Skip to main content
NCBI home page
The Plant Cell logoLink to The Plant Cell
. 2010 Sep 14;22(9):3093–3104. doi: 10.1105/tpc.109.072652

Crystal Structures of a Populus tomentosa 4-Coumarate:CoA Ligase Shed Light on Its Enzymatic Mechanisms

Yonglin Hu a,1, Ying Gai b,1, Lei Yin a,1, Xiaoxue Wang b, Chunyan Feng b, Lei Feng a, Defeng Li a, Xiang-Ning Jiang b, Da-Cheng Wang a,2
PMCID: PMC2965553  PMID: 20841425

The crystal structure of Chinese white poplar 4-coumaric acid:coenzyme A ligase, the enzyme that catalyzes the synthesis of an important precursor of lignin, was determined in the apo, adenosine monophosphate-, and adenosine 5′-(3-(4-hydroxyphenyl)propyl)phosphate-complexed forms. Enzymatic mechanisms were proposed for the protein. Residues responsible for substrate specificity were identified.

Abstract

4-Coumaric acid:CoA ligase (4CL) is the central enzyme of the plant-specific phenylpropanoid pathway. It catalyzes the synthesis of hydroxycinnamate-CoA thioesters, the precursors of lignin and other important phenylpropanoids, in two-step reactions involving the formation of hydroxycinnamate-AMP anhydride and then the nucleophilic substitution of AMP by CoA. In this study, we determined the crystal structures of Populus tomentosa 4CL1 in the unmodified (apo) form and in forms complexed with AMP and adenosine 5′-(3-(4-hydroxyphenyl)propyl)phosphate (APP), an intermediate analog, at 2.4, 2.5, and 1.9 Å resolution, respectively. 4CL1 consists of two globular domains connected by a flexible linker region. The larger N-domain contains a substrate binding pocket, while the C-domain contains catalytic residues. Upon binding of APP, the C-domain rotates 81° relative to the N-domain. The crystal structure of 4CL1-APP reveals its substrate binding pocket. We identified residues essential for catalytic activities (Lys-438, Gln-443, and Lys-523) and substrate binding (Tyr-236, Gly-306, Gly-331, Pro-337, and Val-338) based on their crystal structures and by means of mutagenesis and enzymatic activity studies. We also demonstrated that the size of the binding pocket is the most important factor in determining the substrate specificities of 4CL1. These findings shed light on the enzymatic mechanisms of 4CLs and provide a solid foundation for the bioengineering of these enzymes.

INTRODUCTION

The phenylpropanoid pathway is one of the most important secondary metabolism pathways in plants. It diverts carbon flows from primary metabolism pathways, in the form of Phe, to an array of diverse products, in response to internal and external stresses. These products function in the production of aroma, fruit flavor, and color and as molecular signals, antimicrobials, flower pigments, antioxidants, and UV protectants. Many of these products bear great importance for human life and the ecosystem. For example, many flavonoids possess antioxidant and antitumor properties and have been used as health-promoting agents for thousands of years. The most prominent product of this pathway is probably lignin, one of the most abundant naturally occurring polymers, second only to cellulose. It is estimated that 25 to 30% of the annually sequestered carbon dioxide is deposited in the form of lignin. Lignin confers the rigidity and mechanical strength needed for plant growth and renders plants impermeable to water and resistant to pathogen invasion. On the other hand, lignin is not readily degradable and is a major source of pollution of the pulping industry. It also reduces the digestibility and quality of forage grass, thus reducing livestock productivity.

Recently, the use of biomass as a renewable carbon source for the generation of biofuels and biomaterials has become increasingly important in the quest for sustainable development. The composition of the raw materials, especially of the lignin content, largely determines the efficiency and industrial value of the biomass conversion (Boudet et al., 2003; Ragauskas et al., 2006). Using genetic engineering to optimize plant properties for biomass utilization will play an essential role in this endeavor and requires an in-depth understanding of the structure-function relationships of the enzymes involved in lignin biosynthesis.

Many enzymes are involved in the phenylpropanoid pathway. The pathway begins with the deamination of Phe, a reaction catalyzed by Phe ammonia lyase, to form trans-cinnamic acid, which is subsequently hydroxylated to 4-coumaric acid (4-hydroxyl cinnamic acid) by cinnamate 4-hydroxylase. 4-Coumaric acid is then activated by 4-coumarate:CoA ligase (4CL; EC 6.2.1.12) into 4-coumaroyl CoA. This compound is converted by hydroxycinnamoyltransferase to 4-coumaroyl quinate or 4-coumaroyl shikimate, which are subsequently hydroxylated by 4-coumarate 3-hydroxylase into cafferoyl quinate or cafferoyl shikimate. The products are then converted into cafferoyl CoA thioesters by hydroxycinnamoyltransferase, and the quinate and shikimate are recycled. The cafferoyl-CoA is then methylated by cafferoyl-CoA O-methyltransferase to form feruloyl-CoA (Hoffmann et al., 2004). The 4CL proteins play essential roles in this pathway because the products of their enzymatic activities, hydroxycinnamoyl-CoA thioesters, serve as precursors for the synthesis of lignins, flavonoids, stilbenes, and other phenylpropanoids.

The enzyme 4CL is expressed differentially in different tissues and development stages in wood, and multiple isoforms with different substrate specificities exist in plants (Voo et al., 1995; Ehlting et al., 1999; Lindermayr et al., 2002; Zhang et al., 2003; Hamberger and Hahlbrock, 2004), suggesting that 4CL may play important roles in the regulation of the phenylpropanoid pathway by regulating the synthesis rates of the different hydroxycinnamoyl CoA thioesters (Knobloch and Hahlbrock, 1975). In Populus tremuloides (quaking aspen or trembling aspen), for example, two isoforms of 4CL have been identified. One of them, designated 4CL1, was found to be expressed in the developing xylem of woody stems and is associated with the biosynthesis of lignin, whereas the other, 4CL2, takes part in the biosynthesis of phenylpropanoids other than lignin in the epidermal cells of stems and leaves (Hu et al., 1998). On the other hand, the activity of 4CL determines the overall carbon flow to the phenylpropanoid pathway. For these reasons, 4CL has been the focus of genetic engineering studies to improve the quality of plant products, and various degrees of success have been achieved by attenuating the expression of 4CL and thus the production of lignin (Lee et al., 1997; Kajita et al., 2002; Li et al., 2003).

