OUP user menu

Coral mucus-associated bacteria: a possible first line of defense

Maya Shnit-Orland, Ariel Kushmaro
DOI: http://dx.doi.org/10.1111/j.1574-6941.2008.00644.x 371-380 First published online: 1 May 2009


Interactions among microorganisms found in coral mucus can be either symbiotic or competitive. It has been hypothesized that microbial communities found on the surface of coral play a role in coral holobiont defense, possibly through production of antimicrobial substances. Selected microorganisms isolated from the mucus layer of a number of coral species were grown using agar-plating techniques. Screening for antimicrobial substances was performed using overlay and drop techniques, employing several indicator microorganisms. Between 25% and 70% of cultivable mucus-associated bacteria from scleractinian corals demonstrated bioactivity. Higher percentages of activity were evident in mucus-associated cultivable bacteria from massive and solitary corals, as compared with bacteria from branching or soft corals. Isolates related to the genera Vibrio and Pseudoalteromonas demonstrated high activity against both Gram-positive and Gram-negative bacteria. Gram-positive bacteria (Bacillus, Planomicrobium) demonstrated lower levels of activity, primarily against other Gram-positive bacteria. In some cases, inhibitory effects were confined to the cell fraction, suggesting the involvement of a cell-bound molecule, sensitive to temperature and most likely proteinaceous in nature. These results demonstrate the existence of microorganisms with antimicrobial activity on the coral surface, possibly acting as a first line of defense to protect the coral host against pathogens.

  • antibacterial activity
  • corals
  • mucus-associated bacteria
  • symbiosis


In the oligotrophic marine environment, there are ecological niches rich in nutrients and diverse in bacterial populations. One such niche is the mucus layer found on the surface of sessile marine organisms, such as corals. Recent studies have revealed the presence of a dynamic microbial biota living on the surface and in the tissue of many coral species (Ritchie & Smith, 1997; Kushmaro et al., 1999; Rohwer et al., 2001, 2002; Kooperman et al., 2007). However, the role of these microorganisms in the reef ecosystem and their contribution to coral well-being, remain, for the large part, unclear. Microorganisms may be saprophytic or pathogenic, or may provide other important functions in the ecosystem (Kushmaro et al., 1996; Santavy & Peters, 1997; Harvell et al., 1999). For instance, microorganisms found on coral surfaces may provide the host with protection from pathogens (Rohwer et al., 2002) and/or supply nutrients, including nitrogen and phosphorous, that are not provided by the coral-symbiotic zooxanthellae (Sorokin, 1973, 1978; Anthony, 1999, 2000; Rosenfeld et al., 1999; Anthony & Fabricius, 2000). Mucus secreted by coral invariably becomes enriched with microorganisms (Coffroth, 1990). The changing microbial fauna found on the coral surface may be the result of mechanical sloughing or competitive interactions among microorganisms (Rohwer et al., 2001, 2002; Johnston & Rohwer, 2007). It has been hypothesized by Rohwer (2002) that microbial communities found on the surface of corals may play a role in the coral defense mechanism, possibly by occupying niches or through the production of antimicrobial substances.

Indeed, numerous studies have reported the antimicrobial activity of a variety of ‘extracts’ from various organisms, such as sponges (Kelman et al., 2001), soft corals (Kelman et al., 1998; Puglisi et al., 2002; Harder et al., 2003) and scleractinian corals (Koh, 1997; Geffen & Rosenberg, 2005; Marquis et al., 2005). Antimicrobial activity reported in sponges was found, in some cases, to result from the activities of their prokaryotic symbionts (Chelossi et al., 2004). Kelman (2006) showed that the majority of extracts from six soft corals demonstrated high antimicrobial activity against marine bacteria isolated from surrounding seawater. On the other hand, these authors also showed that extracts from six stony corals exhibited little or no antimicrobial activity. They, therefore, suggested that stony corals may have developed different means by which to combat potential microbial pathogenic infections. One of these approaches may depend on the mucus-associated microbial community, as demonstrated by Ritchie (2006), who found that mucus obtained from the coral Acropora palmata inhibited the growth of potentially invasive microorganisms and that 20% of cultured bacteria from such mucus displayed antimicrobial activity. Accordingly, a novel mucus-mediated medium was reported and found to be selective for isolates that produce antibiotics. Thus, the presence of microbial populations on the mucus surface of many invertebrates may play a part in their defense strategies. According to the ‘Coral Probiotic Hypothesis’ (Reshef et al., 2006), coral lives in a symbiotic relationship with a diverse metabolically active microbial population, such that when environmental conditions are altered, the microbial biota undergoes changes that aid coral holobiont fitness. It is thus possible that in some cases, corals can fend off disease or develop resistance to certain microbially-driven diseases by actively or passively modifying the profile of their associated microbial communities (Reshef et al., 2006).

