Emerging patterns in the regulation and evolution of marine invertebrate settlement and metamorphosis
Emerging patterns in the regulation and evolution of marine invertebrate settlement and metamorphosis
Abstract and Keywords
Life histories of many marine invertebrate groups are diverse and complex, yet frequently include a pelagic larval phase that must make a developmental, morphological, behavioural, and ecological transition into a benthic adult phase. The mechanisms underlying this life history transition are largely unknown, but with the recent application of modern molecular and genomic methodologies, some exciting patterns are beginning to emerge. New data suggest that hormones and specific neurotransmitter systems such as nitric oxide and histamine signalling play a critical role in larval settlement and metamorphosis. This chapter, which reviews available literature on mechanisms underlying this major life history transition in selected marine invertebrate groups, discusses new avenues of research that broaden the understanding of how gene regulatory networks are orchestrated during complex developmental programs.
The biophysical properties of seawater and the connectivity of marine habitats impose specific opportunities and constraints on reproduction and dispersal in the ocean. Marine invertebrates have evolved a diversity of reproductive strategies (Fig. 3-1). One distinct mode of reproduction is via dispersing larvae, which must settle out of the plankton and undergo a metamorphic transition into the benthic adult form (Strathmann 1990) Despite the commonality of this life cycle transition in the marine environment, the molecular, cellular, and physiological mechanisms that regulate settlement and metamorphosis as individuals move from the plankton to the benthos are poorly understood. This in turn limits our ability to consider mechanisms and pathways of the evolution of the plankton–benthos transition.
Phenotypes are linked with genotypes via physiological and developmental (proximate) mechanisms (see also Chapter 1). Understanding mechanisms of life history evolution (the ultimate level of inquiry) requires a fundamental understanding of such proximate mechanisms across relatively closely related taxa (Wray 1994). Research of proximate mechanisms of marine invertebrate life histories so far has largely been biased towards early life history stages due to the difficulty of rearing and experimentally manipulating larval and juvenile stages in the laboratory for extensive periods of time. Consequently, life history models and empirical data have focused on the relationship between egg investment and larval type, and how this association affects larval dispersal and survival and subsequent evolutionary changes in life histories (Vance 1973a,b, Christiansen and Fenchel 1979, Caswell 1981, Perron 1986, Roughgarden 1989, McEdward 1997). Specific developmental mechanisms underlying metamorphic competence and settlement have received less attention (but see Hentschel and Emlet 2000, Day and Rowe 2002, Marshall and Keough 2003, Toonen and Tyre 2007).
Understanding mechanisms underlying complex life cycles and their evolution in marine habitats requires an understanding and consideration of both ecology and development (Moran 1994). From an ecological perspective, dispersal potential and the identification of suitable settlement sites are larval life history traits with significant fitness effects (for a review see Chia and Rice 1978). Mechanisms regulating the development of feeding, dispersal and sensory structures that enable larvae to survive in the plankton and to detect suitable settlement sites are likely to be under strong selection, creating an important link between such developmental mechanisms and larval ecology. For example, trade-offs between dispersal potential and survival are mechanistically rooted in processes regulating the development of these larval structures, including mechanisms of plasticity (see Chapter 17). Similarly, the timely recognition of settlement cues, hence metamorphic competence, depends on the differentiation of larval sensory structures. The development of these structures is likely integrated with (p.30)
Intraspecific studies of the morphological, physiological, and molecular components regulating life cycle transitions (see above) provide comparative data to enable examination of evolutionary changes among species, and for testing of hypotheses on how species may have evolved the diversity of life histories we observe in extant marine invertebrate taxa. Thus, we can use closely related species that vary in particular life history characters and determine how such differences at the level of the life history are regulated at the molecular level.
In this chapter, we discuss proximate mechanisms underlying metamorphic competence and settlement, and how changes in such mechanisms can lead to changes in life histories over evolutionary time. We first describe metamorphic competence and settlement in the context of marine invertebrate life cycles and summarize mechanisms that regulate these processes. We then analyse energy allocation trade-offs between larval and juvenile structures for echinoids (sea urchins and sand dollars). Echinoid larvae are characterized by extreme indirect development (i.e. distinctly separate programs for larval and juvenile morphogenesis and differentiation). Finally we summarize the diversity of chemical cues that are known to modulate the settlement responses among marine invertebrate larvae and discuss hypotheses on the evolution of settlement strategies.
