MAGI-1a Functions as a Scaffolding Protein for the Distal Renal Tubular Basolateral K+ Channels
2008; Elsevier BV; Volume: 283; Issue: 18 Linguagem: Inglês
10.1074/jbc.m707738200
ISSN1083-351X
AutoresMasayuki Tanemoto, Takafumi Toyohara, Takaaki Abe, Sadayoshi Ito,
Tópico(s)Plant Stress Responses and Tolerance
ResumoAs the K+ recycling pathway for renal Na+ reabsorption, renal tubular K+ channels participate in the fluid and electrolyte homeostasis. Previously, we showed that the Kir5.1/Kir4.1 heteromer, which is a heteromeric assembly of two inwardly rectifying K+ channels, composes the principal basolateral K+ channels in distal renal tubules and that two motifs in the carboxyl-terminal portion of the Kir4.1 subunit regulate its functional expression. In this study, by using yeast two-hybrid screening, we identified a new isoform of membrane-associated guanylate kinase with inverted domain structure 1 (MAGI-1a-long) as a scaffolding protein for the basolateral K+ channels. MAGI-1a-long interacted with the PSD-95/Dlg/ZO-1 (PDZ)-binding motif of Kir4.1 by its fifth PDZ domain, and a high salt diet, which could suppress mineralocorticoid secretion, facilitated the interaction. The phosphorylation of serine 377 in the PDZ-binding motif disrupted the interaction, and the disruption of the interaction altered the intracellular localization of the channels from the basolateral side to perinuclear components. These results demonstrate that the phosphorylation-dependent scaffolding of the basolateral K+ channels by MAGI-1a-long participates in the renal regulation of the fluid and electrolyte homeostasis. As the K+ recycling pathway for renal Na+ reabsorption, renal tubular K+ channels participate in the fluid and electrolyte homeostasis. Previously, we showed that the Kir5.1/Kir4.1 heteromer, which is a heteromeric assembly of two inwardly rectifying K+ channels, composes the principal basolateral K+ channels in distal renal tubules and that two motifs in the carboxyl-terminal portion of the Kir4.1 subunit regulate its functional expression. In this study, by using yeast two-hybrid screening, we identified a new isoform of membrane-associated guanylate kinase with inverted domain structure 1 (MAGI-1a-long) as a scaffolding protein for the basolateral K+ channels. MAGI-1a-long interacted with the PSD-95/Dlg/ZO-1 (PDZ)-binding motif of Kir4.1 by its fifth PDZ domain, and a high salt diet, which could suppress mineralocorticoid secretion, facilitated the interaction. The phosphorylation of serine 377 in the PDZ-binding motif disrupted the interaction, and the disruption of the interaction altered the intracellular localization of the channels from the basolateral side to perinuclear components. These results demonstrate that the phosphorylation-dependent scaffolding of the basolateral K+ channels by MAGI-1a-long participates in the renal regulation of the fluid and electrolyte homeostasis. Kidney is the essential organ for the fluid and electrolyte homeostasis, and the derangement of its function results in life-threatening diseases, including hypertension, the most common disease in industrialized societies (1Wang Y. Wang Q.J. Arch. Intern. Med. 2004; 164: 2126-2134Crossref PubMed Scopus (464) Google Scholar, 2Psaltopoulou T. Orfanos P. Naska A. Lenas D. Trichopoulos D. Trichopoulou A. Int. J. Epidemiol. 2004; 33: 1345-1352Crossref PubMed Scopus (125) Google Scholar, 3Gu D. Reynolds K. Wu X. Chen J. Duan X. Muntner P. Huang G. Reynolds R.F. Su S. Whelton P.K. He J. Hypertension. 2002; 40: 920-927Crossref PubMed Scopus (502) Google Scholar). The Mendelian form abnormalities have provided clues to pathophysiology of many diseases, and recent genetic analysis revealed that the Na+ reabsorption in distal renal tubules participates in the pathogenesis of hypertension (4Wilson F.H. Disse-Nicodeme S. Choate K.A. Ishikawa K. Nelson-Williams C. Desitter I. Gunel M. Milford D.V. Lipkin G.W. Achard J.M. Feely M.P. Dussol B. Berland Y. Unwin R.J. Mayan H. Simon D.B. Farfel Z. Jeunemaitre X. Lifton R.P. Science. 2001; 293: 1107-1112Crossref PubMed Scopus (1217) Google Scholar, 5Lifton R.P. Gharavi A.G. Geller D.S. Cell. 2001; 104: 545-556Abstract Full Text Full Text PDF PubMed Scopus (1364) Google Scholar). The process of the Na+ reabsorption in distal renal tubules is the result of coordinated electrolyte transports across the renal epithelia (5Lifton R.P. Gharavi A.G. Geller D.S. Cell. 2001; 104: 545-556Abstract Full Text Full Text PDF PubMed Scopus (1364) Google Scholar, 6Meneton P. Loffing J. Warnock D.G. Am. J. Physiol. 