The Interactive Fly
Zygotically transcribed genes
Distinct functional units of the Golgi complex in Drosophila cells
AP-1 and clathrin are essential for secretory granule biogenesis in Drosophila
Rab3-GAP controls the progression of synaptic homeostasis at a late stage of vesicle release
A striking variety of glycosylation occur in the Golgi complex in a protein-specific manner, but how this diversity and specificity are achieved remains unclear. This study shows that stacked fragments (units) of the Golgi complex dispersed in Drosophila imaginal disc cells are functionally diverse. The UDP-sugar transporter Fringe-connection (Frc) is localized to a subset of the Golgi units distinct from those harboring Sulfateless (Sfl), which modifies glucosaminoglycans (GAGs), and from those harboring the protease Rhomboid (Rho), which processes the glycoprotein Spitz (Spi). Whereas the glycosylation and function of Notch are affected in imaginal discs of frc mutants, those of Spi and of GAG core proteins are not, even though Frc transports a broad range of glycosylation substrates, suggesting that Golgi units containing Frc and those containing Sfl or Rho are functionally separable. Distinct Golgi units containing Frc and Rho in embryos could also be separated biochemically by immunoisolation techniques. Tn-antigen glycan is localized only in a subset of the Golgi units distributed basally in a polarized cell. It is proposed that the different localizations among distinct Golgi units of molecules involved in glycosylation underlie the diversity of glycan modification (Yano, 2005).
The pattern of glycosylation is extremely diverse, yet is highly specific to each protein. How can this specificity (and diversity) be achieved? There are >300 glycosylenzymes in humans and >100 in Drosophila, but is their enzymatic specificity sufficient to explain the precise modification of all substrates? One possible mechanism that might also contribute to the specific (and diverse) pattern of glycosylation would be the localization/compartmentalization of glycosylenzymes (Yano, 2005).
The Golgi complex, where protein glycosylation takes place, has been regarded as a single functional unit, consisting of cis-, medial-, and transcisternae in mammalian cells. However, the three-dimensional reconstruction of electron microscopic images of the mammalian Golgi structure has suggested the existence of more than one Golgi stack, with the individual stacks being connected into a ribbon by tubules bridging equivalent cisternae. Furthermore, during mitosis, the Golgi cisternae of mammalian cells become fragmented without their disassembly. In Drosophila, Golgi cisternae are stacked but are not connected to form a ribbon at the embryonic and pupal stages even during interphase, although there has been no evidence to date to indicate functional differences among the Golgi fragments (Yano, 2005).
The Golgi complex is a stack of cis-, medial-, and transcisternae in mammalian cells. In contrast, Golgi markers often do not overlap with each other in Saccharomyces cerevisiae, in which the Golgi cisternae are not stacked but disassembled. The Golgi cisternae of Drosophila are stacked but are not connected to form a ribbon at the embryonic and pupal stages even during interphase. To determine whether Drosophila imaginal disc cells have assembled or disassembled Golgi cisternae, the localizations were compared of the cis-cisternal marker dGM130, the transcisternal marker Syntaxin16 (Syx16), and the Golgi-tethered 120-kDa protein, which is commonly used to detect the Golgi complex in Drosophila. The 120-kDa protein was identified by immunoaffinity purification and protein sequencing as a Drosophila homolog of the vertebrate 160-kDa medial Golgi sialoglycoprotein (MG160), which resides uniformly in the medial-cisternae of the Golgi apparatus in vertebrate cells. An antibody specific for the 120-kDa protein also stained numerous Golgi fragments in imaginal disc cells. More than 80% of immunoreactivity for the 120-kDa protein was colocalized with both dGM130 and Syx16, suggesting that 120-kDa protein-positive fragments of the Golgi complex indeed comprise assembled cisternae; these fragments are referred to as 'Golgi units.' The distributions of the 120-kDa protein, dGM130, and peanut agglutinin (PNA), another transcisternal marker, also show that the markers are closely apposed but not identical, suggesting that the Golgi units are polarized. Interestingly, most of the PNA-positive transcisternae are oriented toward the basal side of the cell, within the Golgi complex, whereas most of the GM130-positive cis-cisternae are oriented toward the apical side of the cell. The cis-to-trans polarity of each Golgi unit thus appears to be correlated with the apico-basal polarity of the disc cells (Yano, 2005).
Drosophila mutant larvae defective in the UDP-sugar transporter Frc manifest a highly selective phenotype: the lack of Notch glycosylation in the presence of normal GAG synthesis (Goto, 2001). This limited phenotype is unexpected, given that Frc exhibits a broad specificity for UDP sugars used in the synthesis of various glycans including N-linked types, GAGs, and mucin types. However, given that the frcR29 allele studied previously (Goto, 2001) is hypomorphic, whether the selective glycosylation defect might be a consequence of partial loss of Frc activity was examined. With the use of imprecise excision, a new allele was generated, frcRY34, the presence of which results in the death of most larvae during the second-instar stage, much earlier than the death induced by frcR29. Real-time PCR analysis revealed that the amount of frc transcripts in the second-instar larvae of frcRY34 or frcR29 mutants was 4.2% and 24.4% of that in the wild type, respectively. About 1 kb of the gene, including the transcription initiation site, was deleted in the frcRY34 allele. Together, these observations suggest that frcRY34 is essentially a null allele (Yano, 2005).
Clonal cells of the frcRY34 mutant exhibited normal levels of GAGs, as detected by immunostaining with the 3G10 antibody, whereas the amount of GAGs was reduced in clones of tout-velu (ttv) mutant cells. Given that GAGs are required for signaling by Hedgehog (Hh), Wingless (Wg), and Decapentaplegic (Dpp), the expression of corresponding target genes [patched (ptc) for Hh signaling and Dll for Wg and Dpp signaling] was examined in the wing discs of the frcRY34 mutant. Expression of ptc and that of Dll in the ventral compartment of the wing discs were unaffected in the mutant clones, suggestive of normal GAG function (Yano, 2005).
Given that Notch glycosylation by Fringe (FNG), a fucose-specific beta1,3-N-acetylglucosaminyltransferase, requires Frc activity, Notch glycosylation was examined in the frcRY34 mutant. The frcRY34 mutant clones in the dorsal compartment, but not those in the ventral compartment, of the wing discs induced wg expression at their borders, suggesting that Notch glycosylation is impaired in the frcRY34 mutant. The ectopic expression of Wg induced by the frcRY34 mutant clones is likely responsible for the observed induction of Dll expression in the dorsal compartment (Yano, 2005).
To determine why the loss of a UDP-sugar transporter with a broad specificity selectively affects Notch glycosylation, the subcellular localization of Frc was investigated. Frc tagged with the Myc epitope was expressed in imaginal discs under the control of the arm-Gal4 driver. The Gal4-induced expression of Frc-Myc rescues the frc mutant phenotype, suggesting that Frc-Myc is functional and properly localized. Immunostaining of imaginal discs of wild-type larvae expressing Frc-Myc with antibodies to Myc and to the 120-kDa protein revealed that Frc is localized to only a small subset of Golgi units. This differential immunostaining of different Golgi units is not likely to be due to differential penetration of the antibodies or cripticity of the epitopes. The penetration of antibodies would not vary within the cell, because the Golgi units were distributed evenly throughout the cell, not in a biased manner. Moreover, it is unlikely that degradation of the epitopes during the immunostaining experiments due to contaminating proteases might alter the cripticity of the epitopes in different Golgi units, since the percentage of different Golgi units among the anti-120-kDa-positive Golgi units was statistically constant in several independent experiments. Thus, it is hypothesized that the Golgi units might be functionally heterogeneous, and that those containing Frc might modify some proteins, including Notch, but not others (Yano, 2005).
To test this hypothesis, the localizations of various molecules involved in protein modification in the Golgi complex were compared with that of Frc. It was found that Sfl is also restricted to a subset of Golgi units, but that its distribution does not overlap with that of Frc. This differential localization of Sfl and Frc might thus explain the observation that frc mutant clones in wing discs do not show any defect in GAG synthesis by Sfl (Yano, 2005).
The Spi-processing enzyme Rho is also localized to a subset of Golgi units distinct from those containing Frc, in addition to its presence in other compartments. This result indicates the existence of at least two types of Golgi units, those containing Rho and those containing Frc. To determine whether these two types of Golgi units differ functionally, the glycosylation state and function of Spi was examined in frc mutants (Yano, 2005).
Given that the extent of Notch glycosylation, as detected by wheat germ agglutinin (WGA), is markedly reduced in frc mutants compared with that in the wild-type background, whether the WGA-reactive glycan of Spi is also affected by frc mutation was also examined. Myc epitope-tagged Spi was expressed in the wild type or the frcRY34 mutant. Spi-Myc was then precipitated from larval homogenates with antibodies to Myc and was examined for its glycosylation by SDS/PAGE and subsequent blot analysis with WGA. The reactivity of the Spi glycan with WGA was similar in the frc mutant and in the wild type. Whether the frcRY34 mutation affects the Spi glycan was examined by mobility shift analysis. The electrophoretic mobility of glycosylated Spi from the wild type is similar to that from the frc mutant. Deglycosylation of Spi by neuraminidase, peptide-N-glycosidase (PNGase) F, and O-glycanases also increases its mobility to the same extent in wild-type and frc mutant larvae, suggesting that the core protein is not affected by the frc mutation. Together, these results indicate that the function of Frc is not necessary for formation of the Spi glycan. It is also concluded that the function of the Rho-Spi pathway is not affected by frc mutation (Yano, 2005).