Most 4CL proteins are found in higher plants. Recently, Silber et al. (2008) identified a 4CL gene family in the moss Physcomitrella patens. 4CL is a member of the so-called ANL superfamily. This superfamily derives its name from three subfamilies: the acyl-CoA synthetase subfamily, to which 4CLs belong; the adenylation domain of nonribosomal peptide synthetases; and the luciferase subfamily (Gulick, 2009). It was formerly referred to as the adenylate-forming superfamily (Babbitt et al., 1992) because all members of this superfamily share a conserved adenylation partial reaction in their enzymatic mechanisms, even although the overall reactions catalyzed by these enzymes are very diverse. Although ANL enzymes contain 10 highly conserved segments of amino acid sequences, designated A1 to A10 (Marahiel et al., 1997), the overall sequence identities, especially for proteins from different subfamilies, are rather low. So far, the crystal structures of several non-4CL proteins from this superfamily have been determined and have been extensively reviewed (Gulick, 2009), but none of the 4CLs has been characterized by x-ray crystallography. Thus, the crystal structure of a 4CL protein would greatly increase our understanding of this class of enzymes.

In this article, we report the three-dimensional structural characterization of Populus tomentosa (Chinese white poplar) 4CL1 by x-ray crystallographic and mutagenesis analyses. 4CL1 is one of the two 4CL proteins identified from this species. Like in P. tremuloides, two 4CL genes, designated 4CL1 and 4CL2, have been identified in P. tomentosa. P. tomentosa 4CL1 is homologous to P. tremuloides 4CL1, with a sequence identity of 97% (Figure 1). We solved the crystal structures of the apo, AMP-complexed, and adenosine 5′-(3-(4-hydroxyphenyl)propyl)phosphate (APP)-complexed forms of P. tomentosa 4CL1 at 2.4, 2.5, and 1.9 Å resolution, respectively. The compound APP used in the cocrystallization is a mimic of adenosine 5′-coumaroyl phosphate, the product of the adenylate-forming step of 4CL. APP differs from adenosine 5′-coumaroyl phosphate in that it is a phosphate ester instead of a phosphate-carboxylate anhydride, and the carbon-carbon double bond is reduced to a carbon-carbon single bond (Figure 2). Based on these crystal structures and structure-based mutagenesis studies, we identified residues essential for the enzymatic mechanisms of 4CL1. We also identified the substrate binding pocket of 4CL1 and the structural basis of substrate specificity. Our studies lay a solid foundation for understanding the enzymatic mechanisms and bioengineering of this important enzyme.

Figure 1.

Figure 1.

Open in a new tab

Sequence Alignment of Several ANL Enzymes Referred to in This Article.

P. tomentosa 4CL1, P. tremuloides 4CL1, G. max 4CL1, Arabidopsis 4CL4, P. patens 4CL1 (36 residues removed from the N terminus to simplify the figure), and S. enterica Acs. White letters on black background designate completely conserved residues. White letters on gray background designate highly conserved residues. Black letters on gray background designate conserved residues. Residues involved in hydroxycinnamate binding are indicated with triangles, while those involved in enzymatic functions are labeled with stars.

Figure 2.

Figure 2.

Open in a new tab

Intermediate and an Intermediate Analog of 4CL1.

Molecular structure of adenosine 5′-coumaroyl phosphate, the intermediate of the 4CL enzymatic reaction, and its analog APP, which was used in the crystallization.

RESULTS

Overall Structure

The P. tomentosa 4CL1 protein consists of 536 residues. The crystal structure of the apo-protein of 4CL1 was solved by the single anomalous dispersion method in combination with the molecular replacement method using firefly luciferase (Conti et al., 1996) as the model. It was then used as a model to solve the AMP-4CL1 and APP-4CL1 complex structures using the molecular replacement method. All structures are of excellent quality (Table 1). The crystallographic R factors are 0.192, 0.204, and 0.176 and the free R factors are 0.243, 0.241, and 0.196, respectively, for the apo, AMP-complexed, and APP-complexed forms.

Table 1.

Statistics on Data Collection, Structure Refinement, and Model Assessment

Crystal SeMet
apo-4CL1 apo-4CL1 AMP-4CL1 APP-4CL1
Space group P212121 P212121 I213 I213
Cell parameters (Å) a = 51.738, b = 78.739, c = 118.787 a = 51.740, b = 78.736, c = 118.789 a = b = c = 162.34 a = b = c = 161.51
Wavelength (Å) 0.9791 (peak) 1.000 1.5418 0.97916
Resolution (last shell) (Å) 50–3.0
(3.11–3.0) 50–2.4
(2.48–2.4) 57–2.5
(2.64–2.5) 34–1.9
(2.0–1.9)
Total reflections 144,516 253,168 124,694 527,176
Unique reflections 18,855 19,643 24,708 55,010
Completeness
(last shell) 100.0% (99.9%) 97.1%
(87.2%) 99.9%
(99.5%) 100.0%
(100.0%)
Redundancy (last shell) 7.7 (5.8) 12.9 (12.5) 5.0 (3.9) 9.6 (8.9)
I/σ(I) (last shell) 13.8 (3.9) 38.1 (8.3) 17.8 (3.2) 17.2 (6.2)
Rsym (last shell) 12.6% (36.3%) 8.2% (28.7%) 8.5% (38.1%) 8.8% (35.0%)
Refinement Statistics
 R 19.2% 20.4% 17.7%
 Free R 24.3% 24.1% 19.6%
 Protein atoms 3,946 4,048 4,055
 Water molecules 243 124 555
 Heteroatoms 0 23 78
RMSDs from ideal model
 Bond (Å) 0.007 0.007 0.005
 Angles 1.36° 1.31° 1.33°
Ramachandran plots
 Most favored region 90.4% 90.5% 91.2%
 Additionally allowed region 9.2% 9.0% 8.4%
 Generously allowed region 0.4% 0.4% 0.4%
 Disallowed regions 0% 0% 0%
Open in a new tab

The overall structure of 4CL1 consists of two distinctive globular domains. The larger domain consists of the N-terminal 434 residues, while the smaller one consists of residues 435 to 536 (Figure 3A). In this article, these two domains will be referred to as the N- and C-domains, respectively.

Figure 3.

Figure 3.

Open in a new tab

Overall Structure of 4CL1.

(A) Structure of apo-4CL1. The C-domain is colored yellow. The three subdomains of the N-domain, referred to as N1, N2, and N3, are colored blue, purple, and green, respectively, following the scheme used by Conti et al. (1996).

(B) Stereoview of the 4CL1-APP complex. The N- and C-domains are shown as green and yellow, respectively. APP is shown as balls, with C, O, and N atoms colored in pink, red, and blue, respectively. The three-dimensional visual information of stereoviews can be retrieved using stereoscopes or using naked eyes by focusing left and right eyes on left and right stereograms, respectively.