2004; 287: F593-F601Crossref PubMed Scopus (158) Google Scholar). In the epithelia, Na+ efflux occurs actively via basolateral Na+/K+-ATPases, whereas apical pathways passively mediate Na+ influx. Different pathways mediate the apical Na+ influx in each segment of the tubules as follows: Na+-K+-2Cl- cotransporters in the thick ascending limbs of Henle's loop (TAL), 2The abbreviations used are: TAL, thick ascending limb; DCT, distal convoluted tubule; PDZ, PSD-95/Dlg/ZO-1; MDCK, Madin-Darby canine kidney; HEK, human embryonic kidney; FITC, fluorescein isothiocyanate; GST, glutathione S-transferase; PKA, protein kinase A; RT, reverse transcription; Ni-NTA, nickel-nitrilotriacetic acid; WT, wild type; CT, carboxyl terminus; GFP, green fluorescent protein. 2The abbreviations used are: TAL, thick ascending limb; DCT, distal convoluted tubule; PDZ, PSD-95/Dlg/ZO-1; MDCK, Madin-Darby canine kidney; HEK, human embryonic kidney; FITC, fluorescein isothiocyanate; GST, glutathione S-transferase; PKA, protein kinase A; RT, reverse transcription; Ni-NTA, nickel-nitrilotriacetic acid; WT, wild type; CT, carboxyl terminus; GFP, green fluorescent protein. Na+-Cl- cotransporters in the distal convoluted tubules (DCT), and Na+ channels in the connecting tubules and the collecting ducts. Because Na+/K+-ATPases induce intracellular K+ influx with simultaneous Na+ efflux in these segments, K+ excretion, i.e. K+ recycling, is essential for the continuance of Na+ reabsorption (7Giebisch G. Kidney Int. 2001; 60: 436-445Abstract Full Text Full Text PDF PubMed Scopus (50) Google Scholar). Previously, we showed that the Kir5.1/Kir4.1 heteromer, which is a heteromeric assembly of two inwardly rectifying K+ channels, composes the principal pathway for the basolateral K+ recycling in the distal portion of TAL and the DCT (8Tanemoto M. Kittaka N. Inanobe A. Kurachi Y. J. Physiol. (Lond.). 2000; 525: 587-592Crossref Scopus (130) Google Scholar, 9Lourdel S. Paulais M. Cluzeaud F. Bens M. Tanemoto M. Kurachi Y. Vandewalle A. Teulon J. J. Physiol. (Lond.). 2002; 538: 391-404Crossref Scopus (140) Google Scholar, 10Tanemoto M. Abe T. Onogawa T. Ito S. Am. J. Physiol. 2004; 287: F1148-F1153Crossref PubMed Scopus (29) Google Scholar). The functional expression of this heteromer is regulated by the following two motifs in the carboxyl-terminal (CT) portion of the Kir4.1 subunit: (i) dihydrophobic motif for functional cell-surface expression and (ii) PSD-95/Dlg/ZO-1 (PDZ)-binding motif for basolateral localization (11Tanemoto M. Abe T. Ito S. J. Am. Soc. Nephrol. 2005; 16: 2608-2614Crossref PubMed Scopus (26) Google Scholar). Therefore, the renal tubules are believed to regulate the basolateral K+ recycling by the mechanism that recognizes these motifs, and the mechanism is a key factor for the renal regulation of Na+ and K+ homeostasis (6Meneton P. Loffing J. Warnock D.G. Am. J. Physiol. 2004; 287: F593-F601Crossref PubMed Scopus (158) Google Scholar, 12Brown D. Breton S. Kidney Int. 2000; 57: 816-824Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar). This study reports that an isoform of membrane-associated guanylate kinase with inverted domain structure 1 (MAGI-1a) functions as a scaffolding protein for the channels that compose the basolateral K+ recycling pathway. Yeast Two-hybrid Screening and Cloning—The CT of rat Kir4.1, which is a fragment of 105 amino acid residues, was fused to a GAL4 DNA-binding domain, and 500,000 clones of a rat kidney cDNA library were screened by using the Matchmaker GAL4 two-hybrid system (Takara Bio Clontech) according to the manufacturer's instructions. The sequence of several clones showed high homology to the sequence of the CT domain of mouse MAGI-1a. The full-length of rat MAGI-1a isoforms, MAGI-1a-long and MAGI-1a-short, was cloned by PCR from rat kidney cDNA with the primer pairs of 5′-ATGTCGAAAGTGATCCAGAA-3′ and 5′-TCATGGAGTCATGCCAGGGAAGG-3′. For expression analysis of MAGI-1a isoforms in rat tissues, we performed PCR on Rat MTC Panel I (Clontech) using the primer pairs of 5′-ATGTCGAAAGTGATCCAGAA-3′ and 5′-CCGAGGGTCTAACCATGATG-3′. Construct of Fusion Proteins and Deletion/Point Mutants—Mammalian cell expression vectors that contain green fluorescent protein (GFP)-tagged Kir4.1 (GFP-Kir4.1), its deletion/point mutants, and MAGI-1a isoforms were constructed as described previously (11Tanemoto M. Abe T. Ito S. J. Am. Soc. Nephrol. 