Resistance to AFN-1252 Arises from Missense Mutations inStaphylococcus aureus Enoyl-acyl Carrier Protein Reductase(FabI)*

Received for publication, August 22, 2013, and in revised form, November 1, 2013 Published, JBC Papers in Press, November 4, 2013, DOI 10.1074/jbc.M113.512905

Jiangwei Yao, John B. Maxwell, and Charles O. Rock1

From the Department of Infectious Diseases, St. Jude Children’s Research Hospital, Memphis, Tennessee 38105

Background: AFN-1252 is a FabI inhibitor developed to treat Staphylococcus aureus.Results: AFN-1252 resistance arises from a single missense mutation that produces a FabI(M99T) protein.Conclusion: The interaction between Met-99 and AFN-1252 accounts for the staphylococcal selectivity of the drug.Significance:The AFN-1252-resistant strains remain susceptible to drug concentrations typically achieved in antibiotic therapy.

AFN-1252 is a potent antibiotic against Staphylococcus aureusthat targets the enoyl-acyl carrier protein reductase (FabI). A thor-ough screen for AFN-1252-resistant strains was undertaken toidentify the spectrum of mechanisms for acquired resistance. Amissense mutation in fabI predicted to encode FabI(M99T) wasisolated 49 times, and a single isolate was predicted to encodeFabI(Y147H). AFN-1252 only bound to the NADPH form ofFabI, and the close interactions between the drug and Met-99and Tyr-147 explained how themutations would result in resis-tant enzymes. The clone expressing FabI(Y147H) had a pro-nounced growth defect that was rescued by exogenous fatty acidsupplementation, and the purified protein had less than 5% ofthe enzymatic activity of FabI. FabI(Y147F)was also catalyticallydefective but retained its sensitivity to AFN-1252, illustratingthe importance of the conservedTyr-147hydroxyl group inFabIfunction. The strains expressing FabI(M99T) exhibited normalgrowth, and the biochemical properties of the purified proteinwere indistinguishable from those of FabI. The AFN-1252 Ki

app

increased from4nM in FabI to 69 nM in FabI(M99T), accountingfor the increased resistance of the correspondingmutant strain.The low activity of FabI(Y147H) precluded an accurate Ki mea-surement. The strain expressing FabI(Y147H) was also resistantto triclosan; however, the strain expressing FabI(M99T)wasmoresusceptible. Strainswithhigher levels ofAFN-1252resistancewerenot obtained. The AFN-1252-resistant strains remained sensitiveto submicromolar concentrations of AFN-1252, which blockedgrowth through inhibition of fatty acid biosynthesis at the FabIstep.

Enoyl-acyl carrier protein (ACP)2 reductase (FabI) catalyzesthe NAD(P)H-dependent reduction of trans-2-enoyl-ACP to

acyl-ACP, as one of the four essential steps required for everycycle of two-carbon elongation in the biosynthesis of fatty acidsin bacteria (1). The FabI of Escherichia coli plays a determinantrole in pulling each round of elongation to completion (2). Thediscovery that the antibiotic activity of triclosan (Fig. 1A)against E. coli was due to the selective inhibition of FabI (3, 4)validated FabI as a bona fide drug target and stimulated thebiochemical and structural characterization of FabI (5–12) aswell as the development of multiple small molecule FabI inhib-itors as potential antibacterial agents (11–15). It was initiallythought that such inhibitors would be broad-spectrum antibac-terials because triclosan is effective in killing almost all bacteria.However, FabI is not the only enoyl-ACP reductase isoform inbacteria. Streptococcus pneumoniae uses a flavoprotein calledFabK that is unrelated in structure to FabI and is refractory totriclosan inhibition (16, 17). Thus, the inhibitory action of tri-closan in this bacterium arises by an unknown mechanismunrelated to fatty acid synthesis. Two other FabI-related pro-teins also carry out enoyl-ACP reduction, FabL (5) and FabV(18). Although these enzymes are related to FabI, bacteria thatexpress these proteins are resistant to triclosan (5, 16, 18).Despite the fact that FabI inhibitors will not have broad-spec-trum activity, development of FabI-directed drugs has contin-ued because FabI is essential in the important human pathogenStaphylococcus aureus (6). The rapid spread of S. aureus resis-tant to multiple drugs is a major threat to our health care sys-tem, and therapies directed against new targets are needed tocombat this emerging pathogen (19, 20).AFN-1252 (Fig. 1A) was developed to selectively target

S. aureus FabI through structure-guided optimization (21–24).AFN-1252 is a potent, orally administered anti-staphylococcaldrug. It inhibits FabI with an apparent IC50 of about 10 nM, andthe drug has 4 ng/ml antibacterial potency against S. aureus (25,26). The advancement of AFN-1252 into human Phase II clinicaltrials emphasizes the importance of understanding the biochemi-cal mechanism(s) of genetically acquired resistance. In previouswork, bacterial clones with increased AFN-1252 resistance wereisolated that contained missense mutations in the fabI gene pre-dicted to encode FabI(M99T) and FabI(Y147H) proteins (25,26). This study conducts a thorough screen for AFN-1252-re-sistant mutants to identify all mechanisms for genetically

* This work was supported, in whole or in part, by National Institutes of HealthGrants GM034496 (to C. O. R.) and CA21765 (Cancer Center Support Grant).This work was also supported by the American Lebanese Syrian AssociatedCharities.

1 To whom correspondence should be addressed: Dept. of Infectious Dis-eases, St. Jude Children’s Research Hospital, 262 Danny Thomas Pl., Mem-phis, TN 38105. Tel.: 901-595-3491; Fax: 901-595-3099; E-mail: charles.rock@stjude.org.

2 The abbreviations used are: ACP, acyl carrier protein; apoACP, ACP lacking the4�-phosphopantetheine prosthetic group; FabI, enoyl-ACP reductase; MIC,minimum inhibitory concentration; FASII, bacterial type II fatty acid synthase.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 288, NO. 51, pp. 36261–36271, December 20, 2013© 2013 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 JOURNAL OF BIOLOGICAL CHEMISTRY 36261

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

acquired resistance, examines the biochemical properties of theresistant enzymes, and evaluates the effect of their expressionon bacterial physiology. Although strains expressing FabImutants had increased resistance to AFN-1252 inhibition, theystill had submicromolar MICs for AFN-1252 due to on-targetinhibition of FabI.

EXPERIMENTAL PROCEDURES

Bacteria and Molecular Biology—All strains used in thiswork were derivatives of S. aureus strain RN4220 (27). TheMICs were determined by growing strains to an optical densityat 600 nm (A600) of 1.0 and diluted 30,000-fold in Luria-Bertani(LB)medium. A 10-�l aliquot of diluted cells was added to eachwell of a U-bottom 96-well plate containing 100 �l of LBmediumwith the appropriate concentration of compound. Theplate was incubated at 37 °C for 20 h, and cell growthwas deter-mined using a Fusion plate reader at 600 nm. Cells grown inmedium containing dimethyl sulfoxide (DMSO; 1%) were usedas the control for 100% growth. LB medium containing 0.1%Brij-58 at 37 °C with or without supplementation with 133 �M

anteiso-15:0, 66�M anteiso-17:0, and 10�M lipoate was used totest the influence of exogenous fatty acids on growth.The fabI gene from S. aureuswas cloned into pET15b (Nova-

gen), overexpressed, and purified as described previously (6).The plasmids expressing the FabI(M99T) and FabI(Y147H)mutants were generated by mutagenizing the pET15b-fabIplasmid using the QuikChange mutagenesis kit from AgilentTechnology, following the manufacturer’s protocol. Themutagenesis primers for the M99T mutant were 5�-tgtatat-cattcaatcgcatttgctaatacggaagacttacgcgg and 5�-ccgcgtaagtcttc-cgtattagcaaatgcgattgaatgatataca. The primers for the Y147Hmutant were 5�-tagcattgttgcaacaacacatttaggtggcgaattcgc and5�-gcgaattcgccacctaaatgtgttgttgcaacaatgcta. Expression andpurification of themutant proteins were the same as those usedfor the wild-type protein. The FabI(M99T,Y147H) doublemutant was generated by sequential mutagenesis of theFabI(M99T) plasmid with the FabI(Y147H) primers. The wild-type and mutant fabI genes were cloned into the plasmidpCL15, an S. aureus expression vector, and the plasmids weretransformed into strain RN4220 as described previously (26).Selection of AFN-1252-resistant Mutants—AFN-1252-resis-

