Garry Adams

Scholarly Interests

I have long been intrigued with the biological interface between hosts and etiological agents, and the resulting patterns of morphological lesions, especially lesions caused by infectious disease agents. The morphologic, cytokine/chemokine, transcriptomic, proteomic and metabolomic patterns of host and pathogen responses provoke a series of fundamental questions, e.g. what is the molecular pathogenesis of these lesions?; does the host genome manipulate the pathogen or vice versa?; what host and pathogen cellular pathways are perturbed and dysfunctional in the disease processes?; what is the cause of death?; can we apply genomic pathology (the convergence of ‘omics’ and morphology) and systems biology to more fully understand infection biology as the basis for improved prediction of host and pathogen mechanistic genes and pathways critical to health and clinical illness? In response to these questions, my research is focused on the: 1) investigation of the comparative molecular pathogenesis of zoonotic intracellular bacterial pathogens in natural animal models, particularly salmonellosis, brucellosis and mycobacterial diseases, 2) development of vaccines and host gene expression-based diagnostics for zoonotic and select agent caused diseases, and especially 3) development of in silico host:pathogen interactome predictive models based upon bi-directional in vivo host (bovine/murine) and Salmonella enterica Typhimurium interactions at the target organ interface, enteric Peyer’s patches. The vision to develop improved animal models for salmonellosis originated in the early 2000s in association with my collaborators, Drs. Andreas Bäumler and Renee Tsolis. From our very successful and continuing collaboration grew the idea to develop a computational infection biology model based on temporal neonatal calf in vivo microarray-based transcriptomic and proteomic profiling of the acute Salmonella infection process. We expanded our research team by including Drs. Sara Lawhon, Kenneth Drake and Harold ‘Skip’ Garner, focusing on an envisioned systems biology analysis of both host and pathogen comprehensive transcriptomic and proteomic datasets derived from our in vivo biological model. We next computationally fused the datasets based on actual Salmonella proteomic data (performed by Dr. Mary Lipton at Pacific Northwest National Laboratory) and computationally predicted bovine host structural proteins to identify maximum likelihoods of host and pathogen protein:protein interactions as the basis for our preliminary in silico interactome model to predict mechanistic genes and linked perturbed cellular pathways. We then established in vitro phenotypes of S. Typhimurium deletion mutants (provided by Dr. Helene Andrews-Polymenis) of the predicted mechanistic genes for attachment, invasion or survival in murine RAW 264.7 macrophages to confirm almost half of the modeled predictions. Finally, over half of the model-predicted and in vitro-confirmed S. Typhimurium genes were validated to have in vivo phenotypes in our bovine ligated ileal loop model. We subsequently performed siRNA knockdowns in RAW 264.7 macrophages of predicted host proteins interacting with Salmonella proteins and validated in vitro phenotypes in almost half of the predicted host genes. Our lab proposes to significantly enhance the predictability of the in silico interactome model by adding the NEXTGen sequencing transcriptomics, advanced proteomics and metabolomics. By fusing higher quality genomic, transcriptomic, proteomic and metabolomic datasets derived from in vivo S. Typhimurium comparative infection models in mice, calves and non-human primates, we propose to further enhance the predictability and validation of in silico interactome model as a tool for the ultimate practice of “precision human medicine” as envisioned by the 2011 report by the National Academy of Sciences. As our publications document, we have also developed in silico interactome modeling tools for intracellular bacterial pathogens, Brucella melitensis and Mycobacterium avium ssp. paratuberculosis, all based on in vivo datasets. Our team has published 42 peer-reviewed manuscripts on salmonellosis using animal models, confirming our combined abilities to perform complex series of in vivo and in vitro experiments and developing high level in silico predictive models as we propose here. My long term role as co-developer of our in vivo based in silico interactome predictive model and personal collaborations with each of our team of investigators and core directors enables us to envision developing the in silico interactome predictive model for the biomedical research community.