Characterization of the Protein Kinase Activity of TRPM7/ChaK1, a Protein Kinase Fused to the Transient Receptor Potential Ion Channel*

Channel-kinase TRPM7/ChaK1 is a member of a recently discovered family of protein kinases called (cid:1) -ki-nases that display no sequence homology to conventional protein kinases. It is an unusual bifunctional protein that contains an (cid:1) -kinase domain fused to an ion channel. The TRPM7/ChaK1 channel has been characterized using electrophysiological techniques, and recent evidence suggests that it may play a key role in the regulation of magnesium homeostasis. However, little is known about its protein kinase activity. To characterize the kinase activity of TRPM7/ChaK1, we expressed the kinase catalytic domain in bacteria. ChaK1-cat is able to undergo autophosphorylation and to phosphorylate myelin basic protein and histone H3 on serine and threonine residues. The kinase is specific for ATP and cannot use GTP as a substrate. ChaK1-cat is insensitive to staurosporine (up to 0.1 m M ) but can be inhibited by rot- tlerin. Because the kinase domain is physically linked to an ion channel, we investigated the effect of ions on ChaK1-cat activity. The kinase requires Mg 2 (cid:2) (opti-mum at 4–10 m M ) or Mn 2 (cid:2) (optimum at 3–5 m M ), with activity in the presence of Mn 2 (cid:2) being 2 orders of magnitude higher than

A new family of protein kinases that do not display sequence homology to conventional eukaryotic protein kinases has been recently identified (1,2). When mammalian and Caenorhabditis elegans elongation factor-2 kinases (eEF-2 1 kinases) were cloned, it was found that they do not display sequence homology to any conventional eukaryotic protein kinase (1). However, their catalytic domains appeared to be homologous to the catalytic domain of myosin heavy chain kinase A from Dictyoste-lium (3)(4)(5). Two more protein kinases with the same type of catalytic domain have been subsequently identified in Dictyostelium and have been called myosin heavy chain kinases B and C (6,7). This new family of protein kinases was named ␣-kinase, because the existing evidence suggests that these protein kinases phosphorylate amino acids located within ␣-helices (2). This is different from conventional protein kinases that phosphorylate amino acids located within loops, turns, or regions with irregular structure (8). The ␣-kinase catalytic domain is characterized by several conserved motifs, which are different from the distinguishing sequence motifs that are found in conventional protein kinases (2). Surprisingly, the recently determined three-dimensional structure of the ␣-kinase catalytic domain revealed that, despite the lack of sequence homology, ␣-kinases have a fold that is very similar to conventional eukaryotic protein kinases (9).
Five more human proteins with the ␣-kinase domain have been identified and cloned (2,10,11). Unexpectedly, it was found that two of these proteins, which we initially named melanoma and kidney ␣-kinases and subsequently renamed channel-kinases 1 and 2 (ChaK1 and ChaK2), contained domains that are homologous to members of the transient receptor potential (TRP) family of ion channels (10,11). Three other laboratories independently cloned ChaK1 and named it ChaK (9), TRP-PLIK (12), or LTRPC7 (13). The TRP family of ion channels consists of various cation channels with diverse cellular functions that can be subdivided into three subfamilies: TRP-classic (TRPC), TRP-vanilloid (TRPV), and TRP-melastatin (TRPM) (reviewed in Refs. 14 -16). According to recently suggested unified nomenclature for TRP channels (16), ChaK1 and ChaK2 are now called TRPM7 and TRPM6 respectively. To emphasize the unique bifunctional nature of TRPM7 and TRPM6, we will be referring to them as TRPM7/ChaK1 and TRPM6/ChaK2 in this paper.
