点击显示 收起
【摘要】 TRPV4, a renally expressed nonselective cation channel of the transient receptor potential (TRP) family, is gated by hypotonicity. Kinases of the WNK family influence expression and function of the thiazide-sensitive Na + -Cl - cotransporter, and monogenic human hypertension has been linked to mutations in the gene coding for WNK4. Along with TRPV4, WNK isoforms are highly expressed in the distal nephron. We show here that coexpression of WNK4 downregulates TRPV4 function in human embryonic kidney (HEK-293) cells and that this effect is mediated via decreased cell surface expression of TRPV4; total abundance of TRPV4 in whole cell lysates is unaffected. The effect of the related kinase WNK1 on TRPV4 function and surface expression was similar to that of WNK4. Disease-causing point mutations in WNK4 abrogate, but do not eliminate, the inhibitory effect on TRPV4 function. In contrast to wild-type WNK4, a kinase-dead WNK4 point mutant failed to influence TRPV4 trafficking; however, deletion of the entire WNK4 kinase domain did not blunt the effect of WNK4 on localization of TRPV4. Deletion of the extreme COOH-terminal putative coiled-coil domain of WNK4 abolished its effect. In immunoprecipitation experiments, we were unable to detect direct interaction between TRPV4 and either WNK kinase. In aggregate, these data indicate that TRPV4 is functionally regulated by WNK family kinases at the level of cell surface expression. Because TRPV4 and WNK kinases are coexpressed in the distal nephron in vivo and because there is a tendency toward hypercalcemia in TRPV4 -/- mice, we speculate that this pathway may impact systemic Ca 2+ balance. In addition, because WNK kinases and TRPV4 are activated by anisotonicity, they may comprise elements of an osmosensing or osmotically responsive signal transduction cascade in the distal nephron.
【关键词】 hypertension calcium balance hypotonicity osmoregulation
THE TRANSIENT RECEPTOR POTENTIAL (TRP) channels represent a large family of cation channels regulated by diverse afferent inputs; a subset of TRP channels respond to environmental stimuli such as temperature and tonicity ( 6 ). Although TRPV4 [also known as VR-OAC ( 18 ), OTRPC4 ( 28 ), VRL-2 ( 10 ), and TRP12 ( 37 )] had been described in other contexts, its functional role was established when it was identified as the mammalian homolog of the Caenorhabditis elegans osmosensory protein OSM9 ( 18, 20, 28 ). On the basis of its expression in the blood-brain barrier-deficient osmosensing nuclei of the hypothalamus ( 18 ), in conjunction with the abnormal water metabolism exhibited by mice harboring targeted deletions of the TRPV4 gene ( 19, 24 ), the channel is believed to be instrumental in physiological or pathophysiological regulation of systemic water balance. However, in addition to a role in systemic osmoregulation, TRPV4 also plays a role in osmoregulation at the cellular level. In human airway epithelial cells ( 3 ) and Chinese hamster ovary cells ( 4 ), TRPV4 expression was essential for the regulatory volume decrease that followed hypotonic cell swelling. TRPV4 also likely plays a direct role in kidney physiology. TRPV4 mRNA was expressed most abundantly in kidney, primarily in kidney cortex ( 10, 18, 28, 37 ). TRPV4 protein is abundantly expressed along the nephron distal to the junction of the descending and ascending thin limb of Henle's loop, with the sole exception of the cells of the macula densa ( 29 ); this distribution is consistent with a role in local osmosensing ( 7 ).
The WNK family of kinases was described by Xu et al. ( 38 ) in a screen for novel mitogen-activated protein (MAP)/extracellular signal-regulated protein kinase (ERK) kinase (MEK) family members in rat brain. Interest in these putative kinases increased when mutations in the WNK1 and WNK4 genes were causally linked to severe hypertension ( 35 ) in a subset of patients diagnosed with familial hyperkalemia and hypertension (also known as Gordon's syndrome or pseudohypoaldosteronism type II). Because of the clinical and biochemical abnormalities associated with this disease, aberrant regulation of the thiazide-sensitive Na + -Cl - cotransporter of the distal convoluted tubule was suspected. Yang et al. ( 43 ) and Wilson et al. ( 36 ) subsequently showed that WNK4 downregulates expression of the thiazide-sensitive Na + -Cl - cotransporter, and some, but not all, disease-causing point mutants did not. In contrast, WNK1 exhibited no effect with respect to thiazide-sensitive Na + -Cl - cotransporter function; however, the kinase blunted the inhibitory effect of WNK4 ( 43 ). Biochemically, WNK kinases are activated by a variety of stimuli, including hypertonicity and hypotonicity ( 17, 38 ). Because WNK kinases and TRPV4 are osmotically responsive and expressed in the distal nephron ( 35, 41 ), among other sites, we speculated that WNKs may participate in the regulation of this cation channel.
METHODS
Cell surface biotinylation and immunoblotting. Protein extracts (20 µg) were used for immunoblotting with polyclonal rabbit anti-TRPV4 ( 41 ) at 1:1,000 dilution as previously described ( 41 ). The secondary antibody was goat anti-rabbit horseradish peroxidase at 1:4,000 dilution, and visualization was via Chemiluminescence Plus reagent (PerkinElmer Life Science). For anti-V5 immunoblotting, primary antibody dilution of 1:5,000 was used; secondary antibody was goat anti-mouse horseradish peroxidase at 1:6,000 dilution. For cell surface biotinylation 48 h after transient transfection, human embryonic kidney (HEK-293) cell monolayers were washed three times with ice-cold PBS, incubated with 0.5 mg/ml Sulfo-NHS-LC-Biotin (Pierce Biotechnology, Rockford, IL) for 30 min at 4°C, quenched by incubation with 100 mM glycine in ice-cold PBS for 30 min at 4°C, and then washed three times with ice-cold PBS. Monolayers were lysed in lysis buffer [125 mM NaCl, 50 mM Tris (pH 7.5), 0.1% SDS, 0.5% sodium deoxycholate, 1% Nonidet P-40, 1 mM sodium orthovanadate, 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride, 1 µg/ml leupeptin, 1 µg/ml aprotinin, 1 µg/ml pepstatin A, 25 mM -glycerophosphate, and 2 mM sodium pyrophosphate] for 30 min at 4°C. The protein concentrations were determined by the Bradford method (Bio-Rad). ImmunoPure streptavidin beads (40 µl; Pierce Biotechnology) were added to 3 mg of biotinylated protein, and the mixture was incubated at 4°C for 4 h. The beads were washed five times with ice-cold PBS and eluted with 1 x SDS sample buffer. The eluted proteins were immunoblotted with anti-V5 antibody as described above.
