* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download full text pdf
Survey
Document related concepts
Adaptive evolution in the human genome wikipedia , lookup
Epigenomics wikipedia , lookup
Deoxyribozyme wikipedia , lookup
Cancer epigenetics wikipedia , lookup
Behavioural genetics wikipedia , lookup
History of genetic engineering wikipedia , lookup
Group selection wikipedia , lookup
Heritability of IQ wikipedia , lookup
Epigenetics of neurodegenerative diseases wikipedia , lookup
Population genetics wikipedia , lookup
Dual inheritance theory wikipedia , lookup
Epigenetic clock wikipedia , lookup
Microevolution wikipedia , lookup
Koinophilia wikipedia , lookup
Behavioral epigenetics wikipedia , lookup
Epigenetics wikipedia , lookup
Nutriepigenomics wikipedia , lookup
Transcript
Non-Genetic Inheritance Mini-review • DOI: 10.2478/ngi-2013-0003 • NGI • 2013 • 27-32 Non-genetic Inheritance in Evolutionary Theory: a Primer Abstract Evolutionary biology traditionally equates inheritance with transmission of genes from parents to offspring. However, recent literature calls for considering ‘non-genetic inheritance’ in evolutionary theory. These calls have met with substantial scepticism. What is more, they appear to have caused further confusion both with respect to what inheritance is and what types of inheritance mechanisms are evolutionarily consequential. Building on previous work, we make use of the Price Equation to outline a general discussion of how non-genetic inheritance can affect phenotypic change within populations, exemplified by epigenetic inheritance. This shows that integrating non-genetic inheritance in evolutionary theory will require specific attention to the developmental processes that shape the relationship between the fitness of parents and the phenotype of their offspring. Tobias Uller1*, Edward Grey Institute, Department of Zoology, University of Oxford. 1 Heikki Helanterä2 Centre of Excellence in Biological Interactions, Department of Biosciences, University of Helsinki 2 Keywords Heredity • Price Equation • Maternal effects • Parental effects • Plasticity © 2013 Tobias Uller et al., licensee Versita Sp. z o. o. This work is licensed under the Creative Commons Attribution-NonCommercialNoDerivs license (http://creativecommons.org/licenses/by-nc-nd/3.0/), which means that the text may be used for non-commercial purposes, provided credit is given to the author. Introduction In biology, inheritance or heredity typically refers to the phenomenon that offspring resemble their parents. Many modern definitions of inheritance specifically refer to the transmission of genes between generations as the causal mechanism behind this similarity (e.g., [1,2]). Indeed, the reason that offspring resemble their parents more than they resemble unrelated individuals can often be linked to differences in the DNA sequences that they inherited from their parents. This goes both for inheritance of species-typical phenotypes (e.g., why human babies look human rather than chimpanzee and vice versa) and for inheritance of differences between lineages within populations (e.g., why each of the authors of this paper resembles their father). In many areas of biology, including evolutionary theory, the focus is on the latter of these inheritance concepts, i.e., inheritance of differences in phenotypes within populations rather than inheritance of species-typical phenotypes [3]. DNA is clearly important for the inheritance of differences in phenotypes, but any mechanism that contributes to parent-offspring similarity within populations is potentially of evolutionary relevance. This could be due to causal effects of the parental phenotype on offspring phenotype, which are often referred to as parental effects [4]. These include molecules associated with DNA, amount and location of egg mRNA, yolk nutrients, and different forms of behavioural interactions. In all those cases, the contribution from the parent can depend on Received 10 July 2013 Accepted 22 August 2013 various aspects of the parental phenotype. But inheritance also involves stable transmission of biomolecules down generations, where this transmission does not depend on the parental phenotype. For example, the amount and sequence of DNA that is transferred from parent to offspring is typically not affected by the phenotype of the parent. The stability during transmission is one of the crucial features of DNA that makes it warranted to speak