Citation: Holmes EC (2004) Adaptation and Immunity. PLoS Biol 2(9): e307. https://doi.org/10.1371/journal.pbio.0020307
Published: September 14, 2004
Copyright: © 2004 Eddie C. Holmes. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abbreviations: C, cytosine; dN, rate of nonsynonymous substitution per site; dS, rate of synonymous substitution per site; G→A, guanine-to-adenine; HIV, human immunodeficiency virus; MHC, major histocompatibility complex; U, uracil
The ongoing battle between hosts and pathogens has long been of interest to evolutionary biologists. Because hosts and pathogens act as environments for each other, their intertwined struggle for existence is both continual and rapid. At the molecular level, this cycle of environmental change and evolutionary response means that mutations are continually being tried out by natural selection. It is therefore little wonder that the host and pathogen genes that control infection and immunity frequently show high levels of genetic diversity and present some of the best examples of positive selection (adaptive evolution) reported to date (Yang and Bielawski 2000). In particular, rates of nonsynonymous substitution per site (resulting in an amino acid change; dN) often greatly exceed those of synonymous substitution per site (silent change; dS), as expected if most mutations are fixed because they increase fitness (Figure 1).
(A) Classification of substitutions. Nonsynonymous substitutions (red) are those that change the amino acid sequence of the protein encoded by the gene, while the degeneracy of the genetic code ensures that synonymous substitutions (yellow) result in the same amino acid sequence.
(B) Calculation of dN/dS. By assuming that synonymous mutations are neutral and fixed by random genetic drift, it is possible to determine the mode of selection acting on nonsynonymous mutations. If all nonsynonymous substitutions were neutral, then their rate of occurrence per site (dN) would be the same as that of synonymous substitutions per site (dS), so that dN/dS equals one. A lower ratio of nonsynonymous to synonymous substitutions per site (dN/dS < 1) means that some proportion of the nonsynonymous mutations are deleterious and removed by purifying selection. Conversely, positive selection fixes advantageous nonsynonymous mutations faster than genetic drift fixes synonymous mutations (dN/dS > 1), although this is usually restricted to a small proportion of amino acid sites within any gene. In the hypothetical example of five gene sequences shown here, with dS given above the diagonal and dN below the diagonal, there is no evidence for positive selection because mean dN/dS the (0.577) is less than one.
At the host level, most studies of the selection pressures acting on immune system genes have concentrated on genes implicated in the adaptive immune response against microbial pathogens, particularly those producing antibodies (Sitnikova and Nei 1998; Sumiyama et al. 2002), or genes encoding reconnaissance molecules known as the major histocompatibility complex (MHC), which control the action of T-cells (Hughes and Nei 1988; Yeager and Hughes 1999) (Box 1). As its name suggests, the role of the adaptive immune response is to stimulate and ‘memorise’ immunity to specific pathogens. As microbial pathogens such as viruses are both abundant and rapidly evolving, positive selection on components of the adaptive immune response is often very strong (Yeager and Hughes 1999). Far less attention has been directed toward the less specific innate (‘nonadaptive’) immune response, even though this response requires a wide array of genes and acts as the front line of immune defence (Box 1). Would we expect the same strength of positive selection on a generalized pathogen control system? This is a question of fundamental importance because the luxury of adaptive immunity is not available to most organisms, having probably evolved along with the vertebrates (Bartl et al. 1994), whereas the more widespread innate immune system is often depicted as a primitive characteristic.
Molecular Evolution of the Innate Immune System
The genes involved in innate immunity have recently come under the molecular evolutionists' gaze. One important group are the defensins, a large class of short antimicrobial peptides that constitute an effective immune response team in organisms as diverse as plants and primates (Boman 1995). Because defensins are cationic (positively charged), they are able to interact with negatively charged molecules on the surface of microbes and permeate their membranes. Sequence analyses of defensins and similar antimicrobial peptides have revealed the telltale signatures of positive selection, with dN greater than dS in many comparisons (Hughes 1999; Duda et al. 2002; Maxwell et al. 2003). Other genes of the innate immune system also seem to be subject to powerful positive selection. One dramatic example described in this issue of PLoS Biology is the APOBEC3G gene of primates (Sawyer et al. 2004). This case is especially striking because rather than killing pathogens through protein or cellular interactions, like most immune genes, APOBEC3G works by manipulating the genome sequence of the invading microbe.
