PfPKB, a novel protein kinase B-like enzyme from Plasmodium falciparum: I. Identification, characterization, and possible role in parasite development.

Extracellular signals control various important functions of a eukaryotic cell, which is often achieved by regulating a battery of protein kinases and phosphatases. Protein Kinase B (PKB) is an important member of the phosphatidylinositol 3-kinase-dependent signaling pathways in several eukaryotes, but the role of PKB in protozoan parasites is not known. We have identified a protein kinase B homologue in Plasmodium falciparum (PfPKB) that is expressed mainly in the schizonts and merozoites. Even though PfPKB shares high sequence homology with PKB catalytic domain, it lacks a pleckstrin homology domain typically found at the N terminus of the mammalian enzyme. Biochemical studies performed to understand the mechanism of PfPKB catalytic activation suggested (i) its activation is dependent on autophosphorylation of a serine residue (Ser-271) in its activation loop region and (ii) PfPKB has an unusual N-terminal region that was found to negatively regulate its catalytic activity. We also identified an inhibitor of PfPKB activity that also inhibits P. falciparum growth, suggesting that this enzyme may be important for the development of the parasite.

Plasmodium falciparum, a unicellular protozoan responsible for the most lethal form of human malaria, has re-emerged as a leading cause of mortality in developing countries, especially in the population of young children age 5 and under. Widespread drug resistance exacerbates the problem and limits our options for effective malaria control. Current efforts to produce effective vaccines have not yet resulted in any significant success. Identification of drugs that interfere with parasite development could be a useful way to inhibit parasite growth in humans. Detailed knowledge of molecular mechanisms that control the life cycle of malaria parasite could provide crucial information needed to achieve this goal.
The life cycle of the malaria parasite is a complex but a well synchronized series of events. After invasion of the erythro-cytes, the parasite can either propagate asexually or undergo sexual differentiation. The role of extracellular signals and molecular events in parasite life cycle are not well understood. It is well known that the fate of most eukaryotic cells is controlled by specific cell signaling pathways. Therefore, it is reasonable to assume that cell-signaling cascades may be very important for the development of Plasmodium. Systematic analysis of molecules involved in cell signaling events that occur during the course of P. falciparum development need to be pursued. A recently published genome sequence (1) and earlier studies suggest that several homologues of eukaryotic signaling proteins, such as protein kinases and phosphatases, are conserved in P. falciparum (2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13). The major challenge is to understand how these enzymes integrate in cellular machinery of Plasmodium and what roles they play in parasite development. The answer to these questions could lead to identification of important signaling pathways involved in the development of this parasite.
We are interested in understanding the role of phosphatidylinositol 3-kinase (PI3K) 1 and protein kinase B (PKB) in P. falciparum. The PI3K-PKB-mediated signaling is important for proliferation and survival of several eukaryotic cell types such as fibroblasts and neuronal cells (14). After PI3K activation, 3Ј-phosphorylated phosphoinositides (phosphatidylinositol 3,4-diphosphate or phosphatidylinositol 3,4,5-trisphosphate) are generated, and they bind to the PH domain present at the N terminus of PKB, resulting in its translocation to the cell surface and phosphorylation-dependent activation (15)(16)(17). This regulatory phosphorylation is carried out by phosphatidylinositol-dependent kinase 1 (PDK1), which also regulates other members of the AGC group of protein kinases like PKA and isoforms of PKC (18,19). However, maximal catalytic activity is achieved only upon phosphorylation of a C-terminal site. PKB 2 targets a wide variety of cellular targets ranging from transcription factors, anti/pro-apoptotic proteins, enzymes involved in glycogen metabolism, etc. (15)(16)(17). Function of neither PI3K nor PKB in Plasmodium or any other protozoan parasite is understood.
We report identification and biochemical characterization of a PKB homologue from P. falciparum, PfPKB. Although it shares significant homology with mammalian-PKB catalytic domain, PfPKB lacks a phosphoinositide interaction domain (PH domain). Biochemical studies suggested that PfPKB autophosphorylation is pivotal for its activation, unlike the mammalian enzyme, which needs phosphorylation by PDK1. PfPKB was expressed at highest levels during the schizont/merozoite stage of the parasite lifecycle. Studies using a pharmacological inhibitor suggested that PfPKB activity may be important for parasite development.
