AN ABSTRACT OF THE THESIS OF
Jeannine Kay Larabee for the degree of Master of Science in Zoology presented on
October 24, 1997. Title: Sulfide Metabolism in Lucinafloridana, a Sulfur SymbiontContaining Marine Clam.
Redacted for Privacy
Abstract approved:
Amy E.
derson
Lucina floridana is an intertidal clam that contains intracellular, sulfur-oxidizing
symbiotic bacteria in its gills. These bacteria are chemolithoautotrophic, using the energy
from the oxidation of reduced sulfur compounds to power carbon fixation. Although
sulfide is an energy-rich compound, it is also toxic to aerobic metabolism and autooxidizes
in the presence of oxygen. Flow-through experiments were designed to examine the
metabolic response to sulfide of the intact symbiosis. Three questions were addressed: 1)
What are the major sulfide oxidation products in the gill and in the hemolymph? 2) How
are sulfide and its major oxidation products distributed between the hemolymph and the
gill? and 3) What is the metabolic poise (use of aerobic or anaerobic metabolism) in the
gills during sulfide oxidation? The production of sulfide in the gills in response to various
sulfur starvation regimes was also examined.
Sulfide and its major oxidation products were determined using HPLC analysis of
incompletely oxidized sulfur compounds. The metabolic poise was assessed by measuring
the concentrations of substrates and products of anaerobic metabolism.
Sulfide was produced in the gills of Lucina floridana after five days of sulfur
starvation. The addition of invertebrate feed, a source of organic carbon for the clams, to
the seawater significantly reduced the concentration of sulfide in the gills following sulfur
starvation, suggesting that sulfide production was a consequence of symbiont oxidation of
reduced sulfur compounds in the gills to provide energy for carbon fixation. Sulfide was
oxidized to thiosulfate which was preferentially distributed in the hemolymph relative to
the gills, a pattern that is consistent with further oxidation of thiosulfate by the bacteria in
the gills for energy generation. This study was the first to investigate the metabolic poise
of a lucinid in response to varying oxygen and external sulfide treatments; evidence
suggests that anaerobic metabolism may be an important metabolic mode for Lucina
floridana.
Sulfide Metabolism in Lucina floridana, a Sulfur
Symbiont-Containing Marine Clam
by
Jeannine Kay Larabee
A THESIS
submitted to
Oregon State University
in partial fulfillment of
the requirements for the
degree of
Master of Science
Presented October 24, 1997
Commencement June 1998
Master of Science thesis of Jeannine Kay Larabee presented on October 24, 1997
APPROVED:
Redacted for Privacy
Major Professor, representing Zoology
Redacted for Privacy
epartment o Zoology
Redacted for Privacy
Dean of Graduat
chool
I understand that my thesis will become part of the permanent collection of Oregon State
University libraries. My signature below authorizes release of my thesis to any reader.
Redacted for Privacy
Jeannine Kay Lar
e, Author
ACKNOWLEDGMENTS
I thank my advisor, Amy Anderson, for the support and encouragement she has
given me throughout my graduate career. She has contributed immensely to my
intellectual development as a scientist. She has also been very generous with her time
spent working with me.
I also thank the other member of the Anderson lab, Sarah Deel, for her help during
my graduate career. In addition, I thank Randy Bender, Brady Allen, Iki Hahn, Lena
Gerwick, Nora Demers, Sarah Fryer, and Chris Bayne, members of the Bayne lab, for
acting as a surrogate lab during the 1995/1996 school year and for their advice and
discussions about my thesis research.
I thank George Somero and Chris Bayne, members of my committee, for the ideas,
advice, and support they provided during my graduate career. They have always been
available to discuss my thesis research. I thank Michael Hyman for serving on my
committee. I also thank Barb Taylor for her support and input regarding my future career
plans. I thank the Somero, Bayne, Moore, Morris, Taylor and Mason labs for the use of
various items of equipment during my thesis research. I thank Allen Gibbs and the other
members of the Gibbs lab for their encouragement during the writing of this thesis.
The lam discussions and microbrews with fellow night owl Barb Byrne have aided
this project tremendously; I appreciate her friendship and support. I also gratefully
acknowledge the support and friendship of Carla Richardson, Iki Hahn, Mike Lemaster,
Laura Knittel, Darren Lerner, Ignacio Moore and Jen Penn.
I thank my parents, Gayle and Ken Larabee, my sister, Michelle Larabee, and my
in-laws, Mike and Cheryl Greene, for their constant encouragement during my graduate
career.
I thank Mike Greene for his perspective and for his continual support, especially
during the difficult times of this project, but most of all, for his love; I thank Moab and
Loki for their invaluable advice and encouragement. I love all three of you.
TABLE OF CONTENTS
Page
Introduction
1
Methods
13
Results
20
Discussion
39
References
59
LIST OF FIGURES
Page
Figure
Effect of sulfur starvation regime on sulfide (H2S) concentrations (.1M)
in the gill of Lucina floridana
22
Sulfite (S032), thiosulfate (S2032), and sulfide (H2S) concentrations
(.1M) in the hemolymph and gills of Lucina floridana from flowthrough experiments run in the presence of different external [H2S] in
seawater without invertebrate feed
25
Sulfite (S032), thiosulfate (S2032), and sulfide (H2S) concentrations
(gM) in the hemolymph and gills of Lucina floridana from flowthrough experiments run in seawater containing invertebrate feed with
and without added H2S
29
4
Correlation plot of succinate (mM) vs. aspartate (mM) for anaerobic
and aerobic control Lucina floridana gills
35
5
Succinate:aspartate ratios for Lucina floridana gills from anaerobic and
aerobic control treatments and from flow-through experiments run in
the presence of different external [112S] with and without invertebrate
feed (I.F.)
38
1
2
3
LIST OF TABLES
Page
Table
1
Effect of different five-day sulfur starvation regimes on sulfide
concentrations in the gill of Lucina floridana
(H2S)
21
Sulfite (S032), thiosulfate (S2032), and sulfide (H2S) concentrations in
hemolymph (H) and gills (G) of Lucina floridana from flow-through
experiments run in the presence of different external [H2S] in seawater
without invertebrate feed
24
Sulfite (S032), thiosulfate (S2032), and sulfide (H2S) concentrations in
hemolymph (H) and gills (G) of Lucina floridana from flow-through
experiments conducted in seawater containing invertebrate feed with
and without added H2S
28
The distribution of sulfite (S032), thiosulfate (S2032), and sulfide
(H2S) between the hemolymph and the gill for Lucina floridana pooled
across all flow-through treatments
32
5
Aspartate and succinate concentrations (mM) and succinate:aspartate
ratios (S:A) in anaerobic and aerobic gills of Lucina floridana
34
6
Aspartate and succinate concentrations and succinate:aspartate ratios
(S:A) in gills of Lucina floridana from flow-through experiments run
in the presence of different external [H2S] in seawater with and without
invertebrate feed
37
2
3
4
DEDICATION
This thesis is dedicated to my mother and father, Gayle and Kenneth Larabee, in
appreciation of their ever-lasting support for all my endeavors.
Sulfide Metabolism in Lucina floridana, a Sulfur
Symbiont-Containing Marine Clam
Introduction
Invertebrate/sulfur bacteria symbioses
Marine invertebrate/sulfur bacteria symbioses were discovered by researchers
investigating the source of nutrition for the animals of the Galapagos deep-sea
hydrothermal vents (Cavanaugh et al., 1981; Felbeck, 1981). Sulfur-based symbioses have
since been found in a wide range of reducing, hypoxic marine habitats from the deep sea to
the intertidal zone (for reviews see Somero et al., 1989 and Fisher, 1990), including
hydrocarbon seeps, mangrove swamps, eelgrass beds, deep-sea basins, sewage outfalls and
pulpmill effluent sites. The common feature among these habitats is the presence of
hydrogen sulfide that is produced as a result of geothermal processes (hydrothermal vents)
or from anaerobic degradation of decaying organic matter (non-hydrothermal vent sites),
the source of which may be anthropogenic or natural, by sulfate-reducing bacteria that use
sulfate as an alternate electron acceptor (instead of oxygen). The symbiotic sulfuroxidizing bacteria are chemolithoautotrophic; that is, they oxidize an inorganic chemical,
either hydrogen sulfide or one of its oxidation products, to provide energy (ATP) and
reducing equivalents (NAD(P)H) to power carbon fixation via the Calvin-Benson cycle.
While bacterial hydrogen sulfide oxidation is an energetically favorable process, hydrogen
sulfide spontaneously autooxidizes in the presence of oxygen (Chen and Morris, 1972) and
is also toxic to aerobic metabolism (Nicholls, 1975; National Research Council, 1979). As
the sulfur-oxidizing bacteria are endosymbiotic (and intracellular in many hosts), the hosts
2
must be able to supply hydrogen sulfide to their symbionts while both minimizing sulfide
autooxidation and preventing inhibition of aerobic metabolism. The hosts for these sulfurbased symbioses span 6 invertebrate phyla (Platyhelminthes, Nematoda, Vestimentifera,
Mollusca, Pogonophora and Annelida) and more than 15 families, some of which may be
exclusively symbiotic (for review see Fisher, 1990). Marine invertebrate/sulfur bacteria
symbioses are well represented in the phylum Mollusca; the majority of molluscan species
containing chemolithoautotrophic symbionts are bivalves (Fisher, 1990). About half of
these bivalve species belong to the family Lucinidae, including the subject of my thesis
research, Lucina floridana.
Sulfide toxicity and options for symbiotic organisms facing sulfide toxicity
Hydrogen sulfide (the term sulfide, also abbreviated as H2S, is used to represent all
three forms of sulfide: H2S, HS-, and S2) is the most reduced form of sulfur ("sulfur" is a
general term used to represent all inorganic forms of sulfur while the term "elemental
sulfur" refers specifically to the elemental form) and thus contains a relatively large store
of potential energy to be gained via oxidation. However, sulfide is also highly diffusible (as
H2S) and toxic, exerting its toxic effects through two mechanisms. First, sulfide can bind
to the iron center on the heme portion of the cytochrome oxidase complex of the
mitochondrial electron transport system which leads to a reversible inhibition of aerobic
metabolism (Nicholls, 1975; National Research Council, 1979). Sulfide also reversibly
binds to bacterial terminal oxidases, inhibiting bacterial aerobic metabolism as well
(Nicholls, 1975; Wilson and Erecinska, 1979; Kraus et al., 1992). Second, sulfide can
bind to the heme portion of hemoglobin (Hb) molecules (both circulating and cytoplasmic
3
Hbs) in most organisms, causing a reduction in the affinity of Hb for oxygen by several
orders of magnitude (Carrico et al., 1978; National Research Council, 1979).
Sulfur symbiont-containing marine invertebrates must deliver sulfide to their
symbionts while minimizing sulfide autooxidation and toxicity. At least four potential
options are available: 1) sulfide-insensitive cytochrome oxidase and Hb variants, 2) sulfide
binding proteins, 3) sulfide detoxification mechanisms, and 4) reliance on anaerobic
metabolism by the host and/or the symbiotic bacteria. To date, no sulfide- insensitive
cytochrome oxidase variants have been found, though there is evidence for sulfide
insensitive hemoglobins (see below; for reviews see Somero et al., 1989 and Childress and
Fisher, 1992).
Sulfide binding proteins, which may be cytoplasmic or circulatory, function to
minimize autooxidation of sulfide and prevent inhibition of aerobic metabolism (for
reviews see Somero et al., 1989 and Childress and Fisher, 1992). Circulating sulfide
binding proteins are present in two taxa, the phylum Vestimentifera and the molluscan
family Vesicomyidae. In the vestimentiferans the animals' circulating Hb is capable of
binding sulfide on a site different from the oxygen binding site (Arp et al., 1987; Arp et al.,
1990), while the vesicomyids appear to use a non-Hb molecule that may contain Zn2+ in
the active site (Childress et al., 1991b; Childress et al., 1993). Cytoplasmic sulfide binding
proteins have been found in a vesicomyid (Childress et al., 1993) and in several members
of two other bivalve families, the Solemyidae and the Lucinidae. In most cases these
animals' cytoplasmic Hb serves as a sulfide binding protein by forming ferric Hb sulfide;
these organisms possess one form of Hb for binding sulfide and one or more forms of Hb
capable of binding oxygen (Doeller et al., 1988; Kraus and Wittenberg, 1990; Wittenberg
4
and Kraus, 1991; Kraus et al., 1996). Another lucinid, Myrtea spinifera, has a non-Hb
cytoplasmic sulfide binding protein (Dando et al., 1985).
Another strategy for avoiding the toxic effects of sulfide is the oxidation of sulfide
to a non-toxic intermediate. In Solemya reidi, a shallow water clam that contains
symbiotic sulfur bacteria, the mitochondria are capable of oxidizing sulfide (Powell and
Somero, 1985) to thiosulfate (S2032), a non-toxic partially reduced sulfur compound
(O'Brien and Vetter, 1990). Thus, mitochondrial oxidation of sulfide detoxifies two
sulfide molecules and produces a soluble intermediate used by the bacteria for further
energy generation (O'Brien and Vetter, 1990). Mitochondrial sulfide oxidation is linked
to ATP production; evidence suggests that sulfide oxidation is coupled to oxidative
phosphorylation through cytochrome c (Powell and Somero, 1986). Mitochondrial sulfide
oxidation may be partially uncoupled from ATP synthesis to serve as a sulfide
detoxification mechanism (O'Brien and Vetter, 1990). Sulfide oxidation may be a
common capability of mitochondria as it has been demonstrated in many organisms,
including fish, rats, annelids and molluscs (Bartholomew et al., 1980; Bagarinao and
Vetter, 1990; Tschischka and Oeschger, 1995; Volkel and Grieshaber, 1995).
