Cells in situ are often polarized and have multiple plasma membrane domains. To establish and maintain these domains, polarized transport is essential, and its impairment results in genetic disorders. Nevertheless, the underlying mechanisms of polarized transport have not been elucidated. Drosophila photoreceptor offers an excellent model for studying this. We found that Rab10 impairment significantly reduced basolateral levels of Na+K+ATPase, mislocalizing it to the stalk membrane, which is a domain of the apical plasma membrane. Furthermore, the shrunken basolateral and the expanded stalk membranes were accompanied with abnormalities in the Golgi cisternae of Rab10-impaired retinas. The deficiencies of Rab10-GEF Crag or the Rab10 effector Ehbp1 phenocopied Rab10 deficiency, indicating that Crag, Rab10 and Ehbp1 work together for polarized trafficking of membrane proteins to the basolateral membrane. These phenotypes were similar to those seen upon deficiency of AP1 or clathrin, which are known to be involved in the basolateral transport in other systems. Additionally, Crag, Rab10 and Ehbp1 colocalized with AP1 and clathrin on the trans-side of Golgi stacks. Taken together, these results indicate that AP1 and clathrin, and Crag, Rab10 and Ehbp1 collaborate in polarized basolateral transport, presumably in the budding process in the trans-Golgi network.

Epithelial cells contain two well-differentiated plasma membrane domains. These are absorptive or secretory apical membranes and basolateral membranes, and are essential for cell viability. These structures are formed and maintained by polarized vesicle transport of lipids and proteins from the trans-Golgi network (TGN) to specific plasma membrane domains (Rodriguez-Boulan et al., 2005; Rodriguez-Boulan and Macara, 2014). Impairment of polarized transport results in genetic disorders, including microvillus inclusion disease, polycystic kidney disease and retinal degeneration (Charron et al., 2000; Cotton et al., 2013; Hollingsworth and Gross, 2012; Müller et al., 2008; Ryan et al., 2010; Schneeberger et al., 2015; Vogel et al., 2017; Xiong and Bellen, 2013). The Drosophila retina is suitable as a genetics-based model for studying the mechanisms of polarized transport (Pocha et al., 2011; Satoh et al., 2013, 2016; Xiong et al., 2012; Yano et al., 2012). Drosophila photoreceptors are differentiated from epithelial cells (Cagan and Ready, 1989; Wolff and Ready, 1993). During their morphogenesis, axons extend from the basolateral membrane, and the apical plasma membrane is subdivided into the central photosensitive rhabdomere and the surrounding stalk membrane (Karagiosis and Ready, 2004; Ogi et al., 2019; Pellikka et al., 2002).

Proteins in the Rab family of small GTPases regulate the specificity between donor and acceptor membranes in the processes of vesicle budding, docking, tethering and fusion during transport (Pfeffer, 2013; Stenmark, 2009). Each Rab protein is activated by specific guanine-nucleotide exchange factors (GEFs) and recruits many effector proteins (Barr and Lambright, 2010; Ishida et al., 2016). In Drosophila photoreceptors, Rab6 and Rab11 regulate the apical transport of Rh1, which is the rhodopsin expressed in the R1-R6 outer retinal photoreceptor cells. In Rab6-deficient photoreceptors, Rh1 does not accumulate in the rhabdomeres and the transmembrane protein Crumbs does not considerably localize in the stalk membrane. Instead, both localize in multi-vesicular bodies (MVBs) and transport to both the rhabdomere and the stalk membrane is inhibited; however, basolateral transport is not affected (Iwanami et al., 2016). Deficiency of any constituent of the Rab11, dRip11 and MyoV trio-complex causes specific rhabdomere transport defects and substantial accumulation of Rh1-bearing vesicles in the cytoplasm. It does not, however, affect transport towards the basolateral and stalk membranes (Li et al., 2007; Satoh et al., 2005). Rab8, Rab11, EhbpL1, Syn3 and MyoVb have been identified as factors involved in apical transport in mouse intestinal epithelial cells (Nakajo et al., 2016; Sato et al., 2014, 2007; Sobajima et al., 2015) and Caco2 cells (Vogel et al., 2017, 2015). Mutations in the MYO5B gene cause microvillus inclusion disease in humans (Müller et al., 2008; Schneeberger et al., 2015; Vogel et al., 2015).

The Rab proteins involved in the polarized transport of basolateral transmembrane proteins have not yet been clearly characterized in either mammalian systems or fly photoreceptors. Studies using Madin Darby canine kidney (MDCK) cells indicated that Rab8 and Rab10 are involved in basolateral transport (Ang et al., 2003; Henry and Sheff, 2008; Huber et al., 1993; Schuck et al., 2007). In preceding studies with Rab8-knockout mice, however, it was shown that Rab8 is essential for apical, but not basolateral, transport in their intestines (Nakajo et al., 2016; Sato et al., 2014, 2007). Because the Rab10-knockout mouse is embryonically lethal, basolateral transport in its epithelial tissue has not been investigated (Lv et al., 2015). Similarly, in Drosophila, together with Rab11 and the exocyst complex, Rab8 is essential for apical membrane addition during epithelial tissue formation in early embryos (Holly et al., 2015; Mavor et al., 2016). The Rab10-GEF Crag is essential for the basal transport of basement membrane (BM) components, such as laminin, collagen IV and perlecan, in the follicle cells of Drosophila ovary (Denef et al., 2008). Nevertheless, the distribution patterns of the apical receptor Notch, the adherens junction (AJ) protein DE-cadherin, and the basolateral adhesion molecules Fasciclin II, Fasciclin III and Neuroglian (NRG) were indistinguishable in wild-type and Crag-deficient follicle cells. Hence, Crag and Rab10 have been proposed to specifically regulate the transport of BM components, which are giant extracellular matrix proteins, but not the transport of apical or basolateral transmembrane proteins (Denef et al., 2008; Lerner et al., 2013). In addition to Crag-null mutant cells, Rab10-RNAi or Rab10 dominant-negative protein expression results in the mis-transport of BM components to the apical lumen (Lerner et al., 2013). One study reported a role for Rab8 in apical rather than basolateral transport in early embryos (Mavor et al., 2016). However, another report showed that Rab8 and its GEF, stratum, are also involved in the transport of BM components (Devergne et al., 2017). Ehbp1 was first identified as the binding protein for the EH domain and actin filaments in cultured adipocytes (Guilherme et al., 2004), and thereafter it was discovered as the Rab10 effector involved in the recycling pathway in Caenorhabditis elegans (C. elegans) (Shi et al., 2010; Wang et al., 2016). A recent fly study indicated that the overexpression of Rab10, Crag or Ehbp1 increased the secretion of components to the BM, resulting in rigid eggs (Isabella and Horne-Badovinac, 2016). Therefore, Ehbp1 is probably a Rab10 effector in Drosophila. In summary, Rab10, Crag and Ehbp1 collaborate in the polarized secretion of components to the BM.

A clathrin adaptor complex, AP1B is another essential component involved in polarized transport (Bonifacino, 2014; Deborde et al., 2008; Rodriguez-Boulan and Macara, 2014). Several studies have shown that AP1B is involved in basolateral transport (Fölsch, 2015; Nakatsu et al., 2014), but others indicated that it is also required to sort various anterograde apical cargoes in mice and C. elegans (Gillard et al., 2015; Shafaq-Zadah et al., 2012; Zhu et al., 2015). Clathrin participates in basolateral transport in mammalian epithelial cells (Deborde et al., 2008). In Drosophila, loss of the AP1γ subunit of photoreceptors, which is a subunit of the sole fly AP1 complex, results in the mis-transport of Na+K+ATPase to the stalk membrane (Satoh et al., 2013).

Na+K+ATPase significantly decreases on the basolateral membrane and mislocalizes to the stalk membrane in Rab10-deficient photoreceptors

GDP-locked Rab10 (Rab10T23N) functions as a dominant-negative protein in the Drosophila ovary (Lerner et al., 2013); however, the transgenic flies with UASp-Rab10T23N used in that study are not suitable for retinal expression. Therefore, we generated our own transgenic flies using UAST-Rab10T23N. We crossed them to longGMR-Gal4 flies to drive expression starting at the morphogenetic furrow in the imaginal eye disk in the third instar larvae, then investigated the phenotypes using the section at the depth where a couple of photoreceptor nucleus within ommatidia were observed. We found that the overexpression of Rab10T23N induced a significant reduction in the amount of the α-subunit of Na+K+ATPase (Na+K+ATPase-α) on the basolateral membrane and mislocalized Na+K+ATPase-α to the stalk membrane (Fig. 1A,C). The lengths of the sections of the stalk and basolateral membranes in Rab10T23N-overexpressing photoreceptors were longer (1.79×) or shorter (0.80×) than those of the control photoreceptors, respectively (Fig. 1G). In contrast, the rhabdomere proteins Rh1 (NINAE), TRP (Montell and Rubin, 1989) and Eys secreted from the stalk membrane (Husain et al., 2006) localized at normal levels in the rhabdomeres and the inter-rhabdomere space (IRS) (Fig. 1B,D). In Rab10T23N-overexpressing ommatidia, AJs visualized with phalloidin (Fig. 1A,C, arrows) normally separate the stalk and the basolateral membrane as in the wild type; however, they were positioned on the outer circumferences of ommatidia rather than in proximity to the center of the ommatidia, which is the regular position in the wild type (Fig. 1A,C). The IRS in Rab10T23N-overexpressing ommatidia had a more pronounced star shape than that in the control ommatidia, as a result of the shrinkage of the basolateral membrane and expansion of the stalk membrane (Fig. 1B,D). These results indicate that basolateral transport was severely inhibited and that some of the proteins destined for the basolateral membrane were mis-transported to the stalk membrane in Rab10T23N-overexpressing photoreceptors. Our study also investigated whether the localization of Na+K+ATPase-α can similarly be affected by the expression of wild-type or the constitutively active Q68L mutant of Rab10. Unlike overexpression of Rab10T23N, overexpression of Rab10Q68L did not alter the basolateral localization of Na+K+ATPase-α; however, the overexpression of wild-type Rab10 resulted in a weaker but similar mislocalization of Na+K+ATPase-α on the stalk membrane when compared with the results observed with the expression of Rab10DN (Fig. S1A,B).

Fig. 1.

Basolateral transport is inhibited in Rab10-deficient photoreceptors. (A-D) Immunostaining of retinas from newly emerged longGMR-Gal4/+ (A,B) and longGMR-Gal4/UAS-Rab10T23N (C,D) flies using anti-Na+K+ATPase-α (blue), phalloidin (red), anti-Crb (green) (A,C), anti-Rh1 (blue), anti-TRP (red) and anti-Eys (green) (B,D) antibodies. Arrows indicate adherence junctions. (E,F) Immunostaining of late pupal retinas from coinFLP-Gal4/UAS-Rab10RNAi, UAS-GFP flies using anti-Na+K+ATPase-α (blue), anti-Crb (red) (E), anti-TRP (blue) and anti-Eys (red) antibodies (F). GFP shows cells with Rab10RNAi. (G,H) Lengths of stalk and basolateral membranes of photoreceptors in longGMR-Gal4/+ (wild type) and longGMR-Gal4/UAS-Rab10T23N fly retinas (Rab10T23N) (G), and in wild-type and Rab10RNAi photoreceptors in coinFLP-Gal4/UAS-Rab10RNAi, UAS-GFP flies (H). Data are mean±s.d. with individual data points indicated. (I,J) Immunoblotting of retinas from longGMR-Gal4/+ (wild type) and longGMR-Gal4/UAS-Rab10T23N flies (Rab10DN) using anti-α-tubulin, anti-Na+K+ATPase-α and anti-Na+K+ATPase-β antibodies (I). Plot of the relative amounts of Na+K+ATPase-α and Na+K+ATPase-β proteins in Rab10DN-expressing retinas compared with the wild type, normalized to the amount of α-tubulin protein (J). Data are mean±s.d. with individual data points indicated. ***P<0.001, *P<0.05 (Student's t-test). Scale bar: 5 μm.

