CaMKII tethers to L-type Ca2+ channels, establishing a local and dedicated integrator of Ca2+ signals for facilitation

Abstract

Ca2+-dependent facilitation (CDF) of voltage-gated calcium current is a powerful mechanism for up-regulation of Ca2+ influx during repeated membrane depolarization. CDF of L-type Ca2+ channels (Ca(v)1.2) contributes to the positive force-frequency effect in the heart and is believed to involve the activation of Ca2+/calmodulin-dependent kinase II (CaMKII). How CaMKII is activated and what its substrates are have not yet been determined. We show that the pore-forming subunit alpha(1C) (Ca(v)alpha1.2) is a CaMKII substrate and that CaMKII interaction with the COOH terminus of alpha1C is essential for CDF of L-type channels. Ca2+ influx triggers distinct features of CaMKII targeting and activity. After Ca2+-induced targeting to alpha1C, CaMKII becomes tightly tethered to the channel, even after calcium returns to normal levels. In contrast, activity of the tethered CaMKII remains fully Ca2+/CaM dependent, explaining its ability to operate as a calcium spike frequency detector. These findings clarify the molecular basis of CDF and demonstrate a novel enzymatic mechanism by which ion channel gating can be modulated by activity.

Figure 1.

Phosphorylation of the α_1C subunit by CaMKII. (A) Autoradiogram showing phosphorylation of α_1C by CaMKII. Lysates from HEK cells transfected with α_1C and β2b (lanes 2–4) or nontransfected cells (lane 1) were immunoprecipitated with an anti-α_1C antibody (lanes 1, 3, and 4) or control IgG (lane 2) and then incubated with purified α-CaMKII in the presence of Ca²⁺/CaM and Mg²⁺/ATP³² as described in Materials and methods. 200 nM of the CaMKII inhibitor peptide AIP-2 (Calbiochem) was included (lane 4) to demonstrate kinase specificity. Phosphorylated α_1C is indicated by an arrowhead; autophosphorylated CaMKII, retained after the kinase reaction despite extensive washing of the immunoprecipitate, is indicated with a double arrowhead. An anti-α_1C immunoblot of the samples used in the kinase reaction is shown below the autoradiogram. (B) Schematic of α_1C. Thick black lines highlight regions used to generate GST fusion proteins. (C) GST fusion proteins enriched from bacterial lysates using glutathione–sepharose were incubated with purified α-CaMKII in the presence of Ca²⁺/CaM and Mg²⁺/ATP³² as described in Materials and methods. After extensive washes, proteins were eluted using SDS-PAGE sample buffer. Autoradiogram of fusion proteins separated by SDS-PAGE after phosphorylation by CaMKII. C-term refers to the more distal COOH-terminal fusion protein, containing aa 1669–2171. Above the autoradiogram is the Coomassie blue–stained band for each fusion protein, indicating nearly equal loading of substrate for all fusion proteins.

Figure 2.

CaMKII coimmunoprecipitates and colocalizes with α_1C. (A) Biotinylated calmodulin overlay of rat cardiac sarcolemmal membranes after immunoprecipitation with an anti-α_1C antibody. Purified α-CaMKII was run as a control to demonstrate effectiveness of CaM overlay. An anti-α_1C antibody (but not control IgG) coimmunoprecipitated a protein identified as the δ isoform of CaMKII by biotinylated CaM overlay and apparent molecular mass. (B) Anti-GFP immunoblot after immunoprecipitation of GFP-CaMKII by control IgG (lane 4) or anti-α_1C antibody (lane 5) from lysates of HEK 293 cells transiently transfected with GFP-CaMKII and α_1C.

Figure 3.

Activity-dependent interaction of CaMKII with the cytoplasmic determinants of α_1C. Immunoblots using an mAb (CBα2) for CaMKII after a GST pull-down assay with 20 nM of native (top), Ca²⁺/CaM-activated (middle), or Ca²⁺/CaM/autophosphorylated α-CaMKII (bottom). GST fusion proteins contained various cytoplasmic regions of α_1C just as in Fig. 1 C. Panel above the immunoblots shows a representative Ponceau stain of each fusion protein. Although only one Ponceau staining profile is shown, all blots were run in parallel, and equal loading of all fusions proteins was independently verified.

Figure 4.

Localization of the CaMKII binding site on the COOH terminus of α_1C. (A) Diagram of α_1C fusion proteins used in GST pull-down assays with autophosphorylated α-CaMKII, exhibiting robust (+), partial (±), and no (−) binding. (B) Immunoblot with CBα2 after GST pull down of 20 nM of purified autophosphorylated α-CaMKII, using α_1C aa 1581–1690 fused to GST. Pull-down assay performed in the presence of 40 μM of the indicated peptide or the peptide diluent DMSO. (C) Quantification after immunoblot with CBα2 of GST pull-down assays of purified autophosphorylated α-CaMKII, using α_1C aa 1581–1690 (wild type [WT]), a ¹⁶⁴⁴TVGKFY¹⁶⁴⁹ → EEDAAA mutant (Mut6), or GST alone shows that Mut6 blocks CaMKII binding. Panel above the immunoblots shows a representative Ponceau stain of each fusion protein. *, P < 0.001 for a one-way analysis of variance followed by Dunnett's test to identify specific pair-wise differences between the means for Mut6 versus WT and GST versus WT (n = 4–8). Inset shows an exemplar immunoblot with CBα2. (D) An exemplar immunoblot with an anti-CaM antibody, showing that CaM binding is not affected by the Mut6 mutation. Panel above the immunoblots shows a representative Ponceau stain of each fusion protein.

