FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4294667079> ?p ?o ?g. }

• W4294667079 endingPage "1194" @default.
• W4294667079 startingPage "1188" @default.
• W4294667079 abstract "Reproductive isolation (RI) is often considered an essential component of speciation; however, its definition varies, and it is challenging to measure, making it difficult to compare across studies. To help overcome these difficulties, Westram et al. (2022) suggest to define RI as ‘a quantitative measure of the effect of genetic differences on gene flow’, specifically at neutral loci. Here, we consider this definition of RI in the context of epigenetic variation. We define epigenetics as molecular interactions with DNA that influence gene expression without changes in the underlying nucleotide sequence (see Box 1 for background on epigenetic variation). Furthermore, we consider a process as epigenetic, regardless of whether it is transmitted across generations or not. As we will see, intergenerational transmission of epigenetic state is not a prerequisite to influence RI. In the following, we establish a framework through which to better quantify epigenetic influences on population divergence and speciation by building upon the model of RI at neutral loci established by Westram et al. (2022). Two examples of epigenetic marks are DNA methylation, in which a methyl group is covalently bonded by a cytosine to form methyl-cytosine, and histone modifications, in which the tail of a histone (a component of DNA packaging nucleosomes) undergoes one of many forms of modification, for example, methylation, acetylation or phosphorylation (Figure B1). These tags can alter processes such as DNA packaging and gene expression, by influencing the action of DNA interacting proteins such as histones and transcription factors (Gibney & Nolan, 2010). These epigenetic marks are phylogenetically widespread (Feng & Jacobsen, 2011; Glastad et al., 2019) and can influence adaptive phenotypes (Baerwald et al., 2016; Rangani et al., 2012). Among epigenetic marks, DNA methylation is the best understood and most widely studied, particularly in the context of natural populations (Hu & Barrett, 2017; Kilvitis et al., 2014; Vandegehuchte & Janssen, 2014). DNA methylation state can be passed from parent to offspring (Yagound et al., 2020) and differs among populations in divergent environments (Dubin et al., 2015; Wogan et al., 2020; for examples of both see Heckwolf et al., 2020; Verhoeven & Preite, 2014). Epigenetics can underlie phenotypic plasticity. A large body of work has examined phenotypic plasticity (West-Eberhard, 2003), including its effects on gene flow and speciation (Fitzpatrick, 2012; Klemetsen, 2010; Otte et al., 2016; Thibert-Plante & Hendry, 2011). However, focussed studies that have directly examined epigenetic variation as a mechanism producing barriers to reproduction (Lafon-Placette & Köhler, 2015; Smith et al., 2016) and speciation (Greenspoon et al., 2022) are only now emerging. In their examination of RI, Westram et al. (2022) focus on two genetic loci, a locus under divergent selection between two environments and a neutral locus linked with the locus under selection. In an analogous fashion, we consider an epigenetic locus and a neutral genetic locus linked to it. A well-studied example of an epigenetic locus is the cytosine-guanine dinucleotide (CpG), which can be methylated to form mCpG (Feng et al., 2010; Zemach et al., 2010). One mechanism through which methylation state can change is in response to an external signal from the environment, and such change in methylation state can alter the phenotype of an individual without a change in nucleotide sequence (Angers et al., 2010; Schmid et al., 2018). We shall call this process induction. This induction of the methylated state can also have a variable degree of stability over time, being lost during the life of an individual, or during epigenetic remodelling in germ cells and early embryos of some organisms (Feng et al., 2010; Santos & Dean, 2004). We shall call this process erasure. We can, therefore, imagine an epigenetic locus whose state (in this example methylated or unmethylated) encodes a phenotype and a neutral genetic locus that is linked with such epigenetic locus (Figure 1a), similar to a genetic locus (Figure 1b). Consequently, we consider variation at the epigenetic locus being produced by induction and erasure, rather than mutation (Figure 1). As we will see, these differences can have consequences for RI. When two habitats provide strong divergent environmental induction, two populations