The aim of this study was to assess the antibacterial potential of coral mucus-cultivable bacteria and to further expand our understanding of their possible contribution to the coral holobiont.

Materials and methods

Bacterial sampling and isolation

Samples were collected from the coral mucus layer of different coral species, as well as from the water column and benthic area around the corals, from the reef adjacent to the Inter-University Institute for Marine Science in the Gulf of Eilat (29°51′N, 34°94′E), at depths of 3–19 m. The coral samples collected included six stony corals (Platygyra sp., Porites sp., Fungia granulosa, Favia sp., Stylophora sp. and Pocillopora sp.) and two soft corals (Rhytisma fulvum fulvum and Xenia sp.). Samples were collected at three time points; August 2004, March 2005 and August 2005. Bacterial isolates obtained in August 2004 were used for antibacterial screening via the overlay technique, while isolates obtained during March and August 2005 were screened using the drop technique. Several isolates were screened using both techniques. Mucus samples were collected from seemingly healthy corals from the upper portion of the coral colony or polyp. No bleaching events occurred during or between sampling periods. Mucus samples were collected as previously described by Barneah (2007), using plastic bacteriological loops to rub off the coral surface mucus layer. The loops (three for each sample) were transferred to sterile 15-mL plastic vials containing a minimal amount of seawater to produce a sample of <1 mL, and sealed underwater. Seawater and sediment from the benthic area adjacent to the main coral colony were also sampled. Serial dilutions were performed using autoclaved artificial seawater followed by spreading 100 μL of the dilute over marine agar (Himedia Laboratories, Mumbai, India) plates at 50% and 10% concentrations. Seawater samples were treated in the same way as the mucus samples, diluted and spread over agar plates. Sediment samples were combined with autoclaved artificial sea water, firmly mixed, followed by treating the aqueous upper phase as stock for subsequent dilution, and lawn spread over marine agar plates. After incubation for 2–5 weeks at room temperature, bacteria exhibiting unique colony morphology (in comparison with other colonies on the same agar plates and at the same dilution) were subcultured for purification under the same growth conditions, as previously described by Ritchie (2006).

Antibacterial screening

Screening for antibacterial activity was performed against common indicator bacteria, including Bacillus cereus, Escherichia coli, Serratia marcescens and Staphylococcus aureus, as well as the marine pathogen, Vibrio coralliilyticus. Two techniques were used for the screening process, namely overlay with soft agar or the drop technique. The first method is a modification of the overlay technique described by Geis (1983), and involved pouring 8 mL of soft agar [0.7 g agar agar (Hispanagar, Spain), 2 g Luria–Bertani (LB) broth powder (Hy-labs, Rehovot, Israel) 100 mL−1 deionized water] mixed with indicator bacteria over marine agar plates containing 5-week-old colonies of marine isolates from coral mucus. For this method, only B. cereus and E. coli served as indicator bacteria. Antibacterial activity was defined by measuring the radius of the inhibition zone of the growth of indicator bacteria around the colony after 18 h of incubation at the optimal growth temperature for the indicator bacteria (30 and 37 °C for B. cereus and E. coli, respectively).

