Title: The Src Homology 3 Domain of the β-Subunit of Voltage-gated Calcium Channels Promotes Endocytosis via Dynamin Interaction
Abstract: High voltage-gated calcium channels enable calcium entry into cells in response to membrane depolarization. Association of the auxiliary β-subunit to the α-interaction-domain in the pore-forming α1-subunit is required to form functional channels. The β-subunit belongs to the membrane-associated guanylate kinase class of scaffolding proteins containing a Src homology 3 and a guanylate kinase domain. Although the latter is responsible for the high affinity binding to the α-interaction domain, the functional significance of the Src homology 3 domain remains elusive. Here, we show that injection of isolated β-subunit Src homology 3 domain into Xenopus laevis oocytes expressing the α1-subunit reduces the number of channels in the plasma membrane. This effect is reverted by coexpressing α1 with a dominant-negative mutant of dynamin, a GTPase involved in receptor-mediated endocytosis. Full-length β-subunit also down-regulates voltage-gated calcium channels but only when lacking the α-interaction domain. Moreover, isolated Src homology 3 domain and the full-length β-subunit were found to interact in vitro with dynamin and to internalize the distantly related Shaker potassium channel. These results demonstrate that the β-subunit regulates the turnover of voltagegated calcium channels and other proteins in the cell membrane. This effect is mediated by dynamin and depends on the association state of the β-subunit to the α1-pore-forming subunit. Our findings define a novel function for the β-subunit through its Src homology 3 domain and establish a link between voltage-gated calcium channel activity and the cell endocytic machinery. High voltage-gated calcium channels enable calcium entry into cells in response to membrane depolarization. Association of the auxiliary β-subunit to the α-interaction-domain in the pore-forming α1-subunit is required to form functional channels. The β-subunit belongs to the membrane-associated guanylate kinase class of scaffolding proteins containing a Src homology 3 and a guanylate kinase domain. Although the latter is responsible for the high affinity binding to the α-interaction domain, the functional significance of the Src homology 3 domain remains elusive. Here, we show that injection of isolated β-subunit Src homology 3 domain into Xenopus laevis oocytes expressing the α1-subunit reduces the number of channels in the plasma membrane. This effect is reverted by coexpressing α1 with a dominant-negative mutant of dynamin, a GTPase involved in receptor-mediated endocytosis. Full-length β-subunit also down-regulates voltage-gated calcium channels but only when lacking the α-interaction domain. Moreover, isolated Src homology 3 domain and the full-length β-subunit were found to interact in vitro with dynamin and to internalize the distantly related Shaker potassium channel. These results demonstrate that the β-subunit regulates the turnover of voltagegated calcium channels and other proteins in the cell membrane. This effect is mediated by dynamin and depends on the association state of the β-subunit to the α1-pore-forming subunit. Our findings define a novel function for the β-subunit through its Src homology 3 domain and establish a link between voltage-gated calcium channel activity and the cell endocytic machinery. Cellular processes including muscle contraction, endocrine secretion, synaptic transmission, and gene expression (1Arikkath J. Campbell K.P. Curr. Opin. Neurobiol. 2003; 13: 298-307Crossref PubMed Scopus (421) Google Scholar), depend on the regulated influx of calcium through voltagegated calcium channels (VGCCs). 3The abbreviations used are: VGCC, voltage-gated calcium channel; AID, α-interaction domain; CaVβ, β-subunit of VGCCs; CaVα1, pore-forming α-subunit of VGCCs; CaV1.2, cardiac isoform of the α-subunit of VGCCs; SH3, Src-homology 3 domain; GK, guanylate kinase domain; β2a-SH3, Srchomology 3 domain from β2a isoform; GST, glutathione S-transferase; GST-Dyn829–842, dynamin peptide encompassing residues 829–842; Qon, charge movement; PRD, proline-rich domain; WT, wild type; HA, hemagglutinin; pC, picocoulomb; cps, counts/second. 