Identification, characterization, and inhibition of Plasmodium falciparum beta-hydroxyacyl-acyl carrier protein dehydratase (FabZ).

The emergence of drug-resistant forms of Plasmodium falciparum emphasizes the need to develop new antimalarials. In this context, the fatty acid biosynthesis (FAS) pathway of the malarial parasite has recently received a lot of attention. Due to differences in the fatty acid biosynthesis systems of Plasmodium and man, this pathway is a good target for the development of new and selective therapeutic drugs directed against malaria. In continuation of these efforts we report cloning and overexpression of P. falciparum beta-hydroxyacyl-acyl carrier protein (ACP) dehydratase (PffabZ) gene that codes for a 17-kDa protein. The enzyme catalyzes the dehydration of beta-hydroxyacyl-ACP to trans-2-acyl-ACP, the third step in the elongation phase of the FAS cycle. It has a Km of 199 microM and kcat/Km of 80.4 m-1 s-1 for the substrate analog beta-hydroxybutyryl-CoA but utilizes crotonoyl-CoA, the product of the reaction, more efficiently (Km = 86 microM, kcat/Km = 220 m-1 s-1). More importantly, we also identify inhibitors (NAS-91 and NAS-21) for the enzyme. Both the inhibitors prevented the binding of crotonoyl-CoA to PfFabZ in a competitive fashion. Indeed these inhibitors compromised the growth of P. falciparum in cultures and inhibited the parasite fatty acid synthesis pathway both in cell-free extracts as well as in situ. We modeled the structure of PfFabZ using Escherichia coli beta-hydroxydecanoyl thioester dehydratase (EcFabA) as a template. We also modeled the inhibitor complexes of PfFabZ to elucidate the mode of binding of these compounds to FabZ. The discovery of the inhibitors of FabZ, reported for the first time against any member of this family of enzymes, essential to the type II FAS pathway opens up new avenues for treating a number of infectious diseases including malaria.

Malaria continues to exact the highest mortality and morbidity rate next only to tuberculosis. "The scourge of the tropics," malaria is endemic to around 100 countries in the world.
Approximately 500 million cases of malaria are reported every year, and around 3000 children die of malaria every day (1). Emerging resistance to chloroquine and other currently prescribed drugs limits treatment of malaria today, in particular cerebral malaria, caused by Plasmodium falciparum (2,3). The situation definitely warrants express remedial actions: extensive research on P. falciparum to identify drug targets and, ultimately, the development of a new armamentarium of antimalarials.
Our recent demonstration of the occurrence of the type II fatty acid synthesis (FAS) 1 pathway in the malaria parasite and its inhibition by triclosan, an inhibitor of the rate-limiting enzyme of type II FAS, enoyl-acyl carrier protein (ACP) reductase, proved the pivotal role played by this pathway in the survival of the malarial parasite (4,5). The essential role of fatty acids and lipids in cell growth and differentiation and the different type (type I) of fatty acid biosynthetic pathway occurring in the human host, which is distinct from type II FAS of the malaria parasite, makes this pathway an attractive drug target for treating malaria (6,7).
The type II fatty acid biosynthesis pathway, found in most bacteria and plants, is typified by the existence of distinct enzymes encoded by unique genes for catalyzing each of the four individual chemical reactions required to complete successive cycles of fatty acid elongation (Refs. 4, 8, and 9 and Scheme I). This is in contrast to the type I FAS characterized by a multifunctional enzyme catalyzing all the steps of the pathway (10). The third step of the elongation cycle, the dehydration of ␤-hydroxyacyl-ACP to trans-2-acyl-ACP, is carried out by FabA or FabZ (11). While FabZ only catalyzes the dehydration reaction, FabA is a bifunctional enzyme and carries out an additional isomerization reaction of trans-2-acyl-ACP to cis-3-acyl-ACP, a reaction essential to unsaturated fatty acid synthesis (12). In Streptococcus pneumoniae, instead of FabA, a new enzyme, FabM (trans-2,cis-3-decenoyl-ACP isomerase), catalyzes the isomerization step of trans-2-decenoyl-ACP to cis-3decenoyl-ACP following the dehydration step by FabZ (13). FabZ is the primary dehydratase that participates in the elongation cycles of saturated as well as unsaturated fatty acid biosynthesis, whereas FabA is more active in the dehydration of ␤-hydroxydecanoyl-ACP (11,14). This dichotomy allows the synthesis of unsaturated as well as saturated fatty acids, which eventually dictate the composition of the cell membranes. Dehydratase activity is crucial for the supply of trans-2-acyl-ACP to FabI, which pulls each cycle of elongation to completion (11,15). FabZ thus presents itself as a suitable, yet unexplored target for the design of antimalarials.
Here we report the cloning, expression, and characterization of P. falciparum FabZ (PfFabZ), the gene sequence of which has been deposited earlier by us in a public data base (GenBank™ accession number AY118082). FabZ was expressed as a soluble, active protein, and its purification was achieved by a single step purification protocol. We report its molecular and enzymatic properties. Further we have identified two lead compounds with inhibitory activity toward P. falciparum FabZ, which represents a significant advance in this area as no inhibitors of FabZ from any source are known to date.

