Advertisement

The Spectrin Cytoskeleton Influences the Surface Expression and Activation of Human Transient Receptor Potential Channel 4 Channels*

  • Adam F. Odell
    Correspondence
    To whom correspondence should be addressed (current address): University of Leeds, Leeds, United Kingdom. Tel.: 44-113-343-7728; Fax: 44-113-343-6603
    Affiliations
    School of Biomedical Sciences, Faculty of Health, University of Newcastle, Level 5, MSB, University Drive, New South Wales 2308, Australia
    Search for articles by this author
  • Dirk F. Van Helden
    Footnotes
    Affiliations
    School of Biomedical Sciences, Faculty of Health, University of Newcastle, Level 5, MSB, University Drive, New South Wales 2308, Australia
    Search for articles by this author
  • Judith L. Scott
    Affiliations
    School of Biomedical Sciences, Faculty of Health, University of Newcastle, Level 5, MSB, University Drive, New South Wales 2308, Australia
    Search for articles by this author
  • Author Footnotes
    * This work was supported by the National Health and Medical Research Council (NHMRC) of Australia, the Australian Research Council, and the Hunter Medical Research Institute. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
    The on-line version of this article (available at http://www.jbc.org) contains supplemental Fig. S1 and Table S1.
    2 An NHMRC Principal Research Fellow and Brawn Senior Fellow.
Open AccessPublished:November 29, 2007DOI:https://doi.org/10.1074/jbc.M709729200
      Despite over a decade of research, only recently have the mechanisms governing transient receptor potential channel (TRPC) channel function begun to emerge, with an essential role for accessory proteins in this process. We previously identified a tyrosine phosphorylation event as critical in the plasma membrane translocation and activation of hTRPC4 channels following epidermal growth factor (EGF) receptor activation. To further characterize the signaling events underlying this process, a yeast-two hybrid screen was performed on the C terminus of hTRPC4. The intracellular C-terminal region from proline 686 to leucine 977 was used to screen a human brain cDNA library. Two members of the spectrin family, αII- and βV-spectrin, were identified as binding partners. The interaction of hTRPC4 with αII-spectrin and βV-spectrin was confirmed by glutathione S-transferase pulldown and co-immunoprecipitation experiments. Deletion analysis identified amino acids 730-758 of hTRPC4 as critical for the interaction with this region located within a coiled-coil domain, juxtaposing the Ca2+/calmodulin- and IP3R-binding region (CIRB domain). This region is deleted in the proposed δhTRPC4 splice variant form, which failed to undergo both EGF-induced membrane insertion and activation, providing a genetic mechanism for regulating channel activity. We also demonstrate that the exocytotic insertion and activation of hTRPC4 following EGF application is accompanied by dissociation from αII-spectrin. Furthermore, depletion of αII-spectrin by small interference RNA reduces the basal surface expression of αhTRPC4 and prevents the enhanced membrane insertion in response to EGF application. Importantly, depletion of αII-spectrin did not affect the expression of the δ variant. Taken together, these results demonstrate that a direct interaction between hTRPC4 and the spectrin cytoskeleton is involved in the regulation of hTRPC4 surface expression and activation.
      Recently, the seven mammalian members of the canonic or classic TRPC
      The abbreviations used are: hTRPC4
      human transient receptor potential channel 4
      EGF
      epidermal growth factor
      EGFR
      epidermal growth factor receptor
      GST
      glutathione S-transferase
      CIRB
      Ca2+/calmodulin- and IP3R-binding region
      NHERF
      Na+/H+-exchanger regulatory factor 1
      STK
      Src-family tyrosine kinase
      aa
      amino acid(s)
      CMV
      cytomegalovirus
      HA
      hemagglutinin
      siRNA
      small interference RNA
      RIPA
      radioimmune precipitation assay
      GFP
      green fluorescent protein
      ERK
      extracellular signal-regulated kinase.
      3The abbreviations used are: hTRPC4
      human transient receptor potential channel 4
      EGF
      epidermal growth factor
      EGFR
      epidermal growth factor receptor
      GST
      glutathione S-transferase
      CIRB
      Ca2+/calmodulin- and IP3R-binding region
      NHERF
      Na+/H+-exchanger regulatory factor 1
      STK
      Src-family tyrosine kinase
      aa
      amino acid(s)
      CMV
      cytomegalovirus
      HA
      hemagglutinin
      siRNA
      small interference RNA
      RIPA
      radioimmune precipitation assay
      GFP
      green fluorescent protein
      ERK
      extracellular signal-regulated kinase.
      subfamily (TRPC1-7) have emerged as strong candidates for regulating non-excitable Ca2+ entry. They can be broadly classified into two subgroups based on homology and activation mechanisms. The TRPC3/6/7 subgroup responds to the generation of diacylglycerol following receptor-mediated activation of phospholipase C, whereas members of the TRPC1/4/5 subgroup demonstrate both store-operated and receptor-mediated properties. The store-operated phenotype is typified by the reduction in thapsigargin-induced Ca2+ entry in endothelial cells derived from the TRPC4 null mouse (
      • Freichel M.
      • Suh S.H.
      • Pfeifer A.
      • Schweig U.
      • Trost C.
      • Weissgerber P.
      • Biel M.
      • Philipp S.
      • Freise D.
      • Droogmans G.
      • Hofmann F.
      • Flockerzi V.
      • Nilius B.
      ,
      • Tiruppathi C.
      • Freichel M.
      • Vogel S.M.
      • Paria B.C.
      • Mehta D.
      • Flockerzi V.
      • Malik A.B.
      ). An involvement in receptor-mediated Ca2+ entry was also demonstrated in the TRPC4 null mouse as well as in HEK293 cells treated with TRPC4 antisense oligonucleotides, which exhibited impaired carbachol-mediated oscillations (
      • Freichel M.
      • Suh S.H.
      • Pfeifer A.
      • Schweig U.
      • Trost C.
      • Weissgerber P.
      • Biel M.
      • Philipp S.
      • Freise D.
      • Droogmans G.
      • Hofmann F.
      • Flockerzi V.
      • Nilius B.
      ,
      • Tiruppathi C.
      • Freichel M.
      • Vogel S.M.
      • Paria B.C.
      • Mehta D.
      • Flockerzi V.
      • Malik A.B.
      ,
      • Wu X.
      • Babnigg G.
      • Zagranichnaya T.
      • Villereal M.L.
      ). TRPC4 has also recently been identified as an essential component of EGF receptor-stimulated Ca2+ influx (
      • Yang H.
      • Mergler S.
      • Sun X.
      • Wang Z.
      • Lu L.
      • Bonanno J.A.
      • Pleyer U.
      • Reinach P.S.
      ); however, the precise nature of the signaling pathway regulating activation of Ca2+ entry following receptor stimulation and store-depletion remains unclear. The organization of TRPC channels into signaling complexes is proposed as critical in regulating the activation of TRPC channels and controlling their physiological functions (
      • Ambudkar I.S.
      ). Through interactions with various protein and lipid molecules, some of which remain unknown, TRPC channels are assembled into functional signaling microdomains (
      • Kiselyov K.
      • Shin D.M.
      • Kim J.Y.
      • Yuan J.P.
      • Muallem S.
      ). Definitive identification and characterization of the microdomains involved in TRPC-regulated Ca2+ entry are necessary to fully elucidate the mechanisms of store- and receptor-operated Ca2+ entry. Recent efforts have identified several potential components of the signaling machinery involved in TRPC4-mediated Ca2+ entry. These include the Na+/H+-exchanger regulatory factor 1 (NHERF), various IP3R subtypes, Ca2+-calmodulin, immunophilins, protein 4.1, Homer and stromal interacting protein 1 (
      • Sinkins W.G.
      • Goel M.
      • Estacion M.
      • Schilling W.P.
      ,
      • Yuan J.P.
      • Kiselyov K.
      • Shin D.M.
      • Chen J.
      • Shcheynikov N.
      • Kang S.H.
      • Dehoff M.H.
      • Schwarz M.K.
      • Seeburg P.H.
      • Muallem S.
      • Worley P.F.
      ,
      • Kiselyov K.
      • Kim J.Y.
      • Zeng W.
      • Muallem S.
      ,
      • Cioffi D.L.
      • Wu S.
      • Alexeyev M.
      • Goodman S.R.
      • Zhu M.X.
      • Stevens T.
      ,
      • Yuan J.P.
      • Zeng W.
      • Huang G.N.
      • Worley P.F.
      • Muallem S.
      ). Most proteins have been found to bind to discrete domains within TRPC4, such as the C-terminal CIRB domain for which Ca2+-calmodulin and IP3Rs compete (
      • Zhu M.X.
      • Tang J.
      ,
      • Kiselyov K.
      • Mignery G.A.
      • Zhu M.X.
      • Muallem S.
      ). Removal of the CIRB domain from the βTRPC4 splice variant was reported to enhance carbachol-mediated Ca2+ entry in HEK293 cells, and led to the suggestion of the less plentiful β variant being the active form of the channel (
      • Schaefer M.
      • Plant T.D.
      • Stresow N.
      • Albrecht N.
      • Schultz G.
      ,
      • Trost C.
      • Bergs C.
      • Himmerkus N.
      • Flockerzi V.
      ,
      • Tang J.
      • Lin Y.
      • Zhang Z.
      • Tikunova S.
      • Birnbaumer L.
      • Zhu M.X.
      ). Additionally, the role of NHERF in controlling the surface expression and localization of TRPC4 is also well documented (
      • Tang Y.
      • Tang J.
      • Chen Z.
      • Trost C.
      • Flockerzi V.
      • Li M.
      • Ramesh V.
      • Zhu M.X.
      ,
      • Obukhov A.G.
      • Nowycky M.C.
      ,
      • Mery L.
      • Strauss B.
      • Dufour J.F.
      • Krause K.H.
      • Hoth M.
      ), and recently a role for the spectrin cytoskeleton in the control of endothelial ISOC was identified where a spectrin-protein 4.1 interaction influenced Ca2+ entry following thapsigargin treatment (
      • Wu S.
      • Sangerman J.
      • Li M.
      • Brough G.H.
      • Goodman S.R.
      • Stevens T.
      ).
      Spectrins are ∼200-400 kDa proteins that form extended, flexible protein dimers of ∼200-260 nm in length and 3-6 nm across and contain actin-binding domains at each end. Spectrin dimers are composed of α- and β-subunits, which associate laterally to form anti-parallel heterodimers. These heterodimers then assemble head-to-head to form heterotetramers, which are organized in a polygonal network of 5-7 extended spectrin molecules linked to short actin filaments (
      • Bennett V.
      • Baines A.J.
      ). This spectrin-actin network is coupled to the plasma membrane primarily by association with ankyrin, which in turn is bound to dimers of the Na+/H+ exchanger (
      • Hryniewicz-Jankowska A.
      • Czogalla A.
      • Bok E.
      • Sikorsk A.F.
      ). Additional connections are formed between band 4.2, protein 4.1, the MAGUK p55, and glycophorin C, which associate at the spectrin-actin junction (
      • Bennett V.
      • Baines A.J.
      ). Spectrin is now recognized as a ubiquitous scaffolding protein that acts in conjunction with a variety of adaptor proteins to organize membrane microdomains on both the plasma membrane and on intracellular organelles (
      • Wechsler A.
      • Teichberg V.I.
      ,
      • Bloch R.J.
      • Bezakova G.
      • Ursitti J.A.
      • Zhou D.
      • Pumplin D.W.
      ,
      • Bennett V.
      ). Spectrin appears to organize and control the localization and function of integral membrane proteins by creating meshworks with other proteins (
      • Bennett V.
      • Baines A.J.
      ). A role for TRPC4 in mediating the interaction between the spectrin cytoskeleton and Ca2+ entry was recently proposed, with disruption of the protein 4.1-binding site on TRPC4 preventing activation of the endothelial ISOC channel (
      • Cioffi D.L.
      • Wu S.
      • Alexeyev M.
      • Goodman S.R.
      • Zhu M.X.
      • Stevens T.
      ,
      • Cioffi D.L.
      • Wu S.
      • Stevens T.
      ).
      Here, we have identified a direct coupling between αII-spectrin and hTRPC4. This is dependent on a coiled-coil domain located near the proline-rich, protein 4.1-binding motif of hTRPC4 and is mediated by the C-terminal spectrin repeats of αII-spectrin. Previous experiments demonstrated enhanced insertion of hTRPC4 into the plasma membrane following activation of the EGF receptor (
      • Odell A.F.
      • Scott J.L.
      • Van Helden D.F.
      ). Here we report that this insertion is accompanied by the dissociation of spectrin-hTRPC4 interaction. Channels lacking the spectrin-binding site failed to undergo membrane insertion and were unable to be activated by EGF. This suggests that a direct association with the spectrin cytoskeleton may influence the signaling events involved in controlling the level of active hTRPC4 channels expressed at the cell surface.