Molecular Cloning and Mutagenesis of PfPKB cDNA-Human PKB␤ sequence was used to BLAST search (in tBLASTN mode) the Plasmodium genome sequence at Sanger Center, The Institute of Genomic Research, and Stanford University (chromosome 12). A contig highly homologous to PKB was found on chromosome 12. The longest open reading frame (ORF) spanning this region was named PfPKB (described in detail under "Results"). PCR was performed using primers (sequences given below) based on the PfPKB ORF ( Fig. 1). Total RNA was isolated from asynchronous P. falciparum 3D7 cultures, and reverse transcription was performed using random hexamers provided in the Thermoscript reverse transcription-PCR kit (Invitrogen). Complimentary or genomic DNA was used as the template in PCR reactions, which were performed using platinum Taq polymerase (Invitrogen) with the following cycling parameters: 94°C for 2 min initial denaturation followed by 30 cycles at 94°C for 30 s, 42°C for 30 s, 72°C for 2 min and final extension at 72°C for 10 min. The following primer sets were used for cloning the full-length PfPKB gene (PfPKB_1 and Pf-PKB_3) and its catalytic domain, ⌬PfPKB (PfPKB_2 and PfPKB_3): forward (PfPKB_1), 5Ј-ATGATCATATACATGTACCATATCTATGCCC-C-3Ј; forward, (PfPKB_2), 5Ј-CGTAACTCTATGTCCTTATCATATGAA-AGGAAA-3Ј; reverse, (PfPKB_3), 5Ј-TCATTTTTGTTGACCTGATTTT-TCTCATAATAGTTG-3Ј.
Expression and Purification of Recombinant PfPKB and Its Variants-PfPKB or ⌬PfPKB was amplified using primers (described above) containing overhangs for BamHI and XhoI restriction enzymes to facilitate cloning into a GST fusion protein expression vector pGEX-4T1 (Amersham Biosciences). Recombinant expression vectors containing either the PfPKB gene or its variants were transformed in Escherichia coli BL21-RIL strain (Stratagene). Cultures were grown in LB media containing 100 g/ml ampicillin and 25 g/ml chloramphenicol. When cultures were in mid-logarithmic phase (A 600 value of 0.6), expression of proteins was induced by 1 mM isopropyl-1-thio-␤-D-galactopyranoside either at 37°C for 4 h or at 18°C for 14 h. Bacterial cells were harvested by centrifugation at 4000 ϫ g for 30 min suspended in Buffer A (50 mM Tris, pH 7.4, 2 mM EDTA, 1 mM dithiothreitol, 1% Triton X-100, and protease inhibitor mixture (Roche Applied Science)) followed by sonication for 8 cycles of 1 min each on ice. Cell lysates were clarified by centrifugation at 20,000 ϫ g for 30 min at 4°C. Subsequently, the supernatant was incubated with glutathione-Sepharose resin (Amersham Biosciences) with end-to-end shaking for 6 h at 4°C and was washed with Buffer A. The recombinant GST fusion proteins were eluted in 50 mM Tris, pH 8.0, containing 10 mM glutathione, dialyzed against 50 mM Tris, pH 7.5, 1 mM dithiothreitol, and 10% glycerol, concentrated using 10-kDa cutoff concentrator (Millipore), and analyzed by SDS-PAGE. Protein visualization and assessment were done on a gel documentation system. For expression of the N-terminal region (NTR) as a His 6 -tagged or GST fusion protein, the region corresponding to NTR with additional catalytic domain residues (1-126) of PfPKB was amplified using primers 5Ј-GAAGTGAATTCGATGATCATATACATG-TAC-3Ј and 5Ј-GGATTCCTCGAGTTTTCCATATGATCCTTC-3Ј, which possess overhangs for BamHI and XhoI, and was subsequently cloned in pET28 and pGEX4T1 vectors. BL21-RIL cells bearing recombinant plasmids were grown in LB medium containing either kanamycin (25 g/ml) or ampicillin (100 g/ml) and chloramphenicol (50 g/ml). Protein expression was induced by treatment of bacterial cultures with 1 mM isopropyl-1-thio-␤-D-galactopyranoside at 18°C for 12 h. Recombinant His 6 NTR protein was extracted in denaturating extraction buffer (50 mM Tris, pH 7.4, 300 mM NaCl, and 8 M urea), and the clarified lysate was incubated with nickel nitrilotriacetic acid-agarose (Qiagen) with end-to-end shaking for 3 h at room temperature. The resin was poured into a column, washed with extraction buffer, and eluted in extraction buffer containing 25-500 mM imidazole. Fractions containing the protein of interest were pooled and dialyzed against buffers containing decreasing amounts of urea (25 mM Tris, pH 7.4, 150 mM NaCl, 1 mM dithiothreitol, 0 -6 M urea) to refold the recombinant protein.