The N-domain of 4CL1 can be divided into three subdomains, as observed in the structure of firefly luciferase (Conti et al., 1996). These three subdomains, designated N1, N2, and N3, respectively, are formed by noncontinuous residues from the N-domain (Figure 3A). N1 and N2, the two larger subdomains, have very similar structures. Each of these two subdomains has a central eight-stranded β-sheet that is flanked by two α-helices on one side and four α-helices on the other. Both of the central β-sheets are composed of six parallel β-strands and two antiparallel β-strands at both ends of the sheet. The topologies of these two subdomains are also very similar (Conti et al., 1996). The N3 subdomain is smaller than the other two. The majority of this subdomain is a distorted β-barrel formed by eight β-strands; only one α-helix and one short 310-helix are found in this subdomain. The β-barrel is approximately shaped like a bowl, with the concave region facing the center of the N-domain. Two short β-strands from this subdomain are packed against the antiparallel β-strands of the central β-sheets of the other two subdomains.

The center of the C-domain of 4CL1 is a three-stranded mixed β-sheet. Located on one side of the sheet are two α-helices and a short, two-stranded, antiparallel β-sheet; on the other side is a single α-helix.

Structural comparisons revealed that the overall fold is highly conserved among ANL enzymes. In fact, the N-domains from all of these proteins can be superposed on that of 4CL1 with root of mean square deviations (RMSDs) of 1.9 to 2.2 Å for 355 to 391 Cα atoms, whereas the C-domains can be superposed with RMSDs of 0.9 to 1.7Å for 71 to 103 Cα atoms. The structure of human acyl-CoA synthetase medium-chain family member 2A complexed with butyric CoA thioester is the most similar to that of 4CL1 in APP-complexed forms. A total of 481 Cα atoms from both domains of these two structures can be superposed with an RMSD of 1.9 Å.

The APP- and AMP-bound 4CL1 crystals are isomorphous (Table 1). The 4CL1 structures in these two complexed forms are essentially identical, as evidenced by an RMSD of 0.4 Å. For this reason, we will not distinguish AMP- and APP-complexed structures in subsequent discussions unless specified otherwise.

Hydroxycinnamate Binding Pocket

In the 4CL1 structure, a cavity was identified in the larger N-domain. In the APP-complexed structure, a group of strong electron density peaks that corresponded well to an APP molecule (Figure 4A) were located inside the cavity. Structural comparisons show that this cavity superimposes well with the ligand binding pockets of other ANL enzymes, demonstrating unambiguously that this cavity is the ligand binding site of 4CL1.

Figure 4.

Figure 4.

Open in a new tab

4CL1 Ligand Binding Site and APP-4CL1 Interactions.

(A) The σA-weighted 2Fo−Fc omit map around the bound APP molecule, contoured at 6σ. The map is superimposed on the final model of the ligand. The C, O, and N atoms of the ligand are colored pink, red, and blue, respectively.

(B) Stereoview of 4CL1-APP interactions. APP is colored pink, red, and blue, for C, O, and N atoms. Residues from the C-domain are colored yellow. Hydrogen bonds between APP and 4CL1 are shown as red dashed lines. The letter m marks the meta-positions of the phenyl ring where substitutions take place.

(C) Stereoview of the binding site with a perspective different from that of (B), highlighting residues Pro-337 and Val-338.

(D) Structural changes at the hydroxycinnamate binding pocket of 4CL1 upon the binding of APP. Apo and APP-complexed 4CL1 structures are shown in yellow and green, respectively. Bound APP is shown as spheres. Some residues were removed to highlight the changes in Tyr-236 and S304GGAP308.

The hydroxycinnamate and AMP binding pockets are located in two relatively independent parts of the cavity, and they have distinct protein-substrate interaction patterns. The hydroxycinnamate binding pocket of 4CL1 is clearly defined. Along the direction perpendicular to the 4-hydroxyphenyl group of APP, the binding pocket is defined by the Tyr-236 side chain on one side and the main chain of Tyr-330 and Gly-331 on the other. On the lateral direction of the 4-hydroxyphenyl group, it is defined by the main chain of Gly-305 on one side and residues Pro-337 and Val-338 on the other (Figures 4B and 4C).

This hydroxycinnamate binding pocket has two distinct structural features. The first is the closeness between and parallel arrangement of the side chain of Tyr-236 and main chain of Tyr330-Gly331. The side chain of Tyr-330 points away from the pocket, so only the main chain of this residue contributes to substrate binding. Gly-331, on the other hand, would have the side chain point toward the bound substrate if it were mutated to any other non-Gly residues. The 4-hydroxyphenyl group of APP is inserted between and forms very close contacts with the side chain of Tyr-236 and main chain of Tyr-330 and Gly-331 (Figure 4B). The two 4-hydroxyphenyl groups from APP and Tyr-236 are almost parallel to each other at an interplane distance of ~4 Å (Figures 4B and 4C). The main chain of residues Tyr-330 and Gly-331 runs parallel to, and forms tight van der Waals interactions with, the 4-hydroxyphenyl group of APP on the other side, with interatomic distances of as small as 3.4 Å (between the carbonyl carbon atom of Gly-331 and the C3 atom of the propyl moiety). The side chain of Val-338 also interacts with the 4-hydroxyphenyl group from this side (Figures 4B and 4C).

The second important structural feature of this binding pocket is found along the direction lateral to the 4-hydroxyphenyl group of APP, where a highly bulged loop is formed by residue Gly-336, Pro-337, and Val-338. This loop is located between a 310-helix and a β-strand (Figure 4C). It adopts the same conformation in apo and complexed 4CL1 structures. This unusual structure is caused by the short distance between the two ends of the loop and a cis-peptide bond between Gly-336 and Pro-337. This bulged loop forms one side of the binding pocket along the lateral directions, while the main chain groups of Gly-305, Gly-306, and Gly-329 are on the other side (Figure 4B).

In all the main chain–substrate interactions, the main chain always interacts with substrates from the direction perpendicular to the peptide planes, thus excluding hydrogen bond formation and ensuring that only hydrophobic interaction is involved in substrate binding. In fact, the 4-hydroxyphenyl group of APP is held almost exclusively by the hydrophobic groups of 4CL1, with only the 4-hydroxyl group forming hydrogen bonds with the side chain of Ser-240 and Lys-303.

AMP Binding Pocket

The AMP moieties in the AMP- and APP-complexed structures mainly interact with the N-domain. A5, one of the 10 highly conserved sequences (Fulda et al., 1994; Marahiel et al., 1997), which corresponds to Q328GYGMTEA335 in 4CL1, provides the majority of the surface of the binding pocket. This binding motif is almost completely conserved in 4CLs (Figure 1).

The adenine group of the bound AMP is held by extensive van der Waals interactions, from the side chain phenyl group of Tyr-330 on one side of the ring and the main chain and side chain groups of Gly-306 and Ala-307 from the other side. In addition, atom N6 of the adenine donates a hydrogen bond to the carbonyl oxygen of Gly-329. Atom N1 accepts a hydrogen bond from a nearby water molecule (Figure 4B).

