InteractiveFly: GeneBrief

vps25: Biological Overview | Developmental Biology | Effects of Mutation | Evolutionary Homologs | References

Gene name - vps25

Synonyms - CG14750, lethal (2) 44Db

Cytological map position - 44D4--5

Function - signal transduction

Keywords - endocytic sorting

Symbol - vps25

FlyBase ID: FBgn0022027

Genetic map position - 2R

Classification - tandem winged helix domains

Cellular location - cytoplasmic

NCBI links: Precomputed BLAST | EntrezGene

Cell-cell signaling coordinates proliferation of metazoan tissues during development, and its alteration can induce malignant transformation. Endocytosis regulates signaling by controlling the levels and activity of transmembrane receptors, both prior to and following ligand engagement. Vps25, a component of the endosomal sorting complex required for transport (ESCRT) machinery that regulates endocytic sorting of signaling receptors, has been identified as an unconventional type of Drosophila tumor suppressor. vps25 mutant cells in the eye disc undergo autonomous neoplastic-like transformation, but they also stimulate nonautonomous cell proliferation. Endocytic trafficking defects in vps25 cells cause endosomal accumulation of the signaling receptor Notch and enhanced Notch signaling. Increased Notch activity leads to ectopic production of the mitogenic JAK-STAT pathway ligand Unpaired, which is secreted from mutant cells to induce overproliferation of the surrounding epithelium. The data show that defects in endocytic sorting can both transform cells and, through heterotypic signaling, alter the behavior of neighboring wild-type tissue (Vaccari, 2005).

Clones mutant for vps25 in the leg disc mimic those expressing an activated form of the Dpp receptor (Thickveins, Tkv). Activated Tkv signaling represses expression of Wingless, allowing Hedgehog to activate expression of the Dpp ligand in ventral cells and generate ventral leg outgrowths. vps25 mutant clones located on the ventral side of the leg imaginal disc similarly upregulate expression of the Dpp ligand, observed with a dpp-lacZ transgene. In contrast, the dpp-lacZ transgene is not activated in vps25 mutant clones in wing discs, where the antagonistic Dpp-Wg regulatory loop does not operate and dpp-lacZ is not induced by Dpp signaling. However, mutant cells in wing discs still mimic overexpressed Tkv, revealed by upregulation of the Dpp target gene, omb-lacZ. These results indicate that Dpp receptor (Tkv) signaling is upregulated in vps25 mutant cells and is responsible for generating ectopic DV organizer activity and consequent ventral leg outgrowths (Thompson, 2005).

Metazoan animals must coordinate the proliferation of tissues to permit normal development of individual organs and the organism as a whole. Coordination of tissue growth occurs through signal transduction networks that regulate the underlying cell cycle machinery. These signal transduction networks rely on surface receptors to communicate with neighboring cells. Proper regulation of this network is essential, since aberrant cell-cell communication arising from alterations in receptor number or function can lead to uncoordinated proliferation and the formation of benign and malignant tumors (Vaccari, 2005).

While the importance of spatiotemporally controlled ligand and receptor synthesis in controlling cell-cell signaling has long been known, a recently appreciated mode of signal regulation involves endocytosis. Endocytosis of ligand-bound, activated receptors can turn off signaling by removing the receptor from the cell surface, or it can promote signaling by bringing an activated receptor in proximity to internally localized signal transduction components. In addition to these mechanisms, endocytosis can control the steady-state level of nonactivated receptors on the cell surface, which will determine the cell’s sensitivity to ligand (Vaccari, 2005).

While the stimuli for internalization vary among receptors, ubiquitination of endocytosed receptors is an emerging common theme. Ubiquitination may play a role in promoting receptor internalization per se, as well as the eventual fate of the internalized receptor. Internalized receptors generally traffic along one of two itineraries: either they are returned to the cell surface via a recycling endosome, or they are transported to the lysosome for degradation. Ubiquitination often promotes the latter route. Many ubiquitinated receptors reach a prelysosomal endocytic compartment called the multivesicular body (MVB). Inward budding of the limiting membrane then carries these receptors into internal vesicles of the MVB lumen, which permits the entire receptor to be degraded upon subsequent fusion with the lysosome (Vaccari, 2005).

The cellular machinery that sorts ubiquitinated proteins for lysosomal degradation has been identified through yeast genetic screening. Mutations in a group of genes called 'class E vps' (vesicular protein sorting) genes give rise to an expanded endosomal compartment that accumulates ubiquitinated transmembrane proteins destined for degradation. Eleven of the identified class E vps proteins form four complexes, termed Hrs/Stam and ESCRT (endosomal sorting complex required for transport) -I, -II, and -III. Extensive biochemical and cell biological experiments (Babst, 2005) have led to a model in which Hrs (see Drosophila Hrs) binds ubiquitinated cargo in the early endosome and then recruits the ESCRT-I complex from the cytoplasm (See Drosophila Tumor suppressor protein 101 as an example of an ESCRT-I protein). Endosomal recruitment brings ESCRT-I in proximity to and activates ESCRT-II, which receives the ubiquitinated cargo. ESCRT-II then induces the assembly of ESCRT-III, to which ESCRT-II passes its cargo in kind. The net result of this process is to concentrate lysosome-destined cargo into a membrane subdomain that will invaginate inward, isolating transmembrane receptors from the general cellular cytoplasm. Since receptor signaling can occur from endosomal environments as well as from the cell surface, sorting by ESCRT proteins may modulate the amplitude and kinetics of various signal transduction pathways (Vaccari, 2005).

This study shows that the ESCRT-II complex member Vps25 acts as an unconventional type of Drosophila tumor suppressor. In addition to cell-autonomous loss of cell polarity and proliferation control, mutations in vps25 cause ectopic mitogenic signaling. In mutant tissue, Notch is trapped in an endosomal compartment, and inappropriate Notch activation results. This ectopic Notch activity induces production of the secreted signal Unpaired (Upd), which stimulates excess cell division in neighboring wild-type cells. A mutation in a single gene thus causes both persistent proliferation of mutant cells and also induction of overproliferation in surrounding tissue (Vaccari, 2005).

A model is presented for tissue transformation in vps25 mosaic epithelia. In wild-type epithelial cells, Notch is endocytosed and degraded via MVB sorting in endosomes. In vps25 mutant cells, Notch is endocytosed but fails to be degraded due to impaired MVB sorting; thus, it accumulates in enlarged endosomes. vps25 mutant cells also fail to polarize, to exit the cell cycle and to differentiate; they are later eliminated by apoptosis. Due to ectopic Notch activation, vps25 mutant cells produce and secrete Upd. Via the JAK-STAT pathway, the ectopic Upd promotes extra growth of the neighboring wild-type epithelium. This heterotypic signaling process echoes aspects of the tumor-host interactions observed during malignant transformation of mammalian tissues (Vaccari, 2005).

Class E vps proteins have been studied in cultured vertebrate cells (Babst, 2005), but the early lethality of mutant mice and cell cycle arrest seen in tissue-specific inactivation have hampered functional analyses in mammals. As in yeast and mammals, loss of ESCRT-II function in flies causes accumulation of ubiquitinated proteins in an enlarged endosomal structure, indicating that the cell biological role of ESCRT-II is conserved across phylogeny. Trapping of Notch in an early endosomal compartment in hrs mutants, and in an Hrs-positive compartment in vps25 mutants, is also consistent with the ordering of class E protein functions in yeast. ESCRT-II physically interacts with both ESCRT-I and ESCRT-III, and identical phenotypes on endosomal organization and sorting are seen in yeast mutant for any ESCRT complex member. Thus, the phenotype described here probably represents that of general ESCRT absence in flies (Vaccari, 2005).

It was surprising to find strong differences between the Drosophila mutant phenotypes of vps25 and hrs. In yeast, deletion of the hrs homolog vps27 causes an endosomal and cargo accumulation phenotype that is indistinguishable from deletion of ESCRT complex members. However, the immunoreactivity to Notch, ubiquitin, and the early endosomal marker Avalanche (Avl), which is largely regular in hrs cells (Jekely, 2003), is highly irregular in vps25 cells. Loss of vps25, which acts downstream of hrs in the endosomal sorting pathway in yeast (Babst, 2002b) thus causes a more severe disruption in endosomal organization than hrs. It remains possible that other Drosophila gene products can partially substitute for Hrs or that the N-terminally truncated protein produced by the single hrs allele (Lloyd, 2002) retains some function. The distinct endosomal phenotypes of hrs and vps25 are significant given their different affects on signaling pathways and dramatically disparate functions in controlling both autonomous and nonautonomous proliferation (Vaccari, 2005).

The autonomous overproliferation of vps25 mutant cells strongly resembles that of tissues mutant for other Drosophila neoplastic tumor suppressor genes, including the endocytic syntaxin-encoding avl (Bilder, 2004; Lu, 2005). In addition to the immense increase of cell numbers seen in mutant eye discs, vps25 cells resemble avl cells in misdistribution of polarized proteins and cellular junctions, and both genes are required for the endocytic trafficking and degradation of Notch (Lu, 2005). However, their effects on Notch localization and Notch pathway activity are quite different. In vps25 cells, Notch is trapped internally in the early endosome, and high levels of Notch activity are seen. In avl tissue, where Notch is trapped at the cell surface, ectopic pathway activation is not seen despite the elevated pools of Notch protein. Similarly, ectopic Notch pathway activation is not seen in cells mutant for other characterized neoplastic tumor suppressor genes (D. Bilder, unpublished data cited in Vaccari, 2005). While a requirement for Notch in the cell-autonomous overproliferation of vps25 mutant tissue has not been directly tested, the avl phenotype suggests that altered activities of other mistrafficked membrane proteins, perhaps including the apical membrane determinant Crumbs (Lu, 2005), are responsible for the persistent proliferation of mutant cells (Vaccari, 2005).

By contrast, the nonautonomous tissue growth induced by vps25 (but not avl) cells is spurred by Notch-mediated production of the extracellular ligand Upd. Overexpression of Upd, either alone or in response to ectopic Notch expression, can cause hyperproliferation of cells anterior to the morphogenetic furrow resembling the phenotype of vps25 mosaic clones. Moreover, Upd overexpression phenotypes, like vps25 mutant phenotypes, are suppressed by STAT92E heterozygosity. Importantly, Notch also acts through Upd to regulate eye disc tissue growth during wild-type development. Thus, activation of Upd by Notch in vps25 mutant eye cells is an appropriate cellular response in an inappropriate developmental context that leads to tumorous tissue growth (Vaccari, 2005).

In mammalian immune cells, ectopic Notch pathway activity can lead to lymphomas, while, during normal development, Notch activation induces production of interleukin-4 and related molecules, which, like Upd, signal through the JAK-STAT pathway. Intriguingly, a recent report describes a critical role for Interleukin-8 expression in neovascularization of Ras-transformed human cells. While a number of cytokines are among the secreted factors that can be produced in tumors, there are no examples in which mammalian Upd orthologs specifically have been shown to modify the tumor environment of Notch-induced malignancies, for instance by promoting angiogenesis or recruiting stroma. Nevertheless, it is speculated that heterotypic signaling by Upd-like factors might alter the proliferation rates of untransformed cells relative to tumor cells, ultimately favoring tumor expansion. Further investigations will be required to establish whether the vps25 mutant phenotype represents a novel inductive mechanism relevant to mammalian tumor-host interactions (Vaccari, 2005).