To confirm that the Golgi units containing Frc and those containing Rho are distinct, whether these Golgi units could be selectively isolated by using antibodies to Myc (for Myc-tagged Frc) or HA (for HA-tagged Rho) was examined. Because it was very difficult to collect enough of the imaginal discs, the starting material was switched to embryos, and whether Frc and Rho localize to distinct Golgi units was examined in embryos. Frc-Myc and Rho-HA were coexpressed in the embryos by the arm-Gal4 driver, and immunostaining with antibodies to Myc and to HA revealed that the Golgi units containing Frc-Myc (45.4% of total Golgi units) and those containing Rho-HA (43.0% of total Golgi units) are largely distinct: only 11.6% of total Golgi units were positive for both Frc-Myc and Rho-HA. Immunoisolation was attempted from embryonic lysates by using either antibody to Myc or HA, and how much Frc-Myc and Rho-HA were coisolated in each immunoisolate was examined. When Frc-Myc was immunoisolated with an antibody to Myc, the recovery of Frc-Myc was 5.7 times greater than that of Rho-HA. Moreover, when Rho-HA was immunoisolated with an antibody to HA, the recovery of Rho-HA was 18.3 times greater than that of Frc-Myc. The immunoblot analysis of these immunoisolates with the anti-120-kDa antibody confirmed that the Golgi units were concentrated in these immunoisolates. These results support the notion that Frc-Myc-containing fraction is distinct and could be separated from Rho-HA-containing fraction (Yano, 2005). Whether these distinct Golgi units contain different constituents was examined. Fringe (Fng) is one of the candidate molecules that may be colocalized with Frc. Therefore, expression of ectopically expressed Fng was examined in Rho- and Frc-containing immunoisolates. It was found that expression of Fng in Frc-containing immunoisolates was 26 times greater than in Rho-containing immunoisolates, supporting the idea that Fng is localized in the Frc-positive Golgi units rather than the Rho-positive Golgi units. Immunostaining analysis confirmed that FNG was colocalized mostly with Frc (88.1% of the FNG-positive Golgi units), but not with Rho (16.6% of the FNG-positive Golgi units), by immunostaining analysis (Yano, 2005).
The data suggest that different Golgi units perform different functions, a notion that is also supported by the observation that Tn antigen (O-linked N-acetylgalactosamine) was detected in only a subset of Golgi units in imaginal eye disc cells. In addition, most of these Tn antigen-positive Golgi units were found to be distributed in the basal region of the disc cells, suggesting that the differential distribution of Golgi units might contribute to the apicobasal polarity of glycan distribution (Yano, 2005).
In contrast to the larval stage, Frc is required for GAG synthesis at the early embryonic stage. To determine why the Frc requirement for GAG synthesis differs between the embryonic and larval stages, embryos were stained expressing Frc-Myc with antibodies to Sfl and to Myc. Sfl was found to be colocalized with Frc, likely explaining the importance of Frc for GAG synthesis at the embryonic stage. In addition, this embryonic requirement of Frc for GAG synthesis excludes the possibility that the selective defects in Notch and not in GAG synthesis observed in frc mutant larvae are caused by the selective Frc-dependent transport of a subset of UDP-sugars used only for glycosylation of Notch but not for GAGs synthesis (Yano, 2005).
These results provide evidence for the existence of functionally distinct Golgi units in Drosophila cells. Such functional heterogeneity of Golgi units is likely responsible for the diversity of protein glycosylation. At least two types of Golgi units containing either Frc or Sfl were shown to be present in larval disc cells. Two distinct sets of proteins, exemplified by Notch and GAG core proteins, might thus be selectively transported to Frc- or Sfl-containing Golgi units, respectively, where they undergo glycosylation by different sets of molecules (Yano, 2005).
The variety of Golgi units might be established by separate transport of secretory proteins and glycosylenzymes from the endoplasmic reticulum (ER) to the distinct Golgi units. In yeast, glycosylphosphatidylinositol (GPI)-anchored proteins exit the ER in vesicles distinct from those containing other secretory protein. Given that the GAG core protein Dally in Drosophila is anchored to the membrane by GPI, it is possible that Dally and Notch are loaded into distinct vesicles as they exit the ER (Yano, 2005).
Combinations of glycosylenzymes and transporters, such as Sfl and Frc, contained in Golgi units of Drosophila differ not only between embryos and larval disc cells but also among cell types. For example, it was found that Frc is localized to all Golgi units in salivary gland cells at the larval stage. It has also been shown that all of the Golgi complexes dispersed in oocytes may have the ability to process the Gurken precursor protein, which is usually cleaved in a subset of the Golgi complexes residing in the dorso-anterior region. The Golgi units may thus be altered in a manner dependent on development, cell type, and signaling processes (Yano, 2005).
The functional diversity of Golgi units also might contribute to the polarized distribution of glycans along the apicobasal axis of cells. It was found that Tn antigen is synthesized in the basal Golgi units of larval disc cells. Furthermore, certain types of glycans are distributed along the apicobasal axis of pupal ommatidia. These glycans might thus be synthesized differentially in the Golgi units that are asymmetrically distributed along the apicobasal axis and then be secreted at either the apical or basal cell surface (Yano, 2005).
Whereas Golgi units are dispersed throughout Drosophila cells, the Golgi complex in mammalian cells is thought to be a single entity that is located in the pericentriolar region through its association with the microtubule-organizing center in interphase and which is fragmented at the onset of mitosis. The Golgi fragments apparent in mammalian cells during mitosis are highly similar to the Golgi units of Drosophila cells in both electron and confocal microscopic images. The mammalian Golgi complex during interphase may therefore be comprised of functionally distinct units that are associated with the microtubule-organizing center and connected with each other (Yano, 2005).
During the cell cycle, the Golgi, like other organelles, has to be duplicated in mass and number to ensure its correct segregation between the two daughter cells. It remains unclear, however, when and how this occurs. This study shows that in Drosophila S2 cells, the Golgi likely duplicates in mass to form a paired structure during G1/S phase and remains so until G2 when the two stacks separate, ready for entry into mitosis. Pairing requires an intact actin cytoskeleton which in turn depends on Abi/Scar but not WASP. This actin-dependent pairing is not limited to flies but also occurs in mammalian cells. It is further shown that preventing the Golgi stack separation at G2 blocks entry into mitosis, suggesting that this paired organization is part of the mitotic checkpoint, similar to what has been proposed in mammalian cells (Kondylis, 2007).
During the cell cycle, the Golgi, like other organelles, has to duplicate in mass and/or number to ensure its correct segregation between the two daughter cells. It remains unclear, however, when and how this occurs.
The process of Golgi duplication and inheritance in mammalian cells is still debated, as different modes of Golgi biogenesis have been proposed. One reason why this issue is not yet settled could be due to the elaborate organization of the Golgi stacks, which are interconnected to form a single-copy organelle capping the nucleus, thus impeding clear visualization of organelle duplication and segregation. Therefore, this study has exploited the relatively small number of Golgi stacks in Drosophila tissue-cultured S2 cells to revisit this issue (Kondylis, 2007).
In S2 cells, the Golgi stacks are found in close proximity to transitional endoplasmic reticulum (tER) sites, forming tER-Golgi units (Kondylis, 2003; Herpers, 2004). Their number nearly doubles at G2 phase. In an effort to identify factors mediating this process, focus was placed on cytoskeletal elements that have been involved in the organization of the mammalian Golgi apparatus. Microtubules are involved in mammalian Golgi ribbon maintenance, as their depolymerization leads to its reorganization into individual Golgi stacks in close proximity to ER exit sites (Kondylis, 2007 and references therein).
F-actin has also been implicated in the organization of the mammalian Golgi apparatus; its depolymerization leads to a compact appearance of this organelle without disruption of cisternal stacking. A key regulator of actin polymerization is the Arp2/3 complex. Its F-actin nucleation activity is triggered both by Wiskott-Aldrich syndrome protein (WASP) and WASP family verprolin-homologous (WAVE/Scar) proteins, which are in turn regulated by Rho small GTPases. WASP exists in an autoinhibited state that is released by the cooperative action of Cdc42, PI(4,5)P, and other SH3-containing proteins. In contrast, WAVE/Scar proteins, together with Sra-1, Kette (Nap1), Abi, and HSPC300, form a stable complex, which is itself regulated by Rac (Kondylis, 2007 and references therein).
Rho GTPases have recently been implicated in maintaining Golgi architecture. Cdc42 has been localized on the Golgi membrane and shown to recruit the Arp2/3 complex around this organelle via ARHGAP10. Furthermore, suppression of the brain-specific Rho-binding protein Citron-N in neurons was shown to lead to fragmentation of the Golgi apparatus, and Rho1 was proposed to exert its local effect on F-actin by regulating ROCK and profilin activity (Kondylis, 2007 and references therin).
This study shows that drug-induced F-actin depolymerization in S2 cells leads to doubling of the number of tER-Golgi units independent of anterograde transport. Using live cell imaging, electron microscopy, and three-dimensional (3D) electron tomography, this study shows that each Golgi is organized as a pair of stacks held together by an actin-based mechanism, both in Drosophila and in human cells. In S2 cells, this is mediated by Abi and Scar, suggesting a novel role for the Rac signaling cascade in Golgi architecture. Last, it was shown that the Golgi stacks undergo separation at G2 through modulation of Abi and Scar, and that blocking this separation prevents cells from entering mitosis, supporting the existence of a G2/M checkpoint related to Golgi structural organization (Kondylis, 2007).
