Role of adaptor proteins and clathrin in the trafficking of human kidney anion exchanger 1 (kAE1) to the cell surface Mutita Junking1, Nunghathai Sawasdee1, Natapol Duangtum1,2, Boonyarit Cheunsuchon3, Thawornchai Limjindaporn1,4, Pa-thai Yenchitsomanus1,* 1

Division of Molecular Medicine, Department of Research and Development, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand 2 Graduate Program in Immunology, Department of Immunology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand 3 Department of Pathology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand 4 Department of Anatomy, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand

*Corresponding author: Pa-thai Yenchitsomanus Division of Molecular Medicine, Department of Research and Development, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand Tel: 66-2-419-2777 Fax: 66-2-411-0169 E-mail: [email protected] [email protected]

Keywords: Kidney anion exchanger 1 (kAE1), distal renal tubular acidosis, protein-protein interaction, basolateral sorting protein, protein trafficking Synopsis Mutations of SLC4A1 encoding kidney anion exchanger 1 (kAE1) result in distal renal tubular acidosis (dRTA). The mutant kAE1 shows basolateral trafficking defect in α-intercalated cells. It is not clear how the normal intracellular trafficking of kAE1 occurs. We identified the proteins involved in intracellular sorting and trafficking of kAE1 in kidney cell lines and human kidney tissues. Adaptor protein complexes (AP-1 mu1A, AP-3 mu1, and AP4 mu1) and clathrin play crucial role in intracellular sorting and trafficking of human kAE1. Abstract Kidney anion exchanger 1 (kAE1) plays an important role in acid-base homeostasis by mediating chloride/bicarbornate (Cl-/HCO3-) exchange at basolateral membrane of α-intercalated cells in distal nephron. Impaired intracellular trafficking of kAE1 caused by mutations of SLC4A1 encoding kAE1 results in kidney disease – distal renal tubular acidosis (dRTA). However, it is not known how the intracellular sorting and trafficking of kAE1 from trans-Golgi network (TGN) to the basolateral membrane occur. Here, we studied the role of basolateral-related sorting proteins, including mu1 subunit of adaptor protein (AP) complexes, clathrin, and protein kinase D, on kAE1 trafficking in polarized and non-polarized kidney cells. By using RNA interference, coimmunoprecipitation, yellow fluorescent protein-based protein fragment complementation assay, and immunofluorescence staining, we demonstrated that AP-1 mu1A, AP-3 mu1, AP-4 mu1 and clathrin (but not AP-1 mu1B, PKD1 or PKD2) play crucial roles in intracellular sorting and trafficking of kAE1. We also demonstrated colocalization of kAE1 and basolateral-related sorting proteins in human kidney tissues by double immunofluorescence staining. These findings indicate that AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin are required for kAE1 sorting and trafficking from TGN to the basolateral membrane of acid-secreting α-intercalated cells.