The second method, referred to as the drop technique, involved incubation of marine isolates in 100% marine broth (Himedia Laboratories) for 72 h at 26 °C with agitation (120 r.p.m.). After sufficient growth, 10 μL of the marine isolates were placed on agar plates containing 100 μL of an overnight culture of indicator bacteria spread as a lawn, followed by incubation for 18 h at the optimal temperature for that indicator. Isolates obtained from seawater and sediment were used as negative controls for antibacterial screening, using the drop technique. Supernatant, produced by filtrating marine isolate cultures through sterile 0.2-μm filters, was also tested for antibacterial activity using this technique. Activity was detected by the formation of an inhibition zone where the drop was placed (but not outside the drop area) and recorded qualitatively, based on degree of transparency. Data is expressed on an arbitrary scale of 0–5, with 0 representing no inhibition and 5 corresponding to maximal transparency, i.e. no growth of the indicator bacteria. Each experiment was performed three to five times in duplicate or triplicate.

All indicator bacteria were cultured on LB agar (Hy-labs) plates, except for V. coralliilyticus, grown on TCBS agar (Himedia Laboratories), at their optimal growth temperatures (30 °C for V. coralliilyticus and S. marcescens and 37 °C for S. aureus).

Antimicrobial activity characterization

Selected marine isolates that demonstrated activity against B. cereus were chosen for further characterization of their antibacterial activity. Isolates demonstrating high antibacterial activity and/or unique colony morphology and/or isolates originating from corals known to produce antimicrobial activity were selected for subsequent characterization. Additional tests were performed to determine optimal growth temperature, sensitivity to commercial antibiotics and temperature of inactivation, as described below. Each experiment was performed two to three times in duplicate. Optimal growth temperatures were determined by culturing the active isolates in 100% marine broth tubes with agitation (120 r.p.m.) at three growth temperatures (18, 26 and 37 °C). Growth and antibacterial activity were reassessed using the drop technique, as described above.

Sensitivity to common commercial antibiotics was tested according to disc diffusion assay (Awerbuch et al., 1988), using paper discs containing the following concentrations of antibiotics (Oxoid, Basingstoke, Hants, UK): TE30 – tetracycline, 30 μg; SXT25 – sulphamethoxazole trimethoprim, 25 μg; NA30 – nalidixic acid, 30 μg; E15 – erythromycin, 15 μg; C30 – chloramphenicol, 30 μg; MET5 – methicillin, 5 μg; AMP10 – ampicillin, 10 μg. After spreading 100 μL of the tested isolates on a 50% marine agar plate, antibiotic discs were placed on top of the agar for a 24-h period of incubation.

Determination of inactivation temperatures were performed after culturing the active isolates for an incubation period of 48 h at an optimal growth temperature of 26 °C and allowing them to reach stationary phase. Inactivation temperatures were determined by incubation for 10 min at 26, 30, 40, 42, 45, 48, 50, 55, 60, 80 or 100 °C, and reassessing antibacterial activity against B. cereus using the drop technique.

Bacterial molecular identification

Following incubation in 100% marine broth, DNA was extracted from pure cultures using an UltraClean Microbial DNA Isolation Kit (MoBio Laboratories, Solana Beach, CA), according to the manufacturer's instructions. DNA was PCR amplified with a Mastercycler gradient thermocycler (Eppendorf, Westbury, NY), using specific 16S rRNA gene primers for bacteria, namely, forward primer 8F, as described by Felske (1997) but modified by shortening three base pairs (CAC) at the 5′-end, and reverse primer 1512R, as described by Felske (1997). Primers were obtained from Sigma-Genosys. The reaction mixture included 12.5 μL ReddyMix (PCR master mix containing 1.5 mM MgCl2 and a 0.2 mM concentration of each deoxynucleoside triphosphate; ABgene, Surrey, UK), 1 pmol each of the forward and reverse primers, 1–2 μL of the sample preparation, and water to bring the total volume to 25 μL. An initial denaturation hot start of 4 min at 94 °C was followed by 30 cycles of the following incubation pattern: 94 °C for 40 s, 54 °C for 50 s and 72 °C for 2 min. A final extension at 72 °C for 20 min concluded the reaction. PCR products were purified using the Wizard SV Gel kit and the PCR CleanUp System (Promega, Madison, WI). DNA sequencing was performed using an ABI PRISM dye terminator cycle sequencing ready reaction kit with AmpliTaq DNA polymerase FS and an ABI model 373A DNA sequencer (Perkin-Elmer). Sequences were then compared with GeneBank database (NCBI blast) to find best matches and aligned using molecular evolutionary genetics analysis3.1 software (mega) (Kumar et al., 2004). A phylogenetic tree was constructed by the neighbor-joining method (Saito & Nei, 1987), based on sequences of 256 bp.

