Twisted Shells, Spiral Cells, and Asymmetries: Evo-Devo Lessons Learned from Gastropods
- 225 Downloads
Gastropods (snails) are members of the phylum Mollusca and have been studied in a range of evolutionary developmental biology contexts. As members of the third major branch of the bilaterians, Spiralia (Lophotrochozoa) and the most speciose group outside of the arthropods, gastropods present a microcosm of body plan diversity. Historically, gastropods served as major players in comparative embryology research, and the cellular homologies of the spiral cleavage pattern continue to be important in understanding how changes in early development lead to diverse larval and adult phenotypes. The hallmarks of the gastropod body plan (torsion, asymmetry, external coiled shell) and various aspects of their development are being explored in the light of new understanding of gastropod phylogeny and conserved molecular developmental mechanisms. This chapter explores some of the latest Evo-Devo lessons that snails have taught us and suggests areas for renewed attention.
KeywordsSpiral cleavage Biomineralization Left-right asymmetry Torsion mRNA localization
“Ontogeny does not recapitulate phylogeny: it creates it.”
Walter Garstang (1922)
At the end of the nineteenth century, gastropods were among the leading models used in developmental biology. Comparative embryology used spiral-cleaving embryos, including the gastropods Crepidula (the slipper snail) and Tritia (the mud snail, formerly Ilyanassa), to explore early development, patterns of spiral cleavage, and cellular homologies, explicitly linking development and evolution. Classical examinations of cell lineages by Edwin Grant Conklin, Frank Rattray Lillie, and Edmund Beecher Wilson showed that despite their highly divergent adult body plans, the spiral cleavage pattern is conserved across molluscs, annelids, and some other groups, demonstrating that early development could be used to understand animal relationships and evolution. Gastropods were instrumental in Garstang’s challenge to Haeckel’s biogenetic law. The biogenetic law stated that ontogeny recapitulates phylogeny and required that embryos pass through a series of ancestral stages during development, adding novelties only at the end of development. Garstang used examples from gastropod larvae to illustrate that changes could take place at any point in a developmental series. However, gastropods and spiralians were neglected as developmental biologists began to embrace experimental embryology and developmental genetics took off, prompting a shift to more genetically tractable systems. It was not until the latter part of the twentieth century when molecular phylogenies forced a rethinking of bilaterian relationships and molecular developmental techniques became more widespread that interest in these animals was reinvigorated. This has been accelerated by the development of techniques including high-throughput sequencing and genome editing that can be used in a range of species, providing powerful tools for non-model organisms.
As the most speciose group of animals outside of the arthropods, gastropods exhibit a wide range of adult phenotypes, with largely conserved early development, presenting a microcosm of spiralian diversity. The combination of spiral cleavage, varied adult morphology, and a well-represented fossil record has made snails and their relatives a useful group for exploration of the evolutionary links between early development and adult morphology. In this chapter, we take a broad view of gastropod evolutionary developmental biology. We highlight key examples, discuss some of the most intriguing puzzles, and contemplate the future of the field. Due to space limitations, we are unable to cover all aspects of gastropod evolutionary developmental biology; however, we aim to provide a diverse sampling of the lessons offered by snail Evo-Devo, past and present.
What Is a Gastropod?
Spiral Forms Most Beautiful: Spiral Cleavage, Embryonic Organization, and Pattern Formation
The specification of the different cell quadrants depends on the initial cleavage divisions, which can be characterized as equal or unequal. This is of particular significance for the D quadrant, the source of the dorsal organizer that directs the development of adjacent cells, as well as the source of much of the mesoderm in gastropods and other spiralians (i.e., 4d). Initial specification of the D quadrant is either via induction by progeny of the first quartet of micromeres, in the case of equal cleavage (Fig. 3a, b), or by differential apportioning of cytoplasmic determinants, in the case of unequal cleavage (Fig. 3c, d) (Freeman and Lundelius 1992). Unequal cleavage is accompanied by shifting of the cleavage spindles (Fig. 3c), the formation of a polar lobe (Fig. 3d), or some combination of these mechanisms. For example, in Tritia obsoleta, a large polar lobe is allocated to one of the daughter cells during the first two cleavages, producing an early asymmetry and determining the identity of the larger D quadrant blastomere at the four-cell stage (Fig. 3d). Removal of the polar lobe of T. obsoleta before the first cleavage produces a radialized embryo, demonstrating the role of the polar lobe and D quadrant in inducing the dorsoventral axis (Clement 1952). In equally cleaving embryos, such as the equally cleaving limpet Patella vulgata, induction of the D quadrant occurs after the formation of the third quartet of micromeres (van den Biggelaar and Guerrier 1979). Prior to this, all macromeres are capable of becoming the D quadrant, and removal of the first quartet cells produces a radialized embryo (van den Biggelaar and Guerrier 1979; Henry et al. 2017a).