      EXPERIMENTAL PROCEDURES

      Molecular Biology–The C terminus of hTRPC4 (CT, aa 686-977) and various truncations (CT1, aa 686-784; CT2, aa 686-859; CT3, aa 759-859; CT4, aa 759-977; and CT5, aa 846-977) were amplified by PCR from pCMVSport-hTRPC4 (kindly provided by Dr. J. W. Putney (
      • McKay R.R.
      • Szymeczek-Seay C.L.
      • Lievremont J.P.
      • Bird G.S.
      • Zitt C.
      • Jungling E.
      • Luckhoff A.
      • Putney Jr., J.W.
      )) and cloned into the BamH1 and Sal1 restriction sites of pBTM116 in frame with LexA DNA-binding domain. A full list of primers can be found insupplemental Table S1. Truncations of hTRPC4 were co-transformed with pACT2-αII-spectrin into L40 yeast and assessed for reporter gene activation. For GST fusion protein construction, the C-terminal fragments of hTRPC4 (CT and CT1-CT5) were restriction-digested from pBTM116 with EcoR1 and Sal1 and subcloned into the corresponding sites in pGEX-4T-1. The C terminus of αII-spectrin (aa 1998-2474) was amplified by PCR using the primers outlined insupplemental Table S1 and the insert ligated into the Sal1 restriction site of pBTM116. The C terminus of αII-spectrin was amplified by PCR from clone 18 and ligated into the EcoR1 and Sal1 restriction sites of pAS2.1. The C-terminal fragments of βV-spectrin and αII-spectrin were amplified by PCR from clones 52 and 18, respectively, using the primers outlined insupplemental Table S1 and the PCR fragments cloned into pGEX-4T-1 vector for expression of GST fusion proteins.
      For the generation of the αII-spectrin GST fusion protein truncations, various fragments were amplified by PCR from pBluescript-αII-spectrin (kindly supplied by Dr. J. S. Morrow). Seesupplemental Table S1 for primers where, shown in brackets, are the corresponding beginning and final αII-spectrin amino acids of the truncations. PCR inserts were digested with the relevant restriction enzymes, cloned into pGEX-4T-1 and constructs assessed by DNA sequencing and the induction of protein synthesis. The further αII-spectrin truncation, αII-(951-2472), was generated by PCR from pBluescript-αII-spectrin. PCR products were ligated into Sal1 and Not1 restriction sites of pGEX-4T-1.
      To generate the bacterial constructs of full-length αII-spectrin for GST-tagged protein generation, a 1000-bp PCR fragment corresponding to the 5′-end of the αII-spectrin was amplified and ligated into pBluescript digested with Kpn1 and Xba1 restriction sites using the following primer pair; For, ACGGGTACCAGTCGACATGGACCCAAGTGGGGTCAAAGTGCTG and Rev, TGCTCTAGAGCAGCAAGATCTCTCTCGAGACCCTC. For the generation of full-length untagged αII-spectrin GST fusion protein, an Sal1 site was used after the 5′ Kpn1 restriction site. This produced pBluescript constructs containing the 5′ end of αII-spectrin with either Kpn1 or Sal1 restriction sites. The pBluescript-αII-spectrin construct also contained a Not1 restriction site at the 3′ end. Full-length αII-spectrin was subsequently subcloned from pBluescriptII-αII-spectrin with Sal1 and Not1 and ligated into pGEX-4T-1. All constructs were confirmed by DNA sequencing and assessed for full-length protein expression by Western blotting.
      The β and δ splice variants of hTRPC4 as well as the M1 (MUT1), M2 (MUT2), and M3 (MUT3)-hTRPC4 deletion mutants were generated using a modified linker-scanning mutation approach. Briefly, primers were designed flanking the regions to be excised and two PCR fragments generated encompassing the 5′ and 3′ segments on either side of the proposed deletion. These two PCR products were combined and utilized as the template for a second round of PCR amplification with the forward (containing the FLAG epitope), and reverse primers were utilized in the first PCR round covering the start and stop codons of hTRPC4. The primers are outlined insupplemental Table S1. The resultant products were assessed for the absence of intended C-terminal regions by restriction digestion and DNA sequencing. Correct expression of the protein product was assessed by transient transfection and Western blotting with anti-FLAG antibody.
      Yeast Media and Manipulation–The L40, Y187, and Y190 Saccharomyces cerevisiae yeast strains were grown at 30 °C on YPD media (10 g/liter yeast extract, 20 g/liter peptone, 0.1 g/liter adenine, 2% dextrose) or synthetic minimal media with the appropriate supplements. Synthetic minimal media consisted of Dropout base powder (1.7 g/liter yeast nitrogen base, 20 g/liter dextrose, 5 g/liter ammonium sulfate, BIO 101, Vista, CA) supplemented with the appropriate synthetic media formulation (CSM, BIO 101). The synthetic (YC) complete media was then autoclaved at 121 °C for 15-20 min with or without agar (25 g/liter) and stored at 4 °C in the dark. Tryptophan was removed to select the bait vectors, pBTM116, pFBL23, and pAS2.1, and leucine was omitted to select for the cDNA library pACT2 vector. Lysine and uracil were omitted from all YC synthetic media to maintain selection of the integrated HIS3 and lacZ reporter genes, respectively, in L40 and Y190 yeast. Histidine was removed when selecting for interacting bait and prey proteins. In all yeast YC media, lysine and uracil are absent. All the initial characterizations of bait constructs and library screens were performed in the L40 strain of Saccharomyces cerevisiae (
      • Hollenberg S.M.
      • Sternglanz R.
      • Cheng P.F.
      • Weintraub H.
      ). Two bait vectors were utilized to investigate hTRPC4 binding partners: the LexA-based bait vector, pBTM116, and the GAL4-based pAS2.1 vector (Clontech). The prey vector used was the pACT2 GAL4-based DNA-activation domain vector (Clontech). The human brain cDNA library was purchased already cloned into the GAL4 DNA-activation domain of pACT2 prey plasmid and titered according to the manufacturer's instructions.
      Antibodies–Anti-M2 FLAG antibody and anti-α-tubulin antibodies were obtained from Sigma-Aldrich. Anti-PY20 antibody was supplied by BD Transduction Laboratories, and the 4G10 and anti-β1-integrin antibodies were from Upstate Biotechnology. Anti-TRPC4 rabbit polyclonal antibody was from Alomone Laboratories (Israel). The anti-TRPC4 (C20) goat and anti-FYN (FYN3) rabbit polyclonal antibodies were purchased from Santa Cruz Biotechnology. Anti-EGFR mouse monoclonal antibody was kindly supplied by Dr. M. Hibbs. Anti-HA monoclonal, rabbit polyclonal anti-αII-spectrin, and rabbit anti-phospho-ERK1/2 monoclonal antibodies were from Cell Signal Technologies. The rabbit polyclonal anti-NHERF antibody was from Abcam Laboratories. All secondary antibodies, anti-GST antibody and streptavidin-horseradish peroxidase were from Amersham Biosciences, GE, except anti-goat AlexaFluor488, which was from Molecular Probes.
      Cell Culture, Transfection, and Immunofluorescence Microscopy–Cell lines were maintained in Dulbecco's modified Eagle's medium containing 2 mm l-glutamine, 10% fetal bovine serum at 37 °C in 5% CO2. Cell culture reagents were obtained from Invitrogen. HEK293 cells stably expressing FLAG-hTRPC4 constructs were maintained in the above media containing 400 μg/ml Zeocin (Invitrogen). COS-7 and HEK293 cells were transfected with expression plasmids with Lipofectamine 2000 (Invitrogen) according to manufacturer's instructions, and 35-60% transfection efficiency was routinely achieved. siRNA depletion of αII-spectrin was performed by reverse transfection using Lipofectamine RNAiMAX reagent (Invitrogen). Control and αII-spectrin (sc-36549) siRNA were obtained from Santa Cruz Biotechnology. For immunofluorescence detection, transfected cells were fixed in 4% paraformaldehyde and stained with anti-TRPC4 C20 antibody followed by anti-goat AlexaFluor488 secondary antibody. Nuclei were visualized by 4′,6-diamidino-2-phenylindole staining. Cells were mounted with Anti-Fade (DakoCytomation) and observed under 63× oil-immersion objective on a Zeiss Axiovert 100M confocal microscope and analyzed with Zeiss LSM510.
      In Vitro Binding Assays, Immunoprecipitation, and Immunoblotting–GST fusion proteins were generated as outlined in a previous study (
      • Odell A.F.
      • Scott J.L.
      • Van Helden D.F.
      ). For GST-pulldown assays equal volumes of immobilized GST fusion proteins were incubated with mammalian cell lysates containing 1-3 mg of protein at 4 °C for 2 h or overnight. Beads were washed four times with lysis buffer, and bound proteins were resolved by 8% SDS-PAGE and visualized by Western blotting on nitrocellulose as indicated. For preparation of lysates of COS-7, HEK293, and A431 cells, cells were washed with ice-cold phosphate-buffered saline, lysed with RIPA buffer, and equal amounts of protein subjected to immunoprecipitation with protein G-Sepharose (GE, Amersham Biosciences) as outlined previously (
      • Odell A.F.
      • Scott J.L.
      • Van Helden D.F.
      ). For immunoprecipitation from rat brain membranes, lysates were pre-cleared with protein G-Sepharose followed by centrifugation. Immune complexes were washed four times with lysis buffer and resolved, along with lysates by 8% SDS-PAGE followed by Western blotting onto Hybond ECL nitrocellulose membrane (GE, Amersham Biosciences). Membranes were blocked and probed with indicated antibodies for 2 h or overnight and visualized with ECL Plus reagent on Typhoon Imaging System (GE, Amersham Biosciences). For quantification, the protein bands were analyzed with ImageQuaNT software.
      Biotinylation of Cell Surface Proteins–Following treatment, indicated cells were biotinylated with EZ-Link Sulfyl-NHS-Biotin (Pierce Biotechnology) as outlined previously (
      • Odell A.F.
      • Scott J.L.
      • Van Helden D.F.
      ). Lysates were then prepared by addition of modified RIPA buffer, and equal quantities of protein were either incubated with 50% NeutrAvidin-agarose slurry (Pierce Biotechnology) or immunoprecipitated with the appropriate antibody for 2 h at 4 °C. Beads were washed three times with lysis buffer, subjected to SDS-PAGE, and analyzed by Western blotting.
      Purification of Rat Brain Microsomes–Adult Sprague-Dawley rats were killed by decapitation according to Institutional Animal Ethics, and cortex tissues were dissected and homogenized in 10 volumes of ice-cold homogenization buffer (320 mm sucrose, 1 mm EDTA (pH 7.5)). The homogenate was centrifuged at 1,000 × g for 10 min, and the supernatant at 15,000 × g for 30 min at 4 °C. The pellet (P2) was resuspended in RIPA buffer, and insoluble material was removed by centrifugation. 1 mg of soluble rat brain protein was used for immunoprecipitation with 4 μg of the indicated antibodies, and 25 μl of protein G-Sepharose overnight at 4 °C. Washed immunocomplexes were analyzed as above.
      Ca2+ Imaging–Ca2+ imaging was performed as outlined previously (
      • Odell A.F.
      • Scott J.L.
      • Van Helden D.F.
      ). Briefly, cells were cultured on glass coverslips and loaded with 2-4 μm fura-2 acetoxymethyl ester in Dulbecco's modified Eagle's medium at 37 °C for 30 min and incubated in HEPES-buffered saline (140 mm NaCl, 5 mm KCl, 0.5 mm MgCl2, 5.5 mm HEPES, 10 mm glucose, 2 mm CaCl2, pH 7.4) for 30 min prior to stimulation. In experiments performed in Ca2+-free medium, cells were placed in HEPES-buffered saline without CaCl2 containing 200 μm EGTA. Changes in [Ca2+]i were monitored as 340/380 nm fluorescence ratio, and single cell imaging was performed on transfected HEK293 cells illuminated with Lambda DG-4 lamp, captured with an Orca ER camera, and analyzed with MetaFluor program (Universal Imaging Devices) under 40× objective lens. All statistics were performed using Prism, one-way analysis of variance, and the Tukey post test.

