Vertical Transmission of Diverse Microbes in the Tropical Sponge … · Here we report that a diverse group of microbes, including both bacteria and ... Samples were preserved in

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 2007, p. 622–629 Vol. 73, No. 20099-2240/07/$08.00�0 doi:10.1128/AEM.01493-06Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Vertical Transmission of Diverse Microbes in theTropical Sponge Corticium sp.�

Koty H. Sharp,1† Boreth Eam,1 D. John Faulkner,2‡ and Margo G. Haygood1,2*Marine Biology Research Division, Scripps Institution of Oceanography, University of California, San Diego, La Jolla,

California 92093,1 and Center for Marine Biotechnology and Biomedicine, Scripps Institution of Oceanography,University of California, San Diego, La Jolla, California 920932

Received 28 June 2006/Accepted 4 November 2006

Sponges are host to extremely diverse bacterial communities, some of which appear to be spatiotemporallystable, though how these consistent associations are assembled and maintained from one sponge generation tothe next is not well understood. Here we report that a diverse group of microbes, including both bacteria andarchaea, is consistently present in aggregates within embryos of the tropical sponge Corticium sp. The majortaxonomic groups represented in bacterial 16S rRNA sequences amplified from the embryos are similar tothose previously described in a variety of marine sponges. Three selected bacterial taxa, representing pro-teobacteria, actinobacteria, and a clade including recently described sponge-associated bacteria, were testedand found to be present in all adult samples tested over a 3-year period and in the embryos throughoutdevelopment. Specific probes were used in fluorescence in situ hybridization to localize cells of the three typesin the embryos and mesohyl. This study confirms the vertical transmission of multiple, phylogenetically diversemicroorganisms in a marine sponge, and our findings lay the foundation for future work on exploring verticaltransmission of specific, yet diverse, microbial assemblages in marine sponges.

Many marine invertebrates engage in long-term, specific asso-ciations with microorganisms. Sometimes these symbiotic associ-ations are highly specialized, like the association between thesquid Euprymna scolopes and a single species of light-producingbacterium (Vibrio fischeri), where the host utilizes strain-specificmechanisms to acquire a single symbiont species from the sur-rounding environment and the symbiont induces developmentalchanges in the host (25, 29). In contrast, the gutless marine oli-gochaete genus Olavius hosts multiple bacterial symbionts (5, 11).Not surprisingly, in marine sponges, the choanocytes, which filterseawater, tend to accumulate diverse microbial assemblages.More remarkably, microbes are often observed in the interiortissues (mesohyl) of sponges by microscopy. Indeed, complexcommunities found in some sponges lead one to view these asmacroscopic microbial consortia organized by a scaffold of spongecells. The microbial community within some sponges is domi-nated by a single bacterial or archaeal species. Recent biochem-ical and genomic research on specific sponge-associated pro-karyotes has shown that, in several sponges, bacteria and archaeaare involved in production of bioactive compounds, autotrophiccarbon fixation, or translocation of nutrients and antioxidants tothe sponge host (21, 32, 33, 40, 44). Genomic data from Cenar-chaeum symbiosum, a crenarchaeon found in the sponge Axinellamexicana (32), suggest that the archaeal symbiont can live chemo-lithoautotrophically by ammonia oxidation (17). Other sponges

appear to have very few interior microbes at all (19, 43). Themetabolic diversity of the various eubacteria and archaea found inmarine sponges is likely to contribute significantly to nutrientcycling within sponges and their survival in the ecosystems theyinhabit.

In addition to understanding the functional roles of bacteriaand archaea in sponges, a central objective of sponge microbiol-ogy is to gain a better understanding of the diversity and predict-ability of sponge-prokaryote associations. Small-subunit rRNA-based molecular approaches (20, 22) indicate that severalunrelated groups of microbes are consistently found in diversesponges. This pattern suggests reliable mechanisms of transmis-sion or recruitment of multiple microbes, but whether the mi-crobes are selected by sponges from environmental populationsor transmitted directly between sponge generations is unknown.

Vertical transmission of microbial symbionts, characteristic oflong-term obligate associations, is documented in many animalphyla, including bivalves (8, 10, 16, 26, 36), bryozoans (18), andascidians (23). Ultrastructural studies showing bacteria in spongereproductive tissues provide strong evidence that vertical trans-mission of bacteria is a common phenomenon in sponges (34, 35,42). Recently, Enticknap et al. (12) used fluorescence in situhybridization (FISH) to localize an alphaproteobacterium withinthe developing embryos of the sponge Mycale laxissima. The spe-cies found in M. laxissima embryos belongs to a larger group ofbacteria found in several other marine sponges (12, 47). Micros-copy revealed the presence of cyanobacterial symbionts in theeggs and sperm of Chondrilla australiensis (41, 42). Oren et al.used fluorescence microscopy and electron microscopy to dem-onstrate the presence of cyanobacteria in larvae of the Red Seasponge Diacarnus erythraenus. 16S rRNA gene sequence analysisfrom D. erythraenus larval DNA extracts showed the presence ofunicellular cyanobacteria closely related to the known symbioticcyanobacteria in Aplysina aerophoba and Chondrilla nucula (30).

* Corresponding author. Present address: Department of Environ-mental and Biomolecular Systems, OGI School of Science & Engi-neering, Mail Code OGI 100, Oregon Health & Science University,20000 NW Walker Road, Beaverton, OR 97006-8921. Phone: (503)748-1993. Fax: (503) 748-1464. E-mail: [email protected].

† Present address: Smithsonian Marine Station, 701 Seaway Drive,Fort Pierce, FL 34949.

‡ Deceased.� Published ahead of print on 22 November 2006.

622

on Decem

ber 19, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

However, in most sponges, microbes in or on the oocytes, sperm,embryos, and larvae have yet to be investigated via rRNA genesequence analysis. To date, in situ hybridization has not been usedto confirm transmission of a diverse set of bacteria in spongeembryos.

