Polygenic Adaptation: Integrating Population Genetics and Gene Regulatory Networks

Abstract

Quantitative genetics models have highlighted the diversity of genetic architectures underlying polygenic traits. This diversity has an impact on how the traits respond to directional selection and on its molecular signatures on the genome.Genome-wide scans for selection have revealed examples of polygenic selection driving local adaptation of populations in several species. This polygenic selection disproportionately targets regulatory regions, hinting for an important role of gene regulatory networks in evolution.Gene regulatory network inference helps identifying and grouping together genes and regulatory elements that participate to the same biological processes. It also helps discovering how the structure of regulatory networks can put constraints on which genes and regulatory regions can be leveraged by polygenic selection.The introduction of gene regulatory network information in the omnigenic model highlights the pervasive pleiotropy in the genome. The general interconnection between all genes within the regulatory network might strongly limit the action of selection. The adaptation of populations to local environments often relies on the selection of optimal values for polygenic traits. Here, we first summarize the results obtained from different quantitative genetics and population genetics models, about the genetic architecture of polygenic traits and their response to directional selection. We then highlight the contribution of systems biology to the understanding of the molecular bases of polygenic traits and the evolution of gene regulatory networks involved in these traits. Finally, we discuss the need for a unifying framework merging the fields of population genetics, quantitative genetics and systems biology to better understand the molecular bases of polygenic traits adaptation. The adaptation of populations to local environments often relies on the selection of optimal values for polygenic traits. Here, we first summarize the results obtained from different quantitative genetics and population genetics models, about the genetic architecture of polygenic traits and their response to directional selection. We then highlight the contribution of systems biology to the understanding of the molecular bases of polygenic traits and the evolution of gene regulatory networks involved in these traits. Finally, we discuss the need for a unifying framework merging the fields of population genetics, quantitative genetics and systems biology to better understand the molecular bases of polygenic traits adaptation. Many adaptive traits (i.e., traits involved in adaptation to local environments) are polygenic traits (see Glossary). Also called quantitative or complex traits, they are determined by multiple genes and regulatory loci. Population genetics studies have helped improve our understanding of the effects of directional selection on polygenic traits [1.Latta R.G. Differentiation of allelic frequencies at quantitative trait loci affecting locally adaptive traits.Am. Nat. 1998; 151: 283-292Crossref PubMed Scopus (134) Google Scholar, 2.Kremer A. Le Corre V. Decoupling of differentiation between traits and their underlying genes in response to divergent selection.Heredity. 2012; 108: 375-385Crossref PubMed Scopus (61) Google Scholar, 3.Chevin L.-M. Hospital F. Selective sweep at a quantitative trait locus in the presence of background genetic variation.Genetics. 2008; 180: 1645-1660Crossref PubMed Scopus (116) Google Scholar, 4.Höllinger I. et al.Polygenic adaptation: from sweeps to subtle frequency shifts.PLoS Genet. 2019; 15e1008035Crossref PubMed Scopus (30) Google Scholar], which depends upon the trait’s underlying genetic architecture. Since Fisher’s seminal work on phenotypic traits [5.Fisher R.A. XV., The correlation between relatives on the supposition of Mendelian inheritance.Trans. R. Soc. Edinb. 1919; 52: 399-433Crossref Scopus (2271) Google Scholar], quantitative genetics researches have revealed a continuum of genetic architectures for traits in many organisms, including yeasts, insects, worms, plants, and mammals [6.Holland J.B. Genetic architecture of complex traits in plants.Genome Stud. Mol. Genetics. 2007; 10: 156-161Google Scholar, 7.Mackay T.F.C. et al.The genetics of quantitative traits: challenges and prospects.Nat. Rev. Genet. 2009; 10: 565-577Crossref PubMed Scopus (738) Google Scholar, 8.Timpson N.J. et al.Genetic architecture: the shape of the genetic contribution to human traits and disease.Nat. Rev. Genet. 