It can if there is frequency-dependent selection against susceptibility, rather than selection for resistance. If only one or a few alleles at HLA were to confer resistance to a disease, then during an epidemic these alleles would sweep through the population (that is, individuals with these alleles would be more likely to survive). This process would reduce HLA polymorphism. Instead, at least in this case, a few alleles confer susceptibility and the majority confer resistance. As these "susceptibility" alleles are reduced in frequency during the course of an epidemic, the proportion of susceptible hosts and the likelihood that the virus can spread are reduced. This resistance occurs without destroying the genetic variation that is essential for defenses against other diseases. The major histocompatibility complex (MHC) appears to have been selected for high levels of diversity, such that each disease organism can only interact unfavorably with a minority of MHC alleles. D. R. Green and I have described this process as a kind of genetic herd-immunity, in which host population diversity can protect individual hosts against a wide variety of diseases (2).

This "selection against susceptibility" can be contrasted with single-gene resistance that can be traced to altered receptor molecules, such as Duffy-negative (3), which confers resistance to vivax malaria, and mutants of CCR5, which may confer resistance to a variety of diseases (4). Such single-gene resistance can sweep through the host population, as has happened with Duffy-negative. MHC, unlike these receptor loci, has been shaped by evolution to retain its polymorphism even in the face of strong directional selection.

Christopher Wills
Department of Biology,
University of California, San Diego,
La Jolla, CA 92093, USA
E-mail: cwills{at}ucsd.edu

REFERENCES

1. M. Carrington et al., Science 283, 1748 (1999) [Abstract/Free Full Text] .
2. C. Wills and D. R. Green, Immunol. Rev. 143, 263 (1995) [CrossRef] [Web of Science] [Medline] .
3. R. Horuk et al., Science 261, 1182 (1993) [Abstract/Free Full Text] .
4. M. Carrington et al., Am. J. Hum. Genet. 61, 1261 (1997) [CrossRef] [Web of Science] [Medline] ; J. C. Stephens et al., ibid. 62, 1507 (1998) .

Response: Wills makes a good point. Most observers agree that the enormous diversity at MHC loci in vertebrates owe their retention to selective pressures produced by fatal disease outbreaks among the ancestors of modern species (1). Because selective effects of pathogens require exposure to the agent, frequent susceptible alleles will suffer more severe consequences when their frequency is high and less when rare; that is, frequency- dependent selection occurs (2).

The simplest explanation for heterozygous advantage (overdominance) for MHC loci would be an increased repertoire for individual recognition of viral epitopes encountered by pathogen exposure (1). How more susceptible alleles operate is less clear, but could involve a failure of allele products to effectively recognize or present viral epitopes to CD8-T lymphocytes (probably not the case for B35 and C04, because several HIV-1-specific monomers are presented by these two HLA class I products) (3). Alternatively, potential explanations involve differential killer cells facilitation (4), or even the concept that certain viral epitopes provide a decoy, whereby host cytolytic T lymphocytes are occupied with a pathogen target that is ineffective in pathogen clearances. Newly available technologies and reagents to test these hypotheses might help resolve these important questions (5).

Stephen J. O'Brien
Mary Carrington
Laboratory of Genomic Diversity,
National Cancer Institute,
Frederick, MD 21702, USA
E-mail: obrien{at}mail.ncifcrf.gov

REFERENCES

1. P. Parham and T. Ohta, Science 272, 67 (1996) [Abstract] ; A. Hughes and M. Yeager, Annu. Rev. Genet. 32, 415 (1998) [CrossRef] [Web of Science] [Medline] .
2. C. Wills and D. R. Green, Immunol. Rev. 143, 263 (1995) .
3. H. Tomiyama et al., J. Immunol. 158, 5026 (1997) [Abstract] ; S. Rowland-Jones et al., Nature Med. 1, 59 (1995) [CrossRef] [Web of Science] [Medline] ; erratum, ibid., p. 598; E. Rosenberg et al., Science 278, 1447 (1997) [Abstract/Free Full Text] ; O. O. Yang et al., J. Virol. 71, 3120 (1997) [Abstract] ; F. Gotch et al., Immunol. Lett. 51, 125 (1996) [CrossRef] [Web of Science] [Medline] ; T. Harrer et al., AIDS Res. Hum. Retroviruses 12, 585 (1996) [Web of Science] [Medline] .
4. M. Carrington et al., Science 283, 1748 (1999) .
5. A. J. McMichael, Cell 93, 673 (1998) [CrossRef] [Web of Science] [Medline] ; ___ and C. A. O'Callaghan, J. Exp. Med. 187, 1367 (1998); S. Rowland-Jones et al., Adv. Immunol. 65, 277 (1997) [Web of Science] [Medline] .