tant mutants were selected on LB-agar plates containing 40ng/ml AFN-1252. An aliquot of 200 �l of A600 � 1 (1 � 108cells) S. aureus strain RN4220 was spread on a 10-cm LB-agarplate containing AFN-1252. The plate was incubated at 37 °Cfor 48 h, and on average, 5–8 resistant colonies were found perplate. Colonies that grew on the plate were restreaked ontoanother AFN-1252-containing LB agar plate to purify individualclones. The sequences of the fabI genes of the confirmedmutantswere amplified via PCR and sequenced. Strains RN4220,MWF32(FabI(M99T)), and MWF33 (FabI(Y147H)) were subjected toadditional rounds of selection on LB-agar plates containing 0.2–1�g/ml AFN-1252 as described above. No resistant colonies wereisolated at this AFN-1252 concentration.Synthesis of Crotonyl-ACP—The S. aureus acpP gene was

amplified via Taq polymerase (Invitrogen) from S. aureusgenomicDNA and cloned into the pET-15b plasmid (Novagen)at the NdeI and BamHI restriction sites. The resulting plasmid

was transformed into BL21 DE3 (Novagen) cells. Following a3-h induction of log phase cells, the cells were lysed, and theamino-terminal His6-tagged ACP was purified using nickel-ni-trilotriacetic acid affinity chromatography. A yield of 40 mg ofpure apoACP was obtained from each liter of culture. The pro-tein was dialyzed against 20 mM Tris, pH 8.0, to remove theimidazole from purification. The His6 tag was cleaved fromapoACP by treatment with thrombin (5 units/mg purified pro-tein) overnight at 4 °C. Thrombin and the cleaved histidine tagwere separated from the ACP via size exclusion chromatogra-phy. The resulting apoACP had a Gly-Ser-His addition at theamino terminus of the protein. Crotonyl-ACP was biosynthe-sized by incubating 250 �M crotonyl-CoA with 200 �M apo-SaACP, 10 mM MgCl2, and 1 �M purified Streptococcus pneu-moniae ACP synthase in 50 mM Tris-HCl, pH 7.0, at 30 °C for2 h (28). Completion of the reactionwas determined by analysisusing a 15% polyacrylamide gel containing 0.5 M urea gel toseparate apoACP from crotonyl-ACP. The His-tagged ACPsynthasewas removedusing a 0.2-ml nickel-nitrilotriacetic acidcolumn. The flow-through containing the crotonyl-ACP wasconcentrated with an Amicon Ultra centrifugal filter (3 kDacut-off). The MgCl2 and CoA were then removed using Zeba-Spin desalting columns.Real-time PCR for Analysis of Gene Expression—The strains

were grown to an A600 of 1.0. Cells were washed with 1 ml ofRNAlater solution from Ambion and treated with Lysostaphinfor 15 min at room temperature. Total RNA was isolated fromthe bacterial cells using the RNAqueous kit (Ambion) per themanufacturer’s instructions, including the LiCl precipitation.The pelleted RNAwas resuspended in nuclease-free water, anda 10-�g aliquot was treated with Ambion’s TurboDNA-free kitto remove genomic DNA. First-strand cDNA was generatedfrom 500 ng of total RNA using random primers and Super-Script II reverse transcriptase (Invitrogen) according to themanufacturer’s protocol. Quantitative real-time PCR was per-formed using the ABI Prism 7700 sequence detection system.Experimental samples were run in triplicate; negative controls(distilled water) and RNA samples without the reverse tran-scription step were run in duplicate. PCRs contained SYBRGreen PCR Master Mix (Applied Biosystems), a 150 nM con-centration of each primer, and cDNA synthesized from 10 ng oftotal RNA. The expression level of various housekeeping genes(proC, gyrB, gmk, glyA, rpoD, rho, recF, and pyk) was checked,and gyrB was determined to be the calibrator least changed byAFN-1252 treatment and subsequently was used as the controlto normalize mRNA levels between samples. Template curvesof seven points ranging from 50 ng to 50 pg of total RNA inputwere run for each primer set and analyzed for linearity andrelative efficiency of the PCR as compared with the control.Dissociation curves were generated after each real-time PCRrun to check for the presence of nonspecific amplification.Analysis of [14C]AFN-1252 Binding to FabI—Gel filtration

was used to determine which FabI form bound AFN-1252.[14C]AFN-1252 (specific activity 58.1 mCi/mmol) was a gener-ous gift from Nachum Kaplan of Affinium Pharmaceuticals.[14C]AFN-1252 (25 �M) was incubated with FabI (25 �M), andeither 200�MNADP�, 200�MNADPH, or nonicotinamides in20 mM Tris-HCl, pH 7.8, for 20 min at room temperature. The

Molecular Mechanism of AFN-1252 Resistance

36262 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288 • NUMBER 51 • DECEMBER 20, 2013

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

mixtures were then chromatographed over a 24-ml SuperdexG200 column equilibrated with 20mMTris HCl, 150mMNaCl,pH 7.8, at 4 °C. Then the sample (100 �l) was applied to thecolumn, the column was eluted with 20 mM Tris-HCl, 150 mM

NaCl, pH 7.8, at 1 ml/min, and 200-�l fractions were collected.The fractions were added to 3ml of ScintSafe scintillation solu-tion and counted in a Beckman LS6500 scintillation counter.ACP Immunoblotting—Cultures (50 ml) of strains RN4220,

MWF32, and MWF33 were grown in LB medium to an A600 of0.5 at 37 °C. AFN-1252 (final concentration of 100 ng/ml forstrain RN4220 and 2�g/ml for themutant strains)was added tothe culture. Control samples had only DMSO (1%) used to dis-solve the drug. The treated cultures were incubated for 30 min;the cells were harvested and lysed; and the cell extracts werefractionated on a 15%polyacrylamide, 0.5 M urea gel and immu-noblotted using anti-ACP antibodies (29) to determine theACP species (26). Protein from the same number of cells wasloaded into each lane by normalizing the extract volume to theA600 of the culture.Acetate Labeling—[14C]Acetate labeling experiments were

conducted to measure the effect of AFN-1252 on lipid biosyn-thesis in strains RN4220, MWF32, and MWF33 when treatedwithAFN-1252. Starter cultures were grown overnight, and thestarter culture was back-diluted to an initial A600 of 0.1 andgrown untilA600 reached 0.5 at 37 °C and 225 rpm shaking. Theculture was split into 10-ml aliquots, and each aliquot was incu-bated with the appropriate concentration of AFN-1252 and 10�Ci of [14C]acetate for a single doubling time (30min for strainsRN4220 and MWF32, 120 min for strain MWF33). The cellswere collected by centrifugation and washed twice withphosphate-buffered saline. Lipids were harvested from pel-leted cells using the method of Bligh and Dyer (30). [14C]Ac-etate incorporation was measured by counting the lipidextracts on an LS6500 multipurpose scintillation counter.Measurements were made in duplicate, and the averageswith S.E. values were reported.Analysis of FabI Mutants by Heterologous Expression in

E. coli—The Met-99 residues of the S. aureus fabI gene in theBluescript vector (6) was mutagenized to other amino acidsusing the QuikChange site-directed mutagenesis kit (Agilent)and primers generated by the QuikChange Primer Design pro-gram (Agilent). The changes were verified by DNA sequencing.The Bluescript vectors harboring the mutant fabI genes weretransformed into E. coli strain JP1111 (fabI(Ts)) via electropo-ration. Strain JP1111 harboring pBluescript expression vectorcontaining the mutant fabI gene were streaked on LB-carbeni-cillin plates at 42 °C to determine if the S. aureus FabI mutantssupported cell growth. These same vectors were also trans-formed into E. coli strain ANS1 (tolC) to rank the AFN-1252resistance phenotype of the mutants by determining their abil-ity to shift the MIC for AFN-1252 as described previously (26).Strain ANS1 cells harboring the empty vector without theS. aureus fabI gene were included as control.FabI Enzymatic Assays—The FabI enzymatic progress was

determined by measuring the conversion of NADPH toNADP� at 340 nm. The enzyme reactions were performed in a100-�l volume in Costar UV half-area 96-well plates with aSpectraMax 340 instrument taking 340-nm readings at 10-s

intervals at 30 °C. For velocity measurements, FabI enzymebuffered in 100 mM glutamate, pH 7.8, was added to either 50�M crotonyl-ACP or 200 �M crotonyl-CoA and 250 �M

