcohol dehydrogenase type 3 (ADH3) and aldehyde dehydrogenase type 2 (ALDH2). We also investigated acetaldehyde concentrations in the breath of 52 of the 78 patients. All the patients with multiple LVL were both drinkers and smokers. Multivariable logistic analysis showed that the inactive ALDH2 allele (ALDH2-2) was the strongest contributing factor for the development of multiple LVL (odds ratio 17.6; 95% confidence intervals 4.7-65.3). After alcohol ingestion, acetaldehyde in the breath was elevated to a significantly higher level in all patients with the ALDH2-2 allele than in those without it. The high levels of breath acetaldehyde were significantly modified by the slow-metabolizing ADH3-2 allele. These results reveal strong evidence for a gene-environmental interaction between the ALDH2-2 allele and alcohol consumption, for the risk of developing multiple LVL, resulting in the development of second ESCC in patients with HNSCC. Ultimately, increased local acetaldehyde exposure thus appears to be a critical determinant of the phenomenon of 'field cancerization'.

The human OGG1 (hOGG1) gene encodes a DNA glycosylase that is involved in the excision repair of 8-hydroxy-2'-deoxyguanine (8-OH-dG) from oxidatively-damaged DNA. To determine whether hOGG1 plays a role in risk for orolaryngeal cancer, we screened normal orolaryngeal tissue specimens for hOGG1 expression and assessed the role of the hOGG1 Ser326Cys polymorphism in risk for orolaryngeal cancer. hOGG1 expression was determined by reverse transcription-polymerase chain reaction of total RNA from aerodigestive tract tissues, and hOGG1 genotyping was performed by polymerase chain reaction-restriction fragment length polymorphism analysis of buccal cell DNA isolated from 169 Caucasian orolaryngeal cancer cases and 338 race-, sex- and age-matched controls. hOGG1 mRNA was detected in all aerodigestive tract tissues tested including tonsil, tongue, floor of mouth, larynx and esophagus. Significantly increased risk for orolaryngeal cancer was observed for both the hOGG1 326(Ser)/326(Cys) (odds ratio [OR] = 1.6, 95% confidence interval [CI] = 1.04-2.6) and hOGG1 326(Cys)/326(Cys) (OR = 4.1, 95% CI = 1.3-13) genotypes. Although no significant difference in risk for orolaryngeal cancer was observed for hOGG1 genotypes in never-smokers, increased risk for orolaryngeal cancer was observed for subjects with the homozygous polymorphic hOGG1 326(Cys)/326(Cys) genotype in smokers (>100 cigarettes lifetime; OR = 4.8, 95% CI = 1.3-18). Similarly, although no association was observed in never drinkers of alcohol, significantly increased risk was observed for the hOGG1 326(Cys)/326(Cys) genotype in alcohol drinkers (>1 shot/week; OR = 6.9, 95% CI = 1.6-29). These results suggest that hOGG1 may play an important role in the repair of 8-OH-dG adducts in the aerodigestive tract and that the hOGG1 Ser326Cys polymorphism plays an important role in risk for smoking- and alcohol-related orolaryngeal cancer.

Oesophageal cancer is one of the most common and lethal malignancies in the world. Despite many efforts, treatment is still ineffective for most cases; thus, the development of preventive strategies is crucial for decreasing the burden presented by this disease. Environmental factors, particularly nitrosamines, are thought to be involved in the genesis of oesophageal tumours, and knowledge about the expression of enzymes capable of activating pre-carcinogens in human oesophagus is very important for the development of preventive measures. We analysed the expression of CYP1A1, CYP1A2, CYP2A6/2A7, CYP2E1 and CYP3A4 mRNA in oesophageal mucosa of 50 patients by semi-quantitative RT-PCR. In five patients, who suffered from squamous cell carcinoma, we measured Nnitrosodimethylamine and N-nitrosodiethylamine metabolism in normal and tumorous tissue. CYP2A6/2A7 mRNA was expressed in 61% and CYP2E1 mRNA in 96% of the patients, but in the latter a lower degree of inter-individual variation was observed. These enzymes were expressed either in the distal or middle portions of the oesophagus of 90% of the patients. CYP1A1, CYP1A2 and CYP3A4 mRNA expression was not detected in any portion of the oesophagus. Oesophageal microsomes activated N-nitrosodimethylamine with a low degree of inter-individual variation and microsomes prepared from the tumour of a patient who strongly expressed CYP2A6/2A7 mRNA activated N-nitrosodiethylamine. We conclude that the human oesophagus expresses CYP2A6/2A7 and CYP2E1 and can activate nitrosamines. Notably, the expression of these enzymes is preferentially localized to the most common sites where tumours arise.

We determined inactivation of the CDKN2A (p16(INK4a) and p14(ARF)) gene in 21 cases of oesophageal squamous cell carcinoma (OSCC). The tumours were also analysed for mutations in exons 5-8 and allelic losses in the p53 gene. In addition, we screened the CDKN2B (p15 INK4b), CDKN2C (p18 INK4c), CDK4 and p53R2 genes for mutations in the tumour tissues. Besides concomitant alterations in the CDKN2A and p53 loci in more than half of the cases, our results showed that in 18 OSCC (86%) the CDKN2A (p16(INK4a) and p14(ARF) ) gene was affected through mutations, homozygous/hemizygous deletions and promoter hypermethylation. Eight out of 10 tumours with mutations or promoter hypermethylation specific to the CDKN2A/p16 INK4a gene showed loss of the wild-type allele. One tumour with a single base deletion in the N-terminus (codon 8) of the CDKN2A/p16(INK4a) gene carried a novel germ-line mutation or a rare polymorphism (Ile51Met) in exon 2 of the CDK4 gene. Promoter hypermethylation in the CDKN2A/p14 ARF gene was detected in 11 tumours. In the p53 gene 15 mutations were detected in 14 tumours. We detected an inverse relationship between CDKN2A/p16 INK4a inactivation and frequency of loss of heterozygosity at the p53 locus (OR 0.09, 95% CI 0.01-0.98; Fisher exact test, P-value approximately 0.03). Screening of nine exons of the p53R2 [Human Genome Organisation (HUGO) official name RRM2B] gene resulted in identification of a novel polymorphism in the 5' untranslated region, which was detected in four cases. Our results suggest that the CDKN2A (p16(INK4a) and p14(ARF) ) and p53 genes involved in the two cell cycle pathways are major and independent targets of inactivation in OSCC.

