Marine Biology (2004) 144: 583–592 DOI 10.1007/s00227-003-1227-0
R ES E AR C H A RT I C L E
I. J. Domart-Coulon Æ C. S. Sinclair Æ R. T. Hill S. Tambutte´ Æ S. Puverel Æ G. K. Ostrander
A basidiomycete isolated from the skeleton of Pocillopora damicornis (Scleractinia) selectively stimulates short-term survival of coral skeletogenic cells Received: 21 May 2003 / Accepted: 12 September 2003 / Published online: 5 November 2003 Springer-Verlag 2003
Abstract Endolithic fungi bore through the extracellular calcium carbonate skeleton of reef-building scleractinian corals, both healthy and dead, and effect net erosion of coral reefs. Potential fungal interactions with coral tissue were investigated using an in vitro approach suggested by earlier observations of skeletal repair cones at the site of fungal perforation in Porites sp. A fungal strain was isolated from the skeleton of a long-term culture of healthy, tissue-covered, Pocillopora damicornis Linnaeus colonies maintained in a recirculating system in Monaco. As coral soft tissue spontaneously dissociated in vitro, the skeleton became exposed and hyaline hyphae emerged radially from 15% of the total clipped branches. In this study, which was performed between January 2001 and March 2003, 35 skeleton–hypha explants were embedded in agar-based solid medium, yielding 60% hyphal growth. A fungal strain (F19-3-1) of the dominant (80%) morphology was isolated and propagated in agar-based solid medium. The strain was Communicated by J.P. Grassle, New Brunswick I. J. Domart-Coulon Æ S. Tambutte´ Æ S. Puverel Centre Scientifique de Monaco, Avenue Saint Martin, Monaco Ville, MC 98000, Monaco C. S. Sinclair Department of Biological Sciences, Towson University, 8000 York Road, Towson, MD 21252, USA R. T. Hill Center of Marine Biotechnology, University of Maryland Biotechnology Institute, Baltimore, MD 21202, USA G. K. Ostrander (&) Department of Biology, Johns Hopkins University, 3400 N. Charles Street, Baltimore, MD 21218, USA E-mail: gofi
[email protected] Fax: +1-410-5164100 G. K. Ostrander Department of Comparative Medicine, Johns Hopkins University, 459 Ross Building, Baltimore, MD 21218, USA G. K. Ostrander National Aquarium in Baltimore, 501 E. Pratt Street, Baltimore, MD 21202, USA
identified by 18S and 26S rDNA gene sequence analysis as a basidiomycete in the genus Cryptococcus. Cocultures were used to provide experimental exposure of coral soft tissue to the fungus. The fungus extended the survival of coral cells by 2 days, selectively maintaining skeletogenic cell types. This effect may be interpreted as stimulation by the fungus of a short-term coral defense response.
Introduction A diverse community of endolithic microorganisms is associated with the calcium carbonate extracellular skeleton of reef-building scleractinian corals. Cyanobacteria, heterotrophic filamentous fungi, and phototrophic green algae of the genus Ostreobium sp. have been detected within coral skeleton (Wainwright 1963; Kohlmeyer 1969; Kendrick et al. 1982). Boring through the aragonitic biomineral and occasionally spreading into the skeletal pores, i.e. the space left after skeleton evacuation by the polyps, they may have a major impact on net coral bioerosion (Tribollet et al. 2002). Bioerosion was thought to be the entire extent of the fungal– coral association, and therefore fungal interactions with tissue were not deemed significant. However, a scanning electron microscopy study of Porites sp. skeletal microarchitecture revealed calcium carbonate repair cones at the site of fungal perforation of the skeletal surface (Le Campion-Alsumard et al. 1995). It was suggested that Porites sp. coral tissue resisted the attempted exit of fungal hyphae from the skeleton by locally depositing new layers of skeletal material: once fungi pierced through the repair cones, coral polyps withdrew upward, evacuating the previously occupied skeletal pores. A potential parasitic attack of the coral by these unidentified filamentous fungi was suggested (Le CampionAlsumard et al. 1995). A diverse population of endolithic fungi has since been confirmed within the carbonate skeleton of live,
