Integrative and Comparative Biology Advance Access originally published online on June 22, 2007
Integrative and Comparative Biology 2007 47(5):759-769; doi:10.1093/icb/icm050
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Key transitions during the evolution of animal phototransduction: novelty, "tree-thinking," co-option, and co-duplication
Ecology, Evolution and Marine Biology. The University of California-Santa Barbara, CA 93106
Correspondence: 1E-mail: plachetzki{at}lifesci.ucsb.edu
 |
Synopsis |
|---|
Top
 Synopsis IntroductionTree-thinking and the evolution...SummaryAcknowledgmentsReferences |
|---|
Biologists are amazed by the intricacy and complexity of biological interactions between molecules, cells, organisms, and ecosystems. Yet underlying all this biodiversity is a universal common ancestry. How does evolution proceed from common starting points to generate the riotous biodiversity we see today? This "novelty problem"—understanding how novelty and common ancestry relate—has become of critical importance, especially since the realization that genes and developmental processes are often conserved across vast phylogenetic distances. In particular, two processes have emerged as the primary generators of diversity in organismal form: duplication plus divergence and co-option. In this article, we first illustrate how phylogenetic methodology and "tree-thinking" can be used to distinguish duplication plus divergence from co-option. Second, we review two case studies in photoreceptor evolution—one suggesting a role for duplication plus divergence, the other exemplifying how co-option can shape evolutionary change. Finally, we discuss how our tree-thinking approach differs from other treatments of the origin of novelty that utilized a "linear-thinking" approach in which evolution is viewed as a linear and gradual progression, often from simple to complex phenotype, driven by natural selection.
|
Introduction |
|---|
|
Top
SynopsisIntroductionTree-thinking and the evolution...SummaryAcknowledgmentsReferences |
|---|
"Novelties come from previously unseen association of old material. To create is to recombine."—Jacob (1977
)
"Gene duplication emerged as the major force of evolution".
—Ohno (1970
Two different, but interdependent, processes likely contribute to the majority of novel evolutionary changes and each will leave different signatures that are distinguishable in comparative analyses. Some novel traits have their mutational origins in duplication, subsequently followed by differential divergence. Gene duplications are perhaps the best-studied duplication events, but "duplication" (either by copying or by fission) happens at other levels of biological organization, including protein domains, groups of interacting proteins, chromosomes, genomes, cells, organs, castes, species, and ecosystems. Other novel traits originate as new combinations of existing traits, a process sometimes termed bricolage, or tinkering (Jacob 1977). As with duplication, recombinational novelties occur at multiple levels of biological organization: domains fuse to form new proteins, proteins gain new interactions with other proteins, cells/species merge as in the endosymbiotic origin of eukaryotes, and new ecological interactions emerge by dispersal of species. Recognizing and distinguishing duplication and co-option requires a historical perspective.
This historical or "tree-thinking" perspective (Ohara 1997), which focuses on duplication and co-option, has not always been the dominant mode of thinking about the origin of novelties. Many evolutionists have instead suggested linear histories for the origin of complex traits, whereby the evolution of traits is viewed as a linear and gradual progression, often from simple to complex phenotype, and driven by natural selection. For example, Salvini-Plawen and Mayr (1977) constructed linear histories of eye evolution, which they termed "morphological sequences of differentiation" by collecting examples of eyes of differing complexity from closely related species (Fig. 1A). Additionally, Nilsson and Pelger (1994) constructed a linear conceptual model for the gradual origin of lens eyes (Fig. 1B). These linear histories are important for understanding how, and how quickly, natural selection might incrementally mold complexity; yet at the same time linear thinking discourages questions about the origins of novelty. Under linear models, all that is needed to proceed in a gradual and linear fashion from simple to complex traits is nondescript heritable variation and selection. As such, heritable variation—which is abundant in nature—is taken for granted and the focus is placed almost exclusively on selection. When considering linear, gradual models, little attention is therefore paid to how variations originate or to whether certain types of variation (e.g., duplication or co-option) are more common fuel for evolution. In contrast, the possibility of identifying the different types of variation that have been involved in the origins of novelties is precisely what a historical, tree-thinking perspective can add.
|
Duplication/divergence and co-option in the origin of novelties
Certainly, many novelties originate by duplication and subsequent differential divergence. In fact, duplication and divergence has become widely accepted as the primary means of origin of new genes (Zhang et al. 2003
; Taylor and Raes 2004). The mutational mechanisms causing duplication are well understood, and include retrotransposition, segmental, and genome duplications (Hurles et al. 2004). The original mutational events, however, are only part of the story, the other part being the subsequent differential divergence of the newly arisen gene copies. Divergence may involve natural selection on one or both copies, or neutral processes that fix complementary loss of gene functions (Zhang et al. 2003). Besides genes, duplication may occur also at other levels of biological organization. For example, co-expressed gene networks can be duplicated (Conant and Wolfe 2006). In this case, the mechanism of duplication may often be whole genome duplications, unless all genes in the network are copied independently or co-expressed genes are genomically adjacent, thereby allowing segmental duplication of the network. In yeast, differential divergence of duplicated co-expression networks has occurred by differential loss of network interactions (Conant and Wolfe 2006). Duplication also occurs in tissues and organs. For example, serial homologs like vertebrae, teeth, segments, limbs, and eyes may be considered duplicates (Arthur et al. 1999; Minelli 2000). The mechanistic origins of duplicated organs are not as well understood as are duplications that occur at the genetic level. One likely mechanism is fission of a formerly homogeneous field of cells into multiple fields, a developmental mechanism involved in the duplication of arthropod eyes (Friedrich 2006).
