Calmodulin Is Involved in the Ca2+-dependent Activation of Ceramide Kinase as a Calcium Sensor*

We recently demonstrated that the activation of ceramide kinase (CERK) and the formation of its product, ceramide 1-phosphate (C1P), are necessary for the degranulation pathway in mast cells and that the kinase activity of this enzyme is completely dependent on the intracellular concentration of Ca2+ (Mitsutake, S., Kim, T.-J., Inagaki, Y., Kato, M., Yamashita, T., and Igarashi, Y. (2004) J. Biol. Chem. 279, 17570-17577). Despite the demonstrated importance of Ca2+ as a regulator of CERK activity, there are no apparent binding domains in the enzyme and the regulatory mechanism has not been well understood. In the present study, we found that calmodulin (CaM) is involved in the Ca2+-dependent activation of CERK. The CaM antagonist W-7 decreased both CERK activity and intracellular C1P formation. Additionally, exogenously added CaM enhanced CERK activity even at low concentrations of Ca2+. The CERK protein was co-immunoprecipitated with an anti-CaM antibody, indicating formation of intracellular CaM·CERK complexes. An in vitro CaM binding assay also demonstrated Ca2+-dependent binding of CaM to CERK. These results strongly suggest that CaM acts as a Ca2+ sensor for CERK. Furthermore, a CaM binding assay using various mutants of CERK revealed that the binding site of CERK is located within amino acids 422-435. This region appears to include a type 1-8-14B CaM binding motif and is predicted to form an amphipathic helical wheel, which is utilized in CaM recognition. The expression of a deletion mutant of CERK that contained the CaM binding domain but lost CERK activity inhibited the Ca2+-dependent C1P formation. These results suggest that this domain could saturate the CaM and hence block Ca2+-dependent activation of CERK. Finally, we reveal that in mast cell degranulation CERK acts downstream of CaM, similar to CaM-dependent protein kinase II, which had been assumed to be the main target of CaM in mast cells.

Sphingolipids and their metabolites have emerged as a new class of lipid mediator of various cell functions (1)(2)(3). Ceramide (N-acylsphingosine; Cer), 2 the precursor for all sphingolipids, functions as a lipid second messenger in a variety of cellular events (4). Many stimuli, such as tumor necrosis factor ␣, Fas ligand, ␥-irradiation, anti-tumor reagents, and heat shock, cause an elevation in intracellular Cer content following the hydrolysis of sphingomyelin by endogenous sphingomy-elinase. These changes can result in cell cycle arrest, cell differentiation, and apoptosis (5,6). Cer is converted to sphingosine through the action of ceramidase, and in turn the sphingosine is metabolized to sphingosine 1-phosphate (S1P) by sphingosine kinase. S1P has been found to regulate cell growth (7) and motility (8,9). Interestingly, S1P inhibits apoptosis induced by Cer and Fas ligand (10), indicating that the balance of Cer/sphingosine/S1P can affect cell phenotype. The Cer/sphingosine/S1P pathway had been considered to be the major metabolic pathway of Cer. However, in 2002, the enzyme ceramide kinase (CERK), which metabolizes Cer to ceramide 1-phosphate (C1P), was cloned (11), revealing a new pathway for Cer metabolism. CERK activity was initially described as a Ca 2ϩ -stimulated lipid kinase activity that was co-purified with brain synaptic vesicles (12); it has since been reported in HL60 cells (13) and neutrophils (14). Additionally, CERK is thought to be involved in phagolysosome formation in polymorphonuclear leukocytes and to promote liposome fusion (15).
The product of CERK activity, C1P, has been reported to have mitogenic effects (16), although exogenously added C1P is rapidly hydrolyzed by a phosphatase (17). C1P was found to be a direct activator of cytosolic phospholipase A2 and to be involved in arachidonic acid release (18,19). Additionally, our previous report revealed that CERK was a mediator of Ca 2ϩ -dependent degranulation in mast cells (20). In both arachidonic acid release and mast cell degranulation, the intracellular elevation of Ca 2ϩ is a crucial event that acts as a regulatory mechanism of CERK activity. However, there are no apparent Ca 2ϩ binding domains in the primary structure of CERK (such as an EF-hand or C2 domain), and the Ca 2ϩ regulatory mechanism has not been well understood. Calmodulin (CaM) has been recognized as a calcium sensor that interacts with and regulates multiple protein targets (21). When intracellular Ca 2ϩ levels rise, four Ca 2ϩ ions bind to CaM and the Ca 2ϩ ⅐CaM complex binds to the target protein, initiating various signaling cascades. In this manner, CaM is known to regulate ion channels, cell cycle, and cytoskeletal organization and to influence development (22,23).
