I am an unbeliever...in overdominance

RPM pointed me to this new paper, Major Histocompatibility Complex Heterozygosity Reduces Fitness in Experimentally Infected Mice:

...Our results show that MHC effects are not masked on an outbred genetic background, and that MHC heterozygosity provides no immunological benefits when resistance is recessive, and can actually reduce fitness. These findings challenge the HA hypothesis and emphasize the need for studies on wild, genetically diverse species.

MHC are a group of loci which are critical in the adaptive immune systems of "higher" organisms. They are also among the most polymorphic (genetically diverse) loci, and the source of that diversity seems to be a form of balancing selection. Various alleles on the MHC (e.g., the HLA) exhibit so much time depth that they span species in terms of their coalescence to the last common ancestor. To be concrete about it, you may share an HLA lineage with a chimpanzee which you do not share with your sibling; ergo, on that allele you are more closely related to that chimpanzee than you are to your sibling. Over time neutral processes (see the 5 links) naturally turn over alleles so that they change. Functional constraint can prevent these dynamics from allowing the genome to evolve so that alleles replace each over time, but what you have with the HLA are many, many, variants preserved over millions of years.

So what's going on? The data above seems to reject the hypothesis of heterozygote advantage; e.g., the fitness of genotype Aa is greater than aa and AA. Rather, I accept W.D. Hamilton's contention that the polymorphism can best be explained by long term negative frequency dependent selection. In this case the fitness of the allele is inversely proportional to its frequency within the population (e.g., fitness of allele = 1/{frequency of allele}). Initially rare mutants spread because of the fitness they confer, but as they increase in frequency they converge upon the mean population fitness. If stochastic forces drive the frequency down again then it is once more favored. In practical biological terms as an allele becomes common pathogens quickly adapt to it as part of their environmental background. This results in a mix of alleles in the genetic background always extant at low to moderate frequencies being buffeted by pathogen selection pressures. Extremely rare alleles which might otherwise go extinct due to neutral processes are preserved in the genetic background because of their utility in containing pathogens who have no defenses for uncommon immune strategies. This explains the persistence of extremely ancient trans-species alleles at low frequencies on the HLA; whereas in a normal evolutionary process they would have become extinct, negative frequency dependent fitness results in a sharp increase in their selective value as they approach zero, always moving them away from the extinction boundary.

Of course a great deal of extant polymorphism results in heterozygosity as a natural outcome of segregation during sexual reproduction. That does not mean that that heterozygosity is in itself being selected for, and rather it may exhibit short term fitness load. Moving beyond the MHC loci I have already posted on why I am skeptical of the ubiquity of overdominance.