3.2 Introduction to marine invertebrate life histories
Marine invertebrate life histories are diverse and can be categorized based on mode of gamete release (e.g., broadcast or spermcast spawning, copulation, pseudocopulation), location of development (e.g., planktonic, benthic, brooding), presence or absence of larval stages (e.g., indirect or direct), feeding requirements of these larvae (non-feeding or feeding), and associated presence or absence of a settlement and/or metamorphic transition. Marine invertebrate groups have evolved diverse combinations of these broad categories, allowing for some generalizations. In Figure 3-1 we present a simplified diagram of this diversity, while keeping in mind that variation is present in each of these modes. In this chapter, we focus our discussion on indirect developing species that include a distinct feeding or non-feeding planktonic larval stage, and which then undergo a metamorphic transition associated with settlement into the benthic habitat.
Indirect development in the planktonic environment includes a distinct larva, with or without feeding structures. Larvae with the ability to feed are referred to as planktotrophic (see Box 3-1 for definition of feeding development and planktotrophy) (p.31) (p.32) because they require nutrition derived from feeding on plankton in order to reach the juvenile stage. Larvae that can reach metamorphosis without the necessity for exogenous food frequently lack feeding structures; they are called lecithotrophic (see box for definition of non-feeding development and lecithotrophy), as they are sustained instead by maternal nutrients provided in the egg. In some rare cases, larvae have the ability to feed but do not require food to become juveniles. These larval types are referred to as facultative planktotrophs (Emlet 1986, Hart 1996) and may represent a transitional mode between planktotrophic and lecithotrophic development. Maternal investment in the egg is a relatively strong predictor of developmental mode across invertebrate phyla, where eggs with higher provisioning typically develop into non-feeding larvae (Vance 1973a,b). The relative amount of energy investment varies among lineages, such that an absolute threshold value does not apply broadly (reviewed in Strathmann 1985).
In these biphasic life cycles, selection on the planktonic larval stage may act quite differently from that on the benthic juvenile/adult stages. This situation creates trade-offs between life cycle phases that will ultimately shape the evolution of indirect development. Across all developmental strategies, the larval period may be under strong selection because larvae face various mortality risks including predation and exposure to adverse environmental conditions such as drastic salinity, temperature, and pH changes, all of which can have detrimental effects. Lecithotrophic larvae are, by definition, maternally provisioned with adequate resources to complete the pre-component stage of the larval period. Therefore, the development time of these larvae is typically shorter than for planktotrophic larvae, and environmental factors likely to affect development are not related to food acquisition. Survival and growth of pre-competent planktotrophic larvae are largely dependent on food availability in the environment in addition to abiotic factors. Larval distribution into favorable feeding environments is influenced by adult spawning behavior as well as by larval behavior. In many species, feeding larvae exhibit morphological plasticity to adjust feeding and digestive structures to increase energy acquisition in variable feeding environments (reviewed in Chapter 17).
Metamorphosis is an intricate part of indirect development and so is metamorphic competence, the phase directly preceding settlement (Hadfield et al. 2001, Bishop et al. 2006a, Heyland and Moroz 2006, Hodin 2006). Competent larvae respond to highly specific environmental cues by rapidly settling into the juvenile habitat (Hadfield et al. 2001). In species with indirect development, juvenile structures differentiate during the larval phase and are likely creating energy allocation trade-offs between these two distinct developmental compartments. We will discuss regulatory mechanisms underlying such trade-offs in sea urchin and sand dollar larvae, and the consequences of these trade-offs for life history evolution. It should be noted that although such a developmental pattern is relatively common among marine invertebrates, there are numerous taxa (e.g., many lophotrochozoan groups) without such a clear distinction between juvenile and larval structures. In these groups we generally see a more gradual transition from larva to juvenile (further discussed below).
