acyl carrier protein synthases from gram-negative, gram ...envismadrasuniv.org/physiology/pdf/acyl...

JOURNAL OF BACTERIOLOGY, July 2006, p. 4737–4748 Vol. 188, No. 130021-9193/06/$08.00�0 doi:10.1128/JB.01917-05Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Acyl Carrier Protein Synthases from Gram-Negative, Gram-Positive,and Atypical Bacterial Species: Biochemical and Structural

Properties and Physiological ImplicationsKelly A. McAllister, Robert B. Peery, and Genshi Zhao*

Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285

Received 15 December 2005/Accepted 21 April 2006

Acyl carrier protein (ACP) synthase (AcpS) catalyzes the transfer of the 4�-phosphopantetheine moiety fromcoenzyme A (CoA) onto a serine residue of apo-ACP, resulting in the conversion of apo-ACP to the functionalholo-ACP. The holo form of bacterial ACP plays an essential role in mediating the transfer of acyl fatty acidintermediates during the biosynthesis of fatty acids and phospholipids. AcpS is therefore an attractive targetfor therapeutic intervention. In this study, we have purified and characterized the AcpS enzymes fromEscherichia coli, Streptococcus pneumoniae, and Mycoplasma pneumoniae, which exemplify gram-negative, gram-positive, and atypical bacteria, respectively. Our gel filtration column chromatography and cross-linkingstudies demonstrate that the AcpS enzyme from M. pneumoniae, like E. coli enzyme, exhibits a homodimericstructure, but the enzyme from S. pneumoniae exhibits a trimeric structure. Our biochemical studies show thatthe AcpS enzymes from M. pneumoniae and S. pneumoniae can utilize both short- and long-chain acyl CoAderivatives but prefer long-chain CoA derivatives as substrates. On the other hand, the AcpS enzyme from E.coli can utilize short-chain CoA derivatives but not the long-chain CoA derivatives tested. Finally, ourbiochemical studies show that M. pneumoniae AcpS is kinetically a very sluggish enzyme compared with thosefrom E. coli and S. pneumoniae. Together, the results of these studies show that the AcpS enzymes from differentbacterial species exhibit different native structures and substrate specificities with regard to the utilization ofCoA and its derivatives. These findings suggest that AcpS from different microorganisms plays a different rolein cellular physiology.

Fatty acids, essential components of bacterial membranelipids and lipopolysaccharides, are synthesized via a pathwayconsisting of enzymes catalyzing a repeated cycle of conden-sation, reduction, dehydration, and reduction reactions (8, 38,40). In these reactions, holo-acyl carrier protein (holo-ACP)plays an essential role as an acyl carrier for fatty acid precur-sors, growing acyl intermediates, and nascent fatty acid prod-ucts (8, 11, 29, 38, 40). ACP is a small, abundant, acidic proteinin bacteria (8, 38). The newly synthesized ACP, or apo-ACP, isnot functional in fatty acid biosynthesis. The conversion ofapo-ACP to the functional holo-ACP is mediated by acyl car-rier protein synthase (AcpS), which transfers the 4�-phospho-pantetheine moiety from coenzyme A (CoA) onto a serineresidue of apo-ACP, resulting in the conversion of apo-ACP toholo-ACP (8, 11, 13, 28, 29, 32, 38). The holo-ACP formedthen mediates the transfer of acyl intermediates by the cova-lent attachment of all acyl intermediates via their carboxylgroup to the thiol group of the 4�-phosphopantetheine groupof holo-ACP (8, 11, 13, 28, 29, 31, 32, 38, 40). Thus, AcpS alsoplays an essential role in fatty acid biosynthesis.

Homologues of AcpS and ACP have been identified in allbacterial and Mycoplasma genomes sequenced to date (3, 6, 18,19, 24, 25, 42, 43). The first biochemical identification of thegene product of an unknown dpj gene as AcpS in Escherichiacoli (29) is significant. This landmark work has led to the

subsequent identification of a number of AcpS-like enzymesfrom different bacterial species, which are required for thebiosynthesis of polyketides, enterobactin siderophore, and oth-ers, and the demonstration of the potential cross-functionalityof ACP and AcpS in different biosynthetic systems (4, 9, 12–14,26, 28, 29, 32, 36, 44, 48). The E. coli acpS gene encodes asmall, highly basic protein with a molecular mass of 14 kDa(27, 29). Streptococcus pneumoniae and Bacillus subtilis eachalso possess an AcpS enzyme with a similar size (5, 32, 33).Both E. coli and S. pneumoniae enzymes have been purifiedand characterized (13, 29, 32). Purified E. coli AcpS appears tobe a homodimer (29), whereas S. pneumoniae AcpS is a ho-motrimer as revealed by gel filtration column chromatographyand cross-linking studies (32). The trimeric structure of S.pneumoniae enzyme was also confirmed by the x-ray crystal-lography (5). The x-ray crystallography studies of B. subtilis andStaphylococcus aureus enzymes also demonstrate a homotri-meric structure (10, 35). Thus, AcpS enzymes appear to exhibitdifferent native structures. AcpS enzymes from different bac-terial species can also utilize ACPs that are required in otheraspects of cellular metabolism (4, 9, 13, 26, 28, 29, 31, 32, 44,48). The broad substrate specificity of AcpS suggests that theseenzymes may be able to participate in other types of cellularmetabolism beside fatty acid biosynthesis in the cell.

E. coli and S. pneumoniae AcpS enzymes have been charac-terized (5, 13, 29, 32), yet their substrate specificities withregard to the utilization of CoA derivatives have not beensystematically examined. It is also known that the lipid com-position of the bacterial membrane varies among differentspecies (8, 38). In addition, Mycoplasma species, unlike gram-

* Corresponding author. Mailing address: Lilly Research Laborato-ries, Cancer Research, DC 0434, Eli Lilly and Company, Indianapolis,IN 46285. Phone: (317) 276-2040. Fax: (317) 276-1414. E-mail: [email protected].

4737

negative and gram-positive bacteria, do not possess a fatty acidbiosynthetic pathway, and their source of fatty acids is exclu-sively derived from their immediate environment throughtransport (18, 30). Finally, a hypothetical pathway to lipidbiosynthesis was proposed for Mycoplasma on the basis ofsequence comparisons with the known lipid biosynthetic en-zymes present in other bacterial species (18). However, thisproposed pathway has not been tested. Therefore, it is notclear whether the function of AcpS in different bacterial spe-cies is required only for fatty acid biosynthesis, lipid biosyn-thesis, or both. It is also not clear whether AcpS plays anessential role in the acylation of fatty acids derived from theenvironment before being incorporated into the membrane,especially in Mycoplasma pneumoniae and S. pneumoniae, sincethey are the major human pathogens of the respiratory tract.To further understand the physiological function of AcpS en-zymes in different bacterial species and to assess their structureand activity relationship, we have purified the AcpS enzymes ofE. coli, M. pneumoniae, and S. pneumoniae and characterizedthese enzymes with regard to their native structures and sub-strate specificities regarding the utilization of CoA derivatives.We show that the AcpS enzymes from different bacterial spe-cies exhibit different native structures and substrate specifici-ties. These findings suggest that these AcpS enzymes play adifferent role in fatty acid and lipid biosyntheses in these or-ganisms.

MATERIALS AND METHODS

Materials. Unless specified otherwise, fine chemicals were purchased fromSigma Chemical Company (St. Louis, Mo.). Fast-protein liquid chromatographyresins and columns used for purification were purchased from Amersham Bio-sciences (Piscataway, NJ). Expression vectors and expression strains were pur-chased from Novagen (Madison, WI). Luria-Bertani (LB) broth medium waspurchased from Bio 101, Inc. (Vista, CA). Polyacrylamide gels and reagents werepurchased from Invitrogen (Carlsbad, CA). The Bradford protein assay reagentwas purchased from Bio-Rad (Hercules, CA), and ethylene glycolbis (succinimi-dylsuccinate) (sulfo-EGS) was purchased from Pierce (Rockford, IL).

