Experimentally induced endosymbiont loss and re-acquirement in the hydrothermal vent bivalve Bathymodiolus azoricus

Journal of Experimental Marine Biology and Ecology 318 (2005) 99 – 110 www.elsevier.com/locate/jembe

Experimentally induced endosymbiont loss and re-acquirement in the hydrothermal vent bivalve Bathymodiolus azoricus Enikf Ka´da´r a,T, Raul Bettencourt a,b, Valentina Costaa, Ricardo Serra˜o Santosa, Alexandre Lobo-da-Cunhac, Paul Dandod a

IMAR Centre of the University of Azores, Department of Oceanography and Fisheries, Rua Cais de Santa Cruz, 9900 Horta, Portugal b Medical School, University of Massachussets, Worcester, MA 01605, USA c Laboratory of Cell Biology, Institute of Biomedical Sciences Abel Salazar, University of Porto, and CIIMAR, Portugal d School of Ocean Sciences, University of Wales-Bangor Anglesey, LL59 5AB, United Kingdom Received 29 November 2004; received in revised form 2 December 2004; accepted 12 December 2004

Abstract Invertebrates harbouring endosymbiotic chemoautotroph bacteria are widely distributed in a variety of reducing marine habitats, including deep-sea hydrothermal vents. In these species mechanisms of symbiont transmission are likely to be key elements of dispersal strategies that remained partially unresolved because the early life stages are not available for developmental studies. To study cessation and re-establishment of symbiosis in the host gill a laboratory experiment was conducted over 45 days in a controlled set-up (LabHorta) that endeavour re-creation of the hydrothermal vent chemical environment. Our animal model was the vent bivalve Bathymodiolus azoricus from the Menez Gwen vent site of the Mid Atlantic Ridge (MAR). Animals were exposed to conditions lacking inorganic S supply for 30 days, which is vital for their symbionts, and then re-acclimatized in sulphide-supplied seawater for an additional 15 days. Gradual disappearance of bacteria from the symbiont-bearing gill cells was observed in animals kept in seawater free of dissolved sulphide for up to 30 days, and was evidenced by histological, ultrastructural observations and Polymerase Chain Reaction tests. Following re-acclimatisation in S-supplied seawater, proliferation of sulphur-bacteria in the gill bacteriocytes confirms the functionality of our sulfide-feeding system in supporting chemoautotrophic symbionts. It may also indicate a horizontal endosymbiont acquisition, i.e. from the environment to the host by means of phagocytosis-like mechanism involving special bpit-likeQ structures on the apical cell membrane. The present work reports the first laboratory set-up successfully used to maintain the hydrothermal vent bivalve B. azoricus for prolonged periods of time by supplying inorganic sulphur as an energy source for its bacterial endosymbionts. Survival of symbiont bacteria is a critical factor influencing the host physiology and thus the methods

* Corresponding author. Tel.: +351 292 200 417; fax: +351 292 200 400. E-mail address: [email protected] (E. Ka´da´r). 0022-0981/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jembe.2004.12.025

100

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

reported here represent great potential for future studies of host-symbiont dynamics and for post-capture experimental investigations. D 2004 Elsevier B.V. All rights reserved. Keywords: Hydrothermal vent; Bivalves; Sulfur-oxidizer bacteria; Endosymbiont; Bathymodiolus azoricus

1. Introduction Deep-sea hydrothermal vent ecosystems are functionally dependent on the association between macroorganisms and their symbiotic chemolithoautotrophic bacteria, the presence of reduced chemicals from the vent effluent and the oxygen in seawater (Fisher et al., 1989). These conditions are fulfilled in the areas where the mixing of hydrothermal fluid and seawater allows the availability of both reduced (sulphide, methane) and oxidized (oxygen, nitrate) compounds (Sarradin et al., 1999). Bacterial chemosynthesis utilises the energy obtained from the oxidation of the reduced chemicals from hydrothermal fluid for the fixation of CO2 required for primary production (Yamanaka et al., 2003). Bathymodiolus azoricus is the dominant species of many of the hydrothermal vents of the midAtlantic Ridge (MAR), but the genus has worldwide distribution. Thirteen species of the genus Bathymodiolus have been described to date in both the Pacific and Atlantic Oceans (Von Cosel and Olu, 1998; Von Cosel et al., 1999; Gustafson et al., 1998). Bathymodiolus spp. exhibits a mixotrophic existence being capable to obtain carbon and other nutrients from several sources. Bathymodiolus thermophilus from the Pacific ingests particulate organic matter (Page et al., 1991) and also has endosymbiotic chemoautotrophic bacteria housed within the gills (Fiala-Medioni et al., 1986; Fiala-Medioni et al., 1994; Fiala-Medioni et al., 2002; Fisher et al., 1988). B. azoricus hosts two metabolically distinct (methanotrophic and thiotrophic) procaryotic endosymbionts (Pond et al., 1998). According to FialaMedioni et al. (1986), Bathymodiolus sp. can also absorb and incorporate free amino acids. This mixotrophy allows the bivalve to have a broad distribution relative to the effluent compared to other vent species, being able to supplement its metabolic needs by heterotrophy (Fisher et al., 1989). Moreover, larvae of these species appear to be capable of