Several hydrogen bonds are formed between the ribose oxygen atoms and the side chains of Arg-432, Lys-434, and Lys-438 (Figure 4B). The phosphate group forms hydrogen bonds with the side chains of Thr-333 and Gln-443 and with the main chain NH group of Thr-333 (Figure 4B). The binding of the ribose and phosphate groups involves both N- and C-domain residues. While residues from the N-domain (i.e., Thr-333 and Arg-432) show little change in their positions between the apo- and APP/AMP-complexed structures, C-domain residues (i.e., Lys-434, Lys-438, and Gln-443) are only able to interact with APP/AMP when 4CL1 changes from the open conformation in the apo form to the closed conformation in complexed forms (see discussions below for interdomain movement). These three residues are located at the N terminus of the C-domain and are highly conserved in ANL enzymes. They may be responsible for the ligand binding–induced conformational change of 4CL1.

Ligand Binding–Induced Conformational Changes

Conformational changes, both localized and global, were observed in 4CL1 structures upon the binding of APP or AMP.

A large cleft was observed between the N- and C-domains in apo-4CL1 (Figure 3A) and is closed by an 81° rotation of the C-domain relative to the N-domain (Figure 3B). Because no major internal changes are found in either the N- or C-domain, the closure of the interdomain cleft upon the binding of AMP or APP is a rigid-body movement. Similar phenomena were observed in other ANL enzymes. In fact, the N- and C-domains of apo and AMP/APP-complexed structures can be superimposed with RMSDs of 1.1 and 0.9 Å, respectively. It has been known that ANL enzymes use this type of interdomain movement to bring different catalytic residues from the C-domains to substrate binding sites to catalyze the respective partial reactions (Gulick, 2009).

Some localized structural changes at the hydroxycinnamate binding pocket were also observed upon APP binding. The Tyr-236 side chain in APP-complexed 4CL1 rotates ~90° around the Cβ-Cγ bond, and its main chain atoms shift ~1 Å relative to its position in apo and AMP-complexed structures (Figure 4D). The rotation of the side chain would evacuate the space needed to accommodate hydroxycinnamate substrates, while the shift in main chain atoms would enable the side chain hydroxyl group to form a hydrogen bond with the carbonyl oxygen of Val-277. The side chain of His-234 also undergoes a conformational change upon the binding of APP, with the torsion angle χ1 changing from −175° in the AMP-complexed structure to −75° in the APP-complexed structure (Figure 4D). A similar conformational change in Alcaligenes sp 4-chlorobenzoat:CoA ligase has been proposed to play important roles in regulating the access of CoA to the active center (Reger et al., 2008). In AMP- and APP-complexed 4CL1 structures, His-234 adopts the adenylate-forming and thioester-forming conformations, respectively. The structural changes of His-234 and Tyr-236 only occur when 4CL1 binds its hydroxycinnamate substrates because these residues have identical conformations in apo and AMP-complexed structures.

Another prominent structural change between the apo structure and APP- and AMP-complexed structures was observed in segment S304GGAP308, which moved ~4 Å toward the substrates in APP- and AMP-complexed structures compared with the apo structure. The Ala307-Pro308 peptide bond also underwent a cis- to trans-conversion upon substrate binding, to move the Ala-307 side chain out of way (Figure 4D).

Mutagenesis Studies on Residues Involved in Substrate Binding

To investigate the importance of the residues lining the 4CL1 hydroxycinnamate binding pocket, we prepared multiple mutants, such as Y236A, Y236F, Y236W, G331A, G305A, S240A, and K303A, and assayed their enzymatic activities.

Mutations such as G331A and Y236W that introduce steric conflicts at the binding site completely abolish 4CL1 activities, probably because these mutations abolish the affinity of 4CL1 for its substrates. The Y236A mutation, on the other hand, enhances 4CL1 activities (Figure 5), probably because it decreases steric conflicts and increases the protein’s affinity toward its substrates. These results are in agreement with the tight arrangement between the Gly-331 main chain, the Tyr-236 side chain, and the 4CL1 substrates. Interestingly, Y236F greatly diminishes 4CL1 activities toward caffeic acid and ferulic acid but does not affect the activity toward 4-coumaric acid, even although some 4CL proteins have Phe at this position. This is probably because the mutation disrupts a hydrogen bond between Tyr-236 and Val-277 in 4CL1 and makes substrate binding less thermodynamically favorable, especially for bulky substrates.

Figure 5.

Figure 5.

Open in a new tab

Enzymatic Activity Assays of 4CL1 and Its Mutants.

(A) Wild-type 4CL1 activities toward 4-coumaric, caffeic, and ferulic acids.

(B) Activities of 4CL1 and mutants toward 4-coumaric acid. Results in (B) to (D) were the average of five measurements with the standard deviations.

(C) Activities of 4CL1 and mutants toward caffeic acid.

(D) Activities of 4CL1 and mutants toward ferulic acid.

As part of the binding pocket, residue Gly-305 is located inside a loop region with very high temperature factors (60 to 70 Å2) in apo structure but much lower in AMP- and APP-complexed structures (20 to 30 Å2). This loop undergoes significant structural changes upon substrate binding (Figure 4D). This indicates that steric conflicts introduced by mutations in this region can be more easily relieved. Consequently, the G305A mutation diminishes, but does not completely abolish, 4CL1 activities toward 4-coumaric and caffeic acids (Figure 5).

Hydrogen bond formation also plays important roles in 4CL1-APP interactions. Mutations of Ser-240 and Lys-303, which form hydrogen bonds with the hydroxyl group of APP, abolish 4CL1 activities (Figure 5). Residue Ser-240 is highly conserved in 4CLs, but Lys303 is not. Its hydrogen bond with the substrate is also very weak (length of 3.52 Å). Therefore, we propose that the hydrogen bond with Ser-240 is essential for substrate binding in 4CLs and that the hydrogen bond with Lys-303 is important but not required.

Catalytic Centers

ANL enzymes are known to have two distinct catalytic centers, both located in the C-domain, for the two partial reactions. They are brought to the substrate binding sites to catalyze respective partial reactions by interdomain movements (Gulick, 2009). A strictly conserved Lys residue, which is Lys-523 in 4CL1 (Figure 1), has been identified as the catalytic center of the adenylation half reaction (Conti et al., 1997; May et al., 2002; Gulick et al., 2003; Jogl and Tong, 2004; Nakatsu et al., 2006). Mutation (Branchini et al., 2000; Horswill and Escalante-Semerena, 2002) and acetylation (Starai et al., 2002) of this residue abolish the substrate adenylation half reactions of ANL enzymes but have much smaller impacts on the thioester-forming activities of these enzymes. Enzymatic assays showed that mutation of Lys-523 to Ala completely abolished the activities of 4CL1 (Figure 5), demonstrating the importance of this residue in the enzymatic mechanism.

