www.elsevier.com/locate/jembe

Journal of Experimental Marine Biolo

Experimentally induced endosymbiont loss and re-acquirement in

the hydrothermal vent bivalve Bathymodiolus azoricus

Enikf Kadara,T, Raul Bettencourta,b, Valentina Costaa, Ricardo Serrao Santosa,

Alexandre Lobo-da-Cunhac, Paul Dandod

aIMAR Centre of the University of Azores, Department of Oceanography and Fisheries, Rua Cais de Santa Cruz, 9900 Horta, PortugalbMedical School, University of Massachussets, Worcester, MA 01605, USA

cLaboratory of Cell Biology, Institute of Biomedical Sciences Abel Salazar, University of Porto, and CIIMAR, PortugaldSchool 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

0022-0981/$ - s

doi:10.1016/j.jem

* Correspondi

E-mail addr

gy and Ecology 318 (2005) 99–110

ee front matter D 2004 Elsevier B.V. All rights reserved.

be.2004.12.025

ng author. Tel.: +351 292 200 417; fax: +351 292 200 400.

ess: Enikokadar@notes.horta.uac.pt (E. Kadar).

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110100

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 func-

tionally dependent on the association between macro-

organisms 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 mid-

Atlantic Ridge (MAR), but the genus has worldwide

distribution. Thirteen species of the genus Bathymo-

diolus 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 (meth-

anotrophic and thiotrophic) procaryotic endosym-

bionts (Pond et al., 1998). According to Fiala-

Medioni 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). More-

over, 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. non-

symbiotic (Craddock et al., 1995).

B. azoricus is found at different depths in hydro-

thermal vents along theMAR: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 hydro-

thermal 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 non-

exclusive 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. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110 101

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 ultra-

structural changes in the gill of B. azoricus main-

tained 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 main-

tenance strategies enabling the study of B. azoricus

for prolonged periods essential for in depth post-

capture 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,

Azores (38.58N 28.78W). Continuous semi-dark con-

ditions 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 sulphide-

supplied 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 satu-

ration. 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 color-

imetry 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 sulphide-

free seawater since initial collection and for 15 days;

group 3, mussels maintained in sulphide-free sea-

water 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 pressur-

ised nitrogen several times before sampling. A

sample of 1.6-ml water was taken from above the

0

10

20

30

40

50

60

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17Time (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.

0

10

20

30

40

50

60

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 differen

depths above the mussel clump from the experimental tanks.

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110102

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% paraf-

ormaldehyde made up with a fixation buffer contain-

ing: 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.

2.4. DNA extraction and PCR amplification

Nucleic acids from whole gill tissues were ex-

tracted 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 proteinaseK 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) extrac-

tions 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

1�reaction buffer, following manufacturer’s instruc-

tions. Thermo-cycling conditions were performed

according to standard conditions. Briefly, an initial

denaturation step at 94 8C for 2 min (bhot startQ) wasfollowed 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: Bathymodio-

lus endosymbiont gene for 16S rRNA (sense

5VAGAGTTTGTTCATGGCTCAGA3Vand antisense

.

t

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110 103

5VGAAGGCACCAATCCATCTCTG3V); Bathymo-

diolus endosymbiont ATP sulpfurylase gene (sense

5VGCTTTTCAGACCCGCAACCCC3V and antisense

5VCTTGGTGCCGGAGAGCA GTAC3V). PCR prod-

ucts were examined on 0.8% agarose gel electro-

phoresis according to standard protocol.

