sign in sign up
<<
Home , Body plan, Evolution of fish, Hox gene, Human Genome Project, Onthophagus taurus

PERSPECTIVES

Scientist 1 May (2000). Box 5 | European Parliament directive on patenting 12. Eisenberg R. S. Intellectual property issues in . Trends Biotechnol. 14, 302–307 (1996). Directive 98/44/EC of the European Parliament and of the Council of the 6 July 1998 on the legal 13. US Code, title 35, § 102(a). 14. US Code, title 35, § 103. protection of biotechnological inventions, Official Journal L 213, 30/07/1998 p. 0013–0021 Article 5: 15. Parchomovsky, G. Publish or perish. Michigan Law Rev. • The body, at the various stages of its formation and development, and the simple 98, 926–952 (2000). 16. Lichtman, D., Baker, S. & Kraus, K. Strategic disclosure in discovery of one of its elements, including the sequence or partial sequence of a , cannot the patent system. Vanderbilt Law Rev. (in the press). constitute patentable inventions. 17. Bishop, J. E. Plan may blow lid off secret gene research. Wall Street Journal 28 September (1994). • An element isolated from the or otherwise produced by means of a technical process, 18. In re Bell, 991 Federal Reporter, 2d series 781 (Court of including the sequence or partial sequence of a gene, may constitute a patentable invention, even if Appeals for the Federal Circuit, 1993). 19. In re Deuel, 51 Federal Reporter, 3d series 1552 (Court of the structure of that element is identical to that of a natural element. Appeals for the Federal Circuit, 1995). 20. Marshall, E. Drug firms to create public of genetic • The industrial application of a sequence or a partial sequence of a gene must be disclosed . 284, 406–407 (1999). in the patent application. 21. US Code, title 35, § 157(c). 22. US Code, title 35 § 102(e). 23. Shapiro, C. & Varian, H. R. Information Rules: A Strategic concern may motivate some institutions to ests might do more to enlighten public policy Guide to the Network Economy (Harvard Business School Press, Cambridge, Massachusetts, 1998). defer publication in precisely the circum- debates about the importance of the public 24. Venter, J. C. Clinton and Blair shouldn’t destroy our stances that it motivates other institutions to domain in genomics research than appeals to research. Wall Street Journal 21 March (2000). 25. Marshall, E. Talks of public–private deal end in acrimony. make prompt disclosure. The difference ethical imperatives. Science 287, 1723–1725 (2000). depends on whether they believe that pre- 26. Bentley, D. R. Genomic sequence data should be released Rebecca S. Eisenberg is the Robert & Barbara immediately and freely in the public domain. Science 274, empting future patents is good or bad. Apart Luciano Professor of Law at the University of 533–534 (1996). 27. Adams, M. D. & Venter, C. J. Should non-peer-reviewed from concern about preserving their own Michigan Law School, Ann Arbor, Michigan 48109, raw DNA sequence data be forced on the scientific patent rights, public research sponsors and USA. e-mail: [email protected] community? Science 274, 534–536 (1996). publicly funded research performers may 1. Venter, J. C. et al. Shotgun of the human Acknowledgements . Science 280, 1540–1542 (1998). This research has been supported by a grant from the United worry that premature public disclosure could 2. King R. T. Jr Code green: Gene quest will bring glory to States Department of Energy. prevent them from complying with their some; Incyte will stick with cash. The Wall Street Journal 10 February (2000). Links mandate under the Bayh–Dole Act to pro- 3. Healy, B. Special report on gene patenting. N. Engl. J. Med. 327, 664–667 (1992). mote technology transfer and product devel- 4. Eisenberg, R. S. Public research and private development. COMPANIES Celera | Monsanto | Merck | Incyte opment by patenting research results. Indeed, Virginia Law Rev. 82, 1663–1727 (1996). | 5. Waterston, R. & Sulston, J. E. The Human : this concern was cited by former NIH director Reaching the finish line. Science 287, 53–54 (1998). FURTHER INFORMATION Human Genome in support of the decision to 6. Marshall, E. Claim and counterclaim on the human Project | Joint statement by and genome. Science 288, 242–243 (2000). | The SNP Consortium | The file patent applications on the first ESTs iden- 7. Nowak, R. The gold bug: Helicobacter pylori; claimed to tified by when he was at NIH25. be the first free-living organism genome fully sequenced. Bermuda rules | National Human Genome Science 267, 173–174 (1995). Research Institute policy on patenting of In fact, it does not seem that publication of 8. Wade, N. 10 Drug makers join in drive to find diseases’ human genomic sequence | Interim utility raw genomic DNA sequence will prevent the genetic roots. The New York Times 15 April (1999). 