Detailed lineages of the D quadrant have only recently been produced. For example, Lyons et al. (2012) described the lineage of the 4d mesentoblast in Crepidula, providing evidence for the origins of the primordial germ cells and other tissues. Detailed knowledge of this important and highly conserved cell lineage is valuable for understanding the embryological basis of the origins of novel structures within a stereotyped cleavage pattern. Similar detailed lineage studies have revealed differences in the origins of mesoderm and endoderm and identity of the organizer cell and specific cell fates in other gastropods (and other spiralians), for example, in T. obsoleta (Chan and Lambert 2014), demonstrating the flexibility of development within the conserved spiral cleavage pattern.
Understanding cellular origins is only the first step to understanding the molecular processes governing the development of spiralian animals. In gastropods, the embryonic organizer has been identified as the 3D and/or the 4d cell where this has been investigated (Henry et al. 2017a). The Erk-1/2 MAPK pathway has been implicated in the establishment of the organizer in many gastropods. For example, MAPK is activated in the 3D and 4d cells followed by the first quartet micromeres in T. obsoleta (Lambert and Nagy 2001). Disruption of MAPK activity disrupts D quadrant specification and organizer activity, producing radialized embryos (Lambert and Nagy 2001). Activation of MAPK in equally cleaving gastropods is generally restricted to the 3D cell. However, the identity of upstream members of the signaling cascade and the role of MAPK in D quadrant specification, developmental mode, and other species-specific differences remain open questions.
Spiral cleavage was previously described as mosaic or determinate, with each cell having a specific developmental fate. However, cell fates are specified by a combination of cell autonomous and conditional mechanisms. During spiral cleavage, developmentally significant mRNAs localize to specific quartets of micromeres (Lambert and Nagy 2002; Rabinowitz and Lambert 2010). This process has been examined in detail in T. obsoleta. During prophase, mRNAs are located diffusely throughout the cytoplasm or localized to the cell cortex. At interphase, specific mRNAs concentrate in the pericentriolar matrix adjacent to the nucleus localizing to the centrosome. As division proceeds, these mRNAs decouple from the centrosome, shifting to the cytoplasm or cell cortex, and may be transferred asymmetrically to specific daughter cells, resulting in a particular pattern of distribution associated with the birth of various quartets of cells (Lambert and Nagy 2002). The patterns of localization are complex and distinguish each tier of micromeres with their own specific complement of mRNAs. These mRNAs have subsequently been shown to play a role in specifying the fates of these cells. Coupled with the asymmetric signals from the D quadrant organizer, specific cell fates are distinguished both along the animal-vegetal and the dorsoventral axes.
The timing and specification of the D quadrant and changes in the pattern of spiral cleavage may be tied to evolutionary patterns in development of gastropods. Equal cleavage is thought to be the ancestral cleavage type (Fig. 3a, b, Freeman and Lundelius 1992; van den Biggelaar and Haszprunar 1996). Shifts in the timing of D quadrant specification have been used to understand gastropod phylogenetic relationships, and acceleration of the timing of organizer determination (via asymmetric cleavage divisions, Fig. 3c, d) correlates to more derived gastropod groups (Freeman and Lundelius 1992; van den Biggelaar and Haszprunar 1996). This acceleration coincides with the presence of feeding larvae and is thought by many to have been a requirement for the accelerated development of juvenile structures in pre-metamorphic stages. For example, the production of a polar lobe allocating cytoplasmic determinants allows cell fates (including those of the D quadrant) to be specified at an earlier stage in development (Freeman and Lundelius 1992). This in turn may have been important in accelerating development of mesodermally derived structures and possibly the evolution of feeding larvae.
Asymmetries Left and Right
Handedness is largely genetically determined. In the pond snail, Lymnaea stagnalis, genetic crosses and the manipulation of cellular contents demonstrate that dextral coiling is dominant to sinistral coiling (Freeman and Lundelius 1982), and a single maternally expressed locus confers handedness. More recently, the locus of shell coiling in L. stagnalis was identified as the diaphanous-related formin gene, Ldia2 (Davison et al. 2016). Formin is a protein affecting actin nucleation and elongation. Ldia2 is asymmetrically localized at first cleavage in L. stagnalis, demonstrating asymmetry at the expression level well before being morphologically apparent in the embryo (Fig. 4e–g). Pharmacological inhibition disrupts chiral twisting at the third cleavage, supporting the identification of formin as a molecular cause of early asymmetry in L. stagnalis (Davison et al. 2016).
