Cation Chloride Cotransporters Interact with the Stress-related Kinases Ste20-related Proline-Alanine-rich Kinase (SPAK) and Oxidative Stress Response 1 (OSR1)

Abstract

Cells respond to stress stimuli by mounting specific responses. During osmotic and oxidative stress, cation chloride cotransporters, e.g. Na-K-2Cl and K-Cl cotransporters, are activated to maintain fluid/ion homeostasis. Here we report the interaction of the stress-related serine-threonine kinases Ste20-relatedproline-alanine-rich kinase (SPAK) and oxidative stress response 1 (OSR1) with the cotransporters KCC3, NKCC1, and NKCC2 but not KCC1 and KCC4. The interaction was identified using yeast two-hybrid assays and confirmed via glutathione S-transferase pull-down experiments. Evidence for in vivo interaction was established by co-immunoprecipitation of SPAK from mouse brain with anti-NKCC1 antibody. The interacting region of both kinases comprises the last 100 amino acids of the protein. The SPAK/OSR1 binding motif on the cotransporters consists of nine residues, starting with an (R/K)FX(V/I) sequence followed by five additional residues that are essential for binding but for which no consensus was found. Immunohistochemical analysis of choroid plexus epithelium revealed co-expression of NKCC1 and SPAK on the apical membrane. In contrast, in choroid plexus epithelium from NKCC1 null mice, SPAK immunostaining was found in the cytoplasm. We conclude that several cation chloride co-transporters interact with SPAK and/or OSR1, and we hypothesize that this interaction might play a role during the initiation of the cellular stress response. Cells respond to stress stimuli by mounting specific responses. During osmotic and oxidative stress, cation chloride cotransporters, e.g. Na-K-2Cl and K-Cl cotransporters, are activated to maintain fluid/ion homeostasis. Here we report the interaction of the stress-related serine-threonine kinases Ste20-relatedproline-alanine-rich kinase (SPAK) and oxidative stress response 1 (OSR1) with the cotransporters KCC3, NKCC1, and NKCC2 but not KCC1 and KCC4. The interaction was identified using yeast two-hybrid assays and confirmed via glutathione S-transferase pull-down experiments. Evidence for in vivo interaction was established by co-immunoprecipitation of SPAK from mouse brain with anti-NKCC1 antibody. The interacting region of both kinases comprises the last 100 amino acids of the protein. The SPAK/OSR1 binding motif on the cotransporters consists of nine residues, starting with an (R/K)FX(V/I) sequence followed by five additional residues that are essential for binding but for which no consensus was found. Immunohistochemical analysis of choroid plexus epithelium revealed co-expression of NKCC1 and SPAK on the apical membrane. In contrast, in choroid plexus epithelium from NKCC1 null mice, SPAK immunostaining was found in the cytoplasm. We conclude that several cation chloride co-transporters interact with SPAK and/or OSR1, and we hypothesize that this interaction might play a role during the initiation of the cellular stress response. Environmental change in a variety of physiological contexts elicits cellular responses via the activation of cell surface receptors. Depending on the nature of the stimulus, different protein families are involved. Osmotic and oxidative changes, for instance, have been shown to activate cation chloride cotransporters as well as stress signaling pathways. Cation chloride cotransporters are intrinsic membrane proteins that move Na+, K+, and Cl− ions across plasma membranes in a tightly coupled electroneutral fashion. They mediate secondary active transport and are driven by the gradients generated by the Na+/K+-ATPase. The cation chloride cotransporter family consists of at least seven members. A group of three Na+-dependent inward cotransporters comprises one Na-Cl cotransporter (NCC) 1The abbreviations used are: NCC, Na-Cl cotransporter; NKCC1 and -2, Na-K-2Cl cotransporters; KCC1-4, K-Cl cotransporters; SPAK, Ste20-related proline-alanine-rich kinase; OSR1, oxidative stress response 1; PP1, phosphatase 1; GST, glutathione