Mutations of the solute carrier family 4 member 1 (SLC4A1) gene encoding kidney anion (chloride/bicarbonate ion) exchanger 1 (kAE1) can cause genetic distal renal tubular acidosis (dRTA). Different SLC4A1 mutations give rise to mutant kAE1 proteins with distinct defects in protein trafficking. The mutant kAE1 protein may be retained in endoplasmic reticulum (ER) or Golgi apparatus, or mis-targeted to the apical membrane, failing to display its function at the baso-lateral membrane. The ER-retained mutant kAE1 interacts with calnexin chaperone protein; disruption of this interaction permits the mutant kAE1 to reach the cell surface and display anion exchange activity. However, the mechanism of Golgi retention of mutant kAE1 G701D protein, which is otherwise functional, is still unclear. In the present study, we show that Golgi retention of kAE1 G701D is due to a stable interaction with the Golgi-resident protein, coat protein complex I (COPI), that plays a role in retrograde vesicular trafficking and Golgi-based quality control. The interaction and co-localization of kAE1 G701D with the γ-COPI subunit were demonstrated in human embryonic kidney (HEK-293T) cells by co-immunoprecipitation and immunofluorescence staining. Small interference RNA (siRNA) silencing of COPI expression in the transfected HEK-293T cells increased the cell surface expression of transgenic kAE1 G701D, as shown by immunofluorescence staining. Our data unveil the molecular mechanism of Golgi retention of kAE1 G701D and suggest that disruption of the COPI-kAE1 G701D interaction could be a therapeutic strategy to treat dRTA caused by this mutant.
A heritable inability to acidify urine is a hallmark of genetic distal renal tubular acidosis (dRTA), a rare disease that is dominantly or recessively inherited . Patients affected by dRTA typically present with metabolic acidosis, hypokalemia, nephrocalcinosis, and failure to thrive in childhood [1–4]. Mutations in solute carrier family 4 member 1 (SLC4A1) gene encoding the kidney anion (chloride/bicarbonate ion) exchanger 1 (kAE1) can cause dRTA [5–7]. The kAE1 protein is expressed at the baso-lateral membrane of α intercalated cells in the cortical and medullary collecting tubules of the nephron, and provides a major exit route for bicarbonate ion (HCO3−) in exchange with chloride ion (Cl−) during the apical secretion of hydrogen ion (H+) into the tubular lumen [8–10]. Newly synthesized kAE1 proteins in endoplasmic reticulum (ER) co-translationally acquire a high-mannose oligosaccharide that is converted into a complex oligosaccharide when the protein reaches the Golgi apparatus [11,12]. The short cytosolic carboxy-terminal domain of kAE1 interacts with adaptor proteins at the trans-Golgi network (TGN). These adaptors proteins play a role in kAE1 trafficking to the baso-lateral membrane [13,14].
kAE1 proteins naturally form dimers at the baso-lateral plasma membrane. In epithelial Madin-Darby canine kidney (MDCK) cells, dominant-dRTA mutant kAE1 proteins can form heterodimers with wild-type kAE1, yet they are predominantly retained in ER or mis-targeted to the apical membrane [15,16]. In contrast, the heterodimeric form of wild-type kAE1 and a recessive-dRTA mutant kAE1 protein can traffic to the baso-lateral membrane [15,17–19]. Interestingly, some dRTA mutant kAE1 proteins are functional, on the basis of observations from patients’ red blood cells or when these mutant kAE1 proteins are expressed in Xenopus oocytes [20–22]. The recessive-dRTA kAE1 G701D mutation is common in Southeast Asia, either in the homozygous state or in the compound heterozygous state with kAE1 Southeast Asian ovalocytosis (SAO), E522K, S773P, or A858D mutations [23–26]. The mutant kAE1 G701D protein is predominantly retained in the Golgi apparatus of renal epithelial cells [18,27]. However, it is functionally detected at the surface of Xenopus oocytes when co-expressed with the erythroid-specific chaperone-like glycophorin A . Previous studies have started to identify strategies to overcome the intra-ER or intra-Golgi retention of partially active mutated membrane proteins. Baso-lateral membrane targeting of ER-retained mutant kAE1 R901X and R589H proteins ectopically expressed in MDCK cells was observed after disruption of their interactions with calnexin, a lectin chaperone . This treatment was, however, ineffective in the case of the mutant kAE1 G701D protein, which is retained in the Golgi apparatus by yet unknown mechanisms.
