Development (Cambridge)

We have designed a system for targeted gene expression that allows the selective activation of any cloned gene in a wide variety of tissueand cell-specific patterns. The gene encoding the yeast transcriptional activator GAL4 is inserted randomly into the Drosophila genome to drive GAL4 expression from one of a diverse array of genomic enhancers. It is then possible to introduce a gene containing GAL4 binding sites within its promoter, to activate it in those cells where GAL4 is expressed, and to observe the effect of this directed misexpression on development. We have used GAL4-directed transcription to expand the domain of embryonic expression of the homeobox protein even-skipped. We show that even-skipped represses wingless and transforms cells that would normally secrete naked cuticle into denticle secreting cells. The GAL4 system can thus be used to study regulatory interactions during embryonic development. In adults, targeted expression can be used to generate dominant phenotypes for use in genetic screens. We have directed expression of an activated form of the Dras2 protein, resulting in dominant eye and wing defects that can be used in screens to identify other members of the Dras2 signal transduction pathway.

We have constructed a series of strains to facilitate the generation and analysis of clones of genetically distinct cells in developing and adult tissues of Drosophila. Each of these strains carries an FRT element, the target for the yeast FLP recombinase, near the base of a major chromosome arm, as well as a gratuitous cellautonomous marker. Novel markers that carry epitope tags and that are localized to either the cell nucleus or cell membrane have been generated. As a demonstration of how these strains can be used to study a particular gene, we have analyzed the developmental role of the Drosophila EGF receptor homolog. Moreover, we have shown that these strains can be utilized to identify new mutations in mosaic animals in an efficient and unbiased way, thereby providing an unprecedented opportunity to perform systematic genetic screens for mutations affecting many biological processes.

In a large-scale screen, we isolated mutants displaying a specific visible phenotype in embryos or early larvae of the zebrafish, Danio rerio. Males were mutagenized with ethylnitrosourea (ENU) and F2 families of single pair matings between sibling F1 fish, heterozygous for a mutagenized genome, were raised. Egg lays were obtained from several crosses between F2 siblings, resulting in scoring of 3857 mutagenized genomes. F3 progeny were scored at the second, third and sixth day of development, using a stereo-microscope. In a subsequent screen, fixed embryos were analyzed for correct retinotectal projection. A total of 4264 mutants were identified. Two thirds of the mutants displaying rather general abnormalities were eventually discarded. We kept and characterized 1163 mutants. In complementation crosses performed between mutants with similar phenotypes, 894 mutants have been assigned to 372 genes. The average allele frequency is 2.4. We identified genes involved in early development, notochord, brain, spinal cord, somites, muscles, heart, circulation, blood, skin, fin, eye, otic vesicle, jaw and branchial arches, pigment pattern, pigment formation, gut, liver, motility and touch response. Our collection contains alleles of almost all previously described zebrafish mutants. From the allele frequencies and other considerations we estimate that the 372 genes defined by the mutants probably represent more than half of all genes that could have been discovered using the criteria of our screen. Here we give an overview of the spectrum of mutant phenotypes obtained, and discuss the limits and the potentials of a genetic saturation screen in the zebrafish.

Cadherins are a family of glycoproteins involved in the Ca2+-dependent cell-cell adhesion mechanism which is detected in most kinds of tissues. Inhibition of the cadherin activity with antibodies induces dissociation of cell layers, indicating a fundamental importance of these molecules in maintaining the multicellular structure. Cadherins are divided into subclasses, including E-, N- and P-cadherins. While all subclasses are similar in molecular weight, Ca2+- and proteasesensitivity, each subclass is characterized by a unique tissue distribution pattern and immunological specificity. Analysis of amino acid sequences deduced from cDNA encoding these molecules showed that they are integral membrane proteins of 723–748 amino acids long and share common sequences; similarity in the sequences between subclasses is in a range of 50–60 % when compared within a single animal species.

L cells, with very little endogenous cadherin activity, transfected with the cadherin cDNA acquired high cadherin-mediated aggregating activity. Their colony morphology was altered by the ectopic expression of cadherins from the dispersed type to the compact type, providing direct evidence for a key role of cadherins in cell-cell adhesion. It has been suggested that cadherins bind cells by their homophilic interactions at the extracellular domain and are associated with actin bundles at the cytoplasmic domain.

