Diversity and bioactive potential of endospore-forming bacteria cultured from the marine sponge Haliclona simulans


  • R.W. Phelan,

    1.  Department of Microbiology, University College Cork, Cork, Ireland
    2.  Biomerit Research Centre, University College Cork, Cork, Ireland
    Search for more papers by this author
  • J.A. O’Halloran,

    1.  Department of Microbiology, University College Cork, Cork, Ireland
    2.  Biomerit Research Centre, University College Cork, Cork, Ireland
    Search for more papers by this author
  • J. Kennedy,

    1.  Marine Biotechnology Centre, Environmental Research Institute, University College Cork, Cork, Ireland
    Search for more papers by this author
  • J.P. Morrissey,

    1.  Department of Microbiology, University College Cork, Cork, Ireland
    2.  Marine Biotechnology Centre, Environmental Research Institute, University College Cork, Cork, Ireland
    Search for more papers by this author
  • A.D.W. Dobson,

    1.  Department of Microbiology, University College Cork, Cork, Ireland
    2.  Marine Biotechnology Centre, Environmental Research Institute, University College Cork, Cork, Ireland
    Search for more papers by this author
  • F. O’Gara,

    1.  Department of Microbiology, University College Cork, Cork, Ireland
    2.  Biomerit Research Centre, University College Cork, Cork, Ireland
    3.  Marine Biotechnology Centre, Environmental Research Institute, University College Cork, Cork, Ireland
    Search for more papers by this author
  • T.M. Barbosa

    1.  Department of Microbiology, University College Cork, Cork, Ireland
    Search for more papers by this author

Fergal O’Gara, BIOMERIT Research Centre, Department of Microbiology, University College Cork, College Road, Cork, Ireland. E-mail: f.ogara@ucc.ie
Teresa M. Barbosa, Department of Microbiology, University College Cork, College Road, Cork, Ireland. E-mail: t.barbosa@ucc.ie.


Aims:  Despite the frequent isolation of endospore-formers from marine sponges, little is known about the diversity and characterization of individual isolates. The main aims of this study were to isolate and characterize the spore-forming bacteria from the marine sponge Haliclona simulans and to examine their potential as a source for bioactive compounds.

Methods and Results:  A bank of presumptive aerobic spore-forming bacteria was isolated from the marine sponge H. simulans. These represented c. 1% of the total culturable bacterial population. A subgroup of thirty isolates was characterized using morphological, phenotypical and phylogenetic analysis. A large diversity of endospore-forming bacteria was present, with the thirty isolates being distributed through a variety of Bacillus and Paenibacillus species. These included ubiquitous species, such as B. subtilis, B. pumilus, B. licheniformis and B. cereus group, as well as species that are typically associated with marine habitats, such as B. aquimaris, B. algicola and B. hwajinpoensis. Two strains carried the aiiA gene that encodes a lactonase known to be able to disrupt quorum-sensing mechanisms, and various isolates demonstrated protease activity and antimicrobial activity against different pathogenic indicator strains, including Clostridium perfringens, Bacillus cereus and Listeria monocytogenes.

Conclusions:  The marine sponge H. simulans harbours a diverse collection of endospore-forming bacteria, which produce proteases and antibiotics. This diversity appears to be overlooked by culture-dependent and culture-independent methods that do not specifically target sporeformers.

Significance and Impact of Study:  Marine sponges are an as yet largely untapped and poorly understood source of endospore-forming bacterial diversity with potential biotechnological, biopharmaceutical and probiotic applications. These results also indicate the importance of combining different methodologies for the comprehensive characterization of complex microbial populations such as those found in marine sponges.


With the world-wide increase in antibiotic resistance and the emergence of multidrug-resistant strains of bacteria, the scientific community is facing new challenges to identify and develop novel therapeutic approaches to help combat the spread of infectious diseases. Recently, this has resulted in an increased focus on as-yet under explored environments, such as marine sponges, which are among the animal kingdom’s most important producers of bioactive metabolites (Kennedy et al. 2007; Taylor et al. 2007). While many of these compounds have been isolated from marine sponges themselves, there is ever increasing evidence to suggest that a proportion of them actually originate from bacterial symbionts of these marine invertebrates (Hentschel et al. 2006; Taylor et al. 2007; Muscholl-Silberhorn et al. 2008; Lee et al. 2009). As a result, a large number of culture-dependent and culture-independent studies have been undertaken to increase our understanding of the abundance, diversity and specificity of the microbial population associated with marine sponges, with a view to identifying and potentially exploiting novel bioactivities (Hentschel et al. 2001; Kennedy et al. 2008, 2009; Muscholl-Silberhorn et al. 2008; Lee et al. 2009; Menezes et al. 2009).

The isolation of endospore-forming bacteria from different sponge samples has been frequently reported, but their abundance seems to vary widely between sponges (Muscholl-Silberhorn et al. 2008; Zhu et al. 2008; Lee et al. 2009; Menezes et al. 2009), and detailed characterization of the isolates is frequently overlooked. Some reports describe specific sponges as Bacillus-rich habitats, where the culturable population is largely dominated by bacteria from this genus, while others have either failed to identify Bacillus or describe them as only a minor component of the sponge-associated bacteria (Muscholl-Silberhorn et al. 2008; Zhu et al. 2008; Lee et al. 2009; Menezes et al. 2009). This group of bacteria is also frequently missed by culture-independent studies (Kennedy et al. 2008; Zhu et al. 2008; Montalvo and Hill 2011).