The quality control systems of the Golgi apparatus depend on proteins that bind to the mutant proteins and impede their traffic to the actual site . Coat protein complex I (COPI) and ER class I α-mannosidase (ERManI) contributes to a Golgi-based quality control module that captures ER-associated protein degradation (ERAD) substrates, such as mis-folded or mutant proteins that escaped from ER, and facilitates their loading into vesicles via the association with the γ-COP subunit (COPG).
Thus, COPI and ERManI form a multifunctional gatekeeper in protein quality control . Worthy of note, the novel Golgi retention motif (KXD/E) of endomembrane protein 12 (EMP12), which localizes in the Golgi apparatus, interacts with COPG, and mutations of this motif release EMP12 to the TGN and downstream vacuolar compartments . Previously, it has been shown that β-COP (the other COPI subunit protein) interacts with the Ca2+-activated chloride channel (CaCC), anoctamin-1 (ANO1), thus impairing the cell surface localization of the latter . Based on these premises, we hypothesized that COPI protein could play a role in the Golgi retention of kAE1 G701D. In the present study, we demonstrate that it is indeed the case. We show that kAE1 is tightly bound to COPI, and that it can be released from Golgi and reach the baso-lateral membrane upon suppression of COPI synthesis.
Materials and methods
pcDNA-kAE1-WT-Myc, containing the 9E10 sequence of Myc epitope (WT, wild type), was constructed by site-directed mutagenesis to insert the sequence TAC CCA TAC GAT GTT CCA GAT TAC GCT between positions 557 and 558 at the third extracellular loop of kAE1. pcDNA-kAE1-WT-Myc was used as a template to generate pcDNA-kAE1-G701D-Myc by site-directed mutagenesis. Our previous work demonstrated that the presence of Myc epitope between positions 557 and 558 of kAE1 does not alter its trafficking as compared with the native mutant protein .
pcDNA3.1/Zeo-YFP-Zip and pcDNA3.1/Zeo-YFP-Zip, encoding two separate fragments of yellow fluorescent protein (YFP) fused with leucine zipper protein inserted between the NotI/XbaI sites of the pcDNA3.1/Zeo vectors (Invitrogen, Waltham, MA, USA), were used for fluorescence complementation assay. The plasmids bear the sequences encoding YFP fragments, either YFP – amino acids 1 to 158 – or YFP – amino acids 159 to 240 – respectively, and Zip cDNA encoding the leucine zipper. A sequence encoding flexible 10 amino-acid linker (GGGGS)2 was inserted between the sequences encoding YFP fragment and leucine zipper. The plasmids were a kind gift from Prof. Stephen Michnick, Department of Biochemistry, Faculty of Medicine, University of Montreal, Canada. The cDNA encoding wild-type kAE1 (WT), mutant kAE1 G701D, and adaptor proteins (AP1mu1A, AP1mu1B, AP3mu1, and AP4mu1) were amplified by polymerase chain reaction (PCR) and cloned into pcDNA3.1-YFP-zipper and pcDNA3.1-YFP-zipper constructs, as previously reported . The following plasmids, including pcDNA3.1-YFP-kAE1-WT, pcDNA3.1-YFP-kAE1-G701D, pcDNA3.1-YFP-AP-1 mu1A, pcDNA3.1-YFP-AP-1 mu1B, pcDNA3.1-YFP-AP-3 mu1, and pcDNA3.1-YFP-AP-4 mu1 were then generated. These constructs could express YFP-kAE1-WT, YFP-kAE1-G701D, YFP-AP-1 mu1A, YFP-AP-1 mu1B, YFP-AP-3 mu1, and YFP-AP-4 mu1 fusion proteins, respectively. The insert sequences and their in-frame junction were confirmed by DNA sequencing in all constructs.
Cell culture and transfection
HEK-293T cells were maintained in complete Dulbecco's Modified Eagle Medium (DMEM, Gibco Life Technologies, Waltham, MA, USA) supplemented with 10% fetal bovine serum (Perbio, Cramlington, UK), 100 units/ml penicillin and 100 µg/ml streptomycin. The cells were cultured in 25-cm2 flask at 37°C with 5% CO2 and sub-cultured twice per week, following a standard trypsinization protocol. One day before transfection, the HEK-293T cells were collected by trypsinization and seeded in 6-well plates. The cultured cells were transiently transfected with pcDNA3.1 vector (sham transfection) or its derivative constructs according to the designed experiments by Lipofectamine (Invitrogen, Waltham, MA, USA) transfection method following the manufacture's protocol (Invitrogen). After transfection, the cells were further incubated 36–48 h to allow maximal expression and then used for further analyses.