2005; 16: 2608-2614Crossref PubMed Scopus (26) Google Scholar). Polyhistidine-tagged CT81 segments with and without mutation were constructed by subcloning PCR-amplified DNA fragments into the expression vector pcDNA4/HisMaxC (Invitrogen). Transient Expression in Mammalian Cell Lines—Human embryonic kidney (HEK)293T cells and Madin-Darby canine kidney (MDCK) cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (both from Invitrogen) and were transfected with the expression vectors by using Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol. MDCK cells were plated on polycarbonate Millicell transwell filters (Millipore, Bedford, MA) prior to transfection. The further analysis of expressed proteins was usually conducted at 48–72 h after transfection as described previously (13Tanemoto M. Fujita A. Higashi K. Kurachi Y. Neuron. 2002; 34: 387-397Abstract Full Text Full Text PDF PubMed Scopus (57) Google Scholar). Antibodies—A polyclonal anti-MAGI-1 antibody was raised in rabbits against the synthetic peptide SFTADSGDQDEPTLQEATL that corresponds to amino acids 258–267 of MAGI-1a-long (amino acids 39–58 of MAGI-1a-short). A polyclonal anti-K+-channel Kir4.1 antibody (Sigma), BD Living Colors™ Av peptide antibodies (Takara Bio Clontech), Alexa Fluor 594 anti-rabbit IgG (Fab′)2 (Molecular Probes, Eugene, OR), and fluorescein isothiocyanate (FITC)-labeled anti-rabbit IgG (Dako, Glostrup, Denmark) were purchased. Protein Precipitation and Immunoblotting Analysis—Protein precipitation and immunoblotting analysis were performed as described previously (8Tanemoto M. Kittaka N. Inanobe A. Kurachi Y. J. Physiol. (Lond.). 2000; 525: 587-592Crossref Scopus (130) Google Scholar). The antibody-pretreated protein A-Sepharose (Amersham Biosciences) was used for immunoprecipitation. The protein A-Sepharose pretreated with preimmune rabbit IgG was used for negative control. Ni-NTA-agarose (Qiagen, Hilden, Germany) was used for polyhistidine-tagged fragments precipitation. The precipitated proteins were analyzed by immunoblotting and detected by SuperSignal West Dura Extended Duration Substrate (Pierce). The intensity of the detected bands was calculated using NIH Image software (National Institutes of Health, Bethesda). Immunohistochemical Analysis—Adult Sprague-Dawley rats were anesthetized in accordance with the regulations of the Animal Care Committee of our institute, and then perfused transcardially with 0.9% saline followed by perfusion with 4% paraformaldehyde in 0.1 m sodium phosphate (pH 7.4). After dehydration, the kidney was embedded in paraffin, and thin sections of 3–5-μm thickness were obtained. After deparaffinization and rehydration, the sections were immunostained and then observed using a confocal microscope (model LSM 5 PASCAL, Carl Zeiss Co., Ltd., Jena, Germany) as described previously (10Tanemoto M. Abe T. Onogawa T. Ito S. Am. J. Physiol. 2004; 287: F1148-F1153Crossref PubMed Scopus (29) Google Scholar). In brief, after incubation in blocking buffer containing 5% normal goat serum, the sections were incubated with the anti-Kir4.1 antibody followed by incubation with an excess (1:20 dilution) of Alexa Fluor 594 anti-rabbit IgG (Fab′)2 to saturate the epitopes of the primary antibody. After being washed extensively with phosphate-buffered saline, the sections were subsequently incubated with the anti-MAGI-1 antibody followed by incubation with FITC-labeled anti-rabbit IgG at 1:100 dilution. The sections were observed by a confocal microscope after extensive washing with phosphate-buffered saline. The saturation of the first primary antibody was confirmed by the preliminary experiment that did not detect any labeling with FITC by subsequent incubation with control rabbit IgG. Salt Loading for Animals—At 8 weeks of age, the diet of rats was switched from a 0.3% NaCl to an 8% NaCl (high salt)-containing diet for 2 weeks. Control rats were fed with a 0.3% NaCl diet throughout the experiments. Glutathione S-Transferase (GST) Pulldown Analysis—GST fusion proteins of several CT segments of MAGI-1a were constructed by subcloning PCR-amplified DNA fragments into the bacterial expression vector pGEX-5X3 (Amersham Biosciences), and each fusion protein was purified on glutathione-Sepharose™ 4B (Amersham Biosciences) according to the manufacturer's protocol. GFP-Kir4.1 resolved as described above was incubated with these purified GST fusion proteins and then analyzed by immunoblotting as described previously (13Tanemoto M. Fujita A. Higashi K. Kurachi Y. Neuron. 