Abstract:
E. coli Nissle 1917 (EcN) is a probiotic bacterium whose efficacy as therapeutic in inflammatory bowel diseases (IBDs) to prolong remission phases has been clinically proven. IBDs are multifactorial diseases that occur in genetically predisposed hosts who develop an aberrant immune response towards shifted, dysbiotic commensal microbiota. One of many molecular mechanisms by which EcN regulates intestinal microbiota and the immune response to reestablish homeostasis involves its flagellin (FliC). FliC is the monomeric subunit of the bacterial flagellum, a whip-like appendage used for bacterial motility. Additionally, bacterial flagellin is a microbe associated molecular pattern (MAMP) which is recognized by the host via a pattern recognition receptor (PRR), Toll-like receptor 5 (TLR5). TLR5-FliC recognition requires residues in the conserved part of flagellin that are also crucial for bacterial motility by stabilizing the flagellum core. Nevertheless, the flagellin of pathogenic bacteria, such as Campylobacter jejuni can evade the immune response while maintaining motility.
A previous study suggested that FliC deficient EcN fails protect from weight loss induced by dextran sodium sulfate (DSS) in a mouse model of IBD. This failure was linked to TLR5 signaling and specifically implicated a role for the long hypervariable region (HVR) as a discriminatory factor between EcN FliC and FliC of other commensal but not probiotic E. coli strains. The HVR was thought to form the outer sheath covering the flagellar core, which is formed by the conserved domains of the flagellum. Unpublished crystallographic structure determination revealed that the HVR of EcN forms two structurally distinct features that were novel at the time of discovery, namely (1) a flexible but stabilized Linker region connecting the conserved region and the HVR, and (2) an additional D4 domain in the HVR.
Aim of this study was to characterize the influence of the novel features of EcN FliC on TLR5 recognition, motility and its probiotic properties in DSS colitis. To probe this, modifications of FliC to eliminate the distinct structural features (Linker, D4 domain) and mutation of selected residues speculated to uncouple TLR5 recognition and motility as for C. jejuni were generated. These EcN mutants were tested for their ability to stimulate TLR5 on various cell types (reporter cells, epithelial cells, BMDCs, THP-1). Further they were assessed in terms of flagellum formation and motility in soft agar. Lastly, the EcN strains lacking parts or the whole HVR should be tested in vivo in a DSS colitis model for their probiotic properties.
These analyses revealed that the TLR5 recognition and motility could not be uncoupled and that the HVR or its parts are dispensable for TLR5 recognition. Deleting parts or the whole HVR also did not alter the human β defensin production of a colon adenocarcinoma cell line C2BBe1 or cytokine secretion patterns of THP-1 macrophages or bone marrow derived dendritic cells (BMDCs). While deletions within the HVR also did not influence flagellin export and flagellum formation, deletion of the Linker or the D4 domain rendered the bacteria immotile. Intriguingly, deletion of the complete HVR led to functional flagella formation and thus motility similar to the wildtype (WT) strain. When testing these deletion mutants in vivo, 3 out of 4 experiments failed to establish a protective role for EcN WT from colitis. Hence, no conclusions can be drawn on how the HVR influences EcN FliC’s probiotic properties.
In conclusion, EcN FliC’s long HVR does not seem to provide advantages in evasion of innate immune cell and/or TLR5 recognition, neither by directly influencing TLR5 recognition nor by compensating for destabilizing mutations in the conserved region. However, its integrity is required for outer sheath formation and bacterial motility and thus likely for effective colonization as a perquisite for EcN to convey its probiotic properties in vivo.
Flagellum