of DNA as an inheritance system in a more strict sense than most parental effects [5]. However, DNA is not unique in this regard. Recent findings that also epigenetic variants can be stably transmitted down generations mean that long-run transgenerational inheritance of differences in phenotypes could be epigenetic rather than genetic [6,7]. Furthermore, it is possible that the action of the parent ‘reconstructs’ the environmental conditions that are conducive to the expression of the same phenotype in the next generation, thereby generating transgenerationally stable differences between lineages [4]. This is true, for example, when animals return to their natal habitat type (e.g., host plant) for reproduction [8], thereby setting the stage for transgenerationally stable differences between lineages. In addition to the classic treatments of the biology of inheritance (see e.g. discussion in Ch. 7 in Ref. [9]), there is now a small but growing literature that discuss conceptual challenges that arise from non-genetic inheritance (accessible introductions include [6,10-13]). We will not review this literature here. Instead, our aim is to show how non-genetic inheritance * E-mail: [email protected] 27 Unauthenticated Download Date | 6/17/17 9:26 AM T. Uller, H. Helanterä can be integrated into evolutionary theory using a mainstream mathematical approach to phenotypic evolution. In doing so, we do not wish to imply that this is all there is to integrating non-genetic inheritance in evolutionary thinking (indeed, we argue elsewhere that more fundamental conceptual shifts are needed, [11,14]). But we do believe that it will be helpful to see that the consequences of non-genetic inheritance for phenotypic evolution can be approached without the need to endorse more radical claims about the need for an extended evolutionary synthesis. Because the Price Equation does not make any assumptions of the underlying mechanisms of parent-offspring similarity it can of course be used to derive the standard single-locus and quantitative genetic formulations of evolutionary change that are found in evolutionary biology and genetics textbooks [23]. However, none of the terms in Eq. (1) are obviously referring to the covariance between parental and offspring phenotypes, which makes this difficult to see. Assuming that the parentoffspring regression is linear, we can rewrite the equation as 1 (2) ∆z = β o Cov w, z + E ∆z + Cov( w, e) w Non-genetic inheritance and the Price Equation ∆z = 1 ( Cov ( w, z ) + E ( w∆z ) ) w ( ) ( βz ( ) ) ) , z , which is referred to as the heritability of the phenotype (using selected parents, [25]; the slope of the regression is equal to the phenotypic covariance between parents and offspring divided by the total phenotypic variance in parents. Also note that in quantitative genetics, heritability is often defined in terms of genetic and phenotypic variance instead of a regression; see chapter 7 in Ref [23] for discussion). Thus, if the remaining terms are zero, this equation corresponds to the breeder’s equation, usually written as R=h2S [28], which can be found in every undergraduate textbook on evolution. The second term, E ( ∆z ) , represents the expected phenotypic change in the absence of fitness differences. This is often referred to as ‘transmission bias’ [25], but it is important to realise that this term can be affected by a variety of developmental processes, including those that are independent of the parents [23]. Mathematically this is represented by the intercept of the regression and could be, for example, changes that occur because of intergenerational environmental change that affects phenotypic development (i.e., phenotypic plasticity) [23]. The third term consists of the covariance between fitness and the scatter around the regression line (i.e., the residuals, e), Cov( w, e) [25]. Equation 2 assumes that the regression is linear. If we are not making this assumption, Equation 2 becomes (1) where ∆z is the change in the average phenotype (e.g., height) in the population, w is the mean number of descendants per individual, cov w, z is the covariance between fitness and trait value, and E w∆z is the expected value of the product of fitness and the average phenotypic difference between parent and offspring in the absence of selection [20]. The terms within the parenthesis can be interpreted as the change due to differential reproduction and survival and the change that