Abstract
BACKGROUND: Bacterial pneumonia remains a leading cause of morbidity and mortality worldwide despite the widespread availability of antibiotics. Novel pneumonia therapies and biomarkers are urgently needed to improve outcomes and advance personalized therapy. Using an established baboon model of S. pneumoniae pneumonia, we sought to characterize the temporal dynamics of pneumonia host responses to identify novel potential diagnostic and therapeutic molecular targets. METHODS: We performed whole blood transcriptomics, unbiased proteomics, and peripheral cytokine measurements serially in baboons inoculated with S. pneumoniae (n = 23) or saline (n = 10) and modeled the peripheral blood host response using principal components analysis and complex sparse logistic regression. Differentially expressed genes were analyzed for pathway analysis. RESULTS: Inoculated animals developed characteristic signs and symptoms of pneumonia. A 39-gene signature was derived that classified S. pneumoniae infection with high accuracy (auROC 0.9 and 0.99 at 24 and 48 h post-inoculation, respectively). Similar performance was observed for 48-h biomarker signatures derived from peripheral blood plasma proteomic and cytokine measurements (both auROC >0.9). The gene signature retained strong diagnostic performance (auROC = 0.88) when transformed to human orthologs and applied to patients with acute respiratory illness (n = 34) or healthy controls (n = 20). Pathway analysis at 48 h identified down-regulation of mitophagy and glucocorticoid signaling in peripheral blood. CONCLUSIONS: We report novel peripheral blood gene and protein expression signatures of S. pneumoniae pneumonia that could improve pneumonia diagnosis and found distinct pathways that may be amenable to modulation. Our findings illustrate how non-human primate models of bacterial pneumonia can successfully translate biomarker discoveries to patients.