In the work presented here, we investigated the mechanism of the Ca 2ϩ -dependent activation of CERK and found that CaM was in fact involved. Utilizing point mutation analysis, we also identified the CaM binding site in CERK.
Cell Culture and Metabolic Labeling-Rat basophilic leukemia (RBL-2H3) cells and the previously established CERK transfectants RBL-CK3 and CK4 (20) were cultured as monolayers in Eagle's minimum essential medium (Sigma) supplemented with 10% fetal bovine serum, penicillin, and streptomycin. Chinese hamster ovary (CHO-K1) cells were maintained in Ham's nutrient mixture F12 medium (Sigma), supplemented with 10% fetal bovine serum, penicillin, and streptomycin. Human embryonic kidney 293FT cells were purchased from Invitrogen and cultured according to the manufacturer's directions. Cells were transfected with the indicated DNA using Lipofectamine Plus (Invitrogen) for CHO-K1 cells and Lipofectamine 2000 (Invitrogen) for human embryonic kidney 293FT cells, respectively.
For metabolic labeling, RBL-CK3 cells (10 6 ) were incubated with 2 Ci (2 l) of carrier-free [ 32 P]orthophosphoric acid (PerkinElmer) in 5 ml of Eagle's minimal essential medium for 90 min at 37°C. After the incubation, the cells were harvested and washed twice with Tyrode's buffer (25 mM PIPES (pH 7.2), 119 mM NaCl, 5 mM KCl, 0.4 mM MgSO 4 , 5.6 mM glucose, 1 mM CaCl 2 , and 0.1% bovine serum albumin). As indicated, W-7 was added to the Tyrode's buffer, and cells were preincubated for 10 min. Subsequently, the cells were stimulated at 37°C with 0.1 M A23187, a calcium ionophore, for 30 min. Reactions were terminated by adding 7 volumes of chloroform/methanol (1:1, v/v). Two phases were generated by adding 1.6 volumes of 1 M KCl. The organic phase was dried and subjected to a mild alkaline treatment to remove glycerophospholipids, as described elsewhere (13). Lipids were re-extracted by the method of Bligh and Dyer (24). [ 32 P]C1P was separated on HPTLC and imaged as above for the CERK assay.
Immunochemical Studies-For immunoprecipitations, harvested cells were incubated for 10 min at 4°C in IP buffer (50 mM Tris-HCl (pH 7.2), 150 mM NaCl, 5 mM EDTA, 0.5% Triton X-100, 10% glycerol, and Complete TM protease inhibitor mixture). After centrifugation at 15 ϫ 10 3 rpm for 5 min, primary antibodies were added to the supernatants, and samples were incubated overnight at 4°C. Antibody complexes were precipitated by incubating for 3 h at 4°C with 20 l of 50% protein G-Sepharose FF (Amersham Biosciences) in IP buffer. Prior to their use in CERK assay, the beads were pelleted and washed three times in IP buffer. For CERK assays the beads were repeatedly washed with phosphate-buffered saline to remove detergent and other reagents.
For use in Western blotting, the beads were taken up in 20 l of SDS loading buffer containing 2% 2-mercaptoethanol, and the proteins were separated by SDS-PAGE according to the method of Laemmli (26). Proteins were transferred onto a polyvinyldifluoride membrane according to the method described by Towbin et al. (27). After a 1-h incubation with 3% skim milk in Tris-buffered saline containing 0.2% Tween 20 (T-TBS), the membrane was incubated with antibodies at 4°C overnight. After a wash with T-TBS, the membrane was incubated for 2 h with horseradish peroxidase-conjugated anti-rabbit IgG (Amersham Biosciences). After another wash with T-TBS, the membranes were developed using an ECL kit (Amersham Biosciences) as recommended by the manufacturer and were then visualized using x-ray film.
