Deleterious mutations and the evolution of sex
In spite of decades of intense debate, the evolutionary reasons for sex are still unknown. In the light of the 'two-fold cost of sex' (Maynard Smith 1971; Williams 1975), a plausible short-term advantage to sex must be found to explain its maintenance. At present, two hypotheses seem to predominate (Crow 1994; Hurst and Peck 1996): the Parasite Hypothesis (e.g. Hamilton 1980), emphasising the selection pressure created by the presumed abundance of parasites, and the Deterministic Mutation Hypothesis (further referred to as DMH; e.g. Kondrashov 1988), which emphasises the necessity of sex to eliminate recurring deleterious mutations. As the number of theoretical solutions to the problem has risen to about 20 (Kondrashov 1993), the lack of knowledge of the genetics of natural populations that is needed to discriminate between the various ideas, has become inconvenient.Progress on this topic can be made in two ways: either by testing the unique assumptions of a specific hypothesis, or by finding and testing the discriminating prediction. The latter has proven to be very hard: most predictions made by specific hypotheses are not unique (Hurst and Peck 1996). Especially the predictions of the Parasite Hypothesis resemble those of the DMH (Hamilton et al. 1990). For instance, both hypotheses predict the predominance of sexual reproduction in saturated environments, where selection is truncation-like due to limited space or nutrients (i.e. if selection ranks individuals with respect to fitness, and subsequently truncates the number of individuals by eliminating those with low fitness). Thus, it may be more revealing to test the assumptions that are unique to the various theories.The present thesis has aimed at testing the DMH. The value of this hypothesis is not only its theoretical generality (deleterious mutations are inextricably bound up with DNA replication), but also its simple basis of only two discriminating assumptions: (1) a sufficiently high rate of deleterious mutations, and (2) synergistic interaction between deleterious mutations. These two assumptions have been addressed by performing experiments designed to generate estimates for the relevant parameters involved in the assumptions.MethodologyFour of the five chapters describe attempts to reveal the relationship between deleterious mutation number and fitness. More specifically, the question is raised whether mutations show synergistic interaction, i.e. whether mutations amplify each other's negative fitness effect. However, the lack of direct information on the number of deleterious mutations carried by specific individuals prevents a quick answer. Since a deleterious mutation is defined by its negative effect on fitness, information on the number of deleterious mutations can only be inferred from fitness measurements. In the chapters 2 and 3, two simple tests are proposed to derive relevant information on the relationship between mutation number and fitness from a sexual cross between two parents.The test proposed in chapter 2 ('means test') is based on comparing the mean log fitness of parents and offspring: the concave relationship expected under synergism predicts a higher mean log fitness of the offspring, if both parents carry a sufficiently different mutation number. (The logarithm of fitness is considered, since the absence of mutation interaction means multiplicative mutation effects.) If both parents carry similar mutation numbers, synergism predicts a lower mean log fitness of the sexual offspring. Thus, for the 'means test' information on the difference in parental mutation number is needed. In chapter 3, an even simpler test ('skewness test') is proposed, which only considers the fitness of the offspring from a cross. The skewness of the log fitness distribution of the offspring reflects information on the nature of epistasis (i.e. interaction) between alleles at loci controlling fitness that segregate in the cross. This is so, because genetic recombination between the parental genomes produces a symmetrical distribution of the number of mutations among the offspring from a cross. If these mutations do not interact, they act additively at a log-scale and will produce a symmetrical log fitness distribution as well. Synergistic interaction is reflected by negative skewness at a log-scale.Application of the means test to Chlamydomonas and AspergillusIn chapter 2, the means test is applied to a number of crosses between strains of Chlamydomonas moewusii. Parental mutation number is manipulated by treating strains with different doses of ultraviolet radiation to cause additional deleterious mutations. Fitness of parents and offspring is measured by measuring the two logistic parameters of the asexual part of the life cycle in batch culture: maximum growth rate ( r ) and carrying capacity ( K ). These two parameters are thought to be relevant predictors of the fitness of asexual unicells under different ecological conditions: r in a disturbed, and K in a saturated and constant environment, but together they provide a complete description of the fitness (Bell 1991). The