S-transferase; HEK, human embryonic kidney; PBS, phosphate-buffered saline; HA, hemagglutinin; ERK, extracellular signal-regulated kinase; JNK, c-Jun NH2-terminal kinase; GCK, germinal center kinase; MAPK, mitogen-activated protein kinase; PASK, proline-alanine-rich Ste20-related kinase. and two Na-K-2Cl cotransporters (NKCC1 and -2). Na+-independent outward transport of K+ and Cl− is mediated by four K-Cl cotransporters (KCC1–4) with distinct functional properties (KCC1 (1Gillen C.M. Brill S. Payne J.A. Forbush B.I. J. Biol. Chem. 1996; 271: 16237-16244Abstract Full Text Full Text PDF PubMed Scopus (341) Google Scholar), KCC2 (2Payne J.A. Am. J. Physiol. 1997; 273: C1516-C1525Crossref PubMed Google Scholar), KCC3 (3Hiki K. D'Andrea R.J. Furze J. Crawford J. Woollatt E. Sutherland G.R. Vadas M.A. Gamble J.R. J. Biol. Chem. 1999; 274: 10661-10667Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar, 4Mercado A. Song L. George A.J. Delpire E. Mount D.B. J. Am. Soc. Nephrol. 1999; 10: 38AGoogle Scholar), and KCC4 (4Mercado A. Song L. George A.J. Delpire E. Mount D.B. J. Am. Soc. Nephrol. 1999; 10: 38AGoogle Scholar)). Relevant functions of the cotransporters include: 1) epithelial transport such as Na+ reabsorption in the thick ascending limb of Henle and distal convoluted tubule of the kidney, K+ reabsorption in choroid plexus, K+ secretion in the inner ear, and Cl− secretion in the sweat and salivary glands, intestine, and airway epithelia; 2) regulation of Cl− gradient and thereby modulation of excitability of neurons in both the central and peripheral nervous systems; and 3) cell volume regulation (reviewed in Refs. 5Haas M. Forbush B.I. Annu. Rev. Physiol. 2000; 62: 515-534Crossref PubMed Scopus (332) Google Scholar, 6Russell J.M. Physiol. Rev. 2000; 80: 211-276Crossref PubMed Scopus (737) Google Scholar, 7Delpire E. News Physiol. Sci. 2000; 15: 309-312PubMed Google Scholar, 8Delpire E. Mount D.B. Annu. Rev. Physiol. 2002; 64: 803-843Crossref PubMed Scopus (188) Google Scholar). Several lines of evidence also suggest a role for cation chloride cotransporters in cell proliferation, survival, and differentiation, especially for NKCC1 (9Matthews J.B. Hassan I. Meng S. Archer S.Y. Hrnjez B.J. Hodin R.A. J. Clin. Invest. 1998; 101: 2072-2079Crossref PubMed Scopus (62) Google Scholar, 10Panet R. Marcus M. Atlan H. J. Cell. Physiol. 2000; 182: 109-118Crossref PubMed Scopus (65) Google Scholar) but also for KCC3 (11Shen M.-R. Chou C.-Y. Hsu K.-F. Liu H.-S. Dunham P.B. Holtzman E.J. Ellory J.C. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14714-14719Crossref PubMed Scopus (83) Google Scholar). Identification of proteins that directly interact with cation chloride cotransporters and thereby regulate or mediate their activity is important to understand their different physiological functions. However, the sole evidence for direct protein interaction comes from a study that showed in vitro phosphorylation of NKCC1 by the stress-related kinase c-Jun NH2-terminal kinase (JNK) (12Klein J.D. Lamitina S.T. O'Neill W.C. Am. J. Physiol. 1999; 46: C425-C431Crossref Google Scholar). Additionally, strong conclusive evidence for the interaction of phosphatase 1 (PP1) with NKCC1 is provided by mutation studies of the PP1 motif in heterologous expression studies (13Darman R.B. Flemmer A. Forbush B.I. J. Biol. Chem. 2001; 276: 34359-34362Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar). Here we performed a yeast two-hybrid screen to search for key proteins that interact with cation chloride cotransporters. We identified the serine-threonine kinases SPAK and OSR1 as direct interactors of several members of the cation chloride cotransporter family. The interaction was substantiated by GST pull-down and co-immunoprecipitation studies. Furthermore, we provide immunohistochemical evidence for the interaction by a colocalization experiment in brain sections. The results presented here support a possible role for the cation chloride cotransporter's interaction with the stress kinases as an initial sensor in the oxidative and/or osmotic stress response pathway. The amino terminus of KCC3a (N′KCC3) was amplified by PCR from mouse brain cDNA using a sense primer that contained an EcoRI site and an antisense primer that contained a stop codon followed by aBamHI site. The fragment was fused to the Gal4 binding domain in the pGBDUc2 vector and transformed into PJ69-4A cells (14James P. Halladay J. Craig E.A. Genetics. 