Other well-characterized molecular markers of asymmetry, nodal and pitx, are also expressed asymmetrically. In embryos of Lottia gigantea and Biomphalaria glabrata, left-right asymmetric embryonic expression corresponds to the dextral (L. gigantea) or sinistral (B. glabrata) pattern of cleavage and shell coiling, respectively (Fig. 4a–d, h, i) (Grande and Patel 2009). Asymmetrical nodal expression is established in early cleavage and is upstream of pitx expression (Grande and Patel 2009). Blocking nodal activity by pharmacological means leads to the loss of shell coiling and symmetrical pitx expression in Biomphalaria embryos (Grande and Patel 2009). The relationship between the direction of cleavage and shell coiling has been demonstrated in L. stagnalis. Physically altering the position of the micromeres at third embryonic cleavage, upstream of initial nodal expression in the embryo, changes the location of nodal expression and subsequent adult handedness (Kuroda et al. 2009). Right-handed (dexiotropic) embryos that are mechanically shifted to have left-handed (leiotropic) cleavage grow up to have left-handed (sinistral) shell coiling, while right-shifted embryos of sinistrally cleaving embryos show right-handed (dextral) shell coiling as adults. It remains to be seen how nodal and pitx expression is regulated and what upstream mechanisms, including physical interactions between cells, mediate handedness in cleavage and shell coiling.
Larval Biology and Life History Evolution
Life history and larval development, particularly the origins of larval feeding, have been the subject of interest to evolutionary and developmental biologists alike. Differences in dispersal ability due to the presence or absence of a planktonic larval form may have implications for speciation and extinction events, and the study of larval morphology has formed the basis for various evolutionary scenarios. The ancestral gastropod larva is argued to be represented by a swimming, nonfeeding larva. Although phylogenetic hypotheses are equivocal on the subject (e.g., Zapata et al. 2014), indirect development with larval planktotrophy likely emerged at least once in the apogastropoda and may have evolved multiple times in gastropods. While rare in other groups (e.g., echinoderms), molecular phylogenetic and morphological evidence shows that it is possible for planktotrophy to re-evolve following a loss in a number of gastropod groups. For example, detailed phylogenetic analyses and larval morphology show at least one instance where larval planktotrophy re-emerged after a loss in the calyptraeid family of gastropods (Collin 2004), which includes the slipper snails (Crepidula spp.), the cup and saucer snails (Crucibulum spp.), and the hat shells (Calyptraea spp.). While all calyptraeids brood their offspring, developmental mode is highly diverse, including feeding (planktotrophic) larvae, nonfeeding (lecithotrophic) larvae, direct development with crawl-away juveniles, and direct development via consumption of nutritive embryos (Lesoway et al. 2014). The loss of feeding larvae had been thought to be irreversible owing to the low likelihood of regaining complex feeding and swimming structures. For example, echinoderm larvae show dramatic reduction of larval characters in lecithotrophic forms (Strathmann 1978). However, encapsulated development has the potential to allow for retention of larval structures in the calyptraeids and other groups where feeding larvae are thought to have re-emerged. This includes the retention of features such as the ciliary bands required for larval swimming and feeding (i.e., prototroch, metatroch, and food groove) in direct developers, including species that feed on nutritive embryos during encapsulated development (Collin 2004; Lesoway et al. 2014). Further, gastropod larvae do not undergo the same catastrophic metamorphosis seen in sea urchins and other echinoderms, and metamorphosis is often reduced externally to shedding of the larval velum. In gastropods, many juvenile structures begin their development precociously during larval stages; hence the re-emergence of planktotrophy following a loss appears to be more widespread in the gastropods than thought possible in other groups.
Hard Parts and Body Plan Puzzles
Comparative proteomics and transcriptomic screens have found that shell secretion is highly variable within gastropods (and molluscs more generally). Transcriptomic comparisons of the nacreous secretome of the vetigastropod Haliotis asinina and the owl limpet L. gigantea and three species of the bivalve Pinctada show relatively low levels of conservation of secreted components, suggesting that molluscan shell secretion genes have evolved very rapidly (Jackson et al. 2010). Of the genes that are shared among these molluscs, several are known from other biomineralizing animals, suggesting some level of conservation. Indeed, the level of evolutionary conservation of a “biomineralization GRN” is a tantalizing question for further research.