The genus Bacillus comprises a diverse collection of Gram-positive, aerobic to facultative, endospore-forming, rod-shaped bacteria. Although routinely described as saprophytic soil organisms, they are also found in many other environments, including the gastrointestinal tract and water habitats (Ivanova et al. 1999; Barbosa et al. 2004, 2005; Kennedy et al. 2008). This ubiquitous nature is frequently attributed to air or water dispersal of spores, which are extremely resilient structures with exceptional longevity in the environment. However, emerging evidence suggests that the ecological niche for Bacillus strains might be considerably wider than that traditionally associated with soil. Particular species or strains may have evolved in response to specific environmental cues and are likely to be more adapted for particular habitats (Tam et al. 2006; Hong et al. 2009a; Alcaraz et al. 2010).

Species of the genus Bacillus are renowned for either the production of chemical agents with antimicrobial properties or enzymes that are of biotechnological interest (Westers et al. 2004; Stein 2005). Interest in endospore-forming bacteria, particularly Bacillus, has seen a comeback in recent years, as spore probiotic preparations are currently being used in human therapy, animal production and aquaculture (Barbosa et al. 2005; Hong et al. 2005; Lalloo et al. 2009).

Thus, the aims of this study were to isolate and characterize the culturable spore-forming population of the marine sponge Haliclona simulans, with the aim of determining their abundance and diversity, and to examine their potential as a source for novel bioactives, including those that might be effective against food pathogens.

Materials and methods

Sample collection and isolation of sponge-associated sporeformers

Samples of H. simulans (class Demospongiae, order Haplosclerida and family Chalinidae) were collected during scuba diving in Gurraig Sound Kilkieran Bay, Galway, on the west coast of Ireland, at a depth of 15 m (53°18·944′N and 09°40·140′W) as previously described (Kennedy et al. 2008). Whole sponge samples were kept in sea water, transported to the laboratory on ice and immediately stored at −80°C until further processing.

A sample of sponge tissue (1 g) was homogenized by grinding with a sterile porcelain mortar and pestle in 9 ml of sterile artificial sea water [ASW, 3·33% (w/v) artificial sea salts – Instant Ocean Brand], corresponding to an initial 1 in 10 dilution. The sponge homogenate was subsequently serially diluted in ASW to 10−5, before and after heat or ethanol treatment to select for sporeformers, and 100-μl aliquots of the different dilutions were plated onto Marine agar (MA) 2216 (Difco) and SYP-SW Agar [1% (w/v) starch, 0·4% (w/v) yeast extract, 0·2% (w/v) peptone and 3·33% (w/v) artificial sea salts – Instant Ocean Brand, 1·5% (w/v) agar; (Kennedy et al. 2009)]. For heat treatment, aliquots of the initial sponge homogenate were incubated at 65°C for 30 min. For ethanol treatment, the sponge homogenate was further diluted 1 : 1 in ethanol [final concentration, 50% (v/v)] and incubated for 1 h at room temperature with constant agitation (Barbosa et al. 2005). Plates were incubated at room temperature and 30°C, respectively, for up to 6 weeks. Colonies were examined and picked daily for the first week and weekly thereafter.

Colonies representing different morphologies (colour, texture, shape and size) under the different selective conditions were purified by restreaking as required on MA plates. All isolates were checked for catalase activity by resuspension of a fresh colony in a 3% solution of hydrogen peroxide (Sigma), before storage at −80°C in Marine broth (MB) with 25% glycerol.

Strains MMA7 and CH8a identified and characterized in this work were obtained from a previous study that targeted the total culturable bacterial population from H. simulans and were isolated in modified MA and chitin agar (Kennedy et al. 2009), respectively.

Growth properties, sporulation and antibiotic susceptibility testing

Sponge-associated strains were routinely grown and maintained aerobically, on Difco™ MA and MB (Difco 2216), at 30°C, unless otherwise stated. To check for the ability to grow under different conditions, sponge isolates were initially grown for 24 h at 30°C in MB, and 5 μl of these cultures was spotted onto the different plates and incubated under the specified conditions. To test for temperature tolerance, plates were incubated at 30, 40 and 50°C, respectively. Salt requirement and tolerance was established by monitoring the growth of the isolates at 30°C on LB plates supplemented with 0, 1, 2, 4, 6, 8, 10, 12 and 15% (w/v) NaCl or 0, 25, 50, 100, 150 and 200% ASW (100% sea water contains 3·33% artificial sea salts). Susceptibility to tetracycline and erythromycin was determined by spotting the MB cultures of the different isolates onto Muller–Hinton (MH, Merck, Darmstadt, Germany) plates supplemented with different concentrations of the relevant antibiotic, while susceptibility to penicillin G and ampicillin was determined using the disc diffusion method. In both cases, plates were incubated at 30°C. Haemolytic activity was determined by streaking colonies of the isolates from fresh MA plates onto Columbia 5% sheep blood agar plates (Fannin Healthcare, Dublin, Ireland), which were incubated at 30°C. Spore production was investigated by phase-contrast microscopy of isolates grown on Difco sporulation medium (DSM) and on MA. For all of the tests described earlier, plates were examined after 24- and 48-h incubations.

