PfPKB, a Protein Kinase B-like Enzyme from Plasmodium falciparum

Intracellular cell signaling cascades of protozoan parasite Plasmodium falciparum are not clearly understood. We have reported previously (Kumar, A., Vaid, A., Syin, C., and Sharma, P. (2004) J. Biol. Chem. 279, 24255–24264) the identification and characterization of a protein kinase B-like enzyme in P. falciparum (PfPKB). PfPKB lacks the phosphoinositide-interacting pleckstrin homology domain present in mammalian protein kinase B. Therefore, the mechanism of PfPKB regulation was expected to be different from that of the host and had remained unknown. We have identified calmodulin (CaM) as the regulator of PfPKB activity. A CaM binding domain was mapped in the N-terminal region of PfPKB. CaM, in a calcium-dependent manner, interacts with this domain and activates PfPKB. CaM associates with PfPKB in the parasite and regulates its activity. Furthermore phospholipase C acts as an upstream regulator of this cascade as it facilitates the release of calcium from intracellular stores. This is one of the first multicomponent signaling pathways to be dissected in the malaria parasite.

Plasmodium falciparum is responsible for most cases of human malaria worldwide. This parasite invades both hepatocytes as well as erythrocytes in human host, but it is the erythrocytic phase of its life cycle that causes severe pathogenesis of malaria. After invading erythrocytes, the parasite undergoes well defined developmental changes inside the erythrocyte host. The parasite adopts a ringlike morphology and acquires necessary nutrients from the host during the trophozoite stages. Subsequently nuclear division gives rise to multinucleated schizont containing ϳ24 merozoites. These merozoites when released after schizont rupture invade fresh erythrocytes to start another cycle of asexual development. Although it is known that Plasmodium can utilize host G-protein signaling (2) and alters phosphorylation of erythrocyte cytoskeletal proteins during infection (3), parasite signaling pathways have remained largely uncharacterized. Given the importance of cell signaling cascades in proliferation and differentiation of eukaryotic cells, dissection of signal transduction mechanisms may provide useful insights about the development of this protozoan parasite. Plasmodium genome analysis revealed that there are close to 65 protein kinases, major mediators of cell signaling, in P. falciparum (4,5). Apart from a few of these kinases (6 -10), the function and mechanism of regulation and identity of cellular targets of most of these enzymes is largely unknown.
We recently identified a protein kinase B-like enzyme in P. falciparum (PfPKB) 3 . Despite sharing significant sequence homology (ϳ70%) with the catalytic domain of PKB, PfPKB lacks a pleckstrin homology (PH) domain present at the N terminus of the mammalian enzyme. The N-terminal region (NTR) of PfPKB is inhibitory as its deletion results in PfPKB catalytic activation (1). The NTR does not exhibit similarity with any other protein in the non-redundant protein data base. PKB binds phosphoinositides via the PH domain, which is crucial for its membrane translocation and catalytic activation. Whereas PKB is activated by phosphoinositide-dependent kinase 1-mediated phosphorylation at Thr-308 in its activation loop (11), autophosphorylation of PfPKB at its analogous Ser-271 residue results in its activation (1). PfPKB is expressed mainly in the schizont/merozoite stages of P. falciparum. Using a pharmacological inhibitor, we had proposed a role of PfPKB during schizont-to-ring transition of the parasite (1). Despite this information, it had remained unknown how PfPKB is regulated in P. falciparum. In the present study we identified calmodulin as the upstream regulator of PfPKB activity in vitro and in vivo. These findings resulted in identification of a novel signaling pathway in the malaria parasite.