CaM Affinity Binding Assays-CHO cells overexpressing CERK or CERK mutants were harvested and lysed with lysis buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM MgCl 2 , 1 mM dithiothreitol, 0.5% Triton X-100, and Complete TM protease inhibitor mixture). After the lysates were diluted with 4 volumes of dilution buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM MgCl 2 , and 1 mM dithiothreitol), they were centrifuged (20,000 ϫ g) for 10 min at 4°C to remove debris. The supernatants (50 l) were added to 50 l of CaM-Sepharose 4B beads that had been equilibrated with binding buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM MgCl 2 , and 0.1% Triton X-100) containing 2 mM CaCl 2 or 4 mM EGTA, and the mixture was incubated for 1 h at 4°C. Unbound proteins were removed by washing three times with binding buffer. Proteins bound to the CaM-Sepharose beads were then analyzed by Western blotting as described above. The blot was probed with an anti-CERK antibody that had been previously prepared (21).
Degranulation Assays-RBL-2H3, -CK3, or -CK4 cells were harvested and washed with Tyrode's buffer. After a 10-min preincubation in the presence or absence of the indicated concentration of W-7 or KN-93, the cells were stimulated at 37°C with 0.1 M A23187 for 30 min. Cells were then centrifuged at 800 ϫ g for 5 min, and the ␤-hexosaminidase activity was measured in both the supernatant and cell pellet, using p-nitrophenyl N-acetyl ␤-D-glucosaminide (Sigma) as a chromogenic substrate (28). Degranulation was expressed as a percentage of the ␤-hexosaminidase activity released into the supernatant.

RESULTS
CaM Is Involved in the Ca 2ϩ -dependent Activation of CERK and the Ca 2ϩ -dependent Formation of C1P-We previously reported that CERK is involved in Ca 2ϩ -dependent degranulation in mast cells and that C1P elevation following CERK activation is required for the elevation of intracellular Ca 2ϩ (20). This suggests that the activation of CERK by Ca 2ϩ is required for degranulation. Sugiura et al. (11) previously reported that nanomolar-order concentrations of Ca 2ϩ were sufficient to activate CERK in vitro. However, in our previous studies, immunoprecipitated CERK required micromolar-order concentrations of Ca 2ϩ for activation. The major difference between these two experiments was the source of the enzyme. Sugiura used whole cell lysates from CERK-overexpressing cells, whereas we used the immunoprecipitated enzyme. These facts suggested the presence of a cofactor for CERK that enhances its Ca 2ϩ sensitivity, a cofactor that is lost during the immunoprecipitation.
CaM is a well known Ca 2ϩ sensor that modulates the activity of a number of metabolic enzymes (21). We investigated whether CaM might be involved in the Ca 2ϩ activation of CERK, initially by examining the effects of the CaM antagonist W-7 on CERK activity in vitro. We used a previously established cell line, RBL-CK3, that stably expresses CERK and displays an activity ϳ250 times greater than that of its parent RBL-2H3 cells (20). Enzymes were prepared from the RBL-CK3 cells as whole cell lysates (lysate-CERK) or immunoprecipitated CERK. As shown in Fig. 1, W-7 had almost no effect on the activity of the immunoprecipitated CERK, whereas it decreased the CERK activity in the lysate in a dose-dependent manner. W-7 is known to bind to CaM and to inhibit Ca 2ϩ /CaM-regulated enzyme activity. It is possible that the immunoprecipitation process might cause the displacement of CaM associated with the CERK, in which case the W-7 would have no effect on the activity. However, in the enzyme assays of the cell lysate, W-7 increasingly excluded the binding of Ca 2ϩ from Ca 2ϩ /CaM⅐CERK complexes, resulting in a dose-dependent decrease in the Ca 2ϩ -dependent activation of CERK. Even in excess Ca 2ϩ , the presence of W-7⅐CaM complexes might interfere with the association of CERK and Ca 2ϩ . The baseline activity (determined in the absence of CaCl 2 ) of the lysate CERK was actually greater than that of the immunoprecipitated CERK, but it exhibited only moderate activation by exogenous CaCl 2 compared with the immunoprecipitated CERK, suggesting that endogenous Ca 2ϩ ⅐CaM complexes had already activated the CERK in the lysate to some extent.
To further evaluate the possible involvement of CaM in the activation of CERK, purified CaM from bovine brain was added to reaction mixtures of the immunoprecipitated CERK, and CERK assays were performed. Free Ca 2ϩ concentrations were adjusted using a CaCl 2 /EGTA mixture as calculated by the MAX Chelator program. In the presence of 4 M CaM and 600 nM Ca 2ϩ , the CERK activity increased two and half-fold ( Fig. 2A, black bar). This increase was completely blocked by 50 M W-7 ( Fig. 2A, gray bar), indicating that the activation was entirely dependent on CaM. In the absence of Ca 2ϩ , CaM and W-7 showed almost no effect. The moderate level of activation seen in the control lacking the added CaM was sensitive to W-7 ( Fig. 2A, white and gray  bars), suggesting that some CaM might remain in complex with CERK after the immunoprecipitation and might act as a Ca 2ϩ sensor.
