The Science
MC1R variants: why the tanning response isn't universal
The redhead article tells the clean version: MC1R is a switch, α-MSH throws it toward protective eumelanin, and a cluster of “red-hair-colour” variants leave it stuck. That version is true and incomplete. The gene is not a switch with two positions. It is a dimmer with a few dozen stops, most of them occupied by people who do not have red hair, and a future photoprotective drug would have to work — or fail — across all of them.
This is the file on that spectrum. Not another origin story. A map of who the receptor actually is, in living populations, and what that does to the dream of a tanning drug for the people who burn.
R and r are not the same brokenness
Geneticists split the common European variants into two working bins.
R, the strong red-hair-colour alleles: R151C, R160W, D294H, and a shorter list of rarer changes (D84E, R142H). In a UK Biobank analysis, penetrance as homozygotes ran from modest to near-certain — D294H above 90 percent for red hair, R151C the most frequent strong allele in that cohort. In Australian twin families, people with two strong copies (R/R) had red hair about two-thirds of the time; a mixed R/r genotype about one time in ten; a single R against a wild-type copy almost never. Red hair is recessive. The receptor damage is not.
r, the weak alleles: V60L, V92M, R163Q. Common, mild, and easy to miss if you are only looking for redheads. V60L sits around 12 percent allele frequency in European samples. V92M, in the functional work that matched genotype to dish, produced a cAMP response as strong as wild-type — or stronger. R163Q is unremarkable in Europe and extremely common in East Asia, where it is not a red-hair allele. Calling every MC1R change “the redhead gene” is how the spectrum gets flattened into a costume.
The bins are a convenience, not a law. I155T has been filed as R and as r depending on the paper. Frameshifts and stop codons sit off the common list and simply delete the protein. The gene is unusually polymorphic in people of European ancestry; more than eighty protein- changing variants have been catalogued, most of them rare. The common nine do most of the population-level work.
Two ways to break a receptor
The phenotype looks similar. The breakage is not.
Some R alleles — R151C, R160W, D84E — are trafficking failures. The protein is made and then held inside the cell. There is less receptor on the surface for α-MSH, or for any analog, to find. In co-expression experiments those same alleles drag wild-type receptor off the surface with them: a dominant-negative effect, which is why a heterozygote is not simply “half a working copy.”
Others — D294H, R142H — reach the membrane and then fail to couple. The keyhole is there. The lock does not turn. A ligand can occupy the site and still not raise cAMP.
That distinction is not academic for a drug. An agonist cannot rescue a receptor that never arrived. It might, in principle, shout louder at a receptor that arrived and couples poorly — or it might not, if the coupling defect is downstream of occupancy. The Arizona program was aimed at MC1R. It was not aimed at the several different ways MC1R stops working.
Geography is the other axis
Strong R alleles are a northern-European pile-up. Combined R frequency has been measured falling from something like one in five alleles in Britain and Iceland, through the Low Countries and France, to under three percent in Greece. That cline is why red hair, freckling, and the whole photoprotection brief have a latitude. It is also why a trial enrolled in Arizona or Australia — high UV, European ancestry — is a different pharmacogenetic sample than a trial enrolled in Athens.
The quiet alleles have their own maps. R163Q is the extreme: a weak European r that becomes a majority allele in parts of East Asia, where it does not make red hair and does not mean the same thing for tanning as R151C means in Scotland. V60L is common in southern Europe among people who are not fair and not red. A “melanocortin tanning drug” imagined as a product for Celtic skin is already a niche. Imagined as a product for everyone who burns, it has to survive contact with genotypes the original papers barely measured.
African and East Asian populations, taken as groups, carry far fewer of the strong European R alleles. That is not the same as saying MC1R does not matter there. It is saying the European variant catalogue is not a world catalogue, and a development program that only genotypes the usual suspects will not know whom it has enrolled.
The risk is not only the hair
MC1R variants raise melanoma risk beyond what skin colour accounts for. A large French case-control study put the odds ratio at about 2.7 for R alleles and 1.5 for r alleles — and, because r is so common, the population attributable fractions came out almost even. The quiet alleles, in bulk, explain as much of the extra melanoma as the loud ones. People who carry variants and do not have red hair still show the association. Darker-pigmented Europeans are not exempt.
Part of that is the thinner eumelanin shield. Part is the receptor’s second job: MC1R signalling is wired into DNA repair in the melanocyte, so a weak allele is a weaker parasol and a slower repair crew. Cultured melanocytes homozygous or compound-heterozygous for strong R alleles respond poorly to α-MSH and are more readily killed by ultraviolet. Pheomelanin, the fallback pigment, is not just a weak shield. Under UV it is a source of oxidative chemistry of its own.
The redhead piece called this a double burden. The spectrum version is worse, because the burden is not confined to the people you can pick out of a crowd. Freckling, fair skin without red hair, a family that “just burns” — those are r and R/+ genotypes, and they are where most of the extra risk actually lives.
What this does to a future agonist
Labeled, because this part is inference from the genetics, not a trial result.
A photoprotective MC1R agonist is a ligand. It assumes a receptor. The people the original Arizona brief most wanted to protect — those who cannot tan — are enriched for genotypes in which that assumption fails.
Wild-type homozygotes would likely respond. They also tan. They are the easiest customers and the least interesting medically.
Weak-allele carriers are the open question. If the protein is on the surface and couples at all, a selective agonist might still push eumelanin. V92M, functionally near-normal in the dish, is not the same problem as R151C. A development program that does not stratify by allele will average those two stories into noise.
Strong-allele homozygotes and compound heterozygotes are the ceiling. If the receptor is trapped inside the cell, no amount of ligand in the blood finds it. If coupling is dead, occupancy is not signalling. These are the patients a marketing story would put on the box, and the patients a pharmacologist would enroll last.
Heterozygotes with a dominant-negative R copy sit in between, and they are numerous. A trial that reports “MC1R variant carriers” as a single bin is mixing people whose receptors might still work with people whose good copy is being pulled off the membrane.
None of this is an argument against the drug. It is an argument against pretending the drug is for “people who burn” as a class. Personalisation here is not a slogan. It is whether the target exists in the person you hope to treat. Afamelanotide’s EPP label never had to answer that at population scale. A general-public photoprotective would. The sponsor problem is partly an invoice. It is also a genotype table nobody has budgeted to fill.
How a tan protects still stands. Selectivity still stands. The missing sentence is that MC1R is not one target. It is a family of slightly different proteins, unequally distributed, some of which cannot hear the signal a drug would send. The redhead is the illustration. The spectrum is the constraint.