Biochem. J. (2014) 460, 157–163 (Printed in Great Britain)

157

doi:10.1042/BJ20140239

John E. CRONAN*1 *Departments of Microbiology and Biochemistry University of Illinois, Urbana, IL 61801, U.S.A.

ACPs (acyl carrier proteins) play essential roles in the synthesis of fatty acids, polyketides and non-ribosomal polypeptides. ACP function requires the modification of the protein by attachment of 4 -phosphopantetheine to a conserved serine residue. The phosphopantetheine thiol acts to tether the starting materials and intermediates as their thioesters. ACPs are small highly soluble proteins composed of four α-helices. The helices form a bundle that acts as a hydrophobic sleeve that sequesters the acyl chains and activated thioesters from solvent. However, in the synthesis of fatty acids and complex lipids the enzymes of the pathway must access the thioester and the proximal carbon atoms in order to perform the needed chemistry. How such access is provided without exposure of the acyl chains to solvent has

been a longstanding question due to the lack of acyl-ACP– enzyme complexes, a situation generally attributed to the brevity of the interactions of acyl-ACPs with their cognate enzymes. As discussed in the present review the access question has now been answered by four recent crystal structures, each of which shows that the entire acyl chain plus the 4 -phosphopantetheine prosthetic group partitions from the ACP hydrophobic sleeve into a hydrophobic pocket or groove of the enzyme protein, a process termed chain flipping.

INTRODUCTION

considerable disruption of acyl-ACP structure that would expose the acyl chain to solvent. On the basis of the crystal structure of the yeast multifunctional fatty acid synthase, in which the ACP domains appear caught in the act, Leibundgut et al. [6] put forth a ‘switchblade’ mechanism in which the PPant prosthetic group and the acyl chain initially buried within the hydrophobic ACP core would entirely flip into the hydrophobic active site of a fatty acid synthetic enzyme (Figure 1). That is, rather than discretely altering the ACP structure to allow access of an enzyme to the proximal part of the acyl chain, the entire acyl chain plus much of the prosthetic group is transferred from the ACP core into the enzyme active site. Although this chain transfer mechanism has been called the switchblade mechanism, this term implies that an appreciable energetic force (the switchblade spring) is exerted to expel the acyl chain from the ACP hydrophobic pocket. There is no evidence or proposed mechanism for such an active expulsion. Moreover, the switchblade analogy suggests that the acyl chain goes from a sequestered environment to a very different and unstructured environment. This is not the case in the extant crystal structures where the chain goes from one hydrophobic environment to another. Given the precedent of the base-flipping mechanism of DNA modification and repair enzymes [7, 8], chain flipping seems a more appropriate term and will be used herein. The chain-flipping mechanism proposed for the yeast synthase [6] was supported by the higher-resolution structures obtained for two biotin synthetic enzymes [9,10]. However, it could readily be argued that these structures could be exceptions to the rule because both enzymes catalyse acyl chain transformations targeted seven carbons downstream of the thioester bond (discussed below). This raised the question of whether or not enzymes that need to interact only with the thioester-proximal portion of the acyl chain would

The mechanism of saturated fatty acid synthesis is strongly conserved between bacteria and eukaryotes (archaea synthesize isoprenoid-based lipids), although the catalytic entities are found in markedly different protein arrangements [1–4]. In most bacteria, plus mitochondria and chloroplasts, fatty acid synthesis is catalysed by a series of discrete proteins each of which catalyses a distinct reaction. In contrast, in eukaryotes (and a few bacteria) the individual enzyme reactions are catalysed by distinct domains of very large polyfunctional proteins [1–4]. In both cases the growing fatty acid chain is tethered to the thiol of the PPant (4 phosphopantetheine) prosthetic group attached to a small protein (or domain) called the ACP (acyl carrier protein) [3]. In bacteria the acyl chain of a finished ACP-bound acyl chain is used in synthesis of complex lipids such as membrane phospholipids. Although in recent years high-resolution structures of ACP and numerous other bacterial fatty acid synthetic proteins have been obtained (together with generally lower resolution structures of the eukaryotic polyfunctional protein complexes), structures containing ACP-bound substrates have not been available. This situation was not for lack of effort, and the failure to observe such complexes has been ascribed to the transient and weak nature of the interactions [3–5]. As discussed below, it was known that the growing fatty acid chain is sequestered within a bundle of ACP helices. Although this arrangement nicely sequesters the acyl chain from solvent, lipid synthesis requires that the fatty acid synthetic and acyl transfer enzymes are able to gain access to the carbon atoms proximal to the acyl-ACP thioester in order to catalyse the needed chemistry. In the absence of structural data we could only guess how access could be provided without

Key words: acyl carrier protein (ACP), chain flipping, phosphopantetheine thiol, saturated fatty acid synthesis.

