PAX6

In one line: PAX6 is the master switch for building an eye — one gene so ancient and so universal that a mouse's copy, switched on in a fruit fly, grows a fly's eye where a leg should be.

---

The gene

PAX6 sits on the short arm of human chromosome 11, at band 11p13. It doesn't build an eye itself — it's a transcription factor, a protein that lands on DNA (through two binding domains, a paired domain and a homeodomain) and switches on the hundreds of downstream genes that actually do the building. It behaves like the topmost switch of the eye's genetic program, which is why Walter Gehring called it a master control gene.

The hunt

The strangest thing about PAX6 is that it was discovered three separate times, in three species, by people who didn't know they'd found the same gene. In the fruit fly it was eyeless — a mutant born with no eyes. In the mouse it was Small eye — shrunken eyes when one copy failed, no eyes at all when both did (Hill et al., 1991). In humans it was aniridia — children born without the coloured ring of the iris (Ton et al., 1991). In 1994, Gehring's lab showed all three were the same gene (Quiring, Walldorf, Kloter & Gehring, Science).

The mechanism

Then, in 1995, they did the opposite of breaking it. Using targeted expression, Gehring's lab forced eyeless on in tissues that should never hear it — the wing, the antenna, the leg — and complete, functional eyes grew there (Halder, Callaerts & Gehring, Science). The result that stunned biologists: the mouse Pax6 gene did it too. A mouse gene, switched on in a fly, still grew an eye.

But look closely, and those eyes are fly eyes — faceted compound eyes made of fly ommatidia, not mouse eyes. PAX6 doesn't carry the blueprint for an eye. It throws a switch, and the host's own downstream network — a cascade of many genes — builds whatever kind of eye that animal makes. The switch is universal; the eye it builds is local.

The stakes, and the frontier

In humans PAX6 is haploinsufficient: two working copies are needed, and losing the function of just one is enough to cause aniridia — an underdeveloped iris, an unfinished fovea at the centre of the retina, and reduced vision. (This is isolated aniridia, caused by mutations inside PAX6 itself. It should not be confused with WAGR syndrome, where a larger deletion of 11p13 removes PAX6 and the neighbouring WT1 gene, adding a risk of Wilms tumour.)

And the eye is only part of what PAX6 builds. The same switch is at work in the developing brain, the olfactory system (the nose), and the endocrine pancreas — where it helps specify the islet cells that manage blood sugar, including the glucagon-producing alpha cells (St-Onge et al., 1997; Sander et al., 1997). The dose matters: losing one copy leaves the iris underdeveloped, but in mice, losing both copies produces an embryo with no eyes and no nose at all (Hill et al., 1991; Grindley et al., 1995) — a reminder of how much of the body this one gene helps lay out.

The universality carries a deeper message. For most of the last century, biologists thought eyes had been invented independently dozens of times across the animal kingdom (Salvini-Plawen & Mayr estimated ~40 to 65 separate origins). One shared gene, conserved from flies to mice to us, told a different story — what's now called deep homology: a single ancient genetic program for building eyes, even though the eyes themselves (compound, camera, pinhole) each evolved their own form.

For children with aniridia, the frontier is therapeutic. Many PAX6 mutations introduce a premature "stop" in the gene; one experimental strategy is to coax the cell to read through that stop (the drug ataluren was tested in the Phase 2 STAR trial). It did not meet its primary endpoint, and gene-therapy approaches remain preclinical — the search to restore a broken copy is real, and still open.

Sources

Full claim-by-claim evidence is in references.md. Primary anchors:

• Halder, Callaerts & Gehring, Induction of ectopic eyes by targeted expression of the eyeless gene in Drosophila. Science 267:1788–1792 (1995).
• Quiring, Walldorf, Kloter & Gehring, Homology of the eyeless gene of Drosophila to the Small eye gene in mice and Aniridia in humans. Science 265:785–789 (1994).
• Ton et al., Cell 67:1059–74 (1991); Hill et al., Nature 354:522–5 (1991).
• Shubin, Tabin & Carroll, Deep homology and the origins of evolutionary novelty. Nature 457:818–823 (2009); Salvini-Plawen & Mayr, Evol Biol 10:207–263 (1977).
• GeneReviews: PAX6 Aniridia Syndrome (NBK1360); ClinicalTrials.gov NCT02647359 (STAR).

Accuracy note: "Master control gene" is Gehring's framing — PAX6 acts at the top of a network of eye-selector genes, not entirely alone. "Deep homology" means the shared thing is the genetic switch, not the eye as a single ancestral organ. And isolated aniridia (PAX6 alone) is distinct from WAGR syndrome (an 11p13 deletion removing PAX6 and WT1).