This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/tra.12172

This article is protected by copyright. All rights reserved

Introduction Distal renal tubular acidosis (dRTA) is a kidney disease characterized by a failure of acid secretion by the αintercalated cells of the cortical collecting duct of the distal nephron. This failure leads to several clinical manifestations including muscle weakness, growth retardation, metabolic bone disease, nephrocalcinosis, nephrolithiasis, chronic pyelonephritis, and renal failure (1, 2). Familial dRTA can be inherited in both autosomal dominant (3) and autosomal recessive (AR) patterns with a broad spectrum of clinical severity. A common cause of AR dRTA prevalent in tropical regions – namely tropical dRTA (4, 5) – and AD dRTA observed in temperate regions is attributable to mutations of solute carrier family 4, anion exchanger, member 1 (SLC4A1) encoding human anion exchanger 1 (AE1). These mutations often result in impaired trafficking of mutant kAE1 proteins to function at basolateral membrane of the α-intercalated cells (5, 6). Human AE1 is a member of the bicarbonate transporter superfamily involved in maintenance of acid-base homeostasis in the human body. The SLC4A1 encodes both erythroid AE1 (eAE1) and kidney AE1 (kAE1) isoforms. The former is a major integral protein having roles in Cl-/HCO3- exchange and cytoskeletal anchorage of red cell membrane whereas the latter is a basolateral Cl-/HCO3- exchanger of the acid-secreting α-intercalated cells of the distal nephron (7). Although kAE1 is important for acid-base homeostasis, there is little information regarding kAE1 sorting and trafficking in kidney cells. kAE1 is a multi-spanning membrane protein which both the cytosolic N- and C-termini are essential for binding to other proteins and for basolateral trafficking (8). Several motifs including tyrosine-based and di-leucine motifs located within the cytosolic domains are involved in targeting of kAE1 to basolateral membrane (8-10). The cytosolic C-terminus of kAE1 binds to several proteins including carbonic anhydrase II, glyceraldehyde-3 phosphate dehydrogenase (GAPDH), and nephrin (11-13). Our group has recently reported the interaction between C-terminal region of kAE1 (Ct-kAE1) and adaptor-related protein complex 1 mu1A (AP-1 mu1A) subunit (14). Suppression of endogenous AP-1 mu1A in human embryonic kidney (HEK) 293T decreased membrane localization of kAE1, suggesting a role of AP-1 mu1A in the kAE1 trafficking in kidney cells (14). Additionally, kAE1 co-localized with gamma-adaptin of AP-1 complexes in mouse kidney sections; and suppression of AP-1 mu1A subunit interfered with surface delivery of kAE1 (15). The adaptor protein (AP) complexes play crucial roles in cargo protein selection and vesicle formation in postGolgi trafficking pathways. In mammalian cells, two forms of AP-1 complexes, AP-1 mu1A and AP-1 mu1B, exist. Both share the same subunits excepting for the medium chain (16), which is a central subunit that mediates cargo protein selection via directly binding with tyrosine-based sorting motifs on the cytoplasmic domain of the cargo protein (17-20). Functionally, the ubiquitous AP-1 mu1A facilitates vesicle transport between the TGN and early endosomes, and the epithelial-specific AP-1 mu1B is required for polarized trafficking of many basolateral membrane proteins (21, 22). AP-3 was found to regulate the exit of exogenous vesicular stomatitis virus glycoprotein (VSVG) that contains basolateral sorting signal, from the Golgi complex in HeLa cells (16). Simmen et al. have proposed a role for AP-4 in basolateral sorting of receptors for LDL, transferrin, and mannose-6phosphate in MDCK cells, but the organelles associated with this sorting event have not fully been clarified (23). Clathrin may also be involved in basolateral protein sorting, as AP complexes have binding sites for clathrin (24). Suppression of clathrin depolarized most basolateral proteins and promoted their mis-sorting into apical carrier vesicles while clathrin knock-down did not affect apical proteins. These results demonstrate the necessity for clathrin in basolateral protein trafficking in epithelial cells (25). In polarized epithelial cells, membrane protein is transported from TGN towards either apical or basolateral membrane domain. Protein kinase D1 (PKD1) is a serine/threonine kinase that binds to TGN through its first cysteine-rich domain in a DAG-dependent process (26, 27). Inactivation of PKD1 inhibited basolateral transport vesicle formation, resulting in passive entry of basolateral cargo into apical transport carrier. However, inactivation of PKD neither alters the expression of proteins on the apical plasma membrane nor results in accumulation in TGN (28). Two additional isoforms of PKD, PKD2 and PKD3, were identified. Similarly, PKD2 and PKD3 are also necessary for transport of basolateral proteins from TGN. Interestingly, PKD1 and PKD2 are responsible for transporting different basolateral cargo molecules from TGN. The exocyst (Sec6) that binds to basolateral transmembrane protein was arrested in TGN of cells which deficient in PKD1 kinase activity, while expression of a kinase-inactive form of PKD2 resulted in retardation of β-integrin and E-cadherin transport (28). To identify the proteins that play roles in basolateral-related sorting and trafficking of kAE1, we used RNA interference technique to knock-down mRNAs of the proteins functioning in this respect in polarized MDCK cells

This article is protected by copyright. All rights reserved

permanently expressing kAE1, in which kAE1 trafficking could be observed. The interactions between kAE1 and sorting or trafficking proteins were examined in vitro by co-immunoprecipitation and in situ by yellow fluorescent protein-based protein fragment complementation assay (YFP-PCA) in human kidney cell lines (HEK 293T). Moreover, to confirm their interactions in human kidney, we also investigated the co-localization between kAE1 and sorting proteins in human kidney tissue by using immunofluorescence-staining technique.