2018; 19: 110-124Crossref PubMed Scopus (148) Google Scholar]. These architectures range from the Mendelian model, where a trait is determined by a single gene, to the infinitesimal model [9.Barton N.H. et al.The infinitesimal model: definition, derivation, and implications.Theor. Popul. Biol. 2017; 118: 50-73Crossref PubMed Scopus (89) Google Scholar], where a seemingly infinite number of loci determine a trait. In between, many traits present an oligogenic or polygenic determinism: flowering time in plants, ethanol tolerance in yeasts, or lipid traits in humans [10.Pais T.M. et al.Comparative polygenic analysis of maximal ethanol accumulation capacity and tolerance to high ethanol levels of cell proliferation in yeast.PLoS Genet. 2013; 9 (Publisher: Public Library of Science)e1003548Crossref PubMed Scopus (55) Google Scholar, 11.Peiffer J.A. et al.The genetic architecture of maize height.Genetics. 2014; 196: 1337-1356Crossref PubMed Scopus (178) Google Scholar, 12.Salome P.A. et al.Genetic architecture of flowering-time variation in Arabidopsis thaliana.Genetics. 2011; 188: 421Crossref PubMed Scopus (125) Google Scholar, 13.Shi H. et al.Contrasting the genetic architecture of 30 complex traits from summary association data.Am. J. Hum. Genet. 2016; 99: 139-153Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar]. Many of these traits are considered to be under directional selection [14.Daub J.T. et al.Evidence for polygenic adaptation to pathogens in the human genome.Mol. Biol. Evol. 2013; 30: 1544-1558Crossref PubMed Scopus (104) Google Scholar, 15.He F. et al.The footprint of polygenic adaptation on stress-responsive cis-regulatory divergence in the Arabidopsis Genus.Mol. Biol. Evol. 2016; 33: 2088-2101Crossref PubMed Scopus (27) Google Scholar, 16.Zan Y. Carlborg O. A polygenic genetic architecture of flowering time in the worldwide Arabidopsis thaliana population.Mol. Biol. Evol. 2019; 36: 141-154Crossref PubMed Scopus (12) Google Scholar]. Independently, systems biology studies initiated a switch in the perception of molecular bases of polygenic traits, from a gene-first to an interaction-first model [17.Kirschner M.W. The meaning of systems biology.Cell. 2005; 121: 503-504Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar]. Quantitative genetics represents the molecular bases of polygenic phenotypes as a collection of independent loci, each coding for a fraction of the phenotype, with potentially some interactions (1A). Conversely, systems biology focuses primarily on these interactions. In particular, gene regulatory networks (GRNs) represent the regulatory relationships between genes, their products, and their regulators (1C). Because they integrate and organize information from different levels (genomics, epigenomics, transcriptomics, proteomics, ...), they allow new information on the molecular bases of polygenic traits to be obtained [18.Sonawane A.R. et al.Network medicine in the age of biomedical big data.Front. Genet. 2019; 10Crossref PubMed Scopus (43) Google Scholar]. These methods can now be applied to most species, as long as the required data are available (assembled genomes, gene expression data, and/or proteomic data). A model describing the genetic architecture of polygenic traits has been recently proposed that integrates information from GRN and quantitative genetics: the omnigenic model [19.Boyle E.A. et al.An expanded view of complex traits: from polygenic to omnigenic.Cell. 2017; 169: 1177-1186Abstract Full Text Full Text PDF PubMed Scopus (989) Google Scholar,20.Liu X. et al.Trans effects on gene expression can drive omnigenic inheritance.Cell. 2019; 177: 1022-1034.e6Abstract Full Text Full Text PDF PubMed Scopus (100) Google Scholar] (Figure 1B ). It describes the genetic contribution to a trait as a combination of direct effects from core genes and indirect effects from peripheral genes acting in trans. This model shows how a modification in a trans gene can be amplified by the GRN, up to the point where trans genes explain a majority of trait variance. However, this model postulates that most of the genome participates in each trait, potentially severely limiting the action of directional selection on polygenic traits [19.Boyle E.A. et al.An expanded view of complex traits: from polygenic to omnigenic.Cell. 2017; 169: 1177-1186Abstract Full Text Full Text PDF PubMed Scopus (989) Google Scholar]. In the next sections, we will summarize the different models developed to explore the genetic and molecular bases of polygenic traits and their adaptation. We will discuss how they affect our perception of polygenic trait evolution and their advantages and limits. Finally, we will propose development of a new