dispersing hundreds of kilometres along a continuous ridge system and across spreading centres presenting a larval stage with feeding that might facilitate dispersal between ephemeral vent habitats, i.e. nonsymbiotic (Craddock et al., 1995). B. azoricus is found at different depths in hydrothermal vents along the MAR: Menez Gwen (37835VN– 388N, 840–870 m depth) and Lucky Strike (37800VN– 37835VN, 1700 m depth), Rainbow (36814VN, 2300 m depth) and at Broken Spur (29810VN, 3000 m depth) (Southward et al., 2001). Menez Gwen (MG) was the source of our experimental specimens, because it is the shallowest site on the Azores Triple Junction (ATJ) of the MAR, and thus mussels survive without the specialised pressure gear. It was discovered on the DIVA 1 cruise in May 1994 (Von Cosel et al., 1999). Patches of mussels at MG range between densities of 400–700 individuals m 2 clustering in a dilute hydrothermal medium (88–100% seawater, according to Sarradin et al., 1999) with an average of 8.2–8.6 8C and pH of b6.2–8.0 (in situ measurements during mussel collection). Total hydrogen sulphide concentrations have been reported as ranging between 0 and 60 AM l 1 (Sarradin et al., 1999). Earlier ultrastructural investigations on the gill of B. thermophilus (Fiala-Medioni et al., 1986; Le Pennec and Bejaoui, 2001; Dubilier et al., 1998) revealed that the filaments are mainly composed of bacteriocytes with a large number of intracellular bacteria, except for two small, frontal and distal areas that remained ciliated. However, the presence of a functional feeding groove and the labial palps, as well as studies of gut contents (Le Pennec and Bejaoui, 2001) confirm the ability of this bivalve to capture particulate organic matter and thus having a nonexclusive dependence on its endosymbiotic bacteria. Pond et al. (1998) concluded that methanotrophs and thiotrophs contribute equally to the nutrition of B. azoricus from Menez Gwen. The preponderance of the autotrophic vs. heterotrophic nutritional processes to total nutrition is suggested to be dependent on the

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

ecological conditions to which the organisms are subjected. Consequently, we predict that given the functional plasticity of the bathimodiolid ctenidium, changes in nutrient supply would induce changes in the gill, which in turn would give clues on how endosymbiont bacteria are incorporated into the host. Therefore, we set out to elucidate putative ultrastructural changes in the gill of B. azoricus maintained in seawater unsupplied with inorganic S and methane during a period of 30 days after which animal were re-acclimatized to conditions similar to the vent environment. In order to address adaptational strategies involved in B. azoricus remarkable feeding flexibility, and also to discriminate specific responses to nutrient supply variations, we have developed a simplified system in which sulphide was not supplied. Hydrogen sulphide (H2S) is generally believed to be the energy source for the establishment of sulphur oxidizing symbiotic communities (Arndt et al., 2001) and thus it was regarded as the limiting factor in our experiment. In this context, cytomorphological changes caused by nutrient alterations are described. The functional laboratory set-up described here was valuable in providing new insights into maintenance strategies enabling the study of B. azoricus for prolonged periods essential for in depth postcapture experimental investigations.

2. Materials and methods 2.1. Animal collection, maintenance and experimental conditions Mussels were collected at an average depth of 850 m from Menez Gwen vent site on the MAR, using the Remote Operated Vehicle (ROV), Victor 6000, aboard the R/V, L’Atalante, during the SEAHMA I cruise (29th July–14th August, 2002). The mussels were immediately transferred to fresh, cooled seawater held in plastic cool boxes during shipment and then were relocated into 40-l volume tanks housed within a refrigerated unit (ambient temperature 8–11 8C, water temperature 8.5 8C) at atmospheric pressure. Each tank contained 30 animals and was maintained as a closed system. Seawater for the aquaria was supplied from a reservoir containing sand-filtered seawater from an unpolluted bay in Horta,

101

Azores (38.58N 28.78W). Continuous semi-dark conditions were maintained throughout the experiment. Animals were kept in sulphide-free seawater for 30 days, followed by an additional 15-day exposure to sulphide-supplied seawater. Meanwhile, this sulphidesupplied tank housed 30 mussels already introduced within 24 h upon collection, to serve as experimental controls and potential symbiont infection source. Aeration using ordinary aquarium air diffusers enabled a water oxygen level below 50% of saturation. Sulphide was supplied from a 20 mM stock solution prepared as 5-l volume batches using commercially available Na2S crystals dissolved in filtered seawater (0.45 Am pore size cellulose acetate membrane, MilliporeR) which was then adjusted to pH 8.8–9.2 with 0.5 N HCl. The stock solution was dispensed using a Masterflex pump (model 77120-62) at a rate of 2 ml min 1, for 15 min, every 2 h to the seawater tanks, through a diffusion tube at the base of the tank. The average tank-concentration of sulphide was 15 AM (cycle range). The water in the aquaria was changed every 7–10 days when the animals were transferred to a new tank (previously supplied with sulphide in order to minimize stress-induced reactions) for removal of pseudo-faeces and other organic matter that caused oxygen depletion, and thus, excessive H2S build up. Total dissolved sulphide present in a mixture of H2S+HS +S2 was measured twice a day by colorimetry according to Cline (1989). Water temperature, pH and dissolved oxygen concentrations were also measured daily. The following groups of five animals were analysed: group 0, animals freshly dissected immediately after collection; group 1, mussels quickly transferred to sulphide-supplied seawater upon collection and dissected after 15-day exposure; group 2, mussels maintained in sulphidefree seawater since initial collection and for 15 days; group 3, mussels maintained in sulphide-free seawater since initial collection and for 30 days; group 4, mussels from group 3 re-acclimatized to sulphide supplied seawater. 2.2. Sulphide monitoring Two-milliliter syringes were flushed with pressurised nitrogen several times before sampling. A sample of 1.6-ml water was taken from above the