The genomes of primates contain a family of nine APOBEC genes that encode enzymes involved in the editing of RNA and/or DNA through the deamination of cytosine (C), so that this nucleotide mutates to uracil (U). This is essential for various aspects of cellular function. APOBEC1, for example, is involved in the C→U editing of apolipoprotein B mRNA (therein christening the family), while another family member, the activation-induced deaminase, has a vital role in adaptive immunity in that it assists in the diversification of antibodies. Two more enzymes, APOBEC3G and APOBEC3F, form part of the innate immune system; they function as antiviral agents and are being intensively studied in the context of infection with the human immunodeficiency virus (HIV), the cause of AIDS. In particular, APOBEC3G targets the reverse transcription step of the HIV life cycle, in which the viral genomic RNA is converted into proviral DNA, which is then integrated into the host genome (Mangeat et al. 2003). APOBEC3G-induced deamination at this stage results in monotonous guanine-to-adenine (G→A) nucleotide changes, a phenomenon called G→A hypermutation that had long been noted by HIV researchers without a clear understanding of its cause. We now know that G→A hypermutation is part of the innate immune response to retroviral infections.
Although there is still some debate over exactly how APOBEC3G leads to viral eradication, the most likely scenario is that G3A hypermutation results in the generation and incorporation of a multitude of deleterious mutations that fatally disrupt viral functions. This strategy is likely to work well for retroviruses like HIV because their genomes are so compact that individual sequence regions often perform multiple functions. Under these cramped conditions, most mutations are likely to severely disrupt some aspect of viral function and thereby reduce fitness (Holmes 2003). Indeed, it has been estimated that the deleterious mutation rate in viruses that replicate using RNA polymerases (either reverse transcriptase in the case of retroviruses or RNA-dependent RNA polymerase for other RNA viruses) is on the order of one error per replication cycle, so that many of the viral progeny produced by replication are defective (Elena and Moya 1999). HIV is normally able to overcome this burden of deleterious mutation because of its remarkable reproductive power; each day, on the order of 1010 virions are produced in a single infected individual (Perelson et al. 1996), so that enough fit and able recruits will make it through to the next generation.
Treating RNA Virus Infections Through Lethal Mutagenesis
The high mutation rates of RNA viruses mean that adaptively useful genetic variation is produced frequently. The rub, however, is that fitness-enhancing mutations are a small minority, and the preponderance of deleterious mutations means that RNA viruses live on the edge of survival (Domingo 2000). By increasing the rate at which deleterious mutations appear, APOBEC3G pushes viruses over this edge, causing a form of ‘lethal mutagenesis’ that results in their destruction; the rate of mutation becomes so high that no genome can reproduce itself faithfully, and the population crashes. Intriguingly, researchers designing new antiviral drugs have also begun to realise that forcing viruses into this sort of ‘error catastrophe’ might be an effective way to treat them (Figure 2). There are a growing number of studies in which mutagens, such as ribavirin and 5-fluorouracil, are applied to viral infections in vitro and in vivo, including HIV, in the hope that these will induce so many deleterious mutations that the virus suffers an error catastrophe and is cleared (Loeb et al. 1999; Sierra et al. 2000; Crotty et al. 2001; Ruiz-Jarabo et al. 2003). The results produced to date are highly encouraging, particularly when these error-inducing drugs are combined with more conventional treatment strategies that aim to reduce the rate of viral replication (Pariente et al. 2001). The discovery that a natural antiviral agent, APOBEC3G, probably works in much the same way should provide even more encouragement.
(A) In a viral population prior to the application of mutagens, the mean error rate (white) is on the order of one per genome per replication (mutations marked by asterisks). (B) If a mutagen such as ribavirin is then applied to an infected patient, the mean error rate of the virus (black) is increased so that the population crosses a threshold of ‘error catastrophe’; after this point fitness declines dramatically and the population crashes. This drug-induced lethal mutagenesis seems to work more efficiently when it is used in combination with drugs that reduce the rate of viral replication.
Sadly, however, the pathogens have fought back. The anti-HIV properties of APOBEC3G were discovered because most viral strains escape its neutralising properties. Lentiviruses like HIV possess a gene that encodes an protein called Vif (‘viral infectivity factor’) that counters APOBEC3G (Sheehy et al. 2002). Hence, it is probably only in naturally occurring Vif-defective mutants that APOBEC3G is effective against HIV. Furthermore, because positive selection on APOBEC3G has operated for at least 30 million years and lentiviruses in general, and HIV in particular, are likely to be more recently evolved than this, it is clear that a broad range of retroviral pathogens have been responsible for the adaptive evolution of this particular immune gene (Sawyer et al. 2004). Given the frequency with which the remnants of past retroviral infections are found in the mammalian genome (Smit 1999), in the form of usually defunct endogenous retroviruses (Box 1), it is likely that our genomes are continually bombarded with retroviruses like HIV but that the majority are cleared by innate immune mechanisms like APOBEC3G. It is possible that the retroviruses that successfully infect us are those, like HIV, that have managed to evolve strategies to avoid the destructive capacities of APOBEC3G.
