A pantothenate kinase from Staphylococcus aureus refractory to feedback regulation by coenzyme A.

The key regulatory step in CoA biosynthesis in bacteria and mammals is pantothenate kinase (CoaA), which governs the intracellular concentration of CoA through feedback regulation by CoA and its thioesters. CoaA from Staphylococcus aureus (SaCoaA) has a distinct primary sequence that is more similar to the mammalian pantothenate kinases than the prototypical bacterial CoaA of Escherichia coli. In contrast to all known pantothenate kinases, SaCoaA activity is not feedback-regulated by CoA or CoA thioesters. Metabolic labeling of S. aureus confirms that CoA levels are not controlled by CoaA or at steps downstream from CoaA. The pantothenic acid antimetabolite N-heptylpantothenamide (N7-Pan) possesses potent antimicrobial activity against S. aureus and has multiple cellular targets. N7-Pan is a substrate for SaCoaA and is converted to the inactive butyldethia-CoA analog by the downstream pathway enzymes. The analog is also incorporated into acyl carrier protein and D-alanyl carrier protein, the prosthetic groups of which are derived from CoA. The inactivation of acyl carrier protein and the cessation of fatty acid synthesis are the most critical causes of growth inhibition by N7-Pan because the toxicity of the drug is ameliorated by supplementing the growth medium with fatty acids. The absence of feedback regulation at the pantothenate kinase step allows the accumulation of high concentrations of intracellular CoA, consistent with the physiology of S. aureus, which lacks glutathione and relies on the CoA/CoA disulfide reductase redox system for protection from oxidative damage.

CoA is an essential cofactor that functions as the major acyl group carrier in intermediary metabolism and is synthesized in five steps from pantothenic acid (vitamin B 5 ) (1, 2). The first step in the pathway is the ATP-dependent phosphorylation of pantothenate by pantothenate kinase (3). All organisms characterized to date share a common mechanism to regulate the phosphorylation of pantothenic acid. Feedback inhibition of pantothenate kinase by CoA and/or its thioesters governs the flux through the pathway and determines the upper threshold for the intracellular cofactor levels in bacteria (4 -6), plants (7), and animals (8,9). The primary sequences of the prototypical prokaryotic Escherichia coli pantothenate kinase (EcCoaA) 1 and the eukaryotic pantothenate kinases have little sequence similarity (Fig. 1). Nonetheless, the enzymes share the common property of being feedback-regulated by CoA and its thioesters, and this regulatory mechanism is primarily responsible for controlling the intracellular CoA concentration (4,6,9). The bacterial enzyme is more effectively regulated by CoA (6,7), whereas the mammalian and Aspergillus nidulans pantothenate kinases are most potently inhibited by acetyl-and malonyl-CoA (7, 9 -11). The mechanism of the feedback inhibition by CoA has been extensively investigated in E. coli and arises from competitive inhibition of ATP binding (5,6). The details of this interaction are revealed by the high-resolution x-ray structures of the EcCoaA-ATP␥S (12), EcCoaA-CoA (12), and EcCoaA-ADP-pantothenate (13) complexes. Although the E. coli CoaA is considered the model bacterial pantothenate kinase, this isoform is not universally expressed in bacteria (14). Pseudomonas aeruginosa and Helicobacter pylori do not have recognizable pantothenate kinases in their genomes, although all other components of the biosynthetic pathway are present. Staphylococcus aureus and Bacillus anthracis possess an alternate isoform of pantothenate kinase that is more closely related to the mammalian enzymes than it is to EcCoaA (Fig. 1). SaCoaA was recently expressed and demonstrated to possess pantothenate kinase activity (15).
CoA is also the source of the 4Ј-phosphopantetheine prosthetic group present in a number of proteins that function as acyl/aminoacyl/peptidyl group carriers. Examples are the carrier proteins (ACP) of fatty acid synthases, polyketide synthases, and non-ribosomal peptide synthases (16). The transfer of the 4Ј-phosphopantetheine moiety of CoA to a conserved serine residue of these carrier proteins is catalyzed by dedicated phosphopantetheinyl transferases. Some of these transferases possess a restricted substrate specificity such as E. coli AcpS and EntD (17), whereas others are more promiscuous like Bacillus subtilis Sfp (17,18). E. coli AcpS specifically modifies the ACP involved in fatty acid biosynthesis, although it also accepts substrates such as Dcp and some ACPs of type II polyketide synthases (17,19,20). Dcp is a small carrier protein required for the transfer of D-alanine to lipoteichoic acid in Gram-positive bacteria (21)(22)(23).
