rhomboid

Protein Interactions

The ventral midline provides the site for Spitz expression and processing. The Drosophila EGF receptor (DER) is activated by secreted Spitz to induce different cell fates in the ventral ectoderm. Processing of the precursor transmembrane Spitz to generate the secreted form is shown to be the limiting event. The ectodermal defects in single minded (sim) mutant embryos, in which the midline fails to develop, suggests that the ventral midline cells contribute to patterning of the ventral ectoderm. The Rhomboid and Star proteins are also expressed and required in the midline. The ectodermal defects of spitz, rhomboid or Star mutant embryos can be rescued by inducing the expression of the respective normal genes only in the midline cells. Rho and Star thus function non-autonomously, and may be required for the production or processing of the Spitz precursor. Secreted Spitz is the only sim-dependent contribution of the midline to patterning the ectoderm, since the ventral defects observed in sim mutant embryos can be overcome by expression of secreted Spitz in the ectoderm. While ectopic expression of secreted Spitz in the ectoderm or mesoderm gives rise to ventralization of the embryo, increased expression of secreted Spitz in the midline does not lead to alterations in ectoderm patterning. A mechanism for adjustment to variable levels of secreted Spitz emanating from the midline may be provided by Argos, which forms an inhibitory feedback loop for DER activation. The production of secreted Spitz in the midline, may provide a stable source for graded DER activation in the ventral ectoderm (Golembo, 1996).

Ectopic expression of rho during wing development leads to the formation of extra veins. Gene dosage studies among ventrolateral genes suggest that RHO may facilitate Spitz-EGF-R signaling, resulting in activation of ras (Sturtevant, 1993).

Levels of Drosophila EGF-R mRNA are strongly down-regulated in epidermal cells likely to have recently undergone high levels of EGF-R signaling. The cells in which EGF-R mRNA levels are down-regulated express the rhomboid gene, which is thought to locally amplify EGF-R signaling. Widespread EGF-R mRNA down-regulation can be induced by ubiquitous expression of rhomboid or by eliminating the Gap1 gene. These results suggest that cells engaged in intense EGF-R/RAS signaling limit the duration of the signal through a combination of short-acting negative feedback mechanisms such as receptor internalization followed by a longer lasting reduction in receptor transcript levels. Control of EGF-R mRNA levels by altering transcription or mRNA stability is a new tier of regulation to be considered in analysis of EGF-R signaling during development (Sturtevant, 1994).

Spatially restricted processing of Spitz may be responsible for EGF-R graded activation. The Rhomboid and Star proteins have been suggested, on the basis of genetic interactions, to act as modulators of EGF-R signaling. No alteration in EGF-R autophosphorylation or the pattern of MAP kinase activation by secreted Spitz is observed when the RHO and Star proteins are coexpressed with EGF-R in S2 cells. In embryos mutant for rho or Star the ventralizing effect of secreted Spitz is epistatic, suggesting that RHO and Star may normally facilitate processing of the Spitz precursor (Schweitzer, 1995).

Targeted expression of rhomboid in the seven 'pair-rule' stripes of paired resulted in a reiterated expansion of ventral nervous system defective (NK-2) and orthodenticle, suggesting that ectopic rhomboid is sufficient to transform neuroectodermal cells into ventral epidermis precursor cells (cells at the ventral midline the secrete the centralmost ventral denticles). This suggests that nearly all neuroectodermal cells can potentially respond to spitz signaling and activate ventral epidermal developmental pathways. The transformed cells also expressed wingless. These cells did not express single minded, indicating that rho is not required for normal expression of sim in CNS midline precursor cells. The rho gene has previously been found to play an important role in the development of ventral epidermis precursor cells and differentiation of midline neurons and glia. It now seems that rho function is not essential for early expression of sim or vnd, suggesting that initial formation of mesectodermal and ventral ectodermal cells may not require the spitz signaling pathway (Xiao, 1996).

Rhomboid encodes a protein that is localized on the apical surface of the dorsal-anterior follicle cells surrounding the oocyte. Loss of rho function causes ventralization of the eggshell and the embryo, whereas ectopic expression leads to dorsalization of both structures. Thus, spatially restricted RHO is necessary and sufficient for dorsal-ventral axis formation. It is proposed that the spatially restricted RHO protein could function in selective activation of the epidermal growth factor receptor in the dorsal follicle cells and subsequently the specification of the dorsal follicle cells (Ruohola-Baker, 1993).

Rhomboid and Star facilitate presentation and processing of the Drosophila TGF-alpha homolog Spitz

Activation of the Drosophila epidermal growth factor receptor (Egfr) by the transmembrane ligand, Spitz (Spi), requires two additional transmembrane proteins: Rhomboid and Star. Genetic evidence suggests that Rhomboid and Star facilitate Egfr signaling by processing membrane-bound Spi (mSpi) to an active, soluble form. To test this model, an assay based on Xenopus animal cap explants was used in which Spi activation of Egfr is both Rhomboid and Star dependent. Spi is on the cell surface but is kept in an inactive state by its cytoplasmic and transmembrane domains; Rhomboid and Star relieve this inhibition, allowing Spi to signal. Spi is likely to be cleaved within its transmembrane domain. However, a mutant form of mSpi that is not cleaved still signals to Egfr in a Rhomboid and Star-dependent manner. These results suggest strongly that Rhomboid and Star act primarily to present an active form of Spi to Egfr, leading secondarily to the processing of Spi into a secreted form (Bang, 2000).

Rhomboid and Star-mediated Egfr signaling was analyzed by an assay in which Xenopus animal cap explants were isolated from embryos injected with in vitro synthesized Egfr, spi, rhomboid, and Star mRNA. This assay is based on the fact that Egfr activates the Ras pathway, which should lead to an up-regulation in the expression of the Ras target gene Xenopus Brachyury (XBra) in animal caps. Animal caps from injected embryos were allowed to develop until sibling embryos were late gastrulae (stage 11.5), when they were analyzed for XBra expression by RNAse protection assay (RPA). Expression of XBra can be induced in animal caps by Egfr but only under the same conditions that are required for the activation of Egfr in Drosophila. Thus, expression of XBra is not induced in animal caps that express Egfr alone, Egfr along with mSpi, or Egfr along with just Rhomboid and Star. In contrast, a high level of XBra expression is induced when animal caps express Egfr along with mSpi, Rhomboid, and Star. The requirement for Rhomboid and Star for Egfr activation can be overcome in the animal cap assay, as in Drosophila, by expressing sSpi, an engineered form of Spi that contains just the extracellular domain. In addition, Egfr activation can be blocked, as in Drosophila, by introducing the Egfr inhibitor, Argos (Bang, 2000).

By themselves, Rhomboid and Star each weakly promote mSpi activation of Egfr, however, together they are strongly synergistic. Thus, both Rhomboid and Star may be required to achieve maximal levels of Egfr activation, but for lower levels of signaling, either one alone may be sufficient. It is possible that Rhomboid and Star are obligate cofactors but that there are homologous proteins present in the animal cap that fulfill the role of the missing component, albeit weakly. Alternatively, this result may reflect a way in which various levels of receptor activation may be achieved. In some settings, such as the Drosophila wing veins, rhomboid and Star are codependent, whereas in the eye, Star is sufficient and rhomboid function appears to be dispensible (Bang, 2000).

Next, a determination was made whether Rhomboid and Star are required for Egfr activity by acting in the signaling cell, the receiving cell, or in both cells. To do this, activation of XBra was measured in sandwiches that were made by combining an animal cap expressing Egfr with another animal cap expressing mSpi, in the presence or absence of Rhomboid and Star. When Rhomboid and Star are present in the receptor-expressing cells, mSpi fails to activate Egfr. However, when Rhomboid and Star are present in the ligand-expressing cells, mSpi strongly activates Egfr. It has been suggested that Rhomboid and Star may act as cell adhesion molecules to bring together the receptor and ligand into a cell surface complex. To test this idea, sandwiches were made in which rhomboid and Star were expressed in both the sending and receiving cells. Interestingly, this configuration attenuates the level of Egfr signaling, with the strongest repression occuring when both Rhomboid and Star are present on both sides of the sandwich. It is an intriguing possibility that an interaction between Rhomboid and/or Star in trans may dampen the level of signal received by Egfr, providing another possible mechanism by which the level of Egfr activation could be finely tuned. Together, these results argue against models in which Rhomboid and Star regulate receptor function or act as cell adhesion molecules and support a model in which Rhomboid and Star potentiate Egfr activation by acting in the signaling cell (Bang, 2000).

It was asked whether Rhomboid and Star potentiate Egfr signaling by changing the activity of its ligand, as suggested by the observation that sSpi does not require Rhomboid and Star to activate Egfr, whereas mSpi does. To address this question, a series of chimeras were made by replacing portions of human TGF-alpha, a vertebrate homolog of Spi, with the corresponding regions from mSpi. Human TGF-alpha alone strongly activates the human EGFR in the animal cap assay. Strikingly, when the cytoplasmic (C) and transmembrane (TM) domains of TGF-alpha are replaced with those of mSpi (TGF-alpha/SpiTMC), the chimeric molecule activates the human EGFR only when Rhomboid and Star are present. In contrast, chimeric molecules in which the TGF-alpha C or TM domains are replaced separately with those of mSpi (TGF-alpha/SpiC and TGF-alpha/SpiTM, respectively) are constitutively active. Thus, together the mSpi TM and C domains are sufficient to confer Rhomboid and Star dependence on TGF-alpha. This result suggests that the C and TM domains maintain Spi in an inactive state, and that their ability to do so is transferrable to another EGFR ligand. As predicted by this interpretation, a membrane-bound form of Spi that activates Egfr signaling in the absence of Rhomboid and Star can be generated by replacing the mSpi TM and C domains with those of TGF-alpha (Spi/TGF-alphaTMC). In addition, SpiDelta53C, a Spi mutant in which 53 carboxy-terminal residues are deleted and 17 cytoplasmic residues remain, exhibits some Rhomboid and Star-independent activity, providing further evidence that the C domain plays an inhibitory role. Together these results argue strongly that the C and TM domains of mSpi act to maintain an inactive state, with ligand activation occuring upon interaction with Rhomboid and Star (Bang, 2000).