Fig. 1.

Basolateral transport is inhibited in Rab10-deficient photoreceptors. (A-D) Immunostaining of retinas from newly emerged longGMR-Gal4/+ (A,B) and longGMR-Gal4/UAS-Rab10T23N (C,D) flies using anti-Na+K+ATPase-α (blue), phalloidin (red), anti-Crb (green) (A,C), anti-Rh1 (blue), anti-TRP (red) and anti-Eys (green) (B,D) antibodies. Arrows indicate adherence junctions. (E,F) Immunostaining of late pupal retinas from coinFLP-Gal4/UAS-Rab10RNAi, UAS-GFP flies using anti-Na+K+ATPase-α (blue), anti-Crb (red) (E), anti-TRP (blue) and anti-Eys (red) antibodies (F). GFP shows cells with Rab10RNAi. (G,H) Lengths of stalk and basolateral membranes of photoreceptors in longGMR-Gal4/+ (wild type) and longGMR-Gal4/UAS-Rab10T23N fly retinas (Rab10T23N) (G), and in wild-type and Rab10RNAi photoreceptors in coinFLP-Gal4/UAS-Rab10RNAi, UAS-GFP flies (H). Data are mean±s.d. with individual data points indicated. (I,J) Immunoblotting of retinas from longGMR-Gal4/+ (wild type) and longGMR-Gal4/UAS-Rab10T23N flies (Rab10DN) using anti-α-tubulin, anti-Na+K+ATPase-α and anti-Na+K+ATPase-β antibodies (I). Plot of the relative amounts of Na+K+ATPase-α and Na+K+ATPase-β proteins in Rab10DN-expressing retinas compared with the wild type, normalized to the amount of α-tubulin protein (J). Data are mean±s.d. with individual data points indicated. ***P<0.001, *P<0.05 (Student's t-test). Scale bar: 5 μm.

To determine whether Rab10 is involved in basolateral transport in photoreceptors, we tried to generate null mutant alleles of Rab10, using the CRISPR/Cas9 system and end-out method. However, we failed to maintain the candidate alleles, probably because Rab10 localizes too close to the edge of X chromosome, where the FM7 balancer cannot suppress recombination. Thus, we expressed a Rab10RNAi construct using a coinFLP-Gal4 system with ey-FLP (Bosch et al., 2015). We found clear mislocalization of Na+K+ATPase-α to the stalk membrane in Rab10RNAi-expressing photoreceptors. However, shrinkage of the basolateral membrane was not prominent (Fig. 1E,H). TRP and Eys localized normally in the rhabdomeres and IRS (Fig. 1F). These results strongly suggest that Rab10 is essential for the basolateral transport of Na+K+ATPase-α in fruit fly photoreceptors.

In addition to the mis-transport of Na+K+ATPase to the stalk membrane, immunoblotting indicated that the protein content of the α- and β-subunits of Na+K+ATPase in Rab10T23N-expressing retinas were 0.48 and 0.54 times lower than those in the control retinas, respectively (see Materials and Methods, Fig. 1I,J). To assess whether this reduction of Na+K+ATPase is caused by lysosomal degradation, we used a null allele of light, lt1 or the Rab7 dominant-negative protein Rab7T22N to inhibit lysosomal degradation, because Light, a HOPS subunit and Rab7 are involved in the tethering and fusion of late endosomes with lysosome (Balderhaar and Ungermann, 2013; Spang, 2016). In lt1 homozygous photoreceptors or the photoreceptors expressing Rab7T22N alone by longGMR-Gal4, Na+K+ATPase is detected only in the basolateral membrane (Fig. S1C,E), indicating that Na+K+ATPase is not degraded by lysosomes in the wild-type photoreceptor. Meanwhile, considerable amounts of Na+K+ATPase accumulated in the cytoplasmic organelle with Rh1, presumably MVBs, in lt1 homozygous photoreceptors expressing Rab10T23N or the photoreceptors expressing both Rab10T23N and Rab7T22N together (Fig. S1D,F). Thus, Na+K+ATPase is not only mis-transported to the stalk membrane, but some part of it is degraded in the lysosomes.

Upon Rab10T23N expression, the basolateral membrane shrinks and the stalk membrane elongates

Using transmission electron microscopy, the lateral membranes of the neighboring photoreceptors in wild-type ommatidia were found to be aligned in parallel, and the AJs were situated on the innermost points of the parallel lateral membranes (Fig. 2A and Fig. S2A). However, in Rab10T23N-expressing retinas, stalk membranes, instead of basolateral membranes, of the neighboring photoreceptors aligned in parallel and AJs were located on the circumference of ommatidia and were often accompanied by extra segments of AJs (Fig. 2B, inset; Fig. S2B). The differences between the wild-type- and Rab10T23N-expressing retinas in terms of the appearance of their membranes and the positions of the AJs likely resulted from the changes in the lengths of their membrane sections. The stalk and basolateral membrane lengths of longGMR-Gal4/+ photoreceptors were 4.65 μm±0.43 μm and 13.1 μm±0.97 μm, respectively, and those of the stalk and basolateral membranes of longGMR-Gal4/Rab10T23N photoreceptors were 7.28 μm±0.54 μm and 8.44 μm±0.41 μm, respectively (Fig. 2G). Thus, in Rab10T23N-expressing photoreceptors, the cross-sectional lengths of the stalk and basolateral membranes were 1.57 times longer and 0.65 times shorter than those of the control photoreceptors, respectively.

Fig. 2.

Shrinkage of the basolateral membrane and elongation of the stalk membrane in Rab10-deficient photoreceptors. (A,B) Electron micrographs of longGMR-Gal4/+ (A) and longGMR-Gal4/UAS-Rab10T23N (B) ommatidia from newly emerged flies. Pink and green lines indicate the stalk and basolateral membranes, respectively. Red indicates the AJs. (C,D) Golgi stacks in longGMR-Gal4/+ (C) and longGMR-Gal4/UAS-Rab10T23N (D) photoreceptors. (E,F) Typical longGMR-Gal4/+ (E) and longGMR-Gal4/UAS-Rab10T23N (F) photoreceptors. Arrows indicate ER. (G) Lengths of the stalk and basolateral membranes of the photoreceptors in longGMR-Gal4/+ (wild-type) and longGMR-Gal4/UAS-Rab10T23N retinas (Rab10T23N). Data are mean±s.d. with individual data points indicated. (H) Lengths of the ER membranes of the photoreceptors in longGMR-Gal4/+ (wild-type) and longGMR-Gal4/UAS-Rab10T23N retinas (Rab10T23N). Data are mean±s.d. with individual data points indicated. (I,J) Plots of the average number of Golgi stacks (I) and the number of Golgi stacks with three cisternae (J) from three wild-type and three Rab10T23N-expressing retinas. (K) Number of cisternae in Golgi stacks in wild-type and Rab10T23N-expressing retinas. Twenty-six and 18 Golgi stacks for the wild-type and Rab10T23N-expressing retinas, respectively, were investigated. ***P<0.001, **P<0.01, *P<0.05 (Student's t-test). Scale bars: 2 μm in A,B; 300 nm in C,D; 1 μm in E,F.

Fig. 2.

Shrinkage of the basolateral membrane and elongation of the stalk membrane in Rab10-deficient photoreceptors. (A,B) Electron micrographs of longGMR-Gal4/+ (A) and longGMR-Gal4/UAS-Rab10T23N (B) ommatidia from newly emerged flies. Pink and green lines indicate the stalk and basolateral membranes, respectively. Red indicates the AJs. (C,D) Golgi stacks in longGMR-Gal4/+ (C) and longGMR-Gal4/UAS-Rab10T23N (D) photoreceptors. (E,F) Typical longGMR-Gal4/+ (E) and longGMR-Gal4/UAS-Rab10T23N (F) photoreceptors. Arrows indicate ER. (G) Lengths of the stalk and basolateral membranes of the photoreceptors in longGMR-Gal4/+ (wild-type) and longGMR-Gal4/UAS-Rab10T23N retinas (Rab10T23N). Data are mean±s.d. with individual data points indicated. (H) Lengths of the ER membranes of the photoreceptors in longGMR-Gal4/+ (wild-type) and longGMR-Gal4/UAS-Rab10T23N retinas (Rab10T23N). Data are mean±s.d. with individual data points indicated. (I,J) Plots of the average number of Golgi stacks (I) and the number of Golgi stacks with three cisternae (J) from three wild-type and three Rab10T23N-expressing retinas. (K) Number of cisternae in Golgi stacks in wild-type and Rab10T23N-expressing retinas. Twenty-six and 18 Golgi stacks for the wild-type and Rab10T23N-expressing retinas, respectively, were investigated. ***P<0.001, **P<0.01, *P<0.05 (Student's t-test). Scale bars: 2 μm in A,B; 300 nm in C,D; 1 μm in E,F.

Golgi cisternae are de-stacked by dominant-negative Rab10 photoreceptors

Increased numbers of the endoplasmic reticulum (ER) membrane were found in the photoreceptors of longGMR-Gal4/Rab10T23N heterozygous flies in comparison with those of the control flies (longGMR-Gal4/+) (Fig. 2E,F, arrows). These ER membranes look clustered or stacked together. The lengths of the ER membrane cross-sections in the Rab10T23N-expressing photoreceptors were 1.39× longer than those of the control photoreceptors (Fig. 2H). This result is consistent with a previous report that found that Rab10 regulated ER dynamics and morphology (English and Voeltz, 2013; Lv et al., 2015).

Furthermore, we encountered abnormalities in the Golgi stacks. In Drosophila retina with wild-type photoreceptors, each Golgi stack consisted of two or three tightly stacked cisternae and some vesicles, whereas in Rab10T23N-expressing photoreceptors, the Golgi cisternae were not stacked correctly (Fig. 2C,D). We compared the numbers of Golgi stacks and their cisternae in Rab10T23N-expressing and control photoreceptors. The total number of Golgi stacks was approximately the same in both cases (Fig. 2I); however, the numbers of Golgi cisternae and Golgi stacks with more than three cisternae were significantly reduced in Rab10T23N-expressing photoreceptors (Fig. 2J,K). We investigated whether Rab10T23N expression influences Golgi stack polarity by immunostaining for the cis-Golgi marker GM130, the trans-Golgi and TGN marker Rab6, and the TGN marker Golgin245 (Fig. S3A,B). These markers remained polarized in their distribution, suggesting that Golgi stack polarity was not affected in longGMR-Gal4/Rab10T23N photoreceptors. Nevertheless, the Golgi cisternal organization was impaired in Rab10T23N-expressing photoreceptors.

