DISCUSSION:

Through our attempts to rescue the DN-RhoA phenotype, we learnt that it was not a straightforward ‘block then rescue’ phenomenon. There might be multiple context dependencies involved in how the interaction of RhoA with its downstream targets leads to the upwelling of the vegetal cell mass we call vegetal rotation.

The first attempt made to rescue the DN-Rho-phenotype involved the use of constitutively active RhoA. This molecule, due to the fact that it is always active and bypasses the regulatory pathways, can start signalling cascades at the wrong places and at wrong times that can easily lead to wide range chaos within the embryo or at least in the region affected by the injection. The result as we saw was death of embryos at higher doses and significant tissue damage and necrosis at lower doses. Since we still do not know every signalling pathway that a versatile molecule such as Rho might be involved in, it was already risky to tinker with its signalling networks by introducing a form that bypasses regulation. However, if it was not for these experiments, we could never know if or not it would work.

As for the use of wildtype Rho in rescue, one might think it can overwhelm the dominant negative Rho present in a cell on the same principle as the dominant negative Rho overwhelms endogenous Rho, by competing with it in binding & interacting with the molecules endogenous Rho interacts with. However, there can be many problems with this approach as well. First of all any exogenous RNA is an extra load on the cells’ protein synthesizing machinery. This means exogenous RNA competes with the cellular RNAs for the same translating machinery that makes vital housekeeping proteins. A common observation that exogenous-RNA-injected embryos develop slower than the wildtype embryos can easily be the result of the extra load on the cells’ protein synthesizing system. Considering the fact that dominant negative Rho inhibits the endogenous Rho in orchestrating vegetal rotation movements only at very high doses (starting with 250pg/Blastomere at 4-cell stage), it is obvious that the amount of wildtype Rho required to counter the effect of the dominant negative should be proportionately as much more than the dominant negative as it is more than the cellular Rho. This means a huge amount of wtRho RNA injected into the embryo. Given the death rate as high as 40% of the injected embryos in our trials (see chart 2, results section) at only three times more wildtype RNA than the dominant negative used, any more RNA would certainly be more detrimental. However, we do not know at this point what exactly caused the death and necrosis in embryos, a high dose of dominant negative, or that of the wild type, or both. In earlier experiments (last term), it was found that high doses of dominant negative Rho RNA (more than 350pg/blas), caused tissue necrosis and death of embryos, but the death rate was never more than 15-20%. Although using higher dose of dominant negative RNA makes the difference between partially rescued and inhibited embryos more apparent (chart 2, results), it would be a better approach to use no more dominant negative than the minimum effective amount, and instead to increase the amount of wildtype Rho RNA.

Nonetheless, a high dose of wildtype Rho might also cause aberrant cytoskeletal dynamics and signal transduction, leading to cell death and tissue necrosis. A 2004 screen for FGF targets found a G-protein coupled receptor (GPCR4) that produced phenotypes resembling DN-RhoA phenotype (at later developmental stages), when both blocked or overexpressed (9). The blastopore in both cases failed to close. This suggests a possible involvement of the GPCR in controlling vegetal rotation, but it cannot be ascertained unless the embryos were sectioned at between stages 10.5 to 11, when the vegetal rotation actually occurs. LPA-GPCR signalling is a candidate pathway for upstream activation of RhoA. LPA (lysophosphatidic acid) is a modified phospholipid that acts as a ligand, binding to a G-protein-coupled receptor (GPCR). The GPCR in turn activates an associated G-protein. The α-subunit of that G-protein activates a Rho-GEF, which in turn leads to the activation of Rho itself. Most LPA-GPCR pathways though involve structural analogs of LPA that serve as ligands, rather than LPA itself. An example of this generic pathway in Xenopus is that of XLPA₂ (Xenopus LPA), which is a GPCR discovered in a recent study where overexpression of XLPA₂ improved wound healing (5). However, in embryos coinjected with DN-XRhoA mRNA along with XLPA₂ mRNA, there was no improvement in wound healing (5). This lead to the conclusion that XRhoA might be downstream of XLPA₂ in that context. There could be a similar pathway where GPCR4 might be an upstream activator of RhoA. However, the most important lesson to learn from the GPCR4 inhibition and overexpression study is that the signals that orchestrate gastrulation are highly regulated and any imbalance due to inhibition or overexpression of the players involved can lead to similar morphogenetic defects. Therefore it is not very unlikely for a very high dose of wildtype Rho to also screw things up as the dominant negative does. Hence a complete rescue of the DN-XRhoA phenotype might be immensely difficult if not impossible.

The results of myosin inhibition using Blebbistatin were also different from what was initially expected. Since we already know from our experiments with ROK (Rho activated kinase) inhibitor that blocking the action of ROK leads to the seizure of vegetal rotation movements, we expected myosin, which is a downstream target of ROK to have the same effect. However, since vegetal rotation did not fail on blocking myosin, it seems myosin is not the only target of ROK that is involved in vegetal rotation. While vegetal rotation in blebbistatin-treated embryos was not severely affected, archenteron formation was and it was no deeper than a small notch. I just came across another study that also used Blebbistatin to block myosin function (16). The investigators blocked myosin in order to study its role in bottle cell formation. Although they had used more than 6 times the concentration of blebbistatin we used (we used 16.7mM, they used 100mM), the bottle cells still contracted their apices to form a faint notch. This suggests that the blebbistatin concentration we used might not have been high enough to block vegetal rotation effectively. Moreover, since injecting an inhibitor into the blastocoel is a very crude procedure, and as observed, a puncture in the blastocoel roof heals slower than one in a blastomere at 4-cell stage, it is also likely that some of the inhibitor leaked out of the puncture-hole in the blastocoel roof.

Alternatively it is also possible that there are other players involved in vegetal rotation than only myosin. As mentioned earlier in the introduction, ROK also activates LIMK, which in turn inhibits cofilin, thus allowing the stabilization of actin fibres and preventing their breakdown (13, 14). Actin fibres are involved not just in forming stress fibres (actomyosin), but also form a gel-like network at the leading edge of a cell that can extend forward, through the growth of actin fibres, to produce a lamellipodium (4). Extending actin fibres at the leading edge and breaking down actin at the rear can serve as an alternate mode of locomotion to the one that involves stress fibres. Cadherins might also play a role in such movement by making and breaking contact between adjacent cells.

A study that sheds light on the importance of LIMK in gastrulation and the dynamic regulation thereof involved the overexpression and inhibition of Slingshot SSH, an a phosphatase that removes phosphates from both LIMK and cofilin, inhibiting LIMK and activating cofilin (15). In case of both overexpression and inhibition of SSH, that is reciprocally both the inhibition and loss of regulation of LIMK, gastrulation was defective and the blastopore failed to close. Since the investigators (15) did not section the embryos, there can be no remarks made about what they might have looked like from the inside. This proves that LIMK mediated actin stabilization is vital for proper gastrulation movements to occur. However, if LIMK alone can orchestrate gastrulation movements in the absence of myosin or partially so is still open to debate and nothing can be ascertained unless both myosin and LIMK are perturbed.

In future investigations, we can use higher doses of blebbistatin to determine if it blocks vegetal rotation. Alternatively explants can also be used instead of whole embryos to determine if blocking myosin function affects vegetal rotation. Moreover, we can also keep the embryos longer and observe their later development to see if the effect can later be recovered. Furthermore we can also inhibit LIMK alongside myosin to see the effect on vegetal rotation. Or we can establish ineffective doses of both myosin and LIMK, that do not show any adverse effect on vegetal rotation on their own, but when put together inhibit vegetal rotation.