Dennis C. Chang, Ph.D.
Consultant
Clarion Healthcare Consulting

Anonymous

Can you explain the genetics and evolution of skin color? Is skin color useful for grouping people into ethnic groups?

Dennis C. Chang
Consultant,
Clarion Healthcare Consulting,
former HHMI predoctoral fellow

Skin color varies tremendously across human populations (as the song goes, from “ebony” to “ivory”). As one of the most visible human genetic traits, it has become intimately tangled with the issues of racial identity.

A classification scheme based on skin color (or any other crude criterion) does not hold up to scientific scrutiny. To explain why this is so, I will answer the following three questions:

• What is the biological basis for skin color?
• How and why did skin color become so diverse?
• What does skin color have to do with racial groupings?

What Is the Biological Basis for Skin Color?

Skin Structure and Cells (See Chapter 1 of Robins 1991.)

Skin is divided into multiple layers with distinct physical and cellular features. The two major subdivisions are the dermis and the epidermis. The dermis is the inner layer containing blood vessels, nerves, glands, hair follicles, and connective tissue. The epidermis is the outer layer, consists predominantly of cells called keratinocytes, and is the primary determinant of skin color. The epidermis has four layers: the stratum basale, stratum spinosum, stratum granulosum, and struatum corneum. In simple terms, the innermost layer (stratum basale) contains actively dividing cells and is thus the source of new, young skin cells. The older keratinocytes are continuously pushed outward, and in each subsequent layer, the cells get flatter and more jam-packed with the structural protein keratin. The outermost layer (stratum corneum) consists of dead, keratin-filled cells that are so tightly flattened against each other that they are often fused to one another.

Within the stratum basale, only 90 percent of the cells are keratinocytes; most of the remaining 10 percent are melanocytes, cells that produce the pigment melanin. Melanocytes possess specialized, membrane-bound organelles called melanosomes, which are responsible for synthesizing melanin. As a melanosome fills up with melanin, it is transported down fingerlike projections (called dendrites) of the melanocyte and finally transferred to a neighboring keratinocyte. The long dendrites of the melanocyte allow it to contact (and thus transfer melanosomes to) a large number of surrounding keratinocytes. Skin color is (mostly) determined by the size, number, and distribution of melanosomes within keratinocytes. For example, a dark-skinned person (e.g., of African descent) is likely to have many large melanosomes (containing lots of melanin) dispersed throughout the cytoplasm of each keratinocyte, while a light-skinned person (e.g., of northern European descent) may have much smaller melanosomes (and thus less melanin) clustered together in each keratinocyte. The size of the melanosomes seems to determine whether they cluster together or remain dispersed; small melanosomes tend to cluster, while large melanosomes do not. Skin color is primarily determined by how melanosomes are made by the melanocytes. It does not depend on the number of melanocytes, which is about the same in both pale- and dark-skinned people.

Melanocytes also provide melanin to hair cells and the iris of the eyes. However, each set of melanocytes comprise distinct populations that are regulated by different pathways. In other words, the melanocytes associated with hair follicles are regulated differently from those in the epidermis. Therefore, a fair-skinned person may have dark hair and vice versa. (You probably know many fair-skinned, dark-haired people, but the “vice versa” is less common; one example is dark-skinned Australian Aborigines, who often have blonde hair in childhood that darkens at puberty.)

Melanin (See Chapter 2 of Robins 1991.)

Although human skin color is affected by many pigments, including carotene and hemoglobin, the overwhelmingly dominant pigment is melanin. Or rather, I should say the dominant pigments are melanins, since “melanin” denotes not a single chemical compound but a class of related pigments. Melanocytes produce melanins from the amino acid tyrosine using a suite of enzymes, beginning with tyrosinase, found in the melanosomes. The importance of melanin for skin coloration is illustrated by the fact that black Africans have a higher level of tyrosinase activity than white Caucasians. Moreover, individuals with a mutation in the tyrosinase gene (preventing the synthesis of melanin) display type 1, or oculocutaneous, albinism: almost complete lack of pigmentation in their skin, hair, and irises (Rees 2003).

There are primarily two types of melanin: eumelanin and pheomelanin (sometimes spelled phaeomelanin). Each is a polymer of different chemical derivatives of tyrosine. (Sometimes mixed-type melanin can form by polymerization of both eumelanin subunits and pheomelanin subunits, but this third type is relatively rare, and I will disregard it.) Eumelanin is brown to black in color, while pheomelanin is yellow to reddish-brown in color. In addition to color differences intrinsic to the pigment, eumelanin and pheomelanin also form melanosomes with different shapes and structural properties: eumelanin-dominant melanosomes tend to be larger and oblong, while pheomelanin-dominant melanosomes tend to be smaller and spherical.

