Part-of relations in anatomy ontologies: a proposal for RDFS and OWL formalisations.

Part-of relations are central to anatomy. However, the definition, formalisation and use of part-of in anatomy ontologies is problematic. This paper surveys existing formal approaches, as well as the use of part-of in the Open Biological Ontologies (OBO) anatomies of model species. Based on this analysis, we propose a minimal ontology for anatomy which is expressed in the Semantic Web languages RDFS and OWL-Full. The paper concludes with a description of the context of this work in capturing cross-species tissue homologies and analogies.

Early nephron formation in the developing mouse kidney.

This paper reports 3-dimensional confocal microscopy observations on how nephrogenic aggregates form from the NCAM- and Pax2-positive caps (4-5 cells deep) of condensed metanephric mesenchyme surrounding the duct tips of the mouse kidney. Aggregates of 6-8 cells are first seen at approximately E12.5-12.75 immediately proximal to this cap, closely abutting the duct surface. As the tip advances, NCAM expression is maintained in the cap but is otherwise restricted to aggregates whose cells rapidly epithelialise, forming tubules that invade the duct epithelium. Pax2 expression studies shows how the rind of nephrogenic blastemal cells forms: as duct tips extend towards the kidney surface, the associated Pax2+ cells form patches of cells on the kidney surface. These observations revise our knowledge of the timing and process of nephron initiation.

Expression and genomic characterization of protein phosphatase inhibitor-1: a novel marker for mesothelium in the mouse.

Protein phosphatase inhibitor-1 (inhibitor-1 or I-1) is involved in signal transduction and is an endogenous inhibitor of protein phosphatase-1. The mouse I-1 protein sequence has been deduced from cDNA and is strongly homologous to the published rat sequence. A mouse genomic library was screened, and the I-1 gene was characterized and localized by fluorescent in situ hybridization (FISH) to chromosome 15F. Protein expression in a range of embryonic and adult tissue was analysed using confocal microscopy. Inhibitor-1 is expressed by: the coelomic epithelium; the epithelial bounding layer of cells of the kidney, lung, liver, heart, intestine and gonad; and the surface ectoderm. The blast cells of the kidney do not express I-1. We conclude that I-1 is a marker for mesothelium.

A bioinformatics approach to investigating developmental pathways in the kidney and other tissues.

Over the past few years, large amounts of data linking gene-expression (GE) patterns and other genetic data with the development of the mouse kidney have been published, and the next task will be to integrate these data with the molecular networks responsible for the emergence of the kidney phenotype. This paper discusses how a start to this task can be made by using the kidney database and its associated search tools, and shows how the data generated by such an approach can be used as a guide to future experimentation. Many of the events taking place as the kidney develops do, of course, also take place in other tissues and organisms and it will soon be possible to incorporate relevant information from these systems into analyses of kidney data as well as the new information from microarray technology. The key to success here will be the ability to access over the internet data from the textual and graphical databases for the mouse and other organisms now being established. In order to do this, informatic tools will be needed that will allow a user working with one database to query another. This paper also considers both the types of tools that will be necessary and the databases on which they will operate.

Abnormal collagen assembly, though normal phenotype, in alginate bead cultures of chick embryo chondrocytes.

The collagens produced by chick embryo chondrocytes cultured in alginate beads were investigated both biochemically and ultrastructurally. The cartilage phenotype is maintained for at least 14 days, as indicated by the production of the cartilage-specific collagens II, IX, and XI and the absence of collagen I. There were differences in the distributions of collagens among the three different compartments analyzed (cells and their associated matrix, further-removed matrix (released by alginate solubilization), and culture medium), with large amounts of collagen IX (mainly in proteoglycan form) in the culture medium. Inhibition of lysyl oxidase activity by beta-aminopropionitrile led to an overall decrease in collagen production. In contrast to the biochemical observations, collagen ultrastructure in the extracellular matrix of alginate cultures was not in the form of the expected 64-nm banded fibrils, but rather in the form of segment-long-spacing-like crystallites. This abnormal structure is likely to be a result of alginate disrupting normal assembly. We conclude that, in this system, the native fibrillar structure of the collagenous matrix is not essential for the maintenance of the differentiated phenotype of chondrocytes.