DNA extraction, PCR amplification, DNA sequencing and phylogenetic analysis

Total genomic DNA and plasmid DNA of sponge-associated isolates were extracted from 24-h MB cultures as previously described (Barbosa et al. 2005). All primers used in this study were obtained from Eurofins MWG Operon (Ebersberg, Germany). Colony lysates used as templates for PCR amplification were prepared by resuspending a single colony in 25 μl of sterile distilled water and heating at 95°C for 15 min. Unless otherwise stated, PCR mixtures (50 μl) contained 3 μl of colony lysate or 1–2 μl of genomic DNA as template, 1× BioTaq PCR Buffer (Bioline), 1·5 mmol l−1 of MgCl2, 0·2 mmol l−1 of dNTPs, 0·5 μmol l−1 of each primer and 2·5 U of BioTaq DNA polymerase (Bioline). The universal eubacterial primers 27f (5′-AGA GTT TGA TCM TGG CTC AG-3′, M=C or A) and 1492r (5′-GGT TAC CTT GTT ACG ACT T-3′) (Lane 1991) were used to amplify small-subunit rRNA (16S rRNA) gene sequences of the sponge-associated spore-forming isolates. PCR was carried out under the following cycling conditions: initial denaturation at 94°C for 5 min, followed by 30 cycles of 94°C for 30 s, 52°C for 30 s and 72°C for 45 s, with a final extension at 72°C for 10 min. 16S rRNA gene partial sequences obtained with primer 27f [c. 700–1000 nucleotides (nt) of the 5′ end] were compared with sequences in the GenBank nucleotide sequence database (http://blast.ncbi.nlm.nih.gov/Blast.cgi) using Blast (Altschul et al. 1997; Zhang et al. 2000). Approximately 650 nt of the 5′ end of the 16S rRNA gene (nt 117–751 in the Escherichia coli 16S rRNA gene, accession number X80725) were aligned using ClustalX (Thompson et al. 1997), and neighbour-joining phylogenetic trees were constructed using mega 4 (Tamura et al. 2007). Bootstrap tests were performed 1000 times.

PCR amplification of the lactonase gene, aiiA, was achieved with the primer pair NF (5′- CCG GAT CCA TGA CAG TAA ARA ARC T-3′) and CR (5′- TTT CTG CAG CTA TAT ATA YTC HGG G-3′; R, G or A; Y, T or C; H, A, T or C) (Lu et al. 2006). The PCR and cycling parameters were as described previously for the 16S rRNA gene with the exception of the annealing temperature, which was 55°C.

When required, PCR products were purified using the QIAquick PCR Purification kit (Qiagen GmbH, Hilden, Germany). Sequencing of purified PCR products was performed by GATC Biotech AG (Konstanz, Germany).

Protease activity

Protease activity was screened by spotting 10 μl of a fresh overnight MB culture on a 3% reconstituted skimmed milk (SKM) plate, which was incubated at 30°C, and checked for clear halos after 24- and 48-h incubations.

Antimicrobial activity screening assays

Indicator strains (Table 1) were routinely grown aerobically at 37°C, with the exception of Candida glabrata, which were grown at 30°C, and the clostridia strains, which were grown in an anaerobic jar. Antimicrobial activity of the sponge isolates against the indicator strains was assessed with a colony overlay assay adapted from the method described by Barbosa et al. (2005). Aliquots of 10 μl from 24-h overnight cultures in MB were spotted onto 20 ml MA plates (100 mm), allowed to dry and incubated at 30°C for 24–48 h. Plates were then overlaid with 6 ml soft agar (0·75%) of the required media (according to the indicator strains) previously inoculated with a fresh overnight culture of the different indicator strains. Plates were incubated at 37°C (30°C for C. glabrata), and zones of inhibition around the spots after 5 h and/or 24 h were scored as positive.

Table 1.   Bacterial strains used in this study
BacteriaGrowth media*/origin†
  1. *Standard media were LB, Luria–Bertani; RCM, Reinforced Clostridial media (Merck); SDM, Sabouraud dextrose media (Merck); BHI, Brain heart infusion (Merck).

  2. †Bacterial strains were obtained from the Microbiology Department Culture Collection, University College Cork (MDCC UCC), and from the Bacillus Genetic Stock Centre (BGSC).

Bacillus cereus NCIMB 9373LB/MDCC UCC
Bacillus subtilis NCDO 1769LB/MDCC UCC
Bacillus megaterium ATCC 19213 (BGSC 7A2)LB/BGSC
Staphylococcus aureus NCDO 949BHI/MDCC UCC
Enterococcous faecium NCIMB 11508BHI/MDCC UCC
Listeria innocua DPC 3567BHI/MDCC UCC
Listeria monocytogenes EGDeBHI/MDCC UCC
Clostridium perfringens NCDO 1799RCM/MDCC UCC
Clostridium sporogenes NCDO 1791RCM/MDCC UCC
Escherichia coli NCIMB 15943LB/MDCC UCC
Enterobacter aerogenes NCIMB 10102LB/MDCC UCC
Salmonella Typhimurium LT2LB/MDCC UCC
Pseudomonas aeruginosa PAO1LB/MDCC UCC
Candida glabrataSDM/MDCC UCC

Nucleotide sequence accession numbers

All 16S rRNA gene sequences from the sponge-associated sporeformers described in this study have been deposited in the GenBank nucleotide sequence database under the accession numbers JF803846JF803875.