can form distinct epigenetic-dependent phenotypes, which produce RI. If the epigenetic marks are completely erased between generations, the state of the epigenetic locus only effects the fitness of the migrants between habitats and not their offspring, thus producing a uniform and global (genome-wide) reduction in gene flow to all neutral loci across the genome. In this context, epigenetic variation will act like a geographic barrier, as described by Westram et al. (2022). In contrast, if the epigenetic mark is not erased between generations, there will be additional selection against recombination surrounding the epigenetic locus. This will cause a local (restricted to parts of chromosomes) reduction in gene flow, influencing neutral loci dependent on their linkage with the epigenetic locus. To highlight key similarities and differences between the effects of purely environmentally induced epigenetic versus genetically based barriers to neutral gene flow, we analyse the two-population model with divergent selection acting on a single target locus, as presented by Westram et al. (2022). Epigenetic state is usually mediated by genetic variation to some extent (Adrian-Kalchhauser et al., 2020), suggesting that much observed epigenetic variation may be due to underlying genetic variation. However, here we focus on epigenetic mechanisms independent of genetic sequence, as it is the simplest starting point and best corresponds to the framework of Westram et al. (2022). The key feature which dictates the effect of epigenetics on RI is transmission (v). When v = 0, there is no transmission of epigenetic state to the next generation; thus, rv = 1, and me consequently reduces to q + p(1 − s). Under such circumstances, the epigenetic barrier to neutral gene flow is entirely due to selection acting against the individuals migrating from population A into population B. As recombination is not involved, the proximity of the neutral allele n to the target locus does not bear on the strength of the epigenetic barrier (Figure 3). Selection would, thus, be acting on the genome as a whole and have a uniform, genome-wide effect on reducing neutral gene flow, analogous to geographic isolation as described by Westram et al., 2022), yet it is determined by the epigenetic state of the organism. In contrast, when v = 1, there is perfect intergenerational transmission of epigenetic state; thus, rv = r, and estimates of me for epigenetic and genetic differences are the same. In other words, a perfect transmission of an epigenetic modification in population B is equivalent to Mendelian inherence of genetically based variation (Figure 3). Between these two extremes, epigenetic marks may exhibit a range of transmissibility and span from geographic-like to genetic-like effects on RI (Figure 3). Thus, there may be instances of increased local RI centred around sites lacking genetic variation, though they will be weaker than those due to genetic polymorphisms. Whether epigenetic variation as considered in our model should be considered as a contributor to RI is an open issue. As described by Westram et al., RI must be based on genetic differentiation. Following this definition, an environmentally induced epigenetic state cannot produce RI since there is no genetic differentiation between populations. Thus, the reduction in gene flow due to selection directly against the epigenetic state of migrants should not be considered to reduce me, despite such epigenetic differences having the same effect (1 − s in Equation 2) as a genetic difference for reducing gene flow in the parental migrant generation. Moreover, when the environmentally induced epigenetic state is even weakly transmissible, there will be a local reduction in gene flow around the epigenetic locus despite a lack of genetic differentiation, which may be interpreted as evidence of RI in empirical observations. In such circumstances, selection on the migrant generation due to an epigenetic difference would be included in the estimation of me and associated with RI akin to a genetic difference. We leave it to the reader to decide their position on this question, as it underscores the difficulty in conceptually defining RI. While our model suggests that RI produced by epigenetic loci will generally be weaker than that produced by genetic loci, epigenetic barriers to gene flow are not necessarily negligible. The rapid loss of maladaptive epigenetic states due to low transmissibility (v < 1) means that, with strong gene flow, populations can be more divergent at epigenetic loci than equivalent genetic loci (q and p in Equation 2), producing greater interpopulation differences for selection to act on. There are observations of greater epigenetic than genetic differentiation among locally adapted populations (Dubin et al., 2015; Foust et al., 2016; Gugger et al., 2016; Herrera & Bazaga, 2016; Heckwolf et al., 2020; Johnson & Kelly, 2020; Lira-Medeiros et al., 2010; Platt et al., 2015; Richards et al., 2012; Wogan et al., 2020; for counterexamples see Herden et al., 2019; Keller et al., 2016; Robertson et al., 2017). The major caveat to these observations is that epigenetic differentiation at many loci may be caused by differentiation at genetic loci. Furthermore, large epigenetic divergence does not necessarily underly environmentally induced and locally adaptive phenotypes. There is much room to build additional nuance on top of this modelling framework. Here, we have presented an epigenetic model that is independent of genetic variation; however, as we emphasized above, genetic and epigenetic variation may not be independent. As explored by Greenspoon et al. (2022), epialleles may either promote speciation in accordance with genetic loci, or diminish speciation if they overwhelm genetic differentiation as the primary adaptive mechanism (see Table S1 for suggestions of how to incorporate further model complexity into the study of epigenetic RI). If there is any influence of epigenetics on gene flow, it is not straight forward to make inferences about RI by measuring divergence between populations at neutral loci. As per the definition of RI proposed by Westram et al. (2022), the measurement of RI by introgression of neutral loci among populations attributes all RI to genetic differences. Yet, some RI may be due to unobserved epigenetic mechanisms. What must we measure to discern genetic from epigenetic RI? An important first step in understanding the role of epigenetics in RI is to observe the extent of epigenetic variation among natural populations (for example Gugger et al., 2016; Venney et al., 2021; Wogan et al., 2020). Then, it is necessary to establish if epigenetic differences between populations contribute to RI. The contribution of epigenetic state to RI can be inferred from the degree to which epigenetic state covaries with locally adaptive phenotypes (Baerwald et al., 2016; Kooke et al., 2015; Schmid et al., 2018), pre-zygotic reproductive barriers (Smith et al., 2016) or post-zygotic reproductive barriers (Lafon-Placette & Köhler, 2015). However, correlating epigenetic differences to such traits is inherently limited, as it measures RI due to only one barrier to gene flow. To measure epigenetic RI, we must estimate the transmission (v) of epigenetic marks in natural populations. Heckwolf et al. (2020) examined differences in DNA methylation among salt- and freshwater-adapted populations of Gasterosteus aculeatus. They crossed individuals from the two habitats and exposed these mixed lineages to one of the divergent habitats for either one or two generations. Then, they showed that differentially methylated sites are transmissible (v > 0), as F2 crosses whose parents were exposed to the same divergent habitat, had DNA methylation profiles that were more like the natural population in that habitat, than those whose parents were not. Furthermore, epigenetic transmission can be estimated by directly comparing the epigenetic profile of parents and their offspring (Weyrich et al., 2016; Yagound et al., 2020); however, such observations do not specifically measure the transmission of locally adapted epigenetic marks. The primary caveat to such study designs is that the observed v is not precisely the same as the v in our model, as it is not independent of genetic variation. To more directly study the environmental induction of epigenetic marks, manipulative experiments can be performed with methyltransferase inhibitors (Biergans et al., 2012, 2015; Bossdorf et al., 2010; Herden et al., 2019; Wilschut et al., 2016), RNA interference (Bewick et al., 2019) and CRISPR-Cas9 modification of epigenetic marks (Kang et al., 2019; Vojta et al., 2016). While these manipulative experiments give much greater power to observe the effects of epigenetic change in the absence of genetic change, they are difficult to implement in the natural setting. We must investigate the genetic basis of epigenetic state, to understand the interplay between the two and its effect on RI. There are approaches to statistically control for genetic and geographic structure when testing for epigenetic differentiation among populations (Herrera et al., 2016; Lea et