The two Golgi stacks could be physically linked without displaying membrane continuity or being interconnected, for instance through intercisternal tubular connections, either permanent or transient. Tubules connecting cisternae of adjacent stacks are involved in the formation of the Golgi ribbon in mammalian cells and, recently, GM130 and GRASP65 have been proposed to be required for their integrity. However, the putative tubules connecting the two stacks in the pair would have different molecular requirements, at least in Drosophila, since depletion of dGM130 or dGRASP does not lead to their separation (Kondylis, 2003; Kondylis, 2005; Kondylis, 2007 and references therein).
F-actin could provide a physical link holding the paired Golgi stacks together, or it could help in the formation/maintenance of intercisternal tubules. In addition, short actin filaments have been proposed to link spectrin mosaics leading to the formation of a skeleton that surrounds the Golgi complex. One of its functions could be to hold the two Golgi stacks close enough to allow the formation and fusion of the tubules. It could also surround the tubules themselves, thus providing membrane stability. The localization of Abi and Scar at the periphery of the tER-Golgi units and between the two stacks in a pair is consistent with both proposed functions. These tomography studies so far have not revealed clear membrane continuities between Golgi cisternae, though examples have been found of a tubular network which is shared by the paired stacks (Kondylis, 2007).
tER sites behave similarly to the Golgi, as they also separate at G2 and upon F-actin depolymerization. Because little is known about the mechanism regulating the biogenesis of tER sites, it is difficult to envisage how the two parts could be held together. The spectrin-actin mesh described above could be common to Golgi and tER sites, and Golgi and tER site scission could be achieved in a synchronized fashion. Alternatively, either of these organelles could split first and lead to the scission of the other, perhaps by providing positional information. Recently, the centrosome component centrin 2 that is also localized to tER sites in Trypanosoma has been shown to give such positioning information. A more in-depth study combining immunogold labeling and 3D tomography would be required to elucidate such fine details of tER-Golgi structural organization (Kondylis, 2007).
Drosophila Rho1 is unlikely to have a role in holding the two Golgi stacks together. The overexpression of the Rho1 constitutively inactive mutant or treatment of S2 cells with ROCK or myosin light chain inhibitors (Y27632 and blebbistatin) did not affect the Golgi number. Cdc42 is also unlikely to participate as the depletion of its downstream effector WASP did not lead to Golgi separation, although the overexpression of the Cdc42T17N dominant negative did. However, this effect could be due to nonspecific sequestration of the guanine nucleotide exchange factor involved in maintaining the paired Golgi stacks and may be shared with other small GTPases (Kondylis, 2007).
Interestingly, the results are consistent with a role for Rac GTPases in Drosophila Golgi architecture. Expression of the constitutively inactive form of Rac1 led to a near-doubling in the Golgi number, and depletion of Scar/WAVE or Abi, which are regulated by Rac GTPases, led to a similar phenotype. The identical results obtained in Scar and Abi RNAi suggest that this well-established Scar/WAVE pentameric complex is involved in holding the paired Golgi stacks together by promoting F-actin polymerization. These data indicate that the Rac signaling pathway is involved. However, the Scar/Abi complex has recently been shown to also stimulate Arp2/3 and F-actin polymerization independently of Rac. This would need to be investigated further (Kondylis, 2007).
This study shows that the separation of the paired Golgi stacks occurs at G2, prior to mitosis. A similar phenomenon has already been reported during cell division in Toxoplasma gondii, where a single Golgi stack grows as a duplicated organelle and is separated as the cell divides. However, the mechanism underlying this separation is not known (Kondylis, 2007).
The Golgi doubling in number at G2 phase resembles many aspects of this observed upon F-actin depolymerization. In both cases, a similar increase in Golgi number and decrease in their size are observed. Furthermore, this study has shown that it is the modulation of the F-actin cytoskeleton and the activity of Abi/Scar at G2 that lead to Golgi stack separation. (1) It was found that both Scar and Abi localized to the Golgi, strongly arguing for having a role in actin remodeling around this organelle. (2) The Golgi stacks in G2 cells remain insensitive to F-actin depolymerization. (3) Cells depleted of Abi and Scar that exhibit separated Golgi stacks do not split them further at G2. (4) The overexpression of Abi prevents Golgi separation at G2. This strongly suggests that the F-actin/Abi/Scar-mediated link of the two stacks has been severed in a G2-specific manner, perhaps by kinases such as Polo (Kondylis, 2007).
Because tER sites and the Golgi apparatus ultimately disperse later in mitosis, both in mammalian and Drosophila S2 cells, the Golgi stack separation prior to dispersion might be part of the proposed Golgi G2/M checkpoint. Indeed, reagents that interfere with the GRASP65/55 phosphorylation by Polo and ERK/MEK, respectively, arrest or delay the cell cycle at the G2/M transition. This study shows that blocking Golgi separation at G2 by overexpressing Abi also prevents S2 cells from entering mitosis. This strengthens the relationship between Golgi organization and mitotic entry, although it cannot formally be excluded that the mitotic block observed is partly due to additional effects of Abi overexpression, for instance at the plasma membrane (Kondylis, 2007).
It is proposed that at G2, the paired stacks are separated along with the adjacent tER sites. As the cell enters mitosis, the Golgi membrane and the tER sites disperse, and are segregated into the two daughter cells, where the tER-Golgi units are rebuilt. The Golgi could be rebuilt as a very small paired stack in close association with Scar, Abi, and F-actin, or as a single stack that will duplicate by a mechanism that still needs to be unraveled. Since G1 cells are all sensitive to F-actin depolymerization, this suggests that the formation of the paired Golgi stack starts just after the exit from mitosis and persists until S phase, when the Golgi seems to grow significantly. A more detailed understanding will come from EM study of S and G2 cells (Kondylis, 2007).
One of the remaining questions regards the impact of the Abi/Scar role on Golgi organization during development. Using Scar/WAVE, Abi, Kette, and Sra-1 mutants, as well as transgenic flies carrying inducible RNAi constructs, it will be possible to assess whether any of the observed phenotypes (defects in oogenesis, cell and organ morphology, neuroanatomical malformations, and failure in cell migration) is in part due to defects in Golgi organization (Kondylis, 2007).
The development of air-filled respiratory organs is crucial for survival at birth. A combination of live imaging and genetic analysis was used to dissect respiratory organ maturation in the embryonic Drosophila trachea. Tracheal tube maturation was found to entail three precise epithelial transitions. Initially, a secretion burst deposits proteins into the lumen. Solid luminal material is then rapidly cleared from the tubes, and shortly thereafter liquid is removed. To elucidate the cellular mechanisms behind these transitions, gas-filling-deficient mutants were identified showing narrow or protein-clogged tubes. These mutations either disrupt endoplasmatic reticulum-to-Golgi vesicle transport or endocytosis. First, Sar1 was shown to be required for protein secretion, luminal matrix assembly, and diametric tube expansion. sar1 encodes a small GTPase that regulates COPII vesicle budding from the endoplasmic reticulum (ER) to the Golgi apparatus. Subsequently, a sharp pulse of Rab5-dependent endocytic activity rapidly internalizes and clears luminal contents. The coordination of luminal matrix secretion and endocytosis may be a general mechanism in tubular organ morphogenesis and maturation (Tsarouhas, 2007).
Branched tubular organs are essential for oxygen and nutrient transport. Such organs include the blood circulatory system, the lung and kidney in mammals, and the tracheal respiratory system in insects. The optimal flow of transported fluids depends on the uniform length and diameter of the constituting tubes in the network. Alterations in the distinct tube shapes and sizes cause pronounced defects in animal physiology and lead to serious pathological conditions. For example, tube overgrowth and cyst formation in the collecting duct are intimately linked to the pathology of Autosomal Dominant Polycystic Kidney Disease. Conversely, stenoses, the abnormal narrowing of blood vessels and other tubular organs, are associated with ischemias and organ obstructions (Tsarouhas, 2007 and references therein).
While the early steps of differentiation, lumen formation, and branch patterning begin to be elucidated in several tubular organs, only scarce glimpses into the cellular events of lumen expansion and tubular organ maturation are available. De novo lumen formation can be induced in three-dimensional cultures of MDCK cells. Recent studies in this system revealed that PTEN activation, apical cell membrane polarization, and Cdc42 activation are key events in lumen formation in vitro. In zebrafish embryos and cultured human endothelial cells, capillary vessels form through the coalescence and growth of intracellular pinocytic vesicles. These tubular vacuoles then fuse with the plasma membranes to form a continuous extracellular lumen. Salivary gland extension in Drosophila requires the transcriptional upregulation of the apical membrane determinant Crumbs (Crb), but the cellular mechanism leading to gland expansion remains unclear (Tsarouhas, 2007 and references therein).
