Tucked inside a single fucoidan molecule can hide dozens of different chemical links and branching patterns — a knot so tangled that no one microbe can untie it alone. For years, that left a puzzling gap in how scientists understood the ocean's carbon cycle. Now a team led by researchers at MIT and ETH Zurich has cracked the mystery, and the answer is teamwork.
Fucoidan is a tough carbohydrate that brown algae and diatoms use to build their protective outer layers. Because it resists decay, when it sinks into the deep ocean it drags carbon down with it, storing it for long stretches. That makes it a quiet heavyweight in the ocean's carbon cycle. But exactly how it gets broken down in nature had stayed a mystery.
The new study, published in the open-access journal Nature and led by Andreas Sichert, a former MIT postdoc now at ETH Zurich, and Otto X. Cordero, an associate professor at MIT, reveals that marine bacteria split the job. "No single bacterium can finish the job," says Cordero. "Instead, fucoidan is degraded through teamwork."
The evidence was staggering. When the researchers enriched a fucoidan-eating bacterial community from coastal seawater, they found more than 453 different genes — each one coding for an enzyme that acts on fucoidan — spread across eight bacterial strains. None of those strains alone could fully break the molecule down. But using a fast new mass spectrometry method, the team watched exactly which sugar building blocks each bacterium consumed.
Out of all that genetic chaos emerged a clean pattern with just two roles. Some strains specialized in chewing up fucoidan's fucose-rich backbone, while others stripped away its side branches, which hold rarer sugars like xylose and galactose. When strains from both roles worked together, the effect wasn't just additive — it was synergistic, beating what anyone could predict from the bacteria's individual efforts. In some pairings, the communities came close to completely degrading the molecule.
The surprise didn't stop there. The division of labor made the whole system more predictable, not less. The team built a simple model that sorted bacteria into just two categories — fucose-eaters and side-chain-eaters — and trained it on tiny communities of one to three strains. That bare-bones model then predicted breakdown in communities of up to seven strains, and even held up across nine structurally different fucoidans from other algae.
"A predictive understanding of a complex system need not come from characterizing each of its parts, but from finding the right simplification," Cordero says.
The researchers also propose a concept they call "diversity-limited degradation": when the right mix of complementary specialists is missing, fucoidan lingers longer instead of being eaten. That idea may help explain why some algal carbon stays locked in the ocean for so long. The finding hints that scientists could one day predict how other complex, carbon-rich materials break down in nature — even when the exact chemistry is still a puzzle.