Next, we examined whether the intracellular C1P formation would be affected by W-7. In a previous report, we showed that C1P formation in RBL-CK3 cells could be induced by the calcium ionophore A23187 (20). Therefore, RBL-CK3 cells were preincubated with [ 32 P]orthophospholic acid for 90 min and then stimulated for 30 min with 0.5 M A23187. W-7 treatment decreased the observed Ca 2ϩ -dependent formation of [ 32 P]C1P in a dose-dependent manner (Fig. 2B). This is consistent with the hypothesis that CaM acts as a calcium sensor for CERK.
Ca 2ϩ -dependent Association of CERK with CaM-To examine whether the CaM is bound directly to CERK, co-precipitation assays were performed using CaM-Sepharose 4B. Lysates from CHO cells expressing CERK were incubated with CaM-Sepharose 4B in the presence of 1 mM CaCl 2 or 2 mM EGTA. CERK was co-precipitated with CaM-Sepharose 4B only in the presence of Ca 2ϩ (Fig. 3A). This result supports the contention that CaM binds to CERK and that this binding is Ca 2ϩ -dependent.
Next, we examined whether endogenous CaM is also bound to CERK in situ. CERK was overexpressed in CHO cells, and the cells were treated with the cleavable protein-protein cross-linker, dimethyl 3,3-dithiobispropionimidate. CaM was immunoprecipitated using an anti-CaM antibody, the cross-linker was cleaved with 2-mercaptoethanol, and the sample was processed for Western blotting. CERK was co-immunoprecipitated with CaM only when the anti-CaM antibody was added (Fig.  3B, upper panel), yet the CERK levels were the same in both samples (Fig. 3B, lower panel). These experiments demonstrate that exogenous as well as endogenous CaM binds to CERK in a Ca 2ϩ -dependent manner.
Identification of the CaM Binding Site in CERK-To determine the CaM binding domain in CERK, we prepared a series of truncation mutants of CERK as shown in Fig. 4. Each mutant was expressed in CHO cells, and cell lysates were incubated with CaM-Sepharose in the presence of CaCl 2 or EGTA. Full-length CERK (FL) and the dC450 mutant both bound CaM in the presence of Ca 2ϩ , whereas dC330 did not. The C-terminal fragment (dN330) was also able to bind CaM. Together, these results indicate that CaM binding to CERK is restricted to the C-terminal region between residues 330 -450.
Most CaM targets contain one of two types of recognition motifs, a Ca 2ϩ -independent IQ motif or one of the Ca 2ϩ -dependent motifs 1-5-10 or 1-8-14 or its variants (21). Types 1-5-10 and 1-8-14 motifs are defined by amphipathic helices with a net positive charge, in which hydrophobic residues occur preferentially at positions 1-5-10 or 1-8-14. According to searches using the Calmodulin Target Data base (calcium.uhnres.utoronto.ca/ctdb/ctdb/home.html), there are several candidates for a CaM binding site present in CERK. Particularly, residues 422-435, which were also present in dN330 fragment, showed characteristics of a type 1-8-14B CaM binding motif. The motif corresponding to 1-8-14B is composed of the sequence (FILVW)XXXXXX(FAIL-VW)XXXXX(FILVW), with a net charge of ϩ2 to ϩ4. The hydrophobic amino acids Leu-422, Phe-429, and Leu-435 correspond to the 1st, 8th, and 14th amino acids, respectively (Fig. 4B, asterisks), and the net charge of this region is ϩ4. As shown in Fig. 4B, residues 422-435 of mouse CERK are conserved in human, Drosophila melanogaster, and Caenorhabditis elegans, especially in their hydrophobic nature. To determine whether these residues were involved in CaM binding, we replaced Leu-422, Phe-429, Phe-431, or Leu-435 in the dN330 fragment of CERK with Arg, to disrupt the hydrophobicity. These mutants were expressed and CaM binding assays were performed. Wild type CERK, dN330, and dN330F422R showed similar Ca 2ϩ -dependent CaM binding (Fig. 5A), whereas the dN330F429R mutant displayed only weak binding. Neither dN330F431R nor dN330F435R had any binding activity to CaM (Fig. 5A).
