INTRODUCTION:

 

            Gastrulation is the process that turns a body of cells into an embryo consisting of germ layers. It is in fact the process that puts all the different cell types in a blastula into their appropriate context for further specification and differentiation. In short, gastrulation is responsible for arranging the different cell types of a blastula in a three-dimensional orientation that establishes the proper signalling environment for each cell type to differentiate along its prospective fate. This massive ‘cell-moving’ operation is best understood in amphibians, particularly so in Xenopus laevis, the African clawed frog. Amphibian gastrulation is better understood than mammalian gastrulation because of the external mode of fertilization that is characteristic to the majority of amphibians. External fertilization leads to exposed, ex-utero (outside the female’s body) development, allowing easy access to the developing embryos for invasive experimental procedures (making explants, transplants) as well as for observations.

            The Xenopus blastula, just like the oocyte consists of a pigmented animal hemisphere and an un-pigmented, yolk laden vegetal hemisphere. The animal hemisphere of the blastula contains a large fluid filled cavity occupying the majority of its volume, called blastocoel. The blastocoel is enclosed by a cup shaped blastocoel roof (BCR), made of a few layers (~3-5) of pigmented cells held together in an epithelial-like (lacking a true basement membrane) assemblage. The floor of the blastula is composed of the large, yolky, un-pigmented cells of the vegetal hemisphere. Most of the pigmented cells of the blastocoel roof give rise to the ectoderm. The endoderm on the other hand is derived from the large yolky vegetal cells as well as from the superficial cells of the equatorial region. The mesoderm derives from the cells that exist as a belt forming the edges of the ‘cup-shaped’ blastocoel roof, and a ring along the edges of the blastocoel floor. Part of the mesoderm (axial and paraxial) also derives from a patch of superficial cells in the dorsal equatorial region (1, 2). At the onset of gastrulation, a depression develops on the dorsal side, at the interface between the large yolky cells of the vegetal hemisphere and the small superficial equatorial cells (superficial marginal zone or IMZ). This depression is soon highlighted by pigment as the pigmented cells higher up start to migrate towards it, at the same time this depression starts to expand laterally like a crescent whose two ends meet on the ventral side eventually forms a ring around the vegetal cells called blastopore. As this depression deepens on the dorsal side to form a pit, the downward migrating marginal cells form an overhang over the depression called the dorsal lip of blastopore. Under the dorsal lip of blastopore, the pit further deepens into a cavity that eventually replaces the blastocoel, called the Archenteron (primitive gut).

            The deepening of the archenteron is preceded by a massive upwelling of the large yolky vegetal cells called the Vegetal Rotation. Vegetal rotation is strongest in the region just beneath the dorsal lip of blastopore. It is here that the upwelling vegetal cells shear the local mesodermal-belt in such a way that it gets rolled up and pressed against the blastocoel roof (BCR). The extracellular matrix (ECM) lining the blastocoel roof forms a barrier between the mesoderm at the lower edges of the BCR, and the mesoderm at the periphery of the blastocoel floor (BCF). This tissue-separation results in the formation of a cleft called Brachet’s cleft (1, 3). In addition to the mechanical force of the upwelling vegetal cells that apposes mesoderm of the BCF against the BCR; mesodermal cells are also believed to be attracted to the BCR by a mechanism not as yet fully understood. As the vegetal cells move further up against the BCR, they merge with the mesodermal cells at their leading edge to form the Mesendodermal leading edge. Mesendodermal cells migrate along the BCR towards the animal pole by interacting with the fibronectin (FN) matrix. This mesendodermal migration also serves to enlarge the archenteron. This migration can be blocked by RGD (arginine, glycine, aspartic acid) peptide that competes with FN for interaction with the cellular integrin receptors that facilitate this migration. However, even in the presence of RGD peptide, tissue separation at the Brachet’s cleft is unaffected. As the mesendodermal cells migrate towards the animal pole, eventually all the mesoderm is rolled under the Brachet’s cleft and apposed against the ECM of the BCR. This process of mesoderm apposition and migration along the BCR is called Involution.

