Pesticide and Gene Interactions in Latina Farm Workers
|Institution:||University of California, San Francisco|
Paul Mills , Ph.D., MPH -
|Award Cycle:||2008 (Cycle 14)||Grant #: 14IB-0032||Award: $160,718|
|Etiology>Environment and gene/environment interactions: nature vs. nurture|
Initial Award Abstract (2008)
Susceptibility to cancer in humans is a function of both environmental factors (including chemical exposures) and constitutional factors (including genetics). Previous studies have suggested, though not proven, that organochlorine (OC) pesticides (which include the now-banned DDT) increase breast cancer risk; additionally common variants of certain genes such as the BRCA genes and carcinogen metabolizing genes may increase breast cancer risk. It is then possible that the genetic make-up of a woman may predispose her to adverse effects from chemical exposures and increase her risk for developing breast cancer.
We plan to assess breast cancer risk in a population of female Hispanics residing in rural central in California who have been heavily exposed to pesticides (including the OCs dieldrin, chlordane and heptachlor) by collecting detailed occupational histories, pesticide exposure information, and saliva (DNA) samples to test for variants in genes that may metabolize these pesticides. We plan to identify and analyze approximately 100 Hispanic females recently (2005-2006) diagnosed with breast cancer and 100 controls to evaluate the relationship between OC pesticide exposure, xenobiotic (i.e., a chemical which is found in a person, but which is not normally produced or expected to be present)-metabolizing genes (CYP1A1, CYP1A2, CYP1B1, GSTM1, GSTT1 and GSTP1), and their interaction on breast cancer risk. We will use a newly developed “Job Exposure Matrix” to assess current and historic pesticide use in the locations where these women have lived and worked. This will be used in the analysis to estimate lifetime exposure to OC pesticides.
The project data is designed to determine, (1) if breast cancer risk is increased among women living and working in areas heavily exposed to OC pesticides, (2) whether variants in pesticide metabolizing genes increase breast cancer risk, and (3) if there is an interaction between OC exposure and common variants of carcinogen metabolizing genes on breast cancer risk. This study is very innovative in that it combines the resources of several agencies in California: the California Cancer Registry (CCR), the Survey Research Group (SRG), Fresno State University, the University of California, San Francisco and the Department of Pesticide Regulation (DPR) in conducting novel breast cancer research. In addition, we propose to use new techniques for identifying important gene by environmental interactions in breast cancer causation.
Final Report (2010)
California is the leading state of agricultural produce in the United States. There has been a wide range of research correlating cancer with pesticide usage. A common class of pesticides, organochlorines (OC), resembles the structure of estrogen and are collectively called xenoestrogens (environmental estrogens). Exposure and accumulation of xenoestrogens are known to have profound effects on women’s health including in utero feminization, breast growth and lactation, and breast cancer. The pathway to xenobiotic elimination from the body is a multi-step process resulting in excretion of contaminants through the urine or bile. The first step involves oxidation which is thought to be primarily carried out by the Phase I enzyme family, cytochrome P450 (CYP), and is typically an activating reaction creating a more polar byproduct. The second step involves conjugation with an endogenous ligand through a Phase II enzyme family, glutathione-S-transferase (GST), and is typically a detoxifying reaction. The goals of this study were to determine the association between exposure to organochlorine pesticides and breast cancer risk in the Hispanic (Latina) female population of the intensely agricultural San Joaquin Valley of California by assessing single nucleotide polymorphisms (SNPs) in select xenobiotic-metabolizing genes. Our study involved the use of DNA samples consented from 55 Hispanic participants and utilized three different molecular strategies. The selection of these polymorphisms was based on the potential of each polymorphism to either alter gene inducibility or enzyme catalytic efficiency through a direct change in the amino acid sequence. We genotyped GSTM1 and GSTT1 (wild type/heterozygous vs. homozygous null deletion). The prevalence of the null genotype in this female Hispanic population was examined among the controls (non-breast cancer) and for GSTM1 the prevalence was 9/16 or 0.56. For GSTT1, the prevalence of the null genotype among the controls was only 2/16 or 0.125. We found no association between the GSTM1 null polymorphism and breast cancer risk in this sample (O.R. = 0.99, 95% CI=0.28, 3.51), but we did find a doubling in breast cancer risk among those women who carried the null polymorphism for GSTT1 (O.R. = 2.21, 95% CI=0.39, 12.63).
For GSTP1, we genotyped two polymorphisms two GSTP1 polymorphisms (codon 105 Ile/Val and codon 114 Ala/Val) . At codon 105, the frequency of the A/G null was 7/16 or 43.7% in the controls, while the frequency of the null G/G was 3/16 or 18.7%. Both of these two null genotypes were associated with two fold elevations in breast cancer risk, the O.R. for the A/G = 2.79 (95% CI=0.58-13.3), and for G/G, the O.R. = 2.5 (95% CI=0.37, 16.9). At codon 114, the frequency of the null C/T genotype was only 2/13, or 15.4%. However, this null genotype was also associated with elevated breast cancer risk, O.R.=2.14, (95% CI=0.38,12.2). We measured two CYP1B1 polymorphisms (codon 119 Ala/Ser and codon 432 Val/Leu). For the polymorphism of the CYP1B1 gene at codon 119, we found no alteration in breast cancer risk (O.R. = 0.77, 95% CI=0.14, 3.70). But at codon 432 (Val → Leu), we found elevated risk of breast cancer, (O.R. 2.33, 95% CI= 0.64, 8.54). The women carrying the Val CYP1B1 allele had higher risk than those women with the Leu/Leu genotype.
Interactions are defined in this analysis as the modification of the pesticide-breast cancer association by gene status. Accordingly, the risk of breast cancer associated with the most commonly encountered pesticide which showed elevations in breast cancer risk, namely Endosulfan, was examined within strata of those genetic polymorphisms listed above. Small numbers precluded definitive results but data were available for Endosulfan exposure and genetic polymorphisms on 26 breast cancer cases and 16 control participants. For the interaction of Endosulfan exposure and GSTM1, for example, the odds ratio for breast cancer associated with Endosulfan in those women with the “wild” type was 1.33, indicting the gene is functioning and there is no elevated risk in this subgroup. However, among those women with the ‘null” genotype, indicating an inablility of the gene to function and to detoxify chemicals, the odds ratio associating Endosulfan with breast cancer was 13.5, a rather large increase in risk. However, due to small numbers, neither point estimate was statistically significant, and confidence intervals were very wide.
Similarly, the interaction between Endosulfan and the GSTT1 gene was evaluated. When the odds ratios were calculated for each strata, the p value associated with the test of homogeneity (no-difference) of the odds ratios was very small, on the order of p<0.01, indicating interaction. However, for the interaction of Endosulfan and the GSTP1A gene, there was no evidence of interaction, although the results are once again based on small numbers. Nor was there any interaction detected between Endosulfan and the CYP1B1 gene (A119S, G>T), where the p-value for homogeneity was 0.45. Finally there was no interaction found between the CYP1B1 gene (L432V) and exposure to Endosulfan, p=0.46.,
This study indicates that it is feasible to identify, trace, consent and recruit female Hispanic women in the San Joaquin Valley of California who have recently been diagnosed with breast cancer. It is possible to obtain comprehensive lifestyle information, including work histories and to construct profiles of pesticide exposure. However, it is challenging to obtain saliva samples from this population and very difficult to obtain written informed consent to use saliva samples to extract DNA for research purposes.