Sulfide oxidation is also catalyzed by non-enzymatic mechanisms. Heme-
associated iron, other metal ions, and several low molecular weight organic molecules are
capable of oxidizing sulfide (Chen and Morris, 1972). Elemental sulfur is capable of
oxidizing sulfide to polysulfides (Millero, 1986). As mentioned earlier, sulfide
spontaneously oxidizes upon contact with oxygen (Chen and Morris, 1972). Sulfide
oxidation is also catalyzed by pigment granules, concretions containing proteins, heme,
5
and minerals, in several invertebrate species (Powell and Arp, 1989; Anderson and Hand,
1990).
The inhibition of aerobic metabolism by sulfide contrasts with the classical
definition of anaerobiosis: the onset of anaerobic metabolism due to internal or
environmental hypoxia. Anaerobic metabolism may be an important metabolic mode for
many of the invertebrate hosts as most hosts inhabit environments that have conditions
sufficient for inhibiting aerobic metabolism, high levels of sulfide and low (or fluctuating)
levels of oxygen (Somero et al., 1989; Fisher and Childress, 1992). Organisms inhabiting
reducing environments generally use energetically improved alternative anaerobic
pathways (i.e. a greater ATP yield per mole of glucose catabolized) relative to the
standard glucose to lactate fermentation pathway used by most aerobes (Hochachka and
Somero, 1984). Most bivalves use an anaerobic pathway called the glucose-aspartate cofermentation pathway where aspartate serves as a substrate and succinate serves as an
endproduct for energy production (Hochachka, 1980; DeZwaan, 1983)
Sulfide can inhibit aerobic metabolism in bacteria by reversibly binding to the
terminal oxidase of the electron transport chain (Nicholls, 1975; Wilson and Erecinska,
1978; Kraus et al., 1992) which precludes aerobic (use of 02 as a terminal electron
acceptor) bacterial sulfide oxidation. Recently, several studies have suggested that the
bacterial symbionts of three hosts may be able to use nitrate as an alternate electron
acceptor (Wilmot and Vetter, 1992; Hentschel and Felbeck, 1993; Hentschel et al., 1993;
Hentschel and Felbeck, 1995; Hentschel et al., 1996). Anaerobic (without 02 as a
terminal electron acceptor) oxidation of sulfide represents an intriguing possibility, as
partitioning of oxygen to the host and nitrate to the symbionts would prevent competition
6
between host and symbiont for low levels of habitat oxygen (Hentschel et al., 1993;
Hentschel et al., 1996). While these studies have shown the symbionts to be capable of
using nitrate as an electron acceptor, they have used unrealistically high concentrations of
nitrate (and in most cases, isolated symbionts). Thus, the relative importance of anaerobic
sulfide oxidation by the symbionts in their natural habitat cannot be determined from these
studies.
Chemoautotrophy and the transfer of symbiont-fixed carbon from symbiont to host
Evidence for the autotrophic nature of the bacteria was first suggested by the
presence of enzymes of the Calvin-Benson cycle in the symbiont-containing tissues, the
same pathway used by green plants to fix carbon (Fisher, 1990). That symbiont-fixed
carbon is important to the hosts' carbon budget is suggested by the observation that most
hosts have absent or extremely reduced feeding appendages and guts (Fisher, 1990).
There are several methods that have been used to investigate the importance of symbiontfixed carbon to the host: for example, 1) the detection of molecules unique to bacteria
(such as bacterial cell wall components) in the host or 2) the transfer of radiolabeled
carbon from bacteria to host (Fisher and Childress, 1986; Fisher, 1990; Conway and
McDowell-Capuzzo, 1991; Boetius and Felbeck, 1995). The results from all of these
methods suggest that a large portion of the hosts' carbon requirements are met by the
symbionts (Fisher, 1990; Conway and McDowell-Cappzzo, 1991; Boetius and Felbeck,
1995). However, the strongest evidence for the importance of symbiont-fixed carbon to
the hosts' nutrition are the demonstrations of net autotrophy (i.e. CO2 uptake is greater
7
than CO2 respired) in the presence of sulfide for the clam, Solemya reidi (Anderson et al.,
1987), and the vent tubeworm, Riftia pachyptila (Childress et al., 1991a).
Sulfide metabolism in free-living sulfur bacteria and symbiotic sulfur bacteria
The sulfide oxidation pathways used by the symbiotic sulfur bacteria are not well
understood because researchers have been unable to successfully culture isolated
symbionts; in contrast, the sulfide oxidation pathways used by free-living sulfur-oxidizing
bacteria are well characterized. The metabolic pathways used by free-living sulfuroxidizing bacteria for sulfur metabolism are extremely diverse, that is, there is not one
common set of enzymes used for sulfur oxidation by all species of sulfur-oxidizing bacteria
(Kelly, 1988). For many free-living sulfur-oxidizing bacteria, elemental sulfur (S°)
appears to be an intermediate in the sulfide oxidation pathway and many free-living sulfur-
oxidizing bacteria contain elemental sulfur deposits. There is some debate, however, over
whether production of elemental sulfur from sulfide is a side reaction (i.e. bacteria produce
elemental sulfur to store energy) or the main intermediate in the sulfide oxidation pathway
(Nelson and Hagen, 1995). The presence of elemental sulfur has also been demonstrated
in several symbiotic sulfur bacteria (Dando et al., 1985; Vetter, 1985, Cary et al., 1989;
Deel, 1996). Most free-living sulfur-oxidizing bacteria can oxidize thiosulfate (S2032) to
sulfate (S042) with sulfite (S032") serving as either a free or enzyme-bound intermediate
(Nelson and Hagen, 1995). There is enzymatic evidence that sulfite is also an intermediate
in the sulfur oxidation pathways of symbiotic sulfur bacteria (Fisher, 1990). Evidence
from studies using free-living sulfur-oxidizing bacteria suggests that ATP is generated via
both substrate-level phosphorylation reactions and bacterial electron transport though the
10
the hemolymph of clams incubated in sulfide-containing seawater (Anderson, 1995), 3)
excised gills are capable of thiosulfate uptake from the surrounding medium (Anderson,
1995), and 4) thiosulfate increases the ventilation frequency of whole clams (Anderson,
1995).
Summary of thesis questions and hypotheses
The goal of these studies was to further investigate the physiology and
biochemistry of sulfide metabolism in L. floridana. This research presents data showing
the production of sulfide in the gills of clams exposed to various sulfur starvation regimes.
The potential for the production of sulfide using a type of inorganic fermentation,
disproportionation, a metabolic pathway that has been demonstrated in free-living sulfur
bacteria (Bak and Cypionka, 1987; Kramer and Cypionka, 1989; Thamdrup et al., 1993;
Janssen et al., 1996), is discussed as well as the possible sulfur compound(s) used for the
disproportionation reaction.
These studies also examined the metabolic response of clams exposed to seawater
containing different sulfide concentrations. Three specific questions were addressed:
1) What are the major sulfide oxidation products in the gill and in the hemolymph?
2) How are sulfide and its major oxidation products distributed between the hemolymph
and the gill? and 3) What is the metabolic poise (i.e. are anaerobic or aerobic metabolic
pathways being used) of the gills during sulfide oxidation? Flow-through conditions were
used to answer these questions because of the ability to maintain a relatively constant
sulfide concentration and oxygen tension.
11
I expected that thiosulfate would be a major sulfide oxidation product as has been
suggested in previous studies using lucinids (Cary et al., 1989; Anderson, 1995) and
Solemya reidi (Anderson et al., 1987). Regardless of the sulfide oxidation mechanism(s)
used for thiosulfate production, I hypothesized that thiosulfate would be distributed
preferentially in the hemolymph relative to the gill. The gill houses the symbionts that are
likely able to oxidize thiosulfate, thus, the gill may serve as a thiosulfate "sink."
The metabolic poise during sulfide oxidation had not been investigated for any
lucinid prior to this study. In Solemya reidi, aerobic metabolism is maintained in the
animal tissues at external sulfide concentrations up to 100pM, the sulfide concentration at
which maximum net autotrophy is seen; anaerobic metabolic pathways are initiated at
higher external sulfide concentrations (Anderson et al., 1990). Based on these results, I
expected that aerobic metabolism would be maintained in the gills at low external sulfide
concentrations; as sulfide in the environment increased, aerobic metabolic pathways would
be inhibited and anaerobiosis would occur. I also hypothesized that the concentration of
sulfide that is inhibitory to aerobic metabolism in L. floridana would be lower than that
which inhibits aerobic metabolism in S. reidi, as L. floridana has a lower metabolic rate
(Anderson, pers. comm.) and inhabits sediments with lower concentrations of sulfide
(Childress and Lowell, 1982; Fisher and Hand, 1984). To determine the metabolic poise
of the gills during sulfide oxidation, the concentrations of aspartate and succinate were
measured and compared with the concentrations of these metabolites in control clams
from aerobic or anaerobic conditions.
The results from these experiments are compared with what is known about sulfide
metabolism in several other lucinids and with the model of sulfide oxidation for S. reidi
12
(Anderson et al., 1987). This study contributes to the understanding of the metabolism of
reduced sulfur compounds in a sulfur-symbiont containing bivalve that may rely on
anaerobic metabolism a large portion of the time as it inhabits a hypoxic, sulfidic
environment, conditions that are capable of inhibiting aerobic metabolism.
13
Methods
Collection and maintenance of Lucina floridana
Lucina floridana were collected from seagrass beds at depths of 0.25-1.0 m from
St. Joseph's Bay, Florida. They were shipped overnight in flip-top plastic bags containing
four to eight clams in 10 g sand and 20 ml seawater (SW). The clams were housed in 20
L aquaria at 20°C (room temperature) with 0.8 pm filtered SW obtained from Hatfield
Marine Science Center (HMSC) in Newport, Oregon. Aquaria were fitted with
AquaClear 200 filters, a biological filtration system. Each aquarium contained sediment
from the clams' natural habitat to a depth of 10-12 cm. Supplemental sulfur compounds
were added to the aquaria by placing the sulfur crystals 2-5 cm below the surface of the
sediment. Small sodium sulfide crystals were added weekly (approximately eight to ten 30
mg crystals per aquarium) and small sodium thiosulfate crystals were added bi-weekly
(approximately six to eight 25 mg crystals per aquarium). A week prior to receiving a
shipment of clams, spinach, paper towels and calcium sulfate were layered beneath the
mud to recondition the aquaria per the methods of Anderson et al. (1987). Clams were
allowed to acclimate to laboratory conditions for one week prior to use in experiments.
Sulfur starvation experiments
It was important to use clams with clams relatively low (less than 30uM)
concentrations of sulfur compounds in their gills (including any elemental sulfur stores in
the bacteria) and hemolymph at the start of a flow-through experiment so that these sulfur
compounds contribute minimally to the net sulfur flux during the experiment. To
determine a sulfur starvation regime to prepare clams for the flow-through experiments,
14
clams were starved of exogenous sulfur compounds for various lengths of time and under
various conditions. For all of the sulfur starvation regimes described below, clams were
randomly selected from the aquaria and placed in 200 ml plastic beakers (one clam/beaker)
containing 150 ml of 0.8 pm filtered SW; the SW in each container was changed daily.
Several researchers have noted that clams with large elemental sulfur deposits in
their gills have thick creamy yellow gills while clams with elemental sulfur-depleted gills
have thin dark brown gills (Dando et al., 1985; Vetter, 1985; Deel, 1996). Deel (1996)
has noted that it may require up to six weeks of sulfur starvation for complete depletion of
elemental sulfur stores in L. floridana. After two days of sulfur starvation in unaerated
seawater, the clams' gills became thin and dark brown, indicating a reduction in elemental
sulfur stores; however, the concentrations of some of the other sulfur compounds in the
gill and hemolymph were still relatively high (> 250pM; the data for sulfur starvation
experiments shorter than five days are not shown). After five days of sulfur starvation in
unaerated seawater, all sulfur compounds in the gill and hemolymph were relatively low
except for gill sulfide; sulfur starvation experiments conducted for longer than five days
led to large mortalities (up to 60% for a fifteen day sulfur starvation experiment; the data
for starvation experiments longer than five days are not shown). Because of these
problems with high clam mortality, I decided to sulfur starve clams for five days and try
two other methods to reduce the levels of sulfur compounds in the hemolymph and the
gill: 1) sulfur starvation in the presence of aeration (SW+A) and 2) sulfur starvation in the
presence of aeration in seawater containing invertebrate feed, a mixture of proteins, lipids
and carbohydrates (SW +AIF). For clams sulfur- starved in SW+A, each clam's container
was continuously aerated over the five day sulfur starvation period. For clams sulfur-
15
starved in SW+AIF, invertebrate feed, (Liquify invertebrate feed by Interpet, Bristol, UK),
was added to the seawater at a concentration of 2 drops/L.
Sulfur starvation regime for clams used in the flow-through experiments
Clams were randomly selected from the aquaria and placed in 200 ml plastic
beakers (1 clam per beaker) containing 150 ml of 0.8 pm filtered SW with invertebrate
feed added at a concentration of 2 drops/L. Each container was continuously aerated over
the starvation period and the SW was changed daily. Each clam was sulfur starved for
five days prior to use in the flow-through experiments.