The intense selective pressure on the defensins and APOBEC3G illustrates that although the innate immune response is generalist in its action, it is as highly and intricately evolved as its better-studied ally, the adaptive immune system. Rather than being an evolutionary remnant, it is a dynamic and continually adapting system. Less clear is whether other host proteins act in the same manner as APOBEC3G. In particular, the most common and destructive pathogens faced by humans and other mammals are RNA viruses, such as influenza A, yellow fever, and hepatitis C. In most cases, our ability to survive these viral infections is simply a combination of good luck and good breeding; with the right combination of MHC alleles, itself a function of population history and what we by chance inherit from our parents, some individuals may be more able to fight off viral infections than others. The ubiquity of RNA viruses hints that our genomes might also contain an innate, yet highly adapted, defence system that targets this abundant class of pathogens by manipulating their mutation rate. Although utilizing lethal mutagenesis might one day be an important way to design new drugs against a variety of viral pathogens, it would come as no surprise if nature got there first.
Box 1. Glossary
Adaptive immune system. The pathogen-specific part of the vertebrate immune system. It is comprised of two major arms, antibodies (the humoral response) and T-cells (the cellular response), both of which lay down an immunological memory for future defence.
Innate immune system. The nonspecific part of the vertebrate immune system. It has a wide variety of components, ranging from lysozymes in saliva to cytokines, defensins, interferons, and natural killer cells in a variety of tissues.
Endogenous retroviruses. The (usually) dead remnants of functional retroviruses that are now passed on through the germ line like normal genes. It has been estimated that approximately 5% of the human genome is composed of endogenous retroviruses.
- 1. Bartl S, Baltimore D, Weissman IL (1994) Molecular evolution of the vertebrate immune system. Proc Natl Acad Sci U S A 91: 10769–10770.S. BartlD. BaltimoreIL WeissmanMolecular evolution of the vertebrate immune system.Proc Natl Acad Sci U S A1994911076910770
- 2. Boman HG (1995) Peptide antibiotics and their role in innate immunity. Annu Rev Immunol 13: 61–92.HG BomanPeptide antibiotics and their role in innate immunity.Annu Rev Immunol1995136192
- 3. Crotty S, Cameron CE, Andino R (2001) RNA virus error catastrophe: Direct test by using ribavirin. Proc Natl Acad Sci U S A 98: 6895–6900.S. CrottyCE CameronR. AndinoRNA virus error catastrophe: Direct test by using ribavirin.Proc Natl Acad Sci U S A20019868956900
- 4. Domingo E (2000) Viruses on the edge of adaptation. Virology 270: 251–253.E. DomingoViruses on the edge of adaptation.Virology2000270251253
- 5. Duda TF, Vanhoye D, Nicholas P (2002) Roles of diversifying selection and coordinated evolution in the evolution of amphibian antimicrobial peptides. Mol Biol Evol 19: 858–864.TF DudaD. VanhoyeP. NicholasRoles of diversifying selection and coordinated evolution in the evolution of amphibian antimicrobial peptides.Mol Biol Evol200219858864
- 6. Elena SF, Moya A (1999) Rate of deleterious mutation and the distribution of its effects on fitness in vesicular stomatitis virus. J Evol Biol 12: 1078–1088.SF ElenaA. MoyaRate of deleterious mutation and the distribution of its effects on fitness in vesicular stomatitis virus.J Evol Biol19991210781088
- 7. Holmes EC (2003) Error thresholds and the constraints to RNA virus evolution. Trends Microbiol 11: 543–546.EC HolmesError thresholds and the constraints to RNA virus evolution.Trends Microbiol200311543546
- 8. Hughes AL (1999) Evolutionary diversification of the mammalian defensins. Cell Mol Life Sci 56: 94–103.AL HughesEvolutionary diversification of the mammalian defensins.Cell Mol Life Sci19995694103
- 9. Hughes AL, Nei M (1988) Pattern of nucleotide substitution at major histocompatibility complex class I loci reveals overdominant selection. Nature 335: 367–370.AL HughesM. NeiPattern of nucleotide substitution at major histocompatibility complex class I loci reveals overdominant selection.Nature1988335367370