Human embryonic kidney (HEK-293T) cells were grown in a six-well plate for 24 h. The cells, at 60–70% confluence, were transfected with 1 µg of recombinant plasmids (pcDNA3.1, pcDNA-kAE1-WT-Myc or pcDNA-kAE1-G701D-Myc) and further incubated at 37°C with 5% CO2 for 48 h. Transfected cells were fixed with 4% glutaraldehyde in PBS for 24 h, washed with PBS, and sequentially dehydrated with ethanol (50, 70, 80, 90, 95, and 100%) for 15 min each. After removal of ethanol, the cells were embedded in a 1:1 mixture of epoxy resin and 100% ethanol with gentle shaking in a rotor for 2 h; thereafter, samples were transferred to gelatin capsules filled with 100% embedding epoxy resin and allowed to polymerize in an oven at 45–50°C overnight. The resin blocks were trimmed to pyramid form, and subjected to semi-thin sections of 150–200 nm. The sample blocks were subjected to ultrathin sectioning (60–90 nm) with a diamond knife and the sections were mounted on a nickel grid (250 mesh). Grids were incubated in a drop of 1% aqueous uranyl acetate and lead citrate and then washed with distilled water. The sections were examined by transmission electron microscopy (Tecnai™ G2 20, FEI, Hillsboro, OR, USA).
HEK-293T cells were grown on glass coverslips in a 24-well plate. The cells were transfected with pcDNA-kAE1-WT-Myc, pcDNA-kAE1-G701D-Myc or empty vector as negative control. Two days post-transfection, the cells on the coverslips were washed once with PBS and fixed with 4% paraformaldehyde in PBS at room temperature for 20 min. After washing twice with PBS, the cells were permeabilized in 0.2% Triton X-100 at room temperature for 15 min followed by rinsing twice with PBS. The coverslips were blocked with 1% BSA in PBS for 30 min, then incubated with the primary antibodies for 1 h at 37°C, washed twice with PBS for 5 min, and then incubated with appropriate secondary antibodies for 1 h at 37°C. Finally, the coverslips were washed twice with PBS and mounted on glass slides with 50% glycerol in PBS. Fluorescence images were captured by Zeiss LSM 510 META confocal microscopy. The following primary antibodies were used: rabbit anti-Myc polyclonal antibody (1:500 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA); goat anti-COPG polyclonal antibody (1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA); goat anti-giantin polyclonal antibody (1:500 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA); goat anti-TGN38 polyclonal antibody (1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA). The following secondary antibodies were used: Cy3-conjugated donkey anti-rabbit IgG and Alexa Fluor® 488-conjugated donkey anti-goat (1:1000; Molecular Probes, Waltham, MA, USA).
Co-immunoprecipitation (Co-IP) and immunoblot analysis
Adherent HEK-293T cells were transfected with the plasmid expressing either wild-type kAE1 or kAE1 G701D by the lipofection method. Two days post-transfection, the cells were detached and collected by centrifugation at 3000×g for 5 min. In certain experiments, 15 min before collection, 0.5 M 3,3′-dithiodipropionic acid di(N-hydroxysuccinimide ester) (Sigma Aldrich) dissolved in DMSO was added to the cells . The cells were lysed with 500 µl of IPB+ [1 mM EDTA, 0.5% (v/v), Igepal (Nonidet P-40 detergent), 150 mM NaCl, 0.2% (w/v) bovine serum albumin, 10 mM Tris-HCl, pH 7.5, and protease inhibitors cocktail]. Five micrograms of goat anti-COPG (Santa Cruz Biotechnology, Dallas, TX, USA) were added to the lysate. The mixture was incubated with gentle rotation at 4°C for 6 h, then Protein G Sepharose (GE Healthcare, Chicago, IL, USA) was added and the incubation continued for 24 h. Subsequently, Protein G-Sepharose beads were collected by centrifugation at 7500×g for 5 min and washed thoroughly with washing buffer 1 (0.1 mM EDTA, 0.1% Igepal, 150 mM NaCl, 10 mM Tris-HCl, pH 7.5) and washing buffer 2 (0.1 mM EDTA, 10 mM Tris-HCl, pH 7.5). The protein complexes were eluted with 2× SDS-PAGE sample loading buffer containing 2% (v/v) β-mercaptoethanol and heated at 95°C for 5 min. The immunoprecipitated proteins were separated by SDS-PAGE and transferred to nitrocellulose membrane. The membrane was blocked with 5% skimmed milk, and then incubated with rabbit anti-Myc (Santa Cruz Biotechnology, Dallas, TX, USA), followed by donkey anti-rabbit antibody conjugated-horseradish peroxidase (HRP) (DakoCytomation, Glostrup, Denmark). Chemiluminescence signals were generated by SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific, Waltham, MA, USA). The membrane was exposed to an X-ray film in a cassette and the film was processed through developing reagents. Alternatively, the chemiluminescent bands were revealed by Enhanced Chemiluminescence reagent (ECL; Perkin Elmer) and imaged using the VersaDOC Imaging System (BioRad).
