Crystal Structure of a Bacterial Type III Polyketide Synthase andEnzymatic Control of Reactive Polyketide Intermediates*

Received for publication, June 11, 2004, and in revised form, July 12, 2004Published, JBC Papers in Press, July 20, 2004, DOI 10.1074/jbc.M406567200

Michael B. Austin‡§, Miho Izumikawa¶, Marianne E. Bowman‡, Daniel W. Udwary¶,Jean-Luc Ferrer�, Bradley S. Moore¶**‡‡, and Joseph P. Noel‡§ §§

From the ‡Structural Biology Laboratory, The Salk Institute for Biological Studies, La Jolla, California 92037, the§Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, California 92037-0634, the¶Division of Medicinal Chemistry, University of Arizona, Tucson, Arizona 85721-0207, �Institut de Biologie StructuralJean-Pierre Ebel/Laboratoire de Cristallographie et Cristallogenese des Proteines, 41 rue Jules Horowitz, 38027 GrenobleCedex 1, France, and the **Department of Chemistry, University of Arizona, Tucson, Arizona 85721

In bacteria, a structurally simple type III polyketidesynthase (PKS) known as 1,3,6,8-tetrahydroxynaphtha-lene synthase (THNS) catalyzes the iterative condensa-tion of five CoA-linked malonyl units to form a penta-ketide intermediate. THNS subsequently catalyzes dualintramolecular Claisen and aldol condensations of thislinear intermediate to produce the fused ring tetrahy-droxynaphthalene (THN) skeleton. The type III PKS-catalyzed polyketide extension mechanism, utilizing aconserved Cys-His-Asn catalytic triad in an internal ac-tive site cavity, is fairly well understood. However, themechanistic basis for the unusual production of THNand dual cyclization of its malonyl-primed pentaketideis obscure. Here we present the first bacterial type IIIPKS crystal structure, that of Streptomyces coelicolorTHNS, and identify by mutagenesis, structural model-ing, and chemical analysis the unexpected catalytic par-ticipation of an additional THNS-conserved cysteineresidue in facilitating malonyl-primed polyketide exten-sion beyond the triketide stage. The resulting newmechanistic model, involving the use of additional cys-teines to alter and steer polyketide reactivity, may gen-erally apply to other PKS reaction mechanisms, includ-ing those catalyzed by iterative type I and II PKSenzymes. Our crystal structure also reveals an unantic-ipated novel cavity extending into the “floor” of thetraditional active site cavity, providing the first plausi-ble structural and mechanistic explanation for yet an-other unusual THNS catalytic activity: its previouslyinexplicable extra polyketide extension step whenprimed with a long acyl starter. This tunnel allows forselective expansion of available active site cavity vol-

ume by sequestration of aliphatic starter-derivedpolyketide tails, and further suggests another distinctprotection mechanism involving maintenance of a lin-ear polyketide conformation.

1,3,6,8-Tetrahydroxynaphthalene (THN)1 is biosynthesizedfrom polyketide intermediates in fungi and bacteria (Fig. 1A).In some fungi, THN is reduced to 1,8-dihydroxynaphthaleneand undergoes polymerization to form UV-protective 1,8-dihy-droxynaphthalene-melanin (1). Filamentous bacteria of the ge-nus Streptomyces utilize THN not only for melanin production(2) but also incorporate the THN scaffold into pharmacologi-cally active meroterpenoids such as neomarinone (3).

Fungi and bacteria use dramatically different enzymatic sys-tems to produce THN from five activated thioester-linked ma-lonyl building blocks. In fungi, THN is synthesized by an iter-ative type I polyketide synthase (PKS), which functions as alarge (�230-kDa) multidomain complex (4). Whereas type IPKSs and their fatty acid synthase ancestors have been in-tensely studied, their three-dimensional domain organizationremains obscure due to their size and complexity. In starkcontrast, bacteria use a structurally simple type III PKS archi-tecturally organized as a homodimeric condensing enzyme(�40 kDa) to synthesize THN from five coenzyme A (CoA)-tethered malonyl units, without the use of additional enzymesor catalytic domains (5–8).

The type III PKS enzyme family (9–11), defined by homologyto chalcone synthase (CHS) (12–15), is currently known toinclude at least 15 functionally divergent (11) �-ketosynthases(KSs) of plant (9–11) and bacterial origin (11, 16). As firstrevealed in the crystal structure of alfalfa CHS (17), each typeIII PKS monomer utilizes a Cys-His-Asn catalytic triad (Fig.1B) within an internal active site cavity that is connected to thesurrounding aqueous phase by a narrow CoA-binding tunnel.The catalytic triad and the buried active site cavity condensean acetyl unit (derived from the decarboxylation of a malonyl-CoA) to a preloaded starter molecule attached by a thioesterbond to the catalytic cysteine (18). Subsequent structure-guided mutagenic and biochemical studies of CHS provided a

* This work was supported by National Institutes of Health GrantAI52443 (to B. S. M. and J. P. N). The Stanford Synchrotron RadiationLaboratory Structural Molecular Biology Program is supported by theDepartment of Energy, Office of Biological and Environmental Re-search, and by the National Institutes of Health, National Center forResearch Resources, Biomedical Technology Program. The costs of pub-lication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and structure factors (code 1U0M) have beendeposited in the Protein Data Bank, Research Collaboratory for Struc-tural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

‡‡ To whom correspondence may be addressed: Division of MedicinalChemistry, University of Arizona, Tucson, AZ 85721. Tel.: 520-626-6931; Fax: 520-626-2466; E-mail: moore@pharmacy.arizona.edu.

§§ To whom correspondence may be addressed: Structural BiologyLaboratory, The Salk Institute for Biological Studies, 10010 N. TorreyPines Rd., La Jolla, CA 92037. Tel.: 858-453-4100 (ext. 1442); Fax:858-452-3683; E-mail: noel@salk.edu.

1 The abbreviations used are: THN, 1,3,6,8-tetrahydroxynaphtha-lene; PKS, polyketide synthase; THNS, THN synthase; CHS, chalconesynthase; STS, stilbene synthase; TAL, triacetic acid lactone (4-hy-droxy-6-methyl- 2H-pyran-2-one); 2-PS, 2-pyrone (TAL) synthase; PEG,polyethylene glycol; KS or KAS, �-ketoacyl synthase; MtFabH, an M.tuberculosis KAS III enzyme; MtPKS18, M. tuberculosis PKS18; HPLC,high performance liquid chromatography; MOPSO, (3-N-morpholino)-2-hydroxypropanesulfonic acid).

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 43, Issue of October 22, pp. 45162–45174, 2004© 2004 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org45162

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

clear picture of the conserved type III PKS polyketide extensionmechanism (Fig. 1B) (19–21). Type III PKSs are both iterativeand multifunctional, having evolved to catalyze an impressiverepertoire of functionally divergent and mechanistically com-plex reactions. These remarkable enzymes typically performthree polyketide extensions of their preferred CoA-activatedstarter molecules (ranging in size from acetyl- to caffeoyl-CoA),prior to catalyzing six-membered ring formation via intramo-lecular Claisen-, aldol-, or lactone-forming cyclization reactions(11). Mutagenic analyses of plant PKSs based upon homologywith CHS or upon the more recent crystal structures of daisy2-pyrone synthase (2-PS) (22) and pine stilbene synthase (STS)(23), facilitated identification of much of the structural andmechanistic underpinnings for plant type III PKS substratespecificity and catalysis (11).

Despite these advances, the structural and mechanistic basisfor the bacterial THNS reaction pathway remains frustratinglyobscure. Whereas most type III PKSs cyclize linear tetraketideintermediates to form phloroglucinol- or resorcinol-based prod-ucts, THNS uniquely catalyzes two successive intramolecularcarbon-carbon cyclization reactions of an all malonate-derivedpentaketide chain (Fig. 1A). How THNS controls chain lengthand reactivity while avoiding the many possible derailmentreactions available to its carboxylated polyketide intermediateis one of this enzyme’s most intriguing features. Furthermore,it is unclear which of several possible cyclization routes to THNare utilized by the enzyme (Fig. 1C), despite the perpetuationin the literature of a hypothetical STS-like initial C-2 3 C-7intramolecular aldol condensation (5). Our recent structuraland mechanistic analysis of STS (23), considered in light ofearlier biomimetic studies (24), instead implicates the C-1 thio-ester carbonyl and the C-10 methylene (activated by the C-11carboxylate moiety) as the most likely electrophile and nucleo-phile, respectively, for intramolecular cyclization. Thus, an ini-tial CHS-like C-6 3 C-1 Claisen condensation could lead to aprearomatic phloroglucinol-like intermediate, or a C-10 3 C-5aldol condensation could lead to the corresponding resorcinol-like intermediate (Fig. 1C). It is also possible that the THNSactive site may juxtapose these two reactive carbons, C-1 andC-10, to first facilitate an unprecedented C-10 3 C-1 Claisencondensation. Although the resulting 10-membered polyketidering seems energetically unfavorable relative to a more stablesix-membered ring, in the context of the THNS active site, itcould form and rapidly collapse via any one of five symmetricintramolecular aldol condensations to form the two stable fusedsix-carbon rings of the THN product (Fig. 1C). Significantly, noon-pathway monocyclic intermediates or their stable aromaticderivatives have ever been observed to result from the THNS-catalyzed reaction, suggesting that the unknown initial cyclicintermediate must indeed strongly facilitate enzymatic forma-tion of the second ring.