3The abbreviations used are: VGCC, voltage-gated calcium channel; AID, α-interaction domain; CaVβ, β-subunit of VGCCs; CaVα1, pore-forming α-subunit of VGCCs; CaV1.2, cardiac isoform of the α-subunit of VGCCs; SH3, Src-homology 3 domain; GK, guanylate kinase domain; β2a-SH3, Srchomology 3 domain from β2a isoform; GST, glutathione S-transferase; GST-Dyn829–842, dynamin peptide encompassing residues 829–842; Qon, charge movement; PRD, proline-rich domain; WT, wild type; HA, hemagglutinin; pC, picocoulomb; cps, counts/second. VGCCs are multiprotein complexes containing a pore-forming subunit (CaVα1) and a variable number of auxiliary subunits. Association of the auxiliary β-subunit (CaVβ) to a site shared by all CaVα1, the so-called α-interaction domain (AID), is mandatory to form a fully functional VGCC. Homology modeling (2Hanlon M.R. Berrow N.S. Dolphin A.C. Wallace B.A. FEBS Lett. 1999; 445: 366-370Crossref PubMed Scopus (109) Google Scholar) and the recent high resolution crystal structures of three CaVβ isoforms (3Opatowsky Y. Chen C.C. Campbell K.P. Hirsch J.A. Neuron. 2004; 42: 387-399Abstract Full Text Full Text PDF PubMed Scopus (224) Google Scholar, 4Chen Y.H. Li M.H. Zhang Y. He L.L. Yamada Y. Fitzmaurice A. Shen Y. Zhang H. Tong L. Yang J. Nature. 2004; 429: 675-680Crossref PubMed Scopus (259) Google Scholar, 5Van Petegem F. Clark K.A. Chatelain F.C. Minor Jr., D.L. Nature. 2004; 429: 671-675Crossref PubMed Scopus (350) Google Scholar) identified CaVβ as a novel member of the membrane-associated guanylate kinase class of scaffolding proteins containing a Src homology 3 (SH3) and a guanylate kinase (GK) domain (Fig. 1A). As shown by the crystal structure of CaVβ complexed to AID, the CaVβ-GK binds to the AID, whereas CaVβ-SH3 interacts with GK. Although SH3 domains are known to mediate protein-protein interactions by binding to proline-rich motifs in ligand proteins (6Sparks A.B. Rider J.E. Hoffman N.G. Fowlkes D.M. Quillam L.A. Kay B.K. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 1540-1544Crossref PubMed Scopus (331) Google Scholar), no interactions mediated by the CaVβ-SH3 have been described yet. Moreover, the functional integrity of CaVβ-SH3 domain is uncertain since the residues homologous to the ones critical for binding PXXP motifs in canonical SH3 modules are occluded in the crystal structure of CaVβ. Intriguingly, canine and human cardiac cells express splicing variants encoding short versions of the CaVβ that only encompass the variable N-terminal region and the SH3 domain (7Foell J.D. Balijepalli R.C. Delisle B.P. Yunker A.M. Robia S.L. Walker J.W. McEnery M.W. January C.T. Kamp T.J. Physiol. Genomics. 2004; 17: 183-200Crossref PubMed Scopus (91) Google Scholar, 8Cohen R.M. Foell J.D. Balijepalli R.C. Shah V. Hell J.W. Kamp T.J. Am. J. Physiol. 2005; 288: H2363-H2374Crossref PubMed Scopus (34) Google Scholar) (Fig. 1A, V1 and C1, respectively). Here, we studied the effect of isolated CaVβ-SH3 on calcium channel function and expression. The SH3 domain of the rat β2a isoform of CaVβ (β2a-SH3) was expressed in bacteria, purified, and injected into Xenopus oocytes expressing the cardiac CaVα1 (CaV1.2) subunit isoform. We found that the β2a-SH3 induces removal of channels from the plasma membrane in a dynamin-dependent fashion. This function is preserved by full-length CaVβ in the absence of CaVα1 subunit or when binding to it is disrupted by deleting the AID site. Our results define a novel interaction and outline a new function for the calcium channel β-subunit. Recombinant Proteins—The cDNA encoding the SH3 domain of the rat β2a isoform (Swiss-Prot entry:Q8VGC3) encompassing residues 24–136 was subcloned between BamHI and EcoRI restriction sites by conventional PCR methods into pRSETB vector (Invitrogen) to introduce a polyhistidine tag at the N-terminal. The molecular mass predicted by the amino