102

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

2.4. DNA extraction and PCR amplification

60 50 40 30 20 10 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Time (hours) Fig. 1. Concentration of dissolved sulphide in the water column monitored. Measurements were made before and after the 15-min pumping periods (indicated by arrows) every 2 h.

densest clump of mussels, 25 cm below the surface. Immediately afterward, 0.4 ml of Diamine reagent (N,N-dimethyl-1, 4-phenylendiammoniumdichlorid, MERCK) was drawn up into the syringe and mixed. The samples were left to react for 15 min and then read in a spectrophotometer at 670 nm. Standards between 0 and 100 AM were made by diluting immediately from a 10 mM stock solution. Degassed seawater was used to dilute the standards and make up the stock solution (Cline, 1989). 2.3. Tissue preparation for light and electron microscopy Small (1 mm3) tissue pieces were fixed in modified Trump’s fixative (3% glutaraldehyde and 3% paraformaldehyde made up with a fixation buffer containing: 0.15 M Na-cachodylate, 0.3 M sucrose, 0.2 M NaCl and 0.008 M CaCl according to Distel and Felbeck, 1987). Following primary fixation, samples were washed in 0.1 M cacodylate buffer (pH 7.8), post-fixed in 1% osmium tetroxide in cacodylate buffer for 1 h, dehydrated in ethanol and embedded in Spurr resin (Sigma). Semi-thin (2 Am) sections were obtained using diamond knife on a LKB-BROMMA ultramicrotome and stained with methylene blue. Ultra-thin sections were mounted on copper grids and were double stained with uranyl acetate and lead citrate. Five individuals from each experimental group were investigated (3 blocks per individual) and observations were made on filaments detached from the mid portion of the external demibranchs.

Nucleic acids from whole gill tissues were extracted according to Sambrook et al. (1989) with slight modifications. In brief, 100–150 mg of tissue was homogenized with 400 Al STE buffer and resulting homogenate treated with 20 Al of proteinase K and 40 Al of 10% SDS solution for 2 h at 55 8C. Two consecutive phenol/chloroform/isoamyl alcohol (25:24:1; SIGMA, Molecular Biology Grade) extractions were performed followed by DNA precipitation with 45 Al of 5 M NH4OAC mixed to the aqueous phase. Subsequently, purified DNA was used as template for PCR amplification. 1 Al of Vent polymerase (New England Biolabs) was used in 50 Al total volume reactions including 20 pM of each sense and antisense primers, dNTPs at 0.25 mM and 1reaction buffer, following manufacturer’s instructions. Thermo-cycling conditions were performed according to standard conditions. Briefly, an initial denaturation step at 94 8C for 2 min (bhot startQ) was followed by 35 cycles of denaturation at 94 8C for 1 min, annealing at 50 8C for 1 min and extension 68 8C, followed by a 7-min final extension at 72 8C. Primers design was based on nucleotide sequences available from the NCBI public database for the following genes with GenBank accession nos. AB073122 and AB178052 respectively: Bathymodiolus endosymbiont gene for 16S rRNA (sense 5VAGAGTTTGTTCATGGCTCAGA3Vand antisense

60 50 40 30 20 10 0 1 2

3

4

5

6

7

8

9 10 11 12 13 14 15

Time (days) Fig. 2. Concentration of dissolved sulphide in the water column showing an excessive build up on day 12 followed by water change. Values represent averageFSEM of three samples taken at different depths above the mussel clump from the experimental tanks.

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

5VGAAGGCACCAATCCATCTCTG3V); Bathymodiolus endosymbiont ATP sulpfurylase gene (sense 5VGCTTTTCAGACCCGCAACCCC3V and antisense 5VCTTGGTGCCGGAGAGCA GTAC3V). PCR products were examined on 0.8% agarose gel electrophoresis according to standard protocol.