The electrophysiological properties of the TRPM7/ChaK1 channel have been recently characterized (12,13). It was found that the TRPM7/ChaK1 channel is permeable to Ca 2ϩ and Mg 2ϩ , is inhibited by Mg 2ϩ or Mg-ATP (13), and is inactivated by phosphatidylinositol 4,5-bisphosphate hydrolysis (17). A distinctive current that is believed to be mediated by TRPM7/ ChaK1 and that can be inhibited by Mg 2ϩ , has been characterized in several types of cells (13, 18 -20). A recent study suggests that TRPM7/ChaK1 may represent a novel ion channel mechanism for cellular trace metal ion entry into vertebrate cells (21). Another recent study suggests that TRPM7/ ChaK1 may play a key role in the regulation of Mg 2ϩ homeostasis (22).
Despite the extensive characterization of TRPM7/ChaK1 channel activity, little is known about TRPM7/ChaK1 kinase activity besides its ability to autophosphorylate and to phos-phorylate myelin basic protein (10,12,22). In addition, the functional inter-relationship between the ␣-kinase domain and the ion channel domain in TRPM7/ChaK1 is still unclear.
This work provides the first detailed characterization of the protein kinase activity of channel-kinase TRPM7/ChaK1. We investigated the biochemical properties of the TRPM7/ChaK1 kinase catalytic domain and, in particular, the effect of various divalent metal ions on TRPM7/ChaK1 kinase activity. Our results suggest that, among divalent metal ions that TRPM7/ ChaK1 channel can permeate, only Mg 2ϩ can directly modulate TRPM7/ChaK1 kinase activity in vivo.

EXPERIMENTAL PROCEDURES
Materials-Buffer reagents and other chemicals were obtained from Sigma. Rottlerin and staurosporine were dissolved in Me 2 SO. Radioisotopes were from PerkinElmer Life Sciences. PCR reagents and restriction enzymes were from Invitrogen and Roche Applied Science. Myelin basic protein (MBP) and calmodulin were a kind gift of Dr. Donald Wolff (Department of Pharmacology, UMDNJ). eEF-2 purification as well as expression and purification of eEF-2 kinase were performed as described previously (5).
Expression and Purification of ChaK1-cat-We expressed the Cterminal part of the TRPM7/ChaK1 containing kinase catalytic domain as a fusion with maltose binding protein (ChaK1-cat). Two fusion proteins were produced: ChaK1-cat (short), containing the last 462 amino acids (1403-1864) of TRPM7/ChaK1 (GenBank TM accession number AF346629), and ChaK1-cat (long), containing the last 767 amino acids. All experiments were carried out with both the short and long forms of ChaK1-cat, and essentially no difference was observed in any of the experiments. Experimental data presented in Figs. 1-7 were obtained with the short form of ChaK1-cat.
To produce ChaK1-cat (short form), a DNA fragment corresponding to the TRPM7/ChaK1 kinase domain was obtained by PCR from HeLa marathon-ready cDNA (Clontech) using the following primers: mkh-C3, 5Ј-GTTAGTACACCATCTCAGCCAAGTTGCAAA-3Ј and mkh-CR, 5Ј-TTATAACATCAGACGAACAGAATTAGTTGATTCTGATTCT-3Ј. The PCR fragment was inserted into a pCR II-TOPO vector (Invitrogen) by TA cloning and then subcloned into a pMAL-p2x vector (New England Biolabs) using EcoRI restriction sites to obtain the resulting construct, pMAL-mkhC 3.15.
The expression and purification of ChaK1-cat (short form) was performed as follows. Escherichia coli DH5␣ cells harboring the pMAL-mkhC 3.15 construct were grown at 37°C in 1 liter of LB medium supplemented with 2g/liter glucose (with ampicillin, 100 g/ml) to A 600 of 0.5 and induced with 0.5 mM isopropyl-␤-D-thiogalactopyranoside. Cells were then grown for an additional 6 h at 37°C. All of the following procedures were carried out at 4°C. The cells were pelleted, resuspended in 10 ml of buffer A (50 mM Tris-HCl (pH 8.0), 100 mM NaCl, 10 mM MgCl 2 , 1 mM EDTA, 10 mM ␤-mercaptoethanol, and 20% (w/v) glycerol) containing complete protease inhibitor (Roche Applied Science), and dissolved in an 80-ml solution of 8 M urea in buffer A. The sample was sonicated and centrifuged at 30,000 ϫ g for 30 min. The supernatant was dialyzed overnight against buffer A. After dialysis, the sample was centrifuged again at 30,000 ϫ g for 30 min. The supernatant was loaded onto an amylose (New England Biolabs) column (10 ml) equilibrated with buffer A containing complete protease inhibitor (Roche Applied Science). Elution was performed by a step gradient using buffer A plus 10 mM maltose.