EXPERIMENTAL PROCEDURES
Reagents-pGEX4T1-PfPKB and pET-NTR plasmids used for protein expression and anti-PfPKB rabbit antisera used in these studies have been described earlier (1). Site-directed mutagenesis was carried out using the QuikChange site-di-* This work was supported in part by a senior research fellowship (to P. S.) from The Wellcome Trust, UK and by intramural funding from the Department of Biotechnology, New Delhi to National Institute of Immunology. Paper I in this series is Ref. 1. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  rected mutagenesis kit (Stratagene). The peptides crosstide (GRPRTSSFAEG), calmodulin (CaM) binding domain (CBD) peptide (IGKKRLRNSMSLSYERKKRIR), and scrambled (scr) peptide (MKLSGKRYRNSRLKEIRSRIK) were custom synthesized by Peptron. scr peptide has an amino acid composition similar to CBD, but their arrangement has been scrambled. U73322, U73122, and W7 were purchased from Calbiochem. Anti-CaM monoclonal antibody against Dictyostelium discoideum CaM and purified bovine CaM were also obtained from Calbiochem. Unless indicated, all other fine chemicals were purchased from Sigma. Parasite Culture-P. falciparum strain 3D7 was cultured at 37°C in RPMI 1640 medium using either 10% AB ϩ human serum or 0.5% Albumax II (Invitrogen) (complete medium). Cultures were gassed with 7% CO 2 , 5% O 2 , and 88% N 2 , and synchronization of the parasites in culture was achieved by sorbitol treatment (1,12). Sorbitol synchronization yielded parasites purely in ring form; these rings matured to trophozoites 30 -36 h later. After nuclear division schizonts containing merozoites were observed. Ruptured schizonts with emerging merozoites were seen after ϳ44 h; this was followed by formation of fresh rings as a result of red blood cell invasion. Typically pharmacological inhibitors were incubated with schizonts for 15-60 min (ϳ3% parasitemia) at 5% hematocrit.

Site-directed Mutagenesis and Recombinant Protein Expression-
For expression of PfPKB as a GST fusion protein, pGEX4T1-PfPKB plasmid construct was used (1). Deletion mutant ⌬CBD and S271A mutants were generated by using the above mentioned PfPKB construct and the QuikChange site-directed mutagenesis kit (Stratagene). Recombinant proteins were expressed and purified as described earlier (1). Protein concentration was estimated by performing densitometry of SDS-PAGE gels using NIH Image software.
Immunoblotting and Immunoprecipitation-Parasites were released from infected erythrocytes by 0.05% (w/v) saponin treatment. Cell-free protein extracts from specific parasite stages were prepared by suspending parasite pellets in a buffer containing 10 mM Tris, pH 7.5, 100 mM NaCl, 5 mM EDTA, 1% Triton X-100, 20 M sodium fluoride, 20 M ␤-glycerophosphate, 100 M sodium orthovanadate, and 1ϫ Complete protease inhibitor mixture (Roche Applied Science) using a syringe and a needle. In most experiments 1 mM CaCl 2 was also included in the buffer, and 1 mM EGTA was added to perform experiments in calcium-free buffer. Lysates were cleared by centrifugation at 14,000 ϫ g for 30 min. PfPKB was immunoprecipitated from the schizont or the merozoite lysates using anti-PfPKB antisera (1). 50 -100 g of lysate were incubated with the antisera at 4°C overnight on an end-to-end shaker. Subsequently antigen-antibody complexes were incubated with 50 l of protein A/G-Sepharose for 6 h with shaking at 4°C. Resin was washed with the lysis buffer several times and was finally resuspended in 50 l of 1ϫ kinase assay buffer. After separation of parasite lysates or PfPKB-IP on SDS-PAGE gels, Western blotting was performed as described previously (1).