The pantothenic acid analogs N5-Pan and N7-Pan inhibit E. coli growth (24 -26). These pantothenate antimetabolites are substrates and competitive CoaA inhibitors with respect to pantothenate (13). N5-Pan is further metabolized through the CoA biosynthetic pathway to the cofactor analog ethyldethia-CoA (25). More recently, the incorporation of these analogs into ACP was demonstrated, and the blockade of fatty acid synthesis by the accumulation of inactive ACP is the root cause for growth inhibition in the E. coli model system (26). N5-Pan and N7-Pan are also effective antimicrobial agents against S. aureus (15,19); however, in this case Choudhry et al. (15) found that these compounds are inhibitors but not substrates for SaCoaA. This suggests that the pantothenate antimetabolites have a different mechanism of action in S. aureus than in E. coli. This study reports the unique regulatory properties of SaCoaA and the mechanism of action of the pantothenamide antimetabolites in S. aureus. All other reagents were of analytical grade or better and were obtained from Sigma, unless otherwise stated. N-Pentylpantothenamide and N-heptylpantothenamide were synthesized as described previously (13). Analysis by electrospray (ES-MS) or matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry of protein samples was performed by the Hartwell Center for Bioinformatics and Biotechnology, St. Jude Children's Research Hospital.
LB medium was supplemented with kanamycin (30 g/ml) for the expression of CoaA or with carbenicillin (100 g/ml) and chloroamphenicol (30 g/ml) for the expression of ACP from the respective strains. Overnight cultures of strains BL21 Star(DE3)/pSC4 or Rosetta(DE3)/pRL002 at the early to mid-log phase were used as 1% inoculum into fresh LB medium (0.2-1 liter), and cells were cultured at 37°C until the A 600 reached 0.5-0.7. Expression of ACP was induced at this point by the addition of isopropyl 1-thio-␤-D-galactopyranoside (1 mM), whereas cultures of strain BL21 Star(DE3)/pSC4 were incubated at 42°C for 10 min followed by 20 min of incubation at room temperature prior to the addition of the same inducer (0.1 mM). Cells were grown for 3 h postinduction and harvested by centrifugation (Sorvall SLA-3000, 6000 rpm, 10 min, 4°C). Cell pellets were stored at Ϫ20°C.
Protein Purification and Preparation of ACP-E. coli AcpS was expressed as a His 6 -tagged protein and purified by immobilized metal ion affinity chromatography following the general procedure described below. Protein concentrations were determined by the method of Bradford (27) using ␥-globulin as a standard. Protein purity was estimated by SDS-PAGE analysis on precast 10% bis-Tris gels (Invitrogen).
Strain Rosetta(DE3)/pRL002 cell pellets from 1 liter of culture were resuspended in loading buffer containing 1 mM phenylmethylsulfonyl fluoride, 0.1 mg/ml DNase, 0.1 mg/ml lysozyme (20 ml) and lysed by using a French press. His 6 -tagged ACP was purified by scaling up the procedure described above and obtained with a high purity (Ͼ98%) from the nitrilotriacetic acid column. The protein was dialyzed against dialysis buffer (4 liters) and concentrated by ultrafiltration, and an aliquot (0.5 ml, 12 mg/ml) was treated with thrombin as per the manufacturer's instructions. Thrombin-cleaved apoACP (0.4 ml, 8 mg/ml) was converted to ACP in reaction buffer (50 mM Tris-HCl, pH 8.5, 55 mM MgCl 2 , and 2.5 mM dithiothreitol, 0.8 ml) containing 4 M AcpS and 2 mM CoA (28). After 30 min of incubation at 37°C, the reaction mixture was applied onto a PD-10 column equilibrated in 20 mM Tris-HCl, pH 7.5, 50 mM NaCl, 5 mM ␤-mercaptoethanol, and ACP was then eluted and stored at Ϫ20°C.
