SLC2A9 gene is reported to have 2 splice variants: short isoform (SLC2A9-S) with 512 amino acids and long isoform (SLC2A9-L) with 540 amino acids. Their C-terminal 490 residues are identical and only difference lies at their N-termini. SLC2A9-S and SLC2A9-L have different N-termini of 21 and 50 amino acids, respectively (Fig. 1A). To detect specific expression of SLC2A9-S and SLC2A9-L, antibodies against these different amino acid residues were produced. Western blotting was performed with these antibodies to check the specificity (Fig. 2A–D). COS cells were transfected with mock, SLC2A9-S and SLC2A9-L, and cell lysate was blotted with anti-SLC2A9-S (Fig. 2A) or anti-SLC2A9-L (Fig. 2C) antibodies. These antibodies specifically detected each SLC2A9 isoform without any cross reactivity. The signals corresponding to SLC2A9s were disappeared when Western blotting was performed with antibodies pre-absorbed with immunogens (Fig. 2B and 2D). Immunofluorescent staining was also used to confirm the specificity of these antibodies (Fig. 2E–P). Green fluorescence protein (GFP) tagged SLC2A9-S or SLC2A9-L was transfected into COS cells and cells were stained with anti-SLC2A9-S (Fig. 2E–J) or anti-SLC2A9-L (Fig. 2K–P) antibodies. Anti-SLC2A9-S and anti-SLC2A9-L antibodies did not recognize the COS cells expressing SLC2A9-L (Fig. 2 I) and SLC2A9-S (Fig. 2L), respectively. In contrast, anti-SLC2A9-S and anti-SLC2A9-L antibodies specifically detect COS cells transfected with SLC2A9-S (Fig. 2F) and SLC2A9-L (Fig. 2O), respectively.

Human kidney sections were stained with these antibodies together with fluorescein-Lotus Tetragonolobus lectin (LTL) or rhodamine - Dolichos Biflorus agglutinin (DBA) (Fig. 3); LTL specifically binds to renal proximal tubules and DBA specifically binds to collecting ducts. SLC2A9-S was expressed at the apical membrane of collecting ducts, which were stained with DBA (Fig. 3D–F). In contrast, SLC2A9-L was expressed at the basolateral membrane of proximal tubules, which were positive for LTL (Fig. 3G–J). Some PECAM-1 (Platelet Endothelial Cell Adhesion Molecule-1) positive endothelial cells expressed SLC2A9-S or -L (data not shown). Only weak background signal was obtained when immunostaining was performed using these antibodies pre-incubated with their antigen peptides (data not shown).

In the human kidney, SLC2A9-S was expressed at the apical membrane, whereas SLC2A9-L was expressed at the basolateral membrane in the different renal tubules. This sorting pattern is consistent with the previous report using MDCK cells [7]. However, the paper failed to report the mechanism of differential sorting of these 2 isoforms. Here, we pursued this question in detail. GFP tagged SLC2A9-S or -L was transfected into polarized MDCK cells (MDCK type II). To identify apical or basolateral membrane, cell surface was labeled with biotin. When GFP tagged SLC2A9-S was expressed in MDCK cells, it was distributed in both apical and basolateral membranes (Fig. 4A). Some part of SLC2A9-S was localized in the intracellular compartment. On the other hand, SLC2A9-L was expressed only at the basolateral membrane of MDCK cells (Fig. 4B). These results indicate that intracellular trafficking of SLC2A9 in polarized cells is determined by its N-terminal residues, only differences between SLC2A9-S and -L. Interestingly, SLC2A9-S localized differently in human renal tubule cells and cultured MDCK cells.