In the APP-4CL1 complex structure, the side chains of Lys-438 and Gln-443 are located near to, and form multiple hydrogen bonds with, the phosphate group. We propose that these two residues form the catalytic center for the second partial step. In the catalytic process, we propose that the side chain of Lys-438 would form a hydrogen bond with the carbonyl group of the coumaroyl-AMP conjugate, activating it for nucleophilic substitution by the thiol group of CoA. In the APP-4CL1 structure, the distance between the side chain Nε atom of Lys-438 and the C1 atom of the propyl moiety is 3.9Å (Figure 4B). When a coumaroyl-AMP conjugate, the product of the adenylate-forming step of 4CL, is bound, the distance between the Lys-438 Nε atom and the carbonyl oxygen atom on C1 will be in the hydrogen bonding range. The side chain of residue Gln-443 forms two hydrogen bonds with the phosphate group. We propose that the function of this residue is to stabilize the negative charges on the phosphate group during the cleavage of coumaroyl-AMP and the formation of the final thioester products.

We prepared K438A and Q443A mutants and a K438A/Q443A double mutant and assayed their activities. These mutations completely abolish 4CL1 activity (Figure 5). This verified the essential roles played by these residues in the enzymatic mechanisms of P. tomentosa 4CL1.

DISCUSSION

Most of the enzymes involved in the phenylpropanoid pathway have been characterized by x-ray crystallography (Ferrer et al., 2008); however, previous attempts to crystallize 4CL were not successful. Other indirect techniques, such as homology modeling (Stuible and Kombrink, 2001; Schneider et al., 2003) and molecular engineering (Ehlting et al., 2001; Lindermayr et al., 2003), were used to study the structure-function relationships of 4CL, using the structures of non-4CL ANL enzymes as models. These studies greatly underscore the importance of 4CL structures in our understanding of the functions of these proteins. Because of the low sequence homology between 4CL and non-4CL ANL proteins, these approaches were not very satisfactory. In one of these studies, for example, Schneider et al. (2003) proposed a substrate binding pocket in Arabidopsis thaliana 4CL2 based on homology modeling and extensive mutagenesis and activity assays. The authors correctly identified residues such as Lys-320 and Val-355 (equivalent to Lys-303 and Val-338 of P. tomentosa 4CL1, respectively) as parts of the substrate binding pocket, but they could not identify Tyr-253, the equivalent of P. tomentosa 4CL1 Tyr-236, as an essential residue of the binding pocket. Our finding that the Y236A mutant is more active than wild-type P. tomentosa 4CL1 highlights the limitations of homology modeling and mutagenesis studies in the absence of crystal structure.

The comprehensive, high-resolution structural information along with the mutagenesis studies presented here enable us to identify the residues essential for catalytic activities and substrate binding and the structural elements that determine the substrate specificities of P. tomentosa 4CL1 and other 4CL enzymes.

Substrate Specificity of P. tomentosa 4CL1 Is Determined by the Size of the Substrate Binding Pocket

The P. tomentosa 4CL1 binding pocket for hydroxycinnamate substrate is lined mostly with hydrophobic groups (Figure 4B). This is in agreement with the finding that hydrophobic interaction is the largest contribution to the binding of proteins with their cognate small molecule ligands (Wang et al., 2004). Consequently, the ligand specificity of P. tomentosa 4CL1 is not based on the formation of hydrogen bonds; instead, it is based on the recognition of the size of the ligand by the binding pocket, especially by the structural features discussed above.

The first structural feature (i.e., the close and parallel arrangement between the Tyr-236 side chain and the Tyr330-Gly331 main chain) is proposed to be responsible for the specificity of 4CLs, especially for the exclusion of nonaromatic compounds, because the distance between the Gly-331 and Tyr-236 side chains only allows for the binding of phenyl derivatives. Any other compound would be too thick to fit in. The Y236A mutation, which relieves steric conflicts in this area, enhances P. tomentosa 4CL1 activity, probably because this mutant has high affinities for hydroxycinnamate substrates, but at the cost of specificity.

The two meta-positions of the 4-hydroxyphenyl group of APP are in different environments and that could account for the specificity of the P. tomentosa 4CL1 toward different substituted hydroxycinnamate substrates. One of the meta-positions is stacked tightly against the P. tomentosa 4CL1 structure, leaving no space for substitutions (Figure 4B). The other meta-position is located near the side chain of Val-338, and this position is surrounded by a cavity (Figure 4C), which accommodates the 3-hydroxyl and 3-methoxy groups of caffeic acid and ferulic acid, respectively. This explains why P. tomentosa 4CL1 cannot activate 3,5-disubstituted hydroxycinnamate substrates such as sinapic acid.

Because of the highly conserved nature of the residues surrounding the substrate binding pocket, most 4CLs are predicted to have substrate binding pockets very similar to what we see in P. tomentosa 4CL1. Most characterized 4CL enzymes are inactive toward sinapic acid because their substrate binding pocket is not large enough to bind this compound, as is the case in P. tomentosa 4CL1 structures. Interestingly, two 4CL enzymes, soybean (Glycine max) 4CL1 (Lindermayr et al., 2002) and Arabidopsis 4CL4 (Schneider et al., 2003; Hamberger and Hahlbrock, 2004), were found to be active toward sinapic acid. Sequence comparison shows that both proteins have a deletion mutation at positions that correspond to Val-338 of P. tomentosa 4CL1, which is located in the bulged loop discussed above. Deleting the corresponding residues from G. max 4CL2 and 4CL3 conveyed sinapate activity to the mutants (Lindermayr et al., 2003). Since the bulged loop of G336PV338 is part of the substrate binding pocket of 4CLs, we propose that the deletion of Val-338 would straighten the bulge, remove one side of the substrate binding pocket, and significantly enlarge the pocket, thus enabling the 4CL enzymes to bind to the larger sinapic acid.