Although homology-based analysis of bacterial THNSs pre-dicts a number of unusual active site residues (11, 25), includ-ing four or five exposed cysteine side chain thiol groups thatmight contribute to these enzymes’ unusual physiological reac-tion, sequence divergence between bacterial CHS-like enzymesand the structurally characterized plant PKSs increases thelikelihood of significant model bias and error in homology mod-els derived from distantly related plant PKS structures. In orderto illuminate the mechanistic features responsible for these en-zymes’ catalytic pathway, we undertook a structural and bio-chemical analysis of THNS from the model actinomycete Strep-tomyces coelicolor A3 (2, 26). Toward this end, we present herethe 2.2-Å crystal structure of THNS from S. coelicolor A3 (2),accompanied by additional mutagenic and biochemical investiga-tions into the complicated THNS reaction mechanism.

EXPERIMENTAL PROCEDURES

Expression, Purification, and Mutagenesis—The construction andcharacterization of the S. coelicolor THNS construct in the pHIS8 vectorwas previously reported (8). Mutations were introduced using muta-genic oligonucleotides (Table I) and the QuikChange (Stratagene) sys-tem. The entire coding sequences of mutant constructs were verified bynucleotide sequencing. Wild type and mutant proteins were expressedand purified as previously described (8). For crystallization, THNS wasoverexpressed in Escherichia coli BL21(DE3) cells containing the pLysSplasmid (Invitrogen), purified, buffer-exchanged, and concentrated asdescribed for CHS (19), resulting in final enzyme concentrations of 5–50mg/ml in 12 mM HEPES (pH 7.5), 25 mM NaCl, and 5 mM dithiothreitol.Aliquots were stored at �80 °C.

Crystallization and Data Collection—THNS was crystallized by va-por diffusion in hanging drops consisting of a 1:1 mixture of purifiedprotein solution and crystallization buffer. The crystallization buffercontained 14% (w/v) PEG 8000, 200 mM MgCl2, 100 mM Na�-MOPSObuffer (pH 7.0), 5 mM dithiothreitol, and 3% (w/v) sucrose. A 30-s crystalsoak prior to freezing employed a cryogenic solution differing from thecrystallization buffer due to an increased PEG 8000 concentration (16%(w/v)) and the inclusion of 20% (v/v) glycerol. For heavy atom derivat-ization, native THNS crystals were soaked overnight in crystallizationbuffer differing from the crystallization solution due to an increasedPEG 8000 concentration (16% (w/v)) and the inclusion of 0.1 mM

K2PtCl4 prior to cryogenic freezing as before. Several unsuccessfulheavy atom soaks were screened for anomalous diffraction at the Eu-ropean Synchrotron Radiation Facility, whereas the 2.2-Å native and2.9-Å K2PtCl4 derivative data sets were collected at the Stanford Syn-chrotron Radiation Laboratory. Images were indexed and integratedwith DENZO (27), reflections were merged with SCALEPACK (27), anddata reduction was completed with CCP4 programs (28). THNS crys-tallizes in the P2(1) space group, with unit cell dimensions of a � 76.68Å, b � 69.68 Å, c � 81.14 Å, � � � � 90°, and � � 95.42°.

Structure Determination and Refinement—The S. coelicolor THNSstructure was solved from the 2.9-Å platinum derivative data set.Experimental multiple wavelength anomalous dispersion phases wereobtained using SOLVE (29), based upon two K2PtCl4 sites. The exper-imental electron density maps were improved by bulk solvent densitymodification and automated building with RESOLVE (29). Inspection ofelectron density maps and subsequent model building were performedin O (30). The asymmetric unit contains one physiological THNS dimer.Two copies of a monomeric THNS homology model, generated by MOD-ELLER (31) from an alignment with the alfalfa CHS2 crystal structure(17), were manually inserted into the density-modified electron densitymaps and rebuilt in O. Following the success of a single round ofrefinement in CNS (32) of this rebuilt protein model against the same2.9-Å platinum derivative data set, subsequent iterative rounds ofrefinement in CNS utilized the higher quality 2.2-Å native data set(Table II). Coordinate and parameter files for the PEG and glycerolligands were obtained from the HIC-Up (Hetero-compound InformationCentre-Uppsala) site on the World Wide Web (x-ray.bmc.uu.se/hicup/).PROCHECK (in CCP4) analysis of the final refined crystal structure(see Table II) revealed 86.6% of residue conformations to be in the mostfavored region of the Ramachandran plot, with 11.2% in additionalallowed, 2.2% in generously allowed, and no residues in disallowedregions. THNS residues in generously allowed conformations are lo-cated in poorly ordered surface loops. Coordinates and structure factorsfor THNS have been deposited in the Protein Data Bank (1U0M).Structures were overlaid for comparison using MIDAS (33). Structuralillustrations were prepared with MOLSCRIPT (34) and rendered withPOV-Ray (persistence of vision ray tracer; available on the World WideWeb at www.povray.org).

Enzyme Activities—THNS activity was monitored by reversed-phaseHPLC (8) and TLC. Reversed-phase TLC analysis of flaviolin/TALproduct ratios were obtained from standardized 50-�l room tempera-ture reactions containing 100 mM Tris-HCl buffer (pH 7.5), 10 �M

THNS, and 20 �M [2-14C]malonyl-CoA (PerkinElmer Life Sciences).Reactions were quenched with 5 �l of concentrated HCl and extractedtwice with 50 �l of ethyl acetate. The extracts were dried, dissolved in10 �l of MeOH, applied to Merck C18-silica TLC plates, developed inMeOH/H2O/AcOH (60:40:1, v/v/v), and visualized with radiographyfilm. Apparent kinetic constants kcat and Km for THN production (TableIII) were determined from Eadie-Hofstee plots, using an average ofthree or more independent assays (8).

Flaviolin Labeling and 13C NMR Analysis—E. coli BL21(DE3)pLysScells (Invitrogen) harboring the pHIS8-THNS construct were grown at37 °C in 1.2 liters of Terrific broth containing 50 �g/ml kanamycin and

Polyketide Synthase Control of Reactive Intermediates 45163

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

37 �g/ml chloramphenicol until A600 reached 0.7. Following inductionwith 0.5 mM isopropyl �-D-1-thiogalactopyranoside and the addition of120 mg of [1,2-13C]AcONa (Cambridge Isotope Laboratories, Inc.), cul-tures were shaken at 28 °C overnight, with additional 120-mg aliquots

of [1,2-13C]AcONa added 3 and 6 h after induction for a total of 360 mgof labeled acetate. Cells were removed by centrifugation. The mediasupernatant was acidified with concentrated HCl and extracted withethyl acetate. The crude organic extract was concentrated and sepa-

FIG. 1. Known and possible reactions catalyzed by THNS. A, overall reaction catalyzed by THNS. Portions of the linear pentaketide derivedfrom the starter (first) malonyl unit loaded are pink. B, conserved type III PKS catalytic triad and reaction mechanism for starter loading andpolyketide chain extension by iterative Claisen condensations. The starter moiety is pink, whereas the first acetyl extension is green. C, allplausible Claisen and aldol condensation cyclization routes to THN of the fully elongated linear intermediate of THNS are shown. The mostfavorable initial condensations are green, and secondary cyclizations leading to S- or U-shaped THN cyclization patterns (see “Results”) aredepicted in red and blue, respectively. Predicted substrate labeling patterns are indicated by boldface bonds on the hypothetical monocyclicintermediates and resulting THN.

Polyketide Synthase Control of Reactive Intermediates45164

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

rated by oxalic acid-treated silica flash chromatography employing astepwise chloroform to methanol gradient. The 90% methanol fractionwas further purified by preparative HPLC (YMC ODS-AQ 10 � 250-mmreversed-phase column with a linear solvent gradient of 0.15% triflu-oroacetic acid in water to methanol over 60 min; flow rate � 2.0ml/min). 2.5 mg of enriched flaviolin in Me2SO-d6 was analyzed by 13CINADEQUATE NMR on a Bruker 600-MHz spectrometer.

RESULTS

THNS Crystal Structure—We were unable to solve theTHNS crystal structure by molecular replacement using a2.2-Å native data set and homology models based on plant typeIII PKS structures. Moreover, the in vivo incorporation of sel-enomethionine into the E. coli-expressed S. coelicolor THNSdid not result in stable and active protein preparations. Fortu-nately, a 2.9-Å data set of a K2PtCl4 soak of our native crystalsproduced an anomalous signal suitable for phasing the struc-ture using multiple anomalous diffraction with SOLVE (29).Two platinum sites producing strong peaks on anomalous dif-ference Patterson maps permitted phasing as well as phaseextension and phase improvement by density modification us-ing RESOLVE (29). The quality of the phase information dete-

riorated significantly beyond 3.9 Å, but the initial figure ofmerit-weighted 2Fo � Fc electron density map was interpreta-ble. Whereas RESOLVE automatic model building using thislow resolution phasing information produced only a few erro-neous polyalanine chains, their symmetrical distribution in theasymmetric unit facilitated identification of electron densitycorresponding to the THNS conserved type III PKS �����-foldcore domain and expected physiological homodimeric interface(11). A CHS-derived THNS homology model was then globallypositioned in this experimental electron density map, and eachresidue with interpretable electron density was individuallyrepositioned using O. Following a single round of successfulCNS refinement of our manually placed and partially rebuiltmodel against the 2.9-Å derivative data, the resulting 2Fo � Fc

and Fo � Fc electron density maps indicated a number ofregions needing to be rebuilt. In order to move rapidly andachieve a higher quality final model, subsequent rounds ofrefinement utilized our original native data set extending to2.2-Å resolution. Portions of the model with no apparent elec-tron density (discussed below) were initially deleted, but most

TABLE IMutagenesis of THNS active site cysteines

The mutated codon is underlined.