Basolateral transport is impaired in photoreceptors deficient in Rab10-GEF Crag or the Rab10 effector Ehbp1

Because Rab10 is essential for basolateral Na+K+ATPase transport, we investigated whether the lack of Rab10-GEF Crag has an impact on basolateral transport by using mosaic retinas containing both the wild-type marked by RFP and the Crag-null mutations CragCJ101 and CragGG43. To this end, we used the FLP/FRT method with indirect immunohistochemistry (Satoh et al., 2013; Xu and Rubin, 1993). Similar to the Rab10-deficient photoreceptors, Na+K+ATPase-α mislocalized in the stalk membranes of the homozygous CragCJ101 and CragGG43 photoreceptors (Fig. 3A). The stalk membrane of CragCJ101 or CragGG43 photoreceptors was found to be 1.34 or 1.46 times longer in comparison with the corresponding wild-type photoreceptors of the same CragCJ101 or CragGG43 mosaic retinas (Fig. 3B). The basolateral membrane of the homozygous CragCJ101 or CragGG43 photoreceptors was 0.70 or 0.67 times longer than those of the corresponding wild-type basolateral membranes of the same CragCJ101 or CragGG43 mosaic retinas (Fig. 3B). The AJs were located on the outer circumferences of Crag-deficient ommatidia (see DE-Cad staining in Fig. 3A). The aforementioned defects were reversed by FLAG-tagged wild-type Crag expression (Fig. 3A, right). However, the expression of the rhabdomere proteins Rh1 and TRP, and secretion of Eys were normal in homozygous CragCJ101 and CragGG43 ommatidia. Therefore, Crag is also involved in basolateral transport, presumably through Rab10 activation.

Fig. 3.

Basolateral transport is inhibited in Crag- and Ehbp1-deficient photoreceptors. (A) Immunostaining of CragCJ101 and CragGG43 mosaic retinas, and CragGG43 mosaic retina expressing FLAG::Crag from newly emerged flies using antibodies against the indicated proteins. RFP (red) indicates wild-type cells. (B) Lengths of the stalk and basolateral membranes of the wild-type and Crag-deficient photoreceptors in CragCJ101, CragGG43 and CragGG43 rescue (expressing FLAG::Crag) mosaic retinas. Data are mean±s.d. with individual data points indicated. (C) Immunostaining of Ehbp1A28 and Ehbp1O4 mosaic retinas, and Ehbp1A28 mosaic retina expressing FLAG::Ehbp1 from newly emerged flies using antibodies against the indicated proteins. RFP (red) indicates wild-type cells. (D) Lengths of stalk and basolateral membranes of the wild-type and Ehbp1-deficient photoreceptors in Ehbp1A28, Ehbp1O4 and Ehbp1A28 rescue (expressing FLAG::Ehbp1) mosaic retinas. Data are mean±s.d. with individual data points indicated. ***P<0.001 (Student's t-test). Scale bar: 5 μm.

Fig. 3.

Basolateral transport is inhibited in Crag- and Ehbp1-deficient photoreceptors. (A) Immunostaining of CragCJ101 and CragGG43 mosaic retinas, and CragGG43 mosaic retina expressing FLAG::Crag from newly emerged flies using antibodies against the indicated proteins. RFP (red) indicates wild-type cells. (B) Lengths of the stalk and basolateral membranes of the wild-type and Crag-deficient photoreceptors in CragCJ101, CragGG43 and CragGG43 rescue (expressing FLAG::Crag) mosaic retinas. Data are mean±s.d. with individual data points indicated. (C) Immunostaining of Ehbp1A28 and Ehbp1O4 mosaic retinas, and Ehbp1A28 mosaic retina expressing FLAG::Ehbp1 from newly emerged flies using antibodies against the indicated proteins. RFP (red) indicates wild-type cells. (D) Lengths of stalk and basolateral membranes of the wild-type and Ehbp1-deficient photoreceptors in Ehbp1A28, Ehbp1O4 and Ehbp1A28 rescue (expressing FLAG::Ehbp1) mosaic retinas. Data are mean±s.d. with individual data points indicated. ***P<0.001 (Student's t-test). Scale bar: 5 μm.

Using the FLP/FRT method, we investigated the phenotypes of two null mutations of the Ehbp1 gene, a Rab10 effector. We found that Na+K+ATPase-α mislocalized to the stalk membranes of the homozygous Ehbp1A28 and Ehbp1O4 photoreceptors (Fig. 3C). The cross-sections of the stalk and basolateral membranes of homozygous Ehbp1A28 and Ehbp1O4 photoreceptors were longer (1.43 and 1.31 times) and shorter (0.72 and 0.69 times), respectively, in comparison with the corresponding wild type in Ehbp1A28 and Ehbp1O4 mosaic retinas (Fig. 3D). The AJs were situated on the outer circumferences of the homozygous Ehbp1A28 and Ehbp1O4 ommatidia (Fig. 3C). These defects were reversed by FLAG-tagged wild-type Ehbp1 expression (Fig. 3C, right). The expression of the rhabdomere proteins Rh1 and TRP, and secretion of Eys were normal in homozygous Ehbp1A28 and Ehbp1O4 ommatidia. These phenotypes resembled those of Crag-null homozygous or Rab10T23N-expressing ommatidia (Figs 1 and 3A,B). Therefore, Ehbp1 participates in basolateral transport along with Crag and Rab10.

The basolateral membrane shrinks and the stalk membrane expands in Crag- or Ehbp1-deficient photoreceptors

We observed thin sections of CragCJ101 and CragGG43 mosaic retinas under a transmission electron microscope. Homozygous CragCJ101 and CragGG43 photoreceptors were identified by their lack of ommochrome-containing pigment granules, because the wild-type chromosome was marked by a white gene capable of pigment synthesis. We also observed the partially rescued homozygous Ehbp1A28 whole-eye clones. As for Rab10T23N-expressing ommatidia, there was no parallel alignment of the neighboring photoreceptor lateral membranes in homozygous CragCJ101, CragGG43 and Ehbp1A28 photoreceptors (Fig. 4B,C; Fig. S2D-F). The AJs located on the circumferences of homozygous CragCJ101, CragGG43 and Ehbp1A28 ommatidia were often accompanied by extra segments of AJ (Fig. 4B,C, inset; Fig. S2D-F). The cross-sections of the stalk membranes of homozygous CragCJ101, CragGG43 and Ehbp1A28 photoreceptors were 1.70, 1.79 and 1.76 times longer, respectively, than those of the wild-type photoreceptors (Fig. 4G). The basolateral membranes of CragCJ101, CragGG43 and Ehbp1A28 photoreceptors were 0.57, 0.59 and 0.57 times shorter, respectively, in comparison with the length of those of the wild-type photoreceptors (Fig. 4G). Therefore, both Crag and Ehbp1 are required for the membrane transport to the basolateral membrane. In addition, the ER membranes of homozygous CragGG43 and Ehbp1A28 photoreceptors were 1.79 and 1.71 times longer than those of the wild-type photoreceptors, respectively (Fig. 4H and Fig. S4). Therefore, both Crag and Ehbp1 participate in ER membrane homeostasis, similar to Rab10.

Fig. 4.

Shrinkage of the basolateral membrane and elongation of the stalk membrane in Crag- and Ehbp1-deficient photoreceptors. (A-C) Electron micrographs of the wild-type, CragGG43 and Ehbp1A28 ommatidia from newly emerged flies. The wild-type and CragGG43 ommatidia were derived from a CragGG43 mosaic retina. The CragGG43 ommatidium is characterized by its lack of pigment granules, as we used the w+ gene as a wild-type cell marker. The Ehbp1A28 ommatidium was obtained from a whole-eye homozygous clone of Ehbp1A28. Pink and green lines indicate the stalk and basolateral membranes, respectively. Red indicates the AJs. (D-F) Electron micrographs of the Golgi stacks in the wild-type (D), CragGG43 (E) and Ehbp1A28 (F) photoreceptors. (G) Lengths of stalk and basolateral membranes of the wild-type, CragCJ101, CragGG43 and Ehbp1A28 photoreceptors. Data are mean±s.d. with individual data points indicated. (H) Lengths of the ER membranes of the wild-type, CragGG43 and Ehbp1A28 photoreceptors. Data are mean±s.d. with individual data points indicated. (I,J) Average number of Golgi stacks (I) and the number of cisternae in the Golgi stacks (J) in the wild-type, CragGG43 and Ehbp1A28 photoreceptors. Twenty, 18 and 19 Golgi stacks for wild-type, CragGG43 and Ehbp1A28 photoreceptors, respectively, are investigated (J). ***P<0.001 (Student's t-test). Scale bars: 2 μm in A-C; 300 nm in D-F.

Fig. 4.

Shrinkage of the basolateral membrane and elongation of the stalk membrane in Crag- and Ehbp1-deficient photoreceptors. (A-C) Electron micrographs of the wild-type, CragGG43 and Ehbp1A28 ommatidia from newly emerged flies. The wild-type and CragGG43 ommatidia were derived from a CragGG43 mosaic retina. The CragGG43 ommatidium is characterized by its lack of pigment granules, as we used the w+ gene as a wild-type cell marker. The Ehbp1A28 ommatidium was obtained from a whole-eye homozygous clone of Ehbp1A28. Pink and green lines indicate the stalk and basolateral membranes, respectively. Red indicates the AJs. (D-F) Electron micrographs of the Golgi stacks in the wild-type (D), CragGG43 (E) and Ehbp1A28 (F) photoreceptors. (G) Lengths of stalk and basolateral membranes of the wild-type, CragCJ101, CragGG43 and Ehbp1A28 photoreceptors. Data are mean±s.d. with individual data points indicated. (H) Lengths of the ER membranes of the wild-type, CragGG43 and Ehbp1A28 photoreceptors. Data are mean±s.d. with individual data points indicated. (I,J) Average number of Golgi stacks (I) and the number of cisternae in the Golgi stacks (J) in the wild-type, CragGG43 and Ehbp1A28 photoreceptors. Twenty, 18 and 19 Golgi stacks for wild-type, CragGG43 and Ehbp1A28 photoreceptors, respectively, are investigated (J). ***P<0.001 (Student's t-test). Scale bars: 2 μm in A-C; 300 nm in D-F.

We investigated Golgi stack abnormalities in Crag- and Ehbp1-deficient photoreceptors. Similar to Rab10T23N-expressing photoreceptors, the number of Golgi stacks did not significantly differ from those in the wild type (Fig. 4I). Nevertheless, the cisternae comprising the Golgi stacks in homozygous CragGG43 and Ehbp1A28 photoreceptors were poorly stacked (Fig. 4D-F,J). However, the cis-Golgi marker GM130 and the trans-Golgi/TGN marker Rab6 remained in polarized distribution (Fig. S3C,E). The medial-Golgi marker MPPE and TGN marker Golgin245 also remained in polarized distribution (Fig. S3D,F). Therefore, the polarity of the Golgi stacks is maintained, but Golgi cisternal organization is impaired, in both CragGG43 and Ehbp1A28 photoreceptors.

The basolateral protein Na+K+ATPase mislocalizes to the stalk membrane in AP1 and clathrin knockdown photoreceptors

We have previously reported that the AP1 adaptor complex is involved in the polarized transport of basolateral transmembrane proteins in fly photoreceptors (Satoh et al., 2013). To compare the phenotypes of AP1 deficiency to those of Crag, Rab10 or Ehbp1 deficiency, we revisited the effect of AP1 deficiency using the hypomorphic alleles AP1γe04546 and AP1μEP1112. Few ommatidia in the AP1-deficient clones showed a defective polarity, such as an altered ommatidial organization and an unusual positioning of the rhabdomeres (Fig. S5G). As it is difficult to judge whether the localizations of apical and basolateral proteins are affected in these ommatidia, our study focused on the ommatidia with normal ommatidial organization. In these ommatidia, Na+K+ATPase-α mislocalized to the stalk membrane in both AP1γe04546 and AP1μEP1112 homozygous photoreceptors (Fig. 5A). The stalk membranes in homozygous AP1γe04546 or AP1μEP1112 photoreceptors were 1.41 and 1.17 times longer than those of the wild-type photoreceptors in mosaic retinas, respectively (Fig. 5D,E). On the other hand, the basolateral membranes in homozygous AP1γe04546 or AP1μEP1112 photoreceptors were slightly shorter (0.82 or 0.95 times) than those of the wild-type photoreceptors in AP1γe04546 or AP1μEP1112 mosaic retinas (Fig. 5D,E). The rhabdomeric proteins Rh1 and TRP, and secreted protein Eys show the normal localization in homozygous AP1γe04546 or AP1μEP1112 ommatidia (Fig. 5A). Therefore, the AP1 complex is specifically involved in basolateral transport but not in the apical transport in fly photoreceptors.