A three-dimensional model of the mouse at embryonic day 9.

This paper describes a digital, three-dimensional model of the mouse embryo at E9. The model was made by reconstruction from images of serial histological sections digitally warped to remove distortions and has a resolution of approximately 9 microns. The model can be digitally resectioned in any plane to provide images which resemble conventional histological sections. The main tissues have been identified and delineated by digital painting so that the anatomical components can be visualized and manipulated in 3-D surface- and volume-rendered views. This provides a three-dimensional definition of anatomy that will provide a useful tool for interpreting and understanding spatial data in mouse embryos. The anatomy of the model is discussed where it provides landmarks for interpretation and navigation or where it is unexpected in light of existing descriptions of the E9 mouse embryo. The complete anatomy is not presented in this paper but will be available on CD-ROM. A detailed description of the technical aspects of the construction of the model is included in an appendix. The model is the first of a series that will form the basis for an atlas/database of mouse development. This reconstruction and its associated anatomy are available in a variety of data formats with some supporting software from http:@genex.hgu.mrc.ac.uk/.

Developmental bioinformatics: linking genetic data to virtual embryos.

This paper discusses current efforts to produce databases of gene expression for the major model embryos used in developmental biology. The efforts to build these resources were motivated by the need for immediate internet access to all types of research data, and the production of these databases is a major and new challenge for bioinformatics. Thus far bioinformatics has mainly been concerned with textually oriented resources and data, much of it concerned with gene and protein sequences. Because the genetic basis of developmental biology is integrated with developmental anatomy, these databases require the use of images to link molecular data with spatial information. In order to standardise database formats, digital atlases of some model systems are being produced that include integrated anatomical descriptions and these are being linked to appropriate genetic data. Integrating such image-based, searchable data into databases makes new demands on the field of bioinformatics and we consider here the imaging modalities that are used to obtain information and we discuss in particular the production of 3D images from serial sections. Next, we consider how to integrate textual and spatial descriptions of gene expression and the key tool needed to make this possible, i.e. anatomical nomenclature. A short review of internet resources on developmental biology is also given and future prospects for the development of these databases are discussed.

Computer-aided 3-D reconstruction of serially sectioned mouse embryos: its use in integrating anatomical organization.

This paper reviews recent work on a project that uses a computer-aided approach for making 3-D reconstructions of serially sectioned mouse embryos (the digital mouse). The captured images are aligned using a warping program so that almost perfect alignment of adjacent sections is achieved with minimal deformation. The sections that are viewed on the computer screen are in fact computer-generated grey-level images with a resolution of about 10 microm. The reconstructed embryo may then be resectioned in any plane to simulate as near as possible an exact match on the computer screen to the viewer's own material. Individual anatomical domains may then be painted in different colors, and these domains may be selected by querying the textual database containing anatomical and other information. Further, it is now possible to generate 3-D images of individual anatomically-discrete components or related sets of components of a particular system in isolation from the rest of the embryo, or, if required, against a 'ghost-like' image of the intact embryo, or specific parts of an embryo. In the article, examples are given of the use of the system in interpreting the vascular, gut and paraxial mesoderm systems, while both the advantages and disadvantages of this approach are also discussed. The eventual aim will be to provide 3-D reconstructions of mouse embryos from fertilization up to 14 days postcoitum of development. When completed, this project will allow the accurate spatial mapping of gene-expression and cell lineage data onto the digital Atlas of normal mouse embryonic development.

Development, databases and the Internet.