al., 2015, 2017); however, they typically apply a genome-wide correction for relatedness, which does not capture locus-specific interaction between genetic and epigenetic variation. To capture locus-specific covariation between genetic structure and epigenetic state, one can perform a genome-wide association study on epigenetic state, that is, treating epigenetic state as a molecular phenotype (Figure 4). Dubin et al. (2015) conducted such an analysis among accessions of Arabidopsis thaliana, allowing them to determine the extent to which cis- and trans-acting genetic loci influence geographic variation in DNA methylation. To this end, Dubin et al. (2015) found that variation in methylation at single cytosines is roughly equally influenced by cis- and trans-acting SNPs, whereas per cent gene body methylation is largely determined by trans-acting SNPs. Notably, while it is useful to study asexual species to control for genetic effects on epigenetic state (Verhoeven & Preite, 2014), we must understand the genetic basis of epigenetic variation in sexually reproducing species to study its influence on RI. Ideally, the observations which we have outlined could be made within a single-model system. One would have to observe epigenetic variation in natural populations, assess the transmissibility of locally adaptive epigenetic states and parse the influence of genetic sequence and environmental induction on such states. After this, one can finally measure neutral gene flow with respect to these divergent epigenetic loci, to derive an estimate of epigenetic RI as per Equation (1. While implementing such an analysis is challenging, the studies highlighted above show that we can begin to make progress towards this ideal. We have illustrated using a simple scenario, how epigenetics can influence neutral gene flow. However, it is becoming more apparent that epigenetic and genetic adaptation are inter-dependent processes, rather than one being an epiphenomenon over the other (Danchin et al., 2019; Gardiner et al., 2018; Klironomos et al., 2013; Pimpinelli & Piacentini, 2020). As such, as our models of RI grow in nuance to match the complexity of natural populations, the incorporation of mechanisms such as epigenetics will become essential. Due to the diversity of fidelity, genomic scales, geographic scales and time scales over which epigenetic marks operate, they have the potential to exhibit a vast array of influences on local adaptation and RI. Even if we find that epigenetics rarely affects RI in natural populations, it is worthwhile to understand why natural circumstances tend to eliminate epigenetic barriers to gene flow. By rooting epigenetic modifications within the framework of genetic variation in both empirical and theoretical work, we can bridge the gap between contemporary organismal processes and speciation. All authors contributed to the development of the paper's conceptual framework. NPP wrote the manuscript, and CFC, JF, ZG and PN edited the manuscript. We would like to thank Hélène Collin for discussion and inspiration to pursue the role of epigenetics in local adaptation. Open access funding enabled and organized by ProjektDEAL. The authors have no conflicts of interest to declare. The peer review history for this article is available at https://publons.com/publon/10.1111/jeb.14033. The authors have no raw data to report. R code used to generate Figure 3 is available at https://github.com/planidin/Westram_comment. Table S1 Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article." @default.
• W4294667079 created "2022-09-06" @default.
• W4294667079 creator A5030592051 @default.
• W4294667079 creator A5030626361 @default.
• W4294667079 creator A5059102793 @default.
• W4294667079 creator A5070831863 @default.
• W4294667079 creator A5073363693 @default.
• W4294667079 date "2022-09-01" @default.
• W4294667079 modified "2023-10-14" @default.
• W4294667079 title "Epigenetics and reproductive isolation: a commentary on Westram et al., 2022" @default.
• W4294667079 cites W1597169112 @default.
• W4294667079 cites W1900612108 @default.
• W4294667079 cites W1915859614 @default.
• W4294667079 cites W1955074387 @default.
• W4294667079 cites W1966700840 @default.
• W4294667079 cites W1971895280 @default.
• W4294667079 cites W1974450480 @default.
• W4294667079 cites W1985267478 @default.
• W4294667079 cites W1987902909 @default.
• W4294667079 cites W1994637391 @default.
• W4294667079 cites W2024848067 @default.
• W4294667079 cites W2029650666 @default.
• W4294667079 cites W2039153787 @default.
• W4294667079 cites W2055196978 @default.