3.3 Regulation of larval development and the evolution of feeding modes in echinoids: Energy allocation trade-offs during larval development
In echinoids (e.g., sea urchins and sand dollars), bryozoans, nemertean and crustacean larvae, juvenile tissue begins to differentiate inside the larval body as a distinct structure during the pre-competent period. During the entire larval period, larval feeding supports the growth and differentiation of this juvenile tissue in addition to maintaining larval structures; this can create an energy allocation trade-off between larval and juvenile growth and development (Hart and Strathmann 1994, Heyland and Hodin 2004, Strathmann et al. 2008). In many of these marine invertebrate groups, growth and development of juvenile structures inside the larval body are extraordinarily plastic. In echinoid and bryozoan larvae, for example, the maturation of the juvenile rudiment can be largely decoupled from larval development, depending on environmental food conditions (Strathmann et al. 1992, 2008). Such allocation “decisions” require the integration of ecological with developmental processes (p.33) that ultimately determine the fitness of the organism. Precisely how endogenous and exogenous energy resources are divided between the development of larval versus juvenile structures will depend on the specific environmental conditions in which a larva finds itself (for some examples see Toonen and Pawlik 2001, Strathmann et al. 2008). Moreover, larval life history traits, such as age and size at metamorphosis, as well as juvenile traits such as post-settlement performance and timing of first reproduction, may be linked to maternal investment and availability of nutrition and predation in the larval environment (reviewed in Strathmann 1990, Hanvenhand 1993, Levin and Bridges 1995, Morgan 1995, Hentschel and Emlet 2000). These relationships are schematically depicted for these examples of indirect development in Figure 3-2. The quality of the settling larva can influence the size of the juvenile, which affects the probability of survival for this stage. Extended larval periods may therefore lead to increased juvenile quality, which creates a trade-off between larval survival and juvenile performance. Note, however, that these trade-offs primarily exist for feeding larvae, because non-feeding larvae are limited by endogenous reserves and therefore more prone to declining juvenile quality by staying longer in the plankton or delaying settlement
3.3.1 Hormonal regulation of juvenile development
Many animals with indirect life histories and a drastic metamorphic transition, such as frogs and insects, regulate larval and juvenile development using hormones (reviewed in Chapter 6 for several fish species, in Chapter 7 for amphibians, and in Chapters 5, 13, and 20 for holometabolous insects). The role of hormones in marine invertebrate life histories may be equally critical and widespread, at least for deuterostomes and ecdysozoan larvae (for review see Heyland et al. 2005). New data on thyroid hormone (TH) signaling in basal chordates, that is, urochordates and cephalochordates, suggest a regulatory function of THs in the metamorphosis of these groups as well. Paris et al. (2008) showed that several types of thyroid hormone (T4, T3, and thyroid hormone triiodothyroacetic acid—TRIAC) can induce metamorphosis by binding to the amphioxus thyroid hormone receptor (Paris et al. 2008). Moreover, many TH synthesis and signal transduction genes have been annotated from the recently released Amphioxus genome (Holland et al. 2008). In the ascidian Boltenia villosa, inhibition of TH synthesis leads to a correlated inhibition of adult differentiation after settlement (Davidson et al. 2002), and in three other ascidian species thyroxine effects on larval development as well as TH synthesis have been confirmed (D’Agati and Cammarata 2006). Still, detailed information on how THs regulate the energy allocation trade-off between larval and juvenile structures remains to be elucidated in cephalochordates and urochordates.
THs have been identified as regulators of development and metamorphosis in echinoids (sea urchins and sand dollars) (Chino et al. 1994, Saito et al. 1998, Hodin et al. 2001, Heyland and Hodin 2004, Heyland et al. 2005, Heyland and Moroz 2006, Heyland et al. 2006a,b, Hodin 2006). Specifically, thyroxine and T3 lead to a significant acceleration of juvenile rudiment development as well as a short-arm larval phenotype, suggesting that thyroxine and other THs may function in the coordination of energy investment between larval and juvenile structures. In one sand dollar species with obligatory feeding larvae, Leodia sexiesperforata, TH treatment resulted in larvae metamorphosing in the absence of food (i.e., functionally non-feeding larvae), suggesting that TH may have played an important role in the evolution of non-feeding development (Heyland and Hodin 2004). The molecular mechanisms underlying TH action in echinoids are still poorly understood, but preliminary data suggest that larvae can synthesize TH endogenously (Heyland et al. 2006a,b). In addition to these endogenous sources, TH can originate from algal food (Chino et al. 1994, Heyland and Moroz 2005). While this has broad implications for larval life history evolution (Heyland et al. 2005, Heyland and Moroz 2005, Miller and Heyland 2010) that are further discussed below, the mechanisms of TH synthesis in larvae and the transfer between algae and larva remain to be elucidated.
As in insects, molting cycles in marine arthropods are regulated by hormones (Hartnoll 2001), specifically molt inhibiting hormone (MIH) and various steroid hormones related to 20-hydroxy ecdysone. Crustacean hyperglycaemic hormone (CHH) has a very similar structure to MIH and appears to be involved in the molt regulation via inhibition of ecdysteroid synthesis. Finally, methylfarnesoate can stimulate molting by increasing ecdysteroid production. A review of recent findings suggests that ecdysteroid action in crustaceans is also mediated via nuclear hormone receptors (Hartnoll 2001, Wu et al. 2004).
3.3.2 Hormonal signaling and the evolution of alternative life history modes
Life history trade-offs can result from two life history traits with opposite fitness effects that are mechanistically (i.e., physiologically, developmentally, genetically) linked in the same organism. (p.35) Mechanisms underlying trade-offs can therefore affect the course of evolutionary change in life histories (see also Chapters 1 and 2 for other examples). Here we will discuss one such evolutionary transition in life history mode in more detail—the evolution of non-feeding development from feeding development—and the role that THs might have played in this process.