framework integrating their main aspects in to improve our understanding of polygenic trait adaptation and better its signatures (see for a to of first describes the important obtained from gene regulatory network of the regulatory elements or and the genes The structure is revealed by the in and the is with to The elements are in and the high elements are in The describes important information obtained from population and quantitative genetics including selection and polygenic The describes our proposed model integrating information from the first We for all elements in the regulatory that are for the trait, to a selection that the the and the polygenic The first describes the important obtained from gene regulatory network of the regulatory elements or and the genes The structure is revealed by the in and the is with to The elements are in and the high elements are in The describes important information obtained from population and quantitative genetics including selection and polygenic The describes our proposed model integrating information from the first We for all elements in the regulatory that are for the trait, to a selection that the the and the polygenic population genetics models developed to polygenic trait selection and its molecular signatures at the These models are the effects of each on the are independent of each and interactions can also be that several selection allow the phenotypic to be [1.Latta R.G. Differentiation of allelic frequencies at quantitative trait loci affecting locally adaptive traits.Am. Nat. 1998; 151: 283-292Crossref PubMed Scopus (134) Google Scholar], from where with effects in to the polygenic selection model, where each locus frequency allelic up loci. main can affect the selection model: the of by each the and the shape of the et al.Genetic architecture and sweeps polygenic adaptation to trait Genet. 2018; PubMed Scopus Google Scholar]. an trait by many loci that the of sweeps will on the number of loci involved in the each locus will respond strongly in of allelic frequency [1.Latta R.G. Differentiation of allelic frequencies at quantitative trait loci affecting locally adaptive traits.Am. Nat. 1998; 151: 283-292Crossref PubMed Scopus (134) Google Scholar]. these loci, some are loci with a high contribution to genetic they are to be by allelic frequency A. Le Corre V. Decoupling of differentiation between traits and their underlying genes in response to divergent selection.Heredity. 2012; 108: 375-385Crossref PubMed Scopus (61) Google Scholar]. This by a locus coding for a trait, all loci as background genetic L.-M. Hospital F. Selective sweep at a quantitative trait locus in the presence of background genetic variation.Genetics. 2008; 180: 1645-1660Crossref PubMed Scopus (116) Google Scholar]. Finally, a complex model including interactions that the of a sweep at a locus involved in a quantitative trait depends on the background (i.e., the of the of the population by the of all background I. et al.Polygenic adaptation: from sweeps to subtle frequency shifts.PLoS Genet. 2019; 15e1008035Crossref PubMed Scopus (30) Google Scholar]. sweeps at the high values frequency sweeps will be for that sweeps will for of adaptation to an of variation under selection and effects on a single 2019; PubMed Scopus Google Scholar]. on the results several methods developed for polygenic selection These methods at of selection loci affecting the same biological or trait. of data to gene regulatory interactions genes, and association studies The of these are in J. N.H. polygenic of 2018; PubMed Scopus Google Scholar, et polygenic polygenic and human phenotypic Public 2019; PubMed Scopus Google Scholar, et for polygenic adaptation of in 2019; PubMed Scopus Google Scholar, et al.Polygenic adaptation on is to in association 2019; PubMed Scopus Google Scholar] including the and with their We to point with the genetic architecture of these to to signatures of polygenic selection have on signatures of selection of a as for adaptive The first of or gene are in selection [14.Daub J.T. et al.Evidence for polygenic adaptation to pathogens in the human genome.Mol. Biol. Evol. 2013; 30: 1544-1558Crossref PubMed Scopus (104) Google Scholar], with many loci a high of genetic differentiation A recently developed for local under differentiation A. et gene under selection in biological 2017; PubMed Scopus Google Scholar]. The of a trait has been to divergent selection A population genetic of polygenic Genet. 