Cloning and expression of the acpS and acpP genes. Cloning and expression ofthe S. pneumoniae acpS and acpP genes have been described (32). The cloningand expression of the acpP and acpS genes from M. pneumoniae were carried outas described before (47). All reagents, plasmids, and cell lines used for cloningand expression were the same as those described before (47). To clone the acpSgene, the following PCR primers were designed and used to amplify the acpSgene for cloning into E. coli expression systems. The 5� PCR primer (5�-CGCGGATCCCATATGATTCTA GGCATAGGGATTGATTTAGTC-3�) was de-signed at the ATG start codon of acpS and contains BamHI and NdeI sites forcloning purposes. The 3� PCR primer (5�-CGCGGATCCTCATGGTGTTTGTTGTGCCAAACAGATGGC-3�) was designed at the stop codon of acpS andcontains a BamHI site after the stop codon. Using these primers, acpS was PCRamplified from M. pneumoniae for 25 cycles under the conditions as describedbefore (47). Five PCR products were combined, and a portion of the pooledPCR products was digested with BamHI. The BamHI-digested PCR fragmentwas cloned into pCZA342, a low-copy-number plasmid (2) that had been di-gested with BamHI and dephosphorylated with calf intestinal alkaline phos-phatase. The acpS gene from several pCZA342 clones was sequenced, and aclone containing the consensus acpS gene sequence was used for constructingexpression systems. This pCZA342 clone was digested with NdeI and BamHI.The NdeI-BamHI DNA fragment containing acpS was subcloned into pET-1la(Novagen). The resulting plasmid was designated pLY270.

To clone the acpP gene from M. pneumoniae, the following PCR primers wereused for amplification: the 5� PCR primer (5�-CGCGGATCCCATATGCAAGAGCGTGACATTC-3�) and the 3� PCR primer (5�-CGCGGATCCCTATACCCCTTTTTGACTTA TTA-3�). Using these primers, acpP was PCR amplifiedfrom M. pneumoniae as described above. The PCR products were digested andcloned into pCZA342 and finally pET-11a (Novagen) exactly as described above.The resulting plasmid was designated pLY368.

To clone the acpS gene from E. coli, the following PCR primers were used foramplification: the 5� PCR primer (5�-CGCGGATCCCATATGAGCACCATCGAAGAACGTGTGAAAAAA-3�) and the 3� PCR primer (5�-CGCGGATCCTTACGCCTGGTTT CCGTTAATATAGACAAT-3�). Using these primers,acpP was PCR amplified from E. coli as described above. The PCR productswere digested and cloned into pCZA342 and finally pET-11a (Novagen) exactlyas described above. The resulting plasmid was designated pLY296.

For the expression of M. pneumoniae ACP and AcpS and E. coli AcpS, theexpression plasmids were transformed into E. coli BL21 pLysS as describedbefore (47), and the resulting E. coli expression strains were designated LY368,LY270, and LY296, respectively.

For the expression of E. coli ACP, DK554, an E. coli ACP overexpresser strain,was used as described previously (29).

Purification of the AcpS and apo-ACP proteins of E. coli, M. pneumoniae, andS. pneumoniae. The S. pneumoniae AcpS and apo-ACP proteins were expressedand purified as described previously (32).

For the purification of E. coli AcpS, an E. coli expression strain (LY296)carrying the acpS gene on an expression plasmid as described above was inocu-lated from an overnight culture into LB with 100 �g/ml ampicillin and grown at35°C with shaking at 250 rpm until an optical density of 0.5 to 0.6 at 590 nm wasreached. The culture was induced with 1 mM isopropyl-1-thio-�-D-galactopyra-noside for 3 h. Cells were harvested by centrifugation at 4,500 � g at 4°C for 8min, washed twice in phosphate-buffered saline, and frozen at �80°C. The cellpellet was thawed and resuspended in 50 mM Tris-HCl, pH 7.0, and 100 mM KCl(buffer A) and disrupted by passing it twice through a French press cell. Theresulting cell lysate was centrifuged at 160,000 � g for 40 min at 4°C. Thesupernatant fraction was collected and applied to a Source S column (15S, 2.5 by8 cm) that had been equilibrated with buffer A. The column was washed withbuffer A and eluted with a linear gradient of 0.1 to 1.0 M KCl in buffer A.Fractions (7.0 ml each) were collected, and the presence of AcpS in the fractionswas confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis(SDS-PAGE) analysis (16% Tricine gels). The fractions containing AcpS werepooled, adjusted to final concentrations of 1 mM dithiothreitol (DTT) and 15%glycerol (vol/vol), and stored in small aliquots at �80°C. Protein concentrationswere determined using a Bradford protein assay kit (Bio-Rad) with bovine serumalbumin as a standard.

For the purification of E. coli ACP, an E. coli ACP overexpression strain,DK554, was grown in LB containing glucose (50 mM), kanamycin (50 �g/ml),and pantothenate (25 mM) at 30°C until an optical density of 0.6 at 590 nm wasreached (29). The cells were induced using 1 mM isopropyl-1-thio-�-D-galacto-pyranoside for 3 h, harvested by centrifugation, and washed twice in phosphate-buffered saline prior to storage at �80°C. Cell pellets were thawed and resus-pended in 50 mM Tris-HCl, pH 7.0, and 50 mM KCl (buffer B), and disruptedby passage through a French pressure cell as described above. The resulting celllysate was centrifuged as described above, and the supernatant fraction collectedwas applied to a Source Q column (30 ml) equilibrated with buffer B. Fractions(7 ml each) containing apo-ACP, as judged by SDS-PAGE (4 to 20% Tris-glycine) analysis, were pooled and applied to a Sepharose (900 ml, S-100) sizeexclusion column preequilibrated with 50 mM Tris-HCl, pH 8.0, and 100 mMKCl. Fractions containing apo-ACP were pooled, adjusted to a final concentra-tion of 15% glycerol, and stored frozen at �80°C. Protein concentration wasdetermined as described above.

For the purification of M. pneumoniae AcpS and apo-ACP, E. coli cells (LY270and LY368 carrying M. pneumoniae AcpS and apo-ACP genes on expressionplasmids, respectively) were grown and induced as described above. Cells wereharvested and disrupted in buffer B as described above. The supernatant fractioncontaining M. pneumoniae AcpS was applied to a Source S column (15S, 2.5 by8 cm), whereas the supernatant fraction containing M. pneumoniae apo-ACP wasapplied to a Source Q column (15S, 2.5 by 8 cm), both of which had beenequilibrated in buffer B. The columns were washed with buffer B and eluted witha linear gradient of 0.1 to 1.0 M KCl in buffer B. Fractions (7 ml each) werecollected, and the presence of AcpS and apo-ACP was determined by SDS-PAGE (16% Tricine gels) analysis. The fractions containing AcpS were col-lected, adjusted to a final concentration of 15% glycerol (vol/vol), and frozen at�80°C. The fractions containing M. pneumoniae apo-ACP protein were furtherpurified by using a Sepharose gel filtration column (S-100, 5 by 60 cm) equili-brated in 50 mM Tris-HCl, pH 8.0, and 50 mM KCl. The column was eluted withthe same buffer, and fractions (10 ml each) were collected. Fractions containingapo-ACP were pooled, adjusted to a final concentration of 15% glycerol (vol/vol), and frozen at �80°C. Protein concentrations were determined as describedabove.

