Variation in genome content and predatory phenotypes between Bdellovibrio sp. NC01 isolated from soil and B. bacteriovorus type strain HD100

Abstract

Defining phenotypic and associated genotypic variation among Bdellovibrio may further our understanding of how this genus attacks and kills different Gram-negative bacteria. We isolated Bdellovibrio sp. NC01 from soil. Analysis of 16S rRNA gene sequences and average amino acid identity showed that NC01 belongs to a different species than the type species bacteriovorus. By clustering amino acid sequences from completely sequenced Bdellovibrio and comparing the resulting orthologue groups to a previously published analysis, we defined a 'core genome' of 778 protein-coding genes and identified four protein-coding genes that appeared to be missing only in NC01. To determine how horizontal gene transfer (HGT) may have impacted NC01 genome evolution, we performed genome-wide comparisons of Bdellovibrio nucleotide sequences, which indicated that eight NC01 genomic regions were likely acquired by HGT. To investigate how genome variation may impact predation, we compared protein-coding gene content between NC01 and the B. bacteriovorus type strain HD100, focusing on genes implicated as important in successful killing of prey. Of these, NC01 is missing ten genes that may play roles in lytic activity during predation. Compared to HD100, NC01 kills fewer tested prey strains and kills Escherichia coli ML35 less efficiently. NC01 causes a smaller log reduction in ML35, after which the prey population recovers and the NC01 population decreases. In addition, NC01 forms turbid plaques on lawns of E. coli ML35, in contrast to clear plaques formed by HD100. Linking phenotypic variation in interactions between Bdellovibrio and Gram-negative bacteria with underlying Bdellovibrio genome variation is valuable for understanding the ecological significance of predatory bacteria and evaluating their effectiveness in clinical applications.

Keywords: lytic enzymes; plaque phenotype; predation efficiency; predatory bacteria; prey range.

Fig. 1.

Microscopy of Bdellovibrio sp. NC01 isolated from bioswale soil. (a) Compilation of four images taken using 1000x phase-contrast microscopy to show small, comma-shaped NC01 attack phase cells attached to larger E. coli ML35 cells after 48 h of co-culture. (b) Compilation of two images taken using 1000x phase-contrast microscopy to show rounded structures characteristic of bdelloplasts after 48 h of co-culture. (c) Negative stain electron microscopy of a NC01 attack phase cell. The scale bar is 500 nm.

Fig. 2.

16S rRNA gene phylogenetic tree. We aligned 16S rRNA gene sequences using sina 1.2.11, then constructed a phylogenetic tree using RAxML BlackBox with 100 bootstrap replicates. The phylogenetic tree is rooted on the branch joining Bdellovibrio with Bacteriovorax / Halobacteriovorax , and bootstrap support for each node is shown.

Fig. 3.

Genome-wide pairwise comparisons of Bdellovibrio nucleotide sequences. The genome of Bdellovibrio sp. NC01 is shown as the innermost ring. The next ring is a plot of GC content and the purple and green ring is a plot of GC skew, both of which were generated by BRIG using sliding windows to determine deviation from the average value for the genome. The blue ring and red ring show pairwise nucleotide sequence alignments of NC01 to HD100 and W, respectively. Each individual alignment is shown as a block along the ring, and colour gradation of the blocks reflects the percent similarity of alignments. The outermost ring denotes eight regions lacking alignments, which may indicate unique NC01 gene content.

Fig. 4.

Efficiency of Bdellovibrio sp. NC01 and B. bacteriovorus HD100 predation on E. coli ML35. We quantified viable E. coli ML35 over 72 h on its own (control), combined with Bdellovibrio sp. NC01 (NC01) and combined with B. bacteriovorus HD100 (HD100). The resulting c.f.u. ml^{− 1} data were log transformed and plotted using ggplot in RStudio. Each time point shows the mean and standard deviation for at least five replicates.

Fig. 5.

Population density of Bdellovibrio sp. NC01 and B. bacteriovorus HD100 in the presence and absence of E. coli ML35. We quantified Bdellovibrio cells ml^{− 1} at the start of the predation efficiency assay and Bdellovibrio p.f.u. ml^{− 1} after 72 h of co-culture with E. coli ML35 (diamonds). For comparison, we quantified Bdellovibrio p.f.u. ml^{− 1} over 72 h of incubation in HM buffer alone (circles). The resulting cells ml^{− 1} and p.f.u. ml^{− 1} data were log transformed and plotted using ggplot in RStudio. Each time point shows the mean and standard deviation for all replicates.

Fig. 6.

Plaque phenotypes of Bdellovibrio sp. NC01 and B. bacteriovorus HD100. Using a double agar overlay method, we assessed plaque formation on lawns of E. coli ML35. These images were taken after 6 days of incubation and illustrate reproducible differences in the appearance of plaques formed by HD100 (a) and NC01 (b). The scale bar is 10 mm.