9. European Patent Convention, Article 54. guidelines and written description guidelines issuance of patents on that are subse- 10. US Code, title 35, § 102(b). 11. Palevitz, B. A. Rice genome gets a boost: private for Patent Examiners | European Parliament quently found to lie within that sequence sequencing effort yields rough draft for the public. The directive on patenting under law. The situation in Europe is less certain and awaits clarification of national laws in response to a 1998 directive of the European Parliament on the legal protec- OPINION tion of biotechnological inventions (BOX 5). Although the patent system has not yet resolved many of the legal issues that will determine what portions of the human Evo-devo: the genome may be patented, for the time being there seems to be little threat that disclosure of of a new discipline the human genome in the public domain will leave future researchers who identify and char- Rudolf A. Raff acterize genes with nothing left to patent. The history of documented in the of development, and how the process Conclusion record shows that the evolution of of development itself biases or constrains Complex and interrelated strategies for complex organisms such as and evolution. A revolutionary synthesis of endowing the public domain are at work in the has involved marked changes in developmental and evolution is field of genomics. These strategies arise out of , and the appearance of new in progress. the varied plans of different institutions for features. However, evolutionary change extracting value out of genomic information, occurs not by the direct transformation of Developmental and complicated by the interplay of the public adult ancestors into adult descendants are two disciplines that explore morpholog- domain with the patent system. Public disclo- but rather when developmental processes ical change in organisms over time. sure of genomic information advances some produce the features of each generation However, the processes involved are differ- interests while harming others, with no simple in an evolving lineage. Therefore, ent. Development is genetically pro- distinction between the interests of public and evolution cannot be understood grammed and cyclical. Evolution is non- private institutions. Understanding these inter- without understanding the evolution programmed and contingent. Although a

74 | OCTOBER 2000 | VOLUME 1 www..com/reviews/genetics PERSPECTIVES link between the two processes was recog- mental biologists. But palaeontology pro- phylogenetic distance, making genes identi- nized in the late nineteenth century, an vides insights available in no other way. For fied by developmental mutations most use- effective connection of evolutionary and example, the discovery that the earliest (fos- ful in comparisons of related taxa. awaited the appear- sil) had feet with eight toes rather ance of developmental data that contained a than five2 was a complete surprise, and was The contribution of strong and marked evolutionary signal. This important in providing us with a new view Evolutionary biology is comparative, and happened in the 1980s, when the growth of of what ancestral limbs were actually like, requires tracking events over long time developmental established a link and for giving us clues as to how devel- frames, and across phylogeny. Although phy- between genes and development. As devel- opment evolved. logenetic relationships have not been regard- opmental regulatory genes were cloned and Finally, evolutionary biologists are faced ed as important for the study of develop- sequenced — notably those of the with understanding how small genotypic mental mechanisms, they become crucial , which are important in specification once we begin to consider the evolution of of the identity of segments — it was developmental processes3. New analytical realized that the same regulatory genes were “Development is methods provided by and the shared by animals with different body plans avalanche of gene sequence data have revo- (for example, and ). More genetically programmed lutionized phylogeny. importantly, shared regulatory genes have and cyclical. Evolution Phylogeny imparts three important kinds conserved roles in development, which of information. First, we can determine the some have taken to indicate homologies in is non-programmed direction in which developmental features the