Black band disease (BBD) of corals is known to contribute to the degradation of coral reefs in the wider Caribbean [1
], the Indo-Pacific [2
], including the Great Barrier Reef [3
], and the Red Sea [4
]. It has been reported to affect 70 species of corals, including both scleractinians and gorgonians [2
], and can cause whole colony mortality by rapid tissue lysis. The disease often targets the ecologically important reef-framework coral species.
In appearance, BBD is a dark, well-defined cyanobacterial mat that forms a band, which moves across the coral surface, degrading coral tissue and leaving behind bare coral skeleton. It is a polymicrobial disease dominated by non-heterocystous filamentous cyanobacteria and contains populations of sulfate-reducing bacteria, sulfide-oxidizing bacteria, and heterotrophic bacteria. The BBD microbial community is highly diverse as revealed through microscopy [5
] and, in particular, molecular biological methods [7
It has been proposed that some of the BBD microbial diversity may be due to incorporation of coral surface mucopolysaccharide layer (SML) bacteria into the BBD population [8
]. In turn, the presence of disease may affect the SML bacterial community. Previous studies of coral-associated bacteria revealed that apparently healthy corals contain less diverse bacterial communities than the healthy areas of diseased corals, and that the compositions of the two populations are different [7
]. While it is known that coral-associated bacteria are present in coral tissue and skeleton as well as the SML, studies of the composition of coral-associated bacterial communities have focused on those present in coral tissue [7
], with fewer studies targeting those within the coral SML [10
]. The results of these studies, in general, indicate that members of the gamma- and alpha-proteobacteria dominate coral-associated bacterial populations.
The coral probiotic hypothesis [11
] proposes that coral resistance to disease can be promoted by coral-associated bacteria, which prevent colonization by potential pathogens or outcompete pathogens which may settle on coral. The protective mechanism of coral probiotic bacteria is proposed to include antibacterial activity, and to date there have been a number of studies aimed at demonstrating that corals and their associated bacteria possess such properties [16
]. Koh [18
] showed that the alcohol extract of a large percentage of coral samples had antibacterial activity against a number of heterotrophic bacteria and cyanobacteria, while results obtained by Kim [20
] demonstrated antibacterial activity for both polar and non-polar (coral-derived) fractions with higher activity associated with non-polar fractions. In both studies the extracts were prepared from the coral holobiont (the coral animal, endosymbiotic zooxanthellae, and coral-associated bacteria), and the origin of the active compounds was not determined. Ritchie [16
] found antibacterial activity among the microbial community associated with the mucus of healthy corals.
Deciphering the relationships among and between the bacterial members of the complex BBD microbial consortium could provide insight into the etiology of this polymicrobial disease. Since cyanobacteria are the dominant component of BBD in terms of biomass, we selected this group of microorganisms to investigate their ability to inhibit growth of both BBD and (healthy) coral SML bacterial isolates. In this way, we hypothesize that cyanobacteria may structure the complex BBD microbial community. We also investigate, for comparison, the antibacterial activity of sub-tropical marine cyanobacteria from other marine sources. This study assesses the potential of marine cyanobacteria as a source for novel antibacterial agents that could potentially be applied in human health.
In recent years the study of coral-associated microorganisms has greatly expanded, with a focus on the role of such microorganisms in coral health and disease. A number of these studies have documented antibacterial activity among and between coral-associated bacteria [16
], and also antimicrobial activity of corals themselves [16
Cyanobacteria are well known to produce antibacterial compounds [23
] and are known to be associated with corals [8
]. However, little work has been done to assess the antibacterial activity of coral-associated cyanobacteria, or, for that matter, cyanobacteria in other marine environments. In this work we assessed the activity of marine cyanobacteria on growth of coral-associated bacteria using two different methods: co-cultivation in which cyanobacterial metabolites were allowed to diffuse into agar (without breaking their cell walls), and use of extracts prepared from equal amounts of dried cyanobacterial culture biomass. Our results revealed that co-cultivation of cyanobacteria with BBD or SML heterotrophic bacteria resulted in 4.8% cases of inhibition of BBD bacteria and 10% cases of inhibition of SML bacteria. This result was statistically significant for both BBD and other marine cyanobacteria (P
< 0.05; see legend of Figure 2
Overall, BBD bacteria appear to be more resistant to antibacterial activity of marine cyanobacteria in general. This might be expected since BBD bacteria live in close physical contact with cyanobacteria. Within the total BBD bacterial/cyanobacterial test cases, 82% of BBD bacteria exhibited resistance. However, the 18% that were inhibited were represented by six of the 10 BBD bacterial strains tested. This result may be important in that the complex BBD microbial community may require small populations of some functional members; from this perspective, regulation of growth of certain bacterial members within the organic carbon/nutrient rich BBD environment would be an important part of the disease etiology.
There was a much higher percentage (30% of test cases) of inhibition of SML bacteria by BBD cyanobacteria when compared to inhibition of BBD bacteria. This finding suggests that BBD cyanobacteria may be capable of eliminating potentially beneficial, protective coral-associated bacteria, thus supporting the coral probiotic hypothesis [11
]. Alternatively, BBD cyanobacteria may eliminate SML bacteria within a developing BBD microbial community to reduce competition. Other marine cyanobacteria also showed antibacterial activity in 9% of test cases. Therefore, it can be hypothesized that non-pathogenic cyanobacteria present in healthy corals may also provide protection against disease-causing microorganisms.