There is now a rapidly expanding population of interlinked developmental biology databases on the World Wide Web that can be readily accessed from a desk-top PC using programs such as Netscape or Mosaic. These databases cover popular organisms (Arabidopsis, Caenorhabditis, Drosophila, zebrafish, mouse, etc.) and include gene and protein sequences, lists of mutants, information on resources and techniques, and teaching aids. More complex are databases relating domains of gene expression to embryonic anatomy and these range from existing text-based systems for specific organs such as kidney, to a massive project under development, that will cover gene expression during the whole of mouse embryogenesis. In this brief article, we review selected examples of databases currently available, look forward to what will be available soon, and explain how to gain access to the World Wide Web.

The mouse 14-3-3 epsilon isoform, a kinase regulator whose expression pattern is modulated in mesenchyme and neuronal differentiation.

Kidney development is a complex, little understood process based on inductive interactions and intricate epithelial and mesenchymal morphogenesis. Here, we report the use of subtractive hybridization to clone cDNAs expressed in early nephrogenesis. cDNA made from E14.5 mouse kidney was hybridized with adult mouse liver mRNA employing a technique based on labeling the driver mRNA with photoactivatable biotin and using streptavidin to remove RNA:cDNA complexes. An aliquot of the unhybridized cDNA identified several clones including three isolates that proved to be the epsilon isoform of the 14-3-3 gene family that is, among other functions, implicated in protein kinase C regulation. Northern blot analysis showed a 2.0-kb transcript widely present in mouse embryos from E7.5 onward, but, as expected from the subtractive strategy, absent in adult liver. In situ hybridization was carried out on mouse embryos aged E8.5 to E15.5. These showed that, in the E8.5 embryo, the 14-3-3 epsilon gene was expressed throughout the embryo, but that, within a day, expression was more marked in mesenchyme than elsewhere (e.g., epithelial tissue, where it was generally low), although levels in neural tissue rose again by about E12.5. This difference was maintained until E15.5 when expression levels started to drop in most tissues, with those of the nervous system, tooth, and kidney being exceptions. Perhaps the most intriguing feature of the expression pattern, however, was that, while the gene was strongly expressed in early mesenchyme, the level of expression decreased as the mesenchyme differentiated. This change was particularly noted in mesenchymal condensations that would become cartilage, bone, and myotome-derived muscle, in the presumptive muscle layer of the gut, and in the kidney. In this last case, the gene was strongly expressed in stem cells and mesenchyme, but expression levels dropped markedly as early nephrogenic condensates epithelialized. The results as a whole thus argue for the 14-3-3 epsilon isoform playing roles in neural development and in early mesenchyme, with this latter function being lost or replaced as the tissue differentiates.

Towards a genetic basis for kidney development.

Although it is not easy to investigate the regulatory basis of developmental processes in most mammalian tissues, the mouse kidney has several distinct advantages as a model system. Its development involves a wide variety of developmental processes that include induction, stem-cell regulation, a mesenchyme-to-epithelium transition, epithelial morphogenesis and pattern formation. Further, there are several genetic disorders associated with its development, much of nephrogenesis will take place in vitro and a significant start has been made in elucidating the regulatory molecules involved in its ontogeny. Here, we summarise current knowledge on how the various aspects of kidney development are controlled at the genetic level. For this, we have compiled a table showing when and where the more than forty regulatory genes thus far identified are expressed during nephrogenesis (this table being a subset of a database also containing information on structural and functional proteins expressed during nephrogenesis). The data on the regulatory genes demonstrate, in particular, the importance of the Wilms' tumour gene, WT1, in nephrogenesis, the growth-stimulating interaction between the hepatocyte growth factor and its receptor, c-met, and the differences between uninduced and induced metanephric mesenchyme. In an attempt to highlight those stable developmental pathways which underpin the formation of the kidney and to facilitate future work, we have identified possible checkpoints occurring during nephrogenesis (stages at which a positive signal is needed for development to continue). The data to hand suggest that such checkpoints occur when metanephric mesenchyme is established in the intermediate mesoderm, when induction takes place, when stem cells are activated and before mesenchyme aggregates to form nephrogenic condensations.(ABSTRACT TRUNCATED AT 250 WORDS)

Chick corneal development in vitro: diverse effects of pH on collagen assembly.