SLC2A9-S was expressed at both apical and basolateral membranes in MDCK cells (Fig. 4A). To reveal the role of SLC2A9-S N-terminus on this expression pattern, amino acids were deleted stepwise by 4 residues and its sorting was observed in MDCK cells. Removal of the N-terminal 4∼16 amino acids from SLC2A9-S did not affect its sorting and these deletion mutants were localized both in the apical and basolateral membranes similar to SLC2A9-S wild-type. Localization of the shortest deletion mutant, Short_16AA (Fig. 1B), in MDCK cells was presented in Fig. 5. Short_16AA was expressed both in the apical and basolateral membranes. As noted with SLC2A9-S wild-type, some of Short_16AA was localized at the intracellular compartment. Removal of the N-terminal 20 amino acids unique to SLC2A9-S resulted in very low protein expression (data not shown). These results suggest that N-terminal of SLC2A9-S does not contain an apico-basolateral sorting signal recognized by MDCK cells unlike renal collecting duct cells. N-terminal amino acids unique to SLC2A9-S may be important for protein expression/stability at least in MDCK cells.

Di-leucine (LL) motif is often critical for trafficking of membrane proteins such as basolateral targeting and cell surface/intracellular recycling (endocytosis). SLC2A9-L has a putative di-leucine motif at 33th and 34th leucine and it may be important for basolateral sorting of SLC2A9-L. This LL is an obvious candidate that regulates basolateral targeting of SLC2A9-L in the polarized epithelial cell. To evaluate the role of the LL, 2 mutant constructs were made: Long_33-34A, in which the LL was substituted with alanine-alanine, and Long_32-36del, in which 5 amino acids around the LL were deleted. When these mutants were expressed in MDCK cells, they were targeted to the basolateral membrane similar to SLC2A9-L wild-type (Fig. 6). These results indicate that the LL sequence in SLC2A9-L does not work as a typical di-leucine motif and neither LL sequence nor its surrounding amino acids are critical for its basolateral sorting.

To identify the important amino acid residues of SLC2A9-L for its basolateral membrane sorting, several N-terminal deletion mutants were made (Fig. 1). N-terminus of SLC2A9-L was deleted stepwise by 5 amino acids and those mutants were transfected into MDCK cells. When 5, 10, 15, 20 or 25 amino acids were removed from SLC2A9-L N- terminal, these mutants went to the basolateral membrane similar to SLC2A9-L wild-type. Localization of Long_25AAdel, in which 25 amino acids from SLC2A9-L N-terminal were removed, was shown in Fig. 7A. Removal of the N-terminal 30 amino acids from SLC2A9-L (Long_30AAdel), however, resulted in both apical and basolateral membrane expression (Fig. 7B) similar to the localization of SLC2A9-S. Based on these results, amino acids 27^th–31^st of SLC2A9-L (PGPGR) appeared to be important in basolateral targeting of this transporter. To test this hypothesis, alanine substitution and deletion mutant constructs were made. To our surprise, Long_27-31A mutant, in which 27^th–31^st amino acids were replaced with alanine, was expressed exclusively at the basolateral membrane (Fig. 7C). Deletion of those five amino acids (Long_27-31 del) did not affect its basolateral localization (Fig. 7D), either.

When 21 amino acids from 30th to 50th were deleted, the deletion construct, Long_30-50del, was expressed only at the basolateral membrane similar to SLC2A9-L wild-type (Fig. 8A). On the other hand, when 26 amino acids from 25th to 50th were deleted, the deletion construct, Long_25-50del, was localized both at the apical and basolateral membranes (Fig. 8B). The difference between Long_30-50del and Long_25-50del is PPGPG sequence (26th–30th), which was also removed in Long_30AAdel, which was localized both in the apical and the basolateral membrane (Fig. 7B) However we have already shown that PGPGR (27th–31st) was dispensable for basolateral trafficking of SLC2A9-L (Fig. 7, C and D). It is possible to argue that 26^th amino acid proline is important, but N-terminal deletion mutant upto the 26^th proline, Long_25AAdel, was sorted correctly to the basolateral membrane (Fig. 7A). Taken together, N-terminal amino acids unique to SLC2A9-L was important in its basolateral membrane targeting. Unfortunately, we were not able to determine which part(s) of these 50 amino acids are critical for the sorting mechanism.