Assay of Kinase Activity-GST fusion proteins of PfPKB (or its variants) or 10 l of immunoprecipitated PfPKB were assayed in a buffer containing 50 mM Tris, pH 7.5), 10 mM magnesium chloride, 1 mM dithiothreitol, and 100 M [␥-32 P]ATP (6000 Ci/mmol) using a small peptide substrate ("crosstide") or histone II AS as the phosphate acceptor substrate. Reactions were terminated by spotting the reaction mixture on P81 phosphocellulose paper (Whatmann), and phosphate incorporation was measured by scintillation counting of the P81 paper. When histone was used as the substrate, reactions were stopped by boiling the reaction mixture in SDS-PAGE loading buffer. After electrophoresis, phosphate incorporation in histone was visualized by using a Fuji FLA5000 phos- phorimaging system. For typical CaM activation experiments, recombinant proteins were incubated with purified bovine CaM in the presence or absence of CaCl 2 15 min prior to addition of the phosphoacceptor substrate (100 M crosstide or 5 g of histone) and ATP. Unless indicated, the concentration of CaM and CaCl 2 used in kinase assays was 5 and 100 M, respectively. For experiments described in Fig. 1B, Ca 2ϩ /CaM was preincubated with GST-PfPKB 1 h and ATP prior to the addition of crosstide. 1 unit of PfPKB activity is equivalent to 1 mol of phosphate/min/mg. All experiments were done at least three times.

RESULTS
Calmodulin Activates PfPKB Activity in a Calcium-dependent Manner-We have previously reported that PfPKB shares significant homology with the catalytic domain of PKB. However, it lacks the N-terminal PH domain present in PKB (Fig.  1A). When NTR of PfPKB is deleted it results in its catalytic activation, which is strictly dependent on autophosphorylation of Ser-271 in its activation loop (1). Based on these observations, it was reasonable to assume that PfPKB may be activated by interaction of regulatory molecules with the inhibitory NTR.
Preliminary experiments performed in our laboratory suggested that the activity of recombinant PfPKB can be regulated by factors present in parasite protein lysates in a calcium-dependent manner. Because the activity of recombinant PfPKB was not directly modulated by calcium, it was reasonable to speculate that calcium may control PfPKB activity with the aid of one of its effector molecules (data not shown). CaM is a FIGURE 2. CaM interacts with a CBD in the NTR of PfPKB. A, presence of a putative CBD in the N-terminal region of PfPKB, which has several basic (blue) and hydrophobic (brown) residues. Shown is a sequence comparison of PfPKB CBD with the CaM binding regions of smooth muscle light chain kinase (smMLCK ), CaM-dependent kinase II (CamKII), and myristoylated alanine-rich C kinase substrate (MARCKS). Key hydrophobic residues involved in CaM anchoring are underlined (16,27). An RXRXS type pseudosubstrate motif embedded in CBD of PfPKB is also indicated. B, deletion of CBD impairs activation of PfPKB by Ca 2ϩ /CaM. Equal amounts of GST-PfPKB or its CBD-deleted version, GST-⌬CBD, were incubated with or without 5 M CaM and 100 M CaCl 2 . Kinase assays were performed as described for Fig. 1. C, 2.5 g of GST-PfPKB or GST-⌬CBD were incubated with CaM-agarose resin in a buffer containing 1 mM CaCl 2 (inset, top panel) or 1 mM EGTA (inset, bottom panel). Proteins bound to CaM were eluted with 5 mM EGTA and were detected by performing immunoblotting using anti-GST antibody. Quantitation of protein bands from the blots was done by densitometry, and data are presented as percentage of total input protein that was bound to Ca 2ϩ /CaM. The inset shows a representative immunoblot with eluted protein (B) and the input protein (I). No detectable binding of PfPKB or ⌬CBD was observed in the absence of calcium. D, 10 M CaM was incubated with 10 M CBD peptide (lane 2) or alone (lane 1) in a buffer containing 1 mM CaCl 2 (top panel) or 1 mM EGTA (bottom panel). Lane 3 represents a