NADPH buffered in 20 mM Tris-HCl, pH 7.8. Assays for E. coliFabI were the same except that NADH was substituted forNADPH instead. For determination of Km of NADPH for thewild type and mutant FabI enzymes, 50 nM enzyme was addedto 50 �M crotonyl-ACP and 15, 25, 50, 75, 100, 150, or 250 �M

NADPH. The reaction was mixed for 10 s by the plate reader,and data were acquired at 10-s intervals for 10 min. Initialvelocities were calculated from the initial linear phase of theprogress curve. All kinetic experiments were run in duplicates,and parameters were fit to the duplicate data sets. Determina-tion of Km of crotonyl-ACP was similar, except with 200 �M

NADPH and 2.5, 5, 7.5, 10, 15, 25, and 50 �M crotonyl-ACP.The affinity of AFN-1252 for FabI was determined by mea-

suring the initial velocity with 50 nM enzyme, 200 �MNADPH,50�M crotonyl-ACP, and 15.6, 31.2, 62.5, 125, 250, 500, or 1000nMAFN-1252. The affinity of triclosan for FabI was determinedsimilarly except that FabI was incubated with triclosan and 200�M NADP� for 10 min before starting the assay to account forthe slow binding nature of triclosan (3). Under standard steadystate conditions, inhibitors are kept at concentrations 10-foldor more above the concentration of the enzyme, so that thatformation of the enzyme�inhibitor complex does not alter theconcentration of free inhibitors.However, AFN-1252 had affin-ity values in the nanomolar range, whereas nanomolar concen-trations of FabI are necessary to generate a detectable signal inour kinetic assay. Thus, the data were fit to theMorrison quad-ratic equation for fitting tight binding inhibitors (Equation 1),where v0 is the velocity with no inhibitor, vi is the velocity at thegiven concentration of inhibitor, [E]T is the total concentrationof FabI in the reaction, [I]T is the total concentration of inhibi-tor in the reaction, and Ki

app is the apparent dissociation con-stant of the inhibitor (31).

vi

v0� 1 �

��E�T � �I�T � Kiapp� � ���E�T � �I�T � Ki

app�2 � 4�E�T�I�T

2�E�T

(Eq. 1)

Briefly, the Morrison equation allows the determination ofaffinity in terms of free and bound concentrations of enzymeand inhibitor, accounting for the impact of enzyme�inhibitorbinding on the free concentration of inhibitor.For determination of the slow binding mechanism of tri-

closan to FabI(M99T), 50 nM enzymewas added to a solution of400 �M NADPH, 120 �M crotonyl-ACP, and 0.25, 0.5, 0.75, 1,1.25, 1.5, or 2�M triclosan. The reaction progresswasmeasuredover 5 min. Data were analyzed via the method of determiningthe mechanism of slow binding inhibitors described by Cope-land (32). Briefly, the progress curve was fit to Equation 2,where [P] is the concentration of product at time t, vi is theinitial velocity, vs is the steady state velocity, and kobs is the rateconstant for the change from initial velocity to steady statevelocity.

�P� � vst �vi � vs

kobs�1 � exp� � kobst�� (Eq. 2)

Molecular Mechanism of AFN-1252 Resistance

DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 JOURNAL OF BIOLOGICAL CHEMISTRY 36263

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

Then kobs was plotted against the concentration of inhibitor todetermine themechanism of slow binding. If the relationship islinear, then the inhibitor is a simple slow binding inhibitor,where the on rate for the binding of the inhibitor is slow. If therelationship is hyperbolic, then the inhibitor is a slow tightbinding inhibitor, where the EI complex isomerizes into an EI*complex at a slow rate.

RESULTS

Selection of AFN-1252-resistant Strains—AFN-1252 was apotent antibiotic against S. aureus strain RN4220with aMIC of3.9 ng/ml in LB broth (Fig. 1B). Individual clones of S. aureusstrain RN4220 resistant to AFN-1252 inhibition were selectedon LB-agar plates containing 40 ng/ml AFN-1252. Mutant col-onies were purified, and the sequence of the fabI genes was deter-mined in 50 of the resistant clones. Forty-nine resistant clones hadthe same single missense mutation in fabI (T296C) predicted toencode a FabI(M99T) protein, and one isolate had a missensemutation (T439C) predicted to encode FabI(Y147H). StrainMWF32 (FabI(M99T)) was typical of all 49 mutant strains withan AFN-1252 MIC of 250 ng/ml (Fig. 1B). Strain MWF33(FabI(Y147H)) was consistently 2-fold more resistant to AFN-

1252 than strain MWF32 with a MIC of 500 ng/ml (Fig. 1B).Strain PS01 (accD) did not require a functional FabI forgrowth and was completely refractory to AFN-1252 inhibition.This strain cannot synthesize fatty acids due to the accD dele-tion, and the fact that it was not affected by AFN-1252 illus-trated that the drug did not have an off-target effect inS. aureus.The wild-type and mutant fabI genes were cloned into the

S. aureus expression vector pCL15, and strain RN4220 was trans-formed with these expression constructs. Plasmid-driven expres-sion of the wild-type FabI increased the MIC for AFN-1252 from3.9 to 15.6 ng/ml, consistentwith FabI as the cellular target for thedrug (Fig. 1C). Expression of either FabI(M99T) or FabI(Y147H)increased the MIC to 500 or 1000 ng/ml, respectively (Fig. 1C).This analysis confirmed that FabI(Y147H) was 2-fold moreresistant to AFN-1252 than FabI(M99T) in vivo. Together, theselection and expression experiments show that the expressionof FabI(M99T) or FabI(Y147H) mutants were necessary andsufficient to confer increased resistance to AFN-1252. Thecross-resistance of strains MWF32 and MWF33 against otherFASII-targeted drugs was tested (Fig. 1D). Strain MWF33exhibited increased resistance to the prototypical FabI inhibitor

FIGURE 1. Susceptibility of AFN-1252-resistant S. aureus strains to FabI-directed inhibitors. A, structures of triclosan and AFN-1252. *, the carbon inAFN-1252 that was radiolabeled. B, MIC of AFN-1252 in LB media against S. aureus strains RN4220 (F), MWF32 (FabI(M99T)) (E), MWF33 (FabI(Y147H)) (f), andPS01 (accD) supplemented with 133 �M anteiso-15:0 and 66% anteiso-17:0 in 0.1% Brij-58 to permit growth of this fatty acid auxotroph (�). C, MICs forAFN-1252 against S. aureus strain RN4220 harboring either the empty plasmid pCL15 (F), plasmid pCL15 expressing FabI (E), plasmid pCL15 expressingFabI(M99T) (f), or plasmid pCL15 expressing FabI(Y147H) (�). D, MICs for triclosan against S. aureus strains RN4220 (F), MWF32 (FabI(M99T)) (E), MWF33(FabI(Y147H)) (f), and PS01 (accD) supplemented with 133 �M anteiso-15:0 and 66 �M anteiso-17:0 in 0.1% Brij-58 (�). Cell growth was monitored by A600,and the samples with no drug present were set at 100% growth. Error bars, S.E.

