A Methanotrophic Marine Molluscan (Bivalvia, Mytilidae) Symbiosis: Mussels Fueled by Gas JAMES J. CHILDRESS, C. R. FISHER, J. M. BROOKS, M. C. KENNICUTI II, R. BIDIGARE, A. E. ANDERSON An undescribed mussel (family Mytilidae), which lives in the vicinity of hydrocarbon seeps in the Gulf of Mexico, consumes methane (the principal component of natural gas) at a high rate. The methane consumption is limited to the gills of these animals and is apparently due to the abundant intracellular bacteria found there. This demonstrates a methane-based symbiosis between an animal and intracellular bacteria. Methane consumption is dependent on the availability of oxygen and is inhibited by acetylene. The consumption of methane by these mussels is associated with a dramatic increase in oxygen consumption and carbon dioxide production. As the methane consumption of the bivalve can exceed its carbon dioxide production, the symbiosis may be able to entirely satisfy its carbon needs from methane uptake. The very light (813C = -51 to -57 per mil) stable carbon isotope ratios found in this animal support methane (813C = -45 per mil at this site) as the primary carbon source for both the mussels and their symbionts. B ACTERIAL ENDOSYMBIONTS USING reduced sulfur compounds as energy sources and fixing CO2 by means of the Calvin-Benson cycle have been implicat- ed as the major source of primary produc- tion around the deep-sea hydrothermal vents (1-3). These bacteria are found in the gills of the clams and within the trunk of the vestimentiferan tubeworms that live near the vents. Since the initial discovery, such sym- bioses have been found in a variety of other taxa (2, 4) as well as a variety of other habitats (5-7) characterized by the availabil- ity of both reduced sulfur compounds and 02. Shortly after the initial discovery of these symbioses, investigators began look- ing for symbioses based on other reduced compounds found in some of these environ- ments. We present here evidence of a meth- ane-based symbiosis between an animal and intracellular bacteria. Methane is the princi- pal component of natural gas. Childress et al. have shown CH4 con- sumption by the bacteria-containing tissue of the vent tubeworm Riftia pachyptila (8), but the intact animal does not take up CH4 (9). Other researchers have suggested on the basis of the internal membranes seen in some pogonophoran and mussel symbionts that some symbioses consume CH4 (2, 10); however, there has been no demonstration of CH4 uptake by these symbioses. On the basis of their observations of unusually light stable carbon isotope ratios, Kuhm et al. have suggested that the clams and worms of the Oregon subduction zone consume CH4 (11). However, the absence of any support- ing data on the Oregon organisms and the fact that the same species found elsewhere have sulfur-based symbioses (2) make this suggestion highly speculative at best. Arp et al. have also failed to demonstrate signifi- cant CH4 uptake in the vent clam Calypto- gena magnifica (12). Thus, the results de- Table 1. Gas exchange rates in tissue pieces from intact mussels of an undescribed species from a hydrocarbon seep off Louisiana. The tissue pieces were incubated in 20-ml glass syringes in membrane-filtered (0.45 pum) seawater. The whole animals were measured in a flowing stream of water. All gases were analyzed by gas chromatography (9). All measurements were made at 7.5°C. Gill protein content averaged 15.3% of wet weight by the Lowry method with bovine serum albumin as a standard. Numbers in parentheses are the 95% confidence intervals. 02 range CH4 range Gas exchange rates (pmol/g wet weight per hour) Conditions Tissue n (pmol/ (RmoI/ liter) liter) CO2 02 CH4 Aerobic Gill 6 90-200 +1.44(t0.35) -1.35(±0.39) Aerobic, Gill 10 90-200 44-190 +2.09(±0.38) -2.10(±0.22) -1.36(±0.23) CH4 Aerobic, Gill 3 100-210 58-169 +1.36(±0.20) -1.33(±0.43) -0.10(±0.11) CH4, acetylene Hypoxic, Gill 4 5-30 200-435 +1.05(±0.47) -.0.15(±0.08) -0.17(±0.25) CH4 Aerobic, Foot 2 130-200 110-207 +0.45, +0.94 -0.86, -0.43 +0.05, -0.11 CH4 Aerobic, Mantle 2 140-200 84-226 +0.65, +0.38 -0.54, -0.53 -0.14, +0.07 CH4 Aerobic Whole 2 120-200 +0.40, +0.34 -0.29, -0.24 animal Aerobic, Whole 2 50-150 20-205 +0.50, +0.37 -1.19, -1.20 -0.74, -0.90 CH4 animal scribed in this report were unexpected. The very first measurements and all subsequent ones have shown that these mussels con- sume CH4 at a high rate. Whole animal experiments have now been conducted on nine individual mussels. In addition, 28 separate experiments have been carried out on excised gill tissue from eight different individual mussels. In every case where 02 was not limiting and no inhibitor was being used, the rate of CH4 consumption was high, generally approaching that of 02 con- sumption. The mussels used in this study (13) were collected in two trawls near hydrocarbon seep sites on the Louisiana slope of the Gulf of Mexico (6) (27°41'N, 91032'W) at bot- tom depths of 600 to 700 m. The same trawls also retrieved vestimentiferan tube- worms (14) and pogonophorans. Immedi- ately after capture, the blood gas contents of seven vestimentiferans, which were collected in a clump at the first site, were analyzed by gas chromatography (9). Five of these ani- mals contained H2S (113, 42 and 27 ,umol/ liter and two with trace amounts), three contained CH4 (142 and 110 ,umol/liter and one with a trace amount), and two contained CO (24 and 16 ,umol/liter). Thus, this appears to be a habitat with CH4 as well as sulfide available at high concentrations. As confirmation, a piston core taken along the trawl track (27°40.5'N, 90031.6'W) showed visible oil staining and H2S. Meth- ane concentrations in the sediments can be very high since gas hydrates are present in this region (15). Our initial measurements of 02 and CH4 consumption by whole mussels showed ex- traordinarily rapid rates of initial CH4 and 02 consumption, making experimentation difficult. Therefore, we focused our efforts at sea on studying the metabolism of gill pieces (Fig. 1, A, B, and C, and Table 1). The gills were rapidly dissected free of the animals, rinsed in 0.45-p.m membrane-filtered sea- water (MFSW), cut into 0.4- to 0.5-g pieces, and kept at 7.5°C in MFSW until used. The gills' ciliary activity continued for more than 12 hours after dissection under these conditions; however, we completed all experiments within 7 hours. The separated pieces of gill were incubated at 7.5°C in 20- ml glass syringes closed with plastic valves. The syringes were filled with MFSW that had been partially decarbonated (total [CO2] =600 ,umol/liter), adjusted to pH 8.3, and equilibrated with appropriate gases. Imme- J. J. Childress, C. R. Fisher, A. E. Anderson, Oceanic Biology Group, Marine Science Institute, and Depart- ment of Biological Science, University of California, Santa Barbara, CA 93106. J. M. Brooks, M. C. Kennicutt II, R. Bidigare, Depart- ment of Oceanography, Texas A&M University, College Station, TX 77843. SCIENCE, VOL. 233 1306 on November 13, 2015 www.sciencemag.org Downloaded from on November 13, 2015 www.sciencemag.org Downloaded from on November 13, 2015 www.sciencemag.org Downloaded from