1996; 144: 1425-1436Crossref PubMed Google Scholar) according to standard yeast handling procedures (Yeast Handbook, Clontech). Yeast cells that contained N′KCC3-pGBDUc2 were transformed with a mouse brain library in pACTII (Clontech). A total of 13 × 106clones was screened. The first round of nutritional selection was carried out on −Ura−Leu−Ade plates, and in the second and third round clones were restreaked on triple dropout (−Ura−Leu−His) + 2 mm 3-amino-1,2,4-triazole plates. β-Galactosidase activity of the surviving clones was tested in a liquid assay. Yeast cells were lysed with 10 units of lyticase (Sigma) in lysis buffer (1.2m sorbitol, 100 mmNaH3PO4, pH 7.4), and clones were PCR amplified using pACTII-specific primers with AmpliTAQ Gold or the Expand Long Template PCR System (Roche Molecular Biochemicals) according to manufacturer instructions. Amplified fragments were sequenced with ABI Prism Big Dye Terminator (Applied Biosystems, Foster City, CA). Shortened KCC3a-bait constructs were made by PCR usingEcoRI- or BamHI-containing primers or for shorter constructs (40 bp or less) by using annealed oligonucleotides (adapters) with the same restriction sites. For the kinase (fish) constructs, EcoRI and XhoI sites were added to PCR primers or to adaptors to allow for subcloning into pACTII. Small scale transformations of PJ69-4A yeast were performed to test for interaction between specific bait and fish constructs. Interaction was tested on triple dropout (−Ura−Leu−His) + 2 mm3-amino-1,2,4-triazole plates; transformation efficiency was tested throughout on double dropout (−Ura−Leu) control plates. The domain of SPAK was from the library in pACTII using a sense primer that contained a restriction site and of the sequence and an antisense primer that contained bp specific to SPAK, a stop and an restriction site. The PCR fragment was with and into at sites. The vector was into cells using (Roche Molecular Cells were lysed with buffer J. Molecular A Scholar) on for a and at for 10 at were used in GST pull-down KCC3 constructs were into the vector at the EcoRI sites. GST proteins were in on and with cell several proteins were with glutathione and to Molecular Biochemicals) was used at a of from were in lysis buffer that contained mm pH 2 mm and (Roche Molecular were on for for on for an additional 2 and at for The was at in the volume of the brain was to 1 by of lysis was by of anti-NKCC1 Delpire E. J. Clin. Invest. 1996; 98: PubMed Scopus Google Scholar) or 10 of to the brain at followed by the of of for 2 at The were three with 1 of lysis buffer at 2 The were in of buffer that contained at for and to or were in buffer at for and on a The were and membranes were for 2 at in in with of was performed at at in were in for in secondary in for 1 at and for 2 in were by A to of the mouse SPAK was fused to The protein was in E. and with The was from GST using were by of the into two The from the first was as J. Mount D.B. Delpire E. 2001; Scopus Google Scholar). sequence between SPAK and OSR1 showed of in the region to the were performed with the on and proteins at different showed that the an two of for SPAK for OSR1 not from and NKCC1 null were and in at were from an mouse that been with for and pH for were in at for 2 were with in for a of that contained choroid plexus or salivary were on were with in a that contained in PBS, for at with followed by with SPAK NKCC1 Delpire E. J. Clin. Invest. 