SCOOTER82a_Livingstone_Frequencies of Hemoglobin Variants ...

thousands of blood preparations I examined.
The question is then why it has only increased in
this one malarious area. Migration is surely one
of the forces contributing to its distribution, but
the populations of Papua New Guinea are ethnically
quite distinct from the peoples of the
Malay Peninsula. The migration would have to
have been very ancient and did not result in
hemoglobin E being dispersed among the same
peoples. This seems to be more evidence for the
later diffusion of hemoglobin E, but it would also
seem to raise the possibility that malaria selection
is somewhat different in this area from that
in Africa or the Middle Eastand India.
The very high frequencies of hemoglobin E in
Southeast Asia raise a similar problem. This allele
is also a common mutant and, as has been previously
stated, is now found occasionally in
Europe and also among the Eti-Turks. Given its
high frequencies in Asia, one would assume that
it would have increased more than it has in other
malarious areas. Again the possibility of differences
in malaria selection arises, although the
same species of human malaria parasite occurs
in both Southeast Asia and the Africa-Middle
East area. Hemoglobin C is found in high frequencies
in West Africa, and hemoglobin OArab
in much of the Middle East and North Africa in
low frequencies. These are both the same glu­
Iys change found in hemoglobin E but at different
positions. Since the clinical symptoms of
homozygosity for these two hemoglobin variants
are similar to those of hemoglobin E, they would
appear to have the same fitness disadvantage
and probably the same advantage in heterozygotes;
but with the exception of some hemoglobin
C frequencies in West Africa, these two variants
nowhere approach in frequency the
hemoglobin E frequencies in Southeast Asia or
Assam. In all cases, these glu-Iys mutants seem
to be out-competed by hemoglobin S and are in
the process of being replaced by it. The fact that
hemoglobin S is not found in Southeast Asia but
in high frequencies in Africa and the Middle East
could account for these differences.
Finally, the distributions of a-thalassemia in
Southeast Asia and Africa seem so different as to
require further factors to explain them. In most
human populations, the a locus is duplicated, so
that most individuals have four functional sites
that produce a chains. a-thalassemia is a mutation
that results in the absence of function of
either one or both of the a genes on a single
chromosome; the single deletion is a2-thalassemia
(-a) and the double deletion, al-thalassemia
(- -). Homozygosity for the double deletion
results in an inviable fetus with hydrops
fetalis, while simultaneous heterozygosity for
the single and double deletions (-al - -)
causes hemoglobin H disease, which can result
in varying degrees of decreased fitness. Thus,
the two genes are in competition. Most estimates
of the fitness values of all the genotypes
imply that there are no stable polymorphism frequencies
with both alleles present in a population
(Wills and Londo 1981, Livingstone 1983).
Most estimates also indicate that the al-thalassemia
gene would be replaced, which has led to
Wills and Londo calling it a "fugitive" allele.
Yokoyama (1983) made a similar analysis of the
a-thalassemias and concluded that a stable polymorphism
can exist with both the single and
double deletions present. However, his estimates
of the fitness values of the various genotypes
seem to be very unrealistic. Of the eight
sets of fitnesses that result in a stable polymorphism,
half (four) estimate the fitness of hemoglobin
H disease (heterozygous for both the sin-
9
gle and double deletion (-al - -)) to be
greater than normals or individuals with four a
genes. This is impossible. For the other four sets,
individuals with two a genes in the form of heterozygosity
of the double deletion and the normal
(- - I aa) have an increased fitness compared
to normals, while those with two deletions
who are homozygous for the single deletion
(-al-a) have a decreased fitness. Given the
phenotypic similarity of these genotypes, these
estimates seem highly unrealistic. In addition,
the decrease in fitness of the single deletion
homozygote is as great in most cases as those
with hemoglobin H disease, which again seems
to me to have little evidence in its favor. Thus,
more reasonable estimates of these fitness values
still lead to the conclusion that no possible
stable polymorphism could exist. Nevertheless,
although a2-thalassemia (-a) is much more
widespread and occurs in higher frequencies, a
l-thalassemia (--) is found in polymorphic frequencies
in the Mediterranean and is very common
in Southeast Asia. I have shown that althalassemia
will inhibit the increase of a2-thalassemia
if it attains polymorphic frequencies first
and that may have happened in Southeast Asia
(Livingstone 1983). But this is still one more
striking difference between Southeast Asia and
other areas with many malaria polymorphisms in
the Old World.
Only a few of the major problems raised by
current knowledge of the distribution of these
red cell traits have been discussed. A perusal of
many of the frequencies recorded here raises
many other questions. It is hoped that this compilation
of the data will be useful in solving some
of them and in relating this certain segment of
human genetic variation to human demographic,
cultural, and epidemiological history.