Fig. 5.

Basolateral Na+K+ATPase mislocalizes to the stalk membrane in photoreceptors with AP1 and clathrin knockdown. (A) Immunostaining of AP1γe04546 and AP1μEP1112 mosaic retinas from newly emerged flies using antibodies for the indicated proteins. RFP (red) indicates wild-type cells. (B) Immunostaining of retinas from coinFLP-Gal4/UAS-ChcDN, UAS-GFP flies using anti-Na+K+ATPase-α (red) and anti-Rh1 antibodies (blue). GFP and asterisks show cells expressing ChcDN. (C) Immunostaining of retinas from newly emerged coinFLP-Gal4/UAS-ChcRNAi, UAS-GFP flies using anti-Na+K+ATPase-α (red) and anti-Rh1 antibodies (blue). GFP and asterisks show the cells with ChcRNAi. (D,E) Lengths of the stalk and basolateral membranes of the wild-type and homozygous AP1γe04546 (D) or AP1μEP1112 (E) photoreceptors within the same AP1γe04546 or AP1μEP1112 mosaic retinas. Data are mean±s.d. with individual data points indicated. ***P<0.001, **P<0.0 (Student's t-test). Scale bar: 5 μm.

Fig. 5.

Basolateral Na+K+ATPase mislocalizes to the stalk membrane in photoreceptors with AP1 and clathrin knockdown. (A) Immunostaining of AP1γe04546 and AP1μEP1112 mosaic retinas from newly emerged flies using antibodies for the indicated proteins. RFP (red) indicates wild-type cells. (B) Immunostaining of retinas from coinFLP-Gal4/UAS-ChcDN, UAS-GFP flies using anti-Na+K+ATPase-α (red) and anti-Rh1 antibodies (blue). GFP and asterisks show cells expressing ChcDN. (C) Immunostaining of retinas from newly emerged coinFLP-Gal4/UAS-ChcRNAi, UAS-GFP flies using anti-Na+K+ATPase-α (red) and anti-Rh1 antibodies (blue). GFP and asterisks show the cells with ChcRNAi. (D,E) Lengths of the stalk and basolateral membranes of the wild-type and homozygous AP1γe04546 (D) or AP1μEP1112 (E) photoreceptors within the same AP1γe04546 or AP1μEP1112 mosaic retinas. Data are mean±s.d. with individual data points indicated. ***P<0.001, **P<0.0 (Student's t-test). Scale bar: 5 μm.

It has been reported that clathrin is involved in basolateral transport in MDCK cells (Deborde et al., 2008). We investigated the effect of a lack of clathrin heavy chain (Chc). Both Chc RNAi and Chc dominant-negative overexpression caused mislocalization of Na+K+ATPase-α to the stalk membrane (Fig. 5B,C). Therefore, clathrin also participates in basolateral transport in fly photoreceptors. However, many cells in Chc-deficient clones show the severe defects on apico-basal cell polarity, which prevented the measurement of the length of the membranes; however, the stalk-basolateral proportion phenotypes caused by Chc or AP1 deficiency were apparently milder than those caused by Rab10, Crag or Ehbp1 deficiency (Fig. 5). As we found the polarity defect in Chc- or AP1-deficient clones, we investigated whether Rab10-, Crag- or Ehbp1-deficient clones also had polarity defects (Fig. S5). Rab10T23N or Rab10RNAi expression did not yield the polarity defects, but Crag- or Ehbp1-deficient clones had mild polarity defects, although they were significantly weaker than those in AP1 or Chc deficiency (Fig. S5J).

Crag, Rab10 and Ehbp1 localize to the Golgi stacks

Mislocalization of basolateral proteins to the stalk membrane of the Rab10-, Crag- or Ehbp1-deficient clones suggests that Rab10, Crag and Ehbp1 have a function closely associated with the sorting of membrane proteins. As the sorting takes place at the trans-side of the Golgi or TGN, we investigated whether they localize on the Golgi stacks. In Drosophila photoreceptors, Golgi stacks are well developed at the early pupal stages, and their diameters might be >1 µm (Satoh et al., 2005). We explored the association of Rab10 and Crag with Golgi cisternae by generating transgenic flies with UAST-Myc::Rab10 and UAST-FLAG::Crag constructs, and expressing them with GMR-Gal4 or elav-Gal4, which induces high expression in young pupal retinas. Myc::Rab10 colocalized with the cis-Golgi marker GM130 and the trans-Golgi/TGN marker Rab6 (Fig. 6A). Myc::Rab10 associated with Golgi stacks over a wide range from the cis-Golgi side to the trans-Golgi side (Fig. 6B). FLAG::Crag colocalized with Rab6 but not with GM130, indicating that FLAG::Crag associated only with the trans-side of the Golgi stacks (Fig. 6C,D). These results suggested that Rab10 is recruited to early Golgi compartments but is activated only in late Golgi compartments by Rab10-GEF Crag.

Fig. 6.

Crag, Ehbp1, Chc and AP1 colocalize on the trans-side of the Golgi stacks. (A,C,E,G,I,K,M,O) Triple staining of Golgi stacks of young pupal retinas or laminas using indicated antibodies or fluorescent proteins. Arrows indicate the relative position of the staining. (B,D,F,H,J,L,N,P) Plots show signal intensities from image on the left. Signal intensity was measured along the arrow (representing 1.5 μm) in inset; graph shows the overlap between channels (A,B) Wild-type eye expressing Myc::Rab10 driven by GMR-Gal4. Anti-Myc (red), anti-GM130 (green) and anti-Rab6 antibodies (blue). (C,D) Wild-type eye expressing FLAG::Crag driven by elav-Gal4. Anti-FLAG (red), anti-GM130 (green) and anti-Rab6 antibodies (blue). (E,F) Wild-type eye expressing GalT::CFP driven by elav-Gal4. GalT::CFP (green), anti-Ehbp1 (red) and anti-GM130 antibodies (blue). (G,H) Anti-Rab11 (red), anti-Ehbp1 (green) and anti-Rab6 (blue). (I,J) Wild-type eye expressing FLAG::Crag driven by elav-Gal4. Anti-FLAG (red), anti-Ehbp1 (green) and anti-p120 antibodies (blue). (K,L) Wild-type eye expressing FLAG::Crag driven by elav-Gal4. Anti-FLAG (red), anti-Chc (green) and anti-Rab6 antibodies (blue). (M,N) Anti-Chc (red), anti-Ehbp1 (green) and anti-p120 (blue). (O,P) Wild-type eye expressing AP1::VFP driven by elav-Gal4. AP1::VFP (green), anti-Chc (red) and anti-Rab11 antibodies (blue). Scale bar: 1 μm.

Fig. 6.

Crag, Ehbp1, Chc and AP1 colocalize on the trans-side of the Golgi stacks. (A,C,E,G,I,K,M,O) Triple staining of Golgi stacks of young pupal retinas or laminas using indicated antibodies or fluorescent proteins. Arrows indicate the relative position of the staining. (B,D,F,H,J,L,N,P) Plots show signal intensities from image on the left. Signal intensity was measured along the arrow (representing 1.5 μm) in inset; graph shows the overlap between channels (A,B) Wild-type eye expressing Myc::Rab10 driven by GMR-Gal4. Anti-Myc (red), anti-GM130 (green) and anti-Rab6 antibodies (blue). (C,D) Wild-type eye expressing FLAG::Crag driven by elav-Gal4. Anti-FLAG (red), anti-GM130 (green) and anti-Rab6 antibodies (blue). (E,F) Wild-type eye expressing GalT::CFP driven by elav-Gal4. GalT::CFP (green), anti-Ehbp1 (red) and anti-GM130 antibodies (blue). (G,H) Anti-Rab11 (red), anti-Ehbp1 (green) and anti-Rab6 (blue). (I,J) Wild-type eye expressing FLAG::Crag driven by elav-Gal4. Anti-FLAG (red), anti-Ehbp1 (green) and anti-p120 antibodies (blue). (K,L) Wild-type eye expressing FLAG::Crag driven by elav-Gal4. Anti-FLAG (red), anti-Chc (green) and anti-Rab6 antibodies (blue). (M,N) Anti-Chc (red), anti-Ehbp1 (green) and anti-p120 (blue). (O,P) Wild-type eye expressing AP1::VFP driven by elav-Gal4. AP1::VFP (green), anti-Chc (red) and anti-Rab11 antibodies (blue). Scale bar: 1 μm.

We also investigated the localization of Ehbp1 using an anti-Ehbp1 antibody generously donated by Dr Hugo Bellen (Baylor College of Medicine, Houston, TX; Giagtzoglou et al., 2012). Ehbp1 localized on the trans-side of the Golgi stacks beyond GM130 and the trans-Golgi marker GalT::CFP (Fig. 6E,F). The recycling endosome (RE) marker Rab11 localized on the far trans-side of the Golgi stacks, as previously shown (Satoh et al., 2005). Ehbp1 specifically localized between Rab6 and Rab11 (Fig. 6G,H). The closer localization of FLAG::Crag than Ehbp1 to a medial-Golgi marker p120 (Fig. 6I,J) supported the theory that FLAG::Crag might activate Rab10, which, in turn, recruits Ehbp1 to the trans-side of the Golgi stacks.

Next, we studied the relationships of FLAG::Crag, Ehbp1, AP1 and Chc localization. FLAG::Crag and Chc were intensively colocalized (Fig. 6K,L). Chc also colocalized with Ehbp1 but extended to the cis area more than Ehbp1 (Fig. 6M,N). AP1μ::VFP extended more to the trans-side of the Golgi stacks than Chc but did not overlap with Rab11 (Fig. 6O,P). These results indicated that FLAG::Crag, Ehbp1, AP1 and Chc mostly colocalized to the trans-side of Golgi stacks but not to the Rab11-positive REs.

FLAG::Crag colocalizes intensively with Myc::Rab10 but not with Myc::Rab11

We examined the localization of Myc::Rab10 and FLAG::Crag in the photoreceptors of newly emerged flies. Myc::Rab10 and FLAG::Crag intensively colocalize, but localizations to the Golgi stacks were unclear in the late pupal photoreceptors (Fig. 7A). Because neither Myc::Rab10 nor FLAG::Crag colocalized with the ER marker Cnx (CNX99A), they were not on the ER membrane but in the cytoplasm (Fig. 7C, and data not shown). Crag has been reported to function as Rab11-GEF in adult photoreceptors (Xiong et al., 2012); therefore, we co-expressed Myc::Rab11 and FLAG::Crag to localize them. In contrast to Myc::Rab10, colocalization of Myc::Rab11 with FLAG::Crag was limited (Fig. 7B). FLAG::Crag showed a diffuse cytoplasmic pattern; however, Myc::Rab11 localized at the bases of the rhabdomeres, presumably in post-Golgi vesicles. Myc::Rab11 was also found in some cytoplasmic dots, which were probably Golgi-associated REs. This localization of Myc::Rab11 resembled that of endogenous Rab11 (Fig. 7B). Therefore, most of the Crag colocalized with Rab10 rather than Rab11 in late pupal retinas, although we cannot exclude the possibility that the small amount of Crag colocalizes with Rab11.