development of body architecture and contingent.” are evolving. Second, knowing the diver- among different body plans1. gence times of branches in a tree allows evo- Developmental biology has once again lution rates to be inferred. (There is, at pre- become relevant to understanding both modificationsEPIGENETICS are translated into phenotypic sent, controversy about using extrapolations evolutionary mechanisms and the patterns changes during evolution, and how micro- of rates of gene evolution to determine of evolutionary history that are revealed by evolutionary changes contribute to the important divergences that pre-date visible palaeontology and PHYLOGENETIC studies. MACRO-EVOLUTIONARY events on the timescale fossil evidence; the divergence among ani- observed in the fossil record. Their interests mal phyla is such a case7.) Third, phyloge- Cardinal issues also converge on those of evolutionary netic trees allow homologies to be inferred What constitutes the fundamental problems developmental biologists in asking whether or, conversely, show that apparently homolo- for a science of evolutionary developmental developmental processes themselves bias the gous features are not so. The consequences biology (evo-devo) depends on whether the possible directions of evolution by con- can be profound, as seen, for example, in scientist is a developmental biologist, a straining the relationship between allelic and studies of the evolution of and tetra- palaeontologist or an evolutionary biologist. phenotypic variation. Any limitation pod limbs. Modern fish and tetrapods build Some of the main issues (and controversies) imposed by developmental programmes on their fins or limbs using different parts of the are summarized in BOX 1. Developmental the would affect the kinds of shared ancestral . So to avoid mistaken genetics now dominates a wide swathe of morphological variation that are possible, comparisons of gene expression pattern in biology, and powerful genetic and molecu- and its response to selection3. Leroi4 has non-homologous features8, it is important lar tools have made it possible to define the argued strongly that micro-evolution and to understand the evolutionary relationship machinery of development in terms of gene macro-evolution result from the same between structures that are being compared action and the operation of regulatory processes. Orr5 showed that mutations of in different organisms. Furthermore, phylo- genes. These studies revealed that regulatory both large and small effect can be fixed (see genetics shows us that, to understand better genes are conserved across phyla, which glossary) in evolution. Haag and True6 note the variation in developmental mechanisms, provides an impetus to think about the evo- that genes identified by mutations which and to map the origins of novel features, we lutionary dimension of development. The cause developmental can, in must widen the sample of organisms on experimental tools have led to an under- some cases, have similar effects during evo- which our developmental models are based. standing of the development of a few heavi- lution. However, this correlation drops with This has been especially noticeable in the ly studied , and allowed us to com- pare developmental features among a range Box 1 | Current issues and controversies in evo-devo of species. For developmental biologists, the principal and inter-related problems are • How do developmental constraints bias the direction of evolution? how development has evolved, and how • How do micro-evolutionary processes relate to macro-evolutionary differences? developmental evolution has resulted in • Do genes identified by mutations that affect development within a species correspond to genes changes in particular structures or features that produce differences between species? of body organization. • What are the roles of modules in development and evolution? Palaeontologists would seem to be • How should we make an appropriate phylogenetic sampling of organisms for evo-devo studies? unlikely partners in any enterprise with developmental biologists. Palaeontologists • Can gene expression patterns be used to establish homologies between developmental features focus on the appearance of novel features of distantly related organisms? and new body plans during evolutionary • Why is there a conflict between molecular clocks and the fossil record in timing the history — a view that constitutes an overlap metazoan radiation? of interests, if not of methods, with develop- • Were Pre- metazoan ancestors similar to larvae or to miniature adults?

