Cargando…

2595. Murine Models for the Host Response to Typical and Atypical Pneumonia

BACKGROUND: The etiology of pneumonia is difficult to diagnose, with typical bacterial, atypical bacterial, and viral infections being the most common. However, diagnostics that discriminate these infectious etiologies are limited. We, therefore, focused on the host response to identify possible dia...

Descripción completa

Detalles Bibliográficos
Autores principales: Khan, Kirin, Betancourt-Quiroz, Marisol, Zaas, Aimee K, Treece, Amy E, Que, Loretta G, Henao, Ricardo, Suchindran, Sunil, McClain, Micah T, Woods, Chris W, Ginsburg, Geoffrey S, Tsalik, Ephraim L
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Oxford University Press 2019
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6810146/
http://dx.doi.org/10.1093/ofid/ofz360.2273
Descripción
Sumario:BACKGROUND: The etiology of pneumonia is difficult to diagnose, with typical bacterial, atypical bacterial, and viral infections being the most common. However, diagnostics that discriminate these infectious etiologies are limited. We, therefore, focused on the host response to identify possible diagnostic markers and better understand these infections. However, atypical bacterial pneumonia is challenging to identify in humans precisely because of this diagnostic difficulty. Therefore, we utilized murine models to define host response differences between typical bacterial, atypical bacterial, and viral pneumonia. METHODS: Mice were intranasally inoculated with S. pneumoniae (n = 38), M. pneumoniae (n = 27), H1N1 pr8 (n = 19), or saline as a control (n = 42). RNA was extracted from peripheral blood collected at 24, 48, 72, 120, or 168 hours and subjected to microarray analysis. Diagnostic signatures were generated using lasso logistic regression and accuracy was assessed using nested leave-one-out cross-validation with feature selection repeated within each iteration. Differentially expressed genes were used to perform gene set enrichment analysis. These murine-derived signatures were externally validated in silico in 487 human subjects found across 5 publicly available data sets. RESULTS: We generated pathogen-specific murine disease signatures that performed with 91–100% accuracy. Pathway analysis revealed that animals with pneumococcal pneumonia had a robust immune response by 48 hours that continued to 72 hours post-infection. In contrast, animals infected with M. pneumoniae did not show evidence of a strong immune response until 72-hours post-infection. Additionally, the immune response to M. pneumoniae bared greater similarity to the viral response than it did to the host pneumococcal response. H1N1-infected mice showed an anti-viral response at 120 hours that resolved by 168 hours post-infection. The AUC values resulting from independent human validation of our murine signatures ranged from 89 to 98%. CONCLUSION: There are discrete host responses to typical bacterial, atypical bacterial, and viral etiologies of pneumonia in mice. These signatures validate well in humans, highlighting the conserved nature of the host response to these pathogen classes. [Image: see text] [Image: see text] DISCLOSURES: Ephraim L. Tsalik, MD MHS PhD, Immunexpress: Consultant; Predigen, Inc.: Officer or Board Member, Research Grant.