Figure 1.
Predation of HI strains compared to predatory HD100.
Time-lapse microscopy of individual wild-type HD100 cells (A(a)) and cells of the markerless deletion mutant of bd0108 (A(b)) preying upon the E. coli strain pMAL-p2_mCherry with a fluorescent periplasm. At T=0 there is attachment of free swimming Bdellovibrio in both strains, with invasion occurring within 2 frames (5 minutes) for most cells observed. Arrows indicate both free swimming and attached Bdellovibrio cells. By 1 hour the Bdellovibrio is established in the periplasm and begun to grow as a filament. By 8 hours the growing filaments of both the wild-type and deletion strains have septated to form single progeny and the prey bursts releasing them. Predation was carried out on a 1% Agarose pad with images acquired every 2.5 minutes.
B. Reduction of E. coli numbers in a predatory lysate comparing wild-type HD100 or spontaneously generated HIs with the markerless deletion of bd0108 strains. The spontaneously generated HI strains included some with a variety of point mutations or the common 42bp deletion in bd0108 and some with a wild-type copy of bd0108. HI cultures readily prey upon E. coli in liquid cultures, reducing prey numbers by several logs in 2 days with predation by the population as a whole appearing to be slightly slower than wild-type HD100. All HI strains were grown axenically, independently of prey cells before the experiment.
Figure 2.
Luminescent prey assay of predation efficiency for Host Dependent and Host Independent strains.
For Host Dependent cells (A) with a genomic copy of the wild-type bd0108 gene, carrying the pSUP404.2 plasmid encoding either the wild-type bd0108 gene or the 42 bp deletion variant of bd0108 had no effect on predation compared to those carrying the empty pSUP404.2 plasmid. This assay shows the typical result of logarithmically faster reduction in luminescence with more Bdellovibrio initially added and shows that the extra copies of bd0108 or the 42 bp deletion variant have no significant effect on this. For Host Independent cells (B) there is not the same proportional decrease in luminescence when more cells are added supporting the hypothesis that a proportion are growing axenically as HIs rather than predatorily. HI mutants carrying the pSUP404.2 plasmid encoding the wild-type bd0108 gene, however, restore the typical initial-cell-number to drop in luminescence relationship seen in the HD wild-type. This suggests that the presence of a wild-type bd0108, but not the 42 bp deletion variant, in trans restores the HI cells to a predatory lifestyle.
Figure 3.
Plaque assay of predation by Host Independent mutants compared to HD100.
Dilution of predatory cultures of Bdellovibrio strains on double layer YPSC overlay plates containing prey E. coli S17-1 in the top layer. Dilution was carried out in Ca/HEPES with 10 µl of the resulting dilution series spotted at positions indicated in the diagram. Plates show there is no significant effect on HD100 having extra copies of the bd0108 gene or the 42 bp deletion variant of bd0108 on plaquing ability. Complementation tests for the HID13 strain (ATG->ATA mutation in bd0108) with the plasmid pSUP404.2 containing the wild-type bd0108 gene showed an increase in plaquing, however this was not observed for the Δbd0108 deletion strain. Plates are representative of at least three independent repeats.
Figure 4.
PFU to CFU ratio for pooled spontaneous HI strains and the markerless bd0108 deletion HI mutants of Bdellovibrio.
In the spontaneous generated HI strains; HID2 and HID26 (wild-type for bd0108) HID6, HID13, HID18, (point mutations in bd0108) and HID22 (bd0108∆42bp), and the markerless deletion mutants of bd0108, complementation tests shows that these strains carrying the pSUP404.2 plasmid encoding the wild-type bd0108 is significantly deficient in the capacity to form HI colonies on PY agar plates compared to the capacity to form plaques on double layer overlay plates. This was significant compared to cells containing the pSUP404.2 plasmid alone, or the pSUP404.2-Δ42 with the 42 bp deletion variant of bd0108. The ratio of CFU:PFU was reduced from 2.16 x 10-1 in strains containing the pSUP404.2 alone to 3.21 x 10-2 in strains containing pSUP404.2-108 (p = 0.026) but were not significantly reduced relative to strains containing pSUP404.2-Δ42, 2.56 x 10-1 (p = 0.31). The difference between strains carrying the pSUP404.2-108 and the pSUP404.2-Δ42 was also significant (P<<0.01).
Figure 5.
Expression of bd0108 and co-transcription of surrounding hit locus genes using matched amounts of total RNA.
In spontaneously generated HI strains; HID2, HID26 (with wild-type bd0108) HID13 andHID22 (bd0108∆42bp) there was transcription of bd0108 (A). The primers flank the point of the 42 bp deletion in bd0108 so for strain HID22 the PCR product is smaller. HID13, which has a mutated first codon (ATG->ATA) had lower amounts of expression of bd0108 compared to other strains. RT-PCR analysis across the intergenic regions shows co-transcription of the gene pairs; bd0108 and bd0109 (B), bd0109 and bd110 (C), and bd0110 and bd0110 (D) in HI strains; HID13, HID22 and wild-type Attack Phase HD100. Primers are indicated by arrows above and below the gene cartoons.