The epithelial cells of the Drosophila tracheal network form tubes of different sizes and cellular architecture, and they provide a genetically amenable system for the investigation of branched tubular organ morphogenesis. Tracheal development begins during the second half of embryogenesis when 20 metameric placodes invaginate from the epidermis. Through a series of stereotypic branching and fusion events, the tracheal epithelial cells generate a tubular network extending branches to all embryonic tissues. In contrast to the wealth of knowledge about tracheal patterning and branching, the later events of morphogenesis and tube maturation into functional airways have yet to be elucidated. As the nascent, liquid-filled tracheal network develops, the epithelial cells deposit an apical chitinous matrix into the lumen. The assembly of this intraluminal polysaccharide cable coordinates uniform tube growth. Two luminal, putative chitin deacetylases, Vermiform (Verm) and Serpentine (Serp), are selectively required for termination of branch elongation. The analysis of verm and serp mutants indicates that modifications in the rigidity of the matrix are sensed by the surrounding epithelium to restrict tube length. What drives the diametric expansion of the emerging narrow branches to their final size? How are the matrix- and liquid-filled tracheal tubes cleared at the end of embryogenesis (Tsarouhas, 2007)?
This study used live imaging of secreted GFP-tagged proteins to identify the cellular mechanisms transforming the tracheal tubes into a functional respiratory organ. The precise sequence and cellular dynamics were characterized of a secretory and an endocytic pulse that precede the rapid liquid clearance and gas filling of the network. Analysis of mutants with defects in gas filling reveals three distinct but functionally connected steps of airway maturation. Sar1-mediated luminal deposition of secreted proteins is tightly coupled with the expansion of the intraluminal matrix and tube diameter. Subsequently, a Rab5-dependent endocytotic wave frees the lumen of solid material within 30 min. The precise coordination of secretory and endocytotic activities first direct tube diameter growth and then ensure lumen clearance to generate functional airways (Tsarouhas, 2007).
Two strong, hypomorphic sar1 alleles were identified in screens for mutants with tracheal tube defects. In wild-type embryos, the bulk of luminal markers 2A12, Verm, and Gasp has been deposited inside the DT lumen by stage 15. However, in zygotic sar1P1 mutants (hereafter referred to as sar1), luminal secretion of 2A12, Verm, and Gasp was incomplete. The tracheal cells outlined by GFP-CAAX partially retained those markers in the cytoplasm. sar1 zygotic mutant embryos normally deposited the early luminal marker Piopio by stage 13. Luminal chitin was also detected in sar1 mutants at stage 15. However, the luminal cable was narrow, more dense, and distorted compared to wild-type. To test if the sar1 secretory phenotype in the trachea is cell autonomous, Sar1 was reexpressed specifically in the trachea of sar1 mutants by using btl-GAL4. Such embryos showed largely restored secretion of 2A12, Verm, and Gasp. Thus, it is concluded that tracheal sar1 is required for the efficient secretion of luminal markers, which are predicted to associate with the growing intraluminal chitin matrix (Tsarouhas, 2007).
sar1 mRNA has been reported to be abundantly maternally contributed. At later stages, zygotic expression of sar1 mRNA is initiated in multiple epithelial tissues. To monitor Sar1 zygotic expression in the trachea, a Sar1-GFP protein trap line was used. Embryos carrying only paternally derived Sar1-GFP show a strong zygotic expression of GFP in the trachea. An anti-Sar1 antibody was used to analyze Sar1 expression in the trachea of wild-type, zygotic sar1P1, and sar1EP3575Δ28 null mutant embryos were generated. Both zygotic mutants showed a clear reduction, but not complete elimination, of Sar1 expression in the trachea. To test the effects of a more complete inactivation of Sar1, transgenic flies were generated expressing a dominant-negative sar1T38N form in the trachea. In btl > sar1T38N-expressing embryos, early defects were observed in tracheal branching and epithelial integrity as well as a complete block in Verm secretion. In contrast to btl > sar1T38N-expressing embryos, zygotic sar1P1 mutant embryos show normal early tracheogenesis with no defects in branching morphogenesis and epithelial integrity (Tsarouhas, 2007).
In summary, tracheal expression of Sar1 is markedly reduced in zygotic sar1 mutant embryos. While maternally supplied Sar1 is sufficient to support early tracheal development, zygotic Sar1 is required for efficient luminal secretion (Tsarouhas, 2007).
Given the conserved role of Sar1 in vesicle budding from the ER, its subcellular localization in the trachea was determined by using anti-Sar1 antibodies. Sar1 localizes predominantly to the ER (marked by the PDI-GFP trap). Continuous COPII-mediated transport from the ER is required to maintain the Golgi apparatus and ER structure. To test if zygotic loss of Sar1 compromises the integrity of the ER and Golgi in tracheal cells, sar1 mutant embryos were stained with antibodies against KDEL (marking the ER lumen) and gp120 (highlighting Golgi structures). In sar1 mutant embryos, a strongly disrupted ER structure and loss of Golgi staining was observed in dorsal trunk (DT) cells at stage 14. Additionally, TEM of stage-16 wild-type and sar1 mutant embryos showed a grossly bloated rough ER structure in DT tracheal cells. Consistent with its functions in yeast and vertebrates, Drosophila Sar1 localizes to the ER and is not only required for efficient luminal protein secretion, but also for the integrity of the early secretory apparatus (Tsarouhas, 2007).
To analyze tracheal maturation defects in sar1 mutant embryos, sar1 strains were generated and imaged that carry either btl > ANF-GFP btl-mRFP-moe or btl > Gasp-GFP. In sar1 mutants, luminal deposition of both ANF-GFP and Gasp-GFP is reduced. Like endogenous Gasp in the mutants, Gasp-GFP was clearly retained in the cytoplasm of sar1 embryos. ANF-GFP was also retained in the tracheal cells of sar1 mutants, but to a lesser extent. Strikingly, sar1 mutants failed to fully expand the luminal diameter of the DT outlined by the apical RFP-moe localization. This defect was quantified by measuring diametric growth rates in metamere 6 for wild-type and sar1 mutant embryos. While early lumen expansion commences in parallel in both genotypes, the later diametric growth of sar1 mutants falls significantly behind compared to wild-type embryos. The DT lumen in sar1 mutants reaches only an average of 70% of the wild-type diameter at early stage 16. Identical diametric growth defects were detected in fixed sar1 mutant embryos expressing btl > GFP-CAAX by analysis of confocal yz sections or TEM. Reexpression of sar1 in the trachea of sar1 mutant embryos not only rescued secretion, but also the lumen diameter phenotype at stage 16. In contrast to the diametric growth defects, DT tube elongation in sar1 embryos was indistinguishable from that in wild-type. This demonstrates distinct genetic requirements for tube diameter and length growth. It also reveals that the sar1 DT luminal volume reaches less than half of the wild-type volume. Prolonged live imaging showed that sar1 mutants are also completely deficient in luminal protein and liquid clearance. Up to 80% of the rescued embryos also completed luminal liquid clearance, suggesting that efficient tracheal secretion and the integrity of the secretory apparatus are prerequisites for later tube maturation steps. Taken together, the above-described results show that tracheal Sar1 is selectively required for tube diameter expansion. Additionally, subsequent luminal protein and liquid clearance fail to occur in sar1 mutants (Tsarouhas, 2007).
Do the tracheal defects of sar1 reflect a general requirement for the COPII complex in luminal secretion and diameter expansion? To test this, lethal P element insertion alleles were examined disrupting two additional COPII coat subunits, sec13 and sec23. Mutant sec13 and sec23 embryos were stained for luminal Gasp and for Crb and α-Spectrin to highlight tracheal cells. At stage 15, embryos of both mutants show a clear cellular retention of Gasp. Furthermore, stage-16 sec13 and sec23 embryos show significantly narrower DT tubes when compared to wild-type. The average diameter of the DT branches in metamere 6 was 4.8 μm and 4.4 μm in fixed sec13 and sec23 embryos, respectively, compared to 6.3 μm in wild-type. Therefore, sec13 and sec23 mutants phenocopy sar1. The phenotypic analysis of three independent mutations disrupting ER-to-Golgi transport thus provides a strong correlation between deficits in luminal protein secretion and tube diameter expansion (Tsarouhas, 2007).
The live-imaging approach defines the developmental dynamics of functional tracheal maturation. At the organ level, three sequential and rapid developmental transitions were identified: (1) the secretion burst, followed by massive luminal protein deposition and tube diameter expansion, (2) the clearance of solid luminal material, and (3) the replacement of luminal liquid by gas. Live imaging of each event additionally revealed insights into the startlingly dynamic activities of the tracheal cells. ANF-GFP-containing structures and apical GFP-FYVE-positive endosomes rapidly traffic in tracheal cells during the secretion burst and protein clearance. The direct live comparison between wild-type and mutant embryos further highlights the dynamic nature of epithelial activity during each pulse (Tsarouhas, 2007).
This study identified several mutations that selectively disrupt distinct cellular functions and concurrently interrupt the maturation process at specific steps. This clearly demonstrates the significance of phenotypic transitions in epithelial organ maturation and establishes that secretion is required for luminal diameter expansion and endocytosis for solid luminal material clearance (Tsarouhas, 2007).