NATURE REVIEWS | GENETICS VOLUME 1 | OCTOBER 2000| 75 PERSPECTIVES

study of the insect head9. The head of There is an unexpected theme to the melanogaster, the most-studied “… have a dorsal developmental regulatory systems underly- insect, is highly specialized but its develop- ing such organs as the heart18,eyes19 and ment is not typical of head development in , appendages20 of insects and vertebrates, insects. So the evolution of insect head a and paired indicating that many phyla may share development can only be understood by homologous precursors to these organs. investigating other groups, using molecu- muscle groups, which However, it is important to be sceptical lar–genetic tools originally devised for the are present from trout about apparent homologies, however seduc- study of Drosophila. to tyrannosaur.” tive. Although many developmental regula- tory genes are conserved across phyla, con- Developmental regulatory genes served genes and gene pathways can be and The richest source of data, at present, comes are co-opted to new functions. For example, from empirical evolutionary studies of the of the body axis. However, evolu- only about 16 basic eukaryotic signal trans- developmental regulation of , tion of Hox gene regulation in vertebrates has duction pathways21 must control the devel- of individual adult body features and of been different from insects. The expression of opment of about 35 phyla, each with a early development. individual Hox genes in insects is linked to unique body plan. Among closely related Studies on the evolution of development segment number, although downstream taxa, such as insects, the same developmental have revolved around the astonishing find- responses in individual segments, leading to regulatory genes probably control homolo- ing that principal regulatory genes are con- distinct segment identities, differ among gous features. However, as phylogenetic dis- served across phyla. Genes of the HOX CLUSTER taxa14. Therefore, although the third thoracic tances increase, the probability of co-option are integrated into animal axial differentia- segment in both taxa expresses the same Hox to non-homologous roles grows, and inter- tion, and are even present in CNIDARIANS,such gene code, a second pair of is produced pretations become more controversial. This as corals10. Detailed examination of expres- in butterflies, compared with HALTERES in flies. is potentially most frustrating precisely sion patterns of individual Hox genes has In vertebrates, the Hox gene expression pat- where we seek homologies between phyla. been used to unravel the individualization of tern is linked to segment identity rather than Nonetheless, some deeply conserved gene body segments and appendages segment number15,16. So all cervical vertebrae expression patterns probably remain for us from a primitive pattern of equivalent seg- have the same Hox gene code, whether there to tease out. ments. Homologies are being drawn among be seven as in or 14 as in the chick. Although we expect to find a larger insect groups that have highly divergent A radical change in Hox gene expression, number of common mechanisms in similar mouthparts to infer how these ecologically involving changes in Hox gene expression organisms, we are discovering that changes driven modifications evolved9. Comparisons domains, correlates with the great expansion in genetic regulatory systems have also been also reveal homologies among insect, crus- of thoracic identity in the axial skeleton in marked among quite closely related taxa. tacean and chelicerate (notably spider) seg- body plan evolution (FIG. 1)17. This For instance, all vertebrates show internal ments11,12, as well as insights into the origins broad comparison between insects and verte- left–right asymmetry, but there are impor- of segmental differentiation in these arthro- brates shows that there is considerable flexi- tant differences in how this is genetically pods and in more primitive arthropod rela- bility in the mode of regulatory evolution, controlled in various vertebrates22. tives such as the velvet worm, Peripatus (an and that analogous effects can result from Although all tetrapods have similar limb onychophoran13). quite different evolutionary modifications of structures, the expression of regulatory Hox genes also regulate the development complex regulatory systems. genes in the developing frog is different from that observed in and mammals23. Finally, although the gene reg- a b Chick Python ulatory machinery used to develop the ver- tebrate fore- and hindlimbs is the same, Cervical T specific genes control fore- and hindlimb h identity24. F o l On the basis of results from develop- r Thoracic a mental genetic studies done in model sys- Flank a n c tems, such as Drosophila, mutations in k i genes controlling early development would c be expected to be deleterious, as they are Lumbar bound to affect all of later development. Early development should therefore evolve Hindlimb Hindlimb slowly or not at all. However, studies of many organisms give the counter-intuitive Figure 1 | Hox gene expression in the evolution of — a dramatic modification of the result — early development evolves freely, vertebrate body axis. a | The skeleton of a python stained with Alcian blue () and allowing highly divergent ontogenies to Alizarin red (). b | Schematic diagram comparing domains of Hox gene expression in chick and evolve among closely related species3.By snake : HoxB5, green; HoxC8, blue; HoxC6, red. Hox genes are involved in the this method distinct developmental modes regionalization of the lateral plate into forelimb, flank and hindlimb, to specify limb position. The expansion of HoxC8 and HoxC6 domains in python correlates with the expansion of thoracic and larval features have evolved among sea identity and can account for the absence of . (Adapted with permission from Nature (REF. 17) urchins, , ascidians, salamanders, © (1999) Macmillan Magazines Ltd.) frogs, and even polyembryonic

76 | OCTOBER 2000 | VOLUME 1 www.nature.com/reviews/genetics PERSPECTIVES insects, where a single gives rise to 2,000 ab separate embryos through a completely new developmental pathway25. These stud- ies show that early development can evolve as radically as later development, and that it also can contribute marked evolutionary novelties.

