Why do you have blue eyes?

Gene ExpressionBy Razib KhanSep 23, 2007 1:30 AM


Sign up for our email newsletter for the latest science news

Well, you may not have blue eyes, but many people do. The post below suggests that there is still a lot of confusion on how eye color is inherited, but now in 2007 we are coming close to clearing up many issues. A paper which came out early this year, A Three-Single-Nucleotide Polymorphism Haplotype in Intron 1 of OCA2 Explains Most Human Eye-Color Variation (Open Access), suggests that about

3/4 of the eye color variation in Europeans (from pale blue to dark brown) can be explained by polymorphism around the OCA2 gene

. In other words, eye color comes close to being a monogenic Mendelian trait when it comes to inheritance, but not quite.

The diagram below is probably close to what you learned in high school:

Standard Model

Brown heterozygote parent


Brown heterozygote parentBlue

Blue Blue (Blue phenotype)

Blue Brown (Brown phenotype)


Blue Brown (Brown phenotype)

Brown Brown (Brown phenotype)

In this model the expression of blue eyes is recessive, you need two copies. Heterozygotes, those who carry one copy of each allele, express brown eyes but can have blue eyed offspring. Blue eyed people can only have blue eyed offspring because they have to be homozygotes, carry two copies. A physiological explanation also offers itself up in this case, the blue eye allele is simply a copy which has lost function and results in the lack of production of melanin in the iris. The brown eye allele on the other hand functions normally. Even if only one copy is functioning to produce melanin, that is enough in terms of dosage to produce a brown coloration. Only with two copies which are non-functioning is there a total loss of melanin in the iris.

Reality is more complex. The diagram below is adapted from the paper referenced above. I've limited the data to those where the number of individuals in their sample was greater than 10, and those combinations where a proportion of individuals expressed blue eyes. For ease of inspection I've continued to color code the genetic variation. Instead of a single clearly defined gene with two alleles, that is, two genetic variants on a well defined physical location, the diagram below illustrates the combinations of two haplotypes, termed diplotypes. This is important because there is possibly no one blue eye gene responsible for all the variation, rather, a host of tightly linked loci may operate as a sort of genetic network. In any case, I have colored the haplotypes so as to show their preponderant average effect.

Adapted from Table 5 of A Three-Single-Nucleotide Polymorphism Haplotype in Intron 1 of OCA2 Explains Most Human Eye-Color Variation


Blue 1 / Blue 1 1,77262.5%28.0%9.5%

Blue 1 / Brown 4 13847.1%20.3%32.6%

Blue 1 / Brown 3 15427.9%22.1%50.0%

Blue 1 / Brown 3 15427.9%22.1%50.0%

Brown 3 / Brown 4 1225.0%8.3%66.7%

Brown 2 / Brown 3 2920.7%31.0%48.3%

Blue 1 / Brown 2 36417.6%38.5%44.0%

Blue 1 / Brown 8 2537.9%23.3%68.8%

Brown 2 / Brown 4 185.6%11.1%83.%

Brown 8 / Brown 8 224.5%0%95.5%

As you can see there is a strong statistical trend, but the relationship between genetic variation and phenotypic variation is not deterministic. The authors of the paper state, "These data are consistent with the...haplotype 1 [Blue 1] acting as a highly penetrant recessive blue-eye-color allele." Highly penetrant basically means "a lot of the time, but not always." As the lower rows illustrate even alleles which are predominantly predictive of brown eyes can together result in offspring with blue eyes! Conversely, a minority of homozygotes for the most powerful predictor of blue eyes still express brown eyes. What's going on here?

Please note that I said that 3/4 of the eye color variation could be explained by variation around OCA2, that still leaves room for genes of smaller effect to round out the balance.

The sample space of the genome is enormous, and one could easily imagine a large number of low frequency alleles at other loci besides OCA2 contributing to loss of function, or gain of function. Reading the paper closely the molecular-physiological nature of OCA2's modulation of the expression of this trait doesn't seem fully elucidated. There are many ways to alter regulatory pathways, so it should not surprise that other genes could account for the remaining 1/4 of the variation. Some scientists would assert that the whole concept of penetrance is a band-aid which exists to mask the reality that a fully understanding of the polygenic nature of a trait has not been achieved.

Note: What I have labeled Blue 1 is extant at around a frequency of 80% in northwest Europe. It seems to have been subject to powerful selection within the last 10,000 years, and is associated with the third longest haplotype within the European HapMap sample.

1 free article left
Want More? Get unlimited access for as low as $1.99/month
Already a subscriber? Log In or Register
1 free articleSubscribe
Want unlimited access?

Subscribe today and save 70%


Already a subscriber? Log In or Register
More From Discover
Recommendations From Our Store
Shop Now
Stay Curious
Our List

Sign up for our weekly science updates.

To The Magazine

Save up to 70% off the cover price when you subscribe to Discover magazine.

Copyright © 2021 Kalmbach Media Co.