YFP-based protein fragment complementation assay (YFP-PCA)
HEK-293T cells were seeded on sterile glass coverslips in a 24-well plate and allowed to adhere for at least 24 h. The cells were then co-transfected with different pairs of the plasmid constructs, including pcDNA3.1-YFP-zipper and pcDNA3.1-YFP-zipper (as a negative control), pcDNA3.1-YFP-kAE1-WT and pcDNA3.1-YFP-AP-1 mu1A, pcDNA3.1-YFP-kAE1-WT and pcDNA3.1-YFP-AP-1 mu1B, pcDNA3.1-YFP-kAE1-WT and pcDNA3.1-YFP-AP-3 mu1, pcDNA3.1-YFP-kAE1-WT and pcDNA3.1-YFP-AP-4 mu1, pcDNA3.1-YFP-kAE1-G701D and pcDNA3.1-YFP-AP-1 mu1A, pcDNA3.1-YFP-kAE1-G701D and pcDNA3.1-YFP-AP-1 mu1B, pcDNA3.1-YFP-kAE1-G701D and pcDNA3.1-YFP-AP-3 mu1, and pcDNA3.1-YFP-kAE1-G701D and pcDNA3.1-YFP-AP-4 mu1 by using Lipofectamine 2000 (Invitrogen, Waltham, MA, USA) according to the manufacturer's protocol. Two days post-transfection, the cells on the coverslips were washed with PBS and fixed with 4% paraformaldehyde, then mounted with 50% glycerol in PBS. Fluorescence images were captured by Zeiss LSM 510 META confocal microscopy (Carl Zeiss, Oberkochen, Germany).
SMARTpool®: ON-TARGETplus COPG siRNA 5 nmol (Thermo Scientific Dharmacon, Epsom, UK) directed against COPG mRNA was purchased from Dharmacon™ (Chicago, IL, USA). Transfection of siRNA or siControl was carried out by using Lipofectamine 2000 (Invitrogen, Waltham, MA, USA) according to the manufacturer's instructions. HEK-293T cells were allowed to adhere on a 24-well plate for 1 day before transfection. Forty pmol of SMARTpool® siRNA were used for each well. Silencing efficiency was checked by western blotting at day 1, 2, and 3 post-transfection. The cells were examined by co-immunoprecipitation and immunofluorescence staining at day 3 post-transfection.
All data were reproduced in three independent experiments and reported as the mean ± SD. Statistical differences between the groups were tested with an unpaired t-test using StatView version 5.0, and P value less than 0.05 was considered to be statistically significant.
kAE1 G701D localizes in Golgi apparatus of kidney cells
We first confirmed the localization of kAE1 G701D in non-polarized HEK-293T cells at ultrastructural level. The cells were transfected with pcDNA-kAE1-WT-Myc or pcDNA-kAE1-G701D-Myc and processed for electron microscopy. While the parental un-transfected cells and the cells expressing wild-type kAE1 showed a normal morphology of the Golgi apparatus, the cells expressing kAE1 G701D presented with abnormal dilated round-shaped membranes containing electron-dense material in the Golgi area (Figure 1A). We used confocal fluorescence microscopy to see if these abnormal structures referred to the specific accumulation of the mutant protein. To this end, the transfected cells were double-stained with anti-Myc antibody to reveal the transgenic protein and with anti-giantin to label the Golgi apparatus. As shown in Figure 1B, the wild-type kAE1 protein localized at the plasma membrane (upper panels), whereas kAE1 G701D mainly concentrated in a para-nuclear area in the cytoplasm and co-localized with the Golgi marker, giantin (lower panels). The transfected cells were also double-stained with anti-Myc antibody and trans-Golgi network marker, anti-TGN38. As shown in Figure 1C, the wild-type kAE1 protein always localized at the plasma membrane (upper panels), whereas in a large proportion of the cell population kAE1 G701D mainly concentrated in a para-nuclear area in the cytoplasm and co-localized with the TGN38 (lower panels). From these data, we can conclude that the mutant kAE1 G701D protein is not efficiently trafficked to the plasma membrane, in contrast with the wild-type counterpart, and instead it is retained in the Golgi apparatus.
