The PSIIM Program in Systems Immunology and Infectious Disease Modeling represents a new multidisciplinary research initiative focused on the immune system, with an emphasis on quantitative, computer-based, microscopic and macroscopic modeling of immune functions, integration of these modeling efforts with data sets derived from global analyses of cell components, and the development and application of advanced imaging methods to the analysis of immune responses in vivo in models systems, and ultimately, man. The program is designed to deal with the existing lack of any large-scale effort to understand the engineering of the immune system from the biochemical through the organismal levels and to generate predictive models based on such understanding. The overall goal of the PSIIM will be the development of a new level of integrated understanding of how the immune system functions and how it interacts with pathogens. The primary imperative would be accumulation of the specific information necessary to devise robust quantitative, predictive models of immune behavior in various circumstances, including exposure to infectious agents, following vaccine administration, or in autoimmune diseases. A key effort will be the implementation of the type of measurement rigor needed for effective modeling at all levels of immune organization, from the biochemical to the whole animal. During the past year, the Immunology Team has continued its major project involving a top-down analysis of the immune and non-immune tissue response of mice to various strains of influenza virus. Highly standardized preparations of both mildly pathogenic (Tx91) and highly pathogenic (PR8) viruses have been used at varying infectious doses in a single inbred strain of mouse and several hundred highly qualified microarray transcriptional analyses have been conducted with RNA isolated from infected lung tissues of these mice at varying time points post-inoculation. Early findings suggest marked differences in the types of immune cells and immune factors present at the site of infection at early time points in mice given the Tx91 vs. the PR8 viruses, with some of the data pointing to possible explanations for the marked differences in pathogenicity of the two infectious agents in terms of wound healing responses and tissue damage at the epithelial level. During this past year we have also developed methods for 9+ color flow cytometric separation of hematopoietic cell subsets (lymphoid and myeloid) and non-hematopoietic cells from infected lungs, validated the separation methods as not contributing to substantial changes in RNA patterns, and begun to generate microarray data on these purified subpopulations to increase the resolution of the analysis. In addition, immunohistochemical studies of viral protein expression have been completed on infected tissues, as have flow cytometric studies of the cells present in infected lungs, to correlate with the microarray data. A variety of informatic tools have been employed to organize and analyze the emerging data sets, especially a k-means clustering algorithm that allows study of the RNA response in terms of coregulated gene sets with linked function (modules). Work has begun on generation of viral re-assortments to map the genomic regions involved in the marked pathogenicity difference between PR8 and Tx91, which will enable us to relate viral genetic variation to differences in host response and to begin construction of Bayesian models of the network of interactions involved in each infectious process.
Showing the most recent 10 out of 19 publications