To produce ChaK1-cat (long form), a DNA fragment corresponding to the whole C-terminal part of TRPM7/ChaK1 starting right after the sixth trans-membrane domain was obtained by PCR using TRPM7/ChaK1 cDNA in pCR II-TOPO as a template and the following primers: chak1-C5-blunt, 5Ј-AATGTGTATTTACAAGTGA-AGGCAATTTC-3Ј, and chak1-C3-XbaI, 5Ј-GCTCTAGATTATAACAT-CAGACGAACAGAATTAGTTG-3Ј. KOD Hot Start DNA polymerase (Novagen) was used, which produced PCR product with blunt ends. The PCR product was digested with XbaI and then inserted into pMAL-c2x vector (New England Biolabs) using XmnI and XbaI restriction sites to produce the pMmkC2 construct. The expression and purification of the long form of ChaK1-cat was performed similar to ChaK1-cat (short form), except the expression was carried out at room temperature, which produced soluble protein. Therefore, the denatur-FIG. 1. Analysis of purified ChaK1-cat by electrophoresis and gel filtration. A, SDS-PAGE analysis of purified ChaK1-cat. 1.5 g of affinity-purified enzyme was subjected to electrophoresis on 7.5% SDS-polyacrylamide gels and stained with Coomassie Blue R-250. B, gel filtration chromatography of ChaK1-cat. 15 g of affinity-purified recombinant protein was subjected to gel filtration on a Bio-Silect SEC 250 -5 high pressure liquid chromatography column (Bio-Rad) equilibrated with buffer (25 mM Tris-HCl, pH7.6, 1 mM MgCl 2 , 0.5 mM dithiothreitol, and 150 mM NaCl). Fractions of 0.5 ml each were collected. ChaK1-cat activity was determined in each fraction using MBP as a substrate. Inset, the column was calibrated with standard molecular weight marker proteins (thyroglobulin, 670,000; IgG, 158,000; ovalbumin, 44,000; myoglobin, 17,000; and vitamin B 12 , 1350). ation/renaturation step involving the dissolution in urea and dialysis was omitted. Bacteria were lysed by incubation with lysozyme (1 mg/ml) for 1 h on ice and subsequently sonicated.
Protein Phosphorylation Assay-Assays for ChaK1-cat activity were carried out in an assay buffer containing 50 mM HEPES-KOH (pH 7.0), 10 mM MgCl 2 , 5 mM dithiothreitol, 100 M ATP, 2 Ci of [␥-33 P]ATP (specific activity of 3000 Ci/mmol), and 0.5 g of recombinant ChaK1cat. The total volume of the reaction was 50 l. After incubation at 30°C for 10 min, Laemmli sample buffer was added, and the samples were boiled. The samples were then analyzed by SDS-PAGE and autoradiography. In the case where MBP or histone H3 was phosphorylated by ChaK1-cat, the assays were performed as described above with the addition of 1 g of MBP or 1 g of histone H3.