results show a significantly lower mean log fitness of the offspring relative to their parents in the 'high UV cross' for both r and K . Since in the high UV cross the parental fitnesses have become more similar than in the control cross, it is argued that their mutation number has probably become more similar as well. Therefore, the lower mean log offspring fitness in this cross reveals evidence for synergistic interaction between UV-induced mutations with respect to these fitness parameters.In chapter 5 , the means test is applied to strains of the asexual filamentous fungus Aspergillus niger. In this study, marker mutations are used as a model system to study epistasis between slightly deleterious mutations. Strains carrying different marker combinations are constructed by using the parasexual cycle of A. niger. The benefit of using marker mutations is that individual mutation number is accurately known. Here, the means test is applied by comparing the mean log fitness of a strain with for instance four markers and the wild-type strain, with the mean log fitness of all strains with intermediate numbers of these four markers: synergism is reflected by a higher mean log fitness of the 'intermediate strains'. The marker mutations involved are two colour, two resistance, and five auxotrophic mutations. To convert the auxotrophic mutations into slightly deleterious ones, two components of the fitness are measured on supplemented medium: the mycelial growth rate, which correlates strongly with the rate of spore production, and the maximum spore production when nutrients and space are limited. The results show highly significant interactions between markers for both fitness components, but no prevalent synergism or antagonism. Thus, it is likely that some marker combinations show antagonistic interaction, while other combinations interact synergistically. A tendency towards synergism is demonstrated for the maximum spore production, but not for the mycelial growth rate.Application of the skewness test to Chlamydomonas, fungi and plantsIn chapter 3 , the skewness test is applied to two Chlamydomonas crosses. Here, the focus of study is interaction between naturally accumulated instead of UV-induced (chapter 1) mutations. For this purpose, two strains are crossed that have been kept without sex in the laboratory for over 60 years. It is argued that the frequent transfer of samples of cells to fresh medium will have caused the accumulation of numerous deleterious mutations due to the action of Muller's ratchet (i.e. the random loss of the least-mutated cells; Muller 1964). As a control, two strains are crossed that presumably had not accumulated many mutations, since they were isolated only recently. The results show significant negative skewness of the distribution of log K in the 'accumulation cross', together with a highly significant increase of the genetic variance for K in this cross relative to the control cross. These results suggest that deleterious mutations have accumulated in the parents of this cross, and that they show synergistic interaction with respect to K. The lack of significant skewness of the log r distribution suggests no interaction between mutations there, although this result is less robust than the results for K since the power of detecting skewness is rather low.Inferring conclusions from skewness on the nature of epistasis depends on the absence of other sources of skewness. For example, a skewed error variance that is large relative to the genetic variance may affect the skewness. Furthermore, the skewness test has originally been developed for haploids. In diploid organisms, dominance, causing the heterozygote to have a different fitness than the mean of both homozygotes, may cause skewness as well. In chapter 6 , these alternative sources of skewness are studied in an analysis of data gathered from the literature on a number of plant and fungal species. The fitness characters involved include the mycelial growth rate for fungi, and earliness, resistance against various pathogens, seed number and pollen fitness for plants. The results on the mycelial growth rate are ambiguous, but the skewness observed in the plant fitness components is almost exclusively negative, indicating synergistic interaction between the loci controlling these characters. It is argued that, due to the likelihood of a concave relationship (optimum curve) between fitness components and total fitness, the synergism found is likely to apply to total fitness as well. An optimum relationship between a fitness component and total fitness may arise from trade-offs between different fitness components. A special case is the synergism found between resistance loci: this result gives also support to the significance of the Parasite Hypothesis if selection for resistance is the main determinant of fitness (Hamilton et al. 1990). It is demonstrated that the negative skewness observed is probably not due to a skewed error variance or dominance. The latter result justifies a more general application of the skewness test to diploids. Taken together, these results provide substantial empirical support