Developmental Genetics

Overview

Figure 1.

Figure 1.

zoom

The vertebrate body plan is established during embryogenesis through the coordinated actions of signaling pathways involved in germ layer induction, patterning, and morphogenesis. Once the cells that will give rise to the three germ layers are induced and patterned, it is through the movements of gastrulation that these specified cells are arranged into an embryonic axis. An important feature of gastrulation, in addition to the morphogenetic movements, is the signalling activity of the organizer where the dorsal mesodermal precursors are positioned.

In our lab we are addressing the study of the body axis specification by embryological and genetic approaches using zebrafish as a model organism. Given the main role of members of the TGFβ family in the specification of the body axis and the organizer we are givin special attention to the activity of these family of proteins during these stages of embryogenesis. In addition to this we have been traditionaly working on the development of the notochord, a tissue derived from the dorsal most part of the early embryo specified by the activity of the organizer.

Specification, maintenance and regeneration of the Organizer

Embryological approach

Organizer formation and activity were first studied in amphibian embryos and results primarily from experiments in Xenopus have led to a model where the overlap of dorsal determinants and mesoderm inducers in the dorsal vegetal Nieuwkoop Center induces the organizer from dorsal ectoderm (Nieuwkoop, 1969; Gerhart, 1999). The Nieuwkoop Center acts as early as the 64-cell stage in frogs whereas the organizer acts much later during gastrulation (Gimlich and Gerhart, 1984). Because the organizer is present during a stage of very extensive morphogenetic movements, it raises the question of whether the organizer is a fixed population of cells induced at a discrete stage in development under the action of the Nieuwkoop Center or whether it is a dynamic population of cells constantly being induced and thereby maintained.

Results from chick embryos show that the organizer does not consist of a fixed population of cells and that new organizer tissue is constantly induced during gastrulation (Joubin and Stern, 1999). The mechanism for this constant induction is an interplay of inducers and feedback inhibitors, consisting of members of the TGFβ , BMP, and Wnt signalling pathways, present in different parts of the embryo that act to position the organizer at the anterior tip of the primitive streak throughout gastrulation. During normal development, ADMP present in the organizer prevents surrounding areas from acquiring organizer properties. When the organizer is experimentally ablated, the source of inhibitory ADMP is removed, and a new organizer can be induced in neighboring cells by Vg1 and Wnt8 expressed in the middle of the primitive streak. This regeneration of the organizer in amniotes is an example of regulative development, where inducers and feedback inhibitors enable communication between different parts of the embryo. A similar strategy is used in the establishment of the chick primitive streak. Even when the embryo consists of thousands of cells, the blastoderm is totipotent in its ability to initiate an embryonic axis, so that if the embryo is cut into multiple pieces a primitive streak can form in each (Lutz, 1948; Spratt and Haas, 1960; 1961a; 1961b). This occurs because the developing primitive streak emits antagonistic influences, which include Cerberus and Lefty, which inhibit axis formation and prevent the formation of more than one axis during the initial polarization of the blastoderm (Khaner and Eyal-Giladi, 1986; 1989). When the embryo is cut, these non-streak regions are relieved from the effects of the inhibitors and can form a primitive streak. Regulative development is in contrast to the mosaic development observed in amphibian embryos where localized determinants control the dorsal ventral axis. Already at the 4-cell stage, the dorsal and ventral halves are specified (Cooke and Webber, 1985a). However, even in Xenopus, tissue can be respecified suggesting that the reason for the mosaic development is the segregation of dorsal determinants, not due to a lack of competence to respond or a commitment to a dorsal fate. Also, in frogs, there is evidence of "communication" between dorsal and ventral halves mediated by BMP family members and their inhibitors (Reversade and De Robertis, 2005). Still, embryological studies in Xenopus which directly addressed organizer regeneration following experimental ablation showed that regeneration does not occur and that the organizer consists of a fixed population of signalling cells (Cooke, 1985).

In fish, variable results and conclusions have been reported with respect to organizer regeneration. In our lab we developed an efficient micropipette-based transplating method. By using this method we found that when one completely removes the organizer region, the resulting embryos completely lack axial mesendoderm. By contrast, if only the morphological shield is removed, the resulting embryos regenerate the missing tissue and develop normally. Such morphological shields are sufficient to induce the formation of a complete secondary axis upon transplantation to the ventral or lateral regions of a host embryo. Finally, by subdividing the shield into deep (goosecoid expressing) and superficial (floating head expressing) pieces and separately transplanting fragments, we found that the organizer activity can be divided into separable head-inducing and trunk/tail-inducing activities (Saude and Stemple., 2000).