      RESULTS

      The C Terminus of hTRPC4 Binds to αII-Spectrin and βV-Spectrin–A yeast two-hybrid screen in L40 yeast of a human brain cDNA library generated in pACT2 was performed using a LexA fusion protein of the C terminus of hTRPC4 (hTRPC4-CT, aa 686-977) to identify intracellular proteins that interact with hTRPC4 and regulate its activation. A number of potential binding partners were identified, including the previously isolated hTRPC4-interacting protein, Na+/H+-exchanger regulatory factor or NHERF (1 out of 128 positive clones). A novel putative interaction was also identified between the cytoskeletal proteins, αII-spectrin (2 out of 128 positive clones) and βV-spectrin (10 out of 128 positive clones) and the C terminus of hTRPC4. The interactions between hTRPC4 and both αII-spectrin and βV-spectrin were confirmed by re-introduction of the constructs into L40 yeast in combination with control vectors (Table 1). Isolation and sequence analysis of the interacting spectrin clones revealed that hTRPC4-CT was binding to the C termini of both molecules; aa 1998-2472 of αII-spectrin and aa 3232-3674 of βV-spectrin. The interacting amino acids corresponded to spectrin repeats 19-21 and the C-terminal EF-hand domains of αII-spectrin and spectrin repeats 28-30 and PH domain of βV-spectrin.
      TABLE 1Summary of confirmatory transformations between hTRPC4 and either αII-spectrin or βV-spectrin
      BaitPreyHIS3lacZ
      pBTM116pACT2
      pBTM116-FospACT2-Jun+++
      pBTM116-LaminCpACT2
      pBTM116-CTpACT2
      pBTM116-CT-(686-977)pACT2-αII-(1998-2472)+++
      pBTM116-LaminCpACT2-αII-(1998-2472)
      pBTM116-CT-(686-977)pACT2-βV-(3232-3674)++
      pBTM116-LaminCpACT2-βV-(3232-3674)
      The interactions between hTRPC4-CT and αII-spectrin and βV-spectrin were confirmed by co-transformation of the interacting cDNAs into L40 yeast expressing the C terminus of hTRPC4 (pBTM116-CT). Neither clone isolated from the yeast two-hybrid screen interacted with the control LexA-LaminC protein when co-expressed with pBTM116-LaminC; however, both αII-spectrin and βV-spectrin activated expression of the lacZ and HIS3 reporter genes in L40 yeast, thereby confirming the specificity of their interaction with hTRPC4 (Table 1). Also shown in Table 1 are the negative (any assay with an empty vector) and positive control (pBTM116-Fos and pACT2-Jun) reactions. To further confirm the interaction between hTRPC4 and the spectrin molecules, the C terminus of hTRPC4 was cloned into an alternative bait vector, the GAL4 DNA-binding domain-based vector, pAS2.1 (pAS2.1-CT), and co-expressed with the pACT2-spectrin plasmids in Y187 yeast. Under these conditions, both αII-spectrin (Fig. 1A) and βV-spectrin (data not shown) induced activation of the lacZ reporter gene. There was no lacZ expression when either prey construct was co-expressed with empty pAS2.1 vector (Fig. 1A). Similarly, pAS2.1-CT failed to activate the reporter gene in the absence of spectrin fusion proteins (data not shown). Negative control transformations did not activate lacZ reporter gene expression (Fig. 1A). In addition, a domain exchange assay was performed between hTRPC4-CT transferred to pACT2 and αII-spectrin bait generated in pAS2.1. Upon co-transformation into Y187 yeast strain, only pAS2.1-αII-spectrin and pACT2-hTRPC4-CT bait-prey combinations induced reporter gene activation as assessed by colony-lift filter assay (Fig. 1A). No other control transformations induced lacZ gene activation (Fig. 1A). Similar results were obtained upon co-transformation of pBTM116-αII-spectrin and pACT2-CT in L40 yeast (data not shown). Taken together, these results confirm that the C terminus of hTRPC4 interacts specifically with αII-spectrin and βV-spectrin regardless of bait-prey combination and yeast strain utilized.
      Figure thumbnail gr1
      FIGURE 1The C terminus of hTRPC4 interacts with αII-spectrin and βV-spectrin. A, colony-lift filter assay was performed between αII-spectrin and hTRPC4. Y187 yeast were transformed with the indicated bait/prey plasmid combinations and selected on YC-Trp-Leu drop-out media. Single colonies were selected and re-streaked onto the same selective media prior to analysis of lacZ gene activation by colony-lift filter assay. B, GST-pulldown analysis of the interaction between hTRPC4 and αII-spectrin/βV-spectrin. Purified and immobilized GST, GST-αII-spectrin-CT, or GST-βV-spectrin were incubated with 1% Nonidet P-40-solubilized cell lysates of HEK293 cells transiently expressing FLAG-hTRPC4. Bound proteins were analyzed by Western blotting using anti-FLAG M2 antibody. Also shown is a Coomassie stain of GST fusion proteins. C, purified GST or GST-hTRPC4-CT were incubated with RIPA-solubilized HEK293 cell lysates, and bound proteins were analyzed by Western blotting with anti-αII-spectrin antibody.
      GST fusion protein pulldown assays were also performed to confirm and biochemically characterize the interaction between αII-spectrin, βV-spectrin, and hTRPC4. Recombinant GST fusion proteins of αII-spectrin (GST-αII-spectrin-CT) and βV-spectrin (GST-βV-spectrin-CT) regions were isolated from the two-hybrid screen, generated in BL21 Escherichia coli, and incubated overnight with Nonidet P-40 lysates of HEK293 cells transiently transfected with FLAG-tagged hTRPC4. Bound proteins were examined for the presence of hTRPC4 by SDS-PAGE and immunoblotting with anti-M2 FLAG antibody. As shown in Fig. 1B, both C-terminal αII-spectrin and βV-spectrin constructs isolated from the yeast two-hybrid screen were able to bind FLAG-hTRPC4 under the conditions examined. No significant interaction was detected between GST alone and hTRPC4 (Fig. 1B). The affinity of GST-αII-spectrin for FLAG-hTRPC4 was greater than that of βV-spectrin despite similar fusion protein input, suggesting a stronger association between the former proteins (Fig. 1B). This, combined with the known role of αII-spectrin in vesicle trafficking (
      • Sikorski A.F.
      • Terlecki G.
      • Zagon I.S.
      • Goodman S.R.
      ,
      • Iga M.
      • Inui M.
      • Sobue K.
      ) and the interaction of αII-spectrin with numerous Ca2+ signaling proteins, provided the impetus for further investigation of this putative interaction. Thus, the interaction between hTRPC4-CT and endogenous full-length αII-spectrin was also investigated in reverse pulldown assays using the immobilized C terminus of hTRPC4 (GST-hTRPC4-CT) incubated with RIPA-solubilized HEK293 extracts. Bound proteins were analyzed by SDS-PAGE and immunoblotting for αII-spectrin. As shown in Fig. 1C, full-length αII-spectrin (230 kDa) was present in precipitates from GST-hTRPC4-CT but not in those performed with GST alone. Equal quantities of fusion protein were included in each reaction.
      The Interaction between hTRPC4 and αII-Spectrin Is Mediated by Spectrin Repeats in the C Terminus of αII-Spectrin–The interaction of full-length αII-spectrin with hTRPC4 was also studied in GST pulldown assays. Full-length αII-spectrin (1-2474, 230 kDa) and truncated forms of the molecule (aa 951-2472 and 1998-2474) expressed as GST fusion proteins were able to bind FLAG-hTRPC4 (Fig. 2A). However, a decrease in the affinity of αII-spectrin for hTRPC4 was observed as the length of fusion protein increased (Fig. 2A), possibly indicating competition between other proteins in the HEK293 lysates and FLAG-hTRPC4 for αII-spectrin-binding sites. An interaction between full-length αII-spectrin and TRPC4 was also sought through co-immunoprecipitation experiments of FLAG-tagged murine TRPC4 from transiently transfected HEK293 cells. Proteins bound to anti-FLAG antibodies incubated with either control or FLAG-mTRPC4-transfected HEK293 cells were immunoblotted for the presence of αII-spectrin. Under these conditions αII-spectrin was shown to co-immunoprecipitate from cells expressing FLAG-mTRPC4 and not from control HEK293 cells transfected with empty vector (Fig. 2B) indicating the interaction is preserved between human and mouse. Additionally, NHERF was also present in the FLAG-mTRPC4 precipitates, indicating the formation of a complex between mTRPC4, αII-spectrin, and NHERF in HEK293 cells (Fig. 2B).
      Figure thumbnail gr2
      FIGURE 2hTRPC4 forms a complex with αII-specrin. A, GST, GST-αII-spectrin (aa 951-2472), GST-αII-spectrin (aa 1998-2472), and full-length GST-αII-spectrin (aa 1-2472) were incubated with lysates from HEK293 transiently expressing FLAG-hTRPC4, and bound proteins were analyzed by immunoblotting with anti-FLAG antibody. Total lysate, corresponding to 6% total input, is shown as a positive control. Also shown is a Coomassie stain of the GST fusion proteins. B, mTRPC4 was immunoprecipitated with anti-FLAG antibody from untransfected HEK293 cell lysates and cells expressing mTRPC4 and immunoblotted with anti-αII-spectrin, anti-FLAG, and anti-NHERF antibodies.