Isolation of sporeformers from the marine sponge H. simulans

Following either heat or ethanol treatments for the selection of sporeformers, sponge samples were plated on different media and incubated aerobically at different temperatures together with untreated samples. The majority of colonies became visible only after 3 days of incubation, with numbers increasing slightly between the days 3 and 8, before stabilizing thereafter through the subsequent weeks. Colony numbers on day 8 of incubation reflected a total bacterial count in the order of 105 for untreated plates, with the numbers on treated plates being in the order of 103, indicating that the relative abundance of presumptive culturable sponge-associated sporeformers is c. 1% of the total culturable bacterial population. No major differences in colony numbers were observed between the different selection treatments, media and incubation temperatures, despite slightly lower numbers for samples treated with ethanol, as compared to those treated by heat. Interestingly, a large number and variety of yellow/orange-pigmented colonies were observed on many plates (Fig. 1a).

Figure 1.

 Morphological diversity of the sporeformers isolated from Haliclona simulans. (a) Colonies on Marine agar (MA) plates from representative strains, including some that were strongly pigmented. (b) Cells and spores morphological diversity. Phase-contrast microscopy of strains grown on Difco sporulation medium (DSM) or Difco MA. All microphotographs are at the same scale. Isolates for which an image is not provided were unable to grow under the tested conditions.

A total of 117 colonies representing distinct morphologies were picked from the different selective conditions and purified by restreaking on MA agar plates. All presumptive sporeformers were rod-shaped, even though some displayed an irregular rod morphology (Fig. 1b). While all isolates were catalase positive, not all were able to grow or sporulate on DSM (Nicholson and Setlow 1990) or sporulated only inefficiently (Fig. 1b).

Phylogenetic analysis of sponge-associated sporeformers

Twenty-eight isolates representing different colony morphologies from within each of the different selection conditions were subsequently selected from the initial 117 isolates for further characterization. Two additional isolates, strains CH8a and MMA7, which were previously isolated from the same sponge material and which were determined to be aerobic rod-shaped sporeformers, were also examined during this study.

Analysis of the 16S rRNA gene sequences from the 30 isolates revealed a large diversity, with the isolates being distributed through a number of different Bacillus and Paenibacillus species (Table 2, Fig. 2). This included recently described species that have been primarily or exclusively reported from marine habitats including B. aquimaris, B. algicola, B. baekryungensis and B. hwajinpoensis (Yoon et al. 2003, 2004; Ivanova et al. 2004). Sequence comparison revealed that a number of the closest matches to the sponge sporeformers were marine strains isolated from sea water, alga, sediments, fish gut and sponges from different geographical regions. Isolates #24 and #43, despite showing a 98% sequence identity to Bacillus sp. L91, only shared 96% identity with its closest typed species, the alkali-tolerant B. lehensis, suggesting that they may represent a novel species. This analysis also highlighted the limitations of 16S rRNA-based phylogenetic analysis for the identification of closely related bacteria with virtually identical 16S rRNAs. Although comparative phylogenetic analysis with type strains of the different species provided some information for potential species association, unambiguous species identification of some isolates was not possible using this data (Table 2, Fig. 2). To gain a better understanding of the diversity of the sporeformers present in H. simulans, additional tests were performed, such as growth and sporulation characteristics, cell morphology and the presence of plasmid DNA.

Table 2.   Identification of sponge-associated sporeformers by 16S rRNA gene sequencing analysis
Strain No.Closest typed species*% ID
  1. *Closest typed species using Blast search of the GenBank nucleotide sequence database. Closest hits may exist for specific isolates, but those were to strains that have not been classified to the species level or to isolates whose taxonomic name was not validly published at the time of the search. Where more than one species hit are shown for any given isolate, they are provided by order of listing on the BLAST search. % ID, percentage of sequence identity.

1Bacillus aquimaris/B. vietnamensis99
2Paenibacillus amylolyticus99
3B. altitudinis/B. stratosphericus/B. pumilus/B. aerophilus100
4B. hwajinpoensis/B. baekryungensis100
8P. xylanexedens/P. pabuli/P. amylolyticus99
10B. hwajinpoensis/B. baekryungensis99
11B. cereus group: B. cereus/B. thuringiensis100
12B. cereus group: B. weihenstephanensis/B. mycoides100
14B. altitudinis/B. pumilus/B. aerophilus100
15B. hwajinpoensis/B. baekryungensis100
22P. amylolyticus100
23B. aquimaris/B. vietnamensis99
24B. lehensis/B. patagoniensis/B. oshimensis96
27B. hwajinpoensis/B. baekryungensis100
42B. algicola99
43B. lehensis/B. patagoniensis96
49B. licheniformis99
51B. cereus group: B. cereus/B. thuringiensis100
52B. cereus group: B. cereus/B. thuringiensis100
56B. altitudinis/B. pumilus/B. aerophilus100
73B. barbaricus/B. arsenicus99
74B. hwajinpoensis/B. baekryungensis100
78B. hwajinpoensis/B. baekryungensis100
83B. hwajinpoensis/B. baekryungensis100
84B. pumilus/B. safensis100
91B. licheniformis99
146B. hwajinpoensis/B. baekryungensis100
147B. licheniformis100
MMA7B. subtilis/B. amyloliquefaciens/B. mojavensis100
CH8aB. cereus group: B. cereus/B. thuringiensis/B. anthracis100
Figure 2.