The identity of each purified ACP or AcpS protein preparation was confirmedby N-terminal sequencing analysis as described before (47). The identity of each

4738 MCALLISTER ET AL. J. BACTERIOL.

purified ACP preparation was also confirmed by electrospray mass spectropho-tometry (ESMS) analysis as described before (32). In addition, each purifiedACP preparation was also subjected to ESMS analysis to verify their primarypresence in the apo form.

Determination of the native structures of purified AcpS proteins. For thedetermination of the native structure of each purified AcpS protein, analyticalgel filtration column chromatography was carried out. A sample of 100 �l of apurified AcpS preparation (�300 �g/ml) was applied to a Superdex gel filtrationcolumn (S-75, HR 1.0 by 30 cm) that was equilibrated with 50 mM Tris-HCl, pH7.5, 50 mM KCl, and 1 mM DTT. The column was calibrated with proteinmolecular weight standards (Sigma). On the basis of the elution profiles of themolecular weight standards, a standard curve was generated, which was used tocalculate the molecular weight of each protein according to its elution profiles.

To further verify the native oligomerization state of AcpS, each purified AcpSpreparation was subjected to a cross-linking analysis as described below. EachAcpS protein preparation was first dialyzed against 4 liters of 20 mM HEPES, pH7.5, and 1 mM DTT at 4°C for 18 h. Then, 100 �l of each protein preparation (50�M) was mixed with 0.5, 2.5, or 5.0 mM of sulfo-EGS and the reaction mixtureswere incubated at room temperature for 30 min. The reactions were terminatedby the addition of 50 mM Tris-HCl, pH 7.5, at room temperature for 30 min. Thesulfo-EGS-treated and untreated AcpS preparations (10 �l each) were analyzedby SDS-PAGE (10 �l each gel) and stained with Coomassie blue R-250. On thebasis of the relative mobilities of the molecular weight markers (Invitrogen)versus their molecular weights, a standard curve was generated, which was thenused to calculate the molecular weight of sulfo-EGS-treated and untreated AcpSprotein preparations (Invitrogen).

Biochemical characterization of purified AcpS proteins. The ability of thepurified AcpS enzymes to convert their native apo-ACP substrates to holo-ACPwas assessed by ESMS analysis. The reaction mixtures containing 50 mM Tris-HCl buffer, 10 mM MgCl2, 1 mM DTT, 300 �M CoA, 5 �M AcpS, and apo-ACP(100 �M of S. pneumoniae ACP, 47 �M of E. coli apo-ACP, or 120 �M of M.pneumoniae apo-ACP) were incubated at room temperature for 1 h beforeESMS analysis as described previously (32).

The conversion of apo-ACP to acyl-ACP in the presence of different CoAderivatives was monitored by a high-pressure liquid chromatography (HPLC)-based assay (32) that was used for subsequent enzyme characterization andkinetic analysis. Reaction mixtures (100 �l each) were injected onto a reverse-phase column (Vydac Selectapore 300M) and separated by a 25 to 100% ace-tonitrile gradient containing 0.1% trifluoroacetic acid (31). Acyl-ACP productswere detected by their absorption at 220 nm, and the product formation wasdetermined by comparing the peak areas of the acyl-ACP (products) to those ofthe substrate (apo-ACP) peak and the product peak.

For the determination of the kinetic parameters (Km and kcat) of AcpS, theactivity of each enzyme was assayed using an HPLC method (32) at a fixedconcentration of one substrate and various concentrations of the other as de-scribed below. For the determination of the kinetic parameters of AcpS withrespect to CoA, each enzyme was assayed at a fixed concentration of the apo-ACP (1 �M) and various concentrations of CoA (0.5 to 100 �M). For thedetermination of the kinetic parameters of AcpS with respect to apo-ACP, eachenzyme was assayed at a fixed concentration of the CoA (100, 40, and 20 �M forE. coli, M. pneumoniae, and S. pneumoniae AcpS, respectively) and variousconcentrations of apo-ACP (0 to 20 �M). The concentrations of apo-ACP usedfor each AcpS enzyme were determined on the basis of their lack of significantinhibition of each enzyme activity. In all cases, the reaction mixtures containing50 mM Tris-HCl at different optimal pHs (8.0, 7.2, and 7.0 for E. coli, M.pneumoniae, S. pneumoniae AcpS enzymes, respectively), 10 mM MgCl2, and 1mM DTT were incubated at 37°C for 1 h, in quadruplicate, and the reactionswere terminated by the addition of 50 mM EDTA.

To examine the substrate specificity of AcpS enzymes, the activity of eachenzyme was assayed in the presence of a fixed concentration of their nativeapo-ACP and different concentrations of CoA derivatives: CoA, acetyl-CoA,malonyl-CoA, acetoacetyl-CoA, butyryl-CoA, crotonyl-CoA, decanoyl-CoA,myristoyl-CoA, and palmitoleoyl-CoA. The ability of S. pneumoniae AcpS toutilize myristoleoyl-CoA and palmitoyl-CoA was also tested. For the determina-tion of optimal activity of each enzyme for CoA derivatives, the activity of eachAcpS was measured in the presence of CoA derivatives and different enzymeconcentrations as follows: 0.001 to 1 �M of E. coli AcpS, 0.1 to 1.0 �M of M.pneumoniae AcpS, and 1.0 to 20 nM of S. pneumoniae AcpS. On the basis ofthese results, the appropriate enzyme concentrations for the assay were deter-mined for the assessment of the kinetic parameters of each enzyme for the CoAderivatives. For the assessment of E. coli AcpS substrate specificity, the enzymeactivity was assayed in the presence of 1 �M of E. coli apo-ACP and 0.5 to 50 �Mof CoA derivatives. For the assessment of M. pneumoniae AcpS substrate spec-

ificity, the enzyme activity was assayed in the presence of 1 �M of M. pneumoniaeapo-ACP and the following concentrations of different CoA derivatives: 0 to 100�M (acetyl-CoA, crotonyl-CoA, butyryl-CoA), 0 to 30 �M (decanoyl-CoA, my-ristoleoyl-CoA, palmitoleoyl-CoA), and 0 to 250 �M (malonyl-CoA, acetoacetyl-CoA). For the assessment of S. pneumoniae AcpS substrate specificity, theenzyme activity was assayed in the presence of 1 �M S. pneumoniae apo-ACPand the following concentrations of different CoA derivatives: 0 to 50 �M exceptfor malonyl-CoA, of which the enzyme required much higher concentrations (0to 400 �M) for activity. From the progress curves generated, kinetic parameters(Km and kcat) were calculated by fitting the curves obtained to the Michaelis-Menten equation using Sigma-Plot.

RESULTS

Cloning and expression of the acpS and acpP genes of E. coliand M. pneumoniae. To further understand the physiologicalrole of AcpS in gram-negative (E. coli), gram-positive (S. pneu-moniae), and also atypical (M. pneumoniae) bacteria and toexamine their structure and activity relationship, we wanted tocharacterize the AcpS enzymes from these phylogeneticallywell-spaced and evolutionarily diverse organisms with regardto their native structures and substrate specificities. Since theacpS gene from M. pneumoniae has not been identified andcharacterized, we first searched the acpS and also acpP genesfrom the genome of M. pneumoniae by using the S. pneumoniaeand E. coli acpS and acpP gene sequences as queries in theBLAST program (1). We identified a gene consisting of 360base pairs, which encodes a protein with a predicted molecularmass of 13,774 Da (accession number NP_109986). As shownnext, we have confirmed the identity of this gene as acpS. TheM. pneumoniae acpS gene appears to be clustered with thegenes in the order of unknown-fmt-acpS, since rnc and rpsT,located downstream and upstream to the acpS cluster, respec-tively, are transcribed from the opposite directions with respectto the acpS cluster (18). Thus, the acpS cluster appears toconsist of fmt, involved in protein synthesis (methionyl-tRNAformyltransferase), and a gene of unknown function. In thisregard, the genomic organization of acpS in M. pneumoniae isquite different from that of acpS in E. coli and S. pneumoniae(27, 32). The acpS gene in E. coli consists of an operon with itsupstream pdxJ gene that is required for vitamin B6 biosynthesis(27, 41). It is also interesting to note that in M. pneumoniae,rnc, a gene encoding RNase III, is located immediately down-stream of the acpS cluster and is transcribed from an oppositedirection (18), whereas in E. coli, rnc is located in an operonimmediately upstream to the acpS operon (27, 41). Thus, rncand acpS appear to be located in the same vicinity on thechromosomes of both organisms. The acpS cluster in S. pneu-moniae, aroG-aroF-acpS-alr-recG, consists of the genes thatare required for aromatic amino acid biosynthesis, cell wallbiosynthesis, and DNA recombination (32).

