Joined: May 2002
In the "nonmotile appendages can have a dispersal-related function despite being nonmotile" category:
Knutton S, Shaw RK, Anantha RP, Donnenberg MS, Zorgani AA. The type IV bundle-forming pilus of enteropathogenic Escherichia coli undergoes dramatic alterations in structure associated with bacterial adherence, aggregation and dispersal. Mol Microbiol 1999 Aug;33(3):499-509
BFP, a plasmid-encoded type IV bundle-forming pilus produced by enteropathogenic Escherichia coli (EPEC), has recently been shown to be associated with the aggregation of bacteria and dispersal of bacteria from bacterial microcolonies. In standard 3 h HEp-2 cell assays, EPEC adhere in localized microcolonies; after 6 h, bacterial microcolonies are no longer present, indicating that bacterial aggregation and dispersal occurs in vitro during EPEC adhesion to cultured epithelial cells. To examine the role of BFP in EPEC aggregation and dispersal, we examined HEp-2 cell adhesion of strain E2348/69 and defined E2348/69 mutants by immunofluorescence and immunoelectron microscopy. BFP was expressed initially as approximately 40 nm diameter pilus bundles that promoted bacteria-bacteria interaction and microcolony formation. BFP subsequently underwent a striking alteration in structural organization with the formation of much longer and thicker ( approximately 100 nm diameter) pilus bundles, which frequently aggregated laterally to form even thicker bundles often arranged in a loose three-dimensional network; EPEC dispersal from bacterial microcolonies was associated with this transformation of BFP from thin to thick bundles. Bacterial dispersal and transformation of BFP from thin to thick bundles did not occur with a bfpF mutant of strain E2348/69. It is concluded that BFP promotes both the formation and the dispersal of EPEC microcolonies, that the dispersal phase requires BfpF and that dispersal is associated with dramatic alterations in the structure of BFP bundles.
As dispersal of bacteria from microcolonies occurred between 3 h and 6 h, we examined cells at intermediate times in order to follow the dispersal process and any associated change in BFP morphology. At 4 h, by both immunofluorescence (Fig. 6A, arrow) and scanning electron microscopy (Fig. 7A, arrows), one could start to see the formation of thick BFP bundles within some bacterial microcolonies and, by scanning electron microscopy, bacteria appeared to have been lost from regions of the microcolony in which thick BFP bundles had formed (Fig. 7A). At 5 h, significant dispersal of bacteria from many microcolonies had occurred, although this varied from colony to colony.
Fig. 7. Scanning electron micrographs of HEp-2 cells infected with EPEC strain CVD206 for 4 h (A), 5 h (B) and 6 h © showing stages in bacterial dispersal. After 4 h, thick BFP bundles are forming within this bacterial microcolony, and bacteria look as though they may have been lost from these regions of the microcolony (A, arrows). After 5 h, dispersal of bacteria from some microcolonies is almost complete, and only a few small bacterial aggregates remain attached to thick BFP bundles (B, arrows); in contrast, there is no evidence of BFP transformation or bacterial dispersal in the microcolony seen on the right (B, asterisk). After 6 h, bacterial dispersal is virtually complete; one of the few remaining bacteria from this microcolony is anchored to the thick BFP bundles by thin bundles (C, arrows). Scale bars: A, 1 m; B, 2 m; C, 0.5 m.
BfpF and bacterial dispersal
BfpF has been shown to be required for the dispersal of EPEC from bacterial microcolonies (Bieber et al., 1998). We therefore examined a bfpF mutant of E2348/69 in order to determine whether the observed morphological transformation of BFP was affected by this component of the BFP operon. In 3 h assays, we confirmed previous observations that a mutation in bfpF resulted in increased localized adhesion and that bfpF mutants are hyperpiliated compared with the wild-type strain. Furthermore, the mutation did not affect the ability of this strain to produce A/E lesions. Other than the size of microcolonies, this phenotype showed no alteration after 6 h (Fig. 8); bacteria remained hyperpiliated (Fig. 8A), adherent bacteria produced A/E lesions (Fig. 8C and D), but there was no transformation of thin BFP bundles to thick bundles (Fig. 8A and B) and no dispersal of bacteria from microcolonies (Fig. 8A, B and D).