Notch and its ligands are distributed widely throughout development, yet Notch activity is highly localized to specific times and places. Many posttranslational mechanisms are involved in restricting Notch activity, including ligand presentation, modification of the receptor by sugars, and proteolytic processing to create active receptor forms. Each of these processes both provides a potential point of regulation and ensures that inappropriate activation does not occur, which is critical due to the potent effects of Notch-mediated signals. The vps25 phenotype described here highlights another mechanism that prevents inappropriate activation: endocytic sorting of receptor that is being cleared from the cell surface. In wild-type cells, Notch is continuously internalized and degraded in the lysosome to maintain steady-state levels of surface expression and therefore receptor availability. For this process to accurately control signaling, the cell must ensure that the unliganded Notch is not activated during internalization. The results indicate that prevention of Notch activation requires MVB sorting, and they suggest that rapid transit through the endosomal environment is required to prevent this inappropriate activation (Vaccari, 2005).

How could the endosomal accumulation of Notch in vps25 cells lead to ectopic Notch signaling? The terminal proteolytic cleavage in Notch activation is mediated by the membrane bound γ-secretase activity provided by Presenilin and its associated proteins. The site of γ-secretase activity is controversial, with some evidence pointing toward the cell surface and other evidence pointing to an endosome. Interestingly, ubiquitination of Notch has recently been linked to both its internalization and its activation. A partially processed form of mammalian Notch requires ubiquitination for efficient γ-secretase cleavage and activation, while several Drosophila ubiquitin ligases seem to influence an endosomal sorting decision specifying degradation rather than activation of unliganded Notch. In this regard, it is notable that although hrs and vps25 mutant cells both contain elevated levels of ubiquitinated proteins, only vps25 mutants show ectopic Notch signaling. The differences in Notch signaling activity could in theory arise from differences in the amount of Notch trapped in the different endocytic mutants. However, very high amounts of Notch are present in avl mutant cells, which do not show ectopic Notch signaling (Lu, 2005). Therefore, the possibility is favored that ectopic Notch activity may be due to the locus of endocytic trapping, which differs between avl, hrs, and vps25 mutant cells. Possible mechanisms for inappropriate activation include coaccumulation of Notch and its ligands, prolonged exposure to γ-secretase, or eventual dissociation of the heterodimer in the endosomal environment. These possibilities are not mutually exclusive, and the altered organization of the vps25 endosome in addition to the absence of flux through the compartment is likely to contribute to inappropriate activation of Notch signaling. Future studies will discriminate among these possible mechanisms of ectopic Notch activation (Vaccari, 2005).

While this discusssion has concentrated on the Notch pathway, it is clear that many molecules are trapped in vps25 endosomes and that vps25 mutations are phenotypically pleiotropic due to alterations in a number of signaling pathways. For instance, STAT92E suppression of vps25 phenotypes is less complete than STAT92E suppression of overproliferation mediated by Upd alone, suggesting that additional factors contribute to vps25-induced tissue overgrowth. One candidate that merits exploration is the MAPK signaling cascade, since vps25 mutants enhance gain-of-function alleles of EGFR (Elp) and MAPK (Sem). The latter evidence is consistent with the persistent MAPK signaling described in several class E mutant tissues, including hrs in flies and TSG101 in mammals (Vaccari, 2005).

Thus, the complexity of the vps25 mutant phenotype emphasizes that endosomal sorting is a point of contact between diverse signaling pathways, and a likely regulatory nexus for normal development and for pathology. Since human tumors benefit from the coordinated disruption of multiple signaling pathways, subversion of endosomal sorting may be one susceptible route toward malignant transformation. Increasing evidence implicates defects in trafficking of specific receptors in the ontogeny of mammalian tumors. Moreover, the ESCRT-I complex member TSG101 was originally isolated for a tumor suppressive function in cultured cells, although such a role in vivo has not been established. The accessibility of Drosophila tissues, along with the availability of mutations that block specific steps of endocytic traffic, will help to elucidate how endocytosis affects metazoan signaling and the consequent effects on cell proliferation during development as well as tumorigenesis (Vaccari, 2005).


Tumor suppressor properties of the ESCRT-II complex component Vps25 in Drosophila

The Drosophila gene vps25 possesses several properties of a tumor suppressor. (1) vps25 mutant cells activate Notch and Dpp receptor signaling, inducing ectopic organizers in developing eyes and limbs and consequent overproliferation of both mutant and nearby wild-type cells. (2) As the mutant cells proliferate, they lose their epithelial organization and undergo apoptosis. Strikingly, when apoptosis of mutant cells is blocked, tumor-like overgrowths are formed that are capable of metastasis. vps25 encodes a component of the ESCRT-II complex, which sorts membrane proteins into multivesicular bodies during endocytic trafficking to the lysosome. Activation of Notch and Dpp receptor signaling in mutant cells results from an endocytic blockage that causes accumulation of these receptors and other signaling components in endosomes. These results highlight the importance of endocytic trafficking in regulating signaling and epithelial organization and suggest a possible role for ESCRT components in human cancer (Thompson, 2005).

In a genetic screen for genes controlling tissue growth, a mutation was recovered in vps25 that caused tissue outgrowths in eyes, wings, and legs of adult Drosophila when clones of mutant cells were induced during larval stages. Interestingly, these tissue overgrowths were not associated with overproliferation of the vps25 mutant clone, which instead occupied only a tiny proportion of the overgrown tissue. The size of control and vps25 mutant clones can be compared in adult eyes, where they are marked by an orange eye color encoded by the white+ transgene carried on the Piggybac transposon. These results show that growth of vps25 mutant clones is impaired cell-autonomously, but that these clones nonetheless stimulate growth of surrounding tissue non-cell-autonomously. The basis for these phenotypes was examined (Thompson, 2005).

The ability to stimulate growth of surrounding tissue is one property of an 'organizer'. Cells that form organizers release intercellular signals that can act on the cell itself and on nearby cells to drive cell proliferation. Indeed, the eye and limb outgrowths caused by vps25 mutant clones are similar to those obtained when ectopic dorsoventral (DV) axis organizers are formed in these tissues. However, the mechanisms by which DV organizers are established are different in eyes, wings, and legs (Thompson, 2005).

In the developing eye imaginal disc, the DV organizer is established at the boundary between the dorsal and ventral compartments by activation of Notch signaling. Notch signaling induces expression of the secreted signal Unpaired (Upd, a cytokine), which mediates the function of the DV organizer in driving eye growth. When clones mutant for vps25 were induced in the eye imaginal disc, Upd was ectopically expressed within the clones, indicating that vps25 mutant clones establish an ectopic DV organizer. These results suggest that Notch signaling is ectopically activated in vps25 mutant clones (Thompson, 2005).

In the developing wing imaginal disc, the DV organizer is, like the eye, established along the DV compartment boundary by the activation of Notch signaling. In this case, however, the target of Notch signaling is the secreted signal Wingless (Wg). When vps25 mutant clones were induced in the wing disc, Wg expression was ectopically activated, once again indicating that these clones have activated Notch signaling that establishes an ectopic DV organizer, leading to tissue outgrowths (Thompson, 2005).

In the developing leg, the DV axis is organized by Dpp and Wingless signals that are expressed near the anterior-posterior compartment boundary in response to Hedgehog signals. Wingless is restricted ventrally, while Dpp is stronger dorsally, due to mutual antagonism between the two signals. Clones mutant for vps25 mimic those expressing an activated form of the Dpp receptor Thickveins. Tkv signaling is upregulated in vps25 mutant cells and is responsible for generating ectopic DV organizer activity and consequent ventral leg outgrowths (Thompson, 2005).

Since ectopic Notch and Dpp signaling is sufficient to explain how vps25 mutant clones produce tissue outgrowths, attempts were made to analyze how Notch and Dpp signaling could have become activated in vps25 mutant cells. vps25 encodes a component of the ESCRT-II (Endosomal Sorting Complexes Required for Transport-II) complex, one of three protein complexes discovered in yeast to mediate a critical step during endocytic trafficking of transmembrane proteins to the lysosome (reviewed in Katzmann, 2002 and in Raiborg, 2003). Downregulation of transmembrane signaling receptors by endocytosis and degradation in the lysosome has long been suspected to be of pivotal importance in determining the level of signaling activity in cells. On their journey to the lysosome, signaling receptors are first delivered to endosomes and then sorted from the outer membrane of the endosome into internal vesicles by an inward vesiculation event of unusual topology that gives rise to the multivesicular body (MVB). The contents of the MVB can then be degraded upon fusion with the lysosome (Thompson, 2005).

The precise function of ESCRT complexes is to generate the internal vesicles of MVBs, loaded with transmembrane proteins bound for the lysosome. A key signal that determines entry of membrane proteins into MVBs is ubiquitylation. In both yeast and Drosophila, loss of ESCRT activity interferes with MVB biogenesis and causes accumulation of ubiquitylated proteins on enlarged endosomes (Jekely, 2003; Katzmann, 2002; Lloyd, 2002). vps25 mutant cells also exhibit these classical MVB sorting defects, with large amounts of ubiquitylated proteins detected on endosomes. Among the proteins that accumulate on endosomes in vps25 mutant cells are Notch and the Dpp receptor, Tkv. These results show that downregulation of Notch and Tkv receptors is prevented in vps25 mutant cells due to an endocytic blockage at the point of entry into MVBs (Thompson, 2005).

Is upregulation of Notch and Tkv sufficient to explain why Notch and Dpp signaling is activated in vps25 mutant cells? In the case of Tkv, it is known that receptor overexpression can have strong effects upon signaling, but the leg outgrowths observed require more explanation. The answer is likely to lie in the regulatory wiring of the leg imaginal disc, where expression of the Dpp ligand can be induced by activated Tkv signaling itself. This forms a positive feedback loop that drives signaling to high levels. This mechanism explains how vps25 mutant cells in the leg disc can induce expression of the Dpp ligand and generate leg outgrowths (Thompson, 2005).

In the case of Notch, simple upregulation of receptor levels has relatively mild effects on signaling in vivo. However, Notch signaling is unique in that its activating ligands and other signal transducing components are also transmembrane proteins. Indeed, upregulation of the Notch ligand Delta is found in vps25 mutant cells. Whereas overexpression of Delta alone causes cis-inhibition of Notch signaling, co-overexpression with Notch causes activation. Hence, accumulation of all of these Notch signaling components within the same endosomal compartment provides a likely explanation for why Notch signaling is activated in vps25 mutant cells (Thompson, 2005).

The effect of interfering with MVB sorting in Drosophila was first examined with mutants in hrs, which encodes an ancillary factor that recognizes ubiquitylated membrane proteins and delivers them to ESCRT complexes (Jekely, 2003; Lloyd, 2002). In yeast, all components of the ESCRT complexes and their ancillary factors (such as Hrs/Vps27) have a similar 'class E' vacuolar protein sorting (vps) mutant phenotype (reviewed in Katzmann, 2002). In Drosophila, both hrs and vps25 mutants cause MVB sorting defects, but vps25 appears to have a much stronger phenotype, particularly with respect to Notch signaling. This raises the question of whether regulation of Notch signaling is a general function of the ESCRT complexes or a specific function of Vps25 (Thompson, 2005).