One way in which Rhomboid and Star could lead to ligand activation is by promoting proteolytic processing, thus converting mSpi into a form similar to sSpi. Given the possibility that only low levels of Spi are required to activate Egfr, a more sensitive assay was used to determine whether Rhomboid and Star promote proteolysis of Spi. Conditioned medium was prepared from dissociated animal cap cells from embryos injected with RNA encoding sSpi, or mSpi, or coinjected with RNAs encoding mSpi, Rhomboid, and Star. Egfr-injected animal caps were incubated in the conditioned medium and then analyzed for expression of XBra. The conditioned medium from animal caps expressing sSpi or mSpi/Rhomboid/Star contains an activity that activates Egfr, whereas that from animal caps expressing mSpi alone does not. In addition, the conditioned medium activity is Egfr dependent, as it is ineffective on uninjected animal caps. These results suggest that Rhomboid and Star activate mSpi by promoting its cleavage and secretion (Bang, 2000).

Next, it was determined whether proteolytic processing is required for Rhomboid and Star activation of mSpi. To do this, potential sites for processing of mSpi were removed by deleting the sequences encoding the 15 amino acids (aa) between the Spi EGF and TM domains (Spi-15aa). This region was selected because cleavage of TGF-alpha is known to take place within an analogous interval. When tested in the animal cap assay, Spi-15aa strongly activates Egfr in a Rhomboid and Star-dependent manner. In contrast, conditioned medium prepared from animal caps expressing Spi-15aa, Rhomboid, and Star does not contain any activity that activates Egfr, indicating that Spi-15aa is not cleaved. Taken together, these results suggest that cleavage of mSpi depends on the sequence deleted in the Spi-15aa mutant; however, mSpi does not need to be cleaved to activate Egfr signaling. Thus, Rhomboid and Star may act to present mSpi to Egfr and subsequently facilitate or allow its cleavage (Bang, 2000).

The results obtained with the Spi-15aa deletion mutant suggest that mSpi, like TGF-alpha, is processed to generate a soluble form. To examine the nature of this processing further, the possibility was tested that the processing includes a cleavage within the transmembrane domain of mSpi. This possibility is suggested by the results obtained with the Spi/TGF-alpha chimeras, showing that the Spi transmembrane domain is important for Rhomboid and Star-dependent activation. Moreover, another multimembrane-spanning protein, Presenilin-1, mediates proteolyis of the beta-amyloid precursor protein and Notch, both of which are cleaved within their transmembrane domains. If processing does lead to a cleavage in the membrane, it was reasoned that this would release the intracellular domain of Spitz in a Rhomboid/Star-dependent manner. To detect this cleavage, a chimeric molecule was generated in which the mSpi C domain is replaced with the myc-tagged, intracellular domain of the Xenopus Notch receptor (Spi/NICD). The endogenous, gamma-secretase-dependent Notch cleavage site is not present in the Spi/NICD chimeric molecule. If proteolytic processing of this molecule occurs within the Spi TM domain in a Rhomboid/Star-dependent manner, NICD may be released, translocate to the nucleus, and activate target genes. As a Notch target gene Xenopus Enhancer-of-split-related-1 (Esr-1) was analyzed in animal caps that were coinjected with the neuralizing factor noggin, because Esr-1 is normally expressed in neural tissue and its induction by NICD is more robust in a noggin background (Bang, 2000).

When tested in the animal cap assay, Spi/NICD activates Egfr, but only in the presence of Rhomboid and Star, indicating that the Spi/NICD chimeric molecule still exhibits Rhomboid and Star-dependent Spi activity. This result also indicates that the myc-tagged Xenopus NICD can effectively replace the Spi C domain, suggesting that the ability of the C domain to maintain Spi in an inactive state depends more on its structure than on its primary sequence. Significantly, Spi/NICD also activates the Notch target gene, Esr-1, in a Rhomboid and Star-dependent manner. This result suggests that Rhomboid and Star promote a proteolytic processing event within the Spi-TM domain that releases NICD. In addition, because Esr-1 induction is Rhomboid and Star dependent in the absence of Egfr, Rhomboid and Star can function independent of Egfr (Bang, 2000).

By analogy to the beta-amyloid precursor protein and Notch, whose activites are regulated by multiple, interdependent cleavage events, the possibility was tested that the 15 amino acids between the Spi EGF and TM domains that are required for production of soluble Spi are also required for the cleavage of the Spi/NICD chimeric molecule within its TM domain. Thus, the sequence encoding these 15 amino acids was depleted in the Spi/NICD chimera to produce Spi-15aa/NICD. This deletion mutant still strongly activates Egfr in a Rhomboid and Star-dependent manner, but no longer induces Esr-1, indicating that NICD is not released, and thus cleavage of this mutant does not occur. Thus, these results provide further independent evidence for the contention that Rhomboid and Star-dependent cleavage of mSpi requires the amino acids deleted in the Spi-15aa mutant, but mSpi need not be cleaved to activate Egfr signaling. Finally, these results suggest that there is a Rhomboid and Star-dependent cleavage event of mSpi within its TM domain. One possible explanation for these observations is that mSpi is cleaved both within the TM domain and within the 15 amino acids between the TM and EGF domains. Alternatively, a single cleavage of mSpi could occur within its TM domain that depends on the 15 amino acid interval (Bang, 2000).

Several models could account for the Rhomboid and Star-dependent effects observed. One model is that Rhomboid and Star are required to direct mSpi to the proper compartment for signaling to occur. The results from the biotinylation experiments suggest strongly that Rhomboid and Star are not required for transport of mSpi to the cell surface, but it remains a possibility that Rhomboid and Star could play a role in localizing mSpi to specific cell surface microdomains. An alternative class of models is that mSpi is at the cell surface and ready to signal, but that Rhomboid and Star are required for bringing mSpi into an active conformation. One version of this model is that Rhomboid and Star activate mSpi by promoting its oligomerization. However, this idea is difficult to reconcile with the observation that sSpi is active and either does not require oligomerization or oligomerizes independently of Rhomboid and Star. In addition, soluble EGF, which is similar to sSpi, binds as a monomer to the extracellular domain of the EGFR in a 1:1 ratio, suggesting that membrane-bound EGFR ligands may also bind the receptor as monomers. For these reasons, an alternative model is favored in which mSpi is present at the membrane in an inactive dimeric or oligomeric complex. Rhomboid and Star would be required to either prevent formation of this complex or to alter its conformation such that mSpi could be presented as an active form. This model is precedented by observations suggesting that a number of receptor tyrosine kinases exist as inactive dimers that are activated when specific inter-subunit conformational changes occur upon ligand binding. Thus, formation of an inactive mSpi complex would be mediated by its C and TM domains and inhibited by an interaction between these domains and Rhomboid and Star. This model explains both why removal of these domains relieves the requirement for Rhomboid and Star, and transfer of these domains to TGF-alpha confers Rhomboid and Star dependence. Such a model also predicts that sSpi would be Rhomboid and Star independent (Bang, 2000).

How do Rhomboid and Star promote cleavage of mSpi? Rhomboid and/or Star could play a passive role by making mSpi accessible to proteolysis upon presentation. Alternatively, Rhomboid and/or Star may actively facilitate Spi proteolysis either by activating or recruiting a protease or transporting Spi to the appropriate subcellular compartment. It is also possible that Rhomboid and/or Star could themselves have proteolytic activity, as has been proposed for Presenilin-1. A protease responsible for Spi cleavage has yet to be identified. Finally, although this study strongly suggests that presentation of Spi is inhibited by its C-domain, the question of whether proteolysis of Spi is also affected by the C-domain has not been addressed. For instance, proteolytic release of the extracellular domain of membrane bound neuregulin is dependent on its cytoplasmic domain. Future experiments will be aimed at determining whether Rhomboid and Star play a passive or an active role in the proteolysis of mSpi (Bang, 2000).

Regulated intracellular ligand transport and proteolysis control EGF signal activation in Drosophila

The membrane proteins Star and Rhomboid-1 have been genetically defined as the primary regulators of EGF receptor activation in Drosophila, but an understanding of their molecular mechanisms has remained elusive. Both Star and Rhomboid-1 have been assumed to work at the cell surface to control ligand activation. This study demonstrates that they control receptor signaling by regulating intracellular trafficking and proteolysis of the ligand Spitz. Star is present throughout the secretory pathway and is required to export Spitz from the endoplasmic reticulum to the Golgi apparatus. Rhomboid-1 is localized in the Golgi, where it promotes the cleavage of Spitz. This defines a novel growth factor release mechanism that is distinct from metalloprotease-dependent shedding from the cell surface (Lee, 2001).

In the absence of Star, Spitz is retained in the ER. This explains why the domain of EGF receptor activation is much narrower than the expression pattern of Spitz, and why ectopic expression of full-length Spitz does not activate the receptor. Star, a protein with a single TMD, chaperones Spitz into the Golgi apparatus and the subsequent secretory pathway. The principal interaction between Spitz and Star occurs between the lumenal domains of the two proteins. Two models can be envisaged: Star could specifically block the ER retention signal; alternatively, Star could actively export Spitz from the ER, and in doing so, counteract retention (Lee, 2001).

Drosophila genetics indicate that Star and Rhomboid-1 are both prime regulators of EGF receptor activity: they both appear to be necessary and they cannot replace each other. It has not been possible to separate their functions. The results described here explain their codependency and synergy, and also provide a clear mechanistic distinction between them. An important issue is whether Star is necessary for Rhomboid-1-dependent proteolysis itself, as an enzymatic cofactor. This possibility can be ruled out, based on the Spi:TGFalpha-C chimera: it leaves the ER independently of Star and is efficiently cleaved by Rhomboid-1 in the absence of Star, implying that the primary function of Star is to export Spitz from the ER, thereby allowing it access to Rhomboid-1. Note, however, that the Rhomboid-1-dependent cleavage of the Spi:TGFalpha-TMC chimera suggests a possible secondary, nonessential role for Star as an adaptor, delivering substrate to Rhomboid-1. The data also do not rule out a role for Star in promoting efficient Spitz secretion (Lee, 2001).

