William P. Bennett

Molecular epidemiology of human cancer risk: gene-environment interactions and p53 mutation spectrum in human lung cancer.

The p53 tumour suppressor gene is at the crossroads of a network of cellular pathways including cell cycle checkpoints, DNA repair, chromosomal segregation, and apoptosis. These pathways have evolved to maintain the stability of the genome during cellular stress from DNA damage, hypoxia, and activated oncogenes. The high frequency of p53 mutations in human cancer is a reflection of the importance of p53 involvement in this network of pathways during human carcinogenesis. An electronic database containing p53 mutations from more than 9000 cancers (http://www.iarc.fr/p53/homepage.html) can be used to generate hypotheses for further clinical, epidemiological, and laboratory investigations. For example, one can hypothesize that (a)p53 mutations vary in their pathobiological significance; (b) cellular content influences the selection of p53 mutations in clonally derived cancers; (c) the location and type of mutation within the p53 gene provide clues to functional domains in the gene product; and (d) the p53 mutation spectrum can be a molecular link between aetiological agents and human cancer. This review will focus on the role of p53 and cancer susceptibility genes in the molecular pathogenesis and epidemiology of human lung cancer. Copyright

Five missense variants in the amino-terminal domain of the glucocorticoid receptor: no association with puerperal psychosis or schizophrenia.

Steroid hormone administration causes behavior changes in many and psychosis in a few. The clinical features suggest that genetic variants of the glucocorticoid receptor or cofactors could produce susceptible subpopulations who react adversely to hormonal cascades. To investigate this possibility, coding and splice site sequences of the glucocorticoid receptor were scanned for single nucleotide polymorphisms in genomic DNA samples from 100 schizophrenics (86 Caucasians and 14 African-Americans) and 40 Caucasians with puerperal psychosis. Five amino acid substitutions were found in the amino-terminal domain at frequencies of 0.6 to 3.8% in Caucasians: R23K, F29L, L112F, D233N, and N363S. In addition, four silent nucleotide changes were found: E22E, K293K, D677D, and N766N; a transversion in intron 4 occurred beyond the splice junction. None of these variants can be linked to these disorders at present. However, the N363S variant contributes a new potential phosphorylation site and has been associated with increased body mass and reduced bone mineral density [Huizenga et al., 1998], so it is possible that the other missense variants confer traits that currently are unrecognized. Comparisons to natural glucocorticoid receptor mutants in the familial glucocorticoid resistance syndrome and steroid resistant leukemias suggest that amino acid substitutions at highly conserved residues may cause severe functional defects and serious illness, while changes at less conserved sites produce lesser alterations and milder disease.

Elevated frequency of ATM gene missense mutations in breast cancer relative to ethnically matched controls

Studies of families of patients with ataxia telangiectasia (A-T) show an increased risk of breast cancer in heterozygous A-T carriers. However, expected increased levels of mutations in the ATM gene among unselected breast cancer patients have not been found to date. Previous methods of mutation detection were biased toward the detection of truncating mutations, and single nucleotide substitutions were likely to have been underreported. In this study, genomic DNA from 43 breast cancer patients and 43 control individuals were scanned for mutations in the entire ATM coding region (exons 4-65) and adjacent intronic splice regions (three megabases total) using detection of virtually all mutation-single-strand conformation polymorphism (SSCP), a modification of SSCP with sufficient redundancy to detect virtually all mutations. Excluding a polymorphism found commonly in cases and controls, there were missense changes in 12 breast cancer patients, one of whom also had a protein truncating mutation, versus six controls (P=0.09). When all structural changes common to the cases and controls were excluded, missense or truncating changes were found in 10 cases compared to two in controls (P=0.013). The background of missense changes in controls is high. There is a trend towards elevation of all structural changes in cases, but the results are not statistically significant. Cohort-specific structural changes are significantly more prevalent in the breast cancer patients. The data are compatible with certain missense mutations in ATM predisposing to breast cancer.

Radon, secondhand smoke, glutathione-S-transferase M1 and lung cancer among women

Tobacco smoke and ionizing radiation induce oxidative stress by transmitting or generating reactive oxygen species (ROS). We hypothesized that glutathione‐S‐transferase M1 (GSTM1) null homozygotes would have decreased ability to neutralize ROS that might increase their susceptibility to lung cancer. A case‐only design was used with lung cancer cases pooled from 3 previously completed case‐control studies using archival tissue samples from 270 lung cancer cases to genotype GSTM1. Radon concentrations were measured with long‐term α‐track radon detectors. Secondhand smoke (SHS) was measured with questionnaires and interviews. Unconditional logistic regression was used to calculate the interaction odds ratios (OR) and 95% confidence intervals (95% CI). Radon concentrations >121 Bq m−3 were associated with a >3‐fold interaction OR (OR = 3.41; 95% CI = 1.10, 10.61) for GSTM1 null homozygotes compared to GSTM1 carriers; the linear trend was significant (p trend = 0.03). The SHS and GSTM1 interaction OR was also elevated (OR = 2.28; 95% CI = 1.15–4.51) among never‐smokers. This may be the first study to provide evidence of a GSTM1 and radon interaction in risk of lung cancer. Additionally, these findings support the hypothesis that radon and SHS promote neoplasia through shared elements of a common pathway.