We noticed that SLC2A9-L mutants with large deletion in the N-terminal amino acids were not only targeted to both apical and basolateral membranes, but also localized in the intracellular compartment. To identify which compartment they were expressed, MDCK cells expressing SLC2A9-L, Long_25AAdel or Long_30AAdel, were stained with intracellular organellar markers, calnexin (ER), golgin 97 (Golgi) and lysotracker (lysosome). Almost all SLC2A9-L and Long_25AAdel were localized at the basolateral cell membrane without co-localizing with any intracellular markers tested (Fig. 9A and B). In contrast, Long_30AAdel, which went to the intracellular compartment, was co-stained with lysotracker (Fig. 9C).

This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of Ministry of Education, Culture, Sports, Science and Technology-JAPAN. The protocol was approved by Institute of Laboratory Animals, Graduate School of Medicine, Kyorin University (The Experimental Animal Ethics Committee in Kyorin University, permit number 90).

To generate antibodies which would detect SLC2A9-S and SLC2A9-L differently, we used N-termini of SLC2A9, which have different amino acid sequences (Fig. 1A), as immunogens. pGEX constructs including the different segment of SLC2A9s were transformed into E.coli BL-21. The expression of GST fusion protein was induced with 0.1 mM IPTG and a protein extract was prepared with lysozyme treatment, freeze-thaw cycle and sonication in PBS. The extract was incubated with Glutathione Sepharose™ 4B beads (GE Healthcare Bio-Sciences Corp. Piscataway, NJ) for 6 hrs at 4°C. After incubation, these beads were washed 4 times with PBS and eluted with glutathione elute buffer (50 mM Tris-HCl, 10 mM Reduced Glutathione, pH 8.0). Eluted proteins were dialyzed with PBS and injected into rabbits together with TiterMax Gold Adjuvant (Sigma-Aldrich, St. Louis, MO). After 5th injection of the antigen, serum was collected.

cDNAs of SLC2A9-S and SLC2A9-L were purchased from OriGene Technologies (Rockville, MD) and Fisher Scientific (Pittsburgh PA), respectively. To construct GFP fusion proteins, SLC2A9s were amplified by polymerase chain reaction (PCR). This construct was subcloned as a EcoRI/KpnI fragment into the pEGFP-N2 vector (Clontech Laboratories, Mountain View, CA), in which the insert was fused to the amino terminus of GFP. To construct GST fusion proteins for producing antibodies, SLC2A9s were amplified by PCR with primers that included EcoRI and XhoI restriction sites. The PCR fragment was subcloned into pGEX-6P-1 vector, in which the insert was fused to the carboxyl terminus of glutathione S-transferase. All PCR primer sequences are available on request.

COS-7 cells and MDCK cells (American Type Culture Collection (ATCC), Manassas, VA) were cultured in a humidified incubator under 5% CO₂ in Dulbecco’s Modified Eagle’s Medium (Sigma-Aldrich, St. Louis, MO) supplemented with 10% FBS, 2 mM L-glutamine, 50 U/mL penicillin and 50 µg/ml streptomycin. DNA transfection into COS cells or MDCK cells was performed with Lipofectamine 2000 (Life Technologies, Carlsbad, CA) according to the manufacturer’s instructions. Two days after transfection, following assays were performed.

Reactivity of produced antibodies were examined by Western blotting and immunofluorescence. For Western blotting analysis, transfected COS cells were incubated with lysis buffer containing 1% Triton X-100, 150 mM NaCl, 5 mM EDTA, and 25 mM Tris-HCl, pH 7.4 for 30 min at 4°C. Insoluble material was removed through centrifugation at 10,000 g for 30 min at 4°C. After centrifugation, 30µg of protein were mixed with SDS-PAGE sample buffer. The samples were separated by SDS-PAGE and analyzed by Western blotting. For antigen pre-absorption test, antibodies were incubated with GST fusion proteins and Glutathione Sepharose™ 4B beads (GE Healthcare Bio-Sciences Corp. Piscataway, NJ). Overnight incubation at 4°C, supernatant was used for Western blotting. For immunofluorescence, COS cells were seeded on cover slips before transfection. Two days after transfection, cells were fixed with cold methanol and immunofluorescence was performed with anti-SLC2A9 antibodies as shown in immunohistochemistry part.