103

3. Results 3.1. Sulphide levels in the water column Sulphide concentration reached a dynamic equilibrium at about 40 AM in approximately 12 h (Fig. 1)

Plate 1. Macroscopic and histological changes in B. azoricus exposed to sulphur-free seawater for 30 days followed by re-acclimatization to seawater containing about 10 AM dissolved sulphide. (a) Gross appearance of ctenidium in a freshly collected animal (from group 0) vs. (b) a symbiont starved mussel (group 3). Note colour change from brown to white. Scale bars: 1 cm. (c) Light micrograph of the gill filaments from group 0 mussels. Gill filaments are divided into 3 functionally different zones: ciliated frontal zone (CZ), transitory (TZ) and bacteriocyte zones (BZ). Scale bar 50 Am. (d) Bacteriocytes (B) in the gill of group 0 mussels. Scale bar 10 Am. (e) Gill filaments from an animal kept in H2S-free seawater for 15 days (group 2), containing several amoebocytes (arrows) within the lumen. The thickness of the bacteriocyte layer is substantially reduced. Scale bar: 50 Am. (f) Very thin bacteriocytes and an increase in the number and size of amoebocytes (arrows) can be observed in gills of group 3 mussels. Scale bar: 10 Am. (g) Gill filaments from group 4 mussels (i.e. reacclimatize to H2S-supplied seawater for 15 days following the 30-day sulphide free treatment). Scale bar: 50 Am. (h) Bacteriocyte zone of a re-acclimatize animal. The bacteriocytes are slightly thicker than in animals kept in sulphur-free seawater, but gill filaments have not regained their normal thickness. Scale bar: 10 Am.

104

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

Plate 2. Electron micrographs of gill filaments of B. azoricus dissected freshly upon collection (group 0). (a) Bacteriocytes contain both types of symbionts at the apical region; the smaller rod-shaped ones are sulfur-oxidizer bacteria (Sb) and the larger are methanotrophic bacteria (Mb). The central nucleus (N) and several large lysosomes (L) with membranous content are also visible. Scale bar: 2 Am. (b) Symbiont bearing vacuoles with non-mixing bacterial content. A double cell membrane (Gram-) and a central clear zone with DNA strands are visible in the sulfur-oxidizer bacteria (Sb), and rich membranous content characterize the larger oval shaped methanotrophic bacteria (Mb). Scale bar: 0.5 Am.

when pumped at 2 ml min 1, for 15 min every 2 h in tanks with seawater. In the presence of animals, levels were around 10 AM, whereas anaerobic conditions, possibly due to proliferation of free living bacteria, produced elevated levels of dissolved sulphide in the water column on day 12, as shown in Fig. 2. Noteworthy is the full water change to avoid buildup of excessive sulphide levels incurring significant mussel loss. 3.2. Morphological and histological changes resulted from exposure of mussels to sulphide-free seawater followed by re-acclimatization Freshly collected B. azoricus exhibited brown gills with thick demibranchs due to the proliferation of

symbionts in the host tissue (Plate 1a). Thirty days after being transferred to sulphur- and CH4-free seawater, mussels appeared with significant changes in colour and thickness of their ctenidium. Gills have become white and feeble as compared to their natural aspect (Plate 1b). Sections of filaments from gill lamellae, revealed a single-layered wall around the central lumen, and exhibited three typical zones that were previously defined for other symbiont-bearing bivalves (Distel and Felbeck, 1987; Fiala-Medioni et al., 1986). The three typical zones, shown in Plate 1c are the frontal ciliated zone, the transitional zone of non-ciliated, non-pigmented and bacteria-free cells and the bacteriocyte zone composed of cells hosting two types of symbiotic bacteria. These cells constituted the major part of the filament (Plate 1d).

Plate 3. Electron micrographs of bacteriocytes from B. azoricus gills under different treatments. (a) and (b) Group 1 mussels, placed in sulphidesupplied seawater within 24 h of collection, methanotrophs are no longer present but several empty vacuoles (asterisks) and sulfur-oxidizer bacteria (Sb) are observed in bacteriocytes. Scale bar: 2 Am. (b) Sulfur-oxidizer bacteria (Sb) do not show morphological alterations. Scale bar: 1Am. (c) and (d) Bacteriocytes from animals kept in H2S-free seawater for 30 days (group 3) have a spongy appearance, containing symbiontfree vacuoles (*) and large size lysosomes (L). Scale bars: 0.5 Am and 0.3 Am, respectively. (e) Pit-like structures (arrows) are evident on the apical cell surface of bacteriocytes from animals re-acclimatized in seawater containing about 10 AM dissolved sulphide for 15 days (group 4). Copious amounts of sulfur-oxidizer bacteria (Sb) fill the apical zone of the bacteriocytes. Scale bar: 1 Am. (f) Sulfur-oxidizer bacteria (Sb) could be observed in close association with the pit-like structures on the surface of bacteriocytes in re-acclimatized mussels. Scale bar: 0.2 Am.

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

105

106

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

Mussels kept in sulphide-free seawater revealed gill filaments that, although exhibiting the single cell layer intact, were much slimmer in the bacteriocyte zone and presented a more vacuolated cytoplasm (Plate 1e–f). An increase in both the number and the size of amoebocytes from the central lumen of filaments was also observed (Plate 1e–f). Even though the gill filaments and the bacteriocytes did not recover to their original aspect by day 15 of re-acclimatization to sulphide-supplied seawater (Plate 1g, h), many bacteriocyte cells contain copious number of bacteria. However, light micrographs revealed that the filaments remained shrunken and the incidence of binfectedQ cells seemed lower as compared to freshly collected animals (Plate 1h). 3.3. Cytomorphological changes in the bacteriocytes of symbiont-starved and re-acclimatized bivalves Ultrastructural observations on the bacteriocytes from freshly collected animals revealed that the major volume of cells was composed of symbionts that occupied the apical region (Plate 2a). These vacuoles contained two distinctive types of bacteria: the smaller, rod-shaped and more abundant, sulphur oxidizers, and the larger, oval shaped, methanotrophic bacteria with rich membranous content, probably, type I methanotrophs (Plate 2b). The two types of bacteria did not show signs