Flow-through experiments
The SW used for all flow-through experiments was filtered through a 0.8 pm filter,
buffered with 20mM Hepes (Sigma Chemical) and the pH adjusted to 8.0 with HC1.
Invertebrate feed (IF) was added (at a concentration of 2 drops/L) to SW for the flowthrough experiments conducted in SW with IF. A sulfide stock solution was prepared by
de-gassing SW with helium gas for thirty minutes to remove oxygen. Sodium sulfide
crystals (Sigma Chemical) were added to the de-gassed SW solution at one of several
concentrations and the pH was adjusted to 8.0 with HC1. An aerated SW stock solution
was prepared by aerating and stirring SW for thirty minutes. A Rainin Rabbitt peristaltic
pump was used to pump the sulfide stock solution and a Manostat peristaltic pump was
used to pump the aerated SW solution; both solutions were pumped to an in-line, stirred
mixing chamber (flow rates ranged from 3-4 ml/min). The mixed, partially oxygenated
solution (2.7-3.5 ml 02 /L; 45-55% 02 saturation) was pumped from the mixing chamber
to the flow-through chamber which contained four to five clams per flow-through
16
experiment. The shells of clams used for flow-through experiments were cleaned with a
3% hydrogen peroxide solution before placement into the flow-through chamber, a
plexiglas tube 35 cm in length and 7.5 cm in diameter which contained 250 ml of aerated
SW at the start of each flow-through experiment. The flow-through chamber was covered
with a dark cloth to simulate burrow conditions during the flow-through experiments.
Control treatments were conducted in sulfide-free seawater (SFSW) both with and
without IF. For the two control (SFSW) flow-through treatments, the procedure for
preparing the sulfide stock solution was followed except that sodium sulfide crystals were
not added; this was done so that all treatment groups experienced similar oxygen tensions
in the flow-through chamber. Experimental treatments consisted of different external
sulfide concentrations (external [112S]) in SW both with and without IF. A period of one
hour was required for complete equilibration in the chamber with the treatment solution;
clams were incubated for eight hours under control or experimental treatments. 02
tension was monitored continuously (Microelectrodes 02 probe and meter). Water
samples for external [H2S] determinations were taken from the mixed solution as it passed
through the mixing chamber and the sulfide concentration was determined by HPLC (see
below). Four treatments were used in SW without IF (sulfide concentrations (pM) listed
in parentheses as mean ± standard error of the mean (S.E.M.)): SFSW (0.843 ± 0.102),
6pM H2S (5.94 ± 0.335), 13pM H2S (12.97 ± 1.47), and 40pM H2S (40.02 ± 2.49).
Two treatments were used in SW with IF: SFSW (0.230 ± 0.071) and 21M H2S (2.37 ±
0.137).
17
HPLC analysis of gills and hemolymph from flow-through experiments
At the conclusion of each flow-through experiment, each clam was removed from
the flow-through chamber: both gills were excised and the hemolymph was collected. The
wet weights for each gill ranged between 75-100 mg and the volume of hemolymph
ranged between 100-300 pliclam. One gill was freeze-clamped using liquid nitrogen and
stored at -80°C for aspartate and succinate determinations; the other gill and the
hemolymph were derivatized with monobromobimane (mBB, Calbiochem), a fluorescent
compound that reacts with incompletely oxidized sulfur compounds, per the methods of
Fahey and Newton (1987) and Vetter et al. (1989) with the following modifications.
Homogenized gill and hemolymph samples were reacted with 27mM mBB at a 10:1
volume (hemolymph: mBB, at 3-fold excess mBB of the expected concentration of each
sulfur compound). The mBB-derivatized samples were filtered using a 0.45 pm filter
(Gelman Vericel Fp-450) and frozen at -80°C for later HPLC analysis. For clams from
the flow-through experiment conducted in SFSW without IF, the mantle tissue was also
removed and derivatized with mBB according to the above procedure.
High performance liquid chromatography (HPLC) analysis of mBB derivatized
samples was performed using an OMS Tech gradient mixer and integrater. Samples were
separated using a 25 cm Beckman C-18 silicia reversed phase column. An Alltech HPLC
pump was used to pump an increasing hydrophobic gradient of HPLC grade methanol
(Fisher Chemical) and 2.5mM methanesulfonic acid (pH 3.5, Sigma Chemical) at a flow
rate of 1.28 ml/min. The samples were detected using a Spectrovision fluorometer with a
390-395 nm excitation filter and a 475-480 nm detection filter. The lower sulfide
detection limit of the HPLC system is between 50-75 nM H2S.
18
Metabolic poise controls: anaerobic and aerobic control treatments
To prepare the aerobic control clams, six clams were randomly selected from the
aquaria and each clam was placed in a separate 1L beaker containing 1L continuously
artificial seawater (ASW). One clam from the aerobic treatment group died during the
treatment and was not used. For the anaerobic controls, six clams were randomly selected
from the aquaria and were placed in a 1L sealed flask containing 1L ASW that had been
de-gassed with helium for thirty minutes. Clams were maintained under these aerobic or
anaerobic conditions for thirty hours, at which point the gills were excised and freezeclamped for later use in the aspartate and succinate assays (see next section).
Metabolic poise assays: determination of succinate and aspartate of gills
To determine the metabolic poise of the tissues, the gills from clams from both the
flow-through experiments and the aerobic and anaerobic control treatments were used for
measurements of aspartate and succinate. Each frozen gill was ground under liquid
nitrogen in a mortar and pestle. The sample was diluted 1:15 (wt: vol) with chilled 7%
perchloric acid in a microcentrifuge tube. Each sample was homogenized twice for one
minute each on ice using a high speed tissue homogenizer (Janke and Kunkel Ultraturax).
The sample was then centrifuged at 10,000g for fifteen minutes at 4°C. The supernatant
was decanted and neutralized to pH 7 with KOH and potassium bicarbonate, stored
overnight at 4°C, and re-centrifuged for 5 minutes at 10,000g. The resulting supernatants
were stored at -80°C for use in the succinate and aspartate determinations.
Succinate and aspartate were analyzed according to the methods of Bergmeyer
(1984) per the modifications of Anderson et al. (1990). Briefly, consumption of each
19
metabolite was linked through a series of enzymatic reactions that were coupled to the
consumption of NADH. Thus, the amount of succinate or aspartate is proportional to the
decline in NADH absorbance at 340 nm. Both metabolites were measured at 25°C using
a Cary 319 spectrophotometer.
Statistical analysis
One-way analysis of variances (ANOVAs), Tukey post-hoc tests, paired t-tests,
correlation analyses and Fina tests were performed using the Systat statistical computing
package. The significance of the Pearson correlation coefficient was determined using the
methods of Sokal and Rohlf(1995).
20
Results
Effects of different sulfur starvation regimes on sulfide concentrations in the gill of
Lucina floridana
Three different five-day sulfur starvation regimes were conducted to determine
which treatment led to the lowest concentration of gill sulfide: 1) sulfur starvation in
seawater without invertebrate feed and without aeration (SW), 2) sulfur starvation in
seawater without invertebrate feed but in the presence of aeration (SW+A), and 3) sulfur
starvation in seawater with invertebrate feed and in the presence of aeration (SW +AIF).
Three clams were used for each of the first two sulfur starvation regimes. The third sulfur
starvation regime was conducted twice, using three clams each time, and the data were
pooled for statistical analysis.
Sulfur starvation in SW+AIF significantly reduced the concentration of sulfide in
the gill compared to sulfur starvation in SW (Table 1, Fig. 1). The reduction of gill sulfide
after sulfur starvation in SW+AIF was large relative to sulfur starvation in SW+A but the
two treatments did not differ significantly (P = 0.09). Subsequent to these initial studies,
sulfur starvation in SW+AIF conditions was chosen as the sulfur starvation regime for all
clams used for the flow-through experiments because this sulfur starvation regime led to
the lowest concentrations of gill sulfide (Table 1, Fig. 1). A one-way ANOVA was
significant with respect to sulfur starvation treatment (Table 1). Comparisons among the
means were performed using the Tukey test, a post-hoc test that maintains the
experimentwise error rate at a by adjusting the significance level to a value less than a for
21
each comparison between two means (Sokal and Rohlf, 1995). The Tukey test was
chosen as it is robust against unequal sample sizes (Day and Quinn, 1989).
Table 1. Effect of different five-day sulfur starvation regimes on sulfide (H2S)
concentrations in the gill of Lucina floridana. Values represent mean ± standard error
of the mean (S.E.M.). Means with different letters are significantly different from each
other (P < 0.01). See text for identification of sulfur starvation regimes.
Sulfur starvation regime
H2S (1-01)
SW
n=3
SW+A
n=3
SW+AIF
n=6
ANOVA
P-value
423.41 ± 57.36a
294.46 ± 43.28°
144.86 ± 42.37b
P = 0.008
22
700
1
600
500
-i- -
400
SW
SW+A
SW+A I F
a
(3)
a,b
(3)
visi
i 300
b
200
(6)
100
0
Sulfur starvation regime
Figure 1. Effect of sulfur starvation regime on sulfide (H2S) concentrations (i.iM) in
the gill of Lucina floridana. Values represent mean ± S.E.M. Means with different
letters are significantly different from each other (P < 0.01). See text for identification of
sulfur starvation regimes.
Effects of different external sulfide concentrations on sulfite, thiosulfate and sulfide
concentrations in the gills and hemolymph of Lucina floridana from flow-through
experiments run in seawater without invertebrate feed
The effects of different external sulfide concentrations (external [H2S]) on sulfite,
thiosulfate and sulfide concentrations in the gill and hemolymph of Lucina floridana from
flow-through experiments run in SW without invertebrate feed (IF) are shown in Table 2
and Figure 2. A control treatment was conducted in sulfide-free seawater (SFSW).
Experimental treatments were conducted at three external [H2S]: 6pM, 13pM and 40pM.
23
Data were pooled from the two control (SFSW) treatments (nine clams total). Three
clams were used for the flow-through experiment run at 61M sulfide, and four clams each
were used for the 13pM and 40pM sulfide flow-through experiments.
External [thS] had no significant effect on concentrations of sulfite, thiosulfate or
sulfide in gill or hemolymph (Table 2, Fig. 2). Sulfite was present at 22.15 ± 2.60pM in
mantle tissue from the SFSW treatment group (mean ± S.E.M.; Fig 2). Mantle tissue from
the SFSW treatment group contained thiosulfate at a concentration of 14.09 ± 10.30pM
(mean ± S.E.M.; Fig. 2). Sulfide was present in mantle tissue from the SFSW treatment
group at a concentration of 6.90 ± 0.28pM (mean ± S.E.M.; Fig. 2).
Table 2. Sulfite (S032), thiosulfate (S2032), and sulfide (H2S) concentrations in hemolymph (H) and gills (G) of Lucina
floridana from flow-through experiments run in the presence of different external [H2S] in seawater without invertebrate feed.
Values represent mean ± S.E.M. P-values: sulfur compound in the hemolymph or gill among external [H2S].
External [H2S] (PM)
n=3
n=4
40
n=4
3.88 ± 0.72
3.32 ± 0.40
4.17 ± 1.92
5.48 ± 0.39
P = 0.62
G
23.64 ± 6.46
55.0 ± 10.0
47.94 ± 16.73
37.64 ± 4.93
P = 0.12
H
50.25 ± 16.14
38.82 ± 18.38
94.22 ± 62.88
95.40 ± 19.65
P = 0.51
G
13.95 ± 3.49
13.74 ± 11.22
17.14 ± 3.92
16.98 ± 7.01
P = 0.67
6.11 ± 1.45
6.86 ± 1.66
8.04 ± 4.22
6.31 ± 1.11
P = 0.99
136.00 ± 41.78
158.13 ± 128.49
124.80 ± 64.33
139.12 ± 73.76
P = 0.98
Tissue
Type
S032- (pM)
H
52032" (pM)
ANOVA
P-value
13
Sulfur
Compound
H25 (PM)
G
SFSW
n=9
6
25
Figure 2. Sulfite (S032) , thiosulfate (S2032), and sulfide (H2S) concentrations (11M)
in the hemolymph and gills of Lucina floridana from flow-through experiments run
in the presence of different external [H2S] in seawater without invertebrate feed.
Bars represent mean ± S.E.M. Sample size for each treatment is in parentheses above
each bar. Mantle tissue does not contain the symbionts; it serves as a control for
concentrations of the sulfur compound (S032-, S2032-, or H2S) in a non-symbiotic tissue.
26
100
I hemolymph
gill
IM mantle
80
(4)
(3)
60
A
40
(9)
20
(9)
0
(4)
(3)
[
I
SFSW
13
6
40
200
150
a
100
50
350
300
I hemolymph
Ma gill
mantle
(3)
250 H
(4)
200
m
(4)
(9)
150
100
50
0
(9)
(3)
SFSW
(4)
(3)
6
(4)
13
40
External sulfide concentration (IAM)
Figure 2
27
Effects of different external [H2S] on sulfite, thiosulfate and sulfide concentrations in
Lucina floridana gills and hemolymph from flow-through experiments run in seawater
containing invertebrate feed
I hypothesized that flow-through experiments run in SW with IF, a mixture
of lipids, proteins and carbohydrates, would lead to lower concentrations of sulfide in the
gills of Lucina floridana relative to flow-through experiments run in SW without IF. This
hypothesis was based on results from the sulfur starvation experiments (Table 1 and Fig.