Mutant Mutagenic oligonucleotide sequence

C106S 5�-G ATC ATC TAC GTC TCC TCC ACG GGC TTC ATG ATG CC-3�C168S 5�-C GTG GCC TGC GAG TTC TCC TCG CTG TGC TAC CAG CC-3�C171S 5�-GC GAG TTC TGC TCG CTG TCC TAC CAG CCC ACC GAC CTC G-3�C184S 5�-GC GTG GGC TCC CTG CTC TCC AAC GGC CTC TTC GGC G-3�

TABLE IICrystallographic data and refinement statistics for the S. coelicolor THNS structure

NativePlatinum derivative

�1 �2 �3

Wavelength (Å) 0.773 1.0718 1.0713 0.9050Resolution (Å) 2.2 2.95Space group P2(1) P2(1)Unit cell dimensions (Å) a � 76.7 a � 76.5

b � 69.7 b � 69.9c � 81.1 c � 81.2

Unit cell dimensions (degrees) � � 95.4 � � 95.3Total reflections 139,869 92,789 102,360 87,984Unique reflections 37,666 32,802 35,270 31,799Completenessa (%) 89.8 (20.6) 87.8 (52.0) 93.9 (76.2) 84.9 (46.7)I/�a 25.5 (4.9) 15.0 (2.3) 13.1 (1.8) 13.9 (1.8)Rsym

a,b 5.8 (11.9) 9.9 (37.9) 10.4 (50.7) 10.6 (49.1)Heavy atom sites 2Phasing figure of merit 0.49Refinement statistics

Rcrystc/Rfree

d (%) 25.2 / 29.3Protein atoms 5308Ligand atoms 50Water molecules 148Root mean square deviation bond lengths (Å) 0.008Root mean square deviation bond angles (degrees) 1.4Average B-factor, protein (Å2) 54.0Average B-factor, solvent (Å2) 67.0

a The number in parenthesis is for the highest resolution shell.b Rsym � ��Ih � �Ih�/ �Ih, where �Ih is the average intensity over symmetry equivalent reflections.c R-factors (Rcryst) � ��Fo � Fc�/�Fo, where summation is over the data used for refinement.d Rfree, same definition as for Rcryst, but it includes only 5% of data excluded from refinement.

TABLE IIISteady state kinetic constants for THN production for wild type THNS and the Cys to Ser mutants

Wild type C106S C168S C171S C184S

kcat � 102 (min�1) 27.4 2.4 NAa NDb 27.9 1.3 47.5 5.4

Km (�M) 2.3 0.6 NA ND 8.5 1.0 1.4 0.5

kcat/Km (s�1 M�1) 1,970 NA ND 549 5,530a NA, no activity.b ND, not determined.

Polyketide Synthase Control of Reactive Intermediates 45165

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

were eventually replaced, since later electron density mapsgreatly improved following iterative rounds of building andrefinement using the 2.2-Å native data set.

Although previous plant PKS-based homology modelingfailed to predict the many subtle backbone differences observedthroughout this first bacterial type III PKS crystal structurereported here, there are no drastic rearrangements of the con-served �����-fold or dimer interface in THNS. Our crystalstructure shows that homology-based assignments of theTHNS catalytic triad and other active site residues are essen-tially accurate, as previously supported by in vitro assays ofStreptomyces griseus THNS point mutants (7).

Interestingly, THNSs feature unusual (�25-residue) exten-sions of their C termini, not found in plant or most otherbacterial CHS-like sequences examined to date. These addi-tional residues may facilitate in vivo protein-protein interac-tions with upstream and downstream enzymes, including theP450 or ORF3 proteins encoded next to the THNS gene inseveral gene clusters obtained from Streptomyces species (2, 6).The elimination of this C-terminal extension was previouslyshown to have no detrimental effect upon the in vitro activity ofS. coelicolor THNS (8). Although we crystallized the full-lengthprotein, this C-terminal extension is completely disordered inthe electron density maps calculated from the most recentTHNS model obtained after multiple rounds of rebuilding andrefinement. Moreover, several adjacent solvent-exposed shortloops on the ����� core domain are also somewhat disordered,with interpretable electron density for many of these residuesemerging only in the final stages of refinement. Whereas theconformational homogeneity of these exposed loops is no doubtlocally perturbed by the adjacent disordered C termini of theTHNS physiological dimer, the THNS unit cell also lacks sta-bilizing crystal lattice contacts in this “upper” half of the phys-iological dimer. Our refined THNS crystallographic B (temper-ature) factors indicate that this segregation of lattice contactsallows subtle global rigid body pivoting movements that uni-formly increase conformational heterogeneity as a function ofthe distance from stabilizing crystal contacts (Fig. 2A).

In contrast to these particular surface-exposed regions adja-cent to the disordered C termini, the electron density mapsnear the active site cavity were unambiguous and provided nohint of conformational disorder. Interestingly, the electron den-sity associated with the interior of THNS clearly indicated theunexpected presence of two different small molecules boundnear the traditional “floor” of the type III PKS active site cavity(if the Cys-His-Asn catalytic triad is considered to project fromthe active site “ceiling” formed by the upper �����-fold coredomain; see Fig. 2A). Subsequent refinement confirmed theseserendipitous small molecules to be a polyethylene glycol(PEG) heptamer, present in the crystallization solution, and aglycerol molecule, subsequently introduced to the PEG-boundTHNS crystal during a final 30-s cryoprotection treatmentimmediately prior to freezing in liquid nitrogen. Although nei-ther of these bound molecules are a physiological substrate,intermediate, or product of the THNS reaction, the bound lo-cation of PEG is surprising and relevant to the functionalinterpretation of our static structure.

One end of this PEG heptamer protrudes into the THNSactive site cavity, but most of the molecule is sequestered in anewly discovered and buried tunnel that extends into the tra-ditionally solid “floor” of the type III PKS active site cavity(Figs. 2, A and B, and 3, A and B). Not predicted by homology,this novel cavity is formed in THNS by the subtle tertiaryrepositioning (relative to plant type III PKS structures; see Fig.2B) of structural elements that come together “under” theactive site cavity, including the outward rigid body displace-

ment of three conserved helices (Fig. 2, A and B). One of thesehelices forms the lower threshold of the type III PKS-conservedCoA-binding tunnel’s intersection with the active site cavity(Fig. 2B), whereas the other two helices form an adjacent andstructurally conserved helix-loop-helix motif, whose solvent-exposed outer surface contains the phosphate-binding basicresidues that mediate CoA complex formation in all CHS-likeenzymes examined to date (11). Significantly, this latter helix-loop-helix motif also participates in THNS crystal lattice con-tacts, which probably select for and stabilize the observed PEG-binding conformation of THNS. Whereas this PEG-boundtunnel conformation is probably not a static apo (non-ligand-bound) conformation in solution, it supports earlier predictionsthat this area of type III PKSs might be subject to dynamicmotion in solution (11). This novel tunnel also suggests the firstmechanistic hypothesis for the cryptic in vitro activity of THNSwhen primed with long acyl starter molecules (35).

Remarkably, the S. griseus THNS (RppA) catalyzes an addi-tional (fifth) polyketide extension step in vitro when primedwith octanoyl-CoA, thus generating a hexaketide from thislarger, nonphysiological starter molecule (35). This unprece-dented result contradicts existing (and well supported) modelsof both physiological and “unnatural” type III PKS catalysis,which correlate polyketide size limits to active site cavity vol-ume (i.e. the promiscuous incorporation of a significantly largerstarter within a given active site cavity should result in fewerpolyketide extension steps). Nonetheless, we were able to du-plicate this previous S. griseus THNS result using our S. coeli-color enzyme (data not shown). In CHS-like enzymes, linearreaction intermediates are covalently attached to either thecatalytic cysteine or bound CoA through thioester linkages atthe C-1 carboxyl carbon. The distance between Cys138 and thenovel PEG-binding tunnel precludes all but the longest (i.e.pentaketide) of the malonyl-primed linear physiological reac-tion intermediates from accessing the THNS PEG-binding tun-nel. However, the PEG-like aliphatic tail of polyketide inter-mediates generated from an octanoyl-CoA starter is easily longenough to access, bind to, and stabilize the open conformationof this tunnel should it become even transitorily available dueto dynamic fluctuations during priming and/or polyketidechain extension (Fig. 3A). Notably, the depth of the observedtunnel is such that it could progressively envelop the aliphatictail of a growing C-1-tethered polyketide intermediate, thuseffectively (and selectively) increasing the active site volumefor the octanoyl-primed reaction. Such progressive bindingwould maintain a relatively taut linear polyketide intermedi-ate conformation inconsistent with the juxtaposition of atomsnecessary for early termination of polyketide extension by thetypical intramolecular Claisen or lactone cyclization routes.Efforts to probe the mechanistic importance of this novel THNScavity are under way, but initial indications suggest, as ex-pected due to its distance from the nucleophilic Cys138, that thePEG tunnel is not important for the physiological production ofTHN (data not shown).