Transient transfection and fura 2 ratiometry. HEK-293 cells were transiently transfected with Lipofectamine PLUS (Life Technologies) in accordance with the manufacturer's directions using 15 µl of PLUS reagent, 30 µl of Lipofectamine, and 10 µg of plasmid DNA reagent per 100-mm dish of cells. After 48 h, the cells were harvested, washed, and resuspended in 10 ml of Hanks' balanced salt solution [HBSS; 130 mM NaCl, 4.7 mM KCl, 1.25 mM CaCl 2, 1.18 mM MgSO 4, 5 mM glucose, and 15 mM HEPES (pH 7.5)] supplemented with 2 µM fura 2-AM and 100 µl of 2% Pluronic F-127 (20% stock solution in DMSO; Molecular Probes, Eugene, OR) per 100-mm dish and then incubated for 45 min at 37°C. The cells were pelleted at 1,000 g for 5 min at 25°C, resuspended with 1-2 ml of HBSS to achieve final concentration of 2-8 x 10 7 cells/ml, and maintained on ice for 30 min. The suspended fura 2-loaded cells (50 µl) were assayed for intracellular Ca 2+ concentration in a cuvette filled with prewarmed (37°C) HBSS (in the presence or absence of extracellular Ca 2+ ) under constant gentle stirring (2-ml final volume) as previously reported ( 41 ). Fluorescent emission was monitored at 510 nm and recorded at 1-s intervals in the presence of alternating excitation at 340 and 380 nm using a fluorescence spectrophotometer (model F-2500, Hitachi Instruments, Naperville, IL). Calibration of the fura 2 signal was performed as previously described ( 27 ) using a fura 2·Ca 2+ dissociation constant of 224 nM ( 11 ). For each experiment, data from three separate cuvettes of treated cells were averaged; experiments were repeated at least three times. Murine WNK4 and rat WNK1 cDNAs were the kind gifts of Dr. David Ellison and Dr. Chao-Ling Yang, and Dr. Melanie Cobb, respectively. Rat WNK1 protein (NP_446246 ) diverges from mouse (NP_941992 ) and human (NP_061852 ) WNK1, in that it lacks 247 amino acids spanning residues 792-1038 of the canonical human WNK1 sequence. After alignment with the human WNK1 gene (NC_000012 ; nucleotides 732993-888219), this corresponds precisely to the absence of exons 11 and 12; alternative splicing of these exons in mRNA from rat kidney has been described elsewhere ( 25 ). Nonetheless, all full-length rat WNK1 protein sequences returned via National Center for Biotechnology Information protein search ( http://www.ncbi.nlm.nih.gov/ ) similarly lack these two exons (Q9JIH7, NP_446246 , and AAF74258 ). Most in vitro and cell culture studies have used this identical clone originally isolated by Xu et al. ( 38 ).
Mutated amino acid residues in murine WNK4 differ from their human counterparts on the basis of numbering of the conceptual translation products from NM_175579 for mouse and NM_032387 for human. The human disease-causing WNK4 mutants hWNK4 E562K and hWNK4 Q565E correspond to mouse mWNK4 E559K and mWNK4 Q562E, respectively, in the present study. Putative kinase-dead murine WNK4 is mWNK4 D318A. These point mutants, as well as kinase-dead WNK1 K233M and WNK1 S382A, were generated via site-directed mutagenesis (QuikChange, Stratagene) in accordance with the manufacturer's directions. The SMART online resource ( http://smart.embl-heidelberg.de/ ) was used for determination of conserved motifs in rat WNK1 and murine WNK4. Three putative coiled-coil domains were detected in rat WNK1 using the Coils2 program [based on the algorithm of Lupas et al. ( 21 )] at residues 194-217, 563-597, and 1814-1841. No coiled-coil domains were detected in murine WNK4, although there was substantial sequence homology between the isoforms at the first and third putative coiled-coil domains.
Image processing and statistical analysis. For quantitation of autoradiograms, exposed films were scanned (Canon LiDE80) and data were reduced using ImageJ ( http://rsb.info.nih.gov/ij/; National Institutes of Health) and Excel (Microsoft). For all depicted scans of enhanced chemiluminescence exposures of immunoblots, contrast was improved by decreasing the maximum input level from 255 to 175 (Adobe PhotoShop CS) to mimic the true appearance of the exposed film. In Fig. 3 A, some intervening lanes were digitally removed to preserve consistency with Fig. 3, B and C; all lanes in Fig. 3 A are from the same exposure of the same autoradiogram. Where data are shown, all experiments were performed a minimum of three times. Values are means ± SE (Excel, Microsoft), and, where indicated, the number of independent experiments is shown. Where multiple comparisons were performed, statistical significance was attributed using the Student's t -test [for correlated samples using raw data or for independent samples using normalized data (VassarStats; http://faculty.vassar.edu/lowry/VassarStats.html )], in accordance with the false discovery rate procedure, where P i < d i ( 9 ) (where P i is significance level associated with comparison i and d i is critical significance level); this latter approach is used to test the validity, in parallel, of more than one null hypothesis.
Fig. 3. Effect of WNK4 and WNK1 on TRPV4 expression and trafficking to the plasma membrane. HEK-293 cells were transiently transfected with TRPV4 in conjunction with WNK4, WNK1, WNK4 + WNK1, or empty expression vector and then subjected to cell surface biotinylation. Biotinylated proteins were isolated with avidin-agarose beads, resolved via SDS-PAGE, and subjected to anti-TRPV4 immunoblotting; whole cell extracts were prepared in parallel for comparison ( A ). B and C : densitometric quantitation of 4-10 such immunoblot experiments (depending on condition), with normalization to TRPV4 expression or TRPV4 cell surface localization in the absence of coexpressed WNK kinase (Vector). ppt, Precipitable.