      Given the strong interaction between αII-spectrin and hTRPC4, mutation deletion analysis of αII-spectrin cDNA was used to define the region of the spectrin molecule responsible for binding to hTRPC4. The αII-spectrin fragment isolated from the yeast two-hybrid screen contained the final three spectrin repeats and the C-terminal EF1 and EF2 Ca2+-binding domains (see Fig. 3A, schematic). Previous reports have implicated spectrin repeats in mediating a variety of protein-protein interactions. For example, the interaction between the spectrin repeat-containing protein α-actinin-2 and the Kv1.5 potassium channel was dependent upon the presence of spectrin repeats but independent of the EF1 and EF2 domains of α-actinin (
      • Cukovic D.
      • Lu G.W.
      • Wible B.
      • Steele D.F.
      • Fedida D.
      ). Therefore, it was of interest to investigate the role of the spectrin repeats in binding to hTRPC4. A series of C-terminal truncated αII-spectrin constructs was generated as GST fusion proteins: GST-S1, spectrin repeat 19; GST-S2, spectrin repeats 19 and 20; GST-S3, spectrin repeats 19, 20, and 21; GST-S4, spectrin repeat 21; and GST-S5, containing the EF1 and EF2 domains (Fig. 3A), and incubated with lysates of HEK293 cells transfected with FLAG-hTRPC4. As shown in Fig. 3, A (indicated by the + or under hTRPC4) and B, all fusion proteins containing spectrin repeats (S1, S2, S3, and S4) bound FLAG-hTRPC4. Furthermore, regardless of the number or type of spectrin repeats included, there was no significant difference in the level of hTRPC4 binding observed among the truncations (Fig. 3B). GST fusion proteins containing the EF1 and EF2 domains alone, did not bind hTRPC4, however removal of these domains from the spectrin repeats reduced hTRPC4 binding compared with the αII-spectrin fragment isolated in the original two-hybrid screen (Fig. 3B). These results indicate that the EF domains strengthen or modulate the interaction of αII-spectrin with hTRPC4 rather than directly participate in the binding. Taken together, these results suggest that the spectrin repeats of αII-spectrin are responsible for mediating the interaction with hTRPC4. There appears to be a redundancy in binding, with a single spectrin repeat able to mediate the association with hTRPC4. Moreover, the repeats do not act in an additive manner to strengthen the interaction between hTRPC4 and αII-spectrin.
      Figure thumbnail gr3
      FIGURE 3The interaction of αII-spectrin with hTRPC4 is dependent on the final three spectrin repeats of αII-spectrin and independent of the Ca2+-binding EF domains. A, schematic of GST-αII-spectrin fusion proteins utilized in binding assays in vitro. Indicated is the relative strength of the hTRPC4 association. B, various GST-αII-spectrin mutants were immobilized and incubated with FLAG-hTRPC4-transfected HEK293 lysates. Bound proteins were analyzed by SDS-PAGE and Western blotting with anti-FLAG M2 antibody along with 6% of the pulldown input. Coomassie staining is shown for demonstration of equal fusion protein loading. Right, FLAG immunoreactivity was quantified from four independent assays, collated, and expressed as a percentage of hTRPC4 binding to the complete C terminus of αII-spectrin isolated from the yeast two-hybrid screen.
      The Spectrin Binding Site Is Absent in the δ Splice Variant of hTRPC4–To define the region of hTRPC4 responsible for αII-spectrin binding, DNA constructs encoding fragments of the C terminus of hTRPC4, CT1-hTRPC4 (aa 759-977), CT2-hTRPC4 (aa 846-977), CT3-hTRPC4 (aa 759-859), CT4-hTRPC4 (686-859), and CT5-hTRPC4 (aa 686-784), were generated by PCR and cloned in-frame with the LexA DNA binding domain of pBTM116 (Fig. 4A). Yeast two-hybrid assays were then performed between the hTRPC4 C-terminal truncations and the C-terminal αII-spectrin (Fig. 4A) and βV-spectrin (data not shown) fragments isolated from the library screen. Under these conditions only the CT4-hTRPC4 (aa 686-859) and CT5-hTRPC4 (aa 686-784) fragments interacted with αII-spectrin and βV-spectrin (Fig. 4A), as determined by reporter gene activation, indicating that the interacting domain in hTRPC4 potentially lies between amino acids 686 and 784 (Fig. 4A). Amino acids 686-784 consist of a coiled-coil domain ∼95 amino acids upstream of the EWKFAR TRP box and independent of the recently described protein 4.1 binding site and the established Ca2+-calmodulin/IP3R-binding domain (CIRB domain), the latter of which is absent in the βhTRPC4 splice variant (
      • Schaefer M.
      • Plant T.D.
      • Stresow N.
      • Albrecht N.
      • Schultz G.
      ,
      • Mery L.
      • Magnino F.
      • Schmidt K.
      • Krause K.H.
      • Dufour J.F.
      ).
      Figure thumbnail gr4
      FIGURE 4The δ splice variant of hTRPC4 does not associate with αII-spectrin. A, schematic of pBTM116 fusion proteins and a summary of lacZ gene activation in the presence of pACT2-αII-spectrin. Shown darker is the region of hTRPC4 C terminus responsible for mediating αII-spectrin and βV-spectrin binding. Also indicated to the right is a colony-lift filter assay of lacZ expression from L40 yeast co-transformed with the opposing pBTM116 construct and pACT2-αII-spectrin. B, GST fusion proteins of the C termini of αII-spectrin and βV-spectrin were incubated with lysates from either FLAG-tagged αhTRPC4-, βhTRPC4-, or δhTRPC4-expressing HEK293 cells. Bound proteins were analyzed by Western blotting with anti-FLAG antibody. C, COS-7 cells grown on coverslips were transiently transfected with either FLAG-αhTRPC4 or FLAG-δhTPRC4, fixed, and stained with anti-TRPC4 C20 antibody following by anti-goat AlexaFluor 488 and 4′,6-diamidino-2-phenylindole. Cells were visualized by Zeiss LSM510 confocal microscope under 63× objective.
      The novel spectrin-binding site coincides with a recently described additional calmodulin-binding domain, located 95 amino acids downstream of the CIRB domain responsible for facilitating the actions of Ca2+/calmodulin (
      • Ordaz B.
      • Tang J.
      • Xiao R.
      • Salgado A.
      • Sampieri A.
      • Zhu M.X.
      • Vaca L.
      ). This region was absent in the postulated δhTRPC4 splice variant, which lacks 141 amino acids (aa 730-870) in the C-terminal tail. Because the potential αII-spectrin binding site of hTRPC4 partially overlaps with splice sites for the δhTRPC4 variant, it was of interest to determine whether δhTRPC4 could bind αII-spectrin. FLAG-tagged δhTRPC4 was generated by linker-scanning PCR, and the sequence was verified by DNA sequencing. The βhTRPC4 variant (lacking aa 785-868) was similarly generated for use as a control. A schematic of the regions deleted from each variant is shown in Fig. 4B. Full-length hTRPC4 (α variant) and the β and δ variants were transiently transfected into HEK293 cells, and lysates were incubated with either immobilized GST, GST-αII-spectrin (C-terminal domain), or GST-βV-spectrin (C-terminal domain) fusion proteins. Bound proteins were analyzed by immunoblotting for the presence of FLAG-hTRPC4 variants. As shown in Fig. 4B, αhTRPC4 and βhTRPC4 bound to both GST-αII-spectrin and βV-spectrin, however the δhTRPC4 variant failed to bind to either GST-αII-spectrin or GST-βV-spectrin. These data, taken together with the results obtained from the yeast two-hybrid and GST-pulldown studies with truncated hTRPC4, further defined the spectrin binding site in hTRPC4 to amino acids Gly-730 through Glu-784. Despite lacking the spectrin-binding domain, the cellular distribution of the αhTRPC4 and δhTRPC4 variants were not noticeably different when expressed in COS-7 cells and examined by confocal immunofluorescence microscopy (Fig. 4C), suggesting that this domain does not dramatically affect the cellular targeting of the channel.
      Deletion mutations of the spectrin-binding site of hTRPC4 were generated by linker-scanning PCR. FLAG-tagged, M1-hTRPC4 (ΔG730-E784), M2-hTRPC4 (ΔG730-K758), and M3-hTRPC4 (ΔK758-E784) (see Fig. 4B, schematic) were transiently expressed in HEK293 cells, and solubilized lysates were incubated with immobilized GST-αII-spectrin (Fig. 5A). Immunoblot analysis of bound proteins indicated that both M1-hTRPC4 and M2-hTRPC4 were unable to interact with GST-αII-spectrin. However, the M3-hTRPC4 construct, lacking Lys-758 through Glu-784, bound αII-spectrin with similar affinity to wild-type αhTRPC4 (Fig. 5A), thereby demonstrating that the αII-spectrin binding motif of hTRPC4 is located between amino acids Gly-730 and Lys-758 (Fig. 5A). Gly-730 through Leu-753 is a highly conserved region between hTRPC4 and hTRPC5 with 22 out of 23 residues identical between the two members (Fig. 5B). Other members of the TRPC family display limited homology across the putative spectrin-binding domain, and together this may indicate that a direct interaction with spectrin is unique to the TRPC4 and TRPC5 subfamily of TRPC channels.