Effect of CoA and CoA Thioesters on E. coli and S. aureus CoaAs-E. coli CoaA was expressed as a His 6 -tagged fusion protein from a pET-15b-derived vector and purified by immobilized metal ion affinity chromatography in a single step, as described under "Protein Purification and Preparation of ACP." The pantothenate kinase activity of EcCoaA and SaCoaA was measured in reaction mixtures containing 85 M D-[1-14 C]pantothenate (specific activity, 55 mCi/mmol), 250 M ATP, 25 mM MgCl 2 , 100 mM Tris-HCl, pH 7.5, increasing concentrations of CoA, acetyl-CoA, or malonyl-CoA, as indicated, and either 100 ng of EcCoaA or 20 ng of SaCoaA. Reaction mixtures were analyzed as described previously (5,29).
Analysis of Intracellular and Extracellular ␤-Alanine-derived Metabolites-␤-[3-3 H]Alanine mixtures of increasing concentration were prepared by diluting ␤-[3-3 H]alanine (20 M, specific activity 50 Ci/mmol) with non-radioactive ␤-alanine. This resulted in a 2-fold decrease of the isotope specific activity (from 50 to 0.78 Ci/mmol) for each successively higher concentration. S. aureus strain RN4220 was cultured in 1% tryptone medium overnight at 37°C and used as 1% inoculum into fresh medium (10 ml). 1-ml aliquots were then transferred to 15-ml tubes containing increasing concentrations of ␤-[ 3 H]alanine from 0.5 to 32 M as indicated. After 4 h at 37°C, cells were transferred to 1.5-ml tubes and harvested by centrifugation. The supernatant was removed and stored at Ϫ20°C. The cell pellets were washed with phosphate-buffered saline and resuspended in 50 mM Tris-HCl, pH 7.5, 12.5 mM EDTA, 0.4 mg/ml lysostaphin (50 l). Cells were lysed after 30 min of incubation at room temperature, and the cell debris was removed by centrifugation. The total radioactivity in the cleared lysates was estimated by scintillation counting. The ␤-[ 3 H]alanine-derived products were separated on silica gel H thin layer chromatography plates developed with butanol:acetic acid: water (5/2/4, v/v) and detected by a Bioscan imaging detector (30,31).
Measurement of the MIC-The MICs of N5-Pan and N7-Pan against S. aureus strain RN4220 were determined by a broth microdilution method. Briefly, S. aureus was cultured at 37°C in 1% tryptone to mid-log phase, diluted 30,000 times, and used to inoculate (10 l, 3,000 -5,000 colony-forming units) the wells of a 96-well plate (U-bottom with low evaporation lid) containing fresh 1% tryptone medium (100 l) supplemented with the indicated concentrations of pantothenate analogs or Me 2 SO (negative control). The same experiments were repeated in the presence of 50 M pantothenate. After 20 h of incubation at 37°C, the A 600 of the cell suspensions in the wells was measured with a Fusion TM universal microplate analyzer (Packard Instrument Co.). The optical density measured in negative controls was taken as 100% growth.
Determination of IC 50 Values and Kinetic Parameters for the Pantothenate Analogs-The inhibitory effect of N5-Pan and N7-Pan on the kinase activity of SaCoaA was estimated by measuring the amount of radioactive 4Ј-phosphopantothenate produced in standard reaction mixtures containing 45 M D-[1-14 C]pantothenate (specific activity, 55 mCi/mmol), 100 M ATP, 10 mM MgCl 2 , 50 mM Tris-HCl, pH 7.5 (13), and increasing concentrations of each analog as indicated. Reaction mixtures were incubated at 37°C for 10 min and then stopped and analyzed as described previously (13). The kinetic constants were obtained from duplicate experiments by nonlinear regression analysis of initial velocities using Prism 4 (GraphPad Software). The constants for pantothenate and ATP were determined under standard conditions by fixing the ATP concentration at 250 M and increasing the D- Reaction mixtures were incubated for 30 min at 37°C, and the products were analyzed as described previously (13).