The data clearly show that Rhomboid-1 is a Golgi-localized protein that triggers the proteolytic cleavage of Spitz. Rhomboid-1 could therefore be a novel protease, or it could recruit an unidentified protease; detailed biochemical analyses will be needed to resolve this. Star and Rhomboid-1 are sufficient to cleave Spitz in all cells tested, suggesting that they may be the only components required. The analysis also rules out the involvement of metalloproteases that are responsible for the release of TGFalpha and many other growth factors, further supporting the idea that Rhomboid-1 may itself be a protease. The absence of a genetically identified candidate protease, other than Rhomboid-1, despite much genetic screening, is also consistent with this hypothesis. The principal objection to this parsimonious model is Rhomboid-1's lack of identifiable protease domains. However, there are two recent precedents for multiple transmembrane domain proteins without recognizable protease domains being discovered to be novel proteases: presenilin-1 and SREBP site 2 protease (Lee, 2001).

Despite the distinctions between Spitz and TGFalpha processing, the similarities between flies and mammals may be greater than is at first apparent. For example, mature TACE (an ADAM family metalloprotease that acts on TGFalpha) is predominantly localized in intracellular compartments, suggesting that the cell surface may not be the only location of TGFalpha cleavage. Additionally, there is evidence for TACE-independent TGFalpha processing. Furthermore, TGFalpha also undergoes regulated transport through the secretory pathway, albeit by a distinct mechanism dependent on PDZ domain proteins. Finally, it is worth pointing out that while TGFalpha appears to be the mammalian ligand most similar to Spitz, there are several other analagous human EGF receptor ligands whose regulation is still poorly understood (Lee, 2001).

Intracellular trafficking by Star regulates cleavage of the Drosophila EGF receptor ligand Spitz

Cleavage of the ubiquitously expressed transmembrane form of Spi (mSpi) precedes EGF receptor activation. The Star and Rhomboid (Rho) proteins are necessary for Spi cleavage in Drosophila cells. Complexes between the Spi and Star proteins, as well as between the Star and Rho proteins were identified, but no Spi-Star-Rho triple complex was detected. This observation suggests a sequential activity of Star and Rho in mSpi processing. The interactions between Spi and Star regulate the intracellular trafficking of Spi. The Spi precursor is retained in the periphery of the nucleus. Coexpression of Star promotes translocation of Spi to a compartment where Rho is present both in cells and in embryos. A Star deletion construct that maintains binding to Spi and Rho, but is unable to facilitate Spi translocation, has lost biological activity. These results underscore the importance of regulated intracellular trafficking in processing of a TGFalpha family ligand (Tsruya, 2002).

To identify the domain(s) of Star required for its biological activity, deletion constructs were generated. Because Star is a novel protein with no defined domains (except for the transmembrane domain), no clues were available for the generation of such constructs. Searching the database, a Bombyx mori cDNA sharing homology with Star was identified. Complete sequence of this cDNA identified a type II transmembrane protein of 315 amino acids, which was termed BmS. Notably, the amino-terminal cytoplasmic domain of this protein is only 74 residues long, and the major boxes of homology are located immediately carboxy-terminal to the transmembrane domain. This pattern of homology raised the possibility that most of the amino-terminal domain of Star may be dispensable for its function, and prompted an examination of constructs of Star missing most of the amino- or carboxy-terminal domains. Two truncated Star constructs were generated. NTM contains the entire amino-terminal and transmembrane domains, and is truncated 16 residues after the transmembrane domain. Conversely, in TMC, the amino-terminal 259 residues were deleted, and an initiator methionine was added (Tsruya, 2002).

The biological activity of the constructs was tested in S2 cells. Only the TMC construct was capable of promoting mSpi cleavage, but at a lower efficiency than full-length Star. sSpi could be detected in the medium of cells coexpressing mSpi, TMC and Rho, but not in the medium of cells expressing only mSpi and TMC. Stable transfected lines expressing mSpi and TMC provided a more sensitive assay and produced low levels of sSpi. The lower levels of sSpi, however, indicate that this construct is less potent than the full-length Star protein. The NTM and BmS constructs are not active alone or even in the presence of Rho. In addition, after sequencing the Star⁵⁴ null allele, a termination codon was identified at position Q387 in the extracellular domain. This protein is 67 residues longer than NTM and yet is inactive, indicating that the carboxy-terminal region of Star is essential for its biological function (Tsruya, 2002).

The interaction between Star and mSpi prompted an examination of whether Star also associates with Rho. Star-HA was coexpressed with Rho-TAP in S2 cells. Immunoprecipitation of Rho-TAP showed coprecipitation of Star. In the reciprocal experiment, immunoprecipitation of Star-HA or Star-TAP coprecipitates Rho. To identify the regions of Star necessary for this interaction, the experiment was repeated with the three Star constructs. The NTM and BmS constructs retain the interaction with Rho, whereas the TMC construct shows no detectable interaction or only weak interaction. It appears that sequences within the amino-terminal cytoplasmic domain of Star that are not present in TMC mediate the binding to Rho (Tsruya, 2002).

The capacity of Star to bind both Spi and Rho, raises the possibility that Star may function as a scaffold protein, to form a trimeric protein complex. Examination of possible direct interactions between mSpi and Rho failed to detect any significant coprecipitation. If a trimeric complex is formed, it would be expected that Star would promote coprecipitation of mSpi by Rho. However, no elevation in mSpi coprecipitation with Rho was detected when Star was also expressed. Under these conditions, Star retains the capacity to bind mSpi, regardless of the expression of Rho. Therefore, it appears that whereas mSpi-Star or Star-Rho complexes are formed in the cells, no trimeric complex is present (Tsruya, 2002).

Because coexpression of Rho and Star significantly promotes processing of mSpi, it is possible that short-lived trimeric complexes are formed as a result of cleavage of mSpi. To circumvent this problem, the formation of complexes with a noncleavable mSpi precursor was examined. An mSpi protein in which 16 amino acids between the EGF and transmembrane domains were deleted, does not undergo cleavage in S2 cells. A similar construct fails to be cleaved in the Xenopus assays. Like mSpi, the deleted mSpi protein readily forms complexes with Star, but not with Rho. Again, even in the presence of Star, Rho is not capable of coprecipitating this uncleavable Spi construct, arguing against the presence of even a transient triple complex. This experiment suggests that Star and Rho function sequentially in mSpi processing (Tsruya, 2002).

In view of the apparent sequential roles of Star and Rho, attempts were made to identify the cellular compartments in which they are active. The subcellular localization of Star and Rho may provide a clue as to their order of activity. Star protein has been reported to be in the periphery of the nucleus, consistent with an ER localization. Immunostaining of S2 cells to follow Star overexpression confirms this localization pattern. Immunostaining of Rho in S2 cells overexpressing Rho shows a punctate distribution that does not colocalize with a 120-kD integral Golgi membrane protein, as well as plasma membrane staining. This pattern is in accordance with previous observations in embryos showing punctate endogenous Rho distribution. Following overexpression, plasma membrane staining is also detected. Expression of Star and Rho shows only a restricted overlap. The distinct cellular distribution of these proteins suggests that a key feature in the regulation of mSpi processing may be its cellular trafficking. This possibility is in accord with the sequential activities of Star and Rho implied by the coprecipitation results (Tsruya, 2002).

Expression of Rho with mSpi-GFP does not alter its distribution, in agreement with a lack of coprecipitation. However, coexpression of Rho with mSpi-GFP and Star reduces dramatically the levels of mSpi-GFP observed in the cells, in accordance with the results obtained in anti-Spi blots. The residual mSpi-GFP is found in a punctate staining colocalizing with Rho. These results show that Rho can alter the levels of mSpi in the cells, but only in the presence of Star. They are consistent with sequential activity of Star and Rho, where Star is required first to transport mSpi from the ER to the compartment containing Rho, and Rho subsequently facilitates the cleavage and secretion (Tsruya, 2002).

In conclusion, this work shows that the intracellular localization and trafficking of mSpi is crucial for its regulated cleavage. mSpi is normally retained in a peripheral nuclear compartment where it does not undergo cleavage and may be rapidly degraded. Star, which is also enriched in this compartment, associates with mSpi and translocates it to a compartment in which Rho is enriched, to allow cleavage. The association between Star and Rho may allow Star to efficiently deliver mSpi. The cleavage process thus entails orchestration of trafficking and protein-protein associations to ensure tight regulation of mSpi processing and, hence, Egfr activation (Tsruya, 2002).

A family of Rhomboid intramembrane proteases activates all Drosophila membrane-tethered EGF ligands

Drosophila has three membrane-tethered epidermal growth factor (EGF)-like proteins: Spitz, Gurken and Keren. Spitz and Gurken have been genetically confirmed to activate the EGF receptor, but Keren is uncharacterized. Spitz is activated by regulated intracellular translocation and cleavage by the transmembrane proteins Star and the protease Rhomboid-1, respectively. Rhomboid-1 is a member of a family of seven similar proteins in Drosophila. Four of the rhomboid family members have been examined: all are proteases that can cleave Spitz, Gurken and Keren, and all activate only EGF receptor signaling in vivo. Star acts as an endoplasmic reticulum (ER) export factor for all three. The importance of this translocation is highlighted by the fact that when Spitz is cleaved by Rhomboids in the ER it cannot be secreted. Keren activates the EGF receptor in vivo, providing strong evidence that it is a true ligand. These data demonstrate that all membrane-tethered EGF ligands in Drosophila are activated by the same strategy of cleavage by Rhomboids, which are ancient and widespread intramembrane proteases. This is distinct from the metalloprotease-induced activation of mammalian EGF-like ligands (Urban, 2002).

Within the seven Rhomboid-like molecules in Drosophila, Rhomboids 1-4 are more closely related to each other than they are to any of the remaining three Rhomboids. Rhomboids 6 and 7 are more divergent, although they both retain the residues required for the serine protease activity of Rhomboid-1. Rhomboid-5 has similarities to the others but does not contain the catalytic residues. Whether Rhomboids 1-4 all have proteolytic activity against Spitz, the known substrate for Rhomboid-1, was examined. Spitz is cleaved efficiently by all Rhomboids tested, albeit with some significant differences. A cell culture assay allowed the distinguishing of Spitz cleaved in cell lysates from that which had been secreted into the medium. The amount of cleaved Spitz detected in cells varies with different Rhomboids; no or very little intracellular cleaved Spitz is detected in cells with Rhomboid-1, while cleaved intracellular Spitz is readily detected with Rhomboids 2-4. This is most apparent by comparing the relative levels of full-length and cleaved Spitz in cell lysates. In the presence of Star, the amount of secreted Spitz present in the medium is the same for all core Rhomboids, even when they are made limiting by reducing their levels of expression, indicating that all four Rhomboids have similar levels of proteolytic activity against Spitz (Urban, 2002).