Phosphoamino Acid Analysis-Autophosphorylation of ChaK1-cat and phosphorylation of MBP and histone H3 were done as described above, except that 20 Ci of [␥-33 P]ATP (specific activity of 3000 Ci/ mmol) per reaction was used, and the reaction volume was increased 5-fold. The samples were separated by 8% SDS-PAGE and transferred to a polyvinylidene difluoride Immobilon-P membrane (Millipore) by semi-dry transfer. The portion of membrane with phosphoprotein was excised and incubated in 6 M HCl at 110°C for 1.5 h. After incubation, the samples were dried with a Speedvac evaporator, and the dried material was dissolved in water. Nonradioactive phosphoserine, phosphothreonine, and phosphotyrosine were added to the samples. Phosphoamino acids were separated by two-dimensional electrophoresis on thin-layer cellulose plates 10 ϫ 10 cm (cellulose on polyester, Aldrich). The first dimension was performed in pH 1.9 electrophoresis buffer containing 0.58 M formic acid and 1.36 M acetic acid at 1000 V for 20 min. The second dimension was performed in pH 3.5 electrophoresis buffer containing 0.87 M acetic acid, 0.5% (v/v) pyridine, and 0.5 mM EDTA at 1000 V for 8 min. The TLC plates were stained with 0.2% ninhydrin in ethanol and then exposed to film.

RESULTS
To investigate the protein kinase activity of TRPM7/ChaK1, we expressed its C-terminal part containing kinase catalytic domain as a fusion with maltose binding protein. SDS-PAGE analysis of affinity-purified fusion protein (ChaK1-cat) showed a major band migrating at about the 95-kDa position (Fig. 1A).
To determine whether ChaK1-cat is a multimer in solution, gel filtration was performed. ChaK1-cat eluted as a single kinase activity peak corresponding to a protein with the molecular mass of ϳ300 kDa (Fig. 1B). This molecular mass is higher than the expected molecular mass of ChaK1-cat dimer (190 kDa); however, it is consistent with a dimer if the molecule is elongated. In fact, according to x-ray analysis, ChaK1 crystallizes as a dimer that has an unusually elongated structure (9). Therefore, active ChaK1-cat is likely to be a dimer in solution.
Affinity-purified ChaK1-cat was able to undergo autophosphorylation and also phosphorylated MBP and histone H3 ( Fig.  2A). ChaK1-cat appears to be specific for ATP and cannot utilize GTP as a phosphate donor. Neither autophosphorylation nor phosphorylation of MBP was observed when ATP was substituted in the reaction with GTP (Fig. 2B). Fig. 2C represents the time course of phosphorylation of MBP by ChaK1-cat. Incorporation of 33 P into MBP was linear during the first 15 min. To analyze the effect of autophosphorylation on kinase activity, ChaK1-cat was preincubated with unlabeled ATP for 20 min at 30°C before the addition of MBP and [␥-33 P]ATP. As can be seen from Fig. 2D, preincubation of ChaK1-cat with ATP resulted in inhibition of subsequent autophosphorylation but did not affect the kinetics of MBP phosphorylation, suggesting that autophosphorylation does not affect ChaK1-cat kinase activity.
Phosphoamino acid analysis revealed that autophosphorylation of ChaK1-cat and phosphorylation of MBP occur predominantly on serine (Fig. 3, A and B). Phosphorylation of threonine in both cases was barely detectable (Fig. 3, A and B). In contrast to autophosphorylation and phosphorylation of MBP, phosphorylation of histone H3 occurred at threonine (60%) and serine (40%) (Fig. 3C). As can be seen in Fig. 3D, eEF-2 kinase, another ␣-kinase that is known to phosphorylate its substrate on threonine, indeed phosphorylated eEF-2 exclusively on threonine.
Considering the unique nature of TRPM7/ChaK1 as being protein kinase covalently linked to an ion channel, there is a possibility that TRPM7/ChaK1 kinase activity can be regulated by the ions that permeate the channel. Because TRPM7/ChaK1 channel is permeable not only to Ca 2ϩ but also to other divalent metal ions, such as Mg 2ϩ , Mn 2ϩ , Zn 2ϩ , Ni 2ϩ , and Co 2ϩ (21) and is Mg 2ϩ -regulated (13), we investigated the effect of Mg 2ϩ and other divalent metal ions on ChaK1-cat kinase activity. As can be seen in Fig. 4, A and B, ChaK1-cat requires the presence of Mg 2ϩ for its activity, and the optimal concentration of Mg 2ϩ for both phosphorylation of MBP by ChaK1-cat and its autophosphorylation is ϳ4 -10 mM.