2002; 34: 387-397Abstract Full Text Full Text PDF PubMed Scopus (57) Google Scholar). Protein Kinase A (PKA) Phosphorylation—Experiments involving PKA stimulation were performed using a cAMP mixture containing 5 μm forskolin, 100 μm 8-Br-cAMP, and 100 μm 3-isobutyl-1-methylxanthine. The cAMP mixture was applied for 10 min until just before further preparation. For the inhibition of PKA, a PKA-specific blocker, N-(2([3-(4-bromophenyl)-2-propenyl]amino)-ethyl)-5-iso-quinolinesulfonamide (H89) (Seikagaku Corp., Tokyo, Japan) was applied at a concentration of 1.0 μm from 3 min prior the application of the mixture (14Chijiwa T. Mishima A. Hagiwara M. Sano M. Hayashi K. Inoue T. Naito K. Toshioka T. Hidaka H. J. Biol. Chem. 1990; 265: 5267-5272Abstract Full Text PDF PubMed Google Scholar). Electrophysiological Recordings—Channel activity was analyzed with the patch clamp method in whole-cell configuration as described previously (8Tanemoto M. Kittaka N. Inanobe A. Kurachi Y. J. Physiol. (Lond.). 2000; 525: 587-592Crossref Scopus (130) Google Scholar). In brief, the resistance of patch electrodes, when filled with the pipette solution (140 mm KCl, 1 mm MgCl2, 1 mm CaCl2 and 5 mm HEPES-KOH (pH 7.4)) was adjusted to 1–2 megohms. Currents were elicited on GFP-Kir4.1-transfected HEK293T cells by voltage steps in the bath solution containing 120 mm NaCl, 20 mm KCl, 5 mm EGTA, 2 mm MgCl2, and 5 mm HEPES-KOH (pH 7.3) using a patch clamp amplifier (Axopatch 200B, Axon Instruments, Union City, CA). Data were analyzed by pCLAMP 9 (Axon Instruments), and barium (3 mm)-sensitive components of the currents were recorded as K+ currents. Reverse Transcription (RT)-PCR Analysis—Total RNA was extracted from MDCK cells using TRIzol reagent (Invitrogen) and reverse-transcribed with (dT)12–18 primer by using Superscript II RT (Invitrogen). MAGI-1a expression was detected by PCR using a primer pair 5′-CCAGTAATTGGGAAATCACACC-3′ and 5′-CCGCCTCAGAAACAGACGGAC-3′, which spans introns to distinguish these sequences from the genomic DNA. The PCR products were analyzed by agarose gel electrophoresis and visualized by staining with ethidium bromide. Cytological Observation—Cytological observation was performed as described previously (11Tanemoto M. Abe T. Ito S. J. Am. Soc. Nephrol. 2005; 16: 2608-2614Crossref PubMed Scopus (26) Google Scholar). Especially in experiments with MDCK cells, formation of tight junction between cells after confluent growth was confirmed by the expression of a typical protein of tight junction, ZO-1, on the top of a lateral wall. Immunostaining was performed in phosphate-buffered saline containing 0.05% Triton X-100, 5% bovine serum albumin, and 1% normal goat serum. Statistical Analysis—The intensity of the bands in immunoblotting analysis was expressed as means ± S.D. Comparison was performed by the paired two-tailed Student's t test. Probability values of p < 0.01 were considered to be statistically significant. Identification of MAGI-1a as a Scaffolding Protein for Kir4.1 in the Kidney—Two isoforms of MAGI-1a were identified from rat kidney cDNA (supplemental Fig. S1). The longer one, named MAGI-1a-long (GenBank™ accession number AY598952), had two short inserts to the reported MAGI-1a (before and after the second WW domain). The shorter one, named MAGI-1a-short (GenBank™ accession number AY598951), had two splice-out portions from MAGI-1a-long (in the amino terminus and between PDZ2 and PDZ3) (Fig. 1A and supplemental Fig. S1). As a consequence of the splice-out, MAGI-1a-short lacked the amino-terminal PDZ domain (so-called PDZ0) and the guanylate kinase domain (Fig. 1B). Analysis by PCR on rat tissue cDNA revealed that the kidney expressed both MAGI-1a-long and MAGI-1a-short, and MAGI-1a-short was expressed only in the kidney among the tissues examined (Fig. 1C and supplemental Fig. S2). Expression of MAGI-1a-long-Kir4.1 Complex in Renal Distal Tubules—Using the anti-MAGI antibody, which specifically recognized MAGI-1a-short and MAGI-1a-long expressed in HEK293T cells (Fig. 2A), we confirmed the expression of both MAGI-1a isoforms in the kidney (Fig. 2B). The antibody detected two bands in the lysate from the membrane fraction of rat kidney (Fig. 2B, left panel), the identical size of MAGI-1a-short (lower band) and the predicted size of MAGI-1a-long (upper band). The antibody pre-absorption by the antigenic peptide reduced the detection and almost eliminated these bands (Fig. 2B, right panel). Using the antibody, we next examined the intrarenal interaction between MAGI-1a isoforms and Kir4.1. In the immunoprecipitant with the