Fig. 7.

Ehbp1 localizes to the basolateral membrane. Immunostaining of late pupal retinas using the indicated antibodies. (A) Wild-type eye expressing FLAG::Crag and Myc::Rab10 driven by Rh1-Gal4. Anti-Myc (red) and anti-FLAG antibodies (green). (B) Wild-type eye expressing FLAG::Crag and Myc::Rab11 driven by Rh1-Gal4. Anti-Myc (red) and anti-FLAG antibodies (green). (C) Wild-type eye expressing Myc::Rab10 driven by Rh1-Gal4. Anti-Myc (red), anti-Cnx antibodies (green) and phalloidin (blue). (D) Ehbp1O4 mosaic retina. Anti-Ehbp1 antibody (green) and phalloidin (blue). RFP shows the wild-type photoreceptors. (E) Wild-type eye expressing FLAG::Ehbp1 driven by Rh1-Gal4. Anti-FLAG (red), anti-Rh1 (blue) antibodies and phalloidin (green). (F) CragCJ101 mosaic retina. Anti-Ehbp1 antibody (green) and phalloidin (blue). RFP shows the wild-type photoreceptors. Scale bar: 5 μm.

Fig. 7.

Ehbp1 localizes to the basolateral membrane. Immunostaining of late pupal retinas using the indicated antibodies. (A) Wild-type eye expressing FLAG::Crag and Myc::Rab10 driven by Rh1-Gal4. Anti-Myc (red) and anti-FLAG antibodies (green). (B) Wild-type eye expressing FLAG::Crag and Myc::Rab11 driven by Rh1-Gal4. Anti-Myc (red) and anti-FLAG antibodies (green). (C) Wild-type eye expressing Myc::Rab10 driven by Rh1-Gal4. Anti-Myc (red), anti-Cnx antibodies (green) and phalloidin (blue). (D) Ehbp1O4 mosaic retina. Anti-Ehbp1 antibody (green) and phalloidin (blue). RFP shows the wild-type photoreceptors. (E) Wild-type eye expressing FLAG::Ehbp1 driven by Rh1-Gal4. Anti-FLAG (red), anti-Rh1 (blue) antibodies and phalloidin (green). (F) CragCJ101 mosaic retina. Anti-Ehbp1 antibody (green) and phalloidin (blue). RFP shows the wild-type photoreceptors. Scale bar: 5 μm.

Staining of wild-type adult photoreceptors with the anti-Ehbp1 antibody enabled us to visualize the basolateral membrane and the retinal terminal web (RTW). In Ehbp1-deficient cells, Ehbp1 staining was absent from the basolateral membrane but present in the RTW. Ehbp1 thus localized primarily to the basolateral membrane (Fig. 7D), and RTW staining was not associated with Ehbp1. In wild-type adult photoreceptors, FLAG::Ehbp1 localized to the basolateral membrane, but in contrast to anti-Ehbp1 staining, it also localized to the rhabdomeres but not the RTW (Fig. 7E). Ehbp1 localization could be affected by its fusion with FLAG-tag. Because Ehbp1 has a binding site for actin filaments (Guilherme et al., 2004), and an actin filament is the core of the rhabdomere microvillus, FLAG-tag Ehbp1 fusion could cause this mislocalization. Ehbp1 normally localized to the basolateral membrane in Crag-deficient cells; therefore, it was not dependent on Crag or Rab10 (Fig. 7F).

In the present study, we found that photoreceptors with impaired Rab10 function had shrunken basolateral membranes and expanded stalk membranes. In addition, the quantity of basolateral Na+K+ATPase levels were reduced significantly in comparison with the wild type by Light and Rab7-dependent degradation, and Na+K+ATPase relocated to the stalk membrane in Rab10-deficient retinas. Corresponding to previous reports indicating that Crag is the Rab10-GEF and Ehbp1 is a Rab10 effector, deficiencies of Crag or Ehbp1 phenocopied the Rab10 impairment phenotype in fly photoreceptors. Crag colocalized with Rab10 and Ehbp1 to the trans-side of the Golgi stacks. Taken together, these results indicated that Crag, Rab10 and Ehbp1 collaborated in polarized basolateral transport in fly photoreceptors.

Crag, Rab10 and Ehbp1 are involved in BM transport to the basolateral membrane in Drosophila follicle cells (Denef et al., 2008; Isabella and Horne-Badovinac, 2016; Lerner et al., 2013). Nevertheless, Crag deficiencies do not affect the localization of the basolateral membrane proteins in the follicle cells (Denef et al., 2008). Based on these results, it has been proposed that Crag specifically regulates BM protein transport but is not required for the polarized transport of basolateral membrane proteins in follicle cells. This discrepancy might be explained by the difference in tissue types. It could also reflect relative differences in basolateral membrane protein transport requirements. During morphogenesis, fly photoreceptors increase in size to more than 10 times that of the original epithelial cells. Thus, photoreceptors can readily detect transport defects. On the other hand, the follicle cells grow a little but secrete substantial amounts of BM proteins. These characteristics may account for Crag deficiency targeting BM transport rather than generally affecting basolateral transport in follicle cells (Denef et al., 2008).

Xiong et al. reported that Crag is a Rab11-GEF that regulates light-dependent Rh1 transport in adult photoreceptors (Xiong et al., 2012). They also demonstrated that Rh1 localized normally in the rhabdomeres of the newly emerged flies and dark-reared adult flies, suggesting the existence of another Rab11-GEF in addition to Crag. Recently, we reported that Parcas (Pcs) is a major Rab11-GEF for Rh1 transport in late pupal stage of Drosophila (Otsuka et al., 2019). The lack of Pcs induces Rh1 accumulation in the cytoplasm, but Na+K+ATPase normally localizes on the basolateral membrane in late pupal photoreceptors. Xiong et al. indicated that Crag had a GEF activity not only for Rab11, but also for Rab10: interestingly, GEF activity for Rab10 was higher than that for Rab11. The mammalian Crag homolog DENND4B has previously been shown to have high Rab10-GEF activity (Yoshimura et al., 2010). In addition, we found that Crag substantially colocalized with Rab10 but not with Rab11 in pupal photoreceptors. Therefore, the main function of Crag would be the exchange of GDP to GTP on Rab10 in pupal retinas.

Ehbp1 has been identified as a Rab10 effector in C. elegans (Shi et al., 2010; Wang et al., 2016), and a fly study indicated that Ehbp1 overexpression increased the secretion of components to the BM, as did Crag and Rab10 (Isabella and Horne-Badovinac, 2016). Drosophila Ehbp1, which was originally identified as a Notch signaling factor, regulated asymmetric cell division in extrasensory organs (Giagtzoglou et al., 2012). Ehbp1 localizes to the actin-rich interface of asymmetrically dividing cells. It is situated in a part of the basolateral membrane where it can localize Delta by interacting with Sec15 and Rab11 (Giagtzoglou et al., 2012). In fly retinas, Ehbp1 regulates Scabrous (SCA) secretion during Notch-mediated lateral inhibition and photoreceptor development. AP1γ is associated with Scabrous in Ehbp1-deficient photoreceptors (Giagtzoglou et al., 2013). These two developmental studies did not directly indicate the function of Ehbp1 in polarized transport. However, the basolateral localization of Ehbp1 and the association of AP1γ with Scabrous in Ehbp1-deficient photoreceptors imply that Ehbp1 participates in basolateral transport.

AP1 and Chc are well established as being involved in vesicle budding rather than in the transport, tethering or fusion of post-Golgi vesicles. In agreement with this notion, deficiencies of AP1 and Chc in the present study resulted in Na+K+ATPase mislocalization to the stalk membrane, rather than accumulation in post-Golgi vesicles. This provides a notable contrast to the deficiencies of the Rab11, dRip11 and MyoV complex involving the transport of post-Golgi vesicles or of exocyst complex involving tethering to the rhabdomere base, in which post-Golgi vesicles bearing Rh1 massively accumulate in the cytoplasm. In the absence of Crag, Rab10 and Ehbp1, Na+K+ATPase mislocalized to the stalk membrane, rather than accumulating in post-Golgi vesicles. Together with the Golgi localization of Crag, Rab10 and Ehbp1, Crag, Rab10 and Ehbp1 may regulate basolateral transport during vesicle budding on the trans-side of the Golgi stack. It has been reported that Rab10 recruits one of the Rab10 effectors, Sec16A, which accelerates the formation of vesicles that ferry GLUT4 to the plasma membrane under insulin stimulation (Bruno et al., 2016). Rab10 may be involved in the recruitment of AP1 and Clathrin to the TGN and may facilitate the formation of Clathrin-coated vesicles in a similar way. It is crucial to further study the relationship between Crag, Rab10 and Ehbp1, and AP1 and Clathrin.

Here, we present models of polarized transport in wild-type and Crag-, Rab10- and Ehbp1-mutant photoreceptors (Fig. 8). Membrane proteins are synthesized on the ER and transported to cis-Golgi cisternae, which progressively mature to medial, trans and the TGN. Basolateral proteins are sorted by Crag, Rab10 and Ehbp1 into post-Golgi vesicles, and the remaining TGN matures to REs, where membrane proteins destined for the two apical membranes are sorted into separate post-Golgi vesicles and transported to the stalk and rhabdomeres. In Crag-, Rab10- and Ehbp1-mutant photoreceptors, however, basolateral membrane proteins are not entirely removed at the TGN. Some of them are transported to REs along with the apical cargoes, while some are degraded in the endo-lyosomal system. Basolateral membrane proteins transported to RE are then transported to the stalk membrane. We have previously reported that lack of GPI-anchored protein synthesis causes mislocalization of Na+K+ATPase and Crb on the apical membrane (Satoh et al., 2013). Therein, we postulated that unidentified GPI-anchored protein(s) concentrate Rh1 and exclude Na+K+ATPase and Crb from the post-Golgi vesicles that are destined to the rhabdomeres. Based on this model, impairment of the Crag, Rab10, Ehbp1, AP1 or Clathrin genes failed to recruit and concentrate Na+K+ATPase in the post-Golgi vesicles destined to the basolateral membrane; however, the GPI-dependent sorting excluded Na+K+ATPase from the post-Golgi vesicles destined to the rhabdomeres in these mutants, thus Na+K+ATPase localizes on the stalk membrane, but not on the rhabdomere in Crag, Rab10, Ehbp1, AP1 or Clathrin mutant photoreceptors.

Fig. 8.

Polarized transport in fly photoreceptors. In wild-type photoreceptors, the membrane proteins are synthesized on the endoplasmic reticulum (ER) and transported to cis-Golgi cisternae that mature into medial- and trans-Golgi cisternae, and then to the trans-Golgi network TGN. Basolateral proteins are sorted into post-Golgi vesicles at the TGN in a Crag-, Rab10-, Ehbp1-, AP1- and clathrin-dependent manner. In contrast, membrane proteins destined for the two apical membranes, the stalk and the rhabdomeres, reach the REs together and are then sorted into separate post-Golgi vesicles to be delivered to the stalk and the rhabdomeres. In Crag-, Rab10- and Ehbp1-mutant photoreceptors, the basolateral membrane proteins are not completely removed at the TGN. Some of them are transported to the recycling endosome (RE) along with apical cargoes and then transported to the stalk membrane, and the others are degraded in lysosomes.

Fig. 8.