We also considered whether the observed interaction of typ-ically buried polar residues (Asp176 and Asn185) with the boundglycerol ligand might be mechanistically informative. Althoughagain slightly too distant from Cys138 to interact with physio-logical C-1-tethered linear intermediates, we reasoned that aninitial CHS-like thioester-cleaving intramolecular C-6 3 C-1Claisen condensation could liberate a phloroglucinol cyclizationintermediate whose secondary cyclization to form THN mightbe catalyzed within this polar glycerol-binding lower pocket.However, subsequent point mutations of the residues bindingglycerol (Asp176 and Asn185) did not derail THN production(data not shown). We thus concluded that neither PEG nor

Polyketide Synthase Control of Reactive Intermediates45166

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

glycerol binding were relevant to the malonyl-primed physio-logical THNS reaction and turned our attention to the tradi-tional type III PKS active site cavity for insights into thestructural and chemical control of polyketide length and reac-tivity by THNS.

As predicted by homology modeling, the THNS active sitecavity contains a number of unusual amino acid substitutionsrelative to plant type III PKSs. Notably, these residues areconserved in all of the functionally confirmed bacterial THNSsequences discovered to date (5–8). Funa and co-workers re-

cently identified the importance of Tyr224 for THNS starterspecificity (7). Although mutation to Phe or Trp was toleratedby THNS, nonconservative mutations at this position abro-gated the enzymatic loading of the physiological malonylstarter. These results are not surprising, since the substitutionof bulky side chains in place of the spatially equivalent Gly256

residue of alfalfa CHS was shown to affect both starter speci-ficity and the number of catalyzed polyketide chain extensionsteps (22, 36). As first observed in the crystal structure of 2-PS(22), which loads an acetyl moiety as a starter, variation of

FIG. 2. Overall structure of THNS and comparison with evolutionarily related enzymes. A, view of the THNS physiological homodimerand serendipitously bound small molecules. Oxygen atoms of these CPK-depicted ligands are red, whereas carbon atoms of PEG and glycerol aregold and green, respectively. Blue portions of each monomer correspond to three conserved �-helices whose subtle rearrangement in THNS relativeto previously described plant type III PKSs contributes to PEG tunnel formation. Inset, B-factor plot (blue to red from low to high B-factors) of therefined THNS structure, reflecting the lack of crystal contacts near the disordered 25-residue C termini of each monomer. B, comparison of thethree-dimensional architectures of a plant type III PKS (alfalfa CHS; green) and the bacterial type III PKS (S. coelicolor THNS; rose) determinedhere. Type III PKS-conserved access of CoA-tethered substrates to the active site cavity cysteine is demonstrated by depiction of CHS-bound CoA(ball-and-stick). C, similar comparison of the three-dimensional architectures and unusual active site cavity extensions of the bacterial type IIIPKS determined here (THNS; rose) and a Mycobacterium enzyme MtFabH (blue), a divergent member of the distantly related KAS III enzymefamily involved in fatty acid biosynthesis in bacteria and plants. Water molecules demonstrating the location of the unique MtFabH acyl-bindingcavity extension are red, whereas the PEG molecule occupying the similar but distinct tunnel of THNS is gold. CoA is modeled (based on itsCHS-bound position shown in B). The position of the nucleophilic cysteine from each enzyme’s dyad-related monomer illustrates slight changes indimerization. Important structural differences (see “Results”) are also labeled. The figure was prepared with MOLSCRIPT (34) and rendered withPOV-Ray (persistence of vision ray-tracer; available on the World Wide Web at www.povray.org).

Polyketide Synthase Control of Reactive Intermediates 45167

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

steric bulk at this particularly well placed active site position isquite effective for horizontal modulation of the type III PKSactive site cavity volume. These prior studies also surprisinglyrevealed that a too spacious active site cavity can actuallydecrease the ability of a type III PKS to utilize smaller startermolecules, possibly due to excessive conformational freedom ofthe loaded starter (11). Our current structure confirms thatTyr224 indeed contributes to a steric constriction along a hori-zontal direction of the THNS active site cavity relative to CHS(17) and STS (23), no doubt limiting the conformational free-dom of a Cys138-loaded malonyl starter.

Horizontal constriction by Tyr224 is the THNS active site’sonly similarity to 2-PS, which features two additional bulkysubstitutions relative to alfalfa CHS (T197L and S338I). Thiscomplementary vertical constriction of the 2-PS active sitecavity results in this enzyme’s catalysis of only two polyketideextensions of its small acetyl starter (22). In contrast, THNSfeatures less bulky residues relative to CHS at these (T197Cand S338A) and other (T132C and T194C) sterically importantactive site positions. Thus, the THNS active site cavity com-bines a 2-PS-like horizontal constriction with slight expansionsin volume toward the back and bottom of the cavity (with thislatter vertical expansion possibly accentuated by the dynamicflexibility suggested by the appearance of a PEG binding tun-nel). This architectural strategy seems ideal for maintainingthe ability to utilize a small starter molecule (via horizontalconstriction) while providing adequate volume for the addi-tional polyketide chain extension steps and dual cyclizationreactions necessary for THN production (via vertical expan-sion). Interestingly, another CHS-like enzyme that extendspolyketides past the typical tetraketide length was reportedduring the preparation of this manuscript. This rhubarb aloe-sone synthase catalyzes six polyketide extensions of an acetyl-CoA starter, apparently terminated by a CHS-like C-6 3 C-1Claisen cyclization to produce a phloroglucinol ring (37). Thereported aloesone synthase sequence appears to mirror thearchitectural strategy evident in the THNS active site, by com-

bining a horizontally restricting (2PS-like) G256L substitutionwith a downward expanding T197A substitution, relative toalfalfa CHS.

As chemical reasoning predicted (see Introduction), theTHNS active site does not possess the STS-like “aldol switch”thioesterase-like hydrogen bond network we recently linked tothe intramolecular C-2 3 C-7 aldol condensation cyclizationspecificity of STS (23), reinforcing our conclusion that the phys-iological THNS polyketide is likely to be offloaded via an in-tramolecular Claisen condensation severing the C-1 thioesterlinkage. However, other than suggesting a vertical orientationof the THNS fused ring cyclization product, as opposed to thehorizontal orientation of CHS and STS product complexes, theobserved THNS active site topology does not seem to precludeany of the plausible cyclization pathways leading to THN(depicted in Fig. 1C); nor does this newly described topologyindicate how THNS prevents the multitude of derailmentcyclization reactions available to its reactive pentaketide in-termediate. With retention of the starter malonyl’s carboxy-late, even the first six-carbon polyketide intermediate productshould be susceptible to lactonization. Prior to the structuralelucidation of THNS, these factors led us to expect that theTHNS active site cavity would probably exert dramatic stericrestrictions on the polyketide intermediates’ conformationalfreedom. However, when we began to model cysteine-linkedlinear polyketide intermediates in the THNS active site cavity,we were struck by the low degree of apparent steric control ofpolyketide conformation imposed by the enzyme. In fact, wewere able to manually model a surprising range of polyketideconformations, including those leading to typical lactone derail-ment products. Unfortunately, whereas automated dockingprograms are often useful to address substrate, intermediate,and product binding, our experience with this and previousPKS structures indicates that these programs are unable toaccurately predict productively folded polyketide conforma-tions in type III PKS active sites. Although we pursued co-crystallization and crystal soaks with nearly a dozen linear,

FIG. 3. The THNS active site andcomparison with the plant type IIIPKS, CHS. A, covalently linked octanoyl-primed hexaketide intermediate modeledin the THNS active site, illustrating howlong acyl starters could selectively in-crease available active site cavity volumeby accessing the PEG-binding tunnel. B,side-by-side comparison of the PEG-bound THNS active site (left) and the ac-tive site of the naringenin-bound planttype III PKS CHS (right), with function-ally important residues labeled. The con-served catalytic triad residues are high-lighted in black. The figure was preparedwith MOLSCRIPT (34) and rendered withPOV-Ray (persistence of vision ray-trac-er; available on the World Wide Web atwww.povray.org).

Polyketide Synthase Control of Reactive Intermediates45168

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

monocyclic, and fused ring polyketide and cyclization interme-diate analogs, none of these compounds succeeded in binding tothe active site and displacing PEG and glycerol, even in caseswhere co-crystallization altered the dimensions of the crystal-lographic unit cell (data not shown).

13C NMR Analysis of the THN Cyclization Pattern—We con-ducted labeling experiments to elucidate the folded conforma-tion of linear polyketide intermediates leading to the THNskeleton, hoping to further illuminate the bacterial THNS cy-clization pathway. [1,2-13C]Acetate was fed to THNS-express-ing E. coli, and the resulting THN-derived pigment flaviolinwas isolated and characterized (Fig. 4). 13C NMR spectroscopywas utilized to examine whether the central carbon-carbonbond of THN (the bond common to both fused rings) was de-rived from the polyketide chain itself (indicative of an S-shapedconformation of the linear intermediate) or instead representeda bond formed during cyclization of a U-shaped intermediate(Fig. 1C). Whereas each peripheral carbon atom coupled toeach of its neighbors as expected on the basis of the symmet-rical intermediate THN, a one-bond correlation between theflaviolin bridgehead carbon atoms was not evident by an IN-ADEQUATE experiment (data not shown). This experimentverified that THN is formed from a U-shaped polyketide inter-mediate, which unfortunately only eliminates two of the plau-sible cyclization pathways outlined in Fig. 1C.