2014; PubMed Scopus Google Scholar, F. et polygenic adaptation in 2018; PubMed Scopus Google Scholar, et genetic differentiation of and Genet. PubMed Scopus (100) Google Scholar]. These methods are on association studies that loci with the traits and their on these traits. This the polygenic of the in the populations to be this can then be with its under genetic Because they on these methods be this a information on the genome. to signatures of polygenic selection have on signatures of selection of a as for adaptive The first of or gene are in selection [14.Daub J.T. et al.Evidence for polygenic adaptation to pathogens in the human genome.Mol. Biol. Evol. 2013; 30: 1544-1558Crossref PubMed Scopus (104) Google Scholar], with many loci a high of genetic differentiation A recently developed for local under differentiation A. et gene under selection in biological 2017; PubMed Scopus Google Scholar]. The of a trait has been to divergent selection A population genetic of polygenic Genet. 2014; PubMed Scopus Google Scholar, F. et polygenic adaptation in 2018; PubMed Scopus Google Scholar, et genetic differentiation of and Genet. PubMed Scopus (100) Google Scholar]. These methods are on association studies that loci with the traits and their on these traits. This the polygenic of the in the populations to be this can then be with its under genetic Because they on these methods be this a information on the genome. in polygenic selection can be by the of of the genetic architecture of traits. In on of genes or regulatory a of polygenic trait architecture is these often genes, regulatory scans polygenic selection signatures in regulatory regions, an in selection cis-regulatory elements and expression quantitative trait loci et acting on cis-regulatory regions in humans from of and Genet. 2009; PubMed Scopus Google et al.Polygenic selection the evolution of within a 2019; PubMed Scopus Google Scholar]. this by the genome to and regulatory However, these several limits. a of the genetic bases of traits is the [19.Boyle E.A. et al.An expanded view of complex traits: from polygenic to omnigenic.Cell. 2017; 169: 1177-1186Abstract Full Text Full Text PDF PubMed Scopus (989) Google Scholar], limiting our understanding of their interactions are also often they a of polygenic trait their evolution et the of polygenic adaptation and the role of in 2018; PubMed Scopus Google Scholar]. interactions explain a of the they can selection in the of interactions or in the of to complex effects of on the response to selection is important for selection and adaptation: 2013; PubMed Scopus Google Scholar]. interactions between traits, in genes and regulatory loci, are often they can strongly H. et evolution by Evol. 2017; PubMed Scopus Google Scholar]. However, pleiotropy can also explain an trait might have been in the and response to pathogens in humans The contribution of selection to to and 2014; PubMed Scopus Google highlight the of the adaptive and the of in of and A. The of and their Rev. Genet. 2019; PubMed Scopus Google et and a regulatory network 2018; PubMed Scopus Google Scholar], to better understand the evolution of traits and signatures of polygenic adaptation et al.Polygenic adaptation: a unifying framework to understand Rev. Genet. PubMed Scopus Google Scholar]. the systems biology point of understanding the molecular bases of polygenic selection of how regulatory interactions under The accumulation of data now and on their The different of GRN evolution are for on gene regulatory network Genet. 2017; 33: Full Text Full Text PDF PubMed Scopus Google Scholar] and networks and the evolution of Biol. 2018; PubMed Scopus Google Scholar] and many of selection on are et al.Polygenic selection the evolution of within a 2019; PubMed Scopus Google et expression of polygenic adaptation in Biol. Evol. 2019; Scopus Google et of polygenic cis-regulatory Genet. 2011; PubMed Scopus Google Scholar]. In GRN evolution models developed et gene regulatory a 2019; Scopus Google Scholar], with models how each gene each and A. of gene networks by gene a model and its on genome PubMed Scopus Google A. PubMed Google Scholar]. In these models, affect the regulatory relationships and selection an optimal of expression for each The analysis of and their together with the of GRN models, to about how polygenic selection and how GRN structure might or GRN are important in this and their (Figure and a collection of that together to the of expression of genes, as and These regulatory relationships can be direct or networks from and gene expression to direct regulatory networks from data indirect regulatory the expression of are often as that represent the elements and the interactions between elements can of as or several of as regulatory and The of the of these better understanding of how the elements together