Development of an impedance assay-based model highlighting the crucial role of hypermucoviscosity in hypervirulent K. pneumoniae pathophysiology

Background Hypervirulent Klebsiella pneumoniae (HvKp) cause pyogenic liver abscesses (PLA) in otherwise healthy people. Molecular surveillance of HvKp reveal that the global dissemination of HvKp poses an urgent threat to human health. The key microbiological property associated with the hypervirulence is capsule production resulting in the hypermucoviscous phenotype necessary for infection. Whilst pathogenesis occurs in the liver, HvKp are thought to colonise the gut before causing disease. We sought to establish the pathogenic process as current studies provide no definitive causes to explain the switch from “gut-located HvKp” and “pathogenic liver abscess HvKp”. Methods Using HvKp strains, SGH10 and SGH10-p (loss of the virulence plasmid resulting in a no hypermucoviscosity) and Ecl8 (neither hypercapsulated or hypermucoviscous), we infected Caco2 (ileocolic cells), HepaRG (hepatocytes and cholangiocytes) and HUVEC (endothelial cells) cell lines. To dissect the cell-pathogen interactions, we employed the impedance- based assay, ECIS (Electrical cell-impedance system), to measure the variations in cellular electrical resistance in the presence or absence of pathogen. Pathogen-mediated damage when measured at low or high frequency using ECIS reflects damage to the integrity of the intracellular tight junctions or general cellular integrity respectively. Transepithelial (TEER) assays and western blot confirmation of the levels of tight junction scaffold protein, Zonula Occludens-1 (ZOI) were undertaken to confirm the ECIS data. Results SGH10 adhered and internalised less significantly to Caco2, HepaRG and HUVEC cell lines than SGH10-p and Ecl8. However, ECIS analyses shows that, despite being less adherent and internalised, SGH10 had a significantly greater effect on Caco2 cells by producing TJ disruption at least 2-3hrs relative to SGH10- and Ecl8. As expected, this trend was also mirrored in HepaRG. No differences were noted with HUVEC cells. TEER analyses and western blot analyses for ZO1 confirm the ECIS data in Caco2 cells where SGH10 results in significantly lower ZO1 TJ protein expression relative to SGH10p- and Ecl8. Conclusions Our data demonstrates that SGH10 elicits rapid damage of gut epithelial Caco2 cells by disrupting TJ interactions. This suggests that HvKp do not simply “colonise” the gut to gain access to the hepatic environment, providing new insights into the gut-liver axis in the pathophysiology of HvKp.