Abbreviations: ACP, acyl carrier protein; AcpS, holo-[acyl-carrier-protein] synthase; FabA, 3-hydroxyacyl-ACP dehydratase/isomerase; LpxD, UDP-3O-(3-hydroxymyristoyl) glucosamine N -acyltransferase; PPant, 4 -phosphopantetheine. 1 email [email protected]  c The Authors Journal compilation  c 2014 Biochemical Society

Biochemical Journal

The chain-flipping mechanism of ACP (acyl carrier protein)-dependent enzymes appears universal

www.biochemj.org

REVIEW ARTICLE

158

Figure 1

J. E. Cronan

Structure of decanoyl-ACP and cartoon depiction of the chain-flipping mechanism

(A) The structure of C10 acyl-ACP modified from PDB code 2FAE using Chimera [56]. The PPant prosthetic group and acyl chain is in CPK (Corey–Pauling–Koltun) colouring and the helices are numbered (white lettering). The structure following helix I is the small pseudo helix formed between Gln19 and Glu21 (E. coli numbering) sometimes seen in crystal structures, but not in NMR structures. (B) A cartoon of the structure of (A) where the prosthetic group except for the thiol (yellow sphere) is depicted as blue spheres and the acyl chain as black spheres. (C) The chain-flipping mechanism in which the acyl chain and distal end of the prosthetic group partition from the hydrophobic ACP pocket into the hydrophobic pocket of enzyme X.

also use the chain-flipping mechanism. Two new papers make it probable that this is the case for all ACP–enzyme interactions. Moreover, these papers are concerned with the dominant reactions undergone by acyl-ACPs, fatty acid synthesis and acylation of a water-soluble precursor in the synthesis of a complex lipid. In the former case Nguyen et al. [5] overcame the transient nature of the fatty acid synthesis reactions by use of a mechanismbased probe to obtain precise covalent cross-linking of an acyl-ACP to FabA, the Escherichia coli 3-hydroxyacyl-ACP dehydratase/isomerase of unsaturated fatty acid synthesis. In the latter case Masoudi et al. [11] studied the E. coli LpxD [UDP3-O-(3-hydroxymyristoyl) glucosamine N-acyltransferase], an enzyme of lipid A (endotoxin) synthesis that forms an unusually long-lived complex with its acyl-ACP substrate.

BACKGROUND: ACP STRUCTURE

E. coli ACP, the best-studied member of this protein family, is a small (M r 8860), very acidic (pI of 4.1) and extremely soluble protein that interacts with a greater variety of cellular proteins than does any other monomeric protein [12]. The carboxy groups of the fatty acyl intermediates are in thioester linkage to the thiol of the PPant prosthetic group obtained by transfer from CoA (the PPant moiety is the non-nucleotide portion of CoA), which is in turn linked to ACP Ser36 through a phosphodiester bond. In E. coli these ACP thioesters are the substrates for the enzymes of the pathway, although some pathway intermediates are diverted to introduce the double bond of the unsaturated fatty acid species (the FabA reaction), to provide the 3-hydroxy and short chain fatty acids of lipid A (the LpxD reaction) plus the octanoyl-ACP and pimeloyl-ACP molecules consumed in the synthesis of lipoic acid [13] and biotin [14] respectively. ACP (encoded by the acpP gene) is one of the most abundant E. coli cytosolic proteins and constitutes approximately 0.25 % of the total soluble protein [(6–8)×104 molecules/cell] [15,16]. The abundance of ACP is readily rationalized because it is a cofactor rather than an enzyme and should be compared with cofactors such as NAD and CoA. In 1979 Charles Rock and I predicted the secondary structure of E. coli ACP from its amino acid sequence [17] and hypothesized that the protein forms a helical bundle that protects the acyl chain from solvent. Given the primitive nature of the protein structure prediction tools then available, it  c The Authors Journal compilation  c 2014 Biochemical Society

was extremely good fortune that our model was reasonably ontarget. Moreover, we found support for the helical bundle chain sequestration hypothesis in that ACPs carrying long chain acyl chains (acyl-ACPs) were more stable to denaturation and had smaller Stokes radii than the non-acylated protein [17]. Both of these properties indicated stabilizing hydrophobic interactions between the acyl chain and the hydrophobic surfaces of the postulated amphipathic helices. These results were best explained by a model in which the helices formed a parallel helix bundle that sequestered the acyl chain in a hydrophobic sleeve formed by the helices. This model was strengthened when it became possible to synthesize acyl-ACPs with short acyl chains [18]. It was found that the degree of ACP structural stabilization was directly proportional to acyl chain length up to approximately C8 , and that the first six to eight acyl chain methylene groups were indeed sequestered within the protein structure in that they could not interact with an immobilized hydrophobic ligand [19]. These studies also showed that increased length of the acyl chain gave increased stability of the thioester bond, indicating that the thioester was pulled into the hydrophobic pocket by longer acyl chains, whereas shorter acyl chains allowed greater solvent exposure of the thioester [19,20]. The predicted E. coli ACP helix bundle structure was largely confirmed first by NMR spectroscopy [21–23] and subsequently by X-ray crystallography [24–27]. ACP is composed of a preponderance of acidic residues largely grouped into three αhelices (helices I, II and IV) oriented in an up–down–down topological arrangement to form the helical bundle (Figure 1). Helices I, II and IV are strongly amphipathic in that they have highly acidic outer faces (which impart ACP solubility) and strongly hydrophobic inner faces. A short fourth helix (helix III) of lower stability is found in various positions in the available structures with the extremes being almost parallel with and almost perpendicular to the helical bundle. ACP helix III region is the most dynamic segment of the protein as shown by NMR and crystal structure B-factors and this, plus its proximity to the thioester bond in several acyl-ACP structures, suggested that helix III movements might be a key to accessing the proximal part of the acyl chain. The dynamic structure of ACP seen in the absence of acylation is also seen, albeit to a lesser extent (as expected from the smaller Stokes radii), in its acylated forms [20,24,25,28]. However, short (e.g. C4 ) or polar acyl chains are not stably localized within the