Nucleotide sequence accession number

The 16S rRNA gene sequences from this study have been deposited in the NCBI GeneBank database under accession numbers FJ041063FJ041108.


Throughout the three sampling periods, six different species of stony corals and two species of soft corals were sampled. Sediment and seawater surrounding the corals were sampled as well.

Seventy bacterial isolates originating from stony and eight from soft coral mucus layers were screened for antibacterial activity using the overlay technique. Nineteen (27%) isolates from the stony corals and one (12.5%) from the soft corals displayed antibacterial activity. The percentage of active bacterial isolates obtained from each coral species are presented in Fig. 1. Seventeen isolates (16 from stony corals and one from the soft coral) were found to be active against B. cereus, two isolates from stony corals were active against E. coli and only one isolate (from a stony coral) demonstrated antibacterial activity against both indicator bacteria. The highest percentage of active isolates originated from stony Pocillopora sp. and Platygyra sp. (44% and 38%, respectively).


Percent of active bacterial isolates from massive, branching and soft corals, able to inhibit the growth of indicator bacteria, detected according to the two screening techniques employed. Data are presented as a summary of total antibacterial activity. n, number of screened bacterial isolates.

Using the drop technique, 21 of the 84 (25%) screened isolates from stony corals and three of the 22 (13%) isolates from soft corals displayed high activity against indicator bacteria (Figs 1 and 2). None inhibited the coral pathogen, V. coralliilyticus (data not shown). None of the bacterial supernatants exhibited antibacterial activity using the drop technique. In addition, only one of the 14 (7%) sea water isolates demonstrated antibacterial activity, while none of the 26 sediment isolates were active (Fig. 2). Similar to results obtained by the overlay technique, the drop technique revealed that stony corals produced the highest percentage of active isolates (25%) when compared with soft corals (13%) (Fig. 2). The highest percentage (42%) of active marine isolates originated from the stony coral, Favia sp. (Fig. 1).


A comparison of active bacterial isolates from stony corals, soft coral, seawater and sediment, assessed using the drop technique. n, number of screened bacterial isolates.

The sty-1-10-01 isolate displayed activity only against E. coli when using the overlay technique, whereas using the drop technique, the isolate was found to be active against E.coli, B. cereus and S. aureus. This isolate originated from the stony coral, Stylophora sp., and was found to be related (99% similar) to the Vibrio campbellii 16S rRNA gene sequence (accession number AY738129).

16S rRNA gene sequences of 46 marine bacterial isolates (38 with and eight without antibacterial activity) were sequenced and subsequently aligned to construct a phylogenetic tree (Fig. 3). Most active isolates (89.5%) were found to belong to class Gammaproteobacteria, while the rest belong to the phylum Firmicutes. Nine Gram-negative isolates belonging to Vibrio and Pseudoalteromonas genera (Fig. 3, marked as black triangles) possessed high activity against both Gram-negative and Gram-positive indicator bacteria. Twenty-nine isolates (Fig. 3, marked as black circles) revealed antibacterial activity against Gram-positive bacteria alone.