acid sequence of the CaVβ2a-SH3 His-tagged protein is 16.7 kDa. The His-tagged CaVβ2a-SH3 was expressed in BL-21 (DE-3) E. coli bacteria by a 2-h induction with 0.5 mm isopropyl-β-d-thiogalactopyranoside at 37 °C. Cells were harvested by centrifugation, flash-frozen, and stored until use at –80 °C. Right before protein purification, the cells were resuspended in phosphate buffer (50 mm sodium phosphate buffer and 300 mm NaCl, pH 7.0) containing EDTA-free protease inhibitor mixture (Roche Applied Science) and disrupted by ultrasonication. After centrifugation, the protein was purified from the cleared cell lysate by using a cobalt-based metal affinity chromatography (Talon, BD Biosciences) according to the manufacturer's instructions followed by size-exclusion chromatography onto a Superdex™ S-200 column 26/60 (GE Healthcare Life Sciences) pre-equilibrated with buffer containing 20 mm Tris buffer, 300 mm NaCl, 1 mm EDTA, pH 8.0. The fractions containing the protein were pooled, concentrated up to 1–2 mg/ml by centrifugation using Amicon Ultra tubes with 10,000 molecular weight cut-off (Millipore), aliquoted, flash-frozen, and stored at –80 °C until use. The full-length CaVβ2a was prepared as described (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). The apparent molecular mass of His-tagged CaVβ2a-SH3 determined from the size exclusion chromatography calibration curve was obtained from the partition coefficient value (Kav) calculated from its elution volume as described (10Neely A. Garcia-Olivares J. Voswinkel S. Horstkott H. Hidalgo P. J. Biol. Chem. 2004; 279: 21689-21694Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar), where Kav is equal to (Ve – Vo)/(Vt – Vo) and Ve is the elution volume of the protein; Vo is the void volume of the column calibrated with blue dextran, and Vt is the total bed volume. A set of globular protein standards was used as indicated in Fig. 1. Mass spectrometry analysis was performed in the mass spectrometry laboratory, Zentrums Pharmakologie und Toxikologie, Medizinische Hochschule Hannover. The protein was digested by trypsin, and the peptides were analyzed in an Ultraflex matrix-assisted laser desorption/ionization-time of flight/time of flight mass spectrometer (Bruker Daltonics). The GST-Dyn829–842 peptide was prepared as follows. Two overlapping oligonucleotides were designed according to the dynamin sequence (Swiss-Prot entry: Q05193) to encode the peptide sequence from residues 829–842 (829PPQVPSRPNRAPPG842). After annealing, the oligonucleotides were ligated into pGEX vector (GE Healthcare Life Sciences) to fuse a GST at the N-terminal (GST-Dyn829–842 peptide). The GST alone and GST-Dyn829–842 peptide were expressed in bacteria and purified as described (10Neely A. Garcia-Olivares J. Voswinkel S. Horstkott H. Hidalgo P. J. Biol. Chem. 2004; 279: 21689-21694Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). Dynamin mutation and CaV1.2 Δ AID deletion were done by standard overlapping PCR using complementary oligonucleotides. Binding Assay—Pull-down assays using His-tagged CaVβ2a-derivatives as baits were performed as described (10Neely A. Garcia-Olivares J. Voswinkel S. Horstkott H. Hidalgo P. J. Biol. Chem. 2004; 279: 21689-21694Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). Briefly, His-CaVβ2a derivatives were coupled to cobalt-based agarose for 1 h at 4°C and incubated for another hour with precleared tsA201 cell extract obtained 24–48 h after transfection with dynamin or with YFP-CaV1.2 expression vector. The beads were pelleted and washed five times, and bound fractions were eluted with SDS-PAGE loading buffer and resolved on SDS-PAGE. In the binding assays to dynamin, the gel was transferred to nitrocellulose membrane and subjected to immunoblot analysis using anti dynamin antibody (BD Biosciences). Binding to YFP-CaV1.2 was visualized by fluorescence scanning using a Typhoon imager (GE Healthcare Life Sciences). Xenopus Oocytes Preparation, Injection, and Electrophysiological Recordings—Xenopus laevis oocytes were prepared, injected, and maintained as in a previous report (10Neely A. Garcia-Olivares J. Voswinkel S. Horstkott H. Hidalgo P. J. Biol. Chem. 2004; 279: 21689-21694Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). All capped cRNAs were synthesized using the MESSAGE-machine (Ambion, Austin TX), resuspended in 10 μl of water and stored in 2-μl aliquots at –80 °C until use. The CaV1.2-subunit used in this study bears a deletion of 60 amino acids at the N-terminal end that increase expression (11Wei X. Neely A. Olcese R. Lang W. Stefani E. Birnbaumer L. Receptors Channels. 