| Lung transplantation is effective therapy for many individuals suffering from advanced lung disease. However, long-term survival of recipients is limited by the development of chronic rejection, manifested as obliterative bronchiolitis. Acute rejection is a major risk factor for obliterative bronchiolitis. Noninvasive biomarkers of acute rejection and clinically acceptable predictors of obliterative bronchiolitis have not yet been identified. The work described in this thesis employed the use of microarray analysis for the first time to identify patterns of gene expression associated with acute rejection in human BAL samples and in acute and chronic rejection in a murine model of rejection. Two human BAL studies presented represent an ongoing process of exploration, refinement and expansion of data analysis methods, and the continual development of new hypotheses in trying to better understand acute rejection and its subsequent effect on chronic rejection.; An analysis of a heterotopic mouse model of acute and chronic lung rejection confirmed that there is a predictable and stereotypical process that includes acute rejection of an MHC incompatible graft and ultimately results in chronic rejection of that graft. Expression profiles observed were in accordance with the current paradigm of a predictable sequence of events, beginning with airway injury; an innate immune response followed by an adaptive immune response, including both cell-mediated and humoral components; and eventual loss of airway epithelial cells. These observations confirmed and expanded the list of genes and molecular processes that can be studied as potential surrogate markers or targets for intervention of OB.; The effect of mRNA decay on human microarray experiments was explored by combining BAL and peripheral blood expression information with mRNA decay information. There appeared to be a relationship between the hybridization efficiency, as measured by scaling factor, and the overall RNA quality of the sample. This analysis underlined the need to account for arrays with disparate scaling factors and present calls. |