In addition to duplication plus divergence, evolutionary novelties often originate as new combinations of existing genes, structures, or species. For example, many new genes originate as novel combinations of existing protein domains. As one of many examples, the nervous-system-specific transcription-factor gene Pax-6 originated as a fusion of a Paired domain and a Homeodomain (Catmull et al. 1998). The mechanism of domain fusion likely involves the duplication of those domains, thus illustrating one way that duplication and co-option are interrelated. New combinations of existing elements also occur at other levels of biological organization. Gene-interaction networks gain new interactions with existing genes (Olson 2006). In addition, whole gene networks are often co-opted to function in new contexts (Gompel et al. 2005). Mechanistically, this often occurs by changes in gene regulation. In laboratory experiments, changes in regulation of transcription factor genes that lead to new sites of expression can cause ectopic production of organs like eyes (Halder et al. 1995). Ectopic expression of structures during evolution has been termed heterotopy (Haeckel 1866; West-Eberhard 2003). Interestingly, a co-option event at the level of the developmental regulatory network that leads to ectopic organ expression can be considered a mechanism for the duplication of organs, illustrating another relationship between co-option and duplication. Namely, co-option at one level may be viewed as duplication at another level of biological organization. Finally, co-option is a process that may act at the species level. Fusions of existing species may have led to novel species, for example by endosymbiosis (Thompson 1987; Margulis and Sagan 2002; Kutschera and Niklas 2005).
Although analogous processes of duplication and co-option generalize across numerous levels of biological organization, we focus our attention here on protein-interaction networks. In protein networks—like other biological levels—duplication–divergence and co-option lead to different patterns that can be detected by phylogenetic analyses (Abouheif 1999; Geeta 2003; Serb and Oakley 2005). Despite the focus here on networks, the general approach can be easily translated to other levels of biological organization. Before examining the evolutionary origins of specific phototransduction networks below, we will outline the phylogenetic patterns in protein networks expected to result from origins of novelty by duplication and co-option.
Phylogenetic patterns following duplication and co-option
If two protein networks originated by duplication of an ancestral network, then multiple components of the descendent networks will show patterns of co-duplication. In other words, simultaneous duplication of multiple genes of an ancestral network is a pattern consistent with the origin of the descendent networks by duplication (Fig. 2A). In contrast, the components of a novel protein network that evolved by co-option will have originated by gene duplications that occurred at different times (Fig. 2B). Co-duplication and co-option need not be considered discrete alternatives. Instead, some genes of a network may co-duplicate, whereas others may be co-opted, an intermediate situation that will often be true, especially when examining the evolution of particular networks at multiple time scales, and when examining networks at increasing spatial scales by increasing the number of interactions considered.
|
|
Differentiating co-duplication from co-option requires determining the relative timing of duplications for the genes involved in the network. Co-duplication may be viewed as a null hypothesis, which can be rejected by different hypothesis-testing procedures, in favor of the alternative hypothesis of co-option. A primary means of testing gene co-duplication is by reconciled tree analysis (RTA), the comparison of gene phylogenies to a species phylogeny (Goodman et al. 1979
; Page and Charleston 1998). RTA determines a phylogenetic interval for gene duplication events, with upper bounds at one speciation event and lower bounds at another speciation event. For example, if two sister species possess duplicate genes and an outgroup species possesses only one such gene, we might infer that a gene duplication occurred prior to the origin of the sister species, but after their divergence with the outgroup. Algorithmic approaches are available for more complicated situations (Page and Charleston 1998; Page 2002). If duplications giving rise to the origins of genes in a protein network map to different nodes on a species phylogeny, the null hypothesis of co-duplication is rejected in favor of network assembly by recombination of protein interactions (co-option). Another means of testing a co-duplication hypothesis is by explicitly estimating divergence times of the nodes leading to the network-genes, for example by relaxed molecular clock analysis (Britton et al. 2002; Sanderson 2002; Thorne and Kishino 2002). Such an analysis may offer an additional test after RTA, especially in situations where RTA maps co-duplication events to a large phylogenetic interval due to lack of sampling or extinction of species. Conceptually analogous statistical tests for co-speciation in host and parasite phylogenies also have been described (Huelsenbeck et al. 2000).