Strains and Plasmids
Escherichia coli DH5␣ cells were used during the cloning of the gene. pET-28a(ϩ) vector (Novagen) and BL21(DE3) cells (Novagen) were used for the expression of FabZ.

Cloning of PfFabZ in E. coli
Total RNA was isolated from 10 ml of packed erythrocytes (infected with P. falciparum, 10-12% parasitemia) after saponin lysis by a single step method of RNA isolation (16). The isolated RNA was treated with RQ1 RNase-free DNase (Promega, 1 unit/g of RNA) for 45 min at 37°C and repurified by phenol-chloroform extraction and ethanol precipitation. RT-PCR was performed using a one step RT-PCR kit (Qiagen, Valencia, CA). PCR was performed with the primers (forward, 5Ј-GGAATTCCATATGAATTTAACCTTTCCTAATTATG-3Ј, and reverse, 5Ј-CGGGATCCTTATTTCGATAAGGCAAACGTCATTTC-3Ј with NdeI and BamHI sites underlined, respectively). PCR conditions used were: 1 ϫ (94°C 5 min), 30 ϫ (94°C 1 min, 50°C 1 min, 72°C 1 min), 1 ϫ (72°C 10 min). The primers were designed using the GenBank TM accession number AF237572 to clone the mature protein (without the leader peptide and transit sequence, required for targeting of the protein to apicoplast). The 465-bp RT-PCR product was excised from a 1.2% agarose gel, purified using silica gel particles (QIAEX II gel extraction system, Qiagen), and cloned in pGEMT-Easy vector (Promega). Candidate plasmids containing the correct sized inserts were confirmed by digestion and dideoxy sequencing on an ABI Prism 377 semiadaptive sequencer Version 3.0.

Expression of PfFabZ
The insert present in pGEMT-Easy was reamplified using the abovementioned primers and subcloned in pET-28a(ϩ) vector (Novagen) inframe with the N terminus His 6 tag. The constructs were transformed into E. coli BL21(DE3) cells (Novagen), and cultures were grown at 37°C in Luria Broth (Hi-media) until A 600 of 0.6. These were induced with 1 mM isopropyl-␤-D-thiogalactopyranoside and further incubated at 12°C for 12 h. Cells were harvested at 8000 rpm for 10 min, and the resultant pellet was stored at Ϫ70°C until further use.

Purification of PfFabZ
The pellet was resuspended in lysis buffer containing 20 mM Tris-Cl (pH 7.5), 0.5 M NaCl, and 5 mM imidazole. Cells were disrupted using a probe-type ultrasonicator (Vibra-Cell, Sonics and Materials). Cell debris were removed by centrifugation at 15,000 rpm for 30 min. The supernatant obtained was applied to a Ni-NTA metal affinity column (His-bind resin, Novagen) equilibrated with the lysis buffer. The column was initially washed with lysis buffer and subsequently washed with the same buffer containing 60 mM imidazole. The protein was eluted using a step gradient of 0.3-0.5 M imidazole, and fractions were tested for purity by SDS-PAGE. The protein was applied on a fast protein liquid chromatography desalting column to remove imidazole followed by concentration of the protein using Centriprep-10. Protein concentration was determined from A 280 , assuming the molar extinction coefficient E 280 ϭ 9530 M Ϫ1 cm Ϫ1 , which was calculated using the formula from Ref. 17.
We further truncated the protein by removing 10 residues from the N terminus by using primers (forward, 5Ј-ACGTCCATGGTGCATCATC-ATCATCATCATATTGATATAGAAGATATTAAGAAAATTCTTCCAC-ATAGATATCCTTTCC-3Ј, and reverse, 5Ј-CGGGATCCTTATTTCGAT-AAGGCAAACGTCATTTC-3Ј with NcoI and BamHI sites underlined, respectively), and the PCR conditions used were as described above. The truncated gene was ligated into pET-28a(ϩ) vector. The truncation led to an increase in the protein yield by 3-fold, i.e. from 5 to 15 mg/liter.