Results Polarized MDCK cells stably expressed HA-tagged kAE1 at basolateral membrane To identify sorting proteins that regulate kAE1 trafficking in polarized kidney cells, we generated stable MDCK cell lines that expressed kAE1. The MDCK cells were transduced to express kAE1 by using a replication defective retrovirus, pFB NeO retroviral vector, subcloned with cDNA encoding HA-tagged kAE1 (kAE1-HA). kAE1-HA expression in MDCK cells was determined by immunoblotting and immunofluorescence staining. kAE1HA could be detected in protein lysate with two major bands (Figure 1A) and images of xz scanning sections of polarized cell monolayers revealed localization of kAE1 predominantly in the basolateral domain (Figure 1B). The expression of kAE1-HA in MDCK cells was stable in long-term cultures with gradual decrease to about 50% within 3 months. MDCK-kAE1-HA cells expressed endogenous sorting proteins In polarized epithelial cells, several proteins are involved in sorting and transporting cargo proteins from TGN to basolateral membrane. These proteins include mu subunit of adaptor protein (AP) complexes (AP-1 mu1A, AP-1 mu1B, AP-3 mu1, and AP-4 mu1), clathrin (CLTC), and protein kinase D (PKD1 and PKD2). Before studying the roles of these proteins in kAE1 sorting and trafficking by using RNA interference (siRNA) to knock-down their RNAs, we tested whether mRNAs of these proteins are present in MDCK-kAE1-HA cells or not. We examined their transcripts by reverse-transcription (RT) PCR. Total RNA was extracted from MDCK-kAE1-HA cells and reverse transcribed to cDNA by reverse transcriptase. Then, cDNA was amplified by PCR using primers specific to each cDNA sequence (Table 1). MDCK-kAE1-HA cells were found to express mRNAs of all candidate-sorting proteins examined (Figure 1C). Thus, they could be further studied by the siRNA knock-down method. RNA interference effectively reduced mRNA and sorting proteins To examine effects of basolateral-sorting proteins on kAE1 trafficking, we used siRNA to transiently deplete expression of each sorting proteins in MDCK-kAE1-HA cells. siRNAs targeting to sorting proteins were designed (Table 2) and their efficiencies in suppression of endogenous sorting proteins are shown in Figure 2. As demonstrated by real-time PCR analysis, transfection of MDCK-kAE1-HA cells with 10, 20, and 40 pmol of siRNA against mRNA of each sorting protein generally reduced transcription within 24 h (excepting for AP-4 mu1 which its mRNA was clearly reduced at 48 and 72 h) (Figure 2A). Moreover, siRNA treatment efficiently suppressed the mRNA levels of sorting proteins in a dose- and time-dependent manner, in which the suppression were detected until 72 h after transfection (Figure 2A). Neither siRNA-scramble (siControl), nor Lipofectamine 2000 (data not show) affected mRNA levels of the sorting proteins. Levels of protein expression were also determined by western blot analysis. Transfection of MDCK-kAE1-HA cells with 40 pmol for 72 h of each siRNA clearly decreased adaptor protein complexes (AP-1 mu1A, AP-1 mu1B, AP-3 mu1, and AP-4 mu1) and clathrin (CLTC), and protein kinase D (PKD1 and PKD2) proteins while expression levels of kAE1-HA and β–actin were not changed (Figure 2B). The levels of protein left after the siRNA treatment were variable but generally under 20% of their original levels (Figure 2C). Suppression of subset of sorting proteins reduced cell surface expression of kAE1-HA in non-polarized and polarized MDCK cells The effects of siRNAs against mRNA of each sorting protein on kAE1 localization were observed in non-polarized and polarized MDCK-kAE1-HA cells. HA epitope inserted at the third extracellular loop of kAE1 was used to observe kAE1 expression on cell surface, which could be examined by flow cytometry. Non-polarized and polarized MDCK-kAE1-HA cells were transfected with 40 pmol of siRNA against mRNA of each sorting protein for 72 h, then the non-polarized MDCK-kAE1-HA cells were collected and examined by flow cytometry using rabbit anti-HA antibody and donkey anti-rabbit IgG conjugated with Cy3 fluorescein for detection. The results showed that siRNAs targeting to AP-1 mu1A, AP-3 mu1, AP-4 mu1, and CLTC mRNAs decreased membranous kAE1 expression. Percentage of cell surface expression of kAE1-HA in MDCK cells transfected with siAP-1 mu1A, siAP-3 mu1, siAP-4 mu1, and siCLTC were 38.6+3.6, 64.2+14.1, 58.8+15.0, and 32.6+4.5, respectively while percentage of cell surface expression of kAE1-HA in MDCK cells transfected with siAP-1 mu1B, siPKD1, and siPKD2 were 96.1+1, 95.9+0.8, and 95.4+0.9, respectively (Figure 3A and Figure S1).