Acyl carrier protein chain flipping

helical bundle. Indeed, butanoyl-ACP was crystallized in two forms, one with the acyl chain within the hydrophobic pocket and the other with the chain exposed to solvent [24], whereas NMR data indicate that the C4 chain is exposed to solvent, although the timescale of the experiments was not given [29]. The current picture is that bacterial and plant ACPs and their acylated derivatives can adopt several different structures by sliding the helices relative to one another and by rearranging the prosthetic group, helix III and the loops that link the helices. The dynamics of these processes will depend on the polarity and length of the acyl chain [24,25,30,31]. The plastic structure of ACP allows chains longer than C8 to be accommodated in the hydrophobic pocket, although at the cost of the stability of the thioester bond [32,33]. A straightforward explanation for this loss of stability is that the longer chains can slide up and down (as depicted in Figure 1) within the helix bundle and push the thioester bond into the solvent where it is susceptible to hydrolysis. Consistent with this explanation, the rate of hydrolysis of longer chain acyl-ACPs depends directly on their length [32,33]. The PPant prosthetic group of ACP is generally found to be highly mobile, although occasionally portions of it are sufficiently immobile to be visible in acyl-ACP X-ray and NMR structures.

ACP–ENZYME COMPLEXES

Although the first ACP–enzyme complex was reported in 2000, it did not involve an acyl chain. The structure was that of an enzyme–product complex, ACP complexed to AcpS (holo[acyl-carrier-protein] synthase), the enzyme that transfers the PPant moiety from CoA to ACP Ser36 [26]. This structure appeared 8 years before the next high-resolution structure of an enzyme–ACP species complex and thus had a primacy that coloured interpretations of enzyme–acyl-ACP interactions studied by indirect means (e.g. in vitro activity of mutated enzymes). The ACP–AcpS structure showed that the enzyme interacted largely with ACP helix II, the helix with PPant attached to Ser36 at its N-terminus [26]. Moreover, the interactions were predominately hydrophilic in nature with almost all being salt bridges between helix II acidic residues and AcpS arginine residues, giving rise to the concept that ACP-requiring enzymes will have an arginine/lysine rich ‘positive patch’ and that ACP helix II will be the site of interaction with the patch. The positive patch concept has stood the test of four subsequent acylACP–enzyme crystal structures including the FabA and LpxD structures. However, in some complexes the salt bridges involve acidic residues of ACP helices I and III plus residues in the loops that connect these helices to helix II. In the years between the appearances of the ACP–AcpS structure and the recent acyl-ACP complexes with FabA and LpxD, three other complexes of an acyl-ACP with its cognate enzyme have been reported. However, none of these structures gave information on how enzymes access the thioester proximal acyl chain carbon atoms. The first of these structures was of the FabI enoyl–ACP reductase responsible for reduction of the trans2 double bonds of enoyl–ACP intermediates [34]. Although this enzyme recognizes the thioester-proximal part of the acyl chain, the crystal structure was minimally informative because it lacked electron density for the acyl chain and the ACP side chains. The other two crystal structures, although of good quality, were of biotin synthetic enzymes that catalyse reactions far removed from the thioester bond. These were BioI, a cytochrome P450 that cleaves the chain between carbon atoms 7 and 8 of acylACPs [10], and BioH, an esterase that cleaves the methyl group from C-7 of the O-methyl ester of a C7 ω-carboxyl acyl-ACP

159

[9]. In both cases the reaction occurs in an active site buried deep within the enzyme that is accessed by extension of the prosthetic group and acyl chain (Figure 2). In BioH and BioI the complexes with the acyl-ACP substrates are formed without significant structural changes in either the enzyme or the ACP protein moiety [9,10]. Indeed, it seems that the only structural change required was partition of the acyl chain from the ACP hydrophobic core into the hydrophobic active site channels of the enzymes. Hence, it seems that some transient destabilization of the ACP hydrophobic core upon interaction with the cognate enzyme allows the acyl chain to partition into the hydrophobic active site channels of BioH and BioI. Given the reactions catalysed by these enzymes, a chain-flipping model seemed unavoidable. However, it was unclear if these structures were relevant to enzymes that catalyse more typical reactions adjacent to the thioester where localized restructuring could suffice. The recent acyl-ACP co-crystal structures with FabA and LpxD argue that all enzymes active on acyl-ACPs use the chain-flipping mechanism.