GST-NTR was purified as described above for other GST fusion proteins. Recombinant proteins were estimated and analyzed by densitometry.
Assay of PfPKB Activity-GST fusion proteins of PfPKB (its deletions or its mutants) or 10 l of PfPKB immunoprecipitated from parasite lysate (see below) were assayed for catalytic activity 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 histone H1, histone II AS , or a small peptide substrate of PKB, "crosstide" (22), as phosphate-acceptor substrates. Typically, reactions were carried out for 20 -60 min at 30°C and were terminated by boiling the samples in SDS-PAGE sample buffer. Reaction mixtures were electrophoresed on 15% SDS-PAGE gel (for histone II AS ) followed by autoradiography to visualize phosphate incorporation in histone II AS . In some experiments, GST-NTR (0.1 g) and His 6 -NTR (0.1 g) were preincubated with ⌬PfPKB (0.4 g) before the addition of substrate and ATP.
When peptide substrates were used, reactions were stopped by spotting the samples on P81 phosphocellulose paper followed by washing the paper strips with 75 mM orthophosphoric acid. Phosphate incorporation was assessed by scintillation counting of P81 paper. One unit of enzyme activity represents 1 pmol of phosphate transferred to 1 mg of substrate in 1 min. Data representative of at least three independent experiments are illustrated in all figures.
Generation of anti-PfPKB Serum, Immunoblotting, Immunoprecipitation, and Immunofluorescence-Polyclonal anti-PfPKB serum was raised in rabbits using a synthetic peptide designed from the C-terminal sequence (DFNYNYYEFSGQQK). This peptide was conjugated to keyhole limpet hemocyanin via an additional N terminus cysteine residue (Zymed Laboratories Inc., Inc.) Parasites were released from infected erythrocytes by 0.1% (w/v) saponin treatment. Cell-free protein extracts either from specific parasite stages or asynchronous cultures 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, and 1ϫ Complete Protease inhibitor mixture (Roche Applied Science) using a syringe and a needle. Lysates were cleared by centrifugation at 14,000 ϫ g for 30 min. After separation of lysate proteins on SDS-PAGE gels, proteins were transferred to a nitrocellulose membrane. Immunoblotting was performed using the above-described anti-PfPKB antisera and horseradish peroxidase-labeled anti-rabbit IgG. ECL substrate (Pierce) was used to develop the blots following the manufacturer's instructions.
PfPKB was immunoprecipitated from the schizont-enriched parasite lysate using anti-PfPKB antisera. Erythrocyte-free parasite lysates were prepared as described above except phosphatase inhibitors (20 M sodium fluoride, 20 M ␤-glycerophosphate, and 100 M sodium vanadate) were added to the lysis buffer. 100 g of lysate were incubated with the anti-sera at 4°C overnight on an end-to-end shaker. Subsequently, antigen-antibody complexes were incubated with 50 l of protein A-Sepharose beads for 2 h with end-to-end shaking. Beads were washed with phosphate-buffered saline several times and were finally resuspended in 50 l of 1ϫ kinase assay buffer containing phosphatase inhibitors.
Immunofluorescence microscopy was performed on thin blood smears of parasite cultures that were fixed with methanol or acetone. Smears were blocked in phosphate-buffered saline containing 0.05% saponin and 5% bovine serum albumin. Subsequently, incubations with anti-PfPKB and/or anti-MSP1 1-19 anti-sera (1:100 -1:200 dilution) were performed for 2 h at room temperature or at 4°C overnight. Anti-mouse IgG labeled with fluorescein isothiocyanate and anti-rabbit IgG labeled with Texas Red were used to localize merozoite surface protein 1 (MSP1) and PfPKB, respectively. Parasite nuclei were localized with Hoechst 33342. Stained parasites were visualized using either a NikonTE2000 or a Zeiss confocal microscope. Images were analyzed and merged using Image Proplus or Adobe Photoshop software.