Among echinoid groups, non-feeding development has evolved many times independently from feeding development. This evolutionary transition resulted in heterochronic shifts in development as well as the reduction or complete loss of feeding structures (reviewed in Wray 1995). One consequence of feeding loss may have been the loss of exogenous hormone sources, creating a requirement for endogenous hormone synthesis (Heyland and Moroz 2005). Expanding on our data from Leodia sexiesperforata, we hypothesize that up-regulation of endogenous TH synthesis may have been sufficient to transform an obligatory feeding larva into a facultative feeding larva (Heyland et al. 2004). Therefore the ability of larvae to endogenously synthesize THs may be a pre-adaptation for the evolution of non-feeding development in this group (Heyland and Hodin 2004). Based on that hypothesis, we predict that clades, which frequently underwent the evolutionary transition to non-feeding development (such as the Clypeasteroida) may have a higher capacity to synthesize hormones endogenously in comparison to clades that rarely or never underwent this transition, such as the Diadematoida urchins.
The sand dollar Dendraster excentricus produces a feeding larva and responds to the TH synthesis inhibitor thiourea by delaying the metamorphic transition, an effect that can be rescued with exogenous hormone (Heyland and Hodin 2004). Similar responses have been found for facultative planktotrophic larvae of the sea biscuit Clypeaster rosaceus (Heyland et al. 2006b) and the non-feeding Japanese sand dollar Peronella japonica (Saito et al. 1998). Together these results indicate that larvae of these three species have the capacity to endogenously synthesize THs, which also regulate larval development and metamorphosis. Preliminary experiments with Diadema antillarum larvae show very different results; that is, there is no evidence for endogenous hormone synthesis (Hodin and Heyland unpublished data). While these results are preliminary and further sampling of taxa is needed, they are consistent with the hypothesis that extant representatives of taxa that frequently evolved non-feeding development are more likely to synthesize THs endogenously.
One important question is whether endogenous and exogenous hormone sources have a) identical structures and b) comparable effects on development. The data so far show that THs in algae are the same as those that larvae endogenously synthesize and TH from endogenous and exogenous sources effect development to metamorphosis in a similar way (Heyland and Hodin 2004). Still, the detailed signaling systems involved in the process have not been elucidated. Studies focusing on this question will have to assess whether larval and juvenile development is regulated by a linked signaling system and whether manipulation of endogenous TH synthesis in larvae can lead to quantitative changes in development comparable to changes induced by nutritionally derived hormone.
A recent study suggests that changes in ecdysteroid signaling may have played a role in the evolution of alternative life histories in crustaceans. As in the majority of marine invertebrate larvae, the larval stage of crustaceans is the dispersal stage whereas the juvenile and adult stage is involved in growth and reproduction. In a parasitic copepod species, the function of these life history phases has been reversed. Metanauplius larvae of Caribeopsyllus amphiodiae live parasitically inside burrowing ophiuroids (Amphiodia urtica) before they transform into short-lived free-living adults (Hendler and Dojiri 2009). The pedomorphic life cycle of this copepod probably evolved through a delay of metamorphosis regulated by developmental hormones (Hendler and Dojiri 2009).
In summary, hormonal signaling pathways appear to be consistently involved in the differentiation of juvenile structures in marine invertebrate species, primarily in deuterostome and ecdysozoan larvae. The extent to which these hormonal signaling pathways regulate life history trade-offs in other groups such as the lophotrochozoa and non-bilaterian groups remains to be explored.
(p.36) 3.4 Mechanisms underlying larval settlement and the evolution of alternative settlement strategies: Signal detection and modulation during settlement
Regardless of their nutrition source, both feeding and non-feeding larvae eventually reach metamorphic competence, a developmental stage at which they become responsive to species-specific settlement cues that contain information about the suitability of future settlement sites (Fig. 3-1). The regulation of developmental and physiological processes during competence and settlement requires the coordination of a complex set of exogenous and endogenous factors. For example, echinoid larvae are sensing and processing information about their current environment, while trading off the allocation of energy into larval versus anticipatory juvenile structures (see above). Marine larvae perceive information about future habitats via chemical signals detected either in the water or via direct contact with the substratum. Responses to these exogenous cues need to be coordinated with endogenous signals that regulate the development of the organisms, including hormones, peptides and neurotransmitters. Understanding the source, nature, and transduction of these signals, and their integration with morphogenesis and differentiation, can provide critical insights into the regulation and evolution of marine life histories (Zimmer and Butman 2000).