THE COMPLEX OF FabA WITH A COVALENTLY LINKED ACYL-ACP DEMONSTRATES THE CHAIN-FLIPPING MECHANISM

FabA, an enzyme essential for unsaturated fatty acid synthesis in E. coli, is an unusual enzyme in that it catalyses two reactions within the same active site [35]. Many years ago Bloch [35] (using N-acetylcysteamine model substrates) demonstrated that FabA first dehydrates R-3-hydroxydecanoyl-ACP to trans-2-decenoylACP and then isomerizes this enoyl-ACP to cis-3-decenoyl-ACP. The cis-3-decenoyl-ACP is then elongated by FabB, a specialized 3-oxoacyl-ACP synthase, to 3-oxo,cis-5-dodecenoyl-ACP [36]. Further elongation by the canonical fatty acid synthesis system results in the C16 and C18 unsaturated fatty acids of the membrane phospholipids. Nguyen et al. [5] extended the classical work of the Bloch laboratory to obtain FabA cross-linked to acyl-ACP. In their studies of the equilibria of the FabA-catalysed reactions, Bloch [35] synthesized each of the model substrates and showed that the reactions were fully reversible. However, one batch of the cis3-decanoyl-N-acetylcysteamine substrate was not only inactive, but also rapidly and irreversibly inactivated FabA [37,38]. This batch contained a contaminant that turned out to be the 3-decynol thioester. A portion of the 3-decynoic acid used to synthesize this batch of cis-3 decenoic acid had escaped reduction and was converted into the thioester [38]. When incubated with the pure 3-decynol thioester, FabA was found to catalyse isomerization of the acetylenic thioester to the corresponding 2,3-allenic thioester. The allene then attacked the active site histidine residue to form an inactivating covalent adduct [39,40]. This was the first demonstration of mechanism-activated enzyme inhibition (often called enzyme suicide or suicide inhibition) [37]. Therefore, given ACP acylated with 3-decynoic acid, it seemed straightforward to obtain a cross-linked FabA–ACP species. However, this scheme has two Achilles heels. These were the readily cleavable thioester linkage and the non-enzymatic isomerization of the triple bond to the allene [37], a very reactive species that could non-specifically attack protein side chains leading to heterogenity. A conundrum was that that a thioester is required for allene formation [40– 43], presumably due to the acidity imparted to the α-proton. Indeed, the first attempts to cross-link FabA and ACP using 3-decynoyl-ACP gave only low yields of cross-linked species [44]. Burkart and colleagues then sought a non-hydrolysable 3decynoyl linkage having α-proton acidity sufficiently weak to preclude non-enzymatic isomerization, but retaining sufficient acidity to permit efficient enzyme-catalysed abstraction of the proton required for allene formation [45,46]. In walking this fine  c The Authors Journal compilation  c 2014 Biochemical Society

160

Figure 2

J. E. Cronan

Chain flipping and the hydrophobic pockets of BioH and BioI

(A) Overall structure of the complex of methyl-pimeloyl-ACP with BioH. The ACP molecule is coloured blue with helix 2 labelled. The PPant-linked pimeloyl methyl ester is shown in stick and ball representation with carbon atoms coloured yellow, whereas BioH is shown in grey except-site the labelled capping helices that interact with the ACP moiety which are shown in pink. The catalytic triad residues are shown in stick and ball representation with carbon atoms in cyan (BioH lacked the active serine to prevent hydrolysis of the ester). (B) Structure of the ACP moiety and tetradecanoyl-PPant of the BioI–substrate complex. The stick representations denote the Cα traces of E. coli ACP and PPant-linked fatty acids from the BioI–ACP C14 complex (in blue), apo-ACP (in pink) and C10 acyl-ACP (in orange). The BioI protein surface is displayed in grey with the catalytic haem in red. (C) The BioH hydrophobic channel that directs the methylpimeloyl moiety towards the active site. The structural elements are represented and coloured as in (A). The side chains of all residues at a maximum distance of 4.5 A˚ from the PPant-linked pimeloyl methyl ester are shown. The side chains atoms are in stick and ball representation with the carbon atoms coloured green. (D) The BioI hydrophobic pocket showing the residues that produce the U-shaped conformation of the acyl chain. The C-7 and C-8 atoms of the chain where cleavage occurs are labelled. The PPant and C14 acyl chain are shown in orange. The BioH Figures are from Agarwal et al. [9], whereas the BioI Figures are from Cryle and Schlichting [10].

line they found that a sulfonyl group linking a 3-alkyene moiety to pantetheine gave an α-proton of the required acidity [45,46]. Finally by use of a one-pot chemoenzymatic method pioneered in their laboratory [47] the sulfonyl compound was converted into an acyl-CoA analogue by enzymes of the CoA synthetic pathway which provided the substrate for attachment of the analogue to unmodified apo-ACP by the non-specific PPant transferase, Sfp. This acyl-ACP analogue was purified, added to FabA and the resulting cross-linked species was purified and crystallized. The structure of the cross-linked acyl-ACP analogue–FabA complex [5] shows that the acyl chain had vacated the ACP hydrophobic cavity and completely entered the hydrophobic active-site tunnel (‘wormhole’) formed between the subunits of the dimeric FabA (Figure 3). This was expected from the previous crystal structures of FabA cross-linked to the acetylenic inhibitor [48,49]. More strikingly, the ACP cavity has collapsed to give a structure resembling that of ACP (apo-ACP) that lacks the PPant moiety [27]. The structures of the various acyl-ACP molecules  c The Authors Journal compilation  c 2014 Biochemical Society

plus NMR measurements argue that the chain flipping begins by ACP helix III–FabA interactions that allow FabA to access the acyl chain [5]. THE LpxD ACYLTRANSFERASE COMPLEXED WITH ITS COGNATE ACYL-ACP ALSO EXHIBITS CHAIN FLIPPING

