Prediction of gene expression levels and the role of cis-acting elements in age-related cataract by applying a promoter-based modeling approach

Abstract

Cataract is a dynamical process of lens opacity formation involving many inter- and intracellular regulations, as well as metabolic genes and transcription factors. Using a series of microarray-derived mRNA profiles for human cataractogenesis (Hawse et al. Mol. Vision 2003, 9, 515-537), we develop a promoter-based system-theoretic modeling to demonstrate model-driven prediction of gene expression levels and to identify the role of critical cis-acting elements. In this study, 14 key mRNA expression data from the structural and pathological molecules of age-related cataract samples are used. The first seven genes consist of structural molecules, and the second half of genes are composed of heat shock proteins, filensin, and glutathione peroxidase 3. The presented result demonstrates that mRNA expression levels of structural proteins such as crystallins can be successfully predicted from 5' flanking regulatory DNA sequences. In addition, predicted gene expression levels of heat shock protein, beta-tubulin, and alpha A-crystallin accurately estimate the stimulatory or inhibitory role of distributed cis-acting elements, i.e., c-Myc, GATA-1, GR, NE-E, and Pit-1. Although it is difficult to predict the overall gene expression levels in cataract samples, the present study shows the potential use of promoter-based modeling and prediction of the gene expression levels for age-related cataract.