Star regulates Spitz cleavage by Rhomboid-1 by transporting Spitz to the Golgi apparatus. Strikingly, although Star was essential for ligand secretion into the culture medium in each case, it does not affect the ability of Rhomboids 2, 3 and 4 to catalyse Spitz cleavage. The extensive O-linked glycosylation that is diagnostic of transit through the Golgi apparatus (and which increases the apparent molecular weight of Spitz) was not present in cell lysates. Therefore, in contrast to Rhomboid-1, Rhomboids 2, 3 and 4 causes the accumulation of an intracellular cleaved Spitz that is not transported past the trans-Golgi network and thus not secreted (Urban, 2002).

The ability of Rhomboids 1-4 to catalyse Spitz cleavage in the tissue culture assay suggested that all may be involved in activating the EGF receptor in vivo. This has been clearly demonstrated for Rhomboid-1, was genetically determined in the case of Rhomboid-3, and has been proposed for Rhomboid-2. To investigate this further, the potential activity of Rhomboids 2-4 in vivo was compared by overexpressing them in developing Drosophila tissues. In all cases examined, Rhomboids 2-4 cause similar phenotypes to Rhomboid-1, consistent only with EGF receptor hyperactivation. When expressed in the developing wing, for example, all core Rhomboids produce ectopic and thickened vein phenotypes similar to those observed for Rhomboid-1. This phenotype was modified predictably by mutations in other members of the EGF receptor pathway. Furthermore, as in cell culture assays, all four Rhomboids are synergistic with the co-expression of Star. In all cases, UAS Rhomboids 1 and 3 produce consistently strong wing phenotypes, whereas Rhomboids 2 and 4 are weaker. Similar results were obtained in the eye, follicle cells of the ovary and the embryo. Importantly, no other phenotypes were observed in eyes, wings or embryos expressing Rhomboids, suggesting that they do not affect any other pathways. If, for example, the previously uncharacterized Rhomboids 2 or 4 caused the activation of other signaling pathways, their ectopic expression would lead to additional phenotypes. These observations confirm that Rhomboids 2-4 contain the same proteolytic activity as Rhomboid-1; furthermore, the absence of phenotypes associated with other pathways strongly suggests that Rhomboids 1-4 are all dedicated to regulating EGF receptor signaling. These data demonstrate that Rhomboids 1-4 all share proteolytic activity against the ligand Spitz (Urban, 2002).

A new Spitz-like gene was identified as a cDNA submitted to GenBank by the Berkeley Drosophila cDNA sequencing project. With the subsequent completion of the Drosophila genome sequence, this gene has been annotated as Keren and is the only previously unknown membrane-tethered EGF-like molecule identified by the Drosophila genome project. Keren has been referred to previously as Spitz-2 and Gritz. Like Spitz and Gurken, Keren has a single extracellular EGF repeat and a single transmembrane domain. The amino acid sequence of Keren is more closely related to Spitz than to Gurken (49% identity, 55% similarity to Spitz; 30% identity, 37% similarity to Gurken). While all three ligands were predicted to have N- and O-linked glycosylation signals, Spitz contains a 10 residue insert in its N-terminus, which contains an additional O-linked glycosylation site. Consistent with this, Spitz is the only ligand to be hyperglycosylated in the presence of Star, although deletion of the insert does not fully abolish hyperglycosylation. Rhomboids 1-4 all cleaved Keren in a mammalian tissue culture assay. There was, however, an interesting distinction between Keren and Spitz: Star was not essential for Keren secretion in every case. A significant amount of Keren was secreted in the presence of Rhomboids 3 and 4 alone. Despite this, Star always enhanced the secretion of cleaved Keren, implying that it can interact with Keren (Urban, 2002).

A key to understanding the regulation of Spitz activation by Rhomboid-1 and Star was the observation that the ligand was restricted to the ER in the absence of Star. The intracellular localization of Keren was examined in COS cells. Like Spitz, Keren was only detectable in the ER; it exhibited characteristic perinuclear and reticular staining and co-localized with the ER marker protein disulfide isomerase (PDI) (Urban, 2002).

Star's role in Spitz activation is to export Spitz from the ER to the Golgi apparatus where it encounters the proteolytic activity of Rhomboid-1. Star also promotes the release of Keren into the medium, suggesting its role is similar to that in Spitz processing. This relocalization of Keren is very similar to the relocalization observed for Spitz, suggesting that Keren also needs to be relocalized to the Golgi apparatus for efficient processing and secretion (Urban, 2002).

To test the prediction that Keren is a genuine EGF receptor ligand, it was misexpressed in developing Drosophila tissues. By analogy to similar experiments with Spitz, either full-length, membrane-tethered Keren (mKeren) was expressed, or a truncated form that corresponds to the extracellular, secreted form of Keren (sKeren). In most contexts, full-length Spitz is unable to signal when misexpressed because Star and Rhomboid-1 activity limit its activation, while a truncated form of Spitz, missing its transmembrane domain and C-terminus, signals in a Rhomboid-1- and Star-independent manner. In contrast to Spitz, ectopic expression of mKeren activates the EGF receptor pathway in both eyes and wings. For example, when expressed in the developing wing, mKeren produced wing phenotypes similar to misexpression of other positively acting members of the EGF receptor pathway, ranging from thickened and ectopic wing veins to blistering. In many cases, the activation was so strong that the entire wing was converted to vein-like material. These results indicate that either Keren has membrane-tethered, juxtacrine activity or that it is processed and secreted, possibly by Rhomboids 3 or 4, which would be consistent with the results obtained in the cell culture assay (Urban, 2002).

To test whether the activity of mKeren represents the full potential phenotype of ectopic Keren, or whether proteolytic activation has the potential to activate it further, the effects of sKeren were examined. This form is even more potent than mKeren, causing lethality even when driven by tissue-specific drivers. However, in the embryo, where mKeren has a weak effect, ubiquitous misexpression of sKeren causes lethality and results in significantly widened denticle belts and a reduction in naked cuticle in the ventral epidermis, identical to that caused by the misexpression of sSpitz. The greater potency of sKeren therefore suggests that Keren is proteolytically activated in vivo (Urban, 2002).

Together, these results indicate that Keren is a genuine ligand for the EGF receptor, being able to activate the receptor pathway in vivo. They also suggest that Keren is activated by Rhomboid proteases and Star, although it remains possible that the membrane-tethered form of the ligand has some juxtacrine activity (Urban, 2002).

Gurken function is restricted to oogenesis where it is required to polarize both major axes of the egg. Recent evidence strongly suggests that Gurken undergoes proteolytic processing in vivo: Gurken is released from the oocyte and is internalized by follicle cells, exists exclusively in a cleaved form in oocytes, and an uncleavable mutant form is inactive. Gurken can be processed directly by Rhomboid proteases 1-4. In the tissue culture assay, Rhomboid protease activity is required for Gurken cleavage and secretion. Although Rhomboid-1 is not required in the female germline, the specific expression of Rhomboid-2 in the early oocyte suggests that a Rhomboid might have a role in Gurken processing. Star can translocate Gurken from the ER to the Golgi apparatus in cell culture and, in some cases, enhance Gurken secretion. Together, these results strongly suggest that Gurken activity, like that of Spitz, is at least partially regulated by Star-dependent ER to Golgi transport. The regulation of Gurken activity, however, also depends on the transmembrane protein Cornichon. Recent evidence in yeast and Drosophila suggests that Cornichon is an ER export factor, raising the question of the relative roles and significance of Star and Cornichon (Urban, 2002).

Current data suggest two possible explanations for the multiplicity of Drosophila Rhomboids. First, several members appear to have tissue-specific functions. For example, Rhomboid-3 acts specifically in the developing eye. This duplication of function may reflect the complexity of regulation of EGF receptor activation. Rhomboid-1 is the principal determinant of EGF receptor activity and is regulated transcriptionally. The full gamut of transcriptional control may be too difficult to achieve in a single rhomboid gene. Partitioning this transcriptional regulation among multiple genes of similar biochemical activity could solve this problem. Note that although there is also a requirement for Star, which exists only as a single gene in Drosophila; its expression is not as restricted as that of Rhomboid-1, and its requirement is not absolute, as certain Rhomboids have the ability to release EGF receptor ligands in the absence of Star (Urban, 2002).

A second reason for the multiplicity of Rhomboids is suggested by their distinct characteristics: although all four cleave EGF ligands, they have different abilities to elicit cleavage and secretion of ligands in the absence of Star. As such, multiple Rhomboid-ligand combinations may result in distinct signaling characteristics such as intensity, duration or range (Urban, 2002).

Finally, the present analysis has been limited to Rhomboids 1-4. Rhomboid-5 lacks residues necessary for proteolysis, but the sequences of Rhomboids 6 and 7 suggest that they are proteases, and a key question will be whether these more distant members of the family are also dedicated to the EGF receptor pathway or whether they have distinct functions (Urban, 2002).

Keren, a new ligand of the Drosophila epidermal growth factor receptor, undergoes two modes of cleavage

Spitz (Spi) is the most prominent ligand of the Drosophila EGF receptor. It is produced as an inactive membrane precursor that is retained in the endoplasmic reticulum (ER). To allow cleavage, Star transports Spi to the Golgi, where it undergoes cleavage by Rhomboid. Since some Egfr phenotypes are not mimicked by any of its known activating ligands, an additional ligand (Keren) was identified by database searches. Krn is a functional homolog of Spi since it can rescue the spi mutant phenotype in a Rho- and Star-dependent manner. In contrast to Spi, however, Krn also possesses a Rho/Star-independent ability to undergo low-level cleavage and activate Egfr, as evident both in cell culture and in flies. The difference in basal activity correlates with the cellular localization of the two ligands. While Spi is retained in the ER, the retention of Krn is only partial. Examining Spi/Krn chimeric and deletion constructs implicates the Spi cytoplasmic domain in inhibiting its basal activity. Low-level activity of Krn calls for tightly regulated expression of the Krn precursor (Reich, 2002).