Once the evolutionary patterns of co-duplication and co-option have been established for a protein network, additional questions can be addressed. To what extent do co-duplicated genes respond to natural selection as a unit of interacting proteins rather than as individual genes? Are rates of evolution in co-duplicated genes concordant? Does one descendent network evolve at a different rate than the other? Are duplicated networks subfunctionalized (Conant and Wolfe 2006)? Can we determine the cis-regulatory changes responsible for co-option events (Gompel et al. 2005)? Do rates of evolution change after co-option? Does selection act to allow co-opted genes to specialize to their new roles? Next, we propose that metazoan phototransduction will provide a valuable model for investigating co-duplication, co-option, and the evolutionary outcomes of these novelty-generating mechanisms.
|
Tree-thinking and the evolution of animal phototransduction cascades |
|---|
|
Top
SynopsisIntroductionTree-thinking and the evolution...SummaryAcknowledgmentsReferences |
|---|
Multiple different signaling pathways enable much of life to respond to, and interact with, the light environment (Spudich et al. 2000
). The dominant signaling pathways found in animals are phototransduction cascades initiated by opsin, a member of the G-protein-coupled-receptor (GPCR) class of proteins. Variations of opsin pathways mediate a wide range of acuity for sensing light, from simple photosensitivity to vision (Land and Nilsson 2002). In this section, we explore the evolution and diversification of the animal phototransduction cascade with the purpose of highlighting instances where duplication and co-option may have contributed to key transitions in the evolution of these pathways. Namely, we will focus on two examples. First, we review evidence that the different phototransduction pathways of vertebrate rods and cones originated primarily by duplication. Second, we suggest that the evolutionary origins of two primary transduction pathways of animals (ciliary and rhabdomeric) involved extensive co-option.
Duplication and the origin of vertebrate rod and cone phototransduction cascades
The dual presence of rods and cones allow vertebrate retinas to overcome an inherent tradeoff between acuity and sensitivity: whereas rods are specialized for sensitivity in dim light (scotopic vision), cones are specialized for acuity in bright light (photopic vision). This duplicity theory (Schultze 1866) was first supported by the observation of two morphological classes of vertebrate photoreceptor cells and is now widely accepted and is supported by additional physiological and molecular data. How did the novelty of dual cell types originate in the evolution of the vertebrate retina? In this section, we will first discuss a hypothesis framed in linear thinking. Next, we will take a tree-thinking perspective and argue that rod and cone phototransduction pathways originated largely by duplication. Establishing duplication of pathways requires two things. First, the elements of the pathways must differ. Second, the origin of the elements of the pathway must be coincident in time. Based on available data, both hold true for rod and cone transduction pathways.
One perspective on the origin of rods and cones is that a subset of one cell type gradually transformed into another cell type. In a timeless classic, The Vertebrate Eye and its Adaptive Radiation, Gordon Walls (1942) invoked one such linear model of transformation from primitive to derived stating "They [rods] were derived quite simply from cones by the enlargement of the outer segment and by an increase in the number of visual cells connected to each nerve cell". That vertebrate cones are more ancient than rods is in fact supported by current knowledge that rod opsins are derived within paraphyletic cone opsin clades (Okano et al. 1992). However, Walls’ astute ascertainment of ancestry, entrenched in the linear mode of thinking, tells us little about the types of evolutionary processes or the types of variation that could have given rise to the novelty of vertebrates’ rod cells. In fact, if rods and cones evolved by some means other than by gradual transformation, the linear framework would obscure this phenomenon.
Another perspective is that rods and cones represent evolutionary duplicates or paralogs: just as genes duplicate within lineages, so too may cell types duplicate. Such hypotheses of duplication are supported by the similarity of components. The similarity of amino-acid components of proteins supports duplication of the proteins. In an analogous way, the similarity of expressed protein components of cell types may support duplication of the cell types (Arendt 2003). Although the entire repertoire of expressed proteins in rods and cones is not yet known, we can begin to address a duplication hypothesis for the cell types by investigating their phototransduction pathways. Rods and cones use similar, but different, phototransduction pathways, which may have originated by duplication.
More specifically, rods express one set of paralogous cell-type-specific phototransduction proteins and cone cells express another (Hisatomi and Tokunaga 2002; Nordström et al. 2004). Rod-specific and cone-specific paralogs include those of opsin, the G protein 
-subunit (G), phosphodiesterase (PDE), cyclic-nucleotide-gated ion channels (CNGs), and arrestin (Hisatomi and Tokunaga 2002; Nordström et al. 2004). From a tree-thinking perspective, we can ask if these data indicate duplication or co-option in the origin of rod and cone phototransduction. If the rod phototransduction pathway is a duplicate of the cone pathway, then most components of those pathways should be co-duplicated. In contrast, if one pathway originated largely by co-option, then the components should have originated by duplication at different times. The co-duplication of phototransduction pathway components can be treated as a null hypothesis that could be rejected in favor of co-option.
For the phototransduction pathways of rods and cones, the null hypothesis of co-duplication currently cannot be rejected. Paralogs for each rod-specific and cone-specific gene originated through large-scale segmental duplication of the genome (Nordström et al. 2004). RTA indicates that each of these duplications predates the origin of gnathostomes (jawed vertebrates) and postdates the split of cephalochordates/urochordates and vertebrates (Nordstrom et al. 2004; Fig. 4A). As such, the duplication of multiple genes in these pathways maps to the same phylogenetic interval and is consistent with co-duplication. This phylogenetic interval, however, is large and encompasses branches leading to two extant groups—agnathans [assumed monophyletic (Takezaki et al. 2003)] and chondrichthyans—that could be used to further refine the timing of the gene duplications. If agnathans and chondrichthyans both possess rod and cone paralogs of multiple phototransduction genes, then the phylogenetic interval—thus inferred to be prevertebrate—for these duplications would be reduced, providing a more stringent test of the co-duplication null hypothesis.
