Agouti signaling protein (ASIP): the natural off-switch for skin pigment

Even the system's brakes leave their mark — where ASIP holds the pigment signal back, colour turns pale and a coat turns banded.

Agouti signaling protein — ASIP — is the reason the melanocortin pigment system has an off switch as well as an on switch. Where α-MSH activates MC1R to make dark pigment, ASIP does the opposite: it is the natural blocker that shifts colour the other way. It is the original "agouti" molecule, it gives its name to a whole family of melanocortin antagonists, and it carries one of the most instructive lessons in the system — that a brake is only harmless where it belongs.^[3]

The gene behind a hundred years of mouse genetics

The agouti locus was one of the first genes in mammalian genetics, mapped by its effect on mouse coat colour decades before anyone knew what it encoded. In 1992 and 1993 two groups cloned it and found a gene for a small secreted protein of 131 amino acids, expressed in the hair follicle and epidermis exactly when yellow pigment is being laid down.^[1][2] The same work explained a mystery that had puzzled geneticists since the 1920s. The dominant lethal yellow allele makes mice uniformly yellow, obese, diabetic and tumour-prone; the cloning showed that it is a chromosomal rearrangement that splices a ubiquitous promoter onto the intact agouti coding sequence, so that a protein meant for the hair follicle is made in nearly every tissue of the body.^[2] The pigment and the obesity, it turned out, were the same molecule in two places.

Type-switching the pigment

Pigment cells can make two kinds of melanin: dark, UV-shielding eumelaninwhen MC1R is active, and light red-yellow pheomelanin when it is not. In 1994 agouti protein was shown to be what its genetics predicted: a high-affinity antagonist of the MSH receptor that blocks α-MSH from raising cyclic AMP, the signal through which the receptor drives eumelanin synthesis.^[3] In human melanocytes it does more than compete: pretreating the cells with ASIP leaves them unable to respond to α-MSH at all, which looks like receptor desensitisation rather than simple blockade, and it lowers the receptor's basal activity as an inverse agonist.^[7] The result is "pigment type-switching": the same cell, the same receptor, a different output depending on whether the accelerator (α-MSH) or the brake (ASIP) has the upper hand. A pulse of ASIP while a hair is growing produces the pale sub-terminal band of the classic agouti hair — a brief lift of the accelerator written into the fibre as it forms.^[3]

What the protein looks like

ASIP's business end is its cysteine-rich C-terminal domain, whose NMR structure was solved in 2005. It folds as an inhibitor cystine knot, a tight, disulfide-braced scaffold otherwise known from spider and cone-snail toxins; ASIP and its cousin AgRP are the only mammalian proteins in the class.^[6] The structure also held a surprise: the domain flips between two conformations around a pair of prolines, and only the minor one, with two unusual cis peptide bonds, is active at the melanocortin receptors.^[6] The N-terminal half of the protein has a different job. It is needed for ASIP to interact withattractin, a large membrane protein that acts as a low-affinity co-receptor: mice lacking attractin are immune to the yellow coat and the obesity of the lethal-yellow allele, because the agouti protein can no longer act on the cells it reaches.^[5] AgRP, by contrast, does not need attractin — one of the ways the two cousins have gone their separate ways.

Cousin to the hunger peptide

ASIP's trick — block a melanocortin receptor — was borrowed once more by the body, in the brain. AgRP (agouti-related peptide) is its cousin: it blocks MC4R and MC3R to drive appetite, exactly as ASIP blocks MC1R to lighten pigment. Same fold, same mechanism, different tissue, different job — a neat example of evolution reusing a working idea. It also explains the lethal-yellow mouse. Agouti protein antagonises MC4R as well as MC1R,^[3] and when it is made throughout the body it reaches the hypothalamus and does AgRP's job without AgRP's restraint. The 1997 experiment that nailed this down gave mice a synthetic melanocortin agonist and antagonist into the brain: the agonist suppressed feeding even in lethal-yellow mice, the antagonist blocked that and increased feeding on its own, showing that melanocortin neurons tonically hold appetite down and that the agouti obesity syndrome is what happens when that tone is chronically blocked.^[4] It was the experiment that put MC4R at the centre of body-weight biology, and ASIP was the molecule that pointed to it.

The same disease in people

For thirty years the lethal-yellow mouse looked like a laboratory curiosity. In 2022 it became a human diagnosis. A girl with extreme early-onset obesity was found to carry a tandem duplication at the ASIP locus that places the gene under the control of a ubiquitously active promoter — the human equivalent of the mouse rearrangement — so that ASIP was made in every cell type examined, including hypothalamus-like neurons derived from her own stem cells. Her phenotype matched the mouse's point for point: early obesity, overgrowth, red hair and high insulin. Because this kind of mutation escapes standard genetic screens, the authors went back through a cohort of 1,745 children with obesity and found four more with the identical duplication and the same picture.^[12] It is a monogenic cause of human obesity that no gene panel was looking for, and it rests on the oldest finding in melanocortin genetics: a brake expressed where it does not belong.

ASIP in human hair, skin and cancer risk

Humans do not have banded hairs, but ASIP still shapes how we look. A common variant in the gene's 3′-untranslated region is associated with dark hair and brown eyes,^[9] and a haplotype near ASIP, tagged by two single-nucleotide changes, is associated with red hair, freckling and skin that burns rather than tans — a spectrum of effects strikingly similar to the red-hair variants of MC1R itself.^[10] The same haplotype matters beyond appearance. In more than 40,000 Europeans it raised the risk of cutaneous melanoma by about 45% and of basal cell carcinoma by about a third, and the association held after adjusting for pigmentation, which suggests the brake on MC1R affects more than the colour of the skin it protects.^[11] The genetics page puts these variants alongside MC1R's own.

Not the only brake

ASIP is the classic MC1R antagonist but not the only natural one. Dog breeders had long known a third coat-colour locus that overrides agouti to give a dominant black coat; in 2007 it was identified as a β-defensin, a protein family previously known only from innate immunity, whose product binds MC1R with high affinity and produces black fur in dogs and in transgenic mice.^[8] Human β-defensin 3 antagonises MC1R in cultured human melanocytes in the same way ASIP does.^[7] The melanocortin receptors, it turns out, have more than one family of ligands — and more than one way of being told to stop.

The honest bottom line

ASIP completes the pigment half of the melanocortin story. You cannot fully understand MC1R, tanning, or the drugs that push pigment (like the tanning peptides) without the molecule that pushes the other way. Every melanocortin signal, it turns out, comes with a built-in counterweight — and in the skin, ASIP is it. Its deeper lesson is about place: the same protein that paints a band on a hair will, released everywhere, make a mouse or a child obese. For the receptor it governs, see MC1R — the pigment switch; for the cousin that does the same job in the brain, see AgRP.