Plate 1. Macroscopic and histological changes in B. azoricus exposed to

seawater containing about 10 AM dissolved sulphide. (a) Gross appearance

symbiont starved mussel (group 3). Note colour change from brown to wh

group 0 mussels. Gill filaments are divided into 3 functionally different zon

(BZ). Scale bar 50 Am. (d) Bacteriocytes (B) in the gill of group 0 mussels.

seawater for 15 days (group 2), containing several amoebocytes (arrow

substantially reduced. Scale bar: 50 Am. (f) Very thin bacteriocytes and a

observed in gills of group 3 mussels. Scale bar: 10 Am. (g) Gill filaments fr

15 days following the 30-day sulphide free treatment). Scale bar: 50 Am. (h

slightly thicker than in animals kept in sulphur-free seawater, but gill fila

3. Results

3.1. Sulphide levels in the water column

Sulphide concentration reached a dynamic equili-

brium at about 40 AM in approximately 12 h (Fig. 1)

sulphur-free seawater for 30 days followed by re-acclimatization to

of ctenidium in a freshly collected animal (from group 0) vs. (b) a

ite. Scale bars: 1 cm. (c) Light micrograph of the gill filaments from

es: ciliated frontal zone (CZ), transitory (TZ) and bacteriocyte zones

Scale bar 10 Am. (e) Gill filaments from an animal kept in H2S-free

s) within the lumen. The thickness of the bacteriocyte layer is

n increase in the number and size of amoebocytes (arrows) can be

om group 4 mussels (i.e. reacclimatize to H2S-supplied seawater for

) Bacteriocyte zone of a re-acclimatize animal. The bacteriocytes are

ments have not regained their normal thickness. Scale bar: 10 Am.

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.

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110104

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 build-

up 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

Plate 3. Electron micrographs of bacteriocytes from B. azoricus gills under

supplied seawater within 24 h of collection, methanotrophs are no longe

bacteria (Sb) are observed in bacteriocytes. Scale bar: 2 Am. (b) Sulfur-oxi

1Am. (c) and (d) Bacteriocytes from animals kept in H2S-free seawater for

free vacuoles (*) and large size lysosomes (L). Scale bars: 0.5 Am and 0.3

apical cell surface of bacteriocytes from animals re-acclimatized in seawat

Copious amounts of sulfur-oxidizer bacteria (Sb) fill the apical zone of the

be observed in close association with the pit-like structures on the surface

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 con-

stituted the major part of the filament (Plate 1d).

different treatments. (a) and (b) Group 1 mussels, placed in sulphide-

r present but several empty vacuoles (asterisks) and sulfur-oxidizer

dizer bacteria (Sb) do not show morphological alterations. Scale bar:

30 days (group 3) have a spongy appearance, containing symbiont-

Am, respectively. (e) Pit-like structures (arrows) are evident on the

er containing about 10 AM dissolved sulphide for 15 days (group 4).

bacteriocytes. Scale bar: 1 Am. (f) Sulfur-oxidizer bacteria (Sb) could

of bacteriocytes in re-acclimatized mussels. Scale bar: 0.2 Am.

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110 105

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110106

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 bacterio-

cytes 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 bac-

teria: 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

Fig. 3. PCR detection of B. azoricus endosymbiont bacteria in total geno

rRNA (primer set 1) and ATP-sulfurilase (primer set 2) genes was succe

(group 0, lanes 1 and 2 respectively). Mussels kept in seawater for 30 da

presence of endosymbiont (group 3, lanes 3 and 4). Upon re-acclimatizatio

(lane 5) and ATP sulfurilase (lane 6) is again detected suggesting inter

products are shown (arrows) as well as 1 Kb DNA ladder.

of aggregation within the same vesicle. Despite of

the size and shape modulations (due to sectioning

angle), all sulphur oxidizers shared similar ultra-

structural 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 (exper-

imental control) lost their methane-oxidizing sym-

bionts, 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 lyso-

somes (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

mic DNA preparations from gill tissues. The presence of both 16S

ssfully detected in samples from mussels dissected upon collection

ys failed to reveal the presence of both genes and consequently the

n to sulphide-supplied seawater (group 4) the presence of 16S rRNA

-animal endosymbiont transmission. Expected PCR amplifications

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110 107

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 electro-

phoresis, 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 sulfury-

lase 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 Bathy-

modiolus 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 bac-

teria 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 symbiont-

containing 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

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 Lab-

Horta, 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 AMsulphide 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 sensi-

