Portal de Periódicos da CAPES

The emergence of form: The shape of things to come

2003; Wiley; Volume: 228; Issue: 2 Linguagem: Inglês

10.1002/dvdy.10361

1097-0177

Brian K. Hall,

Language and cultural evolution

Origination of Organismal Form: Beyond the Gene in Developmental and Evolutionary Biology by Gerd B. Müller and Stuart A. Newman A Bradford Book, The MIT Press, Cambridge, MA, 2003, 332 p, $45, £30.95 Brian K. Hall*, * Department of Biology, Dalhousie University, Halifax, NS, Canada. It is hard to imagine a more appropriate journal to review Origination of Organismal Form (henceforth, Origination) than Developmental Dynamics, a journal whose aim is to focus “on morphogenesis: the study of the emergence of form during animal development.” The second volume in the Vienna Series in Theoretical Biology,1 this set of 17 papers had its origin in a 1999 Workshop, organized by the Konrad Lorenz Institute for Evolution and Cognitive Research, which is housed in the late Konrad Lorenz's eighteenth century manor house in Altenberg, on the outskirts of Vienna. The editors' (Gerd Müller and Stuart Newman) aim was to seek “the specific generative mechanisms that underlie the origin and innovation of phenotypic characters,” (p. 3), an endeavor they see as “probably best embodied in the term origination…” (p. 3). My initial reaction was surprise at the term origination.2 Most developmental biologists and morphologists speak of the development, origin, generation, or emergence of form, evolutionary biologists of the evolution of form. I appreciate, however, the important point the editors are making with the term origination. The “origin of form” refers to the starting point, the place or position from which a structure arises. So, we speak of the origin of the muscles of the tetrapod limbs from (or in) paraxial mesoderm, or the origin of the dentary bone in the neural crest. Müller and Newman use origination to emphasize the processes involved in generating form, the transitive verb “originate” denoting causing to begin or initiating. Although the term may not take hold, the underlying premise should, and in many ways has taken hold in the flourishing field of evolutionary developmental biology or evo-devo.3 The premise is twofold: (1) that the emphasis on understanding sources of variation in much of the evolutionary biology of the past 100 or more years has (2) cast aside attempts to understand how phenotypes arise. Charles Darwin (1809–1882) emphasized variation and heredity as two separate but interconnected evolutionary processes. The study of variation was initiated by Darwin (1859) with On the Origin of Species, further enhanced with his monographic The Variation of Animals and Plants under Domestication (Darwin, 1868) and by William Bateson (1894) in Materials for the Study of Variation. Codified by population genetics in the early 20th century, to analyze variation was to study evolution, while evolution was the analysis of genetic variation in populations. The book under review deals with how form and structures arise, not how they vary. As if to reinforce this point, variation does not appear in the index. Analysis of how shape is generated has existed as a separate field (the science of morphology?) since the first third of the 19th century and the works of Lorenz Oken (Okenfuss, 1779–1851) and Wilhelm His the elder (1831–1904). Karl Ernst von Baer (1792–1876) linked form to phylogeny during the same period and to such a degree that Darwin regarded morphology as the “very soul of natural history,” and based his theory of evolution as descent with modification to a large extent on form, both embryonic and adult. Morphology, however, was omitted from the modern synthesis, which gelled in the 1940s (see the papers in Mayr and Provine, 1980 and Bonner, 1982), even though several of the founders were students of form; the paleontologist G.G. Simpson (1902–1984) spent his life in the analysis of form in vertebrate fossils, whereas the botanist G.L. Stebbins (1906–2000) studied the form of plants at more microscopic levels.4 Understanding how structures arise/form was left to embryologists, morphologists, physiologists, functional biologists, and a few engineers, theoretical biologists, and information theorists. De facto camps arose, not around the teachings of grand masters or alpha-males but around the approach deemed most suitable for the study of form. Those mathematically inclined implicitly or explicitly followed D'Arcy Thompson, for whom “the mathematical definition of form has a quality of precision quite lacking in our earlier stage of mere description” (1917, p. 719). Many but certainly not all paleontologists, and certainly not constructional morphologists (see Hall, 2002), followed the British paleontologist D.M.S. Watson in the view that “morphology is a form of logical thought remarkable in that it is not mathematical; indeed, its essential elements, being as they are, qualities, are not susceptible to mathematical expression” (1951, p. 3). To the present day, analysis of change of form has been more amenable to mathematical analysis than has the origin of form; contrast computer-simulation analyses of morphogenesis of snail shells (Raup, 1966; Stone, 1996, 1997) with the molecular genetics approach used to understand how coiled shells arise (Freeman and Lundelius, 