In this study, the vertical transmission of microbial assem-blages was investigated in the tropical Pacific sponge Corticiumsp., which broods its embryos and releases fully developedlarvae into the water column. Previous ultrastructural investi-gations demonstrate that microorganisms are present in thecentral cavity of Corticium candelabrum larvae (6), suggestingthat the sponge maintains microbial associates intergenera-tionally, but the composition and diversity of the transmittedassemblage have yet to be characterized. The aims of this studywere to (i) localize bacteria and archaea in various embryonicstages via FISH and (ii) identify members of the bacterialcommunity in the sponge embryos. FISH revealed that bothbacteria and archaea are present throughout Corticium embryo-genesis, and bacteria are more abundant than archaea in theembryos, as well as in the sponge mesohyl. The bacterial com-munity composition was studied in more detail by constructionof small-subunit rRNA (16S rRNA) clone libraries from dis-sected embryos within adult Corticium sp. Specific FISHprobes were designed to confirm the presence of select se-quences from the embryo clone libraries, and the specificity ofeach probe was tested with rigorous negative-control probes.

MATERIALS AND METHODS

Sponge collection. Sponges of the genus Corticium are widespread throughoutthe tropical Pacific. In Palau, Corticium sp. individuals ranging in size from 0.5cm to 3 cm in diameter occur on and beneath overhangs on reef slopes. TheCorticium sp. collected in Palau occurs in clusters on reef substrate. Figure 1shows an underwater photo of the sponge; it has a black, smooth outer layer withapparent oscules. Slicing the sponge reveals a light gray interior beneath theblack cortex. For this study, large individuals (approximately 2.5 to 3 cm indiameter) were collected by scuba divers from 10 different reef slopes in thePalau archipelago at depths ranging from 3 to 20 m in September 2001, Sep-tember 2002, and March 2004. Sponges were collected in plastic bags at depthand brought to the surface. Samples were preserved in 100% ethanol and 2.5%

glutaraldehyde or frozen solid at �20°C until use. For FISH, whole sponges werefixed in paraformaldehyde (4% in buffer: 20 mM K2HPO4, 0.5 M NaCl, pH 7.4)for 2 h at room temperature and transferred to 70% ethanol for long-termstorage at �20°C. Voucher specimens from each location were frozen in ethanolfor potential future DNA analysis.

FISH. Samples were fixed and stored on the day of collection as describedabove. Individual sponges embedded in paraffin wax were sectioned to 10 �m.Sections were deparaffinized (twice for 5 min in xylene, twice for 5 min in 100%ethanol, and one rinse in Milli-Q water) and air dried. FISH was performed ina humidity chamber in hybridization buffer (0.9 M NaCl, 20 mM Tris-HCl [pH7.4], 0.01% sodium dodecyl sulfate) with 35% percent formamide for 2 h. Allprobes, including the general eubacterial and archaeal probes and the sequence-specific probes designed in this study, were used at a final concentration of 5ng/�l in hybridization buffer. Sequences of all probes used in this study are listedin Table 1. After hybridization, the slides were incubated at 48°C in wash buffer(0.7 M NaCl, 20 mM Tris-HCl [pH 7.4], 50 mM EDTA, 0.01% sodium dodecylsulfate) for 20 min. The wash buffer was rinsed off with Milli-Q water, and slideswere air dried and mounted in VectaShield (Vector Labs, Burlingame, CA).Slides were visualized on an Axioskop epifluorescence microscope (Zeiss) with a40� objective nonimmersion lens.

Bacterial 16S rRNA gene library construction. For embryo DNA extractions,approximately 100 embryos were picked intact from ethanol-preserved spongeswith a sterile 23-gauge syringe needle and rinsed twice in sterile artificial sea-water (ASW) (100 mM MgSO4 · 7H2O, 80 mM CaCl2 · 2H2O, 2.4 M NaCl, 80mM KCl). Genomic DNA was extracted from the picked embryos using aprotocol adapted from Preston et al. (32). ASW was removed and replaced by 1mg/ml lysozyme–TE (10 mM Tris-HCl, 1 mM EDTA, pH 8.0), and the sampleswere incubated at 37°C for 30 min. Proteinase K was added to final concentrationof 0.5 mg/ml, and the sample was incubated at 55°C for 1.5 h, until the solutionwas transparent. To complete lysis, the sample was boiled for 60 s. After lysis, theDNEasy genomic extraction kit (QIAGEN) bacterial DNA protocol was used.

PCR with general eubacterial primers 27f (5�-AGAGTTTGATCMTGGCTCAG-3�) and 1492r (5�-TACGGYTACCTTGTTACGACTT-3�) was done withthe following profile conditions: initial denaturation for 3 min at 95°C; 35 cyclesof denaturation for 30 s at 95°C, annealing for 1 min at 50°C, and elongation for1 min at 72°C; and a final extension step for 7 min at 72°C. Product was analyzedby electrophoresis on a 0.8% agarose gel and purified with a Rapid PCR puri-fication system (Marligen Biosciences). The purified PCR fragment was clonedinto a PCR 2.1 vector (Invitrogen), which was transformed into TOP10 cells(Invitrogen). Transformants were selected using Luria-Bertani plates (10 g/litertryptone, 5 g/liter yeast extract, 10 g/liter NaCl, 15 g/liter agar) containing 5�g/liter kanamycin sulfate, top spread with 50 ng/ml X-Gal (5-bromo-4-chloro-3-indolyl-�-D-galactopyranoside). Inserts were amplified from white coloniespicked from the selective plates in 96-well format PCR, with plasmid-specificprimers (M13f, 5�-GTAAAACGACGGCCAG-3�; M13r, 5�-CAGGAAACAGCTATGAC-3�; Invitrogen). From the three embryo clone libraries, PCR productsof the correct sizes were obtained from a total of 200 colonies. The PCR productswere screened in a restriction digest with the enzymes HhaI and HaeIII, yieldingapproximately 90 total unique restriction patterns. Initial sequencing of theinserts suggested that the 90 unique patterns could be grouped into 19 closelyrelated sequence types. Representatives of each sequence group were selectedfor full twofold sequence coverage, and the resulting sequence contigs wereconstructed and aligned for each clone in Sequencher 4.2 (GeneCodes Corp.,Ann Arbor, MI) and compared to databases at Ribosomal Database Project andNCBI (http://rdp.cme.msu.edu/index.jsp; http://www.ncbi.nlm.nih.gov/BLAST/).