The sudden initiation of an apical secretory burst tightly precedes diametric tube expansion. The completion of both events depends on components of the COPII complex, further suggesting that the massive luminal secretion is functionally linked to diametric growth. How does apical secretion provide a driving force in tube diameter expansion? In mammalian lung development, the distending internal pressure of the luminal liquid on the epithelium expands the lung volume and stimulates growth. Cl− channels in the epithelium actively transport Cl− ions into luminal liquid. The resulting osmotic differential then forces water to enter the lung lumen, driving its expansion. By analogy, the tracheal apical exocytic burst may insert protein regulators such as ion channels into the apical cell membrane or add additional membrane to the growing luminal surface. Since the ER is a crucial cellular compartment for intracellular traffic and lipogenesis, its disruption in sar1 mutants may disrupt the efficient transport of so far unknown specific regulators or essential apical membrane addition required for diametric expansion. Alternatively, secreted chitin-binding proteins (ChB) may direct an increase of intraluminal pressure and tube dilation. Overexpression of the chitin-binding proteins Serp-GFP or Gasp-GFP was insufficient to alter the diametric growth rate of the tubes, suggesting that lumen diameter expansion is insensitive to increased amounts of any of the known luminal proteins. In sar1 mutants, the secretion of at least two chitin-binding proteins, Gasp and Verm, is reduced. Chitin, however, is deposited in seemingly normal quantities, but assembles into an aberrantly narrow and dense chitinous cable. This phenotype suggests that the correct ratio between chitin and multiple interacting proteins may be required for the correct assembly of the luminal cable. Interestingly, sar1, sec13, and sec23 mutant embryos form a severely defective and weak epidermal cuticle. The luminal deposition of ChB proteins during the tracheal secretory burst may orchestrate the construction and swelling of a functional matrix, which, in turn, induces lumen diameter dilation. While this later hypothesis is favored, it cannot be excluded that other mechanisms, either separately or in combination with the dilating luminal cable, drive luminal expansion (Tsarouhas, 2007).
During tube expansion, massive amounts of luminal material, including the chitinous cable, fill the tracheal tubes. This study found that Dynamin, Clathrin, and the tracheal function of Rab5 are required to rapidly remove luminal contents, indicating that endocytosis is required for this process. Several lines of evidence argue that the tracheal epithelium activates Rab5-dependent endocytosis to directly internalize luminal material. First, the tracheal cells of rab5 mutants show defects in multiple endocytic compartments. These phenotypes of rab5 mutants become apparent during the developmental period matching the interval of luminal material clearance in wild-type embryos. Second, tracheal cells internalize two luminal markers, the endogenously encoded Gasp and the dextran reporter, exactly prior and during luminal protein clearance. The number of intracellular dextran puncta reaches its peak during the clearance process and ceases shortly thereafter. Lastly, intracellular puncta of both Gasp and dextran colocalize with defined endocytic markers inside tracheal cells. The colocalization of Gasp and dextran with GFP-Rab7 and of Gasp with GFP-LAMP1 suggests that the luminal material may be degraded inside tracheal cells. Taken together, these data show that the tracheal epithelium activates a massive wave of endocytosis to clear the tubes (Tsarouhas, 2007).
Endocytic routes are defined by the nature of the internalized cargoes and the engaged endocytic compartments. What may be the features of the endocytic mechanisms mediating the clearance of luminal material? The phenotype of chc mutants and the presence of intracellular Gasp in CCVs indicate that luminal clearance at least partly relies on Clathrin-mediated endocytosis (CME). In addition to CME, Dynamin and Rab5 have also been implicated in other routes of endocytosis, suggesting that multiple endocytic mechanisms may be operational in tracheal maturation. The nature of the endocytosed luminal material provides an additional perspective. While cognate uptake receptors may exist for specific cargos such as Gasp, Verm, and Serp, the heterologous ANF-GFP, degraded chitin, and the fluid-phase marker dextran may be cleared by either fluid-phase internalization or multifunctional scavenger receptors. Interestingly, Rab5 can regulate fluid-phase internalization in cultured cells by stimulating macro-pinocytosis and the activation of Rabankyrin-5. The defective tracheal internalization of dextran in rab5 mutants provides further loss-of-function evidence for Rab5 function in fluid-phase endocytosis in vivo. The above-described arguments lead to the speculation that additional Rab5-regulated endocytic mechanisms most likely cooperate with CME in the clearance of solid luminal material (Tsarouhas, 2007).
How is liquid cleared from the lumen? While very little is known about this fascinating process, some developmental and mechanistic arguments suggest that this last maturation step is mechanistically distinct. First, the interval of luminal liquid clearance is clearly distinct from the period of endocytic clearance of solids. Second, the dynamic internalization of dextran and the abundance of GFP-marked endocytic structures decline before liquid clearance. Finally, assessment of liquid clearance further suggests that it requires a distinct cellular mechanism (Tsarouhas, 2007).
Viewing the entire process of airway maturation in conjunction, some general conclusions may be drawn. First, the three epithelial pulses are highly defined by their sequence and exact timing, suggesting that they may be triggered by intrinsic or external cues. Second, the analysis of mutants that selectively reduce the amplitude of the secretory or endocyic pulses demonstrates the requirement for each epithelial transition in the completion of the entire maturation process. These pulses are induced in the background of basal secretory and endocytic activities that operate throughout development. Third, specific cellular activities exactly precede each morphological transition. Finally, the separate transitions are interdependent in a sequential manner. Efficient secretion is a prerequisite for the endocytic wave. Similarly, protein endocytosis is a condition for luminal liquid clearance. This suggests a hierarchical coupling of the initiation of each pulse to completion of the previous one in a strict developmental sequence (Tsarouhas, 2007).
This study provides a striking example of how pulses of epithelial activity drive distinct developmental events and mold the nascent tracheal lumen into an air delivery tube. These findings are likely to be relevant beyond the scope of tracheal development. The uniform growth of salivary gland tubes in flies and the excretory canal and amphid channel lumen in worms also require the assembly of a luminal matrix for uniform tube growth. Luminal material is also transiently present during early developmental stages in the distal nephric ducts of lamprey. Thus, the coordinated, timely deposition and removal of transient luminal matrices may represent a general mechanism in tubulogenesis (Tsarouhas, 2007).
Regulated secretion of hormones, digestive enzymes, and other biologically active molecules requires the formation of secretory granules. Clathrin and the clathrin adaptor protein complex 1 (AP-1) are necessary for maturation of exocrine, endocrine, and neuroendocrine secretory granules. However, the initial steps of secretory granule biogenesis are only minimally understood. Powerful genetic approaches available in Drosophila were used to investigate the molecular pathway for biogenesis of the mucin-containing 'glue granules' that form within epithelial cells of the third-instar larval salivary gland. Clathrin and AP-1 colocalize at the trans-Golgi network (TGN) and clathrin recruitment requires AP-1. Furthermore, clathrin and AP-1 colocalize with secretory cargo at the TGN and on immature granules. Finally, loss of clathrin or AP-1 leads to a profound block in secretory granule formation. These findings establish a novel role for AP-1- and clathrin-dependent trafficking in the biogenesis of mucin-containing secretory granules (Burgess, 2011).
Constitutive secretion of proteins and lipids from the trans-Golgi network (TGN) toward the cell surface is believed to operate in all cells. Constitutive secretion is characterized by the rapid deployment of newly synthesized cargo toward its final cellular destination. Specialized secretory cells such as endocrine, neuroendocrine, and exocrine cells contain an additional pathway termed the regulated secretory pathway. One hallmark of this pathway is the storage of regulated secretory proteins at high concentration in dense-core secretory granules that can be released in response to an external signal. How secreted proteins enter the regulated secretory pathway is a source of debate and may prove to be cargo and cell-type specific. In the case of endocrine and neuroendocrine cells, sorting of secreted cargo is believed to be content driven, with selective aggregation of regulated secretory proteins at the TGN playing a major role in secretory granule biogenesis (Burgess, 2011).
Little is known about the coat proteins that might be required on the cytoplasmic face to promote budding of lumenal regulated secretory cargo from the TGN. Initial studies in AtT20 pituitary cells noted that condensing secretory products accumulate in dilated regions of the TGN that are coated with clathrin. Similarly, in β-cells treated with monensin to perturb intracellular trafficking, proinsulin accumulates in a clathrin-coated compartment related to the TGN. These observations raise the possibility that the formation of regulated secretory granules might require clathrin at the TGN (Burgess, 2011).
Coat proteins selectively incorporate cargo into vesicles and provide a scaffold for vesicle formation. Clathrin and its associated heterotetrameric adaptor proteins (APs) make up a major class of vesicular coats. APs bind to sorting motifs found in the cytoplasmic tails of membrane cargo and function as links between vesicular cargo and the clathrin lattice, although some AP-3 and AP-4 coats lack clathrin. The four different AP complexes (AP-1-4) have distinct sites of action in the cell. Of these, the AP-1 complex has perhaps the most diverse roles, acting at the TGN to promote constitutive secretion (Chi, 2008), at the TGN and endosomes to sort mannose 6-phosphate receptors, and at immature secretory granules of specialized secretory cells to retrieve missorted proteins. Indeed, a coat composed of clathrin and AP-1 is required for maturation and condensation of regulated secretory granules. In contrast to granule maturation, the roles of AP-1 and clathrin in initial stages of secretory granule formation are less well established. AP-1 and clathrin were shown to be required for formation of Weibel-Palade bodies (Lui-Roberts, 2005), secretory organelles that store the hemostatic protein von Willebrand factor. However, a dominant-negative clathrin construct did not interfere with insulin granule production in neuroendocrine cells, suggesting these granules form through a clathrin-independent mechanism. Thus it is not clear how general a role AP-1 and clathrin play in granule biogenesis (Burgess, 2011).