|
Unfortunately, data for phototransduction genes are largely absent from early branching vertebrates. Nevertheless, there is some evidence that differentiated rod and cone phototransduction cascades were present at the origin of vertebrates. Namely, multiple opsin genes are known from agnathans. The most in-depth phylogenetic treatment of the problem of the evolution of basal vertebrate opsin concluded that rod-class and cone-class opsins are present in agnathans (jawless vertebrates), and that rod-opsins branch from within paraphyletic cone-opsin clades (Collin 2006
; Pisani et al. 2006). From this we can conclude that the duplication of the gene that gave rise to the rod-specific opsin class took place prior to the last common ancestor of living vertebrates. One is tempted to argue that this finding alone necessitates that the origin of rod and cone pathways predate living vertebrates. However we still lack data on the rod- and cone-class paralogy assignments for the other phototransduction proteins in agnathans. It remains possible that rod opsin duplicated earlier than the rest of the rod-specific phototransduction genes. For example, in early branching vertebrates, rod opsins and cone opsins could interact with the same genes in the rest of the pathway, which duplicated later. In addition, there is the potential for rod- and cone-specific genes to have duplicated incrementally, not exclusively by genome duplication, within the same phylogenetic interval of rod and cone cell type origins, which were later entrained to their respective cell type-specific pathways. We do not advocate either of these alternative hypotheses for the evolution of vertebrate rods and cones; rather we favor the co-duplication hypothesis. Nevertheless, we recognize that co-duplication can be subjected to additional tests.
The realization that rod and cone phototransduction pathways represent evolutionary duplicates suggests two areas for future work. First, the mode of differentiation of the duplicate pathways may be examined. Was natural selection involved in the differentiation? To what extent does selection act on the entire biochemical pathway as compared to the individual genes? At least one interesting study has been conducted that bears on these questions. Carleton et al. (2005) examined opsin genes and found that there are no amino acids that uniquely define the rod-opsin clade. This result indicates that the origin of the rod pathway probably was not accompanied by a change in the biochemical function of opsin. As such, if selection acted to differentiate all genes in the rod pathway, including opsin, it must have involved changes in rod-opsin expression and not its biochemical interaction with other proteins. Another possibility is that selection acted on other components of the pathway, but was not involved in the maintenance of the duplicated rod opsin.
A second area for future study is to extend our interpretation to the next level of biological organization above the pathway, to the entire cell type. How have the dual processes of co-duplication plus divergence and co-option contributed to the evolution of rod and cone cells? In order to answer this question, we must be able to link the cell's morphological phenotype with its physiological phenotype in an evolutionary explanation (Fig. 4A). In the case of rods and cones, we can begin to make some preliminary connections. It has been shown that the transcription factor Nr2e3 serves as a switch for the terminal differentiation and maintenance of rod cells in the vertebrate retina (Chen et al. 2005; Peng et al. 2005). Comparisons between wild-type and Nr2e3 mutants have revealed that this transcription factor up-regulates rod-specific phototransduction genes in rod-transcriptional contexts and at the same time down-regulates cone-specific genes in cone-specific contexts. In the course of development, vertebrate rod and cone cells—together with other retinal cell types—are derived from a pool of multi-potent progenitor cells (Turner and Cepko 1987; Wetts and Fraser 1988). At some point prior to the evolution of rods and cones, Nr2e3 must have been co-opted to serve a role in either the inhibition or activation of each phototransduction cascade.
Co-option and the major animal phototransduction networks
Animals possess two major classes of photoreceptor cells, ciliary and rhabdomeric. Like rods and cones (which are both ciliary photoreceptor cells), the two major animal classes were first distinguished on the basis of physiological and morphological differences (Eakin 1963). The presence of these major classes has been further supported by molecular data (Arendt 2003). In this section, we first briefly discuss a linear perspective on the origins of the major photoreceptor cell types. We next take a tree-thinking perspective on the origins of ciliary and rhabdomeric cells, and address the null hypothesis of duplication. As discussed subsequently, unlike the origins of rods and cones, co-duplication can be clearly rejected in favor of co-option in the origins of ciliary and rhabdomeric cells.
One perspective on the origins of ciliary and rhabdomeric cells is a gradual linear model. Salvini-Plawen and Mayr (1977) in a landmark paper, include a steady infusion of linear explanations throughout their analysis. Based on morphological data available at the time (see references therein and also Eakin 1979 for a response), these authors provided a scheme of linear modifications that are possible in ciliary and rhabdomeric photoreceptor cells in the course of evolution. Their scheme for ciliary photoreceptors portrays 14 different morphologies from disparate metazoan phyla, each derived by transition from a common prototypical ancestor. Their explanation for diversity among ciliary photoreceptors is displayed as a carousel of varying photoreceptor morphologies revolving around a central ancestral state, each separated by a single linear transition (Fig. 9, Salvini-Plawen and Mayr 1977).