1996; 4: 205-215PubMed Google Scholar). Electrophysiological recordings on CaV1.2-expressing oocytes were performed 2–5 h after protein injection (50 nl of the protein stock solution, 1–2 mg/ml/oocyte) and 5–7 days after cRNA injection using the cut-open oocyte technique with a CA-1B amplifier (Dagan Corp., Minneapolis, MN) as described (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). The external solution contained (in mm)10Ba2+, 96 n-methylglucamine, and 10 HEPES pH 7.0, and the internal solution contained (in mm) 120 n-methylglucamine, 10 EGTA, and 10 HEPES, pH 7.0. Data acquisition and analysis were performed using the pCLAMP system and software (Axon Instruments Inc., Foster City, CA). Currents were filtered at 2 kHz and digitized at 10 kHz. Linear components were eliminated by P/-4 prepulse protocol. The normalized charge movement-voltage plot and the average current-voltage plot were obtained as described (12Dzhura I. Neely A. Biophys. J. 2003; 85: 274-289Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar) using a CA-1B amplifier (Dagan). Currents were filtered at 1 kHz and digitized at 20 kHz. Ionic currents mediated by Shaker potassium channel were recorded 1 day after cRNA injection with two-electrode voltage clamp technique using a Dagan TEV 200A or Warner OC725A and filtered at 10 kHz. For the bafilomycin treatment, oocytes were incubated with 500 nm bafilomycin A1 (Sigma) 24 h prior to protein injection (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). The Shaker channel used, Sh IR (inactivation removed), bears an N-terminal deletion that removes fast inactivation (13Hoshi T. Zagotta W.N. Aldrich R.W. Science. 1990; 250: 533-538Crossref PubMed Scopus (1274) Google Scholar). Surface Expression Measurements in Xenopus Oocytes—Surface expression of CaV1.2 channels bearing the HA epitope (CaV1.2-HA) was measured by immunoassay as described (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). Briefly, 5–7 days after CaV1.2 RNA injection, oocytes were separated in two groups: for electrophysiological recordings and for immunoassay. Oocytes were incubated in blocking buffer containing 1% bovine serum albumin followed by incubation with 1 μg/ml rat monoclonal anti-HA antibody (3F10, Roche Applied Science). After washing, oocytes were incubated with horseradish peroxidase-coupled secondary antibody (goat anti-rat FAB fragments, Jackson ImmunoResearch) and extensively washed, and individual oocytes were placed in 50 μl of SuperSignal enzyme-linked immunosorbent assay femto substrate (Pierce) in 96-well microplates (Optiplate, PerkinElmer Life Sciences) and chemiluminescence-quantified 30 s later with a luminometer (Viktor2, PerkinElmer Life Sciences). β2a-SH3 Reduces the Number of Channels Expressed in the Plasma Membrane—The purified SH3 domain of the CaVβ2a with an expected molecular mass of 16.7 kDa elutes as a mono-disperse peak from a size exclusion chromatography (Fig. 1B). The size exclusion chromatography calibration curve yielded an apparent molecular mass of 20.8 kDa that is compatible with a monomeric conformation of the protein. Mass spectrometry analysis on the purified β2a-SH3 confirmed its identity (data not shown). β2a-SH3 was injected in Xenopus oocytes expressing CaV1.2, and gating and ionic currents were measured using the cut-open oocyte voltage clamp technique. Injection of β2a-SH3 into oocytes causes a dramatic decrease in charge movement (Qon, Fig. 1C) that develops with a time constant of 0.9 h (Fig. 1D). Qon stems from the conformational changes leading to channel opening (14Bezanilla F. Stefani E. Methods Enzymol. 