It was of interest to determine whether Karen processing is regulated in a similar manner to Spi. Triggering of Egfr by Spi at stage 10 generates two prominent domains of activation, in the tracheal placodes and ventral ectoderm. These domains correspond to the sites of Rho expression in the tracheal placodes and midline, respectively. dpERK can be readily detected in these domains, while in spi mutant embryos no dpERK is observed at this stage. The capacity of the Krn precursor to rescue spi mutant embryos was examined. When Krn was ubiquitously expressed in the ectoderm of spi^- embryos (using the 69B-Gal4 driver), complete rescue of the dpERK pattern was observed. Induction of the pathway by Krn at the sites of Rho expression in the midline and tracheal placodes indicates that, like Spi, processing of Krn is dependent upon Rho. To test whether Krn cleavage requires Star, Krn was expressed in Star mutant embryos. No rescue of the phenotype was observed, as monitored by dpERK. It is thus concluded that, like Spi, processing of Krn is Rho and Star dependent (Reich, 2002).

A clearer understanding of Spi cleavage has been gained by studies in cells. Efficient cleavage of Spi occurs only in cells in which both Star and Rho are expressed. Spi is retained in the ER through its intracellular domain. Star binds Spi and translocates it from the ER to the Golgi, where Rho functions as a protease and cleaves Spi. Krn-GFP in Drosophila S2 cells shows partial release from retention, manifested by a vesicular distribution, indicating exit from the ER. The functional implication of this was observed through the ectopic phenotypes in various tissues. For instance, whereas ectopic Spi driven by ubiquitous GAL4 in embryos does not prevent hatching of larvae, ectopic Krn leads to lethality. This highlights the importance of retention, since it allows expression of high amounts of Spi protein in the cell, yet controls Spi activity by preventing Spi from reaching further compartments where cleavage occurs (Reich, 2002).

Chimeric and deletion constructs identify the cytoplasmic domains of Spi and Krn as the domains responsible for their different cleavage profiles. This is a result of different levels of retention in the ER. The mechanism of Spi and Krn retention is not yet clear. Spi has also been shown to be retained in a heterologous system of mammalian cells, implicating the action of conserved molecules or an intrinsic property of the protein. In one model, association of the Spi cytoplasmic domain with an additional protein(s) could mediate retention. In that case, it would be expected that Krn would have lower affinity to this protein(s). In another model, the Spi C-terminus itself could have an intrinsic inhibitory capability through protein folding that sterically prohibits association to proteins -- this would carry Spi further in the secretory pathway. In this case, Krn would be expected to possess a higher affinity to such chaperones, that would allow it to exit the ER without total dependence on Star (Reich, 2002).

Compared with Spi or Krn, the cytoplasmic domain of Grk, the third Egfr ligand with a transmembrane domain, is shorter (only 24 amino acids). Deletion of its cytoplasmic domain did not influence signaling by Grk. Like Krn, overexpression of the full-length Grk protein causes ectopic wing phenotypes. It would be interesting to see to what extent Grk is retained in the ER of S2 cells (Reich, 2002).

Expression of Krn in S2 cells allows the mechanism of low-level cleavage, which is Star and Rho independent, to be followed. What is the protease responsible for this cleavage? The sensitivity of Krn cleavage to inhibitors of serine proteases indicates that cleavage may be mediated by a protease of this family. Unlike Rho, which is expressed in a spatially and temporally regulated manner, the protease is expected to be ubiquitously expressed, since ectopic Krn causes abnormal phenotypes wherever it was expressed. This further elaborates the need for tight transcriptional control on Krn expression (Reich, 2002).

Co-expression of Star with Krn in S2 cells raises the amounts of secreted sKrn in the medium. This is a result of the efficient export of Krn from the ER by Star. Since higher levels of cleavage can be obtained by co-expressing both Rho and Star, it would seem to indicate that the protease involved in low-level cleavage is less efficient than Rho (Reich, 2002).

High-level cleavage of Krn was followed in embryos through the detection of dpERK. The activation profile followed the restricted expression of Rho, since Star is broadly expressed. Only in cell culture could Rho enhance cleavage of Krn without co-expression of Star. This probably occurs because Krn can 'leak' out of the ER, reach compartments where Rho is present and undergo cleavage by Rho independent of Star (Reich, 2002).

Sequence conservation within the transmembrane domains of the Rho protein have suggested that the cleavage site of Spi would reside within the transmembrane domain. There is also evidence to indicate that Rho cleaves Spi within the membrane. The transmembrane regions of Spi and Krn are conserved, and show 50% sequence identity. The transmembrane domain of Krn may thus possess the same recognition sites as that of Spi (Reich, 2002).

In conclusion, Krn was found to be the functional homolog of Spi. Unlike Spi, Krn is capable of undergoing inefficient Star- and Rho-independent cleavage in flies and in cell culture. This is due to differences between the intracellular domains of Krn and Spi, which allow Krn to evade retention in the ER and reach further along in the secretory pathway. This calls for tight transcriptional control of Krn expression, in contrast to Spi, which can be ubiquitously and abundantly expressed (Reich, 2002).

Rhomboid and Star and the activation of Gurken

The mechanism of activation of the Epidermal growth factor receptor (Egfr) by the transforming growth factor alpha-like molecule, Gurken (Grk) has been examined. Grk is expressed in the oocyte and activates the Egfr in the surrounding follicle cells during oogenesis. Expression of either a membrane bound form of Grk (mbGrk), or a secreted form of Grk (secGrk), in either the follicle cells or in the germline, activates the Egfr. In tissue culture cells, both forms can bind to the Egfr; however, only the soluble form can trigger Egfr signaling, which is consistent with the observed cleavage of Grk in vivo. The two transmembrane proteins Star (S) and Brho (rhomboid-2) potentiate the activity of mbGrk. These two proteins collaborate to promote an activating proteolytic cleavage and release of Grk. After cleavage, the extracellular domain of Grk is secreted from the oocyte to activate the Egfr in the follicular epithelium (Ghiglione, 2002).

Grk is cleaved in the germline. An important question is where exactly the cleavage of the Grk precursor occurs? Other studies have concluded that the cleavage of Spitz occurs in the TM and depends on the 15 amino acid stretch located between the EGF and TM domains. The Grk dibasic signal (R₂₄₀ and K₂₄₁) is not the cleavage site because its mutation does not abolish this event. However, mbGrkDelta19AA^myc, in which the 19 amino acid (Y₂₂₄ to V₂₄₂) located between the EGF and TM domains have been deleted, is no longer cleaved, suggesting that this sequence is directly or indirectly involved in the processing (Ghiglione, 2002).

The results do not rule out the hypothesis that Grk cleavage occurs in the TM domain as proposed for Spi. The high conservation between the Spi and Grk TM domains, in addition to aberrant Grk localization observed with different grk alleles affecting this TM domain, reveals its importance. Moreover, the cleaved product of Grk that is released in the medium, after co-expressing mbGrk+S+Rho-1/Brho in S2 cells, has a slightly higher mobility that the engineered secGrk. Thus, it is possible that mbGrk is cleaved within the TM domain and that proteolysis depends on the 19 amino acid interval (Ghiglione, 2002).

These results reflect the importance of the Grk TM domain for proper processing and routing through the secretory pathway. mbGrk processing is probably tightly regulated and leads to efficient Grk secretion, contrary to engineered secGrk, which is poorly secreted from the oocyte and which acts mainly intracellularly (Ghiglione, 2002).

The recent findings that Star and Brho, a Rho-related protein, are expressed in the oocyte led to an investigation of whether they are involved in Grk activation during oogenesis. Star and Rho proteins have been proposed to be involved in the processing and activation of Spi; however, because they have no obvious motifs that predict their biochemical functions, their roles in ligand maturation and/or secretion have remained obscure (Ghiglione, 2002).

The analysis of these proteins in the context of Grk signaling has provided numerous insights into the relationships between these transmembrane proteins. The in vivo data strongly suggest that the expression level of Star and Brho is very high in the oocyte, thus leading to an efficient cleavage and secretion of Grk. However, Star and Rho-1 are probably expressed at low level in the follicle cells. Indeed, the presence of Star in this epithelium using an anti-Star antibody has not been detected, whereas they clearly show a strong staining in the germline. The presence of both endogenous Star and Rho-1 in follicle cells explains why overexpression of mbGrk in this epithelium leads to a weak dorsalization of the eggs. Nevertheless removing one copy of Star is sufficient to completely suppress this phenotype. This confirms the observation that overexpression of mbGrk on its own is not able to activate the Egfr in vivo, as supported by the in vitro study. Overexpression experiments in follicle cells indicate a strong synergy between mbGrk, Star, and Brho, as previously observed for Spi. Further, co-expression of Star and Rho-1/Brho is sufficient for Grk cleavage and secretion in S2 cells, strongly suggesting that they are the only proteins required for this process. In addition, these tissue culture experiments reveal that Star and Rho-1/Brho are not obligate cofactors for this cleavage, because co-expression of mbGrk with Rho-1/Brho is sufficient to catalyze this proteolytic event. Star is not required for Rho-1/Brho-mediated proteolytic cleavage in S2 cells, but the soluble Grk extracellular domain is no longer detected in the medium from these cells, indicating that the function of Srar is necessary for trafficking/secretion of the ligand. However, Star is not able to cleave Grk in absence of Rho-1/Brho. Altogether, these results show that the functions of Rho-1/Brho and Star are distinct, which explains their co-dependence and synergism in vivo (Ghiglione, 2002).

Rho-1/Brho may facilitate Grk proteolysis either by activating or recruiting a yet unknown protease. By analogy to the processing of mammalian Egfr ligands, Grk cleavage may be catalyzed by an ADAM-like metalloprotease. Although these molecules are present in Drosophila, nothing is known yet about their functions. An alternative hypothesis, is that despite the absence of known protease domains in their sequences, Rho-1 and Brho themselves may have proteolytic activity. The subcellular localization of Brho, as observed for mature TACE (ADAM17), is predominantly in intracellular compartments. In addition, and directly relevant to this hypothesis, Presenilins, which define another subfamily of seven-pass transmembrane proteins, have been proposed to encode proteases. In Drosophila, Presenilin may be directly responsible for the proteolysis of the intra-transmembrane domain of Notch (Ghiglione, 2002 and references therein).