Tree-thinking provides another perspective on animal photoreceptor origins. Instead of treating photoreceptor cell type as a single morphological unit and providing descriptive accounts of possible modes of transitions between types, we can examine separately the individual components of the cells, and quantify the timing of their origins. Although the entire suite of expressed proteins in photocells is unknown, especially in nonmodel organisms, we can investigate the phototransduction pathways of different cell types as we have done above for rods and cones. A null hypothesis of duplication for the origin of ciliary and rhabdomeric cells has been described previously (Arendt 2003; Plachetzki et al. 2005). As we will next describe, ciliary and rhabdomeric cell origins reject this null hypothesis of duplication in favor of co-option, providing a counter-example to duplicated rods and cones.
Many similarities exist in the signaling networks of phototransduction in all animals studied to date. In bilaterian animals, three alternative phototransduction cascades are known—two of these cascades correspond to the ciliary and rhabdomeric morphological cell classes (Arendt 2003). All three of these general pathways are based on canonical GPCR signaling, yet they differ in important ways, including the subclass of the opsin protein that initiates each cascade. In one class, photoreceptors with ciliary morphology (such as rods and cones in the example above) generally utilize ciliary opsins (which interact with Gt and PDE) in cell signaling. Ciliary phototransduction leads to the closing of CNGs, the reduction of cation concentration and the hyperpolarization of the cell. Second, the rhabdomeric phototransduction cascade (such as that present in an insect's ommatidia) is also initiated by a class-specific opsin paralog, but it utilizes Gq and phospholipase C (PLC) in cell signaling. Upon activation of the rhabdomeric phototransduction cascade, transient-receptor-potential (TRP) ion channels open allowing the influx of cations and the depolarization of the cell.
An additional class of phototransduction in bilaterian animals has been proposed, based on the presence of a third clade of opsin proteins (known from mollusks, cephalochordates and vertebrates) that utilize Go in cell signaling. This phototransduction pathway has been studied in the dual-retina mantle eye of the scallop Patinopecten (Kojima et al. 1997; Gomez and Nasi 2000), the parietal eye of the reptile Uta (Su et al. 2006) and from the cephalochordate Branchiostoma (Koyanagi et al. 2002). These few investigations have left many unanswered questions. For instance, similar to the ciliary pathway, the Go-mediated pathway in both the vertebrate parietal eye and the scallop retina manifest their physiological signal through CNG ion channels. However, unlike existing data from ciliary and rhabdomeric phototransduction cascades, it would appear that Go phototransduction displays some evolutionary plasticity in the physiological outcome it specifies. As a first example, in the vertebrate parietal eye, Go-signaling leads to a depolarization of the cell but in the scallop this pathway causes hyperpolarization (Su et al. 2006). As another example of Go plasticity during evolution, some members of the Go class of opsins appear to have lost their role in phototransduction altogether and instead play important enzymatic functions as photoisomerases (Shen et al. 1994; Chen et al. 2001), enzymes that regenerate the light-reactive chromophore that binds with opsin.
If ciliary, rhabdomeric, and Go phototransduction cascades originated by duplication, one would expect to find a simultaneous origin by gene duplication of the four major components: opsin, G-protein, PLC/PDE, and ion channel. Testing this co-duplication null hypothesis requires dating the origin of each component separately.
To begin testing the co-duplication hypothesis, we have recently estimated dates for the origin of opsin clades using RTA (Plachetzki et al., in review). Opsins of the three major clades (ciliary, rhabdomeric, Go) were already known to predate bilaterians (Arendt 2003; Terakita 2005). However, no analysis had included opsins from early-branching animals, thus precluding a specific upper bound for the origins of opsin. Our study reported opsins discovered by screening trace genome-sequence data from the cnidarians Hydra magnipapillata and Nematostella vectensis and the poriferan Amphimedon queenslandica (Plachetzki et al., in review), allowing us to date the origins of the three opsin clades. This study had three major results. First, we identified a new class of opsin genes present only in cnidarians. Second, we uncovered ciliary opsins from cnidarians. Third, although our search did uncover opsin-like GPCR's, we were unable to find any true opsins in the genome of the demosponge Amphimedon queenslandica. Based on the best-supported phylogeny, using RTA allowed us to conclude that the ciliary class of opsin genes has an ancient origin in the Eumetazoan ancestor of cnidarians + bilaterian animals, while the rhabdomeric and Go subfamilies are bilaterian innovations with a single sister group of opsins in cndiarians. Based on this timing of the origins of major opsin clades, we will be able to test for co-duplication of opsin and the other components of the signaling cascade.
If G proteins co-duplicated with opsins at the origin of the three animal phototransduction cascades, they should map by RTA to the same phylogenetic interval as the opsins. As we shall discuss, the co-duplication of G-proteins and opsins can be rejected. Assuming co-duplication, the opsin results above allow a specific prediction for the timing of origin of the other transduction components of the animal phototransduction cascade. Under co-duplication, Go and Gq (the Go and rhabdomeric G-proteins) would be the result of a bilaterian-specific duplication and Gt (the ciliary G-protein) would date from a duplication preceding the Eumetazoan ancestor. In contrast to this prediction, the individual G-protein classes are known to predate animals, thus rejecting the co-duplication hypothesis. A previous survey (Suga et al. 1999) identified the full complement of G paralogs known from animal phototransduction from the demospongae, indicating pre-animal origins. Therefore, existing G proteins gained new interactions with opsin proteins during animal evolution. Co-option, not co-duplication was responsible for the origins of the interactions between opsins and G-proteins in the first ciliary, rhabdomeric, and Go pathways (Fig. 4B).