anti-Kir4.1 antibody, a single band was detected from the rat kidney lysate by the anti-MAGI-1 antibody (left panel in Fig. 2C). The size of the band indicated that the isoform coprecipitated with Kir4.1 was MAGI-1a-long. The immunoprecipitant with the anti-MAGI-1 antibody also contained Kir4.1 (right panel in Fig. 2C). The mutual coprecipitation of MAGI-1a-long and Kir4.1 indicated their intrarenal interaction. The intrarenal MAGI-1a-long/Kir4.1 interaction was further supported by immunohistochemical analysis (Fig. 3A). The immunoreactivity against the anti-MAGI-1 antibody was detected in the tubules that expressed Kir4.1, and coimmunostaining with the anti-Kir4.1 antibody showed the colocalization of MAGI-1a isoforms with Kir4.1 on the basolateral side in these tubules. The expression of MAGI-1a isoforms on the basolateral side was confirmed by the single staining with the anti-MAGI-1 antibody (supplemental Fig. S3). Effects of High Salt Diet on Intrarenal MAGI-1a-long/Kir4.1 Interaction—We next examined the effect of high salt diet (salt load), which could suppress mineralocorticoid secretion, on the intrarenal MAGI-1a-long/Kir4.1 interaction (Fig. 3B). The salt load increased the interaction, although it did not increase the amount of Kir4.1 expressed in the kidney. More Kir4.1 was coprecipitated with the same amount of MAGI-1a-long in the kidney of the salt-loaded animals than the control (relative amount: 1.65 ± 0.17, p = 0.005). Interacting Domains of MAGI-1a-long and Kir4.1—Using GST pulldown, we analyzed the regions for the MAGI-1a-long/Kir4.1 interaction. The CT domain of MAGI-1a-long that was identified by the yeast two-hybrid screening could interact with Kir4.1 (Fig. 4A). Sequential deletion of the domain from the CT end showed that the deletion of the fifth PDZ domain (PDZ5) disrupted the interaction, which indicates that the PDZ5 domain is essential for MAGI-1a-long to interact with Kir4.1. The deletion of the PDZ-binding motif in Kir4.1 disrupted the MAGI-1a-long/Kir4.1 interaction (Fig. 4B). A single mutation of serine 377, the critical residue in the motif, to alanine (S377A) or aspartic acid (S377D) also disrupted the interaction. These results indicated that the PDZ-binding motif of Kir4.1 was essential for the MAGI-1a-long/Kir4.1 interaction and confirmed that the interaction was PDZ-dependent. Regulation of the MAGI-1a-long/Kir4.1 Interaction—The PDZ dependence of the MAGI-1a-long/Kir4.1 interaction was further confirmed in HEK293T cells. In HEK293T cells, as well as in the GST pulldown, the Kir4.1 with wild-type (WT) CT portion could interact with MAGI-1a-long, but the deletion of the PDZ-binding motif in Kir4.1 disrupted the mutual interaction between MAGI-1a-long and Kir4.1 (Fig. 5A). In the GST pulldown analysis, the induction of a phosphorylated state-mimicking mutation in serine 377 (S377D) disrupted the MAGI-1a-long/Kir4.1 interaction, which indicated that serine 377 phosphorylation affected the interaction. Because serine 377 of Kir4.1 is a putative site for phosphorylation by PKA, we further examined the effects of PKA phosphorylation on the MAGI-1a-long/Kir4.1 interaction in HEK293T cells (Fig. 5B). The interaction was disrupted during 10 min of incubation with the cAMP mixture, whereas a PKA-specific inhibitor, H89, overcame this disruption. Compared with the amount of Kir4.1 coprecipitated with MAGI-1a-long after simultaneous inhibition with H89, the amount after PKA stimulation was significantly low (0.13 ± 0.03, p < 0.001). Interestingly, the current amplitude of Kir4.1 assessed by patch clamp did not change during the 10-min cAMP mixture application, which sufficiently disrupted the PDZ-dependent interaction with MAGI-1a-long; the amplitude after the 10-min application relative to that before the application was 0.96 ± 0.05 (Fig. 5C). Phosphorylation of Serine 377 in the PDZ-binding Motif—We further examined PKA phosphorylation on the CT81 amino acid residues of Kir4.1 (CT81 segment) (Fig. 6A). Whereas Ni-NTA-agarose precipitated the same amount of the WT and S377A polyhistidine-tagged CT81 segments, the WT segment instead of the S377A segment was more efficiently serine-phosphorylated by the cAMP mixture application in the H89-inhibitory manner. Compared with the intensity of phosphoserine on the WT segment, the intensity on the S377A segment was statistically significantly weak (0.25 ± 0.19, p = 0.004), and the reduction in the intensity by simultaneous H89 inhibition was significant only in the WT segment (WT, 0.29 ± 