RESULTS
HEK-293 cells stably transfected with the tonicity-responsive cation channel TRPV4 exhibit robust Ca 2+ entry on exposure to a hypotonic milieu ( 18, 28 ). This effect is dependent on the presence of extracellular Ca 2+ and is absent in untransfected or vector-transfected HEK-293 cells ( 41 ). For the present series of investigations, we adopted a transient (rather than stable) transfection strategy, which is less subject to confounding effects attributable to over- or underrepresentation of high-expressing clones. HEK-293 cells transiently transfected with TRPV4 exhibited a robust time-dependent response to hypotonicity ( Fig. 1 A ) or the TRPV4 activator ( 34 ) 4 -phorbol 12,13-didecanoate (4 -PDD; Fig. 1 D ), and this effect was abolished in the absence of extracellular Ca 2+. Vector-transfected ( Fig. 1, B and E ) and untransfected ( Fig. 1, C and F ) HEK-293 cells exhibited a much more modest, but reproducible, increment in intracellular Ca 2+ in response to both activators of TRPV4, and these effects were similarly dependent on the presence of extracellular Ca 2+. By immunoblot analysis, we were unable to detect TRPV4 in vector-transfected or untransfected HEK-293 cells, although expression was abundant in the TRPV4 transfectants ( Fig. 1 G ).
Fig. 1. TRPV4-dependent Ca 2+ entry in response to hypotonicity (Hypo) and the TRPV4 activator 4 -phorbol 12,13-didecanoate [4 -PDD (PDD)]. Untransfected human embryonic kidney (HEK-293) cells ( C and F ) or cells transiently transfected with TRPV4 ( A and D ) or empty vector alone ( B and E ) were subjected to hypotonic stress (from 300 to 150 mosmol/kgH 2 O; A-C ) or 10 nM 4 -PDD ( D-F ) in the presence (black trace) or absence (gray trace) of extracellular Ca 2+. G : anti-TRPV4 immunoblot of whole cell lysates prepared from TRPV4- and vector-transfected HEK-293 cells and from untransfected cells.
We used this transient transfection strategy to assess the effect of WNK kinases on TRPV4 function. Coexpression of TRPV4 and WNK4 abrogated the Ca 2+ entry response of TRPV4 to hypotonicity and the TRPV4 activator 4 -PDD ( Fig. 2, A and B ). For rigorous quantitation, comparisons were made using the increment in intracellular Ca 2+ (as assessed via fura 2 ratiometry) at a time of maximal stimulus responsiveness (i.e., at 75 s of treatment). The pooled data ( Fig. 2, C and D ) show that the inhibitory effect of WNK4 was substantial and highly statistically significant.
Fig. 2. Effect of WNK4 and WNK1 on TRPV4-dependent Ca 2+ entry. HEK-293 cells were transiently transfected with TRPV4 in conjunction with vector alone (black trace) or with WNK4, WNK1, or WNK4 + WNK1 (gray traces). Transfectants were then subjected to hypotonicity ( A ) or 4 -PDD ( B ). Each trace represents mean ± SE of 3-4 replicates from a single experiment. Black arrowhead, time of stimulus application; white arrowhead, 75 s of stimulus exposure ( t 75 ); flat baseline before stimulus addition is not shown. Scale bars denote fura 2 ratio [in arbitrary units (U)] as a function of time (s). C and D : increment in fura 2 ratio as an index of intracellular Ca 2+ increase at 75 s after application of hypotonicity or 4 -PDD. Number of experiments is shown inside each bar; there are 3 replicates per condition per individual experiment.
Heterologously expressed WNK1 is reportedly neutral with respect to function of known transport proteins ( 43 ). We compared the effect of WNK1 with that of WNK4. Unexpectedly, WNK1 was virtually as effective as WNK4 at inhibiting the function of TRPV4 in transient transfection in the context of TRPV4 activation by hypotonic stress ( Fig. 2, A and B ) and by 4 -PDD ( Fig. 2, B and D ). Because the conceptual translation of the rat WNK1 cDNA diverges considerably from that of the reported human and murine clones (see METHODS ) and the present studies employed a TRPV4 cDNA of human origin, additional experiments were performed such that the effect of rat WNK1 vis-à-vis murine and human TRPV4 could be assessed in parallel; the effect of rat WNK1 was equivalent, regardless of the species of origin of TRPV4 (data not shown). In some model systems, WNK1 may influence the effect of WNK4 ( 17, 44 ). Therefore, we tested the effect of coexpression of WNK1 and WNK4 with TRPV4. There was no potentiation of the inhibitory response with respect to hypotonicity ( Fig. 2, A and C ) or 4 -PDD ( Fig. 2, B and D ) when both kinases were transfected in combination; however, we could not completely exclude a modest antagonistic effect of WNK1 on the WNK4 effect.
We speculated that WNK4 may influence trafficking of TRPV4 to the plasma membrane, as was demonstrated for several Cl - transport proteins. HEK-293 cells were transiently transfected with TRPV4, in the presence or absence of cotransfection with WNK4 or WNK1. Transfectants were then subjected to cell surface biotinylation, and biotinylated proteins were isolated with avidin-agarose beads. WNK4 and WNK1 downregulated TRPV4 cell surface expression to a degree commensurate with the effect of these kinases on agonist-dependent Ca 2+ entry ( Fig. 3, A and B ). Consistent with the intracellular Ca 2+ data, there was neither abrogation nor potentiation when WNK1 and WNK4 kinases were expressed in concert ( Fig. 3, A and B ). Expression of either WNK kinase alone, or in combination, failed to influence total expression of TRPV4 as determined via anti-TRPV4 immunoblotting of whole cell lysates prepared from the transient transfectants ( Fig. 3, A and C ). Therefore, the WNK effect with respect to TRPV4 function appeared to operate primarily at the level of cell surface localization of the channel.
We next sought to establish the dependence of the WNK4-inducible downregulation of TRPV4 activity on WNK4 kinase activity. A point mutation abolishing WNK4 activity (kinase-dead D318A-WNK4) had been previously described. Heterologous expression of this mutant was used to demonstrate that WNK4 kinase activity was required for WNK4-dependent inhibition of thiazide-sensitive Na + -Cl - cotransporter ( 36, 43 ) but not WNK4-dependent inhibition of ROMK activity ( 16 ). On cotransfection with TRPV4, this "kinase-dead" WNK4 was inert with respect to TRPV4 function, in marked contrast to wild-type WNK4. This effect was evident whether TRPV4 was activated by hypotonicity ( Fig. 4 A ) or 4 -PDD ( Fig. 4 B ). These data suggested that kinase activity was necessary for the WNK4 effect.