occurs as a result of reproduction and the mechanisms of inheritance, respectively (Ch 6 in [23]). The latter can be the result of any process that causes phenotypic change between generations, be it through the actual mechanisms of inheritance or changes in environmental conditions that affect development. Thus, not only parent-offspring covariance, but also developmental responses to environmental change that are shared among members of the population will affect the magnitude of the expected value of phenotypic change between generations [24]. Division by the mean number of descendants means that fitness is relative and not absolute. ( ( z ,z This decomposes change in population mean phenotype into three components (see [23,25,26] and Ch. 14 in [27] for mathematical details and how to get from Eq. 1 to Eq. 2). The first term on the right hand side consists of the covariance between fitness and trait value, Cov w, z , known as the selection differential, times the slope of the parent offspring regression, If we want to understand the consequences of different mechanisms of inheritance for phenotypic evolution, it is sensible to start with a model that does not make any a priori assumptions about those mechanisms [5,16] (perhaps counterintuitively, population genetic approaches can also be useful for modelling evolution under non-genetic inheritance, e.g., [17,18]). The Price Equation allows us to do this as it does not make any assumptions with respect to the underlying biology (indeed, the ‘ancestors’ and ‘descendants’ need not even be biological entities [19,20]; see [21-23] for discussion). Both Helanterä and Uller [15] and Day and Bonduriansky [16] have used the Price Equation to outline some potential evolutionary implications of non-genetic inheritance. Our approach here differs somewhat from both of these. Specifically, our main purpose in this brief paper is to use the Price Equation to explain, in a non-technical way, why non-genetic inheritance will affect phenotypic evolution, not to contrast different mechanisms of inheritance [15] or build mathematical models of specific scenarios [16]. The Price Equation can be written as ( ) ) o ∆z = 1 β o Cov ( w, z ) + E ( ∆z ) + Cov( w, z 0 z ) w z ,z ( ) (3) (see [25],Ch 14 in [27] for derivation). The difference to Eq. 2 is now that β z o , z is the best linear regression given the data, and that Cov( w, z 0 z ) is the covariance between the residuals for the regression of fitness on parental phenotype and the residuals for the regression of offspring phenotype on parental phenotype (i.e., parent-offspring regression; [25]) (Figure 1). Although it is not obvious that the residuals should ever covary, this is possibly quite common. As shown in Figure 1, the reason is that when we describe evolutionary change in this form of the Price Equation we are forcing the slope of the regression to be linear even if it may not be [27]. In fact, empirical 28 Unauthenticated Download Date | 6/17/17 9:26 AM Non-genetic Inheritance in Evolutionary Theory: a Primer Figure 1. Illustration of a situation that gives rise to a spurious response to selection. The relationship between parental phenotype and fitness (top left) and the relationship between parental and offspring phenotype (top right) are non-linear. As a result, the residuals for the best fitting linear regressions are non-random. This results in a negative covariance between the residuals (bottom graph), which is the spurious response to selection, Cov( w, z 0 z ) , in Equation 3. Figure adopted with modifications from [25,27]. studies suggest that the relationship between phenotypes of parents and offspring can be non-linear [24,29]. When we are forcing a linear regression through these data the residuals are biased across parental phenotypic values (Figure 1). Because fitness may also show non-linear relationships with phenotype (probably a common phenomenon), the residuals can be 0 correlated and hence Cov( w, z z ) will be non-zero. Heywood [25] termed this a ‘spurious response to selection’ to emphasize that this is the response that occurs once we have removed the direct (linear) effect of the parental phenotype on both fitness and the phenotype of its offspring. Thus, this response is not adaptive and may even counteract the response to selection. An alternative reason for why Cov( w, z 0 z ) could be non-zero is that, even if one or both regressions are linear, the residuals are correlated