1), where sulfur starvation in seawater containing invertebrate feed (SW+AIF)
significantly reduced gill sulfide concentrations relative to clams starved in seawater
without invertebrate feed (SW). As mentioned earlier, all clams used for the flow-through
experiments were first starved of sulfur compounds in SW+AIF for five days. The goal of
this pre-treatment method was to reduce the concentrations of sulfur compounds in the
hemolymph and gills, including elemental sulfur stores in the bacteria, as bacteria in these
symbioses are known to store elemental sulfur (Dando et al., 1985; Vetter, 1985; Cary et
al., 1989, Deel, 1996) and, as is the case for free-living sulfur bacteria, may have the
capability to mobilize stored elemental sulfur compounds when facing an energy shortage
due to depletion of sulfur compounds in the environment (Jorgensen, 1982; Nelson et al.,
1986; Javor et al., 1990). In contrast to my previous observations showing that clams
became depleted of elemental sulfur after two days of sulfur starvation (as evidenced by
gill color and thickness, see methods), three of the fifteen clams used for the above set of
flow-through experiments were found to have thick, creamy yellow gills upon dissection
following the experiment, reflecting the presence of stored elemental sulfur compounds.
Data from these clams were excluded from the statistical analyses and are not presented.
28
Table 3 and Figure 3 show the effects of external [H2S] on sulfite,
thiosulfate and sulfide concentrations in the gill and hemolymph of Lucina floridana from
flow-through experiments conducted in SW with IF. Two flow-through experiments were
conducted at each of two treatments: SFSW and 21.1.M H2S. The data from the two flow-
through experiments conducted in each treatment were pooled for statistical analysis.
There were no significant differences in sulfite, thiosulfate, or sulfide in the hemolymph or
the gills between clams exposed to SFSW and clams exposed to 2tiM H2S (Table 3, Fig.
3).
Table 3. Sulfite (S032), thiosulfate (S2032), and sulfide (H2S) concentrations in the
hemolymph (H) and gills (G) of Lucina floridana from flow-through experiments
conducted in seawater containing invertebrate feed with and without added H2S.
Values represent mean ± S.E.M. P-values: sulfur compound in the hemolymph or gill
among external [H2S].
External [HA (PM)
SFSW
n=6
ANOVA
P-value
Sulfur
Compound
Tissue
Type
S032- (pM)
H
6.25 ± 2.57
10.89 ± 3.56
P = 0.36
G
37.93 ± 11.40
58.91 ± 11.11
P = 0.23
31.09 ± 18.29
144.93 ± 59.33
P = 0.15
19.03 ± 3.65
28.69 ± 10.20
P = 0.47
7.60 ± 2.53
16.23 ± 3.60
P = 0.11
128.10 ± 52.47
222.01 ± 49.52
P = 0.23
52032" (pM)
G
H2S (PM)
G
2
n=9
29
Figure 3. Sulfite (S032), thiosulfate (S2032) and sulfide (H2S) concentrations (p,M)
in hemolymph and gills of Lucina floridana from flow-through experiments run in
seawater containing invertebrate feed with and without added H2S. Bars represent
mean ± S.E.M. Sample size for each treatment is in parentheses above each bar.
30
SFSW
250
2
I
EgEl
300
gill
(9)
hemolymph
(9)
250
200
(6)
100
SFSW
2
External sulfide concentration (JIM)
Figure 3
31
Effect of invertebrate feed on gill sulfide in clams from the flow-through experiments
In this study, conducting flow-through experiments in SW with IF did not cause a
significant reduction in sulfide concentrations in the gills when compared to flow-through
experiments conducted in SW without IF (P = 0.909).
The decision to conduct flow-through experiments in SW with IF was based on a
comparison between the concentrations of gill sulfide from an early sulfur starvation
experiment where three clams were sulfur-starved for five days in SW+AIF (the sulfur
starvation regime for all clams used in the flow-through experiments) and the
concentrations of gill sulfide from clams from the flow-through experiment conducted in
SFSW without IF. Clams from the latter group had higher (though not significantly, P =
0.08) concentrations of sulfide in their gills, suggesting that nine hours of incubation in
SFSW without IF (the length of the flow-through experiment) was long enough to
stimulate an increase in gill sulfide concentrations.
However, a second set of clams (n=3) were sulfur-starved (in SW+AIF) at the
same time as the flow-through experiments in SW with IF were performed. Gill sulfide
concentrations from this second set of sulfur-starved clams were slightly (though not
significantly) higher than gill sulfide concentrations of clams from the flow-through
experiment conducted in SFSW without IF. Data from the two sulfur starvation
experiments (n=3 for each experiment) in SW+AIF are pooled in Table 1 and Figure 1. In
summation, after five days of incubation in SW+AIF, nine hours of flow-through
incubation in seawater without IF did not appear to be long enough to bring about the
32
expected increase in sulfide concentrations in the gills (as originally hypothesized based on
data from the first set of sulfur-starved clams).
The distribution of sulfite, thiosulfate and sulfide between the hemolymph and the gill
Following the flow-through experiments, there was significantly more thiosulfate
in the hemolymph than in the gill; in contrast the data show significantly more sulfite and
sulfide in the gill than in the hemolymph (Table 2, Table 3, Fig. 2, Fig. 3). A paired t-test
was conducted to compare the concentrations of each sulfur compound between the
hemolymph and the gill (Table 4). Data for each sulfur compound were pooled across the
six flow-through treatments (different external [H2S] treatments, presence or absence of
IF). These results suggest a distinctly different distribution of thiosulfate in the
hemolymph versus the gill from that of sulfide and sulfite (Table 4).
Table 4. The distribution of sulfite (S032), thiosulfate (S2032), and sulfide (H2S)
between the hemolymph and the gill for Lucina floridana pooled across all flowthrough treatments. Values represent mean ± S.E.M. P-values: sulfur compound in the
gill vs. sulfur compound in the hemolymph.
Tissue Type
Hemolymph
n=35
Gill
n=35
ANOVA
P-value
5032- (tiM)
9.03 ± 1.65
50.51 ± 5.75
P << 0.001
52032- (pM)
99.39 ± 26.70
24.83 ± 4.35
P << 0.001
12.78 ± 1.79
184.44 ± 25.86
P << 0.001
H2S (PM)
33
Succinate and aspartate concentrations in gills of anaerobic and aerobic Lucina
floridana: metabolic poise controls
Table 5 shows the concentrations of succinate and aspartate in the gills of Lucina
floridana from aerobic and anaerobic control treatments. Succinate was significantly
higher in anaerobic conditions than in aerobic conditions (Table 5). In contrast, aspartate
was significantly lower in anaerobic conditions than in aerobic conditions (Table 5). These
results suggest that there is an inverse relationship between gill aspartate and succinate. A
correlation plot of these data is shown in Figure 4 (Pearson correlation coefficient =
-0.596, P = 0.053 for test of significance of Pearson correlation coefficient, y = -1.37 (±
0.62)x + 0.21(± 0.04); values given as mean ± S.E.M.). A Pearson correlation coefficient
of 1 or -1 indicates a perfect correlation between two variables and a value of 0 suggests
there is no correlation between the two variables. The inverse relationship found between
gill aspartate and succinate suggests that Lucina floridana uses an anaerobic pathway
common to many marine bivalves, the glucose-aspartate co-fermentation pathway, in
which aspartate serves as a substrate and succinate serves as an endproduct or
intermediate for energy metabolism (Hochachka, 1980; DeZwaan, 1983; Kreutzer et al.,
1985).
The ratio of succinate concentration to aspartate concentration (S:A ratio) for the
control clams is shown in Table 5. The inverse relationship between aspartate and
succinate concentrations in anaerobic and aerobic control gills should be reflected in lower
S:A ratios of aerobic gills relative to anaerobic gills. The S:A ratio of aerobic control gills
is significantly lower than the S:A ratio of anaerobic control gills (Table 5). The ratio of
succinate to aspartate concentration, rather than the absolute concentrations of succinate
34
and aspartate, was used to analyze the metabolic poise in the gills of clams from the flow-
through experiments; ratio analysis provides a measure of the relative flux through the
glucose-aspartate co-fermentation pathway (i.e. anaerobic metabolism) by showing the
relative amount of aspartate to succinate in the gill, yet removes the effects of any
variation in absolute concentrations of succinate and aspartate found within and/or
between treatment groups.
Table 5. Aspartate and succinate concentrations (mM) and succinate:aspartate
ratios (S:A) in anaerobic and aerobic gills of Lucina floridana. Values represent mean
± S.E.M. P-values: aerobic vs. anaerobic conditions.
Value
Aerobic
n=5
Anaerobic
n=6
ANOVA
P-value
Succinate (mM)
0.041 ± 0.021
0.213 ± 0.031
P = 0.002
Aspartate (mM)
0.085 ± 0.023
0.030 ± 0.011
P = 0.04
S:A
0.858 ± 0.546
11.12 ± 3.02
P = 0.01
35
0.40
A
0.35
Anaerobic gill (n=6)
Aerobic gill (n=5)
A
0.30
g, 0.25
A
A
E
r 0.20
a.)
A
0
GO
C,4
Z
C4
A
0.15
A
0.10
A
A
0.05
A
A
0.00
I
0.00
0.05
A
I
0.10
0.15
0.20
Aspartate (mM)
Figure 4. Correlation plot of succinate (mM) vs. aspartate (mM) for anaerobic and
aerobic control Lucina floridana gills. Each value represents a gill from one individual.
Succinate and aspartate concentrations in gills of Lucina floridana from flow-through
experiments run in the presence of different external [112S] in seawater with and without
invertebrate feed.
Succinate and aspartate concentrations and S:A ratios of gills from the flow-
through experiments (with and without IF) are presented in Table 6. The S:A ratios in
gills from the flow-through experiments and from the control anaerobic and aerobic gills
are presented in Figure 5. The S:A ratios of gills from all flow-through treatments are
36
significantly different from the S:A ratio for aerobic control gills, suggesting that the clams
from all flow-through experiments were using anaerobic energy metabolism.
Four flow-through treatments were run in SW without IF: SFSW, 6pM H2S,
1 3pM H2S, and 40pM H2S. Two flow-through treatments were run in SW with IF:
SFSW and 2pM H2S. As was the case for the analysis of the sulfur compound data, data
from clams which had thick, creamy yellow gills upon dissection following the flow-
through experiments (indicating the presence of high levels of elemental sulfur stores) are
not presented and were excluded from the statistical analyses; this was the case for two
clams from the SFSW treatment group run in SW containing invertebrate feed and for one
clam from the 2pM H2S treatment group. The succinate and aspartate data from two
other clams, one from the SFSW (without IF) treatment group and one from the 40pM
112S treatment group, are not presented and are not included in the statistical analyses as
the concentrations of both succinate and aspartate were at the lower detection limits of the
respective assays, making it impossible to obtain an accurate S:A ratio. Additionally,
samples from two other clams, one from the SFSW (with IF) treatment group and one
from the 2pM H2S treatment group, were lost.
Statistical analyses were performed on log-transformed ratios as log-transformed
ratios met the ANOVA assumption of equal variance. An F. test for the homogeneity of
variances among sample groups was performed on untransformed and log-transformed
ratios. The variances of the untransformed ratios were significantly different (P < 0.01;
F. test) whereas the variances of the log-transformed ratios were not significantly
different (P > 0.05; Fmax test). Untransformed ratios are presented in Table 6 and Figure 5.
Comparisons among the means were performed using the Tukey test.
Table 6. Aspartate and succinate concentrations and succinate:aspartate ratios (S:A) in gills of Lucina floridana from flowthrough experiments run in the presence of different external [1-12S] in seawater with and without invertebrate feed. Values
represent mean ± S.E.M.
Seawater without
invertebrate feed
Seawater containing
invertebrate feed
Value
SFSW
n=5
2pM H2S
n=8
SFSW
n=8
6pM H2S
n=3
13pM H2S
n=4
40pM H2S
n=3
Aspartate (mM)
0.029 ± 0.007
0.046 ± 0.013
0.046 ± 0.019
0.016 ± 0.002
0.036 ± 0.012
0.028 ± 0.013
Succinate (mM)
0.101 ± .017
0.199 ± .057
0.152 ± 0.034
0.168 ± 0.063
0.295 ± 0.174
0.271 ± 0.070
S:A
7.19 ± 1.82
8.83 ± 2.61
7.55 ± 2.81
10.56 ± 4.17
7.03 ± 2.74
14.02 ± 6.57
38
30
aerobic control
SFSW + I.F.
211M H2S + I.F.
25
SFSW
b
(3)
61.tM H2S
20
"HMI 13p.M H2S
401.1.M H2S
anaerobic control
15
b
(3)
b
(6)
10
5
a
(5)
0
+ feed
- feed
Figure 5. Succinate:aspartate ratios for Lucina floridana gills from anaerobic and
aerobic control treatments and from flow-through experiments run in the presence
of different external [H2S] in seawater with and without invertebrate feed (I.F.).
Values represent mean ± S.E.M. Sample size is contained in parentheses above each bar.
Means with different letters are significantly different from each other (P < 0.05).
