Plant studies are plant studies and animal studies are animal studies and never the twain shall meet. Though few biologists would endorse that description as a desirable state of affairs, many would agree that it is a fairly accurate one. To help bridge the gap between the two kingdoms, or, more accurately, between the two groups of researchers, a conference on pattern formation in plants and animals was organized by the John Innes Institute, in collaboration with the British Society for Developmental Biology. The ‘Molecular and Cellular Basis of Pattern Formation’ took place in September, the 3rd to the 6th, at the University of East Anglia (Norwich, UK) and was attended by 230 participants, predominantly from the UK and the US, but with significant participation from continental Europe and some attendees from other countries. Saturation Mutagenesis and Analysis of Pattern The conference opened with the 14th Bateson Memorial Lecture from C. Nusslein-Volhard (Tubingen), who described the work from her laboratory on the strict maternal effect mutants of Drosophila that identify genes essential for organizing the early embryonic pattern of the fruit fly. This analysis has identified 30 genes to date, whose actions define four functional groups of maternal effect genes required for pattern formation in four fairly distinct spatial domains. Three of these gene groups are required for organizing the antero-posterior axis (the anterior, posterior, and terminal groups) and the fourth, the dorso-ventral (d-v) axis. Each group is involved with the production of a localized signal but there are significant differences in molecular mechanisms associated with each of these signals. Strikingly, both the d-v system and the terminal system involve membrane transduction systems that mediate localized reactions to external signals originating in the perivitelline fluid. There is, as yet, no plant model system that has been analyzed with the same genetic scrutiny as Drosophila, but Arubidopsis, possessing a small genome and a short generation time, may have the same potential. G. Jurgens (Munich), a former colleague of NussleinVolhard’s, presented the work of his laboratory on the search for mutants affecting pattern in the early embryo of Arubidopsis. Of MOO0 separate lines, derived from
EMS mutagenesis of seeds, 250 mutants, which affect early embryonic pattern without causing early lethality, have been identified. Four broad classes of pattern alteration have been found in these mutants: deleted structures, duplicated structures, transformations and reiterated structures. From complementation studies, it appears that the group of genes that give these pattern defects may be close to saturation. This work is the first attempt to apply saturation mutagenesis to analysis of developmental phenomena in plants and illustrates that the approach, applied to the right kind of system, is likely to be as fruitful as it has been in animals (Drosophila, Caenorhabditis). Cell Polarity The overt development of patterns, whatever their underlying genetic basis, involves the organized spatial behaviours of cells and the first sessions were largely devoted to the discussion of cell properties and cellular interactions in development. Several talks dealt with the development of spatial asymmetries, and in particular, polarity in individual cells. R. Quatrano (Chapel Hill) described the development of polarity in the brown alga, Fucus, which starts as an apolar, spherical cell that develops an axis in response to various non-uniformly distributed environmental signals (light, voltage gradients, p H gradients), eventually accumulating granules in one region and developing a rhizoid there. The development of polarity can be experimentally divided into two phases: the first, axis formation, involving the redistribution and concentration of Ca2+ pumps at the site of future rhizoid development, and the second, axis fixation, which is dependent on microfilaments and their concentration at the site of future rhizoid development. The contribution of the cell wall to these two steps can be determined by its removal; it is found that the cell wall is not needed for the formation of the axis but is for its stabilization. Drug experiments indicate that microtubules (mts) play no significant role in axis formation. In contrast, in the development of polarity in cells of the 8-cell mouse embryo, mts play a significant role. B. Mar0 (Paris) described the various cytoplasmic and surface signs of polarity development in these cells as compaction (the drawing together of cells at the 8-cell stage) takes place and described the role of mts in positioning of the nucleus, which in turn influences the development of surface polarity. Drug experiments indicate that both mts and microfilaments are together necessary for development of full polarity. Similarly, in elongated, vacuolated epidermal plant cells, both mts and actin networks appear to play crucial roles in setting the division plane and, hence, affecting the placement of each cell with respect to its neighbors, as described by C. Lloyd (John Innes). In particular, the ‘avoidance’ by microtubule strands of vertices helps to explain both the staggered arrangement of cell files in the epidermis and the quasi-hexagonal geometry of cells in other