In summary, melanosomes can vary in the amount of melanin and type of melanin they produce; these two factors affect the size and color of melanosomes. The size affects the clustering tendency of melanosomes in keratinocytes. The final color of the skin is affected by all these factors: the size, number, and distribution of melanosomes, as well as the type(s) of melanin contained within them.

The Melanocortin 1 Receptor (MC1R) (See Rouzaud et al. (2005) and Rees (2003, 2004).)

Like virtually everything else, melanosomes are regulated by both genes and environment. In terms of genetic influences, multiple genes are involved. I have already mentioned one of them: tyrosinase, which is more active in black Africans than in native Europeans. Other genes required for the synthesis of melanin are likely to be regulated, too, as are genes required for the transfer of melanosomes to keratinocytes. In recent years, much attention has been given to a gene that does not directly participate in melanin synthesis or melanosome transport. It is the gene that encodes the melanocortin 1 receptor (MC1R), a protein found in the cell membranes of melanocytes and involved in cell-to-cell communication.

When MC1R is active, it initiates a biochemical cascade of events inside the cell, causing the melanocyte to respond by (among other things) producing more eumelanin and less pheomelanin. Thus, anything that increases MC1R’s activity would tend to darken the skin (and/or hair; MC1R is also found on melanocytes that determine hair color). Keratinocytes can secrete chemical signals that bind to the MC1R and stimulate its activity, leading to increased eumelanin production and darker skin.

The two most important MC1R ligands are alpha-melanocyte stimulating hormone (MSH) and adrenocorticotropic hormone (ACTH). Keratinocytes are known to secrete MSH and ACTH in response to ultraviolet (UV) light; this explains tanning. However, environmental conditions (like UV light) are not alone in affecting MC1R activity. Genetics has a lot to do with it as well.

Different people may have different variants or alleles of the MC1R gene. Indeed, there are at least 65 known alleles in the human gene pool (Rees 2004). Several alleles, found most commonly among Caucasians of Celtic descent, seem to impair the function of MC1R rather severely. People in whom both copies of the MC1R gene are impaired tend to have red hair (full of pheomelanin and lacking in eumelanin) and fair skin with freckles and limited tanning ability. Interestingly, some redheads have fully functional MC1R but have a mutation in the gene that gives rise to both MSH and ACTH (it’s called POMC). In these individuals, the MC1R protein is intact, but it cannot be properly activated because its ligands are absent.

Although MC1R is not the only gene important for determining skin (and hair) color, it is clearly a major player. Explaining the variation in MC1R may go a long way toward explaining the variation in skin color in human populations.

How and Why Did Skin Color Become So Diverse?

The Origin and Spread of Modern Humans (See Wells (2002), Tishkoff and Kidd (2004), and Stringer (2002).

Before discussing how genetic variation in skin color arose, I feel that it is important to discuss how different peoples are related to each other. The 6.5 billion humans alive today are all descended from a much smaller group of people who lived in Africa. Within the past 100,000 years, some of the descendants of those ancient Africans gradually migrated around the globe, eventually becoming the “indigenous” peoples of Asia, Europe, Australia, and the Americas. This “Out of Africa” or “Recent African Origin” hypothesis has found strong support from paleontologists, anthropologists, and geneticists and is now widely accepted (Stringer 2002; Wells 2002; Tishkoff and Kidd 2004). There remains some debate over whether or not the recent African emigrants interbred with hominids (e.g., Neanderthals) who had settled Eurasia after a more ancient African origin (Stringer 2002). However, it is clear that we all share a relatively recent common ancestor.

The migration path out of Africa was first sketched out by paleontological data (bones and tools, mainly) and by cultural comparisons among living humans. More recently, genetic analysis of living human populations has helped fill in some of the details (Wells 2002), although the study of our ancient past is very much an ongoing enterprise.