Fig. 4. Disease-causing point mutants of WNK4 partially or completely block the effect of WNK4 on TRPV4 activity. HEK-293 cells were transiently transfected with TRPV4 in conjunction with expression plasmid: vector alone, wild-type WNK4, kinase-dead WNK4 D318A, or disease-causing mutants WNK4 E559K and WNK4 Q562E. Transfectants were then subjected to TRPV4 activators: hypotonicity ( A ) and 4 -PDD ( B ). Vertical axis is increment in intracellular Ca 2+ (as assessed via fura 2 ratiometry) at 75 s of treatment, normalized to TRPV4-transfected cells in the absence of cotransfected WNK kinase (Vector). Coexpression of wild-type WNK4 abrogated the effect of hypotonicity and 4 -PDD. In the case of hypotonicity, the kinase-dead WNK4 mutant WNK4 D318A was devoid of inhibitory activity, whereas the disease-causing point mutants WNK4 E559K and WNK4 Q562E exhibited markedly reduced inhibitory effect ( A ). This pattern was also evident in the setting of 4 -PDD stimulation of TRPV4 ( B ), although not all differences achieved statistical significance.
Point mutations in WNK4 are causative for the severe form of hereditary hypertension seen in familial hyperkalemia and hypertension, an effect attributed to hypofunctioning of the mutants ( 35 ). We therefore tested the ability of two of these disease-causing point mutant forms of WNK4 to inhibit TRPV4 function. We used site-directed mutagenesis to create point mutants in murine WNK4 corresponding to the mutations in the human kindreds ( 35 ) (see METHODS ). In the setting of hypotonic activation of TRPV4, mutant forms of WNK4 (WNK4 E559K or WNK4 Q5622E variant), when cotransfected with TRPV4, were significantly less effective than wild-type WNK4 at inhibiting TRPV4 function ( Fig. 4 A ). Both mutants, however, retained some efficacy, in that they decreased TRPV4-dependent Ca 2+ entry relative to empty vector alone. A similar trend was observed when 4 -PDD served as the activator of TRPV4 ( Fig. 4 B ), although statistical significance was not achieved using this stimulus. In general, the effect of these point mutants was intermediate between the effect of wild-type WNK4 and empty vector control. We conclude that intrinsic kinase activity of WNK4 is potentially instrumental in its downmodulatory effect vis-à-vis TRPV4 function and that disease-causing point mutants of WNK4, in general, are less effective in this respect than wild-type WNK4. Consistent with this model, cell surface expression of TRPV4 was not downregulated by the kinase-dead WNK4, and the effect of the disease-causing WNK4 mutants was diminished with respect to wild-type WNK4 (data not shown; n = 2).
WNK1 exhibits intrinsic kinase activity ( 38 ). To establish a role for this function in the WNK1 effect on TRPV4 localization, we employed two strategies to disrupt WNK1 kinase activity. Lys 233 is believed to be the catalytic lysine for WNK1; mutation of this residue to Met resulted in near-total loss of WNK1 kinase activity ( 38 ). Ser 382 of WNK1 was expected to be phosphorylated on the basis of sequence similarity with other well-studied protein kinases (e.g., MAP kinases). When Ser 382 was mutated to Ala, the resultant protein exhibited negligible kinase activity ( 39 ). We tested the effect of these WNK1 point mutants on TRPV4-dependent Ca 2+ entry. In general, the effect of these mutants was intermediate between those of empty vector control and wild-type WNK1 ( Fig. 5, A and B ). In the TRPV4 response to hypotonicity, both kinase-dead mutants significantly decreased Ca 2+ entry relative to empty vector control ( Fig. 5 A ). The effect of WNK1 K233M did not differ from that of wild-type WNK1; however, WNK1 S382A was less effective than wild-type WNK1 at suppressing Ca 2+ entry in response to hypotonic stress ( Fig. 5 A ). Although a similar trend was observed in the cell response to 4 -PDD, in this setting the effect of the kinase-deficient mutants did not differ dramatically from that of wild-type WNK1 ( Fig. 5 B ). We conclude that intrinsic kinase activity of WNK1 likely plays a modest role, at best, in the regulation of TRPV4 localization.
Fig. 5. Downregulation of TRPV4-dependent Ca 2+ entry is preserved in kinase-dead WNK1 mutants. HEK-293 cells were transiently transfected with TRPV4 in conjunction with expression plasmid: vector alone, wild-type WNK1, kinase-dead WNK1 K233M, or kinase-dead WNK1 S382A. Transfectants were then subjected to TRPV4 activators: hypotonicity ( A ) and 4 -PDD ( B ). Vertical axis is the increment in intracellular Ca 2+ (as assessed via fura 2 ratiometry) at 75 s of treatment, normalized to TRPV4 alone (Vector). Coexpression of wild-type WNK1 abrogated the effect of hypotonicity and 4 -PDD on TRPV4-dependent Ca 2+ entry. In response to both treatments, both kinase-dead WNK1 mutants blunted TRPV4-dependent Ca 2+ entry; only one of the kinase-dead mutants (WNK1 S382A ) in response to a single stimulus (hypotonicity) was significantly less effective than wild-type WNK1.
We speculated that WNK4 and/or WNK1 may exert an effect on TRPV4 localization via a direct interaction with the channel. We employed a series of reciprocal coimmunoprecipitation strategies but were unable to demonstrate a direct interaction between either kinase and TRPV4 (data not shown).
We next sought to define the WNK4 domains responsible for downregulation of TRPV4 expression and function. Mouse WNK4 has a kinase domain spanning approximately residues 171-429 [as analyzed via the SMART website ( http://smart.embl-heidelberg.de/ ); Fig. 6 A ]. Immediately downstream of the kinase domain is an "autoinhibitory domain," which is reasonably well conserved between WNK1 and WNK4. WNK1 has a number of coiled-coil domains potentially mediating protein-protein interactions ( 38 ). It has been suggested that WNK4 has a number of corresponding domains ( 33 ). However, although there is significant homology between WNK1 and WNK4 at two of these putative domains (corresponding to residues 147-167 and residues 1111-1138 in murine WNK4), we were unable to detect coiled-coil domains when the Coils2 program was used to screen WNK4 directly. WNK4 deletion mutants were designed to eliminate the first putative coiled-coil domain and nearly all the kinase domain ( 29-366-mWNK4), the autoinhibitory domain and second putative coiled-coil domain ( 366-794-mWNK4), the COOH terminus, including the third putative coiled-coil domain ( 677-1211-mWNK4), and the extreme COOH terminus containing only the third coiled-coil domain and adjacent residues ( 1086-1211-mWNK4; Fig. 4 A ). The NH 2 -terminal WNK4 mutants 29-366-mWNK4 and 366-794-mWNK4 exhibited preservation of the inhibitory effect vis-à-vis TRPV4 activity in response to hypotonicity ( Fig. 6, B and D ) or 4 -PDD ( Fig. 6, C and E ), even in the absence of an intact kinase domain. The COOH-terminal deletion mutant 677-1211-mWNK4 and the extreme COOH-terminal mutant 1086-1211-mWNK4 were essentially devoid of inhibitory effect ( Fig. 6, B-E ). These data establish the essential nature of the WNK4 extreme COOH terminus for effecting the downregulation of TRPV4 function and, in contrast to our observations using a kinase-dead WNK4 mutant, suggested the dispensability of the kinase domain.