via a third variable [25]. Thus, for a given phenotypic trait, z , there may be a third variable that is correlated with both the residuals of trait z on fitness and the residuals of the parent-offspring regression. Heywood [25] discusses a case with breeding date in birds, where there is a spurious response to selection even when both regressions are linear. This is because a third variable, nutritional status, covaries with both the residuals of breeding date on fitness and the residuals of parental breeding date on offspring breeding date. In summary, Equation 2 describes the change in the population mean phenotype from one generation to the next in terms of the product of the covariance between phenotype and fitness (‘selection differential’) and the parent-offspring regression (‘heritability’) and two terms that can be affected by mechanisms of inheritance and development, including environmental effects (‘transmission bias’ and ‘spurious response to selection’). These components can subsequently be given biological meaning in models that track evolutionary trajectories under different forms of selection or establish evolutionarily stable equilibria (e.g., [26,30]). Quantitative genetics typically assume that the last two terms are zero (hence the Breeder’s Equation), but they may be non-zero even under pure genetic inheritance due to, for example, dominance [25,27]. However, both terms are more likely to be of significance if we include some additional biological reality by allowing non-genetic mechanisms of inheritance. For example, Day and Bonduriansky [16] constructed several models by explicitly separating the evolutionary response in a focal trait in terms of its genetic and non-genetic components, 29 Unauthenticated Download Date | 6/17/17 9:26 AM T. Uller, H. Helanterä and the environment make it more likely that there will be nonlinear relationships between parent and offspring phenotype or biased distribution of residuals of the regressions in Figure 1. In fact, non-linearity may be common whenever there are parental effects as they tend to skew the distribution of phenotypes from that expected under additive genetic variance [23]. For example, biased transmission stability of DNA methylation may result from passive loss of methylation with age. This can result in spurious responses for both reasons mentioned above (Figure 1). Firstly, it could lead to non-linear relationships between phenotypes in parents and offspring and hence residuals may become correlated even in the absence of a causal link. Secondly, age may covary both with the residuals of fitness for a focal trait (e.g., older individuals may be more experienced and thus have higher breeding performance for a given trait value) and the residuals of the parent-offspring regression (e.g., older parents may be less likely to transmit the same epigenetic mark as they themselves inherited because of stochastic loss of DNA methylation with age). This line of reasoning suggests that establishing the pattern of parent-offspring similarity (e.g., if it is linear), and its underlying mechanism (e.g., if there is environment-specific transmission of epigenetic states), is an important task if we are to understand and predict the extent to which epigenetic mechanisms contribute to short- and longterm evolution [16,24]. In summary, we advocate a perspective on inheritance that encompasses all the genomic and non-genomic resources by which parents contribute to offspring development and hence the recurrence of phenotypes (Ch. 4 in [39,40]) . Non-genetic inheritance contributes to more or less persistent recurrence of differences in phenotypes between lineages within populations and the heritability of phenotypes phenotypes can partly be due to non-genetic inheritance. Furthermore, non-genetic inheritance may be more likely than genetic inheritance to cause transmission bias and a spurious response to selection. Nongenetic mechanisms of inheritance are therefore potentially important for both short- and long-term phenotypic evolution. As our understanding of the role of developmental processes in evolution improves, the part of those processes that underlie heredity should take on a central role in evolutionary theory [6,12,14,16,24,41]. for which transmission rules were specified (their model differs somewhat from our discussion here as it did not address the role of the spurious response to selection). These models show that a wide