Gel Filtration and Dynamic Light Scattering Studies of PfFabZ
Purified PfFabZ (2 mg/ml) was injected onto a Superdex TM 200 HR 10-ϫ 300-mm column (Amersham Biosciences) equilibrated in 20 mM Tris, 500 mM NaCl (pH 7.5) connected to an Ä KTA TM design system. The column flow rate was maintained at 0.3 ml/min. The molecular weight of PfFabZ was determined by plotting V e /V 0 versus elution volume for standard proteins. V e corresponds to the peak elution volume of the standard protein, and V 0 represents the void volume of the column determined using blue dextran (M r Ͻ 2,000,000).
Dynamic light scattering studies were performed on a Brookhaven Instruments Dynamic Light Scattering setup that can measure sizes from 2 to 4000 nm. The sample of PfFabZ (1 mg/ml) in 20 mM Tris, 500 mM NaCl (pH 7.5) was centrifuged at 14,000 rpm for 15 min and filtered through a 0.2-filter. The data acquisition time was 3 min. The routines used to fit the data points were cumulants, and non-negative least squares analysis was used to obtain the radius of gyration of PfFabZ.

Spectrophotometric Assay for Determining Enzyme Activity
All experiments were carried out on a Jasco V-530 UV-visible spectrophotometer. FabZ was assayed at 25°C by monitoring the forward as well as the reverse reaction, i.e. increase and decrease in A 260 , respectively, due to the conversion of ␤-hydroxybutyryl-CoA to crotonoyl-CoA and vice versa. The standard reaction mixture in a total volume of 100 SCHEME I. Fatty acid synthesis cycle occurring in E. coli. The condensation of acetyl-CoA and malonyl-ACP in the first reaction of the elongation phase of fatty acid synthesis is catalyzed by ␤-ketoacyl-ACP synthase III (FabH). The reduction of the resulting ␤-ketoester by ␤-ketoacyl-ACP reductase (FabG) forms ␤-hydroxyacyl-ACP, which serves as a substrate for FabZ. The dehydration by FabZ generates ␣,␤-unsaturated acyl-ACP that is a substrate of enoyl-ACP reductase (FabI), the product of which reenters the cycle by condensation with malonyl-ACP catalyzed by ␤-ketoacyl-ACP synthase I/II (FabB/F). For the synthesis of unsaturated fatty acids, the dehydration and subsequent isomerization of ␤-hydroxyacyl-ACP by ␤-hydroxydecanoyl thioester dehydratase (FabA) generates cis-3 unsaturated product, which can again enter the cycle upon condensing with malonyl-ACP by the action of FabB. ACP, acyl carrier protein; CoA, coenzyme A.
l contained 20 mM Tris-HCl buffer, pH 7.5, 500 mM NaCl, 8 g of FabZ, and 100 M ␤-hydroxybutyryl-CoA or crotonoyl-CoA. The K m for each substrate analog was determined by varying its concentration. The kinetic parameters were obtained by fitting initial velocity data to the Michaelis-Menten equation by non-linear regression analysis using Sig-maPlot 2000 software. The forward (k forward ) and reverse (k reverse ) rate constants were determined by using the following first order rate equa-

HPLC Assay
The reaction was performed as described above and stopped by the addition of 3% chloroform, and the compounds were separated by reverse phase HPLC as described earlier (18). Briefly, the mixtures were loaded onto a Sephasil Peptide C18 column (4.6 mm ϫ 25 cm; particle diameter, 5 m; column volume, 4.155 ml; Amersham Biosciences), and the compounds were separated on a gradient of 220 mM phosphate buffer, pH 4.0 and methanol:chloroform (98:2). The retention times of ␤-hydroxybutyryl-CoA, crotonoyl-CoA, butyryl-CoA, and NADH were 14. , and E. coli (E.coli, accession number P21774). The conserved histidine (His-133) and glutamate (Glu-147) active site residues are indicated by *. Conserved residues are shown in bold with black background, negatively charged residues are indicated by Ϫ, positively charged residues by ϩ, aliphatic residues by l, aromatic by a, tiny by t, small by s, big by b, charged by c, polar by p, and hydrophobic by h (all below the alignment). The first arrow denotes the position from where the mature protein starts, and the second arrow denotes the position from where the truncated protein was cloned. The figure was generated using CHROMA (47). All the accession numbers mentioned are from Swiss-Prot data base. B, pairwise sequence alignment of P. falciparum FabZ (accession number Q8I6T4) with the template sequence E. coli FabA (Protein Data Bank code 1MKB). * indicates the active site residues. Residues are indicated as mentioned in A. The first row shows the assignment of helices and ␤-strands to the PfFabZ sequence according to PSIPRED, which has 77% prediction accuracy. The second row gives the secondary structure of the residues in the modeled PfFabZ structure. The fifth row indicates the Protein Data Bank assignment of helices and ␤-strands to EcFabA structure. H ϭ helix; E ϭ extended ␤-strand. The sixth row depicts the consensus between the sequences of PfFabZ and EcFabA. Pred, predicted; Sec. Str., secondary structure.

