TL;DR: It is concluded that ATF6-mediated induction of XBP1 mRNA is important to the production of pXBP1(S), activation of X BP1-target genes, and protection of cells from ER stress.
Abstract: XBP1 is a transcription factor downstream of IRE1, a transmembrane protein in the endoplasmic reticulum (ER) which functions as a sensor and transducer of ER stress. XBP1 mRNA is constitutively expressed at a low level as an intron-containing precursor mRNA (unspliced mRNA), which is subject to IRE1-mediated splicing reaction upon ER stress to produce the active form of XBP1, pXBP1(S). Because the XBP1 promoter carries a perfect ER stress-response element, namely, the cis-acting element responsible for the induction of ER chaperones, and XBP1 mRNA is induced in response to ER stress with a time course similar to that of ER chaperone mRNAs, it is conjectured that transcription factor ATF6, activated immediately upon ER stress, induces the transcription of not only ER chaperone genes but also of XBP1 gene, such that pXBP1(S) produced by the splicing of an increased level of XBP1 mRNA escapes from proteasome-mediated degradation. Here, we examined this notion by determining the induction of XBP1 mRNA and pXBP1(S) in mutant Chinese hamster ovary (M19) cells deficient in Site-2 protease, which executes the last step of ER stress-induced activation of ATF6. We found that the induction of XBP1 mRNA and pXBP1(S) was greatly reduced in M19 cells as compared with wild-type cells, leading to a marked reduction in the extent of induction of XBP1-target gene. M19 cells were much more sensitive to ER stress than wild-type cells. Importantly, overexpression of XBP1 unspliced mRNA in M19 cells reversed all of these phenotypes. We concluded that ATF6-mediated induction of XBP1 mRNA is important to the production of pXBP1(S), activation of XBP1-target genes, and protection of cells from ER stress.
TL;DR: The results suggest that the local production of PtdIns(3)P implicates the fusion of macropinosomes via EEA1 as well as conventional early endosomes and may well be a cell-type specific feature of A431 cells.
Abstract: In the process of receptor-mediated endocytosis, the fusion of endosomes in vitro is known to be inhibited by wortmannin or LY294002; inhibitors of phosphoinositide 3-kinase (PI3K), suggesting that the activity of PI3K is required for the fusion of early endosomes. In macropinocytosis, a process of bulk fluid-phase endocytosis, however, it remains unclear whether PI3K is required for the fusion of macropinosomes, since the macropinosome formation is inhibited by the PI3K inhibitors. In this study, we examined the effect of 3-methlyadenine (3-MA), which shows a distinct specificity to the PI3K classes from wortmannin and LY294002, on the macropinosome formation and fusion in EGF-stimulated A431 cells. Unlike wortmannin or LY294002, 3-MA did not inhibit the uptake of fluorescent dextran by macropinocytosis. However, the fusion of macropinosomes was inhibited by 3-MA. By imaging of live-cells expressing fluorescent protein-fused tandem FYVE domains, we found that PtdIns(3)P appeared on the macropinosomal membrane shortly after the closure of macropinocytic cups and remained on macropinosomes even at 60-min age. The production of PtdIns(3)P and the recruitment of EEA1 to macropinosomes were abolished by the 3-MA treatment. Therefore, it is likely that 3-MA impairs recruitment of EEA1 by inhibiting PtdIns(3)P production and resultantly blocks the fusion of macropinosomes. These results suggest that the local production of PtdIns(3)P implicates the fusion of macropinosomes via EEA1 as well as conventional early endosomes. However, the long association of PtdIns(3)P with macropinosomes may well be a cell-type specific feature of A431 cells.
TL;DR: This work purified an AMSH-binding protein, CHMP3, which is an ESCRT-III subunit, indicating AMSH might also play a role in MVB/late endosomes, and suggests that both the DUB activity of AMSH and its ChMP3-binding ability are required to clear ubiquitinated cargo from endosome.