The acpP gene from M. pneumoniae is 255 base pairs longand encodes a protein consisting of 84 amino acid residueswith a predicted molecular mass of 9,838 Da (accession num-ber S73758). The acpP gene appears to be in a cluster with thegenes in the order of unknown-unknown-acpP-unknown. Thereis a long noncoding region (145 bp) upstream of the first geneof unknown function. In addition, the gene with an unknownfunction located downstream of the acpP cluster is transcribedfrom an opposite direction. Like the acpS gene, the acpP genealso appears to organize into an operon with genes of unknownfunctions.

VOL. 188, 2006 BIOCHEMICAL AND STRUCTURAL PROPERTIES OF AcpS ENZYMES 4739

The subunits of M. pneumoniae AcpS and ACP exhibit mo-lecular weights similar to those of their counterparts in S.pneumoniae and E. coli. The two proteins, AcpS and ACP,share 27 and 33% and 25 and 32% identities with their respec-tive counterparts in E. coli and S. pneumoniae. The pI value ofM. pneumoniae AcpS is estimated to be 9.98, which is identicalto that of E. coli AcpS but much higher than that (6.13) of S.pneumoniae AcpS (32). Thus, M. pneumoniae AcpS, like E. coliAcpS, is significantly more basic than that of S. pneumoniae(27, 29, 32). Like other ACPs (8, 26, 34, 37–39), M. pneumoniaeapo-ACP is very acidic, with a pI value of 4.82.

The acpS and acpP genes from M. pneumoniae and E. coliwere cloned by PCR methodology and expressed in E. coli byusing expression vectors (see Materials and Methods). TheacpS and acpP genes from S. pneumoniae were cloned andexpressed in E. coli as described previously (32).

Purification and identification of the AcpS and ACP pro-teins of M. pneumoniae, S. pneumoniae, and E. coli. The puri-fication and identification of S. pneumoniae AcpS and apo-ACP have been described elsewhere (32). To purify E. coli andM. pneumoniae AcpS proteins, we overexpressed both proteinsin E. coli and purified them to approximately 85% pure by asingle step of anion-exchange column chromatography (datanot shown). Using this purification scheme, we obtained ap-proximately 50 and 13 mg AcpS proteins of E. coli and M.pneumoniae, respectively, from one liter of E. coli cells over-expressing the proteins. The purified AcpS proteins of E. coliand M. pneumoniae exhibited molecular masses of 15 kDa,which were very similar to the predicted values of 14.1 and13.8 kDa, respectively, on the basis of their amino acidsequences (27) (accession number NP_109986). To furtherconfirm the identity of each purified protein, we determinedtheir N-terminal sequences. The N-terminal amino acid se-quences obtained for E. coli and M. pneumoniae AcpS pro-teins were AILGLGTDIV and MILGIGIDLV, respectively,which were identical to the predicted amino acid sequencesexcept that the first Met residue of E. coli AcpS was pro-cessed. Thus, these results have confirmed the identities ofpurified proteins as E. coli and M. pneumoniae AcpS en-zymes.

We also overexpressed M. pneumoniae apo-ACP protein inE. coli and purified both E. coli and M. pneumoniae apo-ACPproteins to apparent homogeneity, as judged by SDS-PAGEanalysis, using a two-step purification scheme consisting ofanion-exchange and gel filtration chromatography (data notshown). Using this scheme, we obtained approximately 23 and50 mg of E. coli and M. pneumoniae apo-ACP proteins, respec-tively, from one liter of E. coli cells overexpressing the pro-teins. To confirm the identities of both proteins, we carried outN-terminal amino acid sequencing and ESMS analyses. TheN-terminal amino acid sequences obtained for E. coli and M.pneumoniae apo-ACP proteins were STIEERVKKIxG andMQERDILLKIKE, respectively, which were identical to thepredicted amino acid sequences except that the first Met res-idue of E. coli ACP was processed and the identities of theresidue between Lys and Gly could not be established due to aweak signal (data not shown). Furthermore, ESMS analysisshowed that the purified ACP proteins of E. coli and M. pneu-moniae exhibited molecular masses of 8,508 and 9,838 Da,respectively, which matched exactly the predicted molecular

masses of both proteins on the basis of their amino acid se-quences (accession numbers AAA24316 and S73758, respec-tively). Thus, the results of these studies have confirmed theidentities of the purified proteins as E. coli and M. pneumoniaeACP proteins.

To establish that the purified E. coli and M. pneumoniaeAcpS proteins were enzymatically active, we performed ESMSanalysis. We found that the molecular mass of E. coli apo-ACPwas increased by 340 Da (from 8,508 to 8,848.6 Da) after theincubation with E. coli AcpS (data not shown). This increase inthe molecular mass of ACP indicates that the transfer of thephosphopantetheine group from CoA to apo-ACP occurred,since the phosphopantetheine group is known to have a mo-lecular mass of approximately 340 Da. Similarly, we found thatthe molecular mass of M. pneumoniae apo-ACP was increasedby approximately 340 Da (from 9,838 to 10,178 Da) after theincubation with M. pneumoniae AcpS (data not shown). Thisincrease in molecular mass again indicates a transfer of thephosphopantetheine group from CoA to apo-ACP. Finally, wehave shown previously that S. pneumoniae apo-ACP exhibiteda molecular mass increase of approximately 340 Da (from8,834 to 9,174 Da) after the incubation with S. pneumoniaeAcpS (32). Together, these results have clearly established thatthe purified AcpS enzymes are enzymatically active.

Determination of native molecular structures of E. coli andM. pneumoniae AcpS enzymes. To better understand the struc-ture and activity relationship of AcpS enzymes, we wanted todetermine the native structures of each purified enzyme. Wehave previously shown that the AcpS enzyme from S. pneu-moniae is a homotrimeric enzyme (5, 32). Interestingly, E. coliAcpS appeared to be a dimeric enzyme as indicated by gelfiltration column chromatography (29). To further confirm thisfinding and to determine the native structure of M. pneumoniaeAcpS, we subjected both purified proteins to analytical gelfiltration column chromatography (Materials and Methods).Since we found that DTT was important for E. coli AcpSactivity, we included it in all the subsequent experiments (32).As a control, we also subjected S. pneumoniae AcpS to gelfiltration column chromatography (Materials and Methods).Under this condition, we found that S. pneumoniae AcpS waseluted in the fractions corresponding to a molecular mass of 39kDa (data not shown), consistent with our previous finding thatthis enzyme is homotrimeric (i.e., 13-kDa/monomer). How-ever, we found that both the E. coli and M. pneumoniae AcpSproteins were eluted in the fractions corresponding to a mo-lecular mass of 29 kDa (data not shown) when subjected toanalysis under identical conditions. These results indicate ahomodimeric structure (13.8 to 14.1 kDa/monomer). To fur-ther confirm these results, we performed cross-linking analysesusing sulfo-EGS as a cross-linker (Materials and Methods).We found that a higher-molecular-mass protein species (34kDa) was present in the S. pneumoniae AcpS preparation afterthe treatment with sulfo-EGS (Fig. 1). This 2.6-fold (from 13.3to 34 kDa) increase in the molecular mass of S. pneumoniaeAcpS indicates a trimeric structure. Similarly, we found thathigher-molecular-mass protein species (32 and 27 kDa) werepresent in the E. coli and M. pneumoniae AcpS preparation,respectively, after the treatment with the cross-linker (Fig. 1).This increase in the molecular mass of E. coli AcpS (15 to 32kDa) or M. pneumoniae AcpS (15 to 27 kDa) indicates a


dimeric structure. Together, the results of these studies haveconfirmed the previous findings regarding the native structuresof E. coli and S. pneumoniae AcpS enzymes and have alsosuggested that M. pneumoniae AcpS is a dimeric enzyme.