Identification of a CoA Binding Tunnel

Although CoA was included in some crystallization conditions, P. tomentosa 4CL1 always crystallized as CoA-free forms. Several ANL enzymes, such as Salmonella enterica acetyl CoA synthetase (Gulick et al., 2003), Alcaligenes sp 4-chlorobenzoat:CoA ligase (Reger et al., 2008), and acyl-CoA synthetase medium-chain family member 2A, have been cocrystallized with CoA or the final CoA thioester products. These structures also happen to be the most similar to APP-complexed P. tomentosa 4CL1. When these structures and P. tomentosa 4CL1 are superposed based on their Cα atoms, their substrates superpose well with the coumaroyl moiety of P. tomentosa 4CL1 (Figure 6A), their pantetheine binding tunnels correspond well with a tunnel in the P. tomentosa 4CL1 structure, and the residues lining the binding tunnels of these proteins superpose well with their counterparts from P. tomentosa 4CL. The pantetheine moieties from these structures fit perfectly into a tunnel of P. tomentosa 4CL1 (Figure 6B). Only some minor steric conflicts are observed on the surface of P. tomentosa 4CL1, which can be easily relieved by local structural adjustments. Furthermore, His-234, which corresponds to His-207 of 4-chlorobenzoat:CoA ligase and is conserved as aromatic residues in ANL enzymes, adopts a conformation identical to that of the thioester-forming complex of 4-chlorobenzoat:CoA ligase in the APP-4CL1 complex (Reger et al., 2008). The AMP-complexed structure of P. tomentosa 4CL1, on the other hand, has this residue in a conformation that corresponds to the adenylate-forming complex of 4-chlorobenzoat:CoA ligase and blocks CoA from approaching the active site. Therefore, it is reasonable to propose that CoA would bind P. tomentosa 4CL1 in a way very similar to what was observed in these proteins.

Figure 6.

Figure 6.

Open in a new tab

CoA Binding Site of 4CL1.

(A) When the thioester-forming structure of 4-chlorobenzoate:CoA ligase is superposed on the APP-4CL1 complex structure, the 4-chlorophenyl group of the thioester product (yellow) of the former structure superposes well with the coumaroyl group (pink) of the latter structure. His-234 and Tyr-236, which play important roles in regulating CoA access to the active center and in hydroxycinnamate binding, respectively, are shown in green.

(B) The thioester product of the 4-chlorobenzoate:CoA ligase thioester-forming structure fit into the pantetheine binding tunnel of APP-4CL1 structure.

This pantetheine binding tunnel is located at the interdomain interface. In the apo-4CL1 structure, the residues that might interact with CoA in the 4CL1-APP complex are located far away from each other, and the tunnel is completely exposed, rendering the protein unable to bind CoA. The Cα atom of Gly-441, for example, is part of the CoA binding pocket in the 4CL1-APP complex, but it moves more than 22 Å in the apo-4CL1 structure. This mechanism, we propose, prevents P. tomentosa 4CL1 from binding CoA prematurely in the absence of the adenylation product.

Possible Enzymatic Mechanisms for P. tomentosa 4CL1

Based on extensive structural and enzymatic studies, a domain alternation theory has been established as the enzymatic mechanism of ANL proteins and has been elegantly reviewed by Gulick (2009). According to this theory, ANL proteins catalyze two partial reactions with two distinct catalytic centers. Both centers are located on the surface of the C-domain and are brought to substrate binding sites by large movements of C-domains relative to N-domains at different stages of the reactions.

Of the two partial reactions of ANL proteins, the adenylate-forming reactions and their catalysis mechanisms are strictly conserved. This partial reaction is invariantly catalyzed by a Lys residue located in the conserved sequence motif A10 (Gulick, 2009). We identified Lys-523 as the catalytic residue of P. tomentosa 4CL1. Although we have not been able to crystallize P. tomentosa 4CL1 in the adenylate-forming conformation, the facts that the adenylate-forming partial reactions are strictly conserved and that the K523A mutation completely abolishes P. tomentosa 4CL1 activity (Figure 5) verify that P. tomentosa 4CL1 has an identical mechanism as other ANL proteins for the adenylate-forming partial reaction.

The second partial reactions catalyzed by ANL proteins are much more diverse. Firefly luciferease, for example, catalyzes the oxidative decarboxylation of the adenylate intermediate in this step (de Wet et al., 1985). As a result, the catalytic centers for this partial reaction are less conserved. Here, we propose that P. tomentosa 4CL1 adopts the catalyzing conformation of the thioester-forming half reactions in the APP-4CL1 complex structure. This proposal is based on the observations that in the APP-4CL1 complex structure, the pantetheine binding tunnel forms at the interdomain (Figure 6B) and that the His-234 side chain adopts the thioester-forming conformation (Figure 4D). Furthermore, the conformation of the APP-4CL1 complex is the same as the ternary complex of S. enterica Acs with a substrate analog and CoA, and the latter has been demonstrated to be the competent conformation for the thioester-forming partial reaction (Gulick et al., 2003). Consequently, residues Lys-438 and Gln-443 are proposed to form the catalytic center of the thioester-forming partial reaction of P. tomentosa 4CL1. The importance of these two residues is verified by the total loss of 4CL1 activity when they are mutated (Figure 5).

Here, we propose an enzymatic mechanism for P. tomentosa 4CL1 based on the above considerations and the general domain alternation theory (Gulick, 2009). The mechanism begins with the binding of ATP and hydroxycinnamate substrates. The binding results in 4CL1 adopting the catalytic conformation for the adenylate-forming partial reaction, in which the side chain of Lys-523 interacts with and directs the carboxylate group of the bound hydroxycinnamates for the nucleophilic attack of the α-phosphate of ATP, resulting in an AMP-hydroxycinnamate conjugate and a PPi molecule. The release of PPi then propels 4CL1 to the catalytic conformation of the thioester-forming partial reaction, as we observed in its APP complex. In this conformation, the pantetheine binding tunnel is formed at the interdomain interface, and the side chain of His-234 swings aside to allow access of CoA to the AMP-hydroxycinnamate conjugate. The AMP-hydroxycinnamate conjugate and CoA are then catalyzed by side chains of Lys-438 and Gln-443 to form the final thioester product. The C-domain rotates again to expose the substrate binding site, and the thioester and AMP are released.

In summary, we determined the crystal structure of P. tomentosa 4CL1 in apo form and in binary complexes with AMP and APP, an intermediate analog. Based on these structures and mutagenesis studies, we identified Lys-438, Gln-443, and Lys-523 as the catalytic residues of 4CL1. We identified the hydroxycinnamate binding pocket and residues that are important for the substrate specificities of 4CL1. This substrate binding model can be used to satisfactorily interpret the specificities of 4CL enzymes, including those of G. max 4CL1 and Arabidopsis 4CL4, toward sinapic acid. We also demonstrated that APP binding induced a conformational change that leads to the formation of a CoA binding pocket at the interdomain interface of P. tomentosa 4CL1 and that the conformation adopted by 4CL1 in the complexes might be competent for catalyzing the thioester formation half reactions. We proposed an enzymatic mechanism based on structural and mutagenesis studies. Because the sequence identities between plant 4CLs are high and the structures of ANL enzymes are highly conserved, the high-resolution crystal structures of 4CL1 will greatly increase our understanding of the structure-function relationships of this important family of proteins. The identification of the residues responsible for P. tomentosa 4CL1 substrate specificities lays a solid foundation for future molecular engineering of 4CL enzymes.