The prevailing view is that people first spread from Africa to the Middle East. Some of the Middle Eastern humans seem to have spread along the southern Asian coastline, learned something of sea craft, and traveled to the Polynesian islands and then to Australia by about 50,000 years ago. Other people spread into Central Asia by about 45,000 to 40,000 years ago. Various subgroups of the Central Asian population spread and settled the four corners of the Asian continent; some of these Asians crossed over a land bridge into North America on the order of 20,000 years ago. Central Asia is also where the ancestors of modern Europeans originated; a group of Central Asians spread into Europe about 30,000 years ago, eventually replacing the Neanderthals who had lived there before them (Wells 2002).

Once humans formed settlements all over the world, the relative geographic isolation of those settlements led to some degree of reproductive isolation. Then, the differences in local climate, ecology, and culture, combined with random mutation and genetic drift, led inevitably to the evolution of ethnic diversity. The single human population that migrated out of Africa eventually became many thousands of tribes, each with slightly different gene pools. Of course, reproductive isolation was never absolute; some genetic mixing of neighbors did occur. Nonetheless, ethnic diversity clearly exists and seems strongly connected to the dispersal of humans to different parts of the globe.

So how is a white-skinned Icelander related to a tan-skinned Taiwanese? Well, the ancestors of both people lived in Central Asia about 40,000 years ago. And how are the Icelander and Taiwanese person related to a black-skinned Rwandan? They all share ancestors who lived in Africa about 100,000 years ago.

Evolutionary Pressures for Light and Dark Skin (See Jablonski and Chaplin (2000) and Chapters 4 and 11 of Robins (1991).)

It is clear that all human beings are related to each other and shared a common ancestor less than 5000 generations ago. Within that time, the genes contributing to skin color have diversified. What would account for those genetic changes?

Our current understanding of genetic change over time (also known as evolution) is that it involves both random processes (mutation and genetic drift) and the nonrandom process of natural selection. Natural selection (sometimes summarized as “survival of the fittest”) is the tendency for organisms that are better adapted to their environments to leave more offspring. Thus, genetic changes that lead to improved adaptation will tend to dominate the gene pool over many generations.

Have different skin colors become predominant in different human populations due to natural selection? To answer this question, we must consider another: What advantages and disadvantages do different skin colors provide in different environments?

One might consider a variety of functions for skin coloration, including camouflage and heat regulation (Robins 1991). However, most scientists agree that in humans, the most important consequence of different skin colors is the differential absorption of UV light. Melanin, especially eumelanin, is an excellent natural sunscreen. Dark skin, with its abundant, large, dispersed melanosomes rich in eumelanin, absorbs more UV light in its dead outer layers than light skin does, thus preventing UV light from penetrating to the living layers beneath.

Blocking UV light has certain benefits, since UV light can be damaging. UV light can cause chemical changes leading to DNA damage, which acutely leads to sunburn and, in the long term, could lead to skin cancer. (Albinos living in tropical climates invariably develop skin cancer before age 20. (Robins 1991).) UV light can also cause the breakdown of folate (or folic acid, one of the B vitamins) and other vitamins. A deficiency in folate can lead to anemia, neurological or psychiatric problems, cardiovascular disease, increased risk of certain cancers, and birth defects (Kelly 1998). All these negative effects of UV light confer a benefit to having dark, melanin-rich skin. Natural selection would therefore favor darker skin, especially in environments where UV light is intense.

However, UV light also has an important benefit: it helps in your body’s synthesis of vitamin D. Vitamin D, in turn, is important for calcium absorption, so a deficiency in vitamin D can lead to bone abnormalities, such as rickets (DeLuca 1988). Vitamin D also plays a role in immune system modulation; thus, a deficiency in vitamin D can contribute to autoimmune diseases such as multiple sclerosis, type 1 diabetes, and rheumatoid arthritis (Ponsonby et al., in press). In environments where UV light is weak, natural selection would tend to favor lighter skin, to increase the body’s ability to capture UV light and make more vitamin D. An exception that proves the rule is the case of the Inuit, who live in the North American arctic (with low UV light levels, especially in the winter) and yet possess relatively dark skin. This seems to be due to the abundance of fish in the Inuit environment and, therefore, in their diet: fish are rich in vitamin D, so the Inuit are saved from having to synthesize this nutrient using UV light.

In summary, the known physiological effects of UV light would suggest that natural selection would tend to favor a skin color (or level of skin melanization) that blocks excess UV light to protect DNA from damage and folate from destruction while allowing enough UV light through to supply the body with vitamin D (Jablonski and Chaplin 2000). Thus light skin cannot be judged to be outright “superior” to dark skin or vice versa.