      Figure thumbnail gr5
      FIGURE 5The αII-spectrin binding domain of hTRPC4 is located between amino acids 730 and 758. A, deletion mutants of hTRPC4 were generated by linker scanning PCR. M1-hTRPC4, M2-hTRPC4, and M3-hTRPC4 were expressed as FLAG-tagged proteins in HEK293 cells and incubated with immobilized GST or GST-αII-spectrin (C terminus). Bound proteins were analyzed by Western blotting with anti-FLAG and anti-GST antibodies. Also shown are the percentages of mutant hTRPC4 bound to αII-spectrin as quantified from three separate experiments and expressed as a percentage of WT-hTRPC4 binding. B, alignment of the final 300 amino acids at the C-terminal of the 6 expressed human TRPC proteins. Note the putative spectrin-binding region highlighted in yellow. Alignment was performed with the ClustalW program.
      αII-Spectrin Associates with hTRPC4 Signaling Complexes and Modulates EGF-induced hTRPC4 Exocytotic Insertion and Activation–The co-localization of signaling proteins within apparent microdomains in cells may be important for the assembly of hTRPC4 channels into signaling complexes and their subsequent activation by EGF receptor stimulation. The role of αII-spectrin in this event was investigated initially by studying αII-spectrin interactions with TRPC4, NHERF, and Fyn in a variety of cell lines and tissues. Fyn immunoprecipitates from rat brain microsomal fractions were found to contain both TRPC4 and αII-spectrin (Fig. 6A), demonstrating the possible presence of a multiprotein complex within the brain. Further, anti-hemagglutinin (HA) precipitates from FLAG-hTRPC4 expressing COS-7 cells transiently transfected with HA-tagged NHERF contained αII-spectrin and EGFR, in addition to hTRPC4, when examined by immunoblotting (Fig. 6B). These results suggest that αII-spectrin and αhTRPC4 are found together in macromolecular complexes within various cell types.
      Figure thumbnail gr6
      FIGURE 6EGF stimulation disrupts the interaction between hTRPC4 and αII-spectrin. A, rat brain microsomes solubilized in RIPA buffer were immunoprecipitated with anti-FYN or anti-rabbit antibody as indicated and Western blotted for αII-spectrin, TRPC4, and Fyn. B, COS-7 cells expressing FLAG-hTRPC4 were transiently transfected with HA-tagged NHERF as indicated, and lysates were immunoprecipitated with anti-HA antibody. Bound proteins were analyzed by Western blotting as indicated. C, COS-7 cells transiently expressing hTRPC4 were serum-starved, stimulated with EGF (100 ng/ml for 15 min), and biotinylated prior to immunoprecipitation of hTRPC4 from RIPA-solubilized lysates. Bound proteins were immunoblotted for the presence of αII-spectrin and biotinylated hTRPC4 as indicated. Shown is a representative of three individual experiments. D, lysates from biotinylated COS-7 cells stimulated with EGF were precipitated with NeutrAvidin-agarose and analyzed by immunoblotting with anti-αII-spectrin and anti-TRPC4 antibodies.
      We have also demonstrated previously that EGF-induced activation of hTRPC4 is dependent on tyrosine phosphorylation of hTRPC4 by the STK, Fyn, and the subsequent association of hTRPC4 with the scaffolding protein NHERF (
      • Odell A.F.
      • Scott J.L.
      • Van Helden D.F.
      ). The participation of αII-spectrin in the EGF-dependent exocytotic movement of hTRPC4 to the plasma membrane was thus examined using COS-7 cells transfected with FLAG-hTRPC4. The cells were stimulated with EGF and biotinylated, and anti-FLAG-precipitated lysates were immunoblotted for the presence of αII-spectrin and biotin-labeled hTRPC4. EGF stimulation (100 ng/ml and 20 min) was found to induce the dissociation of αII-spectrin from hTRPC4 concomitant with an increase in plasma membrane expression of hTRPC4 (Fig. 6C). In addition, when biotinylated proteins from hTRPC4-transfected COS-7 cells stimulated with EGF were isolated with NeutrAvidin-agarose and analyzed for the presence of αII-spectrin, a decrease in the association of αII-spectrin with integral membrane proteins was observed (Fig. 6D).
      The importance of the αII-spectrin-hTRPC4 interaction in hTRPC4 exocytic insertion and channel activation following EGF stimulation was confirmed using COS-7 cells transfected with FLAG-δhTRPC4, which lacks the putative αII-spectrin binding site. Following EGF application, δhTRPC4 failed to undergo further membrane translocation despite an increase in tyrosine phosphorylation (Fig. 7A). This was in contrast to the αhTRPC4 variant, which exhibited a robust membrane insertion and tyrosine phosphorylation response (Fig. 7, A and B). Furthermore, δhTRPC4 channels exhibited elevated and more variable basal plasma membrane expression than αhTRPC4 channels (Fig. 7B) suggesting spectrin binding may influence plasma membrane levels in non-stimulated cells. Importantly, co-expression of δhTRPC4 in HEK293 cells with the EGF receptor failed to produce a further increase in intracellular Ca2+ upon EGF application, in contrast to the αhTRPC4 variant, which showed a larger, sustained increase in intracellular Ca2+ (Fig. 7C). Expression of GFP alone, with the EGF receptor, produced a profile similar to the δhTRPC4 variant (Fig. 7C). Additionally, the stimulation of HEK293 cells expressing both the EGF receptor and various hTRPC4 constructs or GFP, with EGF in the absence of extracellular Ca2+ for 8 min prior to Ca2+ re-addition, produced a larger sustained Ca2+ influx in αhTRPC4-expressing cells (Fig. 7D). The smaller increase in intracellular Ca2+ was similar in cells expressing either δhTRPC4 or GFP (Fig. 7D), suggesting that the δ variant is unable to bind spectrin and thus fails to undergo significant additional membrane insertion or activation in response to EGF. Taken together, these results indicate that spectrin tethers hTRPC4-signaling complexes in intracellular microdomains and influences both the level and activation of hTRPC4 channels following EGF stimulation.
      Figure thumbnail gr7
      FIGURE 7Formation of a functional signaling complex between αII-spectrin and hTRPC4 involved in channel activation. A, the δhTRPC4 variant does not undergo EGF-dependent translocation. Serum-starved COS-7 cells transiently expressing either αhTRPC4 or δhTRPC4 were stimulated with EGF (100 ng/ml for 15 min) prior to biotinylation. Lysates were immunoprecipitated with anti-FLAG antibody and analyzed by Western blotting as indicated. B, shown are the collated data of Fig. 7A from five independent experiments. C, HEK293 cells expressing either GFP, αhTRPC4, or δhTRPC4 were loaded with Fura-2 and stimulated with EGF in the presence of 2 mm Ca2+.Ca2+ transients were measured as the ratio of 340/380 nm wavelength using an ORCA-ER camera. Shown are averaged traces from 40-50 cells. D, Fura-2-loaded HEK293 cells expressing either GFP, αhTRPC4, or δhTRPC4 were stimulated with EGF in the absence of external Ca2+ for 8 min prior to the re-addition of 2 mm Ca2+. Shown are the averaged Ca2+ re-entry profiles from 50-60 individual cells.
      To further examine the role of αII-spectrin in regulating the surface expression of αhTRPC4, siRNA was used to deplete endogenous αII-spectrin from COS-7 cells expressing either αhTRPC4 or δhTRPC4. RNA interference knockdown produced an 86% reduction in αII-spectrin expression (n = 7), without any effect on the expression of EGFR, β1-integrin, transferrin receptor, α-tubulin or either transfected αhTRPC4 or δhTRPC4 as assessed by Western blotting (Fig. 8A). However, the depletion of αII-spectrin resulted in a 51% (n = 12) decrease in the basal surface expression of αhTRPC4 as determined by biotinylation of membrane proteins followed by NeutrAvidin pulldown assays (Fig. 8B). In addition, αhTRPC4 channels when expressed in siRNA-treated cells also failed to undergo EGF-induced membrane insertion (Fig. 8B). In contrast, both the basal surface expression of δhTPRC4 and its response to EGF application were unaffected by αII-spectrin knockdown, in agreement with the previous results. Furthermore, although the treatment of COS-7 cells with αII-spectrin siRNA did not significantly alter the surface expression of transferrin receptor, there was variability in receptor surface levels, showing both increased and unchanged levels in various experiments, upon αII-spectrin depletion (n = 18, S.E. ± 10.6). This variability may reflect the importance of the spectrin cytoskeleton in controlling the general organization of the plasma membrane (