Modeling of P. falciparum FabZ
The structure of FabZ is not available from any source, but P. falciparum FabZ shares 21% sequence identity with E. coli FabA. We modeled P. falciparum FabZ using E. coli FabA (Protein Data Bank code 1MKB) as the template. Modeling was done using MOE (Molecular Operating Environment) (19). Ten intermediate homology models were built as a result of the permutational selection of different loop candidates and side chain rotamers. The intermediate models were averaged to produce the final model by Cartesian average. We used Swiss-Pdb-Viewer to generate the dimeric structure with transformation matrix from the template structure (20).

Synthesis of Inhibitors
Of the four compounds synthesized, NAS-21, NAS-75, and NAS-79 were synthesized using published procedures (21,22). NAS-91 is a novel compound, therefore, its synthesis is described in detail below.
methyl tert-butyl ether 25% NaOCH 3 in methanol (4 ml, 1.5 eq) was added slowly. A solution of 4-nitroacetophenone (1.65 g, 0.01 M) in methyl tertbutyl ether was added to this mixture dropwise over 10 min. The reaction mixture was stirred at room temperature for 18 h. After the completion of the reaction the mixture was quenched with 3 N HCl. The organic layer was collected, washed with water and brine, and dried over Na 2 SO 4 , and the product (1.80 g) was crystallized from chloroform (70% yield).

Preparation of 1-(4-Methoxyphenyl)ethanone [(4-Trifluoromethyl)pyrimidine-2-yl]hydrazone (NAS-75)-
To 4-methoxyacetophenone (150 mg, 1 mM) in 5 ml of acetic acid and 1 ml of water 2-hydrazino-4-(trifluoromethyl)pyrimidine (178 mg, 1 mM) was added. Shortly after addition, a solid formed to which acetic acid was added to maintain stirring for 16 h. The reaction mixture was diluted with 4 ml of water, and subsequently the solid mass was filtered and dried in vacuo to afford 270 mg (90% yield) of the product. ␤-Hydroxyacyl-ACP Dehydratase of Plasmodium falciparum droxyquinoline (2 mM, 360 mg) in 5 ml of N-methylpyrrolidinone, and the slurry was heated with stirring under regular flow of N 2 at 110°C for 45 min. After cooling the reaction mixture to room temperature, 2-bromo-4-chlorophenol (1 mM, 207 mg) and copper (I) chloride (0.5 mM, 50 mg) was added. The flask was degassed and filled with N 2 four to five times. The reaction mixture was heated at 140°C for 70 h under the continuous flow of N 2 . The completion of the reaction was monitored by thin layer chromatography. After completion the reaction mixture was diluted 10 times with ethyl acetate and filtered through celite. The organic layer was washed with water and brine and dried over sodium sulfate. The product in the organic layer thus obtained was purified on a silica gel (100 -200 mesh) column equilibrated in petroleum ether. Adsorbed material was eluted with a gradient of

Inhibition of PfFabZ Activity
The inhibition of PfFabZ activity was monitored by the spectrophotometric assay performed as described above except that a given inhibitor was added prior to the initiation of the reaction by addition of crotonoyl-CoA. The studies were performed in the presence of 1%

Fluorescence Analysis
Fluorescence measurements were performed on a Jobin-Yvon Horiba fluorimeter under computer control. The excitation and emission monochromator slit widths were 3 and 5 nm, respectively. Measurements were performed at 25°C in a 3-ml quartz cuvette, and the solutions were mixed continuously on a magnetic stirrer. The solutions containing PfFabZ were excited at 280 nm, and the emission was recorded from 300 to 500 nm.
For inhibitor binding studies, PfFabZ (22 M) in 20 mM Tris, 500 mM NaCl, pH 7.5 was titrated with different concentrations of the inhibitors (0 -8.3 M). The magnitude of fluorescence decrease (F 0 Ϫ F) upon addition of each inhibitor concentration was fitted to Equation 2 to determine the value of K i , Corrections for the inner filter effect were performed according to Equation 3 (24), where F c and F are the corrected and measured fluorescence intensities, respectively. A ex and A em are the solution absorbances at the excitation and emission wavelengths, respectively. K i represents the dissociation constant of the inhibitor for PfFabZ. ⌬G, the change in free energy upon binding of inhibitor to the protein, was calculated from Equation 4, where R is the gas constant, T is the temperature in Kelvin, and K b is the binding constant.