LpxD was an excellent candidate for co-crystallization with its ACP substrate because it binds its cognate substrate, the ACP thioester of an R-3-hydroxy 14-carbon acid, very tightly (K d value of 59 nM) [50] and a high-resolution structure of LpxD had been determined [51]. Moreover, the first substrate to bind is the acylACP and the last product to dissociate is ACP [50]. By using a mutant LpxD lacking the catalytic histidine residue, LpxD– acyl-ACP co-crystals were obtained and the resulting structure (Figure 4) showed that the entire acyl chain had flipped from the hydrophobic ACP pocket into a hydrophobic channel formed between two monomers of the trimeric LpxD [11]. The PPant

Acyl carrier protein chain flipping

Figure 3

161

Structure of FabA cross-linked to the acyl-ACP analogue

FabA (in shades of blue) is dimeric with the active sites located at the dimer interface. A second acyl-ACP (light red) is behind the plane of the Figure. The acyl chains interact with the unusually hydrophobic helix FabA central helix (α3). Modified from PDB code 4KEH using Chimera [56].

Figure 4

The complex of LpxD with R -3-hydroxytetradecanoyl-ACP

LpxD is a trimer of unusual structure. To simplify the Figure, the lower half of the left-handed parallel LpxD β-helix structure, the N-terminal uridine-binding region and the third molecule of R -3-hydroxytetradecanoyl-ACP located behind the plane of the Figure have been hidden. The acyl chains bind in hydrophobic grooves formed by the LpxD subunit interfaces. Modified from PDB code 4IHF using Chimera [56].

group had also vacated the ACP pocket and is fully extended from its attachment site on the protein (as in the BioI, BioH and FabA structures). However, unlike FabA, the vacated ACP pocket had not collapsed [11]. The PPant moiety packs along the LpxDbound acyl chain in an unusual U-shaped arrangement in which the acyl chain and PPant each make up one side of the U [11]. The result is that the acyl chain becomes buried between the LpxD channel and the PPant moiety (Figure 4).

CONSEQUENCES OF THE CHAIN-FLIPPING MECHANISM

What is the driving force for transfer of an acyl chain from the ACP hydrophobic cavity into the enzyme hydrophobic cavity? Perhaps the driving force is the degree of protection of the acyl chain from solvent. ACP is a highly dynamic structure [3,22,23,31,52] and thus the hydrophobic faces of helices I, II and IV cannot always contact the acyl chain. For example twisting of these  c The Authors Journal compilation  c 2014 Biochemical Society

162

J. E. Cronan

helices would transiently expose the chain to the acidic residues located on the other face of the amphipathic helices. In contrast, the hydrophobic tubes and grooves of the ACP cognate enzymes are much more static (judging from crystallographic B-factors) and would therefore provide more consistent shielding of the acyl chain from solvent. This partition scenario could be aided by displacement of ACP structural elements caused by interaction with enzyme surfaces such as those seen in the FabA [5] and LpxD [11] structures. Indeed the structural agility of ACP may be a major element in facilitating the chain-flipping reaction in that more stable ACPs could be reluctant to give up their cargo. Given this scenario, how can the reaction products return to the ACP cavity to permit further modifications of the acyl chain as required in the synthesis of fatty acids and biotin? In some cases the reaction products themselves may provide the driving force. The BioH and BioI reactions generate a free carboxy group at the terminus of a seven-carbon acyl-ACP, which should result in spontaneous expulsion of the products from the hydrophobic enzyme cavity. Moreover, the ACP sleeve should better suit these charged products because the free carboxy group could protrude from the top of the helix bundle and thereby interact with solvent [9]. For LpxD, the acyl chain becomes attached to the hydrophilic UDP-acylglucosamine and no further interaction with ACP is required. More problematical in the LpxD reaction is release of the deacylated-ACP product. Release of ACP involves a dramatic reorientation of the PPant moiety that was nicely demonstrated by disulfide cross-linking of the PPant thiol to an engineered cysteine residue [11]. The covalent cross-linking of FabA and the acyl-ACP dictated that no product release could occur. However, in the native enzyme the introduced double bond might destabilize an already transient interaction. A putative impediment to reverse flipping of the acyl chain and PPant back into the ACP pocket in the FabA structure is the collapse of the acyl chain-binding pocket of ACP that is not seen in the other three structures. It should be noted that the ACP domain isolated from the mammalian polyfunctional fatty acid synthase has been reported not to sequester acyl chains [53]. However, the domain chosen was based on the bacterial and plant ACPs and the functional ACP of the yeast synthase has been shown to be twice as large as the bacterial and plant ACPs due to four additional helices [6]; hence, this finding may deserve further study. It seems probable that the chain-flipping mechanism will also apply to the ACPs of secondary metabolism, those of polyketide and non-ribosomal peptide synthesis. The structures of many such ACPs have been determined and the overall helical bundle folds are largely conserved with the ACPs of fatty acid synthesis [3]. The polyketide synthases seem likely to more consistently use the chain-flipping model due to their close relationship to the fatty acid synthesis proteins [54] and the fact that their substrates and products tend to be more hydrophobic than those of the non-ribosomal peptide synthases. NMR data indicate that the actinorhodin ACP behaves in a manner similar to the fatty acid ACPs except that that the hydrophobic cavity of the polyketide synthase ACPs is larger which is expected since the polyketide intermediates are bulkier [55].