range of non-genetic inheritance mechanisms can have consequences for the rate and direction of phenotypic (and genetic) evolution. Here we exemplify this with epigenetic inheritance, which has attracted substantial interest recently (e.g., [7,31-33]; see also [16] for a worked example). Epigenetic inheritance, such as DNA methylation, differs from genetic inheritance in several ways (e.g., [6]). Although epigenetic variants can be stably inherited through meiosis in multicellular organisms, such stability seems to be relatively rare and short lived compared to transmission of DNA sequence variation. Like DNA mutation, epigenetic variation can be environmentally induced but, unlike mutation, a broader range of environments are apparently able to modify epigenetic states, perhaps in non-random directions. The degree to which offspring pass on the same ‘epiallele’ as they received from their parents could therefore depend on the similarity of environments across generations, the parental phenotype, and perhaps the epigenetic state itself (discussion in [6,11,34,35]). Each of these has potential consequences for how epigenetic mechanisms contribute to phenotypic evolution. Firstly, transmission of epigenetic variants means that epigenetic mechanisms can cause offspring to resemble their parents, i.e., that the (linear) slope of regression of offspring phenotype on parental phenotype is non-zero. Thus, epigenetic mechanisms contribute to the overall heritability of a character [12,36,37]. The increasing evidence for transgenerational stability of epigenetic variants (at least partly) independently of DNA sequence [7,31] also suggest the potential for cumulative adaptive evolution via epigenetic inheritance. Naturally, the long-term consequences of epigenetic inheritance will depend on the stability of these variants, which itself can be a function of genetic, phenotypic and environmental change. Day and Bonduriansky [16] have shown that differences in the stability of epialleles can affect both evolutionary trajectories and equilibria of genotype and phenotype values within populations (see also [38]). Secondly, the environment-dependence of epigenetic inheritance is likely to cause significant transmission bias, which makes E ( ∆z ) non-zero as well. This affects the predicted evolutionary change from one generation to the next because epigenetic inheritance, or environmental epigenetic effects, causes phenotypes to change more or less predictably from one generation to the next even in the absence of parent-offspring covariance (e.g., due to a common plastic response in the population). Finally, epigenetic inheritance may be more likely than genetic inheritance to generate a spurious response to selection. The stochastic nature of epigenetic inheritance and its dependence on the phenotypic character state of the parent Acknowledgements We are grateful to Neil Youngson for the invitation to contribute this paper, two reviewers for their comments, and to all our colleagues who have taken their time to discuss non-genetic inheritance with us. TU is supported by the Royal Society of London and the European Union’s Seventh Framework Programme (FP7/2007-2011) under grant agreement no 259679. HH is supported by the Academy of Finland (grant number 135970). 30 Unauthenticated Download Date | 6/17/17 9:26 AM Non-genetic Inheritance in Evolutionary Theory: a Primer References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] Merriam-Webster Dictionary entry inheritance: http:// www.merriam-webster.com/dictionary/inheritance Oxford Dictionary, entry heredity: http://oxforddictionaries. com/definition/english/heredity?q=heredity; Mameli M., The inheritance of features, Biol. Philos., 2005, 20, 365-399 Uller T., Parental effects in development and evolution, In: Royle N.J., Smiseth P.T., Kölliker M. (Eds.), The evolution of parental care, Oxford University Press, New York, 2012 Shea N., Pen I., Uller, T., Three epigenetic information channels and their different roles in evolution, J. Evol. Biol., 2011, 24, 1178-1187 Jablonka E., Lamb M.J., Evolution in four dimensions. Genetic, epigenetic, behavioral, and symbolic variation in the history of life, MIT, Cambridge, MA, 2005 Jablonka E., Raz, G., Transgenerational Epigenetic Inheritance: Prevalence, Mechanisms, and Implications for the Study of Heredity and Evolution, Quart. Rev. Biol., 2009, 84, 131-176 Davis J.M., Patterns of variation in the influence