This question was addressed by using a cell culture assay for Notch signaling. In these cells, transfection of the full-length Notch receptor produces a low level of signaling activity. When expression of vps25, another ESCRT-II component (vps22), or an ESCRT-III component (vps32) is knocked down by RNA interference, Notch signaling increases. In contrast, knockdown of hrs does not increase Notch signaling, which is actually decreased. Note that the effects observed in this reporter assay are a specific outcome of Notch signaling, because reporter activity depends strictly on the presence of the Notch receptor, presenilin activity (which can be blocked with the inhibitor DAPT), and the presence of binding sites for the nuclear effector of Notch, Su(H), in the reporter gene. Thus, loss of Vps25 activates Notch signaling by the normal presenilin-dependent Notch cleavage and nuclear signaling mechanism. It is concluded that Vps25 behaves like other core components of the ESCRT machinery, while Hrs is not essential for all aspects of ESCRT complex function. These findings may be explained by the ability of another MVB sorting factor, the ESCRT-I component Vps23/Tsg101, to directly recognize ubiquitylated membrane proteins, independently of Hrs (Katzmann, 2002). Indeed, mutants in Drosophila Tsg101 (Moberg, 2005) have also been found to activate Notch signaling in vivo (Thompson, 2005).

A serious obstacle to the foregoing analysis was that vps25 mutant clones survive poorly in proliferating imaginal disc epithelia. However, mutant clones survive relatively well after arrest of cell proliferation, which occurs in cells behind the morphogenetic furrow of the eye disc and in the follicular epithelium of the ovary. This finding is interesting, because it suggests that the survival defect of a vps25 mutant cell is not simply a general one, but instead is a consequence of cell proliferation (Thompson, 2005).

This result can be explained by considering the phenomenon of 'cell competition', in which faster-growing cells are able to eliminate neighboring slower-growing cells by inducing their apoptosis. If cell competition is responsible for the survival defect of vps25 mutant clones, then clonal growth should be restored by either (1) reducing the competitiveness of neighboring cells, or by (2) blocking apoptosis within the vps25 mutant clone. Both of these predictions were tested in the eye imaginal disc. Slowing the growth of neighboring cells by making them heterozygous for a Minute mutation (encoding a subunit of the ribosome) or expression of the caspase inhibitor, p35, in vps25 mutant cells completely restores growth of the vps25 mutant clone. Strikingly, these mutant clones cause massive tissue overproliferation, producing tumorous discs several times the size of normal eye discs (Thompson, 2005).

The phenotype of these tumorous discs was examined more closely, choosing smaller examples that can be visualized whole with a 40× objective lens. At this magnification, it is clear that some wild-type tissue remains, indicating that tissue overgrowth is caused by secreted signals that are released by vps25 mutant cells and consequently drive proliferation in both mutant and nearby wild-type cells. It is also clear that vps25 mutant cells lose their epithelial organization, forming three-dimensional masses of cells with a rounded, mesenchymal appearance surrounded by distorted wild-type epithelia. In addition, mutant cells are unable to differentiate, as shown by the absence of Elav, a marker of arrested, differentiating cells in the eye. These tumor-like growths are formed not only in eye discs, but also in wing and leg discs when apoptosis is blocked in vps25 mutant clones (Thompson, 2005).

Finally, whether such tumorous discs had the potential to metastasize was tested. A simple test for metastasis in Drosophila is to implant the tumor into the abdomen of an adult female host and examine whether migration of cells to secondary sites can occur. Portions of eye discs containing clones of vps25 mutant cells expressing p35 and RFP were transplanted, and, after 1 week, dead animals in which RFP cells had migrated to other parts of the fly were found in 2/15 cases. Such events did not occur when control discs were transplanted. These results suggest that vps25 mutant cells in which apoptosis has been blocked are capable of growth and metastasis to secondary sites (Thompson, 2005).

This work highlights the potential of Drosophila imaginal disc epithelia as a simple in vivo model system in which forward genetic screening approaches can be applied to identify genes controlling the behavior of clones of cells in tissues. The possibility that this system will deliver discoveries that are relevant for human cancer has been freshly appreciated. The transition from autonomous single-cell life to multicellular organisms defines the origin of metazoans and must have required the invention of controls on cell behavior that subsumed the fate of single cells to the needs of the organism as a whole. It is therefore likely that many of these controls will be shared between all animals. Cancer reflects a reversion toward an autonomous state by cells that have subverted the normal controls on cell behavior. It is clearly of interest to identify mutations that can trigger such transformations, and isolation of mutants in vps25 provides an illustrative example (Thompson, 2005).

Mutation of vps25 endows cells with two key properties involved in the malignant transformation of epithelial cells: activation of signals that drive cell proliferation and loss of epithelial organization. In addition, mutation of vps25 also renders cells sensitive to apoptosis. As far as proliferation is concerned vps25 mutant cells induce ectopic DV organizer activity in developing eye, wing, and leg imaginal disc epithelia. The mutant cells secrete intercellular signals that drive proliferation in both mutant and surrounding wild-type cells. Thus, mutation of vps25 hijacks the normal program of growth control in these tissues. The mechanism by which organizers are established differs in the three tissues, but mutation of vps25 is able to induce all three types of organizer activity. The mutant cells can achieve this because they upregulate both Notch and Dpp signaling, which is sufficient to induce ectopic DV organizers in these tissues (Thompson, 2005).

The root cause of activated Notch and Dpp signaling lies in an endocytic blockage at the point of entry into MVBs, where Vps25 is normally required as part of the ESCRT-II complex to sort transmembrane proteins into the internal vesicles of MVBs for later degradation in the lysosome. In vps25 mutant cells, transmembrane proteins accumulate to high levels in enlarged endosomes, and among these proteins are Notch and the Dpp receptor, Tkv. While upregulation of receptors can increase signaling, the dramatic phenotypes observed in vps25 mutants suggest that induction of ligand is also involved. In the case of the secreted Dpp ligand, its transcription is induced by an indirect positive feedback mechanism that is normally involved in establishing the DV organizer in leg discs. The case of Notch ligands is different, as they are broadly expressed transmembrane proteins that appear to accumulate with their receptor and, very likely, other transmembrane components of the signaling machinery in the enlarged endosomes of vps25 mutant cells, leading to productive signaling (Thompson, 2005).

Note that the cell culture system is devoid of known Notch ligands. Thus, the signaling detected upon overexpression of the full-length Notch receptor is likely to reflect a backgound level of presenilin-mediated cleavage that is independent of ligand. Knockdown of vps25 enhances this background cleavage, and this effect can be reversed by addition of the presenilin inhibitor DAPT. It is concluded that blockade of MVB sorting causes Notch to accumulate with presenilin in the same endosomal compartment, which can lead to ligand-independent cleavage and signaling in this assay (Thompson, 2005).

Metastasis begins when a tumor cell loses its epithelial organization, allowing it to exit the epithelium. In the proliferating imaginal disc epithelia, vps25 mutant cells lose their epithelial organization and drop basally from the epithelium, but this behavior is accompanied by apoptosis of the mutant cells. As a consequence, mutant cells are eventually eliminated from the growing tissue, but neighboring wild-type cells that have overproliferated are left behind as a memory of the organizer activity of mutant cells (Thompson, 2005).

Is apoptosis a consequence of the loss of epithelial organization? This study has shown that loss of epithelial organization and apoptosis occur only when mutant cells are induced in proliferating tissues. In tissues composed of arrested cells, vps25 mutant cells survive well and remain epithelial. These effects can be explained by two well-known phenomena. (1) The process of cell proliferation presents a significant challenge to epithelial integrity, with cells often rounding up within the epithelium during cell division. Consequently, mutants that interfere with a cell's epithelial organization are more likely to manifest in proliferating tissues. (2) Proliferating tissues exhibit cell competition, in which competing cells attempt to induce apoptosis of their neighbors. Apoptosis of vps25 mutant cells is caused by cell competition, because it can be rescued by reducing the competitiveness of neighboring cells. Note, however, that alleviation of cell competition does not rescue the defects in epithelial organization in vps25 mutant cells. The results indicate that loss of epithelial organization causes cells to become sensitive to cell competition, a conclusion supported by analysis of other mutations that affect epithelial integrity (Thompson, 2005).

Why should loss of epithelial organization render cells sensitive to competition? This effect could be viewed as an example of 'antagonistic pleiotropy', in which mutations that cause a potentially malignant effect also trigger a safety mechanism, such as sensitivity to apoptosis (the c-Myc oncogene being a well characterized example. Indeed, single genes whose mutation is sufficient to cause immediate and complete malignant transformation of cells may turn out to be very rare. Consequently, multiple mutations are necessary for tumorigenesis, and this analysis of vps25 provides an illustrative example (Thompson, 2005).

As far as apoptosis is concerned, the process disguises the oncogenic potential of the vps25 mutation, which is revealed upon blockade of apoptosis by a second genetic change. Apoptosis can be prevented either by alleviating competition from neighbors, or by expression of the caspase inhibitor, p35. In both cases, blockade of apoptosis allows the mutant clone to grow to form a large, amorphous mass. The complete lack of epithelial organization is clearly visible within these clones and contrasts with adjacent wild-type tissue, which overproliferates but remains epithelial. Thus, imaginal discs containing such clones are a complex mixture of mutant and wild-type cells that can grow to form a large, tumor-like mass that prevents the onset of proliferation arrest and differentiation at metamorphosis. Finally, when such discs are transplanted into an adult host, the mutant cells are capable of growth and metastasis to a second site within the host's body (Thompson, 2005).

The malignant transformations observed occur because the mutant clone possesses three key properties: it produces signals that drive cell proliferation; apoptosis is prevented; and cells lose their epithelial organization, progressing to metastasis. Note that, in imaginal discs, simply driving proliferation (for example, with activated Notch signaling) and blocking apoptosis causes tissue overgrowth, but epithelial architecture remains intact, and proliferation arrest and differentiation occur normally at metamorphosis. This work supports the view that, at least in Drosophila, loss of epithelial integrity is a critical event that synergizes with aberrant survival and proliferation to generate malignancies. The potential role of genes regulating epithelial organization in human cancer remains to be fully explored, but early results are interesting (Thompson, 2005).

The best-studied ESCRT component in mammals is Tsg101/Vps27. Drosophila Tsg101 has recently been found to have a loss-of-function phenotype similar to vps25 (Moberg, 2005). Tsg101 was first identified by antisense methods as having potential tumor suppressor activity in NIH 3T3 cells, but knockouts of Tsg101 were later shown to cause apoptosis in mouse cell lines and in mouse tissues. The consequence of blocking apoptosis in mouse cells mutant for Tsg101 has not been examined, but the results in Drosophila suggest that this may be a worthwhile line of inquiry, particularly in those tissues in which activated Notch signaling has been implicated in tumor formation (Thompson, 2005).

These findings underscore the importance of cooperative interactions between multiple genetic changes in a Drosophila model of tumorigenesis. In addition, they demonstrate that Drosophila Vps25 possesses several properties of a tumor suppressor. The basis for these properties lies in the requirement for Vps25 in ESCRT-mediated sorting of transmembrane proteins into MVBs -- a function that is conserved throughout the animal kingdom. These findings, therefore, raise the possibility that mutations in human ESCRT components could cooperate with other antiapoptotic mutations in cancer (Thompson, 2005).

Vps25 prevents nonautonomous overproliferation by regulating Notch trafficking

In a mitotic recombination-based screen for genes that control epithelial organization, a mutation called A3 was identifed that causes disruption of cell shapes in both follicular and imaginal epithelia. Follicle cells homozygous for A3 lose their cuboidal morphology and pile upon each other, particularly at the poles of egg chambers. Similarly, in genetically mosaic imaginal discs, A3 cells are round and are arranged in multilayered masses of cells. These masses consist of mutant cells immediately adjacent to the pseudostratified columnar monolayer of surrounding wild-type epithelial cells, demonstrating that loss of epithelial character is strictly cell autonomous (Vaccari, 2005).