Figure 3-3 provides a model of how settlement signals may be processed by the larva and how changes in transduction mechanisms may contribute to the evolution of alternative settlement strategies. For the purpose of this discussion we distinguish between two main components: the larval sensory system and the competence system (Fig. 3-3A). Receptors expressed by cells in the sensory system are capable of detecting specific settlement cues and therefore establish the link between the larval nervous system and the environment. The
3.4.1 The sensory system: Cues, receptors, and signal transduction mechanisms
Exogenous cues capable of inducing settlement are detected by sensory components of the larval nervous system. These cues are diverse and convey information about food availability, physical or chemical substrate composition, or other biotic and abiotic factors in the post-settlement habitat (reviewed in Chia and Rice 1978, Burke 1983b, Bishop et al. 2006b). Cues can be released by complex biofilms, predators, potential prey species, conspecifics, algae, or sediment. Detection of these compounds by competent larvae requires a high degree of specificity, because the consequences of settling into sub-optimal habitat can be detrimental (reviewed in Levin and Bridges 1995). Still, the exact chemical identity of the settlement-inducing cue is rarely known, further complicating studies on signaling transduction systems of the settlement response in marine invertebrate larvae (but see Hadfield and Paul 2001).
The apical sensory organ (ASO) is the central component of the larval nervous system and its involvement in the settlement response has been discussed for a diversity of marine invertebrate species, including mollusks, cnidarians, and echinoderms (Chia and Rice 1978, Hadfield et al. 2001, Page 2002, Kempf and Page 2005, Bishop et al. 2006a, Bishop and Hall 2009). For example, in gastropod larvae the ASO has motor and sensory functions and coordinates the activity of swimming and feeding structures in the larval stage (Page 2002). It also has a critical function in perceiving and modulating sensory input during the competent stage leading to the metamorphic transition (Burke 1980, Couper and Leise 1996, Hadfield et al. 2000, 2001). Using excision experiments in sand dollar larvae, Burke (1983b) was able to show that the apical neuropile and oral ganglion mediate the perception of the natural cue and control the initiation of metamorphosis. Ablation experiments on the tropical nudibranch Phestilla sibogae conclusively showed that the ASO is an important structure for the ability of larvae to metamorphose (Hadfield et al. 2000). Recently, the adoral lobe has been discussed as an important structure in echinoid larval olfaction, specifically in the detection of settlement cues, and preliminary data suggest that chemosensory receptors are involved in the detection of these cues (Bishop and Hall 2009).
The repertoire of signaling molecules used by larvae to transduce settlement cues includes both activating and repressing neurotransmitter actions. For example, serotonin (5HT) has been detected in neurons of larval nervous systems across diverse marine invertebrate phyla (Burke 1983b, Heyland and Moroz 2006) and abundant evidence suggests a modulatory function. Inhibitory effects of 5HT on settlement have been discovered in cnidarians (Zega et al. 2007), ascidians (Zega et al. 2005), and bryozoans (Shimizu et al. 2000, Yu et al. 2007b). In contrast to these findings, 5HT has settlement-inducing effects in several mollusk species (Couper and Leise 1996, Satuito et al. 1999, Urrutia et al. 2004) and one cnidarian species (McCauley 1997). Intriguingly, settlement in barnacle larvae was inhibited by 5HT-like barettin-related compounds isolated from the marine sponge Geodia barretti, suggesting that 5HT-related compounds of naturally exogenous origin may also be involved in interspecies interactions (Hedner et al. 2006). Still, neurotransmitter action in settlement is not restricted to 5HT. A diversity of other amino acids and transmitters appear to modulate this process, including dopamine, noradrenalin, GABA, histamine, glutamine, and many others (reviewed in Hadfield and Paul 2001, Paul and Ritson-Williams 2008). However, it appears that many such compounds act as “artificial” modulators of the settlement process; that is, they are likely acting downstream of the receptor activation and do not represent a specific exogenous signal.
Biofilms are an important feature of marine ecosystems, consisting of communities of bacterial, (p.38) fungal, or algal species. Bacteria use a specific group of signaling molecules for inter- and intra-specific communication, called quorum sensing. Although quorum sensing compounds are restricted to prokaryotic species, increasing evidence suggests that eukaryotes can respond to these compounds, as exemplified by various chemical interactions of bacteria with their eukaryotic hosts (Keller and Surette 2006) as well as the settlement response of algal dispersal stages in response to quorum sensing compounds (Joint 2006). A signaling function of such compounds, especially acylhomoserine lactones (AHSLs), in the settlement response of marine invertebrate species has been previously proposed (Dobretsov et al. 2009). However, their low stability in seawater make experiments with such compounds difficult and convincing evidence for any biological function related to marine invertebrate settlement is lacking (M. Hadfield, pers. comm.). Finally, several recent studies have implicated genes of the innate immune pathway in signal transduction mechanisms of settlement (Davidson and Swalla 2002, Heyland and Moroz 2006, Roberts et al. 2007, Meyer et al. 2009, Williams et al. 2009a). The analysis of such signal transduction pathways in the context of the settlement response may provide interesting new insights.