The gastropod shell is readily fossilized and abundant in the fossil record, leaving a historical record of their evolution. In some cases, the shell also retains evidence of early developmental mode, making gastropods excellent for understanding the evolution of development in deep time. The shell first develops as a pre-metamorphic protoconch (Fig. 6g–i) secreted by the shell field (Fig. 6c, d) during the early embryonic phase and by the mantle tissue at later stages. Post-metamorphosis, the teleoconch (Fig. 6g–i) is secreted by the mantle tissue (Fig. 6f). In snails with a planktonic feeding larva, the protoconch can sometimes be distinguished into earlier embryonic and later larval protoconch by differences in shell sculpturing. The size and shape of these parts of the shell differ with developmental mode (Fig. 6g–i). Indirect-developing snails with a swimming, feeding veliger larva often come from smaller eggs and have a tightly coiled protoconch (Fig. 6h). There may be distinct changes in shell sculpturing between the embryonic and larval protoconch as well as between the protoconch and teleoconch. Direct developers or planktonic nonfeeding larvae (lecithotrophs) frequently have a large, globose protoconch and typically lack protoconchal ornamentation (Fig. 6g). Because remnants of the embryonic and larval shell are retained at the apex of the adult shell, these characters suggest the early development of long dead, fossilized gastropods and have been used to speculate about the evolution of developmental modes. However, these fine structures can often be degraded by weathering, even in living gastropods. Internal casts (steinkern) of fossil gastropod shells, which retain the size and overall shape of the shell, have also been used (more controversially as the external patterning of the shell is not retained) to date the origins of larval planktotrophy in the gastropods to the Ordovician (Nützel et al. 2006). However, evidence for the ancestral mode of development in the gastropods remains equivocal and continues to be a source of debate and disagreement.
Another hallmark of the gastropod body plan is the torted organization of the body (Fig. 1b, c). The dorsal visceropallium of gastropods appears twisted 180° relative to the ventral head and foot, and the anus is deflected anteriorly. The nerve cords are also twisted, although in many species (typically shell-less forms), this arrangement may be absent. Torsion is distinct from shell coiling, and gastropods which have a shell that lacks coiling (i.e., limpet-like, Fig. 5f, l) may still show internal signs of torsion such as crossed nerve cords (Fig. 1b, c). The evolutionary origins of this aspect of the gastropod body plan remain the subject of much debate and speculation. Based on early observations of the process of torsion during the development of basal gastropods, torsion was hypothesized to have originated as a larval adaptation to planktonic life, resulting in the twisted adult body plan (Garstang 1929). Garstang’s torsion hypothesis has proven difficult to test, and no evidence supports the idea that post-torsional veligers are less susceptible to predation than pre-torsional veligers (Pennington and Chia 1985). While appealing as an explanation, Garstang’s hypothesis also conflates the process of developmental torsion during embryogenesis with the evolutionary pattern of a torted adult body plan. Page (2006) reviews the controversy and literature surrounding torsion and notes that this confusion is a source of much of the debate surrounding the origins of the gastropod body plan. Evolutionary scenarios of the gastropod body plan have often relied on a hypothetical ancestral mollusc (the “HAM”) with a posterior mantle cavity and anus (Fig. 1b), with torsion resulting in the modern body plan with an anterior mantle cavity and anus (Fig. 1c). However, this hypothetical ancestor is not based on fossil evidence, and the HAM scenario is tautological (Page 2006). Comparative developmental evidence suggests an alternative hypothesis, with early asymmetry in growth of the mantle producing ontogenetic torsion (Kurita and Wada 2011) and providing a mechanism for the torsion pattern (Page 2006).
The mechanisms of torsion during development are not clear-cut either. Contractile activity of larval retractor muscles has been suggested to be responsible for ontogenetic torsion, based on the position of larval muscle attachments and contractile activity in the early veliger. However, several studies have raised doubt about the ability of the larval retractor muscles to produce torsion during development. Challenges to this hypothesis include a lack of shell stiffness for muscles to work against in embryos undergoing torsion (Hickman and Hadfield 2001), and pharmacological disruption of muscle attachments in embryos undergoing torsion failed to disrupt the process of torsion (Page 2002). Furthermore, pharmacological disruption of nodal signaling in early development blocks both asymmetric growth of the mantle epithelium and torsion, but not the growth of the larval retractor muscles (Kurita and Wada 2011). While torsion is considered by some to be “solved” as an evolutionary question, the relative lack of detailed comparative developmental data and the evidence challenging the traditional torsion hypothesis makes this classical question in Evo-Devo in need of re-evaluation.
