Many motile laboratory strains of E. coli are able to form biofilms on abiotic surfaces such as polyvinylchloride (PVC), polypropylene, polycarbonate, and borosil-icate glass when grown statically in rich medium at room temperature (Pratt and Kolter 1998). Therefore, type 1-mediated biofilm formation may contribute to the ability of UPEC to withstand antibiotic treatments and host antimicrobial defenses in the urinary tract. To gain an understanding of the factors involved in formation of E. coli biofilms, Pratt and Kolter used transposon mutagenesis to generate mutants defective in biofilm formation on abiotic surfaces (Pratt and Kolter 1998). Mutants capable of motility but still severely defective in biofilm formation were isolated and determined to fall within the fim gene cluster. Independent insertions were found within fimB, fimA, fimC, fimD, and fimH. Microscopic analysis of PVC surfaces on which the fim mutants were grown statically in rich medium revealed that fim mutants were so severely defective in initial attachment that most of the surface contained no attached E. coli. This indicated that type 1 pili are essential in initial attachment to abiotic surfaces. Furthermore, FimH was shown to mediate this attachment as a nonmetabolizable mannose analog a-methyl-D-man-noside-inhibited biofilm formation by the type 1-piliated wild type E. coli strain 2K1056. FimH to surface interactions were concluded to be direct and involve nonspecific binding to abiotic surfaces. The role of FimH in mediating attachment to abiotic surfaces was supported by the isolation of both natural and engineered FimH variants that allow E. coli to form biofilms under hydrodynamic flow (HDF) conditions (Schembri and Klemm 2001a). E. coli with the wild type FimH adhesin and commensal E. coli strains could not form biofilms under the same HDF conditions. HDF shear force conditions are thought to better mimic the natural environment during UTI, and interestingly, a G73E variant of FimH that was identified as a functional alteration of FimH involved in biofilm formation was previously identified as a natural FimH variant pathoadaptive for UTI (Sokurenko et al. 1994).
The same FimH mutant library also yielded FimH variants that were capable of mediating autoaggregation of E. coli (Schembri et al. 2001). Phase contrast microscopy of cells expressing the FimH variants showed that they form large tight clusters of cells that might aid in forming microcolonies during biofilm formation. These results were again supported by the isolation of natural FimH variants from UPEC strains that exhibited the ability to autoaggregate, arguing that this pheno-type may be relevant to UTI pathogenesis (Schembri et al. 2001). Furthermore, autoaggregation in general appears to be important for biofilm formation as other self-aggregating cell surface structures such as Antigen 43 (Ag43) and curli are also associated with biofilm formation in E. coli (Vidal et al. 1998; Danese et al. 2000; Kjaergaard et al. 2000; Prigent-Combaret et al. 2000). In support of this idea, DNA microarrays showed that both type 1 pili and Ag43 are more highly expressed in biofilm populations than in planktonic populations (Schembri et al. 2003b). Therefore, the presence of type 1 pili may facilitate the colonization of urinary tract catheters and other implants by mediating biofilm formation and autoaggregation. Finally, E. coli strains carrying transfer constitutive IncF plasmids were shown to form mature mushroom-shaped structures similar to those of P. aeruginosa biofilms in continuous-flow cell cultures in glucose minimal medium at 30°C (Reisner et al. 2003). The presence of type 1 pili was found to be dispensable for the biofilm maturation observed as neither fimbriated nor afimbriated variants had any effect on biofilm maturation by E. coli strains that were plasmid-free or carrying F plasmid.
The regulation of the FimB and FimE recombinases may also function as one regulatory checkpoint for biofilm formation in E. coli. Phase variation of type 1 pili is subject to multiple global regulators including LrpA (Blomfield et al. 1993; Gally et al. 1994), IHF (Blomfield et al. 1997), H-NS (Olsen and Klemm 1994; Donato et al. 1997; O'Gara and Dorman 2000) and LrhA (Olsen and Klemm 1994; Donato et al. 1997; O'Gara and Dorman 2000; Lehnen et al. 2002; Blumer et al. 2005). Both LrpA and IHF are required for the efficient switching by both the FimB and FimE recombinases, whereas H-NS affects only the FimB-mediated recombination event. LrhA, however, affects only the activation of the FimE recombinase that is required to turn off fimA production and subsequently the production of type 1 pili (Blumer et al. 2005). Mutation of LrhA therefore leads to decreased fimE expression and subsequent increased transcription of fimA that ultimately results in increased biofilm formation. Overexpression of LrhA abolishes biofilm formation.
Was this article helpful?