control experiment in which the CBD peptide alone was incubated with the buffer. The mixture was separated by native PAGE. The Coomassie-stained gel shows slower mobility of CaM-CBD complex only in the presence of calcium. E, CBD peptide prevents activation of PfPKB by CaM. GST-PfPKB was incubated with 5 M CaM, which had been preincubated with the indicated concentration of CBD, scr-CBD peptides, water, and kinase assays were performed using 1 mM crosstide as substrate as described under "Experimental Procedures." Activity of CaM-PfPKB in the absence of CBD peptides was considered as 100%. Data are presented as mean Ϯ S.E. of three independent experiments. C.D., catalytic domain. major calcium-binding protein ubiquitously expressed in almost all eukaryotes including Plasmodium and is well conserved across species. P. falciparum homologue of CaM (Gen-Bank TM accession number X56950) shares ϳ97% homology with CaM from other eukaryotes (14). Incubation of CaM with recombinant PfPKB resulted in a dose-dependent increase in its activity as judged by its ability to phosphorylate a small peptide, crosstide, a well established PKB substrate (15) (Fig. 1B). The maximal activation of 1.4 M PfPKB was achieved at ϳ2 M CaM with a K CaM of ϳ0.5 M. Importantly PfPKB activation by CaM was dependent on calcium (Fig. 1C). Because it has been shown previously that PfPKB activation is dependent on autophosphorylation of Ser-271 in its activation loop (1), it was worth exploring the role of phosphorylation of this site in CaMmediated PfPKB activation. Calcium/CaM catalyzed autophosphorylation of PfPKB (Fig. 1D), and mutation of its Ser-271 to Ala (S271A) resulted in almost complete loss of its autophosphorylation and concomitant attenuation of its catalytic activ-ity (Fig. 1E). Collectively these data suggest that Ca 2ϩ /CaM activates PfPKB by promoting its autophosphorylation at Ser-271 (Fig. 1, B-E).
Identification of a CBD in PfPKB-CaM interacts with segments of proteins that form amphipathic ␣-helices and are rich in basic and hydrophobic amino acids (16). On examination of NTR of PfPKB, a 21-amino acid motif possessing a putative CBD was identified ( Fig. 2A). To test whether this motif is the CBD of PfPKB, a deletion mutant of PfPKB lacking these 21 amino acids (⌬CBD) was created. Unlike PfPKB, CaM failed to activate this mutant (Fig. 2B). Direct binding of PfPKB and ⌬CBD was tested by using CaM immobilized on agarose. Whereas PfPKB exhibited significant binding to CaM-agarose, ⌬CBD mutant failed to interact with CaM (Fig. 2C), suggesting that this 21-amino acid stretch is the only CaM binding site in PfPKB. As expected PfPKB did not show binding to CaM in the absence of calcium (Fig. 2C, inset). CaM-CBD interaction was further validated by using a synthetic peptide corresponding to FIGURE 3. A pseudosubstrate motif is present in the CBD. A, ⌬PfPKB, an N terminus-deleted version of PfPKB lacking the first 98 residues that is active independently of CaM, was incubated in a kinase assay mixture with or without CBD peptide. Histone was used as the phosphate acceptor substrate. Inhibition of histone phosphorylation was accompanied by a simultaneous increase in CBD peptide phosphorylation. B, ⌬PfPKB was incubated in a kinase assay mixture, and 50 M CBD or CBD-S98A peptide was used as phosphoacceptor substrate. The phosphorylation of these peptides was measured as described for crosstide in Fig. 1. The average from two determinations done at the same time is shown, and error bars indicate S.E. C, 4 M His 6 -NTR (6xHis-NTR) was preincubated with buffer alone or with 4 M CaM and 100 M CaCl 2 before addition to the kinase assay mixture containing ⌬PfPKB. Kinase activity was determined as described above, and the activity in the absence of NTR was considered as 100%. D, 2.5 M His 6 -NTR was either preincubated with buffer alone (lane 1) or with 100 M CaCl 2 and 2.5 M CaM (lane 2) before addition