Calmodulin antagonists induce cell cycle arrest and apoptosis in vitro and inhibit tumor growth in vivo in human multiple myeloma.

BACKGROUND: Human multiple myeloma (MM) is an incurable hematological malignancy for which novel therapeutic agents are needed. Calmodulin (CaM) antagonists have been reported to induce apoptosis and inhibit tumor cell invasion and metastasis in various tumor models. However, the antitumor effects of CaM antagonists on MM are poorly understood. In this study, we investigated the antitumor effects of naphthalenesulfonamide derivative selective CaM antagonists W-7 and W-13 on MM cell lines both in vitro and in vivo. METHODS: The proliferative ability was analyzed by the WST-8 assay. Cell cycle was evaluated by flow cytometry after staining of cells with PI. Apoptosis was quantified by flow cytometry after double-staining of cells by Annexin-V/PI. Molecular changes of cell cycle and apoptosis were determined by Western blot. Intracellular calcium levels and mitochondrial membrane potentials were determined using Fluo-4/AM dye and JC-10 dye, respectively. Moreover, we examined the in vivo anti-MM effects of CaM antagonists using a murine xenograft model of the human MM cell line. RESULTS: Treatment with W-7 and W-13 resulted in the dose-dependent inhibition of cell proliferation in various MM cell lines. W-7 and W-13 induced G1 phase cell cycle arrest by downregulating cyclins and upregulating p21cip1. In addition, W-7 and W-13 induced apoptosis via caspase activation; this occurred partly through the elevation of intracellular calcium levels and mitochondrial membrane potential depolarization and through inhibition of the STAT3 phosphorylation and subsequent downregulation of Mcl-1 protein. In tumor xenograft mouse models, tumor growth rates in CaM antagonist-treated groups were significantly reduced compared with those in the vehicle-treated groups. CONCLUSIONS: Our results demonstrate that CaM antagonists induce cell cycle arrest, induce apoptosis via caspase activation, and inhibit tumor growth in a murine MM model and raise the possibility that inhibition of CaM might be a useful therapeutic strategy for the treatment of MM.

In vitro substrate phosphorylation by Ca²âº/calmodulin-dependent protein kinase kinase using guanosine-5'-triphosphate as a phosphate donor.

BACKGROUND: Ca2+/calmodulin-dependent protein kinase kinase (CaMKK) phosphorylates and activates particular downstream protein kinases - including CaMKI, CaMKIV, and AMPK- to stimulate multiple Ca2+-signal transduction pathways. To identify previously unidentified CaMKK substrates, we used various nucleotides as phosphate donors to develop and characterize an in vitro phosphorylation assay for CaMKK. RESULTS: Here, we found that the recombinant CaMKK isoforms were capable of utilizing Mg-GTP as a phosphate donor to phosphorylate the Thr residue in the activation-loop of CaMKIα (Thr177) and of AMPK (Thr172) in vitro. Kinetic analysis indicated that the Km values of CaMKK isoforms for GTP (400-500 µM) were significantly higher than those for ATP (~15 µM), and a 2- to 4-fold decrease in Vmax was observed with GTP. We also confirmed that an ATP competitive CaMKK inhibitor, STO-609, also competes with GTP to inhibit the activities of CaMKK isoforms. In addition, to detect enhanced CaMKI phosphorylation in brain extracts with Mg-GTP and recombinant CaMKKs, we found potential CaMKK substrates of ~45 kDa and ~35 kDa whose Ca2+/CaM-induced phosphorylation was inhibited by STO-609. CONCLUSIONS: These results indicated that screens that use STO-609 as a CaMKK inhibitor and Mg-GTP as a CaMKK-dependent phosphate donor might be useful to identify previously unidentified downstream target substrates of CaMKK.

Generation of autonomous activity of Ca(2+)/calmodulin-dependent protein kinase kinase ß by autophosphorylation.