Abstract: The appropriate sorting of vesicular cargo, including cell-surface proteins, is critical for many cellular functions. Ubiquitinated cargo is targeted to endosomes and digested by lysosomal enzymes. We previously identified AMSH, a deubiquitination enzyme (DUB), to be involved in vesicular transport. Here, we purified an AMSH-binding protein, CHMP3, which is an ESCRT-III subunit. ESCRT-III functions on maturing endosomes, indicating AMSH might also play a role in MVB/late endosomes. Expression of an AMSH mutant lacking CHMP3-binding ability resulted in aberrant endosomes with accumulations of ubiquitinated cargo. Nevertheless, CHMP3-binding capability was not essential for AMSH's in vitro DUB activity or its endosomal localization, suggesting that, in vivo, the deubiquitination of endosomal cargo is CHMP3-dependent. Ubiquitinated cargo also accumulated on endosomes when catalytically inactive AMSH was expressed or AMSH was depleted. These results suggest that both the DUB activity of AMSH and its CHMP3-binding ability are required to clear ubiquitinated cargo from endosomes.
TL;DR: Analysis of M19 cells suggests that Rip-mediated activation of ATF6 is important for the homeostasis of the ER in not only ER-stressed but also unstressed cells.
Abstract: Mammalian transcription factor ATF6 is constitutively synthesized as a type II transmembrane protein embedded in the endoplasmic reticulum (ER). It is activated when unfolded proteins are accumulated in the ER under ER stress through a process called regulated intramembrane proteolysis (Rip), in which ATF6 is transported from the ER to the Golgi apparatus where it undergoes sequential cleavage by Site-1 and Site-2 proteases. The cytosolic transcription factor domain of ATF6 liberated from the Golgi membrane enters the nucleus where it activates transcription of ER-localized molecular chaperones and folding enzymes, leading to the maintenance of the homeostasis of the ER. Here, we analyzed M19 cells, a mutant of Chinese hamster ovary cells deficient in Site-2 protease. It was previously shown that M19 cells are defective in the induction of mRNA encoding the major ER chaperone BiP. In M19 cells, ATF6 was not converted from the membrane-bound precursor form to the cleaved and nuclear form as expected. Moreover, some of the ATF6 was constitutively relocated to the Golgi apparatus, where it was cleaved by Site-1 protease, and remained associated with the Golgi apparatus, indicating that the ER of M19 cells was constitutively stressed. Consistent with this notion, the two other ER stress response mediators, IRE1 and PERK, were also constitutively activated in M19 cells. M19 cells showed inefficient secretion of a model protein. These results suggest that Rip-mediated activation of ATF6 is important for the homeostasis of the ER in not only ER-stressed but also unstressed cells.
TL;DR: It is demonstrated here that cytoplasmic domain of HB-EGF is phosphorylated by external stimuli, and that the phosphorylation site is involved in HB- EGF-dependent tumorigenesis.
Abstract: Heparin-binding EGF-like growth factor (HB-EGF) is synthesized as a transmembrane precursor protein that is anchored to the plasma membrane. The extracellular EGF-like domain acts as a mitogen and motogen upon ectodomain shedding, but the functional roles of the transmembrane and cytoplasmic domains are largely unknown. We demonstrate here that cytoplasmic domain of HB-EGF is phosphorylated by external stimuli, and that the phosphorylation site is involved in HB-EGF-dependent tumorigenesis. Treatment of Vero cells overexpressing human HB-EGF with 12-O-tetradecanoylphorbol-13-acetate (TPA) caused ectodomain shedding of HB-EGF and generated two carboxyl (C)-terminal fragments with distinct electrophoretic mobilities. Mutation analysis showed that Ser207 in the cytoplasmic domain of HB-EGF is phosphorylated upon TPA stimulation, generating two C-terminal fragments with distinct phosphorylation states. Treatment of cells with lysophosphatidic acid, anisomycin, and calcium ionophore, all of which are known to induce ectodomain shedding, also caused phosphorylation of HB-EGF. Although ectodomain shedding and phosphorylation of HB-EGF occurred coordinately, Ala substitution of Ser207 had no effect on TPA-induced or constitutive ectodomain shedding. Injection of cells overexpressing HB-EGF into nude mice showed that Ala substitution of Ser207 reduced the tumorigenic activity of HB-EGF, even though the cell surface level and ectodomain shedding of HB-EGF were not affected by the mutation. Moreover, we found that the cytoplasmic domain of another EGFR ligand, transforming growth factor-alpha, is phosphorylated upon TPA stimulation. Thus, the present results suggest a novel role for the cytoplasmic domain of HB-EGF and other EGF family growth factors that is regulated by phosphorylation.