of natal experience of habitat choice, Quart. Rev. Biol., 2008, 83, 363-380 Amundson, R., The changing role of the embryo in evolutionary thought. Cambridge University Press, New York, NY, 2005 Bonduriansky, R., Rethinking heredity, again. Trends Ecol. Evol., 2012, 27, 330-336 Uller T., Non-genetic inheritance and evolution. In: Kampourakis K. (Ed.), Philosophy of Biology: a Companion for Educators, Springer Verlag, 2013 Danchin É., Charmantier A., Champagne F.A., Mesoudi A., Pujol B., Blanchet S., Beyond DNA: integrating inclusive inheritance into an extended theory of evolution, Nat. Rev. Genet., 2011, 12, 475-486 Bonduriansky, R., Day, T. 2009. Nongenetic inheritance and its evolutionary implications. Annu. Rev. Ecol. Evol. Syst., 40,103-125 Uller T., Helanterä H., Heredity and evolutionary theory. In: Huneman P., Walsh D (Eds.), Challenges to evolutionary theory, forthcoming Helanterä H., Uller T., The Price equation and extended inheritance, Philos. Theor. Biol., 2010, 2, e101 Day T., Bonduriansky, R., A Unified Approach to the Evolutionary Consequences of Genetic and Nongenetic Inheritance, Am. Nat., 2011, 178, E18-E36 Cavalli-Sforza L.L., Feldman, M.W., Cultural transmission and evolution. Princeton University Press, Princeton, NJ, 1981 Laland K.N., OdlingSmee F.J., Feldman, M.W., The evolutionary consequences of niche construction: A theoretical investigation using two-locus theory, J. Evol. Biol., 9, 293-316 [19] Price G.R., Selection and covariance, Nature, 1970, 227, 520-521 [20] Price G.R., Extension of covariance selection mathematics, Ann. Hum. Genet., 1972, 35, 485-490. [21] Frank S.A., George Price’s Contributions to Evolutionary Genetics, J. Theor, Biol., 1995, 175, 373-388 [22] Frank S.A., Natural selection. IV. The Price equation, J. Evol. Biol., 25, 1002-1019 [23] Rice S.H., Evolutionary theory. Mathematical and conceptual foundations, Sinauer Associates inc, Sunderland, MA, 2004 [24] Rice S.H., The Place of Development in Mathematical Evolutionary Theory., J. Exp. Zool. B. Mol. Dev. Biol., 2012,318, 480-488 [25] Heywood J.S., An exact form of the breeder’s equation for the evolution of a quantitative trait under natural selection, Evolution, 2005, 59, 2287-2298 [26] Okasha S., Evolution and the levels of selection, Oxford University Press, New York, 2006 [27] Lynch M., Walsh J.B., Evolution and selection of quantitative traits. I. Foundations. 2013, http://nitro. biosci.arizona.edu/zbook/NewVolume_2/newvol2.html [28] Falconer D.S Mackay T.F.C., Introduction to quantitative genetics. Pearson Education Ltd, Harlow, England, 1996 [29] Gimelfarb A., Offspring-parent genotypic regression: how linear is it?, Biometrics, 1986, 42, 67-71 [30] Lehmann L., Feldman, M.W., Foster, K.R., Cultural Transmission Can Inhibit the Evolution of Altruistic Helping, Am. Nat., 2008,172, 12-24 [31] Lim J.P., Brunet, A., Bridging the transgenerational gap with epigenetic memory, Trends Genet., 2013,29, 176186 [32] Daxinger L., Whitelaw, E., Understanding transgenerational epigenetic inheritance via the gametes in mammals, Nat. Rev. Genet., 2012, 13, 153-162 [33] Youngson N. Whitelaw E., Transgenerational epigenetic effects, Annu. Rev. Genom. Hum. Genet., 2009, 9, 233257 [34] Ginsburg S., Jablonka, E., Epigenetic learning in nonneural organisms, J. Biosci., 2009, 34, 633-646 [35] Jablonka E., Epigenetic inheritance and plasticity: The responsive germline, Prog. Biophys. Mol. Biol., in press, DOI: 10.1016/j.pbiomolbio.2012.08.014 [36] Tal O., Kisdi E., Jablonka E., Epigenetic contribution to covariance between relatives, Genetics, 2010, 184, 1037-1050 [37] Furrow R.E., Christiansen F.B., Feldman M.W., Environment-sensitive epigenetics and the heritability of complex diseases, Genetics, 2011, 189, 1377-1387 [38] Geoghegan J.L., Spencer H.G., Exploring epiallele stability in a population-epigenetic model, Theor. Pop. Biol., 2013, 83, 136-144 31 Unauthenticated Download Date | 6/17/17 9:26 AM T. Uller, H. Helanterä [39] Oyama S., A systems view of the biology-culture divide. Duke University Press, 2000 [40] Badyaev A.V., Origin of the fittest: link between emergent variation and evolutionary change as a critical question in evolutionary biology. Proc. R Soc. Lond B., 2011, 278, 1921-1929 [41] Badyaev A.V., Uller, T., Parental effects in ecology and evolution: mechanisms, processes and implications, 2009, Phil. Trans. R. Soc. Lond. B., 2009, 364, 1169 1177 32 Unauthenticated Download Date | 6/17/17 9:26 AM