In vivo, the embryonic chick corneal epithelium lays down a stroma of collagen and proteoglycans whose fibrils are unusual as their diameter distribution peaks sharply about a mean of 20 nm. Such epithelia cultured on Nuclepore filters will also lay down a stroma containing 20 nm diameter fibrils, although there is only limited orthogonal organisation. We report here that collagen fibril morphology is critically dependent on the pH of the medium in which the corneal epithelium is cultured and that normal 20 nm diameter fibrils only assemble in a narrow band around neutral pH (approx. 6.9-7.4). At higher pH (7.6-8.1), fibrils in the distal region of the stroma more closely resemble those seen in non-corneal stroma as their diameters can be up to 200 nm even though fibrils near the basal lamina are only about 10 nm in diameter. At low pH (approx. 6.5), there are again wide fibrils, but with the hieroglyphic cross-sections typical of those seen in heritable disorders of N-terminal procollagen processing. Biochemical analysis by SDS-PAGE and fluorography confirms that N-terminal procollagen processing is deficient at this pH. At very low pH (approx. 5.8-6.2), there is little processing of procollagen and the stroma comprises filamentous material with the occasional banded structures typical of those formed by unprocessed procollagen at high concentration. Gel electrophoresis and peptide mapping showed that the collagens produced by the corneal epithelium of the primary stroma included types I, II and V and that total collagen production, as assessed by incorporation of [3H]proline, increased with pH, although the relative amounts of the different collagens produced remained essentially unchanged. While the biochemical data can account for the altered morphologies in the pH range 5.8 to 7.0, the sensitivity of fibril diameter to small changes around neutral pH remains unexplained, but points to the subtle, charge-based interactions necessary for the formation of 20 nm diameter fibrils in the developing cornea.

Embryonic kidney rudiments grown in adult mice fail to mimic the Wilms' phenotype, but show strain-specific morphogenesis.

12.5- to 14.5-day mouse embryonic kidney fragments were grown syngeneically under kidney or testis capsules in 3 mouse strains for 5-26 weeks to reproduce the Wilms' tumour (WT) phenotype previously reported. About 65% of the transplants grew, but none showed true WT morphology, invaded local tissue or developed metastases. In Swiss mice, most growths contained disorganized stromal and blast cells, but also had more mature structures. CBA growths had cysts, differentiated nephrons and glomeruli, while 129/SV growths gave both types of morphology. In situ mRNA hybridization using the Wilms' tumour predisposition gene (WT1) showed that, unlike the initial rudiments and WT, expression was limited to the glomeruli. The unexpected absence of expression by the apparent blast cells in the growths implies that the system is not a model for WT.

The expression of the Wilms' tumour gene, WT1, in the developing mammalian embryo.

In the developing mouse, the Wilms' tumour gene, WT1, is first expressed in the intermediate mesenchyme lateral to the coelomic cavity (13 somite, early 9 dpc embryo). A few hours later, it is present around all the cavity and in the urogenital ridge (the earliest mesonephric tubules) and the differentiating heart mesothelium. By 11 dpc, expression is in the uninduced metanephric mesenchyme and in the presumptive motor neurons of the spinal cord. By 12.5 dpc, WT1 expression has increased in the induced mesenchyme of the kidney and a day later is particularly marked in the nephrogenic condensations. At 13.5 dpc, WT1 is briefly expressed in some differentiating body-wall musculature, while two days later, there is a small domain of expression in the roof of the fourth ventricle of the brain. By day 20, however, expression has become restricted to the kidney glomeruli. RNA-PCR analysis on 12.5 dpc embryos and on adult tissues shows that WT1 is weakly expressed in both eye and tongue. The expression pattern in human embryos (28-70 days) is very similar to that in the equivalent mouse stages (10-15 dpc). The results indicate that WT1 is mainly present in mesodermally derived tissues, although exceptions are ectodermally derived spinal cord and brain. The data indicate that WT1 plays a role in mediating some cases of the mesenchyme-to-epithelial transition, but its expression elsewhere argues that it has other tissue-specific roles in development.