To examine the distribution of SLC2A9 within the kidney, human tissue sections purchased from BioChain Institute, Inc. (Newark, CA) were stained with TSA Fluorescence System (PerkinElmer, Inc., Waltham, MA) according to the manufacturer’s instructions with minor modifications [26]. In brief, sections were blocked with TNB buffer (0.1 M Tris-HCl, pH 7.5, 0.15 M NaCl and 0.5% Blocking Reagent) containing 5% normal donkey serum. To block endogenous biotin, the sections were further treated with the Avidin–Biotin block (Vector Laboratories, Burlingame, CA), and endogenous peroxidase activity was quenched with 1% H₂O₂. The primary antibody was applied overnight at 4°C. Slides were washed with TNT Wash Buffer (0.1 M Tris-HCl, pH 7.5, 0.15 M NaCl, and 0.05% Tween20) and incubated with biotin-conjugated donkey anti–rabbit IgG (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA), followed by incubation with streptavidin–HRP conjugate (Life Technologies, Carlsbad, CA) together with 1 µg/ml DAPI and 10 µg/ml fluorescein-Lotus Tetragonolobus lectin or 10 µg/ml rhodamine-Dolichos Biflorus agglutinin (Life Technologies, Carlsbad, CA). The antigens were detected using tyramide-FITC or tyramide-Cy3. Sections were mounted in Fluorescence Mounting Medium (Dako, Glostrup, Denmark). Fluorescence was visualized with an Olympus Fluoview FV1000 laser confocal microscope. Images are the product of 3-fold line averaging. Contrast and brightness settings were chosen so that all pixels were within the linear range.

Two days after transfection, MDCK cells were pated on the cell culture inserts (BD, Franklin Lake, NJ) in the presence of 1 mg/ml G418 and were grown to confluence. Cells were washed in ice-cold PBS⁺⁺ (PBS supplemented with 0.1 mM CaCl₂ and 1 mM MgCl₂) and incubated with 2.5 mM Sulfo-NHS-SS-biotin (Thermo Fisher Scientific, Rockford, IL) in PBS⁺⁺ for 30 min. Excess biotin was quenched 3×5 min with 100 mM glycine in PBS⁺⁺. Cell surface membrane was visualized with Alexa Fluor® 594 conjugated streptavidin (Life Technologies, Carlsbad, CA). Cells were mounted in Fluorescence Mounting Medium (Dako, Glostrup, Denmark). Fluorescence was visualized with an Olympus Fluoview FV1000 laser confocal microscope. Images are the product of 3-fold line averaging. Contrast and brightness settings were chosen so that all pixels were within the linear range.

Two days after transfection, MDCK cells were pated on the cell culture inserts (BD, Franklin Lake, NJ) in the presence of 1 mg/ml G418 and were grown to confluence on the cell culture inserts. Cells were incubated with 75 nM LysoTracker (Life Technologies, Carlsbad, CA), in growing medium for 2 hrs. Samples were washed 3 times with PBS⁺⁺ (PBS supplemented with 0.1 mM CaCl₂ and 1 mM MgCl₂) and fixed with 4% paraformaldehyde, stained with DAPI. Cells were washed with PBS and mounted in Fluorescence Mounting Medium (Dako, Glostrup, Denmark). For detection of ER and Golgi, transfected MDCK cells were fixed with cold methanol and immunofluorescence was performed with anti-calnexin antibody (Enzo Life Sciences, Farmingdale, NY) or anti-golgin97 antibody (BD Biosciences, San Jose, CA) as shown in immunohistochemistry part.