Research Project
6654496
DESCRIPTION (provided by applicant): Racial and ethnic disparities in the incidence of preterm labor have been well documented, but the relative impact of genetic predisposition, as compared to environmental and behavioral factors is not known. The objective of this proposal is to test the hypothesis that differential gene expression contributes to disparity in infection-associated preterm labor. A second objective is the creation of two large and novel resources for studying preterm labor: (1) a database of candidate critical genes, and, (2) a human gene expression tissue bank. We suggest that elucidating the pathophysiology of preterm labor requires the power to perform analyses of the fetomaternal interactions required for parturition on a genomic scale. For practical and ethical reasons, a primary study of this type in humans would be severely limited. Therefore, in the first phase of this project, we will use a validated, tightly controlled murine model of infection-induced preterm labor to generate a comprehensive catalogue of gene expression over time in multiple tissues. Preterm pregnant mice will be randomized to treatment groups modeling one of clinical conditions: A. infection with labor; B. infection without labor; C. labor without infection; and D. no infection/no labor. RNA will then be collected from myometrium, ovaries, decidua, placentas and fetal membranes from each of the groups in a time series, and the relative expression of thousands of genes will be analyzed in these samples using DNA microarrays. A similarity metric and a clustering algorithm will be used to categorize individual genes by temporal expression patterns. A novel subtractive strategy will allow us to differentiate transcripts active specifically in infection-induced from those important for infection or labor alone. Inferences will be drawn regarding the involvement of genes not physic represented in the micro arrays by mapping into known functional pathways. The final result of this analysis will be a "short list" of candidate genes with potentially critical roles in infection-induced labor. In the second phase of the project, we analyze human tissues (myometrium, placenta, chorion, amnion, decidua, amniocytes and blood) collected from approximately 1360 patients presenting in term and preterm labor in a cross-sectional case-control study matching for race. Expression analysis in these tissues will be targeted to the human homologues of the enriched list of candidate critical genes identified in the mouse. Differential gene expression in pre term labor with or without infection and premature rupture membranes, as well as among different racial groups, will be characterized. A computational tool known as a support vector machine will be used to generate a rank order list predictive of preterm delivery, taking into account historical, clinical and laboratory variables. This tool will identify the subset(s) of genes most likely to form the genetic basis of preterm labor, and may provide a diagnostic molecular profiling instrument for predicting preterm delivery. The large databases generated in this study will be valuable to any researcher interested in parturition and will be made accessible on the Internet.