to a fairly general occurrence of synergism between deleterious mutations in plants.Implications for the ecology of sexThe different results for the fitness parameters r and K in chapter 3 and mycelial growth rate and maximum spore production in chapter 5, provide support for the notion that synergism between deleterious mutations may depend on the ecological control of fitness: synergism may be only significant for fitness components like competitive ability, that are important in saturated environments. The extreme form of synergism is truncation selection, where fitness is unaffected up to a certain mutation number but drops sharply beyond. It has been notified that truncation-like selection is expected in saturated environments, where fitness is determined by competitive ability rather than exponential growth rate, due to limitation of nutrients and space (Crow and Kimura 1979; Crow 1994; Hamilton et al. 1990). Also on metabolic grounds synergism has been predicted to depend on the ecological control of fitness: only if selection favours a constant optimal concentration of some metabolic intermediate, the effects of mutations on fitness are synergistic (Szathmáry 1993). If selection favours the most rapid synthesis of the end product of the (e.g. growth limiting) pathway, the effects of mutation are predicted to be antagonistic.The carrying capacity ( K ) measured in Chlamydomonas, and the maximum spore production measured in Aspergillus are expected to predict fitness in density- regulated situations, while the maximum growth rate (r) and the mycelial growth rate rather represent fitness if nutrients and space are abundant. The selective finding of synergism for K and (of a tendency towards synergism) for the maximum spore production, but not for r and the mycelial growth rate in these two chapters, suggests that the predictions may be valid. If so, they provide an alternative explanation for the predominance of sex in saturated environments: sex is maintained in these environments since it facilitates selection against deleterious mutations that affect the relevant fitness components there, while sex exerts no advantage by purging the genome from mutations in disturbed environments because the necessary synergism is absent for the relevant fitness components for these environments. So far, the ecological predominance of sex in saturated environments has been explained by the Parasite Hypothesis (Trivers 1985; Hamilton et al. 1990) and by other hypotheses as well (Hurst and Peck 1996).The rate of deleterious mutationsChapter 4 addresses the other assumption of the DMH, i.e. whether the per-genome rate of deleterious mutations is sufficiently high. For this purpose a mutation- accumulation experiment, derived from a classical similar experiment with Drosophila by Mukai (1964), is performed in the asexual filamentous fungus Aspergillus niger. During 60 generations 20 mutation-accumulation (MA) lines with a common ancestor have been maintained under conditions of minimal selection to allow accumulation of mutations. At the end of mutation-accumulation, fitness of ancestral strain and mutation-accumulation lines at generation 30 and 60 have been measured in competition with a reference strain with differently coloured spores. From the expected decrease of mean fitness and increase of the genetic component of the variance in fitness, a lower limit of the deleterious mutation rate ( U ) can be calculated. Unexpectedly, however, the ancestral strain appears to have a lower fitness than the MA lines. But, as expected, the fitness of the MA lines decreases significantly between generation 30 and 60 of mutation- accumulation. It is argued that the low fitness of the ancestral strain is probably due to a variable physiological state of the spores of the ancestral strain. The estimate of U obtained from a comparison between generation 30 and 60 is at least 0.19 deleterious mutations per genome per generation. This is the first estimate of the deleterious mutation rate for a lower eukaryote. The result is consistent with the few estimates available for Drosophila (Mukai et al. 1972) and a number of plant species (Charlesworth et al. 1990; Charlesworth et al. 1994; Johnston and Schoen 1995). However, the confidence intervals of our estimates are rather wide, which makes interpretation speculative.ConclusionsThe contribution made by the work described in this thesis is two-fold. In the first place, two simple experimental tests are proposed for the assumption of the DMH that deleterious mutations show synergistic interaction. These tests have proven to be simple and widely applicable, and may be valuable tools for further studies on the nature of epistasis (Hurst and Peck 1996).Secondly, experimental data are generated that are relevant to testing both assumptions of the DMH. The application of the two tests for the nature of epistasis to a variety of experimental data involving the unicellular alga Chlamydomonas and the filamentous fungus Aspergillus , and to literature data on a number of fungal and plant species, has revealed a fairly general occurrence of synergistic epistasis between