0.05, p < 0.001; S377A, 0.15 ± 0.05, p = 0.305). These results indicated that the serine 377 in the PDZ-binding motif was the residue that was PKA-phosphorylated. The motif with unphosphorylated but not phosphorylated serine 377 could interact with the PDZ5 of MAGI-1a (schematically summarized in Fig. 6B). In the absence of extra activation for PKA, the intensity of phosphoserine on the WT segment changed widely in each experiment (ranging from 0.33 to 1.13). However, the intensity without the extra activation was generally weaker than the intensity with the extra activation (0.72 ± 0.34, p = 0.201). Effect of Kir4.1 Phosphorylation on Intracellular Localization—We further examined the effect of phosphorylation on the intracellular localization of Kir4.1 in renal tubular cells by using a cell line derived from the renal tubule, MDCK cells. RT-PCR analysis by using a specific primer pair for MAGI-1 revealed the expression of MAGI-1 isoform(s) in MDCK cells (Fig. 7A). In MDCK cells, Kir4.1 showed the predominant localization on the basolateral side, but the induction of a phosphorylation-mimicking single mutation at serine 377 (S377D) induced the perinuclear localization (Fig. 7B). These results indicated that the PDZ-dependent interaction with MAGI-1 isoform(s), which is disrupted by the phosphorylation of serine 377 in the PDZ-binding motif, participated in the basolateral localization of Kir4.1 in renal tubular cells. In this study, we revealed that a scaffolding protein, MAGI-1a-long, interacts with Kir4.1 on the basolateral side of distal renal tubules. The interaction is affected by dietary amount of NaCl and regulated by phosphorylation of the PDZ-binding motif of Kir4.1. Previously, we showed that Kir4.1 is a subunit of the basolateral K+ channels (a pathway of basolateral K+ recycling for Na+ reabsorption) in distal renal tubules (8Tanemoto M. Kittaka N. Inanobe A. Kurachi Y. J. Physiol. (Lond.). 2000; 525: 587-592Crossref Scopus (130) Google Scholar, 9Lourdel S. Paulais M. Cluzeaud F. Bens M. Tanemoto M. Kurachi Y. Vandewalle A. Teulon J. J. Physiol. (Lond.). 2002; 538: 391-404Crossref Scopus (140) Google Scholar), and that the PDZ-binding motif of this subunit participates in the proper expression of the channels (10Tanemoto M. Abe T. Onogawa T. Ito S. Am. J. Physiol. 2004; 287: F1148-F1153Crossref PubMed Scopus (29) Google Scholar, 11Tanemoto M. Abe T. Ito S. J. Am. Soc. Nephrol. 2005; 16: 2608-2614Crossref PubMed Scopus (26) Google Scholar). Therefore, our findings in this study indicate that MAGI-1a-long participates in the basolateral localization of the K+ channels in these tubules; this is similar to the PDZ-dependent localization of other membrane proteins by scaffolding proteins (15Fanning A.S. Anderson J.M. Curr. Opin. Cell Biol. 1999; 11: 432-439Crossref PubMed Scopus (274) Google Scholar). We detected the expression of MAGI-1 isoforms in the renal tubules but not in the glomerulus. The intrarenal interaction of MAGI-1a-long with Kir4.1, which is not expressed in the glomerulus but in distal renal tubules (10Tanemoto M. Abe T. Onogawa T. Ito S. Am. J. Physiol. 2004; 287: F1148-F1153Crossref PubMed Scopus (29) Google Scholar, 16Ito M. Inanobe A. Horio Y. Hibino H. Isomoto S. Ito H. Mori K. Tonosaki A. Tomoike H. Kurachi Y. FEBS Lett. 1996; 388: 11-15Crossref PubMed Scopus (115) Google Scholar), confirmed the tubular expression of MAGI-1 isoforms. In a previous report, another MAGI-1-specific antibody also detected tubular but not intraglomerular expression of MAGI-1 (17Laura R.P. Ross S. Koeppen H. Lasky L.A. Exp. Cell Res. 2002; 275: 155-170Crossref PubMed Scopus (77) Google Scholar). However, there are several reports that show intraglomerular expression of MAGI-1 (18Patrie K.M. Drescher A.J. Goyal M. Wiggins R.C. Margolis B. J. Am. Soc. Nephrol. 2001; 12: 667-677PubMed Google Scholar, 19Nishimura W. Iizuka T. Hirabayashi S. Tanaka N. Hata Y. J. Cell Physiol. 2000; 185: 358-365Crossref PubMed Google Scholar, 20Hirabayashi S. Mori H. Kansaku A. Kurihara H. Sakai T. Shimizu F. Kawachi H. Hata Y. Lab. Investig. 2005; 85: 1528-1543Crossref PubMed Scopus (56) Google Scholar). In these previous reports, the antibodies that were raised against other members of scaffolding proteins (19Nishimura W. Iizuka T. Hirabayashi S. Tanaka N. Hata Y. J. Cell Physiol. 2000; 185: 358-365Crossref PubMed Google Scholar, 20Hirabayashi S. Mori H. Kansaku A. Kurihara H. Sakai T. Shimizu F. Kawachi H. Hata Y. Lab. Investig. 