FIG. 1. Corticium sp. underwater. The photo was taken at approx-imately a 30-ft depth, Koror-Babeldaob channel, Palau, Micronesia.Bar � 4 cm.

TABLE 1. List of oligonucleotide probes and primersused in this study

Probe or primer Sequence (5�–3�) Reference(s) orsource

27f AGAGTTTGATCMTGGCTCAG 27EUB338 GCTGCCTCCCGTAGGAGT 2, 28EUBNON ACTCCTACGGGAGGCAGC 28, 45ARCH915 GTGCTCCCCCGCCAATTCCT 37ARCH915NON GTGCTACCCCGCCAATTCCT This study�-CC01 CGACCTCGCGATCTCGCT This studyactino-CC07 CGCTTGACCTCGCGGTGT This studySpC1 CTACACATTCCACCGCTA This study�-CC01NON CGACTTCGCGATCTCGCT This studyactino-CC07NON CGCTTGACCTCGCAGTGT This studySpC1NON CTACTCATTCCACCGCTA This study

VOL. 73, 2007 VERTICAL MICROBE TRANSMISSION IN A TROPICAL SPONGE 623

on Decem

ber 19, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

Phylogenetic analysis. Sequences from each individual clone were edited andassembled in Sequencher 4.2 (GeneCodes Corp., Ann Arbor, MI). The 16SrRNA gene sequences obtained from the clones were run through chimera checkanalysis in the Ribosomal Database Project (9) to confirm that they are nothybrid sequences. The sequences that did not appear to be chimeras were thencompared with other described bacterial sequences through BLAST (1) and theRibosomal Database Project (9). Sequences that matched most closely were usedin an alignment with the 16S rRNA sequences. Sequences were combined withalignments downloaded from Ribosomal Database Project (9) using Sequencher4.2 (Gene Codes Corp., Ann Arbor, MI) and aligned by eye with secondarystructure information (7), yielding 1,300 bp of aligned sequence. Phylogenetictrees were constructed in PAUP� 4.0b10 (38) using a maximum-parsimony (MP)algorithm. Transversions were weighted three times more than transitions (basedon maximum likelihood estimations of the transition-to-transversion ratio), anda heuristic search of 100 repetitions with random addition of sequences wasperformed. MP bootstrapping was performed with 1,000 replicates.

Design and application of species-specific primers and probes. Probes weredesigned targeting three groups of bacterial 16S rRNA gene sequences from thelibrary for use as confirmation of their presence in the sponge and in order tosurvey additional Corticium individuals. Three sequences, representing an alpha-proteobacterial 16S sequence, an actinobacterial sequence, and a deeply branch-ing clade of bacteria, were of particular interest because of their close relation toknown symbionts, producers of bioactive compounds, and appearance as asponge-specific clade, respectively. In order to ensure specificity, the primerswere designed to target a hypervariable region of the 16S rRNA, and for effi-ciency as probes, the primers target regions of extremely high accessibility on the16S rRNA molecule (3). The specific oligonucleotide primers CC01-1216 (5�-CGACCTCGCGATCTCGCT-3�), CC07-1245 (5�-CGCTTGACCTCGCGGTGT-3�), and SpC1 (5�-CTACACATTCCACCGCTA-3�) were designed. PCR, tar-geting the specific sequences, was performed with the general eubacterialoligonucleotide primer 27f (5�-AGAGTTTGATCMTGGCTCAG-3�) paired

with each of the specific oligonucleotide primers. Thermal cycling conditions forthe PCRs with the specific primers were as follows: initial denaturation for 3 minat 95°C; 35 cycles of denaturation for 30 s at 95°C, annealing for 1 min at 65°C,and elongation for 1 min at 72°C; and a final extension step for 7 min at 72°C.Identity of the amplification products from the specific primers was confirmedwith sequencing. ProbeMatch (RDP; http://rdp.cme.msu.edu/probematch/search.jsp) suggests that the three probes do not match the 16S sequence of anymicrobe known in the database. In order to construct probes for FISH, thespecific oligonucleotide primers were 5�-cyanine 3 (CY3) end labeled. Negative-control probes, CY3-labeled single mismatch probes, were designed to confirmthe specificity of the probes (sequences in Table 1).

Nucleotide sequence accession numbers. The GenBank accession numbers forthe 16S rRNA gene sequences cloned from Corticium sp. are DQ247938 toDQ247957.

RESULTS

Bacteria and archaea in Corticium sp. embryos. Corticiumsp. tissue sections were hybridized simultaneously with theCY5-labeled general eubacterial probe (EUB338) and a CY3-labeled general archaeal probe (ARCH915) (sequences shownin Table 1) by FISH. Strong autofluorescence in sponge cellsallows visualization of sponge structures without counterstain-ing, and a probe-conferred signal is identified by comparisonwith negative controls. Early-stage embryos contain conspicu-ous and regularly arranged clusters of bacteria, resemblingbeads on a necklace, lining the inner periphery of the embryos(Fig. 2A). Archaea are present in these aggregates also but are