REFERENCES 1 Cronan, Jr, J. E. (2004) The structure of mammalian fatty acid synthase turned back to front. Chem. Biol. 11, 1601–1602 CrossRef PubMed 2 Cronan, J. E. (2006) Remarkable structural variation within fatty acid megasynthases. Nat. Chem. Biol. 2, 232–234 CrossRef PubMed 3 Chan, D. I. and Vogel, H. J. (2010) Current understanding of fatty acid biosynthesis and the acyl carrier protein. Biochem. J. 430, 1–19 CrossRef PubMed  c The Authors Journal compilation  c 2014 Biochemical Society

4 Byers, D. M. and Gong, H. (2007) Acyl carrier protein: structure-function relationships in a conserved multifunctional protein family. Biochem. Cell Biol. 85, 649–662 CrossRef PubMed 5 Nguyen, C., Haushalter, R. W., Lee, D. J., Markwick, P. R., Bruegger, J., Caldara-Festin, G., Finzel, K., Jackson, D. R., Ishikawa, F., O’Dowd, B. et al. (2014) Trapping the dynamic acyl carrier protein in fatty acid biosynthesis. Nature 505, 427–431 CrossRef PubMed 6 Leibundgut, M., Jenni, S., Frick, C. and Ban, N. (2007) Structural basis for substrate delivery by acyl carrier protein in the yeast fatty acid synthase. Science 316, 288–290 CrossRef PubMed 7 Matje, D. M., Zhou, H., Smith, D. A., Neely, R. K., Dryden, D. T., Jones, A. C., Dahlquist, F. W. and Reich, N. O. (2013) Enzyme-promoted base flipping controls DNA methylation fidelity. Biochemistry 52, 1677–1685 CrossRef PubMed 8 Roberts, R. J. and Cheng, X. (1998) Base flipping. Annu. Rev. Biochem. 67, 181–198 CrossRef PubMed 9 Agarwal, V., Lin, S., Lukk, T., Nair, S. K. and Cronan, J. E. (2012) Structure of the enzyme–acyl carrier protein (ACP) substrate gatekeeper complex required for biotin synthesis. Proc. Natl. Acad. Sci. U.S.A. 109, 17406–17411 CrossRef PubMed 10 Cryle, M. J. and Schlichting, I. (2008) Structural insights from a P450 carrier protein complex reveal how specificity is achieved in the P450(BioI) ACP complex. Proc. Natl. Acad. Sci. U.S.A. 105, 15696–15701 CrossRef PubMed 11 Masoudi, A., Raetz, C. R., Zhou, P. and Pemble, IV, C. W. (2014) Chasing acyl carrier protein through a catalytic cycle of lipid A production. Nature 505, 422–426 CrossRef PubMed 12 Butland, G., Peregrin-Alvarez, J. M., Li, J., Yang, W., Yang, X., Canadien, V., Starostine, A., Richards, D., Beattie, B., Krogan, N. et al. (2005) Interaction network containing conserved and essential protein complexes in Escherichia coIi . Nature 433, 531–537 CrossRef PubMed 13 Zhao, X., Miller, J. R., Jiang, Y., Marletta, M. A. and Cronan, J. E. (2003) Assembly of the covalent linkage between lipoic acid and its cognate enzymes. Chem. Biol. 10, 1293–1302 CrossRef PubMed 14 Lin, S., Hanson, R. E. and Cronan, J. E. (2010) Biotin synthesis begins by hijacking the fatty acid synthetic pathway. Nat. Chem. Biol. 6, 682–688 CrossRef PubMed 15 Lu, P., Vogel, C., Wang, R., Yao, X. and Marcotte, E. M. (2007) Absolute protein expression profiling estimates the relative contributions of transcriptional and translational regulation. Nat. Biotechnol. 25, 117–124 CrossRef PubMed 16 Rock, C. O. and Cronan, Jr, J. E. (1981) Acyl carrier protein from Escherichia coli . Methods Enzymol. 71, 341–351 CrossRef PubMed 17 Rock, C. O. and Cronan, Jr, J. E. (1979) Re-evaluation of the solution structure of acyl carrier protein. J. Biol. Chem. 254, 9778–9785 PubMed 18 Cronan, Jr, J. E. and Klages, A. L. (1981) Chemical synthesis of acyl thioesters of acyl carrier protein with native structure. Proc. Natl. Acad. Sci. U.S.A. 78, 5440–5444 CrossRef PubMed 19 Cronan, Jr, J. E. (1982) Molecular properties of short chain acyl thioesters of acyl carrier protein. J. Biol. Chem. 257, 5013–5017 PubMed 20 Rock, C. O., Cronan, Jr, J. E. and Armitage, I. M. (1981) Molecular properties of acyl carrier protein derivatives. J. Biol. Chem. 256, 2669–2674 PubMed 21 Holak, T. A., Kearsley, S. K., Kim, Y. and Prestegard, J. H. (1988) Three-dimensional structure of acyl carrier protein determined by NMR pseudoenergy and distance geometry calculations. Biochemistry 27, 6135–6142 CrossRef PubMed 22 Kim, Y. and Prestegard, J. H. (1989) A dynamic model for the structure of acyl carrier protein in solution. Biochemistry 28, 8792–8797 CrossRef PubMed 23 Kim, Y. and Prestegard, J. H. (1990) Refinement of the NMR structures for acyl carrier protein with scalar coupling data. Proteins 8, 377–385 CrossRef PubMed 24 Roujeinikova, A., Baldock, C., Simon, W. J., Gilroy, J., Baker, P. J., Stuitje, A. R., Rice, D. W., Slabas, A. R. and Rafferty, J. B. (2002) X-ray crystallographic studies on butyryl-ACP reveal flexibility of the structure around a putative acyl chain binding site. Structure 10, 825–835 CrossRef PubMed 25 Roujeinikova, A., Simon, W. J., Gilroy, J., Rice, D. W., Rafferty, J. B. and Slabas, A. R. (2007) Structural studies of fatty acyl-(acyl carrier protein) thioesters reveal a hydrophobic binding cavity that can expand to fit longer substrates. J. Mol. Biol. 365, 135–145 CrossRef PubMed 26 Parris, K. D., Lin, L., Tam, A., Mathew, R., Hixon, J., Stahl, M., Fritz, C. C., Seehra, J. and Somers, W. S. (2000) Crystal structures of substrate binding to Bacillus subtilis holo-(acyl carrier protein) synthase reveal a novel trimeric arrangement of molecules resulting in three active sites. Structure 8, 883–895 CrossRef PubMed 27 Qiu, X. and Janson, C. A. (2004) Structure of apo acyl carrier protein and a proposal to engineer protein crystallization through metal ions. Acta Crystallogr. D Biol. Crystallogr. 60, 1545–1554 CrossRef 28 Gally, H. U., Spencer, A. K., Armitage, I. M., Prestegard, J. H. and Cronan, Jr, J. E. (1978) Acyl carrier protein from Escherichia coli: characterization by proton and fluorine-19 nuclear magnetic resonance and evidence for restricted mobility of the fatty acid chain in tetradecanoyl-acyl-carrier protein. Biochemistry 17, 5377–5382 CrossRef PubMed