Proper epithelial organization requires polarization of the plasma membrane into apical and basolateral membrane domains separated by adherens junctions. Polarity in A3 mutant cells was assayed by staining with antibodies against the atypical protein kinase C (aPKC), Discs-large (Dlg), and E-cadherin (Ecad), which are markers, respectively, for the apical, lateral, and junctional domains. As opposed to its polarized and restricted localization in wild-type cells, aPKC in A3 cells displays a redistribution throughout the entire cell cortex. By contrast, levels of Dlg and Ecad are reduced when compared to the surrounding wild-type cells. A3 cells thus lose apicobasal polarity and show an expansion of the apical plasma membrane in particular (Vaccari, 2005).

Loss of apicobasal polarity in A3 cells is cell autonomous and is seen both in mosaic clones as well as when eye imaginal discs consisting predominantly of mutant cells are generated by using the eyFLP/cell lethal system. Strikingly, in the latter case, in which almost all of the wild-type cells are removed from the disc, enormous overproliferation of mutant tissue, in addition to epithelial defects, is seen. Analyses of staged larvae demonstrate that A3 mutant eye discs do not grow more rapidly than wild-type eye discs, but that they continue to proliferate during a larval phase that is extended for more than 3 days after wild-type animals have pupated. Larvae containing entirely mutant A3 eye discs are giant and eventually die; they show only initial signs of pupation. The epithelial disorganization and overproliferation of A3 mutant eye discs resemble that caused by mutations in the Drosophila neoplastic tumor suppressor genes scribble, discs-large, lethal giant larvae, and, in particular, avalanche (avl), whose elimination from eye discs causes a nearly identical phenotype (Lu, 2005). These phenotypes demonstrate that A3 mutates an unidentified Drosophila tumor suppressor gene (Vaccari, 2005).

Deficiency mapping was used to localize A3 to the 44D4-5 region on chromosome 2R. A3 fails to complement Df(2R)Exel8047, limiting the A3 region to seven candidate genes. Complementation tests within this region have established that A3 also fails to complement the lethal transposon insertion K08904, which is inserted in the coding sequence of the predicted gene CG14750. Sequencing of A3 homozygous genomic DNA identified a 27 base pair deletion in the second exon of CG14750, suggesting that the A3 epithelial phenotype is caused by loss of CG14750 function. To confirm this, K08904 was recombined onto an FRT chromosome, and it was found that the K08904 mosaic phenotype is indistinguishable from that of A3. Furthermore, provision of transgenic CG14750 from an inducible heat shock promoter restores wild-type morphology to A3 mosaic clones. These data establish that the defects in A3 cells arise from disruption of CG14750 (Vaccari, 2005).

Although the predicted CG14750 protein contains no obvious motifs, BLAST searches reveal that it is homologous throughout its 174 amino acid length to the yeast protein Vps25p. Vps25p is a component of the ESCRT-II (endosomal sorting complex required for transport-II) complex that mediates protein sorting within the endosomal pathway (Babst, 2002b). The crystal structure of Vps25p shows that it contains two 'winged-helix' (WHA/B) domains (Teo, 2004). The K08904 transposon insertion is predicted to cause termination at residue 32, in the midst of the WHA domain, while the A3 deletion removes residues 109-118 at the beginning of the WHB domain. WHA mediates associations with Vps22p and Vps36p to form a functional ESCRT-II complex, while WHB interacts with the ESCRT-III complex protein Vps20p (Hierro, 2004 and Teo, 2004). Since, in yeast, ESCRT-II regulates ESCRT-III formation (Babst, 2002b), both WHA and WHB domains are likely to be required for Vps25 function. Lethal phase analysis of Drosophila trans-heterozygous between A3, K08904, and Df(2R)Exel8047 reveals that all combinations die as sluggish L1 larvae, suggesting that both alleles confer null phenotypes (Vaccari, 2005).

Yeast Vps25p and other ESCRT proteins are required for sorting of ubiquitinated membrane proteins; in mutant yeast, ubiquitinated cargo is not transported to the vacuole. To test whether CG14750 plays a Vps25p-like role in Drosophila, A3 mosaic tissues were stained with an antibody that recognizes ubiquitinated proteins. In wild-type tissue, the anti-ubiquitin signal is present only at low levels. However, in A3 mutant tissue, the anti-ubiquitin signal accumulates to high levels in intracellular puncta that stain with endosomal markers. This phenotype is consistent with a failure to route ubiquitinated cargo for degradation, and provision of transgenic CG14750 restores ubiquitinated protein clearance to the mutant tissue. These results indicate that CG14750 is both structurally and functionally homologous to yeast Vps25p. CG14750 will therefore be referred to as Vps25 (Vaccari, 2005).

To study the impact of vps25 loss on the development of epithelial-derived organs in the adult, genetically mosaic eyes were examined in which mutant clones were produced alongside wild-type cells during larval stages. Cells containing wild-type chromosomes were marked with a pigment-producing transgene such that, following recombination, only vps25 homozygous cells will lack pigment. The eyes resulting from vps25 mosaic animals consist solely of pigmented cells, indicating that vps25 mutant tissue fails to contribute to the adult eye. Surprisingly, despite the absence of mutant tissue, these eyes are dramatically overgrown and bulging. This phenotype is reminiscent of that caused by mutations in certain Drosophila hyperplastic tumor suppressor genes such as hippo, salvador, and warts/lats. However, in these latter cases, it is the mutant tissue that shows extra growth, whereas, in vps25 mosaic eyes, it is genotypically wild-type tissue that is overgrown. By further contrast, in genetically mosaic eye disc clones other neoplastic tumor suppressor genes do not affect the proliferation of neighboring wild-type cells. Thus, vps25 mutants cause excess growth of adult tissue in a manner that is distinct from other previously described Drosophila tumor suppressor genes. Subsequent work has focused on understanding this difference (Vaccari, 2005).

The absence of mutant cells but overproliferation of wild-type tissue in vps25 mosaic animals suggests that the mutant cells are eliminated but stimulate increased cell proliferation in a nonautonomous fashion. This hypothesis was tested by analyzing vps25 mosaic eye imaginal discs with markers for differentiation, proliferation, and cell death. vps25 mutant cells do not form preommatidial clusters and fail to express the neuronal marker Elav, demonstrating that they do not undergo terminal differentiation. Moreover, caspases are activated in this tissue, indicating that mutant cells are undergoing apoptosis. Lack of terminal differentiation and induction of apoptosis can account for the absence of mutant tissue in the adult eye. Nevertheless, vps25 mosaic eye discs overall are significantly larger than wild-type eye discs. The increased size of the eye disc, like that of the adult eye, is rescued by provision of transgenic Vps25. To identify the source of increased disc size, BrdU incorporation assays were carried out to mark proliferating cells. These assays reveal a strong increase in DNA synthesis in many wild-type cells surrounding large mutant clones. This increase is evident over more than ten cell diameters, and it is seen only in cells anterior to the morphogenetic furrow. Tissue size and BrdU incorporation increases are also seen in wing discs, indicating that vps25 mutant tissue stimulates proliferation of surrounding wild-type cells in multiple epithelia (Vaccari, 2005).

The involvement of Vps25p in endocytosis in yeast suggested that proliferation defects in vps25 mosaic Drosophila tissue might arise from defects in trafficking of membrane proteins. To explore this possibility, vps25 mosaic eye discs were analyzed by using immunohistochemistry against the transmembrane receptor Notch, which is highly endocytic. In mutant tissue, alterations were found in subcellular localization of Notch. In wild-type cells, Notch is found at the apical surface as well as in internal puncta. By contrast, in vps25 cells, Notch is depleted from the cell surface, and there is a strong accumulation of Notch in intracellular puncta that are larger than their wild-type counterparts. This accumulation is seen with antibodies directed against both N- and C-terminal epitopes, indicating that both the extracellular and intracellular domains of Notch are trapped in puncta (Vaccari, 2005).

The intracellular accumulation of Notch in vps25 cells could, in theory, result from a defect in either exocytic or endocytic transport. To distinguish between these possibilities, a trafficking assay was performedby labeling cell surface Notch in live eye imaginal discs. In wild-type tissue at 4°C, a restrictive temperature for endocytosis, labeling is predominantly seen on the apical plasma membrane. A 1 hr shift to a permissive temperature results in the relocalization of most of the surface labeling into intracellular puncta. After 6 hr, no labeling is apparent, indicating that most endocytosed Notch has been degraded, due to lysosomal delivery. In vps25 mutant tissue, surface-labeled Notch is removed from the plasma membrane, but it progressively accumulates in large vesicles of endocytic origin resembling the puncta seen in fixed tissue samples. Moreover, vesicular accumulation of Notch in mutant cells persists at late time points when Notch has disappeared from surrounding wild-type cells. Wild-type tissue treated with chloroquine and other inhibitors of lysosomal degradation also display persistent Notch accumulation. These trafficking experiments establish that vps25 cells can internalize Notch, but that a block in subsequent trafficking causes persistent accumulation in an endocytic compartment rather than degradation (Vaccari, 2005).

To characterize the compartment in which Notch accumulates, mutant tissue was costained with markers for endocytic structures. In wild-type cells, most Notch puncta colocalize with late endosomal markers such as Rab7GFP. Only occasional colocalization with Avl, which encodes a syntaxin localized to early endosomes (Lu, 2005) and Hrs, which is required for MVB formation, is seen, suggesting that Notch transits through early endosomes and MVBs before taking up residence in the late endosome prior to degradation. In tissues mutant for hrs, which in yeast acts upstream of ESCRT-mediated sorting, Notch accumulates primarily in the early endosome, as shown by extensive colocalization with Avl and 2XFYVE-GFP. These Avl-positive early endosomes are enlarged in hrs tissue, but they retain a spherical morphology resembling that of wild-type endosomes. By contrast, in vps25 cells, Notch accumulates in an extensive intracellular compartment that displays highly irregular morphology. This compartment is enriched for both Avl and, particularly, for Hrs. Avl and Hrs show largely complementary localization within the vps25 mutant compartment, and only a fraction of Avl and Notch overlap, as in wild-type cells. However, by contrast, substantial amounts of Notch colocalize with Hrs in vps25 mutant tissue. Loss of vps25 thus alters the endocytic distribution of Notch as well as the morphology of the endosomal compartment (Vaccari, 2005).

The endosomal accumulation of Notch in vps25 cells is provocative, since Notch overexpression itself is capable of promoting nonautonomous tissue growth in imaginal discs. It was hypothesized that vps25 cells might contain ectopic Notch activity; this was tested by assaying the expression of E(spl)mβ-lacZ, which is transcribed in response to Notch signaling. In clones of vps25 mutant cells, higher levels of E(spl)mβ-lacZ expression were observed than in the neighboring wild-type cells. This increase does not occur outside of mutant cells, suggesting that it results from effects on the Notch receptor itself, rather than on ligands on neighboring cells. Similar results were obtained by assessing the expression of E(spl)m4-lacZ, another reporter of Notch activity. Interestingly, Notch reporters are not induced in cells mutant for avl and hrs despite the increased levels of Notch protein in these cells (Lu, 2005). Ectopic Notch activity thus correlates with the ability of each endocytic mutant to promote nonautonomous tissue overgrowth (Vaccari, 2005).