Localization of kAE1 G701D in Golgi apparatus of kidney cells.
Defective interaction between kAE1 G701D and adaptor proteins (AP-1 mu1A, AP-1 mu1B, AP-3 mu1, and AP-4 mu1) in HEK 293T cells
It has been reported that the adaptor proteins (AP-1 mu1A, AP-1 mu1B, AP-3 mu1, and AP-4 mu1) bind to kAE1 at trans-Golgi network (TGN) and play an important role in its post-Golgi trafficking . We reasoned that Golgi retention of the kAE1 G701D protein could be due to its imperfect interaction with these adaptor proteins. We tested this hypothesis by analyzing the co-immunoprecipitates isolated from HEK-293T cells co-transfected with either pcDNA-kAE1-WT-Myc or pcDNA-kAE1-G701D-Myc with the plasmid bearing the cDNA for individual adaptor protein. The anti-Myc immunoprecipitates were eluted and analyzed by SDS-PAGE and western blotting to detect the presence of adaptor proteins. Although endogenous AP-1 mu1A, AP-3 mu1, and AP-4 mu1, but not AP-1 mu1B, were all found in the immunoprecipitate of kAE1-WT-myc, none of these adaptor proteins co-immunoprecipitated with kAE1 G701D-Myc (Figure 2A).
Interaction between kAE1 G701D and adaptor proteins.
The in situ interactions between wild-type kAE1 and adaptor proteins were confirmed by YFP-PCA. HEK-293T cells were co-transfected with the recombinant plasmids YFP-kAE1-WT or YFP-G701D together with either YFP-AP-1 mu1A, YFP-AP-1 mu1B, YFP-AP-3 mu1 or YFP-AP-4 mu1. The YFP fusion proteins were co-expressed in HEK-293T cells and their interactions were demonstrated by intracellular yellow (adjusted as green) fluorescent signals. As a negative control, HEK-293T cells were co-transfected to express each of YFP-kAE1-WT or YFP-kAE1-G701D, as well as YFP and YFP and each of the YFP-adaptor proteins (Figure 2B). Co-transfection of each construct of YFP-kAE1-WT and each of the YFP-adaptor proteins showed intracellular green fluorescent signals, except for YFP-AP-1 mu1B (data not shown). However, green fluorescent signals were not observed in co-transfection of the construct of YFP-kAE1-G701D and any of the YFP-adaptor proteins (Figure 2B). Taken together, the data from co-immunoprecipitation and fluorescent complementation assays consistently demonstrate that the kAE1 G701D mutant protein does not interact with the adaptor proteins that normally assist the wild-type kAE1 to exit the TGN .
kAE1 G701D protein interacts with COPI subunit γ (COPG)
Next, we tested the hypothesis that the kAE1 mutant protein could be retained in the Golgi apparatus through interaction with COPG, which has been shown to play a role in the Golgi retention of EMP12 . To this end, we analyzed by western blotting the presence of endogenous COPG in the anti-COPG and anti-Myc immunoprecipitates from HEK-293T cells transfected with pcDNA-kAE1-WT-Myc or pcDNA-kAE1-G701D-Myc. The data showed that endogenous COPG was immunoprecipitated with anti-COPG antibody, indicating that COPG was pulled down by co-immunoprecipitation (Figure 3A, upper). Endogenous COPG is associated with the kAE1 G701D but not with the kAE1-WT protein (Figure 3A, lower). Co-localization between kAE1 G701D protein and endogenous COPG protein was further assessed by immunofluorescence in HEK-293T cells transfected with pcDNA-kAE1-WT-Myc, pcDNA-kAE1-G701D-Myc (or empty vector as a negative control). As expected, the transgenic kAE1 wild-type protein trafficked to the plasma membrane, and did not show co-localization with COPG. In contrast, the kAE1 G701D protein was largely located in the cytoplasm and co-localized as spotted with COPG protein, suggestive of retention in the Golgi apparatus (Figure 3B).
