Integration of Culture-Based and Molecular Analysis of a Complex Sponge-Associated Bacterial Community

The bacterial communities of sponges have been studied using molecular techniques as well as culture-based techniques, but the communities described by these two methods are remarkably distinct. Culture-based methods describe communities dominated by Proteobacteria, and Actinomycetes while molecular methods describe communities dominated by predominantly uncultivated groups such as the Chloroflexi, Acidobacteria, and Acidimicrobidae. In this study, we used a wide range of culture media to increase the diversity of cultivable bacteria from the closely related giant barrel sponges, Xestospongia muta collected from the Florida Keys, Atlantic Ocean and Xestospongia testudinaria, collected from Indonesia, Pacific Ocean. Over 400 pure cultures were isolated and identified from X. muta and X. testudinaria and over 90 bacterial species were represented. Over 16,000 pyrosequences were analyzed and assigned to 976 OTUs. We employed both cultured-based methods and pyrosequencing to look for patterns of overlap between the culturable and molecular communities. Only one OTU was found in both the molecular and culturable communities, revealing limitations inherent in both approaches.


Introduction
Marine sponges (Phylum Porifera) and their microbial symbionts have garnered great interest in recent decades.Two reasons for this interest are 1) the study of sponge-microbe interactions as a model of complex symbioses and 2) pharmaceutical leads from natural products derived from sponges [1].Some of these natural products may be produced, not by the sponge hosts, but by microbial symbionts of the sponges [2,3].While advances in sequencing technology have led to a deeper understanding of sponge microbial communities, the functional roles of these microbial populations has become the primary focus.Several studies have contributed to sponge symbiont function including nitrogen cycling [1,4,5,6], carbon fixation, and vitamin B12 metabolism [7,8].All of these studies suggest bacteria living in association with sponges can provide specific nutrients.However, there are many unanswered questions surrounding the methods by which sponges acquire and maintain their symbionts, the effects of microbes on sponge health, and the specificity of the microbial symbionts to their sponge hosts.A complete understanding of the ecology and genetic diversity of these microbes and their roles in host health, and the production of natural products requires laboratory cultivation.It is now very apparent that many of the major groups of bacteria that are associated with sponges are not being grown in the laboratory.This finding is at once interesting and challenging.
Cultivation provides access to genetic and biochemical characteristics of each microorganism that may not be revealed by 16S rRNA gene metagenomics.Several bacterial strains were isolated from sponge Halichondria sp.collected from the Indo Pacific region with a wide range of antimicrobial activity revealing sponge isolates remain a rich source of novel secondary metabolites [9].Esteves et al. detected for the first time polyketide synthase (PKS) genes in cultured bacteria from the genus Aquimarina (Bacteroidetes) isolated from Irciniidae sponges in the Northeast Atlantic [10].Aside from the production of bioactive compounds, cultured sponge-associated isolates also reveal ecological relevance.The production of the quorum sensing signal molecule N-acyl homoserine lactones (AHLs) that are involved in cell-cell communication and bacterial colonization of higher organisms has been detected in the cultured community of sponge Mycale laxissima [11,12].A further study examined a specific symbiont of M. laxissima, Ruegeria sp.KLH11 and confirmed a complex quorum-sensing network activates flagellar motility and inhibits biofilm formation [13].A culturable a-proteobacterial symbiont was obtained from several sponges from different locations [14], also suggesting the ecologically importance of this cultured bacterium.
Giant barrel sponges of the genus Xestospongia are prolific members of tropical reef environments.Bacterial 16S rRNA gene libraries from the sponges Xestospongia muta and Xestospongia testudinaria show that these closely related sponges are hosts to diverse bacterial communities that are maintained in both sponges from geographically distant locations and are dominated by Chloroflexi, Acidobacteria and Actinobacteria [15,16].In addition, 16S rRNA gene sequences from these sponges are more similar to previously published sponge-derived sequences than the 16S rRNA genes from the surrounding water [16].These findings suggest a stable relationship between the sponges and their bacteria.Culturing efforts may provide additional insight into the bacteria associated with Xestospongia spp.In this study, we have employed a wide range of culture media to increase the diversity of culturable bacteria from X. muta and X. testudinaria.Also, pyrosequencing was used to obtain insights into these bacterial communities to determine whether many of the cultured isolates from these sponges are present at low numbers in the complex bacterial communities associated with these sponges.We found that only one OTU, classified as a Gammaproteobacteria, was present in the cultured and non-cultured bacterial community of Xestospongia spp.