1982); see Collazo and Fraser (1996) for the latter approach. I do not include in this generalization, studies such as the morphogen study published by Alan Turing (1952), which while mathematical, led to a nonmathematical approach in a search for morphogens (e.g., Green, 1990; Gurdon et al., 1995) and to mathematically driven alternatives to the Turing model (e.g., Harris et al., 1984). Computer simulation of the pattern of condensation of tetrapod limb elements, while faithfully reproducing the pattern of the skeletal elements (Ede and Law, 1969; Wilby and Ede, 1975), has not, to my knowledge, been found applicable to the origination of the condensations themselves. To date, it appears that we cannot compute origination, and so in this sense, cannot compute the embryo. Threshold models (including response to morphogen gradient; Gurdon et al., 1995), several of which appear in Origination, appear to hold much more promise in the quest to understand the emergence of form, which, essentially, is to generate a genotype-to-phenotype map for each feature of the phenotype (see the entries in Hall and Olson, 2003). It is perhaps too early to assess the promise of mathematical approaches to form such as that used by Stadler et al. (2001), in which the folding of RNA sequences represents the genotype and secondary structures the phenotype (form). Despite the more recent concentration by evolutionary biologists on variation, form has been an important topic for centuries. Students of form from D'Arcy Thompson (1917) onward waxed lyrical over form and its analysis: “Form is both deeply material and highly spiritual. It cannot exist without a material support; it cannot be properly expressed without invoking some supra-material principle. Form poses a problem, which appeals to the utmost resources of our intelligence, and it affords the means which charm our sensibility and even entice us to the verge of frenzy. Form is never trivial or indifferent; it is the magic of the world” (Dalcq, 1968, p. 91). Gould and Vrba (1982) recapitulated form as a science in their study on “exaptation—a missing term in the science of form.” In an epigram that adorns a poster in many developmental biology laboratories around the world, Lewis Wolpert argued that “it is not birth, marriage, or death, but gastrulation, which is truly the most important time in your life” (Wolpert, 1991, p. 12). Gastrulation is when the form of the organism first appears, a consequence of the cessation of maternal control over cleavage and the assertion of zygotic genomic control over morphogenesis, differentiation, and growth, i.e., over origination. Despite this interest, the origin of form was not seen as a question with an evolutionary perspective, and until recently, not one with an evo-devo perspective. E.S. Russell was an important exception. In his history of embryology, Russell is concerned with “what is the essence of life—organization or activity?” (1916, p. v), where organization is origination and activity is variation, physiology, and function. The titles of the books by Bonner (1974) and Raff (1996) illustrate the modern swing: development as the biology of form (Bonner), and genes, development, and the evolution of form (Raff). Origination treats one pole of the compass of evo-devo. The “blurb” on the back cover places the topics of the book within the framework of evo-devo, [which is] “a new research agenda that concerns the interaction of development and evolution in the generation of biological form.” Thus, the book treats only one aspect of evo-devo, a field that embraces more than form; see Hall (1999), Wilkins (2002), and the entries in Hall and Olson (2003). Origination is avowedly epigenetic and process-oriented. Again, from the back cover: “by placing epigenetic processes, rather than gene sequence and gene expression changes, at the center of morphological origination, this book points the way to a more comprehensive theory of evolution.” The book has ambitious aims, aims that are important for any of us studying morphogenesis. Origination consists of 17 chapters in five sections. The span of topics is wide, illustrating the range of disciplines relevant for, and perhaps even required for, an understanding of how features of the phenotype arise. Paleontology, paleobiology, systematics, molecular biology, and molecular genetics, sit comfortably beside cell and structural biology, morphogenesis, genetic assimilation and theoretical biology. Even if one's interest is purely developmental, developmental biology is informed by knowledge of paleobiological studies on the evolutionary origins of animals and their Baupläne, as much as it is facilitated by investigations at genetic, cellular, and morphogenetic levels. The single chapter in the introduction sets the book in a broader framework by surveying the history of the study of variation and neglect of the origin of form. As discussed above with respect to the title of the book, words and terminology are important. Variation is equated with diversification, origination with innovation. I have wondered for some time whether an emphasis on innovation or novelty, while it emphasizes the need to search for and understand how new features arise—turtle shells, dinosaur/bird feathers, tetrapod limbs—might not lead us astray. How do we determine which features are novelties? The