FIG. 2. Bacteria and archaea within developing Corticium sp. embryos. (A) CY5-labeled general eubacterial probe (EUB338) reveals clustersof bacterial cells (arrowheads) in the inner periphery of the developing embryo. Bar � 10 �m. (B) Simultaneous hybridization with CY5-EUB338(green) and CY3-ARCH915 (red) shows the presence of both archaea and bacteria in the aggregates (arrowheads) within the Corticium sp.embryos. Bar � 10 �m. (C) Negative controls with probes EUBNON and ARCHNON show no hybridization to cells in the aggregates (lines).Bar � 10 �m. (D) CY5-EUB338 hybridizes to a mass of cells (arrowheads) in the central cavities of later-stage embryos. Bar � 100 �m. (E) Bothbacteria (green) and archaea (red) are present in the cavity (arrowhead) of a developing Corticium sp. embryo. Bar � 100 �m. (F) Negativecontrols with probes EUBNON and ARCHNON show no hybridization to central cavity (arrowhead) in the embryo. Bar � 100 �m. (G) CY5-EUB338 shows that the sponge mesohyl is densely packed with eubacterial cells. Bacteria line the choanocyte chambers (ch) but also appear furtherin the sponge interior. Bar � 60 �m. (H) Bacteria (green) and archaea (red) are present throughout the mesohyl and around the choanocytechambers. Bar � 60 �m. (I) Negative controls with probes EUBNON and ARCHNON show no hybridization to cells of the mesohyl (bar � 60 �m).

624 SHARP ET AL. APPL. ENVIRON. MICROBIOL.

on Decem

ber 19, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

less abundant (Fig. 2B). Negative controls for specificity ofboth ARCH915 and EUB338 show no hybridization to cells inthe aggregates (Fig. 2C). In later developmental stages, bacte-ria are in the central cavity of the developing embryo (Fig. 2D).Localization of bacteria in the late-stage embryos is consistentwith previous ultrastructural studies of Corticium sp. swimminglarvae, in which bacterial cells were noted in the larval cavity(6). In addition, hybridization with CY5-ARCH915 and CY3-EUB338 indicates that small numbers of archaea are locatedclose to host tissues and are absent in the cavity (Fig. 2E).CY3-EUB338NON and CY5-ARCH915NON do not hybrid-ize to cells within developing Corticium sp. embryos (Fig. 2F).CY3-EUB338 and CY5-ARCH915 both hybridize to cells

densely packed in choanocyte chambers and throughout themesohyl of the adult sponge (Fig. 2G and H). Negative-controlprobes (CY3-EUBNON and CY5-ARCH915NON) showedno hybridization to cells in the mesohyl or choanocyte cham-bers of adult Corticium sp. (Fig. 2I).

Bacterial 16S rRNA clone libraries. We constructed threeclone libraries of embryo-associated bacterial 16S rRNAgenes. Each library contained 16S rRNA gene sequences am-plified from the embryos within a Corticium sp. individual fromone of three sampling locations. All three of the librariescontained sequences representing diverse bacterial lineages,and none of the libraries was dominated by any one sequence.Two hundred clones were initially analyzed from the three

FIG. 3. Bacterial sequence diversity in clone libraries. Maximum-likelihood phylogenetic tree of the bacterial 16S rRNA sequences obtained fromembryo clone libraries, based on 1,200 bp. Numbers with CC prefix are Corticium sequences from this study. Boxes indicate sequences targeted by specificprobes (this study). Included are sequences previously found in the marine sponges Aplysina and Theonella, shown in boldface and labeled by sponge hostgenus. Sequences from Corticium sp. embryos fall into the Nitrospira, proteobacteria, actinobacteria, and Chloroflexus groups. Many sequences showaffiliation with other sequences found in marine sponges. The lineage of sequences that consists of sequences only to date found in sponges are labeledsponge clades. *, bootstrap support value of 60%. The scale bar represents 10 substitutions per nucleotide position.

VOL. 73, 2007 VERTICAL MICROBE TRANSMISSION IN A TROPICAL SPONGE 625

on Decem

ber 19, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

Corticium sp. libraries. Initial sequencing suggested that se-quences fell into 19 different groups of closely related se-quences; representatives of each group were sequenced for full(twofold) coverage.

Bacterial phylotypes represented in the embryo clone librar-ies (Fig. 3) showed less than 90% identity to any microbial 16SrRNA gene sequences in the Ribosomal Database Project orGenBank (NCBI) databases. Cloned sequences representedmembers of several groups of bacteria, including the Pro-teobacteria, the Actinobacteria, and the Nitrospira and Chlo-roflexus groups. One of the most abundant sequences, repre-senting CC01, is from an alphaproteobacterium closely relatedto the terrestrial plant-symbiotic Rhizobium-Agrobacteriumgroup. Other proteobacterial sequences cluster with thosefrom sulfur-oxidizing chemoautotrophic gammaproteobacte-rial symbionts. Actinobacterial sequences from the Corticiumsp. embryos were distantly related to most known actinobac-terial 16S rRNA sequences but were most closely related tosequences from microorganisms previously isolated from sea-water and marine sponges. Similarly, other sequences from theCorticium sp. embryos fell into distinct clades with those frommicrobes previously found in marine sponges that have beenidentified as Nitrospira, Chloroflexus, and deltaproteobacteria.Other sequences are close matches to those of unclassifiedmicrobes, including the tentative “Poribacteria” clade proposedby Fieseler et al. (14). In addition, there is a clade of sequencesfrom the embryos that fall into a group with sequences fromother marine sponges but are not closely related to any other

sequences currently in RDP or GenBank databases. This cladeis labeled sponge clade 1 (SpC1).

Specific microbes: PCR survey. Three of the cloned se-quences similar to 16S rRNA sequences of previously de-scribed symbionts or bacteria found in other marine spongeswere to chosen for further investigation. The abundance andlocalization of these bacteria, CC01 (alphaproteobacterium),CC07 (actinobacterium), and SpC1, were analyzed with spe-cific reverse primers designed for PCR amplification and for insitu hybridization to rRNA. The specific primers were pairedwith the general eubacterial primer 27f (sequence in Table 1)for PCR amplification. Sequences for each specific oligonucle-otide and its negative control are shown in Table 1. PCR with�-CC01 and actino-CC07 primers yielded only one 16S se-quence from Corticium sp. samples, and PCR with the SpC1primer yielded a group of closely related sequences. A PCRsurvey with the three specific primers on 12 Corticium sp.samples showed that the two bacterial species (primers�-CC01 and actino-CC07) and members of the SpC1 clade arepresent in Corticium sp. populations across broad temporal(3-year) and geographic (100-km) scales (Fig. 4). Sequencinganalysis of each PCR product confirmed the identity as thesequence targeted by the specific primers and confirmed thatthe sequences from each Corticium sp. sample were identical,except in the case of the SpC1 primer pair, in which they wereclosely related (�2%). Negative controls (no template) wererun for each PCR to confirm that the amplification productswere not due to contamination of reagents (not shown).