Checkpoint kinase 2 (hCHK2/hCds1) is a tumor suppressor gene involved in cell-cycle control. A hCHK2/hCds1 polymorphism in codon 84 (A-->G at nucleotide 252) was recently identified in Li-Fraumeni syndrome patients. Because cell cycle regulates DNA repair that is associated with cancer risk, we hypothesized that this new polymorphism exists in the general population and is associated with cancer risk. To test this hypothesis, we evaluated the role of this polymorphism in a case-control study of 215 non-Hispanic white patients with newly diagnosed squamous cell carcinoma of the head and neck (SCCHN) and 229 frequency-matched cancer-free controls. We found that the hCHK2/hCds1 codon 84 variant was rare and less frequent in non-Hispanic white cases (0.0186) than in controls (0.0437; P = 0.033). Although no variant homozygotes were detected in these cases and controls, heterozygosity protected against SCCHN, representing a 60% reduction of risk (adjusted odds ratio = 0.40; 95% confidence intervals, 0.17-0.93) compared with wild-type homozygotes. The variant allele was also rare in other ethnic groups (0.0487, 0.0095 and 0.0541 in 115 African Americans, 105 Hispanic Americans and 111 native Chinese, respectively), and only one variant homozygous individual (a Chinese subject) was identified. These results suggest that this hCHK2/hCds1 codon 84 polymorphism is rare and may have a protective role in the aetiology of SCCHN in non-Hispanic whites. Larger studies are warranted to confirm this finding and further mechanistic studies are needed to understand biological relevance of this polymorphism.

The Ah receptor (AhR) is a ligand-dependent transcription factor that positively regulates the expression of the CYP1A1 gene. We investigated the genetic polymorphisms of the AhR gene including the promoter, and examined the link between these polymorphisms, CYP1A1 inducibility and the lung cancer incidence. The AhR promoter region and the 11 exons of 30 subjects were screened. Among the three polymorphisms found, two [(2417)(A/G) ((157)G/A)] have never been described previously. The (1721)(G/A) and (2417)(A/G) are localized in exon 10 and lead to Arg(554)Lys and Met(786)Val substitutions, respectively. The other polymorphism was found in the 5'-untranslated region, resulting in the substitution of a G by an A at position 157 (157)(G/A). To evaluate the frequency of this allelic variant found, a DNA library of a case-control study of lung cancer (162 controls and 177 patients) was studied. There is no significant association between (1721)(G/A), (157)(G/A) and lung cancer: (1721)(G/A) and (157)(G/A) were detected at the same allele frequency of 0.086 and 0.25, respectively in both controls and patients. (2417)(A/G) was found in only one control of 100 (allele frequency 0.005). Statistical analysis did not show any relationship between both (1721)(G/A) and (157)(G/A) polymorphisms found and CYP1A1 inducibility. Considering the rareness of the (2417)(A/G) allelic variant we were not able to evaluate its association with inducibility. In conclusion, none of the polymorphisms were found to play a key role in the CYP1A1 inducibility or in the susceptibility to develop lung cancer.

Although tobacco smoke has been established as a main risk factor in the development of head and neck squamous cell cancer (HNSCC), genetic polymorphisms of xenobiotic metabolizing enzymes are supposed to modulate an individual's susceptibility to smoking-related HNSCC. N-acetyltransferase (NAT) 1 gene is known to be polymorphic and its protein product is implicated in the activation and detoxification of carcinogens, such as aromatic amines, present in tobacco smoke. We developed a rapid and reproducible LightCycler-assisted real-time polymerase chain reaction (PCR) for NAT1 genotyping, which allowed the parallel differentiation of NAT1*3, *4, *10 and *11 alleles and separately of NAT1*14 and *17 alleles within 60 min without the need for further post-PCR processing. In order to investigate the role of the NAT1 gene polymorphism as a risk-modifying factor in HNSCC, we tested for the presence of NAT1*3, *4, *10, *11, *14 and *17 alleles in a case-control study of 291 HNSCC patients and 300 healthy controls of Caucasian origin. Our findings suggest that in Caucasians, the risk of HNSCC is not associated with NAT1 polymorphism. The overall distribution of NAT1 allele frequencies was not significantly different among cases and controls. The presence of the fast acetylator NAT1*10 and NAT1*11 alleles did not significantly increase the risk of HNSCC and no modifying effect of NAT1*10 was observed among smokers. This new approach in NAT1 genotyping substantially increases throughput of sample analysis and, therefore, enhances opportunities to study NAT1 as a risk factor in different cancers in large-scale studies.

Squamous cell carcinoma of the head and neck (SCCHN), including the oral cavity, pharynx and larynx, is an excellent tumor model to evaluate gene-environment interactions, including alcohol and alcohol-metabolizing enzymes such as alcohol dehydrogenase (ADH). We conducted a hospital-based case-control study including 182 cases with newly diagnosed SCCHN and 202 controls with non-neoplastic conditions of the head and neck that required surgery. The joint effects of lifetime alcohol use and the presence of the ADH3 'rapid' allele (ADH3(*)1) was evaluated in relation to the risk of SCCHN. Logistic regression was used to estimate the interaction between alcohol use and ADH3 genotype with adjustment for tobacco use, age, sex and race. The interaction was evaluated on both the multiplicative and additive scales. The risk of SCCHN was increased nearly 6-fold with consumption of 40 or more alcoholic beverages per week [odds ratio (OR) = 5.9; 95% confidence interval (CI) = 2.0-17.7; adjusted for age, sex, race and years of tobacco use]. We did not find any increase in risk for ADH3*1 homozygotes (OR = 0.9; CI = 0.4-1.9) or heterozygotes (OR = 0.8; CI = 0.4-1.7) relative to ADH3(*)2 homozygotes. There was no suggestion of an interaction between any alcohol use variable and the ADH3(*)1 genotype. For example, the interaction term, including the continuous variable average number of drinks per week and the ADH3 genotypes, was non-significant (P = 0.22). The study does not indicate an important role for the ADH3 (*)1 polymorphism in SCCHN, but larger numbers are needed to more precisely estimate the interaction, if any, with ADH3.