Analysis of Active Site Cysteines in THNS—As predicted byhomology modeling and confirmed by mutation in both S. gri-seus (7) and S. coelicolor (data not shown) THNS enzymes,Cys138 is the ubiquitous covalent attachment site (equivalent toCHS Cys164) necessary for type III PKS polyketide chain ex-tension (11). In addition to this universal catalytic cysteine, allknown bacterial THNSs also feature three additional activesite cysteines at positions 106, 168, and 171 (corresponding toCHS positions 132, 194, and 197, respectively). This abundanceof cysteine residues is not reflected in nonactive site portions ofTHNS and is unrivaled among other known type III PKS en-zymes, which rarely utilize even a second active site cysteineresidue (11). In CHS and most other plant type III PKSs, these

three active site positions are typically occupied by threonineresidues. In plant PKSs, active site threonine hydroxyl groupsare presumed to help steer and conformationally restrictpolyketide intermediates using polar and/or hydrogen bondinginteractions along the growing polyketide chain (11, 17). Thr132

and Thr197 have specifically been shown to help determinecyclization specificity in STS (23) and CHS (22), respectively.Whereas THNS substitution of cysteines for threonines pro-vides a modest increase in active site cavity volume, the reac-tivity of the resulting THNS active site surface with juxtaposedpolyketide intermediates is likely to be dramatically different,since the thiol proton of cysteine is considerably more acidic(theoretical pKa � 8.7) than the alcoholic proton of threonine(pKa � 15) (38).

Previously, the mutation of each additional S. griseus THNSactive site cysteine to alanine was shown to abolish THN pro-duction (7). The S. coelicolor THNS also contains a fifth activesite cysteine, not conserved in other THNSs, at position 184(CHS Gly211). We carried out more conservative serine muta-tions of all four additional S. coelicolor THNS active site cys-teines in order to probe their specific contribution to THNproduction. Notably, whereas the pKa value of serine’s alcoholicproton (�14) (38) is similar to that of threonine, the cysteineand serine side chains are otherwise nearly identical in polar-ity, steric bulk, and hydrogen-bonding ability.

Product specificities of these four isosteric Cys 3 Ser mu-tants are shown in Fig. 5A. Unlike mutations of the nucleo-philic Cys138, none of these mutations abolish malonyl loadingor polyketide chain extension. Furthermore, serine point mu-tants of Cys168, Cys171, or Cys184 (CHS positions 194, 197, and211) are still able to produce THN (Fig. 5A and Table III), albeitwith slightly increased partitioning of the reaction to form thederailment triketide lactone (relative to wild type THNS).These results indicate that the interactions of these threeTHNS active site cysteine side chains with covalently boundpolyketide intermediates are similar to the roles played by theplant PKS threonines (i.e. steric, polar, and/or hydrogen bond-ing). Conversely, the isosteric mutation of Cys106 (CHS position132) to serine abolishes THN production, with complete parti-tioning of the overall reaction to derailment of the triketideintermediate to form triacetic acid lactone (TAL) (Fig. 5A). Thislatter C106S mutant assay result implicates the importance ofthis thiol moiety’s pKa value in wild type THNS, indicating thattransfer or loss of the more labile thiol proton of Cys106 iscrucial for steering the THNS reaction pathway toward theproduction of THN. Moreover, since the C106S mutant pro-duces only a triketide derailment product, this residue’s chem-ical role must occur prior to either THN cyclization event,neither of which can take place before the tetraketide or pen-taketide stage of polyketide chain extension. These resultsimply that partitioning between two competing mechanisticpathways determines the ultimate fate of the physiologicalmalonyl-primed triketide intermediate of THNS (Fig. 5). One ofthese competing branch point triketide reactions facilitates thetwo additional polyketide chain extension reactions and dualcyclization reactions necessary to biosynthesize THN, whereasthe competing branch of this mechanism leads to triketidelactone derailment in the form of TAL production.

Partial derailment to TAL in wild type THNS (and mutantsconserving cysteine at position 106) indicates that the presenceof the Cys106 thiol does not guarantee additional polyketidechain extensions of the reactive triketide intermediate to formTHN. The complete partitioning of the C106S mutant’s reac-tion toward TAL derailment could in theory be achieved in thismutant THNS either by promotion of the derailment reactionor inhibition of the THN mechanistic pathway. Notably, as in

FIG. 4. In vivo THN labeling experiment using heterologouslyexpressed S. coelicolor THNS to determine the S- or U-shapedTHN cyclization pattern (see Fig. 1C). The expected effects of thesymmetric THN product’s spontaneous oxidation to flaviolin upon labeldistribution are depicted, as is our result: the observation of U-THN-derived flaviolin.

Polyketide Synthase Control of Reactive Intermediates 45169

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

other iterative type III PKS enzymes, the third THNSpolyketide extension step is almost certainly catalyzed in aconventional manner by the conserved (and spatially separate)Cys-His-Asn catalytic triad. Indeed, both wild type and C106STHNS enzymes are capable of extension to the tetraketidestage when primed with an acyl starter (data not shown),further implying that the increased intrinsic reactivity of themalonyl-primed triketide is probably the driving force behindTAL derailment of the wild type THNS reaction pathway. THNor TAL product-based kinetic constants, dependent upon therate-limiting steps of each product’s branched multistep path-way, are unlikely to illuminate the product-determining enzy-matic partitioning ratios we focus upon here. This issue isfurther clouded by uncertainty as to the timing of the THNbiosynthetic pathway’s decarboxylation of the starter-derived�-keto acid or subsequent enolate tautomerization of the re-sulting triketide carbanion (Fig. 5B). Although the presumedmechanism of TAL formation employs decarboxylation at (orprior to) the triketide stage (Fig. 5C), it is unclear whether theTAL pathway branches from THN biosynthesis as the result oftriketide decarboxylation (Fig. 5D) or alternatively by failing to

quench the resulting lactonization-prone enolate tautomer(Fig. 5E).

The theoretical pKa of a cysteine residue’s thiol is �8.7 (38),implying protonation at physiological pH. However, proximityto reactive residues and other local environmental effects cansignificantly lower this pKa, as previously demonstrated for theCHS catalytic triad Cys164 (pKa � 5.5) (20). Interestingly, theimplicated THNS position 106 corresponds to CHS Thr132,which is conserved in STS enzymes but repositioned (via pro-tein backbone conformational changes) to form a hydrogen-bonding network that shuttles electrons from the adjacentGlu192 (universally conserved in all type III PKS enzymes) toactivate a hydrolytic water poised next to the catalytic cysteine(23). This emergent thioesterase-like hydrogen bonding net-work promotes the C-2 3 C-7 aldol-based cyclization whichdistinguishes the STS reaction from that of CHS. The CHS/STS“aldol switch” likewise acts by modulating the partitioning of acommon intermediate, in this case a tetraketide, between twocompeting mechanistic paths that are in fact each accessible toboth CHS and STS wild type enzymes. However, the THNScrystal structure shows the immediate environment and back-

FIG. 5. Mutagenic and biosynthetic analysis of additional active site cysteines in THNS and examination of THN/TAL pathwaypartitioning. A, reversed-phase TLC of wild type (WT) and Cys to Ser mutants of THNS. Production of either flaviolin (derived by spontaneousoxidation of THN) or TAL is shown. B, presumed decarboxylation mechanism for starter-derived �-keto acid moiety of THNS intermediates, whichoccurs between starter loading and formation of THN or TAL. Tautomerization of the resulting carbanion to the enolate form necessary for TALformation is also shown. A gray sphere indicates covalent attachment to CoA or Cys138. C, presumed mechanism of TAL formation from thepredicted enolate tautomer of the decarboxylated malonyl-primed THNS triketide intermediate. D, predicted THN/TAL pathway partitioningbranch point if the THN producing pathway decarboxylation occurs after extension to the tetraketide (“late decarboxylation”). The correspondingderailment reaction (decarboxylation) is shown in pink, whereas the competing THN pathway step (polyketide extension) is in blue. E, predictedTHN/TAL pathway partitioning branch point if THN pathway decarboxylation occurs before extension to the tetraketide (“early decarboxylation”).The corresponding derailment reaction (lactonization) is in pink, whereas the competing THN-pathway step (acid quenching of enolate anion) isin blue.

Polyketide Synthase Control of Reactive Intermediates45170

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

bone conformation of Cys106 to be distinctly CHS-like, with noSTS-like interactions of the Cys106 thiol with the adjacentglutamate, nor any other immediate environmental modifica-tion of this cysteine’s theoretical pKa value. This conclusion isqualitatively supported by iodoacetamide labeling experimentsof the various THNS cysteine mutants, which indicate thecatalytic triad cysteine at position 138 to be the only strongnucleophile in the THNS active site at neutral pH (data notshown). The abundance of these thiols in the cysteine-richTHNS active site cavity may facilitate some indirect modifica-tion of the pKa of Cys106, as might the proximity of two nearbypotential helix dipoles (11).

The THNS Cys138 and Cys106 thiol groups are too far apart(�6.4 Å) to allow direct interaction with a Cys138-loaded startermalonyl group. However, modeling of physiological intermedi-ates indicates that Cys106 can easily access the C-5 position ofeither the C-1-tethered diketide or the triketide intermediatesituated at the critical THN/TAL mechanistic branch point(Fig. 5).