FIG. 4. Specific PCR survey. A specific PCR survey demonstrates the presence of two bacterial species (actinobacterium CC07 and alpha-proteobacterium CC01) and SpC1 in all 12 tested Corticium sp. samples collected across the Palau Islands. Sample numbers on the gel picturesand map correspond to the location and year of collection in the table. Circles on the map indicate approximate locations of the 12 collection sitesacross the Palau Islands. Negative controls (no template) were run for each PCR to confirm that the amplification products were not due tocontamination of reagents (not shown). The map was reprinted courtesy of http://www.reefbase.org.

626 SHARP ET AL. APPL. ENVIRON. MICROBIOL.

on Decem

ber 19, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

Specific probes: FISH. The three specific oligonucleotideprimers, CY3 labeled for use as probes (�-CC01, actino-CC07,and SpC1), hybridize to bacterial cells within the peripheralclusters of microbes in early-stage embryos (Fig. 5A, B, and C).Negative control probes (single-base-mismatch probes) weredesigned to test the specificity of each of the three probes. Inall three cases, the negative-control probes demonstrate se-quence specificity: the negative probes do not hybridize to cellsin the aggregates lining the embryos under our experimentalconditions (Fig. 5D, E, and F). The three specific probes hy-bridize to cells within the central cavities of the later-stageembryos (Fig. 6A, B, and C), in addition to the mesohyl of theadult sponge (Fig. 6D, E, and F).

DISCUSSION

Phylogenetic analysis suggests that vertical-transmissionmechanisms are often present in highly coevolved host-mi-crobe associations (31). Previous work on vertical transmissionin sponges has demonstrated the presence of a single microbespecies in developing embryos (12, 30, 42), and molecularmethods indicate the possibility of more-complex assemblagesin the larvae of the sponge M. laxissima (12). Our embryo-based research on Corticium sp. revealed that both bacteriaand archaea are vertically transmitted. A detailed study of thebacterial community showed remarkable diversity of bacteriain Corticium sp. embryos. At least three of these, but likelymore of them, are constant across time and space. The se-quences identified in this study add to the growing collection of16S rRNA sequences found in diverse marine sponges from

diverse regions around the world. Phylogenetic analysis showsthat Corticium sp. hosts a diverse suite of bacteria in its em-bryos, including bacteria from phylogenetic groups previouslyobserved in other marine sponges (20, 22, 39, 47). Because thecomposition of this community is strikingly similar to thatpreviously described for many other sponges, these results laya foundation for future research on transmission of similar,complex microbial communities in other sponges. Similar di-verse microbial communities found in different sponge ordersaround the globe, “sponge specialists” (20, 22), may be mem-bers of specific, highly evolved associations that are maintainedand regulated by reproductive transmission in the host sponge.

While the composition of sequences found in the Corticiumsp. bacterial 16S rRNA gene clone library reflects the makeupof other described bacterial communities in sponges, some ofthe previously well-characterized sponge-associated bacteriaare not closely related to those represented by any of thesequences found in Corticium sp. in this study. For instance,there are many well-documented associations between spongesand cyanobacteria (4, 30, 40, 42), but PCR from Corticium sp.embryo DNA did not yield any cyanobacterial 16S sequences.In addition, very few autofluorescent prokaryotic cells wereobserved in sections of embryos or mesohyl, indicating a strik-ing absence of photosynthetic bacteria in Corticium sp. Though�-CC01 16S sequences were a relatively large portion of thesequences in the embryo clone libraries (approximately 30% ofthe sequences), neither the alphaproteobacterium NW001, iso-lated from the tropical Pacific sponge Rhopaloeides odorabile(12, 46), nor the clade of closely related bacteria found in other

FIG. 5. Specific probes hybridize to cells in aggregates in Corticiumsp. embryos. CY3-labeled probes specific for the alphaproteobacte-rium CC01 (A), the actinobacterium CC07 (B), and the clade SpC1(C) hybridize to cells (arrowheads) in the aggregates in the embryos.Single-base-mismatch probes, negative controls testing probe specific-ity, do not hybridize to cells in the bacterial aggregates (lines) in theembryos (D, E, and F).

FIG. 6. Specific probes hybridize to cells in the central cavities oflater-stage embryos (A–C) and in adult sponge mesohyl (D–F). CY3-labeled probes specific for the alphaproteobacterium CC01 (A), theactinobacterium CC07 (B), and the clade SpC1 (C) hybridize to cells(arrowheads) in the mass in the later-stage larval cavity. Specificprobes also hybridize to cells occurring throughout the mesohyl andsurrounding the choanocyte chambers (panels D, E, and F, respec-tively). ch, choanocyte chambers.

VOL. 73, 2007 VERTICAL MICROBE TRANSMISSION IN A TROPICAL SPONGE 627

on Decem

ber 19, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

sponges (12) was found in Corticium sp. embryo libraries. Inaddition, none of the sequences from the embryos were closelyrelated to those from NW001. A potential explanation is thatNW001 is not present in Corticium sp. However, the clonelibraries in this study were unlikely to have exhausted thediversity of the community in the embryos. Enticknap et al.(12) also noted that, while NW001-like alphaproteobacteriamake up an estimated 50% of the biomass in M. laxissimaembryos, they are underrepresented in 16S rRNA gene clonelibraries, perhaps due to PCR primer bias. Further investi-gation would be necessary to determine whether membersof the NW001-like clade of alphaproteobacteria, a proposed“sponge-specific” lineage, are truly absent in the tropical Pa-cific sponge Corticium sp. Another interesting observation isthat the NW001-like alphaproteobacterium seems to be con-centrated around and in the embryos in the sponge M. laxis-sima but is not highly abundant in the mesohyl (12). The threemicrobial groups examined in this study appear in Corticiumsp. embryos and also are evenly distributed throughout Corti-cium mesohyl (Fig. 5 and 6).