The observation that vps25 mutant tissue, like Notch-overexpressing tissue, can increase proliferation in surrounding cells over a distance suggests that vps25 mutant cells emit a mitogenic signal. A good candidate for such a signal is Unpaired (Upd), a secreted cytokine-like molecule that activates the JAK-STAT pathway in flies. In the second larval instar, Upd is transcribed at the posterior margin of the eye disc in response to Notch signaling; this localized production is thought to promote the proliferation of the entire eye disc. Upd can also be induced by ectopic Notch and can cause nonautonomous proliferation of eye tissue. In wild-type eye discs, Upd expression is largely lost by the third instar. However, in vps25 mutant clones in third instar discs, Upd is found at high levels in a cell-autonomous fashion. Induction of an Upd-lacZ transgene is seen in many vps25 mutant clones, suggesting that Upd expression is predominantly due to increased transcription. To assay whether Upd is indeed involved in the excess proliferation of vps25 mosaic eyes, the effects were tested of reducing levels of STAT92E, which transduces the extracellular Upd signal to the nucleus of receiving cells. Heterozygosity for STAT92E does not reduce the frequency of vps25 clones, nor the presence of Upd in these clones. However, it partially suppresses the tissue overgrowth of both vps25 mosaic imaginal discs as well as adult eyes, implicating the Upd-activated JAK-STAT pathway in the vps25 phenotype (Vaccari, 2005).

To test whether Notch is in fact responsible for the ectopic Upd expression observed in vps25 mutant clones, the levels of Notch in vps25 clones were reduced by means of a Notch RNA interference construct (UAS-N-IR). In vps25 mutant cells expressing UAS-N-IR, Notch does not accumulate intracellularly, and its level is reduced compared to surrounding wild-type cells. Moreover, in striking contrast to the ectopic Upd expression observed in clones of vps25 mutant cells, vps25 mutant cells expressing N-IR display no Upd accumulation. Taken together, the data indicate that excess Notch activity in vps25 cells induces Upd to increase proliferation in wild-type cells (Vaccari, 2005).

vps25 mosaics display non-autonomous cell survival and overgrowth, and autonomous apoptosis

Appropriate cell-cell signaling is crucial for proper tissue homeostasis. Protein sorting of cell surface receptors at the early endosome is important for both the delivery of the signal and the inactivation of the receptor, and its alteration can cause malignancies including cancer. In a genetic screen for suppressors of the pro-apoptotic gene hid in Drosophila, two alleles of vps25, a component of the ESCRT machinery required for protein sorting at the early endosome, were identified. Paradoxically, although vps25 mosaics were identified as suppressors of hid-induced apoptosis, vps25 mutant cells die. However, evidence is provided that a non-autonomous increase of Diap1 protein levels, an inhibitor of apoptosis, accounts for the suppression of hid. Furthermore, before they die, vps25 mutant clones trigger non-autonomous proliferation through a failure to downregulate Notch signaling, which activates the mitogenic JAK/STAT pathway. Hid and JNK contribute to apoptosis of vps25 mutant cells. Inhibition of cell death in vps25 clones causes dramatic overgrowth phenotypes. In addition, Hippo signaling is increased in vps25 clones, and hippo mutants block apoptosis in vps25 clones. In summary, the phenotypic analysis of vps25 mutants highlights the importance of receptor downregulation by endosomal protein sorting for appropriate tissue homeostasis, and may serve as a model for human cancer (Herz, 2006).

The inactivation of signaling pathways is as important for appropriate tissue homeostasis as its activation. Interference with the inactivation process often gives rise to malignant phenotypes, including cancer. Several strategies to restrict signaling exist, including receptor sequestration, receptor inactivation, production of inhibitory signaling proteins and inactivation of intracellular signaling proteins. The phenotypic analysis of vps25 mutants highlights the importance of receptor downregulation by endosomal protein sorting. Lack of vps25 function causes at least three phenotypes: non-autonomous proliferation, non-autonomous resistance to cell death and autonomous apoptosis. The cause of these phenotypes and the potential role of class E Vps proteins for tumorigenesis will be discussed (Herz, 2006).

Vps25 is a component of the ESCRT-II complex required for internalization of cell surface receptors into MVBs at the early endosome. The signal for protein sorting into MVBs is provided by mono-ubiquitylation. In yeast, vps25 mutants cause aberrant endosomal structures and the accumulation of ubiquitylated proteins. A similar phenotype in vps25 clones in Drosophila, suggesting the conserved function of vps25 (Herz, 2006).

The lack of appropriate protein sorting at early endosomes in vps25 clones causes the accumulation of cell surface receptors including N and Dl. Genetic analysis using a dominant-negative N transgene (NDN) suggests that the strong overgrowth phenotype of vps25 mosaics is largely due to inappropriate N signaling, which is known to induce proliferation non-autonomously through activation of the JAK/STAT pathway (Herz, 2006).

It is unclear whether N exerts this function in a ligand-dependent manner. Dl protein also accumulates in vps25 clones, and endocytosis of Dl is required for N activation. Thus, blocking MVB formation in vps25 clones may lead to the accumulation of active Dl, resulting in increased N activity. However, it was also shown that N is required for Dl accumulation in vps25 clones. Therefore, Dl accumulation is either directly or indirectly the result of increased N activity in vps25 clones. This conclusion infers that N activation occurs before Dl accumulation and would argue in favor of a ligand-independent mechanism for N activation in vps25 clones, although Dl may be required for maintaining N activity. N activity is also controlled by several proteolytic cleavages, which lead to translocation of the intracellular domain of N to the nucleus where it regulates the expression of target genes. Thus, a potential ligand-independent mode of N activation may include inappropriate cleavage of N at the vps25 endosome. Further studies are needed to clarify this point (Herz, 2006).

Mutations in erupted, the vps23 homolog that encodes a component of ESCRT-I, give rise to similar phenotypes to those observed for vps25. However, in hrs mosaics in Drosophila, non-autonomous cell proliferation has not been observed, although signaling receptors including N and Dl accumulate in hrs clones. This is a puzzling observation as hrs encodes a class E Vps protein acting immediately upstream of the ESCRT complexes. It is possible that N and Dl are not in an environment in the hrs endosome that permits signaling. Alternatively, it has been shown showed that hrs controls the steady-state levels of non-activated receptors at the plasma membrane. Although this function may apply to vps25, it may also indicate that there are inherent differences between the different class E proteins regarding protein sorting at the early endosome (Herz, 2006).

Paradoxically, although vps25 clones die by apoptosis, the vps25 alleles were identified as being recessive suppressors of GMR-hid-induced cell death. This analysis demonstrates that the wild-type tissue accounts for this suppression even though these cells are exposed to GMR-hid. The initial explanation for this observation was that non-autonomous proliferation mediated by JAK/STAT signaling in vps25 mosaics overrides the apoptotic activity of GMR-hid. However, overexpression of Upd, the ligand of the JAK/STAT pathway, does not significantly suppress GMR-hid, although GMR-upd flies have a similar overgrowth phenotype to vps25 mosaics. This finding excludes non-autonomous proliferation for the suppression of GMR-hid by vps25. However, Diap1 protein levels are increased in tissue abutting vps25 clones. GMR-hid is sensitive to altered levels of Diap1, suggesting that the increase of Diap1 outside of vps25 clones may account for the suppression of GMR-hid. Thus, in addition to non-autonomous proliferation, vps25 clones also increase the apoptotic resistance of adjacent wild-type tissue in a non-autonomous manner. The signaling pathway that can induce non-autonomous survival by increasing Diap1 protein levels is currently unknown (Herz, 2006).

The data suggest that apoptosis in vps25 mutant tissue is not only executed via the Hid/Diap1/Dronc/Ark pathway. vps25 ark clones still died, suggesting that in addition to Ark at least one other cell death pathway is activated in vps25 clones. It has been shown that a Dronc/Ark-independent cell death pathway exists in Drosophila, but this pathway has not been identified. The data in this study implicates JNK as potential mediator of the alternative cell death pathway. vps25 ark/Puc mosaic eye discs are extremely overgrown and the clones occupy a large area of the disc. Anti-cleaved Caspase-3 *-dependent apoptosis is blocked in these clones. Only at the clonal boundaries is Caspase-3* activity still detectable, suggesting that at the interface between vps25 clones and wild-type tissue a third potential apoptotic pathway is activated (Herz, 2006).

The data show that cell competition is not sufficient to induce cell death in vps25 clones. By contrast, given the extremely large size of cell death-inhibited vps25 clones, it appears that vps25 clones have no intrinsic growth disadvantage, and have the capability to overgrow and outcompete the surrounding wild-type tissue if cell death is blocked. Thus, cell competition does not contribute significantly to the apoptotic phenotype of vps25 clones (Herz, 2006).

Hippo signaling is increased in vps25 clones. Hippo signaling can induce cell death, and, consistently, hippo mutants block cell death in vps25 clones. It is unknown how Hippo signaling is activated in vps25 clones. However, in analogy to N, a putative receptor that controls Hippo signaling may be deregulated in vps25 clones and triggers Hippo signaling. This receptor is currently unknown, but has been postulated previously. However, it should be pointed out that ESCRT components have additional functions outside of MVB protein sorting. Certain ESCRT-II members have been shown to bind to the transcriptional elongation factor ELL in order to derepress transcription by RNA polymerase II. Thus, in the absence of Vps25, transcriptional control of components of the Hippo pathway may be deregulated and contribute to cell death (Herz, 2006).

In summary, the data suggest that impaired ESCRT function leads to the accumulation of N and Dl, and possibly of a receptor controlling the Hippo pathway. These receptors control non-autonomous proliferation and autonomous apoptosis, respectively. In addition, a signaling pathway is postulated that induces non-autonomous cell survival by controlling Diap1 protein levels. Further characterization of the vps25 mutant phenotype may help to identify the postulated receptor of the Hippo pathway and the cell survival signaling pathway (Herz, 2006).

Human ESCRT components, most notably TSG101 (Vps23p), have been implicated in tumor suppression. NIH3T3 cells, depleted of Tsg101 by an antisense approach, formed colonies on soft agar and produced metastatic tumors in nude mice. However, the conditional Tsg101 knockout in mouse mammary glands did not cause the formation of tumors over a period of two years, making a role of TSG101 as tumor suppressor controversial. However, Tsg101 mutant cells are very sensitive to apoptotic death, implying that they die before they become harmful to the organism (Herz, 2006).

The phenotypical characterization of vps25 mutants in Drosophila provides an explanation for the failure to confirm TSG101 as tumor suppressor. vps25 clones need to survive over extended periods of time in order to sustain growth. Even though they induce non-autonomous proliferation, after they have died, N signaling is turned off and proliferation stops. Furthermore, the size of the adult eye of vps25 mosaics is only slightly increased when compared with wild type, and does not match the strong overgrowth phenotype of larval imaginal discs, which can be twice as large as wild-type discs. Thus, as long as vps25 clones are not resistant to their own apoptotic death, tissue repair during pupal stages may partially regress the size of the imaginal disc back to almost normal. Instead, it appears that inhibition of cell death is the triggering event for a tumorous phenotype of vps25 clones. vps25/Diap1 and vps25 ark/Puc clones can make up a large fraction of the tissue of imaginal discs, and the entire discs can be five times as large as wild-type discs (Herz, 2006).