tetrapod limb, which arose by transformation of a fish fin? The cetacean flipper, which arose from transformation of a forelimb? The fluke, which arose as a new locomotory organ from the tail region, and/or the hind limbs, which disappeared from adult whales but remain as rudiments in embryos? Different approaches to novelties are evident in the book. In Chapter 4, Müller uses Müller and Wagner (1991, p. 243) to define a morphological novelty as “a structure that is neither homologous to any structure in the ancestral species, nor homonomous to any other structure of the same organism” (p. 60). In Chapter 15, Wagner and Chiu use the same source to define a morphological novelty “as a character derived for a clade (i.e., autapomorphic), hence, not present in the ancestor of a more inclusive clade” (p. 274). In either case, and as discussed by Müller in Chapter 1, distinguishing homology is critical; see also Hall (2003a). We know that change associated with the origin (or loss) of individual features of the phenotype is gradual; e.g., the development/evolution, in whales, of flippers from forelimbs and the developmental/evolutionary loss of the hind limbs (Bejder and Hall, 2002). We know that features develop/evolve with considerable independence from other features—ears, flippers, hind limbs, and flukes in whales. The whale fluke is not present in cetacean ancestors. It is a novel appendage. Flippers were present in whale ancestors, but as forelimbs. Flippers are homologues of forelimbs as anterior paired appendages (and so are not novel) but are novel locomotory organs. Tetrapod limbs are homologues of fish fins as appendages. The digits (and perhaps the radials) of limbs are not represented in fish fins and so are novel skeletal elements. More proximal elements of limbs are represented in fish fins and so are not novelties. At the level of developmental processes, the way digits form is not fundamentally different from how the humerus, radius, or ulna form. Where does innovation stop and tinkering begin? When does variation slide into novelty, or are the two based on such different developmental mechanisms that we cannot ask such a question. Wound healing and regeneration raise the same issues and questions, as to a degree do differentiation and morphogenesis. Such dichotomies/continua go to the heart of how we should view morphological change. Conceptually, distinctive origination processes or tinkering with existing processes could take a feature from an ancestral to a descendant state. There is an interesting parallel here with cell differentiation, in which initiation (origination) and maintenance of the differentiated state are often controlled by different processes, operating at different levels; chondrogenesis is an excellent example. Indeed even the different phases of chondrogenesis—epithelial–mesenchymal interaction, condensation, initiation, termination, and maintenance of differentiation are all subject to separate control (Hall and Miyake, 2000). A hierarchical organization of developmental process is neither unique to innovation nor to novelty. The three chapters in Part II deal with what are identified as problems of morphologic evolution, viz., the Cambrian “explosion” (Conway Morris), convergence and homoplasy (Willmer), and homology (Müller)—big, overarching, and problematic topics that are of considerable importance for developmental biology/biologists. Conway Morris focuses his considerable synthetic skills on the Ediacaran fauna and on larvae as potential candidates in the origin and diversification of metazoans. He cautions against denoting any extant phylum as “primitive” or as a surrogate for the ancestral metazoan. (His reference to the Lower Cambrian agnathan chordate Myllokunmingia fengio as a fish is one of the few missed opportunities for copy-editing in the book.) Conway Morris emphasizes, as he has in past publications (a 1994 article being, perhaps, most relevant for developmental biologists), the need for an interdisciplinary approach, coupling molecular approaches on extant forms, with fossils (both hard and soft parts), and analyses of paleoecology and paleoclimatology. He reinforces the caution that should be exercised in attributing all control of the origination of form to “those all-powerful genes,” and draws our attention to the important study by Graham Budd on whether patterning (Hox) genes drive the origin of morphology or vice versa (Budd, 1999). Again, is it tinkering or innovation? Willmer uses larvae, segmentation, and appendages to determine how many times these stages/features have evolved; just how much convergence and homoplasy was there in the evolution of metazoans? In the late 1880s, presence or absence of segmentation was an important criterion for identifying chordate ancestors. In a very thorough analysis, Arthur Willey effectively silenced much of this speculation and anticipated Willmer in regarding segmentation not as primitive but rather as reflecting a tendency during evolution to duplicate body parts or regions, segmentation having arisen “independently over and over again in different groups of animals” (Willey, 1894, p. 246). Several investigators in anthropology, developmental biology, evolution, and systematics, are beginning to realize that convergence and homoplasy—which both Willmer and I regard as synonymous, convergence