of 2.5 M GST-⌬PfPKB. Glutathione-Sepharose beads were used to pull down the complex of GST-⌬PfPKB and His 6 -NTR. After washing, the proteins bound to the beads were analyzed by SDS-PAGE. Coomassie staining of the gels revealed the presence of NTR bound to PfPKB only in the absence of CaM (lane 1). In the presence of Ca 2ϩ /CaM, no NTR-PfPKB interaction was observed (lane 2). Lane 3 is the input His 6 -NTR used for the experiment. E, a model for PfPKB activation by Ca 2ϩ /CaM. PfPKB is locked in an inactive state as a pseudosubstrate region in the NTR occupies its catalytic cleft. Ca 2ϩ /CaM binding to CBD, which spans the pseudosubstrate motif, causes a conformational change resulting in the dissociation of NTR from the catalytic cleft thereby facilitating the autophosphorylation of Ser-271 of the activation loop. These events result in the catalytic activation of PfPKB. Phos-S271, phosphorylated Ser-271. the CBD sequence. CBD peptide exhibited CaM binding as it caused a mobility shift of CaM on a native PAGE gel in the presence of calcium. When calcium was excluded from the reaction mixture CBD did not interact with CaM as it failed to exhibit a shift in its electrophoretic mobility (Fig. 2D). CBD peptide prevented PfPKB activation in a dose-dependent manner when incubated with CaM ( Fig. 2E) indicating that this peptide competes with the CBD of PfPKB for CaM.
Data presented in Fig. 1 and our previous studies (1) indicate that the NTR is inhibitory for PfPKB in the absence of CaM binding. It is possible that the NTR either keeps PfPKB in an inactive state either by masking its catalytic cleft and/or by physically interacting with its active site. To investigate this, ⌬PfPKB, a deletion mutant of PfPKB that lacks most of the NTR and first nine residues of CBD and is constitutively active (1), was used. Incubation of ⌬PfPKB with CBD peptide inhibited its ability to phosphorylate histone. Interestingly this was accompanied by simultaneous phosphorylation of the CBD peptide (Fig. 3A). These observations indicated that the CBD peptide can also interact with PfPKB active site. A similar inhibition in phosphorylation of crosstide by ⌬PfPKB was observed (data not shown). Upon close examination of CBD sequence, an RXRXS type motif was found embedded in the CBD ( Fig. 2A) that closely resembles a putative substrate motif for AGC family kinases like PfPKB (17) and is not present in ⌬PfPKB. It is possible that this motif may act as a pseudosubstrate and thus have affinity for PfPKB active site. Replacement of the Ser (Ser-98 of PfPKB) in the RXRXS motif to Ala resulted in a loss of phospho-rylation of CBD peptide, indicating that this motif can indeed interact with PfPKB catalytic site (Fig. 3B). As expected, a control peptide with an amino acid composition similar to CBD but in scrambled order (scr-CBD) did not show any phosphorylation of scr-CBD (data not shown). Because there was no evidence of Ser-98 phosphorylation in intact PfPKB, we termed this as the "pseudosubstrate" motif.
Consistent with our previous observations (1), NTR when added to ⌬PfPKB inhibited its activity. Ca 2ϩ /CaM attenuated this inhibition significantly suggesting that NTR interaction with CaM could prevent PfPKB inhibition (Fig. 3C). Furthermore direct interaction between ⌬PfPKB and NTR could be demonstrated. When NTR was incubated with Ca 2ϩ /CaM, it did not interact with ⌬PfPKB suggesting that Ca 2ϩ /CaM can prevent binding of the NTR (Fig. 3D).
Based on the results presented in Figs. 1-3 we propose the following model for PfPKB regulation: PfPKB is most likely held in an inactive conformation by CBD/NTR as the RXRXS pseudosubstrate motif present in this domain occupies the catalytic cleft of the kinase. Ca 2ϩ /CaM binding to CBD induces a conformational change that causes the release of NTR from the catalytic site resulting in PfPKB autophosphorylation and its catalytic activation (Fig. 3E).