      Abdi, K. M., and Bennett, V. (2007) Mol. Biol. Cell, in press

      ,
      • Kizhatil K.
      • Yoon W.
      • Mohler P.J.
      • Davis L.H.
      • Hoffman J.A.
      • Bennett V.
      ,
      • Voas M.G.
      • Lyons D.A.
      • Naylor S.G.
      • Arana N.
      • Rasband M.N.
      • Talbot W.S.
      ). Importantly, αII-spectrin depletion did not alter the autophosphorylation of the EGFR (pEGFR) or the downstream activation of ERK1/2 as determined by immunoblotting for the dually phosphorylated form (pERK1/2) (Fig. 8C). Taken together, these findings suggest that αII-spectrin not only regulates the EGF-dependent insertion of αhTRPC4 channels but is also involved in regulating the basal plasma membrane levels of the channel.
      Figure thumbnail gr8
      FIGURE 8Depletion αII-spectrin by siRNA reduces the surface expression of αhTRPC4 and prevents EGF-stimulated membrane insertion. A, COS-7 cells reverse transfected with either scrambled or αII-spectrin siRNA were transiently transfected with FLAG-αhTPRC4 or FLAG-δhTRPC4 and assessed for the effect on the expression of the indicated proteins by immunoblotting. Right: protein expression for αII-spectrin, αhTRPC4, and δhTRPC4 were quantified and expressed as a percentage of scrambled siRNA levels (n > 4 independent experiments). B, siRNA-treated COS-7 cells transiently expressing either αhTRPC4 or δhTPRC4 were serum-starved, biotinylated, and lysed prior to incubation of cleared lysated with NeutrAvidin-agarose. Bound proteins were separated by SDS-PAGE and analyzed by immunoblotting for TRPC4 and transferrin receptor (TfR) levels. Below: membrane levels were quantified and expressed as a percentage of unstimulated, scrambled siRNA-treated levels (n > 6 independent experiments). C, cell lysates used in Fig. 8B were analyzed by Western blotting for levels of the indicated proteins. Shown are representatives of at least six independent experiments. ***, p < 0.001.