39
Discussion
Disproportionation of sulfur compounds
One of the most interesting findings was the high levels of sulfide found in the gills
of clams from the sulfur starvation experiments. After five days in the absence of
exogenous sulfur compounds, sulfide concentrations in the gill were high in all three
treatment groups and highest in the gills of clams sulfur-starved in SW alone (423pM;
Table 1, Fig. 1). Sulfide has a high rate of autooxidation; the half-life of sulfide in the
presence of oxygen is approximately one hour at 20-25°C (Jorgenson et al., 1978). Thus,
any sulfide present in the gills at the initiation of the sulfur starvation experiment was
likely oxidized after a short time.
Sulfur-reducing bacteria oxidize organic compounds by using the sulfate present in
seawater as an electron acceptor; sulfide is produced as a result of sulfate reduction.
Contamination of the seawater with sulfur-reducing bacteria could lead to high
concentrations of sulfide that could diffuse into the clam and contribute to the relatively
high concentrations of gill sulfide found in clams from the sulfur starvation experiments
(Table 1, Fig. 1). However, in contrast to the relatively high concentrations of sulfide
found in the gills, sulfide concentrations in the hemolymph were relatively low (less than
30pM) after only two days of sulfur starvation in SW and remained low throughout fifteen
days sulfur starvation in SW (hemolymph sulfide concentrations after fifteen days of sulfur
starvation ranged from 15-251.IM). The difference in concentration between sulfide in the
gill and in the hemolymph suggests that the source of gill sulfide was not from diffusion of
sulfide from the external medium into the gills, as sulfide produced by this process would
40
diffuse into the hemolymph as well. Sulfur-reducing bacteria are most common in anoxic,
reducing environments, though recent studies have found that some species of sulfurreducing bacteria inhabit normoxic conditions and are able to respire aerobically (Dulling
and Cypionka, 1990). It is not likely that sulfur-reducing bacteria were present in
substantial numbers in the partially oxygenated flow-through chamber.
Sulfide was also distributed preferentially in the gills relative to the hemolymph for
clams from all flow-through experiments (Table 4). Additionally, for the clams from flowthrough experiments conducted in SFSW without IF, sulfide was distributed preferentially
in the gills relative to the mantle (Fig. 2), once again arguing against external sulfide
produced by sulfur-reducing bacteria as the source of sulfide in the gills. There are several
possible sources of sulfide that was found in the gills of the control flow-through clams
(SFSW with and without IF): 1) diffusion of sulfide from the external SW due to
production by free-living sulfur-reducing bacteria; 2) metabolism of sulfur-containing
amino acids and/or 3) disproportionation of inorganic sulfur compounds by the symbiotic
bacteria in the gills. Sulfide was present in the seawater from the flow-through control
treatments (SFSW) at a concentration of 0.84iM (SFSW without IF) or 0.23pM (SFSW
with IF). However, sulfide concentrations in the gills of clams from flow-through control
treatments were 136pM (SFSW without IF) and 128pM (SFSW with IF) (Table 2, Table
3, Fig. 2, Fig. 3). These extremely low concentrations of sulfide in the mixing chamber
(from which the seawater samples were taken to measure sulfide concentrations, see
methods) cannot account for the relatively high concentrations of sulfide found in the gills,
suggesting that contamination of the seawater by sulfide-producing sulfur-reducing
bacteria is not a likely explanation of the concentrations of sulfide found in the gills.
41
Sulfide is produced as a result of catabolism of sulfur-containing amino acid in
prokaryotes and eukaryotes (Szczepkowski, 1953; Siegel, 1975; Krouse and McCready,
1979; Ferchichi et al., 1986; Mathews and van Holde, 1990; Vismann, 1991). However, it
is unlikely that even a small fraction of sulfide present in the gills of clams from the flow-
through or sulfur starvation experiments was due to catabolism of sulfur-containing amino
acids, as sulfide produced by this process is immediately incorporated into organic
compounds or further oxidized (for review, see Siegel, 1975 and Kelly, 1988; Vismann,
1991). Mantle tissue does not contain the symbionts and thus serves as a control for
concentrations of sulfide in a non-symbiotic tissue (e.g. sulfide production due to
catabolism of sulfur-containing amino acids by host cells). The mantle tissue of clams
from flow-through experiments conducted in SFSW without IF had significantly (P< 0.01)
lower concentrations of sulfide than the gill tissue of clams from the same treatment,
suggesting that the major source of sulfide in the gills was due to bacterial metabolism
(based on the assumption that the rates of catabolism of sulfur-containing amino acids are
similar between the two tissues), possibly by disproportionation of inorganic sulfur
compounds. Thus, the presence of high levels of gill sulfide found after five days of sulfur
starvation suggests that sulfide was produced in the gill during the sulfur starvation
experiments.
The term disproportionation refers to the change in redox state of either a
monoatomic or polyatomic ion: 1) a monoatomic ion in one redox state changing to two
redox states, one higher and one lower than the original redox state (Vairavamurthy et al.,
1993) or 2) the simultaneous oxidation and reduction of a polyatomic ion (Vairavamurthy
et al., 1993). Disproportionation is a type of fermentation, which is an energy producing
42
catabolic pathway that leads to no net change in oxidation state (i.e. the oxidation state of
the substrate equals the sum of the oxidation states of the products) and generally refers to
the use of an organic molecule as both an electron donor and acceptor (instead of using an
electron transport chain to transfer electrons to nitrate or oxygen) (Hochachka and
Somero, 1984; Vairavamurthy et al., 1993). However, inorganic molecules may also
serve as electron donors and acceptors, and this process has been termed inorganic
fermentation (Bak and Cypionka, 1987; Kelly, 1987). Inorganic fermentation of
incompletely oxidized sulfur compounds (such as S032", S2032", S°, etc.) occurs via
disproportionation, and like most fermentations, is associated with small free energy
changes and substrate-level phosphorylation (Bak and Cypionka, 1987; Kelly, 1987;
Kramer and Cypionka, 1989). However, unlike a true fermentation, the bacterial electron
transport system is likely involved in disproportionation of sulfur compounds to overcome
unfavorable redox potentials between substrates (Kramer and Cypionka, 1989; Fuseler
and Cypionka, 1995). Bak and Cypionka (1987) were the first to demonstrate
disproportionation of sulfur compounds using several species of free-living sulfur-reducing
bacteria cultured under anoxic conditions. These sulfur-reducing bacteria are able to
obtain energy for growth by disproportionation of sulfite (eqn. (1)) and/or thiosulfate
(eqn. (2)) according to the following reactions (Bak and Cypionka, 1987):
4 S032- + it --. 3 S042- + HS"
(1)
52032 + H2O --+ S042 + HS + it
thiosulfate --- sulfate + sulfide
(2)
sulfite
---.
sulfate + sulfide
In contrast to the exergonic sulfite and thiosulfate disproportionation reactions,
disproportionation of elemental sulfur (S°) is an endergonic reaction under standard
43
conditions (eqn. (3); Bak and Cypionka, 1987). However, several species of sulfurreducing bacteria are able to obtain energy for growth from disproportionation of
elemental sulfur under anoxic conditions by coupling the reaction to the reduction of
various metals, such as iron, manganese, or magnesium (eqn. (4); Thamdrup et al., 1993;
Lov ley and Phillips, 1994). These metals drive the reaction forward by complexing to (i.e.
removing) one of the products, sulfide, and allowing elemental sulfur disproportionation
to proceed exergonically (Thamdrup et al., 1993; Lovley and Phillips, 1994; Fuseler and
Cypionka, 1995):
4S° + 4H20
sulfur
3H2S + S042" + 2H+
sulfide sulfate
4S° + 3FeCO3 + 4H20
sulfur
S042" + 3FeS + 3HCO3" + 5H+
sulfate iron-sulfide complex
(3)
(4)
A subsequent study determined that disproportionation of elemental sulfur is an
intermediate step in the sulfide oxidation pathway used by species of sulfur-reducing
bacteria that can oxidize sulfide aerobically (Fuseler et al., 1996). Kramer and Cypionka
(1989) investigated the potential for sulfur compound disproportionation in several species
of bacteria, including one sulfur-oxidizing species, and found that the sulfur-oxidizing
species was not able to disproportionate sulfite or thiosulfate (note: they did not
investigate the potential for elemental sulfur disproportionation). The potential for sulfur
compound disproportionation in symbiotic sulfur bacteria (sulfur oxidizers) has not been
investigated.
In this study, the evidence strongly supports the conclusion that the symbionts
produced sulfide as a response to sulfur starvation (Table 1, Fig. 1). A significant finding
was the reduction in gill sulfide found in sulfur-starved clams exposed to IF, a mixture of
44
proteins, lipids, and carbohydrates (Table 1, Fig. 1). If the production of sulfide during
sulfur starvation was due to bacterial disproportionation of sulfur compounds to supply
energy for bacterial carbon fixation (with some of the fixed carbon compounds ultimately
available for host metabolism), then IF may have brought about a reduced rate of sulfur
compound disproportionation by providing organic carbon compounds directly to the clam
for use in energy metabolism. As mentioned earlier, L. floridana has a gut and feeding
appendages (albeit both are reduced in size) and is capable of conventional (heterotrophic)
feeding (Allen, 1958) and is also likely to be capable of active uptake of dissolved organic
matter across its gills, as has been demonstrated for many bivalves (Wright and Manahan,
1989).
If the bacterial symbionts of L. floridana are responsible for the disproportionation
of sulfur compounds, the question arises as to which sulfur compound(s) is(are) used for
this reaction. While sulfate is present at a concentration of 28mM in seawater and at
approximately the same concentration in the hemolymph of most marine bivalves (Sanders
and Childress, 1992), it is unavailable for use in disproportionation reactions as the sulfur
atom is in its most oxidized state (+6). Thiosulfate, sulfite and/or polythionates (Sn062-)
could conceivably serve as substrates for disproportionation reactions (Bak and Cypionka,
1987). However, after two days of sulfur starvation the levels of sulfite and thiosulfate in
the gills and hemolymph were relatively low (data not shown). The most likely candidate
for sulfur disproportionation is elemental sulfur, which is stored in the periplasmic space of
the bacterial symbionts (Vetter, 1985). Deel (1996) has shown that at least six weeks of
sulfur starvation are required for complete depletion of bacterial elemental sulfur stores in
the symbionts of L. floridana; thus, a significant store of elemental sulfur was most likely
45
present (but was not measured) in the bacteria of clams used for the sulfur starvation (and
flow-through) experiments in this study.
While disproportionation of elemental sulfur is an endergonic reaction in the
absence of metals that serve as sulfide-complexing agents (Bak and Cypionka, 1987;
Thamdrup et al., 1993; Lovely and Phillips, 1994), the reaction could conceivably be
driven towards the production of sulfide and sulfate (or sulfite) by the oxidation (i.e.
removal) of the resultant sulfide using any of the pathways for sulfide oxidation present in
sulfur-oxidizing bacteria. Thus, I suggest that disproportionation of elemental sulfur may
be responsible for the high concentrations of sulfide seen in the gills of L. floridana from
the sulfur starvation and flow-through experiments.
Several researchers have observed the production of sulfide and sulfate during
elemental sulfur oxidation by free-living sulfur-oxidizing bacteria, which suggests that
elemental sulfur mobilization may proceed by a disproportionation reaction (Taylor, 1968;
Bacon and Ingledew, 1989). However, other studies have not detected sulfide formation
during elemental sulfur oxidation by free-living sulfur-oxidizing bacteria (Oh and Suzuki,
1977; Beffa et al., 1991). The role of elemental sulfur, its oxidation and/or reduction
pathways, and its coupling to ATP synthesis in free-living, sulfur-oxidizing bacteria are
incompletely characterized and are subject to debate (Brune, 1989; Kelly, 1989; Beffa et
al., 1993; Fuseler and Cypionka, 1995; Nelson and Hagen, 1995; Kelly et al., 1997).
Sulfide production during sulfur starvation has also been observed in the symbiont-
containing gills of another marine bivalve, Calyptogena elongata, maintained under anoxic
conditions (Childress et al., 1993). The authors suggest that bacterial elemental sulfur
46
stores were the likely source of sulfide in the gills because other sulfur compounds were
present at very low concentrations (Childress et al., 1993).
The effect of different external sulfide concentrations on sulfite, thiosulfate and sulfide in
the hemolymph and the gills from flow-through experiments conducted in seawater with
and without invertebrate feed
There was no effect of external [H2S] on sulfide concentrations in the hemolymph
or the gills of clams from the flow-through experiments (Table 2, Table 3, Fig. 2, Fig 3)
There are two possible explanations (that are not mutually exclusive) for the lack of a
treatment effect: 1) disproportionation of sulfur compounds and/or 2) periodic
ventilation. First, if the bacteria can disproportionate sulfur compounds (e.g. elemental
sulfur stores) as a response to low ATP levels, then the rate of sulfide production by these
pathways is likely to be higher in gills that are not receiving an exogenous energy source
(i.e. sulfide or IF) in the flow-through treatment. In short, the sulfide found in the gills
could have been caused by more than one process, depending on the flow-through
treatment: sulfide present in the gills of clams from the control (SFSW) flow-through
treatments could have been due to disproportionation of sulfur compounds, while sulfide
present in the gills of clams exposed to H2S could have been due to diffusion of sulfide
supplied externally, or a combination of both processes.