Fig. 6. WNK4 domains conferring downregulation on TRPV4. Deletion mutants of mWNK4 were made, and their ability to influence TRPV4 Ca 2+ channel activity after transient coexpression was examined in the HEK-293 cell model. A : putative WNK4 domains on wild-type mWNK4. C, coiled-coil domain; K, kinase domain; A, autoinhibitory domain. Deletion mutants ( 29-366-mWNK4, 366-794-mWNK4, 677-1211-mWNK4, and 1086-1211-mWNK4) are shown by black bars below mWNK4; gray regions are absent from deletion mutants. Diagrams are drawn strictly to scale. B and C : effect of each deletion mutant vs. wild-type WNK4 on TRPV4-dependent Ca 2+ entry (expressed as fura 2 ratio) in response to hypotonicity or 4 -PDD treatment for an individual experiment ( n = 3 determinations per condition). D and E : pooled data from multiple experiments with statistical analysis. COOH-terminal portion of WNK4 appears to confer downregulation in TRPV4-dependent Ca 2+ entry; deletion of this region may confer a dominant-negative-acting effect ( D ).
We attempted to establish in preliminary fashion the domain of TRPV4 conferring downregulation in response to WNK4. Human TRPV4 (isoform a or 1) is an 871-amino acid protein with a canonical hexahelical voltage-gated ion channel domain comprising residues 470-717, as assessed via the DAS Transmembrane Prediction Server ( 8 ) ( http://www.sbc.su.se/ miklos/DAS/; Fig. 7 A ). TRPV4 also has as many as six ankyrin binding domains ( Fig. 7 A ), which are believed to be instrumental in protein-protein interaction. Without disturbing the membrane-spanning domains (which would eliminate the cell surface expression we wished to quantitate), we generated a series of deletion mutants of the TRPV4 NH 2 and COOH termini ( Fig. 7 A ). Four of five mutants were abundantly expressed on transient transfection when whole cell lysates were immunoblotted for TRPV4 ( Fig. 7 B ); in all four of these cases, the mutant TRPV4 trafficked to the plasma membrane in appreciable quantities, as assessed via cell surface biotinylation ( Fig. 7 B ). In all but one of these highly expressed mutants, downregulation in response to WNK4 cotransfection was preserved ( Fig. 7 B ). Specifically, the 2-147-TRPV4 mutant, although abundantly expressed, was not downregulated by WNK4 coexpression ( Fig. 7, B and C ). (Data for the 157-299-TRPV4 mutant could not be obtained because of the negligible level of expression of the mutant and are absent from Fig. 7 C.) These data suggested that a feature of the TRPV4 NH 2 terminus separate from the ankyrin binding domains was required for the WNK4 effect. Interestingly, and somewhat paradoxically, deletion of a slightly larger domain (in the 2-157-TRPV4 mutant) did not abrogate WNK4 responsiveness ( Fig. 7, B and C ). In an effort to reconcile this discrepancy, we generated two more deletion mutants of TRPV4: 189-467-TRPV4, absent all six of the putative ankyrin domains, and 1-189-TRPV4, absent the NH 2 terminus upstream of the ankyrin domains ( Fig. 7 A ). Both mutants were subject to downregulation by WNK4 coexpression (data not shown; n = 2). In aggregate, these data supported a role for a unique configuration of the NH 2 terminus (present in the 2-157-TRPV4 mutant but absent from the 2-147-TRPV4 mutant) and/or the membrane-spanning region (by the process of elimination). Because of the gross nature of these deletion mutants and the anticipated effect on channel function, Ca 2+ entry data in response to TRPV4 agonists were not sought.
Fig. 7. Mapping the region of TRPV4 that confers downregulation in response to WNK expression. A : deletion mutants of TRPV4 were generated and transiently transfected into HEK-293 cells, in conjunction with empty vector or with WNK4. Location of ankyrin and membrane-spanning domains in wild-type TRPV4 and in each of the deletion mutants is shown. B : cell surface biotinylation was performed, and expression of avidin-precipitable (cell surface-expressed) full-length or deletion-mutated TRPV4 was compared with abundance in whole cell lysates. One deletion mutant, 157-299-TRPV4, exhibited negligible expression and was excluded from further analysis. C : data from 3-4 experiments (depending on deletion mutant) reduced densitometrically and shown as ratio of avidin-precipitable (i.e., biotinylated and, hence, cell surface-expressed) TRPV4 in the presence of WNK4 coexpression to the absence of WNK4. Therefore, absence of WNK4 effect on a given TRPV4 mutant is depicted by a value of unity (dashed line). Only 1 deletion mutant, 2-147-TRPV4, was resistant to downregulation by WNK4 ( B and C ). Closely related 2-157-TRPV4 exhibited preserved downregulation by WNK4. Two additional mutants were generated that lacked all putative ankyrin binding domains ( 189-467-TRPV4) or the entire NH 2 terminus upstream of these domains ( 1-189-TRPV4; A ); neither deletion blocked WNK4-dependent downregulation of TRPV4 cell surface expression (not shown).
DISCUSSION
In light of their responsiveness to anisotonicity ( 17, 38 ) and their coexpression with the tonicity-responsive TRPV4 in the distal nephron ( 35, 41 ), we hypothesized that WNK kinases would influence function or localization of this channel. In the present study, we note that coexpression of WNK4 or WNK1 with TRPV4 results in marked downregulation of trafficking of the latter to the plasma membrane (without influencing overall abundance of the channel). Correspondingly, we observed a decrease in TRPV4 responsiveness to hypotonicity or the TRPV4 activator 4 -PDD.
WNK4 has been characterized primarily as an inhibitor of Cl - transport. On coexpression in a heterologous model system, WNK4 downregulates expression of the thiazide-sensitive Na + -Cl - cotransporter ( 36, 43 ), as well as the Na + -K + -2Cl - cotransporter and the apical Cl - /HCO 3 - exchanger ( 14 ). With respect to the effect of WNK4 on paracellular Cl - transport, conflicting data have emerged. One group observed that WNK4 had no effect on paracellular Cl - transport ( 42 ), whereas others noted an increase in activity ( 15 ). In addition to Cl - transport, activity of the K + channel ROMK is also decreased by coexpression of WNK4 ( 16 ).