tissues. T. Sachs (Jerusalem) discussed the interplay of a systemic factor (auxin) and intrinsic factors (initial polarity) in strengthening an initial bias in polarity of plant cells. Cell-cell Interactions While polarity and division properties of cells constitute one key element in the cellular basis of patterning, cell-cell interactions, both long- and short-range, constitute another. One approach to studying such cell interactions is to create mosaics or chimaeras for genetic differences that affect a particular property and, depending upon the site in which the gene or genes are expressed and the ultimate phenotypic outcome, determine which tissue is particularly crucial for establishing the patterning difference. Three talks described results using this approach. S. Poethig (University of Pennsylvania) used this approach to study the site of expression of the okra mutation, which affects leaf shape, in cotton. Employing the semigumy mutation (which creates haploid sectors) to form mosaics, he concluded that the leaf patterning step affected by okra occurs when the mutant allele is expressed in the epidermal or subepidermal layers but not in the spongy mesophyll. In contrast, I. Sussex (Berkeley), examining graft chimaeras to analyze the roles of the different tissue layers in setting apical meristem growth, found that the spongy mesophyll layer was critical in setting growth pattern for the shoot. Similarly, S. Hake (USDA, Berkeley), found in mosaics for the mutations in the maize gene Knotted ( K n l ) that the mutation creates its effect through action in the mesophyll layer. K n l is a homeobox-containing gene, and hence a putative transcriptional control gene; intriguingly, the different dominant mutations, with the exception of one that is a tandem duplication, have their mutational lesions within introns, and produce their effects through spatial misexpression patterns. A cell- and tissue-interaction in animal systems that also involves homeodomain proteins was described by N. Hopwood (University of Cambridge). This is the induction of mesoderm in Xenopus embryos, involving the response of animal hemisphere cells to the underlying vegetal cells. Two homeobox genes, encoding the Xenopus homologues of the MyoD and twist genes, are synthesized in the induced mesoderm but in exclusive domains, with MyoD expressed in the myotomes and twi in non-myotome mesoderm (namely the notochord and visceral and lateral mesoderm). They presumably regulate different sets of genes in these different mesodermal derivatives. While the precise nature of the mesoderm-inducing substances that operate in vivo in the amphibian embryo is still something of a matter of debate (though several candidate growth factor homologues are being investigated), the effects of several well-documented molecules that are known to act in vivo were the subject
of several talks. R. Kay (LMB, Cambridge) discussed the roles of the DIF molecules and cAMP in Dictyostelium fruiting body formation. The DIF family, a group of chlorinated alkylphenones, promote stalk cell development in this organism while cAMP promotes spore cell formation. A paradox has been that the DIF gradient in the fruiting body runs in the ‘wrong’ direction, being higher in the pre-spore cells. Kay presented evidence that the different developmental responses may, in fact, reflect DIF dose-dependence in combination with differential cAMP concentrations. The role of retinoic acid (RA) as a morphogen in chick limb development was reviewed by C . Tickle (University College, London). The evidence for an RA gradient in the limb bud is strong as is that for different responses correlated with different dosages. Part of the R A effect may involve differentially regulated proliferative responses, involving production of I G F and TGFbeta. In contrast to the long range patterning effects of RA, building the photoreceptor cell pattern of the Drosophila eye involves shorter range effects, involving cell neighbors. E. Hafen (Zurich) described the role of the sevenless (sev) gene as encoding a receptor on R7 cells necessary for R7 development, in response to a ligand produced by R8 cells, encoded by the bride-ofsevenless (boss) gene. This interaction is closely dependent on the quantities of receptor and ligand; strongly stimulated sev expression, produced by the appropriate genetic engineering and germ-line transformation with the gene construct, can recruit neighboring but non-photoreceptor cells to R7 development, independently of boss.