Following this line of reasoning and based on the results seen from the sulfur
starvation experiments, one might expect that the concentrations of sulfide in the gills of
clams from the flow-through treatment without IF or exogenous H2S would be higher than
the concentrations of sulfide in clams from the flow-through treatment with IF but without
exogenous I-12S, because IF serves as an organic carbon source. However, the
47
concentrations of sulfide in the gills of clams from the two control flow-through
treatments were not significantly different from each other, and as mentioned earlier, this
suggests that, after five days of sulfur starvation in the presence of IF, nine hours in the
absence of IF was not long enough to bring about a " starvation state" and thus the
expected increase in gill sulfide.
Another explanation for the lack of a treatment effect of external [H2S] on sulfide
concentrations in the hemolymph or gills may be due to ventilatory patterns in L.
floridana. L. floridana uses a respiration strategy called periodic ventilation, in which the
clam opens its valves periodically to ventilate the gills to extract oxygen (Anderson,
1995). After the valves close, oxygen in the tissues declines and the clam switches to
anaerobic metabolism (Hochachka, 1980). Most, if not all, bivalves use a pattern of
periodic ventilation though the ventilatory rate varies among bivalves (Barnes, 1980;
Hochachka, 1980). Several studies have shown that clams decrease their ventilatory rate
(i.e. the time spent with the valves in an open position) in response to nutrient-poor and/or
hypoxic habitats, presumably to conserve energy (Bayne, 1976; Wang and Widdows,
1993; Sobral and Widdows, 1997). Anderson (1995) reported that when supplied with
exogenous sulfur compounds, sulfur-starved L. floridana significantly increased their
ventilatory rate, supporting the hypothesis that periodic ventilation is an energy
conservation strategy. Anderson (1995) also suggests that periodic ventilation may serve
as a strategy for sulfide toxicity avoidance for bivalves living in sulfidic habitats.
Anderson et al. (1987) observed that Solemya reidi exposed to sulfide concentrations
above those where net autotrophy occurs appear to decrease their ventilatory rate. Thus,
with no exogenous H2S supplied, L. floridana may decrease its ventilatory rate as an
48
energy conservation strategy, and when external [112S] approaches toxic concentrations,
L. floridana may decrease its ventilatory rate to avoid sulfide toxicity. In this study, L.
floridana may have been able to control their exposure to external sulfide by regulating
their ventilatory rate; this hypothesis could explain the lack of a treatment effect of
external [H2S] on gill and hemolymph sulfide.
Sulfide was present in the hemolymph of clams from all flow-through groups
(Table 2, Table 3, Fig. 2, Fig. 3). Anderson (1995) also reported the presence of sulfide in
the hemolymph of L. floridana in clams exposed to 40 1.1M sulfide. These results are in
contrast to those from studies with S. reidi where sulfide was present in the hemolymph
only at "high" external sulfide concentrations (> 250 l_tM) or when clams were maintained
in hypoxic conditions (Anderson et al., 1987), suggesting that L. floridana has a lower
sulfide oxidation capability than Solemya reidi.
There was no effect of external [H2S] on the concentrations of thiosulfate in the
hemolymph or the gill of clams from the flow-through experiments (Table 2, Table 3, Fig.
2, Fig. 3). This lack of a treatment effect is most likely because of the unchanging H2S
status in the gills or hemolymph regardless of sulfide exposure. That is, assuming most, if
not all, thiosulfate comes from oxidation of sulfide in the hemolymph or the gill, because
there were no significant differences among treatment groups in gill or hemolymph sulfide,
it is not surprising that there were no significant differences in gill or hemolymph
thiosulfate among treatment groups.
In contrast to the preferential distribution of sulfide in the gill, thiosulfate was
preferentially distributed in the hemolymph relative to the gills in clams from all flow-
through treatments (Table 4). Additionally, for clams from the control flow-through
49
experiments (without IF), thiosulfate was preferentially distributed in the hemolymph
relative to the mantle (Fig. 2). The primary product of sulfide autooxidation is thiosulfate
(Chen and Morris, 1972), and as L. floridana is capable of thiosulfate uptake from the
surrounding SW (Anderson, 1995), some of the thiosulfate present in the hemolymph of
the external sulfide treatment groups could have been due to uptake of thiosulfate from
the seawater. However, it is highly unlikely that the majority of thiosulfate present in the
hemolymph was due to uptake of thiosulfate from the external seawater as the
concentrations of thiosulfate in the seawater for all flow-through experiments were low
(less than 5pM for all treatments; data not shown), several orders of magnitude less than
the concentrations of thiosulfate in the hemolymph samples (Table 2, Table 3, Fig. 2, Fig.
3). Elemental sulfur stores might represent another source of sulfur compounds for
thiosulfate; however, to date there is no evidence that free-living sulfur-oxidizing bacteria
or symbiotic sulfur bacteria can convert elemental sulfur directly to thiosulfate (Nelson and
Hagen, 1995; Kelly et al., 1997).
The source of thiosulfate in the hemolymph of the flow-through clams was most
likely due to oxidation of sulfide that diffused from the external SW (for clams supplied
with exogenous sulfide) or from the gills (where it was produced via elemental sulfur
disproportionation as a response to sulfur starvation). After diffusion into the hemolymph,
sulfide could have been oxidized to thiosulfate by autooxidation or by pigment granules.
Mitochondrial sulfide oxidation is a possibility though it is unknown whether the
mitochondria of L. floridana can oxidize sulfide as happens in S. reidi (Powell and
Somero, 1985). Cary et al. (1989) suggest that another lucinid, Lucinoma aequizonata,
50
contains a sulfide oxidase enzyme in its foot (a non-symbiont containing tissue); however,
such an enzyme has never been purified.
Regardless of the mechanism(s), oxidation of sulfide to thiosulfate produces a
soluble, non-toxic, energy-rich, stable compound that may be used for energy generation
by the symbiotic bacteria. The preferential distribution of thiosulfate in the hemolymph
relative to the gills, as found in this study, is in agreement with two other studies
investigating sulfide metabolism in symbiont-containing clams (Anderson et al., 1987;
Cary et al., 1989). In S. reidi, under normoxic conditions, thiosulfate increased to and
was maintained near 300pM in the hemolymph over a twenty-four hour period in the
presence of low levels of external [H2S] (Anderson et al., 1987). As the external [H2S]
increased and became inhibitory to aerobic metabolism or when oxygen tensions dropped
too low, thiosulfate concentrations in the hemolymph increased to over 1mM (Anderson
et al., 1987). Thus, under normoxic conditions and/or low concentrations of external
[H2S], thiosulfate was found to be removed from the hemolymph and further oxidized by
the bacteria (Anderson et al., 1987). Cary et al. (1989) also observed an initial increase in
thiosulfate concentrations in the hemolymph of L. aequizonata when the clams were
maintained in the presence of different external [H2S] concentrations.
The preferential distribution of thiosulfate in the hemolymph, as found in this
study, suggests that thiosulfate is likely further metabolized by the bacteria. Prokaryotes
are capable of transporting thiosulfate as well as other sulfur compounds through their
biological membranes (e.g. Sirko et al., 1995). However, it is less clear as to how
thiosulfate passes through the host membranes (e.g. cell membranes and vacuolar
membranes). Recently, several studies have demonstrated that the epithelial tissues of two
51
species of non-symbiotic marine invertebrates inhabiting sulfidic sediments are permeable
to thiosulfate (Hauschild and Grieshaber, 1995; Wietig et al., 1995). Whether passive
diffusion alone can supply enough thiosulfate to the symbionts or whether the host cells
use another mechanism (e.g. facilitated diffusion or active transport) to supply thiosulfate
to the symbionts is unknown.
In contrast to the preferential distribution of thiosulfate in the hemolymph, sulfite
was preferentially distributed in the gills of clams from all flow-through treatments (Table
4). This result agrees with the model for sulfide oxidation for Solemya reidi which
suggests that sulfite is further oxidized by the bacteria in the gill (Anderson et al., 1987), a
hypothesis that is supported by work on free-living sulfur-oxidizing bacteria. In all freeliving sulfur-oxidizing bacteria examined to date, sulfite serves as an intermediate in the
oxidation pathway of thiosulfate to sulfate (Kelly et al., 1997).
Sulfite may also be produced as a result of catabolism of sulfur containing amino
acids (Szczepkowski, 1953; Siegel, 1975; Krouse and McCready, 1979; Ferchichi et al.,
1986; Mathews and van Holde, 1990; Vismann, 1991). The concentration of sulfite in the
mantle, a non-symbiotic tissue, was lower (though not significantly) than the
concentrations of sulfite in the gills (pooled across all flow-through treatment groups),
suggesting that there was an additional source of sulfite in the gills (e.g. due to oxidation
of thiosulfate by the bacteria) besides that due to catabolism of sulfur-containing amino
acids.
52
The metabolic poise of Lucina floridana
The inverse relationship between aspartate and succinate found in the gills of clams
from the anaerobic and aerobic treatments suggests that the clams were using the glucoseaspartate co-fermentation pathway for anaerobic energy generation, an anaerobic pathway
commonly used by marine bivalves (Hochachka, 1980; DeZwaan, 1983; Kreutzer et al.,
1985). Based upon this inverse relationship, the gills of the clams from all flow-through
experiments were found to be anaerobic (Table 6, Fig. 5). It was expected that most, if
not all, of the gills from clams from the flow-through control treatments would be aerobic,
while at high external [H2S] the clams' mechanism for protection from sulfide toxicity
would have been overwhelmed and the gills would become anaerobic. This is the case for
S. reidi; at lower external [H2S]
100 .tM), the clam is able to maintain aerobic
metabolism while at higher external [H2S] aerobic metabolism is inhibited by sulfide and
anaerobic energy pathways are used (Anderson et al., 1990). However, because the gills
from even the control flow-through treatments (no exogenous sulfide added in SW with
and without IF) were anaerobic, it was impossible to determine the effect of external
[H2S] on the metabolic poise of the gills of L. floridana.
There were three primary differences between the aerobic control treatment and
the flow-through control treatments that could have contributed to the difference in the
metabolic poise of the gills found between the two treatment groups (Fig. 5). First, clams
used for the aerobic control treatments were maintained in artificial seawater (ASW)
whereas clams used for the flow-through control treatments were maintained in natural
seawater (SW). It is not likely that this contributed to the difference in metabolic poise as
ASW is similar in composition as well as in osmolarity to SW. Second, the oxygen
53
tensions differed between the two treatments: clams for the aerobic control group were
maintained in aerated seawater for twenty-four hours, while flow-through control clams
were first maintained in aerated seawater for five days (the length of the sulfur starvation
period) and then placed in approximately 45-55% 02-saturated (2.7-3.5 ml 02/L) seawater
for nine hours (the length of a flow-through experiment). The lower oxygen tension
experienced during the flow-through experiment could have been hypoxic to the clams
(i.e. below their critical p02) and induced anaerobic metabolism though it is not likely that
45-55% 02-saturated seawater was hypoxic to the clams because organisms living in
reducing sediments are usually adapted to maintain aerobic metabolism in the face of
lowered oxygen tensions (relative to the overlying seawater) by having critical p025
around 5-10% 02 saturation (Prosser, 1991). However, the difference in oxygen tensions
between the two treatments cannot be ruled out as a possible explanation for the
difference in the metabolic poise found in the gills.
The third major difference between the aerobic control treatment and the flow-
through control treatments was the feeding state of the clams. Aerobic control clams were
removed from the aquaria in a "fed" state (i.e. sulfur compounds and organic carbon
compounds were present in the mud), whereas flow-through control clams were sulfur-
starved for five days prior to the flow-through experiments. Based on the data of
Anderson (1995) showing that sulfur-starved L. floridana increase their ventilatory rates
in response to sulfur compounds, it is quite conceivable that the sulfur-starved clams had
reduced ventilatory rates and thus increased periods of valve closure (as an energy
conservation strategy) relative to the aerobic control clams. Additionally, any sulfide
produced in the gills during the five-day sulfur starvation period, possibly via
54
disproportionation of elemental sulfur, could have inhibited aerobic metabolism. Thus,
production of sulfide in the gills and a decrease in ventilatory rate, both responses to sulfur
starvation, could have had an additive effect on the initiation of anaerobic metabolic
pathways in the gills and are the most likely causes of the observed differences between
the aerobic control treatment and the flow-through control treatments.
If the clams were anaerobic at the start of the flow-through experiments due to
sulfur starvation, one might have expected that there would have been a low (i.e. less than
toxic) level of sulfide at which aerobic metabolic pathways were once again resumed; in
fact, the gills of clams from all flow-through treatments were anaerobic (Fig. 5). It could
be that L. floridana has a very low sulfide tolerance, and the lowest external [H2S] used in
the flow-through experiments, 2pM, was high enough to inhibit aerobic metabolism. The
clams were maintained in the flow-through conditions for nine hours; thus, the cumulative
effect of constant exposure to even the lowest concentrations of external [H2S] could have
inhibited aerobic metabolism. A more likely explanation is that the relatively high
concentrations of sulfide present in the gills at the start of a flow-through experiment (as a
response to sulfur starvation; Fig. 1) inhibited aerobic metabolism in the gills, which
supports the hypothesis that sulfur starvation is the most likely explanation for the
difference in metabolic poise found between the aerobic control gills and the flow-through
control gills.
