Spinach Solved Dry Eye Before Anything Had Eyes

Restasis — cyclosporine in an emulsion, twice daily — got FDA approval that year. If you respond to it at all, you might see a modest bump in tear production somewhere around month four. Allergan defended it so aggressively that in 2017 the company assigned the patents to the Saint Regis Mohawk Tribe to borrow tribal sovereign immunity against a patent challenge. The patent board rejected the maneuver. The Federal Circuit agreed. Two decades of a drug held less by its efficacy than by its paperwork.

The chemistry that beats it was in the spinach the whole time.

A team at the National University of Singapore, led by David Tai Leong in the Department of Chemical and Biomolecular Engineering with Xing Kuoran as first author, published in Cell in May. What they built is called LEAF — Light-reaction Enriched thylAkoid NADPH-Foundry, an acronym that clearly took longer to assemble than some of the experiments. It is a roughly 400-nanometer particle made of thylakoid grana harvested from spinach: the stacked membranes inside a chloroplast where light gets turned into chemical currency.

The clever part is subtractive. In an intact chloroplast, the light reactions make NADPH and the rest of the machinery immediately spends it fixing carbon. Leong's team stripped out the spending half and kept the earning half. What's left is a nanoscale factory with nothing to do but produce NADPH under ambient light — about twenty percent more of it than loose, unpackaged thylakoids manage.

That matters because dry eye disease, whatever else is driving any individual case, runs substantially on oxidative stress. Reactive oxygen species accumulate on the ocular surface, inflammation follows, the tear film degrades, and the cycle feeds itself. NADPH is what powers the cell's own antioxidant defenses. So the mechanism isn't exotic: put an NADPH generator on a surface that already receives light all day, and let the cornea's existing repair machinery do the work it was always capable of doing.

The numbers are good enough to be annoying. In tears collected from actual dry eye patients, LEAF pushed NADPH up roughly twentyfold and knocked hydrogen peroxide down by more than ninety-five percent. In inflamed corneal cells, NADPH levels recovered within thirty minutes of light exposure. In a mouse model, five days of ambient-light-activated drops restored corneal integrity to near-healthy, outperforming cyclosporine A in a head-to-head. The particles hold for two weeks at room temperature and up to a year frozen.

Now the part the press releases skip.

These are mice. Two preclinical trials, run with ophthalmologists at the Second Affiliated Hospital of Zhejiang University. Human trials are "planned" — a word that has done a great deal of load-bearing in this business. Derek Lowe, reading the paper in In the Pipeline, called the eye a more straightforward shot on goal than the other places you might try transplanting plant machinery into an animal. That is praise, and it is also a low bar.

The problem underneath is the one every topical ophthalmic drug has: things do not stay on an eye. Emulsion drops of the Restasis type clear the ocular surface in a couple of hours, and nobody has yet shown that a 400-nanometer thylakoid particle does meaningfully better. A treatment requiring reapplication before lunch is not obviously an improvement on eyedrops people already can't remember to use. As for whether photosynthetic membranes on a cornea affect what you actually see — the team's answer is that the dose stays low enough not to interfere with colour perception. That's a reassurance, not a study.

So: real mechanism, real numbers, unproven in humans, a pharmacokinetic question sitting unanswered in the middle of it. Standard.

The part worth staying with is where the answer came from. Oxygenic photosynthesis has been running the light reactions for something north of two billion years. Eyes are a Cambrian invention, roughly 540 million years old. The chemistry that fixes the eye predates the eye by a margin that makes the whole comparison feel rude. It was never hidden. It's in the salad.

Evolution ran the search — for longer than there have been animals, across more parallel trials than any pharma pipeline will ever fund — and the results have been legible to us since the 1930s. But noticing was necessary and nowhere near sufficient. Somebody still had to harvest the grana, strip out the spending half, package the remainder at four hundred nanometers, and make it survive a fortnight on a shelf. The search was free. The reading was free. The drop was engineering.

Which is precisely where the money will go, and it's worth being clear about why. Spinach is unpatentable. Two billion years of R&D sit in the public domain and nobody can fence them. What can be fenced is the thin layer of arrangement on top — the extraction method, the subtraction, the packaging, and above all whatever mucoadhesive chemistry eventually solves the residence-time problem. The commons does the healing. The wrapper collects the rent.

Set that against how this piece opened and the two halves lock together. An industry that spent twenty-four years defending an enclosure around a mediocre molecule is about to be handed a therapy derived from a commons no one owns — and its first move will be to build a fresh enclosure around the delivery mechanism. That isn't hypocrisy. It's the only move the structure permits, because the wrapper is the sole ownable part of the whole arrangement.

My prediction: LEAF stalls on residence time for three to five years while somebody works on conjugating it to a mucoadhesive polymer, a startup raises on the Cell paper, and Restasis generics keep quietly outselling all of it. Whoever solves the sticking gets the patent. The spinach gets nothing.

It's used to that.