Polarized transport in fly photoreceptors. In wild-type photoreceptors, the membrane proteins are synthesized on the endoplasmic reticulum (ER) and transported to cis-Golgi cisternae that mature into medial- and trans-Golgi cisternae, and then to the trans-Golgi network TGN. Basolateral proteins are sorted into post-Golgi vesicles at the TGN in a Crag-, Rab10-, Ehbp1-, AP1- and clathrin-dependent manner. In contrast, membrane proteins destined for the two apical membranes, the stalk and the rhabdomeres, reach the REs together and are then sorted into separate post-Golgi vesicles to be delivered to the stalk and the rhabdomeres. In Crag-, Rab10- and Ehbp1-mutant photoreceptors, the basolateral membrane proteins are not completely removed at the TGN. Some of them are transported to the recycling endosome (RE) along with apical cargoes and then transported to the stalk membrane, and the others are degraded in lysosomes.

Future research should be aimed at determining the exact functions of Crag, Rab10 and Ehbp1 in the TGN budding process. This information will help elucidate the basolateral cargo sorting mechanism. In individuals with polycystic kidney disease, sorting and maintenance of the EGF receptor on the basolateral surface of renal epithelial cells is perturbed (Charron et al., 2000; Cotton et al., 2013; Ryan et al., 2010). Determining the mechanism of basolateral transport will provide new insights into the pathogenesis of renal disorders.

Drosophila stocks and genetic background

Fruit flies were grown at 20-25°C on standard cornmeal–glucose–agar–yeast medium either in the laboratory with room light or 12L/12D incubator. The following fly stocks were used: Rh1-Gal4 (Dr Chihiro Hama, Kyoto Sangyo University, Japan), longGMR-Gal4 (Bloomington Stock 8605; indicated as BL8605 in the following stocks), GMR-Gal4 (BL9146), elav-Gal4 (BL5145), coinFLP-Gal4 (BL58751), UAS-Rab7T22N (Dr Emery Gregory, University of Montreal, Montreal, Canada), UAS-FLAG::Ehbp1 (Dr Hugo Bellen, Baylor College of Medicine, Houston, TX), UAS-FLAG::Crag (produced in the present study), UAS-Rab10T23N (produced in the present study), UAS-Myc::Rab10 (produced in the present study), UAS-Myc::Rab11 (produced in the present study), UAS-AP1μ::VFP (BL64260), UAS-Chc RNAi (BL34742), UAS-Chc DN (BL26874), y w CragGG43 FRT19A/FM7c (BL52218), y w CragCJ101 FRT19A/FM7c (BL51341), w; FRT42D Ehbp1O4/CyOGFP (Dr Hugo Bellen), w; FRT42D Ehbp1A28/CyOGFP (Dr Hugo Bellen) and w AP1γe04546, FRT19A/FM7c; eyFLP (Drosophila Genomics and Genetic Resources at Kyoto; 111963). AP1μEP1112 (BL16987) was recombined with FRT82B and investigated.

To generate the Crag-null mosaic retina, females of the genotype ‘y w CragCJ101 FRT19A/FM7c’ or ‘y w CragGG43 FRT19A/FM7c’ were crossed with males of the genotype ‘y w P3RFP FRT19A; eyFLP/SM1’, and their progeny were used for immunostaining. ‘FRT19A; eyFLP/SM1’ was used for electron microscopy.

To generate the Ehbp1-null mosaic retina for immunostaining, males of the genotype ‘w; FRT42D Ehbp1O4/CyOGFP’ or ‘w; FRT42D Ehbp1A28/CyOGFP’ were crossed with females of the genotype ‘w y eyFLP; FRT42D P3RFP’. To generate the partial rescue Ehbp1A28 whole-eye clone for electron microscopy, males of the genotype ‘w; FRT42D Ehbp1A28/CyOGFP; UAS-FLAG::Ehbp1’ were crossed with females of the genotype ‘w; FRT42D GMR-hid; heat shock-Gal4 eyFLP’ and the progenies were exposed to 37°C for 1 h every 3-4 days after egg laying.

Transgenic flies for UAS-FLAG::Crag, UAS-Myc::Rab10, UAS-Myc::Rab11 and UAS-Rab10T23N

The entire coding regions of the Crag, Rab10 and Rab11 genes were amplified from a LP24278 cDNA clone [Drosophila Genomics Resource Center (DGRC)], a Rab10 full-length cDNA and a Rab11 full-length cDNA, respectively (Satoh et al., 1997b). The coding regions were then cloned into pENTER (Invitrogen, Thermo Fisher Scientific) to construct pENTER-Crag, pENTER-Rab10 and pENTER-Rab11. Using the Gateway system, we recombined them into a pTFW vector (DGRC) and generated pUAST-3xFLAG-Crag, pUAST-5xMyc-Rab10 and pUAST-5xMyc-Rab10. To construct pUAST-Rab10T23N, we amplified the entire coding region of Rab10T23N from the genome of the transgenic flies with pUASp-Rab10T23N (BL9787) and cloned it into the pUAST vector. Plasmids were injected into the embryos (BestGene) to generate transgenic lines.

Immunohistochemistry

Fixation and staining were performed as described previously (Satoh and Ready, 2005), except for the fixative PLP (10 mM periodate, 75 mM lysine, 30 mM phosphate buffer and 4% paraformaldehyde). The primary antisera used were as follows: rabbit anti-Rh1 (1:1000) (Satoh et al., 2005), chicken anti-Rh1 (1:1000) (Satoh et al., 2013), rabbit anti-GM130 (1:300) (ab30637, Abcam), rabbit anti-MPPE (1:1000) (a gift from Dr Han, Southeast University, Nanjing, China), mouse monoclonal anti-Na+K+ATPase alpha subunit (1:500 ascites) [Developmental Studies Hybridoma Bank (DSHB)], rat monoclonal anti-DE-Cad (1:20 supernatant; DSHB), rat anti-Crb (1:300) (Dr Ulrich Tepass, University of Toronto, Ontario, Canada), rabbit anti-TRP (Dr Craig Montell, Johns Hopkins University, Baltimore, MD), mouse monoclonal anti-Eys (1:20 supernatant; DSHB), rabbit anti-Chc (1:500) (Dr Satoshi Kametaka, Nagoya University, Nagoya, Japan) (Kametaka et al., 2010), guinea pig anti-Ehbp1 (1:1000) (Dr Hugo Bellen), rabbit anti-Cnx (1:100) (Satoh et al., 2015), rat anti-p120 (Dr Satoshi Goto, Rikkyo University, Tokyo, Japan) (1:15) (Yamamoto-Hino et al., 2012), guinea pig anti-Rab6 (1:300) (Iwanami et al., 2016), rabbit anti-Rab6 (1:300) (Iwanami et al., 2016), rat anti-Rab11 (1:250) (Otsuka et al., 2019), mouse monoclonal anti-Myc (1:12; DSHB), rabbit anti-Myc (1:300) (Medical and Biological Laboratories, Nagoya, Japan; No. 562) and mouse anti-FLAG M2 (1:1000) (F1804, Sigma-Aldrich Japan). Secondary antibodies were anti-mouse, anti-rabbit, anti-rat and/or anti-chicken antibodies labeled with Alexa Fluor 488, 568 and 647 (1:300) (Life Technologies). Sample images were recorded using a Model FV1000 confocal microscope (60×1.42 NA objective lens; Olympus, Tokyo, Japan). To minimize bleed through, each signal in the double- or triple-stained samples was imaged sequentially. Images were processed in accordance with the Guidelines for Proper Digital Image Handling using ImageJ, Affinity photo and/or Adobe Photoshop CS3. The sectional lengths of the stalk (defined as the membrane stained using anti-Crb antibody and used the sum of both sides of the rhabdomere) and basolateral membrane (defined as the membrane stained using anti-Na+K+ATPase-α antibody, but not by anti-Crb antibody) of nine photoreceptors (three cells each for three ommatidia) for three flies were measured using Fiji for more than three independent samples for each genotype. For the plot of the immunostaining intensity across the Golgi units, the lines were drawn through each Golgi stack and intensities were measured using Fiji and plotted using PLOT2 (micw.org). The plots with the typical pattern were presented. The sectional lengths of the stalk and basolateral membrane visualized using Crb and Na+K+ATPase antibodies, respectively, for three independent samples for each genotype were measured using Fiji. For the quantification of polarity defects, cell numbers with and without cell polarity were counted for three independent samples for each genotype.

Electron microscopy

Electron microscopy was performed as described previously (Satoh et al., 1997a). Samples were observed under a JEM1400 electron microscope (JEOL, Tokyo, Japan) and montages were prepared with a CCD camera system (JEOL). The phenotypes were investigated using the section that had the depth in which two or three photoreceptor nuclei within an ommatidia were observed. The sectional lengths of the stalk and basolateral membrane of nine photoreceptors (three cells each for three ommatidia) for three flies, and the sectional lengths of the ER membrane of 18 photoreceptors (six cells each for three ommatidia) for three flies were measured using Fiji for more than three independent samples for each genotype. For quantification of the number of Golgi stacks, we defined Golgi stacks as the organelles with more than three cisternae or the organelles with two cisternae longer than 100 nm in length within a 400×400 nm square. For quantification of the number of Golgi stacks with three cisternae, we defined these as the organelles with more than three cisternae longer than 200 nm within a 200×200 nm square. For quantification of the number of Golgi cisternae, we defined cisternae as having a membrane structure that appears as a closed ring, with a major axis of more than 200 nm in length.

Immunoblotting

For quantitative isolation of the retina, we used freeze-dried flies stored in acetone (Fujita et al., 1987). LongGMR-Gal4/+ and longGMR-Gal4/Rab10T23N flies at 0-7 days were collected and frozen in liquid nitrogen. The heads were collected using two sieves of different mesh sizes. The heads were then immersed in acetone cooled to −80°C and maintained there for longer than 1 week. The acetone was replaced once or twice during this time to remove the water. The retinas were dissected with forceps, and the proteins were extracted in sodium dodecyl sulfate (SDS) sample buffer (0.05 M Tris, 10% v/v glycerol, 5% v/v β-mercaptoethanol and 2.3% v/v SDS in phosphate buffer solution at 80°C for 2 min). We prepared 12 and 14 independent samples for longGMR-Gal4/+ and longGMR-Gal4/Rab10T23N flies.

Immunoblotting was performed as previously described (Satoh et al., 1997a). Mouse anti-α-tubulin (1:200 supernatant), mouse anti-Na+K+ATPase-α (1:10,000 concentrated supernatant) and mouse anti-Na+K+ATPase-β (1:500 supernatant) were used as primary antibodies, and were obtained from DSHB. HRP-conjugated anti-mouse IgG antibody (1:20,000; Life Technologies) was used as a secondary antibody. Signals were visualized by enhanced chemiluminescence (Clarity Western ECL Substrate; Bio-Rad) and imaged with ChemiDoc XRS+ (Bio-Rad). The intensities of the bands were measured using Fiji.

We thank Drs U. Tepass, C. Montell, H. Bellen, S. Goto, J. Han and S. Kametaka for kindly providing fly stocks and reagents. We also thank the Bloomington Stock Center and the Drosophila Genetic Resource Center (Kyoto Institute of Technology) for their fly stocks. We thank Editage (www.editage.jp) for English language editing.

Author contributions

Conceptualization: A.K.S.; Investigation: Y.N., Y.O.; Writing - original draft: T.S., A.K.S.; Supervision: T.S., A.K.S.; Project administration: A.K.S.; Funding acquisition: A.K.S., T.S.