being the only category of homoplasy—are far more abundant than ever imagined; see, for example, the convergent approaches of Lieberman et al. (1996), Meyer (1999), Wake (1991), Parra-Olea and Wake (2001), Hall (2003a), and the chapters in Lockwood and Fleagle (2003). In his chapter, Müller approaches homology as “process” with his proposal for “organizational homology” as a form of biological homology, i.e., homology as the conservative nature of developmental processes as argued by Thomson (1988; and see analyses by Hall, 1989, 1995, and Erwin, 1990) and by Wagner (1989). Unlike some, Müller does not equate homology with synapomorphy (see the definitions of novelty above), his approach to homology being developmental and based on seven premises (p. 64), three of which, he argues, are novel: (1) “constancy of constructional organization despite changes in underlying genetic mechanisms,” a premise whose origins go back to Spemann (1915) and de Beer (1971), both of whom argued that homologues need not share similar genetic mechanisms (see Hall, 1995 for a review), a position accepted by those who take a hierarchical approach to homology; (2) “Homologues act as organizers of the phenotype;” and (3) “as organizers of the evolving molecular and genetic circuitry.” Much as I am in sympathy with this approach to homology, I find the latter two premises difficult, reflecting as they do an “organization-down” rather than a “gene-up” approach. There is no doubt that properties emerge as development proceeds and that those properties cannot be predicted from the properties of the lower levels; the chapters in Part IV attest to this. Epithelial–mesenchymal interactions are a classic example. Nothing about the organization or juxtaposition of epithelium and mesenchyme would lead us to suspect that subsequent interaction between them is one of the most important mechanisms for the origination of form. Such processes—organizational homologues for Müller—do indeed organize the phenotype through initiation and regulation of differentiation and morphogenesis. But there is no feedback to the genotype other than through mutation and certainly no pangenetic mechanisms that would allow any feedback to be heritable. Organizational homologues may well organize the phenotype during ontogeny (premise 2). Seeing them as organizers of the evolving molecular and genetic circuitry is a stretch. Developmental processes, like form, emerge during ontogeny with the emergence of higher-order levels of interactions—cell to cell, cell to matrix, tissue to tissue—as illustrated amply in Part III. The four chapters in Part III, which deal with genes and form, might at first seem to contradict the chapters in Part II; Roy Britten's chapter is entitled “Only Details Determine.” However, Britten is concerned with how the actions of genes are integrated to generate form, expanding on a model he first published in 1998. His nine propositions (pp. 76–79) set the stage for the other three chapters in Part IV, and, more importantly, set the basis for a hierarchical approach to the origination of form, integrating macromolecular and cellular levels. Read his chapter for the propositions. I note only that proposition eight deals with precursor groups of cells, which others have identified as fundamental units of morphology, variously termed modules, cell condensations, or the cell community effect; see Gurdon (1988), Gurdon et al. (1993), and Hall (2003b, c) for overviews. Chapter 7, and the chapters in Part IV provide the cellular contexts in which these interactions occur, Chapter 7 concentrating on the extracellular matrix, Chapter 9 on cell adhesive properties, and Chapters 10 and 11 on gradients, diffusion, and oscillators. Larsen (Chapter 8) also sees the cell as “the pivot between the gene and the development of form” (p. 120), and lists six important factors important for the evolution of form (p. 126). In his chapter, “The Reactive Genome,” Scott Gilbert takes an explicitly ecological approach to the analysis of developmental change, placing such change into physiological life history, and ecological and environment contexts. The next volume in the Vienna Series in Theoretical Biology (see footnote 2) is devoted to this approach, with additional emphasis on endocrine control. In addition to contextualizing cells in development, the chapters in Part IV emphasize the physical determinants of morphogenesis, including a broad historical overview of studies and approaches to cell adhesion (Malcolm Steinberg),5 pattern formation in insects by means of gradients and diffusion (Fred Nijhout), segmentation of somites in response to a biochemical oscillator (Olivier Pourquié), and chapters by Kunihiko Kaneko and Stuart Newman that apply approaches from physics to morphogenesis. For Steinberg, the levels of causality of differential cell adhesion are unidirectional: motile cells → motile adhesive cells → most stable equilibrium structure. Nijhoutdiscusses feedback based on local activation and lateral inhibition founded in reaction-diffusion and genetic circuits, emphasizing the relevance of such a systems approach to “the evolution of robustness in development and the role of heterochrony in evolution” (p. 179), a theme echoed by Pourquié in the context of resetting and pausing morphogenetic clocks by means of biochemical