deleterious mutations. The results suggest also that synergism may be more pronounced in plants than in lower organisms like unicellular algae and fungi. The finding of the possible dependence of synergism on ecological conditions, which is consistent with theoretical predictions and with the ecology of sex, may lead to a refinement of the applicability of the DMH. Also, a first estimate of the rate of deleterious mutations (relevant to the other assumption of the DMH) in a lower eukaryote ( Aspergillus ) is given, together with the methodology to obtain such estimate. The estimate obtained is too low to support the DMH as explanation for sex in lower eukaryotes. However, the uncertainty revealed by a wide confidence interval makes this result tentative.In sum, the results obtained on the nature of epistasis provide conditional support to the DMH: sex may be maintained by facilitating selection against deleterious mutations in plants, and in lower eukaryotes that live in saturated environments. However, conclusions on its plausibility are hindered by the lack of knowledge of the deleterious mutation rate. The few estimates available suggest that the DMH may provide a significant advantage to sex in higher organisms like Drosophila and plants. The constant per-base-pair deleterious mutation rate that is consistent with the present data (including the recent estimate for E. coli ), suggests that its mechanism may not be able to provide the sole explanation for sex in lower eukaryotes. However, also in these organisms the DMH may add to the advantages provided by sex due to other mechanisms, and consequently help its maintenance.The future of the evolution of sexFurther study of the DMH is necessary and should primarily concentrate on estimating the deleterious mutation rate in various organisms. Such estimates would be especially interesting for lower anisogamous eukaryotes, because there the coexistence of sexual and asexual reproduction urges the finding of a short-term sex advantage most explicitly. Furthermore, data on the deleterious mutation rate may discriminate between the DMH and the Parasite Hypothesis unambiguously, while finding synergism can sometimes be interpreted as support for the Parasite Hypothesis as well (if synergism is for resistance against parasites, see discussion chapter 6). Mutation-accumulation experiments, like the one described in chapter 4, provide the most direct estimates. It is conceivable that various fitness components may be subject to different rates of deleterious mutations. Therefore, measuring various components of the fitness of mutation-accumulation lines may reveal information that is relevant to the ecology of sex, analogous to the finding of a possible ecological dependence of synergism.However, the search for synergism should not stop. Knowledge of its generality may help to differentiate the applicability of the DMH. Improvements on the detection of synergism should be possible. Metabolic control theory (Kacser and Burns 1973, 1979) has proven to be a useful tool for predicting the nature of interaction between enzyme mutations (Szathmáry 1993). Predictions from such theoretical studies are well-defined and can easily be tested in simple organisms by comparing the relevant fitness components of double and single mutants (in ways similar to that of the experiment described in chapter 5). Better knowledge of the metabolic conditions for synergism can lead to refined predictions of the ecological predominance of sex that may be unique to the DMH.In general, progress on the problem of the evolution of sex is best helped by finding direct and exclusive support for individual hypotheses. Studying the combined effect of sex advantages provided by different mechanisms might be helpful to reveal possible synergistic or antagonistic interaction between the separate advantages. However, accurate knowledge of the value of individual hypotheses is a prerequisite. The evidence required may either concern the discriminating prediction or the unique assumption(s) of any hypothesis, Since most predictions on the prevalence of sex under specific conditions appear not to be unique to single hypotheses (Hurst and Peck 1996), attempts to find a truly discriminating prediction may be futile. Furthermore, comparing the fitness of sexuals and asexuals in experimental settings meant to test a specific hypothesis is tricky, since it may be difficult to exclude all other possible causes that may generate an advantage to sex. Parameter estimates that are relevant to specific assumptions provide better controlled and more direct support.
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Doctoral thesis biblioteca |
| Language: | English |
| Published: |
Landbouwuniversiteit Wageningen
|
| Subjects: | accessory chromosomes, chromosomes, evolution, origin, phylogenetics, phylogeny, sex chromosomes, accessorische chromosomen, chromosomen, evolutie, fylogenetica, fylogenie, geslachtschromosomen, oorsprong, |
| Online Access: | https://research.wur.nl/en/publications/deleterious-mutations-and-the-evolution-of-sex |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|