1998; 293: 331-352Crossref PubMed Scopus (35) Google Scholar) and, thus, it is proportional to the number of channels. Since decrease in Qon proceeds without changes in the voltage or time dependence (Fig. 1, E–G), it likely reflects a reduction in the number of channels in the cell surface (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). In contrast, injection of full-length CaVβ, does modify channel gating as expected (15Neely A. Wei X. Olcese R. Birnbaumer L. Stefani E. Science. 1993; 262: 575-578Crossref PubMed Scopus (203) Google Scholar) (Fig. 1E). We corroborated that the drop of Qon upon β2a-SH3 injection stems from a decrease in the number of channels in the plasma membrane by immunoassay (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). Channel surface expression was measured in oocytes expressing HA-tagged CaV1.2 channels and compared with Qon measurements on the same group of oocytes as shown in Fig. 2A. Impaired assembly and forward trafficking or enhanced backward trafficking may be responsible for the reduction in the number of channels expressed in the plasma membrane upon β2a-SH3 injection. To discriminate between these two possibilities, we examined the effect of β2a-SH3 when incorporation of new proteins into the plasma membrane was inhibited by bafilomycin. We have previously shown that indeed, bafilomycin treatment interrupts incorporation of new CaV1.2 channels in oocytes and causes a net reduction of channel density due to constitutive turnover (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). Injection of β2a-SH3 in bafilomycin-treated oocytes resulted in 35% reduction in Qon (Fig. 2C) that compares with the 29% observed in control conditions (Fig. 2B). Thus, down-regulation induced by β2a-SH3 was not prevented by bafilomycin, indicating that this domain interferes with the backward trafficking rather than with the incorporation of newly formed channels. β2a-SH3-induced Reduction of Qon Depends on Dynamin—Removal of membrane proteins from the surface implicates endocytosis. Several SH3-containing proteins participate in the regulation of this process by associating with dynamin, a GTPase that excises endocytic vesicles from the plasma membrane (16Hinshaw J.E. Annu. Rev. Cell Dev. Biol. 2000; 16: 483-519Crossref PubMed Scopus (583) Google Scholar, 17Marks B. Stowell M.H. Vallis Y. Mills I.G. Gibson A. Hopkins C.R. McMahon H.T. Nature. 2001; 410: 231-235Crossref PubMed Scopus (373) Google Scholar, 18Gout I. Dhand R. Hiles I.D. Fry M.J. Panayotou G. Das P. Truong O. Totty N.F. Hsuan J. Booker G.W. Cell. 1993; 75: 25-36Abstract Full Text PDF PubMed Scopus (483) Google Scholar, 19Okamoto P.M. Herskovits J.S. Vallee R.B. J. Biol. Chem. 1997; 272: 11629-11635Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar). The proline-rich domain (PRD) of dynamin binds to SH3 domains in the partner protein, and this interaction recruits dynamin to the plasma membrane. Moreover, endocytosis of ion channels and receptors through a dynamin-dependent process has been reported (20Carroll R.C. Beattie E.C. Xia H. Luscher C. Altschuler Y. Nicoll R.A. Malenka R.C. Von Zastrow M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 14112-14117Crossref PubMed Scopus (345) Google Scholar, 21Zeng W.Z. Babich V. Ortega B. Quigley R. White S.J. Welling P.A. Huang C.L. Am. J. Physiol. 2002; 283: F630-F639Crossref PubMed Scopus (66) Google Scholar, 22Shimkets R.A. Lifton R.P. Canessa C.M. J. Biol. Chem. 1997; 272: 25537-25541Abstract Full Text Full Text PDF PubMed Scopus (243) Google Scholar). Therefore, we investigated the potential role of dynamin in β2a-SH3 induced channel internalization. We first verified the presence of endogenous dynamin in Xenopus oocytes by Western blot analysis using anti-dynamin antibody and detected a protein of molecular mass similar to a heterologously expressed HA-tagged dynamin I (Fig. 3A). Coexpressing CaV1.2 with dynamin did not have a direct impact on channel