Utilization of CoA and CoA derivatives as substrates by E.coli, M. pneumoniae, and S. pneumoniae AcpS enzymes. Tofurther assess the physiological role of AcpS in different bacterialspecies, we wanted to examine the abilities of AcpS enzymes toutilize CoA and its derivatives as physiological substrates for fattyacid and lipid biosyntheses in the cell (Fig. 2). To this end, wemeasured the activity of each AcpS enzyme in the presence ofCoA and its acyl derivatives using an HPLC assay previouslydeveloped for the S. pneumoniae AcpS enzyme (32) and de-termined the kinetic properties of each enzyme with respect tothe utilization of CoA and its derivatives. To utilize this HPLCassay for the assessment of a variety of CoA derivatives assubstrates for each AcpS enzyme, we optimized the assay,which resulted in enhanced separations of the longer-chainacyl-ACPs for each enzyme and also apo-ACPs from acyl-ACPs for all CoA derivatives tested except hexanoyl-CoA andoctanoyl-CoA, whose products could not be separated fromthe apo-ACP substrates (data not shown). As a result, thesetwo CoA derivatives could not be analyzed. For all AcpS en-zymes tested, their activities appeared to increase in a dose-dependent manner within a certain range of apo-ACP concen-trations but then appeared inhibited at higher apo-ACPconcentrations. Similar to previously reported results (32), theactivity of S. pneumoniae AcpS increased with the increase ofapo-ACP concentrations until 7.5 �M but then decreased tosome degree at 7.5 to 10 �M (Fig. 3A). At �10 �M of apo-ACP, the activity appeared to increase gradually again (Fig.3A). A similar result was obtained for M. pneumoniae AcpS(Fig. 3C). The activity of this enzyme was inhibited to somedegree at 3 �M of apo-ACP but then increased with a furtherincrease in apo-ACP concentrations (Fig. 3C). In contrast tothe results obtained for M. pneumoniae and S. pneumoniaeenzymes, the activity of the E. coli AcpS enzyme was inhibitedto some degree at 3 to 10 �M of apo-ACP and then completelyinhibited at 20 �M of apo-ACP (Fig. 3B). Due to the inhibition

of AcpS activities by apo-ACP at higher concentrations, thekinetic studies were carried out for each enzyme at the con-centrations of apo-ACP which did not significantly inhibitAcpS activity (Materials and Methods).

All the AcpS enzymes tested appeared to exhibit the Michaelis-Menten kinetics when assayed at different concentrations of CoAor its derivatives and fixed concentrations of apo-ACP at whichthe enzyme activity was not inhibited (Fig. 4B, D, and F; Fig. 5).The Km values of the AcpS enzymes of S. pneumoniae, E. coli,and M. pneumoniae for CoA were determined to be 4.0, 9.3,and 40 �M, respectively (Table 1). These results indicate thatthe affinity of M. pneumoniae AcpS for CoA is much lower thanthose of the E. coli and S. pneumoniae enzymes. The Km valuesof the AcpS enzymes of S. pneumoniae, E. coli, and M. pneu-moniae for apo-ACP were determined to be 1.8, 1.3, and 0.8�M, respectively (Table 1). Thus, the affinities of these en-zymes for apo-ACP are similar. The kcat values obtained for S.pneumoniae and E. coli AcpS enzymes with respect to CoA andapo-ACP were similar but were 150- to 500-fold higher thanthat obtained for M. pneumoniae AcpS (Table 1). Taken to-gether, these results indicate that the AcpS enzyme of M.pneumoniae is kinetically a sluggish enzyme compared withthose of S. pneumoniae and E. coli with regard to the utilizationof their native apo-ACP and CoA substrates.

To assess the ability of each enzyme to utilize CoA deriva-tives as their physiological substrates, we determined the Km

and kcat values of each purified enzyme in the presence ofapo-ACP and different CoA derivatives (Table 1). As shown inTable 1, S. pneumoniae AcpS can utilize short- and long-chainCoA derivatives but prefers the long-chain acyl CoA-deriva-tives (�C10) as evidenced by the lower Km values obtained forthe long-chain CoA derivatives when tested as substrates (Ta-ble 1; Fig. 6, top panel). In contrast, E. coli AcpS can utilizeonly the short-chain CoA derivatives (�C4) but not the long-chain CoA-derivatives (�C4) under the conditions tested (Ta-ble 1; Fig. 6, middle panel). Interestingly, M. pneumoniaeAcpS, like S. pneumoniae AcpS, can utilize both short- andlong-chain CoA derivatives but significantly prefers the long-chain CoA derivatives (Table 1; Fig. 6, bottom panel). The

FIG. 1. Analysis of AcpS native structures by chemical cross-linking. Purified AcpS (50 �M) was treated with 0.5, 2.5, or 5 mM of sulfo-EGSor left untreated, and the resulting AcpS preparations were analyzed by SDS-PAGE as described (see Materials and Methods). Lane M, proteinmolecular mass markers; lanes 1 to 3, purified S. pneumoniae AcpS protein treated with 0.5, 2.5, or 5 mM sulfo-EGS, respectively; lanes 4 to 6,purified E. coli AcpS treated with 0.5, 2.5, or 5 mM sulfo-EGS, respectively; lanes 7 to 9, purified M. pneumoniae AcpS treated with 0.5, 2.5, or5 mM sulfo-EGS, respectively. The arrows point to the cross-linked AcpS species.


affinities of M. pneumoniae AcpS for the long-chain acyl CoAderivatives are significantly higher than those obtained for theshort-chain derivatives, although the kcat values obtained forthe enzyme remained similar. Therefore, these results showthat the AcpS enzymes from these phylogenetically diversebacterial species exhibit significantly different substrate speci-ficities with respect to the utilization of CoA and its derivatives.

DISCUSSION

In this study, we have identified the acpS and acpP genesfrom M. pneumoniae and purified and characterized their geneproducts. We have also compared the kinetic properties andnative structure of this M. pneumoniae AcpS enzyme withthose of the enzymes from E. coli and S. pneumoniae. Theresults of our studies show that the AcpS enzymes from dif-ferent organisms exhibit different native structures and kineticproperties with regard to their utilization of CoA and its de-

rivatives. The results of these studies suggest that AcpS fromdifferent bacterial species may play a different role in cellularphysiology.