Origins of body plans Animal phyla each have visibly distinct body plans — the arrangement of their body parts. Chordates, for instance, have a dorsal central nervous system, a notochord and paired muscle groups, which are present from trout to tyrannosaur. However, gene sequence data show that all phyla (animal and non-animal) are evolutionarily related3. The origin of body plans is an important issue, combining studies of developmental biology, palaeontology and molecular evolu- 26,27 tion . Although the origins of most phyla Figure 2 | in the development of horns in the male dung , taurus. have not yet emerged from the fossil record, a | Small horns, produced by males below threshold size. b | Fully developed horns in a male over threshold fossil remains of MEMBERS of phyla show size. (Adapted with permission from Development (REF.34) © (1999) The Company of Biologists Ltd.) that body plans evolved their features sequentially, and even that some apparent with vertebrates. This anatomical switch is mental mechanism28. Progress in gene intermediate forms between phyla may accompanied by an inversion in the expres- expression studies may allow us to under- occur26,27. One of the main surprises from sion of genes that determine the dorsal and stand even more extreme morphological concerns the long-known ventral axes, indicating that the lineages that transformations, such as how inversion of the dorsal–ventral axis of stem from a common PROTOSTOME–DEUTERO- with pentameral symmetry have evolved and other PROTOSTOMES compared STOME ancestormay share the same develop- from a bilateral ancestor. As the fossil record has not revealed the ancestral animal, or any of the important Glossary ancestors to principal animal (such as CLADISTICS other parts of an embryo or the environment. the protostome–deuterostome ancestor), An approach to inferring evolutionary relationships attempts to infer the properties of these among organisms, on the basis of identifying shared HALTERES ancestors are based largely on the shared features among diverging clades. In Diptera (true flies), the second or hind wings have become modified into a pair of club-like balancing genes and developmental features among HOX CLUSTER organs called halteres. living clades. Current models of METAZOAN A group of linked regulatory genes involved in pattern- ancestors are closely linked to ideas on the ing the animal body axis during development. evolution of development. The larvae of Axial blocks of mesoderm along the vertebrate body most animals are built quite differently MACRO-EVOLUTION axis that further differentiate into dermal skin, bone 29 Evolutionary change above the species level. and muscle. from the adults. It has been argued that Evolutionary changes in populations within a species early animals were similar to larvae of living are termed micro-evolution. PROTOSTOME/DEUTEROSTOME marine , and used gene regula- The two principal divisions of animal phyla, based on tory systems similar to those used to pro- METAZOANS how the forms in the embryo. Multicellular animals. duce modern larvae. Adult body plans and BASAL MEMBERS their different gene regulatory systems Lineages or branches that diverge at the base of a phylo- would have evolved at a later stage with the The course of development in an organism from genetic tree; more primitive lineages. origin of ‘set aside’ cells that produce the embryo to adult. adult body plan within the quite dissimilar BILATERIAN PHYLOGENETICS Animals with bilateral body symmetry. larval body. This model requires that animal The study of evolutionary relationships development acquired a new step, and among organisms. CNIDARIANS demands a great deal of convergent evolu- Radially symmetric animals such as jelly fish, , tion of genetic systems regulating adult and anemonies. Four-legged vertebrate animals. development. The hypothesis is challenged by evidence of how developmental features FIXATION (OF AN ) The rapid diversification of animal life observed in the are phylogenetically distributed. These indi- When an allele replaces all other in a population, fossil record in rocks of early-mid Cambrian age cate that feeding larvae arose after adult so that its frequency is equal to one (that is, 100%). (540–530 million years ago). Many of the major phyla body plans26,30,31, and that set aside cells are that characterize modern animal life evolved at this time. 32 not homologous among all taxa . A second Events in development that depend on interactions with hypothesis therefore states that the ancestral