Figure 5.

CaMKII interaction with the COOH terminus of α_1C is essential for CDF. (A) I _Ba and scaled I _Ca traces during a train of 40 test pulses of V _h from –90 mV to +20 mV at 3.3 Hz recorded from oocytes expressing α_1C I1654A (I/A) or α_1C I1654A/¹⁶⁴⁴TVGKFY¹⁶⁴⁹ → EEDAAA (I/A-Mut6). Bars, 500 nA and 25 ms. (B) Peak I _Ba and I _Ca during trains of 40 repetitive test pulses at 3.3 Hz, normalized to the current amplitude at the beginning of each train (n = 4–5). Values indicate means ± SEM. (C) Changes in peak I _Ba and I _Ca conducted by α_1C I1654A (I/A) or α_1C I1654A/¹⁶⁴⁴TVGKFY¹⁶⁴⁹ → EEDAAA (I/A-Mut6) at indicated stimulation frequencies (n = 4–5) Values indicate means ± SEM. (D) Summary of the recovery from inactivation after a two-step protocol for I/A and I/A-Mut6. The length of the prepulse was individually determined for each oocyte to produce ∼75–90% inactivation. (E) Autoradiograph showing phosphorylation of wild type (WT) or mutant α_1C (Mut6) by CaMKII, performed as in Fig. 1 A. An anti-α_1C immunoblot of the samples used in the kinase reaction confirmed similar expression levels of the WT and mutant α_1C subunits. An anti-CaMKII immunoblot with CBα2 confirmed the identity of the retained 50-kD ³²P-labeled protein as α-CaMKII.

Figure 6.

The binding site for the COOH terminus of α_1C on CaMKII is localized near the catalytic domain. (A) Biotinylated CaM overlay of GST pull downs, using a fusion protein from the COOH terminus of α_1C (aa 1509–1905) on lysates of HEK 293 cells transiently transfected with the CaMKII isoforms (α, β, δ_A, δC, and γ_B; arrows) after thioautophosphorylation. In lanes 6 and 7, lysates of untransfected cells were run with (+) and without (−) purified thiophosphorylated α-CaMKII added to the lysate. (B) Immunoblot using an mAb (CBα2) for CaMKII after GST pull downs, using a fusion protein from the COOH terminus of α_1c (aa 1509–1905) and 20 nM of purified autophosphorylated α-CaMKII. In addition, 20 μM of the indicated peptide was added to each binding reaction. (C) Sequence alignment of CaMKII binding sites from the COOH termini of NR2B and α_1C with the autoregulatory domain from α-CaMKII.

Figure 7.

CaMKII interaction with the COOH terminus of α_1C is not reversed by dephosphorylation or CaM dissociation, and tethered CaMKII requires autophosphorylation or Ca^{2 +} /CaM for activity. (A and B) Immunoblots with CBα2 or a phosphospecific CaMKII mAb after GST pull-down assays, using α_1C aa 1509–1905 and 20 nM of autophosphorylated α-CaMKII. (A) 5 mM EGTA was present in the binding reaction and/or in the wash. (B) Purified recombinant PP1 was added before (PP1-Pre) or after (PP1-Post) the binding reaction in the presence or absence of 5 mM EGTA, as indicated. (C) Time course of reversal of CaMKII autonomous activity after PP1 treatment in solution (n = 4). (D) Activity measurements, using peptide AC-2 as a substrate, of CaMKII recovered in GST pull-down assays, using α_1C aa 1509–1905. Ca²⁺/CaM-dependent and autonomous activity measurements of CaMKII recovered after treatment with recombinant PP1 for 30 min (PP1) or no treatment (−) in the binding assay (n = 4) Values indicate means ± SD.

Figure 8.

Proposed mechanism of CaMKII binding to α_1C to form a local and dedicated Ca^{2 +} spike integrator for CDF. A catalytic core and autoregulatory domain for a prototypical CaMKII inactive subunit is shown on the bottom left (inactive is indicated by green). Ca²⁺/CaM activation and Thr²⁸⁶ autophosphorylation displace the CaMKII autoregulatory domain within the catalytic lobe to activate the subunit (yellow) and to expose an α_1C tethering site. The CaMKII holoenzyme remains bound to the α_1C COOH terminus even after removal of the Ca²⁺/CaM stimulus, and CaMKII dephosphorylation produces an inactive subunit. CaMKII may remain tethered to other cytoplasmic domains of α_1C as well. High depolarization frequencies would produce a threshold level of activated/autophosphorylated CaMKII subunits that increase the P _o of the channel via phosphorylation of the NH₂ and/or COOH termini (top left). At low depolarization frequencies and under the influence of phosphatase action, CaMKII activation would not be produced, favoring a low P _o for α_1C (top right).