The impact of genetic polymorphisms in CYP1A1 on susceptibility to lung cancer has received particular interest in recent years since this enzyme plays a central role in activation of major classes of tobacco carcinogens. Several polymorphisms in the CYP1A1 locus have been identified and their genotypes appear to exhibit population frequencies that depend on ethnicity. We have assessed the role of CYP1A1 genotype in lung cancer risk in the Chinese population via a case-control study. Three polymorphisms, m1 (MSP:I), m2 (exon 7 Ile-->Val) and m4 (exon 7 Thr-->Asn), were determined by PCR-RFLP in 404 controls and 217 lung cancer cases. While no polymorphic alleles were detectable in the m4 site among our study subjects, the allele frequencies for CYP1A1 m1 and CYP1A1 m2 were found to be 35.6 and 25.6% among controls, compared with 42.6 and 34.2% among cases. Multivariate analysis showed an elevated risk for lung cancer in subjects having at least one m1 allele [odds ratio (OR) = 2.0, 95% confidence interval (CI) = 1.4-2.8] or having at least one m2 allele (OR = 1.9, 95% CI = 1.3-2.7). However, this increased risk was limited to squamous cell carcinoma (SCC), but not adenocarcinoma or other histological types of lung cancer. Stratified analysis indicated a multiplicative interaction between tobacco smoking and variant CYP1A1 m1 genotypes on the risk of SCC. The ORs of SCC for the variant CYP1A1 m1 genotype, tobacco smoking and both factors combined were 2.8, 9.1 and 29.9, respectively. When the data was stratified by the pack-year values, this joint effect was consistent and stronger among the heaviest smokers. The interaction between tobacco smoking and the variant CYP1A1 m2 genotypes followed the same pattern. Our findings support the conclusion that CYP1A1 m1 and CYP1A1 m2 polymorphisms are associated with smoking-related lung cancer risk in Chinese.

Squamous cell carcinomas (SCC) induced in hamster buccal pouch (HBP) by 22 weeks of topical N-methyl-N-benzylnitrosamine (MBN) treatment (twice-weekly, 10 mg MBN/ml propylene glycol) were evaluated for: (i) altered expression of p53 using immunohistochemistry (IHC); (ii) mutations in Ha-ras and p53 using PCR/single strand conformation polymorphism (SSCP); (iii) telomerase activity using the telomerase repeat amplification protocol (TRAP). Precancerous lesions were also evaluated using p53 IHC. Hamsters were killed for lesion analysis at either 3 days (group A, eight hamsters, 89 carcinomas) or 7 weeks (group B, six hamsters, 105 carcinomas) following the final MBN application. Between 3 days and 7 weeks post-treatment the proportion of tumors exhibiting p53 IHC activity (at least 10% of nuclei stained using D07 antibodies for detection of both mutant and wild-type p53) fell from 91 to 50%. However, during this same post-treatment period the frequency of tumors analyzed exhibiting confirmed sequence alterations in the conserved exons (E5-E8) of p53 remained constant (5/15 = 33% in group A versus 14/45 = 31% in group B). Heightened expression of wild-type p53 resulting from DNA damage in the immediate post-treatment period is likely to have contributed to the high proportion of group A tumors exhibiting p53 IHC activity. Nearly 80% of the identified p53 mutations were G-->A and C-->T transitions. The identified p53 point mutations occurred at or near (within three codons) of the corresponding hot-spot codons (175, 245, 248 and 273) of human oral SCC. The proportion of group A and group B tumors analyzed exhibiting Ha-ras mutations was 1/15 (7%) and 7/45 (16%), respectively. Only four of the observed eight Ha-ras mutations occurred in codons known to result in activation of this gene. Telomerase activation was demonstrated in 11 of 13 group A tumors (85%) and in 23 of 24 (96%) group B tumors analyzed. The alterations in p53, Ha-ras and telomerase activity observed in this HBP-MBN model are similar in many respects to those observed in the analogous human lesions of the head and neck. This model may be particularly useful for development of cancer chemoprevention regimens and multimodality cancer therapies.

Polymorphic genes for the peroxide scavenger glutathione peroxidase I (GPX1) and 8-hydroxydeoxyguanosine (8-OHdG) DNA glycosylase/apurinic (AP) lyase (hOGG1) map to loci on chromosome 3p which are subject to frequent loss of heterozygosity (LOH) in lung tumours. Levels of the pro-mutagenic, oxidative DNA lesion 8-OHdG, were measured in 37 paired normal and tumorous lung specimens using HPLC with electrochemical detection. Lung tumours were also analysed for 3p LOH by fluorescent PCR with Genescan analysis. No significant difference was observed between 8-OHdG levels in tumour [7.7 +/- 6.7 (mean +/- SE) 8-OHdG/10(6) 2'-deoxyguanosine (dG)] and normal (8.1 +/- 8.8 8-OHdG/10(6) dG) lung tissue. Adduct levels in normal lung tissue DNA were not associated with constitutive hOGG1 genotype although there was a trend towards lower 8-OHdG levels in individuals possessing the ALA6 GPX1 polymorphism. Lung tumours exhibiting 3p LOH (40%) contained higher levels of 8-OHdG adducts (10.9 +/- 2.6 8-OHdG/10(6) dG) (P = 0.05) and lower GPX1 enzyme activity [45.5 nmol glutathione (GSH)/min/mg] (P = 0.09) when compared with tumours without LOH at these sites (5.55 +/- 0.87 8-OHdG/10(6) dG and 63.6 nmol GSH/min/mg, respectively). In conclusion, tumours with 3p LOH at loci associated with hOGG1 and GPX1 appear to have compromised oxidative defence mechanisms as measured by reduced GPX1 enzyme activity and elevated 8-OHdG levels and this may affect the prognosis of lung cancer patients.