Molecular Mechanism of AFN-1252 Resistance

36264 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288 • NUMBER 51 • DECEMBER 20, 2013

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

triclosan compared with the wild-type strain (from 62.5 to 500ng/ml). This result was consistentwith the isolation of amutantfabI allele encoding FabI(Y147H) in a study of triclosan resis-tance (33). In contrast, strainMWF32 had increased sensitivityto triclosan (from 62.5 to 3.9 ng/ml; Fig. 1D). AFN-1252 selec-tion did not give rise to commonly found triclosan-resistantfabI alleles predicted to encode FabI(A95V), FabI(I192S), orFabI(F204S) (3, 9). The fact that triclosan had aMIC of 1 �g/mlagainst strain PS01 (accD) showed that triclosan also had anon-FASII target in S. aureus. The wild-type and AFN-1252-resistant strains MWF32 and MWF33 had the same MIC forFASII drugs that targeted the condensing enzymes of FASII.These included cerulenin (31.3 �g/ml), thiolactomysin (25�g/ml), platensimycin (0.8 �g/ml), and platencin (0.8 �g/ml).Thus, the selected fabI mutations conferred resistance only toAFN-1252 and not to drugs that targeted another pathwayenzyme.Repeated attempts at generating mutants resistant to higher

concentrations ofAFN-1252 by selection onplates containing 1�g/ml AFN-1252, using either the wild-type strain RN4220 orstrainsMWF32 orMWF33, yielded no colonies. Thus, we wereunable to directly select for more resistant clones or to evolve amore resistant clone starting with either of the two resistantclones we obtained in the first round of selection. These dataindicated that these twomutants were the only missensemuta-tions in fabI that led to a significantly increased MIC for AFN-1252.We investigated this hypothesis by generating all possiblemissense mutations at position 99 and used heterologousexpression in E. coli to rank the resistance phenotypes of eachmutant based on the MIC (Table 1). FabI(M99T) was the mostresistant of the single missense mutations that could occur atposition 99. Also, amino acid changes at position 99 that wouldnot be expected from the selection method because theyinvolved either two or three base pair changes also were notmore resistant than FabI(M99T) (Table 1).Phenotypes of the AFN-1252-resistant Strains—The growth

rates of the two fabI mutant strains were compared with wildtype (Fig. 2A). Strain MWF32 expressing the FabI(M99T) pro-tein had a doubling time in LB medium that was indistinguish-able from the parental strain RN4220 (27–30-min doublingtime). In contrast, strain MWF33 (FabI(Y147H)) had a signifi-cantly impaired growth rate (120-min doubling time) (Fig. 2A).If the slow growth rate of strain MWF33 was due to theFabI(Y147H) protein being catalytically defective, then supple-mentation with exogenous anteiso-15:0/17:0 fatty acids andlipoate should cure the deficiency in FASII activity and acceler-ate the growth rate. Supplementation with exogenous fattyacids did restore the doubling time of strain MWF33 to a wild-type rate (35 min) (Fig. 2A). Supplementing strains RN4220 orMWF32 with exogenous fatty acids did not increase theirgrowth rate. The transcription of the fab genes in S. aureus iscontrolled by the FapR repressor (34). This repressor is releasedfrom its DNA binding sites by malonyl-CoA, which accumu-lates when FASII activity is reduced (35). Thus, a hallmark ofFASII inhibition in S. aureus is an increase in the transcriptionof the fab genes (26). ThemRNA levels of plsX in the fapR-plsX-fabD-fabGoperon, fabH in the fabH-fabFoperon, and fabIweremeasured by quantitative RT-PCR to determine if the mutations

compromised FabI function in vivo, whichwould be reflected by acompensatory up-regulation of gene expression in the mutantstrains. All three genes exhibited increased expression in bothmutant strains, indicating that both fabI mutations resulted incatalytically compromised enzymes and triggered the up-regu-lation of the FapR regulon (Fig. 2B). These data suggested thatdeficiencies in the activity of the mutant fabI alleles led to thecompensatory up-regulation of pathway enzymes to maintainthe rate of FASII. In the case of strain MWF33, these compen-satory changes were not sufficient to restore the normal growthof the strain (Fig. 2A). The elevated fab gene expression in strainMWF32 suggested that FabI(M99T) was also catalytically defi-cient in vivo but that the compensatory up-regulation of the fabgenes (including fabI) was sufficient to prevent a deficit in fattyacid production resulting in a normal growth phenotype.Methio-nine is the only residue found at position 99 in S. aureusFabIs, andthe gene expression data indicated that FabI(M99T) was catalyti-cally compromised in vivo. These data provided insight into whywewerenot able to select for aFabI(M99T,Y147H)doublemutantusing higher concentrations of AFN-1252 starting with eitherstrainMWF32 orMWF33. The elimination of these two clearlyimportant drug-protein interactions should give rise to amore resistant enzyme, and indeed when we synthesized theFabI(M99T,Y147H) double mutant and tested it in the E. coli sys-tem, the MIC for AFN-1252 was shifted more than either of thesingle mutants alone (Table 1). Thus, the FabI(M99T,Y147H)was more resistant to AFN-1252 than either FabI(M99T) orFabI(Y147H) but was also predicted to be more catalyticallyimpaired in vivo than the single mutants.We tested this idea by

TABLE 1Complementation and AFN-1252 sensitivity for missense mutationsaffecting positions 99 and 147 of S. aureus FabIComplementationwas scored as growth (Y) or no growth (N) at 42 °Cwhen plasmidexpression of the indicated mutants was transformed into E. coli strain JP1111(fabI(Ts)). The sensitivity to AFN-1252was determined after transforming the plas-mids into E. coli strain ANS1, and the MICs were determined as described under“Experimental Procedures.”

Met-99 substitution Complementation (Y/N) MIC

�g/mlWild type Y 0.11-bp changeM99T Y 0.4M99V Y 0.2M99L Y 0.1M99I Y 0.1M99R Y 0.2M99K Y 0.1

2-bp changeM99S Y 0.2M99P Y 0.2M99A Y 0.2M99Q Y 0.2M99N Y 0.2M99W Y 0.2M99G Y 0.2M99E Y 0.1

3-bp changeM99F Y 0.2M99Y Y 0.2M99H Y 0.2M99C Y 0.2M99D Y 0.1

Y147 mutationsY147H Y 0.4M99T/Y147H Y 3.2Y147F Y 0.2

Molecular Mechanism of AFN-1252 Resistance

DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 JOURNAL OF BIOLOGICAL CHEMISTRY 36265

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

cloning the two single mutants and the double mutant into thepCL15 expression plasmid and transforming strain RN4220.We then compared the growth rates of the strains in the pres-ence of 4 ng/ml AFN-1252 to chemically knock out the chro-mosomally encodedwild-type FabI activity. The strain express-ing FabI(M99T) grew with a doubling time of 72 1 min, andthe FabI(Y147H)-expressing strain had a doubling time of101 6 min. The FabI(M99T,Y147H)-expressing strain had adoubling time of 321 25 min, illustrating the catalytic defi-ciency of the double mutant in the context of the S. aureusFASII enzymes. Thus, we concluded that repeated attempts toselect for the double mutant failed because it was too impairedto support the growth of S. aureus.

Although the strains MWF32 and MWF33 were more resis-tant to AFN-1252, there was still a MIC for the drug. Previouswork concluded that FabI was the only AFN-1252 target inS. aureus (26); therefore, we determined if FASII was still inhib-ited in the mutant strains (Fig. 2C). AFN-1252 most potentlyinhibited [14C]acetate incorporation in strain RN4220 and wasless effective in strainMWF32, and strainMWF33was themostresistant. The ACP pools of AFN-1252-treated wild-type andmutant strains were compared with the ACP pool of the

untreated cells (Fig. 2D). In untreated cells, non-esterifed ACPwas the only clearly discernible protein form. In AFN-1252-treated cells, a series of short-chain acyl-ACP accumulated.Treatment of the two mutant strains with higher concentra-tions of AFN-1252 resulted in the same pattern of acyl-ACPaccumulation, as noted in the wild-type strain. Thus, strainsMWF32 and MWF33 both exhibited increased resistance toAFN-1252, but the MIC for AFN-1252 in these strains was stillattributed to on-target inhibition of the mutant FabI enzymes.Enzymology of FabI Mutant Proteins—The wild-type and the

mutant FabI enzymes were expressed and purified. FabI(Y147H)was not stable to short term storage and precipitated from thestorage buffers, making this enzyme difficult to characterize.Freshly isolated FabI(Y147H) exhibited a marked catalyticdefect (Fig. 3A). The calculated FabI(Y147H) kcat was 2.3 0.8min�1 compared with the FabI kcat of 30.8 5.5 min�1. Thiscatalytic deficiency of FabI(Y147H) was consistent with thefatty acid-dependent growth phenotype of strain MWF33 (Fig.2A). The roles played by Tyr-157 and Lys-164 located in theconserved YX6Kmotif in NADP(H) binding and hydride trans-fer to carbon-3 of the substrate are an established characteristicof this protein family (36). Whether the Tyr-147 hydroxyl