Inhibition of PfPKB Activity and Parasite Growth by Go 6983-Recombinant GST-⌬PfPKB was preincubated for 15 min with varying concentrations of Go 6983 or Go 6976 (Biomol or Calbiochem) prepared in 0.1% Me 2 SO or with 0.1% Me 2 SO. The enzyme activity was assayed as described above. To monitor the effect of Go 6983 on P. falciparum growth, parasites were plated at 1% parasitemia containing mostly rings and were treated either with the indicated concentrations of Go 6983 or 0.1% Me 2 SO. At the indicated time points, thin films were prepared and stained with Giemsa reagent, and the number of intraerythrocytic parasites was counted by using microscopic examination. Parasites were counted from several different fields and also different experiments. The data (Fig. 8) are represented as the ratio of percent of parasite-infected erythrocytes in drug-treated cultures to the ones in the control cultures, which were treated with 0.1% Me 2 SO.

Identification and Molecular
Cloning of the PfPKB Gene-To identify a PKB homologue in P. falciparum, human PKB␤ sequence was used to search the P. falciparum genome sequence in silico. A match on chromosome 12 sequence data base with high sequence homology to the catalytic domain of human PKB (71%) was found. A long and uninterrupted 1.34-kilobase (encoding a polypeptide of 446 amino acids) ORF encompassing this region was designated as PfPKB. PfPKB belongs to the AGC kinase family, which possesses a motif with consensus sequence (S/T)(F/L)CGTP(D/E)Y in its activation loop region (Fig. 1, A and B). Phosphorylation of S/T (at the first position in this sequence) is pivotal for their activation (19). In addition, PfPKB has a hydrophobic patch of amino acids at its C-terminal end (Fig. 1) that is also present at the C-terminal end of several AGC kinases, although its sequence is only loosely conserved (23). While these studies were in progress, annotation of the Plasmodium genome appeared on PlasmoDB (24,25). BLAST searches of these annotated genes resulted in three major "hits" (Tables I and II). All three protein sequences were related to AGC class kinases. The gene with highest sequence homology was closely related to PfPKB. The other two kinase sequences with around 56% homology to the catalytic domain of PKB have previously been identified as the cAMP-dependent protein kinase (13) and the cGMP-dependent kinase (5) homologues from P. falciparum.
PfPKB ORF was amplified using either cDNA or genomic DNA as template, and the amplified PCR products appeared to be identical in size (data not shown). This suggested that Pf-PKB is at least 1.34 kb in size and introns are absent from this contig, which was confirmed by DNA sequencing. The deduced PfPKB sequence (Fig. 1A) shows that it has all kinase subdomains (26) and has a hydrophobic motif at the C terminus present in several AGC class kinases (23). It also has a putative signal sequence at its N-terminal end. Interestingly, PfPKB shares high sequence homology with catalytic domains of PKC isoforms ␣, ␤, ␥, (ϳ64% , Tables I and II). It should be noted that a PKC homologue has not yet been identified in the Plasmodium genome. Our in silico analysis suggested that PfPKB may be the closest relative of PKC in P. falciparum. Strikingly, PfPKB does not contain a PH domain usually found in PKB of higher eukaryotes. This domain is important for binding to phosphoinositides phosphorylated at 3Ј-OH (15)(16)(17). Because yeast and Trypanosoma cruzi homologues of PKB also lack this domain (27)(28)(29), it is possible that PKB in these organisms is regulated via different mechanisms other than interactions with 3Ј-phosphoinositides. During the process of screening for full-length PfPKB, we found clones containing intronic sequences that could result in different-size ORF. This could be a result of alternative splicing or defects in transcripts.