Tumorigenesis requires multiple genetic alterations that transform normal cells progressively into malignant cancer cells. Thus, additional genetic 'hits' may be necessary to inhibit apoptosis of Tsg101 mutant cells, which may then be able to induce a similar growth phenotype to that observed for vps25. Thus, although a tumor suppressor function for Tsg101 was not confirmed in a mouse model, it still is possible that Tsg101 and other mammalian ESCRT members have tumor suppressor properties (Herz, 2006).

Common and distinct genetic properties of ESCRT-II components in Drosophila

Genetic studies in yeast have identified class E vps genes that form the ESCRT complexes required for protein sorting at the early endosome. In Drosophila, mutations of the ESCRT-II component vps25 cause endosomal defects leading to accumulation of Notch protein and increased Notch pathway activity. These endosomal and signaling defects are thought to account for several phenotypes. Depending on the developmental context, two different types of overgrowth can be detected. Tissue predominantly mutant for vps25 displays neoplastic tumor characteristics. In contrast, vps25 mutant clones in a wild-type background trigger hyperplastic overgrowth in a non-autonomous manner. In addition, vps25 mutant clones also promote apoptotic resistance in a non-autonomous manner. This study genetically characterize the remaining ESCRT-II components vps22 and vps36. Like vps25, mutants of vps22 and vps36 display endosomal defects, accumulate Notch protein and--when the tissue is predominantly mutant--show neoplastic tumor characteristics. However, despite these common phenotypes, they have distinct non-autonomous phenotypes. While vps22 mutations cause strong non-autonomous overgrowth, they do not affect apoptotic resistance. In contrast, vps36 mutations increase apoptotic resistance, but have little effect on non-autonomous proliferation. Further characterization reveals that although all ESCRT-II mutants accumulate Notch protein, only vps22 and vps25 mutations trigger Notch activity. It is concluded that the ESCRT-II components vps22, vps25 and vps36 display common and distinct genetic properties. These data redefine the role of Notch for hyperplastic and neoplastic overgrowth in these mutants. While Notch is required for hyperplastic growth, it appears to be dispensable for neoplastic transformation (Herz, 2009).

Appropriate cell/cell signaling requires both coordinated activation and inactivation of cell surface signaling receptors. Usually, the receptors are activated by ligand binding upon which they induce an intracellular response including ubiquitination of the receptor which provides the signal for receptor internalization by endocytosis. Endocytosis also controls the steady-state levels of cell surface receptors independently of ligand occupation. After endocytosis, the cell surface receptors are present at the early endosome. Because the intracellular domain of activated signaling receptors is exposed to the cytosol, the receptors are still able to signal. In fact, signaling from the endosomal location appears to be the preferred mode of several signaling pathways as it brings the receptor in close proximity to intracellular signaling complexes. To fully inactivate the signaling receptors, a second form of internalization at the limiting membrane of the early endosome is necessary to form the multi-vesicular body (MVB). In the MVB, the receptors are completely detached from the cytosol and stop signaling. Finally, the MVB fuses with lysosomes for proteolytic degradation (Herz, 2009).

Genetic studies in yeast have identified fifteen class E vps (vacuolar protein sorting) genes required for MVB formation (Raymond, 1992). These genes encode the components of four ESCRT (Endosomal Sorting Complex Required for Transport) protein complexes. Hrs (Vps27) and STAM (Hse1) form ESCRT-0, which initiates the recruitment of the signaling receptor (the cargo) to the early endosome and delivers it to ESCRT-I. From there, the cargo is transferred to ESCRT-II and then to ESCRT-III. At ESCRT-III, the receptors are internalized into MVBs. Loss of class E vps function in yeast leads to accumulation of ubiquitinated proteins on the limiting membrane of enlarged endosomes. Biochemical studies in mammalian cells have revealed a similar function for endosomal protein sorting (Herz, 2009).

The phenotypic consequences of loss of class E vps genes in the context of a multi-cellular organism have just recently been unveiled. In Drosophila, mutants in hrs, erupted (ept, encoding the ESCRT-I component vps23) and vps25 (a component of ESCRT-II) have recently been described. These mutants are characterized by enlarged endosomes which contain increased protein levels of Notch, Delta, EGFR, Patched, Smoothened, and Thickveins (the Drosophila TGF? type 1 receptor). Despite these common endosomal defects, hrs, ept and vps25 display different phenotypes at the organismal level. While hrs mosaics do not display any obvious adult phenotypes, ept and vps25 mosaics are characterized by overgrown adult eyes and heads, and overgrown larval imaginal discs due to hyperplastic proliferation. Hyperplastic proliferation refers to increased proliferation and overgrowth; however, hyperplastic cells still maintain epithelial polarity and will eventually stop proliferating. Interestingly, this hyperplastic growth does not occur in ept and vps25 mutant tissue itself. Instead, it occurs in wild-type cells immediately abutting the mutant tissue. This non-autonomous hyperplastic proliferation is caused by increased Notch activity at the ept and vps25 endosomes which stimulates neighboring cells to undergo proliferation by activating the Jak/STAT pathway. Increased Notch activity has not been observed in hrs mutants despite the accumulation of Notch protein, explaining the lack of hyperplastic overgrowth in hrs mutants (Herz, 2009).

In addition to non-autonomous hyperplastic growth in genetic mosaics, ept and vps25 mutations can cause neoplastic overgrowth. Neoplastic cells lose epithelial polarity and fail to stop proliferating giving rise to significant overgrowth. ept and vps25 mutants show neoplastic overgrowth if almost the entire imaginal disc is mutant. Neoplastic overgrowth can also be induced in vps25 mosaic tissue, if apoptosis is blocked in vps25 mutant cells. Under both conditions, neoplastic growth occurs in an autonomous manner, i.e. in the mutant tissue. These findings were significant for a better understanding of tumor formation caused by inactivation of Tsg101 (tumor susceptibility gene 101), the human vps23 homolog, which has been implicated in cervical, breast, prostate and gastrointestinal cancers (Herz, 2009).

In addition, although vps25 mutant cells undergo apoptosis, before they die they can increase the apoptotic resistance of neighboring cells through up-regulation of the apoptosis inhibitor Diap1 (Drosophila Inhibitor of Apoptosis Protein 1) (Herz, 2009).

Except for vps25, a genetic analysis of the ESCRT-II components for endosomal protein sorting in metazoan organisms has not been reported. This study characterizes and compares the mutant phenotypes of the individual components of the ESCRT-II complex, vps22 (also called larsen, vps25 and vps36 in Drosophila. The ESCRT-II complex is a heterotetramer composed of two Vps25 subunits, and one subunit each of Vps22 and Vps36. This study shows that mutant cells of the three ESCRT-II components display endosomal defects and accumulate Notch protein. Moreover, imaginal discs predominantly mutant for the three ESCRT-II components show characteristics of neoplastic tissue growth. However, despite these common defects, the phenotypic consequences of loss of vps22, vps25 and vps36 in mosaic animals are distinct. vps22 and vps25, but not vps36 mosaics show non-autonomous hyperplastic growth. In contrast, vps25 and vps36, but not vps22 mosaics strongly increase apoptotic resistance. These differences are caused by selective Notch activation. vps22 and vps25 clones display high Notch signaling activity, while vps36 clones do not, suggesting that hyperplastic growth depends on Notch signaling. However, neoplastic growth may be independent of Notch signaling. Thus, despite their intimate physical relationship, the individual ESCRT-II components are genetically not equivalent (Herz, 2009).

Endosomal defects caused by mutations in the ESCRT-II components vps22, vps25 and vps36 in Drosophila are similar. These mutant endosomes accumulate ubiquitinated proteins and signaling receptors including Notch and its ligand Delta. They also show neoplastic characteristics. However, despite these common endosomal defects, at the organismal level, vps22, vps25 and vps36 mosaic animals display distinct phenotypes. vps22 mosaics are characterized by strong non-autonomous proliferation, but not an increase in apoptotic resistance. vps36 mosaics exhibit the reverse phenotype, i.e. increased apoptotic resistance and no or only weak non-autonomous proliferation. vps25 mosaics combine both phenotypes. Thus, this analysis shows that although these components are part of the same structural complex, they are not genetically equivalent and display distinct genetic properties (Herz, 2009).

While the vps22 allele used in this study is a clear null allele, one might argue that the vps36 allele is not a null and that the observed differences are due to the hypomorphic nature of vps36. However, such an assumption does not explain why vps36 is a strong suppressor of GMR-hid, while a null allele of vps22 that causes a strong overgrowth phenotype, completely fails to suppress GMR-hid. In addition, the common phenotypes (endosomal defects giving rise to enlarged endosomes, accumulation of Notch protein, apoptosis and the neoplastic phenotype) are very similar between vps22 and vps36. Thus, it does not appear that the phenotypic differences observed between vps22 and vps36 are due to allelic strength of the mutants. Rather, they appear to be caused by intrinsic differences of the endogenous genes (Herz, 2009).

It has previously been shown that inappropriate Notch signaling is required for non-autonomous proliferation in vps25 mosaics. The current data confirm this notion for vps22 mosaics. vps22 and vps25 mutants contain increased Notch activity and heterozygosity of Notch suppresses the non-autonomous overgrowth phenotype. In contrast, vps36 mosaics do not activate Notch signaling and hence do not cause non-autonomous overgrowth. Thus, Notch activity is required for non-autonomous hyperplastic overgrowth (Herz, 2009).

It is puzzling that despite their intimate physical association in the ESCRT-II complex, loss of vps22, vps25 and vps36 affects Notch signaling differently. One possibility to explain these differences is that these mutants form distinct endosomal microenvironments which may affect signaling from the endosome differently. The resolution of labeling technologies may not be sufficient to pick up these differences in the endosomal microenvironment, but the fact that genetic differences are observed suggests that microenvironmental differences may exist. There is precedence for such a conclusion. Although hrs mutants contain abnormal endosomes leading to accumulation of Notch protein, they do not trigger Notch activity and hence no significant growth defects. Further support of the idea that Notch needs to be in a particular microenvironment at the early endosome in order to be activated comes from a study that analyzes that act upstream of the ESCRT machinery in the endosomal pathway, namely shibire, avalanche and Rab5. Mutations in these genes also result in accumulation of Notch protein, but do not activate the pathway (Herz, 2009).

Class E vps genes have also been reported to function outside of endosomal protein sorting. As such they are involved in virus budding, transcriptional control, cell cycle progression, mRNA localization and apoptosis. Therefore, it is possible that the observed genetic differences of the ESCRT-II components may be caused by distinct requirements in addition to and independently of endosomal function and possibly independently of the ESCRT-II complex and the remaining ESCRT machinery. Future work will be necessary to dissect the roles of the ESCRT-II components in processes unrelated to endosomal processing (Herz, 2009).

While inappropriate Notch signaling correlates well with non-autonomous hyperplastic growth, it does not correlate with autonomous neoplastic growth. Imaginal discs entirely mutant for vps22, vps25 and vps36 all display overgrowth and loss of cellular architecture, hallmarks of neoplastic behavior. The neoplastic phenotype has been attributed to either increased Notch signaling or to mis-localization of the apical transmembrane protein Crumbs. However, vps36 mutant discs display a very robust neoplastic phenotype, but do not activate the Notch signaling pathway significantly, suggesting that activation of Notch is not required for neoplastic growth in vps36 mutant discs. This observation is consistent with previous findings that mutations in the neoplastic tumor suppressor genes avalanche and Rab5 do not activate Notch signaling. This study has not analyzed a genetic requirement of crumbs for the neoplastic phenotype in vps22, vps25 and vps36 mutants, but that would be an interesting experiment in the future (Herz, 2009).