• W4294667079 cites W2095779419 @default.
• W4294667079 cites W2104337366 @default.
• W4294667079 cites W2123430325 @default.
• W4294667079 cites W2133710762 @default.
• W4294667079 cites W2137303277 @default.
• W4294667079 cites W2149206894 @default.
• W4294667079 cites W2156454908 @default.
• W4294667079 cites W2164403987 @default.
• W4294667079 cites W2167022872 @default.
• W4294667079 cites W2169832073 @default.
• W4294667079 cites W2173177442 @default.
• W4294667079 cites W2205162919 @default.
• W4294667079 cites W2256261253 @default.
• W4294667079 cites W2259674058 @default.
• W4294667079 cites W2262655396 @default.
• W4294667079 cites W2294327742 @default.
• W4294667079 cites W2294638722 @default.
• W4294667079 cites W2301110592 @default.
• W4294667079 cites W2344695809 @default.
• W4294667079 cites W2346367686 @default.
• W4294667079 cites W2417187192 @default.
• W4294667079 cites W2604537249 @default.
• W4294667079 cites W2621593078 @default.
• W4294667079 cites W2893251498 @default.
• W4294667079 cites W2895376279 @default.
• W4294667079 cites W2896826128 @default.
• W4294667079 cites W2908439016 @default.
• W4294667079 cites W2950372281 @default.
• W4294667079 cites W2951479823 @default.
• W4294667079 cites W2953209411 @default.
• W4294667079 cites W2964539967 @default.
• W4294667079 cites W2969465898 @default.
• W4294667079 cites W2984854601 @default.
• W4294667079 cites W2990871165 @default.
• W4294667079 cites W2995941448 @default.
• W4294667079 cites W3012350583 @default.
• W4294667079 cites W3092093148 @default.
• W4294667079 cites W3108848792 @default.
• W4294667079 cites W3169128508 @default.
• W4294667079 cites W4211207648 @default.
• W4294667079 cites W4224995333 @default.
• W4294667079 cites W4294666709 @default.
• W4294667079 cites W4294667079 @default.
• W4294667079 doi "https://doi.org/10.1111/jeb.14033" @default.
• W4294667079 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/36063158" @default.
• W4294667079 hasPublicationYear "2022" @default.
• W4294667079 type Work @default.
• W4294667079 citedByCount "2" @default.
• W4294667079 countsByYear W42946670792022 @default.
• W4294667079 crossrefType "journal-article" @default.
• W4294667079 hasAuthorship W4294667079A5030592051 @default.
• W4294667079 hasAuthorship W4294667079A5030626361 @default.
• W4294667079 hasAuthorship W4294667079A5059102793 @default.
• W4294667079 hasAuthorship W4294667079A5070831863 @default.
• W4294667079 hasAuthorship W4294667079A5073363693 @default.
• W4294667079 hasBestOaLocation W42946670791 @default.
• W4294667079 hasConcept C104317684 @default.
• W4294667079 hasConcept C144024400 @default.
• W4294667079 hasConcept C149923435 @default.
• W4294667079 hasConcept C150194340 @default.
• W4294667079 hasConcept C179463898 @default.
• W4294667079 hasConcept C190727270 @default.
• W4294667079 hasConcept C26772143 @default.
• W4294667079 hasConcept C2775941552 @default.
• W4294667079 hasConcept C2908647359 @default.
• W4294667079 hasConcept C41091548 @default.
• W4294667079 hasConcept C54355233 @default.
• W4294667079 hasConcept C60644358 @default.
• W4294667079 hasConcept C78458016 @default.
• W4294667079 hasConcept C86803240 @default.
• W4294667079 hasConcept C90856448 @default.
• W4294667079 hasConceptScore W4294667079C104317684 @default.
• W4294667079 hasConceptScore W4294667079C144024400 @default.
• W4294667079 hasConceptScore W4294667079C149923435 @default.

• next