Growth Inhibition by N7-Pan and Cerulenin and Effect of Exogenous Fatty Acids-Oleate and palmitate stock solutions (10 mg/ml) were prepared in 20% aqueous ethanol containing sufficient KOH to neutralize the corresponding acids. S. aureus strain RN4220 was cultured in 1% tryptone medium (45 ml) at 37°C until the A 600 reached 0.1-0.2. The cell suspension was then divided into two sets of two 10-ml aliquots each; 10 l of Me 2 SO (negative control) were added to one set of cultures and 10 l of N7-Pan (1.6 M final concentration) to the other. One culture per set was also supplemented with oleate (15 g/ml) and palmitate (40 g/ml) (32), and the same volume of 20% aqueous ethanol was added to the other samples. The cultures were incubated at 37°C, and cell growth was monitored by measuring the absorbance at 600 nm. After 1 h of incubation (A 600 ϭ 0.6 -0.8) a 1-ml aliquot per sample was diluted into 9 ml of fresh medium of identical composition and the A 600 measured at set intervals, as indicated. A similar experiment was conducted to monitor the extent of cell growth inhibition caused by 100 g/ml cerulenin in the presence or absence of fatty acids. However, unlike the experiment with N7-Pan, cell cultures were not diluted after 1 h of incubation in the presence of either Me 2 SO or cerulenin.
Labeling of Fatty Acid and Protein Biosynthesis-An overnight culture of S. aureus strain RN4220 in 1% tryptone was used as 1% inoculum into fresh medium (20 ml) and cells were grown at 37°C until the A 600 reached 0.1. The cell culture was then divided into two aliquots to which 10 l of N7-Pan (1.6 M final concentration) or Me 2 SO (negative control) were added. The cultures were returned to 37°C and incubated for a further 3 h and 15 min. 1-ml aliquots were then removed and labeled with either [1-14 C]acetate (73 M, specific activity 55 mCi/mmol) or a L-3 H-labeled amino acid mixture (1 Ci) for 15 min at 37°C. At the end of the labeling experiment, the A 600 was measured, and cells were harvested by centrifugation. To determine the extent of [1-14 C]acetate incorporation into membrane lipids, the cell pellets were washed with phosphate-buffered saline and resuspended in water (100 l), the cellular lipids were then extracted with chloroform, and the incorporation of the 14 C isotope into this phase was quantitated by scintillation counting. S. aureus cultures with 3 H-labeled amino acids were harvested by filtration through HA filters. The filters were washed with phosphate-buffered saline, dried, and transferred to vials for scintillation counting. In both labeling experiments, the total radioactivity was normalized to the A 600 of the samples and converted to percent of biosynthesis with respect to the negative control.
Isolation and Identification of N7-ACP and N7-Dcp by ES-MS and MALDI-TOF/TOF Analyses-N7-ACP and N7-Dcp were partially purified from the cleared lysate of N7-Pan-treated S. aureus strain RN4220. Briefly, S. aureus was cultured overnight in 1% tryptone medium at 37°C, diluted 100-fold into fresh medium (2 liters), and grown until the A 600 reached 0.1. N7-Pan was then added to a final concentration of 1.6 M, and cells were incubated for 3.5 h before being harvested by centrifugation (Sorvall SLA-3000, 9000 rpm, 10 min, 4°C). The cell paste was resuspended in 10 mM bis-Tris, pH 6.5, containing 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 0.3 mg/ml lysostaphin (6 ml) and stirred for 1 h at room temperature. The cell suspension was cleared by centrifugation (Sorvall 70.1 TI, 50,000 rpm, 1 h, 4°C), and the cleared lysate was applied onto a HiTrap DEAE FF column (5 ml, Amersham Biosciences), equilibrated previously in 10 mM bis-Tris, pH 6.5 (50 ml). The column was washed with the same buffer (50 ml), and proteins were eluted at 2.5 ml/min with a linear gradient of 10 mM bis-Tris, pH 6.5, from 0 to 1 M LiCl over 75 ml. The fractions were analyzed by electrophoresis on 13% polyacrylamide gels containing 0.5 M urea, and the fractions containing faster migrating proteins with respect to the ACP standard were pooled and concentrated by ultrafiltration. Concentrated proteins were then applied onto a Superdex 75 16/60 column (Amersham Biosciences) equilibrated in 10 mM bis-Tris, pH 7.9, and eluted at 1 ml/min.