This article is protected by copyright. All rights reserved

Double knock-down using siRNA against sorting-proteins that affect kAE1 localization were also carried out. MDCK-kAE1-HA cells were doubly transfected with 40 pmol each of siRNA for 72 h. Percentage of cell surface expression of kAE1-HA in MDCK cells transfected with siAP-1 mu1A and siAP-3 mu1 was 38+4.3, siAP-1 mu1A and siAP-4 mu1 was 30+7.0, and siAP-1 mu1A and siCLTC was 28+2.3 (Figure 3B). The levels of sorting-protein expression as determined by western blot analysis were clearly decreased (Figures 3C, 3D, and 3F). The polarized MDCK-kAE1-HA cells transfected with different siRNAs were stained with rabbit anti-HA antibody and donkey anti-rabbit IgG conjugated with Cy3 fluorescein. Reduction of basolateral kAE1 was clearly observed in polarized MDCK-kAE1-HA cells transfected with siRNAs targeting to AP-1 mu1A, AP-3 mu1, AP-4 mu1, and CLTC mRNAs, but it was not changed in the cells transfected with siRNAs targeting to AP-1 mu1B, PKD1, and PKD2 mRNAs (Figure 4A). To observe cellular location of kAE1-HA when AP-1 mu1A or AP-3 mu1 or AP-4 mu1 or CLTC was suppressed, we stained the siRNA transfected MDCK-kAE1-HA cells with the antibody to giantin, which was specific to Golgi apparatus (Figure 4B), or with the antibody to TGN46, which was specific to TGN (Figure 4C). In siControl cells, kAE1-HA was predominantly localized at the cell surface whereas in the cells with siAP-1 mu1A, siAP-3 mu1, siAP-4 mu1, and siCLTC, colocalization of kAE1-HA with giantin were 20.5+3.2%, 20.0+4.0%, 40.0+2.0%, and 51.2+2.0%, respectively (Figure 4D). Colocalization of kAE1-HA with TGN46 in the cells with siAP-1 mu1A, siAP-3 mu1, siAP-4 mu1, and siCLTC were 85.7+1.3%, 55.5+2.0%, 85.7+1.0%, and 56.8+2.4%, respectively (Figure 4D). Interaction of kAE1 and subset of sorting proteins in MDCK cells and HEK 293T cells detected by coimmunoprecipitation To investigate whether kAE1 directly interacts with adaptor protein complexes (AP-1 mu1A, AP-1 mu1B, AP-3 mu1, and AP-4 mu1), clathrin (CLTC), and protein kinases (PKD1 and PKD2), kAE1-HA was coimmunoprecipitated with APs or clathrin or protein kinases by using antibody specific to each AP or clathrin or protein kinase. The co-immunoprecipitated kAE1-HA was detected by western blot method using anti-HA antibody. Figure 5A shows co-immunoprecipitation of kAE1-HA with AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin (CLTC) but not with AP-1 mu1B, PKD1, and PKD2. The interaction was confirmed in human kidney cells, HEK 293T cells, transfected with other kAE1 construct, kAE1-Myc. Full-length kAE1 cDNA sequence was subcloned into a mammalian expression vector (pcDNA3.1) to express kAE1 tagged with Myc epitope by insertion at the third extracellular loop of kAE1, which was transfected in to HEK 293T cells. kAE1-Myc was coimmunoprecipitated with AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin (CLTC) but not with AP-1 mu1B, PKD1, and PKD2 (Figure 5B). The results from both experiments suggest that, by the co-immunoprecipitation technique, kAE1 interacts with four sorting proteins including AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin (CLTC) but it does not interact with AP-1 mu1B, PKD1, and PKD2. In situ interaction of kAE1 and subset of sorting proteins in HEK 293T cells detected by YFP-PCA The in situ interactions between kAE1 and each of the sorting proteins within HEK 293T cells were examined by yellow fluorescent protein (YFP)-based protein fragment complementation assay (PCA). The YFP fusion proteins were individually expressed (data not shown) or co-expressed in HEK 293T cells and their interactions were demonstrated by intracellular yellow fluorescent signals, which were captured by confocal microscopy. HEK 293T cells were co-transfected to express YFP[1] and YFP[2] as a negative control, which showed no fluorescent signal (Figure 5Ca). As a positive control, the HEK 293T cells were co-transfected to express YFP[1]-kAE1 and YFP[2]-kAE1; dimerization of kAE1 fusion proteins resulted in YFP[1] and YFP[2] complementation, generating intracellular yellow-fluorescent signals (Figure 5Cb). Similarly, when the HEK 293T cells were co-transfected to express YFP[1]-kAE1 and either YFP[2]-AP-1 mu1A (Figure 5Bc) or YFP[2]-AP-3 mu1 (Figure 5Ce) or YFP[2]AP-4 mu1 (Figure 5Cf) or YFP[2]-CLTC (Figure 5Cg), intracellular yellow-fluorescent signals were generated. However, intracellular yellow-fluorescent signals were not detected in HEK 293T cells when YFP[1]-kAE1 and YFP[2]-AP-1 mu1B were co-expressed (Figure 5Cd). These results confirmed the co-immunoprecipitation findings that kAE1 could interact with AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin but not with AP-1 mu1B. To examine the location of protein-protein interactions, we next stained the transfected HEK 293T cells with the TGN marker (TGN46). The results showed that, YFP signals (the confocal microscope was now adjusted to capture the signals as green to match with red, resulting in orange) co-localized with TGN46 (Figure 5D), indicating that their interactions occurred at TGN. Co-localization of kAE1 and subset of sorting proteins in human kidney tissues detected by double immunoflourescence staining To investigate co-localization of native kAE1 and sorting proteins in human kidney tissues, we conducted double immunofluorescence staining by using antibodies specific to kAE1 and each of the sorting proteins. The sections of human kidney tissues were fixed and reacted with anti-Ct-kAE1 antibody and antibodies against each of the