Growth Inhibition Assay
The experiments were performed using P. falciparum FCK2 strain (chloroquine-sensitive, IC 50 , 18 nM), an isolate from Karnataka, India. P. falciparum was cultured using standard techniques (25) and routinely synchronized using 5% sorbitol (26). Growth inhibition by the compounds of interest was assessed by [ 3 H]hypoxanthine uptake. Typically uninfected or infected (1-2% parasitemia) red blood cells (2% hematocrit) were added to the culture medium in the wells of a 96-well plate (Nunc), and different concentrations of inhibitor in Me 2 SO were added such that the final concentration of Me 2 SO did not exceed 0.05%. The experiment was started with the synchronized parasite culture in the early trophozoite stage, and [ 3 H]hypoxanthine was added to the culture after incubation with the inhibitors for 48 h. The culture from the 96-well plate was harvested after 31 h (27, 28) using a Nunc cell harvester, and P. falciparum growth was assessed by measuring the incorporation of [ 3 H]hypoxanthine (29) using a liquid scintillation counter (Wallac). The IC 50 was calculated from a plot of relative percent parasitemia versus log concentration of the inhibitor by fitting it to non-linear regression analysis using SigmaPlot 2000 software.

Cell-free Fatty Acid Synthesis by P. falciparum Extracts
The in vitro fatty acid synthesis was performed as described earlier (4,30). Trophozoites isolated from 100-ml cultures with 8 -10% parasitemia were resuspended and sonicated in 0.2 ml of 70 mM potassium phosphate buffer (pH 7.0) for 5 s and centrifuged at 48,000 ϫ g for 1 h

Incorporation of [1,2-14 C]Acetic Acid into Fatty Acids in P. falciparum Cultures
Inhibitors (10 and 100 M) were added to P. falciparum cultures (100 ml, 9 -10% parasitemia) for 70 h after which cultures were resuspended in 7 ml of the complete medium while retaining the same concentrations of the inhibitors. To this [1,2-14 C]acetic acid was added (50 Ci/ml) (4). After 2 h, parasites were isolated, washed thoroughly with phosphate-buffered saline, lysed, sonicated, spotted onto a Whatman No. 3MM paper disc, and counted using the scintillation fluid (4). The data reported are from an average of two independent duplicate experiments.

Docking of Inhibitors with PfFabZ
Preparation of the Ligand and Receptor Molecules-The protein target and the ligands were prepared for docking using AutoDock Version 3.05 (31,32). Charges were assigned using the Kollman algorithm (33). Atomic solvation parameters and fragmental volumes were determined using the Addsol program. AutoTors was used to define torsion angles in the ligand. Polar hydrogen charges of the Gasteiger type (34) were assigned, and the non-polar hydrogens were merged with the carbons. The protein side chains were kept rigid in all the docking simulations.
Docking Simulations-Grid maps for docking simulations were generated with 60 grid points (with 0.375-Å spacing) in x, y, and z direction centered at N⑀-2 atom of His-133 by the AutoGrid program. Lennard-Jones parameters 12-10 and 12-6 (supplied with the program package) were used for modeling H-bonds and van der Waals interactions, respectively. The distance-dependent dielectric permittivity of Mehler and Solmajer