584
tissue-covered, massive reef-building corals, both in healthy and in diseased colonies (Bentis et al. 2000; Priess et al. 2000; Ravindran et al. 2001). Ravindran et al. (2001) detected fungi in the polyp zone of Porites lutea collected in the Arabian Sea, up to 5 mm under the skeletal surface. They estimated that total fungal biomass accounted for 0.04–0.05% of the live coral wet weight. The morphology of fungal hyphae varies with their location within the coral skeleton, and has been related to their various life forms (Priess et al. 2000; Ravindran et al. 2001). Endolithic thin hyphae (1–3 lm wide) have been localized within the carbonate skeleton and occasionally around the polyps, whereas cryptoendolithicswollen hyphae emerged into the skeletal pores evacuated by the polyps, extending fruiting bodies into the seawater (Priess et al. 2000). There is no record of sporulation. Based on their morphological characteristics, several filamentous fungi from live coral skeleton have tentatively been identified to genus. Fungal diversity was shown to be high, with 20 fungal taxa isolated (Kendrick et al. 1982). Fungi belonging to terrestrial genera were detected, mostly Aspergillus sp. ascomycetes (Kohlmeyer et al. 1987; Priess et al. 2000) and some Fusarium sp. deuteromycetes, and unidentified mycelial yeasts (Ravindran et al. 2001). Dark and hyaline nonsporulating, septate, mycelial fungi were frequently detected within the skeleton of healthy-looking, tissue-covered Porites sp. (Ravindran et al. 2001; Le Campion-Alsumard, personal communication). A dark strain, morphologically identified as Curvularia lunata, and an unknown hyaline strain were isolated from broken, live P. lutea pieces plated for 1 week in corn meal agar, and their abundance within live coral was confirmed using immunological probes (Ravindran et al. 2001). In the absence of propagation of these fungi in artificial culture, there has been no molecular identification to confirm genus and species identities. Whether fungi interact with coral cells is unclear, and pathogenicity has not been demonstrated either, except for Aspergillus sydowii in gorgonians (Smith et al. 1996). Scanning electron microscopy has shown that another coral skeleton endolith, the siphonal alga Ostreobium sp., was frequently invaded by undetermined filamentous fungi, attached to the algae via specialized hyphal branches (haustoria) and growing inside them. Their parasitic attack resulted in progressive loss of algal pigmentation and structural deterioration (Le CampionAlsumard et al. 1995; Bentis et al. 2000; Priess et al. 2000). Specialized fungal haustoria have not been documented within coral tissue. So far, the evidence for a potential interaction of coral cells with fungi is indirect, based on the detection of repair carbonate material in the skeletal microstructure (Le Campion-Alsumard et al. 1995; Ravindran et al. 2001). Our objectives were to isolate, culture, and identify marine fungal residents of healthy Pocillopora damicornis and to investigate whether exposure to the fungus produced a coral cytological response.
Materials and methods Corals These experiments were carried out between January 2001 and March 2003. Small colonies of Pocillopora damicornis Linnaeus (Cnidaria: Anthozoa: Scleractinia), indigenous to the Indo-Pacific were maintained in long-term cultures at the Centre Scientifique de Monaco. Five coral colonies were kept in a recirculating system with Mediterranean seawater heated to 27C, pH 8.1–8.2. Colonies were illuminated with four 96 W full spectrum fluorescence bulbs (Custom Sealife) on a 12 h light:12 h dark cycle, and fed with Artemia sp. nauplii twice a week. Coral colonies exhibited 100% tissue cover and a daily 0.3% (dry weight) growth rate, as determined during a 3-month observation period (C. Richard, personal communication).
