In overdominant selection, individuals, or more specifically, MHC alleles, that are selected for are those in the heterozygous form. This is also called heterozygote advantage because it is believed that there is some advantage to being heterozygous at that particular locus (6). Hughes and Nei (1988) were the first to claim that overdominant selection was the mechanism maintaining MHC polymorphism. The logic behind heterozygote advantage at MHC is that the possession of two different alleles, each with specific binding specificities, would increase the number of antigen sequences that can be presented by the MHC molecule. The strength of the T-cell response would then be increased should a cell with a heterozygote genotype be infected (9). To show overdominant selection , Hughes and Nei looked at the substitution patterns for class I and II MHC loci. The important feature of the substitution that they considered was whether the substitutions were synonymous or non-synonymous. In a non-synonymous substitution, the amino acid is altered while it remains the same in a synonymous substitution (5). If positive Darwinian selection is occurring, we would expect non-synonymous substitutions to be the majority because this selection is working to increase novel genotypes, which is the case with non-synonymous substitutions (6).
The antigen recognition site (ARS) determines the specificity of the MHC molecule. If positive selection were occurring at MHC loci, it should be acting on the ARS coding region of the gene (6). Hughes and Nei found that indeed, non-synonymous substitutions were significantly higher than synonymous substitutions at the ARS coding region of all MHC loci. By contrast, they found that in the non-ARS regions, non-synonymous mutations were significantly lower than synonymous (5,10). The results were interpreted to indicate that amino acid substitutions in the ARS region are enhanced by positive Darwinian selection. In other words, selection favoured a novel genotype that resulted from an amino acid substitution in the ARS region. The claim is that the substitution is favoured because the mutant allele almost always produces a heterozygote that has a selective advantage over the homozygotes and more common alleles with respect to their immune responses (6).
There is also heterozygote advantage because being heterozygous prevents the possibility of a parasite "attacking" a particular MHC allele and driving it to extinction. The reason for this is the codominant nature of the MHC loci. Both alleles participate in the immune response, so an allele that is susceptible to some particular pathogen may gain protection through the activity of the other allele should that allele be different and resistant to the infecting pathogen. The net result is that an allele in a heterozygote individual that is susceptible to a specific pathogen will remain in the population (3). This is another mechanism by which MHC polymorphism is maintained. Susceptible alleles will be driven out of the population only when found in a homozygote and infected with the specific pathogen. The susceptible allele could also become resistant should the pathogen evolve in such a way that the original MHC molecule can now recognize it. A problem is that pathogen evolution that would save a rare, susceptible allele, cannot likely occur fast enough to save the allele from extinction (12).
It must be made clear that in overdominant selection, the selection is not for advantageous mutations such as a mutation that would perhaps increase the immune response from a particular allele. Selection for advantageous mutations reduces the level of polymorphism and shortens the persistence time of polymorphic alleles to levels much lower that neutral alleles (5). Overdominant selection is clearly selection for new MHC alleles when they are in a heterozygous state. This selection will increase polymorphism and will also create unusually long persistence times for polymorphic alleles (6,11).