Ca(2+)/calmodulin-dependent protein kinase kinases (CaMKKs) phosphorylate and activate specific downstream protein kinases, including CaMKI, CaMKIV, and 5'-AMP-activated protein kinase, which mediates a variety of Ca(2+) signaling cascades. CaMKKs have been shown to undergo autophosphorylation, although their role in enzymatic regulation remains unclear. Here, we found that CaMKKα and ß isoforms expressed in nonstimulated transfected COS-7 cells, as well as recombinant CaMKKs expressed in and purified from Escherichia coli, were phosphorylated at Thr residues. Introduction of a kinase-dead mutation completely impaired the Thr phosphorylation of these recombinant CaMKK isoforms. In addition, wild-type recombinant CaMKKs were unable to transphosphorylate the kinase-dead mutants, suggesting that CaMKK isoforms undergo Ca(2+)/CaM-independent autophosphorylation in an intramolecular manner. Liquid chromatography-tandem mass spectrometry analysis identified Thr(482) in the autoinhibitory domain as one of the autophosphorylation sites in CaMKKß, but phosphorylation of the equivalent Thr residue (Thr(446)) in the α isoform was not observed. Unlike CaMKKα that has high Ca(2+)/CaM-dependent activity, wild-type CaMKKß displays enhanced autonomous activity (Ca(2+)/CaM-independent activity, 71% of total activity). This activity was significantly reduced (to 37%) by substitution of Thr(482) with a nonphosphorylatable Ala, without significant changes in Ca(2+)/CaM binding. In addition, a CaMKKα mutant containing the CaMKKß regulatory domain was shown to be partially phosphorylated at Thr(446), resulting in a modest elevation of its autonomous activity. The combined results indicate that, in contrast to the α isoform, CaMKKß exhibited increased autonomous activity, which was caused, at least in part, by autophosphorylation at Thr(482), resulting in partial disruption of the autoinhibitory mechanism.

Identification of a novel CaMKK substrate.

Ca(2+)/calmodulin-dependent protein kinase kinase (CaMKK) phosphorylates and activates specific downstream protein kinases including CaMKI, CaMKIV and 5'-AMP-activated protein kinase. In order to examine the variety of CaMKK-mediated signaling pathways, we searched for novel CaMKK substrate(s) using N(6)-(1-methylbutyl)-ATP and genetically engineered CaMKKα mutant, CaMKKα (Phe(230)Gly), that was capable of utilizing this ATP analogue as a phosphate donor. Incubation of rat brain extracts with recombinant CaMKKα (Phe(230)Gly), but not with wild-type kinase, in the presence of N(6)-(1-methylbutyl)-ATP and Ca(2+)/CaM, induced significant threonine phosphorylation of a 50kDa protein as well as CaMKI phosphorylation at Thr(177). The 50kDa CaMKK substrate was partially purified by using serial column chromatography, and was identified as Syndapin I by LC-MS/MS analysis. We confirmed that recombinant Syndapin I was phosphorylated by CaMKKα and ß isoforms at Thr(355)in vitro. Phosphorylation of HA-Syndapin I at Thr(355) in transfected HeLa cells was significantly induced by co-expression of constitutively active mutants of CaMKK isoforms. This is the first report that CaMKK is capable of phosphorylating a non-kinase substrate suggesting the possibility of CaMKK-mediated novel Ca(2+)-signaling pathways that are independent of downstream protein kinases.

Identification and characterization of PRG-1 as a neuronal calmodulin-binding protein.