Disease-causing mutants of human WNK4 include a series of individual missense mutations affecting a small cluster of polar, and primarily positively charged, residues (E559K, D561A, and Q562E). In the present study, mutations in murine WNK4 corresponding to these disease-causing mutations in human WNK4 exhibited a partial loss of inhibitory effect relative to wild-type WNK4. The effect of WNK4 disease-causing mutations in other models has been variable. Mutant WNK4 Q562E exhibited a blunted inhibitory effect vis-à-vis activity of the thiazide-sensitive Na + -Cl - cotransporter; however, mutants WNK4 E559K and WNK4 D561A were as effective as wild-type WNK4 in this capacity ( 36, 43 ). With respect to ROMK, disease-causing mutants produced an unexpected finding: the ability of WNK4 to inhibit ROMK activity was potentiated, rather than blocked, by mutants WNK4 Q562E and WNK4 E559K ( 16 ).
The effect of WNK1 in the present model exhibits similarities to other model systems; however, there are important differences. In the present study, WNK4 and WNK1 blunted the effect of two different activators of TRPV4. In contrast, in isolation, WNK1 was unable to influence function of the thiazide-sensitive Na + -Cl - cotransporter in prior studies by Yang et al. ( 43 ). For these studies, a rat WNK1 cDNA clone, the conceptual translation of which differs significantly from that of the reported human and murine WNK1 clones (see METHODS ), was used. However, because Yang et al. used this identical clone for their studies, we do not believe this is the basis for our contrasting observations.
Conflicting data emerge with respect to the requirement for intrinsic kinase activity of WNKs in downregulating TRPV4. Specifically, WNK4 mutated for a key residue within the kinase domain (D318) resulting in a kinase-dead phenotype was less effective than wild-type WNK4 at blocking TRPV4 function. In contrast, deletion of nearly the entire kinase domain (including this key residue) failed to abolish the WNK4 effect. Kinase activity has been a variable requirement for WNK4 effect on channel and transporter activity. For example, inhibition of the thiazide-sensitive Na + -Cl - cotransporter by WNK4 expression was prevented if the kinase-dead WNK4 mutant was used instead ( 36, 43 ). However, the WNK4 effect on ROMK appeared to be independent of WNK4 kinase activity, because it was observed even with the kinase-dead (D318A) WNK4 mutant ( 16 ). It is conceivable that this charged residue, essential for kinase activity, serves an additional role in WNK4 function. This possibility is further supported by the seemingly modest role of intrinsic kinase activity in the ability of WNK1 to downregulate TRPV4 ( Fig. 5 ).
After initial uncertainty, a role for TRPV4 in central osmoregulation in mammals appears secure ( 19, 24 ). The contribution of the abundant renal expression of TRPV4 to this process remains unresolved ( 18, 28, 29 ). The extent to which the abnormal salt and water metabolism in mice harboring TRPV4 targeted deletions is reflective of hypothalamic vs. renal inactivation of this channel has not been explored. Renal TRPV4 is restricted to tubule segments lacking constitutive apical water permeability (i.e., sites distal to the genu of the loop of Henle) and localizes to the basolateral membrane in these sites ( 29 ). We speculate that TRPV4 may function as a distal nephron sensor of interstitial solute and water and, indirectly, of transcellular resorption via the proximal nephron ( 7 ). In the distal nephron, TRPV4 is coexpressed with WNK1 and WNK4 ( 35, 41 ). Osmotically responsive WNK kinases may influence expression and function of TRPV4 in this tissue.
The syndrome of familial hyperkalemia and hypertension is causally linked to overexpression of WNK1 or to inactivating mutations in WNK4 ( 35 ). Patients with WNK1 mutations exhibit normocalciuria ( 2 ), whereas those with mutations in WNK4 manifest hypercalciuria ( 23 ). Mayan and colleagues, who correctly predicted constitutive overactivity of the thiazide-sensitive Na + -Cl - cotransporter in this syndrome before the elucidation of the molecular phenotype ( 23 ), recently proposed that mutant WNK4 causes hypercalciuria by influencing a Ca 2+ channel or transporter ( 22 ). A WNK4 and TRPV4 functional interaction may be perturbed in the setting of WNK4 mutation.
It is important to emphasize that TRPV4 is not the only candidate Ca 2+ channel in this part of the nephron. Although the role of TRPV5 and TRPV6 in regulation of urinary Ca 2+ excretion has appropriately received much attention ( 12 ), TRPV4 potentially participates in this process as well. As described above, TRPV4 is highly expressed along the distal nephron ( 29 ). In addition, there was a trend toward lower plasma Ca 2+ levels in TRPV4 -/- mice than in TRPV4 +/+ controls [8.8 ± 0.3 vs. 9.8 ± 0.4 (SE) mg/dl, n = 10 ( 24 )]; however, these data did not reach statistical significance in this important, but relatively small, study. Therefore, TRPV4 may influence systemic, as well as local, Ca 2+ balance; both may be perturbed by altered WNK abundance or function. An effect of WNK4 on TRPV5 function has been described in preliminary fashion by Peng and colleagues ( 26 ); however, inconsistent with the aberrant Ca 2+ balance accompanying familial hyperkalemia and hypertension, disease-causing WNK4 mutants were indistinguishable from wild-type WNK4 in in vitro assays of TRPV5 function. Therefore, it is possible that mutant WNK4-induced aberrant regulation of TRPV4 or TRPV6 or another Ca 2+ transport protein is operative in this disease.
The molecular mechanism through which WNKs influence TRPV4 targeting and function is unclear. WNK1 may function as a MAP kinase kinase kinase kinase (i.e., a "MAP4K") in the ERK5 pathway; its downstream effects are sensitive to an inhibitor of MEK5 ( 40 ). Interestingly, very similar to TRPV4, the function of ERK5 and its upstream activators is regulated by anisotonicity ( 1, 32 ), suggesting a physiological signaling "module" encompassing WNKs and TRPV4. WNKs also serve as substrates for the kinase Akt/protein kinase B ( 13, 31 ), which is itself an effector of the anisotonicity-responsive phosphatidylinositol 3-kinase pathway ( 30, 45 ). Because intrinsic kinase activity is not an absolute requirement for the WNK effect on TRPV4, it is conceivable that these large kinases also act via a scaffolding mechanism, perhaps transiently bridging TRPV4 and another effector. Alternatively, WNK activation may indirectly liberate a small molecule intermediate that influences TRPV4 trafficking; this latter model is consistent with our inability to demonstrate a direct interaction between WNK kinases and TRPV4.