One of the striking feature of Rho-related proteins is that amino acid sequence conservation is most prominent in the predicted TM regions that contain some invariant charged residues. This suggests the presence of a hydrophilic pocket that might constitute an enzymatic active site or a channel, as observed in Presenilins. This model is further supported by the recent finding that the TM domain of Spi is important for its functional interaction with Rho-1 (Ghiglione, 2002 and references therein).

rho-related genes have been found in organisms from diverse kingdoms including C. elegans, rat, human, Arabidopsis, sugar cane, yeast and bacteria. The data suggest that Brho, like Rho-1, promotes Egfr signaling by activating TGFalpha-like ligands. Since RTKs have not been found in plants, yeast or bacteria, the rho-related genes in these organisms presumably serve other functions. It will be interesting to determine whether the activities of these Rho-related proteins are similar to those of Rho-1 and Brho, such as promoting the processing of proteins (Ghiglione, 2002).

Mosaic analysis of Star, both in the germline and in follicle cells, together with the Star antisense experiment, demonstrate that Star is required in follicle cells for Spi-dependent Egfr activation, and in the germline for Grk-dependent Egfr activation. Tissue culture experiments suggest that Star is not involved in Grk proteolysis, but instead in post-cleavage trafficking or secretion of the ligand. The intracellular localization of Star is also consistent with a role for Star at a step that follows the Brho-dependent cleavage, because it was found that Star is predominantly very close to, or at the plasma membrane, while Brho localizes to the Golgi. The role of Star, however, is not yet resolved because the results contrast with the ER localization of Star in the oocyte described by others. Interestingly, unlike Rho-1 and Brho, Star is probably involved in other processes as well. For example, Star has been identified as a suppressor of Delta, one of the Notch ligands. Delta encodes a transmembrane protein that is cleaved by the Kuzbanian metalloprotease, and the extracellular fragment antagonizes the function of the membrane-bound Delta protein as an activating Notch ligand. In the case of Notch signaling, a reduction of Star gene activity might lead to a reduced release of the extracellular Delta fragment, and thus enhance Delta signaling (Ghiglione, 2002 and references therein).

Finally, understanding the function of Star and Rho-1/Brho in Grk processing is relevant to studies of the mammalian ligands of the EGFR family as well, because TGFalpha may also be processed in vivo before receptor binding. Thus, although further work is needed to fully understand the biochemical function of Star, and Rho-1/Brho, these studies have provided a number of insights into the mechanism of action of these molecules (Ghiglione, 2002).

Mechanism of intramembrane proteolysis investigated with purified rhomboid proteases

Intramembrane proteases have the unusual property of cleaving peptide bonds within the lipid bilayer, an environment not obviously suited to a water-requiring hydrolysis reaction. These enzymes include site-2 protease, gamma-secretase/presenilin, signal peptide peptidase and the rhomboids, and they have a wide range of cellular functions. All have multiple transmembrane domains and, because of their high hydrophobicity, have been difficult to purify. An in vitro assay has been developed to monitor rhomboid activity in the detergent solubilised state. This has allowed isolation of a highly pure rhomboid with catalytic activity. These results suggest that detergent-solubilised rhomboid activity mimics its activity in biological membranes in many aspects. Analysis of purified mutant proteins suggests that rhomboids use a serine protease catalytic dyad instead of the previously proposed triad. This analysis also suggests that other conserved residues participate in subsidiary functions like ligand binding and water supply. A motif has been identified that is shared between rhomboids and the recently discovered derlins, which participate in translocation of misfolded membrane proteins (Lemberg, 2005).

The choice of the detergent that allows retention of the native structure and the biological function of membrane proteins is critical; in this aspect, the rhomboids differ from other intramembrane proteases. The zwitterionic detergents CHAPSO and CHAPS have been successfully used to solubilise the gamma-secretase complex and signal peptide peptidase, respectively. Recently, the nonionic detergent DDM was used for the purification of the site-2 protease homologue RseP. Rhomboids were not solubilised efficiently by CHAPS, but are active in DDM and TX100. In contrast to the similarity of rhomboid cleavage site in both detergents and intact membranes, gamma-secretase specificity is sensitive to solubilisation conditions and TX100 and DDM disrupt the gamma-secretase activity. It is also noted that, even in detergents compatible with activity, gamma-secretase and SPP are inactive when the detergent concentration is above the critical micelle concentration (CMC). Combined with the lipid requirement observed after purification of the gamma-secretase complex, this suggests that the proteolytic activity in these assays might require mixed lipid-detergent-protein structures or partially reconstituted proteoliposomes, which form spontaneously on dilution to detergent concentrations below the CMC. Rhomboids, in contrast, are active even in presence of 0.5% TX100 and DDM, both of which have a CMC below 0.1%. This suggests that rhomboids may differ from gamma-secretase and SPP in their ability to cleave peptide bonds in detergent micelles (Lemberg, 2005).

The shared WR motif between rhomboids and derlins is intriguing. Both are polytopic membrane proteins with the motif residing in the luminal loop, just N-terminal to the second predicted TMD. They also share a highly conserved residue with a hydroxyl side-group, also in the second TMD. Further support for a possible functional relationship between rhomboids and derlins is the observation that reiterative homology searching with the PSI-Blast algorithm reveals a consistent (though weak) sequence relationship between the derlins and rhomboids, suggesting that they may in fact be distantly related. Since the derlins function in a process that involves translocation of transmembrane proteins across the ER membrane prior to proteasomal degradation, it is difficult to guess what mechanistic features they might share with rhomboids. A possible common theme between the two protein families might be the destabilisation or unfolding of transmembrane helices to allow, in one case, translocation to the cytosol, or in the other, cleavage in the plane of the membrane. It is emphasised, however, that this possibility remains speculative until the mechanism of both rhomboids and derlins are better understood (Lemberg, 2005).

Distinct functional units of the Golgi complex in Drosophila cells

A striking variety of glycosylation occur in the Golgi complex in a protein-specific manner, but how this diversity and specificity are achieved remains unclear. This study shows that stacked fragments (units) of the Golgi complex dispersed in Drosophila imaginal disc cells are functionally diverse. The UDP-sugar transporter Fringe-Connection (Frc) is localized to a subset of the Golgi units distinct from those harboring Sulfateless (Sfl), which modifies glucosaminoglycans (GAGs), and from those harboring the protease Rhomboid (Rho), which processes the glycoprotein Spitz (Spi). Whereas the glycosylation and function of Notch are affected in imaginal discs of frc mutants, those of Spi and of GAG core proteins are not, even though Frc transports a broad range of glycosylation substrates, suggesting that Golgi units containing Frc and those containing Sfl or Rho are functionally separable. Distinct Golgi units containing Frc and Rho in embryos can also be separated biochemically by immunoisolation techniques. Tn-antigen glycan is shown to be localized only in a subset of the Golgi units distributed basally in a polarized cell. It is proposed that the different localizations among distinct Golgi units of molecules involved in glycosylation underlie the diversity of glycan modification (Yano, 2005).

The pattern of glycosylation is extremely diverse, yet is highly specific to each protein. How can this specificity (and diversity) be achieved? There are >300 glycosylenzymes in humans and >100 in Drosophila, but is their enzymatic specificity sufficient to explain the precise modification of all substrates? One possible mechanism that might also contribute to the specific (and diverse) pattern of glycosylation would be the localization/compartmentalization of glycosylenzymes (Yano, 2005).

The Golgi complex, where protein glycosylation takes place, has been regarded as a single functional unit, consisting of cis-, medial-, and transcisternae in mammalian cells. However, the three-dimensional reconstruction of electron microscopic images of the mammalian Golgi structure has suggested the existence of more than one Golgi stack, with the individual stacks being connected into a ribbon by tubules bridging equivalent cisternae. Furthermore, during mitosis, the Golgi cisternae of mammalian cells become fragmented without their disassembly. In Drosophila, Golgi cisternae are stacked but are not connected to form a ribbon at the embryonic and pupal stages even during interphase, although there has been no evidence to date to indicate functional differences among the Golgi fragments (Yano, 2005).

A Drosophila UDP-sugar transporter, Fringe connection (Frc) transports a broad range of UDP-sugars that can be used for the synthesis of various glycans, including N-linked types, GAGs, and mucin types. Interestingly, despite its broad specificity, loss-of-function studies have revealed that Frc is selectively required for Notch glycosylation, but not for GAG synthesis. This observation prompted a study at Frc localization; in this study, it was found that Frc is localized only to a subset of Golgi fragments in Drosophila discs and embryos (Yano, 2005).

Frc, Sfl, a glycosylenzyme of GAGs, and Rho, a processing enzyme of Spi glycoprotein, are localized to distinct Golgi fragments, which are referred to as 'Golgi units,' in Drosophila cells. frc mutants do not exhibit defects in the glycosylation and function of Spi nor do they exhibit defects in glycosylation or function of GAG core proteins. Moreover, biochemically separated distinct Golgi units containing Frc and Rho were isolated by immunoisolation technique. This study clearly shows that there are functionally distinct Golgi units in a Drosophila cell (Yano, 2005).