Regulation of PfPKB by CaM in P. falciparum-Whereas
PfPKB is specifically expressed in schizonts/merozoites (1), CaM is present in all intraerythrocytic stages (18). To determine whether PfPKB interacts with CaM in the parasite, PfPKB was immunoprecipitated from schizonts followed by Western blotting for CaM. CaM was co-immunoprecipitated with PfPKB indicating that these proteins associate in the parasite. In contrast, mock immunoprecipitation experiments performed with either preimmune antisera or red blood cell lysates did not show the presence of CaM (Fig. 4A). The ability of CaM to activate PfPKB in P. falciparum was tested by using W7, a specific CaM inhibitor, which has been used previously to demonstrate its role in erythrocyte invasion (19,20). Incubation of either schizonts or free merozoites (data not shown) with W7 resulted in a significant loss of PfPKB activity indicating that CaM is a PfPKB regulator in vivo. This was further established when addition of purified CaM to the PfPKB-IP from W7-treated parasites led to a significant recovery of PfPKB activity (Fig. 4B). Collectively these observations establish that CaM is a regulator of PfPKB in P. falciparum. Moreover when PfPKB was immunoprecipitated in the absence of calcium, a significant loss of CaM binding resulted (Fig. 4C), suggesting that calcium is necessary for CaM-PfPKB interaction in the parasite.
Phospholipase C-mediated Calcium Release Regulates CaM-PfPKB Interaction-Because experiments described above (Figs. 1 and 4C) suggest that the activation of PfPKB by CaM is dependent on calcium, it was important to investigate the mechanism via which parasitic calcium regulates PfPKB. Intracellular calcium levels of the parasite are tightly regulated in Plasmodium, and inhibitors of phospholipase C (PLC) (21), which block inositol 1,4,5-trisphosphate formation, prevent release of free calcium from intracellular parasite stores (22,23). To investigate the role of PLC in PfPKB activation, schizonts were incubated with either U73122, a specific inhibitor of PLC, or its less potent analogue U73322. U73122 treatment resulted in a significant attenuation of PfPKB activity. In contrast, U73322 only caused a marginal effect (Fig. 5A). The loss of PfPKB activity was accompanied by a reduction in amount of CaM associated with PfPKB in PLC inhibitor-treated parasites (Fig. 5B). Similar inhibition of PfPKB activity was obtained when merozoites were treated with PLC inhibitors (data not shown).
It was important to determine whether PLC-mediated regulation of PfPKB was due to its ability to control intracellular calcium levels. To probe this, parasites were treated with U73122 in the presence of ionomycin, a calcium ionophore that can mobilize calcium. Recovery of PfPKB activity, which was lost due to PLC inhibition (Fig. 6), was observed suggesting that PLC controls PfPKB activity by regulating calcium levels inside the parasite. A cell-permeable intracellular calcium chelator, BAPTA-AM, attenuated PfPKB activity, providing direct evidence that PfPKB is regulated by intracellular calcium (Fig. 6). It is important to indicate that P. falciparum has a PLC homologue that shares significant similarity with the catalytic domain of mammalian PLC. 4 In summary, we identified a novel signaling pathway in the malarial parasite that involves activa-tion of PfPKB by CaM, and PLC serves as an upstream activator of this pathway as it provides the release of calcium necessary for PfPKB activation.

DISCUSSION
Based on the sequence homology we had identified PfPKB as a PKB-like kinase in P. falciparum. The PI3K-PKB pathway is a major player in a wide variety of cellular processes in mammalian cells (24,25). Plasmodium possesses only one PI3K homologue 5 and PfPKB, a protein kinase B-like enzyme in P. falciparum (1). PfPKB shares several common features with the catalytic domain of PKB such as the regulatory motif in the T-loop that has a regulatory Ser-271, which is in a location similar to Thr-308 of PKB (1).