Bacterial cultivation
Sponge samples were processed for bacterial cultivation within 2 hours of collection.A 1-cm 3 section of sponge tissue was pulverized in 9 ml of sterile ASW.The homogenate was used to create a 10-fold dilution series of which 100 ml aliquots were added to solid or liquid media.Dilutions were plated on Marine Agar 2216 (MA2216) (Becton Dickinson) and incubated for 1-2 weeks for plate counts and dominant morphotypes.Initial plate counts and subcultivation of dominant morphotypes were done in field labs in Manado and Key Largo, 3-5 days after sample collection.Pure isolates of dominant morphotypes were transported back to the lab for further identification.Isolation plates were then resealed and incubated to allow for growth of additional slow growing bacteria.Dilutions were plated on the following additional media to increase the diversity of cultivable bacteria: 1/10 strength MA2216 with and without antimicrobials, Actinomycete Isolation Agar (Becton Dickenson) with antimicrobials, FSWFA medium (filter-sterilized seawater supplemented with a methanol extraction of lipids from lyophilized X. muta tissue, solidified with agarose), modified Glycerol-Asparagine Agar (peptone, 2 g/L; L-Asparagine, 0.5 g/L; Na propionate, 4 g/L; K 2 HPO 4 , 0.5 g/L; MgSO 4 67H 2 O, 0.1 g/L; FeSO 4 , 0.001 g/L; Glycerol, 5 g/L; Agar, 18 g/L) with antimicrobials, ISP medium 2 (Becton Dickenson) with and without antimicrobials, NTM [17,18] with vitamins, with and without antimicrobials, R2A medium [19] with antimicrobials, R2A with vitamins (R2AV) with and without antimicrobials, SN medium, Starch Casein Agar [20] with antimicrobials, VL55 medium [21] with and without antimicrobials, and whole seawater medium (WSWA) (filtersterilized seawater with vitamins, solidified with agarose).All media without salt in their original formulations were supplemented with NaCl at 20 g/L.All media containing vitamins were supplemented with 1 ml/L Vitamin Solution 1 [22] and 3 ml/L Vitamin Solution 2 [22].All media containing antimicrobials were supplemented with 10 mg/ml cycloheximide, 10 mg/ml nalidixic acid, and 25 mg/ml nystatin.These plates were incubated for six to nine weeks in dark conditions at 25uC and all distinct colony morphotypes were subcultured for further identification.

Identification and classification of bacterial isolates
DNA was extracted from pure cultures using the Mo Bio UltraClean Microbial DNA Isolation Kit (Mo Bio Laboratories Inc, Carlsbad, CA) according to the manufacturer's instructions.16S rRNA gene fragments were PCR-amplified with the universal primers 27F and 1492R and sequenced on an ABI 3130 XL Genetic Analyzer (Applied Biosystems, Foster City, CA).Sequences were assembled and edited using Pregap4 and Gap4 from the Staden Package (http://staden.sourceforge.net/).Sequences were classified with the RDP classifier.The nearest relatives for each sequence were obtained from the GenBank database using the blastn tool (http://blast.ncbi.nlm.nih.gov/) in May 2011.

