Friday, December 16, 2005

Race is skin deep   posted by Razib @ 12/16/2005 09:27:00 AM

A fascinating paper just came out in Science, SLC24A5, a Putative Cation Exchanger, Affects Pigmentation in Zebrafish and Humans. Heather L. Norton is one of the authors listed, so she knew very well what she was talking about when she suggested that there was far more to skin color variation than MC1R. Here is the eye popper from the text: "Based on the average pigmentation difference between European-Americans and African-Americans of about 30 melanin units (33), our results suggest that SLC24A5 explains between 25 and 38% of the European-African difference in skin melanin index." I have pointed to models before which suggest ~4-5 loci that control skin color, this work does nothing to falsify that, but, it does show that between populations the variation in coloration in humans might be due to alleles of large effect which are differentially fixed. Evaluated over the whole species it is also certainly true that various alleles account for a different percentages of the genotypic component of variation, with many small effect modifier alleles likely hovering in the background of populations which are fixed for alternative alleles of large effect. Another important point is that the authors point out that both Africans and East Asians exhibit the ancestral allelic state, while Europeans are derived (a mutant descendent form), so the implication is that light skin is generated in East Asians by alternative genetic conformations. I have already pointed out that East Asians seem to be under strong selection, and moving toward fixation for Arg163gln MC1R allele (in contrast, Europeans are highly polymorphic for this locus). This points to the reality that evolution, or precisely, selection, explores a large phenotype & genotype space with multiple fitness peaks, and the expected outcomes might be random in nature when the options are constrained to the various peaks. The title of the post comes from the author's observation that "This region [of the HapMap -R], which contains several additional SNPs with high-frequency differences between populations, was the largest contiguous autosomal region of low heterozygosity in the European (CEU) population sample...This pattern of variation is consistent with the occurrence of a selective sweep in this genomic region in a population ancestral to Europeans." The low heterozygosity seems to be due to powerful directional selection which dragged a large number of adjacent SNPs within that haplotype block to fixation.

Finally, I want to add that this is a cool paper partly because of its broad methodological scope. Instead of just scanning the HapMap the authors also confirmed the developmental genetic expression via an animal model, zebrafish, which seems to exhibit the same variation in coloration on this locus as our own species. As they say, nature works with what it's got, we've seen this with Foxp2, which shows up in derived form in humans, birds and whales (highly vocal species).

Anyway, the abstract:

Lighter variations of pigmentation in humans are associated with diminished number, size, and density of melanosomes, the pigmented organelles of melanocytes. Here we show that zebrafish golden mutants share these melanosomal changes and that golden encodes a putative cation exchanger slc24a5 (nckx5) that localizes to an intracellular membrane, likely the melanosome or its precursor. The human ortholog is highly similar in sequence and functional in zebrafish. The evolutionarily conserved ancestral allele of a human coding polymorphism predominates in African and East Asian populations. In contrast, the variant allele is nearly fixed in European populations, is associated with a substantial reduction in regional heterozygosity, and correlates with lighter skin pigmentation in admixed populations, suggesting a key role for the SLC24A5 gene in human pigmentation.

Here is an easier to digest piece in The Washington Post, Scientist s Find A DNA Change That Accounts For White Skin (the title isn't true of course, there are East Asians with white skin). And Nick Wade is on it of course. Thanks to Abhi & Theresa for the tip. You can find a PDF of the paper as "whiteskin" in the forums.

Update: NPR has a nice summary. People seem interested in this topic.

Addendum: Check out William Saletan's idiotic comment:

A single gene makes whites paler than blacks. Until 20,000 to 50,000 years ago, everyone was black; then a mutation in this gene created white people. Reactions: 1) Don't talk about racial genetics; it encourages racism. 2) If color comes down to one gene, doesn't that minimize its significance? 3) Did whiteness spread in Europe because it made people healthier, or because it made them more sexually attractive?

Saletan might have been tongue-in-cheek, but if he is he's only perpetuating public stupidity (something that doesn't need any reinforcement, thank you very much). First, less than half of the color variation even between Europeans and Africans is accounted for this gene, i.e., it is a necessary condition for "whiteness" (in the European sense), but it is not a sufficient condition. Living in an east coast metro area his model of a black-white dichotomy (i.e., the stories clearly suggest that East Asians lack the derived variant, but they aren't black to my knowledge) in any conception of race pops up (if ($race != white) { $race = black}). Second, it makes the standard conflation of skin color with populational identity. This map makes it clear that skin color tends to track latitude far more than the combination of latitude & longitude that would imply a tight correlation with geographical populations. If you step outside the black-white world skin color can not predict populational identity very well (except perhaps at the very fair end of the range where Northern Europeans are alone). And as for reaction #2, what idiots actually said anything like this? That is a contention so bizarre that it seems to have been produced by a 3rd grader (I'm expecting to falsified as to the expectation of chronological age). I wouldn't expect much from Saletan, but he writes the "human nature" column for Slate. How can I get a cushy gig like that???

Addendum II: Check out this post from a few years back on altitude adaptation to see what I mean about stumbling upon different fitness peaks via different strategies. Also, consider that many genes have pleiotropic effects, that is, they are implicated in a multitude of genetic pathways and influence many traits. There are many background assumptions operating when one says that locus x has fitness effect y on population z. Consider that perhaps light skin in the generality is a phenotype that is advantageous at higher latitudes, while dark skin is advantageous at lower latitudes. Since to the first approximation humans are creatures of Africa it makes sense that we would have fixed or constrained toward expression of genes that influence skin color toward a dark optimum. As non-Africans are a subset of Africans it seems plausible that the ancestral dark inducing forms of the genes will be shared. On the other hand, as the constraint for dark skin is released, because selection no longer favors it, the genes will start stumbling randomly in various directions via mutations. In the case of Europeans the MC1R locus seems to have walked in a random fashion and diversified greatly1 (30 alleles of greater than 1% frequency). On the hand, in East Asians the MC1R allele seems to have been selected toward one particular form that differs from the ancestral variant, in other words, constraint that limited the fitness toward those bearing the ancestral dark skin inducing allele(s) was released, but unlike Europeans selection now operated on a different allele and constrained diversification. This new finding makes that more intelligible: Europeans were given license to explore the range of MC1R variants because another locus was sufficient in hastening the induction of a light skin phenotype. Selection operates on genotype via phenotype, so the phenotype is sufficient to allow the individual to be fit and reproduce the genes, it is irrelevant what that particular conformation of genes that results in said phenotype is. An important point I am leaving out of this is that it is likely that MC1R has other fitness effects (recall the finding that redheads might be more sensitive to pain), and it is likely that this new locus is also implicated in other phenotypes, at least indirectly. In other words, loci don't explore the fitness landscape by their lonesome, but only in the context of changes and conditions on other loci.

1 - Negative frequency dependent sexual selection is another way to generate diversity. Hard to falsify though, and usually neutrality is assumed to be a legitimate null hypothesis.