It is clear that the endosomal defects in ESCRT-II mutants not only affect Notch signaling. Other membrane proteins are also affected which may contribute to the neoplastic phenotype. For example, in the case of hrs and vps25, other signaling receptors such as EGFR, Tkv, Ptc and Smo accumulate at endosomes. However, it was also shown that these accumulated proteins are largely derived from the pool of unliganded receptors, suggesting that the endosomal defect affects receptor turnover which does not necessarily cause receptor activation. The only receptor known to be activated at the endosome in a ligand-independent manner is Notch. Future work will be necessary to dissect the role of Crumbs and other signaling pathways for developing the neoplastic phenotypes (Herz, 2009).

Notch signaling activates Yorkie non-cell autonomously in Drosophila

In Drosophila imaginal epithelia, cells mutant for the endocytic neoplastic tumor suppressor gene vps25 stimulate nearby untransformed cells to express Drosophila Inhibitor-of-Apoptosis-Protein-1 (DIAP-1), conferring resistance to apoptosis non-cell autonomously. This study shows that the non-cell autonomous induction of DIAP-1 is mediated by Yorkie, the conserved downstream effector of Hippo signaling. The non-cell autonomous induction of Yorkie is due to Notch signaling from vps25 mutant cells. Moreover, activated Notch in normal cells is sufficient to induce non-cell autonomous Yorkie activity in wing imaginal discs. These data identify a novel mechanism by which Notch promotes cell survival non-cell autonomously and by which neoplastic tumor cells generate a supportive microenvironment for tumor growth (Graves, 2012).

This study identifies a novel role of Notch signaling for non-cell autonomous control of apoptosis via induction of Yki activity in neighboring cells. It has previously been shown that Notch signaling controls cell proliferation both autonomously and non-cell autonomously in the developing eye. The non-cell autonomous component of proliferation control was attributed to Notch-dependent activation of Jak/Stat signaling although that recently came into question. Nevertheless, Jak/Stat activation is not sufficient to mediate the effect of Notch on non-cell autonomous control of apoptosis. This study identified the Hpo/Wts/Yki pathway as a target of Notch signaling for the non-cell autonomous control of apoptosis both in eye and wing imaginal discs. Because the Hpo/Wts/Yki pathway also controls proliferation, it is likely that Notch promotes non-cell autonomous proliferation through both Jak/Stat and Hpo/Wts/Yki activities (Graves, 2012).

It is also interesting to note that this non-autonomous control of the Hpo/Wts/Yki pathway by Notch occurs in a position-dependent manner. For example, vps25 mutant clones or NICD-expressing clones located in the hinge and notum of wing discs triggered non-cell autonomous up-regulation of ex-lacZ, while clones in the wing pouch did not. Additionally, vps25 mutant clones located anterior to the morphogenetic furrow triggered non-cell autonomous up-regulation of ex-lacZ, while clones in the posterior of the eye disc did not. The reason for this position-dependence is unknown. However, the regions which do not induce Hpo/Wts/Yki signaling non-autonomously correspond to the zone of non-proliferating (ZNP) cells in the wing disc and post-mitotic, differentiating cells in the eye disc. Therefore, one potential reason for the position-dependence may be that the post-mitotic nature of the cells in the ZNP of the wing pouch and in the posterior of the eye disc render them inert to growth-promoting signals that trigger the Hpo/Wts/Yki pathway. However, while this is one possibility, there may also be additional mechanisms that influence the response to growth-promoting signals (Graves, 2012).

How Notch exerts this non-autonomous effect is an important and interesting question. Based on its function as a transcriptional regulator, it is possible that increased Notch signaling in vps25 mutant cells could lead to transcription of a secreted or transmembrane protein that communicates to surrounding tissue and induces Yki activity. Expression of proteins known to non-cell autonomously activate Yki signaling such as Fat, Dachsous (Ds) and Four-Jointed as well as ds-lacZ, however, are not altered in vps25 mosaic discs. Identification of this non-cell autonomous signaling mechanism may also be critical for understanding tumorigenesis, as mutations in the Notch pathway, the Hippo pathway and in ESCRT components have been implicated in many different types of human cancer. In conclusion, this study provides a mechanism by which neoplastic cells influence the behavior of neighboring wild-type cells, which may be critical for generating a supportive microenvironment for tumor growth by preventing cell death and promoting the proliferation of wild-type cells (Graves, 2012).

De-regulation of JNK and JAK/STAT signaling in ESCRT-II mutant tissues cooperatively contributes to neoplastic tumorigenesis

Multiple genes involved in endocytosis and endosomal protein trafficking in Drosophila have been shown to function as neoplastic tumor suppressor genes (nTSGs), including Endosomal Sorting Complex Required for Transport-II (ESCRT-II) components vacuolar protein sorting 22 (vps22), vps25, and vps36. However, most studies of endocytic nTSGs have been done in mosaic tissues containing both mutant and non-mutant populations of cells, and interactions among mutant and non-mutant cells greatly influence the final phenotype. Thus, the true autonomous phenotype of tissues mutant for endocytic nTSGs remains unclear. This study shows that tissues predominantly mutant for ESCRT-II components display characteristics of neoplastic transformation and then undergo apoptosis. These neoplastic tissues show upregulation of JNK, Notch, and JAK/STAT signaling. Significantly, while inhibition of JNK signaling in mutant tissues partially inhibits proliferation, inhibition of JAK/STAT signaling rescues other aspects of the neoplastic phenotype. This is the first rigorous study of tissues predominantly mutant for endocytic nTSGs and provides clear evidence for cooperation among de-regulated signaling pathways leading to tumorigenesis (Woodfield, 2013).

While it is well established how de-regulated signaling pathways in ESCRT-II mutant clones mediate non-cell autonomous interactions with neighboring non-mutant cells to contribute to hyperplastic overgrowth and increased cell survival, it was largely unknown which signaling pathways trigger neoplastic transformation autonomously. To address this question, predominantly mutant eye-antennal imaginal discs were generated in which competitive interactions are eliminated so that it was possible to examine the autonomous results of de-regulated signaling (Woodfield, 2013).

Overall, it appears that the same signaling pathways that are induced in mosaic clones are also activated in predominantly mutant tissues. However, two results of this study are noteworthy. First, it is surprising that JNK activity is strongly induced in tissues predominantly mutant for ESCRT-II genes. This is surprising because JNK signaling was believed to be induced by cell competition from neighboring non-mutant cells in mosaic tissues. However, non-mutant tissue is largely eliminated by the ey-FLP/cl method and thus competitive interactions are eliminated. Therefore, it is not known how JNK signaling is induced in these tissues. Nevertheless, JNK signaling is critical for the overgrowth phenotype of predominantly ESCRT-II mutant eye discs as inhibition of this pathway partially blocks cell proliferation. Second, de-regulation of the JAK/STAT signaling pathway is critical for the neoplastic transformation of vps22 mutant discs. Loss of JAK/STAT signaling dramatically normalizes the neoplastic phenotype of vps22 mutant cells. In addition to JNK and JAK/STAT activity, Notch activity was also found to be increased in discs predominantly mutant for ESCRT-II genes. Therefore, a genetic requirement of Notch signaling was tested for neoplastic transformation of ESCRT-II mutant cells. However, loss of Notch was inconclusive because even the wild-type control discs did not grow when Notch was inhibited (Woodfield, 2013).

Interestingly, although ESCRT-II mutant tissues undergo neoplastic transformation, they also show high levels of apoptosis. Animals with predominantly mutant eye-antennal imaginal discs die as headless pharate pupae, a phenotype likely caused by the apoptosis of the imaginal discs before the adult stage. Reduction of JNK signaling in vps22, vps25, or vps36 mutant discs leads to lower levels of apoptosis, supporting a role for JNK signaling in the cell death of the predominantly mutant tissues. More excitingly, JNK also controls proliferation in these tissues, as shown by the reduction of proliferation seen when JNK signaling was down-regulated. This observation is consistent with previous findings that JNK can induce non-cell autonomous proliferation and that apoptosis-induced proliferation is mediated by JNK activity. While inhibition of JNK signaling reduces proliferation in predominantly mutant ESCRT-II mutant discs, it does not affect other aspects of the neoplastic phenotype (Woodfield, 2013).

The role of JAK/STAT signaling in these mutants is complex. In mutant clones of ESCRT-II mosaic discs, Notch-induced secretion of the JAK/STAT ligand Upd triggers non-cell autonomous proliferation. However, autonomous de-regulated JAK/STAT signaling observed in predominately mutant discs is critical for the neoplastic transformation of vps22 mutants. In vps22 Stat92E double mutant discs, organization of cellular architecture is definitively rescued with the layout of the tissue closely resembling that of a wild-type eye-antennal imaginal disc. In addition, apical-basal polarity markers are localized more-or-less correctly in these tissues, indicating that epithelial polarity is more intact. Finally, differentiation in the posterior portion of the eye disc is preserved when JAK/STAT signaling is inhibited. Thus, de-regulation of JAK/STAT signaling in vps22 mutant discs contributes to the cellular disorganization and the lack of differentiation seen in the tissues, which is consistent with a previous study that implicated JAK/STAT signaling in cell cycle control, cell size, and epithelial organization in tsg101 mutant tissues (Woodfield, 2013).

It was recently shown that cells with strong gain of JAK/STAT activity transform into supercompetitors and eliminate neighboring cells with normal JAK/STAT activity by cell competition. However, in mosaic discs, a supercompetitive behavior of ESCRT-II mutant cells has not been observed. In fact, these mutant cells are eliminated by apoptosis. Only if apoptosis is blocked in these cells, is a strong overgrowth phenotype with neoplastic characteristics observed. Thus, apoptosis can serve as a tumor suppressor mechanism to remove cells with potentially malignant JAK/STAT activity (Woodfield, 2013).

How endosomal trafficking specifically regulates JAK/STAT signaling and, thus, how blocking trafficking leads to increases in signaling pathway activity are interesting questions to answer in the future. It is possible that, like endocytic regulation of the Notch receptor, the endosomal pathway tightly regulates Domeless (Dome), the JAK/STAT pathway receptor. It has been shown previously that Dome is trafficked through the endocytic machinery and that this trafficking of Dome can affect the downstream output of the JAK/STAT signaling pathway. It is also possible that Notch-induced Upd secretion causes autocrine JAK/STAT signaling in these mutants. However, technical problems (knocking down Notch function both in wild-type and mutant tissue causes general problems in tissue growth) prevented examination of this possibility (Woodfield, 2013).

It will be important to examine how de-regulated JAK/STAT signaling in ESCRT-II mutants causes neoplastic transformation. JAK/STAT signaling is known to be an oncogenic pathway in Drosophila and in humans but its downstream targets that promote tumorigenesis are not yet clear. JAK/STAT signaling may be feeding into other pathways that promote tumorigenesis, such as dpp signaling, or may be targeting other proteins involved in transformation, such as Cyclin D (Woodfield, 2013).

A number of studies have implicated genes that function in endocytosis and endosomal protein sorting as tumor suppressors in human cancers. Most well known is Tsg101, as early studies showed that downregulation of Tsg101 (see Drosophila TSG101) promotes the growth of mouse 3T3 fibroblasts in soft aga. When these cells were injected into nude mice, they formed metastatic tumors. However, later studies have shown conflicting results, and it is still unclear if Tsg101 functions as a tumor suppressor in metazoans. Importantly, a number of studies have shown changes in expression of ESCRT components in human cancer cells, including changes in expression of ESCRT-I components Tsg101 and Vps37A and ESCRT-III components Chmp1A and CHMP3. Since the primary proteins that function in endocytosis and endosomal trafficking are conserved from yeast to humans, it is likely that these findings in Drosophila may have important implications for human disease (Woodfield, 2013).