DISCUSSION

THNS conservation of the plant PKS fold and dimer inter-face reaffirms the utility of homology-based assignments ofbacterial CHS-like enzymes’ active site residues. However, theconformational differences manifest throughout the THNSstructure, including subtle main chain perturbations withinthe conserved ����� core domain (Fig. 2B), reflect its signifi-cant sequence divergence from the more closely related plantenzymes (Fig. 2D). Whereas pairwise superimpositions of thethree structurally characterized plant PKS core domains (shar-ing �75% amino acid sequence identity) produce C� root meansquare deviation values of less than 1 Å, similar overlays ofTHNS with these plant PKS structures produce root meansquare deviation values closer to 2 Å. Since bacterial type IIIPKS enzymes share only �25% amino acid sequence identitywith both plant PKSs and each other, future crystal structuresof functionally divergent bacterial CHS-like enzymes are likelyto continue to bear mechanistic fruit, by revealing conforma-tional differences with the typical plant PKS folds beyond thepredictive scope of homology-based models.

Previous comparisons of the various structurally character-ized �����-fold enzyme families reveal a clear structural andmechanistic evolutionary relationship between type III PKSenzymes and the noniterative �-ketoacyl synthase (KAS) IIIenzymes of primary metabolism that initiate fatty acid synthe-sis in plants and bacteria (11, 39, 40). Extensive sequencedivergence between these now distinct families obscureswhether type III PKSs emerged in bacteria or were later ac-quired by lateral transfer from plants (11). As suggested bysequence-based phylogeny, the overall conformational diver-gence observed in this first bacterial type III PKS structureclearly represents slight variations on the type III PKS theme,rather than some hypothetical structural intermediate be-tween plant CHS-like enzymes and their KAS III ancestors(Fig. 2, B and C).

This point is relevant to our crystallographic observationdescribed here of a novel PEG-occupied tunnel (distinct fromthe conserved CoA-binding tunnel) extending into the floor ofthe THNS traditional type III PKS active site cavity and cre-ated by subtle tertiary rearrangements of conserved secondarystructure elements. This unanticipated tunnel suggests thefirst mechanistic model for THNS’s previously inexplicable fur-ther polyketide chain extension reaction of an aliphatic acylsubstrate that is significantly larger than the physiologicalreaction’s malonyl starter, seemingly contradicting models thatcorrelate polyketide chain length to type III PKS active sitevolume (11, 17, 22, 36). The THNS PEG-binding tunnel dem-

onstrates how the enzyme could selectively increase (in astarter substrate-specific manner) the active site volume avail-able to the resulting polyketides. The biological incorporationof THN into prenylated naphthoquinones (3, 5) makes it tempt-ing to speculate that this tunnel could, by binding isoprenoidmoieties of prenylated THN derivatives, facilitate some in vivoregulatory inhibition of THNS.

Although clearly reinforced by crystal lattice contacts, thePEG-binding conformation supports previous hypotheses thatthis region of CHS-like enzymes might undergo some degree ofdynamic motion in solution (11). It seems possible that othertype III PKS enzymes could transiently form a similar tunnel.Notably, previous CHS-like enzyme crystal structures (17, 22,23) reveal smaller pockets of buried water in this “subfloor”region, and this area is completely disordered in one crystalform of this family’s noniterative KAS III evolutionary ancestor(41). Whereas typical KAS III enzymes catalyze one malonyl-ACP-mediated extension of a preloaded acetyl starter using thesame catalytic triad conserved in CHS-like enzymes, one diver-gent Mycobacterium tuberculosis KAS III (MtFabH) displaysspecificity for a long aliphatic acyl starter (42). This nonitera-tive ketosynthase binds its acyl starter in a unique pocket atthe dimer interface (42) that is distinct from the THNS PEG-binding tunnel we observe here (Fig. 2C). This MtFabH bindingpocket is created by the appearance and dimerization, in anotherwise typical KAS III fold, of a unique helical motif result-ing from an insertion into a KAS III �-finger loop alreadysignificantly longer than the equivalent �-finger in type IIIPKS enzymes. Long acyl starters gain access to this tunnelfrom the active site cavity through steric reduction of the onlyresidue lining each traditional active site cavity that is contrib-uted by the dyad-related monomer (equivalent to THNS Met111

and CHS Met137). Recently, one of three M. tuberculosis CHS-like enzymes (MtPKS18) was found to catalyze the extension oflong acyl starters in vitro (43). These authors predicted theexistence of a MtFabH-like acyl-binding tunnel in MtPKS18,based upon that enzyme’s MtFabH-like steric reduction of CHSposition 137. However, the primary sequence of MtPKS18clearly clusters with other type III PKSs, with MtPKS18 show-ing more sequence identity with conserved plant PKSs than dothe remaining two M. tuberculosis type III PKS sequences.More specifically, no significant sequence deviation from otherCHS-like enzymes is apparent in the relevant MtFabH-modi-fied �-finger region. Furthermore, even if a MtFabH-like inser-tion were to occur in a corresponding type III PKS �-finger loop,the dimerization of such a hypothetical extension would beprevented by the presence of a type III PKS-conserved long�-helix that is much shorter in KAS III enzymes (Fig. 2, B andC). Therefore, MtPKS18 (or any other hypothetical aliphaticacyl-binding type III PKS enzyme) seems much more likely toaccess a THNS-like tunnel than an MtFabH-like one.

In the Introduction, we pointed out the numerous plausibleintramolecular polyketide cyclization routes leading to THN(Fig. 1C). Our crystal structure supports the proposal that aninitial STS-like C-23 C-7 aldol condensation (23) is unlikely inthe bacterial THNS pathway, although the resulting resorcin-olic acid was previously (but not conclusively) identified asa fungal THNS intermediate cyclization product (44). OurS. coelicolor THNS crystal structure, supported by previousmutations in S. griseus THNS (7), suggests that a “horizontalrestriction/vertical expansion” trend in active site topology isresponsible for maximizing polyketide extension steps whilepreserving type III PKS specificity for small starter molecules.This model is distinct from an older one arising from previousstudies of 2-PS, where multidirectional restriction (22) or hor-izontal restriction alone (36) decreased starter size as well as

Polyketide Synthase Control of Reactive Intermediates 45171

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

the number of polyketide extension reactions. This new adden-dum to the older “steric modulation” model (11) is independ-ently supported by a similar trend in the recently reportedsequence of a rhubarb aloesone synthase, which synthesizes aheptaketide using an acetyl starter (37).

Unfortunately, our crystal structure did not illuminate thespecific bacterial THN cyclization pathway. The failure of nu-merous potential intermediate and product analogues to dis-place PEG and glycerol from the THNS active site indicatesthat further crystallographic analysis of the THNS reactionwill require a different crystal form. However, our NMR anal-ysis of 13C-enriched flaviolin conclusively demonstrated thatthis heterologously expressed S. coelicolor type III PKS medi-ates a U-shaped (rather than S-shaped) THN cyclization pat-tern (Fig. 4). This eliminates two of the theoretical routes toTHN depicted in Fig. 1C. The same labeling pattern was pre-viously observed from similar (but not heterologous) in vivolabeling experiments of naphterpin (46) in Streptomyces sp.Strain CL190 (previously known as Streptomyces aeriouvifer)as well as neomarinone (3) in the marine sediment-derivedactinomycete CNH-099. It is important to note that our desig-nation of S- or U-shaped intermediates is distinct from an oldercategorization of type I or type II PKS-derived (and predomi-nantly U-shaped) asymmetric fused ring polyketides as S- orF-mode (for Streptomyces or fungal) (45). This prior classifica-tion, based upon the number of intact acetate labels incorpo-rated into the first cyclized ring, is impractical for symmetricalproducts such as THN.

The isosteric Cys3 Ser mutations of the THNS cysteine-richactive site indeed implicate either the Cys106 thiol proton or itsconjugate base as crucial for THN formation (Fig. 5A). Theseresults further suggest that reaction partitioning between thecompeting physiological THN and derailment TAL pathwaysoccurs at a single critical branch point acting at the triketidestage of polyketide chain extension. Recently, another type IIIPKS example of enzymatic partitioning between competingreactions at a single mechanistic branch point was discoveredto be similarly responsible for modulating the cyclization spec-ificities of STS and CHS enzymes (23). In this previous case,overall (product-based) kinetic comparisons of these multistepbiosynthetic reactions failed to illuminate the critical (but ap-parently not rate-limiting) mechanistic differences betweenthem. This is not surprising, since the subtle inhibition orpromotion of one of two competing reaction rates can easilymodulate the ratio (partitioning) between competing multisteppathways without affecting either pathway’s overall kcat, un-less the affected branch point reaction also happens to be thatpathway’s rate-limiting step. With our present data, it is im-possible to ascertain whether the specific mechanistic role ofCys106 in facilitating THNS triketide chain extension to thetetraketide stage occurs via promotion of this extension reac-tion or instead via inhibition of the competing derailment re-action. Since modulating either reaction’s rate constants di-rectly impacts the partitioning ratio, while also alteringsubstrate availability for the competing reaction, this questionmight be argued to be something of a semantic issue.

It is not known exactly when in the THN biosynthetic path-way’s decarboxylation of the starter-derived �-keto acid or sub-sequent enolate tautomerization of the resulting triketide car-banion occurs (Fig. 5B). Although the presumed mechanism ofTAL formation employs decarboxylation at (or prior to) thetriketide stage (Fig. 5C), it is unclear whether the TAL path-way branches from THN biosynthesis as the result of triketidedecarboxylation (Fig. 5D) or alternatively by failing to quenchthe resulting lactonization-prone enolate tautomer (Fig. 5E).Due to these complications, we were forced to consider several

mechanistic roles for Cys106 that seemed capable of modulatingthe balance of competing THN/TAL pathway branch pointreactions.