This article is protected by copyright. All rights reserved

sorting proteins as primary antibodies, followed by secondary antibodies coupled to Cy3 or Alexa 488. The stained sections were then examined under the LSM 510 META confocal microscope (Carl Ziess). In contrast with the section incubated with only the secondary antibody (Figure 6 – bottom panel), the native endogenous kAE1 could clearly be detected in red fluorescence at the basolateral membrane of the tubular lining epithelium and with lower intensities inside the cells in the human kidney tissue. AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin were stained in green fluorescence in high intensities also at the basolateral membrane and inside the tubular lining cells (Figure 6). The pictures with higher magnification showed co-localization of endogenous kAE1 (red fluorescence) and endogenous AP-1 mu1A, AP-3 mu1, AP-4 mu1, or clathrin (green fluorescence), yielding yellow fluorescence (Figure 6 – last column). However, endogenous kAE1 and AP-1 mu1B were not found to colocalize to each other (Figure 6 – second low and last column). These results indicate that native kAE1 interacts with native AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin but not with AP-1 mu1B in the human kidney tissue.

Discussion Mutations of SLC4A1 encoding kAE1 are a major cause of hereditary dRTA disease. However, the disease does not result from abnormal anion (Cl-/HCO3-) transport function of kAE1 (3, 6, 29-31). Instead, it is often associated with impaired trafficking of mutant kAE1 protein to function at basolateral membrane of the α-intercalated cells. To date, the sorting machinery and normal process of kAE1 trafficking from TGN to basolateral membrane as well as the trafficking defects associated with dRTA have not clearly been characterized. Recently, our group have reported the binding between kAE1 and AP-1 mu1A, a medium subunit of an adaptor protein complex AP-1, which is a clathrin-associated adaptor protein, involved in kAE1 trafficking (14). Suppression of endogenous AP-1 mu1A in human embryonic kidney (HEK) 293T by siRNA could reduce membrane localization of kAE1, suggesting a role of AP-1 mu1A in the kAE1 trafficking in kidney α-intercalated cells. However, it was observed in our AP-1 mu1A knock-down study that the reduction of cell surface expression of kAE1 was not complete (14), indicating the possible involvement of other sorting or adaptor proteins. This has led us to identify the proteins involved in intracellular sorting and trafficking of kAE1 from TGN to basolateral membrane in kidney cells. We investigated basolateral sorting proteins including mu subunit of four adaptor protein complexes (AP-1 mu1A, AP-1 mu1B, AP-3 mu1, and AP-4 mu1), clathrin heavy chain, and two protein kinases D (PKD1 and PKD2) by knock-down studies using siRNAs. We expressed kAE1 that contains an extracellular HA epitope at position 557, this modification has been reported that it does not affect folding or trafficking of kAE1 (8, 9, 32). Immunoblots showed that kAE1 migrates as two major bands (Figure 1A) indicating that, in MDCK cells, HA-tagged kAE1 was processed to high mannose and complex oligosaccharides. Suppression of AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin heavy chain in MDCK-kAE1-HA cells could cause kAE1 trafficking defect and decrease its basolateral membrane localization (Figures 3 and 4), whereas the reduction of AP-1 mu1B, PKD1, and PKD2 did not affect kAE1 trafficking and localization (Figures 3A and 4). The results of our study using siRNA is agreeable with the findings reported by Almomani et al that siRNA directed against AP-1 mu1A but not against AP-1 mu1B interfered with the surface sorting of kAE1 (15). The role of the AP-1 mu1B complex in kAE1 basolateral sorting was previously examined by expression of wild-type C-terminal kAE1 or kAE1Δ11 (with 11-amino acid deletion) in LLC-PK1 cells, which lack the mu1B subunit (33). Although the AP-1 mu1B complex cannot form in these cells, wild-type kAE1 was correctly targeted to the basolateral membrane (33). AP-1 mu1B has a role in the targeting of some newly synthesized basolateral proteins which exit TGN, by enter recycling endosome (RE) before reaching the cell surface (34). Nevertheless, inactivated RE did not affect cell surface trafficking of newly synthesized kAE1 in MDCK cells (15) indicating that AP-1 mu1B complex is not involved in newly synthesized kAE1 sorting or trafficking. However, the physical interaction between AP-1 mu1B and kAE1 was previously found in MDCK cells which was suggested by Almomani et al that this could be the compensated function for the absence of AP-1 mu1A (15). In our study, supression of AP-1 mu1B did not effect kAE1 trafficking and we did not observe the interaction between kAE1 and AP-1 mu1B indicating that AP-1 mu1B is not responsible for kAE1 sorting. We further investigated whether the incomplete reduction of cell-surface expression of kAE1 may be due to the incomplete knock-down by siRNA or kAE1 traffic through balance action of other adaptor proteins. Double knock-down using siAP-1 mu1A and siAP-3 mu1 or siAP-1 mu1A and siAP-4 mu1 or siAP-1 mu1A and siCLTC were performed. As shown in Figure 3B, double knock-down did not significantly reduced cell-surface expression of kAE1 compared to the knock-down of AP-1 mu1A or CLTC alone, indicating that AP-1 mu1A and clathrin are a major adaptor complex that sorts kAE1 from TGN and knock-down either one or both may similarly affect the whole complex. AP-3 mu1 but not AP-4 epsilon subunit (AP-4 E1) was reported to co-immunoprecipitate with kAE1 (15). Here, we found the interactions between kAE1 and AP-1 mu1A, AP-3 mu1, AP-4 mu1, or clathrin (but not AP-1 mu1B, PKD1 or PKD2) in MDCK and HEK 293T cells by co-immunoprecipitation (Figure 5A) and YFP-PCA (Figure 5B).