Reagents All chemicals were purchased from Sigma Chemicals (St. Louis, Mo.). Commercial cell culture reagents were purchased from Invitrogen Life Technologies (Carlsbad, Calif.), and all reconstituted solutions were filtered on 0.2 lm Nunclon filter units. Calcium-free seawater (23 g NaCl, 0.763 g KCl, 1.89 g MgSO4-7 H2O, 10.45 g MgCl2-6 H2O, 3 g Na2SO4, 0.25 g NaHCO3 l)1 of deionized H2O) and artificial seawater (24.5 g NaCl, 0.75 g KCl, 1.1 g CaCl2, 10.2 g MgCl2-6 H2O, 1 g Na2SO4 l)1 of deionized H2O) were prepared according to Frank et al. (1994). Vertebrate cell culture medium (DMEM) was made isotonic to surface seawater ionic ratios by adding 18.1 g l)1 NaCl, 0.35 g l)1 KCl, 1.1 g l)1 CaCl2, 10.2 g l)1 MgCl2-6 H2O, 1 g l)1 Na2SO4 and was supplemented with 0.052 g l)1 taurine and buffered to pH 8.0 with 5.96 g l)1 Hepes (supplemented DMEM). Coral cell culture medium was modified from Kopecky and Ostrander (1999) and contained (vol/ vol) 12.5% supplemented DMEM, as well as 1.25% fetal calf serum (Gibco no. 10270-098, heat-inactivated lot no. 40G7610 K) in artificial seawater. Antibiotic–antimycotic solution containing 10,000 U ml)1 penicillin G sodium, 10,000 lg ml)1 streptomycin sulfate, and 25 lg ml)1 amphotericin B as Fungizone (Gibco BRL no. 15240-096) was added to a 1% final concentration in the culture initiation step (spontaneous dissociation of tissue from the skeleton). Brief exposure to the antibiotic–antimycotic solution limited contamination of the coral explant primary cultures by superficial tissue microorganisms, but did not prevent emergence of endogenous fungal microorganisms from the skeleton evacuated by tissue. Subsequent fungal cultures and coral cell–fungus co-cultures were grown in medium free of the antibiotic–antimycotic solution. The biological effect of the fungus was only recorded in co-cultures free of the antibiotics–antimycotics.
Fungi isolation and culture Coral cell explant primary cultures were initiated according to Domart-Coulon et al. (2001), modified from Kopecky and Ostrander (1999). For each experiment, three to six fast-growing apical branch tips (0.3–0.5 cm long) were clipped from the same parent colony and collected in a Petri dish filled with calcium-free artificial seawater, thereby preventing exposure of the broken surface to air, to avoid contamination by opportunistic aerial fungi. All further manipulations were carried out in a laminar flow cell culture hood. The five parent coral colonies were used in turn, for a total of 250 clipped coral branches. Culture dishes were disposable, six-well, flat-bottom, polystyrene Evergreen tissue culture plates (Bioblock, Strasbourg). To remove coral surface contaminants, excised tissue-covered coral fragments were incubated for 3 h at room temperature, under 100 rpm agitation in (vol/vol) 3% antibiotics–antimycotics in calcium-free artificial seawater. This rinsing step eliminated most of the coral mucus along with some cells of the external (oral) tissue layer. Access of the
585 antibiotics–antimycotics to the skeleton of the coral was limited to its clipped section. Rinsed coral fragments were then plated into explant primary cultures in 5 ml of cell culture medium, one fragment per individual well. Cultures were incubated at 27C, in air, with a 12 h light:12 h dark illumination cycle. Coral soft tissue spontaneously detached from the skeleton within 2–4 days, dissociating into single cells and positively buoyant, ciliated, multicellular isolates. Once tissue detachment was complete and soft tissue was removed, skeleton became exposed and fungal hyphae emerged radially within 2 weeks in 15% of the plated coral branches, extending over the surface and into the surrounding cell culture medium in a network of hyaline branching filaments. Thirty-five skeleton fragments with protruding fungal hyphae were picked from explant primary cultures with alcohol-sterilized tweezers and embedded in 1% agar solid medium in individual Petri dishes. Solid medium was prepared with volume-to-volume-autoclaved 2% agar in deionized H2O, and 2· concentration artificial seawater (ASW) or cell culture medium without antibiotics–antimycotics (CCM). Fungal cultures were incubated at room temperature in air with natural illumination, and checked for growth of fungal hyphae (mycelium) extending from the coral skeleton into the agar. Control plain agar cultures without coral skeleton (agar-CCM) were prepared with 0.2-lm-filtered cell culture reagents and carried out simultaneously to control for the effect of agar and cell culture medium on the cells. Hyphal growth was recorded in 60% of the skeleton-embedded fungal cultures. Of these, 80% (i.e. 17 skeleton cultures) shared a similar fungal morphology: hyaline and septate hyphae, growing inside the agar medium at the rate of 2 mm month)1. Fungal strain F19-3-1 was isolated from one such skeleton fungal culture by repeated transfer of agar sections of hyphal apices in 1% agar solid CCM medium. It was established and propagated in hyphal fungal cultures without skeleton, in liquid or 1% agar ASW or CCM. Its slow growth limited its propagation to three passages over a 12-month period. Maintenance of its biological effect on coral cells was confirmed throughout these passages.
