Principal Investigator: Victoria Cortessis, Ph.D., Department of Preventive Medicine, USC

Description: The CYP1A2 enzyme activates numerous known carcinogens, including aromatic amines, heterocyclic amines, nitroaromatic compounds, mycotoxins and estrogens. Expression of CYP1A2, the gene that encodes this enzyme, is induced by additional environmental agents considered risk factors for cancer. Until recently, CYP1A2 measurements available for use in epidemiologic studies have been limited to phenotype assays, which have limited utility.

Objectives: We proposed to conduct a pilot study to learn whether we may use genotype assays for two recently identified CYP1A2 polymorphisms as measurements of CYP1A2 inducibility, so that these genotypes and data describing exposure to inducers could be used as measures of CYP1A2 activity in epidemiologic studies of cancer. We had proposed to also study the relationship between these genotypes and both phenotypic measurements made in disease-free individuals and the occurrence of hepatocellular carcinoma (HCC) in an aflatoxin-exposed population, among whom CYP1A2 activity is a putative risk factor. At the suggestion of the reviewing committee, we added the objective of relating the genotypes to two additional disease outcomes for which CYP1A2 activity is a suspected risk factor: bladder cancer and adenoma of the colon (“polyps”).

Progress: We completed genotyping assays at both loci for 459 polyps cases and 507 polyps controls, 590 bladder cancer cases and 687 bladder cancer controls, 250 HCC cases and 250 HCC controls; a total of 5486 reactions for 2743 individuals. We then conducted several statistical analyses, first confirming that tobacco smoking is associated with higher phenotypic indices among unaffected individuals in the bladder and polyps studies, and that these indices are lower for females, suggesting that documented patterns of gene induction occur in these study groups. Among the same groups, we then looked for patterns of elevated CYP1A2 indices among individuals with genotypes classified as “inducible” by authors of the original papers describing the polymorphisms, accounting for exposure to cigarette smoking and other putative inducers. Although we were disappointed by not finding that inducibility (as measured by phenotypic indices) could be predicted by joint genotypes defined by both polymorphisms, there was a suggestion of associations with one form of the more 5’ polymorphism. We found no interesting association between the genotypes and occurrence of polyps, but one of the genotypes does appear to be associated with risk of bladder cancer. We are currently preparing a manuscript reporting these results, which we expect to submit for publication within the next month. Thereafter, we will conduct statistical analyses to explore any relationship between the genotypes and onset of HCC. These findings will be published along with results of additional studies of effects of putative susceptibility genotypes at other loci, probably with the next three months.

---

Principal Investigator: Frank Gilliland, M.D., Ph.D., Department of Preventive Medicine, USC.

Description: Exposure to elevated levels of air pollution has long been hypothesized to increase the risk of respiratory infections. We have recently reported that a modest increase in ozone substantially increases respiratory illness-related school absence rates. In this application, we propose to begin an investigation of the genetic determinants of susceptibility for respiratory illnesses. ICAM-1 is a candidate gene for susceptibility to respiratory infections because ICAM-1 1) is the receptor for rhinovirus, the most common cause of respiratory infections, and 2) has functional polymorphisms. Furthermore, ozone induces elevated expression of ICAM-1 through increased transcription. Taken together, children with the variant alleles may be most susceptible to increased risk of respiratory illnesses following ozone exposure. To assess this hypothesis, we propose to examine associations of intercellular adhesion molecule- 1 (ICAM- 1) gene polymorphisms with rates of school absences from respiratory infections and determine whether associations of respiratory illness absence rate with ozone are modified by ICAM-1 polymorphisms at codon 29 (exon 1), codon 241 (exon 4) and codon 469 (exon 6). We would accomplish this aim by studying approximately 1000 4th grade participants in the Children’s Health Study, a 10-year longitudinal study of respiratory health. These subjects will be genotyped for ICAM1 polymorphisms using existing specimens. A stratified two-stage time series model will be fitted to the absence count data and included distributed lag effects of exposure adjusted for long-term pollutant levels. The current study will also contribute to the knowledge based by determining ICAM-1 allele frequencies in African Americans, Asians, Hispanic whites, and non-Hispanic whites. This pilot study offers the NIEHS center the opportunity to expand to a new area of research – environmental effects on infectious disease.

Progress Report: To date, we have developed assays for the codon 29 and 241 SNPs and assayed 2914 and 1593 participants respectively. We have had difficulty designing an assay for the codon 469 polymorphism, and are currently testing new approaches. Analyses will be conducted upon completion of the genotyping.

---

Description: The primary goal of our project was to validate the utility of a flexiblelatent class model for the analysis of DNA methylation data. The appeal of the latent class model is the ability to incorporate measured covariatesinto the classification of subjects into subgroups. These covariates couldvary by locus, such as the presence or absence of a CpG island in the promoter region of the gene of interest, or they could vary by subject, such as the subject's age. In a simulation study, we compared the misclassification rates of four approaches to classifying subjects based ontheir DNA methylation pattern. Overall, we observed the lowestmisclassification rate for the latent class model. In addition, the parameter estimates from the model appeared unbiased.

Results: The results werepresented at the American Society of Human Genetics Meeting in San Diego,California, October 2001 Siegmund KD, 2001. Am J Hum Genet, Supplement to Volume 69:414) and were used as preliminary data in a grant titled "Statistical Models in Epigenomics" submitted to NIH October 2001 (1R01CA97346-01).

---

Description: Controversy exists as to whether semen quality is declining throughout the world. Published data support and reject the finding that a significant decrease in sperm concentration has occurred worldwide during the past 30-50 years. These variable findings have led most investigators to conclude that if a decline in semen quality exists, there maybe significant geogrpahic differnces.

Progress report: We have obtained sperm bank data from Northern (n=35) and Southern California (n=50), as well as air quality data and temperature data from the two locations for the time period over which the semen samples were collected(1996-2000). 8513 Los Angeles samples and 5574 Palo Alto samples were evaluated. Air quality and temperature data were matched to the specific geographic grid of the donors’ home address zip codes. Air quality parameters evaluated included daily averages for levels of carbon monoxide, ozone, PM10, and nitrogen dioxide. These data were generously provided to us by Sonoma Tech, Inc( Petaluma, CA). A mixed model linear regression was used to model linear relationships between outcome , air quality, and temperature. Our preliminary results indicate that 1) a 5.9 % decrease in average sperm count per year and a 2.5% decrease in total motile sperm count per year occurred during the time period studied. 2) changes in temperature are not related to changes in sperm parameters, 3) and significant negative associations between sperm concentration and motility and ozone levels occurred 10 days prior to the collection date of the semen sample. This time period corresponds to that period when sperm reside in the epididymis and is compatible with ozone’s hypothesized lipid peroxide-mediated toxicity.

Results: Invesitgator received a Senior Investigator Research Training Award to continue to study this topic. The data generated by the pilot grant were used to support the research proposal.

---