Acyl carrier protein chain flipping 29 Kosa, N. M., Haushalter, R. W., Smith, A. R. and Burkart, M. D. (2012) Reversible labeling of native and fusion-protein motifs. Nat. Methods 9, 981–984 CrossRef PubMed 30 Ploskon, E., Arthur, C. J., Kanari, A. L., Wattana-amorn, P., Williams, C., Crosby, J., Simpson, T. J., Willis, C. L. and Crump, M. P. (2010) Recognition of intermediate functionality by acyl carrier protein over a complete cycle of fatty acid biosynthesis. Chem. Biol. 17, 776–785 CrossRef PubMed 31 Chan, D. I., Tieleman, D. P. and Vogel, H. J. (2010) Molecular dynamics simulations of β-ketoacyl-, β-hydroxyacyl-, and trans -2-enoyl-acyl carrier proteins of Escherichia coli . Biochemistry 49, 2860–2868 CrossRef PubMed 32 Upadhyay, S. K., Misra, A., Srivastava, R., Surolia, N., Surolia, A. and Sundd, M. (2009) Structural insights into the acyl intermediates of the Plasmodium falciparum fatty acid synthesis pathway: the mechanism of expansion of the acyl carrier protein core. J. Biol. Chem. 284, 22390–22400 CrossRef PubMed 33 Zornetzer, G. A., Tanem, J., Fox, B. G. and Markley, J. L. (2010) The length of the bound fatty acid influences the dynamics of the acyl carrier protein and the stability of the thioester bond. Biochemistry 49, 470–477 CrossRef PubMed 34 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 carrier protein bound to FabI, the FASII enoyl reductase from Escherichia coli . J. Biol. Chem. 281, 39285–39293 CrossRef PubMed 35 Bloch, K. (1971) β-Hydroxydecanoyl thioester dehydrase. In The Enzymes (Boyer, P. D., ed.), pp. 441–464, Academic Press, New York CrossRef 36 Feng, Y. and Cronan, J. E. (2009) Escherichia coli unsaturated fatty acid synthesis: complex transcription of the fabA gene and in vivo identification of the essential reaction catalyzed by FabB. J. Biol. Chem. 284, 29526–29535 CrossRef PubMed 37 Bloch, K. (1986) The beginnings of enzyme suicide. J. Protein. Chem. 5, 69–73 CrossRef 38 Brock, D. J., Kass, L. R. and Bloch, K. (1967) β-Hydroxydecanoyl thioester dehydrase. II. Mode of action. J. Biol. Chem. 242, 4432–4440 39 Helmkamp, Jr, G. M. and Bloch, K. (1969) β-Hydroxydecanoyl thioester dehydrase. Studies on molecular structure and active site. J. Biol. Chem. 244, 6014–6022 PubMed 40 Schwab, J. M., Ho, C. K., Li, W. B., Townsend, C. A. and Salituro, G. M. (1986) β-Hydroxydecanoyl thioester dehydrase: complete characterization of the fate of the suicide substrate 3-decynoyl-NAC. J. Am. Chem. Soc. 108, 5309–5316 CrossRef 41 Helmkamp, Jr, G. M., Rando, R. R., Brock, D. J. and Bloch, K. (1968) β-Hydroxydecanoyl thioester dehydrase. Specificity of substrates and acetylenic inhibitors. J. Biol. Chem. 243, 3229–3231 PubMed 42 Rando, R. R. and Bloch, K. (1968) Mechanism of action of β-hydroxydecanoyl thioester dehydrase. J. Biol. Chem. 243, 5627–5634 PubMed 43 Morisaki, M. and Bloch, K. (1972) On the mode of interaction of β-hydroxydecanoyl thioester dehydrase with allenic acid derivatives. Biochemistry 11, 309–314 CrossRef PubMed