The protein bands were excised from the gel, and the proteins were reduced and alkylated with iodoacetamide. A tryptic digest was prepared and the unfractionated digest subjected to MALDI-TOF/TOF mass spectrometry using a 4700 proteomics analyzer (Applied Biosystems). The digest was introduced into the instrument in a crystalline matrix of ␣-cyano-4-hydroxycinnamic acid. Data base searches were conducted with the Applied Biosystems GPS explorer software, which uses the Mascot search engine. Protein assignments were made on the basis of both MS and MS/MS spectra. NCBI accession number 070404 was used for protein identification.
The fraction enriched in the two proteins (600 g/ml) was analyzed by ES-MS. Proteins were loaded on a weak anion exchange nanoextraction cartridge (Western Analytical, Murrieta, CA) and washed extensively with water at pH 8 to remove the bound salts. The protein was eluted from the column with 30 l of 70% acetonitrile with 5% formic acid and then diluted with 5% formic acid to a final acetonitrile concentration of 50%. Mass measurements were performed using an LCT ES-MS (Micromass Inc., Beverly, MA) equipped with a Z-spray interface (Micromass Inc.). A flow rate of 200 nl/min was maintained using a VLP200 syringe pump (Harvard Apparatus, Holliston, MA), and the desalted protein was introduced by direct injection. Data were collected for an m/z range of 500 -2500 at a cone voltage of 35 V and a manual pusher time of 70 s. All other instrument settings were those typically used for protein measurements on this instrument. Deconvolution of the protein spectrum was accomplished using the maximum entropy algorithm of the MassLynx software (Micromass Inc.) (33).

RESULTS
Purification and Characterization of SaCoaA-SaCoaA possesses an atypical primary sequence (10,15,34) that shares 18% identity with the murine CoaA isoform 1␤ (MmPanK1␤) and only 13% identity with EcCoaA, the prototype for bacterial pantothenate kinases (Fig. 1). Although SaCoaA is more related to the mammalian enzymes, it is not closely related to either pantothenate kinase isoform. SaCoaA was cloned, expressed as a C-terminal His 6 -tagged fusion protein in E. coli, and purified to homogeneity by immobilized metal ion affinity and gel filtration chromatography (Fig. 2). SDS-PAGE analysis of fractions possessing pantothenate kinase activity revealed the presence of a 29-kDa protein, consistent with the subunit molecular size predicted from the amino acid sequence of SaCoaA (Fig. 2B). Characterization by gel filtration analysis also indicated a native molecular mass of 59 kDa for SaCoaA, consistent with the existence of a homodimer ( Fig. 2A).
Regulation of CoA Biosynthesis in S. aureus-SaCoaA was assayed in vitro in the presence of different concentrations of CoA, acetyl-CoA, and malonyl-CoA and compared with the activity of EcCoaA (Fig. 3A). The ability of SaCoaA to phosphorylate pantothenate was completely unaffected by the presence of these pathway end products at concentrations that clearly inhibited EcCoaA (Fig. 3A). These data suggested that CoA biosynthesis in S. aureus was not controlled at the pantothenate kinase step. S. aureus cells were labeled with radioactive ␤-alanine, a pantothenate precursor, and intracellular and extracellular radiolabeled metabolites were analyzed to determine whether CoA synthesis was regulated at a downstream step in the biosynthetic pathway (Fig. 3B). The intracellular CoA concentration increased proportionally to the extracellular ␤-alanine concentration. Neither radiolabeled pantothenate nor 4Ј-phosphopantetheine accumulated in the medium or inside the cells (Fig. 3, C and D). The absence of pathway intermediates accumulating in either intra-or extracellular compartments indicated that CoA biosynthesis in S. aureus was not regulated by feedback inhibition either at SaCoaA or another downstream step.