FIGURE 2. The effect of fabI mutations on the growth rates of the resistant strains. A, the growth rate of strain MWF32(FabI(M99T)) supplemented with 133�M anteiso-15:0, 66 �M anteiso-17:0, and 10 �M lipoate in 0.1% Brij-58 (FA) (F) or with only 0.1% Brij-58 (E). Shown is the growth rate of strain MWF33(FabIY147H) mutant supplemented with 133 �M anteiso-15:0, 66 �M anteiso-17:0, and 10 �M lipoate in 0.1% Brij-58 (FA) (f) or with only 0.1% Brij-58 (�). Thegrowth experiments shown are representative of triplicate, independent experiments. B, increased expression of the plsX, fabH, and fabI genes in strainsMWF32 and MWF33 compared with the wild-type parent, strain RN4220. The level of gene expression in strain RN4220 was set to 1, and the relative levels ofexpression in the other two strains were compared using gyrA as the calibrator gene. Data were calculated from triplicates. C, AFN-1252 inhibition of[14C]acetate incorporation into the lipids of strains RN4220 (wild type) (F), MWF32 (FabI(M99T)) (E), and MWF33 (FabI(Y147H)) (f). Data were calculated fromtriplicate experiments. D, ACP pool composition in wild-type and resistant strains treated with AFN-1252. ACP species were separated by urea gel electropho-resis and visualized by a Western blot using anti-ACP antibodies. Although higher levels of AFN-1252 were required to inhibit the growth of the two resistant strains,the pattern of acyl-ACP accumulation following treatment was the same in all cases. The gel shown was representative of three experiments. Error bars, S.E.

Molecular Mechanism of AFN-1252 Resistance

36266 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288 • NUMBER 51 • DECEMBER 20, 2013

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

was important for activity was tested by constructing theFabI(Y147F) mutant. The E. coli MIC analysis system con-firmed that the FabI(Y147F)-expressing cells did not haveincreased resistance to AFN-1252 compared with FabI (Table1), suggesting that FabI(Y147F) bound AFN-1252 with approx-imately the same affinity as FabI. Like FabI(Y147H), purifiedFabI(Y147F) was not stable and had a similar low kcat (3.9 0.3min�1). The conserved Tyr-147 may form a hydrogen bondwith the thioester carbonyl of enoyl-ACP to align the substratein the active site and promote the reaction. This interpretationfor the role for Tyr-147 in catalysis was also arrived at based onmolecular dynamics simulations of enoyl-ACP binding toE. coli FabI (37).Whether AFN-1252 bound to the free enzyme or to one of

the nucleotide-bound forms was determined by incubating[14C]AFN-1252 with FabI, FabI plus NADP�, or FabI plusNADPH. The samples were then fractionated with a size exclu-sion column, the location of FabI was determined by UV detec-tion, and [14C]AFN-1252 was determined by scintillationcounting. [14C]AFN-1252 co-elutedwith the protein only in thepresence of NADPH, illustrating that AFN-1252 only bound tothe FabI�NADPH complex (Fig. 3B).The high affinity binding of AFN-1252 presented a challenge

in measuring its binding to FabI or its mutant derivatives. The

usual FabI assay uses enoyl-N-acetylcysteamines or enoyl-CoAs as substrate analogs (11, 38). However, the low affinity ofFabI for enoyl-CoA meant that this assay required low micro-molar FabI concentrations to observe detectable rates. Thesehigh protein concentrations cannot be used to kinetically deter-mine the affinity of nanomolar inhibitors with any accuracy.Therefore, our assays used crotonyl-ACP, which allowed us tomeasure initial rates using 20–50 nM FabI. Nonetheless, AFN-1252 binding was still high enough that the formation of theFabI�NADPH�AFN-1252 complex affected the free concentra-tion of the inhibitor, necessitating the use of the Morrisonquadratic equation (31) to take this effect into account in cal-culating binding constants (see “Experimental Procedures”).The affinity of AFN-1252 for FabI and FabI(M99T) was mea-sured kinetically with saturating concentrations of NADPHand crotonyl-ACP. As expected, the FabI(M99T) mutantenzymewasmore resistant against AFN-1252, with aKi

app of 69nM versus 4 nM for FabI under saturating substrate concentra-tions, accounting for the increased resistance against AFN-1252 at the cellular level (Fig. 3C). The low activity ofFabI(Y147H) (Fig. 3A)meant thatmicromolar concentrationsof enzyme were needed in the assays, precluding an accuratemeasurement of AFN-1252 binding to this mutant enzyme.However, all of the biological data indicated that FabI(Y147H)was

FIGURE 3. Biochemical properties of FabI and its AFN-1252-resistant mutants. A, enzyme velocity versus concentration plots were performed in triplicatefor FabI (F), FabI(M99T) (E), and FabI(Y147H) (f) under saturating substrate concentrations. B, the elution profile of [14C]AFN-1252 from a Superdex G200 gelfiltration column when incubated with FabI (F), FabI plus NADP� (E), and FabI plus NADPH (f). The chromatograms shown are representative of three gelfiltration experiments. C, velocity of FabI (F) and FabI(M99T) (E) at different concentrations of AFN-1252. Duplicate data sets were fit via the Morrison equationfor tight binding inhibitors to determine the Ki

app. D, initial velocities of FabI (F) and FabI(M99T) (E) were determined in triplicate as a function of thecrotonyl-ACP concentration. The apparent Km values were similar, but FabI exhibited substrate inhibition at the higher concentrations. Error bars, S.E.

Molecular Mechanism of AFN-1252 Resistance

DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 JOURNAL OF BIOLOGICAL CHEMISTRY 36267

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

�2-fold more resistant to AFN-1252 than FabI(M99T) (Fig.1B and Table 1).The FabI(M99T) enzyme had turnover rates as fast as the

wild type enzyme (Fig. 3A). The kinetic parameters of theFabI(M99T) mutant were compared with wild type enzyme,and no significant differences were observed. The FabI(M99T)mutant had a Ki

app of 42.1 8.6 �M for NADPH and 5.5 0.8�M for crotonyl-ACP compared with 55 7.5 �M for NADPHand 14.4 5.5 �M for crotonyl-ACP for FabI (Fig. 3D). FabIexhibited substrate inhibition by crotonyl-ACP (Ki

app � 20.1 8.4 �M), whereas the M99T mutant did not, although theobserved substrate inhibition was probably not physiologicallyrelevant because enoyl-ACPs were not normally detected invivo (Fig. 2D), and thus their concentrations were probablybelow the FabI Km.Triclosan Binding in the M99T Mutant—The increased sus-

ceptibility of strainMWF32 to triclosan was explored by exam-ining the time-dependent nature of triclosan binding to FabI.Triclosan binding to E. coli FabI exhibited slow binding kinet-ics, with the full effect of the drug occurring 50–80 s after thereaction was initiated (Fig. 4A). Triclosan binding to S. aureusFabI was significantly faster, reaching equilibrium within 20 s(Fig. 4B). However, triclosan binding to FabI(M99T) was a

slower process, taking close to 100 s from the onset for the fulleffect of triclosan inhibition to bemanifested (Fig. 4C). To com-pare affinities for triclosan to the FabI and FabI(M99T)enzymes, both enzymes were preincubated with triclosan andNADP� for 10min to allow inhibitor to bind before the start ofthe enzymatic reaction. These experiments showed thatFabI(M99T) had a higher affinity for triclosan with aKi

app � 7.9nM compared with Ki

app � 13.1 nM for FabI (Fig. 4D). TheFabI(M99T) enzyme was added to assay mixtures containingdifferent concentrations of triclosan, and the progress curveswere measured (Fig. 4C). The kobs term, which measures therate of the onset of inhibition, was determined from the pro-gress curve for each concentration of triclosan and plottedagainst the concentration of triclosan (Fig. 4C, inset). The kobsversus [triclosan] plot fit best to a straight line indicating revers-ible, slow binding of triclosan to FabI(M99T), with an apparenton rate (k3) of 1.7 � 104 M�1 s�1. This “slow binding” behaviorof triclosan is characteristic of its interactions with many FabIs(7, 12, 39), and the importance of slow off rates to the in vivodrug efficacy has recently become clear (12, 39). However, thebasis for the lower MIC in our closed system was attributed tothe higher affinity of triclosan for the FabI(M99T)�NAD�

complex.