TL;DR: Mouse embryonic stem (mES) cells can be maintained in undifferentiated state in the presence of a cytokine, leukemia inhibitory factor (LIF), and it was shown that STAT3 could bind to the promoter region of Jmjd1a gene, demonstrating that JmJd 1a might be a critical signaling molecule underlying the maintenance of pluripotency in mES cells.
Abstract: Mouse embryonic stem (mES) cells can be maintained in undifferentiated state in the presence of a cytokine, leukemia inhibitory factor (LIF). Many investigators found that STAT3 activation is important for the maintenance of pluripotency by LIF. However, the downstream pathways of STAT3 activation are still unknown. To look for STAT3-downstream target genes, we performed DD-RT PCR in the presence or absence of LIF. Through further confirmation, we finally selected 8 genes whose expressions were significantly dependent upon the presence of LIF. Among them, Jmjd1a was down-regulated after LIF withdrawal, and it was selected for further investigation. Its expression started to decrease 1 day after the removal of LIF, and disappeared on day 3. It was also shown that STAT3 could bind to the promoter region of Jmjd1a gene. These data demonstrate that Jmjd1a might be a critical signaling molecule underlying the maintenance of pluripotency in mES cells.
TL;DR: It is suggested that Poldelta but not Poleta is essential for C. elegans in early embryogenesis but only modestly at later stages, as early embryogenic is characterized by high levels of DNA replication, andPoleta may confer UV resistance in C. larvae.
Abstract: Defects in the human XPV/POLH gene result in the variant form of the disease xeroderma pigmentosum (XP-V). The gene encodes DNA polymerase η (Polη), which catalyzes translesion synthesis (TLS) past UV-induced cyclobutane pyrimidine dimers (CPDs) and other lesions. To further understand the roles of Polη in multicellular organisms, we analyzed phenotypes caused by suppression of Caenorhabditis elegans POLH (Ce-POLH) by RNA interference (RNAi). F1 and F2 progeny from worms treated by Ce-POLH-specific RNAi grew normally, but F1 eggs laid by worms treated by RNAi against Ce-POLD, which encodes Polδ did not hatch. These results suggest that Polδ but not Polη is essential for C. elegans embryogenesis. Polη-targeted embryos UV-irradiated after egg laying were only moderately sensitive. In contrast, Polη-targeted embryos UV-irradiated prior to egg laying exhibited severe sensitivity, indicating that Polη contributes significantly to damage tolerance in C. elegans in early embryogenesis but only modestly at later stages. As early embryogenesis is characterized by high levels of DNA replication, Polη may confer UV resistance in C. elegans, perhaps by catalyzing TLS in early embryogenesis.
TL;DR: It is shown that Merkel cells isolated from rat footpad skin were successfully cultured in a monolayer with keratinocytes, and that nerve cells, but not keratinocyte, may produce some survival factors other than the cytokines above for Merkel cells, although Merkel cells may be a terminally differentiated cell type.
Abstract: Merkel cells have been assumed to guide nerve fibers to the skin. However, there has been little in vitro evidence that supports this hypothesis, because there is no suitable established culture system of Merkel cells. Here we show that Merkel cells isolated from rat footpad skin were successfully cultured in a monolayer with keratinocytes. Keratinocytes did not affect any structural changes in Merkel cells. When nerve cells (NG108-15 or PC12) were added to the culture system, both nerve fibers and cytoplasmic processes of Merkel cells outgrew and cooperatively organized synapse-like structures at their contact points. Nerve cells promoted Merkel cell survival, compared with keratinocytes only. Merkel cell proliferation was not detected in all conditions, even with nerve growth factor, neurotrophin-3, interleukin-6 and tumor necrosis factor-alpha. The data suggest, firstly, that Merkel cells may guide nerve fibers to the skin by interacting with nerve cells; and, secondly, that nerve cells, but not keratinocytes, may produce some survival factors other than the cytokines above for Merkel cells, although Merkel cells may be a terminally differentiated cell type. Our method could open a way to study Merkel cell biology.
TL;DR: Type I myosin and F-actin accumulate at the leading edge area of the forespore membrane and may play a pivotal role in its assembly and suggest a close relationship between Cam2 and Myo1.