We found that Mn 2ϩ dramatically activates ChaK1-cat. Analysis of the effect of Mn 2ϩ at concentrations between 0.1 and 10 mM revealed that the stimulatory effect of Mn 2ϩ is noticeable at 0.5 mM and is maximal at ϳ3.5 mM (Fig. 4, C and  D). Both autophosphorylation and phosphorylation of MBP are strongly activated by Mn 2ϩ (Fig. 4C). Phosphorylation of MBP by ChaK1-cat at 3.5 mM Mn 2ϩ is stimulated ϳ70-fold (Fig. 4D). Autophosphorylation in the presence of Mn 2ϩ also caused a noticeable shift in the electrophoretic mobility of ChaK1-cat. In the presence of Mn 2ϩ , Mg 2ϩ is not required for the phosphorylation reaction, and at 3.5 mM Mn 2ϩ and 10 mM Mg 2ϩ kinase activity was slightly lower than in the presence of 3.5 mM Mn 2ϩ alone (Fig. 4A). Thus, ChaK1-cat requires either Mg 2ϩ or Mn 2ϩ ; however, the kinase activity is ϳ2 orders of magnitude higher in the presence of Mn 2ϩ than in the presence of Mg 2ϩ . As can be seen in Fig. 4E, phosphorylation of histone H3 by ChaK1-cat was also strongly increased when MnCl 2 was added to the reaction mixture.
It is known that some protein kinases can change their specificity for phosphorylated amino acid when a reaction is performed in the presence of Mn 2ϩ instead of Mg 2ϩ . For exam-ple, phosphorylase kinase acts as a serine-specific kinase in the presence of Mg 2ϩ but acts as a tyrosine kinase in the presence of Mn 2ϩ (23). Therefore, we performed phosphoamino acid analysis of autophosphorylated ChaK1-cat as well as MBP and histone H3 phosphorylated by ChaK1-cat in the presence of Mn 2ϩ . The addition of Mn 2ϩ does not affect specificity of ChaK1-cat for phosphorylated amino acid. When autophosphorylation of ChaK1-cat and phosphorylation of MBP were performed in the presence of Mn 2ϩ , phosphorylation of both serine and threonine increased with Ͼ90% of 32 P incorporated into serine (data not shown). Similarly, in the presence of Mn 2ϩ , phosphorylation of both threonine and serine in histone H3 was significantly increased. The activity and specificity of eEF-2 kinase were not affected by the addition of Mn 2ϩ (data not shown).
Other divalent cations that we tested cannot substitute for Mn 2ϩ in activating ChaK1-cat. In fact, Zn 2ϩ and Co 2ϩ strongly inhibit phosphorylation of MBP by ChaK1-cat (Fig. 5). Both Zn 2ϩ and Co 2ϩ at concentrations of 500 M completely inhibited MBP phosphorylation by ChaK1-cat (Fig. 5).
The addition of Ca 2ϩ (0.001, 0.01, 0.1, and 1 mM) or EGTA (2 mM) did not have any effect on the kinase activity of ChaK1-cat (Fig. 6A). However, in the presence of calmodulin, the addition of 0.1 or 1 mM Ca 2ϩ led to significant inhibition of MBP phosphorylation by ChaK1-cat (Fig. 6A). Similarly, 1 mM Ca 2ϩ in the presence of calmodulin inhibited the phosphorylation of histone H3 (Fig. 6B). However, the addition of Ca 2ϩ and calmodulin did not have any effect on the autophosphorylation of ChaK1-cat (Fig. 6, A and B). To further analyze the inhibitory effect of Ca 2ϩ /calmodulin on the phosphorylation of MBP by ChaK1-cat, the reaction was performed at various concentrations of MBP and ChaK1-cat. It was found that the inhibitory effect of Ca 2ϩ /calmodulin did not depend on the concentration of ChaK1-cat (Fig. 6C). However, the magnitude of the inhibitory effect of Ca 2ϩ /calmodulin was clearly dependent on the concentration of MBP, and at high concentrations of MBP, this inhibitory effect was not observed (Fig. 6D). These results suggest that the inhibitory effect of Ca 2ϩ /calmodulin on the phosphorylation of MBP by ChaK1-cat is either because of calmodulin interaction with MBP or because of competition between calmodulin and MBP for binding to the kinase.