FIGURE 4. Inhibition of FabI and its AFN-1252-resistant mutants by triclosan. A, enzymatic progress curves of the E. coli FabI reaction with and withouttriclosan. The traces shown are representative of triplicate experiments. B, enzymatic progress curves of the S. aureus FabI reaction with and without triclosan.The traces shown are representative of triplicate experiments. C, enzymatic progress curve of the FabI(M99T) enzyme at 0 �M (F), 0.25 �M (E), 0.5 �M (f), 0.75�M (�), 1 �M (Œ), 1.25 �M (‚), 1.5 �M (�), or 2 �M triclosan (ƒ). The traces shown are representative of duplicate experiments. The inset plots the kobs calculatedfrom each progress curve against the concentration of triclosan of that progress curve. D, velocity of FabI (F) and FabI(M99T) (E) at different concentrations oftriclosan. Duplicate data sets were fit via the Morrison equation for tight binding inhibitors to determine the Ki

app. Error bars, S.E.

Molecular Mechanism of AFN-1252 Resistance

36268 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288 • NUMBER 51 • DECEMBER 20, 2013

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

DISCUSSION

This work identifies the M99T mutation as the most com-mon mechanism for the acquisition of increased resistance toAFN-1252 in S. aureus. Met-99 resides on a conformationallyflexible loop that covers the FabI active site following the bind-ing ofNADPHand substrate (11). The helical segment contain-ing Met-99 in the FabI�NADPH�AFN-1252 ternary complex isnot a well conserved structural element in bacterial FabIs.Residues 95–117 are disordered in the FabI crystal structureto allow the entry of substrates into the active site and form astructured lid over the active site in the FabI�NADP(H),FabI�NADP(H)�triclosan, and FabI�NADP(H)�AFN-1252 com-plexes (11, 40). Met-99 forms a close interaction with the oxo-tetrahydronaphthyridine portion of AFN-1252 that contrib-utes a significant hydrophobic interaction that is clearlyrequired for high affinity AFN-1252 binding (Fig. 5). An analy-sis of FabI protein sequences deposited in the NCBI MicrobialGenomes database shows that staphylococcal species uni-formly have Met-99, whereas other FabIs have more hydro-philic and smaller residues in this position. This fact suggeststhat the interaction between Met-99 and AFN-1252 accountsfor the selective anti-staphylococcal activity of the drug (22).Although FabI(M99T) did not have a pronounced catalyticdefect using crotonyl-ACP as substrate, the actual in vivo sub-strates are an elongating series of branched-chain enoyl-ACPs.

This blind spot in our analysis may be relevant because theup-regulation of the lipid biosynthetic genes in strain MWF32compared with strain RN4220 illustrates that the FapR geneticregulatory switch is increasing protein expression to compen-sate for a deficiency in the pathway. Although the nature of thisinferred defect is not clear, the cells expressing themutant pro-tein are able to compensate and do not have an observablegrowth phenotype in the laboratory. Further experiments willbe required to determine if strains with the FabI(M99T) defectare attenuated for growth in animal models.FabI(Y147H) was the only other AFN-1252 mutant found in

our study (1 in 50). In contrast to Met-99, Tyr-147 is a con-served residue in eubacterial FabIs. The 3-methylbenzofuranrings of AFN-1252 bind in the catalytic pocket, making stronghydrophobic interactions with Tyr-147 (Fig. 5). The absence ofthis interaction in FabI(Y147H) accounts for the lower affinityof AFN-1252. The low catalytic activity of FabI(Y147H) pre-vented the calculation of a binding constant from kineticexperiments. However, the collective in vivo data suggest thatthe AFN-1252 Ki for FabI(Y147H) is 2-fold higher than forFabI(M99T). In bacterial FabIs, this invariant tyrosine residue isflanked by a serine/threonine and a large hydrophobic residue,such as leucine or phenylalanine. Structurally, the tyrosine res-idue is at a spatially conserved position forming a part of theactive site of the FabI enzyme in the crystal structures of FabIfrom S. aureus (11, 40), E. coli (41), Helicobacter pylori (42),Bacillus subtilis (43), and Bacillus cereus (43). The analysis ofthese structures, our site-directed mutagenesis, and moleculardynamics simulations (37) all point to Tyr-147 promoting thereaction by hydrogen-bonding to the substrate thioester car-bonyl. This hypothesis explains the compromised catalyticactivity of FabI(Y147H) and the marked growth defect of strainMWF33.Understanding the number and nature of AFN-1252 resis-

tance mutations in FabI has important implications to thepotential therapeutic use of AFN-1252. The low number ofAFN-1252-resistant mutations is attributed to many of the keyinteractions that occur between AFN-1252 and the backboneamide of Ala-97 and the hydrogen bond network and stackinginteractions that bridge the drug, protein, andNADPHcofactor(25). It is difficult to envision missense mutations that wouldselectively break these key interactions and still preserve cofac-tor binding and catalysis. Both mutant strains are more resis-tant to AFN-1252, but because AFN-1252 is such a high affinityinhibitor, they remained sensitive to 0.25–0.5 �g/ml of thedrug. This suggests that animals infected with the resistantstrains could be treated by administering higher concentrationsof AFN-1252 because the MIC for AFN-1252 in the resistantstrains is still lower thanmost antibiotics, such as linezolid (1–4�g/ml), a front-line drug used to treat S. aureus (22). Animalmodels testing the efficacy of AFN-1252 deliver doses of thedrug that generate serum levels over 1.5 �g/ml (44). The ideathat AFN-1252-resistant mutants could be successfully treatedby simply delivering the same or higher doses of drug will beimportant to test in animal infection models. It is also signifi-cant that we were unable to obtain S. aureusmutants that wereresistant to higher AFN-1252 concentrations than FabI(M99T)and FabI(Y147H) even when starting with strains MWF32 and

FIGURE 5. View of the FabI�NADPH�AFN-1252 ternary complex. The struc-ture was rendered in PyMOL (Protein Data Bank entry 4FS3) (25). The FabIactive site is depicted as a green ribbon diagram, NADP(H) carbons are pink,AFN-1251 carbons are yellow, and Met-99 and Tyr-147 side chain carbons arecyan. Met-99 interacts with the oxotetrahydronaphthyridine part of AFN-1252, and on the other side of the active site, Tyr-147 interacts with the3-methylbenzofuran portion. The loss of these important van der Waals inter-actions explained the increased resistance of FabI(M99T) and FabI(Y147H) toAFN-1252. Other key interactions between and the FabI�NADP(H) complexare hydrogen bond interactions between the carbonyl of the cis-amide ofAFN-1252 and the 2�-hydroxyl of NADPH and the hydroxyl of Tyr-157 andbetween the pyridyl nitrogen and the N-acyl hydrogen of the oxotetrahydro-naphthyride part of AFN-1252 and the backbone amide of Ala-95. The cis-amide also �-stacks with the nicotinamide ring of NADPH.

