RNA-seq reveals condition-dependent global transcriptional effects of the msf gene in Haemophilus in

Evangeline Williams, BA; Laura Anastor-Walters, MS; Bhaswati Sen, PhD; Steven Lang, BS; Sergey Balashov, PhD; Rachel L. Ehrlich, MS; Garth D. Ehrlich, PhD; Benjamin Janto, PhD; Joshua Chang Mell PhD AFFILIATIONS: Drexel University College of Medicine, Center for Advanced Microbial Processing, Institute for Molecular Medicine and Infectious Disease

Despite advances in modern medicine, severe childhood bacterial infections continue to occur, including infections of the middle ear, or otitis media (OM). The most prevalent otopathogen is Haemophilus influenzae, a human-restricted bacterium that normally colonizes the nasopharynx of healthy children but is also a major cause of OM, as well as other infections. Prior work used a pangenome-wide association study to identify accessory genes associated with virulence using 210 clinical isolates of H. influenzae. A novel virulence factor containing Sel1-like repeats was found, msf (macrophage survival factor), which enhanced pathogenesis in an animal model and increased the survival of bacteria engulfed by macrophages. To further dissect the molecular function of msf, we conducted a global transcriptome analysis by RNA-seq, comparing wild-type (WT) vs. msf∆ mutant bacteria grown across a range of in vitro culture conditions. WT cells had distinct expression profiles at different planktonic growth phases, as expected, whereas biofilm cells showed global decreases in gene expression, except for a small subset of upregulated stress response genes. Remarkably, msf∆ mutants showed only slight differences from WT in planktonic cultures, but biofilm cultures showed dramatically higher expression of genes repressed in WT biofilms. These results indicate that msf plays a role in signaling metabolic dormancy in biofilms. The results also imply that disruption of msf-dependent signaling may inhibit the antibiotic tolerance exhibited by biofilms; further experiments are testing this. We have also now expanded to a pan-transcriptomic approach, comparing gene expression patterns across a panel of diverse clinical isolates (n= 12) chosen to contain the maximize genomic diversity in a minimal number of strains. As expected, strains on the panel express diverse biofilm phenotypes, and new RNA-seq dataset will search for a "core" gene expression signature of H. influenzae biofilm formation that may lead to establishing biomarkers of pathogenic diseases.