Regions formed by basic or hydrophobic amino acids are apparent in helical wheel models of many CaM-binding proteins. The helical wheel projection of residues 420 -437 of CERK is shown in Fig. 5B. The arrow indicates the hydrophobic moment. The right side of the wheel contains predominantly hydrophobic (white circles) and basic (gray circles) amino acids. In contrast, hydrophilic amino acids (black circles) including Cys-426, Asn-430, and Ser-427 are located in the left side of the wheel. The region from residues 420 -437 in CERK forms an amphipathic helix that fits the model of a putative CaM binding site. From the results of the binding assays, the hydrophobic natures of Phe-431 and Leu-435 appear to be important for CaM recognition, whereas amino acids Leu-422 and Phe-429, on the opposite side, seem to play a lesser role. As indicated by the hydrophobic moment, the hydrophobic amino acid in the upper right side of the wheel would appear to be important for the formation of the hydrophobic domain recognized by CaM. However, the double mutation for L422R and F429R diminished CaM binding activity of CERK, indicating that although these amino acids might not be important for CaM recognition, they are important for maintaining the helical structure of the CaM binding site.
We also generated the F431R mutation, which had the most significant impact on the CaM binding of the full-length CERK (Fig. 5A). The CaM binding activity of this mutant was greatly reduced compared with that of wild type CERK (Fig. 5C). Furthermore, the activities of F431R and L435R (not shown) were greatly reduced in comparison to that of the wild type CERK (Fig. 5D). Although diminished, weak binding of CaM to F431R remained detectable (Fig. 5C, lane C). This may indicate the presence of yet another CaM binding site, although the binding was much weaker compared with the full-length construct, suggesting that most of the binding occurs in the region from amino acids 422 to 435. Thus, this region can be considered the CaM binding site in CERK. To evaluate the role of this CaM binding domain in vivo, we examined whether the CaM binding domain could inhibit Ca 2ϩ -dependent formation of C1P. As shown in Fig. 5E, when the cells were transfected with FL and empty vector, calcium ionophore A23187 increased intracellular C1P contents efficiently (p Ͼ0.005 by analysis of variance). These results agreed with our previous report (20) and suggested that CERK was activated by intracellular Ca 2ϩ . However, when the cells were transfected with FL and dN330, which contained the CaM binding domain  A, to begin to identify the CaM binding site of the enzyme, co-sedimentation assays were performed on CERK deletion mutants. Full-length CERK (FL) and the mutants dC450, dC330, and dN330, encompassing the indicated mutations, were each expressed in CHO cells, and CaM co-sedimentation assays were performed as in Fig. 3A. B, the CaM Targeted Data Base (calcium.uhnres.utoronto.ca/ctdb/ctdb/home.html) suggested a possible CaM binding site in the CERK protein at amino acids 422-435, with a recognition motif of type 1-8-14B. Such motifs are defined by hydrophobic residues occurring preferentially at positions 1-8-14. The CERK proteins of mouse, human, Drosophila, and C. elegans are presented with asterisks indicating the corresponding 1st, 8th, and 14th hydrophobic amino acids. (Fig. 5A), calcium ionophore A23187 failed to increase intracellular C1P contents, suggesting that dN330 could saturate the CaM and hence block Ca 2ϩ -induced activation of CERK. These results indicate that the CaM binding domain of CERK indeed interacts with Ca 2ϩ /CaM and that this interaction is important for the Ca 2ϩ -dependent regulation of CERK.
