One idea, tested across all 19 genes

A preprint posted to bioRxiv this month, from the Mitochondrial Medicine Frontier Program at the Children's Hospital of Philadelphia and the University of Pennsylvania, takes one straightforward idea and tests it across an entire family of genetic conditions — FARS2 among them. The work was supported by the CureARS Foundation and the National Institutes of Health.

The idea being tested

FARS2 belongs to a group of 19 genes that all do the same kind of work. Each one makes an enzyme responsible for attaching one specific amino acid to its matching carrier molecule, so that mitochondria can build proteins. FARS2 handles phenylalanine; the other eighteen genes each handle a different amino acid.

The approach under test is called cognate amino acid supplementation. "Cognate" just means "the matching one." If the enzyme that loads phenylalanine is sluggish, does supplying more phenylalanine help it keep pace? Individual reports had hinted at benefit here and there, but nobody had tested it systematically across the whole family — so it wasn't clear whether this was a general principle or a handful of lucky exceptions.

What they did, and what happened

The researchers used C. elegans, a roundworm about a millimetre long that has been a workhorse of genetics for decades. They dialled down each of the 19 genes in turn, producing 19 separate models, then gave each model its matching amino acid and measured growth, movement, lifespan, mitochondrial function and fertility. The results:

  • Growth and movement improved in all 19 models — every single one.

  • The effect was dose-dependent: more amino acid, more improvement. That relationship is one of the standard signs that an effect is real rather than coincidental.

  • A mitochondrial stress signal quietened down. Cells run an alarm system that switches on when mitochondrial proteins aren't being made or folded properly; supplementation turned the volume down on it.

  • Fertility, which was impaired in two of the models, was restored — and one of those two was the worm's counterpart to FARS2.

Consistency is what makes this notable — not one gene responding, but all nineteen.

The authors' conclusion is that this amounts to enough preclinical evidence to justify proper clinical trials across this whole family of conditions, rather than continuing to rely on scattered individual reports.

How much weight to put on it

Worms are not people. The models were built by turning genes down rather than by reproducing specific human variants, and results in simple organisms frequently fail to carry over. Correct dosing and safety in humans still have to be established — which is precisely what the trials the authors are calling for would do. This is a reason to watch the space closely, not a reason to change anything at home. Questions about supplements belong with a clinical team.

One more note on what this is: a preprint is posted publicly before peer review, so the findings haven't yet been through outside scrutiny.

Read the preprint: https://www.biorxiv.org/content/10.64898/2026.02.03.703549v1

Next
Next

8,400 compounds, one yeast stand-in