Inhibition of SaCoaA and S. aureus Growth by Pantothenamide Antimetabolites-The pantothenate analogs N5-and N7-Pan inhibit both S. aureus and E. coli growth (13, 15, 25). The MICs of N5-Pan and N7-Pan against S. aureus strain RN4220 were determined by a broth microdilution method in the presence or absence of pantothenate (Fig. 4). The MIC values of N5-Pan and N7-Pan were estimated to be 0. 16  efficacy of the pantothenamides in the presence of pantothenate supported the idea that pantothenate metabolism is the primary target of these antimetabolites.
N5-Pan and N7-Pan inhibited SaCoaA activity with IC 50 values of 3.5 and 4.8 M, respectively (Fig. 5A), and were found to be both inhibitors and substrates for EcCoaA (13,25); however, Choudhry et al. (15) reported that the pantothenamides were not substrates for SaCoaA. We reinvestigated this issue   (Fig. 5B). The kinetic parameters for the natural substrates and the pantothenate analogs were determined and are listed in Table I. The pantothenamides had higher affinities for SaCoaA than pantothenate and lower k cat values resulting in similar k cat /K m values for all the substrates. These data showed that the pantothenamides are processed as efficiently by SaCoaA as pantothenate.
Effect of N7-Pan on Fatty Acid Biosynthesis-N7-Pan proved to be the most potent inhibitor of S. aureus growth and was used in all subsequent in vivo studies. The addition of 1.6 M N7-Pan (10ϫ MIC) to bacterial cultures at the early exponential phase (A 600 ϭ 0.1-0.2) did not result in an abrupt inhibition of cell growth in comparison to the negative control incubated with Me 2 SO (data not shown). Thus, the effects of the pantothenamides were not immediate, and the detection of a substantial difference in growth was observed when the N7-Pan-treated culture at the mid-log phase was diluted into medium containing the pantothenamide (Fig. 6B). Cerulenin inhibits S. aureus growth by blocking fatty acid biosynthesis (35), and the MIC for this compound was 25 g/ml. Unlike N7-Pan, the addition of 100 g/ml (4ϫ MIC) of cerulenin to a growing S. aureus culture quickly reduced the rate of bacterial replication, which ceased after 1 h of incubation (Fig. 6A). We then tested the effect of adding 15 g/ml oleate and 40 g/ml palmitate to S. aureus cells cultured in the presence of Me 2 SO or 1.6 M N7-Pan or 100 g/ml cerulenin. This combination of unsaturated and saturated fatty acids, known to restore the growth to cerulenin-treated S. aureus (32), also allowed the partial recovery of cells incubated with N7-Pan (Fig. 6). These data clearly pointed to fatty acid synthesis as the primary target for N7-Pan. This conclusion was further corroborated by metabolic labeling experiments with [ 14 C]acetate or a mixture of 3 Hlabeled amino acids. N7-Pan reduced by 85% the incorporation of the label into the lipid fraction and by 25% the incorporation of the label into the protein fraction, consistent with fatty acid synthesis being the primary target for N7-Pan.
Isolation of N7-ACP and N7-Dcp-We next investigated the possibility that the toxicity of N7-Pan was because of the formation of an inactive ACP by the transfer of the phosphopantothenamide from butyldethia-CoA to apoACP. The N7-ACP analog would be unable to form thioester bonds and therefore fail to function in fatty acid biosynthesis. The ACP fraction was partially purified from N7-Pan-treated cells and characterized. Different forms of ACP are resolved at alkaline pH on polyacrylamide gels containing low concentrations of urea (36,37). ACPs carrying acyl chains migrate faster than ACP, and because the N7-Pan hydrocarbon chain transferred to ACP would mimic a C-4 acyl chain, we anticipated that N7-ACP would migrate faster than the ACP standard. Two low molecular weight proteins were found in the ACP fraction, neither of which migrated with the ACP standard (Fig. 7A, inset). The two bands were excised and subjected to protein sequencing by MALDI-TOF/TOF mass spectrometry as described under "Experimental Procedures." The slower migrating band was identified as Dcp, a carrier protein required for the incorporation of D-alanine into membrane-associated lipoteichoic acids of Grampositive bacteria (21,38). The faster migrating band was identified as ACP. The ES-MS spectrum of the fraction clearly showed the presence of a protein with a mass of 8,926.58 Da, consistent with the calculated mass for N7-ACP (8,927.62 Da). A less intense peak with a mass of 9,439.67 Da was consistent with the molecular mass for N7-Dcp (expected mass 9,441.14 Da) (Fig. 7A). Normal ACP was not detected in the N7-Pantreated cell extracts. These data established that the pantothenamide antimetabolites were converted to inactive CoA analogs through the coenzyme biosynthetic pathway and used to form inactive ACP (Fig. 7B), explaining the inhibition of fatty  acid synthesis. Furthermore, the detection of N7-Dcp illustrated that these compounds also