Everything Old Is New Again: Future Directions
Many of the organisms that E.G. Conklin and colleagues explored at the beginning of the last century have re-emerged and are thriving as spiralian developmental models. For example, Tritia and Crepidula have been used to produce modern cell lineage analyses, and the use of artificial mRNA constructs and time-lapse imaging and other experimental manipulations has shed light on complex morphogenetic processes in early development, such as gastrulation (e.g., Lyons et al. 2015). Increasing availability of sequencing data and the advent of accessible gene-editing techniques (e.g., CRISPR/Cas9) has meant that many systems traditionally lacking in genomic resources, including the gastropods, are now flourishing. The first demonstration of successful gene editing using CRISPR/Cas9 in a spiralian was done in C. fornicata (Perry and Henry 2015). Recent years have seen the publication of a flurry of transcriptomic databases, and a handful of gastropod genomes (e.g., Lottia, Aplysia, and Biomphalaria) have become available with more on the horizon. The pond snail Lymnaea and more basal gastropods like Patella and Haliotis continue to provide insight into shell secretion and biomineralization. More recently, the direct-developing congener of C. fornicata, C. atrasolea, has been developed as a laboratory-based model of spiralian development (Henry et al. 2017b). Tractability as an experimental system with readily manipulated embryos, year-round embryo production, and direct development make C. atrasolea ideal for developmental study. Comparisons between Crepidula species with different modes of development (e.g., planktotrophy, C. fornicata; direct development, C. atrasolea; direct development with nutritive embryos, C. navicella, Lesoway et al. 2014) will shed light on the mechanisms that have allowed diversity of developmental mode within this group and may elucidate the origins of larval forms more broadly.
The gastropod body plan continues to hold evolutionary mysteries – the contrast of the highly conserved spiral pattern of development with the diversity of adult phenotypes presents an excellent case study for the evolution of novel forms from conserved developmental origins. In addition to the morphological diversity of gastropods, torsion, the evolution of larval forms and transitions among modes of development, variation in shell shape and the mechanisms underlying biomineralization, the segmentation status of early molluscs, and more are all ready for deeper examination under the Evo-Devo paradigm. Revived and new animal models, new techniques, and comparative genetic and morphological approaches provide the tools to address ongoing controversies and long-standing evolutionary questions making the gastropods an exciting nexus for continued lessons in Evo-Devo.
The authors acknowledge the invaluable support of the National Science Foundation (NSF) and were supported by NSF grant IOS-1558061 to JQH (JJH). MPL was supported by a postdoctoral fellowship from the Fonds de recherche du Québec – Nature et technologies (FRQ-NT).
- Davison A, McDowell Gary S, Holden Jennifer M, Johnson Harriet F, Koutsovoulos Georgios D, Liu MM, Hulpiau P, Van Roy F, Wade Christopher M, Banerjee R, Yang F, Chiba S, Davey John W, Jackson Daniel J, Levin M, Blaxter Mark L (2016) Formin is associated with left–right asymmetry in the pond snail and the frog. Curr Biol 26(5):654–660. https://doi.org/10.1016/j.cub.2015.12.071CrossRefPubMedPubMedCentralGoogle Scholar
- Freeman G, Lundelius JW (1992) Evolutionary implications of the mode of D quadrant specification in coelomates with spiral cleavage. J Evol Biol 5(2):205–247. https://doi.org/10.1046/j.1420-9101.1992.5020205.xCrossRefGoogle Scholar
- Garstang W (1922) The theory of recapitulation: A critical re-statement of the Biogenetic Law. Proc Linn Soc Lond Zool 35:81–101Google Scholar
- Garstang W (1929) The origin and evolution of larval forms. Rep Br Assoc Adv Sci Sect D 96:77–98Google Scholar
- Kniprath E (1981) Ontogeny of the molluscan shell field: a review. Zool Scr 10(1):61–79. https://doi.org/10.1111/j.1463-6409.1981.tb00485.xCrossRefGoogle Scholar
- Kurita Y, Wada H (2011) Evidence that gastropod torsion is driven by asymmetric cell proliferation activated by TGF-β signalling. Biol Lett. https://doi.org/10.1098/rsbl.2011.0263
- Nederbragt AJ, van Loon AE, Dictus W (2002) Expression of Patella vulgata orthologs of engrailed and dpp-BMP2/4 in adjacent domains during molluscan shell development suggests a conserved compartment boundary mechanism. Dev Biol 246(2):341–355. https://doi.org/10.1006/dbio.2002.0653CrossRefPubMedGoogle Scholar
- Shimizu K, Iijima M, Setiamarga DH, Sarashina I, Kudoh T, Asami T, Gittenberger E, Endo K (2013) Left–right asymmetric expression of dpp in the mantle of gastropods correlates with asymmetric shell coiling. EvoDevo 4(1):15. https://doi.org/10.1186/2041-9139-4-15CrossRefPubMedPubMedCentralGoogle Scholar