WNKs are believed to influence trafficking of clathrin-coated endocytic vesicles in a dynamin-dependent fashion ( 16 ). An analogous, although biochemically distinct, mechanism of rapid, agonist-dependent shuttling to and from the cell membrane has recently been described for a TRP channel. Specifically, growth factor-dependent insertion of TRPC5 appears to require phosphatidylinositol 3-kinase activity and the Rho family GTPase Rac1 ( 5 ). As suggested above, phosphatidylinositol 3-kinase may function upstream of WNK activation, biochemically linking these two mechanisms.
In summary, these data support a role for WNK kinases in regulating the subcellular localization of TRPV4. Inasmuch as WNK kinases and TRPV4 are responsive to changes in ambient tonicity, it is tempting to speculate that they jointly participate in an osmosensing or osmotically responsive signal transduction pathway in the distal nephron or elsewhere.
GRANTS
This work was supported by the National Institutes of Health, the American Heart Association, and the Department of Veterans Affairs.
ACKNOWLEDGMENTS
The authors thank Drs. Melanie Cobb, Chao-Ling Yang, and David Ellison for valuable reagents.
【参考文献】
Abe J, Kusuhara M, Ulevitch RJ, Berk BC, and Lee JD. Big mitogen-activated protein kinase 1 (BMK1) is a redox-sensitive kinase. J Biol Chem 271: 16586-16590, 1996.
Achard JM, Warnock DG, Disse-Nicodeme S, Fiquet-Kempf B, Corvol P, Fournier A, and Jeunemaitre X. Familial hyperkalemic hypertension: phenotypic analysis in a large family with the WNK1 deletion mutation. Am J Med 114: 495-498, 2003.
Arniges M, Vazquez E, Fernandez-Fernandez JM, and Valverde MA. Swelling-activated Ca 2+ entry via TRPV4 channel is defective in cystic fibrosis airway epithelia. J Biol Chem 279: 54062-54068, 2004.
Becker D, Blase C, Bereiter-Hahn J, and Jendrach M. TRPV4 exhibits a functional role in cell-volume regulation. J Cell Sci 118: 2435-2440, 2005.
Bezzerides VJ, Ramsey IS, Kotecha S, Greka A, and Clapham DE. Rapid vesicular translocation and insertion of TRP channels. Nat Cell Biol 6: 709-720, 2004.
Clapham DE. TRP channels as cellular sensors. Nature 426: 517-524, 2003.
Cohen D. TRPV4 and the mammalian kidney. Pflügers Arch 451: 168-175, 2005.
Cserzo M, Wallin E, Simon I, von Heijne G, and Elofsson A. Prediction of transmembrane -helices in prokaryotic membrane proteins: the dense alignment surface method. Protein Eng 10: 673-676, 1997.
Curran-Everett D. Multiple comparisons: philosophies and illustrations. Am J Physiol Regul Integr Comp Physiol 279: R1-R8, 2000.
Delany NS, Hurle M, Facer P, Alnadaf T, Plumpton C, Kinghorn I, See CG, Costigan M, Anand P, Woolf CJ, Crowther D, Sanseau P, and Tate SN. Identification and characterization of a novel human vanilloid receptor-like protein, VRL-2. Physiol Genomics 4: 165-174, 2001.
Grynkiewicz G, Poenie M, and Tsien RY. A new generation of Ca 2+ indicators with greatly improved fluorescence properties. J Biol Chem 260: 3440-3450, 1985.
Hoenderop JG, Nilius B, and Bindels RJ. Calcium absorption across epithelia. Physiol Rev 85: 373-422, 2005.
Jiang ZY, Zhou QL, Holik J, Patel S, Leszyk J, Coleman K, Chouinard M, and Czech MP. Identification of WNK1 as a substrate of Akt/protein kinase B and a negative regulator of insulin-stimulated mitogenesis in 3T3-L1 cells. J Biol Chem 280: 21622-21628, 2005.
Kahle KT, Gimenez I, Hassan H, Wilson FH, Wong RD, Forbush B, Aronson PS, and Lifton RP. WNK4 regulates apical and basolateral Cl - flux in extrarenal epithelia. Proc Natl Acad Sci USA 101: 2064-2069, 2004.
Kahle KT, MacGregor GG, Wilson FH, Van Hoek AN, Brown D, Ardito T, Kashgarian M, Giebisch G, Hebert SC, Boulpaep EL, and Lifton RP. Paracellular Cl - permeability is regulated by WNK4 kinase: insight into normal physiology and hypertension. Proc Natl Acad Sci USA 101: 14877-14882, 2004.
Kahle KT, Wilson FH, Leng Q, Lalioti MD, O'Connell AD, Dong K, Rapson AK, MacGregor GG, Giebisch G, Hebert SC, and Lifton RP. WNK4 regulates the balance between renal NaCl reabsorption and K + secretion. Nat Genet 35: 372-376, 2003.
Lenertz LY, Lee BH, Min X, Xu BE, Wedin K, Earnest S, Goldsmith EJ, and Cobb MH. Properties of WNK1 and implications for other family members. J Biol Chem 280: 26653-26658, 2005.
Liedtke W, Choe Y, Marti-Renom MA, Bell AM, Denis CS, Sali A, Hudspeth AJ, Friedman JM, and Heller S. Vanilloid receptor-related osmotically activated channel (VR-OAC), a candidate vertebrate osmoreceptor. Cell 103: 525-535, 2000.
Liedtke W and Friedman JM. Abnormal osmotic regulation in trpv4 -/- mice. Proc Natl Acad Sci USA 100: 13698-13703, 2003.
Liedtke W, Tobin DM, Bargmann CI, and Friedman JM. Mammalian TRPV4 (VR-OAC) directs behavioral responses to osmotic and mechanical stimuli in Caenorhabditis elegans. Proc Natl Acad Sci USA 27: 27, 2003.
Lupas A, Van Dyke M, and Stock J. Predicting coiled coils from protein sequences. Science 252: 1162-1164, 1991.
Mayan H, Munter G, Shaharabany M, Mouallem M, Pauzner R, Holtzman EJ, and Farfel Z. Hypercalciuria in familial hyperkalemia and hypertension accompanies hyperkalemia and precedes hypertension: description of a large family with the Q565E WNK4 mutation. J Clin Endocrinol Metab 89: 4025-4030, 2004.