interfere with D-alanine modification of the lipoteichoic acid component of the cell wallmembrane complex in S. aureus. DISCUSSION A key finding of this study is that SaCoaA differs from all described previously pantothenate kinases (6, 9 -11) in that it is refractory to feedback inhibition by CoA and/or its thioesters (Fig. 3A). SaCoaA is more closely related to the mammalian pantothenate kinases than the EcCoaA (Fig. 1), and this atypical prokaryotic isoform is also found in other bacteria includ-ing B. anthracis. Feedback regulation of EcCoaA in vivo is illustrated in a metabolic labeling experiment using increasing amounts of labeled ␤-alanine in the medium (30). An upper limit to the intracellular CoA concentration is achieved at a ␤-alanine concentration of 4 M, and pantothenate accumulates pointing to restriction of CoA formation at the pantothenate kinase step. Expression of a point mutation of EcCoaA that is refractory to CoA feedback inhibition results in uncontrolled pantothenate phosphorylation (4). A similar experiment with S. aureus (Fig. 3B) shows that it produces CoA in proportion to the input of ␤-alanine with no evidence for regulation at SaCoaA or other downstream steps. The rationale for this distinct molecular property of SaCoaA is understood in the context of the physiology of the organism. Eukaryotes and most bacteria contain glutathione as the major low molecular weight thiol, which, together with the NADPH-dependent glutathione reductase, constitutes the primary thiol/disulfide redox system in nature (39,40). This system is essential for maintaining the intracellular reducing environment and protects the organism from oxidative insults by functioning in the detoxification of peroxides, epoxides, and other products of reactive oxygen. S. aureus lacks glutathione (41). Instead, CoA is the major intracellular thiol and together with a unique CoA disulfide reductase functions as the reducing system that performs the same role as the glutathione/glutathione reductase system (42)(43)(44). The CoA concentrations in S. aureus reach millimolar levels (42), and the lack of CoA feedback regulation in SaCoaA allows CoA levels to rise to an upper limit set by the availability of pantothenate and cysteine. These observations lead to the conclusion that the CoA levels in S. aureus are likely to be limited by the supply of pantothenate produced by the biosynthetic pathway encoded by the panB-E genes (2). These key differences in the regulation of the CoA biosynthetic pathway account for the differential response of S. aureus and E. coli to the pantothenamides in the presence of pantothenate. The simultaneous addition of inhibitors plus 50 M pantothenate to the growth medium resulted in an upward shift of the MICs (8-fold for N5-Pan and 64-fold for N7-Pan) against S. aureus, consistent with pantothenate competitively inhibiting the phosphorylation of the pantothenamides. The same concentration of pantothenate only results in a 2-fold shift in the MIC of N5-Pan against E. coli (26), suggesting that the lack of feedback regulation of SaCoA allows synthesis and accumulation of larger amounts of CoA, thus more effectively overcoming the inhibitory effect of the pantothenamides in S. aureus.
There are multiple targets for the pantothenamides in S. aureus and several important differences between this Grampositive pathogen compared with the E. coli model system (26). N7-Pan (MIC ϭ 0.16 M) is considerably more potent than N5-Pan (MIC ϭ 25 M) in S. aureus, but the efficacy of the two compounds is reversed in E. coli because of the export of N7-Pan from the cell via a TolC-dependent pump (26). Thus, the differences in the cell wall and the outer membrane define the selectivity of pantothenamides in Gram-positive and Gramnegative bacteria. In both systems, the pantothenamides are phosphorylated by CoaA and incorporated into CoA and ACP analogs as outlined in the pathway shown in Fig. 7B. The accumulation of N7(N5)-ACP results in the inactivation of ACP and the inhibition of fatty acid synthesis. Fatty acid synthesis is the most critical pathway blocked in both cases as shown by the ability of exogenous fatty acids to ameliorate the toxic effects of the pantothenamides (Fig. 6) (26). However, exogenous fatty acids are more effective in reversing the growth inhibition in response to the highly selective fatty acid synthesis inhibitor cerulenin than the pantothenamides (Fig. 6A) suggesting the existence of other targets in S. aureus. Also, it is important to emphasize that the effect of the pantothenamides on cell growth is slower than with cerulenin because the pantothenamides must be incorporated into ACP and reduce the concentration of the carrier below what is required for fatty acid synthesis. In contrast, cerulenin is a fast acting covalent modifier of the condensing enzyme step (35) and blocks fatty acid synthesis immediately at the concentrations used in our studies.