oscillations of the gene c-hairy1 in presomitic mesoderm. Kaneko and Newman take us into systems dynamics, which Kaneko (p. 217) links nicely to cell sociology by means of the community effects described by Gurdon (1988) and Gurdon et al. (1993). Kaneka provides a set of rules for cells in transit between stages of differentiation (p. 204) and relates these to the evolutionary emergence of the life cycle and multicellularity. In his chapter, Newman further reinforces the importance of differentiation adhesion and the formation of cell compartments as he analyzes the emergence of higher-level organization during development. Part V takes up evolvability, which is “the potential of a lineage to generate heritable variants that respond productively to external challenges, including those properties of developmental systems that constrain and bias further possibilities for morphological change” (p. 241). In Chapter 14, Nanjundiah evaluates evolvability in the context of the phenotypic plasticity provided by genetic assimilation. The important fundamental underlying concept is that multiple phenotypes can arise from a single genotype (a theme already introduced in Chapter 6). Most often seen in examples of phenotypic plasticity—cannibalistic morphs of amphibian tadpoles, neck teeth in Daphnia, induction of plate morphs in sticklebacks are three examples (Hall, 1999) —new phenotypes can arise in organisms that do not show polyphenism, cyclomorphosis, or seasonal polymorphism, provided that there is unexpressed genetic variation, that an environmental signal elicits a phenotypic change, and that the organism is then subject to selection. In their chapter, Wagner and Chiu use the tetrapod limb as a platform from which to launch an analysis of the interplay of genetic and epigenetic factors in the origination of form on both developmental and evolutionary time scales. Striedter takes a similar conceptual approach to vertebrate brain evolution. Both these chapters rely importantly on identification of homology, and both deal effectively with this topic/problem. The final chapter, by Diego Rasskin-Gutman on boundary constraints, gave me the greatest challenge, in part, because I was more familiar with the approaches in the rest of the book, but in part, because it struck me as wrong, or at least the reverse of how I view the topic. The approach Rasskin-Gutman advocates is not new; it is the basis of any morphometric analysis: identify landmarks (Rasskin-Gutman calls them boundary modules) on the structure of interest, connect adjacent points, and plot/analyze the connections. Of course, the mathematical analysis of the data generated can be enormously complex. Boundaries and connections are well known. The proposal of boundaries as modules is novel, but is it innovative? Do the points reflect any more than where two modules meet? This is the conventional view; use homologous landmarks on homologous elements and you can quantify changes in shape/growth of those homologous elements. In this scenario—practiced by morphometricians, constructional morphologists, and clinicians (oral surgeons being one obvious example)—the landmarks and the boundaries they connect are tools or reference points. They are epiphenomena, spandrels; unavoidable zones of contact between units of construction (Gould and Lewontin, 1979; Hall, 2003d). In this chapter, however, boundaries are treated as playing a much more primary role, addressed “within the framework of the morphospace of connections” (p. 321). Earlier in the chapter, boundaries are (I think) properly placed as guides to identify homologues, essentially Etienne Geoffroy St-Hilaire's (1772–1844) principle of connections. Woodger's “distal to, proximal to, articulated with” (p. 309) seems to me to be an orthodox use of Geoffroy's principle. Woodger (1945) used boundaries as starting points to locate the morphological units, not as morphological entities themselves.6 The interfaces between (1) presumptive ectoderm and endoderm in blastulae, (2) notochord and ectoderm in gastrulae, and (3) epithelium and mesenchyme in post–neurula-stage embryos represent boundaries where inductive interactions take place (respectively, induction of mesoderm, neural ectoderm, and a myriad of cell types). In this context, the boundary itself carries no developmental information. It is neither module nor functional unit. It is the interface between two interacting elements. Boundaries can be important, however. A considerable literature is accumulating on how boundaries to cell condensations are set and how sutures are established and maintained between (uniting) skull bones; see Opperman (2000), Hall and Miyake (2000), and Cole et al. (2003) for overviews. Sutures, like joints, are independent morphogenetic and functional units. But sutures or joints are much more than boundaries that delimit them. I recommend Origination as a valuable summary and analysis of important approaches to how form emerges and is maintained. The quality of the production is excellent and the price modest (which should send a message to other publishers). Many of the approaches discussed could also be used to analyze, perhaps even understand, how form varies. A closer linking of origination and variation may lie within these covers, despite the orientation of the book to the former to the exclusion of the latter.