The larval salivary gland in Drosophila provides an excellent system for molecular genetic analysis of factors required for formation of regulated secretory granules. During the last half of third-instar larval development, prior to pupariation, salivary gland cells initiate production of mucin-type secretory granules termed 'glue' granules. These granules contain highly glycosylated mucin-type glue proteins that are required to adhere the pupal case to a solid substrate during metamorphosis. Of the six known glue proteins (also called salivary gland secretion or Sgs proteins), Sgs1, Sgs3, and Sgs4 contain extended amino acid repeats that are likely sites of oligosaccharide linkage. These proteins, which are synthesized in response to a low-titer pulse of the steroid hormone ecdysone at the mid-third-instar larval stage, are stored until an additional high-titer pulse of ecdysone promotes their release at the onset of pupariation (Burgess, 2011).
Secreted mucin-type glycoproteins are ubiquitous in metazoans and serve important roles in animal physiology. This study analyzed the mechanism of mucin-type glue granule biogenesis in third-instar larval salivary gland cells. It was shown that AP-1 and clathrin localize to the TGN prior to glue production, colocalize with newly synthesized glue proteins during early stages of granule formation, and are found at later stages on maturing glue granules. Genetic disruption or knockdown of AP-1 subunits strongly reduces clathrin localization to the TGN. Moreover, AP-1 and clathrin are required for glue granule formation; loss of AP-1 causes glue cargo to accumulate at the TGN and in small, highly aberrant granules. These results reveal a requirement for AP-1 and clathrin in the formation of mucin-type secretory granules (Burgess, 2011).
To identify coats that might function in granule biogenesis, the subcellular distribution of clathrin heavy chain was examined, as well as subunits of the clathrin adaptor protein complexes AP-1 and AP-3, which reside on intracellular organelles (note that Drosophila lacks AP-4). First clathrin, AP-1, and AP-3 were examined in salivary gland cells at stage 0, just prior to glue production. At this stage, Golgi bodies are easily visualized using antibodies directed against the golgin Lava lamp (Lva), which localizes to the cis-Golgi. Note that the cis-Golgi has a cup-shaped appearance. A monomeric red fluorescent protein fusion to clathrin heavy chain (RFP-Chc) predominantly localized to large puncta adjacent to the concave face of the cis-Golgi, consistent with a previous report showing localization of endogenous Chc to intracellular puncta in these cells (Wingen, 2009). Endogenous AP-1γ showed a similar distribution. A projection constructed from serial confocal sections revealed numerous Golgi units scattered throughout the cytoplasm. There was a one-to-one correspondence between AP-1γ- and Lva-positive structures, with the cis-Golgi cups surrounding AP-1γ in a manner consistent with AP-1 localizing to the TGN. Indeed AP-1γ and RFP-Chc colocalized with the trans-Golgi protein EpsinR (also called Liquid facets-Related or LqfR). In contrast, AP-1 showed only minimal overlap with the recycling endosome regulator Rab11. AP-1γ and RFP-Chc colocalized at the TGN, although AP-1γ distribution appeared slightly more diffuse in salivary gland cells expressing RFP-Chc than in nonexpressing cells. Localization of AP-1 to the TGN is adaptor-protein specific, because a functional monomeric cherry fluorescent protein (mCherry) fusion to AP-3δ (called Garnet in Drosophila) showed no overlap with a Venus fluorescent protein (VFP) fusion to AP-1μ (called AP-47 in Drosophila), but rather colocalized with the late endosome marker Rab7. Given the high degree of colocalization of clathrin and AP-1, it was asked whether AP-1 might be required to recruit clathrin to the TGN (Burgess, 2011).
To test whether AP-1 recruits clathrin to the TGN, use was made of a μ1-adaptin null allele, AP-47SHE-11. To bypass late embryonic lethality caused by this allele, mosaic clones were generated in the salivary gland using FLP-FRT-based recombination. Briefly, the wild-type chromosome carries a copy of green fluorescent protein (GFP) such that homozygous mutant cells are marked by the absence of GFP expression and heterozygous and wild-type cells are marked by one or two copies of GFP, respectively. AP-47SHE-11 clones were generated during embryogenesis and analyzed in third-instar larval salivary glands at stage 0, just prior to glue production. To determine whether other AP-1 subunits can localize to the TGN in the absence of AP-47, the distribution of AP-1γ was examined, and its punctate localization was found to be entirely lost in AP-47SHE-11 mutant cells. Hence AP-47 is required for efficient recruitment or stability of AP-1γ, similar to what was previously observed in μ1-adaptin-deficient mouse embryonic fibroblasts. Not all trafficking markers were affected by the loss of AP-47, as the early endosome marker Rab5 was unperturbed (Burgess, 2011).
Strikingly, in AP-47SHE-11 mutant cells, RFP-Chc localization to the Golgi was dramatically reduced. The effect on RFP-Chc distribution was also observed in salivary gland cells in which expression of a double-stranded RNA was used to knock down expression of AP-1γ by RNA interference (RNAi). Most cells depleted of AP-1γ exhibited strong delocalization of RFP-Chc, with only a few cells retaining weak RFP-Chc localization at the TGN. Hence the TGN is the major site of clathrin localization in these cells, and AP-1 plays a pivotal role in clathrin recruitment. Importantly, Golgi integrity per se (as assessed by distribution of Lva) was not affected by disruption of AP-1 (Burgess, 2011).
This study has provided compelling evidence of a previously unknown function for clathrin and AP-1 in the formation of mucin-type secretory granules. Clathrin and AP-1 were shown to localize to the TGN prior to synthesis of secretory cargo, colocalize with newly synthesized secretory cargo, and are required for secretory granule formation. Hence AP-1 and clathrin play a crucial role in early stages of secretory granule formation in salivary gland cells. Consistent with this idea, clathrin becomes delocalized upon AP-1 depletion, indicating that other adaptors cannot recruit clathrin in the absence of AP-1 at this stage of salivary gland development (Burgess, 2011).
The results suggest that formation of mucin-containing glue granules and Weibel-Palade bodies might be similar. Weibel-Palade bodies have an unusual cigar-shaped appearance and it was proposed that AP-1 and clathrin might participate in their formation at the TGN by allowing lumenal cargo to properly fold and aggregate or by preventing premature scission. Indeed, depletion of AP-1 in endothelial cells results in the formation of small, round von Willebrand factor-containing organelles lacking other Weibel-Palade body markers. The data demonstrate that the requirement for clathrin and AP-1 is not restricted to one specific type of granule. Depletion of clathrin or AP-1 in Drosophila salivary glands resulted in the accumulation of glue protein both at the TGN and in small organelles of aberrant morphology. This finding extends the role of AP-1 and clathrin to the formation of granules containing mucoprotein cargo and suggests a broader requirement for this coat complex in granule production (Burgess, 2011).
How might AP-1 participate in glue granule formation? One possibility is that AP-1 and clathrin are directly involved in packaging glue granule cargo at the TGN. In mammalian cells, several transmembrane proteins are targeted to regulated secretory granules, including peptidyl-α-amidating monooxygenase, muclin, and phogrin. Indeed, phogrin has been shown to bind to AP-1 and AP-2 through well-conserved tyrosine and dileucine sorting motifs present in its cytosolic tail. How AP-1, a cytosolic coat protein, might interact with lumenal glue proteins in salivary cells remains to be determined. Because none of the known granule proteins contains a predicted transmembrane domain, a yet-unidentified transmembrane receptor might mediate this interaction (Burgess, 2011).
A distinct possibility is that AP-1 might be required to maintain a steady-state distribution of proteins that shuttle between the TGN and endosomes such that they are available at the TGN during granule formation. For instance, the protein convertase furin recycles between the TGN and endosomes and is required to process numerous secreted proteins such as von Willebrand factor. Importantly, furin is no longer concentrated at the TGN in μ1A-deficient fibroblasts. Thus failure to recycle transmembrane enzymes that play a crucial role in processing secreted cargo could also contribute to defective granule formation (Burgess, 2011).
Reduced levels of AP-1 resulted in intermediate-sized granules, suggesting AP-1 might have an additional role during glue granule maturation. The development of Drosophila glue granules is characterized by an overall increase in size and decrease in number, consistent with homotypic fusion of smaller granules over time (Farkas, 1999). Whether small and large granules are equally capable of fusing and whether fusion events are temporally regulated is not known. AP-1 might regulate granule maturation by sorting or retrieving membrane proteins required for homotypic fusion and eventual exocytosis. Additionally, AP-1 might function directly on maturing granules to remove missorted proteins, such as lysosomal hydrolases, similar to what has been reported for other types of secretory granules. In support of this view, live imaging revealed a dynamic association of AP-1 with immature granules. Further studies are needed to resolve whether AP-1 functions in the addition and/or removal of proteins from maturing glue granules (Burgess, 2011).
On the basis of the small size of mutant cells, AP-1 likely participates in additional trafficking pathways. In mammalian cells, AP-1A is ubiquitously expressed and required for trafficking between TGN and endosomes, whereas AP-1B is present only in polarized epithelial cells and is required for basolateral sorting from recycling endosomes. The sole AP-1 complex in Drosophila might mediate both functions in a single cell type. Interestingly, depletion of AP-1γ in salivary glands after granule formation caused the basolateral protein Discs large to redistribute to the apical surface, suggesting that AP-1 is required for basolateral targeting of proteins in this tissue. However, an independent analysis of AP-1μ null cells in the dorsal thorax epithelium failed to reveal a similar polarity defect (Benhra, 2011). This discrepancy might be due to cell type-specific requirements for AP-1 or to differences in RNAi versus mutant clones (Burgess, 2011).