Recombinant PfPKB Catalytic Domain (⌬PfPKB) Is Active-To get insight into the catalytic mechanism of PfPKB, the catalytic domain of this enzyme with the C-terminal extension (⌬PfPKB, indicated in Fig. 1A) was expressed as a GST fusion protein in E. coli ( Fig. 2A). Catalytic activity of recombinant ⌬PfPKB was assayed using histone II AS and histone H1 as substrates of this enzyme. GST-⌬PfPKB phosphorylated histone II AS (Fig. 2B) and H1 (data not shown). It has been observed that AGC protein kinases (PKA, PKC, PKB, PKG) prefer basic residues in the vicinity of the target phosphorylation sites (30). Crosstide, a peptide with sequence GRPRTSS-FAEG that has been shown to be a good substrate for PKB (22), was phosphorylated by ⌬PfPKB (K m ϳ 20 M) (Fig. 2C). ⌬Pf-PKB also phosphorylated a myelin basic protein-derived peptide (QKRPSQRSKYL) but with 2-fold less efficiency (data not shown here). Because the difference in these peptides was in the location of the basic residues corresponding to the phosphorylatable serine residues, it reflects the preference of PfPKB for basic residues at appropriate position in its substrates.
PfPKB May Be Regulated by Autophosphorylation of Ser-271 and Additional Phosphorylation of Ser-442-PKB is catalyti-cally activated by phosphorylation of activation loop at Thr-309 by PDK1 (18,23,31), and deletion of the N-terminal PH domain of PKB results neither in its catalytic activation nor in autophosphorylation; it requires phosphorylation by PDK1 to be active (32). In contrast, catalytic activity of recombinant ⌬PfPKB was observed in the absence of any exogenous kinase (Fig. 2, B and C). Instead, it exhibited autophosphorylation (Fig. 2D, lane 1). These data suggest that autophosphorylation may be responsible for regulating PfPKB catalytic activity.
Thr-309 phosphorylation in PKB results in conformational changes crucial for its catalysis (15,23,31). Because Ser-271 in PfPKB is complementary to Thr-309, it may be in a similar strategic location in the kinase catalytic core (Fig. 1B) and may play a role in PfPKB activation. To test this, Ser-271 of PfPKB was replaced by alanine, and the effect on its catalytic activity was assessed. Mutation of this serine to alanine resulted in almost a complete loss of ⌬PfPKB activity as the S271A mutant failed to phosphorylate either histone (Fig. 3A) or crosstide (Fig. 3B). Moreover, this mutation resulted in a concomitant loss in autophosphorylation of ⌬PfPKB (Fig. 3C). It has been observed that replacement of Ser/Thr residues by negatively charged aspartate (Asp) or glutamate (Glu) could mimic their phosphorylated state. Replacement of S271A to S271D resulted in a significant recovery of ⌬PfPKB activity, which was lost due to the S271A mutation (Fig. 3, A and B). Collectively, these data suggest that autophosphorylation of Ser-271 is a prerequisite for PfPKB activation. A ⌬PfPKB mutant defective in ATP binding was generated by replacing a crucial lysine residue, which is conserved in all protein kinases, to methionine. It neither exhibited detectable autophosphorylation nor kinase activity, confirming that PfPKB autophosphorylation is responsible for its activity (data not shown).
PKB and other mammalian AGC family protein kinases share several other functional similarities such as dependence on additional phosphorylation of their C-terminal end to achieve maximal activity (17). This phosphorylation site (Ser-474 in PKB) is usually part of a loosely conserved hydrophobic motif with the sequence FXF(S/T)Y (23). PKA has a negatively charged residue (Asp or Glu) at this position, which complements the function of phosphorylated Ser or Thr (23). PfPKB contains a hydrophobic motif (YYEFSG) at its C terminus (Fig.  1). Ser-442 in this region was replaced by aspartic acid in addition to the above-described S271D mutation to mimic phosphorylation at this position. The double mutant (S271D, S442D) exhibited catalytic activity even higher than that of single mutant, S271D (Fig. 4). These data suggest that acidic environment provided by phosphorylation at this site may be important for maximal activation of PfPKB, an observation consistent with mammalian PKB.