This article is protected by copyright. All rights reserved

The different results on the interactions between kAE1 and AP-1 mu1B or AP-4 may be attributable to different subunit of AP-4 detected. The roles of PKD1 and PKD2 were previously reported to be involved in the transport of basolateral cargo in MDCK cells (28). However, the recruitment and activation of PKD at the TGN require the activation of an unidentified G protein–coupled receptor (GPCR) by cargo protein activation of a trimeric G protein (35). kAE1 does not have a binding motif for G protein–coupled receptor. Thus, basolateral localization of kAE1 in MDCK cells was not changed when either PKD1 or PKD2 was suppressed. In post-Golgi trafficking pathways, adaptor protein complexes are essential in cargo protein selection and vesicle formation. The medium subunit of the adaptor protein complexes binds directly to both tyrosine based motif – YXXΦ (in which X represents any amino acid and Φ represents a large hydrophobic residue) and acidic dileucine motif – [DE]XXXL[LI], at least one of which is present in many cargo proteins (18-20). These motifs found to be implicated in basolateral targeting of kAE1 (24). The C-terminus of kAE1 contains a non-classical tyrosine motif (Y904DEV907) in which Y904 is critical for the basolateral trafficking of kAE1 (8, 31, 33). The involvement of dileucine motif at the C-terminus of kAE1 in its basolateral trafficking was also reported (8, 31). To investigate whether human kAE1 naturally co-localize with the basolateral sorting proteins in human kidney tissue or not, we conducted the double immunofluorescence staining of kAE1 and each basolateral sorting protein in the sections of human kidney-tissue. Human kAE1 protein predominantly localizes at basolateral membrane of the tubular epithelial (α-intercalated) cells of collecting duct of nephrons and it also co-localizes with basolateral-sorting proteins including AP-1 mu1A, AP-3 mu1, AP-4 mu1, and clathrin (Figure 6). However, it does not co-localize with AP-1 mu1B in the human kidney-tissue section (Figure 6). These results are in agreement with the knock-down studies in MDCK-kAE1-HA cells, co-immunoprecipitation and YFP-PCA in HEK293T cells as previously described. Taken together, all the results from the present study and from our previous work on the interaction between kAE1 and kinesin family member 3B (KIF3B) in cultured kidney cells (36), which is also necessary for kAE1 trafficking, we propose the model of intracellular sorting and trafficking of human kAE1 from TGN to the plasma membrane of human kidney α-intercalated cells as shown in Figure 7. Based on this model, kAE1 located at the TGN binds to the adaptor protein complex, one of which is AP-1A consisting of AP-1 mu1A, clathrin-associated adaptor protein. The binding occurs at the C-terminus of kAE1 containing the basolateral sorting motifs. Clathrins are then recruited to coat and take the lipid vesicle that carries kAE1 cargo and adapter protein complexes out of TGN. Once the clathrin coated vesicles are released from TGN, clathrins and adapter protein complexes will separate from the vesicle containing kAE1 cargo protein. Adaptor protein complexes have been reported to play a role for basolateral membrane proteins trafficking between the TGN and endosomes (37). Recycling endosome inactivation does not affect newly synthesized kAE1 trafficking, suggesting that kAE1 does not traffic through recycling endosomes to the plasma membrane (15). The TGN-separated vesicle then binds to kinesin (KIF3B) - the motor protein that carries the vesicle and cargo protein to move along microtubule using energy derived from hydrolysis of ATP (38) towards the basolateral membrane. In conclusion, we have demonstrated the roles of adaptor protein complexes (AP-1 mu1A, AP-3 mu1, and AP-4 mu1) and clathrin heavy chain on intracellular sorting and trafficking of human kAE1 by a siRNA knock-down study, which showed a marked reduction of kAE1 on the plasma membrane and its accumulation in the cytoplasm of polarized MDCK cells. The roles of these basolateral sorting proteins on intracellular sorting and trafficking of human kAE1 are well supported by their direct interactions to kAE1 both in cultured HEK293T cells and in the α-intercalated cells in human kidney tissue. Based on the results of our studies, we propose the model of intracellular sorting and trafficking of human kAE1 in the α-intercalated cells in human kidney. For further studies, it would be interesting to examine the roles of adaptor protein complexes by knock-down experiments on mutant kAE1, especially the Δ11 C-terminal deletion, which is mis-sorted to the apical membrane. This will provide the insight into abnormal trafficking process of mutant kAE1.

