Alien vs. predator: bacterial challenge alters coral microbiomes unless controlled by Halobacteriovorax predators

Abstract

Coral microbiomes are known to play important roles in organismal health, response to environmental stress, and resistance to disease. The coral microbiome contains diverse assemblages of resident bacteria, ranging from defensive and metabolic symbionts to opportunistic bacteria that may turn harmful in compromised hosts. However, little is known about how these bacterial interactions influence the mechanism and controls of overall structure, stability, and function of the microbiome. We sought to test how coral microbiome dynamics were affected by interactions between two bacteria: Vibrio coralliilyticus, a known temperature-dependent pathogen of some corals, and Halobacteriovorax, a unique bacterial predator of Vibrio and other gram-negative bacteria. We challenged reef-building coral with V. coralliilyticus in the presence or absence of Halobacteriovorax predators, and monitored microbial community dynamics with 16S rRNA gene profiling time-series. Vibrio coralliilyticus inoculation increased the mean relative abundance of Vibrios by greater than 35% from the 4 to 8 hour time point, but not in the 24 & 32 hour time points. However, strong secondary effects of the Vibrio challenge were also observed for the rest of the microbiome such as increased richness (observed species), and reduced stability (increased beta-diversity). Moreover, after the transient increase in Vibrios, two lineages of bacteria (Rhodobacterales and Cytophagales) increased in coral tissues, suggesting that V. coralliilyticus challenge opens niche space for these known opportunists. Rhodobacterales increased from 6.99% (±0.05 SEM) to a maximum mean relative abundance of 48.75% (±0.14 SEM) in the final time point and Cytophagales from <0.001% to 3.656%. Halobacteriovorax predators are commonly present at low-abundance on coral surfaces. Based on the keystone role of predators in many ecosystems, we hypothesized that Halobacteriovorax predators might help protect corals by consuming foreign or "alien" gram negative bacteria. Halobacteriovorax inoculation also altered the microbiome but to a lesser degree than V. coralliilyticus, and Halobacteriovorax were never detected after inoculation. Simultaneous challenge with both V. coralliilyticus and predatory Halobacteriovorax eliminated the increase in V. coralliilyticus, ameliorated changes to the rest of the coral microbiome, and prevented the secondary blooms of opportunistic Rhodobacterales and Cytophagales seen in the V. coralliilyticus challenge. These data suggest that, under certain circumstances, host-associated bacterial predators may mitigate the ability of other bacteria to destabilize the microbiome.

Keywords: BALOs; Bacterial challenge; Halobacteriovorax; Microbiome; Vibrio coralliilyticus.

Figure 1. Halobacteriovorax predation of Vibrio spp.

(A) Micrograph of the pathogen, Vibrio coralliilyticus BAA450 being attacked by Halobacteriovorax and rounded V. coralliilyticus bdelloplast (right) with Halobacteriovorax inside (B) double layer plate showing freshly lysed plaques on a lawn of V. coralliilyticus cells (C) Overnight liquid cultures of (1) a co-culture of Halobacteriovorax and V. fortis, (2) V. fortis and 0.2 µm filtrate from Halobacteriovorax culture, and (3) V. fortis alone.

Figure 2. Montastraea cavernosa microbiome manipulation experimental design detailing collection and inoculation of coral cores, treatment tanks and replication, sample preservation, tissue removal, DNA extraction, and microbiome sample processing.

Figure 3. Impact of bacterial challenge on coral associated microbial diversity.

(A) Mean alpha diversity (observed species) plotted for each treatment, and (B) mean beta diversity (Weighted UniFrac distance) by treatment. The asterisks indicate Bonferroni-corrected p values <0.05 for the nonparametric t-test between treatments. In both cases while addition of the pathogen alone increased diversity, predator addition counteracted this effect.

Figure 4. Relative abundance of all taxa found to be differentially present under the four bacterial challenges.

Colored lines denote mean relative abundance for each time point with grey transparent shading indicating the standard deviation. Boxplots show the mean relative abundance averaged across all time points for each treatment.