Funding

This work was supported by Precursory Research for Embryonic Science and Technology [25-J-J4215], by the Japan Society for the Promotion of Science [KAKENHI 15K07050], by the Sumitomo Foundation for Basic Science Research Projects, by the Astellas Foundation for Research on Metabolic Disorders and by a Female Researcher Joint Research Grant from Hiroshima University to A.K.S. [KAKENHI 19K06566] to T.S.

Ang
,
A. L.
,
Fölsch
,
H.
,
Koivisto
,
U.-M.
,
Pypaert
,
M.
and
Mellman
,
I.
(
2003
).
The Rab8 GTPase selectively regulates AP-1B-dependent basolateral transport in polarized Madin-Darby canine kidney cells
.
J. Cell Biol.
163
,
339
-
350
.
Balderhaar
,
H. J.
and
Ungermann
,
C.
(
2013
).
CORVET and HOPS tethering complexes - coordinators of endosome and lysosome fusion
.
J. Cell Sci.
126
,
1307
-
1316
.
Barr
,
F.
and
Lambright
,
D. G.
(
2010
).
Rab GEFs and GAPs
.
Curr. Opin. Cell Biol.
22
,
461
-
470
.
Bonifacino
,
J. S.
(
2014
).
Adaptor proteins involved in polarized sorting
.
J. Cell Biol.
204
,
7
-
17
.
Bosch
,
J. A.
,
Tran
,
N. H.
and
Hariharan
,
I. K.
(
2015
).
CoinFLP: a system for efficient mosaic screening and for visualizing clonal boundaries in Drosophila
.
Development
142
,
597
-
606
.
Bruno
,
J.
,
Brumfield
,
A.
,
Chaudhary
,
N.
,
Iaea
,
D.
and
McGraw
,
T. E.
(
2016
).
SEC16A is a RAB10 effector required for insulin-stimulated GLUT4 trafficking in adipocytes
.
J. Cell Biol.
214
,
61
-
76
.
Cagan
,
R. L.
and
Ready
,
D. F.
(
1989
).
The emergence of order in the Drosophila pupal retina
.
Dev. Biol.
136
,
346
-
362
.
Charron
,
A. J.
,
Bacallao
,
R. L.
and
Wandinger-Ness
,
A.
(
2000
).
ADPKD: a human disease altering Golgi function and basolateral exocytosis in renal epithelia
.
Traffic
1
,
675
-
686
.
Cotton
,
C. U.
,
Hobert
,
M. E.
,
Ryan
,
S.
and
Carlin
,
C. R.
(
2013
).
Basolateral EGF receptor sorting regulated by functionally distinct mechanisms in renal epithelial cells
.
Traffic
14
,
337
-
354
.
Deborde
,
S.
,
Perret
,
E.
,
Gravotta
,
D.
,
Deora
,
A.
,
Salvarezza
,
S.
,
Schreiner
,
R.
and
Rodriguez-Boulan
,
E.
(
2008
).
Clathrin is a key regulator of basolateral polarity
.
Nature
452
,
719
-
723
.
Denef
,
N.
,
Chen
,
Y.
,
Weeks
,
S. D.
,
Barcelo
,
G.
and
Schüpbach
,
T.
(
2008
).
Crag regulates epithelial architecture and polarized deposition of basement membrane proteins in Drosophila
.
Dev. Cell
14
,
354
-
364
.
Devergne
,
O.
,
Sun
,
G. H.
and
Schüpbach
,
T.
(
2017
).
Stratum, a homolog of the human GEF Mss4, partnered with Rab8, controls the basal restriction of basement membrane proteins in epithelial cells
.
Cell Rep.
18
,
1831
-
1839
.
English
,
A. R.
and
Voeltz
,
G. K.
(
2013
).
Rab10 GTPase regulates ER dynamics and morphology
.
Nat. Cell Biol.
15
,
169
-
178
.
Fölsch
,
H.
(
2015
).
Role of the epithelial cell-specific clathrin adaptor complex AP-1B in cell polarity
.
Cell Logist.
5
,
e1074331
.
Fujita
,
S. C.
,
Inoue
,
H.
,
Yoshioka
,
T.
and
Hotta
,
Y.
(
1987
).
Quantitative tissue isolation from Drosophila freeze-dried in acetone
.
Biochem. J.
243
,
97
-
104
.
Giagtzoglou
,
N.
,
Yamamoto
,
S.
,
Zitserman
,
D.
,
Graves
,
H. K.
,
Schulze
,
K. L.
,
Wang
,
H.
,
Klein
,
H.
,
Roegiers
,
F.
and
Bellen
,
H. J.
(
2012
).
dEHBP1 controls exocytosis and recycling of Delta during asymmetric divisions
.
J. Cell Biol.
196
,
65
-
83
.
Giagtzoglou
,
N.
,
Li
,
T.
,
Yamamoto
,
S.
and
Bellen
,
H. J.
(
2013
).
Drosophila EHBP1 regulates Scabrous secretion during Notch-mediated lateral inhibition
.
J. Cell Sci.
126
,
3686
-
3696
.
Gillard
,
G.
,
Shafaq-Zadah
,
M.
,
Nicolle
,
O.
,
Damaj
,
R.
,
Pécréaux
,
J.
and
Michaux
,
G.
(
2015
).
Control of E-cadherin apical localisation and morphogenesis by a SOAP-1/AP-1/clathrin pathway in C. elegans epidermal cells
.
Development
142
,
1684
-
1694
.
Guilherme
,
A.
,
Soriano
,
N. A.
,
Bose
,
S.
,
Holik
,
J.
,
Bose
,
A.
,
Pomerleau
,
D. P.
,
Furcinitti
,
P.
,
Leszyk
,
J.
,
Corvera
,
S.
and
Czech
,
M. P.
(
2004
).
EHD2 and the novel EH domain binding protein EHBP1 couple endocytosis to the actin cytoskeleton
.
J. Biol. Chem.
279
,
10593
-
10605
.
Henry
,
L.
and
Sheff
,
D. R.
(
2008
).
Rab8 regulates basolateral secretory, but not recycling, traffic at the recycling endosome
.
Mol. Biol. Cell
19
,
2059
-
2068
.
Hollingsworth
,
T. J.
and
Gross
,
A. K.
(
2012
).
Defective trafficking of rhodopsin and its role in retinal degenerations
.
Int. Rev. Cell Mol. Biol.
293
,
1
-
44
.
Holly
,
R. M.
,
Mavor
,
L. M.
,
Zuo
,
Z.
and
Blankenship
,
J. T.
(
2015
).
A rapid, membrane-dependent pathway directs furrow formation through RalA in the early Drosophila embryo
.
Development
142
,
2316
-
2328
.
Huber
,
L. A.
,
Pimplikar
,
S.
,
Parton
,
R. G.
,
Virta
,
H.
,
Zerial
,
M.
and
Simons
,
K.
(
1993
).
Rab8, a small GTPase involved in vesicular traffic between the TGN and the basolateral plasma membrane
.
J. Cell Biol.
123
,
35
-
45
.
Husain
,
N.
,
Pellikka
,
M.
,
Hong
,
H.
,
Klimentova
,
T.
,
Choe
,
K.-M.
,
Clandinin
,
T. R.
and
Tepass
,
U.
(
2006
).
The agrin/perlecan-related protein eyes shut is essential for epithelial lumen formation in the Drosophila retina
.
Dev. Cell
11
,
483
-
493
.
Isabella
,
A. J.
and
Horne-Badovinac
,
S.
(
2016
).
Rab10-mediated secretion synergizes with tissue movement to build a polarized basement membrane architecture for organ morphogenesis
.
Dev. Cell
38
,
47
-
60
.
Ishida
,
M.
,
E. Oguchi
,
M.
and
Fukuda
,
M.
(
2016
).
Multiple types of guanine nucleotide exchange factors (GEFs) for Rab small GTPases
.
Cell Struct. Funct.
41
,
61
-
79
.
Iwanami
,
N.
,
Nakamura
,
Y.
,
Satoh
,
T.
,
Liu
,
Z.
and
Satoh
,
A. K.
(
2016
).
Rab6 is required for multiple apical transport pathways but not the basolateral transport pathway in Drosophila photoreceptors
.
PLoS Genet.
12
,
e1005828
.
Kametaka
,
S.
,
Sawada
,
N.
,
Bonifacino
,
J. S.
and
Waguri
,
S.
(
2010
).
Functional characterization of protein-sorting machineries at the trans-Golgi network in Drosophila melanogaster
.
J. Cell Sci.
123
,
460
-
471
.
Karagiosis
,
S. A.
and
Ready
,
D. F.
(
2004
).
Moesin contributes an essential structural role in Drosophila photoreceptor morphogenesis
.
Development
131
,
725
-
732
.
Lerner
,
D. W.
,
McCoy
,
D.
,
Isabella
,
A. J.
,
Mahowald
,
A. P.
,
Gerlach
,
G. F.
,
Chaudhry
,
T. A.
and
Horne-Badovinac
,
S.
(
2013
).
A Rab10-dependent mechanism for polarized basement membrane secretion during organ morphogenesis
.
Dev. Cell
24
,
159
-
168
.
Li
,
B. X.
,
Satoh
,
A. K.
and
Ready
,
D. F.
(
2007
).
Myosin V, Rab11, and dRip11 direct apical secretion and cellular morphogenesis in developing Drosophila photoreceptors
.
J. Cell Biol.
177
,
659
-
669
.
Lv
,
P.
,
Sheng
,
Y.
,
Zhao
,
Z.
,
Zhao
,
W.
,
Gu
,
L.
,
Xu
,
T.
and
Song
,
E.
(
2015
).
Targeted disruption of Rab10 causes early embryonic lethality
.
Protein Cell
6
,
463
-
467
.
Mavor
,
L. M.
,
Miao
,
H.
,
Zuo
,
Z.
,
Holly
,
R. M.
,
Xie
,
Y.
,
Loerke
,
D.
and
Blankenship
,
J. T.
(
2016
).
Rab8 directs furrow ingression and membrane addition during epithelial formation in Drosophila melanogaster
.
Development
143
,
892
-
903
.
Montell
,
C.
and
Rubin
,
G. M.
(
1989
).
Molecular characterization of the Drosophila trp locus: a putative integral membrane protein required for phototransduction
.
Neuron
2
,
1313
-
1323
.
Müller
,
T.
,
Hess
,
M. W.
,
Schiefermeier
,
N.
,
Pfaller
,
K.
,
Ebner
,
H. L.
,
Heinz-Erian
,
P.
,
Ponstingl
,
H.
,
Partsch
,
J.
,
Röllinghoff
,
B.
,
Köhler
,
H.
, et al. 
(
2008
).
MYO5B mutations cause microvillus inclusion disease and disrupt epithelial cell polarity
.
Nat. Genet.
40
,
1163
-
1165
.
Nakajo
,
A.
,
Yoshimura
,
S.-I.
,
Togawa
,
H.
,
Kunii
,
M.
,
Iwano
,
T.
,
Izumi
,
A.
,
Noguchi
,
Y.
,
Watanabe
,
A.
,
Goto
,
A.
,
Sato
,
T.
, et al. 
(
2016
).
EHBP1L1 coordinates Rab8 and Bin1 to regulate apical-directed transport in polarized epithelial cells