The AcpS enzymes from E. coli and S. pneumoniae werepurified and characterized (13, 29, 32). The ability of E. coliAcpS to utilize other types of ACP substrates has been studied(4, 9, 13, 14, 26, 33, 44, 48). The broad substrate specificity ofAcpS with respect to the utilization of different ACPs suggestsit may be required for the cross-functionality of different bio-synthetic systems. Studies have also suggested that E. coli andS. pneumoniae enzymes can utilize some different CoA deriv-atives tested as substrates (14, 32). However, comprehensivecomparative studies of AcpS enzymes with regard to theirabilities to utilize CoA derivatives with different lengths ofcarbon chains, especially those from phylogenetically diversemicroorganisms such as E. coli, S. pneumoniae, and M. pneu-moniae, have not been reported. The findings that the AcpSenzymes from M. pneumoniae and S. pneumoniae prefer long-

FIG. 2. The structures of CoA and its derivatives used as substrates of S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes.


chain acyl CoA derivatives but the enzyme from E. coli canonly utilize short-chain acyl CoA derivatives are surprising andmay have significant physiological implications. First, as M.pneumoniae does not possess a pathway for fatty acid biosyn-thesis, the source of fatty acids is exclusively derived from its

host tissues via transport (18, 30). How M. pneumoniae incor-porates fatty acids imported from its host into lipids is notclear. A hypothetical pathway to lipid biosynthesis has beenproposed for Mycoplasma on the basis of sequence compari-sons (18). This proposed pathway, however, has not been

FIG. 3. Effects of apo-ACP concentrations on activities of S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes. The activity of each AcpSenzyme was measured using an HPLC method in the presence of a fixed CoA and various apo-ACP concentrations (see Materials and Methods).Panels A, B, and C represent the substrate saturation curves obtained for S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes, respectively.


tested yet. The pathway involves the use of acyl CoA fatty acidsfor lipid biosynthesis (18, 30). It is not clear whether the acylACP fatty acid intermediates are used for lipid biosynthesis inthis organism (18, 30). The broad substrate specificity of M.

pneumoniae AcpS for CoA derivatives and its preference forlong-chain CoA derivatives suggest that the biosynthesis oflipids in this organism may be derived from acyl ACP inter-mediates that are generated from imported fatty acids and

FIG. 4. Kinetic analysis of S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes. The activity of each AcpS enzyme was measured usingan HPLC method in the presence of a fixed apo-ACP and various CoA concentrations (panels A, C, and E) or conversely a fixed CoA and variousapo-ACP concentrations (panels B, D, and F) as described in Materials and Methods. Kinetic parameters of each enzyme were generated fromthe substrate saturation curves obtained as described in Materials and Methods. Panels A and B, C and D, and E and F represent substratesaturation curves obtained for the S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes, respectively.


apo-ACP by AcpS. In this regard, AcpS in M. pneumoniae mayplay an important role in the initiation of lipid biosynthesis bygenerating acyl ACP precursors from fatty acids derived fromits host tissues. Thus, this finding suggests a pathway to lipidbiosynthesis via acyl-ACP intermediates in M. pneumoniae.Second, since S. pneumoniae is a human pathogen, it is possiblethat during infection, a significant portion of its nutrients,especially fatty acids, is directly derived from its host tissues.Even though S. pneumoniae possesses the machinery to syn-thesize fatty acids de novo, their biosynthesis is expensive andalso competes directly with the tricarboxylic acid cycle for theacetyl-CoA precursor molecule (7, 16, 17, 19, 42). The use ofthe salvage pathway to import fatty acids from its host tissuesfor lipid biosynthesis may offer a distinct advantage in thesurvival of the organism during infection. Therefore, our find-ing that AcpS from S. pneumoniae exhibits broad substratespecificity with a preference for long-chain fatty acids suggeststhat this enzyme plays an important role in lipid biosynthesisbut maybe also in fatty acid biosynthesis. Finally, the prefer-

FIG. 5. Kinetic analysis of S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes using butyryl-CoA as a substrate. AcpS activity wasmeasured using an HPLC method in the presence of a fixed apo-ACP and various butyryl-CoA concentrations as described in Materials andMethods. Kinetic parameters of each enzyme were generated from the substrate saturation curves obtained as described in Materials and Methods.Panels A, B, and C represent the substrate saturation curves obtained for S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes, respectively.

TABLE 1. Kinetic parameters of E. coli, M. pneumoniae, andS. pneumoniae AcpS enzymes with regard to utilization

of CoA and its derivatives as substrates

Substrate

Value for source of AcpS

E. coli M. pneumoniae S. pneumoniae

Km(�M)

kcat(s�1)

Km(�M)

kcat(s�1)

Km(�M)

kcat(s�1)

Apo-ACP 1.3 0.7 0.8 0.002 1.8 0.6CoA 9.3 1.0 40 0.002 4.0 0.3Acetyl-CoA 7.7 0.4 43 0.009 11 0.2Malonyl-CoA 7.1 0.1 94 0.0004 15 0.1Acetoacetyl-CoA 8.8 0.2 51 0.001 19 0.2Butyryl-CoA 8.1 0.2 35 0.004 7.0 0.3Crotonyl-CoA NADa NAD 5.2 0.005 18.0 0.2Decanoyl-CoA NAD NAD 1.4 0.004 1.5 0.5Myristoyl-CoA NAD NAD 1.8 0.003 1.0 0.5Palmitoyl-CoA NDb ND ND ND 1.2 0.4Myristoleoyl-CoA ND ND ND ND 2.5 0.5Palmitoleoyl-CoA NAD NAD 0.9 0.003 1.6 0.4

a NAD, no activity detected.b ND, not determined.


ence of E. coli enzyme for short-chain CoA derivatives and itsinability to utilize long-chain CoA derivatives suggest that thisenzyme only plays a role in fatty acid biosynthesis in thisorganism.

One of the major differences between the AcpS enzyme ofM. pneumoniae and those of E. coli and S. pneumoniae appearsto be their catalytic efficiencies. The enzymes from E. coli andS. pneumoniae exhibit similar kinetic properties (kcat and Km

for apo-ACP, CoA, and short-chain CoA derivatives). Theaffinity of M. pneumoniae AcpS for apo-ACP is similar to those

of E. coli and S. pneumoniae, but its affinity for CoA is signif-icantly lower than those of the other enzymes. In addition, theturnover number (kcat) of the M. pneumoniae enzyme is 150- to500-fold lower than those of the E. coli and S. pneumoniaeenzymes. Thus, the AcpS enzyme from M. pneumoniae is ki-netically a very sluggish enzyme compared with those from E.coli and S. pneumoniae. The much lower catalytic efficiency ofthe M. pneumoniae AcpS enzyme may be expected, since thisorganism grows much more slowly than E. coli and S. pneu-moniae (18, 30). Finally, the cellular concentrations of apo-

FIG. 6. Catalytic efficiencies of S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes with respect to the utilization of CoA and itsderivatives as substrates. The Km and kcat values for each enzyme were obtained by assaying the activity of each enzyme in the presence of fixedconcentrations of one substrate and various concentrations of another using an HPLC method (see Materials and Methods). The catalyticefficiencies, kcat/Km (M�1 s�1), were calculated on the basis of the Km and kcat values obtained for each enzyme. Panels A, B, and C represent thecatalytic efficiencies of the S. pneumoniae, E. coli, and M. pneumoniae AcpS enzymes, respectively, for CoA and its derivatives, as follows:malonyl-CoA (Mal), acetyl-CoA (ac), acetoacetyl-CoA (acetac), butyryl-CoA (but), crotonyl-CoA (crot), decanoyl-CoA (dec), myristoyl-CoA(myr), palmitoyl-CoA (palm), myristoleoyl-CoA (myre), and palmitoleoyl-CoA (palme).


ACP and CoA in E. coli were estimated to be 400 to 1,800 and90 to 200 �M, respectively (15, 20–23, 32, 45, 46). The affinitiesof the all three enzymes for apo-ACP and CoA appear tocorrelate with the concentrations of these substrates in thecells.