It appears that each cadherin subclass has binding specificity and this molecular family is involved in selective cell-cell adhesion. In development, the expression of each cadherin subclass is spatiotemporally regulated and associated with a variety of morphogenetic events; e.g. the termination or initiation of expression of a cadherin subclass in a given cell collective is correlated with its segregation from or connection with other cell collectives. Antibodies to cadherins were shown to perturb the morphogenesis of some embryonic organs in vitro. These observations suggest that cadherins play a crucial role in construction of tissues and the whole animal body.

Development of a vascular system involves the assembly of two principal cell types – endothelial cells and vascular smooth muscle cells/pericytes (vSMC/PC) – into many different types of blood vessels. Most, if not all, vessels begin as endothelial tubes that subsequently acquire a vSMC/PC coating. We have previously shown that PDGF-B is critically involved in the recruitment of pericytes to brain capillaries and to the kidney glomerular capillary tuft. Here, we used desmin and α-smooth muscle actin (ASMA) as markers to analyze vSMC/PC development in PDGF-B−/− and PDGFR-β−/− embryos. Both mutants showed a site-specific reduction of desmin-positive pericytes and ASMA-positive vSMC. We found that endothelial expression of PDGF-B was restricted to immature capillary endothelial cells and to the endothelium of growing arteries. BrdU labeling showed that PDGFR-β-positive vSMC/PC progenitors normally proliferate at sites of endothelial PDGF-B expression. In PDGF-B−/− embryos, limb arterial vSMC showed a reduced BrdU-labeling index. This suggests a role of PDGF-B in vSMC/PC cell proliferation during vascular growth. Two modes of vSMC recruitment to newly formed vessels have previously been suggested: (1) de novo formation of vSMC by induction of undifferentiated perivascular mesenchymal cells, and (2) co-migration of vSMC from a preexisting pool of vSMC. Our data support both modes of vSMC/PC development and lead to a model in which PDGFR-β-positive vSMC/PC progenitors initially form around certain vessels by PDGF-B-independent induction. Subsequent angiogenic sprouting and vessel enlargement involves PDGF-B-dependent vSMC/PC progenitor co-migration and proliferation, and/or PDGF-B-independent new induction of vSMC/PC, depending on tissue context.

It has been proposed that the Xenopus homeobox gene, XlHbox8, is involved in endodermal differentiation during pancreatic and duodenal development (Wright, C. V. E., Schnegelsberg, P. and De Robertis, E. M. (1988). Development 105, 787–794). To test this hypothesis directly, gene targeting was used to make two different null mutations in the mouse XlHbox8 homolog, pdx-1. In the first, the second pdx-1 exon, including the homeobox, was replaced by a neomycin resistance cassette. In the second, a lacZ reporter was fused in-frame with the N terminus of PDX-1, replacing most of the homeodomain. Neonatal pdx-1–/– mice are apancreatic, in confirmation of previous reports (Jonsson, J., Carlsson, L., Edlund, T. and Edlund, H. (1994). Nature 371, 606-609). However, the pancreatic buds do form in homozygous mutants, and the dorsal bud undergoes limited proliferation and outgrowth to form a small, irregularly branched, ductular tree. This outgrowth does not contain insulin or amylase-positive cells, but glucagon-expressing cells are found. The rostral duodenum shows a local absence of the normal columnar epithelial lining, villi, and Brunner’s glands, which are replaced by a GLUT2-positive cuboidal epithelium resembling the bile duct lining. Just distal of the abnormal epithelium, the numbers of enteroendocrine cells in the villi are greatly reduced. The PDX-1/ β-galactosidase fusion allele is expressed in pancreatic and duodenal cells in the absence of functional PDX-1, with expression continuing into perinatal stages with similar boundaries and expression levels. These results offer additional insight into the role of pdx-1 in the determination and differentiation of the posterior foregut, particularly regarding the proliferation and differentiation of the pancreatic progenitors.