Molecular Mechanism of AFN-1252 Resistance

DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 JOURNAL OF BIOLOGICAL CHEMISTRY 36269

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

MWF33. We reasoned that the FabI(M99T,Y147H) doublemutant would be more resistant to AFN-1252 due to the elim-ination of two key interactions and confirmed this hypothesisby constructing and analyzing the double mutant enzyme.However, the in vivo gene expression data indicate that bothFabI(M99T) and FabI(Y147H) are functionally compromised,leadingus toconclude that theFabI(M99T,Y147H)doublemutantis even more deficient. This idea was corroborated by the findingthat a strain possessing a plasmid expressing FabI(M99T,Y147H)grew very poorly on low concentrations of AFN-1252 thatknocked out the endogenous wild-type FabI. These data suggestthat the double mutant does not arise in our selection schemesbecause a single copy of this defective enzyme cannot support thegrowth of S. aureus.Our study also provides insight into cross-resistance against

drugs targeting the same enzyme.Whereas AFN-1252 binds tothe FabI�NADPH complex (Fig. 5) and triclosan binds to theFabI�NADP� complex (3), the two inhibitors occupy the samesubstrate/product bindingpocket.Of the twoAFN-1252-resistantmutants, FabI(Y147H) has increased resistance against triclosan,whereas FabI(M99T) is actuallymore sensitive to triclosan. Futureinfectionmodels will be required to determine if the poor growthof the strain MWF33 expressing the FabI(Y147H) protein com-promises the virulence of S. aureus or whether the strain canobtain sufficient fatty acids from the host to support coloniza-tion. CG400462, a FabI inhibitor developed by CrystalGeno-mics, gave rise to the same two resistant mutants as we foundwith AFN-1252 (14), suggesting that it binds in a similar man-ner to the FabI�NADPH complex. Other FabI-targeted com-pounds, including CG400549 (13), MUT056399 from Mutabi-lis (15), and triclosan analogs from Xu et al. (9), give rise toresistant mutants (F204C/L/S/F and A95V) more typicallyfound for triclosan (45, 46) that were not detected in our study.Thus, the concern that spread of triclosan resistance wouldinvalidate the use of FabI-targeted drugs (47) is limited to theFabI(Y147H) mutant in the case of AFN-1252.

Acknowledgments—[14C]AFN-1252 was a generous gift from NachumKaplan of Affinium Pharmaceuticals.We thankMatt Frank, Pam Jack-son, and Chitra Subramanian for expert technical assistance and theProtein Production Shared Resource for protein expression andpurification.

REFERENCES1. Rock, C. O., and Jackowski, S. (2002) Forty years of fatty acid biosynthesis.

Biochem. Biophys. Res. Commun. 292, 1155–11662. Heath, R. J., and Rock, C. O. (1995) Enoyl-acyl carrier protein reductase

(fabI) plays a determinant role in completing cycles of fatty acid elongationin Escherichia coli. J. Biol. Chem. 270, 26538–26542

3. Heath, R. J., Rubin, J. R., Holland, D. R., Zhang, E., Snow, M. E., and Rock,C. O. (1999) Mechanism of triclosan inhibition of bacterial fatty acid syn-thesis. J. Biol. Chem. 274, 11110–11114

4. Levy, C. W., Roujeinikova, A., Sedelnikova, S., Baker, P. J., Stuitje, A. R.,Slabas, A. R., Rice, D. W., and Rafferty, J. B. (1999) Molecular basis oftriclosan activity. Nature 398, 383–384

5. Heath, R. J., Su, N.,Murphy, C. K., and Rock, C. O. (2000) The enoyl-[acyl-carrier-protein] reductases FabI and FabL from Bacillus subtilis. J. Biol.Chem. 275, 40128–40133

6. Heath, R. J., Li, J., Roland, G. E., and Rock, C. O. (2000) Inhibition of theStaphylococcus aureus NADPH-dependent enoyl-acyl carrier protein re-

ductase by triclosan and hexachlorophene. J. Biol. Chem. 275, 4654–46597. Lu, H., England, K., am Ende, C., Truglio, J. J., Luckner, S., Reddy, B. G.,

Marlenee, N. L., Knudson, S. E., Knudson, D. L., Bowen, R. A., Kisker, C.,Slayden, R. A., and Tonge, P. J. (2009) Slow-onset inhibition of the FabIenoyl reductase from Francisella tularensis. Residence time and in vivoactivity. ACS Chem. Biol. 4, 221–231

8. Parikh, S., Moynihan, D. P., Xiao, G., and Tonge, P. J. (1999) Roles oftyrosine 158 and lysine 165 in the catalytic mechanism of InhA, the enoyl-ACP reductase from Mycobacterium tuberculosis. Biochemistry 38,13623–13634

9. Xu, H., Sullivan, T. J., Sekiguchi, J., Kirikae, T., Ojima, I., Stratton, C. F.,Mao, W., Rock, F. L., Alley, M. R., Johnson, F., Walker, S. G., and Tonge,P. J. (2008) Mechanism and inhibition of saFabI, the enoyl reductase fromStaphylococcus aureus. Biochemistry 47, 4228–4236

10. Stewart, M. J., Parikh, S., Xiao, G., Tonge, P. J., and Kisker, C. (1999)Structural basis andmechanism of enoyl reductase inhibition by triclosan.J. Mol. Biol. 290, 859–865

11. Schiebel, J., Chang, A., Lu, H., Baxter, M. V., Tonge, P. J., and Kisker, C.(2012) Staphylococcus aureus FabI. Inhibition, substrate recognition, andpotential implications for in vivo essentiality. Structure 20, 802–813

12. Chang, A., Schiebel, J., Yu, W., Bommineni, G. R., Pan, P., Baxter, M. V.,Khanna, A., Sotriffer, C. A., Kisker, C., and Tonge, P. J. (2013) Rationaloptimization of drug-target residence time. Insights from inhibitor bind-ing to the S. aureus FabI enzyme-product complex. Biochemistry 52,4217–4228

13. Park, H. S., Yoon, Y. M., Jung, S. J., Kim, C. M., Kim, J. M., and Kwak, J. H.(2007) Antistaphylococcal activities of CG400549, a new bacterial enoyl-acyl carrier protein reductase (FabI) inhibitor. J. Antimicrob. Chemother.60, 568–574

14. Park, H. S., Yoon, Y. M., Jung, S. J., Yun, I. N., Kim, C. M., Kim, J. M., andKwak, J. H. (2007) CG400462, a new bacterial enoyl-acyl carrier proteinreductase (FabI) inhibitor. Int. J Antimicrob. Agents 30, 446–451

15. Escaich, S., Prouvensier, L., Saccomani,M., Durant, L., Oxoby,M., Gerusz,V., Moreau, F., Vongsouthi, V., Maher, K., Morrissey, I., and Soulama-Mouze, C. (2011) The MUT056399 inhibitor of FabI is a new antistaphy-lococcal compound. Antimicrob. Agents Chemother. 55, 4692–4697

16. Heath, R. J., and Rock, C. O. (2000) A triclosan-resistant bacterial enzyme.Nature 406, 145–146

17. Marrakchi, H., DeWolf, W. E., Jr., Quinn, C., West, J., Polizzi, B. J., So,C. Y., Holmes, D. J., Reed, S. L., Heath, R. J., Payne, D. J., Rock, C. O., andWallis, N. G. (2003) Characterization of Streptococcus pneumoniae enoyl-[acyl carrier protein] reductase (FabK). Biochem. J. 370, 1055–1062

18. Massengo-Tiasse, R. P., and Cronan, J. E. (2008) Vibrio cholerae fabVdefines a new class of enoyl acyl-carrier-protein reductase. J. Biol. Chem.283, 1308–1316

19. Chambers, H. F., andDeleo, F. R. (2009)Waves of resistance. Staphylococ-cus aureus in the antibiotic era. Nat. Rev. Microbiol. 7, 629–641

20. Walsh, C. (2003)Where will new antibiotics come from?Nat. Rev. Micro-biol. 1, 65–70

21. Karlowsky, J. A., Laing, N.M., Baudry, T., Kaplan, N., Vaughan, D., Hoban,D. J., and Zhanel, G. G. (2007) In vitro activity of API-1252, a novel FabIinhibitor, against clinical isolates of Staphylococcus aureus and Staphylo-coccus epidermidis. Antimicrob. Agents. Chemother. 51, 1580–1581

22. Karlowsky, J. A., Kaplan, N., Hafkin, B., Hoban, D. J., and Zhanel, G. G.(2009) AFN-1252, a FabI inhibitor, demonstrates a Staphylococcus-spe-cific spectrum of activity. Antimicrob. Agents Chemother. 53, 3544–3548

