We have shown that of three B. bacteriovorus HD100 sigma factor genes with at least partial rpoE homology, one- bd3314, is likely essential for Bdellovibrio cell life and cannot be deleted. bd0881 and bd0743 can be deleted with the Bdellovibrio retaining the ability to grow predatorily or prey-independently.
In the case of ΔBd0881 the predatory efficiency was reduced, despite the flagellar motility of the mutant being slightly increased, (despite a slight but statistically significant shortening of flagellar filament length) thus the change in predation efficiency may not be due to motility changes but regulation of other predatory genes. The bd0881 gene has an expression pattern across the predatory cycle that is similar to that of the flagellin genes whose expression is required for Bdellovibrio motility. That bd0881 expression is turned off and then resumes at a similar time to flagellin gene expression, during the predatory cycle, implies that Bd0881 may have a role associated with pre-septation developmental maturation of Bdellovibrio around the time that flagella are being built in newly dividing cells. However the Bd0881 sigma factor does not directly regulate the expression of fliC flagellin or mot flagellar motor genes themselves.
Surprisingly, predatory efficiency was not affected in our cultures by the slower swimming speed of the ΔBd0743 sigma factor mutant; this is probably indicative of sufficient mixing of predator and prey at close quarters in lab conditions. The slight increase in flagellar length in ΔBd0743 mutants is likely to have come with the incorporation of a higher percentage of a less rigid flagellin in the flagella causing a less efficient “bow wave” and this may account for the slower swimming. In both the ΔBd0743 and ΔBd0881 mutants, small but significant changes in swimming speed were paradoxically associated with changes apparently in the wrong direction
in flagellar length. This shows that it is not simply flagellar length that governs the thrust produced by flagellar propellers. In previous studies on the six different flagellins that are incorporated into the flagellar propeller of Bdellovibrio
], we found that different flagellin compositions of a single Bdellovibrio
flagellum are possible, and that in the case of a fliC4
mutant, for example, an apparently wild type-length flagellum gave a lower swimming speed than wild type [11
] suggesting an altered filament rigidity. As flagellar filament growth, in a bacterium with six flagellins, is a post-transcriptionally highly controlled process involving diverse chaperones and gate keepers at the base of the flagellum allowing different subunits to be added into the growing flagellum [18
] we cannot expect to tell anything meaningful about these small changes of swimming speed from simple studies of flagellar filament gene expression, so we have decided to leave this aspect of the investigation at this point.
In looking at chaperonin expression regulation by B. bacteriovorus
HD100 sigma factors, we found that, in contrast to bd0881
, deletion of which had no effect, the product of gene bd0743
acts more like the heat shock sigma factor RpoE of other bacteria and represses (directly or indirectly) the level of expression of chaperonin genes groES1 groEL
) in non-heat shock conditions and the level of expression of the groES2
) gene under both heat-shock and non-heat-shock conditions. These data and the finding that the groES2
gene is normally expressed in wild type Bdellovibrio
only during the late stages of predation (2–4 hours) when the Bdellovibrio
are septating and preparing to lyse the exhausted prey bdelloplast, may suggest that a modified chaperonin complex involving GroES2 is used in Bdellovibrio
protein expression and folding that occurs at this point. Ascertaining why this is the case requires more chaperone-specific experimentation, beyond the scope of this study and mutagenesis of bd3349
is underway. That the majority of GroES residues shown to interact with GroEL in E. coli
] are conserved or have conserved substitutions in both of the GroES1 and GroES2 homologues of B. bacteriovorus
HD100 supports the idea that they form genuine alternative chaperonin complexes, making GroEL protein folding chambers with different GroES “lids”. It is a tantalising possibility that Bdellovibrio
has a requirement for a modified chaperonin complex for the folding of unusual Bdellovibrio
proteins required for late-stage prey lysis or Bdellovibrio
attack phase cell maturation. The Bd0743-controlled, late-stage expression of groES2
is a possible mechanism for this. Although the (reannotated) Bdellovibrio groES2
gene product is larger at 117 amino-acids than the bd0097 groES1
gene product which is 100 amino-acids, there is no significant additional homology (above that for GroES1) between Bdellovibrio
GroES2 and the bacteriophage T4 Gp31 GroES-like protein (data not shown). The bacteriophage T4 Gp31 GroES-like protein allows formation of a larger protein folding chamber for unusual phage capsid protein Gp23 to fold. Bdellovibrio
, being a bacterium rather than a phage, does not have any homologues of this protein, so any analogous alternative role for GroES2 in Bdellovibrio
protein folding awaits the outcomes of further mutagenesis studies.