UV light is a complete carcinogen, inducing both basal and squamous cell skin cancers. The work described uses the selective COX-2 inhibitor celecoxib to examine the efficacy of COX-2 inhibition in the reduction of UV light-induced skin tumor formation in hairless mice. UVA-340 sun lamps were chosen as a light source that effectively mimics the solar UVA and UVB spectrum. Hairless mice were irradiated for 5 days a week for a total dose of 2.62 J/cm(2). When 90% of the animals had at least one tumor, the mice were divided into two groups so that the tumor number and multiplicity were the same (P < 0.31). Half of the mice were then fed a diet containing 1500 p.p.m. celecoxib. Tumor number, multiplicity and size were then observed for the next 10 weeks. Ninety-five percent of the tumors formed were histopathologically evaluated as squamous cell carcinoma. COX-2 expression and activity were increased in tumors. After 10 weeks, the difference in tumor number and multiplicity in the drug-treated group was 56% of UV controls (P < 0.001). The results show that the orally administered selective COX-2 inhibitor celecoxib prevents new tumor formation after the onset of photocarcinogenesis and suggest that treatment with celecoxib may be very useful in preventing UV-induced skin tumors in humans.

Poly(ADP-ribose) polymerase (PADPRP) is a nuclear DNA-binding enzyme that can modulate chromatin structure close to DNA replication, recombination and repair regions. Two-allele polymorphism on the PADPRP chromosome 13 pseudogene has been studied in several ethnic subpopulations, and the association of each allele with different types of cancer has been investigated. To study the frequency of the allele in the context of lung cancer, we performed a PCR assay for the PADPRP polymorphism in 288 lung cancer patients and 292 matched controls and examined the frequency of the alleles in different ethnic groups. Our results showed that the allele distribution was significantly different among members of different ethnic groups. Specifically, the A allele was dominant in Mexican-American and Caucasian groups but not in the African-American group. The frequencies of the B allele in Mexican-American, Caucasian and African-American controls were 0.184, 0.218 and 0.606, respectively, with the Caucasian cases and controls showing an almost identical lower B allele frequency (0.199 in cases versus 0.218 in controls), and the African-American cases and controls showing an almost identical but considerably higher frequency (0.578 in cases versus 0. 606 in controls). In contrast, the Mexican-American cases and controls exhibited a considerable difference in the B allele frequency (0.306 in cases versus 0.184 in controls). When we combined subjects with the AB or BB genotype into a susceptible genotype group and compared them with the AA group using univariate analysis, the susceptible genotype was not shown to be associated with a risk of lung cancer in either the Caucasian or African-American subpopulation but was significantly associated with an increased risk (2.29-fold) of lung cancer in the Mexican-American group. When lung cancer was categorized by histologic type, no elevated risk was noted for squamous cell carcinoma in any ethnicity. However, in Mexican-Americans, susceptible genotypes were associated with significantly increased risks of adenocarcinoma (3. 21-fold) and large cell carcinoma (10.79). Our study and others have demonstrated that the PADPRP polymorphism may modify an individual's susceptibility to certain cancers. Assessment of the interaction between genetic constitution and environmental exposure might expand our understanding of carcinogenesis and enhance our ability to evaluate the populational cancer risk.

Risk of oral cancer has been associated with exposure to tobacco smoke, alcohol and with genetic predisposition. The aromatic amines and their metabolites, a class of carcinogens present in tobacco smoke, undergo metabolism (activation or detoxification) through an N- or O-acetylation pathway by the polymorphic enzymes, N-acetyltransferases (NAT)1 or NAT2. The genes that encode these enzymes, NAT1 and NAT2, have a variety of high and low activity alleles and we analyzed these genetic polymorphisms in 62 oral squamous cell carcinoma cases, and 122 healthy control subjects from Japan. NAT1 alleles tested were NAT1*3 (C1095A), NAT1*4 (functional reference allele), NAT1*10 (T1088A,C1095A), NAT1*11(9 bp deletion), NAT1*14 (G560A), NAT1*15 (C559T) and NAT1*17 (C190T). No low activity alleles (NAT1*14, NAT1*15 and NAT1*17) were observed in these Japanese subjects. We observed significantly increased risk [odds ratio 3.72; 95% confidence interval (CI) 1.56-8.90; P < 0.01] associated with the NAT1*10 allele, an allele that contains a variant polyadenylation signal. Stratifying by smoking status we found odds ratios of 5.9 (95% CI 1.13-30.6; P < 0.05) for non-smokers with the NAT1*10 allele and 3.1 (95% CI 1.09-9.07; P < 0.05) for smokers, but these risks were not significantly different from each other. Thus, we did not observe that NAT1*10 alleles confer differential risk among smokers and non-smokers. NAT2 rapid acetylation genotype was not a significant risk factor for oral cancer in this Japanese study population. This is the first study to test for oral cancer risk associated with polymorphism in the NAT1 and NAT2 genes, and these positive findings in our pilot study, while based on small numbers, suggest that the NAT1*10 allele may be a genetic determinant of oral squamous cell carcinoma among Japanese people.

Extensive documentation has validated the role of UV irradiation as a tumor initiator and promoter, inducing both squamous and basal cell carcinomas. Human epidermis is a tissue which undergoes active metabolism of arachidonic acid to prostaglandins which is regulated by the action of prostaglandin H synthase (also known as cyclooxygenase). One mechanism for the promotional activity of UV light may involve its ability to induce prostaglandin formation. Work in our laboratory has demonstrated that acute exposure of human keratinocytes to UVB irradiation results in increased production of prostaglandin E2 (PGE2). When cultured human keratinocytes were examined after irradiation with 30 mJ/cm2 UVB in vitro, Western blot analysis showed a 6-fold increase in COX-2 protein which was evident at 6 h and peaked 24 h after irradiation. Furthermore, when human subjects were irradiated on sun-protected skin with up to four times their minimal erythema dosage (MED) and biopsied 24 h later, upregulation of COX-2 protein expression was observed via immunofluorescence microscopy. RNAase protection assays supported this observation, showing induction of COX-2 message which peaked at approximately 12 h following irradiation in vitro. Furthermore, human squamous cell carcinoma biopsies exhibited strongly enhanced staining for COX-2 protein via immunohistochemistry and Western analysis when compared to normal non-sun-exposed control skin. Together, these data demonstrate acute upregulation of COX-2 via UVB irradiation and suggest the need for further studies of COX-2 expression as a potential pharmacological target mediating human skin tumor development.