The lack of organizer regeneration have been described previously (Shih and Fraser, 1996; Brummet, 1968a; 1968b) but also the opposite was described in old embriology experiments (Oppenheimer 1934; 1936a????). Similarly, with respect to the establishment of the dorsal-ventral axis, different conclusions have been reached over the degree of plasticity and multipotency. (Aanstad and Whitaker, 1999) found that there is a difference in Li sensitivity at the 32-cell stage suggesting that dorsal tissue is specified at this time. However, in lineage tracing experiments, labelled blastomeres disperse randomly up until gastrulation demonstrating that at these early stages, there is not a correlation between position and ultimate fate (Kimmel and Law, 1985; Kimmel et al., 1990; Helde et al., 1994). Similarly, Ho and Kimmel (1993) found that the transplantation of blastula and early gastrula stage cells into new environments always results in the graft's adoption of a fate appropriate to the new environment. Moreover, at least in trout, blastula stage blastoderms can be cut into quarters and each piece can regulate to form an embryonic axis (Luther, 1936; Luther, 1937; Devillers, 1951) Ultimately, the events of gastrulation in all vertebrates generate similar body plans and organizations raising the question of why different strategies are adopted with respect to such an important signalling center. In addition to obvious differences in the geometries and temporal development of different embryos, what are the features of gastrulation that are different and what does this reflect about how gastrulation and its signalling events occur differently?

We will adress these questions and discrepancies by a more accurate embryo manipulation and a deep semi-quantitative analysis of the organizer specification, maintenance, activity and regeneration.

We will first establish more accurately and semi-quantitatively when and where organizer activity is present in the embryo:

Figure 2.

Figure 2.

zoom

- By transplanting cells from different regions around the organizer of a donor shield stage embryo to the ventral side of a host shield stage embryo, we will address where organizer activity is pressent.

Figure 3.

Figure 3.

zoom

- By transplanting cells from different stage embryos (30%, shield and 65% epiboly) into the ventral side of a host shield stage embryo and detemining posteriorly its origin by labelling the dorsal region of the donor we will address the temporal activity of the organizer.

Genetic approach

Dorsal organizer specification depends on nodal signalling and one aspect of our research has been to investigate how the activity of these potent secreted proteins is modulated during early development. Chen and Schier showed that the two zebrafish Nodals, Squint and Cyclops, have fundamentally different ranges of activity with Squint acting at a distance to induce mesoderm, and Cyclops only able to act locally. We find that when the two Nodal antagonists, Lefty-1 and Lefty-2, are simultaneously disrupted, unchecked Nodal signalling results in a severe gastrulation phenotype often lethal by 24 hours of development. By sharp contrast, when the Lefties are disrupted in squint homozygous mutants, the resulting embryos are able to gastrulate and form a complete embryonic axis. Thus Lefties play a much more crucial role in restricting the activity of the long-range Squint than for the short-range Cyclops molecule (Feldman et al., 2002).

TGFβ proteins in early development

Transforming Growth Factor Beta (TGFβ ) proteins are secreted cytokines that play diverse roles in development, growth and disease through both auto- and paracrine actions. They are a diverse and highly conserved superfamily of genes organised into many subfamilies which include the sensu stricto TGFβ s, Bone Morphogenetic Proteins (BMPs), Nodals, Inhibin/Activins, Growth and Differentiation Factors (GDFs) and Myostatins. Secreted TGFβ ligands bind and activate serine-threonine kinase receptors causing an intracellular cascade of phosphorylation that leads to the translocation of transcription factors to the nucleus and the regulation of specific target genes. Complex regulation exists at every stage of the TGFβ pathway including ligand production, processing, secretion, transport and receptor binding as well as the intracellular interpretation of TGFβ signalling. We study the TGFβ s during the early stages of development of zebrafish embryos in order to uncover their roles in specific developmental processes, as well as attempting to better understand their activity, regulation, molecular interactions and transcriptional targets.

Heat shifts induce dysmorphology in Sqt-deficient embryos.

Figure 4.

Figure 4.

zoom

Recent work has included an analysis of the environmental and genetic factors affecting the phenotypes associated with loss of squint (sqt) function (Pei et al., 2007). Sqt is a member of the nodal-related subfamily which regulate the specification of the mesodermal and endodermal germ layers during gastrulation (Dougan et al., 2003, Feldman et al., 1998, Chen and Schier, 2001, Amsterdam et al., 2004), although the extent of the maternal contribution towards determining the dorsal axis remains unclear (Pei et al., 2007, Bennet et al., 2007, Gore et al., 2007). There is a variable requirement for sqt protein in embryogenesis, with two independent mutant lines, both null for sqt, producing a similar spectrum of phenotypes ranging from severe cyclopia to viable maternal-zygotic null fish. Experiments to determine potential environmental conditions that confer an increased requirement for intact sqt signalling demonstrate that alterations of temperature that do not affect WT embryos significantly increase phenotypic penetrance in sqt embryos. Further work demonstrates an association between sqt penetrance and HSP90 function, indicating that Sqt, HSP90A and HSP90B act together under normal developmental conditions to protect embryos against cyclopia.

Novel roles for inhβb in otic vesicle development.

Figure 5.