Fungal DNA extraction Genomic DNA was extracted from F19-3-1 as previously described (Sreenivasaprasad 2000). Briefly, the agar section containing hyphae was removed from the plate with a sterile scalpel and placed in a microcentrifuge tube. Care was taken to remove only agar containing hyphae, and any excess was trimmed before proceeding with the extraction. The tube was briefly dipped in liquid nitrogen to freeze the sample. The frozen sample was then ground into powder with a micropestle. Lysis buffer (200 mM Tris HCl pH 7.5, 250 mM NaCl, 1 mM EDTA, 1% SDS) was added, and the mixture was vortexed. To disrupt the fungal cell walls, three cycles of rapid freezing in liquid nitrogen and thawing at 100C were performed, followed by 10 min incubation at 100C. Cell extracts were recovered by centrifugation at 15,000 g for 3 min and collection of the supernatant. DNA was purified from the supernatant using QiaQuick columns (Qiagen, Valencia, Calif.) following the manufacturers protocol for gel extraction of DNA. This protocol was used to eliminate any remaining agar. DNA was eluted from the QiaQuick column with 1 mM Tris HCl, pH 8.5, at 50C.
Fungal 18S and 26S rDNA sequence analysis Fungal rDNA primers [forward primer 18S: 5¢-GTGAGCCTGCATGTCGTTTA-3¢; reverse primer 18S: 5¢-TCTGGACCTGGTGAGTTTCC-3¢; forward primer 26S: 5¢-GCATATCAATAAGCGGAGGAAAAG-3¢ (Fell et al. 2000); reverse primer 26S: 5¢-GGTCCGTGTTTCAAGACGG-3¢ (Fell et al. 2000)] were used to amplify rDNA gene fragments from strain F19-3-1. Primers for the 18S rDNA were designed using PRIMER3 software (http:// www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi) (Rozen and Skaletsky 2000). Reactions were performed in a GeneAmp 9700 thermocycler (PE Applied Biosystems, Foster City, Calif.).
Five microliters of DNA were added to 15 ll PCR-Master mix containing 2· reaction buffer (TaKaRa, Shiga, Japan), 200 lM dNTPs (TaKaRa), 0.5 lM primers, and 0.5 U Ex Taq (TaKaRa). Reaction conditions were as follows: an initial denaturation of 95C for 3 min; followed by 35 cycles of 95C for 30 s, 52C (18S rDNA) or 55C (26S rDNA) for 30 s, and 72C for 45 s; and a final incubation of 5 min at 72C. PCR (polymerase chain reaction) products were cloned using the original TA cloning kit (Invitrogen Life Technologies) following manufacturer protocols. Plasmids with appropriately sized inserts (18S rDNA—521 bp, 26S rDNA—625 bp) were identified by restriction digest, and insert DNA was sequenced on an Applied Biosystems 3700 DNA analyzer in the DNA Analysis Facility at Johns Hopkins Medical Institution. Fungal identification and phylogenetic analysis Sequence data were analyzed by comparison to rDNA gene sequences in the Ribosomal Database Project (Maidak et al. 1999) and the National Center for Biotechnological Information (NCBI) GenBank database (http://www.ncbi.nlm.nih.gov). All sequences were manually aligned using Phydit software (Chun 1995). The nearest relatives of the organism were obtained by BLAST searches of GenBank (Altschul et al. 1990). Phylogenetic trees were then inferred using the neighbor-joining (Saitou and Nei 1987), Fitch– Margoliash (Fitch and Margoliash 1967), and maximum-parsimony (Kluge and Farris 1969) algorithms in the PHYLIP package (Felsenstein 1993). Evolutionary distance matrices for the neighbor-joining and Fitch–Margoliash methods were generated as described by Jukes and Cantor (1969). Tree topologies were evaluated after 1,000 bootstrap re-samplings of the neighbor-joining data. Fungal morphology In order to detect hyphal compartments and septae, fungal esterase activity was visualized by hydrolysis of the vital stain fluorescein diacetate (FDA) into a fluorescent product (488 nm excitation, 530 nm emission). An agar section of F19-3-1 hyphal fungal culture was flooded with filtered seawater, stained for 15 min with 0.1 mM FDA (Sigma), rinsed extensively in filtered seawater and observed in epifluorescence (·100) on a Leica TCS4D inverted microscope. Skeleton cultures of fungal hyphae were fixed and processed without decalcification for observation in electron microscopy. For scanning electron microscopy, samples were fixed with 2.5% glutaraldehyde in 0.18 M Sorensen buffer pH 7.4 containing 0.6 M sucrose, impregnated