Extracts prepared from tissue specimens of normal, non-tumourous human buccal mucosa, and cultured buccal epithelial cells and fibroblasts, exhibited O6-methylguanine-DNA methyltransferase (MGMT) activity by catalysing the repair of the premutagenic O6-methylguanine lesion in isolated DNA with rates of 0.2 to 0.3 pmol/mg protein. An SV40 T antigen-immortalized buccal epithelial cell line termed SVpgC2a and a buccal squamous carcinoma line termed SqCC/Y1, both of which lack normal tumour suppressor gene p53 function, exhibited about 50 and 10% of the MGMT activity of normal cells, respectively. The normal, experimentally transformed and tumourous buccal cell types showed MGMT mRNA levels which correlated with their respective levels of MGMT activity. Exposure of buccal cell cultures to various organic or water-based extracts of products related to the use of tobacco and betel quid, decreased both cell survival (measured by reduction of tetrazolium dye) and MGMT activity (measured subsequently to the exposures in cellular extracts). Organic extracts of bidi smoke condensate and betel leaf showed higher potency than those of tobacco and snuff. An aqueous snuff extract also decreased both parameters, whereas an aqueous areca nut extract was without effect. The well-established sulph-hydryl-reactive agent Hg2+, a corrosion product of dental amalgam, served as a positive control and decreased MGMT activity following treatment of cells within a range of 1-10 microM. Taken together, significant MGMT activities were demonstrated in buccal tissue specimens and in the major buccal mucosal cell types in vitro. Lower than normal MGMT activity in two transformed buccal epithelial cell lines correlated with decreased MGMT mRNA and lack of functional p53. Finally, in vitro experiments suggested the potential inhibition of buccal mucosal MGMT activity by complex mixtures present in the saliva of tobacco and betel nut chewers.

In order to clarify the expression of cytochrome P450 and glutathione S-transferase in human esophagus, 41 samples of human esophagus with squamous-cell carcinoma were investigated by immunoblot analysis and enzyme assays. Cytochrome P450 1A2/1 was clearly expressed in microsomes, and the amount in samples with tumorous tissue was significantly greater than that in samples without tumourous tissues or in liver; cytochrome P450 2B6 and 3A4/3 were expressed polymorphically. Aryl hydrocarbon hydroxylase activity was detected in microsomes and was greater in samples from smokers than non-smokers. Patients who both smoked and drank alcohol, however, had activity similar to that of patients without these habits. Glutathione S-transferase M1 and A1/2 protein existed polymorphically in cytosol, and glutathione S-transferase P1-1 was detected in all samples. The frequency of expression of the glutathione S-transferase A1/2 protein was greater in patients with M1 protein than in those without; no difference in the expression was seen for glutathione S-transferase P1-1. Neither smoking nor drinking influenced the expression or activity of glutathione S-transferase. Our data support the idea that some carcinogens can be directly activated or inactivated in human esophageal epithelium.

Glutathione S-transferase (GST) enzyme activity, GST isoenzyme composition and glutathione (GSH) concentration were assessed in normal and squamous cell carcinoma specimens of 14 patients with oral or oropharyngeal cancer and 11 patients with laryngeal cancer. Comparing malignant with normal oral/oropharyngeal tissues, no significant differences in GSH content, GST enzyme activity or isoenzyme composition were found. However, some tumours had up to 3-fold increased GST enzyme activities and 11 malignant samples over-expressed GST-pi. GST-pi was present in all normal and malignant oral/oropharyngeal specimens investigated, whereas class alpha and class mu were detected in only a few samples. GST-mu was present in 28% of the patients with oral/oropharyngeal tumors as compared with approximately 60% in the normal population. GST-alpha, -mu and -pi were detected in 91, 64 and 100% of the normal laryngeal tissues respectively. In laryngeal tumours significantly higher levels of GST-pi and GSH but significantly lower amounts of GST-alpha were detected. Levels of class mu GST were generally lower in cancerous tissues, but differences were not significant. In comparison with normal oral/oropharyngeal tissues, normal laryngeal tissues contained almost twice the amount of GST enzyme activity due to higher class alpha enzyme levels. It is concluded that GST-pi is elevated in 11 out of 14 tumours of the oral cavity and values are significantly increased in tumours of the larynx, which may contribute to the inherent anti-cancer drug resistance of these malignancies. In laryngeal tumours the increased GSH levels may confer additional resistance to radiation therapy.

The factors that determine progression of cervical intra-epithelial neoplasia (CIN) to squamous cell carcinoma (SCC) are unknown. Cigarette smoking is a risk factor, suggesting polymorphism at loci that encode carcinogen-metabolizing enzymes such as glutathione S-transferase (GSTT1, GSTM1) and cytochrome P450 (CYP2D6) may determine susceptibility to these cancers. We have studied the frequency of the null genotype at the theta class GSTT1 locus in women with low-grade CIN, high-grade CIN and SCC. The control group comprised women with normal cervical pathology suffering menorrhagia. We found the frequency of GSTT1 null in the control and case groups was not significantly different, though frequency distributions of combinations of the genotype with smoking in mutually exclusive groups in the high-grade CIN group and the other case groups were significantly different. Interactive effects of GSTT1 null with the GSTM1 null and CYP2D6 EM genotypes, and cigarette smoking were also studied by comparing the multinomial frequency distributions of these factors over mutually exclusive categories. These showed no significant differences between the controls and SCC or low-grade CIN. Frequency distributions in high-grade CIN, however, were significantly different to the controls, and both SCC and low-grade CIN; frequency distributions of GSTT1 null with smoking and CYP2D6 EM, individually and in combination, were significantly different. However, inspection of our data does not indicate that GSTT1 null is a major factor mediating risk. Thus, comparison of chi 2 values for the differences between frequency distributions in high-grade CIN and other groups shows that values for combinations of GSTT1 null with other factors are lower than those for equivalent combinations with smoking and CYP2D6 EM. Interestingly, the combination GSTT1 null/GSTM1 null did not appear to influence susceptibility to CIN or SCC.