The optimum skin color depends (among other things) on how much UV light is present in the environment. In fact, there is a correlation between how much UV light a region gets and the darkness of the skin of people indigenous to that region; sunny tropical climates tend to be inhabited by darker skinned people, while latitudes far from the tropics tend to be populated by lighter-skinned people (Jablonski and Chaplin 2000). It is clear that natural selection has played a major role in determining human skin colors, favoring the “best” skin color for particular environments in the ancient history of our species. Of course, in this modern era, the availability of vitamin supplements and sunscreen lotion removes much of the historical advantages or disadvantages of different skin colors.

The Evolution of Human Skin Colors (See Jablonski and Chaplin (2000).)

How then would scientists summarize the evolution of human skin color? That depends on when you begin the story. Humans are closely related to chimpanzees and bonobos (sometimes called pygmy chimpanzees); in fact, we are genetically more similar to chimpanzees than orangutans are (hence the title of the book Jared Diamond wrote about human evolution: The Third Chimpanzee (1998)).

With the current genetic and paleontological data, biologists agree that the human lineage shared a common ancestor with chimps and bonobos about 6 to 8 million years ago. That ancestor lived in Africa and was probably hairy, like modern chimps and bonobos (and various other apes). And, like today’s chimps and bonobos, our shared ancestor likely had white skin.

Given that Africa gets heavy exposure to UV light year round, why wouldn’t our ancestors (and why don’t chimps and bonobos) have dark, melanized skin to protect themselves? Well, melanized skin would be unnecessary for an animal covered in melanized hair.

After our hominid ancestors began losing this hairy coat, they quickly evolved darker skin under the pressure of natural selection in the UV-rich African environment (Harding et al. 2000; Rogers et al. 2004). Today, the genetically diverse African peoples all have dark skin. The people who migrated out of Africa and out of the tropical sun into Central Asia experienced different selective pressures and evolved a lighter skin color. Some of the Central Asian peoples migrated into the more tropical south Asia, and evolved darker skin again. Meanwhile, other Central Asians migrated into northern Europe and evolved even lighter skins.

Genetic analysis of the MC1R gene substantiates and enhances this story. Even though black African peoples are extremely genetically diverse, most (and perhaps all) make exactly the same MC1R protein (Harding et al. 2000). Meanwhile, white Europeans have an abundance of MC1R gene variants, each making a slightly different MC1R protein. As mentioned before, many of these European alleles are actually defective. It appears as though (after hominids lost thick body hair) natural selection has forced people living in Africa to keep a certain MC1R allele, one that is naturally active and therefore causes high eumelanin production. In the high-UV environment of most of Africa, any mutation making MC1R less active would have a negative effect, leaving its possessor susceptible to UV-induced DNA damage and folate depletion. People who migrated into more temperate latitudes experienced a relaxation of this constraint and an opposing selective pressure: mutations decreasing MC1R’s activity (and in the extreme case rendering MC1R nonfunctional) would have a favorable effect, allowing its possessor to make more vitamin D.

I want to point out three important points about the current view of human skin color evolution. First, skin color evolved quickly and readily as an adaptation to local environments. Second, in the course of human evolution, skin color changed multiple times; the evolution of skin color does not follow a simple progression from black to white. Finally, skin color does not track the phylogeny or relatedness among ethnic groups except to the degree that related groups share geographic proximity and, thus, similar environmental pressures.

What Does Skin Color Have to Do with Racial Groupings?

Ethnic diversity clearly exists among human populations, and genetic diversity clearly exists in humans. You may think I am simply repeating myself in the previous sentence, but ethnic and genetic diversity are not exactly overlapping terms. A small fraction of the genetic variation among all humans explains the differences among ethnic groups. However, most genetic diversity exists within individual ethnic groups. According to some scientists, the figure is 85 percent: each human population (or ethnic group) contains about 85 percent of the total genetic diversity of the human race (Wells 2002). The differences between two ethnic groups are considerably smaller.

Two ethnic groups would differ presumably because of different evolutionary histories. For example, as mentioned above, the ancestors of Icelanders and Taiwanese people parted ways about 40,000 years ago. Since that time, the descendants of those (Central Asian) ancestors experienced different chance events (e.g., mutations) and different environmental pressures (leading to natural selection of different genetic combinations). This largely explains the genetic differences between the populations of Iceland and Taiwan. However, while 40,000 years seems like a long time, it is a relatively small fraction of the time humans have been around. The Icelandic lineage may have spent 40,000 years isolated (more or less) from the Taiwanese lineage, but before the lineages diverged, they spent a much longer time together. One line of genetic evidence shows that we are all related to an African woman (named “mitochondrial Eve”) who lived about 150,000 years ago (see Wells (2002) for details). And unlike the biblical Eve, this mitochondrial Eve had a long line of ancestors herself.