 Neighbour-joining phylogenetic tree generated by analyzing partial (>650 nucleotides of 5′ end) 16S rRNA gene sequences of sporeformers isolates (in boldface) from Haliclona simulans. All reference strains are type strains and are followed by GenBank accession numbers in parentheses. Accession numbers for the sponge-associated sporeformers are JF803846JF803875. The tree was constructed using maximum composite-likelihood and pairwise deletion. Percentage bootstrap values (>50% only) from 1000 resamplings are indicated at each node. Bar, 5% estimated sequence divergence.

Characterization of sponge-associated sporeformers

Eleven of the thirty isolates were pigmented and the majority of these belonged to species that had previously been isolated from marine habitats. Many of these strains were quite fastidious in their growth requirements. Some were unable to grow, or grew very poorly, in media routinely used to sustain the growth of Bacillus species, such as LB broth, LB agar and sheep blood agar (Table 3). These included isolates that clustered with B. hwajinpoensis/B. baekryungensis, B. algicola and Bacillus sp. isolates #24 and #43. However, despite the phylogenetic relationship, some isolates displayed properties that were distinct from those reported for the associated type strains (Table 3). For example, the H. simulans B. hwajinpoensis/B. baekryungensis isolates were unable to grow, or grew poorly, at 40°C (Table 3). This contrasted to what has been reported for the type strains of these species (Yoon et al. 2004). Discrepancies were also found for isolates that clustered with better represented species. For example, isolates #3, #14 and #56 cluster with B. altitudinis, B. aerophilus and B. stratosphericus type strains (Table 2, Fig. 2). Nevertheless, their abilities to grow at both 40 and 50°C, and in the presence of 8–10% NaCl (Table 3), contrasted with what has been reported for the respective type strains (Shivaji et al. 2006). These were more consistent with properties reported for B. pumilus (Satomi et al. 2006; Vos et al. 2009).

Table 3.   Characterization of sponge-associated sporeformers
Isolate No.*12328223145684410152774788314611125152CH8a244342499114773MMA7
  1. *Strains are grouped by species relatedness and were tested at least twice and in some instances up to five times. For all assays, growth was monitored at 24 h and 48 h.

  2. †+, indicates a positive result regardless of the different extents of proteolytic activity that were observed; (+), clearing zone on the SKM plate without visible growth.
    TetR (10), resistance to 10 μg ml−1 tetracycline; EryR (10), resistance to 10 μg ml−1 erythromycin; +/−, weak growth or isolated colonies; Ng, no growth; R, rare spores observed; V, variable results; DSM, Difco sporulation medium; MA, marine agar.

TetR (10)++
EryR (10)++
Sporulation on
Growth in SW (%)
Growth in NaCl (%)
Growth at 30°C
30°C broth

The five sporeformers identified by 16S rRNA sequence analysis as belonging to the B. cereus sensu lato group (Table 2, Fig. 2), which includes B. cereus sensu stricto, B. anthracis, B. thuringinesis, B. mycoides, B. pseudomycoides and B. weihenstephanensis, were β-haemolytic (Table 3) and grew as pink colonies with the characteristic surrounding white precipitate on B. cereus mannitol-egg yolk-polymyxin (MYP) selective agar (Merck). None of the isolates showed the typical rhizoidal/mycoidal colony morphology characteristic of B. mycoides/B. pseudomycoides on either MA or LB plates, narrowing a possible identification down to B. cereus, B. weihenstephanensis and B. thuringiensis. 16S rRNA signature sequences have previously been employed to assist in the differentiation of the mesophilic B. cereus and B. thuringiensis from the psychrotolerant B. weihenstephanensis and B. mycoides species (Lechner et al. 1998; Guinebretiere et al. 2008). In our analysis, we found the signature for mesophilic strains in isolates #11, #51 and CH8a and the signature for psychrotolerant strains in isolates #12 and #52. However, a second 16S rRNA signature used to discriminate between psychrotolerant and mesophilic strains (Pruss et al. 1999) was only found in isolates #12 and CH8a, respectively. Isolates #11, #51 and #52 contained hybrid signatures, a feature that was previously proposed to result from the coexistence of different proportions of mesophilic and psychrotolerant 16S rRNA copies within a single isolate (Pruss et al. 1999). Of the five strains, CH8a was the only strain that grew equally well at 30, 40 and 50°C (Table 3). This was also the only isolate susceptible to penicillin G and ampicillin. These observations are supported by the phylogenetic analysis (Fig. 2), where isolates #12 and CH8a are distinct and cluster separately from isolates #11, #51 and #52, which appeared to be more closely related.

While B. pumilus, B. licheniformis and B. cereus group strains grew well on LB without salt, a number of isolates, which included most of the pigmented strains, did not grow or grew very poorly without salt supplementation (Table 3). Additionally, their growth was irregular on the different percentages of NaCl tested, with some isolates displaying a very narrow NaCl concentration growth range (e.g. #24, #42, #43 and #74) (Table 3). Overall, isolates that displayed a weaker tolerance for higher levels of salt included isolate #73, the B. cereus group isolate #12 and the three Paenibacillus isolates (Table 3).