Abstract: Myo1, a heavy chain of type I myosin of the fission yeast Schizosaccharomyces pombe, is essential for sporulation. Here we have analyzed the expression, localization and cellular function of the type I myosin light chain calmodulin, Cam2, encoded by cam2+. Transcription of cam2+ was constitutive and markedly enhanced in meiosis. The cam2 null mutant was viable and completed sporulation normally at 28°C, but formed four-spored asci poorly at 34°C. In those sporulation-defective cells, the forespore membrane was formed abnormally. A Cam2-GFP fusion protein accumulated at the cell poles in interphase cells and at the medial septation site in postmitotic cells, colocalizing with Myo1 and F-actin patches. During the mating process, a single Cam2-GFP dot was detected at the tip of the mating projection. During meiosis-I, the Cam2-GFP dots dispersed into the cell periphery and the cytoplasm. At metaphase-II, intense Cam2-GFP signals appeared near Meu14 rings which were formed at the leading edge of expanding forespore membranes. This localization of Cam2 was dependent upon Myo1; and sporulation defect of cam2Δ at 34°C was alleviated by overexpressing Myo1ΔIQ. These results suggest a close relationship between Cam2 and Myo1. In addition, both F-actin and Myo1 localized with Cam2 in the leading edge region. In summary, type I myosin and F-actin accumulate at the leading edge area of the forespore membrane and may play a pivotal role in its assembly.
TL;DR: The notion that early endosomes can serve as a sorting station for not only internalized plasma membrane proteins in the endocytic pathway but also newly synthesized membrane proteinsIn the post-Golgi secretory pathway is supported.
Abstract: Newly synthesized membrane proteins are sorted in the trans-Golgi network (TGN) on the basis of sorting signals carried in their cytoplasmic domains and delivered to their final destinations in the secretory and endocytic pathways. Although previous studies have suggested the involvement of early endosomes in the biosynthetic pathway of transmembrane proteins, the precise trafficking routes followed by the newly synthesized plasma membrane proteins, such as transferrin receptors (TfRs), after exit from the TGN remain unclear. In this report, first, we demonstrated the advantages of photoactivating PA-GFP, a variant of the Aequorea victoria green fluorescent protein (GFP), with multiphoton laser light rather than single-photon laser light, in terms of photoactivation efficiency and spatial resolution. We then applied the multiphoton photoactivation technique to selectively photoactivate the TfR tagged with PA-GFP (PA-GFP-TfR) at the TGN, and monitored the movement of the photoactivated PA-GFP-TfR in live cells. We observed that the PA-GFP-TfR photoactivated at the TGN are transported to the Tfn(+)EEA1(+) endosomal compartments after exiting the TGN. These data support the notion that early endosomes can serve as a sorting station for not only internalized plasma membrane proteins in the endocytic pathway but also newly synthesized membrane proteins in the post-Golgi secretory pathway.
TL;DR: The results indicated that some p53-null non-small cell lung cancers could be successfully treated when X-ray radiotherapy was administered with subsequent or concurrent cisplatin chemotherapy.
Abstract: Radiotherapy is an effective approach to treating many types of cancer. Recent progress in radiotherapy technology, such as intensity-modulated radiation therapy (IMRT) and three-dimensional (3D) radiotherapy, allow precise energy transfer to the tumor, which has improved local control rates. However, the emergence of tolerant cells during or after radiotherapy remains problematic. In the present study, we first established a cell population from H1299, the p53-null non-small cell lung cancer cell line, by 10 Gy irradiation using 6 MV X-rays. The radio- and chemosensitivity of this cell population (referred to as H1299-IR) was determined using colony formation analyses and MTS assays. Compared with the parental cell line, the radiosensitivity of H1299-IR was apparently the same. H1299 and H1299-IR were both more radio tolerant than the A549 cell line. However, H1299-IR became significantly more sensitive to cisplatin, an antitumor agent. After exposure to 25 mug/ml cisplatin for 2 h, parental cells steadily grew during the MTS assay, whereas the sensitivity of H1299-IR cells doubled both at 24 and 48 h. Microarray analysis of over 30,000 H1299-IR genes (Agilent Technology) revealed that 12 and 15 genes were up- (> 2.0) and down- (< 2.0) regulated, respectively. Rad51d (homologous recombination repair protein) gene was down-regulated 2.8-fold, whereas matrix metalloproteinase 1 (collagenase-1) gene was up-regulated 4.4-fold. These results indicated that some p53-null non-small cell lung cancers could be successfully treated when X-ray radiotherapy was administered with subsequent or concurrent cisplatin chemotherapy.
