From genotype to phenotype

All the patterns discussed thus far have pertained to the long-term evolution of the immune system. It is important to remember, however, that all adaptive evolution is based on phenotypic polymorphism that segregates in populations at some point in time. Indeed, extant natural populations harbour considerable genetic variation for immunocompe-tence. This segregating phenotypic variation is the substrate for short-term evolution. Understanding its genetic basis and the forces governing its persistence is essential for predicting the evolutionary response to natural or artificial perturbations in infectious pressure in natural populations.

In organisms with well-characterized genomes, it is possible to directly test the phenotypic effects of allelic variation in pre-chosen 'candidate' genes though genotype-phenotype association mapping. These studies have been employed most effectively in D. melanogaster. For instance, natural allelic variation in the ref(2)P gene clearly determines resistance to the vertically transmitted Sigma virus in D. melanogaster females in an almost purely Mendelian fashion (Contamine et al., 1989; Bangham et al, 2008). Genetic variation in Sigma viral transmission through males, however, does not map to ref(2) P (Bangham et al, 2008). Variation in the ability of D. melanogaster to suppress bacterial infection has been mapped to polymorphisms in pathogen-recognition factors and signalling genes within the Toll and Imd pathways (Lazzaro et al, 2004, 2006). Expression levels, but not polymorphisms, of AMPs are also associated with resistance to infection (T.B. Sackton, B.P. Lazzaro, and A.G. Clark, unpublished data). These observations, coupled with evaluation of transcriptional activity of the immune system, indicate that signalling flux through the Toll and Imd pathways is a tremendously important determinant of resistance to bacterial infection. In contrast to the antiviral resistance determined by ref(2) P, polymorphisms mapped in the antibacterial association studies each make relatively small contributions to variance in the resistance phenotype, suggesting that resistance to bacterial infection is a combinatorial function of multiple genes of individually small effect. Even in sum, the mapped antibacterial factors do not explain the entirety of the genetic variance, indicating that other unstudied genes also contribute to variation in resistance.

If pathogen infection can be so detrimental to the condition of the host, and host alleles that confer high resistance to infection exist in natural populations, why then does resistance not spread to all individuals? Genetic trade-offs, whereby immunocompetence comes at a cost to another phenotype within an organism, can constrain natural selection from fixing resistant genotypes (Roff and Fairbairn, 2007). Potential costs of resistance include direct damage to host tissues due to immune activity and correlated reduction in investment in other physiological traits, including alternative immune functions, metabolism, and reproduction. Which investment strategy is most favourable will depend on the strength of pathogen pressures and on selection acting on other fitness traits of the organism.

An experimental approach that has been used to study genetic trade-offs is artificial selection for increased resistance to infection and subsequent measurement of correlated changes in other fitness traits. This method identifies costs of resistance, defined as changes in traits that reduce fitness in selected lines compared with unselected lines. Artificially selecting the Indian meal moth, Plodia interpunctella, for increased resistance to granulosis virus infection led to correlated increases in larval development time and pupal weight and a decrease in egg viability in selected lines (Boots and Begon, 1993). Selection in D. melanogaster for resistance to parasitoid or fungal infection led to a correlated decreases in larval competitive ability and adult fecundity, respectively, in the absence of infection (Kraaijeveld and Godfray, 1997, 2008). Costs that are measured in artificial selection lines should be interpreted with caution, however, as selection experiments can sometimes result in the fixation of rare alleles with large phenotypic effects that are not representative of functional genetic variation in natural contexts. For example, A. gambiae mosquitoes selected for refractoriness to Plasmodium infection achieve this through an increased melan-ization response (Collins et al, 1986) and high levels of cellular oxidative free radicals that are extremely damaging to host cells (Kumar et al, 2003). Natural resistance in wild populations of A. gambiae, however, is generally accomplished with a melaniza-tion-independent mechanism (Riehle et al, 2006), and is likely to be less costly or damaging than mechanisms seen in laboratory-selected lines.

A more relevant, but much subtler, measurement of genetic trade-offs is obtained by measuring genetic correlations between traits in naturally occurring, unselected genotypes. This is commonly done by measuring phenotypes in genetic clones or in individuals' with known genetic related-ness and estimating the genetic contribution to the phenotype. In D. melanogaster, genotypes with high resistance to bacterial infection had low fecund ity in the absence of infection in a food-limited environment (McKean et al., 2008). In the pea aphid Acyrthosiphon pisum, clonal lines with high resistance to attack by the parasitoid wasp Aphidius ervi had reduced fecundity (Gwynn et al., 2005). However, in this case, resistance to parasitoids can be conferred by bacterial endosymbionts, so the genetic basis for this trade-off may be mediated by factors outside the host genome. In both examples, the cost of resistance is a decrease in reproductive fitness.