FISH with specific probes confirms that the embryo-basedDNA libraries contained sequences that were present withinembryos, not merely contaminants from the surrounding sea-water. The specific groups are consistently present in Corticiumsp. individuals over both spatial and temporal scales (Fig. 4).Each of the three phylotypes was present in all tested Corti-cium sp. samples from the Palau Islands, spanning nearly 100km and 3 years of sampling, further evidence for a long-termassociation between the sponge host and the microbial assem-blage. Our specific FISH demonstrates that at least three spe-cific bacterial phylotypes from the libraries are consistentlyassociated with Corticium sp. embryos throughout their devel-opment. However, the images show that none of the three phy-lotypes is a dominant portion of the bacterial biomass in theembryos, suggesting that other bacteria—whether or not they arerepresented in the clone libraries—are also present in the em-bryos. If there is a single microbial species that is numericallydominant in the embryos, it is yet to be identified. Alternatively,the bacterial assemblage is diverse, and none of the bacteria areparticularly dominant in number or biomass. Further probing inCorticium sp. embryos for groups of bacteria that appear to besponge specific will broaden our understanding of the diversityand transmission of microbe-sponge associations.

The prokaryotic community in Corticium sp. includes bacte-ria whose closest relatives are the sulfur-oxidizing gammapro-teobacteria in invertebrates of highly reduced environments(Codakia symbionts, Riftia symbionts, scaly snail symbiont).There are others that fall into a clade with the nitrogen-fixingRhizobiales. In addition, many of the sequences from Corticiumembryos are related to actinobacteria, a division of the bacteriaknown for its production of diverse and complex bioactivecompounds. While the archaea in Corticium sp. are yet to beidentified, ammonia-oxidizing archaea have been shown to bequite widespread in many marine environments, including theinterior of sponges (15, 17, 24), and further exploration ofammonia-oxidizing archaea in sponges is necessary to under-stand whether they are widespread in marine sponges. Thoughit is unknown how or if the transmitted microbial communityfunctions as part of the host sponge physiology, it seems likelythat this sponge and others possess a predictable set of eubac-

terial and archaeal partners that construct chemical microen-vironments within the animal host and live in syntrophy, cy-cling nutrients and carbon within the sponge. In addition, somemicrobes may prevent predation of the host via production ofbioactive molecules.

Recent work outlines different possible modes of symbiontincorporation from adult mesohyl into embryos or eggs ofHalisarca dujardini (13). Our study highlights the utility ofusing probe-based technology to visualize microbes in earlysponge embryonic stages, allowing new insight into the pro-cesses governing inoculation of sponge embryos. FISH probesreveal that, during embryogenesis in Corticium sp., bacteriaform aggregates within the embryos, and by later stages ofdevelopment, the bacteria are primarily in the central cavity.This embryonic acquisition of bacteria may indicate a fairlyspecialized mechanism of transfer during (or before) embry-onic development, such as those previously described in thesponge Chondrilla australiensis (34, 35, 42). The presence ofbacterial cells in the central cavity of the embryos is consistentwith the description of bacteria in the cavities of Corticium sp.larvae (6). We also show that Corticium sp. possesses a specific,diverse assemblage of bacteria and archaea, within its embryos,and the microbial assemblage is constant across individualssampled across a 3-year and 100-km spread in the PalauIslands. It is striking that the composition of the embryo-associated community is similar to those found in many othersponge species. While sponge-bacterium symbioses are widelyaccepted to be ancient associations, the data here show for thefirst time that a highly complex microbial assemblage, likethose found in many sponges, is maintained intergeneration-ally. Further use of 16S rRNA-based molecular approaches ondiverse sponges will test whether vertical transmission is acommon strategy for the maintenance of specific symbioses inmarine sponges.

ACKNOWLEDGMENTS

This work was funded by California Sea Grant (R/MP-88), theNational Institutes of Health (grant 5R01CA079678-03), and theScripps Institution of Oceanography Graduate Student Office. K.H.S.was supported by the Los Angeles ARCS Foundation.

Collection efforts in Palau were made possible by Pat and Lori Colinand the Coral Reef Research Foundation. We thank CatherineSincich, Christian Ridley, Melissa Lerch, and Joel Sandler for assis-tance with SCUBA collection of Corticium sp. in Palau. We thankNancy Knowlton and Nick Holland for their review of the manuscript.

REFERENCES

1. Altschul, S. F., W. Gish, W. Miller, E. W. Wyers, and D. J. Lipman. 1990.Basic Local Alignment Search Tool. J. Mol. Biol. 215:403–410.

2. Amann, R. I., B. J. Binder, R. J. Olson, S. W. Chisholm, R. Devereux, andD. A. Stahl. 1990. Combination of 16S rRNA-targeted oligonucleotideprobes with flow cytometry for analyzing mixed microbial populations. En-viron. Microbiol. 56:1919–1925.

3. Behrens, S., C. Ruhland, J. Inacio, H. Huber, A. Fonseca, I. Spencer-Martins, B. M. Fuchs, and R. Amann. 2003. In situ accessibility domains ofsmall-subunit rRNA of members of the domains Bacteria, Archaea, andEucarya to Cy3-labeled oligonucleotide probes. Appl. Environ. Microbiol.69:1748–1753.

4. Bewley, C. A., N. D. Holland, and D. J. Faulkner. 1996. Two classes of metab-olites from Theonella swinhoei are localized in distinct populations of bacterialsymbionts. Experientia (Basel) 52:716–722.