To confirm the production of spores by the sponge-associated sporeformers, strains were initially grown on DSM, but it became evident that not all isolates were able to grow on this medium routinely used to induce sporulation by nutrient deprivation (e.g. #43, Table 3, Fig. 1b). Others grew but did not show visible signs of sporulation (#78), while in other cases, sporulation appeared to be very inefficient, as spores were produced in relatively small numbers (e.g. #4, #74 and #146) (Fig. 1b, Table 3). However, when examining the cell morphology of the isolates grown on MA plates, we observed that with the exception of strain #73, all other sponge isolates were able to produce spores on this medium (Fig. 1b). Additionally, even though no attempt was made to quantify the sporulation efficiency of these strains, it was striking to see that the majority of isolates related to marine-associated species appeared to sporulate to a higher extent on MA than on DSM (Fig. 1b, Table 3). In contrast, B. cereus group isolates #12 and #52 and Paenibacillus #22 appeared to sporulate less efficiently in MA than in DSM (Fig. 2, Table 3).

Finally, no clear evidence for the presence of parasporal crystals could be detected within the sporangium of the B. cereus group isolates #11, #12, #51, #52 and CH8a. On the other hand, the presence of parasporal bodies of uncertain nature could be visualized for some of the other isolates, e.g. #23, and similar observations were also present in nonsporulating cells of other isolates, e.g. #52 (Fig. 2).

Plasmid DNA and antimicrobial resistance profiles

With the method employed in this study, plasmid DNA was detected in c. 30% of the isolates (Table 3), with isolates #11 and #24, harbouring more than one plasmid. All plasmids detected in this study were relatively small, migrating, when uncut, below the top band of the DNA marker (c. 9 kb). However, it is conceivable that larger plasmids could be present that were not detected with the current methodology. The presence of plasmid DNA was not universal within isolates phylogenetically related, again attesting for intraspecies diversity. For example, plasmid DNA was detected in only one of the five B. cereus group isolates (#11) (Table 3). With respect to resistance to antibiotics, only three of the thirty isolates were resistant to erythromycin and/or tetracycline (Table 3). Of these, isolate #24 displayed resistance to both antibiotics.

Bioactivity of sponge-associated sporeformers

To explore the industrial and biopharmaceutical potential of the sponge-associated sporeformers, all isolates were screened for the production of lactonases, proteases and antimicrobial compounds. Degenerated primers designed to amplify the lactonase gene aiiA, which has previously been found in Bacillus isolates (Lu et al. 2006), were used to screen the 30 isolates using PCR. Of these, only strains #11 and CH8a, which both belonged to the B. cereus group, gave products of the expected size (data not shown). Subsequent sequencing of these products showed 98% nucleotide identity to the aiiA genes of B. thuringiensis serovar kyushuensis (AF478052) for isolate #11 and 99% identity to B. thuringiensis clone B (AF350931) for isolate CH8a. The nucleotide sequence identity of the aiiA gene between isolates #11 and CH8a was considerably lower (87%).

A large percentage of the sporeformers demonstrated proteolytic activity on SKM plates, despite the fact that some of the isolates were unable to grow or grew poorly on this medium (Fig. 3a, Table 3). A clearing of the plate that was visible in the area where some of the isolates were spotted (e.g. #42 and #74) may be associated with the production of proteases by the isolates during growth in MB and with subsequent use of these cultures to inoculate the SKM plates. Production of proteases by these isolates was confirmed by repeating the assay on SYP-SW-SKM agar plates, which supported their growth (data not shown). Curiously, some weak activity was also detected for the Paenibacillus isolates on the salt-supplemented medium. As a whole, the B. cereus group isolates appeared to produce the most consistent and strongest activity throughout the five isolates. Strong activity was also detected with the B. pumilus-related isolates and with several of the B. hwajinpoensis/B. baekryungensis isolates (e.g. #4, #10, #83 and #146) (Fig. 3a). However, the latter group displayed the strongest activity when the assay was repeated on salt-supplemented SKM plates (data not shown). While isolate #24 was able to grow in the SYP-SW-SKM medium, it did not display proteolytic activity. Isolate #43 was still unable to grow in this medium, and no proteolytic activity from spotting of MB overnight cultures was detected.

Figure 3.

 Bioactivities of sponge-associated sporeformers. (a) Protease activity was determined on 3% skimmed milk plates. Labelling of plates is top right to left, bottom right to left and central. 1: #1, #2, #3, #4, #8; 2: #10, #11, #12, #14, #15; 3: #22, #23, #24, #27, #42; 4: #43, #49, #51, #52, #56; 5, #73, #74, #78, #83, #84; 6: # 91, #146, #147, MMA7, CH8a. (b) Representative example of the antimicrobial activity displayed by the sponge sporeformers. Bacillus subtilis strain MMA7 activity against B. cereus (1), Staphylococcus aureus (2), Listeria monocytogenes (3) and B. megaterium (4) was screened by a colony overlay assay.

The 30 isolates were then screened in an overlay assay for the production of antimicrobial compounds active against a range of Gram-negative indicators (Salmonella Typhimurium LT2, Pseudomonas aeruginiosa, Enterobacter aerogenes and E. coli) and Gram-positive indicators (Listeria monocytogenes, Staphylococcus aureus, Enterococcous faecium, B. cereus, Clostridium perfringens and Cl. sporogenes) and against the yeast C. glabrata. While none of the isolates showed detectable activity against either the Gram-negative indicator strains or C. glabrata, a significant proportion (c. 30%) was active against many of the Gram-positive indicators (Fig. 3b, Table 4). Activity was particularly evident among the B. pumilus, B. licheniformis and B. subtilis isolates and differed between isolates, suggesting some degree of strain specificity.