The ultimate goal is to identify the genetic architecture underlying trade-offs. Quantitative trait locus (QTL) mapping has been used to locate these genetic regions. This approach relies on contrived crosses between chosen parents to establish pheno-typically variable recombinant progeny. Genetic markers are then genotyped at periodic intervals across the genome, allowing the localization of genomic regions encoding the phenotypic variation without relying on a priori candidate genes. QTL mapping, however, lacks the resolution to identify specific genes or alleles. In the red flour beetle T. casteneum and in the bumble bee Bombus terrestis, simultaneous mapping of immune and fitness traits found that loci associated with immune phenotypes occasionally co-localized with QTLs involved in fecundity, viability, and body size (Zhong et al, 2005; Wilfert et al, 2007a). There are two potential genetic mechanisms that could cause genetic correlations between immune and fitness traits. Genetic correlations can be caused by plei-otropy, where a single gene influences multiple traits. Trade-offs are due to antagonistic pleiotropy, where a single allelic variant of a gene has a positive effect on one trait but a negative effect on the other. Alternatively, allelic variants of distinct genes affecting the two traits may be in l inkage disequilibrium due to physical proximity on a chromosome, and thus these variants are coordi-nately passed to the offspring. Selection acts simultaneously on traits that are correlated by either pleiotropy or linkage disequilibrium. However, only antagonistic pleiotropy places a long-term constraint on selection because recombination can degrade correlations based on linkage disequilibrium. QTL mapping relies on experimentally generated linkage disequilibrium that spans much greater physical distances than are observed in natural populations, so it is relevant to follow QTL-based studies of genetic correlations with field-based studies to determine whether the traits co-segregate in nature.

Trade-offs have been also identified within the immune response. For example, in B. terrestis, lines selected for increased resistance to trypanosome infection also had a higher investment in a phe-noloxidase response coupled with a lower investment in AMP response (Wilfert et al, 2007b). The Egyptian cotton leafworm, Spodoptera littoralis, demonstrated positive genetic correlations among haemocyte density, cuticular melanization, and phenoloxidase activity, but a negative genetic correlation between haemocyte density and lysozyme-like antibacterial activity (Cotter et al., 2004). A different result is obtained from females of the mealworm beetle Tenebrio molitor, where cuticular melanization shows a negative genetic correlation with haemocytes and phenoloxidase, suggesting that the genetic architecture of these correlations can vary between species (Rolff et al., 2005). These results demonstrate that increased investment in one component of the immune response can come at a cost to other immune functions, and indicate the potential for trade-offs within the immune response to place constraints on the evolution of global resistance.

Thus far, all resistance measures have been considered only in a single environment; however, the optimal immune strategy can be expected to vary based on environmental conditions (Lazzaro and Little, 2009). Selective pressures are heteroge-neously distributed in the environment. Abiotic factors such as day length, temperature, and moisture vary between populations, affecting development time, metabolic flux, and other traits, and also altering the composition of pathogen communities and nutrient availability. Allelic variants in some genes respond differently to changes in the environment, termed genotype-by-environment interactions. If a genotype is particularly favoured in certain conditions, local adaptation to the proximate environment can occur. Temperate and tropical populations of D. melanogaster varied significantly in their resistance to the generalist fungal pathogen Beauveria bassiana (Tinsley et al, 2006) and bacterial pathogen Providencia rettgeri (Lazzaro et al, 2008). Considerable genotype-by-environment interaction was observed in resistance of D. melanogaster to P. rettgeri infection across multiple temperatures. Despite that observation, temperature populations were on average more resistant to P. rettgeri than the tropical one at lower temperatures, which potentially reflects adaptation to the local temperature. Spatial heterogeneity in the environment can lead to the maintenance of multiple resistance alleles if local adaptation is sufficiently strong to withstand erosion by migration and gene flow.

The magnitude, or even the existence, of genetic trade-offs can also vary between environments. In natural and laboratory settings, infestation by the mite Macrocheles subbadius negatively affects the fertility and body size of its host, Drosophila nigrospiracula (Luong and Polak, 2007). There is genetic variation for resistance to mites, which in this case is mediated by an avoidance behaviour. It has been demonstrated that, similar to D. melanogaster selected for parasitoid resistance, lines selected for mite resistance also suffer a cost in terms of decreased larval competitive ability. Manipulating the environment with high temperatures and increased larval density to create stressful conditions tends to increase costs of resistance. For instance, in previously considered examples from D. melanogaster, resistance to bacterial infection was correlated with low fecundity only in a food-limited environment (McKean et al, 2008), and larval success of parasitoid-resistant larvae was compromised only under crowded conditions (Kraaijeveld and Godfray, 1997). In all of these cases, selection can act independently on the traits in a non-stressful environment but the traits are constrained to each other under resource-limited conditions. Genetic variation for different allocations of resources between resistance and fitness traits can be maintained by environmental heterogeneity since the optimal investment strategy will be context-dependent (Roff and Fairbairn, 2007). Selection on these variants will be inefficient because trade-offs will only be apparent in certain conditions.

