Functional genomics offers new tests of speciation hypotheses

Abstract

Speciation is a fundamental process shaping biodiversity. However, existing empirical methods often cannot provide key genetic and functional details required to validate speciation theory. New gene modification technologies can verify the causal functionality of genes with astonishing accuracy, helping resolve questions about how reproductive isolation evolves during speciation.

Thinking beyond omics in speciation

Understanding the process of speciation is one of the main goals of evolutionary biology. Theories concerning speciation are among the best developed in biology [e.g. ^1–3], and throughout the -omics (see Glossary) era expectations have been high for validating theories about how reproductive isolation (RI) evolves to create new species. However, while an abundance of empirical work has raised our understanding of various aspects of speciation [2,3], there is still a paucity of experimental studies that explicitly distinguish alternative hypotheses. This is because many studies lack the fine genetic detail, functional link to RI, and direct manipulation experiments to substantiate theory. Motivated by the dawn of a technological revolution in functional genomics (Box1), we outline how testing of classical hypotheses in speciation could benefit from advances in gene modification, beyond current candidate gene validation studies, to help spur new research and discovery in the field.

Glossary.

Chromosomal inversion

a structural rearrangement of DNA sequence where the inverted sequence is reversed relative to the collinear sequence.

Coupling

collective effects of different traits or factors involved in reproductive isolation, which strengthen the barrier to gene flow between the diverging populations.

Dobzhansky-Muller incompatibilities

negative epistatic interactions between different genes, often in hybrids.

Ecological speciation

evolution of RI between populations as a result of ecologically-based divergent natural selection.

Epistasis

a phenomenon in which the effect of an allele at one locus is dependent on an allele (or alleles) at one or more other loci (i.e., a between-locus interaction).

Genetic linkage

a non-random association of alleles at different loci. Also, a term used by classical geneticists to refer to genes that reside on the same chromosome.

Genome scans

method of comparison between populations/species across the genome to identify differentiated genetic regions across the genome.

Genome-wide association studies (GWAS)

studies associating genotypic and phenotypic variation, generally using segregating natural genetic variation.

Heterosis

hybrid vigor, or a phenomenon of enhanced function of a trait in hybrids.

Indel

insertion or deletion of nucleotides in the genome sequence

Knockdown

artificial reduction of gene expression by blocking its transcription or breaking down mRNA.

Knock-in

artificial insertion of the nucleotide sequence into a genome.

Knock-out

artificial permanent deactivation of the gene with loss of its functionality.

Magic traits

a trait subject to divergent ecological adaptation which has a pleiotropic effect causing premating isolation.

Mutation order speciation

accumulation of different incompatible mutations in separate populations subject to the same selective regime.

-omics

fields of biology that names end in -omics and that aim to collect large data sets of biological molecules that translate into the structure, function, and dynamics of an organism. Examples include genomics, transcriptomics, proteomics or metabolomics.

One-allele mating mechanism

a scenario when RI forms due to the same allele spreading in both diverging populations (e.g., a single allele that induces assortative mating with a self-referenced phenotype).

Pleiotropy

a phenomenon when a gene affects more than one phenotype.

Quantitative trait loci (QTL) mapping

a statistical method of associating genetic and phenotypic variation via establishing sets of recombinants via genetic crosses.

Reproductive isolation (RI)

genetically-based barriers to gene flow between populations/species.

Snowball theory

a speciation theory predicting that number of genetic incompatibilities that reduce hybrid fitness grows at a faster than linear rate over time (i.e., snowballs).

Box 1. Functional genetics manipulation methods.

RNA interference

RNA interference (RNAi) is an antiviral eukaryotic cell pathway that after recognizing dsRNA in the cytoplasm, targets and digests the corresponding mRNA strand, therefore temporarily knocking down the gene expression [6]. It is used as a molecular method to alter gene expression by injecting RNA molecules into organisms to neutralize complementary targeted mRNA molecules. RNAi silencing machinery is present in many, though not all, eukaryotic organisms. Major advantages of using RNAi in evolutionary biology applications are: 1) the possibility of studying the functions of essential genes when knock-out causes lethality, and 2) application to study species which are difficult to work with at the embryonic (egg) stage, a prerequisite for some alternative methods (including those discussed below).

CRISPR/Cas9 gene editing

CRISPR/Cas9 is a genome engineering tool adopted from the bacterial immune system which consists of a guide RNA and Cas9 protein (recent ongoing developments utilize various Cas proteins), capable of cleaving double-strand breaks (DSB) at the specified target sites. The produced genetic changes are heritable if they occur in the germline. CRISPR/Cas9 editing toolbox has been used for gene (or nucleotide) insertion and deletion, structural changes, replacement of the genes, precise nucleotide editing and gene regulation (Figure I)[12].

Figure I. Gene editing toolbox.

Pleiotropy and tight linkage in speciation studies

One of the main goals of speciation research is to ascertain the genetic basis of RI. Modern approaches utilize methods such as genome scans, QTL mapping, or GWAS to identify the regions of the genome with high population divergence or association with traits causing RI. However, depending on the frequency of recombination and density of markers, such genomic regions can still contain multiple, sometimes hundreds, of genes and regulatory elements. Such low resolution means that distinguishing effects of pleiotropy from those of tight genetic linkage of several genes is difficult [4]. Pleiotropy is relevant for speciation as it explains how multiple traits needed for RI can resist dissociation via recombination. In contrast, models of speciation via genetic linkage require genes to be first positioned correctly in a genome to coordinate their effects and recombination can still dissociate traits controlled by different genes. Accordingly, although pleiotropy and linkage might act similarly over short time scales, both mutation and recombination affect them differently over longer time periods [5].