Table 4.   Antimicrobial activity of selected sponge-associated sporeformers against different bacterial indicators*
StrainsBacillus cereusBacillus subtilisBacillus megateriumStaphylococcus aureusEnterococcous faeciumListeria innocuaListeria monocytogenesClostridium sporogenesClostridium perfringens
  1. *Strains are grouped by species relatedness and only strains that demonstrated some level of antimicrobial activity against any of the indicator strains are represented on the table. Readings were taken after 5-h and 24-h incubations. All strains were negative against the Gram-negative indicators Pseudomonas aeruginosa, Escherichia coli, Enterobacter aerogenes, Salmonella Typhimurium and also against the yeast Candida glabrata. Most strains were tested at least three times and never less than twice. Those for which the results are indicated as V, or results are given between brackets (), were tested 3–5 times; use of brackets (), depicts the result obtained in the majority of the assays, but at least 1 assay in disagreement; +, clear halo of growth inhibition in at least one time point; multiple +, increased activity as assessed visually by the increased diameter of the inhibition zone; −, no inhibition; +/−, reduction in growth intensity but not complete inhibition; V, results varied.



The perception that Bacillus are soil organisms, with their ubiquitous presence in nature attributed to air or water dispersal of spores, is changing. Cumulative evidence suggests that particular groups of Bacillus might be found in, and/or be better adapted to, specific habitats. This includes the proposed existence of marine species that are able to grow and sporulate within the marine environment (Ettoumi et al. 2009; Alcaraz et al. 2010). In this study, we found that spore-forming bacteria appear to account for c. 1% of the total culturable population of H. simulans. Although this last value is in broad agreement with that from a previous study that targeted the total culturable fraction of the same sponge sample (Kennedy et al. 2009), the previous report failed to reveal the diversity of the sporeformers that we observed in the current study. It is possible that numerically dominant groups of bacteria may have masked the actual diversity of less abundant groups. In a separate study that focussed on the characterization of the nonculturable microflora of this same sponge sample, it is noteworthy that no endosporeformers were detected (Kennedy et al. 2008). This discrepancy in the numbers of sporeformers between culturable and culture-independent studies appears to also be found for other marine sponges (S. Jackson, unpublished data). If sporeformers were present in sponges primarily in the form of spores, which could survive the digestion by the sponge bacterium-digesting archaeocytes, their presence would be missed by culture-independent studies. Indeed, routine DNA extraction techniques would likely fail to extract the DNA stored in the spore core, surrounded by several protective layers. On the other hand, the 1% level reported here for H. simulans sporeformers may in fact be an under-representation of the actual levels present in the sponge if, for example, a proportion of these bacteria are present as vegetative cells. Thus, unlike the robust spores, these vegetative cells would have been killed by the heat and ethanol treatments used to eliminate the background of coresident nonsporulating species.

Regardless of their number, very little is known about the diversity and specific properties of marine sporeformers and, in particular, those that are associated with sponges. In this study, we have used different approaches to assess the spore-forming diversity associated with the marine sponge H. simulans. Although all isolates were unambiguously identified to the genus level, many could not be clearly associated with a specific species because of the now widespread overlap of 16S rRNA sequences between many Bacillus and also Paenibacillus species. Still, the spore-forming population of H. simulans appeared to be more heterogeneous than that described for other marine environments. Ettoumi et al. (2009) found that c. 68% of the culturable spore-forming bacilli from marine sediments and water samples belonged to B. subtilis, B. licheniformis, B. pumilus and B. cereus species, while other studies report even greater homogeneity among the bacilli isolated from the different marine sources (Ivanova et al. 1999; Miranda et al. 2008). The H. simulans sporeformers could be divided into two broad groups. One group contained more conventional ubiquitous species, such as B. licheniformis, B. cereus and B. pumilus (Ivanova et al. 1999; Barbosa et al. 2005; Gontang et al. 2007; Fakhry et al. 2008; Muscholl-Silberhorn et al. 2008; Ettoumi et al. 2009; Hong et al. 2009b; Menezes et al. 2009). The other group, which included the pigmented isolates, comprised more recently characterized species that have been associated mainly, or uniquely, with marine habitats. This included B. aquimaris, B. algicola, B. baekryungensis and B. hwajinpoensis. Isolates related to B. baekryungensis/B. hwajinpoensis (Yoon et al. 2004) were the most highly represented among H. simulans sporeformers (8 of 30 isolates). These species have been found in some marine habitats, including sponges, but they are not routinely reported, and appear to be absent from the microflora of some Bacillus-rich sponges (Yeon et al. 2005; Muscholl-Silberhorn et al. 2008; Zhu et al. 2008; Ki et al. 2009; Menezes et al. 2009). The comparatively low number of Paenibacillus isolates (3 of 30) (Ash et al. 1993) is in agreement with other studies which have shown that although Paenibacillus strains are part of the microflora of some sponges and other marine niches, their abundance and prevalence is considerably lower than that seen for Bacillus (Gontang et al. 2007; Zhu et al. 2008; Ettoumi et al. 2009; Lee et al. 2009; Menezes et al. 2009).

With a single exception, all isolates were able to grow and sporulate in the presence of salts. Consistent with their marine source, many of the isolates that demonstrated a salt requirement for their growth appeared to sporulate better on MA than on DSM plates. Some of the halophilic isolates, although able to grow well in LB supplemented with different percentages of sea water, were only able to grow optimally within a very narrow range of NaCl concentrations. This could indicate a dependence on other salts that may be present in the sea water. This dependency on salt for growth and ability to sporulate on MA implies that these isolates may be true marine bacteria rather than contaminants from terrestrial environments.