Homeotic Genes The role of homeotic genes in pattern formation was the subject of the final two sessions of the conference. R. Garber (Novagen, Madison) described some of the combinatorial controls that regulate the Antennapedia (Antp) gene of Drosophila.Antp, like K n l of maize (see above), was first identified on the basis of dominant mutations (the transformation of antenna1 to leg structures) which are now known to involve mutations outside the coding region and misexpression in new sites. Garber described the elaborate regulatory machinery for expressing Antp protein at particular times and places in the embryo and the existence of degrees of functional redundancy between Antp and Sex combs reduced (Scr). R. Krumlauf (Mill Hill, London) described the murine HOX-2 gene cluster, a group of genes homologous to members of the Drosophila bithorax and Antennapedia complexes, and its remarkably conserved order between mammals and Drosophila and the significance of these genes for patterning in the rhombomeres. M. Akam (University of Cambridge) took up the theme of conserved gene order and functions of the homeotic genes and some instructive differences in segmentation genes and
homeotic gene expression between long- and shortgerm band insects. For those familiar with the animal systems but comparatively ignorant about plant developmental biology, however, the three talks on plant homeotics were particularly instructive. E. Coen (John Innes, Norwich) and E. Meyerowitz (CalTech, Pasadena) reviewed the known floral homeotics in Antirrhinum and Arabidopsis, respectively, and presented combinatorial models that both explain the previous data and predict the outcomes of certain double mutant combinations. The models involve both additive and competitive interactions and the parallels to the Drosophila homeotics were brought out strongly. Z . Schwarz-Sommer (Cologne, Germany) described in detail one of the Antirrhinum homeotics, dejiciens (def A ) , which, when mutated, converts petals to sepals and stamens to carpels. d e f A is one of a family of more than 10 closely related putative transcription factors and appears, from its expression pattern, to be required for maintenance of developmental state (rather than initiation), as is the case for Antp and the genes of the
bithorax complex of Drosophila. As emphasized by Meyerowitz and Schwarz-Sommer, the homeotics identified in these two plant species are activated in response to some underlying ‘positional information’ but the nature of the latter remains elusive. The final talk of the conference was a provocative and witty summation by Ian Sussex, who said that he had initially intended to stress some of the salient developmental differences between plants and animals yet, as the conference proceeded, found that the conventionally-cited differences are considerably less compelling than generally believed. He ended by saying that the task for plant biologists was to convince other biologists that plants are as interesting as animals. For this participant, at least, the conference had already achieved that. Adam S. Wilkins is with the Company of Biologists, Department of Zoology, University of Cambridge, Downing St, Cambridge CB2 3EJ, UK.
Correspondence BfoEssays welcomes correspondence. Readers wishing to comment on subjects or issues raised in our pages should send their letters either to the editor, Dr Adam S. Wilkins, Company of Biologists, Department of Zoology, University of Cambridge, Downing St., Cambridge CB2 3EJ, UK or to the chairman of the editorial board, Dr Kermit L. Carraway, Department of Cell Biology and Anatomy, PO Box 016960, University of Miami School of Medicine, Miami, FL 33101, USA.
An error has appeared on the Contents pages of each of the preceding two issues. In the October issue (vol. 12, no. lo), the title of the ‘What the Papers Say’ contribution by Andrew Collins was given as ‘Topoisomerase I can relax: Novobiocin is a mitochondria1 poison after all.’ The title should have read ‘Topoisomerase I1 can relax.. .’.
In the November issue (vol. 12, no. ll), the reviewer of Molecular Biology of Plant Nuclear Genes was mistakenly listed as T. Baldwin. The reviewer, in fact, was Rosalind Slatter. We offer our apologies to Drs Collins and Slatter for these errors.