NTR of PfPKB Negatively Regulates Its Activity-A BLAST search using the NTR amino acid sequence did not show any significant sequence homology with proteins in the nonredundant protein data base. To study the effect of N-terminal region (NTR) on the activity of PfPKB, full-length PfPKB (containing both the N-terminal region and the catalytic domain) was expressed, and its catalytic activity was compared with that of ⌬PfPKB. PfPKB exhibited only marginal catalytic activity in comparison to its NTR deleted version, ⌬PfPKB (Fig. 5, A and  B). In addition, PfPKB also lacked detectable autophosphorylation activity (Fig. 5C). Collectively, these observations suggest that NTR may modulate PfPKB activity by preventing its autophosphorylation, thus preventing its catalytic activation. To determine if NTR directly causes PfPKB inhibition, it was expressed as a GST fusion or a His 6 -tagged protein. When recombinant NTR was incubated in a kinase assay mix with ⌬PfPKB, its activity (Fig. 5D) as well as its autophosphoryla- a % Homology represents the percent similarity in amino acids that are identical and belong to a similar chemical class. It is indicated as % similarity in the output of the BLAST program. tion (Fig. 5E) was inhibited dramatically. Interaction of NTR with the catalytic region of PfPKB may either cause conformational changes unfavorable for catalysis and/or it may restrict entry of the peptide substrate in the catalytic cleft of the enzyme, resulting in its inability to perform efficient catalysis.
Because the data illustrated in Fig. 3 indicate that phosphorylation of Ser-271 (or its mutation to D) activates ⌬PfPKB, it was worth testing the effect of S271D mutation on activation of full-length PfPKB. S271D mutant exhibited catalytic activity higher than that of PfPKB (Fig. 5F), suggesting that conformational changes due to phosphorylation-mimicking mutation of Ser-271 to Asp could result in an increase in activity of fulllength PfPKB. Collectively, the results described above indicate that NTR may prevent the activation of PfPKB by inhibiting its ability to autophosphorylate.
PfPKB Is Expressed in Schizonts and Merozoites-Western blotting, immunofluorescence, and reverse transcription-PCR techniques were used to examine the expression pattern of PfPKB during P. falciparum blood stage development. For this purpose, polyclonal antisera were raised against a synthetic FIG. 2. Expression of PfPKB catalytic domain and its catalytic activity. A, purified GST-⌬PFPKB, the recombinant PfPKB catalytic domain expressed as a GST fusion protein in E. coli, was confirmed by Western blot using anti-GST antibody. B, GST-⌬PfPKB phosphorylated histone II AS . C, a small peptide, crosstide. D, GST-⌬PfPKB was incubated in a kinase assay mix without any substrate to assess its autophosphorylation. Phosphate incorporation was monitored by autoradiography (B and D) or by scintillation counting (C). peptide derived from the C terminus of PfPKB (Fig. 6A). PfPKB expression was predominantly observed in the schizont stages of the parasite as indicated by a band of ϳ52 kDa, which is consistent with the predicted molecular mass of 50 kDa (Fig.  6B). Reverse transcription-PCR analysis indicated that PfPKB transcripts were mainly present in schizont stages and were observed in the trophozoites at only very low levels (Fig. 6C). PfPKB immunoprecipitated from schizont lysates was able to phosphorylate crosstide, indicating PfPKB is active in the schizonts of P. falciparum (Fig. 6D).
Immunofluorescence studies revealed that PfPKB was present mainly in the mid-late schizont stages of the parasite; other mono-nucleated stages did not show any detectable PfPKB expression (Fig. 6E). To probe if PfPKB is localized at the cell surface, co-localization studies were performed using antisera against MSP1, a marker for merozoite surface. PfPKB exhibited a diffused staining pattern in segmented schizonts, indicating its presence in the cytoplasm, and also exhibited some co-localization with MSP1 (Fig. 6F). In free merozoites PfPKB seems to be localized at the apical end (Fig. 6G), a region that is important for erythrocyte invasion. PfPKB also co-localizes with EBA175, a micronemal protein (data not shown).