Tetracycline resistance is the most widespread form of antibiotic resistance in nature. This is attributable in part to its association with mobile elements such as plasmids and conjugative transposons, which frequently also carry erythromycin resistance determinants (Roberts 2003, 2005, 2008). Although c. 30% of the isolates contained plasmid DNA, the incidence of resistance to tetracycline and erythromycin was relatively low, with only isolate #24 showing resistance to both antibiotics. For isolate #24, tetracycline resistance has been associated with a 5-kb mobilizable plasmid (pBHS24), which has also been found among three other Gram-positive isolates from different environments (Phelan et al. 2011). Although we do not have a phenotype for the remaining plasmids, the presence of plasmid DNA was not consistent across related sponge isolates, again attesting to intrastrain diversity.

With a background setting of increased antibiotic resistance, quorum sensing, which is involved in the regulation of different bacterial virulence factors, has emerged as a preferred target for drug discovery in infectious disease (Bjarnsholt et al. 2010; Thoendel and Horswill 2010). In Gram-negative bacteria, interference with quorum-sensing regulatory systems can be achieved through the production of enzymes that degrade the small signalling molecules, N-acyl homoserine lactones (AHLs) (Lu et al. 2006; Czajkowski and Jafra 2009). Two of the H. simulans isolates proved to have the gene required for the production of the AHL-inactivating enzyme lactonase AiiA.

Bacterial proteases are widely employed in a variety of industrial fields, from biotechnology to food processing (Kasana 2010). Although the proteolytic activity of many Bacillus strains is well characterized (Nijland and Kuipers 2008; Lebrun et al. 2009), the activities of more recently described species, including marine isolates, are less well understood. It is noteworthy that most of the H. simulans strains identified in this study, including many of the halophilic isolates, demonstrated proteolytic activity, which could be associated with novel proteases of commercial value.

Bacillus isolates are also well known for the production of a vast array of structurally unrelated antimicrobial compounds, which include polyketides, nonribosomally synthesized peptides and bacteriocins (Stein 2005; Abriouel et al. 2010). Again, a large number of our H. simulans sporeformers demonstrated antimicrobial activity, including activity against recognized food pathogens. Despite the observation that the main producers were related to B. pumilus, B. licheniformis and B. subtilis species, there was, nevertheless, variation in the spectrum and strength of the activities observed among the different strains from each group. Any of the above-mentioned bioactivities would be of relevance in the selection for potential probiotic candidates among the sponge-associated sporeformers.

Despite these properties, the physiological role of sporeformers within marine sponges remains to be established. It is also not clear whether they are present primarily as spores, vegetative cells or both, or yet whether they are capable of having a full life cycle during which the spores will germinate, proliferate and resporulate. Presumably, spores could survive digestion by the sponge archaeocytes, allowing for subsequent germination in nutritionally more favourable parts of the mesohyl (Taylor et al. 2007). Arguably, by using enrichment procedures for sporeformers, such as the heat and ethanol treatments described here, metabolically active vegetative cells would be eliminated. On the other hand, nontargeted culturable approaches, besides masking the diversity of less abundant groups, would be ineffective in establishing the specific contribution of vegetative cells vs spores for the total spore-forming population of the sponge. Further studies involving the use of fluorescence in situ hybridization will be required to provide further clarity in this regard. Notwithstanding this, it has been shown that spores of marine Bacillus can be enzymatically active and may play a role in bacterial oxidation of Mn(II), an important nutrient in sea water (Francis and Tebo 2002; Francis et al. 2002; Dick et al. 2008). A better understanding of the biochemical and genetics properties of the sponge-associated, endospore-forming population will help to establish their role in this particular niche and facilitate their exploitation for medical and industrial applications in a rational manner.

To conclude, among the microflora associated with the marine sponge H. simulans, we have found a large diversity of endospore-forming bacterial species and a wide intraspecies variety that was not revealed by nontarget culturing or culture-independent approaches. Many of the sponge isolates display potentially interesting biotechnological, biopharmaceutical and probiotic properties.


This work was supported in part by grants awarded by the Beaufort Marine Research Award under the Sea Change Strategy and the Strategy for Science Technology and Innovation (2006–2013), with the support of the Marine Institute, funded under the Marine Research Sub-Programme of the National Development Plan 2007–2013, the Department of Agriculture and Food (FIRM08/RDC/629; DAF RSF 06 321; DAF RSF 06 377), the European Commission (MTKD-CT-2006-042062; O36314), the Science Foundation of Ireland (SFI 07/IN.1/B948; 08/RFP/GEN1295; SFI/RFP/BMT2350), the Irish Research Council for Science, Engineering and Technology (IRCSET) (05/EDIV/FP107), the Health Research Board (RP/2006/271; RP/2007/290; HRA/2009/146) and the Environmental Protection Agency (EPA 2006-PhD-S-21; 2008-PhD-S-2). We thank Dr. Grace McCormack from the National University of Ireland, Galway, for the H. simulans sponge samples. We also thank Martina Blake and Natasha O’Sullivan, UCC Microbiology students, for their help in the isolation and preliminary phylogenetic analysis of the sporeformers.