with 30% glycerol in H2O, and cryofractured in liquid N2 with a metal scalpel. Samples were critical-point dried using liquid CO2, shadowed with gold-palladium and observed at the Centre Commun de Microscopie Electronique of the University of Nice Sophia-Antipolis on a JEOL T300 scanning electron microscope operating at 15 kV. For transmission electron microscopy, samples were fixed with 2.5% glutaraldehyde in modified 0.18 M Sorensen buffer pH 7.4 (containing 0.6 M sucrose, 4.2 mM CaCl2, and 1.25 g l)1 Ruthenium Red) for 75 min at room temperature, followed by 75 min at 4C. They were postfixed for 1 h in 1% OsO4 in ice-cold Sorensen buffer pH 7.4 without Ruthenium Red, dehydrated in ascending concentrations of ethanol, substituted with propylene oxide, and embedded in Epon resin. Thin sections were made with a diamond knife on a Leica microtome, post-stained with uranyl acetate and lead citrate, and viewed at 120 kV on a Philips 420 transmission electron microscope at the Integrated Imaging Center of Johns Hopkins University. Coral cell–fungal co-cultures Ciliated coral multicellular tissue isolates, averaging 600 lm in diameter (described by Kopecky and Ostrander 1999) were sampled in 2- to 4-day-old explant primary cultures. One batch of tissue
586 isolates, originating from the same explant primary cultures (initiated simultaneously from one of the five parent coral colonies) was used for each independent experiment. For co-cultures, 8–15 isolates were co-cultured with a 15–25 mm3 agar section containing fungal hyphae, in 5 ml cell culture medium without antibiotic– antimycotic solution (CCM), in individual wells of a six-well plate (volume-to-surface ratio of 0.5). Each well was considered one independent experiment. Control monocultures were carried out simultaneously with 8–15 isolates cultured in CCM with control plain agar sections without hyphae. Each independent experiment was repeated three to eight times, with batches of tissue isolates derived from distinct coral parent colonies. A total of 50–70 isolates were plated for each condition and control. Data from replicate independent experiments were analyzed with a non-paired Students t-test (Statview 5.0 software). Maintenance in suspension of ciliated isolates was checked versus breakdown into single cells. Isolate number, size (diameter), and ciliary activity were recorded over 1 week with a calibrated binocular (Wild M3Z) to assess the effects of exposure to fungi on coral tissue survival. Results were expressed as a percentage of surviving isolates. For quantification of cell viability, isolates were sampled in 3-day co-cultures with hyphal fungal agar sections, mechanically dispersed into single cells, rinsed once in filtered seawater, and stained with a complementary set of cell death markers. Cells were incubated for 15 min in the dark in 500 ll of ASW containing 1/ 100 Annexin V Alexa Fluor 568 conjugate (Molecular Probes A-13202, 568 nm excitation, 600 nm emission; Molecular Probes, Leiden, The Netherlands) and 50 nM Sytox Green (Molecular Probes S-7020, 488 nm excitation, 530 nm emission). Incorporation of these markers detected early and late plasma membrane modifications of the dying cell: Annexin V specifically stains the surface of apoptotic cells as it labels phosphatidylserine translocated and exposed at the external face of the plasma membrane in early apoptotic events. Sytox Green is a non-specific stain of dead cells as it labels DNA, which is accessible once the plasma membrane has become permeable. Cells were rinsed once in ASW and observed with an epifluorescence microscope (Leica DM-IRBE). For each point, a total of 300–400 cells were counted and pooled from two or more random optical fields, and the proportions of zooxanthellae, vacuolized cells, cells positive for Annexin V, and cells positive for Sytox Green in the total cell population were determined. Vacuolized cells contained large vacuoles, occupying up to two-thirds of the intracellular space, and which acidic content was determined by positive staining with acridine orange (DomartCoulon and Tambutte´, unpublished data). Results from three to five replicate independent experiments were analyzed for statistically significant differences (P