      DISCUSSION

      Stimulation of the EGFR was previously shown to induce Ca2+ entry through the activation of hTRPC4 channels (
      • Yang H.
      • Mergler S.
      • Sun X.
      • Wang Z.
      • Lu L.
      • Bonanno J.A.
      • Pleyer U.
      • Reinach P.S.
      ,
      • Odell A.F.
      • Scott J.L.
      • Van Helden D.F.
      ). A role for NHERF and STKs in hTRPC4 regulation was established in this regulation, however other proteins may also be involved in controlling the expression and activation of hTRPC4 channels at the cell surface (
      • Odell A.F.
      • Scott J.L.
      • Van Helden D.F.
      ). In particular, various cytoskeletal proteins, including actin (
      • Morales S.
      • Camello P.J.
      • Rosado J.A.
      • Mawe G.M.
      • Pozo M.J.
      ,
      • Itagaki K.
      • Kannan K.B.
      • Singh B.B.
      • Hauser C.J.
      ) and β-tubulin (
      • Bollimuntha S.
      • Cornatzer E.
      • Singh B.B.
      ), have recently been implicated as important components of TRPC signaling pathways. Utilizing the yeast two-hybrid system, two members of the spectrin family of cytoskeletal proteins, αII-spectrin and βV-spectrin, were identified as novel hTRPC4-binding proteins. The spectrin-binding site of hTRPC4 was localized to between residues 730 and 758, comprising a coiled-coil domain structure previously associated with protein dimerization. This region was responsible for associating with the final three spectrin repeats of αII- and βV-spectrin. The direct interaction between hTRPC4 and αII-spectrin was important in controlling the translocation of hTRPC4 channels to the plasma membrane following EGF stimulation, resulting in activation of the channel.
      Previous studies have implicated the sub-membranous spectrin cytoskeleton in the control of store-operated Ca2+ entry, with disruption of the spectrin-protein 4.1G interaction in endothelial cells reducing the thapsigargin-activated Ca2+-selective ISOC current (
      • Wu S.
      • Sangerman J.
      • Li M.
      • Brough G.H.
      • Goodman S.R.
      • Stevens T.
      ,
      • Cioffi D.L.
      • Wu S.
      • Stevens T.
      ). TRPC4 was suggested as a component of the channels mediating ISOC, and a direct association between TRPC4 and protein 4.1G was proposed to link this channel with the spectrin and actin cytoskeleton and thereby regulate Ca2+ entry (
      • Cioffi D.L.
      • Wu S.
      • Alexeyev M.
      • Goodman S.R.
      • Zhu M.X.
      • Stevens T.
      ). The putative protein 4.1G-binding domain was predicted to reside between amino acids 675 and 685 of TRPC4, ∼40 amino acids downstream of the invariant EWKFAR region. Importantly, the spectrin-interacting domain of hTRPC4 was found within amino acids 730 and 758, which lies ∼50 residues C-terminal of the proposed protein 4.1-binding region, which is similar to the protein 4.1 binding sequences found in membrane-associated guanylate kinases, such as Discs Large (Dlg) (
      • Cioffi D.L.
      • Wu S.
      • Alexeyev M.
      • Goodman S.R.
      • Zhu M.X.
      • Stevens T.
      ,
      • Dimitratos S.D.
      • Woods D.F.
      • Stathakis D.G.
      • Bryant P.J.
      ). This fosters the proposal that formation of macromolecular signaling scaffolds is required to communicate changes in cellular signaling to the plasma membrane and regulate hTRPC4 channels.
      Similar to the protein 4.1-TRPC4 interaction, the association between hTRPC4 and αII-spectrin appears to be constitutively present in unstimulated cells. However, unlike the protein 4.1-TRPC4 interaction reported previously as required for activation of ISOC (
      • Cioffi D.L.
      • Wu S.
      • Alexeyev M.
      • Goodman S.R.
      • Zhu M.X.
      • Stevens T.
      ), the association of hTRPC4 with αII-spectrin is disrupted by stimulation of cells with EGF and coincides with Ca2+ entry. The disruption of αII-spectrin-hTRPC4 binding may reflect the Ca2+-dependent remodeling of the spectrin cytoskeleton, in a manner similar to the redistribution reported to occur following stimulation of exocytosis in chromaffin cells (
      • Perrin D.
      • Aunis D.
      ). This discrepancy between the association of αII-spectrin-hTRPC4 and protein 4.1-hTRPC4 may indicate that the interaction with protein 4.1, and consequently spectrin and the actin cytoskeleton, coordinates the formation of stable signaling complexes surrounding TRPC4, with this complex likely to be additionally stabilized by a direct association between the channel and αII-spectrin. Cellular stimulation with Ca2+-mobilizing ligands may then induce the dissociation of the αII-spectrin interaction, allowing TRPC4 channels to undergo translocation to the plasma membrane. Here, they may be additionally stabilized by the remaining association with protein 4.1. This coordinated response would allow TRPC4 channel activity to be tightly regulated within the cell and thereby minimize deleterious fluctuations in intracellular Ca2+.
      The localization of the spectrin-binding region within a coiled-coil domain of hTRPC4 appears important for mediating the interaction with the spectrin repeats of αII- and βV-spectrin. Coiled-coil domains are highly versatile and frequently observed in protein interaction modules often associated with oligomerization (
      • Burkhard P.
      • Stetefeld J.
      • Strelkov S.V.
      ). They are found within cytoskeletal proteins, ion channels, and motor proteins, and they mediate the assembly of the soluble NSF attachment protein receptor complexes to regulate vesicle fusion (
      • Lupas A.N.
      • Gruber M.
      ). Importantly, spectrin repeats are also composed of coiled-coil domains allowing spectrins to be classified as membrane-bound coiledcoil proteins (
      • Thomas G.
      ). The formation of an interface between the coiled-coil domains of hTRPC4 and the spectrin repeats of αII-spectrin may therefore be mediating their interaction. Whether this interaction is modulated by protein modifications such as phosphorylation and protease cleavage remains to be determined. Indeed, the cleavage of αII-spectrin by calpain is inhibited by Src-dependent phosphorylation of the SH3 domain in αII-spectrin at Tyr-1146 (
      • Nicolas G.
      • Fournier C.M.
      • Galand C.
      • Malbert-Colas L.
      • Bournier O.
      • Kroviarski Y.
      • Bourgeois M.
      • Camonis J.H.
      • Dhermy D.
      • Grandchamp B.
      • Lecomte M.C.
      ). This may stabilize αII-spectrin and promote its association with other proline-rich proteins to stimulate dissociation from hTRPC4. Alternatively, a modification on the C terminus of hTRPC4 may disrupt αII-spectrin binding and release the channel from the cytoskeleton.
      Intriguingly, a novel hTRPC4 splice variant identified in the NCBI data base, the δ variant, was found to lack the αII-spectrin binding domain and as such an interaction between hTRPC4 with αII-spectrin was absent. Importantly, the absence of this interaction correlated with a failure of these channels to translocate to the plasma membrane in response to EGF, despite undergoing a robust tyrosine phosphorylation response. This may reflect a requirement for the αII-spectrin association to couple the channel to the translocation machinery. Alternatively, αII-spectrin may sequester nascent inactive hTRPC4 channels beneath the cell surface, to await an activating stimulus, which the δ isoform is unable to respond to. The absence of a consistent membrane translocation event in response to EGF resulting in a deficiency in the Ca2+ entry profile of the δ variant, supports both possibilities. Thus, the cellular expression of the δhTRPC4 variant may represent a novel mechanism for modulating the Ca2+ entry profile of cells containing hTRPC4.
      It is important to note that the gross cellular localization of both splice variants is not inherently different, with both channels forming punctuate structures consistent with their localization to lipid raft domains.
      A. F. Odell, D. F. Van Helden, and J. L. Scott, unpublished observations.
      Thus their different Ca2+-signaling profiles must be due to an alternate structural feature of the channel, such as the loss of a protein interaction site, which affects their ability to respond to stimuli. Consistent with this hypothesis, depletion of endogenous αII-spectrin from COS-7 cells resulted in a decrease in both membrane insertion and resting plasma membrane levels of αhTRPC4 while exerting no effect on the δ variant of the channel. However, the absence of αII-spectrin also altered the surface expression of other membrane proteins such as β1-integrin
      A. F. Odell, D. F. Van Helden, and J. L. Scott, unpublished observations.
      and transferrin receptor to a variable and non-significant extent, suggesting that spectrins play a role in modulating the plasma membrane structure in addition to participating in direct protein-protein interactions. The effect on αhTRPC4 surface expression is similar to that reported in nodes of Ranvier, where nodal Na+ channel clusters were reduced when αII-spectrin was disrupted (
      • Voas M.G.
      • Lyons D.A.
      • Naylor S.G.
      • Arana N.
      • Rasband M.N.
      • Talbot W.S.
      ). It was suggested that αII-spectrin is involved in stabilizing the nascent channel clusters at nodal sites. We propose that αII-spectrin may perform a similar function here, where αhTRPC4 channel clusters are both arranged and stabilized at, and adjacent to, the plasma membrane sites of Ca2+ entry. The presence or absence of the spectrin-binding domain may thus control the activity of the channel by allowing cytoskeletal control of channel complex surface expression and stabilization.
      The ability of αII-spectrin to act as a scaffold protein linking hTRPC4 channels to intricate signaling complexes is enhanced by the large number of potential binding partners of spectrin. Through interactions with other cytoskeletal components, including F-actin and ankyrin, as well as signaling molecules such as synapsin I, membrane-associated guanylate kinases, and phospholipids, spectrins are ideally positioned to coordinate cellular responses culminating in Ca2+ entry (
      • Bennett V.
      • Baines A.J.
      ). Spectrins are also known to form complexes with various Ca2+ regulators, including IP3Rs, ryanodine receptors, Na+/K+-ATPase, various ion channels, N-methyl-d-aspartic acid receptor subunits, and the epithelial Na+ channel, ENaC (
      • Wechsler A.
      • Teichberg V.I.
      ,
      • Lencesova L.
      • O'Neill A.
      • Resneck W.G.
      • Bloch R.J.
      • Blaustein M.P.
      ,
      • Nakano M.
      • Nogami S.
      • Sato S.
      • Terano A.
      • Shirataki H.
      ,
      • Kontrogianni-Konstantopoulos A.
      • Frye C.S.
      • Benz Jr., E.J.
      • Huang S.C.
      ,
      • Zuckerman J.B.
      • Chen X.
      • Jacobs J.D.
      • Hu B.
      • Kleyman T.R.
      • Smith P.R.
      ,
      • Devarajan P.
      • Stabach P.R.
      • De Matteis M.A.
      • Morrow J.S.
      ), positioning them as potential critical regulators of Ca2+ levels. Interactions between hTRPC4 and the spectrin cytoskeleton may also provide a functional bridge between the endoplasmic reticulum and plasma membrane, forming a mechanism for linking Ca2+ release from intracellular stores with Ca2+ entry through the plasma membrane. Spectrins have also been identified in TRPC5 and TRPC6 immunoprecipitates from rat brain (
      • Goel M.
      • Sinkins W.
      • Keightley A.
      • Kinter M.
      • Schilling W.P.
      ), and we have now shown that αII-spectrin is a component of a multiprotein signaling complex surrounding hTRPC4, along with NHERF, Fyn, and the EGFR. The latter three proteins are integral in the phosphorylation and activation of hTRPC4 by EGF and the addition of αII-spectrin provides a stabilizing platform for the assembly of the complex at the plasma membrane. Together with the recently identified interactions with Homer and stromal interacting protein 1, this macromolecular channel complex may be critical in the activation of TRPC4 by various stimuli (
      • Yuan J.P.
      • Zeng W.
      • Huang G.N.
      • Worley P.F.
      • Muallem S.
      ,
      • Worley P.F.
      • Zeng W.
      • Huang G.N.
      • Yuan J.P.
      • Kim J.Y.
      • Lee M.G.
      • Muallem S.
      ).

      Acknowledgments

      We acknowledge Drs. J. W. Putney and R. R. McKay for hTRPC4 cDNA and Dr. J. S. Morrow for kindly providing αII-spectrin cDNA.