G-protein coupled receptors (GPCRs) are important transducers of exogenous signals in animals and plants (Schoneberg et al. 2007). They have been abundantly discussed in the context of vertebrate and insect olfaction, and increasing evidence suggests that these seven trans-membrane domain-spanning cell-surface receptors fulfill important functions during the induction of metamorphosis in at least some marine invertebrates. GPCR involvement in metamorphic signal transduction has been demonstrated for hydrozoans, barnacles, and mollusks (Baxter and Morse 1987, Schneider and Leitz 1994, Clare 1996b). Similarly, signal transduction during metamorphic induction frequently involves protein kinase activity, as shown in a wide variety of marine invertebrate species, including cnidarians, crustaceans, and polychaetes (Freeman and Ridgway 1990, Yamamoto et al. 1996, Biggers and Laufer 1999, Siefker et al. 2000). These receptor systems likely act together and may involve calcium signaling (e.g., sea urchin Strongylocentrotus purpuratus (Amador-Cano et al. 2006)). Calcium signaling also appears to be critical for barnacle metamorphosis (Clare 1996a) and cAMP dependent protein kinase activity has been shown to mediate settlement in the polychaete Hydroides elegans. In contrast, data on bryozoans, gastropods, and polychaetes (H. elegans) suggest no involvement of GPCR activity in the settlement process (Holm et al. 1998, Bertrand and Woollacott 2003), further emphasizing the variability of transduction events in marine invertebrate species.
3.4.2 The competence system
Two important factors affect the transition between plankton and benthos: the availability of suitable settlement sites (an ecological consideration) and the morphogenetic progression of juvenile structures inside the larval body (a developmental consideration). Larvae will not be responsive to specific settlement cues before they reach metamorphic competence (by definition, the developmental maturity to respond to these settlement cues). During competence, larvae have to balance the benefit of finding the “perfect” habitat with the risk of mortality. Based on our simplified model (Fig. 3-3), we suggest that larvae have a competence system that regulates these “decisions” and integrates them with signals from the external environment. For the purpose of this discussion, we propose that this competence system has the ability to modulate the sensitivity of the sensory system and regulate larval and juvenile morphogenesis differentially. We will first discuss some known components of this system, and then speculate on how it may be different in species with different settlement strategies.
Several lines of evidence implicate endocrine and neuroendocrine signaling in the regulation of metamorphic competence. For example, EGF-like signaling regulates programmed cell death (PCD) associated with metamorphic competence in the metamorphic transition of the ascidian Herdmania curvata (Eri et al. 1999). A new study on the sea hare Aplysia californica shows that transcripts related to hormone synthesis and signaling, specifically several nuclear hormone receptor genes, are dif (p.39) ferentially expressed pre- and post-settlement (Heyland et al. 2010). Moreover, over 50 secretory products and peptides show increased gene expression levels during the metamorphic transition, and preliminary data suggest that some of these gene expression changes may be linked to behavioral changes (Heyland et al. 2010).
Both nitric oxide (NO) and thyroid hormones (THs) have been shown to regulate metamorphic competence in echinoids (Bishop et al. 2006a). Specifically, NO, an inhibitory signal on the progression of metamorphosis) is antagonized by TH (a metamorphosis-promoting signal). Evidence for such a regulatory mechanism originates from combined treatments of TH and NO as well as histochemical analysis of the larval nervous system during metamorphic competence (Bishop et al. 2006a).
One transmitter that has been recently identified as a natural inducer of settlement in sea urchin larvae is histamine (Swanson et al. 2006, 2007). Swanson and colleagues (2006) first identified histamine from the host plant of newly recruited sea urchin larvae (Holopneustes purpurascens) using gas chromatography-mass spectrometry (GC-MS) and nuclear magnetic resonance (NMR). They also noted that different algae species contain different amounts of histamine. Next, they tested histamine experimentally on metamorphically competent larvae and showed that it can induce settlement. Finally, they were able to provide evidence that the sensitivity of larvae to histamine varies with the length of the competent period (Swanson et al. 2007). Specifically, older larvae showed higher sensitivity to histamine than competent larvae, suggesting that histamine concentration is used as a proxy for larvae to assess their future habitat and that larvae of this sea urchin species expand the range of potential settlement sites by changing their sensitivity to histamine.