A similar approach can be taken to test the null hypothesis of co-duplication between opsins and PLC/PDE and similarly between opsins and TRP/CNG ion channels. In these cases, co-option is evident immediately as the ciliary cascade uses PDE and the rhabdomeric cascade uses PLC, two nonhomologous genes. Co-option is also evident in the origin of the rhabdomeric phototransduction cascade, the only pathway to use TRP ion channels. Despite the differences in how these signaling networks mediate their physiological effects, CNG ion channels are known to play a role in both ciliary and Go phototransduction networks (Su et al. 2006). It may be that the presence of CNG ion channels in ciliary and Go phototransduction indicates the condition of the ancestral phototransduction pathway. CNG ion channels thus represent a conserved component of ciliary and Go pathways, networks that otherwise changed through an opsin-duplication event.
In testing the null hypothesis of co-duplication as a mutational mechanism for the origin of the animal phototransduction pathways, we are confronted with the attendant challenges of dealing with vast phylogenetic time scales. Issues relating to our ability to reconstruct evolutionary events that occur at deep time intervals remain a general problem for phylogenetics. As such, it is possible that co-duplication was, in fact, an important factor in the origins of animal phototransduction pathways, but that this signal has been lost in the course of evolution. We advocate the common solution of applying additional data to the problem. Finally, further examination of the phototransduction pathway in early-branching metazoans might provide additional clues to the origins of these pathways.
|
Summary |
|---|
|
Top
SynopsisIntroductionTree-thinking and the evolution...SummaryAcknowledgmentsReferences |
|---|
Tree-thinking provides a general framework for understanding the types of variation present at the origins of novel traits. Understanding these elements is central if we are to construct more complete evolutionary narratives and a more expansive model for the evolutionary process. In some ways, the linear approach has been an important heuristic for understanding the directionality and rate of evolution, but the linear mode falls short of illuminating processes in evolution that have given rise to novel traits. This is precisely where tree-thinking approaches excel. We have used the evolution of the phototransduction cascade as a model protein-interaction network to explore the utility of tree-thinking in explaining some of the key transitions in the evolution of this pathway. This approach, however, is quite amenable to analyzing other levels of biological organization above the molecular level. Not only can the evolution of elements at a single level be better understood using tree-thinking, but interactions between levels of organization can also be addressed. One important outcome of expanding our understanding of the dual novelty-generating processes of duplication plus divergence and co-option to other levels of biological organization will be to gain insight into how different levels interact in evolutionary processes (e.g., Buss 1987
). For instance, can gene duplications provide a selective context for cell type duplications to occur? Can co-option by a given cell type of an alternative signaling or developmental pathway reinforce the evolution of novel tissues, such as vertebrate or molluscan retinas? Much of current evolutionary biology seeks to address the "novelty problem". We argue that the tree-thinking framework can better facilitate our understanding of both the patterns and the processes of evolution. |
Acknowledgments |
|---|
|
Top
SynopsisIntroductionTree-thinking and the evolution...SummaryAcknowledgmentsReferences |
|---|
We are grateful to Bernd Schierwater and Rob DeSalle for organizing the symposium "Key transitions in animal evolution" at the 2007 SICB meeting. We are also thankful to members of the Oakley lab and two anonymous reviewers for comments on the article. This work is supported by grant DEB0316330 to T.H.O from the NSF.
|
Footnotes |
|---|
From the symposium "Key Transitions in Animal Evolution" presented at the annual meeting of the Society of Integration and Comparitive Biology, January 3–7, 2007, at Phoenix, Arizona.
|
References |
|---|
|
Top
SynopsisIntroductionTree-thinking and the evolution...SummaryAcknowledgmentsReferences |
|---|
Abouheif E. Establishing homology criteria for regulatory gene networks: prospects and challenges. Novartis Found Symp (1999) 222::207–21.[Medline]
Arendt D. Evolution of eyes and photoreceptor cell types. Int J Dev Biol (2003) 47::563–71.[Web of Science][Medline]
Arthur W, Jowett T, Panchen A. Segments, limbs, homology, and co-option. Evol Dev (1999) 1::74–6.[CrossRef][Web of Science][Medline]
Britton T, Oxelman B, Vinnersten A, Bremer K. Phylogenetic dating with confidence intervals using mean path lengths. Mol Phylogenet Evol (2002) 24::58–65.[CrossRef][Web of Science][Medline]
Buss LW. The Evolution of Individuality (1987) Princeton, NJ: Princeton University Press.
Carleton KL, Spady TC, Cote RH. Rod and cone opsin families differ in spectral tuning domains but not signal transducing domains as judged by saturated evolutionary trace analysis. J Mol Evol (2005) 61::75–89.[CrossRef][Web of Science][Medline]
Catmull J, Hayward DC, McIntyre NE, Reece-Hoyes JS, Mastro R, Callaerts P, Ball EE, Miller DJ. Pax-6 origins–implications from the structure of two coral pax genes. Dev Genes Evol (1998) 208::352–6.[CrossRef][Web of Science][Medline]
Chen J, Rattner A, Nathans J. The rod photoreceptor-specific nuclear receptor Nr2e3 represses transcription of multiple cone-specific genes. J Neurosci (2005) 25::118–29.