Materials and Methods Establishment of MDCK cells permanently expressing human kAE1 (MDCK-kAE1-HA) Defective retroviruses containing human kAE1 were generated in HEK 293T cells, which were maintained in complete medium [Dulbecco’s Modified Eagle Medium (DMEM)] (GIBCO) supplemented with 10% fetal bovine 2 serum (GIBCO), 100 Units/ml penicillin and 100 μg/ml streptomycin]. The HEK 293T cells were cultured in 25-cm flask at 37ºC with 5% CO2 and sub-cultured twice per week. On the day before transfection, they were trypsinized 2 by following a standard protocol and were cultured in 25-cm flask with 50% confluence. They were then transfected with three plasmids constructs including pVpack-GP, pVpack-VSVG, and pFB Neo-WT kAE1 HA557 ® (1.5 μg each) by using FuGENE 6 Transfection Reagent (Roche). After incubation for 24-36 h, the viruses in the culture medium were collected and filtered through 0.45 μm filter (Sartorius).

This article is protected by copyright. All rights reserved

2

On the day before infection with the defective retroviruses, MDCK cells were cultured in 25-cm flask with 5070% confluence. In order to make the cells susceptible to retrovirus infection, the cells were added with 1:1,000 of Polybrene (Sigma) and incubated for 2 h. Then, the culture medium containing defective retroviruses prepared as above was added to Polybrene treated MDCK cells and incubated for 24-36 h. After that, the culture medium was removed and replaced with fresh medium containing 0.2 mg/ml G418 (Sigma). The cells were cultured in G418 medium for 1 week and colonies resistant to G418 were isolated for culture and detection of the cells expressing kAE1-HA (MDCK-kAE1-HA cells). kAE1-HA was detected by rabbit anti-HA (Invitrogen) followed by swine anti-rabbit antibody conjugated-horseradish peroxidase (HRP) (DakoCytomation), and β-actin (serving as an internal control) was detected by mouse anti-β-actin antibody (Sigma) followed by rabbit anti-mouse antibody conjugated-HRP (DakoCytomation). Chemiluminescent signals generated by SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) were detected using a G:BOX chemiluminescence imaging system (Syngene). Immunofluorescence staining of polarized MDCK cells MDCK-kAE1-HA cells or wild-type MDCK cells were grown on 24 mm polyester Transwell filter supports (Corning Costar), with a diameter of 12 mm and a pore size of 0.4 μm, for 4 days. The filters were removed from the culture wells, rinsed twice with cold PBS, and fixed in ice-cold methanol for 10 min. After washing twice with PBS, the cells on the filters were permeabilized with 0.2% (v/v) Triton X-100 in PBS for 15 min, washed 3 times in PBS, and blocked with 3% (w/v) BSA/PBS for 30 min prior to incubation with rabbit anti-HA (Invitrogen) for 60 min at room temperature. The cells were then washed three times with PBS and incubated mouse anti-pancadherin (Invitrogen) or goat anti-TGN46 (abcam) or mouse anti-giantin (abcam) for 60 min at room temperature. The cells were then washed and incubated with donkey anti-rabbit IgG conjugated with Cy3 fluoresceine and goat anti-mouse IgG conjugated with Alexa488 fluoresceine or donkey anti-goat IgG conjugated with Alexa-488 fluoresceine (all from Jackson Immunoresearch Laboratories) for 30 min at room temperature follow by Hoechst 33342 (Molecular Probes) in PBS and then washed three times with PBS. The filters were mounted on glass slides with Fluorosave (DakoCytomation). Cellular localization of kAE1-HA was observed by using a laser scanning confocal Zeiss LSM 510 microscope (Carl Zeiss). Reverse transcription and polymerase chain reaction (RT-PCR) To examine mRNA expression of the sorting-protein genes