METHODS

Protein Purification

Populus tomentosa 4CL1 was amplified by PCR from P. tomentosa cDNA, using a primer pair of 5′-CGCAATGGACGCCACAATGAAT-3′ and 5′-ACTGTCTTACGTTGGGTACG-3′, and cloned into a pMD18-T (TaKaRa) vector. The gene was then subcloned by PCR using a primer pair of 5′-CGGGATCCCGCAATGGACGCCACAATGAAT-3′ and 5′-CCCCCGGGGGCATCTTCAGTTA-3′ and digested using endonuclease BamHI and ligated into the BamHI restriction site of a pQE31 overexpression vector (Qiagen). The resulting pQE31-4CL1 plasmid expresses a tag of MRGSHHHHHHTDPAMDAT at the N terminus of P. tomentosa 4CL1. Escherichia coli strain M15(pREP4) transformed with the pQE31-4CL1 plasmid was grown in Luria-Bertani medium and induced with 0.4 mM isopropyl β-D-1-thiogalactopyranoside at 300K. The cells were collected by centrifugation at 2500g and lysed by ultrasonification. The lysate was clarified by centrifugation at 20,000g, and the supernatant was loaded on a Ni-NTA affinity column. 4CL1 was eluted using a solution containing 50 mM Na2HPO4, pH 8.0, 0.3 M NaCl, and 0.25 M imidazole. The protein was further purified using a HiLoad 16/60 Superdex 75 XK size-exclusion column (GE Healthcare) with an elution solution containing 50 mM Tris-HCl buffer at pH 8.0.

Cloning and Mutagenesis

Genes encoding 4CL1 mutants were prepared by site-directed mutagenesis using the pQE31-4CL1 plasmid as template. PCR reactions were performed using the LA Taq DNA polymerase from TaKaRa Bio following the protocols provided by the vendor. The mutants were purified following the same protocol described above for wild-type 4CL1.

4CL1 Crystallization

Crystals of the apo-protein of 4CL1 were obtained by the sitting drop method (McPherson, 1999). Protein solution containing 20 mg/mL of 4CL1 in 50 mM Tris-HCl, pH 7.5, 10% (v/v) glycerol, 0.1 M NaCl, 2 mM DTT, and 1 mM EDTA was mixed with precipitate solution containing 50 mM CAPSO, pH 9.8, 35% (w/v) PEG8000, and 0.25 M NaCl in a 1:1 ratio and sealed with paraffin oil. Needle-like crystals generally appear within 24 h and grow to their full sizes in 3 to 4 d.

Crystals of 4CL1 complexed with AMP or APP were obtained by the hanging drop method (McPherson, 1999). 4CL sample used in the experiments was in the same solution as described above, with the exception of the addition of 10 mM AMP (Sigma-Aldrich) or APP (Laviana) prior to crystallization. The precipitant solution contained 50 mM MES, pH 6.0, and 1.8 M ammonium citrate. Cubic crystals generally appeared within 1 week.

Data Collection

A data set at 2.4 Å resolution was collected on a native apo-4CL1 crystal at a wavelength of 1.0000 Å at Beamline 5A, Photon Factory, KEK, Japan. A data set at 3.0 Å resolution was collected on a SeMet-derivatized apo-4CL1 crystal at the selenium peak wavelength of 0.9790 Å at the Beijing Synchrotron Radiation Facility, China. The program package HKL2000 was used to integrate the frames collected at the synchrotron facilities. The data set for the AMP-complexed 4CL crystal was collected using a Raxis-IV++ ImagePlate area detector mounted on a Rigaku rotating-anode x-ray generator running at 45 kV/45 mA. The data set for the APP-complexed crystal was collected at Beamline 17U, Shanghai Synchrotron Radiation Facility, China. The program MOSFLM was used to integrate the collected frames. The program SCALA from the CCP4 package was used to scale the data.

Structural Determination and Refinement

The structure of apo-4CL1 was determined using phase information from the molecular replacement solution and selenium anomalous scattering. The program PHASER (McCoy et al., 2005) was used to obtain molecular replacement solutions using the N- and C-domains of firefly luciferase (Conti et al., 1996) as two independent ensembles. The program RESOLVE (Terwilliger, 2000) was used to remove model bias. The program SOLVE (Terwilliger and Berendzen, 1999) was used to determine and refine the positions of the selenium atoms, and the program RESOLVE was used to combine the phase information from the molecular replacement solution and from selenium anomalous scattering and to refine the resulting phases using solvent flattening. The graphics program O (Jones et al., 1991) was used to build the residues into the electron density map and for manual adjustment of the model. The structure was refined using the program CNS (Brünger et al., 1998), and 5% of the reflections were randomly chosen for Rfree calculations and were excluded from structural refinement. CNS refinements and manual model rebuilding were performed alternatively until the quality of the model was satisfactory.

The AMP- and APP-complexed 4CL1 structures were solved by the molecular replacement method with the program PHASER (McCoy et al., 2005), using the N- and C-terminal domains of the apo structure as independent ensembles. The protocol described above was followed to refine the model.

Enzymatic Activity Assays

Concentrations of 4CL1 and its mutants were assayed with the Bradford method using BSA as protein standard (Bradford, 1976). The enzymatic activities were assayed according to the protocols of Knobloch and Hahlbrock (1975) with minor modifications. Briefly, a 490-μL reaction buffer containing 5 mM ATP, 0.3 mM CoA, 5 mM Mg2+, 0.2 mM substrate (4-coumaric acid, caffeic acid, or ferulic acid), and 0.2 M Tris-HCl at pH 7.8 was incubated at 300K. Blanks used as background contained 0.2 mM substrate, 5 mM Mg2+, and 0.2 M Tris-HCl at pH 7.8. Then, 2 μg of purified proteins in 10 μL of solution was added to start the enzymatic reactions, and UV absorptions were recorded continuously for 1 h at an interval of 40 s, using the blanks as background. The wavelengths used were 333 nm for 4-coumaroyl-, 362 nm for caffeoyl-, and 346 nm for feruloyl-CoA esters (Stöckigt and Zenk, 1975). The activities of 4CL1 and its mutants were calculated using measured data of the first 6 to 8 min.

Accession Numbers

The coordinates and structural factors of the apo-, AMP-complexed, and APP-complexed P. tomentosa 4CL1 crystals were deposited in the Protein Data Bank (www.rcsb.org) with the accession numbers 3A9U, 3A9V, and 3NI2, respectively. The Protein Data Bank accession code of acyl-CoA synthetase medium-chain family member 2A is 3EQ6. Sequence data for P. tomentosa 4CL1 can be found in the GenBank/EMBL data libraries under accession number AY043495.