␤-Hydroxyacyl-ACP Dehydratase of Plasmodium falciparum
with a predicted molecular mass of 26.2 kDa. As can be seen in Fig. 1A, PfFabZ is more similar to plant FabZs than to the bacterial FabZ. PfFabZ, like its counterparts from plants, has a long N-terminal sequence, which is characteristic of proteins targeted to the apicoplast (36,37). Plasmodium and Toxoplasma FabZ sequences share 40.5% identity. In the case of Toxoplasma gondii the mature protein is predicted to start at position 79 of the protein. The residues are highly conserved in this region between P. falciparum and T. gondii FabZ. Thus, we designed primers for cloning the sequence encoding the mature protein (i.e. without the signal and transit sequence) starting at position 77 of P. falciparum FabZ.
Initial attempts to PCR amplify PfFabZ using Plasmodium DNA resulted in a product of ϳ800 bp ( Fig. 2A). However, upon performing RT-PCR using the same primers on Plasmodium RNA, a band corresponding to ϳ465 bp was obtained, thus demonstrating the presence of an intron in the open reading frame (Fig. 2B). Indeed upon careful examination of the fulllength gene, typical intron processing sites were observed. We therefore cloned the RT-PCR product encoding the mature FabZ in pET28a(ϩ) vector.
Expression and Purification of FabZ-The mature PfFabZ (without the signal and transit sequence) was expressed in E. coli BL21(DE3) cells as a fusion protein with an Nterminal His tag. The removal from the mature protein of an additional 10 residues, which do not contribute to enzyme function, following the transit peptide sequence led to an increase in the protein yield from 5 to 15 mg/liter of culture (truncated PfFabZ). The protein was purified on a Ni-NTA affinity column to homogeneity as seen in Fig. 2C. The purified truncated protein on SDS-PAGE yields a monomeric molecular weight of M r 17,000 Ϯ 1500. PfFabZ eluted at an elution volume of 17.5 ml in a Superdex 200 column corresponding to an apparent molecular weight of M r 34,200 Ϯ 2500. Thus, PfFabZ exists as a dimer under these conditions (2 mg/ml PfFabZ in 20 mM Tris, 500 mM NaCl, pH 7.5) (Fig. 2D). The protein yielded a radius of gyration of 6.79 Ϯ 0.15 nm confirming that it exists as a dimer in solution (Fig. 2E).
Kinetic Characterization of FabZ-PfFabZ was characterized in vitro by spectrophotometric as well as HPLC assay. FabZ catalyzes the reversible dehydration of a ␤-hydroxyacyl-CoA to an enoyl-CoA as depicted in Reaction 1 below.
In vivo, ACP thioesters are utilized, although other thio moieties, including CoA and N-acetylcysteamine, can also be exploited in vitro for the study of the enzymes of fatty acid biosynthesis pathway (12). For our studies, we have used substrate analogs ␤-hydroxybutyryl-CoA and crotonoyl-CoA. The conversion of crotonoyl-CoA to ␤-hydroxybutyryl-CoA was monitored by following the decrease in absorbance at 260 nm. A K m of 86 M was determined by this assay (Fig. 3A and Table  I) with a specific activity of 68.5 nmol/min/mg. The value of k cat was calculated as 0.019 s Ϫ1 with k cat /K m of 220 M Ϫ1 s Ϫ1 . The forward reaction, i.e. the conversion of ␤-hydroxybutyryl-CoA to crotonoyl-CoA, was monitored by following the increase in absorbance at 260 nm, and a K m of 199 M and a specific activity of 55.3 nmol/min/mg were obtained (Fig. 3B). The k cat for the conversion of ␤-hydroxybutyryl-CoA to crotonoyl-CoA was 0.016 s Ϫ1 with k cat /K m of 80.4 M Ϫ1 s Ϫ1 . PfFabZ utilized ␤-hydroxybutyryl-N-acetylcysteamine very poorly with a K m and V max of 2.6 mM and 0.27 nmol/min/mg, respectively.
We also analyzed the conversion of crotonoyl-CoA to ␤-hydroxybutyryl-CoA by HPLC assay and determined a K m of 95 M for crotonoyl-CoA (Fig. 4D). We studied the forward and reverse reactions in the presence of a saturating concentration (300 M) of the respective substrate analog and an equal amount of enzyme. PfFabZ catalyzes the reverse reaction more efficiently as is apparent from the greater amount of the product, ␤-hydroxybutyryl-CoA, formed as compared with crotonoyl-CoA in the forward reaction (Fig. 4, A and B). We have analyzed the activity of PfFabZ using four carbon chain substrate analogs. Hence we do not know whether PfFabZ has isomerase activity in addition to its dehydratase activity.
Enoyl-ACP reductase (FabI) catalyzes the NADH-dependent reduction of trans carbon-carbon double bond (as in crotonoyl-CoA) to produce saturated acyl-CoA (butyryl-CoA). In the FabZ-FabI coupled assay (Fig. 4C), we observe substantial decrease in the ␤-hydroxybutyryl-CoA concomitant with the appearance of butyryl-CoA, demonstrating that FabI in P. falciparum elongation reaction pulls the reaction in the forward direction. The ratio of the rate constants for the forward to the ␤-Hydroxyacyl-ACP Dehydratase of Plasmodium falciparum reverse reaction (k f /k r ) as calculated from spectrophotometric assay is 1:7, reiterating the fact that the equilibrium of the PfFabZ reaction lies toward the hydration reaction. Indeed it has been shown earlier that the equilibrium of the dehydratase-catalyzed reaction of E. coli lies toward ␤-hydroxyacyl-ACP; however, this activity of dehydratase is critical to supplying trans-2-acyl-ACP to FabI, which pulls each cycle of elongation to completion (11,15).