- 10. Loeb LA, Essigmann JM, Kazazi F, Zhang J, Rose KD, et al. (1999) Lethal mutagenesis of HIV with mutagenic nucleoside analogs. Proc Natl Acad Sci U S A 96: 1492–1497.LA LoebJM EssigmannF. KazaziJ. ZhangKD RoseLethal mutagenesis of HIV with mutagenic nucleoside analogs.Proc Natl Acad Sci U S A19999614921497
- 11. Mangeat B, Turelli P, Caron G, Friedli M, Perrin L, et al. (2003) Broad antiretroviral defence by human APOBEC3G through lethal editing of nascent reverse transcripts. Nature 424: 99–103.B. MangeatP. TurelliG. CaronM. FriedliL. PerrinBroad antiretroviral defence by human APOBEC3G through lethal editing of nascent reverse transcripts.Nature200342499103
- 12. Maxwell AI, Morrison GM, Dorin JR (2003) Rapid sequence divergence in mammalian ß-defensins by adaptive evolution. Mol Immunol 40: 413–421.AI MaxwellGM MorrisonJR DorinRapid sequence divergence in mammalian ß-defensins by adaptive evolution.Mol Immunol200340413421
- 13. Pariente N, Sierra S, Lowenstein PR, Domingo E (2001) Efficient virus extinction by combinations of a mutagen and antiviral inhibitors. J Virol 75: 9723–9730.N. ParienteS. SierraPR LowensteinE. DomingoEfficient virus extinction by combinations of a mutagen and antiviral inhibitors.J Virol20017597239730
- 14. Perelson AS, Neumann AU, Markowitz M, Leonard JM, Ho DD (1996) HIV-1 dynamics in vivo: Virion clearance rate, infected cell lifespan, and viral generation time. Science 271: 1582–1586.AS PerelsonAU NeumannM. MarkowitzJM LeonardDD HoHIV-1 dynamics in vivo: Virion clearance rate, infected cell lifespan, and viral generation time.Science199627115821586
- 15. Ruiz-Jarabo CM, Ly C, Domingo E, de la Torre JC (2003) Lethal mutagenesis of the prototypic arenavirus lymphocytic choriomeningitis (LCMV). Virology 308: 37–47.CM Ruiz-JaraboC. LyE. DomingoJC de la TorreLethal mutagenesis of the prototypic arenavirus lymphocytic choriomeningitis (LCMV).Virology20033083747
- 16. Sawyer SL, Emerman M, Malik HS (2004) Ancient adaptive evolution of the primate antiviral DNA-editing enzyme APOBEC3G. PLoS Biol 2: e275.SL SawyerM. EmermanHS MalikAncient adaptive evolution of the primate antiviral DNA-editing enzyme APOBEC3G.PLoS Biol20042e275
- 17. Sheehy AM, Gaddis NC, Choi JD, Malim MH (2002) Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein. Nature 418: 646–650.AM SheehyNC GaddisJD ChoiMH MalimIsolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein.Nature2002418646650
- 18. Sierra S, Dávila M, Lowenstein PR, Domingo E (2000) Response of foot-and-mouth disease virus to increased mutagenesis. Influence of viral load and fitness in loss of infectivity. J Virol 74: 8316–8323.S. SierraM. DávilaPR LowensteinE. DomingoResponse of foot-and-mouth disease virus to increased mutagenesis. Influence of viral load and fitness in loss of infectivity.J Virol20007483168323
- 19. Sitnikova T, Nei M (1998) Evolution of immunoglobulin kappa chain variable region genes in vertebrates. Mol Biol Evol 15: 50–60.T. SitnikovaM. NeiEvolution of immunoglobulin kappa chain variable region genes in vertebrates.Mol Biol Evol1998155060
- 20. Smit AF (1999) Interspersed repeats and other mementos of transposable elements in mammalian genomes. Curr Opin Genet Dev 9: 657–663.AF SmitInterspersed repeats and other mementos of transposable elements in mammalian genomes.Curr Opin Genet Dev19999657663
- 21. Sumiyama K, Saitou N, Ueda S (2002) Adaptive evolution of the IgA hinge region in primates. Mol Biol Evol 19: 1093–1099.K. SumiyamaN. SaitouS. UedaAdaptive evolution of the IgA hinge region in primates.Mol Biol Evol20021910931099
- 22. Yang Z, Bielawski JP (2000) Statistical methods for detecting molecular adaptation. Trends Ecol Evol 15: 496–502.Z. YangJP BielawskiStatistical methods for detecting molecular adaptation.Trends Ecol Evol200015496502
- 23. Yeager M, Hughes AL (1999) Evolution of the mammalian MHC: Natural selection, recombination, and convergent evolution. Immunol Rev 167: 45–58.M. YeagerAL HughesEvolution of the mammalian MHC: Natural selection, recombination, and convergent evolution.Immunol Rev19991674558