of aggregation within the same vesicle. Despite of the size and shape modulations (due to sectioning angle), all sulphur oxidizers shared similar ultrastructural features, i.e. double membranes (gram negative) with DNA strands found in the centre of an electron-translucent area (Plate 2b). The larger methanotrophs also had double membrane and often presented multiple electron-translucent areas with DNA strands and cytoplasmatic membranous material (Plate 2b). Mussels from group 1 (experimental control) lost their methane-oxidizing symbionts, but not the sulphur ones (Plate 3 a,b). Ultrastructural changes that occurred as a result of keeping animals in sulphide-free seawater (groups 3 and 4) were most obvious in bacteriocytes, where the apical zone appeared spongy due to the loss of bacteria from vesicles (Plate 3 c,d). The general morphological changes observed on bacteriocytes were an increased incidence of lysosomes, and also, an increase in their size as compared to those in animals from group 0. The simultaneous shrinkage of the bacteriocytes confers an appearance of the cell as if almost its entire volume was occupied by lysosomes (Plate 3 c). In addition, the lysosomal content appeared as in a more advanced degradation stage with unfolded and more heterogeneous membranous material. In mussels from group 4 bacteriocytes resembled those of experimental controls (group 1) in that the

Fig. 3. PCR detection of B. azoricus endosymbiont bacteria in total genomic DNA preparations from gill tissues. The presence of both 16S rRNA (primer set 1) and ATP-sulfurilase (primer set 2) genes was successfully detected in samples from mussels dissected upon collection (group 0, lanes 1 and 2 respectively). Mussels kept in seawater for 30 days failed to reveal the presence of both genes and consequently the presence of endosymbiont (group 3, lanes 3 and 4). Upon re-acclimatization to sulphide-supplied seawater (group 4) the presence of 16S rRNA (lane 5) and ATP sulfurilase (lane 6) is again detected suggesting inter-animal endosymbiont transmission. Expected PCR amplifications products are shown (arrows) as well as 1 Kb DNA ladder.

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

sulfur-oxidizers were abundant and the larger methane oxidizers were absent (compare Plate 3 a,b with e,f). The apical membrane of bacteriocytes displayed several bpit-likeQ structures (Plate 3, arrows) that seemingly allowed the entry of sulphur symbiont bacteria (Plate 3f). 3.4. PCR amplifications PCR reactions performed with two specific set of primers (set 1 and set 2) generated DNA fragments of the expected size as indicated by agarose gel electrophoresis, demonstrating the presence of two bacterial endosymbiont genes from Bathymodiolus sp. Primer set 1, specifically designed to detect the ribosomal gene 16S rRNA from endosymbiont of Bathymodiolus sp. was successful in PCR tests aimed at the detection of this gene in DNA extracts from gills homogenates from mussel group 0 (Fig. 3, lane 1). The specificity of our PCR-based endosymbiont detection method was further demonstrated with a second primer set specifically designed to detect the Bathymodiolus bacterial endosymbiont ATP sulfurylase gene of deep-sea hydrothermal vent mussels from the Japanese Suiyo Seamount (Fig. 3, lane 2). DNA extracts originated from mussel group 3, failed to reveal the presence of both 16S rRNA and ATP sulfurylase genes (Fig. 3 lanes 3 and 4 respectively). In contrast, gill DNA of mussels from group 4 revealed the presence of endosymbiont (Fig. 3, lanes 5 and 6). While primer set 2 oligonucleotide sequence is based on a gene sequence from a different Bathymodiolus species, our results indicated that not only this gene is detected in DNA preparations of B. azoricus gills (Fig. 3 lane 2), but also confirms the reappearance of sulphur-oxidizing endosymbiont bacteria in mussels from group 4 (Fig. 3 lane 6). Our results therefore, indicate that both primer sets can successfully amplify the bacterial symbiont DNA target of genomic DNA extracted from symbiontcontaining gill tissues of B. azoricus.

4. Discussion and conclusions The flexible feeding regime of B. azoricus based on both mixotrophy (filter feeding and symbiosis) and