The Golgi complex is a stack of cis-, medial-, and transcisternae in mammalian cells. In contrast, Golgi markers often do not overlap with each other in Saccharomyces cerevisiae, in which the Golgi cisternae are not stacked but disassembled. The Golgi cisternae of Drosophila are stacked but are not connected to form a ribbon at the embryonic and pupal stages even during interphase. To determine whether Drosophila imaginal disc cells have assembled or disassembled Golgi cisternae, the localizations were compared of the cis-cisternal marker dGM130, the transcisternal marker dSyntaxin16 (dSyx16), and the Golgi-tethered 120-kDa protein, which is commonly used to detect the Golgi complex in Drosophila. The 120-kDa protein was identified by immunoaffinity purification and protein sequencing as a Drosophila homolog of the vertebrate 160-kDa medial Golgi sialoglycoprotein (MG160), which resides uniformly in the medial-cisternae of the Golgi apparatus in vertebrate cells. An antibody specific for the 120-kDa protein also stained numerous Golgi fragments in imaginal disc cells. More than 80% of immunoreactivity for the 120-kDa protein colocalizes with both dGM130 and dSYX16, suggesting that 120-kDa protein-positive fragments of the Golgi complex indeed comprise assembled cisternae; these fragments will be referred to as 'Golgi units'. The distributions of the 120-kDa protein, dGM130, and peanut agglutinin (PNA), another transcisternal marker, also shows that the markers are closely apposed but not identical, suggesting that the Golgi units are polarized. Interestingly, most of the PNA-positive transcisternae are oriented toward the basal side of the cell, within the Golgi complex, whereas most of the GM130-positive cis-cisternae are oriented toward the apical side of the cell. The cis-to-trans polarity of each Golgi unit thus appears to be correlated with the apico-basal polarity of the disc cells (Yano, 2005).

Drosophila mutant larvae defective in the UDP-sugar transporter Frc manifest a highly selective phenotype: the lack of Notch glycosylation in the presence of normal GAG synthesis (Goto, 2001). This limited phenotype was unexpected, given that Frc exhibits a broad specificity for UDP sugars used in the synthesis of various glycans including N-linked types, GAGs, and mucin types. However, given that the frc^R29 allele studied previously (Goto, 2001) is hypomorphic, whether the selective glycosylation defect might be a consequence of partial loss of Frc activity was examined. With the use of imprecise excision, a new allele, frc^RY34, was generated the presence of which results in the death of most larvae during the second-instar stage, much earlier than the death induced by frc^R29. Real-time PCR analysis revealed that the amount of frc transcripts in the second-instar larvae of frc^RY34 or frc^R29 mutants was 4.2% and 24.4% of that in the wild type, respectively. About 1 kb of the gene, including the transcription initiation site, was deleted in the frc^RY34 allele. Together, these observations suggest that frc^RY34 is essentially a null allele (Yano, 2005).

Clonal cells of the frc^RY34 mutant exhibit normal levels of GAGs, as detected by immunostaining with the 3G10 antibody, whereas the amount of GAGs was reduced in clones of tout-velu (ttv) mutant cells. Given that GAGs are required for signaling by Hedgehog (Hh), Wingless (Wg), and Decapentaplegic (Dpp),the expression was examined of corresponding target genes [patched (ptc) for Hh signaling and Dll for Wg and Dpp signaling] in the wing discs of the frc^RY34 mutant. Expression of ptc and that of Dll in the ventral compartment of the wing discs were unaffected in the mutant clones, suggestive of normal GAG function (Yano, 2005).

Given that Notch glycosylation by Fringe (Fng), a fucose-specific ß1,3-N-acetylglucosaminyltransferase, requires Frc activity, Notch glycosylation was examined in the frc^RY34 mutant. The frc^RY34 mutant clones in the dorsal compartment, but not those in the ventral compartment, of the wing discs induce wg expression at their borders, as has been observed with fng mutant clones, suggesting that Notch glycosylation is impaired in the frc^RY34 mutant. The ectopic expression of Wg induced by the frc^RY34 mutant clones is likely responsible for the observed induction of Dll expression in the dorsal compartment (Yano, 2005).

To determine why the loss of a UDP-sugar transporter with a broad specificity selectively affects Notch glycosylation, the subcellular localization of Frc was examined. Frc tagged with the Myc epitope was expressed in imaginal discs under the control of the arm-Gal4 driver. The Gal4-induced expression of Frc-Myc rescues the frc mutant phenotype, suggesting that Frc-Myc is functional and properly localized. Immunostaining of imaginal discs of wild-type larvae expressing Frc-Myc with antibodies to Myc and to the 120-kDa protein revealed that Frc localizes to only a small subset of Golgi units. Thus, it is hypothesized that the Golgi units might be functionally heterogeneous, and that those containing Frc might modify some proteins, including Notch, but not others (Yano, 2005).

To test this hypothesis, the localizations of various molecules involved in protein modification in the Golgi complex were compared with that of Frc. It was found that Sfl is also restricted to a subset of Golgi units, but that its distribution does not overlap with that of Frc. This differential localization of Sfl and Frc might thus explain the observation that frc mutant clones in wing discs do not show any defect in GAG synthesis by Sfl (Yano, 2005).

The Spi-processing enzyme Rho was also localized to a subset of Golgi units distinct from those containing Frc, in addition to its presence in other compartments. This result indicates the existence of at least two types of Golgi units, those containing Rho and those containing Frc. To determine whether these two types of Golgi units differ functionally, the glycosylation state and function of Spi were examined in frc mutants (Yano, 2005).

Given that the extent of Notch glycosylation, as detected by wheat germ agglutinin (WGA), is markedly reduced in frc mutants compared with that in the wild-type background (Goto, 2001), whether the WGA-reactive glycan of Spi is also affected by frc mutation was examined. Myc epitope-tagged Spi was expressed in the wild type or the frc^RY34 mutant. Spi-Myc was then precipitated from larval homogenates with antibodies to Myc and was examined for its glycosylation by SDS/PAGE and subsequent blot analysis with WGA. The reactivity of the Spi glycan with WGA was similar in the frc mutant and in the wild type. Whether the frc^RY34 mutation affects the Spi glycan was examined by mobility shift analysis. The electrophoretic mobility of glycosylated Spi from the wild type was also similar to that from the frc mutant. Deglycosylation of Spi by neuraminidase, peptide-N-glycosidase (PNGase) F, and O-glycanases also increased its mobility to the same extent in wild-type and frc mutant larvae, suggesting that the core protein is not affected by the frc mutation. Together, these results indicate that the function of Frc is not necessary for formation of the Spi glycan (Yano, 2005).

Spi function was evaluated by examining developmental processes such as photoreceptor recruitment and bract formation, both of which require Spi activation. During eye development, although Spi is not necessary for the primary induction of the photoreceptor R8, it is required for the subsequent recruitment of R1 to R7. Given that photoreceptors R1 to R8 express ELAV and that R1 and R6 express Bar, the expression of these proteins was examined in frc mutants. In mutants harboring the hypomorphic allele frc^R29, all photoreceptors are normally induced, although their direction is irregular as seen in fringe or Notch mutants. Similar results were obtained by clonal analysis of frc^RY34 mutants. Spi function in photoreceptor recruitment thus did not appear to be impaired in the frc mutants. The frc^R29 mutant also formed normal bracts on malformed legs. Tests were performed for genetic interaction between rho and frc mutations in wing vein formation. The rho^ve1 mutant is viable but shows partial loss of L3-5 veins. This phenotype is also apparent in rho^ve1, frc^RY34/rho^ve1, frc⁺ flies, suggesting that Frc does not affect Rho function. From these results, it is concluded that the function of the Rho-Spi pathway is not affected by frc mutation (Yano, 2005).

To confirm that the Golgi units containing Frc and those containing Rho are distinct, whether these Golgi units could be selectively isolated was tested by using antibodies to Myc (for Myc-tagged Frc) or HA (for HA-tagged Rho). Because it is difficult to collect enough of the imaginal discs, the starting material was switched to embryos, and whether Frc and Rho are also localized to distinct Golgi units in embryos was examined. Frc-Myc and Rho-HA were coexpressed in the embryos by the arm-Gal4 driver; immunostaining with antibodies to Myc and to HA revealed that the Golgi units containing Frc-Myc (45.4% of total Golgi units) and those containing Rho-HA (43.0% of total Golgi units) are largely distinct: only 11.6% of total Golgi units were positive for both Frc-Myc and Rho-HA. Immunoisolation was attempted from embryonic lysates by using either antibody to Myc or HA and how much Frc-Myc and Rho-HA were coisolated in each immunoisolate was examined. When Frc-Myc was immunoisolated with an antibody to Myc, the recovery of Frc-Myc was 5.7 times greater than that of Rho-HA. Moreover, when Rho-HA was immunoisolated with an antibody to HA, the recovery of Rho-HA was 18.3 times greater than that of Frc-Myc. The immunoblot analysis of these immunoisolates with the anti-120-kDa antibody confirmed that the Golgi units were concentrated in these immunoisolates. These results support the notion that Frc-Myc-containing fraction is distinct and can be separated from Rho-HA-containing fraction (Yano, 2005).

Whether these distinct Golgi units contain different constituents was examined. Fringe (Fng) is one of the candidate molecules that may be colocalized with Frc. Therefore, expression of ectopically expressed Fng was examined in Rho- and Frc-containing immunoisolates. It was found that expression of Fng in Frc-containing immunoisolates was 26 times greater than in Rho-containing immunoisolates, supporting the idea that Fng is localized in the Frc-positive Golgi units rather than the Rho-positive Golgi units. It was also confirmed by immunostaining analysis that Fng colocalizes mostly with Frc (88.1% of the FNG-positive Golgi units), but not with Rho (16.6% of the Fng-positive Golgi units), by immunostaining analysis (Yano, 2005).

The data suggest that different Golgi units perform different functions, a notion that is also supported by the observation that Tn antigen (O-linked N-acetylgalactosamine) was detected in only a subset of Golgi units in imaginal eye disc cells. In addition, it was found that most of these Tn antigen-positive Golgi units are distributed in the basal region of the disc cells, suggesting that the differential distribution of Golgi units might contribute to the apicobasal polarity of glycan distribution (Yano, 2005).

In contrast to the larval stage, Frc is required for GAG synthesis at the early embryonic stage (Goto, 2001; Selva, 2001). To determine why the Frc requirement for GAG synthesis differs between the embryonic and larval stages, embryos expressing Frc-Myc were stained with antibodies to Sfl and to Myc. Sfl was found to be colocalized with Frc, likely explaining the importance of Frc for GAG synthesis at the embryonic stage. In addition, this embryonic requirement of Frc for GAG synthesis excludes the possibility that the selective defects in Notch and not in GAG synthesis observed in frc mutant larvae are caused by the selective Frc-dependent transport of a subset of UDP-sugars used only for glycosylation of Notch but not for GAGs synthesis (Yano, 2005).