Intracellular Ca2+-dependent cellular responses are often mediated by the ubiquitous protein CaM (calmodulin), which, upon binding Ca2+, can interact with and alter the function of numerous proteins. In the present study, using a newly developed functional proteomic screen of rat brain extracts, we identified PRG-1 (plasticity-related gene-1) as a novel CaM target. A CaM-overlay and an immunoprecipitation assay revealed that PRG-1 is capable of binding the Ca2+/CaM complex in vitro and in transfected cells. Surface plasmon resonance and zero-length cross-linking showed that the C-terminal putative cytoplasmic domain (residues 466-766) of PRG-1 binds equimolar amounts of CaM in a Ca2+-dependent manner, with a relatively high affinity (a Kd value for Ca2+/CaM of 8 nM). Various PRG-1 mutants indicated that the Ca2+/CaM-binding region of PRG-1 is located between residues Ser554 and Gln588, and that Trp559 and Ile578 potentially anchor PRG-1 to CaM. This is supported by pronounced changes in the fluorescence emission spectrum of Trp559 in the PRG-1 peptide (residues 554-588) upon binding to Ca2+/CaM, showing the stoichiometrical binding of the PRG-1 peptide with Ca2+/CaM. Immunoblot analyses revealed that the PRG-1 protein is abundant in brain, but is weakly expressed in the testes. Immunohistochemical analysis revealed that PRG-1 is highly expressed in forebrain structures and in the cerebellar cortex. Furthermore, PRG-1 localizes at the postsynaptic compartment of excitatory synapses and dendritic shafts of hippocampal neurons, but is not present in presynaptic nerve terminals. The combined observations suggest that PRG-1 may be involved in postsynaptic functions regulated by intracellular Ca2+-signalling.

Identification and characterization of wolframin, the product of the wolfram syndrome gene (WFS1), as a novel calmodulin-binding protein.

To search for calmodulin (CaM) targets, we performed affinity chromatography purification of a rat brain extract using CaM fused with GST as the affinity ligand. Proteomic analysis was then carried out to identify CaM-binding proteins. In addition to identifying 36 known CaM-binding proteins, including CaM kinases, calcineurin, nNOS, the IP(3) receptor, and Ca(2+)-ATPase, we identified an ER transmembrane protein, wolframin [the product of the Wolfram syndrome gene (WFS1)] as interacting. A CaM overlay and an immunoprecipitation assay revealed that wolframin is capable of binding the Ca(2+)/CaM complex in vitro and in transfected cells. Surface plasmon resonance analysis and zero-length cross-linking showed that the N-terminal cytoplasmic domain (residues 2-285) of wolframin binds to an equimolar unit of CaM in a Ca(2+)-dependent manner with a K(D) for CaM of 0.15 muM. Various truncation and deletion mutants showed that the Ca(2+)/CaM binding region in wolframin is located from Glu90 to Trp186. Furthermore, we demonstrated that three mutations (Ala127Thr, Ala134Thr, and Arg178Pro) associated with Wolfram syndrome completely abolished CaM binding of wolframin. This observation may indicate that CaM binding is important for wolframin function and that impairment of this interaction by mutation contributes to the pathology seen in Wolfram syndrome.

Activation of SAD kinase by Ca2+/calmodulin-dependent protein kinase kinase.

To search for the downstream target protein kinases of Ca (2+)/calmodulin-dependent protein kinase kinase (CaMKK), we performed affinity chromatography purification of a rat brain extract using a GST-fused CaMKKalpha catalytic domain (residues 126-434) as the affinity ligand. Proteomic analysis was then carried out to identify the CaMKK-interacting protein kinases. In addition to identifying the catalytic subunit of 5'-AMP-activated protein kinase, we identified SAD-B as interacting. A phosphorylation assay and mass spectrometry analysis revealed that SAD-B was phosphorylated in vitro by CaMKK at Thr (189) in the activation loop. Phosphorylation of Thr (189) by CaMKKalpha induced SAD-B kinase activity by over 60-fold. In transfected COS-7 cells, kinase activity and Thr (189) phosphorylation of overexpressed SAD-B were significantly enhanced by coexpression of constitutively active CaMKKalpha (residues 1-434) in a manner similar to that observed with coexpression of LKB1, STRAD, and MO25. Taken together, these results indicate that CaMKKalpha is capable of activating SAD-B through phosphorylation of Thr (189) both in vitro and in vivo and demonstrate for the first time that CaMKK may be an alternative activating kinase for SAD-B.