5. Blazejak, A., C. Erseus, R. Amann, and N. Dubilier. 2005. Coexistence ofbacterial sulfide oxidizers, sulfate reducers, and spirochetes in a gutless worm(Oligochaeta) from the Peru Margin. Appl. Environ. Microbiol. 71:1553–1561.

6. Boury-Esnault, N., A. Ereskovsky, C. Bezac, and D. Tokina. 2003. Larval

628 SHARP ET AL. APPL. ENVIRON. MICROBIOL.

on Decem

ber 19, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

development in the Homoscleromorpha (Porifera, Demospongiae). Invertebr.Biol. 122:187–202.

7. Cannone, J. J., S. Subramanian, M. N. Schnare, J. R. Collett, L. M. D’Souza,Y. Du, B. Feng, N. Lin, L. V. Madabusi, K. M. Muller, N. Pande, Z. Shang,N. Yu, and R. R. Gutell. 2002. The comparative RNA web (CRW) site: anonline database of comparative sequence and structure information for ri-bosomal, intron, and other RNAs. BMC Bioinformatics 3:15.

8. Cary, S. C. 1994. Transovarial inheritance of endosymbiotic bacteria in theprotobranch bivalve, Solemya reidi. EOS 75:60.

9. Cole, J. R., B. Chai, T. L. Marsh, R. J. Farris, Q. Wang, S. A. Kulam, S.Chandra, D. M. McGarrell, T. M. Schmidt, G. M. Garrity, and J. M. Tiedje.2003. The Ribosomal Database Project (RDP-II): previewing a new au-toaligner that allows regular updates and the new prokaryotic taxonomy.Nucleic Acids Res. 31:442–443.

10. Distel, D. L., D. J. Beaudoin, and W. Morrill. 2002. Coexistence of multipleproteobacterial endosymbionts in the gills of the wood-boring bivalveLyrodus pedicellatus (Bivalvia: Teredinidae). Appl. Environ. Microbiol. 68:6292–6299.

11. Dubilier, N., C. Mulders, T. Ferdelman, D. de Beer, A. Pernthaler, M. Klein,M. Wagner, C. Erseus, F. Thiermann, J. Krieger, O. Giere, and R. Amann.2001. Endosymbiotic sulphate-reducing and sulphide-oxidizing bacteria in anoligochaete worm. Nature 411:298–302.

12. Enticknap, J. J., M. Kelly, O. Peraud, and R. T. Hill. 2006. Characterizationof a culturable alphaproteobacterial symbiont common to many marinesponges and evidence for vertical transmission via sponge larvae. Appl.Environ. Microbiol. 72:3724–3732.

13. Ereskovksy, A. V., E. Gonobobleva, and A. Vishnyakov. 2005. Morphologicalevidence for vertical transmission of symbiotic bacteria in the viviparoussponge Halisarca dujardini Johnston (Porifera, Demospongiae, Halisarcida).Mar. Biol. 146:869–875.

14. Fieseler, L., M. Horn, M. Wagner, and U. Hentschel. 2004. Discovery of thenovel candidate phylum “Poribacteria” in marine sponges. Appl. Environ.Microbiol. 70:3724–3732.

15. Francis, C. A., K. J. Roberts, J. M. Beman, A. E. Santoro, and B. B. Oakley.2005. Ubiquity and diversity of ammonia-oxidizing archaea in water columnsand sediments of the ocean. Proc. Natl. Acad. Sci. USA 102:14683–14688.

16. Gros, O., L. Frenkiel, and H. Felbeck. 2000. Sulfur-oxidizing endosymbiosisin Divaricella quadrisulcata (Bivalvia: Lucinidae): morphological, ultrastruc-tural, and phylogenetic analysis. Symbiosis 29:293–317.

17. Hallam, S. J., T. J. Mincer, C. Schleper, C. M. Preston, K. Roberts, P. M.Richardson, and E. F. DeLong. 2006. Pathways of carbon assimilation andammonia oxidation suggested by environmental genomic analyses of marineCrenarchaeota. PLoS Biol. 4:520–536.

18. Haygood, M., and S. Davidson. 1997. Small-subunit rRNA genes and in situhybridization with oligonucleotides specific for the bacterial symbionts in thelarvae of the bryozoan Bugula neritina and proposal of “Candidatus en-dobugula sertula.” Appl. Environ. Microbiol. 63:4612–4616.

19. Hentschel, U., L. Fieseler, M. Wehrl, C. Gernert, M. Steinert, and J. Hacker.2003. Microbial diversity of marine sponges, p. 59–88. In W. E. G. Muller(ed.), Marine molecular biotechnology. Springer-Verlag, Berlin, Germany.

20. Hentschel, U., J. Hopke, M. Horn, A. B. Friedrich, M. Wagner, J. Hacker,and B. S. Moore. 2002. Molecular evidence for a uniform microbial commu-nity in sponges from different oceans. Appl. Environ. Microbiol. 68:4431–4440.

21. Hentschel, U., M. Schmid, M. Wagner, L. Fieseler, C. Gernert, and J.Hacker. 2001. Isolation and phylogenetic analysis of bacteria with antimi-crobial activities from the Mediterranean sponges Aplysina aerophoba andAplysina cavernicola. FEMS Microbiol. Ecol. 35:305–312.

22. Hill, M., A. Hill, N. Lopez, and O. Harriott. 2006. Sponge-specific bacterialsymbionts in the Caribbean sponge, Chondrilla nucula (Demospongiae,Chondrosida). Mar. Biol. 148:1221–1230.

23. Hirose, E. 2000. Plant rake and algal pouch of the larvae in the tropicalascidian Diplosoma similis: an adaptation for vertical transmission of photo-synthetic symbionts Prochloron sp. Zool. Sci. (Tokyo) 17:233–240.

24. Konneke, M., A. E. Bernhard, J. R. de la Torre, C. B. Walker, J. B. Waterbury,and D. A. Stahl. 2005. Isolation of an autotrophic ammonia-oxidizing marinearchaeon. Nature 437:543–546.