            Vegetal rotation is also exhibited by BCR-less embryos in which BCR has been removed at the level of BCF. These embryos however, do not develop a Brachet’s cleft; they rather develop a somewhat flat surface where in an intact embryo the Brachet’s cleft would have formed (3). This suggests that Brachet’s cleft is not formed by local tissue separation (delamination), but by apposition against the BCR (3). Similar vegetal rotation movements have been observed in mid-sagittal slices of BCR-less embryos (held between cover slips), mimicking the exact same movements & rhythm that ruled out the possibility of these movements being the artefact of slicing. In BCR-less embryos as well as the slices thereof, vegetal rotation progresses in a well defined rhythm, starting at the dorsal side where it is the strongest and slowly progressing towards the ventral side; over a 1-2 hour period (3). Furthermore, vegetal rotation has also been observed in the fragments of BCR-less embryo-slices. It is strongest in the dorsal fragments and subsequently weaker in ventral fragments as determined by the trajectories of moving cells in a set interval of time (3).

            Among the various factors suspected to be involved in vegetal rotation, Rho family of monomeric GTPases are believed to be involved in vegetal rotation. These GTPases are switch between active and inactive states by being active in a GTP bound state and becoming inactive when the GTP is cleaved in to GDP and inorganic phosphate. In their activated state, these GTPases are capable of cleaving the high-energy bond holding the third phosphate to the GTP. This is achieved through the intrinsic GTPase activity of the monomeric GTPase. The energy released by breaking the high-energy bond is used to activate effector proteins that regulate a wide variety of cellular activities ranging from signal transduction, to cytoskeletal rearrangement, to activation of transcription factors and gene expression.

            All monomeric GTPases have two kinds factors that regulate their activity, GTPase activating proteins or GAPs and Guanine exchange factors or GEFs. GAPs enhance the intrinsic GTPase activity of the Monomeric GTPase, causing it cleave its GTP into GDP and turn from active to inactive state. This has an inhibitory effect. GEFs on the other hand assist the protein in getting rid off the used-up GDP and to bind a new GTP, thus turning it from inactive to active state. In addition, there are also factors called Guanine dissociation inhibitors or GDIs that prevent the dissociation of the used-up GDP from the protein and thus rendering it unable to become active again. GAPs, GEFs and GDIs act as switches that regulate as well as cross-link various signal transduction cascades (4). XRhoA (Xenopus Rho A) is a member of Rho-family of monomeric GTPases found in Xenopus. This GTPase has already been implicated in convergent extension movements and in cytoskeletal rearrangement and cytokinesis in Xenopus (5, 6).

            Rho Activated Kinase (ROK) is a downstream effector of XRhoA that can phosphorylate a number of targets further downstream in order to transduce the signal initiated by XRhoA. Two of the targets of ROK are interesting in the context of cytoskeletal rearrangement and cell-shape changes; these are the regulatory light chain of myosin (MLC) and the myosin light-chain phosphatase (MLCP). MLC when phosphorylated causes the myosin molecules to assemble into filaments that together with f-actin form contractile stress fibres. These stress fibres can bring about cell-shape changes and cell movement. MLCP tends to remove the phosphate off of MLC, resulting in the dissociation of stress fibres and lack of cell movement. When phosphorylated by ROK, MLCP becomes inactive. Thus ROK promotes the formation of stress fibres and cell motility by phosphorylating MLC (activating) and MLCP (inactivating) (13, 14).

            Besides MLC and MLCP, ROK also phosphorylates another kinase called Lim kinase or LIMK. This turns LIMK on and it in turn phosphorylates an actin dissociation factor called cofilin. Phosphorylation of cofilin turns it off and thus prevents the breakdown of actin fibres it is involved in. Therefore LIMK is an important factor in stabilizing actin fibres and thus also stress fibres that are believed to undertake the cell movements induced by RhoA signalling (13, 14). In addition to ROK, LIMK can also be activated by p21 activated kinase or PAK (14). Moreover, there is also a formin homology protein Dia, which is activated by Rho and plays a role in the growth of actin fibres by synergizing with another actin polymerization promoter, profilin.

            The results of our earlier experiments (last term) have shown that a dominant negative version of RhoA can effectively block the upwelling of vegetal cells that initiates the mass movement of cells during gastrulation (figure 1). This inhibition of vegetal rotation is dose dependent and can produce phenotypes from mild with vegetal rotation only slightly arrested, to extreme phenotypes when the embryo is arrested at blastula stage with no further development. In all cases of vegetal-rotation-inhibition, the blastopore fails to close. Later we found that inhibiting ROK using pharmacological inhibitor also has the same effect of arresting vegetal rotation (figure 2). This suggests that although both ROK and Dia lie downstream of Rho, it is ROK that is contributes much more to the role of RhoA in vegetal rotation than Dia.