163

44 Meier, J. L., Haushalter, R. W. and Burkart, M. D. (2010) A mechanism based protein crosslinker for acyl carrier protein dehydratases. Bioorg. Med. Chem. Lett. 20, 4936–4939 CrossRef PubMed 45 Ishikawa, F., Haushalter, R. W. and Burkart, M. D. (2012) Dehydratase-specific probes for fatty acid and polyketide synthases. J. Am. Chem. Soc. 134, 769–772 CrossRef PubMed 46 Ishikawa, F., Haushalter, R. W., Lee, D. J., Finzel, K. and Burkart, M. D. (2013) Sulfonyl 3-alkynyl pantetheinamides as mechanism-based cross-linkers of acyl carrier protein dehydratase. J. Am. Chem. Soc. 135, 8846–8849 CrossRef PubMed 47 Worthington, A. S. and Burkart, M. D. (2006) One-pot chemo-enzymatic synthesis of reporter-modified proteins. Org. Biomol. Chem. 4, 44–46 CrossRef PubMed 48 Leesong, M., Henderson, B. S., Gillig, J. R., Schwab, J. M. and Smith, J. L. (1996) Structure of a dehydratase-isomerase from the bacterial pathway for biosynthesis of unsaturated fatty acids: two catalytic activities in one active site. Structure 4, 253–264 CrossRef PubMed 49 Moynie, L., Leckie, S. M., McMahon, S. A., Duthie, F. G., Koehnke, A., Taylor, J. W., Alphey, M. S., Brenk, R., Smith, A. D. and Naismith, J. H. (2013) Structural insights into the mechanism and inhibition of the β-hydroxydecanoyl-acyl carrier protein dehydratase from Pseudomonas aeruginosa . J. Mol. Biol. 425, 365–377 CrossRef PubMed 50 Bartling, C. M. and Raetz, C. R. (2008) Steady-state kinetics and mechanism of LpxD, the N-acyltransferase of lipid A biosynthesis. Biochemistry 47, 5290–5302 CrossRef PubMed 51 Bartling, C. M. and Raetz, C. R. (2009) Crystal structure and acyl chain selectivity of Escherichia coli LpxD, the N-acyltransferase of lipid A biosynthesis. Biochemistry 48, 8672–8683 CrossRef PubMed 52 Andrec, M., Hill, R. B. and Prestegard, J. H. (1995) Amide exchange rates in Escherichia coli acyl carrier protein: correlation with protein structure and dynamics. Protein Sci. 4, 983–993 CrossRef PubMed 53 Ploskon, E., Arthur, C. J., Evans, S. E., Williams, C., Crosby, J., Simpson, T. J. and Crump, M. P. (2008) A mammalian type I fatty acid synthase acyl carrier protein domain does not sequester acyl chains. J. Biol. Chem. 283, 518–528 CrossRef PubMed 54 Cronan, J. E. and Thomas, J. (2009) Bacterial fatty acid synthesis and its relationships with polyketide synthetic pathways. Methods Enzymol. 459, 395–433 CrossRef PubMed 55 Evans, S. E., Williams, C., Arthur, C. J., Ploskon, E., Wattana-amorn, P., Cox, R. J., Crosby, J., Willis, C. L., Simpson, T. J. and Crump, M. P. (2009) Probing the Interactions of early polyketide intermediates with the Actinorhodin ACP from S. coelicolor A3(2). J. Mol. Biol. 389, 511–528 CrossRef PubMed 56 Yang, Z., Lasker, K., Schneidman-Duhovny, D., Webb, B., Huang, C. C., Pettersen, E. F., Goddard, T. D., Meng, E. C., Sali, A. and Ferrin, T. E. (2012) UCSF Chimera, MODELLER, and IMP: an integrated modeling system. J. Struct. Biol. 179, 269–278 CrossRef PubMed

Received 19 February 2014; accepted 17 March 2014 Published on the Internet 13 May 2014, doi:10.1042/BJ20140239

 c The Authors Journal compilation  c 2014 Biochemical Society