The host immune response faces a special obstacle in evolving immunity: the immune system must respond to living organisms that are them selves free to evolve. Its pathogen 'environment' is capable of rapid evolution, often much more quickly than the host. Analogous to genotype-by-environment interactions, a genotype-by-genotype interaction occurs when the efficacy of a host resistance genotype is dependent on the genotype of the pathogen. Antagonistic pleiotropy can occur in this context if resistance to one pathogen genotype comes with susceptibility to another. The specificity of these interactions can allow for temporal fluctuations in host and parasite genotypes in a frequency-dependent manner. Such fluctuations are generally difficult to measure experimentally, but have been observed natural populations of the snail host Potamopyrgus antipodarum and trematode parasite Microphallus sp. as well as in the crustacean host Daphnia magna and bacterial parasite Pasteuria ramosa (Dybdahl and Lively, 1998; Decaestecker et al., 2007). In both cases, resistant host genotypes are at an advantage when they are rare because their infective parasite genotypes are also rare, allowing resistant host genotypes to then to rise in frequency. This leads to a time-lagged increase in the infective parasite genotype, causing the host advantage to decline, subsequently reducing the frequency first of the host genotype and then the parasite genotype. This type of co-evolution is probably rare, occurring only when a parasite infects a narrow species range of hosts, allowing for specific, reciprocal adaptation, and when the parasite greatly reduces the fitness of the host such that selective pressure on resistance is high. In reality, many parasites are likely adapting to multiple host and impose only small reductions of fitness, placing more diffuse selective pressures on their hosts.

Environmental heterogeneity in pathogens and pathogen genotypes can lead to spatial adaptation to local pathogen pressures (Woolhouse et al., 2002). Genotype-by-genotype interactions between hosts and pathogens allow for adaptation to proximate pathogen pressures. Experimental evolution has been used to demonstrate the potential for local adaptation. In an experiment where P. ramosa was serially passaged for several generations on D. magna, it evolved high levels of infectivity on the host used for passage and in some cases lost virulence on non-passaged hosts (Little et al., 2006). This indicates that parasites can adapt to current hosts, perhaps at a cost of infecting alternate hosts, in only a few generations. Spatial variation in resistance can be detected by comparing the success of infection between host-parasite combinations that are either sympatric (local) or allopatric (foreign). Although most theoretical models predict that the parasite should be most successful in sympatric infections, in practice both parasite local adaptation and maladaption are observed (Woolhouse et al, 2002). In A. gambiae, a locus that was found to control encapsulation response to the malaria parasite Plasmodium falciparum was strongest against allopatric infections (Niare et al., 2002). Another locus restricting infection intensity was strongest against sympatric infections. Despite the opposite directions of these responses, both findings demonstrate population variation in resistance. In some cases, host resistance and parasite virulence have been observed to covary. The parasitoid Asobara tabida has been reported to have the highest virulence in the Mediterranean and lower virulence in northern Europe (Kraaijeveld and Godfray, 1999). D. melanogaster, the host, was observed to have the highest resistance in the Mediterranean and southern Europe, and low resistance in northern Europe, evidence of adaptation to local parasitoid pressures.

Tremendous variation in immunocompetence exists in extant natural populations. Trade-offs within the immune response and between immuno-competence and other fitness components constrain the ability of natural selection to drive resistant genotypes to fixation. Variation in trade-offs is maintained in part by environmental variation, whereby the costs associated with a particular genotype are context-dependent. Genotype-by-environment interactions and local adaptation can potentially lead to the maintenance of multiple polymorphisms in heterogeneous environments. Furthermore, the pathogen 'environment' is itself evolving. These forces in combination oftentimes limit the evolution of a single globally resistant genotype.

Was this article helpful?

0 0
How To Bolster Your Immune System

How To Bolster Your Immune System

All Natural Immune Boosters Proven To Fight Infection, Disease And More. Discover A Natural, Safe Effective Way To Boost Your Immune System Using Ingredients From Your Kitchen Cupboard. The only common sense, no holds barred guide to hit the market today no gimmicks, no pills, just old fashioned common sense remedies to cure colds, influenza, viral infections and more.

Get My Free Audio Book

Post a comment