Starve a Deep-Sea Clam and Its Bacteria Switch Fuel

The limiting resource at a deep-sea methane seep is not food in any ordinary sense. Vesicomyid clams there have largely given up filter feeding; they are fed instead by sulfur-oxidising bacteria living inside the cells of their gills, and those bacteria run on hydrogen sulfide. The clam acquires it by burrowing its muscular foot into the sediment, where sulfide accumulates as microbes oxidise seeping methane without oxygen, and ferries it up to the gills while drawing oxygen from the water above. The animal is, in effect, a pipeline with a shell around it. Sulfide supply at seeps is also notoriously patchy in both space and time, which raises an obvious question that has been difficult to answer because the experiment has to happen two kilometres down.

A team working at the Haima seep in the South China Sea has now run it. Using a remotely operated vehicle, they moved clusters of clams from the sediment onto a platform about half a metre above the seabed, cutting off the foot's access to sulfide-rich pore water, and left them there for a week at one site and ten days at another. The symbionts responded by shifting their metabolism from oxidising sulfide to oxidising thiosulfate, a milder sulfur compound, while the hosts changed how aggressively they digested the bacteria in their own cells, in a way that tracked how severe the shortage was.

Why it matters: Clam beds are the visible structure of seep ecosystems, and their distribution is already known to shift as seepage waxes and wanes. Knowing what the animals do molecularly when the chemistry fails is the difference between predicting that a community will decline and understanding how much slack it has.

The study, by Lan et al. at the Hong Kong University of Science and Technology, with collaborators at the Southern Marine Science and Engineering Guangdong Laboratory, Ocean University of China and the University of Calgary, appears in Science Advances. Its most consequential methodological choice is the least glamorous: the clams were chemically fixed on the seafloor, before being brought up, because the stress of being hauled to the surface rewrites gene expression.

Two shortages, deliberately unequal

The two sites differed before anything was moved. At the more actively venting site, sediment below 12 centimetres held sulfide at 1.15 to 45.14 micromoles per litre. At the second site, where intact sediment traps methane and anaerobic oxidation generates more sulfide, the range ran from 10 to 3104 micromoles per litre. Lifting clams out of the second one therefore imposed a far steeper deprivation, and the team treats the first as moderate shortage and the second as severe rather than pretending the two are a dose series.

Symbiont abundance separated along that line. Under moderate shortage, neither metagenomic sequencing nor quantitative PCR found any change in how many bacteria the gills held. Under severe shortage, both methods found a significant decline (p = 0.036 and p = 0.004 respectively). Technical reproducibility was good, with variation across PCR replicates under 3%, so the null result at the first site is reasonably interpreted as a real absence of change rather than noise swamping a small effect.

Changing fuel, and who supplies it

Gene expression in the symbionts moved in the same direction at both sites. The whole sulfide-oxidation pathway was turned down and the thiosulfate-oxidation pathway turned up, a pattern the proteomic data corroborated for two of the proteins involved. Only 40 genes were differentially expressed in common between the two experiments, 26 down and 14 up, which is a modest overlap; the strength of the claim rests on those genes sitting coherently within sulfur metabolism rather than on their number.

The more interesting part is where the thiosulfate comes from. Clam gills carry an enzyme, thiosulfate sulfurtransferase, whose job is to convert toxic sulfide into thiosulfate, and it is expressed far more strongly in the gills than elsewhere in the animal. Detoxification waste, on this reading, doubles as a reserve fuel for the lodgers. Supporting that, the symbionts' carbon-fixation genes did not change, implying the energy budget for building sugars was being maintained rather than curtailed.

Host protein data split the two sites in a way that fits the abundance result. Under moderate shortage, proteins governing endosome maturation and fusion with lysosomes were down-regulated, consistent with the host easing off its routine digestion of its own symbionts and so holding the population steady. Under severe shortage, lysosomal proteins went up and symbiont numbers fell. Nutrient transporters moving material from bacteria to host dropped at both sites. Electron microscopy confirmed lysosomal digestion of symbionts in undisturbed clams, so this is a dial being turned, not a process switched on by the experiment.

What the study can't say yet

The design confounds its own variable. Caging a clam half a metre above the seabed denies it sulfide, but it also stops it burrowing, and the authors say outright that the responses may reflect both. Nor can the two experiments be compared directly: they ran in different years, for different durations, at different depletion intensities, and the paper restricts its conclusions to within-site comparisons, treating the site contrast as qualitative support only.

On the thiosulfate supply the authors are careful, writing that the relative contributions of host-derived and environmental thiosulfate cannot be resolved from the present data. Nor was thiosulfate itself measured; the switch is inferred from transcripts, proteins and pathway membership.

The hemoglobin result is the most provisional. Two hemoglobin subunits were strongly expressed in foot and gill blood cells and turned down after transplant, and a predicted structure of the assembled complex docked hydrosulfide slightly more favourably than oxygen. That is a computational prediction of a predicted structure, not a binding measurement. A missing pH value at the second site also leaves open secondary effects on how much sulfide was chemically available.

Quick questions

Why not just measure the sulfide inside the clam? The study did measure sulfide in sediment and bottom water, but internal pools and fluxes are much harder to capture in an animal fixed on the seafloor, which is why the metabolic shift is read off gene and protein levels instead.

Does this mean the clams survive sulfide loss? Within one to ten days they adjust. Whether they persist over months is a separate question the experiment is too short to address, and the more active site was in fact dominated by empty shells.

What is the one-line takeaway? Cut off a deep-sea clam's sulfide supply and its internal bacteria switch to a second sulfur fuel the host itself produces, while the host throttles or accelerates its digestion of them depending on how bad the shortage is.