23. Banevicius, M. A., Kaplan, N., Hafkin, B., and Nicolau, D. P. (2013) Phar-macokinetics, pharmacodynamics and efficacy of novel FabI inhibitorAFN-1252 againstMSSA andMRSA in themurine thigh infectionmodel.J. Chemother. 25, 26–31

24. Kaplan, N., Awrey, D., Bardouniotis, E., Berman, J., Yethon, J., Pauls,H. W., and Hafkin, B. (2013) In vitro activity (MICs and rate of kill) ofAFN-1252, a novel FabI inhibitor, in the presence of serum and in com-bination with other antibiotics. J. Chemother. 25, 18–25

25. Kaplan, N., Albert, M., Awrey, D., Bardouniotis, E., Berman, J., Clarke, T.,Dorsey, M., Hafkin, B., Ramnauth, J., Romanov, V., Schmid, M. B.,Thalakada, R., Yethon, J., and Pauls, H. W. (2012) Mode of action, in vitroactivity, and in vivo efficacy of AFN-1252, a selective antistaphylococcal

Molecular Mechanism of AFN-1252 Resistance

36270 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288 • NUMBER 51 • DECEMBER 20, 2013

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

FabI inhibitor. Antimicrob. Agents Chemother. 56, 5865–587426. Parsons, J. B., Frank, M. W., Subramanian, C., Saenkham, P., and Rock,

C. O. (2011) Metabolic basis for the differential susceptibility of Gram-positive pathogens to fatty acid synthesis inhibitors. Proc. Natl. Acad. Sci.U.S.A. 108, 15378–15383

27. Kreiswirth, B. N., Lofdahl, S., Betley, M. J., O’Reilly, M., Schlievert, P. M.,Bergdoll, M. S., and Novick, R. P. (1983) The toxic shock syndrome exo-toxin structural gene is not detectably transmitted by a prophage. Nature305, 709–712

28. McAllister, K. A., Peery, R. B., and Zhao, G. (2006) Acyl carrier proteinsynthases fromGram-negative, Gram-positive, and atypical bacterial spe-cies. Biochemical and structural properties and physiological implica-tions. J. Bacteriol. 188, 4737–4748

29. Jackowski, S., and Rock, C. O. (1983) Ratio of active to inactive forms ofacyl carrier protein in Escherichia coli. J. Biol. Chem. 258, 15186–15191

30. Bligh, E. G., and Dyer,W. J. (1959) A rapidmethod of total lipid extractionand purification. Can. J. Biochem. Physiol. 37, 911–917

31. Morrison, J. F. (1969) Kinetics of the reversible inhibition of enzyme-catalysed reactions by tight-binding inhibitors. Biochim. Biophys. Acta185, 269–286

32. Copeland, R. A. (2005) Evaluation of enzyme inhibitors in drug discovery.A guide for medicinal chemists and pharmacologists. Methods Biochem.Anal. 46, 1–265

33. Ciusa,M. L., Furi, L., Knight, D., Decorosi, F., Fondi,M., Raggi, C., Coelho,J. R., Aragones, L., Moce, L., Visa, P., Freitas, A. T., Baldassarri, L., Fani, R.,Viti, C., Orefici, G.,Martinez, J. L.,Morrissey, I., andOggioni,M. R. (2012)A novel resistance mechanism to triclosan that suggests horizontal genetransfer and demonstrates a potential selective pressure for reduced bio-cide susceptibility in clinical strains of Staphylococcus aureus. Int. J. Anti-microb. Agents 40, 210–220

34. Schujman, G. E., Paoletti, L., Grossman, A. D., and deMendoza, D. (2003)FapR, a bacterial transcription factor involved in global regulation ofmembrane lipid biosynthesis. Dev. Cell 4, 663–672

35. Schujman, G. E., Altabe, S., and de Mendoza, D. (2008) A malonyl-CoA-dependent switch in the bacterial response to a dysfunction of lipid me-tabolism.Mol. Microbiol. 68, 987–996

36. Massengo-Tiasse, R. P., and Cronan, J. E. (2009) Diversity in enoyl-acylcarrier protein reductases. Cell Mol. Life Sci. 66, 1507–1517

37. Rafi, S., Novichenok, P., Kolappan, S., Zhang, X., Stratton, C. F., Rawat, R.,Kisker, C., Simmerling, C., and Tonge, P. J. (2006) Structure of acyl carrierprotein bound to FabI, the FASII enoyl reductase from Escherichia coli.J. Biol. Chem. 281, 39285–39293

38. Heath, R. J., Yu, Y.-T., Shapiro, M. A., Olson, E., and Rock, C. O. (1998)Broad spectrum antimicrobial biocides target the FabI component of fattyacid synthesis. J. Biol. Chem. 273, 30316–30320

39. Lu, H., and Tonge, P. J. (2010) Drug-target residence time. Critical infor-mation for lead optimization. Curr. Opin. Chem. Biol. 14, 467–474

40. Priyadarshi, A., Kim, E. E., andHwang, K. Y. (2010) Structural insights intoStaphylococcus aureus enoyl-ACP reductase (FabI), in complex withNADP and triclosan. Proteins 78, 480–486

41. Ward, W. H., Holdgate, G. A., Rowsell, S., McLean, E. G., Pauptit, R. A.,Clayton,E.,Nichols,W.W.,Colls, J.G.,Minshull,C.A., Jude,D.A.,Mistry,A.,Timms, D., Camble, R., Hales, N. J., Britton, C. J., and Taylor, I. W. (1999)Kinetic and structural characteristics of the inhibition of enoyl (acyl carrierprotein) reductase by triclosan. Biochemistry 38, 12514–12525

42. Lee, H. H., Moon, J., and Suh, S. W. (2007) Crystal structure of the Heli-cobacter pylori enoyl-acyl carrier protein reductase in complex with hy-droxydiphenyl ether compounds, triclosan and diclosan. Proteins 69,691–694

43. Kim, K. H., Ha, B. H., Kim, S. J., Hong, S. K., Hwang, K. Y., and Kim, E. E.(2011) Crystal structures of enoyl-ACP reductases I (FabI) and III (FabL)from B. subtilis. J. Mol. Biol. 406, 403–415

44. Parsons, J. B., Kukula, M., Jackson, P., Pulse, M., Simecka, J. W., Valtierra,D.,Weiss,W. J., Kaplan, N., and Rock, C. O. (2013) Perturbation of Staph-ylococcus aureus gene expression by the enoyl-acyl carrier protein reduc-tase inhibitor AFN-1252. Antimicrob. Agents Chemother. 57, 2182–2190

45. Fan, F., Yan, K., Wallis, N. G., Reed, S., Moore, T. D., Rittenhouse, S. F.,DeWolf, W. E., Jr., Huang, J., McDevitt, D., Miller, W. H., Seefeld, M. A.,Newlander, K. A., Jakas, D. R., Head,M. S., and Payne, D. J. (2002)Definingand combating the mechanisms of triclosan resistance in clinical isolatesof Staphylococcus aureus. Antimicrob. Agents Chemother. 46, 3343–3347

46. Brenwald, N. P., and Fraise, A. P. (2003) Triclosan resistance in methicil-lin-resistant Staphylococcus aureus (MRSA). J. Hosp. Infect. 55, 141–144

47. Levy, S. B. (2001) Antibacterial household products. Cause for concern.Emerg. Infect. Dis. 7, 512–515

Molecular Mechanism of AFN-1252 Resistance

DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 JOURNAL OF BIOLOGICAL CHEMISTRY 36271

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from

Jiangwei Yao, John B. Maxwell and Charles O. RockEnoyl-acyl Carrier Protein Reductase (FabI)

Staphylococcus aureusResistance to AFN-1252 Arises from Missense Mutations in

doi: 10.1074/jbc.M113.512905 originally published online November 4, 20132013, 288:36261-36271.J. Biol. Chem. 

  10.1074/jbc.M113.512905Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/288/51/36261.full.html#ref-list-1

This article cites 47 references, 16 of which can be accessed free at

by guest on July 27, 2020http://w

ww

.jbc.org/D

ownloaded from