.
J. Cell Biol.
212
,
297
-
306
.
Nakatsu
,
F.
,
Hase
,
K.
and
Ohno
,
H.
(
2014
).
The role of the clathrin adaptor AP-1: polarized sorting and beyond
.
Membranes
4
,
747
-
763
.
Ogi
,
S.
,
Matsuda
,
A.
,
Otsuka
,
Y.
,
Liu
,
Z.
,
Satoh
,
T.
and
Satoh
,
A. K.
(
2019
).
Syndapin constricts microvillar necks to form a united rhabdomere in Drosophila photoreceptors
.
Development
146
,
dev169292
.
Otsuka
,
Y.
,
Satoh
,
T.
,
Nakayama
,
N.
,
Inaba
,
R.
,
Yamashita
,
H.
and
Satoh
,
A. K.
(
2019
).
Parcas is the predominant Rab11-GEF for rhodopsin transport in Drosophila photoreceptors
.
J. Cell Sci.
132
,
jcs231431
.
Pellikka
,
M.
,
Tanentzapf
,
G.
,
Pinto
,
M.
,
Smith
,
C.
,
McGlade
,
C. J.
,
Ready
,
D. F.
and
Tepass
,
U.
(
2002
).
Crumbs, the Drosophila homologue of human CRB1/RP12, is essential for photoreceptor morphogenesis
.
Nature
416
,
143
-
149
.
Pfeffer
,
S. R.
(
2013
).
Rab GTPase regulation of membrane identity
.
Curr. Opin. Cell Biol.
25
,
414
-
419
.
Pocha
,
S. M.
,
Shevchenko
,
A.
and
Knust
,
E.
(
2011
).
Crumbs regulates rhodopsin transport by interacting with and stabilizing myosin V
.
J. Cell Biol.
195
,
827
-
838
.
Rodriguez-Boulan
,
E.
and
Macara
,
I. G.
(
2014
).
Organization and execution of the epithelial polarity programme
.
Nat. Rev. Mol. Cell Biol.
15
,
225
-
242
.
Rodriguez-Boulan
,
E.
,
Kreitzer
,
G.
and
Müsch
,
A.
(
2005
).
Organization of vesicular trafficking in epithelia
.
Nat. Rev. Mol. Cell Biol.
6
,
233
-
247
.
Ryan
,
S.
,
Verghese
,
S.
,
Cianciola
,
N. L.
,
Cotton
,
C. U.
and
Carlin
,
C. R.
(
2010
).
Autosomal recessive polycystic kidney disease epithelial cell model reveals multiple basolateral epidermal growth factor receptor sorting pathways
.
Mol. Biol. Cell
21
,
2732
-
2745
.
Sato
,
T.
,
Mushiake
,
S.
,
Kato
,
Y.
,
Sato
,
K.
,
Sato
,
M.
,
Takeda
,
N.
,
Ozono
,
K.
,
Miki
,
K.
,
Kubo
,
Y.
,
Tsuji
,
A.
, et al. 
(
2007
).
The Rab8 GTPase regulates apical protein localization in intestinal cells
.
Nature
448
,
366
-
369
.
Sato
,
T.
,
Iwano
,
T.
,
Kunii
,
M.
,
Matsuda
,
S.
,
Mizuguchi
,
R.
,
Jung
,
Y.
,
Hagiwara
,
H.
,
Yoshihara
,
Y.
,
Yuzaki
,
M.
,
Harada
,
R.
, et al. 
(
2014
).
Rab8a and Rab8b are essential for several apical transport pathways but insufficient for ciliogenesis
.
J. Cell Sci.
127
,
422
-
431
.
Satoh
,
A. K.
and
Ready
,
D. F.
(
2005
).
Arrestin1 mediates light-dependent rhodopsin endocytosis and cell survival
.
Curr. Biol.
15
,
1722
-
1733
.
Satoh
,
A.
,
Tokunaga
,
F.
,
Kawamura
,
S.
and
Ozaki
,
K.
(
1997a
).
In situ inhibition of vesicle transport and protein processing in the dominant negative Rab1 mutant of Drosophila
.
J. Cell Sci.
110
,
2943
-
2953
.
Satoh
,
A. K.
,
Tokunaga
,
F.
and
Ozaki
,
K.
(
1997b
).
Rab proteins of Drosophila melanogaster: novel members of the Rab-protein family
.
FEBS Lett.
404
,
65
-
69
.
Satoh
,
A. K.
,
O'Tousa
,
J. E.
,
Ozaki
,
K.
and
Ready
,
D. F.
(
2005
).
Rab11 mediates post-Golgi trafficking of rhodopsin to the photosensitive apical membrane of Drosophila photoreceptors
.
Development
132
,
1487
-
1497
.
Satoh
,
T.
,
Inagaki
,
T.
,
Liu
,
Z.
,
Watanabe
,
R.
and
Satoh
,
A. K.
(
2013
).
GPI biosynthesis is essential for rhodopsin sorting at the trans-Golgi network in Drosophila photoreceptors
.
Development
140
,
385
-
394
.
Satoh
,
T.
,
Ohba
,
A.
,
Liu
,
Z.
,
Inagaki
,
T.
and
Satoh
,
A. K.
(
2015
).
dPob/EMC is essential for biosynthesis of rhodopsin and other multi-pass membrane proteins in Drosophila photoreceptors
.
eLife
4
,
e06306
.
Satoh
,
T.
,
Nakamura
,
Y.
and
Satoh
,
A. K.
(
2016
).
Rab6 functions in polarized transport in Drosophila photoreceptors
.
Fly (Austin)
10
,
123
-
127
.
Schneeberger
,
K.
,
Vogel
,
G. F.
,
Teunissen
,
H.
,
van Ommen
,
D. D.
,
Begthel
,
H.
,
El Bouazzaoui
,
L.
,
van Vugt
,
A. H. M.
,
Beekman
,
J. M.
,
Klumperman
,
J.
,
Müller
,
T.
, et al. 
(
2015
).
An inducible mouse model for microvillus inclusion disease reveals a role for myosin Vb in apical and basolateral trafficking
.
Proc. Natl. Acad. Sci. USA
112
,
12408
-
12413
.
Schuck
,
S.
,
Gerl
,
M. J.
,
Ang
,
A.
,
Manninen
,
A.
,
Keller
,
P.
,
Mellman
,
I.
and
Simons
,
K.
(
2007
).
Rab10 is involved in basolateral transport in polarized Madin-Darby canine kidney cells
.
Traffic
8
,
47
-
60
.
Shafaq-Zadah
,
M.
,
Brocard
,
L.
,
Solari
,
F.
and
Michaux
,
G.
(
2012
).
AP-1 is required for the maintenance of apico-basal polarity in the C. elegans intestine
.
Development
139
,
2061
-
2070
.
Shi
,
A.
,
Chen
,
C. C.-H.
,
Banerjee
,
R.
,
Glodowski
,
D.
,
Audhya
,
A.
,
Rongo
,
C.
and
Grant
,
B. D.
(
2010
).
EHBP-1 functions with RAB-10 during endocytic recycling in Caenorhabditis elegans
.
Mol. Biol. Cell
21
,
2930
-
2943
.
Sobajima
,
T.
,
Yoshimura
,
S.-I.
,
Iwano
,
T.
,
Kunii
,
M.
,
Watanabe
,
M.
,
Atik
,
N.
,
Mushiake
,
S.
,
Morii
,
E.
,
Koyama
,
Y.
,
Miyoshi
,
E.
, et al. 
(
2015
).
Rab11a is required for apical protein localisation in the intestine
.
Biol. Open
4
,
86
-
94
.
Spang
,
A.
(
2016
).
Membrane tethering complexes in the endosomal system
.
Front. Cell Dev. Biol.
4
,
35
.
Stenmark
,
H.
(
2009
).
Rab GTPases as coordinators of vesicle traffic
.
Nat. Rev. Mol. Cell Biol.
10
,
513
-
525
.
Vogel
,
G. F.
,
Klee
,
K. M. C.
,
Janecke
,
A. R.
,
Müller
,
T.
,
Hess
,
M. W.
and
Huber
,
L. A.
(
2015
).
Cargo-selective apical exocytosis in epithelial cells is conducted by Myo5B, Slp4a, Vamp7, and Syntaxin 3
.
J. Cell Biol.
211
,
587
-
604
.
Vogel
,
G. F.
,
Janecke
,
A. R.
,
Krainer
,
I. M.
,
Gutleben
,
K.
,
Witting
,
B.
,
Mitton
,
S. G.
,
Mansour
,
S.
,
Ballauff
,
A.
,
Roland
,
J. T.
,
Engevik
,
A. C.
, et al. 
(
2017
).
Abnormal Rab11-Rab8-vesicles cluster in enterocytes of patients with microvillus inclusion disease
.
Traffic
18
,
453
-
464
.
Wang
,
P.
,
Liu
,
H.
,
Wang
,
Y.
,
Liu
,
O.
,
Zhang
,
J.
,
Gleason
,
A.
,
Yang
,
Z.
,
Wang
,
H.
,
Shi
,
A.
and
Grant
,
B. D.
(
2016
).
RAB-10 promotes EHBP-1 bridging of filamentous actin and tubular recycling endosomes
.
PLoS Genet.
12
,
e1006093
.
Wolff
,
T.
and
Ready
,
D.
(
1993
).
Pattern formation of the Drosophila retina
. In
The Development of Drosophila melanogaster (ed. M. Bate and A. Martinez Arias)
, Vol.
2
, pp.
1277
-
1325
.
Cold Spring Harbor Laboratory Press
.
Xiong
,
B.
and
Bellen
,
H. J.
(
2013
).
Rhodopsin homeostasis and retinal degeneration: lessons from the fly
.
Trends Neurosci.
36
,
652
-
660
.
Xiong
,
B.
,
Bayat
,
V.
,
Jaiswal
,
M.
,
Zhang
,
K.
,
Sandoval
,
H.
,
Charng
,
W.-L.
,
Li
,
T.
,
David
,
G.
,
Duraine
,
L.
,
Lin
,
Y.-Q.
, et al. 
(
2012
).
Crag is a GEF for Rab11 required for rhodopsin trafficking and maintenance of adult photoreceptor cells
.
PLoS Biol.
10
,
e1001438
.
Xu
,
T.
and
Rubin
,
G. M.
(
1993
).
Analysis of genetic mosaics in developing and adult Drosophila tissues
.
Development
117
,
1223
-
1237
.
Yamamoto-Hino
,
M.
,
Abe
,
M.
,
Shibano
,
T.
,
Setoguchi
,
Y.
,
Awano
,
W.
,
Ueda
,
R.
,
Okano
,
H.
and
Goto
,
S.
(
2012
).
Cisterna-specific localization of glycosylation-related proteins to the Golgi apparatus
.
Cell Struct. Funct.
37
,
55
-
63
.
Yano
,
H.
,
Yamamoto-Hino
,
M.
,
Awano
,
W.
,
Aoki-Kinoshita
,
K. F.
,
Tsuda-Sakurai
,
K.
,
Okano
,
H.
and
Goto
,
S.
(
2012
).
Identification of proteasome components required for apical localization of Chaoptin using functional genomics
.
J. Neurogenet.
26
,
53
-
63
.
Yoshimura
,
S.-I.
,
Gerondopoulos
,
A.
,
Linford
,
A.
,
Rigden
,
D. J.
and
Barr
,
F. A.
(
2010
).
Family-wide characterization of the DENN domain Rab GDP-GTP exchange factors
.
J. Cell Biol.
191
,
367
-
381
.
Zhu
,
H.
,
Sewell
,
A. K.
and
Han
,
M.
(
2015
).
Intestinal apical polarity mediates regulation of TORC1 by glucosylceramide in C. elegans
.
Genes Dev.
29
,
1218
-
1223
.

Competing interests

The authors declare no competing or financial interests.

Supplementary information