The location and lineage of cells that give rise to endocrine islets during embryogenesis has not been established nor has the origin or identity of adult islet stem cells. We have employed an inducible Cre-ERTM-LoxP system to indelibly mark the progeny of cells expressing either Ngn3 or Pdx1 at different stages of development. The results provide direct evidence that NGN3+ cells are islet progenitors during embryogenesis and in adult mice. In addition, we find that cells expressing Pdx1 give rise to all three types of pancreatic tissue: exocrine, endocrine and duct. Furthermore, exocrine and endocrine cells are derived from Pdx1-expressing progenitors throughout embryogenesis. By contrast, the pancreatic duct arises from PDX1+ progenitors that are set aside around embryonic day 10.5 (E9.5-E11.5). These findings suggest that lineages for exocrine, endocrine islet and duct progenitors are committed at mid-gestation.

The growth and differentiation factor transforming growth factor-β2 (TGFβ2) is thought to play important roles in multiple developmental processes. Targeted disruption of the TGF β2 gene was undertaken to determine its essential role in vivo. TGFβ2-null mice exhibit perinatal mortality and a wide range of developmental defects for a single gene disruption. These include cardiac, lung, craniofacial, limb, spinal column, eye, inner ear and urogenital defects. The developmental processes most commonly involved in the affected tissues include epithelial-mesenchymal interac-tions, cell growth, extracellular matrix production and tissue remodeling. In addition, many affected tissues have neural crest-derived components and simulate neural crest deficiencies. There is no phenotypic overlap with TGFβ1- and TGF β3-null mice indicating numerous non-compen-sated functions between the TGFβ isoforms.

The anatomy of the developing root of Arabidopsis is described using conventional histological techniques, scanning and transmission electron microscopy. The root meristem is derived from cells of the hypophysis and adjacent cells of the embryo proper. The postembryonic organization of the root is apparent in the mature embryo and is maintained in the growing primary root after germination. Cell number and location is relatively invariant in the primary root, with 8 cortical and endo-dermal cell files but more variable numbers of pericycle and epidermal cells. The organisation of cells in lateral roots is similar to that of the primary root but with more variability in the numbers of cell files in each layer. [3H]thymidine labeling of actively growing roots indicates that a quiescent centre of four central cells (derived from the hypophysis) is located between the root cap columella and the stele. This plate of four cells is surrounded by three groups of cells in, proximal, distal and lateral positions. The labeling patterns of these cells suggest that they are the initials for the files of cells that comprise the root. They give rise to four sets of cell files: the stele, the cortex and endodermis, the epidermis and lateral root-cap and the columella. A model of meristem activity is proposed based on these data. This description of Arabidopsis root structure underpins future work on the developmental genetics of root morphogenesis.

Systematic genome-wide mutagenesis screens for embryonic phenotypes have been instrumental in the understanding of invertebrate and plant development. Here, we report the results from the first application of such a large-scale genetic screening to vertebrate development.

Male zebrafish were mutagenized with N-ethyl N-nitrosourea to induce mutations in spermatogonial cells at an average specific locus rate of one in 651 mutagenized genomes. Mutations were transmitted to the F1 generation, and 2205 F2 families were raised. F3 embryos from sibling crosses within the F2 families were screened for develop-mental abnormalities. A total of 2337 mutagenized genomes were analyzed, and 2383 mutations resulting in abnormal embryonic and early larval phenotypes were identified. The phenotypes of 695 mutants indicated involvement of the identified loci in specific aspects of embryogenesis. These mutations were maintained for further characterization and were classified into categories according to their phenotypes. The analyses and genetic complementation of mutations from several categories are reported in separate manuscripts. Mutations affecting pig-mentation, motility, muscle and body shape have not been extensively analyzed and are listed here. A total of 331 mutations were tested for allelism within their respective categories. This defined 220 genetic loci with on average 1.5 alleles per locus. For about two-thirds of all loci only one allele was isolated. Therefore it is not possible to give a reliable estimate on the degree of saturation reached in our screen; however, the number of genes that can mutate to visible embryonic and early larval phenotypes in zebrafish is expected to be several-fold larger than the one for which we have observed mutant alleles during the screen. This screen demonstrates that mutations affecting a variety of developmental processes can be efficiently recovered from zebrafish.