expression, and β2a-SH3 induced reduction of Qon was equivalent (compare Fig. 1C and 3B). We then examined the effect of expressing a dominant-negative mutant of dynamin lacking GTPase activity that inhibits endocytosis (dynamin K44A) (23Herskovits J.S. Burgess C.C. Obar R.A. Vallee R.B. J. Cell Biol. 1993; 122: 565-578Crossref PubMed Scopus (396) Google Scholar). Co-expression with dynamin K44A reduced oocyte survival rate and yielded smaller Qon than CaV1.2 alone or with dynamin WT. Although the causes for these changes are unclear, we still observed that β2a-SH3-induced reduction of Qon was blunted by expression of dynamin K44A (Fig. 3C). To further test the role of dynamin, we fused a 14-amino-acid residue peptide spanning the proline-rich region of dynamin I to GST protein to produce GST-Dyn829–842 peptide. This peptide is known to disrupt the interaction between dynamin and SH-3 domains and to inhibit endocytosis in synaptic vesicles (24Shupliakov O. Low P. Grabs D. Gad H. Chen H. David C. Takei K. Camilli P.De Brodin L. Science. 1997; 276: 259-263Crossref PubMed Scopus (398) Google Scholar, 25Shpetner H.S. Herskovits J.S. Vallee R.B. J. Biol. Chem. 1996; 271: 13-16Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar). Fig. 3D shows that preincubation of β2a-SH3 with GST-Dyn829–842 peptide, but not with GST alone, inhibits its potency to reduce Qon. Furthermore, β2a-SH3 binds in vitro to dynamin (Fig. 4A) and, consistently with the electrophysiological data, this binding is partially blocked by GST-Dyn829–842 peptide but not by GST (Fig. 4B). We tested the ability of β2a-SH3 to bind to the full-length channel. Using a similar pull-down assay, we did not observe binding of β2a-SH3 to CaV1.2 fused to the yellow fluorescent protein (YFP-CaV1.2; Fig. 4C). In contrast and as expected, CaVβ and the functional core of CaVβ (26Opatowsky Y. Chomsky-Hecht O. Kang M.G. Campbell K.P. Hirsch J.A. J. Biol. Chem. 2003; 278: 52323-52332Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar) encompassing the SH3 and GK domains (Fig. 1A, C1–C2) bound to the channel. β2a-SH3 and Full-length CaVβ2a Down-regulate the Distantly Related Shaker Potassium Channel Expressed in Xenopus Oocytes—Because β2a-SH3 promotes channel internalization without binding to the channel protein, we examined the effect of β2a-SH3 onto the distantly related Shaker potassium channel that lacks binding activity to the CaVβ-subunit (27Bichet D. Cornet V. Geib S. Carlier E. Volsen S. Hoshi T. Mori Y. Waard De. M. Neuron. 2000; 25: 177-190Abstract Full Text Full Text PDF PubMed Scopus (310) Google Scholar). Injection of β2a-SH3 to oocytes expressing the Shaker channel resulted in no changes in channel gating (supplemental Fig. S1), but ionic currents were reduced by ∼60% 1 h after protein injection (Fig. 5A). This current reduction was also partially blocked by preincubation of β2a-SH3 with GST-Dyn829–842 peptide but not with GST (Fig. 5B). Full-length CaVβ2a preserves the ability of β2a-SH3 to down-regulate Shaker channels expressed in oocytes. CaVβ2a reduced ionic currents to a similar degree as β2a-SH3 (Fig. 5C) without changes in the voltage dependence (supplemental Fig. S2). This current decrease was also antagonized by GST-Dyn829–842 peptide. Moreover, CaVβ2a bound in vitro to dynamin, and this interaction was inhibited by GST-Dyn829–842 peptide (Fig. 5D). These results indicate that CaVβ still acts through the dynamin-dependent endocytic pathway. Full-length CaVβ2a Reduces the Number of Plasma Membrane CaV1.2 Channels Lacking the AID but Not WT Channels—A corollary from the above results is that free CaVβ may also be able to reduce surface expression of CaV1.2 channels when the CaVα1-CaVβ primary interaction site is disrupted. To test this possibility, we deleted the AID site of CaV1.2 (residues 459–475) to obtain CaV1.2-ΔAID channels. This mutated channel yields gating currents that, with respect to their voltage and time dependence, are indistinguishable from wild type CaV1.2 (Fig. 6, A and B), but as expected, CaVβ2a loses its ability to potentiate