Phylogenetic tree of marine bacterial isolates based on 16S rRNA gene sequences and best matches, according to the NCBI database. Family taxa appear on the right. Color map: ▲, isolates active against Gram-positive and Gram-negative indicator bacteria; ●, isolates active against Gram-positive indicator bacteria; ◻, inactive isolates. The bar represents five substitutions per 100 nucleotide positions. Bootstrap probabilities are indicated at branch nodes (values under 50% were excluded). The phylogenetic tree was constructed based on length sequences of 256 bp.

Two (fav-2-10-05 and fav-2-50-12) out of three isolates originating from Favia sp. and one isolate (sty-2-10-05) from Stylophora sp. demonstrated activity against Gram-positive indicator bacteria. All four of these isolates are related to the genus Shewanella (93–95% similarity, according to the 16S rRNA gene sequence) (Fig. 3).

Sixteen bacterial isolates (Table 1) were selected for further examination and tested for optimal growth temperature, sensitivity to commercial antibiotics and inactivation temperature. Optimal growth temperature was found to be 18–26 °C, and all isolates showed a decrease in activity and growth when grown at 37 °C. Determination of inactivation temperature was performed for eleven isolates that demonstrated activity against B. cereus (Table 2). Antibacterial activity was found to be optimal at 26 °C, slightly decreased at 30 °C for three isolates (pla-1-10-05, sty-2-10-05 and Xen-2-10-04B), and completely lost at 42–45 °C for all but three isolates. Two isolates (pff-3-10-02 and xen-2-10-04B) lost activity at 50 °C, while one isolate (pff-3-10-01) demonstrated partial inactivation of antibacterial activity at 60 °C, yet completely lost activity at 80 °C.

View this table:

Sensitivity of bacterial isolates to commercial antibiotics

IsolateBest match (accession number)% Similarity to best matchAntibiotics (concentrations)
TE (30 μg)SXT (25 μg)NA (30 μg)E (15 μg)C (30 μg)MET (5 μg)AMP (10 μg)
fav-2-10-05Shewanella sp. Fun-119 (DQ107396)989.521171333815
fav2-1-10-13Vibrio coralliilyticus sp. LMG 10953 (AJ316167)991316.510.51327.5RR
fav-2-50-06Vibrio sp. 98CJ11027 (AF246980)99919101023RR
fav-2-50-07Vibrio alginolyticus strain UQM 2770 (AY264938)9910199.51022RR
fav-2-50-10Shewanella sp. Fun-119 (DQ107396)9991716.51131R13
fav-2-50-12Shewanella sp. Fun-119 (DQ107396)989201310.526RR
fug-1-10-16Vibrio sp. LMG 10953=coralitycus (AJ316167)991222.510.512.528RR
pla-1-10-05Pseudoalteromonas sp. 03/034 (AJ874351)9910189.5919.5RR
pla-1-10-07Pseudoalteromonas sp. 03/034 (AJ874351)991013101624814
poc-1-10-23Pseudoalteromonas sp. 03/034 (AJ874351)991418162524R18
por-3-50-04Pseudoalteromonas sp. OC-C3-5 (DQ319016)98819.513.514.525.5R14.5
pff-3-10-01Acinetobacter sp. ATCC 31012 (AF542963)99RRRR18RR
pff-3-10-02Acinetobacter sp. ATCC 31012 (AF542963)991024989R8
sty-1-10-01Vibrio campbellii strain 90-69B3 (AY738129)998.521.59.5921.5RR
sty-2-10-05Shewanella fidelia strain KMM3582T (AF420312)957191811.527RR
xen-2-10-04BStenotrophomonas maltophilia (AY738261)998.5199.59.518RR
  • * First three letters of isolate name represents the coral species from which the isolate was obtained: fav, Favia sp.; fug, Fungia granulosa; pla, Platygyra sp.; poc, Pocillopora sp.; por, Porites sp.; pff, Rhytisma fulvum fulvum; sty, Stylophora sp.; xen, Xenia sp.

  • Applied concentrations are noted in parentheses. Sensitivity is expressed as the size of the inhibition zone (diameter in mm), including the antibiotic disc (6 mm).