Interaction between kAE1 G701D and COP I subunit γ (COPG).
COPI subunit γ (COPG) plays an active role in Golgi retention of kAE1 G701D protein
We investigated on the functional significance of the physical interaction between COPG and kAE1 G701D proteins. COPG protein was knocked down with a specific siRNA to determine its possible involvement in the Golgi retention of kAE1 mutant protein. By 48 h post-transfection, 59.9 ± 14.17% of COPG expression was silenced by the specific siRNA, as determined by western blotting analysis (Figures 4A,B). The effect of COPG knock-down on the possible release of kAE1 G701D protein from Golgi and its interaction with adaptor proteins was investigated by co-immunoprecipitation. siCOPG-transfected HEK-293T cells were transfected with the plasmids coding for either kAE1-WT-Myc or kAE1 G701D-Myc. Transfected HEK-293T cells were then lysed and immunoprecipitated with antibodies specific for each adaptor protein. The immmunoprecipitates were examined by western blotting using anti-Myc antibodies. kAE1-WT-myc co-immunoprecipitated with endogenous AP-1 mu1A, AP-3 mu1, and AP-4 mu1, but not AP-1 mu1B. In the siCOPG-treated condition, kAE1 G701D-Myc was co-immunoprecipitated with endogenous AP-1 mu1A, AP-3 mu1, and AP-4 mu1, but not AP-1 mu1B (Figure 4C). The effect of COPG knock-down on kAE1 G701D protein transport to the plasma membrane was then analyzed by immunofluorescence staining. HEK-293T cells were transfected with the siRNA against COPG mRNA for 24 h, and thereafter transfected with pcDNA-kAE1-WT-Myc, pcDNA-kAE1-G701D-Myc or empty vector. HEK-293T cells were grown on sterile glass coverslips and transfected with siRNA against COPG mRNA or siRNA control for 48 h, and thereafter transfected with pcDNA-kAE1-WT-Myc, pcDNA-kAE1-G701D-Myc, or empty vector for an additional 24 h. At the end, the cover slips were processed for immunofluorescence double staining of the transgenic kAE1 protein and of COPG. Representative images shown in Figure 5A demonstrate the presence on the cell surface of the mutant kAE1 G701D in a high percentage of the cells transfected with the siRNA against COPG. Quantification of this effect using ImageJ software (shown in Figure 5B) indicates that the silencing of COPG increased the percentage of cells positive for membrane kAE1 G701D from 14.27 ± 1.55 to 33.97 ± 5.52, which roughly corresponds to doubling the level of surface expression of the protein. It is worth noting that increases of approximately two-fold were obtained from the data analyzed by immunofluorescence staining (Figure 5B).
Efficiency of siRNA knockdown of COPG in HEK-293T cells and the effect of siRNA on the interaction between kAE1 G701D and adaptor proteins.
Effect of siRNA knockdown of COPG on kAE1 G701D subcellular localization.
The renal isoform of the anion exchanger 1 (kAE1), expressed at the baso-lateral membrane of the α (acid)-secreting intercalated cells of the collecting tube of the kidney, mediates the chloride-bicarbonate exchange needed to compensate for urine acidification. Dominant and recessive mutations in its coding gene SLC4A1 that impair the translocation of the protein from ER to the baso-lateral membrane cause dRTA, regardless of the fact that the protein in itself may be functional [17,18,27]. The mutant protein can be retained in the ER or Golgi complex. For instance, dominant-mutant kAE1 R589H and R901X proteins are retained in the ER through the glycoprotein-calnexin binding, from which they can be released by either preventing the glycosylation of kAE1 or by using small drugs able to disrupt the chaperone interaction . Worthy of note, the latter were, however, ineffective on the recessive mutant kAE1 G701D , consistent with the fact that this mutant does not co-localize with calnexin , or with the ER-retained mutant kAE1 C479W . Instead, the human kAE1 G701D ectopically expressed in MDCK cells was shown to co-localize with the Golgi marker, giantin . Interestingly, this mutant was shown to manifest normal functions when co-expressed with glycophorin A in Xenopus oocytes . The mechanisms of Golgi retention of kAE1 G701D in renal cells had not been elucidated so far. In the present study, we provide evidence for an active role of γ-COP in the Golgi retention of kAE1 G701D. We first confirmed that in a non-polarized HEK-293T cell model kAE1 G701D was retained in the Golgi complex and was flooded in the trans-Golgi network, which in fact showed ultrastructural derangements. We also found that, in contrast