Go 6983 Is an Inhibitor of PfPKB Activity and Inhibits Parasite Growth-We were interested in identifying PfPKB inhib-itors to help us understand the structure-function relationship of this Plasmodium protein kinase and, most importantly, its physiological role in promoting parasite growth and development in the host. Because there are no known or available inhibitors of mammalian PKB, we considered the possibility of inhibitors of other related kinases as putative candidates for PfPKB. For this purpose PKC inhibitors were selected based on following reasons. 1) In silico analysis suggested that catalytic domain of PfPKB was most closely related to PKC in comparison to other AGC kinases (Table II); 2) it appears that P. falciparum does not have a PKC homologue. Go 6983 and Go 6976 (Fig. 7A) are isoform-specific PKC inhibitors that target the ATP binding site (33). Inhibition of any other AGC kinases by these compounds has not yet been reported. The effect of these compounds on in vitro recombinant PfPKB activity was tested. Go 6983 inhibited ⌬PfPKB activity, as judged by its ability to phosphorylate histone II AS (Fig. 7B) and crosstide (Fig. 7C), with an IC 50 of ϳ1 M. In contrast, Go 6976 was unable to inhibit the ⌬PfPKB activity even at 10 M concentration (Fig. 7D). These data suggested that Go 6983 could be a useful tool for manipulating PfPKB activity in P. falciparum for deciphering its role in the life cycle of the parasite. To test this, synchronized P. falciparum cultures were incubated with Go 6983, and parasite growth was monitored at different time FIG. 5. N-terminal region of PfPKB hinders its catalytic activity. PfPKB or its catalytic domain (⌬PfPKB) was assayed for kinase activity using histone II AS (A) or crosstide (B) as substrates. In some experiments, exogenous substrates were not included, and autophosphorylation levels of GST-PfPKB (lane 1) or GST-⌬PfPKB (lane 2) were assessed by autoradiography. C, arrows indicate the mobility of the GST-PfPKB and GST-⌬PfPKB proteins. Recombinant His 6 -NTR or GST-NTR was incubated with ⌬PfPKB, and its activity was monitored by its ability to phosphorylate crosstide (D), or its autophosphorylation was monitored by performing incubations in absence of any substrate (E). Catalytic activity of PfPKB was compared with its S271D mutant by performing kinase assays as described above (F).
points. There were no apparent effects of this drug until parasites reached the late schizont/segmenter stage (40 -44 h). A significant decrease in parasitemia was observed in drugtreated cells subsequent to this point. In Go 6983-treated cells, the number of rings in the following cycle was markedly less compared with the control cultures (Fig. 8A). This observation correlates well with the expression profile of PfPKB in the schizont/merozoite stages of the life cycle (Fig. 6). It is reasonable to state that the major effects of Go 6983 were due to its interaction with Plasmodium cellular machinery and not a  2). B, equal amounts of cell lysate prepared from either uninfected erythrocytes (E), schizonts (S), trophozoites (T), or rings (R) were used for Western blot analysis with anti-PfPKB antisera. C, equal amounts of RNA isolated from either schizont (S) or trophozoite (T) stages of P. falciparum were used for reverse transcription-PCR analysis using primers specific to either P. falciparum elongation factor 1␤ (PfEF1␤), used as a control (12), or PfPKB. Equal amounts of PCR product were electrophoresed on 1% agarose gel. D, PfPKB immunoprecipitate (IP) from schizont-rich parasite lysates was used to phosphorylate crosstide. E, immunofluorescence studies performed using anti-PfPKB antisera on thin blood smears of P. falciparum cultures revealed that only late schizonts but no other asexual stages (Hoechst-stained) were recognized by this anti-sera. Rabbit antisera against PfPKB (red) and mouse antisera against MSP1 (green) was used to localize these proteins in late/segmented schizonts (F) or in free merozoites (G). The right panel in G shows a zoomed image of merozoites stained with MSP1 and PfPKB antisera. result of effects on erythrocytes, since other parasite stages and erythrocytes appeared to have normal morphology.
To further establish the effect of this inhibitor, P. falciparum schizonts were incubated with Go 6983, and development of the parasite was monitored. Although untreated cells formed rings within 4 -6 h, Go 6983 treatment resulted in an almost 60% decrease in formation of new rings (Fig. 8B). These data suggest that this drug targets directly the late schizont stages or the merozoites, which would prevent invasion of erythrocytes and subsequent formation of fresh rings. DISCUSSION Role of signal transduction events in the development of most protozoan parasites is unclear. Especially, information about molecular machinery involved in carrying out these events is lacking. One of our interests is to understand the role of the classical PI3K, PKB pathways in P. falciparum. We have identified a PI3K homologue 3 and a PKB homologue (reported here) in P. falciparum that suggests that signaling pathways regulated by PI3K and PKB are likely to be conserved in this parasite, although they may not necessarily operate in a manner similar to higher eukaryotes (see the discussion below).