PKB is regulated by interaction of 3Ј-phosphorylated phosphoinositides with its N-terminal PH domain (26) and phosphorylation of PKB at Thr-308 by phosphoinositide-dependent kinase 1. In contrast, PfPKB is activated by autophosphorylation of Ser-271. The NTR of PfPKB does not have a PH or any other modular domain, but it prevents PfPKB activation (1). Because NTR failed to exhibit any similarity with proteins in the non-redundant data base, it was difficult to postulate its mode of regulation. Biochemical studies performed with PfPKB indicated that it is directly regulated by Ca 2ϩ /CaM. Whereas PfPKB is expressed mainly in the schizont/merozoite stages, our recent findings suggest that PfPI3K homologue in P. falciparum may be present mainly during trophozoite stages 6 ; this also fits in well with a phosphoinositide-independent mechanism of PfPKB regulation.
CaM interacts with segments on proteins that are rich in basic and hydrophobic residues and have the propensity of forming ␣-helices. Typically hydrophobic residues in CBDs of target proteins are critical for anchoring them to hydrophobic  pockets present in CaM. These residues may be separated by 10, 14, or 16 residues as shown in examples illustrated in Fig. 2A (16,27). The number of CBDs that do not strictly follow these rules has grown; moreover there are CBDs that use only one of these residues for interaction (27,28). These studies reflect the diversity in interaction of CaM with its targets. The CBD of PfPKB has a Leu-6 and Ile-20 that are spaced by 15 residues and may anchor binding to CaM; this is a minor diversion from standard examples shown in Fig. 2A.
Mapping of the CBD led to identification of an RXRXS pseudosubstrate motif in the NTR and facilitated the elucidation of the mechanism of PfPKB activation by CaM. These studies could potentially be utilized to devise tools for intervening with PfPKB activation.
Free calcium levels in the parasite are controlled by PLC as it generates inositol 1,4,5-trisphosphate, which releases calcium from the intracellular stores (21). PLC inhibitor U73122 has been successively used in Plasmodium to block calcium release (22). Using this inhibitor we were able to demonstrate that PLC is the upstream regulator of PfPKB as it mediates calcium release crucial for PfPKB activation. These data provide a direct link between PfPKB and PLC-mediated calcium signaling (Fig. 7).
In silico analyses suggest that Plasmodium may lack typical CaM-dependent protein kinases, and more strikingly a protein kinase C-like enzyme seems to be absent (4). Given its versatility in eukaryotic signaling, lack of a protein kinase C-like enzyme in Plasmodium was indeed surprising. Protein kinase C belongs to the same AGC class of kinases as cAMP-dependent protein kinase and PKB, which share significant homology in their catalytic domain region. For instance, PKB and protein kinase C have ϳ67% similarity in their catalytic domains. In addition to the similarity between PfPKB and PKB, it is important to note that PfPKB also shares reasonable homology (ϳ64%) with mammalian protein kinase C (1). We used this information to identify a protein kinase C inhibitor, Go6983, as an inhibitor of PfPKB. When added to schizont stage cultures, this inhibitor blocked ring formation suggesting that PfPKB may be involved in schizont-to-ring transition (1); no other parasitic stage was affected by this compound. Schizont/merozoite-specific expression (1) may allow PfPKB to play a role in early/late stages of parasite life cycle.
Release of intracellular calcium is critical for various parasitic functions (29). Importantly it appears to be indispensable for successful erythrocyte invasion (30,31). CaM is known to localize at strategic locations in merozoites and control invasion (19,20). However, the lack of identity of Ca 2ϩ /CaM targets leaves a gap in understanding their role in the parasite life cycle. We have shown that CaM interacts and regulates PfPKB in response to upstream events in the schizont. Therefore, it is possible that PfPKB may be one of the major targets via which CaM may control important parasitic processes like invasion. Identification of downstream targets of PfPKB will help further in unraveling the function of this pathway.