The original description of BFP defined a role in bacteria-bacteria interaction and microcolony formation (Girón et al., 1991); recently, it has been shown that BFP also promotes dispersal of bacteria from aggregates (Bieber et al., 1998). Bieber et al. (1998) ended their paper by suggesting that 'to dissociate from the aggregate, bacteria would need to shed or distangle their pilus filaments from each other, a process that may require BfpF-mediated, energy dependent pilus retraction or a conformational change in the pilus quaternary structure'. This study, which has now demonstrated that EPEC aggregation and dispersal occurs in vitro during infection of cultured epithelial cells, suggests that the latter may be the case and that BFP undergoes a dramatic BfpF-dependent change in quaternary structure, the consequences of which are (i) a change from a thin to a thicker BFP bundle structure; (ii) disruption of bacteria-bacteria interactions; and (iii) dispersal of EPEC from bacterial microcolonies. The advantage to the organism of such a mechanism is that dispersal of bacteria primed to produce A/E lesions would be expected to lead to infection of new epithelial sites within the small bowel and, therefore, to a more efficient colonization of the gut. It has been known for some time that EAF plasmids are important in EPEC pathogenicity; BFP, by promoting more efficient colonization, is one likely reason why typical EPEC, which possess EAF plasmids, are more virulent than atypical EPEC, which lack EAF plasmids (Levine et al., 1985), and also more virulent than EPEC BfpF mutants, which lack the ability to disperse from microcolonies (Bieber et al., 1998).
Although we suggest a causal relationship between BfpF function, transformation of BFP morphology and bacterial dispersal, the data do not, in fact, demonstrate such a relationship, and so we cannot rule out the possibility that the converse is true, namely that BfpF is involved in other events that promote dispersal of bacteria from microcolonies, which leads, in turn, to the formation of the thick BFP bundles. BFP are very hydrophobic pili and, as bacteria disperse from an aggregate, it could be that the thin BFP filaments become more accessible to each other and are able to associate to form the thick BFP bundles.
In addition to playing a role in bacterial dispersal, an adhesive role for BFP has also been suggested (Girón et al., 1991). While the aim of this study was to examine the role of BFP in bacterial aggregation and dispersal, some of the observations have relevance to the possible role of BFP in cell adhesion. For example, the data suggest that BFP may be involved in initial EPEC adhesion, but that intimin-mediated intimate attachment is required for subsequent adhesive events. Also, the presence of cell-associated BFP after dispersal of all CVD206 bacteria after 6 h demonstrates that this form of BFP can adhere to the surface of HEp-2 cells. However, the role of BFP in EPEC adhesion to cultured and intestinal epithelial cells is the subject of a separate study to be published elsewhere (S. Knutton et al., manuscript in preparation).
[well, "nonmotility" is debatable, but this doesn't appear to be directional movement, i.e. swimming or crawling]
This study confirmed a role for BfpF in microcolony dispersal (Bieber et al., 1998) and showed that this protein, while not required for thin BFP bundle assembly, is involved either directly or indirectly in transformation from thin to thick BFP bundles. Based on the similarity between BfpF and PilT, the putative nucleotide-binding protein of P. aeruginosa, it has been proposed that BFP may play an analogous role to that proposed for PilT in type IV pilus function, namely as an energy source for the retraction of BFP (Anantha et al., 1998). The proposed function of PilT is based on electron microscopic studies of the distribution of antibody and bacteriophage binding to pili from wild-type and mutant P. aeruginosa strains (Bradley, 1974). The mutant used for these studies was subsequently found to have a mutation in pilT, which is also required for twitching motility (Whitchurch et al., 1991). As proteins in the PilT family are proposed to reside in the cytoplasm, our finding that BfpF appears to have a profound effect on the quaternary structure of BFP outside of the bacteria is surprising. However, as BFP has a marked propensity to intertwine in rope-like bundles, it is possible that retraction of individual pili within a bundle mediated by BfpF leads to thickening of the bundle in much the same way as pulling on an individual fibre leads to thickening of a twine. Thus, our study provides additional evidence that type IV pili such as BFP are not fixed structures, but are capable of dynamic alterations that influence bacterial adherence and pathogenesis.