The metabolic poise of the gills of clams from the flow-through experiments is
evidence that the gills of clams were anaerobic both in the presence and absence of
externally supplied sulfide (Table 6, Fig. 5). One of the five clams from the aerobic
control treatment group appeared to be in transition from aerobic to anaerobic
55
metabolism, as its S:A ratio (2.93) was within the 95% confidence interval (2.5-18.4) for
the S:A ratios from the anaerobic control treatment. Thus, it appears that the gills of L.
floridana may be using anaerobic metabolic pathways a significant portion of the time, as
might be expected due to their use of periodic ventilation. Anderson (1995) has noted
that L. floridana can survive up to forty-eight hours of anaerobiosis and suggests that
anaerobic metabolism may be an important metabolic mode for these clams.
As mentioned earlier, the glucose-aspartate co-fermentation pathway is an
anaerobic metabolic pathway commonly used by marine bivalves (Hochachka, 1980;
DeZwaan, 1983; Kreutzer et al., 1985). The metabolic poise of the bacteria was not
directly measured. While bacteria use many different metabolic pathways for anaerobic
energy generation, including glucose and aspartate co-fermentation (Atlas and Bartha,
1993), it is unknown whether free-living and symbiotic sulfur-oxidizing bacteria (i.e.
chemolithoautotrophs) use this pathway. It may appear that bacterial anaerobiosis would
be incompatible with sulfide oxidation; however, several researchers have suggested that
the symbionts of three marine invertebrates, S. reidi, Riftia pachyptila, and L.
aequizonata, can respire nitrate (Wilmot and Vetter, 1992; Hentschel and Felbeck, 1993;
Hentschel et al., 1993). It has been suggested that the symbionts of L. aequizonata are
obligate nitrate respirers, reducing nitrate even under normoxic conditions (Hentschel and
Felbeck, 1995; Hentschel et al., 1996). While nitrate respiration is an intriguing
possibility, there are several problems with these studies. First, only the symbionts of R.
pachyptila and S. reidi couple nitrate respiration to sulfide (or thiosulfate) oxidation
(Wilmot and Vetter, 1992; Hentschel and Felbeck 1993); sulfide oxidation coupled to
nitrate respiration was not detected in the symbionts of L. aequizonata (Hentschel et al.,
56
1993). Second, the concentrations of nitrate used to measure respiration were several
orders of magnitude higher than the concentrations of nitrate found in the animals' natural
habitat: 301_1M in seawater vs. 501_1M-5mM, the concentrations of nitrate used in most
studies (Wilmot and Vetter, 1992; Hentschel and Felbeck, 1993; Hentschel et al., 1993;
Hentschel and Felbeck, 1995; Hentschel et al., 1996). In the one study where a realistic
concentration of nitrate was used (30pM), a 2% increase in heat production over heat
production due to host anaerobic metabolism was observed (Hentschel et al., 1996); it is
unknown whether this is a statistically significant difference. Third, all studies have either
used isolated symbionts (Wilmot and Vetter, 1992; Hentschel and Felbeck, 1993;
Hentschel et al., 1993; Hentschel and Felbeck, 1995) or excised gills (Hentschel et al.,
1996); the nitrate concentrations experienced by the bacteria in an intact symbiosis may be
lower than the 3011M nitrate present in seawater. While the authors suggest that
partitioning of oxygen to the host tissues and nitrate to the symbionts reduces competition
for low levels of habitat oxygen (Hentschel et al., 1993), most organisms inhabiting
reducing environments are adapted to using anaerobic metabolism (and commonly use
energetically improved alternative anaerobic pathways) (Hochachka and Somero, 1984).
Thus, conversely, it is conceivable that the host may rely on anaerobic metabolism while
partitioning the available oxygen to its symbionts for aerobic oxidation of sulfide.
Comparison of data from this study with the metabolic response to sulfur in other
lucinids
There are several generalizations regarding the metabolic response to sulfur in
lucinids. First, sulfide is present in the hemolymph of L. floridana (this study and
Anderson, 1995), Myrtea spinifera (Dando et al., 1985), and L. aequizonata (Cary et al.,
57
1989), even at low (less than 201.1M) external [H2S]. The presence of sulfide in the
hemolymph may be a common feature among lucinids. The presence of sulfide in the
hemolymph is in contrast to S. reidi, where sulfide is present in the hemolymph only at
relatively high external [H2S] or in oxygen-limited conditions (Anderson et al., 1987), and
suggests that lucinids have a lower sulfide oxidation capability than S. reidi. Second,
thiosulfate appears to be a common sulfide oxidation product present in the hemolymph of
L. floridana (this study and Anderson, 1995) and L. aequizonata (Cary et al., 1989),
which suggests that thiosulfate may be an important metabolic intermediate used for
further energy generation by the symbionts as has been shown for S. reidi.
Conclusions
Sulfide was produced in the gills of L. floridana in response to sulfur starvation;
the production of sulfide in the gills may have been due to disproportionation of bacterial
elemental sulfur stores to provide energy for carbon fixation. Sulfur starvation in the
presence of an organic carbon source (IF) significantly reduced the concentration of gill
sulfide. IF may have allowed the clam to feed heterotrophically rather than relying on its
symbionts for organic carbon compounds. The metabolic response to sulfide was
investigated; sulfide was oxidized to thiosulfate as is the case for S. reidi (Anderson et al.,
1987). However, L. floridana appears to have a lower sulfide oxidation capability than S.
reidi. Thiosulfate was distributed preferentially in the hemolymph, a pattern that is in
agreement with further oxidation of thiosulfate in the gills by the symbiotic bacteria as
occurs in S. reidi (Anderson et al.; 1987). That is, the gill may serve as a thiosulfate
"sink." Sulfite appeared to be a metabolic intermediate of thiosulfate oxidation, a pattern
58
that is consistent with the metabolic pathways used by free-living sulfur-oxidizing bacteria.
This study was the first to investigate the metabolic poise of a lucinid under varying
oxygen and external [thS] conditions. These data suggest that the gills may use anaerobic
metabolism a significant portion of the time regardless of whether sulfide is present in the
environment.
59
References
Allen, J.A. 1958. On the basic form and adaptations to the habitat in the Lucinaceae
(Eulamellibranchia). Philosophical Transactions of the Royal Society of London
B 241: 421-484.
Anderson, A.E. 1995. Metabolic responses to sulfur in lucinid bivalves. American
Zoologist 35: 121-131.
Anderson, A.E., Childress, J.J., and J.A. Favuzzi. 1987. Net uptake of CO2 driven by
sulphide and thiosulphate oxidation in the bacterial symbiont-containing clam
Solemya reidi. The Journal of Experimental Biology 133: 1-31.
Anderson, A.E., Felbeck, H., and J.J. Childress. 1990. Aerobic metabolism is maintained
in animal tissues during rapid sulfide oxidation in the symbiont-containing clam
Solemya reidi. The Journal of Experimental Zoology 256: 130-134.
Anderson, A.E., and S.C. Hand. 1990. Intracellular pigment granules in the gills of
Lucina floridana catalyze the oxidation of sulfide. American Physiological Society
79.12A.
Arp, A.J., Childress, J.J., and R.D. Vetter. 1987. The sulphide-binding protein in the
blood of the vestimentiferan tube-worm, Riftia pachyptila, is the extracellular
haemoglobin. The Journal of Experimental Biology 128: 139-158.
Arp, A.J., Doyle, M.L., DiCera, E., and S.J. Gill. 1990. Oxygenation properties of the
two co-ocurring hemoglobins of the tube worm Riflia pachyptila. Respiratory
Physiology 80: 323-334.
Atlas, R.M., and R. Bartha. 1993. Microbial Ecology: Fundamentals and applications
Benjamin/Cummings Publishing Company, Redwood City. 563 pp.
Bacon, M., and W.J. Ingledew. 1989. The reductive reactions of Thiobacillus
ferrooxidans on sulphur and selenium. FEMS Microbiology Letters 58: 189-194.
Bagarinao, T., and R.D. Vetter. 1990. Oxidative detoxification of sulfide by mitochondria
of the California killifish Fundulus parvapinnis and the speckled sanddab
Citarichthys stigmaeus. Journal of Comparative Physiology B 160: 519-527.
Bak, F., and H. Cypionka. 1987. A novel type of energy metabolism involving
fermentation of inorganic sulphur compounds. Nature 326: 891-892.
Barnes, R.D. 1980. Invertebrate Zoology. 4th ed. Saunders College, Philadelphia. 1089
PP.
60
Bartholomew, T.C., Powell, G.M., Dodgson, K.S., and C.G. Curtis. 1980. Oxidation of
sodium sulphide by rat liver, lungs, and kidney. Biochemical Pharmacology 29:
2431-2437.
Bayne, B.L. 1976. Marine mussels: their ecology and physiology. Cambridge University
Press, Cambridge. 413pp.
Beffa, T., Berczy, M., and M. Aragno. 1991. Chemolithoautotrophic growth on elemental
sulfur (S°) and respiratory oxidation of S° by Thiobacillus versutus and another
sulfur- oxidizing bacterium. FEMS Microbiology Letters 84: 285-290.
Beffa, T., Fischer, C., and M. Aragno. 1993. Growth and respiratory oxidation of
reduced sulfur compounds by intact cells of Thiobacillus novellus (type strain)
grown on thiosulfate. Current Microbiology 26: 323-326.
Bergmeyer, H.U., ed. 1984. Methods of enzymatic analysis, Vol. VII and VIII. Verlag
Chemie, Weinheim.
Boetius, A., and H. Felbeck. 1995. Digestive enzymes in marine invertebrates from
hydrothermal vents and other reducing environments. Marine Biology 122: 105113 .
Britton, J.C., Jr. 1970. The Lucinidae (Mollusca:Bivalvia) of the western Atlantic ocean.
Ph.D. dissertation. George Washington University.
Brune, E. 1989. Sulfur oxidation by phototrophic bacteria. Biochimica et Biophysica
Acta 975: 189-221.
Carrico, R.J., Blumberg, W.E., and J. Peisach. 1978. The reversible binding of oxygen to
sulfhemoglobin. Journal of Biological Chemistry 253: 7212-7219.
Cary, S.C., Vetter, R.D., and H. Felbeck. 1989. Habitat characterization and nutritional
strategies of the endosymbiont-bearing bivalve Lucinoma aequizonata. Marine
Ecology Progress Series 55: 31-45.
Cavanaugh, C.M., Gardiner, S.L., Jones, M.L., Jannasch, H.W., and J.B. Waterbury.
1981. Prokaryotic cells in the hydrothermal vent tube worm, Riftia pachyptila:
possible chemoautotrophic symbionts. Science 213: 340-343.
Chen, K.Y., and J.C. Morris. 1972. Kinetics of oxidation of aqueous sulfide by 02.
Environmental Science and Technology 6: 529-537.
61
Childress, J.J., and W. Lowell. 1982. The abundance of a sulfide-oxidizing symbiosis (the
clam Solemya reidi) in relation to interstitial water chemistry. American Zoologist
63: 45A.
Childress, J.J., Fisher, C.R., Favuzzi, J.A., Kochevar, R., Sanders, N.K., and A.M. Alayse.
1991a. Sulfide-driven autotrophic balance in the bacterial symbiont-containing
hydrothermal vent tubeworm, Riftia pachyptila, Jones. Biological Bulletin 180:
135-153.
Childress, J.J., Fisher, C.R., Favuzzi, J.A., and N.K. Sanders. 1991b. Sulfide and carbon
dioxide uptake by the hydrothermal vent clam, Calyptogena magnifica and its
chemoautotrophic symbionts. Physiological Zoology 64: 1444-1470.
Childress, J.J., and C.R. Fisher. 1992. The biology of hydrothermal vent animals:
physiology, biochemistry, and autotrophic symbioses. Oceanography and Marine
Biology Annual Reviews 30: 337-441.
Childress, J.J., Fisher, C.R., Favuzzi, J.A., Arp, A.J., and D.R. Oros. 1993. The role of a
zinc-based, serum-borne sulphide-binding component in the uptake and transport
of dissolved sulphide by the chemoautotrophic symbiont-containing clam
Calyptogena elongata. The Journal of Experimental Biology 179: 131-158.
Conway, N., and J. McDowell-Capuzzo. 1991. Incorporation and utilization of bacterial
lipids in the Solemya velum symbiosis. Marine Biology 108: 277-291.
Dando, P.R., Southward, A.J., Southward, E.C., Terwilliger, N.B., and R.C. Terwilliger.
1985. Sulphur-oxidizing bacteria and hemoglobin in gills of the bivalve mollusc
Myrtea spinifera. Marine Ecology Progress Series 23: 85-98.
Day, R.W., and G.P. Quinn. 1989. Comparisons of treatments after an anlysis of variance
in ecology. Ecological Monographs 59: 433-463.
Deel, S.E. 1996. Characterization of bacteriocytes from two marine bivalves that contain
sulfur-oxidizing symbiotic bacteria. Non-Thesis M.S. Final Report. Oregon State
University.
DeZwaan, A. 1983. Carbohydrate catabolism in bivalves. In: The Mollusca: Metabolic
Biochemistry and Molecular Biomechanics Vol. 1. (P.W. Hochachka, ed.) pp.
137-175. Academic Press, New York.
Dilling, W., and H. Cypionka. 1990. Aerobic respiration in sulfate-reducing bacteria.
FEMS Microbiology Letters 71:123-128.
62
Doe ller, J.E., Kraus, D.W., Colacino, J.M., and J.B. Wittenberg. 1988. Gill hemoglobin
may deliver sulfide to bacterial symbionts of Solemya velum. Biological Bulletin.