TL;DR: It is likely that the megalin in the ES lumen is a soluble component and may be endocytosed by the ES epithelial cells, and the tectorial membrane, an acellular structure in the cochlea, gave a strongmegalin immunoreaction.
Abstract: The endolymphatic sac (ES) is believed to play an important role in maintaining homeostasis in the inner ear by the absorption and endocytosis of endolymph. Megalin is a 600-kDa multiligand endocytic receptor expressed in certain types of absorptive epithelia including kidney proximal tubules. We analyzed the immunoreactivity for megalin in rat ES by immunofluorescence, immunogold electron microscopy, and immunoblotting. With immunostaining, the luminal substances of the ES were strongly stained for megalin. Megalin was also localized in luminal macrophage-like cells and both types of epithelial cell (mitochondria-rich cells and ribosome-rich cells). In these cells, the megalin was localized in the lumen of endosomes, but was not membrane associated. This localization pattern indicates that the megalin in these cells is not a membrane receptor, but merely one of the constituents that are endocytosed from the lumen of the ES. Immunoblotting indicated that the megalin in the ES is a 210-kDa molecule lacking a cytoplasmic domain. This suggests that the megalin in the ES may be a soluble form, different from the 600-kDa membrane-bound receptor expressed in kidneys. Taken together, it is likely that the megalin in the ES lumen is a soluble component and may be endocytosed by the ES epithelial cells. Furthermore, we found that the tectorial membrane, an acellular structure in the cochlea, gave a strong megalin immunoreaction. Since the cochlea is connected to the ES, the megalin may be transported alone or with the components of the tectorial membrane from the cochlea to the ES lumen through longitudinal flow.
TL;DR: Investigation of temporal and spatial expression patterns of Drosophila beta-sarcoglycan during development by immunohistochemistry finds that subcellular localization of dScgbeta dramatically changes during mitosis through possible association with tubulin, pointing to a complex role of sarcoglycans in non-muscle tissues.
Abstract: One of the proposed roles of sarcoglycan is to stabilize dystrophin glycoprotein complexes in muscle sarcolemma. Involvement in signal transduction has also been proposed and abnormalities in some sarcoglycan genes are known to be responsible for muscular dystrophy. While characterization of sarcoglycans in muscle has been performed, little is known about its functions in the non-muscle tissues in which mammalian sarcoglycans are expressed. Here, we investigated temporal and spatial expression patterns of Drosophila β-sarcoglycan (dScgβ) during development by immunohistochemistry. In addition to almost ubiquitous expression in various tissues and organs, as seen for its mammalian counterpart, anti-dScgβ staining data of embryos, eye imaginal discs, and salivary glands demonstrated cytoplasmic localization during S phase in addition to plasma membrane staining. Furthermore we found that subcellular localization of dScgβ dramatically changes during mitosis through possible association with tubulin. These observations point to a complex role of sarcoglycans in non-muscle tissues.
TL;DR: The mechanical strength of sarcomere structures of skeletal muscle was studied by rupturing single myofibrils of rabbit psoas muscle by submicromanipulation techniques using microbeads coated with alpha-actinin and numerous molecular components having the mechanical strength sufficient to sustain the contractile force produced by the actomyosin system.
Abstract: The mechanical strength of sarcomere structures of skeletal muscle was studied by rupturing single myofibrils of rabbit psoas muscle by submicromanipulation techniques. Microbeads coated with alpha-actinin were attached to the surface of myofibrils immobilized to coverslip. By use of either optical tweezers or atomic force microscope, the attached beads were captured and detached from the myofibrils. During the detachment of the beads, the actin filaments bound specifically to the beads were peeled off from the bulk structures of myofibrils, thus rupturing the peripheral components of the myofibrils bound to the actin filaments. By analyzing the ruptures thus produced in various myofibril preparations, it was found that the sarcomere structure of myofibrils is maintained by numerous molecular components having the mechanical strength sufficient to sustain the contractile force produced by the actomyosin system. The present techniques could be applied to study the mechanical strength of cellular organelles containing actin filaments as their component.
TL;DR: The role of ErbB receptors in heart function and development is introduced and the physiological function of HB-EGF is focused on, focusing on the physiological functions of the heparin-binding EGF-like growth factor.