Genetically based differences in metabolism, related to MspI restriction site and Ile-Val polymorphisms of the cytochrome P450 (CYP) 1A1 gene and the null genotype of glutathione transferase class mu (GSTM1), have been reported to be associated with lung cancer susceptibility. The present study was set up to establish the frequencies of the polymorphic genotypes of CYP1A1 and GSTM1 in Sweden, to evaluate a possible increased incidence of the genotypes associated with higher lung cancer risks among Swedish lung cancer patients and to try to make a combined risk estimate for carriers of multiple risk alleles. In a healthy control group, all under 66 years of age, 53% (174/329) of the subjects were of the GSTM1(-) genotype, while in a hospital control group 49% (39/79) carried the GSTM1(-) genotype. In the investigated lung cancer patients this genotype was found in 56% (165/296) and among those patients diagnosed before 66 years of age the deficient genotype was found in 60% (78/131). The highest proportion of the GSTM1(-) genotype was found in patients diagnosed with adenocarcinoma (63%, 29/46) and small cell carcinoma (72%, 21/29) before 66 years of age and among female squamous cell carcinoma patients (79%, 15/19). The allelic variants in CYP1A1 were equally distributed in lung cancer patients and controls. The m1/m2 and m2/m2 genotypes of the MspI site and the Ile/Val genotype were, however, slightly over-represented in squamous cell carcinoma patients. Among patients with squamous cell carcinoma diagnosed before 66 years of age the m1/m2 genotype was found in 28% (10/36), whereas the same genotype was observed in 16% (52/329) of healthy control subjects. A combined risk of squamous cell carcinoma was indicated for patients, diagnosed before 66 years of age, carrying both GSTM1(-) and m2 alleles (OR = 3.0, 95% CI = 1.2-7.2).

The purpose of this study was to determine whether exposure of chemically transformed golden Syrian hamster oral epidermoid carcinoma cell (HCPC-1) cultures to smokeless tobacco extract (STE) is associated with a decrease in specific angiotensin I-converting enzyme (ACE) activity and whether this decrease potentiates bradykinin-induced cell growth. We found that STE induced a significant concentration- and time-dependent decrease in ACE activity in cultured HCPC-1 cells (P < 0.05). STE alone had no significant effect on cell number. Bradykinin alone induced a slight, but significant, increase in cell number (P < 0.05). These effects were significantly potentiated by STE (P < 0.01). We conclude that STE potentiates bradykinin-induced HCPC-1 cell growth, in part by attenuating specific ACE activity in these cells.

This study investigated the value of an immunohistochemical demonstration of placental glutathione S-transferase (GST-P) and a histochemical demonstration of gamma-glutamyl transpeptidase (GGT) for the detection of putative preneoplastic lesions and squamous cell carcinomas in the Syrian golden hamster buccal pouches treated with 7,12-dimethylbenz[a]anthracene (DMBA). The results showed that GST-P foci appeared earlier than GGT foci, that the GST-P foci were much more numerous than GGT foci and that most of the GGT positive foci were GST-P positive. All tumors stained strongly positive for GST-P. Only 62.5% of tumors stained for GGT and the staining was usually weak and involved only the superficial layer of the epithelium or keratin, suggesting that the enzyme activity in the basal cells had decreased with formation of neoplasms. The length percentage of basal cells that were GST-P positive increased rapidly during the experiment and by week 12, 25% of the basal cells exhibited GST-P staining. Such rapid expansion of the putative initiated cells may explain the early development of chemically induced invasive squamous cell carcinoma (weeks 10-12) in 100% of hamsters.

To better understand drug and carcinogen metabolism pathways in head and neck squamous cell carcinoma we assayed the principal drug- and carcinogen-metabolizing enzyme systems in both tumors and their corresponding adjacent non-tumoral tissues. Cytochromes P450 (1A1/A2, 2B1/B2, 2C8-10, 2E1, 3A4), epoxide hydrolase and glutathione S-transferases (GST-alpha, GST-mu, GST-pi) were assayed by immunoblotting. GST activity, total glutathione, UDP-glucuronosyltransferase, beta-glucuronidase, sulfotransferase and sulfatase, were determined by spectral assays. Results showed the absence of all probed cytochromes P450 in tumors and non-tumoral tissues, including P450 1A1/1A2 known to be involved in tobacco-related carcinogenesis. No statistical difference was noted between tumors and adjacent non-tumoral tissues for most enzymes studied (GST-alpha, GST-mu, GST-pi, GST activity, UDP-glucuronosyltransferase, beta-glucuronidase, sulfotransferase and sulfatase). However, total glutathione concentrations were significantly higher (P < 0.05) in tumors (47 +/- 20 nmol/mg protein) than in non-tumoral tissues (19 +/- 9). On the contrary, epoxide hydrolase was significantly less expressed in tumors (18 +/- 9 micrograms/mg protein) compared to corresponding non-tumoral tissues (37 +/- 9). These data provide new information concerning human head and neck cancer biology that could possibly have clinical implications.