Several recent studies have employed genomic, transcriptomic, and even proteomic approaches to elucidate mechanisms underlying metamorphic competence and settlement. Although these studies have produced a large number of candidate mechanisms, they frequently lack data confirming a functional involvement. In a recent review, Heyland and Moroz (2006) found evidence for signaling pathways related to the stress response, immunity, and apoptosis, which appear to have been co-opted for signaling events during settlement and metamorphosis. Some very recent studies provide evidence from a genomics level that an independent developmental program is activated during the metamorphic transition. Williams et al. (2009a) analysed gene expression changes in the abalone Haliotis asinina using a microarray, and found evidence for marked temporal changes in transcription associated with the attainment of competency, and again with the metamorphic transition of this mollusk species. Specifically, their data suggest that gene activity associated with endogenous attainment of competence is somewhat independent from gene activity associated with exogenous induction of settlement. Jacobs et al. (2006) analysed gene expression changes as a consequence of delayed metamorphosis in the ascidian species Herdmania momus. They discovered that independent of larval age, an autonomous developmental program is activated in response to settlement cues. The fact that both mollusks and ascidians appear to recruit independent developmental regulatory mechanisms during metamorphosis further emphasizes the modularity of this process and has important evolutionary implications, which are further discussed below.
Although settlement involves a drastic change in habitat among marine invertebrate species, morphological changes associated with this transition are not always so radical. Some groups, such as barnacles and echinoids, undergo drastic changes when transferring to the benthos, losing the larval body almost entirely. In these groups, juvenile structures generally become functional only after settlement. In other groups, such as mollusks and polychaetes, the transition from the larva to a juvenile is more gradual, in that functional larval structures are carried over to the juvenile. A recent comparison between the proteome of a barnacle and a polychaete species (Mok et al. 2009), representing two extremes of this spectrum, revealed that these differences are also reflected in protein expression patterns. Barnacle larvae show a much more drastic change in protein expression during the metamorphic transition.
Competent larvae generally respond to species-specific settlement cues, yet important differences exist among species in how they adjust their responsiveness to these cues over the duration of larval competence (reviewed in Bishop et al. 2006a). Some larvae may lower their sensitivity to a specific settlement cue as a function of time spent in the competent period, whereas other types will not change their specificity over time and explore putative juvenile habitats until their requirements are met. These two examples are extremes of a spectrum of settlement strategies, as illustrated graphically in Figure 3-3B.
The evolution of settlement strategies is likely influenced by ecological factors such as predation, food availability, and competition in a given pre- and post-settlement habitat. We predict that differences between such strategies are correlated with changes in the larval sensory and competence system outlined above. Possible causes of changes relevant to evolutionary changes are illustrated in Figure 3-3C. For example, a larva with an opportunistic settlement strategy (Fig. 3-3B, S21–2) may have the ability to shift the affinity of receptor systems to specific settlement cues as a function of the time that it spent in the competent state. Alternatively, opportunistic larvae may be able to adjust endogenous modulators of the settlement response. The evolution of larval selectivity likely depends on the relative frequency at which putative habitats of variable qualities are encountered, as well as the length of the larval search period (Toonen and Tyre 2007). Bishop et al. (2006a) also hypothesized that changes in regulatory mechanisms underlying the decision process are likely linked to the metabolism of the larva.
While direct interactions between the competence and sensory systems have rarely been studied in detail, two recent examples help to illustrate the potential importance of this interaction. Holopneustes purpurascens larvae increase their sensitivity in response to histamine as a function of time spent in the competent stage (Swanson et al. 2006, Swanson et al. 2007). Based on our model in Figure 3-3, this shift in sensitivity corresponds to a shift in the response curve to the left, so that lower concentrations of an exogenous modulator (settlement cue) can illicit the settlement response. Note that in this example, the shift in sensitivity could either result from changes in histamine receptor affinity to its ligand (change in settlement system only) or from interactions between the competence system and the settlement system. Another example for a potential interaction between the competence and settlement system was illustrated by Bishop et al. (2006a). Thyroid hormones, which are synthesized endogenously by competent larvae (Chino et al. 1994, Heyland and Moroz 2005, Heyland et al. 2004, 2006a,b), can inhibit the formation of neurons containing nitric oxide synthase (NOS) in the apical neuropile of sea urchin larvae, thus counteracting the inhibitory role of NO in competent larvae.