Distinguishing pleiotropy from linkage is especially relevant in the “magic traits” concept [4,5], where RI is generated as an inadvertent by-product of divergent selection, and is thought to increase the chance of speciation-with-gene-flow [4]. A number of studies suggest putative examples, but very few achieve the genetic resolution to exclude alternative linkage [5].

Another classic hypothesis where a single gene enables speciation-with-gene-flow is the one-allele mating mechanism. In this model, RI is achieved in diverging populations via the spread of a single allele (e.g., an allele inducing preference for a self-referenced phenotype or to “stay where you were born”) [4]. Investigating this concept requires finer mapping of putative RI loci than has been conducted to date.

New gene-editing methods provide means to test the aforementioned hypotheses. For example, manipulating candidate loci with knockdown via a temporary suppression of its expression (RNAi or CRISPRa; Box1) or permanent gene knockout with the CRISPR/Cas-system has been shown to work successfully in a variety of organisms [6](Box1). This could be applied to determine if a locus affects one or more traits associated with RI or if different loci have discrete functionality (Fig.1A).

Figure 1.

Experimental design and predictions for speciation hypotheses tests with use of functional genetics methods to A) distinguish between pleiotropy and linked loci, and B) fine map the causal loci across inversion. Asterisks (*) indicate loci associated with trait of interest.

Inversions

Chromosomal inversions are common and can be involved in local adaptation and speciation because they strongly suppress recombination among blocks of linked genes [7]. For populations diverging-with-gene-flow, recombination suppression presents a powerful mechanism to allow selected and RI genes to be inherited together. Indeed, patterns of heterosis and Dobzhansky-Muller incompatibilities of genes linked by inversions could have important consequences on the likelihood of speciation [7]. In addition, when inversions form, they generate new mutations at their breakpoints that could serve both as a source of genetic variation and a time stamp of the inversion’s age. Thus, comparing the age of causal variants to breakpoint mutations could reveal if divergent alleles pre-date the inversion or accumulated after its formation.

Traditional methods, while useful for identifying inversions, are less helpful for resolving their causal genetic relationship to RI. Specifically, inversions suppress the recombination in hybrids that is needed to dissect how the multiple genes they can contain contribute and interact to generate RI (Fig.1B). As inversions often contain dozens to hundreds of genes, studies of individual loci within inversions are not feasible. Excitingly, CRISPR/Cas9 has been used in both animal and plant models to experimentally create targeted and large inversions [e.g. ⁸]. In principle, such precise inversion engineering could also be used to reverse an existing inversion. Genetic crossing using this genetically produced “reverted” locus could then be used to create recombinants with a non-inverted variant, thus allowing fine-mapping of traits and their causal loci across the inversion (Fig.1B). Once the causal variants are identified, their age and their relationships within the inversion could be determined.

Genes in a genomic context

Above, we discussed how genetic modification methods can used to explore the role of pleiotropy and linkage in RI evolution, however, it could also be used to investigate how genes function in new genomic contexts (i.e., different genetic backgrounds). For instance, genetic modification could be used to test how hybrid dysfunction evolves during speciation, this often involves negative epistatic interactions between genes. Quantifying the strength of epistasis could help validate classical theoretical concepts that remain poorly empirically supported, such as the snowball theory of hybrid incompatibility [1], and lead to interesting new routes of research via systems biology approaches [9].

Previously, hybrid dysfunction has proved to be difficult to study because of the numerous gene-gene interactions possible and the multilevel nature of genes-to-phenotype expression. However, making controlled genetic knock-ins (Box1) to swap divergent genes between populations (or knock-out, in case of gene duplication) could be a first step towards exploring how epistasis in diverging loci functions and evolves. For example, gene knock-ins could be used to quantify exactly how much epistasis results from a specific number of controlled allele swaps. This could provide important extensions to existing results that assume only pairwise gene interactions, likely an underestimation of reality.

Knock-ins of candidate RI genes into otherwise identical genomes could also be a way to precisely explore the role of selection in speciation. Such “common genome experiments” could help, for example, distinguish ecological from mutation order speciation [2]. In addition, a critical question concerns how different selected genes may become “coupled” together to accentuate their consequences for RI, accelerating rates of divergence [1,10]. Knock-ins and knock-outs of candidate RI genes for taxa at various stages of divergence or of different degrees of hybrid ancestry could aid in studying the coupling process and its role in speciation. Notably manipulation of genome structure, but not gene content, has been shown to cause RI [11], demonstrating the need to also consider the spatial genetic contexts behind speciation.

Engineering holistic approaches

Advancements in science are often achieved by matching new technology with novel or classic ideas. In this context, now may be a moment when speciation research can be moved forward by cutting-edge molecular methods. However, one should keep in mind that no gene exists in isolation, and while describing the functionality of individual genes is essential, it is important to address how genes evolve in their genomic contexts. Therefore, we expect the future of speciation research to be increasingly holistic and incorporate systems biology approaches, where functional gene manipulation methods could be an important key to unlocking new discoveries and answering seminal questions about the origins of biodiversity.