Pyrosequencing of barcoded 16S rRNA gene amplicons
Total DNA was extracted from lyophilized sponge tissue as described by Montalvo et al. [15].Universal primers 27F and 338R were used for PCR amplification of the V1-V2 hypervariable regions of 16S rRNA genes as described by Ravel et al. [23].Briefly, the 338R primer included a unique sequence tag to barcode each sample.The V1-V2 regions of 16S rRNA genes were amplified using AmpliTaq Gold DNA polymerase (Applied Biosystems) and 50 ng of template DNA in a total reaction volume of 50 mL.Reactions were run in a PTC-100 thermal controller (MJ Research).Negative controls without a template were included for each bar-coded primer pair.The presence of amplicons was confirmed by gel elec-trophoresis on a 2% agarose gel and staining with SYBRGreen.PCR products were quantified using a GelDoc quantification system (BioRad) and the Quant-iT PicoGreen dsDNA assay.Equimolar amounts (100 ng) of the PCR amplicons were mixed in a single tube.Amplification primers and reaction buffer were removed from each sample using the AMPure Kit (Agencourt).The purified amplicon mixtures were sequenced by 454 FLX pyrosequencing using 454 Life Sciences primer A by the Genomics Resource Center at the Institute for Genome Sciences, University of Maryland School of Medicine, using protocols recommended by the manufacturer as amended by the Center.Sequences were binned by sample as previously described [23].The following criteria were used to assess the quality of the reads: sequence reads must include a perfect match to the unique sequence tag, be at least 200 bp in length, have no undetermined bases, and be at least a 60% match to a previously described 16S rRNA sequence.

Estimation of microbial diversity and analysis of 454 sequences
Mothur (http://www.mothur.org/)[24] was used to assign sequences to operational taxonomic units (OTUs) and to generate rarefaction and rank abundance curves for observed OTUs.Sequences were clustered using the average neighbor algorithm at a distance of 0.03.The alignment and phylogenetic analysis was performed using ARB software package [25].Sequences were classified using the SILVA rRNA database project [26].

16S rRNA gene sequence accession numbers
The bacterial 16S rRNA gene sequences obtained in this study have been deposited on the public MG-RAST server (http:// metagenomics.anl.gov)under MG-RAST accession numbers 4551688.3 to 4551692.3.All sequences can be downloaded by accessing the Xestospongia spp.Associated Bacteria Study using the MG-RAST Project link (http://metagenomics.anl.gov/linkin.cgi?project = 7756)

Diversity of cultured isolates
A total of 434 cultured bacterial isolates were identified from X. muta and X. testudinaria.From X. testudinaria samples, 50 isolates from three isolation media were identified from 5 sponges collected in 2005.From X. muta samples, 384 isolates from 18 isolation media were identified from 14 sponges collected in 2004, 2005, and 2008.A total of 50 genera were represented; 18 genera from the Actinobacteria, 7 genera from the Firmicutes, 12 genera of Gammaproteobacteria, 11 genera of Alphaproteobacteria, and 2 genera of Bacteriodetes (Table 1, Table S1 and Table S2).
Seventeen genera of Actinobacteria were isolated from X. muta.To our knowledge, this is the most genera of Actinobacteria reported from a single sponge species, as well as the first report of Nesterenkonia sp.isolated from a marine sponge.Actinobacteria were isolated on 15 of the 18 media used in this study, with the most diversity coming from R2A supplemented with vitamins and antimicrobials (Table S1).The cultured Actinobacteria were dominated by Micromonospora sp., which are commonly found in sponges and have been shown to produce natural products [27,28].Micrococcus, Brevibacterium, Kocuria, and Streptomyces followed in abundance in the assemblage of cultured actinomycetes, all of which are commonly found in sponges.Microbacterium spp., which have been found previously in sponges [29], were the dominant Actinobacteria cultured from X. testudinaria, but they were not isolated from X. muta.
The Gammaproteobacteria were dominated by Vibrio spp.and Pseudoalteromonas spp.The Firmicutes were dominated by Bacillus spp.The cultured Alphaproteobacteria were dominated by isolates by a single Pseudovibrio sp. that are 99% identical to the NW001-like Alphaproteobacteria first discovered by Webster and Hill [20], and shown to dominate the culturable community of the marine sponge Rhopaloeides odorabile.These bacteria were further characterized by Enticknap et al. [14] and were found to be present in sponges from several oceans, as well as in the larvae of Mycale laxissima.The presence of these NW001-like Alphaproteobacteria in both X. muta and X. testudinaria sponges, further strengthens indications that these bacteria are sponge-specific symbionts.
All Bacteroidetes belong to the class Flavobacteria.Two novel Flavobacteria were isolated from X. muta.Isolates SN8107F and SN8106F were 95% identical to their closest relatives Coccinimonas marina and Muricauda sp.BB-My12.
Only six of the 20 classified bacterial groups represented in the 454 sequence data, did not include OTUs shared by all four sponges (Bacteroidetes, Firmicutes, Planctomycetes, Betaproteobacteria, TM6, and Verrucomicrobia) and these groups were only present in very low numbers, with each group representing less than 1% of the total sequence data (Fig. 2).