Cigarette smoking, body mass index, gastro-esophageal reflux disease, and non-steroidal anti-inflammatory drug use and risk of subtypes of esophageal and gastric cancers by P53 overexpression.

A number of risk factors for esophageal and gastric cancers have emerged, yet little is known whether risk factors map to molecular tumor markers such as overexpression of the tumor suppressor TP53. Using a US multicenter, population-based case–control study (170 cases of esophageal adenocarcinomas, 147 gastric cardia adenocarcinomas, 220 non-cardia gastric adenocarcinomas, and 112 esophageal squamous cell carcinomas), we examined whether the risk associated with cigarette smoking, body mass index (BMI), gastroesophageal reflux disease (GERD), and non-steroidal anti-inflammatory drug (NSAID) use varied by P53 overexpression. We defined P53 overexpression through immunohistochemistry of paraffin-embedded tumor tissues, using cutpoints based on percent of cells positive. Polytomous logistic regression was used to assess differences between each case group (defined by tumor subtype and P53 expression) and the control group by risk factors. The proportion of cases overexpressing P53 by tumor subtype was 72% for esophageal adenocarcinoma, 69% for gastric cardia adenocarcinoma, 52% for non-cardia gastric adenocarcinoma, and 67% for esophageal squamous cell carcinoma. For most tumor subtypes, we found little difference in risk factors by tumor P53 overexpression. For non-cardia gastric cancer however, an association with cigarette smoking was suggested for tumors that do not overexpress P53, whereas larger BMI was related to adenocarcinomas that overexpress P53 versus no overexpression. Overall, this study did not find a clear relationship between P53 protein overexpression and the known risk factors for subtypes of esophageal and gastric cancers. Further research on these tumors is needed to identify molecular markers associated with variations in the risk factor profiles.

EGFR Somatic Mutations in Lung Tumors: Radon Exposure and Passive-smoking in Former- and Never-smoking U.S. Women

Background: Patients with lung cancer with mutations in EGF receptor (EGFR) tyrosine kinase have improved prognosis when treated with EGFR inhibitors. We hypothesized that EGFR mutations may be related to residential radon or passive tobacco smoke. Methods: This hypothesis was investigated by analyzing EGFR mutations in 70 lung tumors from a population of never and long-term former female smokers from Missouri with detailed exposure assessments. The relationship with passive smoking was also examined in never-smoking female lung cancer cases from the Mayo clinic. Results: Overall, the frequency of EGFR mutation was 41% [95% confidence interval (CI), 32%–49%]. Neither radon nor passive-smoking exposure was consistently associated with EGFR mutations in lung tumors. Conclusions: The results suggest that EGFR mutations are common in female, never-smoking lung cancer cases from the United States, and EGFR mutations are unlikely due to exposure to radon or passive smoking. Cancer Epidemiol Biomarkers Prev; 21(6); 988–92. ©2012 AACR.

An in-frame deletion in the alpha(2C) adrenergic receptor is common in African--Americans

α2 adrenergic receptors are activated by adrenaline and noradrenaline, and three subtypes (ie, A, B, C) have differential affinities for antagonists and medications. The α2c adrenergic receptor (ADRA2C), located on chromosome 4p16.3, is a candidate gene for schizophrenia because it binds clozapine, an atypical neuroleptic useful for treatment-resistant schizophrenia. In addition, ADRA2C binds clonidine which is prescribed for three psychiatric diseases. This report communicates the findings of the genetic scanning of this gene of very tough GC content. The complete coding sequences and splice junctions were scanned with [DOVAM]-S in 104 schizophrenics, and pilot probes of patients with alcoholism (41 patients), cocaine abuse (25 patients), puerperal psychosis (30 patients), attention deficient/hyperactivity disorder (25 patients) and autism (25 patients). Six sequence variants were found, including five silent polymorphisms (allele frequencies 0.6–25%) and an in-frame deletion of a homologous repeat at nucleotides 967–978 (ie, TIDRU1). Genotyping of the normal two repeat unit of the Third Intracytoplasmic Domain Repeat Unit (TIDRU2) and the deleted variant (TIDRU1) revealed that TIDRU1 had allelic frequencies of 39% (11/28) and 3.5% (6/172) in African–American and Caucasian schizophrenics, respectively, and it occurred with equal frequency in controls (44%, 31/70 and 3.0%, 6/198). TIDRU1 occurs at a location similar to the third intracytoplasmic 48-nucleotide repeat unit in the DRD4 that is associated with ADHD. Although these data do not suggest an association of TIDRU1 with schizophrenia, additional studies are needed to see whether TIDRU1 confers a clinical phenotype.