  • R, resistance to antibiotics.

View this table:

Inactivation temperatures of marine isolates

Temperature of inactivation (°C)454245424242505080
  • Inactivation temperature was determined by incubating a stationary culture of the isolates for 10 min at 26, 30, 40, 42, 45, 48, 50, 55, 60, 80 or 100°C, and reassessing antibacterial activity against Bacillus cereus, using the drop technique.

Sensitivities to commercial antibiotics at known concentrations were tested using antibiotic diffusion discs (Table 1). Most isolates were sensitive, at different levels, to sulphamethoxazole trimethoprim, nalidixic acid, erythromycin and chloramphenicol, while only two isolates were sensitive to methicillin. One isolate, pff-3-10-01, originating from the soft coral R. fulvum fulvum, demonstrated resistance to all antibiotics except chloramphenicol. This highly resistant isolate was found to be related to Acinetobacter sp. American Type Culture Collection 31012 (99% similarity according to 16S rRNA gene sequence).


This study has addressed the antibacterial activity occurring in cultivable coral mucus-associated bacteria and showed that between 25% and 70% of cultivable bacteria from scleractinian coral mucus display antibacterial activity (Fig. 1). In addition, because none of the supernatant forms of active cultures ever demonstrated antibacterial activity, we raise the possibility that active bacteria may not necessarily secrete compounds to the environment but may be active through additional mechanisms. We further show that the assessment of bioactivity depends on the isolation and screening method used. Both the overlay technique, which utilizes diffusion of active compounds, and the drop technique, which is not restricted by diffusion, revealed that, on average, 25–27% and 13% of the isolates originating from stony and soft corals, respectively, possess antibacterial activity (Figs 1 and 2). These values are compatible with the findings presented by Ritchie (2006), according to which 20% of cultured bacteria from the mucus layer of the coral A. palmata demonstrated antibiotic activity against indicator strains. Interestingly, Kelman (2006), testing for antimicrobial activity in whole coral extracts, found that extracts from soft coral tissue were more active than scleractinian (stony) coral tissue extracts. These authors suggested that different mechanisms of antimicrobial activity against pathogens may be at play in these stony corals. Such diverse mechanisms could explain the fact that a higher percentage of antimicrobial activity was detected among bacterial isolates collected from stony coral mucus. This raises the possibility that production and secretion of antimicrobial compounds by mucus-associated bacteria is part of the scleractinian coral's defense strategy against pathogens (Rohwer et al., 2002; Reshef et al., 2006; Ritchie, 2006).

The overlay and drop screening techniques reported the same percentage (c. 25%) of antibacterial activity when similar numbers of isolates from scleractinian corals were tested. There was no overlap (with the exception of one isolate) in activity between the isolates tested by the two techniques, reaffirming that these techniques test for different mechanisms of secretion of active compounds. Because none of the drop technique supernatants demonstrated antibacterial activity, it is likely that the antimicrobial substances produced by the coral mucus-associated microorganisms are cell-bound. Our results also clearly show the need for developing improved methods for assessing the antimicrobial potential of coral mucus-associated bacteria. For example, the overlay technique has the drawback of requiring heating of the soft agar to 45–50 °C before pouring over an agar plate containing active bacterial colonies. If there are active compounds sensitive to heat present, such heating may damage their antibiotic properties and cause an underestimation of antimicrobial activity. Nonetheless, we demonstrate that by using several techniques, better understanding of the antibiotic potential of coral mucus-associated bacteria is achieved.