PfPKB shows significant sequence homology to the catalytic domain of protein kinase B in higher eukaryotes. In addition it also shares almost 65% sequence homology with the catalytic domain of protein kinase C, another AGC class Ser/Thr kinase. So far, a PKC homologue has not been identified in P. falciparum. Our analysis suggested that PfPKB is closest to, besides PKB, mammalian PKC␣, ␤, ␥ at the sequence level. PfPKB does not have a PH domain, which plays an important role in subcellular localization and activation of PKB by binding to 3 P. Sharma, manuscript in preparation. 3Ј-phosphorylated phosphoinositides, a product of PI3K in mammalian cells (16,17). It is interesting to note that yeast and T. cruzi homologues also do not have a PH domain (27)(28)(29), suggesting that PKB in these unicellular eukaryotes may not be directly regulated by phosphoinositides. The N-terminal region of PfPKB does not share any significant homology to proteins in various databases. It inhibits PfPKB catalytic activation by preventing its autophosphorylation. Often, regulatory domains are found at the N or C terminus of protein kinases, and binding of effector molecules (e.g. calmodulin interaction with calmodulin-dependent kinases, diacylglycerol, and calcium; in the case of PKC, interaction of phosphoinositides with PKB) to these domains brings about conformational changes that result in modulation of their catalytic activity. There is a possibility that interaction with effector molecules may result in PfPKB activation. Identification of these effectors could provide important clues about its cellular regulation.
In contrast to mammalian PKB that is activated by PDK1, PfPKB appears to be activated by autophosphorylation of Ser-271. This serine residue is complementary to Thr-309 of PKB, and phosphorylation of this site by PDK1 is crucial for PKB activation. Despite sequence conservation in the vicinity of Ser-271, human PDK1 was unable to activate PfPKB (data not shown). Moreover, a PDK1-like enzyme is absent in Plasmodium, suggesting that autophosphorylation is the likely mode of activation of PfPKB. Several AGC kinases (e.g. conventional PKC isoforms) are regulated by autophosphorylation and do not appear to require PDK1 (34). It is interesting to note that PKB homologue from another protozoan parasite, T. cruzi, is also regulated by autophosphorylation (27).
It has been shown that the N-terminal PH domain-deleted versions of PKB do not exhibit any significant catalytic activity or autophosphorylation and its activation is dependent on PDK1-mediated phosphorylation (32). Several AGC kinases possess a loosely conserved hydrophobic patch of amino acids at their C-terminal end. A recent crystal structure of PKB revealed that phosphorylation of Ser-474 in this region promotes conformational stability of the N-lobe of the kinase, resulting in a conformation favorable for catalysis (23,31). Enhancement of PfPKB activity upon mutation of Ser-442 to Asp suggests that PfPKB activity may be enhanced by a similar mechanism.
To decipher the role and importance of PfPKB to P. falciparum, we considered the possibility of using pharmacological inhibitors against PfPKB. One of the major drawbacks in performing these studies was the unavailability of effective mammalian PKB inhibitors. As an alternative, we considered the possibility of inhibitors of other kinases for this purpose. Because PfPKB shares a high homology with protein kinase C isoforms, it was reasonable to use PKC inhibitors. Specific PKC inhibitors Go 6983 and Go 6976 were used to inhibit PfPKB activity. These inhibitors target the ATP binding region on PKC (29). Even though the ATP binding site is well conserved in protein kinases (35), it has been shown that there are subtle differences near this site that can be exploited to design effective and specific inhibitors against these enzymes (36). Several Plasmodium kinases have been targeted using inhibitors of their mammalian counterparts (4,13,37,38), and in some cases this information has been used to generate inhibitors specific to the Plasmodium enzymes (39 -41). Differences in Go 6983 and Go 6976 allow them to distinguish between isoforms of PKC (33); Go 6976 is more effective for conventional PKC isoforms and PKC than Go 6983, which is not very effective against PKC (33). In case of PfPKB, only Go 6983 and not Go 6976 inhibited its activity. This inhibitor selectivity could be useful as a template for designing more specific and effective inhibitors against PfPKB. Go 6983 inhibited parasite growth effectively, and its effects were observed mainly during or after schizont stages, which corroborates well with the specific expression of PfPKB in these stages. The arrest in parasite growth could be due to the role of PfPKB in late schizont stage development or during the invasion of erythrocytes by merozoites. PfPKB gene disruption studies, which are currently being pursued, may provide useful information about the role of this enzyme in P. falciparum life cycle. The identity of the upstream signals/molecular events that regulate PfPKB activity and its downstream targets will be crucial for understanding its cellular function.