107

a dual symbiosis (methanotrophic and thiotrophic) (Pond et al., 1998) enables the mussel to tolerate wide temporal and spatial variability of the environmental factors typical to hydrothermal vents (Johnson et al., 1994). Thus we have selected the bivalve as a model organism to investigate the mechanisms underlying nutritional responses in relation to environmental variations under controlled laboratory conditions. The laboratory set-up that was developed in LabHorta, and based on a sulphide feeding system was successful in maintaining endosymbiosis in the host vent bivalve. This opens the possibility to use endosymbiont bacteria, otherwise unculturable under laboratory conditions, as a new experimental tool in vent research. The results obtained on the cyclic sulphide supply into the experimental aquaria are in agreement with the previously reported 1-h half-life of dissolved sulphide at neutral pH in aerobic seawater (Almgren and Hagstrom, 1974). This regime of pumping a stock Na2S solution on a 2-hourly basis permitted the maintenance of a stable average level of 10 AM sulphide inside the aquaria that is within the range of 0.5–18 for the AS reported for mussel beds in Menez Gwen (Sarradin et al., 1999). Prevention of H2S depletion in experimental tanks was insured by an intermittent supply system essential for maintenance of sulphur oxidizer symbiont bacteria reported sensitive to low levels (Childress et al., 1991). Provided that mussels are placed in seawater supplied with dissolved sulphide within 24 h of collection, they survive over several months (data not shown). Oxygen saturation is another major factor influencing animal maintenance. Under anoxic conditions anaerobic sulphide production takes place, as reported by Arndt et al. (2001) for several sulphur-storing symbioses, and also confirmed in the present study (see day 12 in Fig. 2). Unless fresh water change is ensured, massive mortality occurs in the experimental tanks. However, oxygen levels should not exceed 30% of saturation, as oxidation reactions would deplete H2S from the system. Thus, for a successful long-term experimental set-up, rigorously controlled conditions are obtained by an enduring re-adjustment of the three major interacting factors, pH, oxygen and sulphide concentration. Continuous monitoring of these factors is essential for long-term maintenance of B. azoricus in captivity permitting specific inves-

108

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

tigations to be conducted with a great advantage over the costly and both time- and human-resourceconsuming in situ observations. Behavioural observations as well as toxicological investigations will be possible to be performed under controlled laboratory conditions with the prospect of understanding the mechanisms that enable this species to survive under the extreme chemical and physical conditions typical to hydrothermal vents. Ultrastructural and molecular evidence is presented for the substantial loss of both bacterial symbionts in bacteriocytes of mussels kept in sulphide-free seawater for up to 30 days. Residual endosymbiont bacteria might have survived the sulfide-deprivation treatment, however, our PCR detection system suggests that the endosymbiont bacterial population can vary from one experimental condition to another and this population particularly increases when sulfidedeprived mussels are exposed to an environment containing sulfide and other untreated mussel individuals. This, in it self, suggests the existence of a lateral acquisition of symbionts that does not clearly excludes a vertical transmission. Additionally, PCR has been a widely used method in deep-sea environmental microbiology as a molecular tool in the detection of prokaryotes (Imhoff et al., 2003; Gros et al., 1996, 1998a, 2003; Won et al., 2003). In spite of this lack of endosymbiont detection, it is possible however, that a few dormant cells in the form of endospore may have remained and started dividing as soon as sulphide was supplied. If so, these endospores must be resistant to such an extent that it would withstand standard genomic DNA extraction protocols. A more likely alternative is the re-infection of these mussels by means of a passive contamination through bacteria released in the media by control animals kept in the same sulphide-supplied tank. Additionally, when in a separate experiment, symbiont-starved mussels were transferred to sulphide supplied aquarium without other mussels, did no show signs of bacterial re-infection (data not shown). Control animals (group 1) did not loose their sulphur endosymbionts, as demonstrated in our studies, as opposed to a deficiency in methanotrophs as was expected from the experimental conditions (no methane supply). Recovering endosymbiont bacteria from the autoclaved seawater used to rinse gills prior to bacterial extraction brings about indirect evidence for

their spontaneous leaking from bacteriocytes. Additionally, there is evidence originated from studies by Gros and co-workers on environmental transmission of sulfur-oxidizer bacteria in various symbiont-harbouring bivalves (Gros et al., 2003, 1996, 1998a, 1999). These authors have determined the transmission mechanisms by experimental colonisation of aposymbiotic Codakia larvae and proposed the bpitlikeQ structures of the membrane as a route for bacterial access into the bacteriocyte (Gros et al., 1998b). Such structures were consistently observed in our ultra structure studies of membrane invaginations on bacteriocytes found in re-acclimatized mussels’ gills (Plate 3e–f). A third alternative source of infection, although unlikely, may be the invasion of the gills by free-living sulphur oxidizing bacteria that may have proliferated in the sulphide-supplied seawater. There is evidence for the existence of freeliving sulphur oxidizer bacteria becoming symbiotic when entering the water current within the bivalve gill (Gros et al., 1999). The facultative symbiotic nature of these sulfur-oxidizers is also inferred in Nelson and co-workers studies (Nelson et al., 1995) taking in consideration the overlapping symbiont’s molar growth yield on thiosulfate, with that of free-living chemoautotrophs. However, 16S rDNA-based phylogenetic studies on sulphur-oxidizing bacterial endosymbionts indicated that they are clearly distinct from free-living sulphur-oxidizing bacteria of the genera Beggiatoa, Halothiobacillus and Thiomicrospira (Imhoff et al., 2003). Our PCR-based approach to detect the presence of endosymbionts of B. azoricus was designed in a way such, only specific endosymbiont target genes would be amplified, and thus ruling out the possibility of contaminations from freeliving sulfur-oxidizer bacteria (Fig. 3). In light of our results, we conclude that (a) B. azoricus is able to survive in the absence of sulphide and resulting reduction of its endosymbiont population for extended period of time; (b) endosymbiosis can be experimentally manipulated in that once ceased (nutrient deprivation) may be regained provided that adequate chemical conditions in the environment are met; (c) horizontal endosymbiont transmission is possible via inter-animal contamination. Recent field experiments conducted by Raulfs and co-workers (Raulfs et al., 2004) in which B. termophylus were transplanted away from venting exits at hydrothermal site on the

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

southern East Pacific Rise reached similar conclusions regarding symbiont loss and consequent ultrastructural modifications, and thus corroborate our experimental results. By developing a successful laboratory set-up for the maintenance of B. azoricus that enables preservation and manipulation of endosymbiosis, we provide a basis for a more elaborated ecophysiological research, in order to understand the general principles that govern adaptations to the hydrothermal environment.