We also investigated the effect of monovalent cations on ChaK1-cat kinase activity. Both K ϩ and Na ϩ produced significant inhibitory effects on phosphorylation of MBP by ChaK1cat at concentrations above 100 mM (Fig. 7A), suggesting that high ionic strength inhibits ChaK1-cat.
We analyzed the sensitivity of ChaK1-cat to some known inhibitors of conventional protein kinases. Interestingly, ChaK1-cat appears to be resistant to staurosporine, which did not produce any inhibitory effect even at the concentration of 100 M (Fig. 7B). Another protein kinase inhibitor, rottlerin, inhibits ChaK1-cat with an IC 50 of ϳ35 M (Fig. 7,  B and C). DISCUSSION In this work, we provide the first detailed characterization of the protein kinase activity of the channel-kinase TRPM7/ ChaK1. Among ␣-kinases, the kinase activity of only eEF-2 kinase and Dictyostelium myosin heavy chain kinases A, B, and C has been characterized previously. Interestingly, in contrast to eEF-2 kinase and myosin heavy chain kinases, which phosphorylate their substrates exclusively on threonines (24), we have found that phosphorylation of MBP by ChaK1-cat, as well as its autophosphorylation, occur predominantly on serine (Fig. 3). ChaK1-cat can also efficiently phosphorylate histone H3 on serine and threonine residues (Fig. 3). Thus, TRPM7/ChaK1 can be considered a serine/threonine-specific protein kinase.
According to gel filtration, ChaK1-cat exists in solution as a dimer. This is consistent with the crystallographic analysis of ChaK1. It was found that the kinase domain of ChaK1 forms a dimer in the crystal as a consequence of the exchange between monomers of a 27-residue helical segment (9). The oligomeric structure of ChaK1 in vivo is unclear. It is believed that TRP channels exist in vivo as tetramers (14 -16). Therefore, if kinase and channel domains of TRPM7/ChaK1 are, indeed, functionally and physically linked in vivo, the kinase might exist as a "dimer of dimers." We found that ChaK1-cat is specific for ATP and cannot use GTP as a substrate for the phosphorylation reaction (Fig. 2B). Although there are conventional protein kinases that are specific for ATP, many of them can utilize GTP as well as ATP (25). Analysis of ChaK1 structure (9,26) suggests an explanation of why this kinase is specific for ATP. It appears that one of the GTP oxygens would repel Glu-1718 located in the nucleotidebinding site, making binding of GTP unlikely. Interestingly, this glutamate (Glu-1718) is absolutely conserved among ␣-ki- nases (2,26), and therefore other ␣-kinases are likely to be specific for ATP. In fact, we found that eEF-2 kinase cannot use GTP as a substrate. 2 We found that staurosporine, a compound that interferes with ATP binding and inhibits most conventional protein kinases, does not have any effect on the kinase activity of ChaK1cat at concentrations up to 0.1 mM (Fig. 7B). This result was surprising given the structural similarity between ChaK1 and conventional protein kinases (9). However, detailed structural analysis (26) suggests an explanation for this result. In conventional protein kinases, there is substantial rearrangement of the residues in the active site to accommodate the bulky staurosporine molecule. However, in ChaK1, there is a salt bridge between Glu-1718 and Lys-1646 in the back of the hydrophobic pocket, which limits the flexibility of the binding site and makes staurosporine binding unlikely. Because amino acids making this salt bridge are conserved in all ␣-kinases (2,26), it is likely that other ␣-kinases will also not be inhibited by staurosporine. In fact, it was shown previously that eEF-2 kinase is relatively resistant to staurosporine (27).
