Analysis of blood proteome in influenza-infected patients reveals new insights into the host response and signatures distinguishing mild and severe infections

Klaus Schughart^1,2*Stephen C. Threlkeld³Subhashini Sellers⁴William A. Fischer⁴Jens Schreiber⁵Eva Lücke⁵Mark Heise⁴Amber M. Smith^2,6

• ¹Universitat Munster, Münster, Germany
• ²The University of Tennessee Health Science Center Department of Microbiology Immunology and Biochemistry, Memphis, United States
• ³Baptist Memorial Hospital-Memphis, Memphis, United States
• ⁴The University of North Carolina at Chapel Hill, Chapel Hill, United States
• ⁵Otto-von-Guericke-Universitat Magdeburg, Magdeburg, Germany
• ⁶The University of Tennessee Health Science Center Department of Pediatrics, Memphis, United States

Influenza infections result in a wide spectrum of disease outcomes, ranging from asymptomatic cas-es to fatal illness. While immunopathology contributes to an increased risk of hospitalization, the host factors that drive predisposition to ICU admission remain poorly understood. Here, we performed proteome analyses of sera from influenza virus-infected patients who were experiencing moderate disease without ICU admission or severe disease with ICU admission. A unique aspect of our study is that we monitored expression levels of more than 6,000 proteins whereas previous studies only ana-lyzed a very limited number of protein markers. Comparing the responses in infected versus healthy individuals identified many differentially expressed proteins and related molecular pathways involved in lipid metabolism, iron metabolism, chromatin remodeling, and immune signaling in infected pa-tients. These were amplified in patients with more severe disease, where immune signaling, prolifera-tion/differentiation, and metabolic process pathways were increased. Our results suggest strong im-pacts of macrophage- and neutrophil-related responses. A unique aspect of our analysis is that it allowed us to relate the secreted host response in the blood (proteome) with stimulated responses in blood cells (transcriptome) in the same patients. Many differentially expressed proteins in the serum were not identified as differentially expressed genes in blood cells and therefore represent a not yet described set of biomarkers. Furthermore, we identified many strong correlations between blood cell transcriptomes and blood proteomes, which will allow us to validate or generate unique hypotheses of causal relationships between serum proteins and responses in blood cells during an influenza in-fection.