It is assumed that soil bacteria secrete antimicrobial compounds to their surrounding environment in order to deter pathogens and gain advantage against invasive species in competition over a given niche (Wiener, 1996). Soil may be considered to be a semi-diffusible environment, which constrains active compounds to the microorganism's vicinity. In a diffusible environment, such as the marine environment and more specifically, the mucus layer of corals, the aqueous nature of the medium may cause dilution of the antibiotic (Brown & Bythell, 2005). In such cases, the secretion of antibiotics and other antimicrobial compounds outside the bacterial cell will result in high dilution of the active substances, as well as loss of nutrients and energy invested in the synthesis process. It is more likely that in such diffusible environment, bacteria secrete nondiffusible active compounds and maintain them by attachment to their cell wall or membrane. This arrangement would result in reduced dilution and allow for the contact of such active compounds with potential competitors (Wiener, 1996). The application of overlay and drop techniques apparently indicates that there is more than one type of mechanism protecting the holobiont at any time, and that the overall protection of the coral is, in practice, much higher than previously reported, because a total of c. 50% of cultivable bacteria were found to be active using these techniques. Variations in mechanisms of antibacterial activity (and possible synergistic effects between them) may further enhance the holobiont's ability to protect itself against pathogens, as well as to adapt to environmental changes.

The effect of temperature on antimicrobial activity was tested for several isolates. Antibacterial activity was found to be optimal at 26 °C and slightly decreased at 30 °C, while inactivation temperature was, in most cases, found to be 42–45 °C (Table 2). In one case, partial inactivation occurred at 60 °C with complete loss of activity occurring at 80 °C. Ritchie (2006) found that mucus collected from the coral A. palmata during a period of elevated sea water temperature (28–30 °C) did not show significant antibiotic activity, as opposed to the antibiotic activity detected in the mucus of colonies from normal (22–25 °C) sea temperatures. Ritchie also found that about 20% of cultured bacterial isolates from nonbleached colonies of A. palmata displayed antibacterial activity and that this value was reduced to only 2% during a bleaching event when sea surface temperatures were elevated to a mean of 30 °C (Ritchie, 2006). Our results, in addition to these earlier findings, suggest that the antimicrobial compounds may be temperature sensitive. It is also possible that when the temperature rises above 26 °C, the amount and stability of antimicrobial compounds in coral mucus decreases, leaving the coral more sensitive to pathogens.

The existence of antibacterial metabolites in the natural habitat may lead to resistance of selected species against such compounds. To test this concept, the resistance of selected active isolates to commercial antibiotics was tested (Table 1). Most of the isolates demonstrated resistance to methicillin and ampicillin. Such resistance suggests the existence of a form of these antibiotics or of a similar compound relying on a mode of action in the isolates' niche. Curiously, one isolate, found to be related to Acinetobacter sp. (99%), demonstrated multiple resistance to most antibiotics tested. This isolate, obtained from the soft coral R. fulvum fulvum, also exhibited inhibitory activity at relatively high temperatures.

Those bacteria in the coral mucus layer that produce antimicrobial compounds may also assist the coral holobiont as a first line of defense against pathogens and fouling organisms. Through occupation of the mucus layer niche, such beneficial microorganisms may competitively prevent pathogens from physically entering the coral or attaching to the coral epidermis. The production of antibacterial compounds provides the producing bacteria advantage over pathogen-sensitive strains, possibly assisting in competition over space and nutrition. This hypothesis is in agreement with results presented by Wiener (1996), demonstrating that antibiotic-producing strains possess significant advantage against an invading sensitive strain. We further propose that this trait of the coral mucus-associated bacteria is a desired characteristic for the coral host, increasing its resistance to invading pathogens.