Rottlerin, another compound known to inhibit protein kinases (see e.g. Ref. 28), inhibits both autophosphorylation and phosphorylation of MBP by ChaK1-cat with an IC 50 of ϳ35 M (Fig. 7, B and C). Rottlerin similarly inhibits eEF-2 kinase (27) and therefore may be a general inhibitor of ␣-kinases.
Considering the unique nature of TRPM7/ChaK1 as a protein kinase covalently linked to an ion channel, there is a possibility that TRPM7/ChaK1 kinase activity can be regulated by the ions that permeate the channel. Therefore, we investigated the effect of various metal ions on the ChaK1-cat kinase activity. TRPM7/ChaK1 is permeable to both Ca 2ϩ and Mg 2ϩ (13) as well as other divalent ions such as Mn 2ϩ , Zn 2ϩ , Ni 2ϩ , and Co 2ϩ (21).
The possible effect of Mg 2ϩ on TRPM7/ChaK1 kinase activity is particularly interesting not only because Mg 2ϩ can permeate this channel but also because TRPM7/ChaK1 channel activity is known to be modulated by intracellular Mg 2ϩ (13), and, according to a recent report, TRPM7/ChaK1 may play a key role in the regulation of magnesium homeostasis in vertebrates (22). In addition, it was found recently that mutations in a closely related channel-kinase, TRPM6/ChaK2, lead to hypomagnesemia, a disease characterized by a low Mg 2ϩ concentration in blood serum (29,30), suggesting that this channelkinase also plays an important role in the regulation of Mg 2ϩ homeostasis.
We found that ChaK1-cat kinase has Mg 2ϩ optimum between 4 and 10 mM (Fig. 4, A and B). This magnesium optimum is very similar to other ␣-kinases as well as many conventional protein kinases. Conventional protein kinases have two Mg 2ϩ binding sites, one with high affinity (dissociation constant Ͻ0.1 mM) and another with lower affinity (dissociation constant Ͼ1 mM) (31,32). Occupation of both sites with Mg 2ϩ is required for optimal activity of the majority of conventional protein kinases, and therefore they usually have a Mg 2ϩ optimum between 2 and 10 mM. Cyclic AMP-dependent protein kinase (PKA) is an exception to this rule in that binding of the second Mg 2ϩ is inhibitory (reviewed in Ref. 31), and, therefore, PKA has a Mg 2ϩ optimum below 1 mM. Because the Mg 2ϩ concentration in the cytoplasm is usually around 1 mM, ChaK1 kinase potentially can be regulated in vivo by changes in Mg 2ϩ concentration, particularly if TRPM7/ChaK1 functions as a Mg 2ϩ channel, and therefore can affect the local concentration of Mg 2ϩ in the vicinity of the kinase. Because TRPM7/ChaK1 may also function as a Ca 2ϩ channel (12,13), we investigated the effect of Ca 2ϩ on ChaK1-cat kinase activity. Our results demonstrate that the TRPM7/ChaK1 kinase domain was not sensitive to changes in calcium concentration in the range of 0.001-1 mM (Fig. 6A), and, therefore, it is unlikely that this kinase can be directly regulated by calcium flowing through the channel. However, at high concentrations of Ca 2ϩ , calmodulin inhibited ChaK1-mediated phosphorylation of MBP and histone H3 (Fig. 6, A and B). This effect could either be due to the binding of calmodulin to the substrate or because of competitive inhibition of the enzyme by calmodulin, as it was not observed when substrate concentration was increased (Fig. 6D). A similar effect was described by Wolff et al. (33), who found that binding of Ca 2ϩ /calmodulin to histones 1, 2A, and 2B inhibited histone dephosphorylation by bovine brain phosphatase.