175: 388-396.
Fahey, R.C., and G.L. Newton. 1987. Determination of low molecular-weight thiols using
monobromobimane fluorescent labeling and high-performance liquid
chromatography. Methods in Enzymology 143: 85-96.
Felbeck, H. 1981. Chemoautotrophic potential of the hydrothermal vent tubeworm, Riftia
pachyptila Jones (Vestimentifera). Science 213: 336-340.
Ferchichi, M., Hemme, D., and C. Bouillane. 1986. Influence of oxygen and pH on
methanethiol production from L-methionine by Brevibacterium linens CNRZ 918.
Applied and Environmental Microbiology 51: 725-729.
Fisher, C.R. 1990. Chemoautotrophic and methanotrophic symbioses in marine
invertebrates. Critical Reviews in Aquatic Sciences 2: 399-436.
Fisher, M.R., and S.C. Hand. 1984. Chemoautotrophic symbionts in the bivalve Lucina
floridana from seagrass beds. Biological Bulletin 167: 445-459.
Fisher, C.R., and J.J. Childress. 1986. Translocation of fixed carbon from symbiotic
bacteria to host tissues in the gutless bivalve, Solemya reidi. Marine Biology 93:
59-68.
Fuseler, K. and H. Cypionka. 1995. Elemental sulfur as an intermediate of sulfide
oxidation with oxygen by Desulfobulbus propionicus. Archives of Microbiology
164: 104-109.
Fuseler, K., Krekeler, D., Sydow, U., and H. Cypionka. 1996. A common pathway of
sulfide oxidation by sulfate-reducing bacteria. FEMS Microbiology Letters 144:
129-134.
Hand, S.C. 1987. Trophosome ultrastructure and the characterization of isolated
bacteriocytes from invertebrate-sulfur bacteria symbioses. Biological Bulletin 173:
260-272.
Hauschild, K., and M.K. Grieshaber. 1995. How does the lugworm, Arenicola marina,
get rid of thiosulfate, its main product of sulfide detoxification? Physiological
Zoology 68: 133A.
Hentschel, U., and H. Felbeck. 1993. Nitrate respiration in the hydrothermal vent
tubeworm Riftia pachyptila. Nature 366: 338-340.
63
Hentschel, U., Cary, S.C., and H. Felbeck. 1993. Nitrate respiration in chemoautotrophic
symbionts of the bivalve Lucinoma aequizonata. Marine Ecology Progress Series
94: 35-41.
Hentschel, U., and H. Felbeck. 1995. Nitrate respiration in chemoautotrophic symbionts
of the bivalve Lucinoma aequizonata is not regulated by oxygen. Applied and
Environmental Microbiology 61: 1630-1633.
Hochachka, P.W. 1980. Coupled glucose and amino acid catabolism in bivalve molluscs.
In: Living without oxygen: Closed and Open systems in Hypoxia Tolerance. pp.
27-41. Harvard University Press, Cambridge.
Hochachka, P.W., and G.N. Somero. 1984. Biochemical Adaptation. Princeton
University Press, Princeton. 537 pp.
Janssen, P.H., Schuhmann, A., Bak, F., and W. Liesack. 1996. Disproportionation of
inorganic sulfur compounds by the sulfate-reducing bacterium Desulfocapsa
thiozymogenes gen. nov., sp. nov. Archives of Microbiology 166: 184-192.
Javor, B.J., Wilmot, D.B., and R.D. Vetter. 1989. pH-dependent metabolism of
thiosulfate and sulfur globules in the chemolithotrophic marine bacterium
Thiomicrospira crunogena. Archives of Microbiology 154: 231-238.
Jorgenson, B.B. 1982. Mineralization of organic matter in the sea bed-the role of sulphate
reduction. Nature 296: 643-645.
Jorgenson, B.B., Kuenen, J.G., and Y. Cohen. 1978. Microbial transformations of sulfur
compounds in a stratified lake (Solar lake, Sinai). Limnology and Oceanography
24: 799-822.
Kelly, D.P. 1985. Crossroads for archaebacteria. Nature 313: 734.
Kelly, D.P. 1987. Sulphur bacteria first again. Nature 326: 830.
Kelly, D.P. 1988. Oxidation of sulphur compounds. In: The Nitrogen and Sulfur Cycles
(C. Cole, ed.) pp. 65-98. Cambridge Press, Cambridge.
Kelly, D.P. 1989. Physiology and biochemistry of unicellular sulfur bacteria. In: Biology
of Autotrophic Bacteria (H.G. Schlegel and B. Bowien, eds.) pp. 193-217. Science
Tech Publishers, Madison.
Kelly, D.P., Shergill, J.K., Lu, W.P. and A.P. Wood. 1997. Oxidative metabolism of
inorganic sulfur compounds by bacteria. Antonie van Leeuwenhoek 71: 95-107.
64
Kramer, M., and H. Cypionka. 1989. Sulfate formation via ATP sulfurylase in
thiosulfate- and sulfite-disproportionating bacteria. Archives of Microbiology 151:
232-237.
Kraus, D.W., and J.B. Wittenberg. 1990. Hemoglobins of the Lucina pectinatalbacteria
symbiosis. I. Molecular properties, kinetics, and equilibria of reactions with
ligands. Journal of Biological Chemistry 265: 16043-16053.
Kraus, D.W., Doeller, J.E., and J.B. Wittenberg. 1992. Hydrogen sulfide reduction of
symbiont cytochrome c552 in gills of Solemya reidi (Mollusca). Biological Bulletin
182: 435-443.
Kraus, D.W., Doeller, J.E., and C. Powell. 1996. Sulfide may directly modify cytoplasmic
hemoglobin deoxygenation in Solemya reidi gills. The Journal of Experimental
Biology 199: 1343-1352.
Kreutzer, U.B., Siegmund, B., and M.K. Grieshaber. 1985. Role of coupled substrates
and alternative endproducts during hypoxia tolerance in marine invertebrates.
Molecular Physiology 8: 371-392.
Krouse, H.R., and R.G. McCready. 1979. Biogeochemical cycling of sulfur. In:
Biogeochemical Cycling of the Mineral-Forming Elements (P.A. Trudinger, ed.)
pp. 401-430. Elsevier, Amsterdam.
Lovely, D.K., and E.J.P. Phillips. 1994. Novel processes for anaerobic sulfate production
from elemental sulfur by sulfate-reducing bacteria. Applied and Environmental
Microbiology 60: 2394-2399.
Mathews, C.K., and K.E. van Holde. 1990 Biochemistry. Benjamin/Cummings Publishing
Company, Redwood City.
Millero, F.J. 1986. The thermodynamics and kinetics of the hydrogen-sulfide system in
natural waters. Marine Chemistry 18: 121-147.
National Research Council, Division of Medical Science, Subcommittee on Hydrogen
Sulfide. 1979. Hydrogen Sulfide. University Park Press, Baltimore.
Nelson, D.C., Jorgenson, B.B., and N.P. Revsbach. 1986. Growth pattern and yield of a
chemoautotrophic Beggiatoa sp. in oxygen-sulfide microgradients. Applied and
Environmental Microbiology 52: 225-233.
Nelson, D.C., and K.D. Hagen. 1995. Physiology and biochemistry of symbiotic and freeliving chemoautotrophic sulfur bacteria. American Zoologist 35: 91-101.
65
Nicholls, P. 1975. The effect of sulphide on cytochrome aa3 isosteric and allosteric shifts
of the reduced alpha-peak. Biochemica et Biophysica Acta 396: 24-35.
O'Brien, J., and R.D. Vetter. 1990. Production of thiosulphate during sulphide oxidation
by mitochondria of the symbiont-containing bivalve Solemya reidi. The Journal of
Experimental Biology 149: 133-148.
Oh, J.K., and I. Suzuki. 1977. Isolation and characterization of a membrane-associated
thiosulphate-oxidizing system of Thiobacillus novellus. Journal of General
Microbiology 99: 397-412.
Powell, M.A., and G.N. Somero. 1985. Sulfide oxidation occurs in the animal tissue of
the gutless clam, Solemya reidi. Biological Bulletin 169: 164-181.
Powell, M.A., and G.N. Somero. 1986. Hydrogen sulfide oxidation is coupled to
oxidative phosphorylation in mitochondria of Solemya reidi. Science 233: 563566.
Powell, M.A., and A.J. Arp. 1989. Hydrogen sulfide oxidation by abundant
nonhemoglobin heme compounds in marine invertebrates from sulfide-rich
habitats. The Journal of Experimental Zoology 249: 121-132.
Prosser, C.L. 1991. Environmental and Metabolic Animal Physiology. Wiley-Liss, New
York. 578 pp.
Sanders, N.K., and J.J. Childress. 1992. The use of single ion chromatography to measure
ion concentrations in invertebrate body fluids. Comparative Biochemistry and
Physiology A 98: 97-100.
Segerer, A., Stetter, K.O., and F. Klink. 1985. Two contrary modes of chemolithotrophy
in the same archaebacterium. Nature 313: 787-789.
Siegel, L.M. 1975. Oxidation of reduced sulfur compounds in animals. In: Metabolic
pathways: metabolism of sulfur compounds Vol. 3 (3rd ed.) (D.M. Greenberg,
ed.) pp. 275-286. Academic Press, New York.
Sirko, A., Zatyka, M., Shadowy, E., and D. Hulanicka. 1995. Sulfate and thiosulfate
transport in Escherichia coli K-12: evidence for a functional overlapping of
sulfate- and thiosulfate- binding proteins. Journal of Bacteriology 177: 4134-4136.
Sobral, P., and J. Widdows. 1997. Influence of hypoxia and anoxia on the physiological
responses of the clam Ruditapes decussates from southern Portugal. Marine
Biology 127: 455-461.
66
Sokal, R.R., and F.J. Rohlf. 1995. Biometry 3rd. ed. W.H. Freeeman and Company, New
York. 887 pp.
Somero, G.N., Childress, J.J., and A.E. Anderson. 1989. Transport, metabolism and
detoxification of hydrogen sulfide in animals from sulfide-rich marine
environments. Critical Reviews in Aquatic Sciences 1: 591-614.
Szczepkowski, T. 1958. Cysteine. Nature 182: 934-935.
Taylor, B.F. 1968. Oxidation of elemental sulfur by an enzyme system from Thiobacillus
neapolitanus. Biochimica et Biophysica acta 170: 112-122.
Thamdrup, B., Finster, K., Wurgler-Hansen, P., and F. Bak. 1993. Bacterial
disproportionation of elemental sulfur coupled to chemical reduction of iron or
manganese. Applied and Environmental Microbiology 59: 101-108.
Tschischka, K., and R. Oeschger. 1995. Mitochondrial sulphide oxidation in selected
marine invertebrates. Physiological Zoology 68: 135A.
Vairavamurthy, A., Manowitz, B, Luther III, G.W., and Y. Jeon. 1993. Oxidation state of
sulfur in thiosulfate and implications for anaerobic energy metabolism.
Geochemica et Cosmochimica Acta 57: 1619-1623.
Vetter, R.D. 1985. Elemental sulfur in the gills of three species of clams containing
chemoautotrophic symbiotic bacteria: a possible inorganic energy storage
compound. Marine Biology 88: 33-42.
Vetter, R.D., Matrai, P.A., Javor, B., and J. O'Brien. 1989. Reduced sulfur compounds in
the marine environment: analysis by high-performance liquid chromatography. In:
Biogenic sulfur in the environment (E. Saltzman and W. Cooper, eds.) pp. 243261. ACS Symposium Ser. No. 393. American Chemical Society, Washington,
D.C.
Vismann, B. 1991. Sulfide tolerance: Physiological mechanisms and ecological
implications. Ophelia 34: 1-27.
Volkel, S., and M.K. Grieshaber. 1995. Mitochondrial sulfide oxidation in Arenicola
marina- evidence for an alternative terminal oxidase. Physiological Zoology 68:
132A.
Wang, W.X., and J. Widdows. 1993. Calorimetric studies on the energy metabolism of an
infaunal bivalve, Abra tenuis, under normoxia, hypoxia and anoxia. Marine
Biology 116: 73-79.
67
Wietig, S.L., Julian, D., Belisle, W., Howard, J., and A.J. Arp. 1995. Thiosulfate
permeability of the epithelial tissues of Urechis caupo. American Zoologist 35:
143A.
Wilmot, D.B., and R.D. Vetter. 1992. Oxygen- and nitrogen-dependent sulfur metabolism
in the thiotrophic clam Solemya reidi. Biological Bulletin 182: 444-453.
Wilson, D.F., and M. Erecinska. 1978. Ligands of cytochrome c oxidase. Methods in
Enzymology 53: 191-201.
Wittenberg, J.B. and D.W. Kraus. 1991. Hemoglobins of eukaryote/prokaryote
symbioses. In: Structure and Function of Invertebrate Oxygen Carriers (S.N.
Vinogradov and O.H. Kapp, eds.) pp. 323-330. Springer-Verlag, New York.
Wright, S.H., and D.T. Manahan. 1989. Integumental nutrient uptake by aquatic
organisms. Annual Reviews of Physiology 51: 585-600
Zillig, W., Yeats, S., Holz, I., Bock, A., Gropp, F., Rettenberger, M., and S. Lutz. 1985.
Plasmid-related anaerobic autotrophy of the novel archaebacterium Sulfolobus
ambivalens. Nature 313: 789-791.