According to our phylogenetic analysis, most active isolates include Gram-negative bacteria, more specifically belonging to the genera Vibrio and Pseudoalteromonas of the Gammaproteobacteria class. Our results, however, do not likely reflect a true representation of active bacterial isolates in the coral mucus layer as they only include cultivable bacteria, successfully grown on marine agar plates. Still, it is of note that isolates obtained from different corals, demonstrating different patterns of antibacterial activity, were presumably found to belong to the same species (as determined by the high percentage of similarity in the 16S rRNA gene sequence). This is the case for the Pseudoalteromonas, V. coralliilyticus and Shewanella clades (Fig. 3). For example, in the Pseudoalteromonas sp. clade, eight bacterial isolates originating from four different corals (Pocillopora sp., Platygyra sp., Stylophora sp. and Porites sp.) and seawater were all found to be active against B. cereus. A related biofilm-associated marine bacterium, Pseudoalteromonas tunicata, is known to produce a range of biocidal compounds as well as an antifouling agent (Holmstrom et al., 1998; Burmolle et al., 2006, and references therein). In the case of the V. coralliilyticus sp. clade, five isolates originated from four different corals (Pocillopora sp., Favia sp., Platygyra sp. and F. granulosa). One isolate was found to be inactive, while the other four demonstrated activity against Gram-positive bacteria using the drop technique. According to 16S rRNA gene analysis, most of the isolates were found to be related to V. coralliilyticus (accession numbers AJ316167 and AJ440004), with 97–99% similarity. It is interesting to note that V. coralliilyticus (accession number AJ440004), isolated from crushed tissue of a diseased coral is a known coral pathogen, causing rapid tissue lysis in Pocillopora damicornis (Ben-Haim et al., 2003). The third group, the Shewanella sp. clade, also presents such patterns, with four isolates being derived from two different corals, i.e. Favia sp. (three isolates) and Stylophora sp. (one isolate). Three of the isolates demonstrated antibacterial activity against Gram-positive indicator bacteria, while one isolate was inactive.

The isolates in each of the three clades listed above may be designated as ubiquitous coral mucus-associated bacterial symbionts because they are all related to the same bacterial species, according to 16S rRNA gene sequence comparison, despite originating from different coral species. Rohwer (2002) analyzed uncultured bacteria from three different coral species to show that some bacteria form species-specific associations with their coral hosts, while others do not. This duality also exists in our study, with examples of species-specific relationships, in addition to the generalist coral-associated bacterial symbionts. In the present study, closely related bacteria isolated from the same coral species exhibited different patterns of antibacterial activity. This was the case for the V. alginolyticus sp. and the Planomicrobium sp. clades (Fig. 3). Each clade contains three isolates originating from the same coral species (Favia sp. and F. granulosa sp., respectively), with only two isolates exhibiting antibacterial activity.

The presence of these clades in the mucus of a number of colonies from the same coral species suggests the existence of specific coral mucus-associated bacteria that regulate mucus layer bacterial populations through antimicrobial activities. These active bacteria may thus be referred to as species-specific coral symbionts. This concept is in agreement with previous studies. Ritchie & Smith (1997) suggested that Caribbean coral species contained unique and species-specific mucus-associated microbial communities. Rohwer (2001) showed (using culture-independent methods) that different colonies of the coral Montastraea franksi contained bacterial communities that were similar over distance and time. Webster & Bourne (2007) observed such conserved cold-water coral-associated bacterial groups within and between sites in Antarctica.

The Coral Probiotic Hypothesis suggested by Reshef (2006) posits the existence of a symbiotic relationship between the coral and diverse metabolically active microorganisms. The theory further posits that corals may adapt rapidly to changes in environmental conditions by altering the composition of their symbiotic communities. Therefore, it is likely that this dynamic relationship can shift in response to environmental changes to allow an abundance of microbial species that allow coral holobiont to adapt to its new conditions more efficiently. Recent findings by Ritchie (2006), along with our results presented here, support this hypothesis. We further extend this hypothesis by suggesting that in addition to diffusible antimicrobial substances, nondiffusible cell-bound substances may also be produced by the coral mucus associates. These substances may then act additively or synergistically as a first line of defense against invading pathogens for the coral holobiont.


This work was supported by ISF grants 511/02-1 and 1169/07, and by a Kreitman scholarship to M.S.-O. from the Kreitman school of Advanced Graduate Studies at Ben-Gurion University of the Negev, Israel. The authors wish to thank the IUI, Eilat, Israel, for use of their facilities; N. Siboni and E. Ben-Dov for sample collection, technical support and guidance; and E. Kramarsky-Winter and R. Orland for helpful comments on the manuscript.


  • Editor: Riks Laanbroek


View Abstract