The observation that the abundance of Sgs3-DsRed protein and several Sgs mRNAs is reduced upon AP-1 knockdown suggests the existence of a negative-feedback loop, whereby a block in anterograde secretory trafficking results in down-regulation of secretory genes. A block in secretion at the TGN could potentially induce the unfolded protein response, analogous to what happens upon depletion of the Arf1 GEF GBF1. However, GBF1 functions early in the secretory pathway, and knockdown of two Arf-GEFs that act on the TGN did not elicit a similar response. Alternatively, a block in anterograde trafficking might repress transcriptional activation of secretory genes by Drosophila CrebA and Forkhead (Fkh) by some as-yet-unknown mechanism (Burgess, 2011).
In addition to the AP-1 complex, the Drosophila genome encodes two other Golgi-localized clathrin adaptor proteins, EpsinR/LqfR and Golgi-localized, γ-ear-containing, ADP-ribosylation factor-binding (GGA) protein (Drosophila has only one GGA). LqfR partially colocalizes with AP-1 at the TGN in salivary gland cells and lqfR mutants exhibit small salivary glands, suggesting defects in granule biogenesis. It will be interesting to determine whether LqfR and GGA participate in glue granule biogenesis, especially since these clathrin adaptors might facilitate sorting of other types of cargo. For example, EpsinR has been shown to bind SNARE proteins and could function to provide vesicle identity to nascent glue-containing granules. SNAP-24 was previously identified as a glue granule-specific SNARE, although whether this SNARE mediates homotypic fusion of granules or functions during exocytosis of granules at the plasma membrane is unclear. Given the apparent similarities between glue granule and Weibel-Palade body biogenesis, as well as the high degree of conservation of TGN sorting machinery in Drosophila, the current findings suggest that Drosophila salivary glands are of great utility to further elucidate the mechanisms of biogenesis of regulated secretory granules (Burgess, 2011).
Homeostatic signaling systems stabilize neural function through the modulation of neurotransmitter receptor abundance, ion channel density, and presynaptic neurotransmitter release. Molecular mechanisms that drive these changes are being unveiled. In theory, molecular mechanisms may also exist to oppose the induction or expression of homeostatic plasticity, but these mechanisms have yet to be explored. In an ongoing electrophysiology-based genetic screen, 162 new mutations were tested for genes involved in homeostatic signaling at the Drosophila NMJ. This screen identified a mutation in the rab3-GAP gene. This study shows that Rab3-GAP is necessary for the induction and expression of synaptic homeostasis. Evidence is provided that Rab3-GAP relieves an opposing influence on homeostasis that is catalyzed by Rab3 and which is independent of any change in NMJ anatomy. These data define roles for Rab3-GAP and Rab3 in synaptic homeostasis and uncover a mechanism, acting at a late stage of vesicle release, that opposes the progression of homeostatic plasticity (Müller, 2011).
The function of Rab3-GAP and Rab3 have been analyzed extensively, both biochemically and genetically, in systems ranging from yeast to the mammalian central nervous system, and there are several (~10 on average) copies of Rab3-GTP on an individual synaptic vesicle. It has also been shown that Rab3 binds to several presynaptic proteins, most often in its GTP-bound form. Rab3-GAP is required to promote hydrolysis of Rab3-GTP to Rab3-GDP. It is unknown precisely when and where Rab3-GAP acts upon Rab3-GTP, but evidence suggests that this interaction may occur at the synapse. For example, Rab3-GTP is on the vesicle and delivered to the synapse, where it is found bound to the active zone associated protein RIM. Biochemical data indicate that clathrin-coated vesicles lack Rab3-GTP. In combination, these data place Rab3-GAP activity at or near the release site (Müller, 2011).
The data presented in this study are consistent with a model in which Rab3-GTP acts, directly or indirectly, to inhibit the progression of synaptic homeostasis at a late stage of vesicle release, and that Rab3-GAP functions to inactivate this action of Rab3-GTP. Loss of Rab3-GAP was shown to block both the rapid induction and sustained expression of synaptic homeostasis. Rab3 mutations alone do not block synaptic homeostasis and synaptic homeostasis proceeds normally in the
rab3–rab3-GAP double mutant. Genetically, these data indicate that the presence of Rab3 is required for the block of synaptic homeostasis observed in the
rab3-GAP mutant. Thus, Rab3 is likely to be the cognate GTPase for Rab3-GAP. Furthermore, in genetic terms, Rab3 functions to oppose the progression of synaptic homeostasis and, when it is removed, homeostasis proceeds (Müller, 2011).
The possibility is considered that homeostasis proceeds in the absence of Rab3 because another, redundant Rab takes the place of Rab3. In yeast membrane trafficking, there is evidence for semiredundant Rab function. However, this seems to be an exception because Rabs are hypothesized to have unique binding affinities for downstream effector proteins that are essential for their ability to define discrete membrane domains within membrane trafficking
and secretory systems. In C. elegans, Rab3 and Rab27 are both involved in synaptic vesicle release and they can be activated by a common exchange factor. However, based upon available genetic data, these Rabs do not appear to function redundantly during release. In Drosophila, loss of Rab3 causes a dramatic change in active zone size and organization (Graf, 2009). Thus, a redundant Rab would have to selectively and completely replace Rab3 function during synaptic homeostasis without rescuing synapse development, and this seems unlikely. Finally, since Rab3 is required for the complete block of synaptic homeostasis observed in the
rab3-GAP mutant, it is concluded that Rab3 itself participates in mechanisms that determine whether or not synaptic homeostasis will proceed (Müller, 2011).
Next, the possibility is considered that Rab3 accumulates in a GTP-bound form at the synapse in the rab3-GAP mutant background, and this accumulation could block homeostatic plasticity. This model is attractive because it could explain why loss of Rab3 does not block homeostatic plasticity. However, a direct test of this model failed to provide supporting evidence. A constitutively active rab3 transgene was expressed in the rab3 mutant background. This experiment should mimic the accumulation of Rab3-GTP in a rab3-GAP mutant background. It was found that the constitutively active rab3 transgene (rab3CA) is trafficked to the NMJ and localizes in a manner that is indistinguishable from overexpressed wild-type Rab3. Furthermore,
rab3CA has activity at the synapse because it rescues the defects in active zone organization that are caused by loss of rab3. In addition, constitutively active Rab3A was shown to biochemically interact with Rab3-GAP. However, the expression of rab3CA did not disrupt synaptic homeostasis. Therefore, aberrant accumulation of Rab3-GTP is not the cause of impaired synaptic homeostasis in the rab3-GAP mutant and another model should therefore be considered (Müller, 2011).
Another activity of Rab3-GAP that could be relevant to synaptic homeostasis is its ability to physically bind Rab3-GTP. It is known that Rab3-GTP can bind several synaptic proteins including RIM and Rabphillin. Moreover, this study provides evidence that Rab3's GTPase activity is not limiting during synaptic homeostasis. Therefore, it is proposed that Rab3-GAP competes for Rab3-GTP binding with another protein. If this other protein inhibits synaptic homeostasis when bound to the synaptic vesicle, then displacement by Rab3-GAP binding would be a required step for synaptic homeostasis to proceed. This model can explain all of the experimental data. First, Rab3-GAP would be necessary for synaptic homeostasis. Second, synaptic homeostasis would proceed normally in the
rab3 mutant because the homeostatic inhibitor would no longer localize to the synaptic vesicle. Third, overexpression of rab3CA in the rab3 mutant background would not block synaptic homeostasis because Rab3-GAP would still be able to compete for Rab3-GTP binding, displace the homeostatic inhibitor, and allow homeostatic plasticity to proceed (Müller, 2011).
This model is consistent with a conserved function of Rab proteins throughout the membrane trafficking and secretory pathways of organisms ranging from yeast to mammals. Rab proteins, in their GTP-bound state, function to nucleate the assembly of 'effector' protein complexes that define membrane microdomains. However, throughout the literature, more is known about the assembly of these Rab-dependent complexes than is known about their disassembly. The model assumes that the binding of Rab3-GAP to Rab3 and the Rab3CA mutant protein is sufficient to disrupt effector binding (including the proposed homeostatic inhibitor). It is generally believed that Rab-GAPs interact with their cognate GTPases with lower affinity than the effector proteins. However, although Rab3-GAP has a relatively low affinity for Rab3A, Rab3-GAP effectively competes with an effector (Rabphillin) for binding to Rab3A and this is true even when Rab3A harbors the Q81L mutation. These data support the possibility that Rab3-GAP could compete for effector binding and disrupt a Rab3-GTP dependent scaffold. This is not the only manner in which Rab3-GAP differs from other Rab-GAPs. Rab3-GAP is somewhat unique in that it does not contain additional protein-protein interaction motifs. Thus, unlike IQ-GAP proteins for instance, it seems unlikely that Rab3-GAP has unique functions that are independent
of Rab3, consistent with our double-mutant analysis of rab3 and rab3-GAP (Müller, 2011).