Abstract: The epidermal growth factor (EGF)-ErbB signaling network is composed of multiple ligands of the EGF family and four tyrosine kinase receptors of the ErbB family. In higher vertebrates, these four receptors bind a multitude of ligands. Ligand binding induces the formation of various homo- and heterodimers of ErbB, potentially providing for a high degree of signal diversity. ErbB receptors and their ligands are expressed in a variety of tissues throughout development. Recent advances in gene targeting strategies in mice have revealed that the EGF-ErbB signaling network has fundamental roles in development, proliferation, differentiation, and homeostasis in mammals. The heparin-binding EGF-like growth factor (HB-EGF) is a member of the EGF family of growth factors that binds to and activates the EGF receptor (EGFR/ErbB1) and ErbB4. Recent studies using several mutant mice lacking HB-EGF expression have revealed that HB-EGF has a critical role in normal heart function and in normal cardiac valve formation in conjunction with ErbB receptors. HB-EGF signaling through ErbB2 is essential for the maintenance of homeostasis in the adult heart, whereas HB-EGF signaling through EGFR is required during cardiac valve development. In this review, we introduce and discuss the role of ErbB receptors in heart function and development, focusing on the physiological function of HB-EGF in these processes.
TL;DR: It is concluded that ER stress-induced reduction of ATF6 represents a general feature of the ATF6 activation process, and reduction of disulfide bridges and transport of reduced monomer occurred in response to glucose starvation.
Abstract: Mammalian transcription factor ATF6 is constitutively synthesized as a type II transmembrane protein embedded in the endoplasmic reticulum (ER). Upon ER stress ATF6 is transported to the Golgi apparatus where it is cleaved to release its cytoplasmic domain. This is then translocated into the nucleus where it activates transcription of ER-localized molecular chaperones and folding enzymes to maintain the homeostasis of the ER. We recently found that, owing to the presence of intra- and intermolecular disulfide bridges, ATF6 occurs in unstressed ER in monomer, dimer and oligomer forms. Disulfide-bonded ATF6 is reduced on treatment of cells with various chemical ER stress inducers, and only the reduced monomer ATF6 reaches the Golgi apparatus. In this study, we evoked ER stress under more physiological conditions, namely, glucose starvation, and analyzed its consequence for ATF6 activation. Glucose starvation activated ATF6 and induced the ER chaperone BiP, albeit weakly. ATF6 was thus dissociated from BiP, transported to the Golgi apparatus, and cleaved. Glucose starvation enhanced the synthesis of ATF6 approximately two-fold, probably via transcriptional induction. Importantly, reduction of disulfide bridges and transport of reduced monomer occurred in response to glucose starvation. We conclude that ER stress-induced reduction of ATF6 represents a general feature of the ATF6 activation process.
TL;DR: Results suggest that Cdc50p-Drs2p plays an important role in the Arf1p-mediated formation of CCVs for the retrieval pathway from early endosomes to the TGN.
Abstract: Drs2p, the catalytic subunit of the Cdc50p-Drs2p putative aminophospholipid translocase, has been implicated in conjunction with the Arf1 signaling pathway in the formation of clathrin-coated vesicles (CCVs) from the TGN. Herein, we searched for Arf regulator genes whose mutations were synthetically lethal with cdc50Δ, and identified the Arf GAP gene GCS1. Most of the examined transport pathways in the Cdc50p-depleted gcs1Δ mutant were nearly normal, including endocytic transport to vacuoles, carboxypeptidase Y sorting, and the processing and secretion of invertase. In contrast, this mutant exhibited severe defects in the early endosome-to-TGN transport pathway; proteins that are transported via this pathway, such as the v-SNARE Snc1p, the t-SNARE Tlg1p, and the chitin synthase III subunit Chs3p, accumulated in TGN-independent aberrant membrane structures. We extended our analyses to clathrin adaptors, and found that Gga1p/Gga2p and AP-1 were also involved in this pathway. The Cdc50p-depleted gga1Δ gga2Δ mutant and the gcs1Δ apl2Δ (the β1 subunit of AP-1) mutant exhibited growth defects and intracellular Snc1p-containing membranes accumulated in these cells. These results suggest that Cdc50p-Drs2p plays an important role in the Arf1p-mediated formation of CCVs for the retrieval pathway from early endosomes to the TGN.