tive 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). Oxy-

gen saturation is another major factor influencing

animal maintenance. Under anoxic conditions anae-

robic 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-

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110108

tigations to be conducted with a great advantage over

the costly and both time- and human-resource-

consuming in situ observations. Behavioural observa-

tions 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 sea-

water for up to 30 days. Residual endosymbiont

bacteria might have survived the sulfide-deprivation

treatment, however, our PCR detection system sug-

gests that the endosymbiont bacterial population can

vary from one experimental condition to another and

this population particularly increases when sulfide-

deprived mussels are exposed to an environment

containing sulfide and other untreated mussel indi-

viduals. 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 environ-

mental 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 proto-

cols. 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, sym-

biont-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 meth-

ane 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. Addi-

tionally, there is evidence originated from studies by

Gros and co-workers on environmental transmission

of sulfur-oxidizer bacteria in various symbiont-har-

bouring bivalves (Gros et al., 2003, 1996, 1998a,

1999). These authors have determined the trans-

mission mechanisms by experimental colonisation of

aposymbiotic Codakia larvae and proposed the bpit-likeQ 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 sea-

water. There is evidence for the existence of free-

living 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 phylo-

genetic studies on sulphur-oxidizing bacterial endo-

symbionts 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 endo-

symbiont target genes would be amplified, and thus

ruling out the possibility of contaminations from free-

living 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 experimen-

tally 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. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110 109

southern East Pacific Rise reached similar conclusions

regarding symbiont loss and consequent ultrastruc-

tural modifications, and thus corroborate our exper-

imental results.

By developing a successful laboratory set-up for

the maintenance of B. azoricus that enables preser-

vation and manipulation of endosymbiosis, we

provide a basis for a more elaborated ecophysio-

logical research, in order to understand the general

principles that govern adaptations to the hydro-

thermal environment.

Acknowledgements

The research was undertaken under the scope of

the research project SEAHMA (Seafloor and sub-

seafloor hydrothermal modelling in the Azores Sea)

funded by FCT (PDCTM/P/MAR/15281/1999). Post-

doctoral 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 sym-

bionts. Physiological Zoology 64, 1444–1470.

Cline, J.D., 1989. Spectrophotometric determinations of hydrogen

sulphide in natural waters. Limnology and Oceanography 14,

454–459.

Craddock, C., Hoeh, W.R., Lutz, R.A., Vrijenhoek, R.C., 1995.

Extensive gene flow among Mytilid (Bathymodiolus thermo-

philus) 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., Kenni-

cutt, 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 hydro-

thermal vent mussel, Bathymodiolus thermophilus. American

Zoologist 29, A81.

Gros, O., et al., 1996. Environmental transmission of a sulfur-

oxidizing 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 differ-

entiation 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 develop-

ment and endosymbiont transmission mode in the symbiotic

clam Lucinoma aequizonata (Bivalvia : Lucinidae). Invertebrate

Reproduction and Development 36, 93–103.

E. Kadar et al. / J. Exp. Mar. Biol. Ecol. 318 (2005) 99–110110

Gros, O., et al., 2003. Detection of the free-living forms of sulfide-

oxidizing gill endosymbionts in the lucinid habitat (Thalassia

testudinum environment). Applied and Environmental Micro-

biology 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 rDNA-

based phylogeny of sulphur-oxidising bacterial endosymbionts

in marine bivalves from cold-seep habitats. Marine Ecology-

Progress 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 Environ-

mental 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 hydro-

thermal 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 accre-

tionary 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, Mid-

Atlantic 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 thiotro-

phic 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 cur-

rently active hydrothermal vents in the Indian Ocean. Journal

of the Marine Biological Association of the United Kingdom 83,

841–848.