Overlap of 454 sequences and cultured isolates
Gammaproteobacteria were the only group of cultured isolates also represented in the 454 sequences.Four Gammaproteobacteria of the class Oceanospirillales were isolated from two X. muta sponges and two X. testudinaria sponges on MA2216 and also seen in the 454 sequences of XtC and XmF.These isolates were 91-93% identical to their closest relatives in GenBank, Spongiobacter nickelotolerans and strain MOLA 531, and 97-99% identical to each other.These sequences formed a single OTU and may represent a novel group of culturable sponge symbionts (Fig. 4).Spongiobacter sp. have been found in association with sponges and corals [30,31,32].One study examined coral-associated bacteria in their role in biogeochemical cycling of sulfur and found that the genus Spongiobacter sp. was one of the four genera isolated on media with either dimethlysulfoniopropionate (DMSP) or dimethyl sulfide (DMS) as the sole carbon source from two coral species suggesting that these bacteria have the potential to metabolize sulfur compounds and may contribute to coral health [33].Consistent with other studies, this cultured strain could potentially provide nutrients to the sponges.Although there was only one shared OTU between the cultured and pyrosequenced community, Spongiobacter sp. is a well documented isolate of sponges [1] and strain MOLA is also isolated from a sponge, which again supports sponge-specific symbionts.Further study of these bacteria is warranted, as well as looking for these organisms in other marine sponges.