CERK Mediates Mast Cell Degranulation in a CaM-dependent Mechanism-Our previous study indicated the involvement of CERK in mast cell degranulation (20). In the present study, we have shown that Ca 2ϩ /CaM is involved in the activation of CERK. Next we sought to examine the relationship between Ca 2ϩ /CaM-dependent activation of CERK and degranulation of mast cells. Pretreatment of RBL-2H3 cells with W-7 decreased Ca 2ϩ -dependent degranulation (Fig. 6A), indicat-ing that Ca 2ϩ /CaM is involved in the degranulation pathway. This is in agreement with other reports (29,30). In the degranulation of RBL-2H3 cells, CaMKII, myosin-light chain kinase, and calcineurin have each been proposed to be the target of CaM (30). Another report has implied the existence of a CaM-dependent, but CaMKII-independent, pathway in mast cell degranulation (29). To examine the relationship between other CaM-dependent enzymes and CERK in Ca 2ϩ -dependent degranulation in RBL-2H3, the effect of the CaMKII inhibitor KN-93 was examined. In this experiment, we used two CERK-overexpressing clones, RBL-CK3 and RBL-CK4, that we had previously established (20). Calcium ionophore-induced degranulation was decreased by KN-93 treatment in a dose-dependent manner (Fig. 6B, open circles), with the inhibitory effects reaching a plateau at 50 M. Interestingly, though, the CERK-overexpressing cells exhibited partial resistance to KN-93 (Fig.  6B, black and white squares) even at otherwise inhibitory levels. Because the overexpression of CERK has been shown to increase the extent of degranulation significantly (20), the present data might suggest that CERK contributes to the CaMKII-independent pathway. In our previous study, the inhibition of CERK activity caused a decrease in Ca 2ϩdependent degranulation and the exogenous introduction of CERK into permeabilized cells was enough to cause a degranulation (20). Together, these results indicate that CERK acts as a target of Ca 2ϩ /CaM, leading to degranulation in mast cells, and that the pathway involved might be separate from that of CaMKII, which was previously recognized as one of the intracellular targets of CaM. However, the details of any relationships between CERK and other CaM-dependent proteins are not clear from our experiment and must be part of a future study.

DISCUSSION
CaM can bind with high affinity to a relatively small ␣-helical region of amino acids, called an amphipathic helix, in many proteins. The binding mechanism between CaM and its target proteins is roughly classified as hydrophobic interaction. To determine the CaM binding site in the CERK protein, we introduced mutations to disrupt the helical structure and the hydrophobicity of its amphipathic helix. These mutants lost their CaM binding ability, allowing us to determine the CaM binding site. The disruption of the CaM binding also affected enzymatic activity, so we were unable to establish the precise connection between the CaM binding and the CERK activity in vitro. However, when the cells were  transfected with dN330 that contained the CaM binding domain, Ca 2ϩdependent formation of C1P was inhibited, suggesting that dN330 could saturate the CaM and hence block Ca 2ϩ -induced activation of CERK. Our results firmly establish that CaM is involved in Ca 2ϩ -dependent CERK activation.
In the presence of Ca 2ϩ the activity of the immunoprecipitated CERK was not affected by W-7 (Fig. 1), indicating that activation independent of CaM can also occur. Regardless of the stimuli though, the range of intracellular concentrations of Ca 2ϩ rises to several micromoles. Thus, under restricted intracellular conditions, CaM can facilitate Ca 2ϩ -dependent CERK activation in response to any rise in Ca 2ϩ . Ca 2ϩ ionophore-induced C1P formation was greatly decreased with W-7 treatment (Fig. 2B), indicating that the contribution of CaM is more critical than was previously perceived from the in vitro experiment.
CERK was originally cloned based on sequence homology to sphingosine kinase 1 (SPHK1) (11). Recently, translocation of SPHK1 was reported to be dependent on CaM (31) However, the activity of SPHK1 was not affected by W-7 treatment, indicating that CaM was not acting as an activator of this enzyme. The present study provides the first evidence of an enzyme in the sphingolipid signaling pathway using CaM as a Ca 2ϩ sensor.
In response to cell differentiation and extracellular stimuli, the intracellular sphingolipid content changes. Cer is produced immediately by the action of sphingomyelinase and functions as a second messenger in a variety of cellular events, including apoptosis and cell differentiation. Protein kinase C, phosphatidylinositol 3-kinase, phospholipase D, ceramide-activated protein kinase, and ceramide-activated phosphatase, all of which act as important enzymes in various biological processes, are regulated by Cer (4,(32)(33)(34). Thus, the control of intracellular Cer levels is very important in many signaling pathways. Ceramidase, glucosyltransferase, and sphingomyelin synthase have long been known to metabolize Cer. These enzymes have been cloned, and their catalytic topology has been determined (35)(36)(37)(38)(39)(40)(41). Of the Cer metabolic enzymes, only CERK seems to be able to metabolize Cer in the cytosolic region. The above-mentioned potential targets of Cer (protein kinase C, phosphatidylinositol 3-kinase, phospholipase D, ceramide-activated protein kinase, and ceramide-activated phosphatase) are all cytosolic proteins. Considering these facts, it is plausible that CERK acts as a key enzyme in the regulation of intracellular Cer content. Unlocking the mechanism of its activation through Ca 2ϩ /CaM, as reported here, advances our understanding of this enzyme and will facilitate future study into the function of intracellular Cer and C1P as second messengers.