We observed, to a lesser extent, stimulation of growth of some bacterial isolates during co-cultivation with cyanobacteria, in particular with other marine cyanobacteria. While there were only two cases of stimulation by BBD cyanobacteria (both of SML bacteria), there were 10 cases of stimulation by other marine cyanobacteria. This is in agreement with the findings of Morrow et al.
] who reported that two marine Lyngbya
species stimulated, but did not inhibit, the growth of marine bacteria.
Antimicrobial activity was found mostly among extracts obtained with non-polar rather than hydrophilic solvents, which is in agreement with the findings of others [28
]. This can be attributed to the fact that lipophilic compounds more easily cross the cell membrane, thus are more likely to exert an affect. BBD cyanobacteria appear to produce a variety of antibacterial compounds that are not excreted, thus their effect on members of the BBD consortium would be limited. We believe that the co-cultivation experiments more closely mimic the ecological conditions within BBD since microscopic observations of BBD samples reveal healthy, highly motile cyanobacteria and an absence of lysing filaments.
Comparison of the coral-associated bacterial isolates used in these experiments with a recently conducted meta-analysis of 84 BBD bacterial clone libraries [29
] revealed that none of our isolates was a match to BBD sequences deposited from these sources in GenBank. This is not unusual, as pointed out recently in a study of coral-associated bacteria by Rypien et al.
] that used a combination of culture-based and molecular methods. We do note that our test bacteria were represented by members of three genera, Marinobacter
sp., and Vibrio
sp., which have been reported in BBD clone libraries [29
]. Of these, the BBD Marinobacter
isolate used in the current study (strain B5) and the Alteromonas
isolate (strain B6) were inhibited by two and seven strains of BBD cyanobacteria respectively (Table 1
). Neither was inhibited by other marine cyanobacteria, and neither was stimulated by any (BBD or other marine) cyanobacterial strain tested. The Vibrio
sp. strains were inhibited to different extents by both BBD and other marine cyanobacteria.
For the known coral pathogens, co-cultivation in only one of the 60 tests (20 cyanobacteria × 3 bacterial pathogens) resulted in inhibition. This result was obtained for the bacterial bleaching pathogen Vibrio shiloi
. Nissimov et al.
] found, in an investigation of the antibacterial properties of bacteria from coral mucus, that 5.7% of bacterial isolates inhibited Vibrio shiloi
. Therefore, if this bacterium is responsible for bleaching of corals under certain environmental conditions, as proposed by Rosenberg et al.
], it appears that protection of coral would be not be conferred by cyanobacterial members of the coral holobiont, and perhaps not by bacterial members.
It was surprising that while co-cultivation did not inhibit growth of the other two coral pathogens tested (Aurantimonas coralicida and Serratia marcescens) there was inhibition in 27% (16 of 60) of the tests conducted using lipophilic cyanobacterial extracts. Therefore cyanobacteria (both BBD and other marine) may potentially be involved in protecting corals from these specific pathogens.
Our results indicate that both pathogenic (BBD) and non-pathogenic marine cyanobacteria can affect the growth of coral-associated bacteria, and that while most of this activity is manifested as inhibition of growth, growth stimulation also occurs. While we did not, in the present study, identify the cyanobacterial compounds responsible for the observed activity, other studies in our laboratory suggest that at least some of the activity could be due to the cyanotoxin microcystin. We have previously shown that BBD cyanobacteria produce microcystins [32
] and that freshly collected BBD samples contain microcystins [32
]. All of the cyanobacteria used in the current study were previously shown to produce microcystin-LR in the laboratory [33
], with the exception of two Leptolyngbya
strains that were not tested (strains 96-2 and 9-1; see Table 3
). We have also shown that exposure of healthy coral fragments to low concentrations (1 μg L−1
) of purified microcystin-LR resulted in increased bacterial growth in coral fragments, observed using Scanning Electron Microscopy [35
]. This observation was investigated by exposing bacterial isolates to three concentrations (1, 100 and 500 μg L−1
) of purified microcystin-LR and monitoring growth. These experiments were carried out using 24 of the 26 bacteria used in the present study (the two strains not tested were SML strains 1-1 and 1-14, both Vibrio
spp.). Results were variable, with microcystin exposure both inhibiting and stimulating bacterial growth. When comparing the microcystin results with the results of the current study, only 17% of the cases of inhibition or stimulation by co-cultivation, and 23% of the cases of inhibition or stimulation in response to exposure to lipophilic extracts, were recorded for strains that responded to microcystin. Separating microcystin effects from other compounds would require detailed chemical analyses. Another potential source of the active compounds documented in this study consists of bacteria associated with the non-axenic cyanobacterial cultures. However, we deem this source to be negligible since all cyanobacteria were grown autotrophically for these experiments, thus their biomass would be orders of magnitude higher than the contaminants. Microscopic examination of the cyanobacterial cultures did not reveal the presence of bacteria; they could be detected only by plating cyanobacteria onto heterotrophic media.