Finally, the possibility is considered that the rab3-GAP mutation could create a ceiling effect where baseline transmission is normal but release cannot be potentiated under any condition. Several pieces of data argue against this possibility. First, by elevating extracellular calcium, quantal content can be increased in the rab3-GAP mutant, indicating that there is no restriction on the absolute number of quanta that can be released. Second, during a stimulus train (20 Hz, 0.4 mM extracellular calcium), the
rab3-GAP mutant plateaus at a higher EPSP amplitude compared to wild-type. Again, there is no evidence for a ceiling effect in rab3-GAP. Finally, it has been demonstrated that mutations that cause a severe defect in baseline transmission can still undergo homeostatic compensation. Taken together, these data argue against a simple ceiling effect and support the conclusion that Rab3-GAP is directly involved in the mechanisms of synaptic homeostasis (Müller, 2011).
A question that cannot be addressed is which step in the model is modified during the induction of synaptic homeostasis. One interesting possibility is that the interaction of Rab3-GTP with the homeostatic repressor is regulated. For example,
if this interaction is stabilized, then homeostatic plasticity would be opposed and, conversely, if the interaction is weakened, then homeostasis would be allowed to proceed. A recent study in
C. elegans provided evidence that an unknown retrograde signal, from muscle to motoneuron, causes increased expression of YFP-Rab3 at the presynaptic terminal. Although no evidence was found for a similar phenomenon at the Drosophila NMJ during synaptic homeostasis, these data support the possibility that the Rab3/Rab3-GAP signaling complex could be a downstream, regulated target of a homeostatic, retrograde signal at the NMJ (Müller, 2011).
Interpreting the current data requires consideration of a recent study examining the effects of a rab3 mutation on synapse organization at the Drosophila NMJ (Graf, 2009). In the Graf study, it was discovered that a
rab3 mutation causes a dramatic accumulation of both the active zone-associated protein Bruchpilot (Brp, T-bars, the
Drosophila homolog of CAST/ELKS) and presynaptic calcium channels at a subset of active zones (Graf, 2009). Based on these and other data it was suggested that Rab3 promotes the nucleation of new active zones, and without this activity, active zones coalesce (Graf, 2009). It was possible to clearly dissociate any morphological reorganization of the NMJ from a blockade of synaptic homeostasis. The
rab3 mutants have altered NMJ morphology, but normal synaptic homeostasis, whereas rab3-GAP mutants have normal NMJ morphology and a defect in synaptic homeostasis) (Müller, 2011).
It is also important to consider why rab3 mutants do not show excessive homeostatic compensation. It is predicted that Rab3 and Rab3-GAP will not control the magnitude of the homeostatic response, just whether or not it is allowed to proceed. Additional negative feedback signaling mechanisms would be responsible for determining the magnitude of the homeostatic response. This would explain why no excessive homeostatic compensation was observed in the absence of Rab3. Thus, Rab3 and Rab3-GAP provide an additional layer of control on synaptic homeostasis, ensuring that modulation of release probability only occurs when, and perhaps where, appropriate (Müller, 2011).
Ultimately, it would be important to understand how presynaptic vesicle release is modulated during homeostatic plasticity. It is known from previously published data that a homeostatic increase in vesicle release is due to a change in presynaptic release probability without a change in active zone number. Mechanistically, the full functionality of presynaptic calcium channels is necessary for synaptic homeostasis. However, it remains unknown whether synaptic homeostasis involves a change in calcium channel number versus calcium channel function. Given these prior data, one possibility is that the homeostatic signaling system, identified in this study, acts upon presynaptic calcium channels to prevent a change in calcium influx. In this respect, the involvement of RIM and RIM binding protein are intriguing since RIM binds to Rab3-GTP and has been proposed to influence calcium-channel function (Müller, 2011 and references therein).
Rab3-GAP is the first protein to be implicated in the homeostatic modulation of presynaptic release that directly interacts with a resident synaptic vesicle protein. This fact, and analysis of baseline synaptic transmission in the
rab3-GAP mutant raise the possibility that the homeostatic modulation of presynaptic release also includes mechanisms that are independent of increased calcium influx. For example, a defect was observed in presynaptic release probability in the rab3-GAP mutant that occurs only when recording is performed in low extracellular calcium. This defect could reveal a function of Rab3-GAP during vesicle release, or it could reflect the activity of the proposed homeostatic repressor on baseline synaptic transmission. One possibility that could explain the decrease in release probability is that synaptic vesicles reside at a greater physical distance from the calcium channel in the
rab3-GAP mutant. When recording in low extracellular calcium, the calcium microdomains at the active zone would not effectively trigger the release of these more distant vesicles. This model would suggest that enhanced coupling of the synaptic vesicle and the calcium channel is part of the homeostatic modulation of presynaptic release. By extension, the action of the homeostatic repressor would be to prevent a tight association of the synaptic vesicle with the calcium channel. It is interesting to speculate that the homeostatic repressor could be Rabphillin. It has been shown that Rabphillin can compete with Rab3-GAP for binding to Rab3-GTP. Rabphillin has two C2 domains that could confer calcium-dependence to this protein-protein interaction and, by extension, homeostatic plasticity.
Ultimately, a molecular change that influences the functionality of the calcium sensor for vesicle fusion cannot be ruled out. Regardless, the data identify a homeostatic mechanism that functions at a late stage of vesicle release to modulate presynaptic release probability (Müller, 2011).
In combination with a previously published genetic screen, thirteen mutations have now been identified that disrupt the expression of synaptic homeostasis without severely altering baseline synaptic transmission. Among these genes are rab3-GAP and dysbindin. The mutant phenotypes forrab3-GAP and dysbindin are remarkably similar. In both cases, loss of function mutations have little effect on baseline transmission under standard recording conditions (0.4 mM extracellular calcium). However, decreasing extracellular calcium reveals a significant decrease in release probability. In agreement, an increase was also observed in short-term synaptic facilitation in both mutations. Furthermore, neither mutation has an effect on synapse morphology or active zone number. In dysbindin mutants, these effects were shown to be downstream or independent of presynaptic calcium influx. It is tempting to place Dysbindin into the proposed model for homeostatic plasticity.
One possibility is that Dysbindin functions to stabilize the close association of synaptic vesicles with the presynaptic calcium channel. The absence of Dysbindin would therefore phenocopy the rab3-GAP mutant but function through a different set of molecular interactions on the synaptic vesicle. The similarity between the phenotypes of dysbindin and rab3-GAP are also interesting because dysbindin has been linked to schizophrenia in human. The intriguing possibility that Dysbindin interacts with Rab3-Rab3-GAP signaling will be the subject of future studies (Müller, 2011).
Burgess, J., et al. (2011). AP-1 and clathrin are essential for secretory granule biogenesis in Drosophila. Mol. Biol. Cell 22(12): 2094-105. PubMed Citation: 21490149
Chi, S., Cao, H., Chen, J. and McNiven, M. A. (2008). Eps15 mediates vesicle trafficking from the trans-Golgi network via an interaction with the clathrin adaptor AP-1. Mol. Biol. Cell 19: 3564-3575. PubMed Citation: 18524853
Goto, S., Taniguchi, M., Muraoka, M., Toyoda, H., Sado, Y., Kawakita, M. and Hayashi, S. (2001). UDP-sugar transporter implicated in glycosylation and processing of Notch. Nat. Cell Biol. 3: 816-822. 11533661
Graf, E. R., et al. (2009). Rab3 dynamically controls protein composition at active zones. Neuron 64(5): 663-77. PubMed Citation: 20005823
Herpers, B. H. A. and Rabouille, C. (2004). mRNA localization and ER-based protein sorting mechanisms dictate the use of tER-Golgi units involved in gurken transport in Drosophila oocytes. Mol. Biol. Cell 15: 5306-5317. PubMed citation; Online text
Kondylis, V. and Rabouille, C. (2003). A novel role for dp115 in the organization of tER sites in Drosophila: J. Cell Biol. 162: 185-198. PubMed citation: 12876273
Kondylis, V., Spoorendonk, K. M. and Rabouille, C. et al. (2005). dGRASP localization and function in the early exocytic pathway in Drosophila S2 cells. Mol. Biol. Cell 16: 4061-4072. PubMed citation: 15975913
Kondylis, V., et al. (2007). The golgi comprises a paired stack that is separated at G2 by modulation of the actin cytoskeleton through Abi and Scar/WAVE.
Dev. Cell 12(6): 901-15. PubMed citation: 17543863
Lui-Roberts, W. W., et al. (2005). An AP-1/clathrin coat plays a novel and essential role in forming the Weibel-Palade bodies of endothelial cells. J. Cell Biol. 170: 627-636. PubMed Citation: 16087708
Müller, M., Pym, E. C., Tong, A. and Davis, G. W. (2011). Rab3-GAP controls the progression of synaptic homeostasis at a late stage of vesicle release. Neuron 69(4): 749-62. PubMed Citation: 21338884
Tsarouhas, V., et al. (2007). Sequential pulses of apical epithelial secretion and endocytosis drive airway maturation in Drosophila. Dev. Cell 13: 214-225. PubMed Citation: 17681133
Wingen, C., Stumpges, B., Hoch, M. and Behr, M. (2009). Expression and localization of clathrin heavy chain in Drosophila melanogaster. Gene Expr. Patterns. 9: 549-554. PubMed Citation: 19577664
Yano, H., et al. (2005). Distinct functional units of the Golgi complex in Drosophila cells. Proc. Natl. Acad. Sci. 102(38): 13467-72. 16174741
Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.