from MDCK-kAE1-HA cells, total RNA was prepared by Trizol reagent (Invitrogen) and chloroform extraction. mRNA was quantified by RT-PCR using a pair of specific primers for each sorting protein gene (Table 1). The PCR reaction was performed by using GoTaq® DNA polymerase (Promega) in a thermocycler (Biometra). Briefly, mRNA (1 μg) were reverse transcribed by using III SuperScript First-Strand Synthesis System (Invitrogen) as specified by the manufacturer. Amplification of cDNA by PCR was carried out in a reaction mixture containing 1 μl of cDNA, 5x colorless GoTaq® reaction buffer (Promega), 10 mM dNTP mix (Promega), 10 pmol of each primer and 0.25U GoTaq® DNA polymerase (Promega). cDNA sample was heated at 95°C for 5 minutes and then subjected to 30 amplification cycles of denaturation at 95°C for 30 seconds, primer annealing at 58°C for 30 seconds, and extension at 72°C for 30 seconds, followed by a final extension at 72°C for 10 minutes. PCR products were visualized after agarose-gel electrophoresis and stained with ethidium bromide solution. The reaction without adding cDNA was usually conducted in parallel as a negative control. RNA interference Small interfering RNA (siRNA) directed against each sorting-protein mRNA was designed by using BLOCK-iTTM RNAi Designer (Invitrogen). The target sequences for siRNA of sorting-protein mRNAs are shown in Table 2. Transfection of each siRNA or siControl was carried out by using Lipofectamine 2000 (Invitrogen) as detailed by the 4 manufacturer. MDCK-kAE1-HA cells (10 cells) were grown on 24 mm polyester Transwell filter supports for 2 days before transfection. The double-stranded siRNA were transfected. Then, the cells were harvested after transfection for 3 days for further examination by immunofluorescence staining and flow cytometry. For double knock-down experiment, MDCK-kAE1-HA cells were transfected with siAP-1 mu1A in combination with siAP-3 or siAP-4 or siCLTC for 3 days. Then, the cells were harvested for examination by flow cytometry. To examine the efficiency of mRNA knock-down by siRNA, total RNA was prepared from the transfected cells by Trizol reagent (Invitrogen) and chloroform extraction. mRNA was quantified by real-time RT-PCR using a pair of ® specific primers as showed in Table 1. The assay was performed by using LightCycler 480 SYBR Green I Master ® (Roche) and a LightCycler 480 Instrument equipped with a 96-well thermal cycler cassette (Roche). Briefly, RNA sample (1 µg) was reverse transcribed by using SuperScriptIII First-Strand Synthesis System (Invitrogen) as specified by the manufacturer. cDNA templates were then added into PCR reaction containing a pair of primers

This article is protected by copyright. All rights reserved

and Taq DNA polymerase. The PCR reaction mixture was subjected to an initial denaturation at 95°C for 10 minutes, followed by 50 cycles of PCR (95°C for 15 seconds, 58°C for 15 seconds, and 72°C for 30 seconds, per cycle). The mRNA levels were normalized against that of human beta-actin (forward primer: 5’AGAAAATCTGGCACCACACC-3’ and reverse primer: 5’-CTCCTTAATGTCACGCACGA–3’) by using the comparative Ct (delta delta Ct) method. The data set of each experiment was generated from triplicate reactions, which was expressed as an average value (mean + SE). The statistical difference of data sets was calculated by Student’s t-test. A p-value

Role of adaptor proteins and clathrin in the trafficking of human kidney anion exchanger 1 (kAE1) to the cell surface.

Kidney anion exchanger 1 (kAE1) plays an important role in acid-base homeostasis by mediating chloride/bicarbornate (Cl-/HCO3-) exchange at the basola...
3MB Sizes 1 Downloads 4 Views