Figure 5.

zoom

Ongoing work includes analysis of the roles of inhibin beta B (inhβb) in the development of the otic vesicle. inhβb dimerises with other Inhibin beta proteins to form an Activin, or with the Inhibin alpha protein to form an Inhibin. Disruption of inhβb production in Xenopus affects the induction and patterning of mesoderm in a graded fashion, (Piepenburg et al., 2004). By using low doses of inhβb Morpholino (MO), to partially escape the early phenotype, we are uncovering new roles for this gene in later stages of development. Expression of inhβb within the otic vesicle is restricted to a few cells at the anterior end of the vesicle, likely corresponding to the anterior macula. Embryos with decreased inhβb production have impaired otic vesicle development and increased expression of follistatin - an extracellular activin binding protein which regulates its signalling. Morphological changes include epithelial lesions in the otic vesicles of inhβb morphants, suggesting a patterning role for inhβb in otic vesicle development.

Notochord Development

Background

Figure 6.

Figure 6.

zoom

The notochord is an essential organ for vertebrate development. It serves both as a skeletal element and as a source of signals that pattern surrounding tissues. Systematic genetic screens, to identify recessive zygotic-effect mutations affecting embryogenesis, have led to the identification of many genes affecting notochord development (Stemple DL, 2005). Generally speaking, mutations in these loci prevent notochord differentiation, leading to embryos of significantly diminished stature. Of the seven mutant loci most intensively studied in our lab, we have successfully determined the identity of six of the affected genes and anticipate identifying the seventh in the near future.

The Laminin Mutants

Figure 7.

Figure 7.

zoom

Positional cloning revealed that three loci, bashful, grumpy and sleepy, encode individual α1, β1, and γ1 chains of the extracellular matrix protein Laminin-1. Despite many in vitro studies, there are surprisingly few details known of the in vivo requirements for the major isoforms of Laminin. We found that the failure to produce either the β1 or γ1 chain results in complete loss of Laminin-1 immunoreactivity and a widespread disruption of basement membrane. In particular, the basement membrane surrounding the notochord fails to form and notochord cells fail to differentiate (Parsons et al., 2002b).

Efforts to understand how the Laminin/basement membrane notochord-differentiation signal is transduced, led us to disrupt several Laminin receptors expressed in chordamesoderm. Disruption of one such receptor, Dystroglycan, causes a severe form of muscular dystrophy in zebrafish embryos and has opened a new research area for the laboratory (Parsons et al., 2002a). Although we have yet to identify the notochord differentiation signal, positional cloning of the remaining locus, doc, should soon provide us with clues as to the nature of the signal.

Figure 8.

Figure 8.

zoom

The COPI Mutants

Positional cloning revealed that another three loci, sneezy, dopey and happy, encode 3 subunits of the COPI vesicular coat complex. The COPI complex is likely to be essential for all eukaryotic cell as it is responsible for maintenance of the Golgi apparatus and the recycling of transport machinery through the Golgi and back to the endoplasmic reticulum. Why then do the zebrafish COPI mutants have such a specific notochord defect? We find that the maternal supply of COPI components is sufficient for most cells during the first two days of embryogenesis, but the notochord has an increased demand for COPI mediated transport. We find that COPI mRNAs are specifically up regulated in the chordamesoderm just prior to notochord differentiation. As with the Laminin mutants, we find that the COPI mutants also fail to develop a peri-notochordal basement membrane. This is an autonomous defect of notochord cells; hence it most likely results from a failure to transport the necessary basement membrane components. This basement membrane failure may underlie the similarities in the notochord differentiation defect seen in both the Laminin and COPI classes of mutant (Coutinho, et al., 2004).

References

  • Environmental and genetic modifiers of squint penetrance during zebrafish embryogenesis.

    Pei W, Williams PH, Clark MD, Stemple DL and Feldman B

    Developmental biology 2007;308;2;368-78

  • Essential and overlapping roles for laminin alpha chains in notochord and blood vessel formation.

    Pollard SM, Parsons MJ, Kamei M, Kettleborough RN, Thomas KA, Pham VN, Bae MK, Scott A, Weinstein BM and Stemple DL

    Developmental biology 2006;289;1;64-76

  • Structure and function of the notochord: an essential organ for chordate development.

    Stemple DL

    Development (Cambridge, England) 2005;132;11;2503-12

  • Differential requirements for COPI transport during vertebrate early development.

    Coutinho P, Parsons MJ, Thomas KA, Hirst EM, Saúde L, Campos I, Williams PH and Stemple DL

    Developmental cell 2004;7;4;547-58

  • Lefty antagonism of Squint is essential for normal gastrulation.

    Feldman B, Concha ML, Saúde L, Parsons MJ, Adams RJ, Wilson SW and Stemple DL

    Current biology : CB 2002;12;24;2129-35

  • Zebrafish mutants identify an essential role for laminins in notochord formation.

    Parsons MJ, Pollard SM, Saúde L, Feldman B, Coutinho P, Hirst EM and Stemple DL

    Development (Cambridge, England) 2002;129;13;3137-46

  • Axis-inducing activities and cell fates of the zebrafish organizer.

    Saúde L, Woolley K, Martin P, Driever W and Stemple DL

    Development (Cambridge, England) 2000;127;16;3407-17

* quick link - http://q.sanger.ac.uk/uc9fs23v