Figure 6.
Evidence of interaction between Bd0108 and Bd0109.
A. Fluorescence quenching assay of W fluorescence of Bd0109 by addition of (naturally non-fluorescent) Bd0108 protein. Bd0109 at a concentration of 2.41 µM was titrated with Bd0108 at 1.95 mM in non-reducing buffer. Increasing concentrations of Bd0108 resulted in increased quenching of fluorescence indicating interaction between the two proteins.
B. Protease protection assay with Bd0109 and Bd0108 using chymotrypsin. The monomer of Bd0108 is around 17 kDa and the monomer of Bd0109 is around 65 kDa. Extra bands begin to appear around 35 and 43 kDa (indicated by asterisks) in the Bd0109 lanes treated with chymotrypsin which represent degradation products of Bd0109. These products cannot be seen in the lanes containing a complex of the two proteins (labelled complex) indicating that Bd0108 is interacting with Bd0109 to protect it from chymotrypsin digestion. Times indicated at the top are in minutes from adding the chymotrypsin.
Figure 7.
Alignment of Bd0109 with other RHS elements.
Multiple sequence alignment constructed using Clustal W showing that the main regions of homology between the Bdellovibrio Bd0109 predicted protein and other RHS elements is the core pFAM RHS repeat (PF05593) region which consists of a repeating YD element. The whole Bd0109 sequence contains 13 YD or YE sequences and several other Y residues. The sequences aligned are from the genera indicated with the names, with the second Dickeya sequence being RhsB.
Figure 8.
Electron micrographs of HI Bdellovibrio showing the presence and absence of cell surface structures.
Representative electron micrographs of cells stained with 2% PTA pH7.0 to show long straight pili structures. A) An example of a HID22 (bd0108∆42bp) cell showing a pilus greater than 1 µm that occurred in 15% of the isolates. Average representatives of HID22 cells (B), and of a HI strain containing a wild-type bd0108 gene (C). The Δbd0108 markerless deletion strain showing no pilus-like structures (D). E) A table showing 4 HI isolates and their average presence/absence and length of pili. Scale bars = 500 nm.
Figure 9.
In trans expression of fluorescently tagged Bd0108 and Bd0109 in E. coli S17-1 heterologous host.
A. E. coli S17-1 expressing a C-terminal mCherry fusion of Bd0108 protein in trans. Fluorescence can be seen in the periplasm of the cells, with large amounts towards the poles of the cell where cells are plasmolysed.
B. E. coli S17-1 expressing a C-terminal mTFP fusion of Bd0109 protein in trans. Fluorescence can be seen in the periplasm of the cells, with large amounts towards the poles of the cell where cells are plasmolysed.
Figure 10.
RT-PCR expression of the sRNA downstream of bd0108 and expression of a control gene dnaK.
Expression of a small, non-coding RNA downstream of bd0108 was shown to be different in HI strains carrying different mutations in bd0108 and attack phase HD100. Expression was highest in HID13 (ATA->ATG start codon mutation) and the strains with the markerless deletion of bd0108. Strain HID22 (bd0108∆42bp), had slightly lower expression, while those strains with a wild-type bd0108; HD100 and HID2, show much lower expression of the sRNA. Expression of dnaK was uniform across the samples indicating a matched amount of total RNA was used for the experiment.
Figure 11.
Model for possible interactions of Bd0108/Bd0109 controlling the extrusion and retraction of pili.
A. Operonal structure of the bd0108 hit locus and surrounding genes, predicted to have a role in the formation of a Type IVb pilus. Genes are colour coded to correspond to their predicted function in the pilus diagrams underneath.
B. In wild-type cells bd0108 and bd0109 are co-expressed, the mRNA is then translated into proteins containing a signal sequence recognised by the Sec system, the signal is cleaved, and the proteins are transported into the periplasm where the mature Bd0108 protein transiently interacts with Bd0109 to sequester it. When Bd0109 is unbound, it could then anchor at either the cell wall, or with the mature pilus fibre. Both scenarios are possible due to Bd0109’s structural cleft binding carbohydrate that is present in both cell wall and the mature and glycosylated pili. Bd0109 mediates successful pilus extrusion/retraction and signal back into the cytoplasm. In wild-type pilus formation Bd1290 pre-pilins are held in the inner membrane and are assembled into the pilus fibre possibly by the flp pilus ATPases Bd0110 and Bd0111. The balance of sequestering and release of Bd0109 by Bd0108 in the periplasm permits to successful extrusion and retraction of the pilus fibre upon environmental cues.
C. In the absence of Bd0108 protein, Bd0109 is not sequestered and is free to mediate more frequently with pilus extrusion and retraction, resulting in very few pili extruded beyond the cell surface and cues for HI growth signalled to the cell.
D. In HI strains containing the 42 bp deletion variant of bd0108, the gene is still expressed. The truncated form of Bd0108 alters the dynamics of the Bd0109 functionalisation are altered (possibly by over-sequestration of Bd0109) and hyper-extruded pili are seen on the surface more frequently. Hyper-extruded pili or no pili send similar internal signals to regulate prey independent growth.