Homology Modeling of PfFabZ-The sequence of P. falciparum FabZ was BLAST searched and compared for its similarity with FabZ from other organisms using ClustalW (38). As shown in Fig. 1A the sequence of PfFabZ is most similar to the corresponding enzyme from plants and bacteria consistent with its evolutionary linkage to a photosynthetic bacterium and its location in the apicoplast of the parasite. We also note that in PfFabZ, His-133 and Glu-147 are conserved with FabZ from all the organisms reported till date. These residues are perhaps involved in its enzyme activity as they occupy positions similar to the catalytically active His-70 and Asp-84 of E. coli FabA. So far FabZ from only a few sources has been characterized, and the structure of the enzyme has not been reported from any organism. PfFabZ shares 70% amino acid sequence similarity with E. coli FabA, whose structure and function is well studied (Fig. 1B). FabA performs dehydratase as well as isomerase activity, and the two residues involved in the activity are His-70 and Asp-84Ј (39). While histidine is conserved in the two cases, the other residue, a glutamic acid, occupies a position in FabZ identical to that of an aspartic acid of FabA. Much of the 21% identity between PfFabZ and E. coli FabA is localized in the active site region, and the two protein sequences have 70% overall similarity. An examination of the homology-modeled structure of PfFabZ shows that the sequence of PfFabZ is consistent with the ␣ ϩ ␤ "hot dog" fold. The modeled structure of PfFabZ enables us to propose two identical active sites made by the residues from different subunits at the dimer interface, and this is extrapolated from the similar feature of the homologous E. coli FabA (39). Thus, PfFabZ is a physiological dimer. According to the PfFabZ model, the dimensions of the monomer are ϳ55 ϫ 35 ϫ 35 Å, and those of the dimer are ϳ64 ϫ 35 ϫ 47 Å. This correlates with the dynamic light scattering data, which give a radius of gyration of 6.79 Ϯ 0.15 nm for the dimer of native PfFabZ indicating that it exists as a dimer in solution. Gel filtration data are also consistent with these interpretations.
Inhibitors of FabZ Reaction-Due to the importance of FabZ in the cellular fatty acid biosynthesis, it is a potent target for the development of antimalarials. While inhibitors of several of the enzymes of FAS II are known and extensively studied (40,41), including FabA, which is known for its classic susceptibility to mechanism-based inactivation by 3-decynoyl-N-acetylcysteamine, a synthetic substrate based analog (42,43), no inhibitor of FabZ is known to date. We started by designing compounds that have a wide range of functional groups of different series for inhibition of FabZ (Scheme II), viz (i) hydrazones having trifluoromethylpyrimidine rings as well as methoxyphenyl/methyl phenyl (NAS-75 and NAS-79), (ii) diketone having para-nitrophenyl, trifluoromethyl, and an active methylene group (NAS-21), which can easily tautomerize to its keto-enol form, (iii) diaryl ethers having a phenolic group (proton donor), sp 2 hybrid nitrogen in which the lone pair electron is not involved in aromatic ring (proton acceptor), and a chlorine atom, which orients the molecule for forming a complex appropriately with the enzyme. Apart from these properties, molecular geometry has also been taken into consideration as diaryl ethers are extremely planar. Also the molecule has 16 e Ϫ clouds, which may be involved in van der Waals interactions of inhibitors with FabZ, e.g. NAS-91.
The effect of various inhibitors was estimated by spectrophotometric as well as HPLC assay. A decrease in the rate of enzyme activity was observed in the presence of both NAS-91 and NAS-21. The K i for the inhibition of PfFabZ reaction by NAS-91 and NAS-21 was calculated as 1.31 Ϯ 0.09 and 1.46 Ϯ 0.12 M, respectively, by Lineweaver-Burk plots (Fig. 5). We determined a K i of 0.9 Ϯ 0.1 and 1.2 Ϯ 0.15 M for NAS-91 and NAS-21, respectively, by the Dixon method (Fig. 6). Both the inhibitors showed competitive kinetics with respect to crotonoyl-CoA and demonstrated equal inhibitory activity toward PfFabZ. The other two compounds synthesized, NAS-75 and NAS-79, did not cause any inhibition of the PfFabZ activity. NAS-91 has one of the rings (4-chlorophenol) similar to the well known antibacterial triclosan (5-chloro-2-(2,4-dichlorophenoxy)phenol) that has been shown to inhibit enoyl-ACP reductase (FabI) of P. falciparum (4,5). However, we did not observe any inhibition of PfFabI by NAS-91 and NAS-21 even at concentrations as high as 220 M (the maximum concentra- ␤-Hydroxyacyl-ACP Dehydratase of Plasmodium falciparum tion of these inhibitors, especially NAS-91, achievable in the presence of 1% Me 2 SO). Also triclosan did not inhibit PfFabZ. Thus, despite some similarities in the structures of NAS-91 and triclosan, they inhibit specifically PfFabZ and PfFabI, respectively.