Chen P, et al. A photic visual cycle of rhodopsin regeneration is dependent on Rgr. Nat Genet (2001) 28::256–60.[CrossRef][Web of Science][Medline]
Collin S, Trezise A. Response to Pisani et al. Current Biology (2006) 16::R320–R320.[CrossRef][Web of Science]
Conant GC, Wolfe KH. Functional partitioning of yeast co-expression networks after genome duplication. PLoS Biol (2006) 4::109.[CrossRef]
Eakin RM. Lines of evolution of photoreceptors. In: General physiology of cell specification—Mazia E, Tyler A, eds. (1963) New York: McGraw-Hill. p 393–423.
Eakin RM. Evolutionary Significance of Photoreceptors In: Retrospect. In: American Zoologist (1979) 19:647–53.
Friedrich M. Continuity versus split and reconstitution: exploring the molecular developmental corollaries of insect eye primordium evolution. Dev Biol (2006) 299::310–329.[CrossRef][Web of Science][Medline]
Geeta R. Structure trees and species trees: what they say about morphological development and evolution. Evol Dev (2003) 5::609–21.[CrossRef][Web of Science][Medline]
Gomez MP, Nasi E. Light transduction in invertebrate hyperpolarizing photoreceptors: possible involvement of a Go-regulated guanylate cyclase. J Neurosci (2000) 20::5254–63.
Gompel N, Prud’homme B, Wittkopp PJ, Kassner VA, Carroll SB. Chance caught on the wing: cis-regulatory evolution and the origin of pigment patterns in Drosophila. Nature (2005) 433::481–7.[CrossRef][Web of Science][Medline]
Goodman M, Czelusniak J, Moore G, Romero A. Fitting the gene lineage into its species lineage, a parsimony strategy illustrated by cladograms. Syst Zool (1979) 28::132–63.[Abstract]
Haeckel E. Generelle Morphologie der Organismen: Allegmeine Grudzuge der organischen Formen-Wissenschaft, mecanisch begrundet durch die von Charles Darwin reformite Descendez-Theorie (1866) 2:. Berlin: Beorg Reimer.
Halder G, Callaerts P, Gehring WJ. Inducion of ectopic eyes by targeted expression of the eyeless gene in Drosophila. Science (1995) 267::1788–92.
Hisatomi O, Tokunaga F. Molecular evolution of proteins involved in vertebrate phototransduction. Comp Biochem Physiol B Biochem Mol Biol (2002) 133::509–22.[CrossRef][Medline]
Huelsenbeck JP, Rannala B, Larget B. A Bayesian framework for the analysis of cospeciation. Evolution Int J Org Evolution (2000) 54::352–64.[CrossRef][Web of Science][Medline]
Hurles ME, Willey D, Matthews L, Hussain SS. Origins of chromosomal rearrangement hotspots in the human genome: evidence from the AZFa deletion hotspots. Genome Biol (2004) 5::R55.[CrossRef][Medline]
Jacob F. Evolution and tinkering. Science (1977) 196::1161–6.
Kojima D, Terakita A, Ishikawa T, Tsukahara Y, Maeda A, Shichida Y. A novel Go-mediated phototransduction cascade in scallop visual cells. J Biol Chem (1997) 272::22979–82.
Koyanagi M, Terakita A, Kubokawa K, Shichida Y. Amphioxus homologs of Go-coupled rhodopsin and peropsin having 11-cis- and all-trans-retinals as their chromophores. FEBS Lett (2002) 531::525–8.[CrossRef][Web of Science][Medline]
Kutschera U, Niklas KJ. The modern theory of biological evolution: an expanded synthesis. Naturwissenschaften (2004) 91::255–76.[Web of Science][Medline]
Kutschera U, Niklas KJ. Endosymbiosis, cell evolution, and speciation. Theory Biosci (2005) 124::1–24.[CrossRef][Medline]
Land MF, Nilsson D-E. Animal Eyes (2002) Oxford: Oxford University Press.
Margulis L, Sagan D. Acquiring genomes: a theory of the origins of species (2002) New York: Basic Books.
Minelli A. Limbs and tail as evolutionarily diverging duplicates of the main body axis. Evol Dev (2000) 2::157–65.[CrossRef][Web of Science][Medline]
Nilsson DE, Pelger S. A pessimistic estimate of the time required for an eye to evolve. Philo T Ro Soc B (1994) 256::53–58.
Nordström K, Larsson TA, Larhammar D. Extensive duplications of phototransduction genes in early vertebrate evolution correlate with block (chromosome) duplications. Genomics (2004) 83::852–72.[CrossRef][Web of Science][Medline]
Ohara RJ. Population thinking and tree thinking in systematics. Zool Scr (1997) 26::323–9.[CrossRef][Web of Science]
Ohno S. Evolution by gene duplication (1970) New York: Springer.
Okano T, Kojima D, Fukada Y, Shichida Y, Yoshizawa T. Primary structures of chicken cone visual pigments: vertebrate rhodopsins have evolved out of cone visual pigments. Proc Natl Acad Sci USA (1992) 89::5932–6.
Olson EN. Gene regulatory networks in the evolution and development of the heart. Science (2006) 313::1922–7.