NATURE REVIEWS | GENETICS VOLUME 1 | OCTOBER 2000| 77 PERSPECTIVES

resource allocation to one body part affects a b the size of other parts, indicating that inter- actions occur that control relative growth and may provide developmental constraints. Scaling of body parts also can be greatly changed in response to artificial selection, providing a link between micro-evolution and development. It remains unclear how maps to phenotype. It is crucial to discover internal constructional features imminent in devel- opmental processes that constrain variation and determine how selection affects - isms. Constraints have been suggested to lie in the function of regulatory genes and in interactions among elements of a developing regulatory system3,35. The emerging unifying Figure 3 | Evolution of developmental mode in closely related species. a | Ventral view of an eight- theme is that developing systems are com- armed pluteus larva of the indirect-developing , Heliocidaris tuberculata. Arms (ar), hindgut (hg) posed of genetically discrete modules that and mouth (m) are all features of the feeding larva. The rudiment (r) represents the developing juvenile interact EPIGENETICALLY with each other during adult. b | Larva of H. erythrogramma. A ciliary band (cb) is present, but no mouth or larval gut. Most of the body in this larva corresponds to the juvenile rudiment, and feeding larval features have been discarded in development. Modules include individual favour of a highly modified direct development of the adult form. Scale bar, 100 µm. (Reprinted with elements of a developing system, such as the permission from Development (REF. 40) © (1999) The Company of Biologists Ltd.) oral ectoderm of the sea urchin larva, or the limb field of a vertebrate embryo. A modular structure generates constraint because some BILATERIAN animal was small, but possessed an Evolution biased by development interactions between modules may be diffi- adult-like body plan. Planktonic larvae would As in developmental biology, much of cult to de-couple. Paradoxically, have evolved later, perhaps as part of the research in evolutionary developmental biol- also allows marked evolutionary change, CAMBRIAN ‘EXPLOSION’. ogy is empirically driven. This is not surpris- because many inter-modular interactions The debate is not merely an exercise in ing given the lack of a general theory of can be dissociated in timing (), speculative . Both views require that development, and the diversity of develop- or in other ways that allow viable, albeit extensive has taken mental patterns. However, development may changed patterns of development3,36.The place. Either embryonic forms evolved con- make a crucial contribution to evolutionary link between the traits identified in selection vergently in several lineages, or complex fea- theory. Modern evolutionary biology has studies and the modules that seem to be tures of adult body plans and what are gen- focused on the role of , units of development still needs to be clari- erally regarded as shared, deeply embedded which operates external to the organism, and fied. The traits used in selection studies can developmental regulatory gene systems, views organisms as unconstrained in varia- be complex characters composed of several evolved independently with the invention of tion. Micro-evolutionary processes are con- underlying modules. For example, selection set aside cells. This latter hypothesis seems sidered sufficient to explain macro-evolu- on tail length in mice would potentially less probable, particularly in the light of con- tionary history4. However, developmental involve several constituent developmental vergent evolution of larval forms. The issue processes are emergent, and not predictable modules, such as SOMITES. Experimental sys- is still unresolved. from the properties of genes or cells; there- tems (such as butterfly patterning), in A controversial debate surrounds the fore, starting with a particular ONTOGENY, which a link has been found between the divergence times of phyla7. Typically these some phenotypes might be readily achieved units of micro-evolution and developmental have been extrapolated from the differences and others impossible. Developmental modules (and their regulation), provide a in gene sequence between taxa for which mechanisms are crucial, both to large-scale crucial link between development and stud- divergence times are known. Extrapolations evolutionary changes, and also to small-scale ies of selection37. deeper into time estimate diver- evolutionary processes. The genetic mechanisms that permit gence