and how in yeasts, insects, plants, or some they are This that or are to many most are to (Figure are also that they are of that are strongly to the of the a The often to biological a collection of that together to the of expression of genes, as and These regulatory relationships can be direct or networks from and gene expression to direct regulatory networks from data indirect regulatory the expression of are often as that represent the elements and the interactions between elements can of as or several of as regulatory and The of the of these better understanding of how the elements together and how in yeasts, insects, plants, or some they are This that or are to many most are to (Figure are also that they are of that are strongly to the of the a The often to biological GRN under selection. In a a where single to A. PubMed Google Scholar]. a for polygenic as the accumulation of which can then adaptation in environments A. as a general of complex gene PubMed Scopus Google Scholar]. results in from that a modification in a network or impact gene expression or its also polygenic adaptation: in an evolution in phenotypic adaptation in response to in adaptation from in different genetic on the et al.Genetic polygenic adaptation in Biol. 2019; PubMed Scopus Google Scholar]. polygenic selection by the number of available targets (Figure selection will be to different of in and integrating in polygenic for by all coding for the same to selection is GRN that the response of genes to selection. The omnigenic model that each trait might be by all genes [19.Boyle E.A. et al.An expanded view of complex traits: from polygenic to omnigenic.Cell. 2017; 169: 1177-1186Abstract Full Text Full Text PDF PubMed Scopus (989) Google Scholar]. pleiotropy severely polygenic a trait all However, this model for some traits, as human might be for lipid traits as and levels in humans H. et al.Contrasting the genetic architecture of 30 complex traits from summary association data.Am. J. Hum. Genet. 2016; 99: 139-153Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar] are by a number of loci. can the effects of the omnigenic model on polygenic adaptation. are that of of genes strongly with each to the of the a In each a genes are with genes from genes in a are from the of the network and can be leveraged by polygenic selection (Figure Finally, to be that most genes are to genes are to many (Figure as on of networks [18.Sonawane A.R. et al.Network medicine in the age of biomedical big data.Front. Genet. 2019; 10Crossref PubMed Scopus (43) Google R. networks in cell 2005; 118: PubMed Scopus Google Scholar]. In the of are the which the of a in the and the which its in its The of a gene is with its of genetic diversity et al.The between network and regulatory variation within a population of Biol. Evol. 2017; PubMed Scopus Google et selection has on genes with interactions within the human Biol. Evol. Scopus Google Scholar], as by and The at the under in expression are to be by polygenic adaptation their modification affect traits. Conversely, the majority of genes are at the of the involved in or in their affect (Figure These genes the high which are for a regulatory for the that high genes, which are in the are by selection et of selection on genes involved in regulatory a Evol. Biol. 2011; PubMed Scopus Google Scholar]. This explain the that genes et selection on genes in the human 2005; PubMed Scopus Google Scholar]. be targets for polygenic adaptation. GRN structure the for polygenic selection to on many genes and gene regulatory limiting the effects of selection on of systems biology models is their on the evolution of networks of genes with each important that be to a of how polygenic selection can impact most GRN models cis-regulatory regions, as and which can impact the expression of several genes at The role of these regions in the adaptation of complex traits has been and that up to of with polygenic traits in regions et the from association to J. Hum. Genet. 2013; Full Text Full Text PDF Scopus Google et the maize 2016; PubMed Scopus Google Scholar]. are for polygenic selection regulatory et of polygenic cis-regulatory Genet. 2011; PubMed Scopus Google et of selection in human 2014; PubMed Scopus Google Scholar], and in a role in evolution and of Biol. 2007; PubMed Scopus Google Scholar]. However, integrating the of cis-regulatory in polygenic adaptation models might be the role of regulatory on gene expression is to It depends upon many including at the regulatory locus and of in the cell et in and 2017; PubMed Scopus Google Scholar]. information might be in the which allow of the effects of regulatory regions on gene expression at the and of their on et of the effects of gene expression levels on 2016; Full Text Full Text PDF PubMed Scopus Google Scholar], in with the development of the regulatory of and their role J. et of effects on expression and Genet. 