A new finding in S. aureus is that N7-Pan is also incorporated into Dcp (Fig. 7). Dcp is required for the biosynthesis of D-alanyl-lipoteichoic acid, a macroamphiphile component of the Gram-positive cell wall-membrane complex (23). The hydrophilic backbone of teichoic acids is made up of polymers of glycerol or ribitol joined by phosphate groups, and the extent of esterification with D-alanine determines the net anionic charge and the properties of the cell wall (23). D-Alanine is bound to the phosphopantetheinyl prosthetic group of Dcp as a thioester (21,22,38), the formation of which is catalyzed by D-alanine-Dcp ligase (38). Dcp inactivation would block D-alanine incorporation into the cell wall (22,23,38). This biochemical pathway is critical for the functions of the cell wall related to pathogenesis (23), but in the laboratory environment the lack of D-alanine modification is not lethal. The inhibition of Dalanine incorporation into teichoic acids by inactivation of the dlt operon in S. aureus increases not only the sensitivity of this pathogen to positively charged antimicrobial peptides such as defensins but also to vancomycin (45,46). Similarly, inactivation of the gene encoding Dcp (dltC) in Streptococcus mutans abolishes D-alanine incorporation into teichoic acid resulting in FIG. 7. Mass spectrophotometry and gel electrophoretic analysis of N7-ACP and N7-Dcp. A, ES-MS spectrum of purified ACP fraction isolated from N7-Pan-treated S. aureus. The expected mass for N7-ACP is consistent with the major peak at 8,926.58 Da, whereas the expected mass for N7-Dcp is consistent with the peak at 9,439.67 Da. Peaks at the position of normal ACP and Dcp were not observed in the N7-Pan-treated extract. Gel electrophoresis (inset) of the protein fraction showed two major low molecular weight bands that were identified as indicated by MALDI-TOF/TOF sequencing as described under "Experimental Procedures."B, metabolism of the pantothenamides to CoA analogs and prosthetic group transfer to the carrier proteins (CP) ACP or Dcp catalyzed by AcpS. The covalent modification of ACP and Dcp with ethyldethia-CoA (N5-CoA) or butyldethia-CoA (N7-CoA) results in inactivation of these proteins. increased acid sensitivity (47). Another important difference between S. aureus and E. coli is the role of CoA in cellular metabolism. Maintaining the reducing environment in the cell and detoxifying reactive oxygen species is not critical for survival in the laboratory, but they are important in the survival of the bacteria in the oxidative environment within animal hosts. This idea has been suggested previously (42)(43)(44), but the importance of this redox system in pathogenicity has not been directly tested.
In summary, the pantothenamide antimetabolites are substrates for SaCoaA and are incorporated into downstream targets ACP and Dcp via their CoA analogs (Fig. 7B). The proximal cause for cell growth inhibition is the blockade of fatty acid synthesis because of the accumulation of inactive ACP, but the inhibition of D-alanine incorporation into the cell wall because of the inactivation of Dcp and the elimination of the CoA-dependent intracellular redox system are two additional significant cellular processes that are compromised by the pantothenamides. These findings are in sharp contrast to a previous report by Choudhry et al. (15), who conclude that N7-Pan and N5-Pan are inhibitors of SaCoaA but are not substrates for this enzyme. It is likely that the indirect spectrophotometric assay employed by these investigators was not sensitive enough to detect N7-Pan phosphorylation and that they did not investigate its metabolism in intact cells.