The identification of genetic traits that predispose individuals to environmentally induced cancers is one of the challenges in the assessment of individual cancer risk. For this reason, individual variations in the expression of enzymes involved in biotransformation reactions have been extensively studied. One such polymorphic enzyme is GSTM1, which belongs to the class Mu of glutathione S-transferases (GSTs), and is only expressed in 55-60% of Caucasians. Previous data suggest that smokers lacking GSTM1 activity may be at greater risk of developing lung cancer. In this study, we used a polymerase chain reaction-based method to examine this issue in a Finnish study population. We found that 44% of a control group of 142 individuals lacked the GSTM1 gene, i.e. they had the GSTM1 null genotype; the rest were either homozygous or heterozygous for the expressed GSTM1 alleles. In a group of 36 patients with non-neoplastic pulmonary diseases, an identical distribution was observed. However, among 138 lung cancer patients the distribution of the GSTM1 genotypes deviated from that found in the healthy controls (53% nulled; odds ratio 1.5, 95% confidence interval 0.9-2.3). Furthermore, when the lung cancer patients were analysed by tumour type, a statistically significant increase in the GSTM1 null genotypes (62%; n = 71) was seen in the squamous cell carcinoma group, with an odds ratio of 2.1 (95% confidence interval 1.2-3.8). These data support the suggestion that GSTM1 null genotype may act as a risk modifier in lung cancer.

Forty-eight human squamous cell lung carcinomas of previously untreated patients were analyzed for resistance to doxorubicin and for the presence of topoisomerase II (Topo II), metallothionein (MT), thymidylate synthase (TS) and catalase (Cat). Significant correlations exist between resistance to doxorubicin measured by the in vitro short-term test and overexpression of MT and TS measured by immunohistochemistry. No significant correlation was found between resistance and expression of Topo II or Cat. No significant interrelationship between smoking habits of patients and expression of Topo II, MT, TS or Cat was found.

The level of the DNA repair enzyme O6-alkylguanine-DNA alkyltransferase (MGMT) was examined in benign and malignant skin tumors induced with different initiating and promoting agents and from both SENCAR and Sensitive SENCAR Inbred (SSIN) mice. The MGMT levels in the tumors were approximately one-half the level observed in normal surrounding epidermis and in keratinocytes from untreated controls. In addition, a carcinoma-producing cell line, VT 17DT, derived from papillomas in SENCAR mice had no detectable MGMT activity (Mer- phenotype), whereas in the non-tumor forming line, 3PC, MGMT activity was comparable to that in papillomas. The comparatively low level of MGMT in papillomas may contribute to their ease of conversion to squamous cell carcinomas by N-ethyl-N-nitrosourea or n-methyl-N'-nitro-N-nitrosoguanidine. MGMT activity was also determined in the epidermis of non-exposed mice of various stocks and strains. Epidermal MGMT activity was similar to levels in the corresponding livers and was, in general, parallel with stock/strain susceptibility to tumor formation. This is the first report that examined MGMT activity in skin tumors and normal keratinocytes in the mice of several stocks and strains.

The activities of several enzymes involved in reactive oxygen production and detoxification were quantified in murine skin during the ontogeny of chemically induced skin cancer. Relative to solvent-treated controls, the specific activities of epidermal superoxide dismutase (SOD), catalase (CAT) and glutathione peroxidase (GPX) were reduced approximately 45, approximately 60 and approximately 24% respectively, 24 h after the fourth or tenth topical application of 1 microgram of 12-O-tetradecanoylphorbol-13-acetate (TPA) to the dorsal skin of SENCAR mice. The specific activity of epidermal xanthine oxidase (XO) increased approximately 350% during the same period. SOD and CAT specific activities in papillomas and carcinomas generated in an initiation-promotion protocol were approximately 15 and approximately 40% respectively of the activities measured in age-matched, non-treated mice. CAT and SOD activities were also significantly suppressed in the skin adjacent to the papillomas for several weeks following the cessation of TPA promotion, but eventually recovered to the levels measured in age-matched controls. XO specific activities in papillomas and squamous cell carcinomas (SCC) were approximately 85-350% greater than the activities determined in skin adjacent to the tumors. The increases in XO and the decreases in SOD and CAT activities measured in the tumors were independent of continued treatment with TPA, and thus characteristic of the tumor phenotype. GPX activities in papillomas were comparable to normal, untreated skin, but reduced approximately 22-41% in SCC. Collectively, these studies demonstrate that TPA orchestrates changes in the activities of several enzymes involved in reactive oxygen metabolism that are characteristic of the papilloma and SCC phenotype.

The extensive metabolizer phenotype of debrisoquine has been associated with increased risk of lung cancer, and it has been proposed that a molecular test for this phenotype is feasible. DNA restriction fragment length polymorphisms of the human debrisoquine 4-hydroxylase gene locus (CYP2D6), and the metabolic phenotype for debrisoquine have been studied in a group of healthy volunteers, a group of lung cancer patients and two control groups (chronic obstructive pulmonary disease patients and patients with cancers at sites other than the lung). Confirmation of four distinct XbaI allelic fragments (44, 29, 16/9 and 11.5 kb), previously identified among caucasians, was obtained. The 29 kb alleles were the most frequently observed in both poor and extensive metabolizers of debrisoquine. Alleles of 44 kb were found with approximately equal frequency among both poor and extensive metabolizers. The data are consistent with the hypothesis that the 11.5 and 44 kb fragments are associated with mutant alleles of the CYP2D6 gene, but the power of phenotype prediction by these alleles was less than that previously reported for a European (Swiss-German) population. Similarly, the data also show that 8% of 29 kb homozygotes are poor metabolizers (indicating that at least 28% of 29 kb fragments are also associated with mutant alleles) and are not therefore informative for predicting the debrisoquine phenotype. The 16/9 allele may represent either wild-type or mutant alleles. Restriction fragments of 44 kb were found more frequently among cancer patients and chronic obstructive pulmonary disease patients (30%) than among the healthy volunteer group (7%). Genotypes observed were not related to lung tumor histology. Furthermore, at least three EcoRI alleles were found to be in linkage disequilibrium with the 'mutant' 44 kb allele. These data suggest that the 44 kb allele can comprise three distinct haplotypes, in contrast to studies of a European population. These studies indicate that no single mutant CYP2D6 allele as determined by EcoRI appears to be associated with lung cancer, despite the findings that these patients are invariably of the extensive metabolizer phenotype.