EMS mutagenesis of seeds, 250 mutants, which affect early embryonic pattern without causing early lethality, have been identified. Four broad classes of pattern alteration have been found in these mutants: deleted structures, duplicated structures, transformations and reiterated structures. From complementation studies, it appears that the group of genes that give these pattern defects may be close to saturation. This work is the first attempt to apply saturation mutagenesis to analysis of developmental phenomena in plants and illustrates that the approach, applied to the right kind of system, is likely to be as fruitful as it has been in animals (Drosophila, Caenorhabditis). Cell Polarity The overt development of patterns, whatever their underlying genetic basis, involves the organized spatial behaviours of cells and the first sessions were largely devoted to the discussion of cell properties and cellular interactions in development. Several talks dealt with the development of spatial asymmetries, and in particular, polarity in individual cells. R. Quatrano (Chapel Hill) described the development of polarity in the brown alga, Fucus, which starts as an apolar, spherical cell that develops an axis in response to various non-uniformly distributed environmental signals (light, voltage gradients, p H gradients), eventually accumulating granules in one region and developing a rhizoid there. The development of polarity can be experimentally divided into two phases: the first, axis formation, involving the redistribution and concentration of Ca2+ pumps at the site of future rhizoid development, and the second, axis fixation, which is dependent on microfilaments and their concentration at the site of future rhizoid development. The contribution of the cell wall to these two steps can be determined by its removal; it is found that the cell wall is not needed for the formation of the axis but is for its stabilization. Drug experiments indicate that microtubules (mts) play no significant role in axis formation. In contrast, in the development of polarity in cells of the 8-cell mouse embryo, mts play a significant role. B. Mar0 (Paris) described the various cytoplasmic and surface signs of polarity development in these cells as compaction (the drawing together of cells at the 8-cell stage) takes place and described the role of mts in positioning of the nucleus, which in turn influences the development of surface polarity. Drug experiments indicate that both mts and microfilaments are together necessary for development of full polarity. Similarly, in elongated, vacuolated epidermal plant cells, both mts and actin networks appear to play crucial roles in setting the division plane and, hence, affecting the placement of each cell with respect to its neighbors, as described by C. Lloyd (John Innes). In particular, the ‘avoidance’ by microtubule strands of vertices helps to explain both the staggered arrangement of cell files in the epidermis and the quasi-hexagonal geometry of cells in other
tissues. T. Sachs (Jerusalem) discussed the interplay of a systemic factor (auxin) and intrinsic factors (initial polarity) in strengthening an initial bias in polarity of plant cells. Cell-cell Interactions While polarity and division properties of cells constitute one key element in the cellular basis of patterning, cell-cell interactions, both long- and short-range, constitute another. One approach to studying such cell interactions is to create mosaics or chimaeras for genetic differences that affect a particular property and, depending upon the site in which the gene or genes are expressed and the ultimate phenotypic outcome, determine which tissue is particularly crucial for establishing the patterning difference. Three talks described results using this approach. S. Poethig (University of Pennsylvania) used this approach to study the site of expression of the okra mutation, which affects leaf shape, in cotton. Employing the semigumy mutation (which creates haploid sectors) to form mosaics, he concluded that the leaf patterning step affected by okra occurs when the mutant allele is expressed in the epidermal or subepidermal layers but not in the spongy mesophyll. In contrast, I. Sussex (Berkeley), examining graft chimaeras to analyze the roles of the different tissue layers in setting apical meristem growth, found that the spongy mesophyll layer was critical in setting growth pattern for the shoot. Similarly, S. Hake (USDA, Berkeley), found in mosaics for the mutations in the maize gene Knotted ( K n l ) that the mutation creates its effect through action in the mesophyll layer. K n l is a homeobox-containing gene, and hence a putative transcriptional control gene; intriguingly, the different dominant mutations, with the exception of one that is a tandem duplication, have their mutational lesions within introns, and produce their effects through spatial misexpression patterns. A cell- and tissue-interaction in animal systems that also involves homeodomain proteins was described by N. Hopwood (University of Cambridge). This is the induction of mesoderm in Xenopus embryos, involving the response of animal hemisphere cells to the underlying vegetal cells. Two homeobox genes, encoding the Xenopus homologues of the MyoD and twist genes, are synthesized in the induced mesoderm but in exclusive domains, with MyoD expressed in the myotomes and twi in non-myotome mesoderm (namely the notochord and visceral and lateral mesoderm). They presumably regulate different sets of genes in these different mesodermal derivatives. While the precise nature of the mesoderm-inducing substances that operate in vivo in the amphibian embryo is still something of a matter of debate (though several candidate growth factor homologues are being investigated), the effects of several well-documented molecules that are known to act in vivo were the subject
of several talks. R. Kay (LMB, Cambridge) discussed the roles of the DIF molecules and cAMP in Dictyostelium fruiting body formation. The DIF family, a group of chlorinated alkylphenones, promote stalk cell development in this organism while cAMP promotes spore cell formation. A paradox has been that the DIF gradient in the fruiting body runs in the ‘wrong’ direction, being higher in the pre-spore cells. Kay presented evidence that the different developmental responses may, in fact, reflect DIF dose-dependence in combination with differential cAMP concentrations. The role of retinoic acid (RA) as a morphogen in chick limb development was reviewed by C . Tickle (University College, London). The evidence for an RA gradient in the limb bud is strong as is that for different responses correlated with different dosages. Part of the R A effect may involve differentially regulated proliferative responses, involving production of I G F and TGFbeta. In contrast to the long range patterning effects of RA, building the photoreceptor cell pattern of the Drosophila eye involves shorter range effects, involving cell neighbors. E. Hafen (Zurich) described the role of the sevenless (sev) gene as encoding a receptor on R7 cells necessary for R7 development, in response to a ligand produced by R8 cells, encoded by the bride-ofsevenless (boss) gene. This interaction is closely dependent on the quantities of receptor and ligand; strongly stimulated sev expression, produced by the appropriate genetic engineering and germ-line transformation with the gene construct, can recruit neighboring but non-photoreceptor cells to R7 development, independently of boss.