Acknowledgements The research was undertaken under the scope of the research project SEAHMA (Seafloor and subseafloor hydrothermal modelling in the Azores Sea) funded by FCT (PDCTM/P/MAR/15281/1999). Postdoctoral fellowship support was jointly offered by the Portuguese Science Foundation (FCT) and by IMAR, Portugal to Eniko Kadar (IMAR/FCT-PDOC-012/ 2001-EcoToxi). The authors acknowledge the ROV team and the Atalante crew for their contribution in sampling and the technical team that helped running LabHorta, the laboratory set-up for the animal maintenance. We are also indebted to Sergio Stefanni for experimental advice and to Tony Ip for his help in providing primers and allowing the use of his lab facilities at the University of Massachusetts. The experiments carried out for this study comply with the current pertinent laws in Portugal. [SS]

References Almgren, T., Hagstrom, A., 1974. The oxidation rate of sulphide in sea water. Water Research 8, 395 – 400. Arndt, C., Gaill, F., Felbeck, H., 2001. Anaerobic sulphur metabolism in thiotrophic symbioses. Journal of Experimental Biology 204, 741 – 750. Childress, J.J., Fisher, C.R., Favuzzi, J.A., Sanders, N.K., 1991. Sulphide and carbon-dioxide uptake by the hydrothermal vent clam, Calyptogena magnifica, and its chemoautotrophic symbionts. Physiological Zoology 64, 1444 – 1470. Cline, J.D., 1989. Spectrophotometric determinations of hydrogen sulphide in natural waters. Limnology and Oceanography 14, 454 – 459.

109

Craddock, C., Hoeh, W.R., Lutz, R.A., Vrijenhoek, R.C., 1995. Extensive gene flow among Mytilid (Bathymodiolus thermophilus) populations from hydrothermal vents of the Eastern Pacific. Marine Biology 124, 137 – 146. Distel, D.L., Felbeck, H., 1987. Endosymbiosis in the lucinid clams Lucinoma aequizonata, Lucinoma annulata and Lucina floridana—a reexamination of the functional-morphology of the gills as bacteria-bearing organs. Marine Biology 96 (1), 79 – 86. Dubilier, N., Windoffer, R., Giere, O., 1998. Ultrastructure and stable carbon isotope composition of the hydrothermal vent mussels Bathymodiolus brevior and B. sp. Affinis brevior from the North Fiji Basin, western Pacific. Marine Ecology-Progress Series 165, 187 – 193. Fiala-Medioni, A., Metivier, C., Herry, A., Lepennec, M., 1986. Ultrastructure of the gill of the hydrothermal-vent Mytilid Bathymodiolus Sp. Marine Biology 92, 65 – 72. Fiala-Medioni, A., Michalski, J.C., Jolles, J., Alonso, C., Montreuil, J., 1994. Lysosomic and lysozyme activities in the gill of bivalves from deep hydrothermal vents. Comptes Rendus De L Academie Des Sciences Serie Iii-Sciences De La Vie-Life Sciences 317, 239 – 244. Fiala-Medioni, A., McKiness, Z.P., Dando, P., Boulegue, J., Mariotti, A., Alayse-Danet, A.M., Robinson, J.J., Cavanaugh, C.M., 2002. Ultrastructural, biochemical, and immunological characterization of two populations of the Mytilid mussel Bathymodiolus azoricus from the Mid-Atlantic Ridge: evidence for a dual symbiosis. Marine Biology 141, 1035 – 1043. Fisher, C.R., Childress, J.J., Arp, A.J., Brooks, J.M., Distel, D., Favuzzi, J.A., Felbeck, H., Hessler, R., Johnson, K.S., Kennicutt, M.C., Macko, S.A., Newton, A., Powell, M.A., Somero, G.N., Soto, T., 1988. Microhabitat variation in the hydrothermal vent mussel, Bathymodiolus-Thermophilus, at the rose garden vent on the Galapagos rift. Deep-Sea Research. Part A, Oceanographic Research Papers 35, 1769 – 1791. Fisher, C.R., Brooks, J.M., Childress, J.J., Felbeck, H., Hessler, R.R., Johnson, K.S., Macko, S.A., Moe, A., Nelson, D., Somero, G.N., 1989. Microhabitat requirements of the hydrothermal vent mussel, Bathymodiolus thermophilus. American Zoologist 29, A81. Gros, O., et al., 1996. Environmental transmission of a sulfuroxidizing bacterial gill endosymbiont in the tTropical lucinid bivalve Codakia orbicularis. Applied and Environmental Microbiology 62, 2324 – 2330. Gros, O., et al., 1998a. Putative environmental transmission of sulfur-oxidizing bacterial symbionts in tropical lucinid bivalves inhabiting various environments. FEMS Microbiology Letters 160, 257 – 262. Gros, O., Frenkiel, L., Moueza, M., 1998b. Gill filament differentiation and experimental colonization by symbiotic bacteria in aposymbiotic juveniles of Codakia Orbicularis (Bivalvia: Lucinidae). Invertebrate Reproduction and Development 34, 219 – 231. Gros, O., Duplessis, M.R., Felbeck, H., 1999. Embryonic development and endosymbiont transmission mode in the symbiotic clam Lucinoma aequizonata (Bivalvia : Lucinidae). Invertebrate Reproduction and Development 36, 93 – 103.

