Researchers rebuilt the gene annotation of Acanthamoeba from 52 per cent accuracy to 98 per cent, uncovering roughly 300 mitochondrial proteins with no human or yeast counterpart — a shortlist of drug targets for an infection that current drugs barely touch.

Acanthamoeba is a free-living amoeba found in soil and water that occasionally finds its way into a human eye, often by way of contact lens solution. What follows can be an infection of the cornea that resists treatment and ends in blindness. The same organism can infect the brain, where the outcome is frequently death. The drugs available for both are, by the frank assessment of researchers who work on it, inadequate.

“Our current drugs are untargeted and toxic, and we just need more options for treating these infections,” said Jon Stefely of the Morgridge Institute for Research and the University of Wisconsin School of Medicine and Public Health, first author of two new studies on the organism, one of which appears in Cell.

The problem begins before treatment. “It’s a big challenge,” Stefely said. “The lack of readily available diagnostics is a problem.” An infection that cannot be identified promptly cannot be treated promptly, and corneal damage accumulates quickly.

The genome was half wrong

Behind both difficulties lies a more basic gap: nobody had a reliable parts list for the organism. The existing genome annotation — the map of where genes begin and end and what proteins they encode — was roughly 52 per cent accurate. For a research programme that depends on knowing what proteins exist in order to find ones worth targeting, a map that is wrong half the time is close to useless.

The team rebuilt it, bringing accuracy to 98 per cent and locating almost 16,000 genes. The improvement is the foundation for everything else in the work, and it is the kind of result that rarely makes headlines despite being the precondition for the headlines that follow.

With a usable annotation in hand, the comparisons became informative. About one-third of those 16,000 genes are unique to the amoeba when set against the human genome and that of baker’s yeast — a substantial reservoir of biology with no counterpart in either a patient or the standard laboratory model organism.

Why uniqueness is the point

That statistic is not a curiosity; it is the strategic core of the project. Antimicrobial drug development against eukaryotic pathogens — fungi, parasites, amoebae — runs into the problem that the target organism’s cells work much as human cells do. A drug that disables an essential protein in the pathogen is likely to disable the equivalent protein in the patient, which is why the existing treatments are, as Stefely puts it, untargeted and toxic.

“The idea,” Stefely said, “is that if you have a target protein in a biochemical pathway that’s completely unique to the microbe, it’s a better target than something that has a homolog in humans.” Proteins with no human equivalent offer the possibility of a drug that hits the amoeba and leaves the cornea alone.

The work produced two specific shortlists. The team identified 20 protein families involved in building cyst walls that are absent in humans. Cysts are the dormant, armoured form the amoeba adopts under stress, and they are a major reason the infection is so difficult to clear: a drug that kills the active form may leave cysts intact to re-emerge later. Proteins required to construct that wall are, in principle, attractive targets.

The second study, focused on the organism’s mitochondria, yielded roughly 300 proteins unique when compared with human and yeast mitochondria. “There are roughly 300 proteins in Acanthamoeba mitochondria that are unique when compared to human and yeast mitochondria,” Stefely said. Mitochondria are the cell’s energy-producing compartments, and because they are essential, drugs that disrupt them tend to be effective — the obstacle has always been that human mitochondria are similar enough to be collateral damage.

An amoeba that breathes both ways

The mitochondrial work turned up a functional finding alongside the catalogue. Acanthamoeba mitochondria can switch between oxygen-rich and oxygen-deprived modes of operation, running aerobic or anaerobic pathways depending on what is available.

That flexibility helps explain where the organism can live. The surface of the cornea is relatively oxygen-poor, particularly under a contact lens; so is the sediment of a deep lake. An organism able to reconfigure its energy metabolism according to oxygen supply can occupy both, which is a partial answer to why a soil and freshwater amoeba is capable of colonising an eye.

The comparative genomics also places the organism in deep evolutionary time. Acanthamoeba retains many proteins and pathways thought to have been present in the last common ancestor of animals, plants, fungi and protists — a cell that lived roughly two billion years ago. Lineages that have lost less are useful windows onto what early eukaryotic biology looked like, which gives the annotation value well beyond infectious disease.

What the release does not say

The work described here is foundational rather than therapeutic, and the gap between a list of unique proteins and a usable drug is wide. Of the 20 cyst-wall protein families and 300 unique mitochondrial proteins, the release does not indicate which are essential to the organism’s survival — a protein can be unique and still be dispensable, in which case blocking it achieves nothing. Nor does it report any compound screening, any inhibitor, or any test in infected tissue.

Two documentation gaps are worth noting: the release lists no funding sources and includes no conflict-of-interest statement. Its description of the publication venue is also unclear, referring to a second study appearing in “Cell Press Blue,” which does not correspond to a journal title; only one DOI is given, for the mitochondrial proteome paper in Cell.

The release credits Vamsi Mootha of the Broad Institute with conceiving MitoCarta, the human mitochondrial protein inventory that this work’s approach echoes, without describing him as an author on either study. That inventory took years of iteration before it began yielding clinical insight, which is a reasonable expectation to set here too.

---

Notes

1. Stefely, J., et al. “Mitochondrial proteome of Acanthamoeba delineates aerobic and anaerobic pathways dynamically regulated by oxygen.” Cell, 2026. DOI: 10.1016/j.cell.2026.08.056.
2. Morgridge Institute for Research. “Eye-eating amoeba resists treatment — understanding its biochemistry could help.” EurekAlert!, 1 October 2026. https://www.eurekalert.org/news-releases/1144842.