1996; 98: PubMed Scopus Google Scholar) at a of or at a of in at in PBS, were with in at in the They were with and with CA). additional brain from and NKCC1 null were with according to standard and with were with a with an CA). During analysis of yeast two-hybrid clones that three of nutritional selection for we performed liquid assays on clones that a one was with a activity of with the of not analysis revealed that this the domain binding of the serine-threonine kinase SPAK, which was first from brain by H. K. K. 1998; PubMed Scopus Google Scholar). the of with a domain at the terminus and the domain at the amino SPAK the germinal center kinase the of which is 1 A search identified a OSR1 that an with SPAK The OSR1 was identified on a search for in M. J. 1999; PubMed Scopus Google Scholar). OSR1 and SPAK have a The amino is the domain and in the domain of the of a in the of the OSR1 the region of SPAK that to we tested SPAK constructs of with KCC3a in yeast two-hybrid The of the constructs is shown in 1 the domain at the by amino acids and a fragment of residues the interaction In contrast, the by amino acids the interaction the terminus of SPAK is essential for the binding of SPAK to The fragment that tested for interaction the amino acids In we found that the region of OSR1 that is to SPAK also with KCC3a in yeast two-hybrid we identified the amino acids of KCC3a that provide the site of KCC3 interaction with the kinase 2 We KCC3a constructs of in the vector pGBDUc2 and that the fragment that with SPAK nine amino acids residues of a of the amino of cation chloride cotransporters, we found of the interacting region of KCC3 in members of the cation chloride NKCC1, and is the of and residues in cotransporters 2 In contrast, no was found in KCC2 and the yeast two-hybrid we tested for the of to interact with SPAK and For both interaction with NKCC1, and NKCC2 but not with KCC1 and KCC4 2 For we tested an fragment and two fragments of 10 2 and amino of with SPAK and OSR1 with the same analysis that the stress-related kinases SPAK and OSR1 interact with several members of the cation chloride with the same and that nine amino acids are to We the of which of the nine amino acids are essential for the interaction by at of KCC3 and For both KCC3a and NKCC2 the interaction with SPAK or OSR1 in yeast two-hybrid assays was at and and the mutation of the in NKCC2 but not of in KCC3a into at and yeast survival, at and not the of the motif from the or to the of which that the five amino acids of the motif are for the of a consensus motif for by interacting and members of the cotransporter family to and no consensus found. are in the in A. in this are results from additional amino with of the into a at 1 or of the into at not However, of the at 1 into a interaction we also tested the of the motif by the sequence of We to of yeast which that the motif not the interaction not the of the motif we performed a search of protein the of the of amino in we that on the motif found in amino In protein families the motif was the for proteins the (R/K)FX(V/I) motif was found in residues as a in is the that cation chloride cotransporters the motif with the protein families We also tested the motif was the cation chloride cotransporters. motif was found with a in amino as by the in interaction between several cation chloride cotransporters and the stress kinases SPAK and OSR1 and the binding sites using yeast two-hybrid we confirmed the interaction in GST pull-down we tested the of bait bait and a of bait fused to GST, to interact in vitro with an SPAK shown in both bait A amino and the shorter bait amino the a of bait of to which that the interaction is not the of of the an initial of the specific interaction of SPAK and the KCC3 cotransporter and are with yeast two-hybrid that a to mutation in bait the We generated a residues of The was first by in an was in a mouse brain using but not of the with the a is in as the SPAK also OSR1 but with we used the SPAK in assays at to the of SPAK, we performed cellular on brain and SPAK in both the membrane and the were also with to the of SPAK from proteins shown in the in the membrane which that a of SPAK is tightly to membrane the protein is in a cell in the GST pull-down the interacting is in in the that the interaction is and not physiological this we performed co-immunoprecipitation from brain that that the was in We used anti-NKCC1 to the cotransporter from mouse brain and from choroid plexus and for the