with the wild-type counterpart, the mutant kAE1 G701D does not interact with the adaptor proteins AP-1 mu1A, AP-1 mu1B, AP-3 mu1, and AP-4 mu1 that play a role in the trafficking of glycoproteins from TGN to the plasma membrane. This fact also indicates that kAE1 does not even reach the TGN. Based on literature search and bioinformatics analysis, we hypothesized that kAE1 G701D was retained in the Golgi complex possibly through its binding to COPG, a subunit of COPI. The latter is known to play a role in the retrograde protein trafficking and to be involved in a Golgi-based quality control module that captures mis-folded or mutant proteins that have escaped the ER. In fact, COPI contributes with ERManI to the establishment of a multifunctional gatekeeper in the mechanism of protein quality control . In support of our hypothesis is the recent finding that COPG can retain EMP12 by interacting with the KXD/E motif . The latter is regarded as a Golgi retention motif, as when it was mutated EMP12 could traffic to TGN and vacuolar compartments . Immunofluorescence demonstrated that kAE1 G701D co-localizes with COPG in the Golgi complex, and co-immunoprecipitation studies confirmed that the two proteins physically interact. Mechanistically, the role of COPG in retaining kAE1 G701D within the Golgi complex was definitively proven by gene silencing experiments. The knock-down of COPG permitted in fact the release of kAE1 from Golgi complex and its subsequent appearance on the plasma membrane. Likely, the retention of kAE1 in the Golgi complex operated by COPG serves to favor the elimination of the mutant protein via proteasome or lysosome [29,34]. Proteins within the Golgi complex are selected into vesicles at the TGN based on transmembrane span length [35,36]. Hence, there is a possibility that the kAE1 G701D mutant protein causes localized mis-folding of the transmembrane around the kAE1 G701D protein, which affects selection to budding vesicles, and then is sensed by the COPI coat. This may due to a specific interaction with a sequence exposed around G701D transmembrane or loop or region of the protein.
We propose the model of COPG-mediated Golgi retention of kAE1 G701D and of its release upon COPG suppression as illustrated in Figure 6. According to this model, wild-type kAE1 is synthesized and trafficked to the baso-lateral membrane of the cell while kAE1 G701D interacts with COPG and is therefore retained in the Golgi complex and eventually degraded by the proteasome. Disruption of this interaction through COPG knock-down allows the mutant kAE1 protein to reach the baso-lateral membrane, where it can perform its function. This finding paves the avenue to possible interventions with drugs able to disrupt such interaction and, therefore, treatment of dRTA.
Proposed model for COPG-mediated retention of kAE1 G701D in the Golgi apparatus.
coat protein complex I
distal renal tubular acidosis
endomembrane protein 12
ER class I α-mannosidase
human embryonic kidney cells
kidney anion exchanger 1
Madin-Darby canine kidney
small interference RNA
solute carrier family 4 member 1
yellow fluorescent protein
YFP-based protein fragment complementation assay
N.D. performed the sample preparation for all experiments, performed the experiments, and data analysis. M.J. supervised development of work, data interpretation, and drafted the manuscript. S.P. performed the co-immunoprecipitation and data interpretation. N.S. constructed the plasmids. A.C. performed the co-immunoprecipitation and data interpretation. K.C. performed analysis on electron microscopy and interpreted data. T.L. supervised development of work and data interpretation. C.I. supervised development of work, data interpretation, and manuscript evaluation, and P.Y. supervised development of work, data interpretation, and manuscript evaluation.
This work was financially supported by Mahidol University and The Thailand Research Fund [IRG5980006]. P.Y. and M.J. are supported by Chaloemprakiat Grant, Faculty of Medicine Siriraj Hospital. N.D. is supported by a scholarship from the Office of Higher Education Commission (OHEC), Thailand. M.J. is supported by the TRF Grant for New Researcher [TRG5780173].
We thank Prof. Stephen Michnick, Department of Biochemistry, Faculty of Medicine, University of Montreal, Canada for providing plasmids pcDNA3.1/Zeo-YFP-Zip and pcDNA3.1/Zeo-YFP-Zip.
The Authors declare that there are no competing interests associated with the manuscript.
These authors contributed equally to this work.