Mayan H, Vered I, Mouallem M, Tzadok-Witkon M, Pauzner R, and Farfel Z. Pseudohypoaldosteronism type II: marked sensitivity to thiazides, hypercalciuria, normomagnesemia, and low bone mineral density. J Clin Endocrinol Metab 87: 3248-3254, 2002.
Mizuno A, Matsumoto N, Imai M, and Suzuki M. Impaired osmotic sensation in mice lacking TRPV4. Am J Physiol Cell Physiol 285: C96-C101, 2003.
O'Reilly M, Marshall E, Speirs HJ, and Brown RW. WNK1, a gene within a novel blood pressure control pathway, tissue-specifically generates radically different isoforms with and without a kinase domain. J Am Soc Nephrol 14: 2447-2456, 2003.
Peng JB, Jiang Y, and Warnock DG. WNK4 kinase enhances ECaC-mediated calcium transport (Abstract). J Am Soc Nephrol 15: 62A, 2004.
Roullet JB, Luft UC, Xue H, Chapman J, Bychkov R, Roullet CM, Luft FC, Haller H, and McCarron DA. Farnesol inhibits L-type Ca 2+ channels in vascular smooth muscle cells. J Biol Chem 272: 32240-32246, 1997.
Strotmann R, Harteneck C, Nunnenmacher K, Schultz G, and Plant TD. OTRPC4, a nonselective cation channel that confers sensitivity to extracellular osmolarity. Nat Cell Biol 2: 695-702, 2000.
Tian W, Salanova M, Xu H, Lindsley JN, Oyama TT, Anderson S, Bachmann S, and Cohen DM. Renal expression of osmotically responsive cation channel TRPV4 is restricted to water-impermeant nephron segments. Am J Physiol Renal Physiol 287: F17-F24, 2004.
Tilly BC, Edixhoven MJ, Tertoolen LG, Morii N, Saitoh Y, Narumiya S, and de Jonge HR. Activation of the osmo-sensitive chloride conductance involves p21 rho and is accompanied by a transient reorganization of the F-actin cytoskeleton. Mol Biol Cell 7: 1419-1427, 1996.
Vitari AC, Deak M, Collins BJ, Morrice N, Prescott AR, Phelan A, Humphreys S, and Alessi DR. WNK1, the kinase mutated in an inherited high-blood-pressure syndrome, is a novel PKB (protein kinase B)/Akt substrate. Biochem J 378: 257-268, 2004.
Wang X, Merritt AJ, Seyfried J, Guo C, Papadakis ES, Finegan KG, Kayahara M, Dixon J, Boot-Handford RP, Cartwright EJ, Mayer U, and Tournier C. Targeted deletion of mek5 causes early embryonic death and defects in the extracellular signal-regulated kinase 5/myocyte enhancer factor 2 cell survival pathway. Mol Cell Biol 25: 336-345, 2005.
Wang Z, Yang CL, and Ellison DH. Comparison of WNK4 and WNK1 kinase and inhibiting activities. Biochem Biophys Res Commun 317: 939-944, 2004.
Watanabe H, Davis JB, Smart D, Jerman JC, Smith GD, Hayes P, Vriens J, Cairns W, Wissenbach U, Prenen J, Flockerzi V, Droogmans G, Benham CD, and Nilius B. Activation of TRPV4 channels (hVRL-2/mTRP12) by phorbol derivatives. J Biol Chem 277: 13569-13577, 2002.
Wilson FH, Disse-Nicodeme S, Choate KA, Ishikawa K, Nelson-Williams C, Desitter I, Gunel M, Milford DV, Lipkin GW, Achard JM, Feely MP, Dussol B, Berland Y, Unwin RJ, Mayan H, Simon DB, Farfel Z, Jeunemaitre X, and Lifton RP. Human hypertension caused by mutations in WNK kinases. Science 293: 1107-1112, 2001.
Wilson FH, Kahle KT, Sabath E, Lalioti MD, Rapson AK, Hoover RS, Hebert SC, Gamba G, and Lifton RP. Molecular pathogenesis of inherited hypertension with hyperkalemia: the Na-Cl cotransporter is inhibited by wild-type but not mutant WNK4. Proc Natl Acad Sci USA 100: 680-684, 2003.
Wissenbach U, Bodding M, Freichel M, and Flockerzi V. Trp12, a novel Trp related protein from kidney. FEBS Lett 485: 127-134, 2000.
Xu B, English JM, Wilsbacher JL, Stippec S, Goldsmith EJ, and Cobb MH. WNK1, a novel mammalian serine/threonine protein kinase lacking the catalytic lysine in subdomain II. J Biol Chem 275: 16795-16801, 2000.
Xu BE, Min X, Stippec S, Lee BH, Goldsmith EJ, and Cobb MH. Regulation of WNK1 by an autoinhibitory domain and autophosphorylation. J Biol Chem 277: 48456-48462, 2002.
Xu BE, Stippec S, Lenertz L, Lee BH, Zhang W, Lee YK, and Cobb MH. WNK1 activates ERK5 by an MEKK2/3-dependent mechanism. J Biol Chem 279: 7826-7831, 2004.
Xu H, Zhao H, Tian W, Yoshida K, Roullet JB, and Cohen DM. Regulation of a transient receptor potential (TRP) channel by tyrosine phosphorylation. SRC family kinase-dependent tyrosine phosphorylation of TRPV4 on TYR-253 mediates its response to hypotonic stress. J Biol Chem 278: 11520-11527, 2003.
Yamauchi K, Rai T, Kobayashi K, Sohara E, Suzuki T, Itoh T, Suda S, Hayama A, Sasaki S, and Uchida S. Disease-causing mutant WNK4 increases paracellular chloride permeability and phosphorylates claudins. Proc Natl Acad Sci USA 101: 4690-4694, 2004.
Yang CL, Angell J, Mitchell R, and Ellison DH. WNK kinases regulate thiazide-sensitive Na-Cl cotransport. J Clin Invest 111: 1039-1045, 2003.
Yang CL, Zhu X, Wang Z, Subramanya AR, and Ellison DH. Mechanisms of WNK1 and WNK4 interaction in the regulation of thiazide-sensitive NaCl cotransport. J Clin Invest 115: 1379-1387, 2005.
Zhang Z, Yang XY, Soltoff SP, and Cohen DM. PI3K signaling in the murine kidney inner medullary cell response to urea. Am J Physiol Renal Physiol 278: F155-F164, 2000.
作者单位：1 Division of Nephrology and Hypertension, Department of Medicine, 2 Division of Nephrology, Department of Pediatrics, Oregon Health and Science University, and 3 Portland Veterans Affairs Medical Center, Portland, Oregon