shown in we SPAK in the from brain and to a from choroid plexus with the we from mouse brain and protein from mouse as well as the SPAK to the for SPAK and from the of the in the of interaction in 2000; PubMed Scopus Google Scholar) have shown that SPAK mitogen-activated protein kinase but not extracellular signal-regulated or we performed a co-immunoprecipitation experiment using the to the of with We confirmed the of by direct analysis of brain with 2 the of the co-immunoprecipitation with which a that in with the plexus cells a distinct of NKCC1 expression on the apical the membrane is of the expression of NKCC1 on the apical membrane of choroid plexus, we tested the expression of SPAK in this we that SPAK the apical membrane but not the membrane of choroid plexus epithelial cells this expression that of is the that in choroid plexus from NKCC1 J. R. 1999; PubMed Scopus (332) Google Scholar), which a for NKCC1 SPAK immunostaining is found in the the of the choroid plexus cells in NKCC1 null with mice, we brain with in were between the choroid of the two and the of NKCC1 expression the membrane expression of we the of the NKCC1, the is at the apical membrane of choroid plexus epithelial cells of the was not in the NKCC1 mouse We the expression of SPAK in the salivary a epithelium that NKCC1 at the membrane M. Forbush B.I. Annu. Rev. Physiol. 2000; 62: 515-534Crossref PubMed Scopus (332) Google Scholar, 6Russell J.M. Physiol. Rev. 2000; 80: 211-276Crossref PubMed Scopus (737) Google Scholar). In this the NKCC1 the membrane of the epithelial cells immunostaining for SPAK was found to to the membranes The apical which the and a of total membrane were of NKCC1 and SPAK the choroid plexus and salivary of SPAK at epithelial membranes that Furthermore, the of SPAK immunostaining in NKCC1 null that of SPAK to the membrane of choroid plexus epithelium the expression of of SPAK on NKCC1 the physiological of the with yeast GST and co-immunoprecipitation experiments. The activity of cation chloride cotransporters is during the cellular response to stress The of the proteins in this is In the study we identified two stress-related group SPAK and OSR1, as direct interactors of the cation chloride cotransporters. kinases interact with the amino terminus of NKCC1, and KCC3a and but not of KCC1 and a of to a I. A. Biol. 2001; Full Text Full Text PDF PubMed Scopus Google Scholar), group kinases into two the kinases and the the of this SPAK and OSR1, with one and to the that as a a region of we found that this specific region of the terminus with the cotransporters, that this fragment a role to specific and/or to a specific signaling have been or the physiological or the of SPAK and OSR1 have not been However, immunohistochemical of the SPAK in Ste20-related kinase that is in cells that are active in The was in choroid plexus epithelial cells of the distal cells of the in kidney, and epithelial cells of the H. K. K. 1998; PubMed Scopus Google Scholar). expression is in with the expression of cation chloride cotransporters found in Refs. 5Haas M. Forbush B.I. Annu. Rev. Physiol. 2000; 62: 515-534Crossref PubMed Scopus (332) Google and 8Delpire E. Mount D.B. Annu. Rev. Physiol. 2002; 64: 803-843Crossref PubMed Scopus (188) Google Scholar). We found a colocalization of SPAK and NKCC1 in epithelial for both proteins was on the apical membrane in choroid plexus, and on the membrane in salivary The that in the of SPAK from the membrane to the in choroid plexus the colocalization and interaction of SPAK and NKCC1 in this is possible that the of SPAK in NKCC1 is by a secondary on the of the sequence analysis using that the kinase OSR1 M. J. 1999; PubMed Scopus Google Scholar) is to SPAK, with an amino of The of interaction of SPAK and OSR1 at first a of their In this analysis that OSR1 expression is strong in such as the intestine, and M. J. 1999; PubMed Scopus Google Scholar), expression was to or in H. K. K. 1998; PubMed Scopus Google Scholar). that the expression in which the interaction analysis of in expression at the cellular for the binding motif of SPAK and OSR1 we found that nine