Another major difference between AcpS enzymes fromthese phylogenetically diverse bacterial species appears to liein their native structures. E. coli AcpS was reported to be ahomodimer with a molecular mass of 28 kDa (29). The resultsof our gel filtration and cross-linking studies have further con-firmed this finding. S. pneumoniae AcpS was shown to be ahomotrimer by column chromatography, cross-linking, and x-ray crystallography studies (5, 32). Since the gel filtration col-umn chromatography studies were carried out in the presenceof DTT for the E. coli and M. pneumoniae AcpS enzymes(Materials and Methods), similar experiments were also car-ried out in the presence of DTT for the S. pneumoniae enzyme,whose native structure had been determined previously in theabsence of DTT (32). Under these conditions (see Materialsand Methods), S. pneumoniae AcpS was again found to behomotrimeric, but E. coli and M. pneumoniae enzymes ap-peared to be homodimeric. Thus, these AcpS enzymes exam-ined appear to exhibit different native structures. Since M.pneumoniae is phylogenetically more closely related to gram-positive bacteria than gram-negative bacteria (30) and theAcpS enzymes from B. subtilis and S. aureus, both of which aregram-positive bacteria, were also shown to be homotrimeric(30, 33), it is interesting that the AcpS enzyme from M. pneu-moniae exhibits a dimeric rather than a trimeric structure.Thus, it appears that the AcpS enzymes from gram-negativebacteria and Mycoplasma species exhibit a dimeric structure,but those from gram-positive bacteria exhibit a trimeric struc-ture. Clearly, this possibility can be investigated by determiningthe native structures of AcpS enzymes from other gram-nega-tive and atypical bacterial species.

There does not appear to be a correlation between the AcpSnative structure and its kinetic properties (Km, kcat, and sub-strate specificities). Although the E. coli and S. pneumoniaeAcpS enzymes differ in their native structures, they appear toexhibit similar Km and kcat values for their substrates, apo-ACPand CoA (Table 1). On the other hand, although M. pneu-moniae and E. coli enzymes exhibit the same native structures,they differ significantly in kinetic activity (350- to 500-fold)(Table 1). Although E. coli and M. pneumoniae AcpS enzymesexhibit the same dimeric structures, yet they differ significantlyin their substrate specificities. In addition, M. pneumoniae andS. pneumoniae AcpS enzymes exhibit similar substrate speci-ficities, but they differ in their native structures. Therefore, theenzyme structure does not appear to have much effect on thekinetic properties of the enzymes.

Finally, the critical role that the AcpS enzymes from M.pneumoniae and S. pneumoniae may play in the acylation offatty acids derived from human tissues for their lipid biosyn-thesis suggests that AcpS is a more attractive antimicrobialtarget for discovery of novel antibiotics than bacterial fatty acidbiosynthetic enzymes (16). Since M. pneumoniae and S. pneu-moniae are the major human pathogens of the upper respira-tory tract, fatty acids from human tissues are readily availablefor these organisms. Therefore, during infection, the biosyn-thesis of fatty acids may not be required for the survival of the

organisms. In this regard, targeting fatty acid biosynthetic en-zymes may not be effective in the inhibition of bacterial growthin vivo, since fatty acids are readily available from the host andAcpS can function in lipid biosynthesis. Therefore, targetingAcpS may offer a more effective approach to inhibiting bacte-rial growth in vivo by blocking both fatty acid and lipid bio-synthesis.

ACKNOWLEDGMENTS

We thank John Cronan for providing an E. coli strain overexpressingACP used in this study. We also thank Mel Johnson and John Rich-ardson for help with N-terminal sequencing and mass spectrophoto-metric analysis, respectively. We also thank Sheng-bin Peng, HaroldWatson, and Anthony S. Fischl for stimulating discussions and criticalreview of the manuscript.

REFERENCES

1. Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990.Basic local alignment search tool. J. Mol. Biol. 215:403–410.

2. Baltz, R. H., P. Matsushima, R. B. Peery, J. Hoskins, M. Young, F. H. Norris,B. S. DeHoff, P. Rockey, G. Porter, S. G. Burgett, P. R. Rosteck, P. L.Skatrud, and S. R. Jaskunas. 1997. Abstr. 37th Intersci. Conf. Antimicrob.Agents Chemother., abstr. S-46, p. 391.

3. Blattner, F. R., G. Plunkett III, C. A. Bloch, et al. 1997. The completegenome sequence of Escherichia coli K-12. Science 277:1453–1462.

4. Carreras, C. W., A. M. Gehring, C. T. Walsh, and C. Khosla. 1997. Utiliza-tion of enzymatically phosphopantetheinylated acyl carrier proteins andacetyl-acyl carrier proteins by the actinorhodin polyketide synthase. Bio-chemistry 36:11757–11761.

5. Chirgadze, N. Y., S. L. Briggs, K. A. McAllister, A. S. Fischl, and G. Zhao.2000. Crystal structure of Streptococcus pneumoniae acyl carrier proteinsynthase: an essential enzyme in bacterial fatty acid biosynthesis. EMBO J.19:5281–5287.

6. Cole, S. T., R. Brosch, et al. 1998. Deciphering the biology of Mycobacteriumtuberculosis from the complete genome sequence. Nature 393:537–544.

7. Cronan, J. E., Jr., and D. Laporte. 1996. Tricarboxylic acid cycle and glyoxy-late bypass, p. 206–216. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham,E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M.Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella:cellular and molecular biology. ASM Press, Washington, D.C.

8. Cronan, J. E., Jr., and C. O. Rock. 1996. Biosynthesis of membrane lipids, p.612–636 In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B.Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E.Umbarger (ed.), Escherichia coli and Salmonella: cellular and molecularbiology. ASM Press, Washington, D.C.

9. Crosby, J., D. H. Sherman, M. J. Bibb, W. P. Revill, D. A. Hopwood, and T. J.Simpson. 1995. Polyketide synthase acyl carrier proteins from Streptomyces:expression in Escherichia coli, purification and partial characterization. Bio-chim. Biophys. Acta 1251:32–42.

10. Daubaras, D. L., E. M. Wilson, T. Black, C. Strickland, B. M. Beyer, and P.Orth. 2004. Crystallization and preliminary X-ray analysis of the acyl carrierprotein synthase (AcpS) from Staphylococcus aureus. Acta Cryst. D60:773–774.

11. Elovson, J., and P. R. Vagelos. 1968. Acyl carrier protein. X. Acyl carrierprotein synthetase. J. Biol. Chem. 243:3603–3611.

12. Finking, R., J. Solsbacher, D. Konz, M. Schobert, A. Schafer, D. Jahn, andM. A. Marahiel. 2002. Characterization of a new type of phosphopantethei-nyl transferase for fatty acid and siderophore synthesis in Pseudomonasaeruginosa. J. Biol. Chem. 277:50293–50302.

13. Flugel, R. S., Y. Hwangbo, R. H. Lambalot, J. E. Cronan, Jr., and C. T.Walsh. 2000. Holo-(acyl carrier protein) synthase and phosphopantetheinyltransfer in Escherichia coli. J. Biol. Chem. 275:959–968.

14. Gehring, A. M., R. H. Lambalot, K. W. Vogel, D. G. Drueckhammer, andC. T. Walsh. 1997. Ability of Streptomyces spp. acyl carrier proteins andcoenzyme A analogs to serve as substrates in vitro for E. coli holo-ACPsynthase. Chem. Biol. 4:17–24.

15. Heath, R. J., and C. O. Rock. 1996. Regulation of fatty acid elongation andinitiation by acyl-acyl carrier protein in Escherichia coli. J. Biol. Chem.271:1833–1836.

16. Heath, R. J., and C. O. Rock. 2004. Fatty acid biosynthesis as a target fornovel Antibacterials. Curr. Opin. Investig. Drugs 5:146–153.