times that range from 600 to 1,200 The evolution of body shape poses the such dissociations probably lie in the combi- million years ago, which is too broad to be difficult problem of how the scaling of body natorial structure of eukaryotic promoters, useful. Once again, there is hope that the parts is regulated during development. which allow gene expression to be modified fossil record will resolve the argument by Potential constraints in interactive or co- in various ways, and to be readily co-opted providing some unexpected data. varying systems during development high- to new functions. The developmental mech- Microfossils, thought to represent marine light the mechanistically definable limita- anisms of inter-modular dissociation are animal embryos, are emerging from late tions that development imposes on the not well understood. So we have the amaz- rocks, and if they turn out to micro- and macro-evolution of body form. ing but unexplained observation that differ- be widespread in time and diverse in pre- In insects, the growth of body features, such ent developmental pathways can converge served forms, they may show us what as horns, is linked to body size through com- on similar outcomes. For example, changes ancient larvae looked like and provide a mon regulation by juvenile hormone (FIG. 2), in embryonic modules produce different better minimal date for the origins of ani- suggesting mechanisms for evolutionary pathways during early development of simi- mal development33. variation34. Experimentally varying the lar sea urchins (FIG. 3)38, and induction of the

78 | OCTOBER 2000 | VOLUME 1 www.nature.com/reviews/genetics PERSPECTIVES

eye lens in some frogs depends on induc- a Links 19. Halder, G., Callaerts, P. & Gehring, W. J. Induction of ectopic eyes by targeted expression of the eyeless gene tion by the optic cup in some species, but in Drosophila. Science 267, 1788–1792 (1995). FURTHER INFORMATION Faculty research not in others39. 20. Shubin, N., Tabin, C. & Carroll, S. , genes and the Modularity is a characteristic of multi- interests at the University of Indiana | evolution of animal limbs. Nature 388, 639–648 (1997). Evolution and Development | Raff lab 21. Gerhart, J. & Kirschner, M. Cells, Embryos, and Evolution cellular life, and modules themselves must (Blackwell, Malden, 1997). homepage have evolved. Individual developmental 22. Burndine, R. D. & Schier, A. F. Conserved and divergent ENCYCLOPEDIA OF LIFE SCIENCES mechanisms in left–right axis formation. Genes Dev. 14, modules initially may have arisen by inte- 763–776 (2000). Evolutionary developmental biology: 23. Christen, B. & Slack, J. All limbs are not the same. Nature gration of genetic processes that regulated Homologous regulatory genes and processes 395, 230–231 (1998). separate events. Later, as more complex 24. Rodriguez-Esteban, C. et al. The T–box genes Tbx4 and ontogenies evolved, more individualized Tbx5 regulate limb outgrowth and identity. Nature 398, 1. Slack, J. M. W., Holland, P. W. H. & Graham, C. F. The 814–818 (1999). modules may have arisen by packaging ele- zootype and the . Nature 361, 490–492 25. Grbic, M., Nagy, L. M. & Strand, M. R. Development of ments from within larger integrated units (1993). polyembryonic insects: a major departure from typical 2. Coates, M. I. & Clack, J. A. Polydactyly in the earliest insect embryogenesis. Dev. Genes Evol. 208, 69–81 into separate modular entities, each an inde- tetrapod limbs. Nature 347, 66–69 (1990). (1998). pendent target of selection36. 3. Raff, R. A. The Shape of Life: Genes, Development and 26. Budd, G. E. & Jensen, S. A critical reappraisal of the the Evolution of Animal Form (Chicago Univ. Press, fossil record of the bilaterian phyla. Biol. Rev. 75, Chicago, 1996). 253–295 (2000). Challenges 4. Leroi, A. M. The scale independence of evolution. Evol. 27. Conway Morris, S. Why molecular biology needs Dev. 2, 67–77 (2000). palaeontology. Development S1–S13 (1994). The synthesis of the sciences of biological 5. Orr, H. A. The of : the 28. De, Robertis, E. M. & Sasai, Y. A. A common plan for change promises new and powerful solu- distribution of factors fixed during . dorsoventral patterning in . Nature 380, 37–40 Evolution 52, 935–949 (1998). (1996). tions to long-standing problems, and a new 29. Peterson, K. J., Cameron, R. A. & Davidson, E. H. 6. Haag, E. S. & True, J. R. From mutants to mechanisms? Bilaterian origins: significance of new experimental understanding of the basis of evolution. Assessing the candidate gene paradigm in evolutionary observations. Dev. Biol. 219, 1–17 (2000). biology. Evolution (Submitted.) Along with the controversies listed in BOX 1, 30. Jenner, R. A. Evolution of animal body plans: the role of 7. Smith, A. B. Dating the origin of metazoan body plans. metazoan phylogeny at the interface between pattern a number of crucial fundamental challenges Evol. Dev. 1, 138–142 (1999). and process. Evol. Dev. 2, 208–221 (2000). 