Conclusion
It is not surprising that the deep pyrosequencing analysis of the sponges X. muta and X. testudinaria revealed many major groups of bacteria that do not yet have any cultured representatives.Bacterial symbionts can have ancient relationships with their hosts and live within specialized structures or highly enriched host tissue, making them difficult to culture [34].Sponges represent the oldest metazoan phylum [35] and the continuance of bacteria Table 1.Cultured isolates from X. muta and X. testudinaria.The bacterial phyla or classes are bolded and the genera are listed below each phylum or class.The number of isolates cultured and the number of OTUs from X. muta and X. testudinaria are listed for each bacterial group.doi:10.1371/journal.pone.0090517.t001associated with sponges can suggest a long and stable relationship.Several studies have highlighted the large taxonomic differences between the culture dependent and culture independent communities [34,36,37,38].However, several groups have increased culturing efforts of microbes associated with sponges.Wilkinson et al. pioneered this effort and found a similar sponge-specific bacterial isolate that was found in 9 out of 10 sponges samples from both the Mediterranean and Great Barrier Reef [39].Several studies have noted cultured sponge-specific isolates from many marine sponges in difference locations [40,41].A recent survey of the largest cultured collection of bacteria associated with marine sponges uncovered sponge specific isolates that were previously reported as ''uncultured'' however; nearly 88% of the isolates were dominated by the commonly cultivable class Gammaproteobacteria [42].Gammaproteobacteria were also one of the dominant cultured members in the present study.Our previous work in molecular analysis of bacteria associated with Xestospongia sponges [15,16] already partially revealed the remarkable bacterial diversity associated with these sponges, specifically the abundance of Chloroflexi and Actinobacteria.The pyrosequencing reported here was consistent with our previous findings with respect to the abundance bacterial groups and extended this work with an additional 27 OTUs that were present in all four sponge individuals, but were not previously detected in the 16S rRNA clone library analysis [16], including sequences classified as candidate division OD1 (Table S3).The presence of OTUs found in both Xestospongia individuals and different species from geographically distant location, may not just suggest a possible sponge-specific community but a sponge-species-specific community.Lee et al. observed several Red Sea sponges possessed a highly sponge-specific or even sponge-species-specific microbial community that are resistant to environmental disturbance [43].Another study used pyrosequencing on 32 sponges in 8 different locations and found a minimal core bacterial community that existed in all sponges but a large species-specific bacterial community associated with the sponges [44].The bacterial community associated with A. corrugata was assessed between two seasons and although there were slight shifts in the community the overall species-specific bacteria remained [45].The current study contributes to the increasing knowledge of sponge bacterial symbionts and the growing notion of a sponge-species-specific bacterial community.There is still much work needed to understand the roles of these bacterial symbionts and recent work using metaproteogenomics provided new insights into the interactions of sponge-associated microbial communities [46].
It is more surprising that the extensive culture-based work undertaken here resulted in the isolation of many groups of bacteria that are not detected by even the deep pyrosequencing community analysis.There are two possible explanations for this.First, the cultured bacteria may be present as only very minor constituents of the sponge-associated bacterial community.Many of these bacterial groups were isolated repeatedly from different individuals of the same sponge species, and often from both Xestospongia spp.sponges.This suggests that they are consistently associated with the sponges and may play an important role in the bacterial community, even if they are present at very low numbers.A recent study examined the cultured and uncultured bacteria associated with soil samples and observed soil bacteria captured by culturing were in low abundance or absent in the culturedependent community, revealing a fraction of a ''rare biosphere'' that would remain undiscovered by pyrosequencing at a given  sampling depth [47].The second possibility is that the 16S rRNA community analysis [16] and the pyrosequencing analysis may have significant biases resulting in the lack of detection of entire bacterial groups.This concern has been previously discussed [48,49,50,51].Approaches to investigate this possibility include FISH targeting of bacteria found in culture based work (e.g. the Vibrio, Pseudovibrio, Micrococcus, and Micromonospora spp., that were often cultured from X. muta and X. testudinaria), and PCR detection with specific primers designed to target these groups, followed by RT-PCR or FISH to assess their numbers.Cassler et al. used RT-PCR to quantify the cultured and uncultured bacteria associated with a marine sponge and found three uncultured members in large quantities and only detected one of three cultured members at a trace level [52].The present study highlights cultured members isolated from different individuals of the same sponge species as well as different species within the genus Xestospongia.Therefore using RT-PCR for cultured groups associated with Xestospongia spp. that have shown a consistent associated may reveal higher quantities of the bacteria.This work and other recent studies [53,54] emphasize the importance of extensive culture-based studies for isolation of representatives of the novel bacterial groups found in marine sponges.In this study a very diverse group of bacteria was cultured from Xestospongia sponges, but very little overlap was found between the culturable community and the community revealed by molecular approaches.Several novel isolates were cultured, however, the majority of the bacteria that were cultured are closely related to organisms that have been found in other environments, while the majority of the sequences in the molecular analysis have been found only in sponges, or sponges and corals.As previously noted, culturing is adding new insight about sponge bacteria symbionts.However, there is still much effort needed to culture representatives of many of the key groups of bacteria found in marine sponges and one of the challenges for the field of sponge microbiology, as highlighted by Taylor et al. [54], is to successfully culture representatives of these groups. Photobacterium

Figure 4 .
Figure 4. Phylogenetic tree of shared OTUs in the culturebased and culture-independent analysis of Xestospongia spp.Sequences from this study are bolded and their closest relatives are included.Bootstrap values (neighboring-joining method, 100 replicates) are indicated by closed circles (values .90%)and open circles (values .75%).The scale bar represents 10% sequence divergence.doi:10.1371/journal.pone.0090517.g004