The highly toxic environmental contaminant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) is a potent carcinogen and tumor promoter, and affects cellular proliferation and differentiation both in vivo and in vitro. This report presents data showing that TCDD enhances the proliferation of two human squamous carcinoma cell lines in monolayer culture by inhibiting growth arrest at high cell density. SCC-15G and SCC-25 cells were treated with 0-100 nM TCDD in culture medium, and examined for changes in proliferation and differentiation. TCDD stimulated increases in cell number and DNA synthesis of both cell lines, and inhibited differentiation of SCC-15G cells in a dose-dependent manner. The minimum effective concentrations for increases in proliferation were 0.1 nM in SCC-15G cells and 1 nM in SCC-25G cells. The saturation density of SCC-15G cells grown in 10 nM TCDD was approximately double that of untreated controls, while the saturation density of SCC-25 cells was 50% above controls. TCDD-induced increases in proliferation were detectable only in cells exposed at subconfluent density, then assayed after control cultures had reached high-density growth arrest. There was no difference in cell number or DNA synthesis between control and TCDD-treated cultures when cells were both treated and assayed during the logarithmic phase of growth, nor in cultures treated after the cells had reached high density growth arrest. Therefore, TCDD-induced proliferation resulted from failure of treated cells to undergo normal density-dependent growth arrest rather than from direct mitogenic stimulation of the cells. Differentiation (envelope competence and keratin staining) of SCC-15G cells was inhibited by TCDD, despite the fact that in these cultures cell density was twice that of the controls. The sensitivity of SCC-15G cells to modulation of growth and differentiation by TCDD provides the basis for a model to examine the biological mechanisms of TCDD-induced alterations in proliferation and differentiation of epidermal cells.

Retinoic acid and retinyl acetate induce tissue transglutaminase to high levels in cultured SCC-4 keratinocytes, increasing the enzyme specific activity over 50-fold under optimal conditions. Pretreatment of the cells for a day with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), 3-methylcholanthrene (3-MC) or benzo[a]pyrene almost completely prevented the induction observed upon subsequent treatment with retinoic acid for 2 days. Similar aromatic compounds that do not induce aryl hydrocarbon hydroxylase (pyrene, dibenzofuran) did not exhibit this suppressive effect. The concentration dependence on TCDD for induction of aryl hydrocarbon hydroxylase was nearly identical to that for its suppression of transglutaminase induction, with half-maximal effects observed at approximately 20 pM in each instance. Similarly, the concentrations of 3-MC giving half-maximal stimulation of the hydroxylase and suppression of the transglutaminase were comparable (0.9 and 0.3 microM, respectively), although this agent was almost five orders of magnitude less potent than TCDD. These observations reveal a loss of cellular sensitivity to vitamin A mediated by the Ah receptor.

Expression of glutathione S-transferase placental form (GST-pi) in human lung carcinoma tissue taken at autopsy or biopsy was investigated immunohistochemically. All of 34 cases of squamous cell carcinomas, including poorly, moderately and well-differentiated examples were shown to stain positively for GST-pi. Poorly differentiated adenocarcinomas were, however, negatively stained (0/5 cases), while moderately and well differentiated adenocarcinomas were found to stain with GST-pi at rates of 69% (9/13 cases) and 71% (5/7 cases), respectively. Six cases of small cell carcinomas examined were all negative. The results indicate that GST-pi may be a useful marker for non-small cell type lung cancer, especially squamous cell carcinoma which is in agreement with findings for rat lung neoplastic lesions reported previously.

Selenium, glutathione peroxidase, glutathione reductase and glyoxalase I have been measured in normal and neoplastic human adult lung tissues. Interindividual variations of enzyme activities and selenium content in both tumour and non-tumour tissues were considerable. From the measurements of glutathione peroxidase activity with both hydrogen peroxide and cumene hydroperoxide it was deduced that human tumour and non-tumour lung tissues are devoid of the selenium-independent enzyme. In general, a significant increase in the activity of glutathione peroxidase and glutathione reductase was found in tumour. Glyoxalase I in tumour was as high as in non-tumour samples. Mean selenium concentration tended to be higher in tumour than in non-tumour specimens. When a comparison was made between normal and neoplastic tissue of the same individual, glutathione peroxidase, activity was found to be higher in tumour in 19 cases out of 24 and glutathione reductase in 17 out of 22. In 15 cases out of 18 the selenium levels were found to be higher in tumour. It was concluded that changes in the factors involved in anti-oxidative protection actually occur in human lung tumour tissues.

Immunohistochemical staining using anti-rat glutathione S-transferase placental form (GST-P) rabbit antibody and enzyme histochemical staining for gamma-glutamyltranspeptidase (gamma-GT) were investigated in lesions appearing during lung carcinogenesis induced by N-nitrosobis(2-hydroxypropyl)amine (BHP) in rats. Rats were given BHP at a concentration of 2000 p.p.m. in drinking water, and were killed after 12 weeks of BHP intake, after 12 weeks of BHP intake followed by 12 weeks of tap water intake or after 20 weeks of continuous BHP intake. It was found that bronchiolo-alveolar hyperplasias, adenomas, adenocarcinomas, squamous metaplasias and squamous cell carcinomas had been induced by BHP. All of the squamous metaplasias and squamous cell carcinomas were shown to stain with GST-P but not with gamma-GT. On the other hand, the hyperplasias, adenomas and adeno-carcinomas stained with gamma-GT to various degrees and in different areas, but did not stain with GST-P. The incidence of gamma-GT phenotype and the average percentage of gamma-GT-positive areas in hyperplasias and adenomas suggested that adenocarcinomas might develop from hyperplasias and adenomas. These results suggest that GST-P is a marker for squamous lesions while gamma-GT is a marker for adenomatous lesions in rat lung carcinogenesis. Furthermore, squamous metaplasias appear to be preneoplastic lesions of squamous cell carcinomas while gamma-GT-positive hyperplasias or adenomas are preneoplastic lesions of peripheral adenocarcinomas.