8. Mabee, P. M. Developmental data and phylogenetic remain. These include: gaining an under- 31. Rouse, G. W. The epitome of hand waving? Larval . Am. Zool. (in the press). feeding and hypotheses of metazoan phylogeny. Evol. standing of how regulatory gene networks 9. Rogers, B. T. & Kaufman, T. C. Structure of the insect Dev. 2, 222–233 (2000). head in ontogeny and phylogeny: a view from govern ontogeny; what makes developing 32. Valentine, J. W., Jablonski, D. & Erwin, D. H. Fossils, Drosophila. Int. Rev. Cytol. 174, 1–84 (1997). molecules and embryos: new perspectives on the systems robust enough to tolerate muta- 10. Finnerty, J. R. & Martindale, M. Q. Ancient origins of axial Cambrian explosion. Development 126, 851–859 (1999). patterning genes: Hox genes and paraHox genes in the tions that change the course of development 33. Chen, J. Y. et al. Precambrian animal diversity: putative . Evol. Dev. 1, 16–23 (1999). phosphatized embryos from the Doushantuo Formation so that developmental evolution is possible; 11. Abzhanov, A., Popadic, A. & Kaufman, T. C. Chelicerate of . Proc. Natl Acad. Sci. USA 97, 4457–4462 Hox genes and the of arthropod segments. and how the rules that govern ontogeny (2000). Evol. Dev. 1, 77–89 (1999). 34. Stern, D. L. & Emlen, D. J. The developmental basis for constrain the production of new variation 12. Averof, M. & Akam, M. Hox genes and the diversification in insects. Development 126, 1091–1101 of insect and body plans. Nature 376, in phenotypes. Developmental genetics, (1999). 420–423 (1995). 35. Kirschner, M. & Gerhart, J. . Proc. Natl Acad. comparative developmental biology, 13. Grenier, J. K., Garber, T. L., Warren, R., Whitington, P. M. Sci. USA 95, 8420–8427 (1998). & Carroll, S. Evolution of the entire arthropod Hox gene palaeontology and genomics are adding a 36. Wagner, G. P. Homologues, natural kinds and the set predated the origin and radiation of the evolution of modularity. Am. Zool. 36, 36–43 (1996). vast number of new data sets. Questions on onychophoran/arthropod . Curr. Biol. 7, 547–553 37. Brakefield, P. M. et al. Development, plasticity and (1997). the nature of homology (a subject made evolution of butterfly patterns. Nature 384, 14. Weatherbee, S. D. et al. function in butterfly even more rich and strange by the emer- 236–242 (1996). wings and the evolution of patterns. Curr. 38. Raff, R. A. & Sly, B. J. Modularity and dissociation in the gence of evolutionary developmental biolo- Biol. 9, 109–115 (1999). evolution of gene expression territories in development. 15. Gaunt, S. J. Conservation in the Hox code during gy), the origins of novelties and ultimately a Evol. Dev. 2, 102–113 (2000). morphological evolution. Int. J. Dev. Biol. 38, 549–552 complete understanding of evolution lie 39. Jacobson, A. G. & Sater, A. K. Features of embryonic (1994). induction. Development 104, 341–359 (1988). before this young discipline. 16. Burke, A. C., Nelson, C. E., Morgan, B. A. & Tabin, C. 40. Raff, E. C. et al. A novel ontogenetic pathway in Hox genes and the evolution of vertebrate axial embryos between species with different modes of Indiana Molecular Biology Institute and morphology. Development 121, 333–346 (1995). development. Development 126, 1937–1945 (1999). Department of Biology, Indiana University, Jordan 17. Cohn, M. & Tickle, C. Developmental basis of limblessness and axial patterning in snakes. Nature 399, Hall, 1001 East Third Street, Bloomington, 474–479 (1999). Acknowledgements Indiana 47405 USA. e-mail: 18. Fishman, M. C. & Olson, E. N. Parsing the heart: genetic I thank T. Frankino and K. Wilson for helpful discussions and [email protected] modules for organ assembly. 91, 153–156 (1997). suggestions.

We welcome correspondence Has something in the journal caught your attention? If so, please write to us about it by sending an email to: [email protected] and flag it for the attention of the Nature Reviews Genetics editors. Correspondence to the journal will be selected by the editors for publication on the Nature Reviews Genetics website at http://www.nature.com/reviews/genetics/ where it will be linked to the relevant article.

NATURE REVIEWS | GENETICS VOLUME 1 | OCTOBER 2000| 79