It summary, these results provide evidence for the existence of functionally distinct Golgi units in Drosophila cells. Such functional heterogeneity of Golgi units is likely responsible for the diversity of protein glycosylation. At least two types of Golgi units containing either Frc or Sfl are present in larval disc cells. Two distinct sets of proteins, exemplified by Notch and GAG core proteins, might thus be selectively transported to Frc- or Sfl-containing Golgi units, respectively, where they undergo glycosylation by different sets of molecules (Yano, 2005).

The variety of Golgi units might be established by separate transport of secretory proteins and glycosylenzymes from the endoplasmic reticulum (ER) to the distinct Golgi units. In yeast, glycosylphosphatidylinositol (GPI)-anchored proteins exit the ER in vesicles distinct from those containing other secretory protein. Given that the GAG core protein Dally in Drosophila is anchored to the membrane by GPI, it is possible that Dally and Notch are loaded into distinct vesicles as they exit the ER (Yano, 2005).

Combinations of glycosylenzymes and transporters, such as Sfl and Frc, contained in Golgi units of Drosophila differ not only between embryos and larval disc cells but also among cell types. For example, Frc is localized to all Golgi units in salivary gland cells at the larval stage. It has also been shown that all of the Golgi complexes dispersed in oocytes may have the ability to process the Gurken precursor protein, which is usually cleaved in a subset of the Golgi complexes residing in the dorso-anterior region. The Golgi units may thus be altered in a manner dependent on development, cell type, and signaling processes (Yano, 2005).

The functional diversity of Golgi units also might contribute to the polarized distribution of glycans along the apicobasal axis of cells. It was found that Tn antigen is synthesized in the basal Golgi units of larval disc cells. Furthermore, certain types of glycans are distributed along the apicobasal axis of pupal ommatidia. These glycans might thus be synthesized differentially in the Golgi units that are asymmetrically distributed along the apicobasal axis and then be secreted at either the apical or basal cell surface (Yano, 2005).

Whereas Golgi units are dispersed throughout Drosophila cells, the Golgi complex in mammalian cells is thought to be a single entity that is located in the pericentriolar region through its association with the microtubule-organizing center in interphase and which is fragmented at the onset of mitosis. The Golgi fragments apparent in mammalian cells during mitosis are highly similar to the Golgi units of Drosophila cells in both electron and confocal microscopic images. The mammalian Golgi complex during interphase may therefore be comprised of functionally distinct units that are associated with the microtubule-organizing center and connected with each other (Yano, 2005).

Rhomboid family pseudoproteases use the ER quality control machinery to regulate intercellular signaling

Intramembrane proteolysis governs many cellular control processes, but little is known about how intramembrane proteases are regulated. iRhoms are a conserved subfamily of proteins related to rhomboid intramembrane serine proteases that lack key catalytic residues. This study has used a combination of genetics and cell biology to determine that these 'pseudoproteases' inhibit rhomboid-dependent signaling by the epidermal growth factor receptor pathway in Drosophila, thereby regulating sleep. iRhoms prevent the cleavage of potential rhomboid substrates by promoting their destabilization by endoplasmic reticulum (ER)-associated degradation; this mechanism has been conserved in mammalian cells. The exploitation of the intrinsic quality control machinery of the ER represents a new mode of regulation of intercellular signaling. Inactive cognates of enzymes are common, but their functions are mostly unclear; these data indicate that pseudoenzymes can readily evolve into regulatory proteins, suggesting that this may be a significant evolutionary mechanism (Zettl, 2011).

Phylogenetic analysis of rhomboids identified a subgroup of rhomboid-like proteins that lack essential catalytic residues (Lemberg, 2007). These 'iRhoms' are mysterious because, despite their predicted lack of protease activity, they are present in all sequenced metazoans, and their high degree of sequence identity implies selective pressure. Little is known about their function, but human iRhom1/Rhbdf1 has been reported to be necessary for the survival of some epithelial cancer cells (Yan, 2008) and may be linked to GPCR-mediated EGFR transactivation (Zou, 2009; Zettl, 2011).

iRhoms exemplify a more general phenomenon: the existence of conserved but catalytically inactive cognates of enzymes. The widespread occurrence of this type of predicted protein has only been apparent since genome sequences have been available, and their function is largely unknown. They are not generally encoded by pseudogenes and are therefore presumed to be expressed. Bioinformatic analysis has led to the suggestion that inactive enzyme cognates are disproportionately involved in regulatory processes (Zettl, 2011).

In Drosophila, active rhomboids are cardinal regulators of epidermal growth factor receptor (EGFR) signaling, and although it is unclear whether this activity is conserved in mammals, recent evidence supports some evolutionary conservation (Adrain, 2011). Until now, nothing has been reported about iRhom function in Drosophila. This study used a combination of genetics and cell biology to discover that Drosophila iRhom regulates EGFR signaling. The EGFR pathway in Drosophila has become a model for understanding how signaling pathways are regulated with the precision that is necessary to control their multiple developmental and physiological roles. Drosophila iRhom is expressed predominantly in neuronal cells, and its loss causes a 'sleep'-like phenotype that is indistinguishable from gain-of-EGFR signaling in the central nervous system. Consistent with this, genetic interactions show that Drosophila iRhom counteracts the function of active rhomboids, specifically acting to inhibit EGFR signaling. As well as revealing the biological role of Drosophila iRhom, this study investigated the cellular mechanism of iRhom function. Both fly and mammalian iRhoms, which are localized in the ER, inhibit secretion and intracellular levels of specific client proteins by promoting proteasomal degradation. Overall, these data imply that iRhoms regulate secretion of specific client proteins, which include the EGF family of growth factors. They can target clients for proteasomal removal by ER-associated degradation (ERAD). In this way, fundamental cellular quality control machinery is exploited as a mechanism for regulating growth factor signaling (Zettl, 2011).

Intramembrane proteases are potentially dangerous enzymes: they catalyze the irreversible cleavage of proteins that trigger important cellular processes. Their regulation is therefore paramount, but little is known about how this is achieved under physiological conditions. Nevertheless, several lines of evidence suggest that intramembrane proteases rely heavily on regulated segregation of substrate and enzyme. For example, in Drosophila, the type II membrane protein Star regulates access of substrates to rhomboids. Similarly, the trafficking of the site-2 protease substrates SREBP or ATF6 to the Golgi apparatus, the location of S2P, is highly regulated. This study reports a new mechanism, the specific destabilization of substrates in the ER, ultimately preventing access to an active rhomboid. The point is not that these diverse control strategies share common mechanisms, but that they all comply with the regulatory logic of segregating substrate and enzyme. This contrasts with the regulation of soluble proteases, in which there is greater emphasis on regulating enzyme activity. This distinction reflects the greater ability to restrict and control the cellular location of membrane proteins than soluble proteins (Zettl, 2011).

ER-associated degradation (ERAD) was first discovered as a mechanism for removing misfolded components of the T cell receptor complex. It is now clear that it contributes more generally to ER quality control and homeostasis. The current data extend the biological consequences of ERAD by demonstrating that it has been recruited in metazoans as a way of regulating intercellular signaling. Although many ERAD components have been identified, the molecular details of how proteins are targeted for recognition and retrotranslocation remain unclear. iRhoms are polytopic ER membrane proteins that can bind to at least two rhomboid substrates (this work and Nakagawa, 2005). They could enhance ERAD actively by introducing clients into the retrotranslocation machinery or passively by prolonging ER retention, thereby increasing the probability of exposure to ERAD. It is not known how many proteins iRhoms affect; genetic and cell biological data show significant specificity, but the fact that mammalian iRhoms target EGF ligands that are not rhomboid substrates implies that they have evolved additional clients (Zettl, 2011).

The data also demonstrate the physiological significance of iRhoms in Drosophila: they are specific regulators of EGFR signaling. In mammals, the data also show that iRhoms can inhibit secretion of EGF family ligands, but in contrast to flies, the physiological significance is not yet clear. The Drosophila EGF receptor is probably the most genetically well-characterized growth factor receptor in any system, and it is striking how many distinct proteins contribute to its control. The current experiments demonstrate that iRhom is a new type of regulator of EGFR activity - the most upstream control element in the pathway yet discovered. A theme that has emerged is the importance of negative feedback as a way of limiting the extent and/or amplitude of signaling. The fact that iRhom is transcriptionally activated in the developing eye imaginal disc in cells with active EGFR signaling implies that, at least in that context, iRhom could also participate in a feedback loop. Intriguingly, its expression pattern shows that, in the eye at least, its expression peaks at a time when EGFR signaling needs to be switched off (Zettl, 2011).

Degradation of EGFR ligands could therefore represent a robust way of preventing inappropriate EGFR signaling. Although the role of iRhom can be detected in the development of the wing and eye under genetically sensitized conditions, the only prominent phenotype caused by loss of Drosophila iRhom is behavioral. This is consistent with its expression pattern, which is largely restricted to the nervous system. An important open question is whether the activity phenotype of iRhom mutant flies is a consequence of a developmental or physiological defect (Zettl, 2011).

Loss of iRhom causes flies to undergo excessive periods of inactivity during the daytime, when wild-type flies are active. In fact, iRhom mutants appear to be the first loss-of-function mutations to have an excess sleep-like phenotype. Drosophila has become a genetic model to investigate the neuronal and molecular mechanisms of sleep, and the data support the recent report that abnormally high EGFR activity in the CNS leads to inactivity in flies (Foltenyi, 2007; Zettl, 2011 and references therein).

The iRhoms have evolved from rhomboid proteases that lost their catalytic activity but retained their location in the secretory pathway and the ability to bind their substrates. This allowed them to acquire new functions as specific regulators of secreted proteins, without retaining a direct mechanistic interaction with the proteases. Because many mutations will lead to loss of catalytic activity, the evolution of regulatory proteins from 'dead' enzymes might be quite common. The expression pattern, cellular location, and substrate binding capacity of such proteins provide an ideal platform for the subsequent acquisition of specific regulatory properties. Consistent with this idea is the striking existence of inactive cognates of most proteases -- indeed of many enzymes of all families. Little or nothing is known about most of these pseudoenzymes, the widespread existence of which has only become apparent as genomes have been extensively sequenced. But it is an attractive idea that many will have regulatory functions related to the enzymes from which they evolved (Zettl, 2011).

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

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