Page RDM. Tangled trees: phylogeny, cospeciation and coevolution (2002) Chicago: The University of Chicago Press.
Page RDM, Charleston MA. Trees within trees: phylogeny and historical associations. Trends Ecol Evol (1998) 13::356–9.[CrossRef]
Peng GH, Ahmad O, Ahmad F, Liu J, Chen S. The photoreceptor-specific nuclear receptor Nr2e3 interacts with Crx and exerts opposing effects on the transcription of rod versus cone genes. Hum Mol Genet (2005) 14::747–64.
Pisani D, Mohun SM, Harris SR, McInerney JO, Wilkinson M. Molecular evidence for dim-light vision in the last common ancestor of the vertebrates. Curr Biol (2006) 16::R318–9. author reply R320.[CrossRef][Web of Science][Medline]
Plachetzki DC, Serb JM, Oakley TH. New insights into the evolutionary history of photoreceptor cells. Trends Ecol Evol (2005) 20::465–7.[CrossRef][Medline]
Salvini-Plawen LV, Mayr E. (1977) New York: Plenum Press. On the evolution of photoreceptors and eyes. Evolutionary Biology 10.
Sanderson MJ. Estimating absolute rates of molecular evolution and divergence times: A penalized likelihood approach. Mol Phylogenet Evol (2002) 19::101–9.
Schultze M. Anatomie und Physiologie der Retina. Achiv Fur Mikroskopisches Anatomie (1866) 2::175–286.
Serb JM, Oakley TH. Hierarchical phylogenetics as a quantitative analytical framework for Evolutionary Developmental Biology. Bioessays (2005) 27::1158–66.[CrossRef][Web of Science][Medline]
Shen D, Jiang M, Hao W, Tao L, Salazar M, Fong HK. A human opsin-related gene that encodes a retinaldehyde-binding protein. Biochemistry (1994) 33::13117–25.[CrossRef][Medline]
Spudich JL, Yang CS, Jung KH, Spudich EN. Retinylidene proteins: structures and functions from archaea to humans. Annu Rev Cell Dev Biol (2000) 16::365–92.[CrossRef][Web of Science][Medline]
Su CY, Luo DG, Terakita A, Shichida Y, Liao HW, Kazmi MA, Sakmar TP, Yau KW. Parietal-eye phototransduction components and their potential evolutionary implications. Science (2006) 311::1617–21.
Suga H, Koyanagi M, Hoshiyama D, Ono K, Iwabe N, Kuma K, Miyata T. Extensive gene duplication in the early evolution of animals before the parazoan-eumetazoan split demonstrated by G proteins and protein tyrosine kinases from sponge and hydra. J Mol Evol (1999) 48::646–53.[CrossRef][Web of Science][Medline]
Takezaki N, Figueroa F, Zaleska-Rutczynska Z, Klein J. Molecular phylogeny of early vertebrates: monophyly of the agnathans as revealed by sequences of 35 genes. Mol Biol Evol (2003) 20::287–92.
Taylor JS, Raes J. Duplication and divergence: the evolution of new genes and old ideas. Annu Rev Genet (2004) 38::615–43.[CrossRef][Web of Science][Medline]
Terakita A. The opsins. Genome Biol (2005) 6::213.[CrossRef][Medline]
Thompson JN. Simbiont-induced speciation. Biol J Linn Soc (1987) 32::385–93.[CrossRef][Web of Science]
Thorne JL, Kishino H. Divergence time and evolutionary rate estimation with multilocus data. Syst Biol (2002) 51::689–702.[CrossRef][Web of Science][Medline]
Turner DL, Cepko CL. A common progenitor for neurons and glia persists in rat retina late in development. Nature (1987) 328::131–6.[CrossRef][Medline]
Walls GL. The vertebrate eye and its adaptive radiation (1942) Bloomfield Hills, MI: Cranbrook Press.
West-Eberhard MJ. Developmental plasticity and evolution (2003) Oxford: Oxford University Press.
Wetts R, Fraser SE. Multipotent precursors can give rise to all major cell types of the frog retina. Science (1988) 239::1142–5.
Zhang P, Gu Z, Li WH. Different evolutionary patterns between young duplicate genes in the human genome. Genome Biol (2003) 4::56.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Vopalensky and Z. Kozmik Eye evolution: common use and independent recruitment of genetic components Phil Trans R Soc B, October 12, 2009; 364(1531): 2819 - 2832. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D.-E. Nilsson The evolution of eyes and visually guided behaviour Phil Trans R Soc B, October 12, 2009; 364(1531): 2833 - 2847. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Larhammar, K. Nordstrom, and T. A. Larsson Evolution of vertebrate rod and cone phototransduction genes Phil Trans R Soc B, October 12, 2009; 364(1531): 2867 - 2880. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. E. Grus and J. Zhang Origin of the Genetic Components of the Vomeronasal System in the Common Ancestor of all Extant Vertebrates Mol. Biol. Evol., February 1, 2009; 26(2): 407 - 419. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Sakarya, K. S. Kosik, and T. H. Oakley Reconstructing ancestral genome content based on symmetrical best alignments and Dollo parsimony Bioinformatics, March 1, 2008; 24(5): 606 - 612. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||