Fluorometric Analysis of PfFabZ-Inhibitor Interactions-Pf-FabZ exhibited excitation and emission wavelengths of 280 and 310 nm, respectively. The addition of both NAS-91 and NAS-21 led to a decrease in the fluorescence intensity at 310 nm. The magnitude of initial rapid fluorescence decrease (F 0 Ϫ F) caused by the inhibitors showed a saturation behavior (Fig. 7). The data were fit using Equation 2 to obtain a binding constant of 1.6 Ϯ 0.04 ϫ 10 6 and 1.2 Ϯ 0.03 ϫ 10 6 M Ϫ1 for NAS-91 and NAS-21, respectively. Hence the calculated value of change in the free energy upon binding of the two inhibitors, NAS-91 and NAS-21, to PfFabZ are Ϫ8.36 and Ϫ8.19 kcal/mol (Table II).
Inhibition of Plasmodium Growth in Vitro-Of the four compounds analyzed for PfFabZ inhibition, only NAS-21 and NAS-91 showed inhibition. Hence we tested only these compounds in in vitro growth inhibition assays using synchronized P. falciparum cultures. Both NAS-21 and NAS-91 showed a rapid effect, killing the parasites within 48 h (one erythrocytic cycle) with IC 50 values of 100 and 7.4 M, respectively (Fig. 8A). In all the analyses of inhibition data, the in vivo efficacies of NAS-91 and NAS-21 are lower by a factor of 4 -6 and 50 -60, respectively. The observed lower potency of the inhibitors in in vivo assays as well as the lower potency of NAS-21 as compared with NAS-91 in these assays could be due to several reasons. The inhibitors may be differentially bound or metabolized by components of the culture medium or by intracellular and/or intraorganellar compartments, and/or they may cross the erythrocytic, parasitic, or apicoplast membranes to varying extents. Also they may be challenged by the cellular efflux system to varying degrees. There are a number of precedences for even more dramatic differences between the efficacies with the isolated enzymes and the in vivo systems including those in the area of the inhibitors of type II fatty acid synthesis itself.   (44)); however, micromolar concentrations (minimum inhibitory concentration ϭ 0.6 -1 M (45)) are required for inhibiting the growth of bacteria in cultures. Furthermore the possibility of additional unidentified targets for these inhibitors should also be considered as has been noted for triclosan in Bacillus subtilis (46). Inhibition of Parasite Fatty Acid Synthesis-The incorporation of [2-14 C]malonyl-CoA into fatty acids by cell-free P. falciparum extracts was studied in the presence and absence of inhibitors (Fig. 8B). Although NAS-91 and NAS-21 inhibited this incorporation, they had different effects on the parasite fatty acid synthesis pathway. This can be attributed to their differential mode of binding to PfFabZ, or alternatively they may be affecting differentially some other enzyme of the fatty acid biosynthesis pathway. The in situ incorporation of [1,2-14 C]acetic acid into fatty acids by P. falciparum was also used as an index of the fatty acid synthesis. The incorporation of [1,[2][3][4][5][6][7][8][9][10][11][12][13][14] C]acetic acid in the parasite fatty acids was reduced by 46 and 26%, respectively, in the presence of 10 M NAS-91 and NAS-21 (Table III).
Docking of Inhibitors with PfFabZ-We also docked the in-hibitors with PfFabZ to seek an explanation for their affinities (Fig. 9). NAS-91 and NAS-21 gave similar ⌬G values as calculated using AutoDock, which is consistent with our experimental enzyme inhibition data (Table II). The inhibitors docked with the active site region of the enzyme are shown in Fig. 9, A and B. These complexes have the best energies of binding. These studies depict that NAS-91 blocks the entrance of the active site and localizes near residues Lys-137, Ile-139, Pro-182, and Leu-184 from one of the subunits and Lys-165Ј, Asn-166Ј, and Asn-167Ј of the other subunit. Charge group interactions might play a role in fixing the orientation of the phenolic ring of NAS-91. NAS-21 is found in the tunnel-like active site surrounded by residues His-133, Val-143, Val-177, and Trp-179 from one subunit and Glu-147Ј and Leu-168Ј of the other subunit.  a Dashes represent that in the specific assay, the compound was not added.

␤-Hydroxyacyl-ACP Dehydratase of Plasmodium falciparum
The inhibitory activity of NAS-91 is probably related to the orientation of proton donor and acceptor groups in the transform, which imparts them with an affinity for the complementary hydrophilic and hydrophobic patches, respectively, on the enzyme. In NAS-21, there is a possibility of keto-enol tautomerism, which might play a role in the inhibition of the enzyme. The poor activity of NAS-75 and NAS-79 is probably due to the fact that these molecules, apart from having the planar rings, have the sp 3 hybrid tetrahedral carbon, which makes the system rigid creating steric hindrance, which in turn makes it difficult for these complexes to form the complex with the enzyme.
In conclusion, we not only report expression and molecular properties of FabZ from P. falciparum but also two small organic molecules that inhibit a member of the FabZ family of enzymes paving the way for the development of not only antimalarials but also anti-infectives targeting this enzyme of type II FAS. The availability of purified PfFabZ would contribute toward a better understanding of P. falciparum fatty acid biosynthesis and in developing high throughput screening assays for the identification of new inhibitors as potent antimalarials.