amino acids of NKCC2 and KCC3 are for the The motif the amino residues (R/K)FX(V/I) followed by five additional residues with the of the (R/K)FX(V/I) motif in cation chloride cotransporters with protein we found that the cotransporters the motif with the the of the interaction and that cation chloride cotransporters are one of the of SPAK and of the NKCC1 protein revealed that the binding motif in the amino We that the second binding site also with SPAK in a yeast two-hybrid not the second binding site with the binding site for PP1 on NKCC1 (13Darman R.B. Flemmer A. Forbush B.I. J. Biol. Chem. 2001; 276: 34359-34362Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar). of this PP1 site the Cl− activation of NKCC1 with a of the protein (13Darman R.B. Flemmer A. Forbush B.I. J. Biol. Chem. 2001; 276: 34359-34362Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar). we not an of the PP1 and SPAK/OSR1 on cation chloride cotransporters, we that a between SPAK/OSR1 and PP1 is to a for cotransporters. SPAK and OSR1 are the cotransporters and thereby their However, we found that of the two SPAK binding in NKCC1, which the binding of the kinase in yeast two-hybrid and GST pull-down no on the activity of the cotransporter and as in not Furthermore, we that SPAK and OSR1 not interact with KCC1 and both of which have been shown to activated by (1Gillen C.M. Brill S. Payne J.A. Forbush B.I. J. Biol. Chem. 1996; 271: 16237-16244Abstract Full Text Full Text PDF PubMed Scopus (341) Google Scholar, D.B. A. Song J. George Delpire E. J. Biol. Chem. 1999; 274: Full Text Full Text PDF PubMed Scopus Google Scholar, A. Song L. Mount D.B. J. Biol. Chem. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar), that kinases play a role in cell volume on phosphorylation or of the cotransporters. Several lines of evidence to a role for SPAK and OSR1 in cell and/or cell signaling and 2000; PubMed Scopus Google Scholar) several on the domain of SPAK, an domain that is a site and a sequence and this kinase to signaling pathways. The 2000; PubMed Scopus Google Scholar) showed that a of SPAK that to SPAK is in the the protein is in the cytoplasm. SPAK, OSR1 an domain as well as a both kinases as signaling in the and in the is the that SPAK expression is not to the cytoplasm. studies with mouse SPAK as well as with the H. K. K. 1998; PubMed Scopus Google Scholar) showed of protein in the and in the in both studies of the not to of the which with the membrane or with one of Furthermore, studies in choroid plexus that a of the kinase is tightly with the plasma membrane. of the is the A study performed evidence for the role of in H. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar). the study also that the was of a of functional of the and the KCC3 also of peripheral and a of peripheral Mount D.B. J. Song L. R. R. A. George J. A. J. Delpire E. 2002; PubMed Scopus Google Scholar) that is of the in showed that the of that in is not by the H. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar). that activity is in The of the kinase in is by of expression during and in the H. K. K. 1998; PubMed Scopus Google Scholar). also a role for the kinases in cell of to activate one of the three well stress in SPAK was in cellular stress response 2000; PubMed Scopus Google Scholar). SPAK the but not the or pathways. with SPAK in possible that and SPAK a in vivo in which is stress Several to group members as of kinase kinases SPAK/OSR1 and cation chloride cotransporters in the activation of the as to In support of this is the that in the thick ascending limb of Henle a of the to an in expression that by J.A. J. Clin. Invest. 2000; PubMed Scopus Google Scholar). a the J.A. J. Clin. Invest. 2000; PubMed Scopus Google Scholar) directly the activation of the kinase on a in the of the an of the Na-K-2Cl also in is to that the interaction between SPAK/OSR1 and cation chloride cotransporter is in stress is and of the Am. J. Physiol. 2000; PubMed Google Scholar), the initial cellular and study the that the between cation chloride cotransporters and the stress kinases SPAK and OSR1 a that the cellular response to