Targets of Activity

Rhomboid plays a role in the processing of Gurken signal during oogenesis. The zinc finger transcription factor CF2 is a mediator of Egfr activated dorsoventral patterning in Drosophila oogenesis. Dorsal ventral polarity is established by the Gurken signal from the oocyte to the anterior dorsal follicle cells; this signal uses the Egf receptor pathway. CF2 is suppressed by Egfr signaling in the anterior dorsal domain of the follicular epithelium where the dorsal signal is received. In turn CF2 depletion expands the expression of dorsal genes such as rhomboid. Thus CF2 is a candidate for regulation of rhomboid in follicle cells (Hsu, 1996).

Dorsoventral (D/V) patterning in Drosophila oogenesis is initiated by the transmission of a TGF-alpha-like ligand, Gurken (Grk), from the oocyte to the anterodorsal follicle cells, activating the Egfr signaling pathway. The zinc-finger transcription factor CF2 is a negative regulator of the rhomboid (rho) gene that encodes an essential membrane-bound component of the dorsalizing pathway. Expression of CF2 itself is negatively regulated by the activated Egfr. CF2 is the target of down-regulation by the MAPK kinase cascade; this down-regulation is independent of the Rho function. These results suggest that D/V patterning involves a two-step signaling process: the initial Egfr signal (which represses CF2 and induces rho expression), and the subsequent Egfr + Rho signal, which determines the dorsal cell fates. CF2 down-regulation occurs at the post-translational level through a mechanism involving coupled cytoplasmic retention and degradation (Mantrova, 1998).

Among stage 10-12 wild-type egg chambers examined for CF2 protein expression, about 10% show elevated cytoplasmic levels of CF2 in the anterodorsal region. This pattern is particularly striking when the elevated cytoplasmic level of CF2 is contrasted to the empty anterodorsal nuclei. Such a pattern indicates that the Egfr signaling cascade may negatively regulate CF2 function by controlling its subcellular localization. In support of this hypothesis, cytoplasmic accumulation of CF2 can be enhanced in the presence of ectopically expressed constitutively active Raf. In addition, CF2 is presumed to be subject to rapid degradation, since most of the wild-type egg chambers show a complete elimination of CF2 protein in the anterodorsal follicle cells. This mechanism is just the opposite of the one at work in Dorsal protein regulation; Dorsal protein function is activated by translocating into the nucleus and the protein is stable throughout its steady state in the cytoplasm (Mantrova, 1998).

Examination of the CF2 amino acid sequence reveals one optimal consensus MAPK phosphorylation site at residue 40, where the threonine residue is presumed to be the phosphorylation target. The importance of the MAPK site was tested by site-specific mutagenesis, quantifying the effect in cultured cells. The A40 mutation does not by itself alter the stability of CF2 in the absence of activated MAPK, at least when its half-life is measured in the S2 cells. Three types of subcellular localization of CF2 are observed; nuclear (N), cytoplasmic (C), and both (N+C). When expressed alone from an actin 5c gene promoter, CF2wt protein is localized to the nucleus in over 92% of all CF2-expressing cells. In the presence of either Ras1V12 or MAPKSem, there is a significant increase in cytoplasmic accumulation of CF2wt: the percentages of cytoplasmic CF2-containing cells (N + C and C) increase to 44% with Ras1V12 and 52% with MAPKSem. In contrast, the percentages of cells expressing cytoplasmic CF2A40 mutant protein remain largely unchanged at the background level in the presence of Ras1V12 or MAPKSem (Mantrova, 1998).

It is proposed that the Egfr-mediated D/V patterning is a two-step signaling process, demarcated by the appearance of Rho function. This model is based on three previous observations: (1) down-regulation of CF2 precedes rho expression; (2) overexpression of CF2 can suppress rho expression; and (3) Egfr alone cannot induce dorsal cell fates without Rho. In this report, the Ras/MAPK signaling pathway has been placed upstream of CF2. But what are the changes in the signaling cascade, if any, brought on by Rhomboid? It has been shown that Rho by itself cannot induce dorsal fates without Egfr, but it can dorsalize the egg chambers when ectopically expressed in the ventral follicle cells, despite the fact that Egfr in the ventral follicle cells is not pre-activated by Grk. Interestingly, the level of CF2 in the ventral follicle cells is not affected by ectopically expressed Rho. This indicates that the signal induced by Rho is at least different from that of Egfr alone, with respect to CF2 regulation. It has been suggested that Rho is involved in processing a second ligand. If this is correct, then Rho may induce a signaling cascade distinct from Ras-Raf-MEK-MAPK, or may modify the specificity of the existing cascade. Indirect evidence has emerged recently that one or more signaling pathway(s) parallel to that of Egfr may in fact exist. Resolving the events downstream of Rho should be the next step in unraveling this complex developmental signaling process (Mantrova, 1998 and references).

Protein Interactions

It is speculated that Dodo might facilitate the degradation of phosphorylated CF2 by inducing a conformational change that is more accessible to ubiquitination. This was tested in the cultured Drosophila cells. A constitutively active Drosophila Ras (RasV12) transgene was present in all assay conditions because it has been demonstrated that CF2 degradation in the cultured cells requires activated Ras-MAPK signaling (Mantrova, 1998). In this background, wild-type CF2 is stabilized in the presence of the proteasome inhibitor lactacystin, whereas the levels of the A40 mutant are largely unchanged with or without lactacystin. CF2A40 has a higher mobility than wild-type CF2 in the SDS gel. This is most probably due to conformational changes because the integrity of the open reading frames of the transgenes has been verified by sequencing. The same mobility difference is also discernible with the bacteria-produced proteins. The ubiquitin-proteasome pathway in the cultured Drosophila cells is not efficient. This is evident in the experiments described here because lactacystin increases the steady-state level of wild-type CF2 by only ~twofold, and no ubiquitinated CF2 proteins were detectable in these assay conditions. To detect ubiquitinated CF2, the cultured cells were transfected with the ubiquitin coding sequence under the control of the metallothionein gene promoter. The cell extracts were then probed with anti-CF2 antibody. This detection strategy with the highly specific CF2 antibody is favored instead of immunoprecipitation with anti-CF2 antibody (or anti-ubiquitin) followed by western blotting with anti-ubiquitin (or anti-CF2) because the latter method is notoriously non-specific (Hsu, 2001).

When supplied with exogenous ubiquitin, discrete bands of larger CF2-containing complexes are detected. These bands are most probably the CF2 proteins with different extents of ubiquitination. Interestingly, the A40 mutant is ubiquitinated to a smaller extent and these ubiquitinated species are not affected with or without active proteasomes, indicating that these species are not recognized by the proteolytic machinery. In addition, their levels are not increased by added dodo. In contrast, wild-type CF2 can be ubiquitinated to a larger extent and these species are degraded when proteasomes are active. More importantly, the level of these proteasome-sensitive species is markedly increased in the presence of exogenous dodo. It is concluded that Dodo protein can indeed enhance the level of CF2 ubiquitination, resulting in increased efficiency of proteolysis (Hsu, 2001).

dodo: Biological Overview | Evolutionary Homologs | Developmental Biology | Effects of Mutation | References

Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.