2018; PubMed Scopus Google Scholar]. data are available for some model and be available to Finally, some integrate data networks on and gene expression data, in to regulatory interactions also interactions between regulatory regions and genes [18.Sonawane A.R. et al.Network medicine in the age of biomedical big data.Front. Genet. 2019; 10Crossref PubMed Scopus (43) Google et in with 2017; PubMed Scopus (43) Google Scholar]. These allow development of new GRN evolution models integrating cis-regulatory In to the of models of polygenic trait a new framework This new framework the GRN architecture (Figure Here, we in to a unifying we which genes and to polygenic traits, and how they also the complex regulatory interactions that their expression (see and These to all loci involved in the genetic architecture of a trait, also to these loci can be by polygenic selection. We as in the omnigenic model X. et al.Trans effects on gene expression can drive omnigenic inheritance.Cell. 2019; 177: 1022-1034.e6Abstract Full Text Full Text PDF PubMed Scopus (100) Google Scholar] to the regulatory information by to the However, we cis-regulatory regions in the are as this information about the biological role of these regions to be et in with 2017; PubMed Scopus (43) Google Scholar]. In we that GRN are on and in regulatory Finally, we that the of the different regulatory in the GRN are in to for their of by polygenic for the selection of the same of a trait the selection of genetic in different we this at the of the which molecular signatures on the genome and how can we is a of GRN that how polygenic selection are often involved in several is the of pleiotropy of the regulatory in of the as and how their population genetics model the evolution of genetic their complex interactions in However, model the evolution of these regulatory relationships their genetic that model the molecular evolution of regulatory regions in the of improve our understanding of the molecular bases of local on polygenic traits genes involved in the traits, at the and levels [1.Latta R.G. Differentiation of allelic frequencies at quantitative trait loci affecting locally adaptive traits.Am. Nat. 1998; 151: 283-292Crossref PubMed Scopus (134) Google Scholar]. regulatory interactions affect this be to polygenic methods at polygenic selection the of each gene on the trait. the between these genes also be in these for the selection of the same of a trait the selection of genetic in different we this at the of the which molecular signatures on the genome and how can we is a of GRN that how polygenic selection are often involved in several is the of pleiotropy of the regulatory in of the as and how their population genetics model the evolution of genetic their complex interactions in However, model the evolution of these regulatory relationships their genetic that model the molecular evolution of regulatory regions in the of improve our understanding of the molecular bases of local on polygenic traits genes involved in the traits, at the and levels [1.Latta R.G. Differentiation of allelic frequencies at quantitative trait loci affecting locally adaptive traits.Am. Nat. 1998; 151: 283-292Crossref PubMed Scopus (134) Google Scholar]. regulatory interactions affect this be to polygenic methods at polygenic selection the of each gene on the trait. the between these genes also be in these are We the and for and This work by the ). of by a in a the of the in the its local number of to a a of selection that an of a trait all In the of a polygenic trait, can to the in frequency of at several independent loci. in gene regulatory represent relationships of or between between or loci to In the of loci, the of a at of the loci depends on the presence or of in the network that at the of or regulators the network in or A network has a high is a high between from the same and or between from different in gene regulatory represent elements of the genes or or of of a gene or locus that several seemingly traits. also called quantitative or complex trait, a polygenic trait is determined by several genes and/or regulatory loci. of polygenic traits are on the number of loci underlying at of the oligogenic traits are determined by a of genes and loci, at the all the genome to omnigenic traits. of gene regulatory networks that to the and the of the a of selection that an of a trait.

Topics & Concepts