We also considered mechanisms utilizing either the proto-nated thiol or charged thiolate form of Cys106, since its pKa

value has not yet been experimentally determined and mayapproach physiological pH due to influences in the THNS ac-tive site cavity, such as the abundance of other cysteines,proximity to two helical dipoles, or even the positioning ofnearby polyketide intermediates. We arrived at three compet-ing and chemically plausible mechanistic hypotheses for therole of Cys106 involving triketide intermediates in the THNSreaction mechanism that are consistent with the current bio-chemical and structural data (Fig. 6).

We initially considered a fourth possible mechanistic sce-nario, involving Cys106 protonation of the diketide or triketide(starter malonyl-derived) terminal carboxylate moiety. Such aprotonation event, should it occur, would indeed inhibit decar-boxylation of the triketide intermediate (necessary for TALformation) and thus has the potential to alter reaction parti-tioning at the THN/TAL triketide branch point. However, be-cause protonated carboxylates are significantly more acidic(pKa 4–5) (46) than the typical �8.7 pKa of a cysteine thiol, thishypothetical proton transfer seems much less probable thanthe following three alternative proposals.

Whereas TAL formation clearly requires decarboxylation ator prior to the triketide stage, it is unclear if premature decar-boxylation is in fact the watershed branch point reaction doom-ing THN pathway intermediates to TAL derailment. It is pos-sible that “early” decarboxylation sometimes (or always)accompanies THN biosynthesis but that protonation of thedecarboxylated triketide’s enolate anion (the tautomeric nu-cleophile leading to TAL lactone formation; see Fig. 5B) by theCys106 thiol quenches this reactive, triketide enolate anion,thus allowing polyketide chain extension to continue past thetriketide stage (Figs. 5E and 6A).

A second partitioning scenario involves the Cys106-mediatedcatalysis of intermediate polyketide cis-enolization (Fig. 6B). Al-though polyketides are biosynthetically formed (Fig. 1B) andusually depicted as �-ketones, polyketide ketones readily un-dergo acid- or base-catalyzed tautomerization to form anionic andprotonated enols (47). Whereas a comparison of the fundamentalbond energies of keto and enol tautomers reveals the keto form tobe slightly more thermodynamically stable, keto/enol equilibriaare influenced by a number of factors, including chain conjuga-tion, pH, and polarity of solvent (47). The enol form of �-ketonesin aqueous solution is unusually favorable due to a combinationof enol �-bond conjugation and internal hydrogen bonding withthe adjacent �-keto moiety (47). Notably, it is not known whateffect the steric confines of the active site cavity has on tautomer-ization rates and keto/enol equilibria of the various short livedPKS linear polyketide intermediates.

When we positioned a modeled C-1-tethered carboxyl-bear-ing triketide’s C-5 ketone near the Cys106 thiol moiety ofTHNS, we noticed that the location of Cys106 at the back edgeof the active site cavity, in combination with the earlier notedsteric constriction imposed by the bulky Tyr224 side chain ofTHNS, predisposes the seven-carbon triketide to adopt a nearlysyn-periplanar conformation around the C-5 position. Further-more, this cis-like conformation also positions both the C-5carbonyl oxygen and adjacent C-6 methylene protons outwardand near the Cys106 thiol group. Hypothetical hydrogen bondsbetween the Cys106 thiol and both the triketide’s C-5 carbonyloxygen and its adjacent C-6 methylene proton would facilitatea proton shuttle as depicted in this second mechanistic pro-posal for Cys106 (Fig. 6B). This ideally positioned thiol moiety

Polyketide Synthase Control of Reactive Intermediates45172

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

can now donate its proton to the C-5 oxygen while picking upthe acidic C-6 methylene proton in a catalytic cycle that regen-erates the thiol catalyst while accomplishing both acid/basecatalytic events necessary for enolization. Since the polyketideconformation is already close to cis periplanarity, no significantbarrier remains to the completion of enolization via C-5–C-6�-bond formation. Moreover, the pKa of a polyketide methyleneis �9 (47), nearly identical to that of the nearby thiol, therebysupporting this hypothetical proton shuttle model. In additionto the reduction in rotational freedom associated with orbitalrehybridization of the C-6 methylene position from sp3 to sp2,this hypothetical C-5–C-6 cis-�-bond formation would addition-ally alter the bond angles involving the C-6 position from 109 to120°. Although these subtle geometric effects may significantlyimpact conformational equilibria within the steric confines ofthe THNS active site, C-5–C-6 enolization more importantlydirectly impacts the reactivity of the C-5 keto group and C-6methylene moiety positions, both of which probably play rolesin TAL-forming derailment of the physiological malonyl-primed THNS reaction (Fig. 5B).

Our third mechanistic proposal for the THN/TAL partition-ing role of Cys106 involves the formation of a hemithioketal“protecting group” at the C-5 ketone position of the C-1-teth-ered triketide carboxylate intermediate (Fig. 6C). In this sce-nario, nucleophilic addition of the thiolate version of Cys106 tothe C-5 carbonyl carbon provides a stabilizing second cysteinetether to the growing intermediate, thus steering the reactivenascent chain toward an eventual aldol condensation reactionof the linear pentaketide. The C-5 hemithioketal group addi-tionally serves to stabilize the adjacent �-keto acid and preventspontaneous decarboxylation. Whereas we presume thatCys106 is most likely protonated at physiological pH, hemith-ioketal formation also requires the carbonyl oxygen to pick upa proton. It is conceivable that nucleophilic attack on C-5 byCys106 could be accompanied by transfer of the thiol proton(perhaps via a water molecule) to this C-5 carbonyl oxygen,lowering a potential energetic barrier to this proposed reaction.Our modeling in the THNS active site predicts a malonate-likeintramolecular interaction between the triketide’s C-7 terminalcarboxylate and the resulting hemithioketal C-5 alcohol group(Fig. 6C) that is likely to stabilize both moieties during the

remaining polyketide extension steps necessary for THNproduction.

Immobilizing the reactive starter-derived end of the triketideto Cys106 at the back of the active site cavity would presumablyfacilitate the subsequent binding of malonyl-CoA, probably in-creasing the efficiency and therefore the rate of subsequentpolyketide extension steps, thus allowing tetraketide formationto out-compete TAL derailment of the lactonization-pronetriketide intermediate. Surprisingly, our modeling also indi-cates that the second of the two remaining THN pathwayextensions of this hemithioketal-tethered triketide (to producea C-1 thioester- and C-9 hemithioketal-tethered pentaketide)generates enough “slack” in the polyketide chain to allow aCHS-like C-63 C-1 intramolecular Claisen condensation thatwould terminate chain extension by cleaving the C-1 thioesterbond to Cys138. The increased conformational freedom of theresulting phloroglucinol-like cyclization intermediate wouldthen allow formation of the (U-shaped) THN skeleton via one oftwo available secondary aldol condensations (Fig. 1C). Theresulting local disruption of the terminal carboxylate’s stabili-zation of this hypothetical hemithioketal, in combination withfacile aromatization of the THN fused ring skeleton, wouldprobably reverse the hemithioketal attachment to liberate theTHN product.

Our recent structural and mechanistic elucidation of the STS“aldol switch” mechanism (23) suggests that most type III PKSenzymes, which typically catalyze the formation of single phlor-oglucinol- or resorcinol-like rings (11), probably modulatecyclization specificity by taking advantage of the differentintrinsic reactivity of their tetraketide intermediates’ thio-ester-linked (C-63 C-1 Claisen-promoting) or free carboxylate(C-2 3 C-7 aldol-promoting) C-1 carbons (23). In contrast tothese typical type III reactions, THNS produces a longer andmore reactive polyketide intermediate, from which it producesthe bicyclic THN skeleton with fidelity. Our current structuralelucidation of THNS reveals a novel cavity extension that pro-vides an explanation for the peculiar extra in vitro polyketideextension of a large aliphatic starter in THNS (35), whereasour mutagenic analysis of THNS implicates an additional ac-tive site cysteine residue at position 106 that is vital for THNbiosynthesis. Aside from THNS, only Pseudomonas PhlD en-

FIG. 6. Illustration of possible enzymatic mechanisms (see “Results”) for Cys106 structural control of polyketide reactivity inTHNS. Relevant reactive C-1-tethered intermediates of the THN or TAL derailment pathway (see Fig. 5B) are modeled in gold in the THNS activesite, with protons and keto/enol oxygens shown in white and red, respectively. Protons were added to the non-hydrogen-containing THNS crystalstructure model to reflect the presumed protonation of these additional active site cysteines at physiological pH. A, Cys106-mediated protonationof the decarboxylated triketide enolate. B, Cys106-mediated proton shuttle leading to cis-enolization of the triketide carboxylate. Enolizationdecreases polyketide rotational freedom while directly altering the cyclization reactivity of both affected keto/enol and methylene positions. C,Cys106-mediated hemithioketal formation at the triketide carboxylate stage of polyketide elongation. Here, the putative hemithioketal interme-diate product of polyketide protection is shown. This mechanism would provide a second polyketide attachment site, thus preventing cyclization-promoting folded conformations of reactive polyketide intermediates. The figure was prepared with MOLSCRIPT (34) and rendered with POV-Ray(persistence of vision ray-tracer; available on the World Wide Web at www.povray.org).