17. Heath, R. J., S. W. White, and C. O. Rock. 2001. Lipid biosynthesis as a targetfor antibacterial agents. Progress Lipid Res. 40:467–497.

18. Himmelreich, R., H. Hilbert, H. Plagens, E. Pirkl, B.-C. Li, and R.Herrmann. 1996. Complete sequence analysis of the genome of the bacte-rium Mycoplasma pneumoniae. Nucleic Acids Res. 24:4420–4449.

19. Hoskins, J., W. E. Alborn, Jr., J. Arnold, et al. 2001. Genome of the bacte-rium Streptococcus pneumoniae strain R6. J. Bacteriol. 183:5709–5717.


20. Jackowski, S., and C. O. Rock. 1984. Metabolism of 4�-phosphopantetheinein Escherichia coli. J. Bacteriol. 158:115–120.

21. Jackowski, S., and C. O. Rock. 1983. Ratio of active to inactive forms of acylcarrier protein in Escherichia coli. J. Biol. Chem. 258:15186–15191.

22. Jackowski, S., and C. O. Rock. 1981. Regulation of coenzyme A biosynthesis.J. Bacteriol. 148:926–932.

23. Jackowski, S., and C. O. Rock. 1986. Consequences of reduced intracellularcoenzyme A content in Escherichia coli. J. Bacteriol. 166:866–871.

24. Kalman, S., W. Mitchell, R. Marathe, C. Lammel, J. Fan, R. W. Hyman, L.Olinger, J. Grimwood, R. W. Davis, and R. S. Stephens. 1999. Comparativegenomes of Chlamydia pneumoniae and C. trachomatis. Nat. Genet. 21:385–389.

25. Kunst, F., N. Ogasawara, I. Moszer, et al. 1997. The complete genomesequence of the gram-positive bacterium Bacillus subtilis. Nature 390:249–256.

26. Kutchma, A. J., T. T. Hoang, and H. P. Schweizer. 1999. Characterization ofa Pseudomonas aeruginosa fatty acid biosynthetic gene cluster: purification ofacyl carrier protein (ACP) and malonyl-coenzyme A:ACP transacylase(FabD) J. Bacteriol. 181:5498–5504.

27. Lam, H.-M., E. Tancula, W. B. Dempsey, and M. E. Winkler. 1992. Suppres-sion of insertions in the complex pdxJ operon of Escherichia coli K-12 by lonand other mutations. J. Bacteriol. 174:1554–1567.

28. Lambalot, R. H., A. M. Gehring, R. S. Flugel, P. Zuber, M. LaCelle, M. A.Marahiel, R. Reid, C. Khosla, and C. T. Walsh. 1996. A new enzyme super-family: the phosphopantetheinyl transferases. Chem. Biol. 3:923–936.

29. Lambalot, R. H., and C. T. Walsh. 1995. Cloning, overproduction, andcharacterization of the Escherichia coli holo-acyl carrier protein synthase.J. Biol. Chem. 270:24658–24661.

30. Leon, O., and C. Panos. 1981. Long-chain fatty acid perturbations in Myco-plasma pneumoniae. J. Bacteriol. 146:1124–1134.

31. Majerus, P. W., A. W. Alberts, and P. R. Vagelos. 1965. Acyl carrier protein.IV. The identification of 4�-phosphopantetheine as the prosthetic group ofthe acyl carrier protein. Proc. Natl. Acad. Sci. USA 53:410–417.

32. McAllister, K. A., R. B. Peery, T. I. Meier, A. S. Fischl, and G. Zhao. 2000.Biochemical and molecular analyses of the Streptococcus pneumoniae acylcarrier protein synthase, an enzyme essential for fatty acid biosynthesis.J. Biol. Chem. 275:30864–30872.

33. Mootz, H. D., R. Finking, and M. A. Marahiel. 2001. 4�-Phosphopantetheinetransfer in primary and secondary metabolism of Bacillus subtilis. J. Biol.Chem. 276:37289–37298.

34. Morbidoni, H. R., D. De Mendoza, and J. E. Cronan, Jr. 1996. Bacillussubtilis acyl carrier protein is encoded in a cluster of lipid biosynthesis genes.J. Bacteriol. 178:4794–4800.

35. Parris, K. D., L. Lin, A. Tam, R. Mathew, J. Hixon, M. Stahl, C. C. Fritz, J.Seehra, W. S. Somers. 2000. Crystal structures of substrate binding to Ba-cillus subtilis holo-(acyl carrier protein) synthase reveal a novel trimericarrangement of molecules resulting in three active sites. Struct. Fold Des.8:883–895.

36. Quadri, L. E. N., P. H. Weinreb, M. Lie, M. M. Nakano, P. Zuber, and C. T.Walsh. 1998. Characterization of Sfp, a Bacillus subtilis phosphopantetheinyltransferase for peptidyl carrier protein domains in peptide synthetases. Bio-chemistry 37:1585–1595.

37. Rawlings, M., and, J. E. Cronan, Jr. 1992. The gene encoding Escherichiacoli acyl carrier protein lies within a cluster of fatty acid biosynthetic genes.J. Biol. Chem. 267:5751–5754.

38. Rock, C. O., and, J. E. Cronan, Jr. 1996. Escherichia coli as a model for theregulation of dissociable (type II) fatty acid biosynthesis. Biochem. Biophys.Acta 1302:1–16.

39. Shen, Z., and D. M. Byers. 1996. Isolation of Vibrio harveyi acyl carrierprotein and the fabG, acpP, and fabF genes involved in fatty acid biosynthe-sis. J. Bacteriol. 178:571–573.

40. Smith, S. 1994. The animal fatty acid synthase: one gene, one polypeptide,seven enzymes. FASEB 8:1248–1259.

41. Takiff, J. E., T. Baker, T. Copeland, S.-M. Chen, and D. L. Court. 1992.Locating essential Escherichia coli genes by using mini-Tn10 transposons: thepdxJ operon. J. Bacteriol. 174:1544–1553.

42. Tettelin H., K. E. Nelson, et. al. 2001. Complete genome sequence of avirulent isolate of Streptococcus pneumoniae. Science 293:498–506.

43. Tomb, J.-F., O. White, A. R. Kerlavageet, et al. 1997. The complete genomesequence of the gastric pathogen Helicobacter pylori. Nature 388:539–547.

44. Tropf, S., W. P. Revill, M. J. Bibb, D. A. Howood, and M. Schweizer. 1998.Heterologously expressed acyl carrier protein domain of rat fatty acid syn-thase functions in Escherichia coli fatty acid synthase and Streptomyces coeli-color polyketide synthase systems. Chem. Biol. 5:135–146.

45. Vallari, D. S., and S. Jackowski. 1988. Biosynthesis and degradation bothcontribute to the regulation of coenzyme A content in Escherichia coli. J.Bacteriol. 170:3961–3966.

46. Vallari, D. S., S. Jackowski, and C. O. Rock. 1987. Regulation of pantothe-nate kinase by coenzyme A and its thioesters. J. Biol. Chem. 262:2468–2471.

47. Zhao, G., T. I. Meier, R. B. Peery, P. Matsushima, and P. L. Skatrud.1999. Biochemical and molecular analyses of the C-terminal domain ofEra GTPase from Streptococcus pneumoniae. Microbiology 145:791–800.

48. Zhou, P., G. Florova, and K. A. Reynolds. 1999. Polyketide synthase acylcarrier protein (ACP) as a substrate and a catalyst for malonyl ACP biosyn-thesis. Chem. Biol. 6:577–584.


acyl carrier protein synthases from gram-negative, gram ...envismadrasuniv.org/physiology/pdf/acyl...

Documents