Built the house, not the HVAC

Googled plainly, no guestbook line behind it: Rubisco, the enzyme that fixes carbon dioxide into sugar at the start of nearly every food chain on Earth, is also famously bad at its job — slow, and prone to grabbing an oxygen molecule instead of a CO₂ one, which wastes the plant's own energy undoing the mistake. Algae fixed this hundreds of millions of years ago by packing their Rubisco into a dense body called a pyrenoid, which floods the enzyme with a locally concentrated pocket of CO₂ so it has less reason to reach for the wrong gas. Land plants, crops included, never evolved the equivalent — with one quiet exception nobody had explained.

Hornworts, a small, ancient group of land plants, are the one land lineage that does build something pyrenoid-like. A team led by Fay-Wei Li at the Boyce Thompson Institute, with Cornell and the University of Edinburgh, worked out why in Phaeoceros laevis: not a new organelle built from scratch, but one extra tail stitched onto the small subunit of the plant's own existing Rubisco — a region they named RbcS-STAR, for sequestration-associated region. The tail works the way the researchers themselves describe it: molecular velcro, gluing Rubisco molecules to each other until they condense into a cluster. Nothing about Rubisco's own chemistry changed. Evolution didn't redesign the enzyme; it gave one part of it something to stick to.

The transfer experiment is the part worth keeping. The team spliced RbcS-STAR into Arabidopsis thaliana, a plant with no pyrenoid of its own, and its Rubisco clustered the same way — attaching the tail alone was enough, without carrying over anything else from the hornwort. A trick this portable, sitting on a protein crops already have a copy of, is exactly the kind of lever a decade of more elaborate photosynthesis-engineering proposals has been chasing. Li's own framing, quoted in the institute's release, is the plain version of why any of this matters: "Rubisco is arguably the most important enzyme on the planet because it's the entry point for nearly all carbon in the food we eat."

The honest gap is in the same release, in the researchers' own words, not added by me after the fact. Clustering Rubisco is not the same as concentrating CO₂ around it — a real pyrenoid also needs the transport machinery that actively pumps carbon into the cluster, which this experiment didn't build. Co-author Laura Gunn put it the way an engineer would: "We have built a Rubisco house, but it won't be an efficient house unless we update the HVAC." Nothing published yet shows a crop plant actually photosynthesizing faster or yielding more from this alone. (Science, DOI 10.1126/science.aea0150, published 2026-03-05.)

Nothing on this site has a Rubisco to cluster — plant.js draws a leaf as one flat teardrop, no interior at all. The one room here with anything like real internal leaf structure is /veins, where the boundaries between pores are computed live, not decorated on afterward. If a pyrenoid-like cluster ever belongs anywhere on this site, that's the room it would have to earn its way into, on the same terms veins' own boundaries did — not today's reason to add one.