2005; 85: 1528-1543Crossref PubMed Scopus (56) Google Scholar) or that recognized several proteins with sizes different from MAGI-1 (18Patrie K.M. Drescher A.J. Goyal M. Wiggins R.C. Margolis B. J. Am. Soc. Nephrol. 2001; 12: 667-677PubMed Google Scholar) were used, because they could detect MAGI-1. Therefore, intraglomerular immunoreactivity with these antibodies could be against some other proteins, such as a synaptic scaffolding molecule, but not MAGI-1 (21Kawajiri A. Itoh N. Fukata M. Nakagawa M. Yamaga M. Iwamatsu A. Kaibuchi K. Biochem. Biophys. Res. Commun. 2000; 273: 712-717Crossref PubMed Scopus (53) Google Scholar). It is also possible that isoform(s) of MAGI-1 without the amino-terminal portion are expressed in the glomerulus, because we raised the anti-MAGI-1 antibody against an amino acid sequence on the amino-terminal portion of MAGI-1a in this study. In the renal tubules, the MAGI-1 expression was detected on the luminal side in addition to the basolateral side where MAGI-1a-long was colocalized with Kir4.1. Because MAGI-1a-short is the renal MAGI-1 isoform that did not interact with Kir4.1, it would be the isoform expressed on the luminal side. Although we could not clarify the precise localization of the isoforms, they probably have different intracellular localizations, i.e. MAGI-1a-short on the luminal side and MAGI-1a-long on the basolateral side. The PDZ0 and guanylate kinase domain, the domains that only MAGI-1a-long contains, might be responsible for the different localization (15Fanning A.S. Anderson J.M. Curr. Opin. Cell Biol. 1999; 11: 432-439Crossref PubMed Scopus (274) Google Scholar). The different localization would make it possible for MAGI-1a-long but not MAGI-1a-short to interact with Kir4.1 in vivo in the kidney. The mechanism for in vivo interaction of Kir4.1 with the basolaterally expressed MAGI-1a-long but not the luminally expressed MAGI-1 isoforms has not been clarified. However, we propose two possible mechanisms. (i) PDZ5 of the luminally expressed MAGI-1 isoforms is preoccupied in vivo by other proteins (18Patrie K.M. Drescher A.J. Goyal M. Wiggins R.C. Margolis B. J. Am. Soc. Nephrol. 2001; 12: 667-677PubMed Google Scholar, 22Dobrosotskaya I.Y. James G.L. Biochem. Biophys. Res. Commun. 2000; 270: 903-909Crossref PubMed Scopus (138) Google Scholar). (ii) The basolateral K+ channels are specifically delivered to the basolateral side by the mechanism that is not clarified yet (12Brown D. Breton S. Kidney Int. 2000; 57: 816-824Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar, 23Rothman J.E. Nature. 1994; 372: 55-63Crossref PubMed Scopus (2003) Google Scholar). The MAGI-1a-long/Kir4.1 interaction was disrupted by phosphorylation. Phosphorylation is thought to take place on the serine 377 in the PDZ-binding motif, and only the PDZ-binding motif with the unphosphorylated serine 377 could interact with the PDZ5 of MAGI-1a-long (Fig. 6B). Either the phosphorylation or the amino acid mutation (to alanine or aspartic acid, which mimics the phosphorylated serine (24Beguin P. Nagashima K. Nishimura M. Gonoi T. Seino S. EMBO J. 1999; 18: 4722-4732Crossref PubMed Scopus (147) Google Scholar, 25Lin Y.F. Jan Y.N. Jan L.Y. EMBO J. 2000; 19: 942-955Crossref PubMed Scopus (95) Google Scholar)) of the serine 377 disrupted the interaction. In line with our findings, phosphorylation of PDZ-binding motifs in other proteins also changes their interaction with scaffolding proteins and affects their intracellular localization (26Daniels D.L. Cohen A.R. Anderson J.M. Brunger A.T. Nat. Struct. Biol. 1998; 5: 317-325Crossref PubMed Scopus (161) Google Scholar, 27Songyang Z. Fanning A.S. Fu C. Xu J. Marfatia S.M. Chishti A.H. Crompton A. Chan A.C. Anderson J.M. Cantley L.C. Science. 1997; 275: 73-77Crossref PubMed Scopus (1219) Google Scholar). Interestingly, although the PKA phosphorylation disrupted the MAGI-1a-long/Kir4.1 interaction, it did not change the channel activity of Kir4.1 for a short duration. We previously showed that the PDZ-dependent interaction with scaffolding protein(s) facilitates the localization of the K+ channels on the intracellular compartments just beneath the cell surface but not on the extracellular surface (10Tanemoto M. Abe T. Onogawa T. Ito S. Am. J. Physiol. 2004; 287: F1148-F1153Crossref PubMed Scopus (29) Google Scholar, 11Tanemoto M. Abe T. Ito S. J. Am. Soc. Nephrol. 