      Supplementary Material

      References

        • Freichel M.
        • Suh S.H.
        • Pfeifer A.
        • Schweig U.
        • Trost C.
        • Weissgerber P.
        • Biel M.
        • Philipp S.
        • Freise D.
        • Droogmans G.
        • Hofmann F.
        • Flockerzi V.
        • Nilius B.
        Nat. Cell Biol. 2001; 3: 121-127
        • Tiruppathi C.
        • Freichel M.
        • Vogel S.M.
        • Paria B.C.
        • Mehta D.
        • Flockerzi V.
        • Malik A.B.
        Circ. Res. 2002; 91: 70-76
        • Wu X.
        • Babnigg G.
        • Zagranichnaya T.
        • Villereal M.L.
        J. Biol. Chem. 2002; 277: 13597-13608
        • Yang H.
        • Mergler S.
        • Sun X.
        • Wang Z.
        • Lu L.
        • Bonanno J.A.
        • Pleyer U.
        • Reinach P.S.
        J. Biol. Chem. 2005; 280: 32230-32237
        • Ambudkar I.S.
        Trends Pharmacol. Sci. 2006; 27: 25-32
        • Kiselyov K.
        • Shin D.M.
        • Kim J.Y.
        • Yuan J.P.
        • Muallem S.
        Handb. Exp. Pharmacol. 2007; 179: 559-574
        • Sinkins W.G.
        • Goel M.
        • Estacion M.
        • Schilling W.P.
        J. Biol. Chem. 2004; 279: 34521-34529
        • Yuan J.P.
        • Kiselyov K.
        • Shin D.M.
        • Chen J.
        • Shcheynikov N.
        • Kang S.H.
        • Dehoff M.H.
        • Schwarz M.K.
        • Seeburg P.H.
        • Muallem S.
        • Worley P.F.
        Cell. 2003; 114: 777-789
        • Kiselyov K.
        • Kim J.Y.
        • Zeng W.
        • Muallem S.
        Pflugers Arch. 2005; 451: 116-124
        • Cioffi D.L.
        • Wu S.
        • Alexeyev M.
        • Goodman S.R.
        • Zhu M.X.
        • Stevens T.
        Circ. Res. 2005; 97: 1164-1172
        • Yuan J.P.
        • Zeng W.
        • Huang G.N.
        • Worley P.F.
        • Muallem S.
        Nat. Cell Biol. 2007; 9: 636-645
        • Zhu M.X.
        • Tang J.
        Novartis Found Symp. 2004; 258 (discussion 58-62, 98-102, 263-106): 44-58
        • Kiselyov K.
        • Mignery G.A.
        • Zhu M.X.
        • Muallem S.
        Mol. Cell. 1999; 4: 423-429
        • Schaefer M.
        • Plant T.D.
        • Stresow N.
        • Albrecht N.
        • Schultz G.
        J. Biol. Chem. 2002; 277: 3752-3759
        • Trost C.
        • Bergs C.
        • Himmerkus N.
        • Flockerzi V.
        Biochem. J. 2001; 355: 663-670
        • Tang J.
        • Lin Y.
        • Zhang Z.
        • Tikunova S.
        • Birnbaumer L.
        • Zhu M.X.
        J. Biol. Chem. 2001; 276: 21303-21310
        • Tang Y.
        • Tang J.
        • Chen Z.
        • Trost C.
        • Flockerzi V.
        • Li M.
        • Ramesh V.
        • Zhu M.X.
        J. Biol. Chem. 2000; 275: 37559-37564
        • Obukhov A.G.
        • Nowycky M.C.
        J. Cell. Physiol. 2004; 201: 227-235
        • Mery L.
        • Strauss B.
        • Dufour J.F.
        • Krause K.H.
        • Hoth M.
        J. Cell Sci. 2002; 115: 3497-3508
        • Wu S.
        • Sangerman J.
        • Li M.
        • Brough G.H.
        • Goodman S.R.
        • Stevens T.
        J. Cell Biol. 2001; 154: 1225-1233
        • Bennett V.
        • Baines A.J.
        Physiol. Rev. 2001; 81: 1353-1392
        • Hryniewicz-Jankowska A.
        • Czogalla A.
        • Bok E.
        • Sikorsk A.F.
        Folia Histochem. Cytobiol. 2002; 40: 239-249
        • Wechsler A.
        • Teichberg V.I.
        EMBO J. 1998; 17: 3931-3939
        • Bloch R.J.
        • Bezakova G.
        • Ursitti J.A.
        • Zhou D.
        • Pumplin D.W.
        Soc. Gen. Physiol. Ser. 1997; 52: 177-195
        • Bennett V.
        Curr. Opin. Cell Biol. 1990; 2: 51-56
        • Cioffi D.L.
        • Wu S.
        • Stevens T.
        Cell Calcium. 2003; 33: 323-336
        • Odell A.F.
        • Scott J.L.
        • Van Helden D.F.
        J. Biol. Chem. 2005; 280: 37974-37987
        • McKay R.R.
        • Szymeczek-Seay C.L.
        • Lievremont J.P.
        • Bird G.S.
        • Zitt C.
        • Jungling E.
        • Luckhoff A.
        • Putney Jr., J.W.
        Biochem. J. 2000; 351: 735-746
        • Hollenberg S.M.
        • Sternglanz R.
        • Cheng P.F.
        • Weintraub H.
        Mol. Cell. Biol. 1995; 15: 3813-3822
        • Sikorski A.F.
        • Terlecki G.
        • Zagon I.S.
        • Goodman S.R.
        J. Cell Biol. 1991; 114: 313-318
        • Iga M.
        • Inui M.
        • Sobue K.
        Biochem. Biophys. Res. Commun. 1997; 231: 852-855
        • Cukovic D.
        • Lu G.W.
        • Wible B.
        • Steele D.F.
        • Fedida D.
        FEBS Lett. 2001; 498: 87-92
        • Mery L.
        • Magnino F.
        • Schmidt K.
        • Krause K.H.
        • Dufour J.F.
        FEBS Lett. 2001; 487: 377-383
        • Ordaz B.
        • Tang J.
        • Xiao R.
        • Salgado A.
        • Sampieri A.
        • Zhu M.X.
        • Vaca L.
        J. Biol. Chem. 2005; 280: 30788-30796
      1. Abdi, K. M., and Bennett, V. (2007) Mol. Biol. Cell, in press

        • Kizhatil K.
        • Yoon W.
        • Mohler P.J.
        • Davis L.H.
        • Hoffman J.A.
        • Bennett V.
        J. Biol. Chem. 2007; 282: 2029-2037
        • Voas M.G.
        • Lyons D.A.
        • Naylor S.G.
        • Arana N.
        • Rasband M.N.
        • Talbot W.S.
        Curr. Biol. 2007; 17: 562-568
        • Morales S.
        • Camello P.J.
        • Rosado J.A.
        • Mawe G.M.
        • Pozo M.J.
        Cell Signal. 2005; 17: 635-645
        • Itagaki K.
        • Kannan K.B.
        • Singh B.B.
        • Hauser C.J.
        J. Immunol. 2004; 172: 601-607
        • Bollimuntha S.
        • Cornatzer E.
        • Singh B.B.
        Vis. Neurosci. 2005; 22: 163-170
        • Dimitratos S.D.
        • Woods D.F.
        • Stathakis D.G.
        • Bryant P.J.
        BioEssays. 1999; 21: 912-921
        • Perrin D.
        • Aunis D.
        Nature. 1985; 315: 589-592
        • Burkhard P.
        • Stetefeld J.
        • Strelkov S.V.
        Trends Cell Biol. 2001; 11: 82-88
        • Lupas A.N.
        • Gruber M.
        Adv. Prot. Chem. 2005; 70: 37-78
        • Thomas G.
        BioEssays. 1998; 20: 600
        • Nicolas G.
        • Fournier C.M.
        • Galand C.
        • Malbert-Colas L.
        • Bournier O.
        • Kroviarski Y.
        • Bourgeois M.
        • Camonis J.H.
        • Dhermy D.
        • Grandchamp B.
        • Lecomte M.C.
        Mol. Cell. Biol. 2002; 22: 3527-3536
        • Lencesova L.
        • O'Neill A.
        • Resneck W.G.
        • Bloch R.J.
        • Blaustein M.P.
        J. Biol. Chem. 2004; 279: 2885-2893
        • Nakano M.
        • Nogami S.
        • Sato S.
        • Terano A.
        • Shirataki H.
        Biochem. Biophys. Res. Commun. 2001; 288: 468-475
        • Kontrogianni-Konstantopoulos A.
        • Frye C.S.
        • Benz Jr., E.J.
        • Huang S.C.
        J. Biol. Chem. 2001; 276: 20679-20687
        • Zuckerman J.B.
        • Chen X.
        • Jacobs J.D.
        • Hu B.
        • Kleyman T.R.
        • Smith P.R.
        J. Biol. Chem. 1999; 274: 23286-23295
        • Devarajan P.
        • Stabach P.R.
        • De Matteis M.A.
        • Morrow J.S.
        Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 10711-10716
        • Goel M.
        • Sinkins W.
        • Keightley A.
        • Kinter M.
        • Schilling W.P.
        Pflugers Arch. 2005; 451: 87-98
        • Worley P.F.
        • Zeng W.
        • Huang G.N.
        • Yuan J.P.
        • Kim J.Y.
        • Lee M.G.
        • Muallem S.
        Cell Calcium. 2007; 42: 205-211