The modern human lineage had at least 100,000 years (possibly much longer) to generate genetic variety before separating into the different “ethnic groups” seen today. The relatively recent divergence of human ancestors into, for example, European, Asian, and “native” American groups means that there has just been enough time to give rise to a small degree of genetic diversity. As a result, two people of different ethnic groups may be more similar than two people of the same ethnic group. Differences among ethnic groups, although real, are relatively minor when compared with differences among individuals within an ethnic group.

Skin color in particular is an exceedingly poor way of grouping people, since (as noted in the previous section) it is a trait that may change rapidly with appropriate natural selection and does not correspond well with genetic relatedness. For example, a fair-skinned Asian (say from northern China) is more closely related to darker-skinned Polynesians than to fair-skinned Scandinavians. This raises another important concern when grouping people according to genetic similarity: the groupings arrived at when comparing one set of genes may be quite different from those obtained with a different set of genes. For example, alleles related to skin color were subject to natural selection according to UV light exposure (and other factors), while other alleles may have had very different selective pressures. Therefore, there is no reason to expect a grouping of people by skin color would be the same as a grouping of people by, say, blood type or sugar metabolism.

Modern genetics suggests that the crude bases commonly used to define ethnic categories are weak, if not useless. Genetic differences among ethnic groups, if measured carefully, could be useful (e.g., if different ethnic groups respond differently, on average, to a particular medication), but roughly grouping a person into “black,” “white,” “Asian,” and “Hispanic” groups does not allow anyone to make reliable predictions about that person’s physiology or behavior.

References

With the exceptions of Harding et al. (2000), Jablonski and Chaplin (2000), and Rogers et al. (2004), the following references are reviews rather than original research papers. The interested reader may use these reviews as starting points for exploring the enormous literature on skin physiology and human evolution.

DeLuca, H.F. The vitamin D story: a collaborative effort of basic science and clinical medicine. FASEB Journal 2: 224–236, 1988.

Diamond, J. Guns, Germs, and Steel: The Fates of Human Societies. New York: W.W. Norton & Co., 1999.

Harding, R.M., Healy, E., Ray, A.J., Ellis, N.S., Flanagan, N., Todd, C., Dixon, C., Sajantila, A., Jackson, I.J., Birch-Machin, M.A., and Rees, J.L. Evidence for variable selective pressures at MC1R. American Journal of Human Genetics 66: 1351–1361, 2000.

Jablonski, N.G., and Chaplin, G. The evolution of human skin coloration. Journal of Human Evolution 39: 51–106, 2000.

Kelly, G.S. Folates: supplemental forms and therapeutic applications. Alternative Medicine Review 3: 208–220, 1998.

Ponsonby, A.-L., Lucas, R.M., and van der Mei, I.A.F. In press. UVR, vitamin D and three autoimmune diseases—multiple sclerosis, type 1 diabetes, rheumatoid arthritis. Photochemistry and Photobiology, in press.

Rees, J.L. Genetics of hair and skin color. Annual Review of Genetics. 37: 67–90, 2003.

Rees, J.L. The genetics of sun sensitivity in humans. American Journal of Human Genetics 75: 739–751, 2004.

Robins, A.H. Biological Perspectives on Human Pigmentation. U.K.: Cambridge University Press, 1991.

Rogers, A.R., Iltis, D., and Wooding, S. Genetic variation at the MC1R locus and the time since loss of human body hair. Current Anthropology 45: 105–108, 2004.

Rouzaud, F., Kadekaro, A.L., Abdel-Malek, Z.A., and Hearing, V.J. MC1R and the response of melanocytes to ultraviolet radiation. Mutation Research 571: 133–152, 2005.

Stringer, C. Modern human origins: progress and prospects. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 357: 563–579, 2002.

Tishkoff, S.A., and Kidd, K.K. Implications of biogeography of human populations for ‘race’ and medicine. Nature Genetics 36: S21–S27, 2004.

Wells, R.S. The Journey of Man. Princeton, NJ: Princeton University Press, 2002.

08/19/09 21:34