Polyketide Synthase Control of Reactive Intermediates 45173

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

zymes conserve this additional cysteine residue (11), perhapssuggesting a similar mechanism in these related bacterial typeIII PKSs. Our subsequent modeling of the THNS reactionprompted three novel and plausible proposals (Fig. 6) for themechanistic role of Cys106 in partitioning between the compet-ing THN/TAL pathways. These mechanistic proposals mighthave implications for the biosynthesis of polyaromatic naturalproducts by type I and II iterative PKSs. Generally, thequenching of reactive enolate anions, selective catalysis of eno-lization, or intermediate hemithioketal immobilization by cys-teine side chains positioned along an elongated polyketidebackbone, in the context of a solvent-excluded active site, couldactively limit and focus intramolecular bond formation to spe-cific subsets of polyketide keto tautomers.

Acknowledgments—Portions of this research were carried out at theStanford Synchrotron Radiation Laboratory, a national user facility op-erated by Stanford University on behalf of the United States Departmentof Energy, Office of Basic Energy Sciences. We also acknowledge theEuropean Synchrotron Radiation Facility for access to the synchrotronradiation facilities in Grenoble, France. We thank J. Kalaitzis andT. O’Hare (University of Arizona) for help in acquiring the NMR data andin designing the THNS cysteine mutant oligonucleotides, respectively.

Note Added in Proof—After acceptance and publication online, thestructure of MtPKS18 appeared (see Sankaranarayanan, R., Saxena,P., Marathe, U. B., Gokhale, R. S., Shanmugam, V. M., and Rukmini, R.(2004) Nat. Struct. Mol. Biol., in press) confirming the prediction madeabove.

REFERENCES

1. Feng, B., Wang, X., Hauser, M., Kaufmann, S., Jentsch, S., Haase, G., Becker,J. M., and Szaniszlo, P. J. (2001) Infect. Immun. 69, 1781–1794

2. Ueda, K., Kim, K. M., Beppu, T., and Horinouchi, S. (1995) J. Antibiot. (Tokyo)48, 638–646

3. Kalaitzis, J. A., Hamano, Y., Nilsen, G., and Moore, B. S. (2003) Org. Lett. 5,4449–4452

4. Fujii, I., Mori, Y., Watanabe, A., Kubo, Y., Tsuji, G., and Ebizuka, Y. (2000)Biochemistry 39, 8853–8858

5. Funa, N., Ohnishi, Y., Fujii, I., Shibuya, M., Ebizuka, Y., and Horinouchi, S.(1999) Nature 400, 897–899

6. Cortes, J., Velasco, J., Foster, G., Blackaby, A. P., Rudd, B. A. M., andWilkinson, B. (2002) Mol. Microbiol. 44, 1213–1224

7. Funa, N., Ohnishi, Y., Ebizuka, Y., and Horinouchi, S. (2002) Biochem. J. 367,781–789

8. Izumikawa, M., Shipley, P. R., Hopke, J. N., O’Hare, T., Xiang, L., Noel, J. P.,and Moore, B. S. (2003) J. Ind. Microbiol. Biotechnol. 30, 510–515

9. Schroder, J. (1997) Trends Plant Sci. 2, 373–37810. Schroder, J. (2000) Recent Adv. Phytochem. 34, 55–8911. Austin, M. B., and Noel, J. P. (2003) Nat. Prod. Rep. 20, 79–11012. Kreuzaler, F., and Hahlbrock, K. (1972) FEBS Lett. 28, 69–7213. Kreuzaler, F., and Hahlbrock, K. (1975) Arch. Biochem. Biophys. 169, 84–9014. Heller, W., and Hahlbrock, K. (1980) Arch. Biochem. Biophys. 200, 617–61915. Reimold, U., Kroeger, M., Dreuzaler, F., and Hahlbrock, K. (1983) EMBO J. 2,

1801–180516. Moore, B. S., and Hopke, J. N. (2001) ChemBioChem. 2, 35–3817. Ferrer, J. L., Jez, J. M., Bowman, M. E., Dixon, R. A., and Noel, J. P. (1999)

Nat. Struct. Biol. 6, 775–784

18. Lanz, T., Tropf, S., Marner, F. J., Schroder, J., and Schroder, G. (1991) J. Biol.Chem. 266, 9971–9976

19. Jez, J. M., Ferrer, J. L., Bowman, M. E., Dixon, R. A., and Noel, J. P. (2000)Biochemistry 39, 890–902

20. Jez, J. M., and Noel, J. P. (2000) J. Biol. Chem. 275, 39640–3964621. Suh, D. Y., Kagami, J., Fukuma, K., and Sankawa, U. (2000) Biochem. Bio-

phys. Res. Commun. 275, 725–73022. Jez, J. M., Austin, M. B., Ferrer, J., Bowman, M. E., Schroder, J., and Noel,

J. P. (2000) Chem. Biol. 7, 919–93023. Austin, M. B., Bowman, M. E., Ferrer, J., Schroder, J., and Noel, J. P. (2004)

Chem. Biol., in press24. Harris, T. M., and Harris, C. M. (1986) Pure Appl. Chem. 58, 283–29425. Moore, B. S., Hertweck, C., Hopke, J. N., Izumikawa, M., Kalaitzis, J. A.,

Nilsen, G., O’Hare, T., Piel, J., Shipley, P. R., Xiang, L., Austin, M. B., andNoel, J. P. (2002) J. Nat. Prod. 65, 1956–1962

26. Bentley, S. D., Chater, K. F., Cerdeno-Tarraga, A. M., Challis, G. L., Thomson,N. R., James, K. D., Harris, D. E., Quail, M. A., Kieser, H., Harper, D.,Bateman, A., Brown, S., Chandra, G., Chen, C. W., Collins, M., Cronin, A.,Fraser, A., Goble, A., Hidalgo, J., Hornsby, T., Howarth, S., Huang, C. H.,Kieser, T., Larke, L., Murphy, L., Oliver, K., O’Neil, S., Rabbinowitsch, E.,Rajandream, M. A., Rutherford, K., Rutter, S., Seeger, K., Saunders, D.,Sharp, S., Squares, R., Squares, S., Taylor, K., Warren, T., Wietzorrek, A.,Woodward, J., Barrell, B. G., Parkhill, J., and Hopwood, D. A. (2002) Nature417, 141–147

27. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–32628. Dodson, E. J., Winn, M., Ralph, A. (1997) Methods Enzymol. 277, 620–63329. Terwilliger, T. C., and Berendzen, J. (1999) Acta Crystallogr. Sect. D Biol.

Crystallogr. 55, 849–86130. Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1993) Acta Crys-

tallogr. Sect. D Biol. Crystallogr. 49, 148–15731. Sali, A., and Blundell, T. L. (1993) J. Mol. Biol. 234, 779–81532. Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-

Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read,R. J., Rice, R. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr.Sect. D Biol. Crystallogr. 54, 905–921

33. Ferrin, T. E., Huang, C. C., Jarvis, L. E., and Langridge, R. (1988) J. Mol.Graph. 6, 13–27

34. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946–95035. Funa, N., Ohnishi, Y., Ebizuka, Y., and Horinouchi, S. (2002) J. Biol. Chem.

277, 4628–463536. Jez, J. M., Bowman, M. E., and Noel, J. P. (2001) Biochemistry 40,

14829–1483837. Abe, I., Utsumi, Y., Oguro, S., and Noguchi, H. (2004) FEBS Lett. 562, 171–17638. Kyte, J. (1995) Structure in Protein Chemistry, p. 59, Garland, New York39. Qiu, X., Janson, C. A., Konstantinidis, A. K., Nwagwu, S., Silverman, C.,

Smith, W. W., Khandekar, S., Lonsdale, J., and Abdel-Meguid, S. S. (1999)J. Biol. Chem. 274, 36465–36471

40. Davies, C., Heath, R. J., White, S. W., and Rock, C. O. (2000) Struct. Fold. Des.8, 185–195

41. Qiu, X., Janson, C. A., Smith, W. W., Head, M., Lonsdale, J., and Konstantini-dis, A. K. (2001) J. Mol. Biol. 307, 341–356

42. Scarsdale, J. N., Kazanina, G., He, X., Reynolds, K. A., and Wright, H. T.(2001) J. Biol. Chem. 276, 20516–20522

43. Saxena, P., Yadav, G., Mohanty, D., and Gokhale, R. S. (2003) J. Biol. Chem.278, 44780–44790

44. Fujii, I., Mori, Y., Watanabe, A., Kubo, Y., Tsuji, G., and Ebizuka, Y. (1999)Biosci. Biotechnol. Biochem. 63, 1445–1452

45. Thomas, R. (2001) Chembiochem. 2, 612–62746. Brown, H. C., McDaniel, D. H., and Haaflinger, O. (1955) in Determination of

Organic Structures by Physical Methods (Braude, E. A., and Nachod, F. C.,eds) Academic Press, Inc., New York

47. Smith, M. B., and March, J. (2001) March’s Advanced Organic Chemistry:Reactions, Mechanisms, and Structure, 5th Ed., John Wiley & Sons, Inc.,New York

Polyketide Synthase Control of Reactive Intermediates45174

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Jean-Luc Ferrer, Bradley S. Moore and Joseph P. NoelMichael B. Austin, Miho Izumikawa, Marianne E. Bowman, Daniel W. Udwary,

Control of Reactive Polyketide IntermediatesCrystal Structure of a Bacterial Type III Polyketide Synthase and Enzymatic

doi: 10.1074/jbc.M406567200 originally published online July 20, 20042004, 279:45162-45174.J. Biol. Chem. 

  10.1074/jbc.M406567200Access 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/279/43/45162.full.html#ref-list-1

This article cites 42 references, 7 of which can be accessed free at

by guest on March 22, 2018

http://ww

w.jbc.org/

Dow

nloaded from