Molecular Biology Core

University Of Colorado Denver, Aurora CO

Investigators

Linked publications & trials

Abstract

The overall goal of the Program is to define critical cellular risk factors that lead to the development of ALI in certain individuals and not in others. The Molecular Biology Core will perform Affymetrix oligonucleotide array analysis using the human U95 A array. These arrays will enable the examination of gene expression for approximately 12,000 full-length human genes. The studies are designed to evaluate the range of normal expression in human neutrophils, and the response of normal volunteers to endotoxin. Furthermore, patients both at risk for and with ALI will be examined. Each project is vertically integrated by sharing the first two Specific Aims. Through these specific aims, each project will aid in determining neutrophil responses that will characterize 'high' and 'low' responders to endotoxin. Once accomplished, the Molecular Biology Core will perform expression analysis. By sequentially examining normal volunteers, then those treated with LPS (both peripheral neutrophils and bronchoalveolar lavage (BAL) neutrophils, we will be able to define a range of normal expression patterns of peripheral circulating neutrophils and BAL neutrophils, including in vivo responses to LPS. Patients at risk for ALI, and those with both mild and severe ALI will be examined. The overall design enables the definition of the range of normality of expression patterns, both at baseline and with LPS challenge. Coexpression of genes and cluster analysis will be performed, and the data presented to all members of the PPG in the monthly meetings. These data will then be utilized for follow-up and verification studies, as outlined in the proposal. Verification strategies include: Real time PCR, in situ hybridization, and immunochemistry. Once this database is established, we will be able to further analyze expression data for discrimination analysis. An important goal of this proposal is to discern sets of genes that differentiate between nornal and 'at risk' vs. mild and severe ALI. Building discrimination models has a long history in statistical pattern recognition and machine learning, and there are many methods that might be successfully applied to gene expression data. Discrimination analysis allows for the development of algorithms, which predict, using expression analysis, class distinction. By examining the expression patterns of human neutrophils in patients at risk and diseased states, we hope to identify patterns of gene expression changes, which predict a more favorable outcome to this disease and give further insight into the pathogenesis of acute lung injury.