25. Koropatnick, T. A., J. T. Engle, M. A. Apicella, E. V. Stabb, W. E. Goldman,

and M. J. McFall-Ngai. 2004. Microbial factor-mediated development in ahost-bacterial mutalism. Science 306:1186–1188.

26. Krueger, D. M., R. G. Gustafson, and C. M. Cavanaugh. 1996. Verticaltransmission of chemoautotrophic symbionts in the bivalve Solemya velum(Bivalvia: Protobranchia). Biol. Bull. 190:195–202.

27. Lane, D. S. 1990. 16S and 23S rRNA sequencing, p. 115–148. In E. Stacke-brandt and M. Goodfellow (ed.), Nucleic acid techniques in bacterial sys-tematics. John Wiley, New York, NY.

28. Loy, A., M. Horn, and M. Wagner. 2003. probeBase—an online resource forrRNA-targeted oligonucleotide probes. Nucleic Acids Res. 31:514–516.

29. Nyholm, S. V., and M. J. McFall-Ngai. 2004. The winnowing: establishing thesquid-Vibrio symbiosis. Nat. Rev. Microbiol. 2:632–642.

30. Oren, M., L. Steindler, and M. Ilan. 2005. Transmission, plasticity and themolecular identification of cyanobacterial symbionts in the Red Sea spongeDiacarnus erythraenus. Mar. Biol. 148:35–41.

31. Peek, A. S., R. A. Feldman, R. A. Lutz, and R. C. Vrijenhoek. 1998. Cospe-ciation of chemoautotrophic bacteria and deep sea clams. Proc. Natl. Acad.Sci. USA 95:9962–9966.

32. Preston, C. M., K. Y. Wu, T. F. Molinski, and E. F. Delong. 1996. Apsychrophilic crenarchaeon inhabits a marine sponge: Cenarchaeum symbio-sum gen. nov., sp. nov. Proc. Natl. Acad. Sci. USA 93:6241–6246.

33. Regoli, F., C. Cerrano, E. Chierici, S. Bompadre, and G. Bavestrello. 2000.Susceptibility to oxidative stress of the Mediterranean demosponge Petrosiaficiformis: role of endosymbionts and solar irradiance. Mar. Biol. 137:453–461.

34. Sciscioli, M., E. Lepore, M. Gherardi, and L. S. Liaci. 1994. Transfer ofsymbiotic bacteria in the mature oocyte of Geodia cydonium: (Porifera,Demospongiae): an ultrastructural study. Cah. Biol. Mar. 35:471–478.

35. Sciscioli, M., E. Lepore, M. Mastrodonato, L. S. Liaci, and E. Gaino. 2002.Ultrastructural study of the mature oocyte of Tethya aurantium (Porifera:Demospongiae). Cah. Biol. Mar. 43:1–7.

36. Sipe, A. R., A. E. Wilbur, and S. C. Cary. 2000. Bacterial symbiont trans-mission in the wood-boring shipworm Bankia setacea (Bivalvia: Teredinidae).Appl. Environ. Microbiol. 66:1685–1691.

37. Stahl, D. A., and R. Amann. 1991. Development and application of nucleicacid probes, p. 205–248. In E. S. A. M. Goodfellow (ed.), Nucleic acidtechniques in bacterial systematics. John Wiley & Sons Ltd., Chichester,England.

38. Swofford, D. L. 2003. PAUP�: phylogenetic analysis using parsimony (� andother methods), version 4.0b 10. Sinauer Associates, Sunderland, MA.

39. Taylor, M. W., P. J. Schupp, I. Dahllof, S. Kjelleberg, and P. D. Steinberg.2004. Host specificity in marine sponge-associated bacteria, and potentialimplications for marine microbial diversity. Environ. Microbiol. 6:121–130.

40. Unson, M. D., N. D. Holland, and D. J. Faulkner. 1994. A brominatedsecondary metabolite synthesized by the cyanobacterial symbiont of a marinesponge and accumulation of the crystalline metabolite in the sponge tissue.Mar. Biol. 119:1–11.

41. Usher, K. M., and A. V. Ereskovskyz. 2005. Larval development, ultrastruc-ture and metamorphosis in Chondrilla australiensis Carter, 1873 (Demo-spongiae, Chondrosida, Chondrillidae). Invertebr. Reprod. Dev. 47:51–62.

42. Usher, K. M., J. Kuo, J. Fromont, and D. C. Sutton. 2001. Vertical trans-mission of cyanobacterial symbionts in the marine sponge Chondrillaaustraliensis (Demospongiae). Hydrobiologia 461:15–23.

43. Vacelet, J., and C. Donadey. 1977. Electron microscope study of the associ-ation between some sponges and bacteria. J. Exp. Mar. Biol. Ecol. 30:301–314.

44. Vacelet, J., A. Fiala-Medioni, C. R. Fisher, and N. Boury-Esnault. 1996.Symbiosis between methane-oxidizing bacteria and a deep-sea carnivorouscladorhizid sponge. Mar. Ecol. Prog. Ser. 145:77–85.

45. Wallner, G., R. Amann, and W. Beisker. 1993. Optimizing fluorescent in situhybridization with rRNA-targeted oligonucleotide probes for flow cytomet-ric identification of microorganisms. Cytometry 14:136–143.

46. Webster, N. S., and R. T. Hill. 2001. The culturable microbial community ofthe Great Barrier Reef sponge Rhopaloeides odorabile is dominated by analpha-proteobacterium. Mar. Biol. 138:843–851.

47. Webster, N. S., K. J. Wilson, L. L. Blackall, and R. T. Hill. 2001. Phyloge-netic diversity of bacteria associated with the marine sponge Rhopaloeidesodorabile. Appl. Environ. Microbiol. 67:434–444.

VOL. 73, 2007 VERTICAL MICROBE TRANSMISSION IN A TROPICAL SPONGE 629

on Decem

ber 19, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from