It is unlikely that competent larvae have fixed responses to specific environmental stimuli. Instead, one would predict that settlement behavior will vary depending on the internal environment of the organism as well as the external environment under which it encounters the signal. Developmental plasticity that allows larvae to tolerate and colonize variable environments may play a crucial role in determining both the distribution and evolution of marine invertebrate species. In particular, transcriptional plasticity is a potentially powerful mechanism for fast phenotypic responses without trade-offs that compromise fitness under different ecological conditions, because the phenotypic variability is genome-encoded, yet can be switched to suit on a generational basis (Levine and Davidson 2005). For larvae of the tropical abalone Haliotis asinina, for example, variation in the settlement cue induces differential transcriptional responses in a whole suite of genes and tissues that are involved in both attainment of competency and metamorphosis into the post-larval form (Williams and Degnan 2009). The widespread nature of the differences in response suggests a hormonal role via the neuroendocrine system, because hormones can act simultaneously on numerous genes and tissues of an organism (hormonal pleiotropy); the nervous system can respond directly to signals from the environment by producing chemical signals that in turn lead to the production of hormones that can later affect gene expression (p.41) (Kucharski et al. 2008). We suggest this as a valuable focus for further study of the mechanisms and evolution of marine invertebrate settlement and metamorphosis, both within and between species. Comparative studies among species have already demonstrated variation in life history strategies (Krug 2009), but the mechanisms underlying specific changes remain elusive. Several studies have looked at the nervous systems of planktotrophic and lecithotrophic gastropod larvae and found interesting clade-specific differences between both sensory and non-sensory components (Page 2002). Still, a close comparison of asteroid larvae with different developmental modes (feeding and non-feeding) provides evidence for a high degree of conservation in the neurogenesis and the organization of the nervous system (Elia et al. 2009).
In summary, available data suggest that settlement strategies are variable, and presumably evolutionarily labile, as a consequence of specific ecological factors in the marine environment. Although mechanisms underlying such strategies are variable, some lineages appear to have evolved hormonal signaling mechanisms independently. Future studies of such mechanisms will therefore be valuable in contributing to our understanding of specific constraints and opportunities for the evolution of complex developmental programs.
3.6 Future directions
The relative paucity of data describing the mechanisms underlying settlement and metamorphosis for most marine invertebrates provides the most critical hurdle towards generating a broad and comprehensive picture for our understanding of the development, ecology, and evolution of these critical events in life histories. Some preliminary data suggest that hormonal signaling systems may play a role in the energy allocation between larval and juvenile structures and that these hormones may originate from nutritional sources in some cases. We have also described settlement strategies pursued by various marine invertebrate species. While chemo-reception on the level of the larval nervous system appears to play a pivotal role in the transduction of environmental signals, the actual cues used in this process appear to be almost as diverse as the larvae responding to them. Some modulators such as NO and 5HT appear to be the same in different groups and it will be interesting to further investigate these cases of parallel or convergent evolution. For the most part, however, more detailed information on these mechanisms is needed.
Larval development, ecology, and evolution have been most thoroughly studied in echinoderm larvae. For example, comparative analysis of two Australian sea urchin species with dramatically different developmental modes revealed that major developmental changes occurred before the trophic transition to non-feeding development (Sly et al. 2003). Some of these evolutionary differences in development correlate with larval ecology, suggesting that larval ecology represents an important factor shaping the evolution of development in echinoderms (Wray 1994). Unfortunately, these two species differ in a multitude of ways (e.g., developmental and morphological), precluding a clear approach for discerning how or why one strategy diverged from the other. Another productive use of the comparative approach could involve study of closely related species with more subtle differences in particular aspects of the life history (e.g., egg size range within obligate planktotrophic species). Within-species studies dissecting the signaling network and developmental architecture underlying plastic responses will also likely provide a productive means of identifying mechanisms for generating phenotypic diversity.
1. Marine invertebrate life histories are diverse and frequently involve microscopic dispersal stages (larvae) that are morphologically, physiologically, developmentally and ecologically different from the adult stage. Such larval stages can be either feeding or non-feeding, and undergo a drastic metamorphosis at the end of the planktonic period that transforms them into the juvenile stage.
2. Mechanisms underlying larval development and metamorphosis are poorly understood for marine invertebrate species. Available data suggests that both neuronal and hormonal signaling systems are (p.42) involved in the modulation of environmental input. Both systems overlap during larval development. Neuronal signaling involves a wide variety of neurotransmitters and is functional on the level of the metamorphically competent larva responding to specific settlement cues. Hormonal signaling appears to be functional in coordinating the larval and juvenile developmental program. Novel genomics and proteomics approaches reveal a diversity of signaling pathways awaiting functional characterization.
3. Thyroid hormone signaling in echinoid larvae is involved in the regulation of pre-metamorphic development by coordinating energy allocation between larval and juvenile structures. Changes in this signaling system, including switching from an exogenous to an endogenous signal, may have facilitated the evolution of non-feeding development.
4. Understanding the mechanisms underlying life history evolution in marine invertebrates will require detailed mechanistic studies on regulatory mechanisms of larval development and metamorphosis. Larval ecology is one, if not the most, important factor shaping the evolution of these mechanisms.