2005; 16: 2608-2614Crossref PubMed Scopus (26) Google Scholar). Because of this intracellular localization, the PKA phosphorylation would not change the channel activity for a short duration. The phosphorylation would regulate the process of intracellular localization that affects the channels ready to be expressed on the cell surface. The channels on the compartments just beneath the cell surface could be functional on demand (7Giebisch G. Kidney Int. 2001; 60: 436-445Abstract Full Text Full Text PDF PubMed Scopus (50) Google Scholar). Serine residues in the CT portion of Kir4.1 other than serine 377, such as serine 343, could be PKA-phosphorylated (28Pearson R.B. Kemp B.E. Methods Enzymol. 1991; 200: 62-81Crossref PubMed Scopus (869) Google Scholar), and might affect the interaction. Less phosphorylation on a S377A-mutated segment under PKA inhibition than PKA stimulation might reflect PKA phosphorylation of these residues (14Chijiwa T. Mishima A. Hagiwara M. Sano M. Hayashi K. Inoue T. Naito K. Toshioka T. Hidaka H. J. Biol. Chem. 1990; 265: 5267-5272Abstract Full Text PDF PubMed Google Scholar). However, the difference is too small (p = 0.305) to be considered as the different phosphorylation state of any residues, and serine 377 in the PDZ-binding motif is thought to be the residue that is responsible for the phosphorylation-dependent regulation of the interaction. In this study, the activity of PKA, which would have changed widely in vitro according to the condition of the cells without extra activation for it, disrupted the MAGI-1a-long/Kir4.1 interaction in HEK293T cells. Therefore, the intracellular signal cascades that affect the activity of PKA in vivo could regulate the activity of the basolateral K+ recycling in distal renal tubules. However, in distal renal tubules, several kinases that can phosphorylate serine residues, such as SGK1, are known to be expressed (29Alvarez de la Rosa D. Coric T. Todorovic N. Shao D. Wang T. Canessa C.M. J. Physiol. (Lond.). 2003; 551: 455-466Crossref Scopus (47) Google Scholar, 30Vitari A.C. Deak M. Morrice N.A. Alessi D.R. Biochem. J. 2005; 391: 17-24Crossref PubMed Scopus (395) Google Scholar). These kinases might phosphorylate Kir4.1 more efficiently under the conditions that are suitable for their activity. The phosphorylation by these kinases would affect the activity of the basolateral K+ recycling and regulate the renal Na+ reabsorption in vivo in the kidney (29Alvarez de la Rosa D. Coric T. Todorovic N. Shao D. Wang T. Canessa C.M. J. Physiol. (Lond.). 2003; 551: 455-466Crossref Scopus (47) Google Scholar, 30Vitari A.C. Deak M. Morrice N.A. Alessi D.R. Biochem. J. 2005; 391: 17-24Crossref PubMed Scopus (395) Google Scholar). Supporting this notion, in this study, the animal fed with the diet that could reduce the activity of SGK1 (31Hou J. Speirs H.J. Seckl J.R. Brown R.W. J. Am. Soc. Nephrol. 2002; 13: 1190-1198Crossref PubMed Scopus (44) Google Scholar) had the increased intrarenal MAGI-1a-long/Kir4.1 interaction. Although the results of this study suggest the MAGI-1 isoform(s) participate in the localization of the basolateral K+ channel in distal renal tubules and MDCK cells, other scaffolding proteins that contain PDZ domain(s) might participate in the localization. On the basolateral side of MDCK cells and mouse distal renal tubules, the dystrophin-associated protein complex that contains several scaffolding proteins with PDZ domain(s), including α-syntrophin, is reported to exist (32Loh N.Y. Newey S.E. Davies K.E. Blake D.J. J. Cell Sci. 2000; 113: 2715-2724PubMed Google Scholar, 33Kachinsky A.M. Froehner S.C. Milgram S.L. J. Cell Biol. 1999; 145: 391-402Crossref PubMed Scopus (97) Google Scholar). Because α-syntrophin has the PDZ domain that can interact with the PDZ-binding motif of Kir4.1 (34Connors N.C. Adams M.E. Froehner S.C. Kofuji P. J. Biol. Chem. 2004; 279: 28387-28392Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar), these proteins might also contribute to the localization of the basolateral K+ channels in distal renal tubules. The contribution of these scaffolding proteins for the localization of the K+ channels remains to be clarified by their direct interaction in distal renal tubules. In conclusion, MAGI-1a-long functions as a scaffolding protein for the distal renal tubular basolateral K+ channels (Fig. 8). Phosphorylation of the channels impedes the channel scaffolding by MAGI-1a-long and could decrease the basolateral K+ recycling. These findings suggest that the regulatory pathway for the activity of kinases in renal tubules participates in the distal renal tubular Na+/K+ regulation. Download .zip (.32 MB) Help with zip files
Referência(s)