ionic currents (Fig. 6C). As recently corroborated by chemiluminescent enzyme immunoassay (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar), surface expression of CaV1.2 channels in oocytes is not altered by injection of CaVβ2a protein (Fig. 6D). In contrast, in oocytes expressing CaV1.2-ΔAID, injection of CaVβ2a reduces Qon to the same extent as did β2a-SH3 in oocytes expressing wild type CaV1.2 (Fig. 6E). To further prove that channels lacking the AID are indeed expressed in the plasma membrane and that CaVβ decreases the number of channels at the cell surface, we performed the surface expression assay with CaV1.2-ΔAID HA-tagged channels. CaVβ2a injection decreased Qon and chemiluminescence signal in CaV1.2-ΔAID-expressing oocytes (Fig. 6F). We here show that the SH3 domain of the β-subunit of the voltage-gated calcium channels promotes internalization of membrane proteins in a dynamin-dependent manner. In addition, we found that CaVβ-SH3 binds in vitro to dynamin and, since this association is inhibited by GST-Dyn829–842 peptide, we propose that this interaction is mediated by the PRD of dynamin. Nevertheless, simulated docking predictions indicate that CaVβ-SH3 is unlikely to interact with PXXP motifs unless a considerable structural rearrangement occurs (4Chen Y.H. Li M.H. Zhang Y. He L.L. Yamada Y. Fitzmaurice A. Shen Y. Zhang H. Tong L. Yang J. Nature. 2004; 429: 675-680Crossref PubMed Scopus (259) Google Scholar). The dynamin PRD-CaVβ-SH3 interaction may be mediated by non-canonical PXXP binding residues in CaVβ-SH3 or, alternatively, exposition of canonical residues may be tunable by a yet unknown regulatory protein or event. The interaction between recombinant β2a-SH3 and dynamin may reflect an in vivo phenomenon given that a SH3-only form of the CaVβ protein is expressed in cardiac cells (8Cohen R.M. Foell J.D. Balijepalli R.C. Shah V. Hell J.W. Kamp T.J. Am. J. Physiol. 2005; 288: H2363-H2374Crossref PubMed Scopus (34) Google Scholar). Binding of CaVβ-SH3 to the protein being sequestered is not required since no interaction between the β2a-SH3 and the whole CaV1.2 channel was observed, and certainly, no association occurs with the Shaker channel. CaVβ-SH3 has been reported to associate only with isolated regions or truncated CaVα1 channels (28Maltez J.M. Nunziato D.A. Kim J. Pitt G.S. Nat. Struct. Mol. Biol. 2005; 12: 372-377Crossref PubMed Scopus (51) Google Scholar, 29Takahashi S.X. Miriyala J. Tay L.H. Yue D.T. Colecraft H.M. J. Gen. Physiol. 2005; 126: 365-377Crossref PubMed Scopus (55) Google Scholar). Thus, it is conceivable that other cytoplasmic regions within the whole channel hinder this association. In the presence of the full-length CaVβ-subunit, calcium channels lacking the AID site, but not WT channels, are down-regulated, as though binding to CaVβ prevents the channel complex to be internalized. Since association of the CaVβ to VGCCs ensures normal channel activity, this would constitute an efficient quality control mechanism in which the same protein ensures functional fitness and survival of the channel in the plasma membrane (Fig. 7). Our recent finding that the CaVα1-CaVβ interaction is reversible at the level of the plasma membrane (9Hidalgo P. Gonzalez-Gutierrez G. Garcia-Olivares J. Neely A. J. Biol. Chem. 2006; 281: 24104-24110Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar) supports this mechanism. The ability of this auxiliary subunit to influence internalization of other membrane proteins anticipates that replacement of complete signal transduction assemblies may be triggered by the presence of free CaVβ. Although the whole picture is certainly still incomplete, our findings outline a novel signaling pathway for the regulation of intracellular calcium concentration. We thank Dr. Christoph Fahlke and Dr. David Naranjo for insightful discussion, Ute Scholl for kindly providing us with the dynamin constructs, Dr. Matthias Gaestel for kindly sharing the luminometer Victor2, and Dr. Andreas Pich for the mass spectrometry. Download .pdf (.1 MB) Help with pdf files