ESCRT-II in yeast

Sorting of ubiquitinated endosomal membrane proteins into the MVB pathway is executed by the class E Vps protein complexes ESCRT-I, -II, and -III, and the AAA-type ATPase Vps4. This study characterizes ESCRT-II, a soluble approximately 155 kDa protein complex formed by the class E Vps proteins Vps22, Vps25, and Vps36. This protein complex transiently associates with the endosomal membrane and thereby initiates the formation of ESCRT-III, a membrane-associated protein complex that functions immediately downstream of ESCRT-II during sorting of MVB cargo. ESCRT-II in turn functions downstream of ESCRT-I, a protein complex that binds to ubiquitinated endosomal cargo. It is proposed that the ESCRT complexes perform a coordinated cascade of events to select and sort MVB cargoes for delivery to the lumen of the vacuole/lysosome (Babst, 2002b).

The multivesicular-body (MVB) pathway delivers transmembrane proteins and lipids to the lumen of the endosome. The multivesicular-body sorting pathway has crucial roles in growth-factor-receptor downregulation, developmental signalling, regulation of the immune response and the budding of certain enveloped viruses such as human immunodeficiency virus. Ubiquitination is a signal for sorting into the MVB pathway, which also requires the functions of three protein complexes, termed ESCRT-I, -II and -III (endosomal sorting complex required for transport). This study reports the crystal structure of the core of the yeast ESCRT-II complex, which contains one molecule of the Vps protein Vps22, the carboxy-terminal domain of Vps36 and two molecules of Vps25, and has the shape of a capital letter 'Y'. The amino-terminal coiled coil of Vps22 and the flexible linker leading to the ubiquitin-binding NZF domain of Vps36 both protrude from the tip of one branch of the 'Y'. Vps22 and Vps36 form nearly equivalent interactions with the two Vps25 molecules at the centre of the 'Y'. The structure suggests how ubiquitinated cargo could be passed between ESCRT components of the MVB pathway through the sequential transfer of ubiquitinated cargo from one complex to the next (Hierro, 2004).

ESCRT-I, -II, and -III protein complexes are sequentially recruited to endosomal membranes, where they orchestrate protein sorting and MVB biogenesis. In addition, they play a critical role in retrovirus budding. Structural understanding of ESCRT interaction networks is largely lacking. The 3.6 Å structure of the yeast ESCRT-II core presented in this study reveals a trilobal complex containing two copies of Vps25, one copy of Vps22, and the C-terminal region of Vps36. Unexpectedly, the entire ESCRT-II core consists of eight repeats of a common building block, a 'winged helix' domain. Two PPXY-motifs from Vps25 are involved in contacts with Vps22 and Vps36, and their mutation leads to ESCRT-II disruption. Purified ESCRT-II binds directly to the Vps20 component of ESCRT-III. Surprisingly, this binding does not require the protruding N-terminal coiled-coil of Vps22. Vps25 is the chief subunit responsible for Vps20 recruitment. This interaction dramatically increases binding of both components to lipid vesicles in vitro (Teo, 2004).

Down-regulation of plasma membrane receptors via the endocytic pathway involves their monoubiquitylation, transport to endosomal membranes and eventual sorting into multi vesicular bodies (MVB) destined for lysosomal degradation. Successive assemblies of endosomal sorting complexes required for transport (ESCRT-I, -II and III) largely mediate sorting of plasma membrane receptors at endosomal membranes, the formation of multivesicular bodies and their release into the endosomal lumen. In addition, the human ESCRT-II has been shown to form a complex with RNA polymerase II elongation factor ELL in order to exert transcriptional control activity. The crystal structure of Vps25 is reported at 3.1 Å resolution. Vps25 crystallizes in a dimeric form and each monomer is composed of two winged helix domains arranged in tandem. Structural comparisons detect no conformational changes between unliganded Vps25 and Vps25 within the ESCRT-II complex composed of two Vps25 copies and one copy each of Vps22 and Vps36. This structural analysis present a framework for studying Vps25 interactions with ESCRT-I and ESCRT-III partners. Winged helix domain containing proteins have been implicated in nucleic acid binding and it remains to be determined whether Vps25 has a similar activity which might play a role in the proposed transcriptional control exerted by Vps25 and/or the whole ESCRT-II complex (Wernimont, 2004).

Mammalian ESCRT-II complex

CHMP6 (charged multivesicular body protein 6) is a human orthologue of yeast Vps (vacuolar protein sorting) 20, a component of ESCRT-III. Various CHMP6 orthologues in organisms ranging from yeast to humans contain the N-myristoylation consensus sequence at each N-terminus. Metabolic labelling of HEK-293 (human embryonic kidney) cells showed the incorporation of [3H]myristate into CHMP6 fused C-terminally to GFP (green fluorescent protein) (CHMP6-GFP). Interactions of CHMP6 with another ESCRT-III component CHMP4b/Shax [Snf7 (sucrose non-fermenting 7) homologue associated with Alix] 1, one of three paralogues of human Vps32/Snf7, and with EAP20 (ELL-associated protein 20), a human counterpart of yeast Vps25 and component of ESCRT-II, were observed by co-immunoprecipitation of epitope-tagged proteins expressed in HEK-293 cells. The in vitro pull-down assays using their recombinant proteins purified from Escherichia coli demonstrated direct physical interactions which were mediated by the N-terminal basic half of CHMP6. Overexpressed CHMP6-GFP in HeLa cells exhibited a punctate distribution throughout the cytoplasm especially in the perinuclear area, as revealed by fluorescence microscopic analysis. Accumulation of LBPA (lysobisphosphatidic acid), a major phospholipid in internal vesicles of an MVB (multivesicular body), was observed in the CHMP6-GFP-localizing area. FLAG-tagged EAP20 distributed diffusely, but exhibited a punctate distribution on co-expression with CHMP6-GFP. Overexpression of CHMP6-GFP caused reduction of transferrin receptors on the plasma membrane surface, but caused their accumulation in the cytoplasm. Ubiquitinated proteins and endocytosed EGF continuously accumulated in CHMP6-GFP-expressing cells. These results suggest that CHMP6 acts as an acceptor for ESCRT-II on endosomal membranes and regulates cargo sorting (Yorikawa, 2005).


Search PubMed for articles about Drosophila vps25

Babst, M., Odorizzi, G., Estepa, E. J., and Emr, S. D. (2000). Mammalian tumor susceptibility gene 101 (TSG101) and the yeast homologue, Vps23p, both function in late endosomal trafficking. Traffic 1: 248-258. 11208108

Babst, M., et al. (2002a). Emr, Escrt-III: an endosome-associated heterooligomeric protein complex required for mvb sorting, Dev. Cell 3: 271-282. 12194857

Babst, M., Katzmann, D. J., Snyder, W. B., Wendland, B. and Emr, S. D. (2002b). Endosome-associated complex, ESCRT-II, recruits transport machinery for protein sorting at the multivesicular body. Dev. Cell 3(2): 283-9. 12194858

Babst, M. (2005). A protein's final ESCRT. Traffic 6: 2-9. 15569240

Bilder, D. (2004). Epithelial polarity and proliferation control: links from the Drosophila neoplastic tumor suppressors. Genes Dev. 18(16): 1909-25. 15314019

Graves, H. K., Woodfield, S. E., Yang, C. C., Halder, G. and Bergmann, A. (2012). Notch signaling activates Yorkie non-cell autonomously in Drosophila. PLoS One 7: e37615. PubMed ID: 22679484

Herz, H. M., et al. (2006). vps25 mosaics display non-autonomous cell survival and overgrowth, and autonomous apoptosis. Development 133(10): 1871-80. PubMed Citation: 16611691

Herz, H. M., Woodfield, S. E., Chen, Z., Bolduc, C. and Bergmann, A. (2009). Common and distinct genetic properties of ESCRT-II components in Drosophila. PLoS One. 4(1): e4165. PubMed Citation: 19132102

Hierro, A., et al. (2004). Structure of the ESCRT-II endosomal trafficking complex. Nature 431: 221-225. 15329733

Jékely, G. and Rørth, P. (2003). Hrs mediates downregulation of multiple signalling receptors in Drosophila EMBO reports 4: 1163-1168. 14608370

Katzmann, D. J., Odorizzi, G. and Emr, S. D. (2002). Receptor downregulation and multivesicular-body sorting. Nat. Rev. Mol. Cell Biol. 3(12): 893-905. 12461556

Lloyd, T. E., Atkinson, R., Wu, M. N., Zhou, Y., Pennetta, G. and Bellen, H. J. (2002). Hrs regulates endosome membrane invagination and tyrosine kinase receptor signaling in Drosophila. Cell 108(2): 261-9. 11832215

Lu, H. and Bilder, D. (2005). Endocytic control of epithelial polarity and proliferation in Drosophila. Nat. Cell Biol. 7(12): 1132-9. 16258546

Moberg, K. H., Schelble, S., Burdick, S. K. and Hariharan, I. K. (2005). Mutations in erupted, the Drosophila ortholog of mammalian tumor susceptibility gene 101, elicit non-cell-autonomous overgrowth. Dev. Cell 9(5): 699-710. 16256744

Raiborg, C., Rusten, T. E. and Stenmark, H. (2003). Protein sorting into multivesicular endosomes. Curr. Opin. Cell Biol. 15(4): 446-55. 12892785

Raymond, C. K., Howald-Stevenson, I., Vater, C .A. and Stevens, T.H. (1992). Morphological classification of the yeast vacuolar protein sorting mutants: evidence for a prevacuolar compartment in class E vps mutants. Mol. Biol. Cell 3: 1389-1402. PubMed Citation: 1493335

Teo, H., Perisic, O., Gonzalez, B. and Williams, R. L. (2004). ESCRT-II, an endosome-associated complex required for protein sorting: crystal structure and interactions with ESCRT-III and membranes. Dev. Cell 7(4): 559-69. 15469844

Thompson, B. J., Mathieu, J., Sung, H. H., Loeser, E., Rorth, P. and Cohen, S. M. (2005). Tumor suppressor properties of the ESCRT-II complex component Vps25 in Drosophila. Dev. Cell 9(5): 711-20. 16256745

Vaccari, T. and Bilder, D. (2005). The Drosophila tumor suppressor vps25 prevents nonautonomous overproliferation by regulating notch trafficking. Dev. Cell 9(5): 687-98. 16256743

Wernimont, A. K. and Weissenhorn, W. (2004). Crystal structure of subunit VPS25 of the endosomal trafficking complex ESCRT-II. BMC Struct. Biol. 4(1): 10. 15579210

Woodfield, S. E., Graves, H. K., Hernandez, J. A. and Bergmann, A. (2013). De-regulation of JNK and JAK/STAT signaling in ESCRT-II mutant tissues cooperatively contributes to neoplastic tumorigenesis. PLoS One 8: e56021. PubMed ID: 23418496

Yorikawa, C., et al. (2005). Human CHMP6, a myristoylated ESCRT-III protein, interacts directly with an ESCRT-II component EAP20 and regulates endosomal cargo sorting. Biochem. J. 387(Pt 1): 17-26. 15511219

Biological Overview

date revised: 12 January 2018

Home page: The Interactive Fly © 2017 Thomas Brody, Ph.D.