TRPM7 is permeable to other divalent metal ions, and it was recently suggested that it can provide an ion channel mechanism for cellular entry of trace metal ions such as Mn 2ϩ , Zn 2ϩ , and Co 2ϩ (21). We investigated the effect of these metal ions on the kinase activity of ChaK1-cat and discovered that Mn 2ϩ dramatically stimulates phosphorylation of MBP and histone H3 by ChaK1 (Fig. 4). Zn 2ϩ and Co 2ϩ , in contrast to Mn 2ϩ , inhibited ChaK1-cat activity (Fig. 5). Interestingly, the effect of divalent metals on ChaK1-cat kinase activity is very similar to the effect of these metal ions on several conventional protein kinases, particularly tyrosine kinases (34 -40). It was reported for various tyrosine kinases, such as insulin receptor tyrosine kinase (35) and EGF receptor tyrosine kinase (34,36), that their activity can be dramatically stimulated in the presence of Mn 2ϩ . There are also serine/threonine kinases that can be activated by Mn 2ϩ (37,40). This effect is most likely because of the binding of Mn 2ϩ instead of Mg 2ϩ to the second metalbinding site (31,39). At the same time, it was demonstrated that binding of Zn 2ϩ or Co 2ϩ at this second binding site is inhibitory in tyrosine kinases (39). A striking similarity between the effects of divalent metal ions on ChaK1-cat activity and tyrosine kinases suggests that the catalytic mechanism of phosphotransfer in ␣-kinases and conventional protein kinases may be similar.
Is it possible that the effects of divalent metal ions that we observed are involved in the regulation of TRPM7/ChaK1 ki- nase activity in vivo? Because Mn 2ϩ , Zn 2ϩ , and Co 2ϩ produce their effects on ChaK1-cat activity at a very high concentration that is not observed at physiological conditions (Figs. 4 and 5), and Ca 2ϩ has no effect on ChaK1-cat activity (Fig. 6), these ions are unlikely to be involved in the modulation of TRPM7/ChaK1 kinase activity in vivo. Therefore, according to our results, among divalent metal ions, only Mg 2ϩ can potentially regulate TRPM7/ChaK1 kinase activity in the cell, which is also consistent with a recent report by Schmitz et al. (22). If, indeed, TRPM7/ChaK1 functions in vivo as a Mg 2ϩ channel, then it is possible that the kinase activity of TRPM7/ChaK1 is regulated by Mg 2ϩ flux through its ion channel.
According to the recent report, the TRPM7/ChaK1 channel plays an important role in the regulation of vertebrate cellular Mg 2ϩ homeostasis (22). It was suggested that the TRPM7/ ChaK1 channel acts as a primarily Mg 2ϩ -permeant channel and represents cellular Mg 2ϩ uptake mechanism (22). However, the function of TRPM7/ChaK1 kinase is unclear. The kinase domain is not required for TRPM7/ChaK1 channel gating, although it may play some role in modulation of channel activity (22). The kinase domain can also play a role in mediating the interaction of TRPM7/ChaK1 with other cellular proteins such as phospholipase C␤ (12,17). Clearly, the identification of TRPM7/ChaK1 kinase substrates is essential to understanding the physiological role of this kinase. The unique properties of TRPM7/ChaK1 protein kinase described in this paper, such as its dramatic activation by manganese and its insensitivity to staurosporine, will be important tools that may help to distinguish the activity of TRPM7/ChaK1 from other kinases and to identify its natural substrates. FIG. 7. Effect of monovalent metal ions and protein kinase inhibitors on ChaK1-cat activity. A, purified recombinant ChaK1cat was incubated with MBP in a reaction mixture containing 4 mM MnCl 2 , [␥-33 P]ATP, and different concentrations of K ϩ or Na ϩ . Kinase reactions were carried out as described under "Experimental Procedures." The samples were analyzed by SDS-PAGE and autoradiography. The graph was obtained by the quantification of the bands corresponding to phosphorylated MBP on the autoradiogram using the Kodak 1D imaging program. B, effect of different concentrations of rottlerin or staurosporine on ChaK1-cat activity. The reaction was performed using purified recombinant ChaK1-cat and MBP. The samples were analyzed by SDS-PAGE and autoradiography. C, the graph shows the effect of various concentrations of rottlerin on ChaK1-cat activity. The bands corresponding to phosphorylated MBP (on the autoradiogram shown in B) were quantified using the Kodak 1D imaging program.