Homeotic Genes The role of homeotic genes in pattern formation was the subject of the final two sessions of the conference. R. Garber (Novagen, Madison) described some of the combinatorial controls that regulate the Antennapedia (Antp) gene of Drosophila.Antp, like K n l of maize (see above), was first identified on the basis of dominant mutations (the transformation of antenna1 to leg structures) which are now known to involve mutations outside the coding region and misexpression in new sites. Garber described the elaborate regulatory machinery for expressing Antp protein at particular times and places in the embryo and the existence of degrees of functional redundancy between Antp and Sex combs reduced (Scr). R. Krumlauf (Mill Hill, London) described the murine HOX-2 gene cluster, a group of genes homologous to members of the Drosophila bithorax and Antennapedia complexes, and its remarkably conserved order between mammals and Drosophila and the significance of these genes for patterning in the rhombomeres. M. Akam (University of Cambridge) took up the theme of conserved gene order and functions of the homeotic genes and some instructive differences in segmentation genes and
homeotic gene expression between long- and shortgerm band insects. For those familiar with the animal systems but comparatively ignorant about plant developmental biology, however, the three talks on plant homeotics were particularly instructive. E. Coen (John Innes, Norwich) and E. Meyerowitz (CalTech, Pasadena) reviewed the known floral homeotics in Antirrhinum and Arabidopsis, respectively, and presented combinatorial models that both explain the previous data and predict the outcomes of certain double mutant combinations. The models involve both additive and competitive interactions and the parallels to the Drosophila homeotics were brought out strongly. Z . Schwarz-Sommer (Cologne, Germany) described in detail one of the Antirrhinum homeotics, dejiciens (def A ) , which, when mutated, converts petals to sepals and stamens to carpels. d e f A is one of a family of more than 10 closely related putative transcription factors and appears, from its expression pattern, to be required for maintenance of developmental state (rather than initiation), as is the case for Antp and the genes of the
bithorax complex of Drosophila. As emphasized by Meyerowitz and Schwarz-Sommer, the homeotics identified in these two plant species are activated in response to some underlying ‘positional information’ but the nature of the latter remains elusive. The final talk of the conference was a provocative and witty summation by Ian Sussex, who said that he had initially intended to stress some of the salient developmental differences between plants and animals yet, as the conference proceeded, found that the conventionally-cited differences are considerably less compelling than generally believed. He ended by saying that the task for plant biologists was to convince other biologists that plants are as interesting as animals. For this participant, at least, the conference had already achieved that. Adam S. Wilkins is with the Company of Biologists, Department of Zoology, University of Cambridge, Downing St, Cambridge CB2 3EJ, UK.
Correspondence BfoEssays welcomes correspondence. Readers wishing to comment on subjects or issues raised in our pages should send their letters either to the editor, Dr Adam S. Wilkins, Company of Biologists, Department of Zoology, University of Cambridge, Downing St., Cambridge CB2 3EJ, UK or to the chairman of the editorial board, Dr Kermit L. Carraway, Department of Cell Biology and Anatomy, PO Box 016960, University of Miami School of Medicine, Miami, FL 33101, USA.
An error has appeared on the Contents pages of each of the preceding two issues. In the October issue (vol. 12, no. lo), the title of the ‘What the Papers Say’ contribution by Andrew Collins was given as ‘Topoisomerase I can relax: Novobiocin is a mitochondria1 poison after all.’ The title should have read ‘Topoisomerase I1 can relax.. .’.
In the November issue (vol. 12, no. ll), the reviewer of Molecular Biology of Plant Nuclear Genes was mistakenly listed as T. Baldwin. The reviewer, in fact, was Rosalind Slatter. We offer our apologies to Drs Collins and Slatter for these errors.