110

E. Ka´da´r et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110

Gros, O., et al., 2003. Detection of the free-living forms of sulfideoxidizing gill endosymbionts in the lucinid habitat (Thalassia testudinum environment). Applied and Environmental Microbiology 69, 6264 – 6267. Gustafson, R.G., Turner, R.D., Lutz, R.A., Vrijenhoek, R.C., 1998. A new genus and five new species of mussels (Bivalvia, Mytilidae) from deep-sea sulphide/hydrocarbon seeps in the Gulf of Mexico. Malacologia 40, 63 – 112. Imhoff, J.F., Sahling, H., Suling, J., Kath, T., 2003. 16S rDNAbased phylogeny of sulphur-oxidising bacterial endosymbionts in marine bivalves from cold-seep habitats. Marine EcologyProgress Series 249, 39 – 51. Johnson, K.S., Childress, J.J., Beehler, C.L., Sakamoto, C.M., 1994. Biogeochemistry of hydrothermal vent mussel communities— the deep-sea analog to the intertidal zone. Deep-Sea Research Part I-Oceanographic Research Papers 41, 993 – 1011. Le Pennec, M., Bejaoui, N.A., 2001. The conquest of reduced deep marine ecosystems by Mytilid mussels. Bulletin de la Societe Zoologique de France 126, 121 – 127. Nelson, D.C., Hagen, K.D., Edwards, D.B., 1995. The gill symbiont of the hydrothermal vent mussel Bathymodiolus-thermophilus is a psychrophilic, chemoautotrophic, sulphur bacterium. Marine Biology 121, 487 – 495. Page, H.M., Fialamedioni, A., Fisher, C.R., Childress, J.J., 1991. Experimental-evidence for filter-feeding by the hydrothermal vent mussel, Bathymodiolus-thermophilus. Deep-Sea Research Part A-Oceanographic Research Papers 38, 1455 – 1461. Pond, D.W., Bell, M.V., Dixon, D.R., Fallick, A.E., Segonzac, M., Sargent, J.R., 1998. Stable-carbon-isotope composition of fatty acids in hydrothermal vent mussels containing methanotrophic and thiotrophic bacterial endosymbionts. Applied and Environmental Microbiology 64, 370 – 375. Raulfs, E.C., Macko, S.A., Van Dover, C.L., 2004. Tissue and symbiont condition of mussels (Bathymodiolus thermophilus)

exposed to varying levels of hydrothermal activity. Journal of the Marine Biological Association of the United Kingdom 84 (1), 229 – 234. Sambrook, J., Fritsch, E.J., Maniatis, T., 1989. Molecular cloning: A laboratory manual. 2nd ed. Cold Spring Harbour Laboratory Press, Cold Spring Harbour, New York. Sarradin, P.M., Caprais, J.C., Riso, R., Kerouel, R., Aminot, A., 1999. Chemical environment of the hydrothermal mussel communities in the lucky strike and Menez Gwen vent fields, Mid-Atlantic Ridge. Cahiers De Biologie Marine 40, 93 – 104. Southward, E.C., Gebruk, A., Kennedy, H., Southward, A.J., Chevaldonne, P., 2001. Different energy sources for three symbiont-dependent bivalve molluscs at the Logatchev hydrothermal site (Mid-Atlantic Ridge). Journal of the Marine Biological Association of the United Kingdom 81, 655 – 661. Von Cosel, R., Olu, K., 1998. Gigantism in Mytilidae. A new Bathymodiolus from cold seep areas on the Barbados accretionary Prism. Comptes Rendus De L Academie Des Sciences. Serie Iii, Sciences De La Vie-Life Sciences 321, 655 – 663. Von Cosel, R., Comtet, T., Krylova, E.M., 1999. Bathymodiolus (Bivalvia : Mytilidae) from hydrothermal vents on the Azores triple junction and the Logatchev hydrothermal field, MidAtlantic Ridge. Veliger 42, 218 – 248. Won, Y.J., Hallam, S.J., O’Mullan, G.D., Pan, I.L., Buck, K.R., Vrijenhoek, R.C., 2003. Environmental acquisition of thiotrophic endosymbionts by deep-sea mussels of the genus Bathymodiolus. Applied and Environmental Microbiology 69, 6785 – 6792. Yamanaka, T., Mizota, C., Fujiwara, Y., Chiba, H., Hashimoto, J., Gamo, T., Okudaira, T., 2003. Sulphur-isotopic composition of the deep-sea mussel Bathymodiolus marisindicus from currently active hydrothermal vents in the Indian Ocean. Journal of the Marine Biological Association of the United Kingdom 83, 841 – 848.