Galton Institute Home Page March 2006 Newsletter Contents Newsletter Index

The Powers of Natural Selection

9. Balanced Polymorphism and Overdominance

W. M. S. Russell

There is another way in which balanced polymorphism is brought about, viz. when the heterozygote of a major gene is more advantageous than either homozygote. This is called fitness overdominance, and of course it ensures that both of two alleles are preserved in the gene pool of the population. We have seen that neither warfarin resistance in rats nor melanism in moths in industrial areas – nor, for that matter, what we might call demelanism in smoke-free areas – can complete their apparently transient polymorphism, because the total elimination of the disadvantageous allele is prevented by heterozygote advantage. It can thus preserve in the gene pool at a certain frequency wholly harmful and even lethal alleles. This was one of Fisher’s many important discoveries (1930).

Balanced polymorphism due to overdominance has been shown in a number of species, for instance the fruit-fly Drosophila pseudoobscura and the marine copepod Tisbe reticulata (Wallace, 1968). In the grouse locust Apotettix eurycephalus the heterozygotes for several different alleles were in each case at a 7% advantage over one homozygote and 5% or 6% over the other (Sheppard, 1975). Ford (1975) made an ingenious suggestion for how overdominance arises. If an allele of a major gene (as often happens) has several effects, and some are advantageous and some disadvantageous, while the other allele of this gene has exactly the opposite effects, the advantageous effects will become dominant, and the disadvantageous ones recessive. The heterozygote will then only show the advantageous effects, and so will be better off than either homozygote. This situation has actually been demonstrated in the moth Ephestia kühniella, where a pair of alleles most obviously controlling the colour of the testis control a number of other characters as well, which show dominance or recessiveness as predicted by the theory.

Overdominance explains the appreciable frequencies of certain diseases in human populations. This has been shown for Tay-Sachs disease (Myrianthopoulos and Aronson, 1971), and may well be found true of other diseases of the Ashkenazi ‘Jews’. But the most striking case in man is that of the human haemoglobins (Russell and Russell, 1983).

In most people, foetal haemoglobin is replaced by normal adult haemoglobin (A). A single gene mutation causes the production of an abnormal haemoglobin (S), differing from the normal type in respect of a single amino acid. In the homozygote, most haemoglobin is of the S type, the remainder being unreplaced foetal haemoglobin. When oxygen pressure is slightly low, the S molecules form rigid aggregates, distorting the erythrocytes into a sickle shape, causing circulation blockage, cell destruction, hence severe anaemia. Hence the S type and its allele are called ‘sickle cell’. In the heterozygote, called a sickle cell trait carrier, less than half the haemoglobin is S, the remainder being A, and sickling only occurs at very low oxygen pressures, such as at high altitudes or in unpressurised aircraft. In normal life, the heterozygotes are perfectly healthy.

However, nearly all the homozygotes die before reproducing, so there is powerful selection against the sickle cell allele. Despite this, heterozygotes reach frequencies of 20% or even 40% in parts of Africa. This is because the sickle cell trait protects the heterozygote from malaria, which is always carried by different species of the mosquito genus Anopheles (Spielman and D’Antonio, 2002). This balanced polymorphism, and the high heterozygote frequencies, are naturally found only in regions of malaria.

The Atlantic slave trade displaced West African populations with high sickle cell trait frequencies. In non-malarious regions of the New World the frequencies in populations of African origin have declined, since selection against the S homozygote is no longer balanced by selection in favour of the heterozygote, as shown in Table 1 (simplified from Allison, 1971).

Table 1: Percentage Frequencies of the Sickle Cell Trait in the New World

Malarious Regions

Honduras 23.3
Surinam 20.3
Jamaica 11.5
St.Lucia 14.0

Non-Malarious Regions

Curacao 7.2
St.Vincent 8.7
Dominique 9.5
Barbados 7.0

Note: In the 1990s, Anopheles aquasalis arrived in Barbados, but fortunately the very small colony that could breed there has not proved a malaria hazard (Spielman and D’Antonio, 2002).

S haemoglobin is only one of over 100 abnormalities. Of these, thalassaemia is a heterogeneous condition involving several genes, but a number of abnormal haemoglobins are related to multiple alleles of the sickle gene. They are generally alike in producing more or less severe anaemia in the homozygote, while protecting the heterozygote against the local form of malaria. In non-malarious regions, such as England, Sweden and Japan, the abnormal haemoglobin mutants are eliminated by selection; but appreciable frequencies of the heterozygotes for the various abnormal haemoglobins are found in all malarious regions.

In Saudi Arabia, the persistent activity of foetal haemoglobin apparently enables all the sickle cell homozygotes to survive, and trait frequencies already high are expected to increase there (Gelpi, 1979).

In West Africa, by combining studies of genetics, languages and local traditions, Livingstone (1962) showed that ‘the frequency of the sickle cell trait exactly corresponds, with a few centuries’ lag, to the introduction of farming’ (Russell and Russell, 1983). Livingstone observed that the clearing practices of the farmers created stagnant sunlit waters where the mosquito Anopheles gambiae could breed. So ‘farming, by introducing its vector, introduced falciparum malaria, and created the conditions for selection of the sickle cell trait’ (Russell and Russell, 1983). This is a spectacular example of behavioural selection (Russell and Russell, 1990).

The sickle cell trait itself is found over much of Africa, in Mediterranean Europe, in the Near and Middle East, and over much of the Indian sub-continent. In South East Asia, on the other hand, malarious regions have high frequencies of a different abnormal haemoglobin, called E, in heterozygous form. Livingstone (1964) draws the inference: ‘the penetration of agriculture into the tropical regions of Asia came from two centres, China and the Middle East’; the first movement is connected with the spread of haemoglobin E, the second with that of haemoglobin S. ‘Where these two movements met in the vicinity of Calcutta is also the border between the high frequencies of these genes, with S on the West and E to the East’ (Livingstone).