(564e) Gene Network Inference and the Immune Response: A Novel Approach to Identifying Biology's Risk Assessment Strategies

Gene network inference and the immune response: A novel method to identifying biology's risk assessment strategies

            Daily, our immune systems protect us from the various pathogens to which we are exposed. When a pathogen, such as a virus, successfully invades our body, the immune system can employ various biological processes to contain the infection but these processes are also capable of invoking a great deal of tissue damage (e.g., inflammation). Thus, an effective immune response must assess the risk pathogens pose to our health and carefully manage the magnitude of its response to minimize collateral damage. A mismanaged response has been implicated to result in a more sever pathological outcome during virus infections^1–3 and is the underlying cause of allergic reactions⁴. Thus, a better understanding of the mechanisms and strategies behind how the immune system detects a pathogen and manages its response will greatly facilitate the design of immune-intervention therapies and may identify novel, host-based drug targets.

            Recently, our group has applied gene network inference algorithms to dynamic, in vivo gene expression data collected from influenza infected mice to determine which biological processes track the state of the infection, how they communicate with other processes responsible for virus clearance and whether the communication between the various processes is dependent on the virulence of the pathogen. We applied a series of gene set enrichment tests, including our recently developed cell-type enrichment platform, CTen, to identify if a module may be regulated by a particular biological process or if the module represents conserved gene expression resulting from lymphocyte infiltration. Lastly, by analyzing the correlation between gene modules (as opposed to the traditional approach of analyzing within a module to infer gene to gene connectivity), we developed a network abstraction of the immune response to determine how various biological processes are coordinated during infection.

            We found that several major aspects of the immune response, from virus detection events (such as Toll-like receptor signaling) to infection management events (e.g., intracellular death via apoptosis or targeted cell death using lymphocytes), were apparent in microarray data as modules of coexpressed genes. Using flow cytometry, we validated that CTen correctly identified gene expression that is the result of the influx and exodus of macrophages, B cells and T cells.

            Focusing of the dynamics of gene expression in each module, we found that some host pathways employ ultrasensitive mechanisms (low levels of pathogen do not evoke a response, but a strong response occurs after a threshold level is reached). Given the high noise environment in which the immune system operates, we hypothesize that these mechanisms aid in effective noise rejection and play a critical role in the risk assessment apparatus.

            Then, expanding our analysis to the network level, we found that the connectivity pattern between the modules was very consistent with the known behavior of the immune response, but, we also observed several  modules which appear to be critical for the successful transition from the innate to the adaptive immune response. Many of these modules were related to processes which are not generally associated with the immune response while others had no known biological function. We are now performing additional experiments to determine if these apparent interactions represent causal relationships.

            In sum, we present a new approach to profiling the immune response which offers several opportunities for the development of immunomodulatory therapies. We validated that immune cell infiltration can be determined from tissue array data using our CTen platform, identified control strategies employed by the immune response to gauge when an invading pathogens necessitates a response, and found several targets to be considered for targeted intervention.

REFERENCES