Acknowledgments

We thank the staff at Beamline 3W1A, Beijing Synchrotron Radiation Facility, Beamline 17U, Shanghai Synchrotron Radiation Facility, and Beamline 5A (KEK, Tsukuba, Japan) for their assistance with data collections. We thank X.D. Zhao and X.M. Zheng of the Core Facilities of Institute of Biophysics, Chinese Academy of Sciences, for their technique support. D.-C.W. is supported by the “973” and “863” Research Programs of China (2006CB806502, 2006CB10903/901, and 2006AA02A322). X.-N.J. is supported by the “973” Research Program (G1999016005) and by a grant from the Natural Science Foundation of China (NSF30630053). Y.G. is supported by a grant from China Postdoctoral Science Foundation (20090450015).

References

  1. Babbitt P.C., Kenyon G.L., Martin B.M., Charest H., Slyvestre M., Scholten J.D., Chang K.H., Liang P.H., Dunaway-Mariano D. (1992). Ancestry of the 4-chlorobenzoate dehalogenase: Analysis of amino acid sequence identities among families of acyl:adenyl ligases, enoyl-CoA hydratases/isomerases, and acyl-CoA thioesterases. Biochemistry 31: 5594–5604 [DOI] [PubMed] [Google Scholar]
  2. Boudet A.M., Kajita S., Grima-Pettenati J., Goffner D. (2003). Lignins and lignocellulosics: A better control of synthesis for new and improved uses. Trends Plant Sci. 8: 576–581 [DOI] [PubMed] [Google Scholar]
  3. Bradford M.M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248–254 [DOI] [PubMed] [Google Scholar]
  4. Branchini B.R., Murtiashaw M.H., Magyar R.A., Anderson S.M. (2000). The role of lysine 529, a conserved residue of the acyl-adenylate-forming enzyme superfamily, in firefly luciferase. Biochemistry 39: 5433–5440 [DOI] [PubMed] [Google Scholar]
  5. Brünger A.T., et al. (1998). Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54: 905–921 [DOI] [PubMed] [Google Scholar]
  6. Conti E., Franks N.P., Brick P. (1996). Crystal structure of firefly luciferase throws light on a superfamily of adenylate-forming enzymes. Structure 4: 287–298 [DOI] [PubMed] [Google Scholar]
  7. Conti E., Stachelhaus T., Marahiel M.A., Brick P. (1997). Structural basis for the activation of phenylalanine in the non-ribosomal biosynthesis of gramicidin S. EMBO J. 16: 4174–4183 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. de Wet J.R., Wood K.V., Helinski D.R., Deluca M. (1985). Cloning of firefly luciferase cDNA and the expression of active luciferase in Escherichia coli. Proc. Natl. Acad. Sci. USA 82: 7870–7873 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Ehlting J., Büttner D., Wang Q., Douglas C.J., Somssich I.E., Kombrink E. (1999). Three 4-coumarate:coenzyme A ligases in Arabidopsis thaliana represent two evolutionarily divergent classes in angiosperms. Plant J. 19: 9–20 [DOI] [PubMed] [Google Scholar]
  10. Ehlting J., Shin J.J., Douglas C.J. (2001). Identification of 4-coumarate:coenzyme A ligase (4CL) substrate recognition domains. Plant J. 27: 455–465 [DOI] [PubMed] [Google Scholar]
  11. Ferrer J.L., Austin M.B., Stewart C., Jr., Noel J.P. (2008). Structure and function of enzymes involved in the biosynthesis of phenylpropanoids. Plant Physiol. Biochem. 46: 356–370 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fulda M., Heinz E., Wolter F.P. (1994). The fadD gene of Escherichia coli K12 is located close to rnd at 39.6 min of the chromosomal map and is a new member of the AMP-binding protein family. Mol. Gen. Genet. 242: 241–249 [DOI] [PubMed] [Google Scholar]
  13. Gulick A.M. (2009). Conformational dynamics in the Acyl-CoA synthetases, adenylation domains of non-ribosomal peptide synthetases, and firefly luciferase. ACS Chem. Biol. 4: 811–827 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gulick A.M., Starai V.J., Horswill A.R., Homick K.M., Escalante-Semerena J.C. (2003). The 1.75Å crystal structure of acetyl-CoA synthetase bound to adenosine-5′-propylphosphate and coenzyme A. Biochemistry 42: 2866–2873 [DOI] [PubMed] [Google Scholar]
  15. Hamberger B., Hahlbrock K. (2004). The 4-coumarate:CoA ligase gene family in Arabidopsis thaliana comprises one rare, sinapate-activating and three commonly occurring isoenzymes. Proc. Natl. Acad. Sci. USA 101: 2209–2214 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hoffmann L., Besseau S., Geoffroy P., Ritzenthaler C., Meyer D., Lapierre C., Pollet B., Legrand M. (2004). Silencing of hydroxycinnamoyl-coenzyme A shikimate/quinate hydroxycinnamoyltransferase affects phenylpropanoid biosynthesis. Plant Cell 16: 1446–1465 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Horswill A.R., Escalante-Semerena J.C. (2002). Characterization of the propionyl-CoA synthetase (PrpE) enzyme of Salmonella enterica: Residue Lys592 is required for propionyl-AMP synthesis. Biochemistry 41: 2379–2387 [DOI] [PubMed] [Google Scholar]
  18. Hu W.J., Kawaoka A., Tsai C.J., Lung J., Osakabe K., Ebinuma H., Chiang V.L. (1998). Compartmentalized expression of two structurally and functionally distinct 4-coumarate:CoA ligase genes in aspen (Populus tremuloides). Proc. Natl. Acad. Sci. USA 95: 5407–5412 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Jogl G., Tong L. (2004). Crystal structure of yeast acetyl-coenzyme A synthetase in complex with AMP. Biochemistry 43: 1425–1431 [DOI] [PubMed] [Google Scholar]
  20. Jones T.A., Zou J.Y., Cowan S.W., Kjeldgaard M. (1991). Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47: 110–119 [DOI] [PubMed] [Google Scholar]
  21. Kajita S., Ishifuji M., Ougiya H., Hara S., Kawabata H., Morohoshi N., Katayama Y. (2002). Improvement in pulping and bleaching properties of xylem from transgenic tobacco plants. J. Sci. Food Agric. 82: 1216–1223 [Google Scholar]
  22. Knobloch K.H., Hahlbrock K. (1975). Isoenzymes of p-coumarate: CoA ligase from cell suspension cultures of Glycine max. Eur. J. Biochem. 52: 311–320 [DOI] [PubMed] [Google Scholar]
  23. Lee D., Meyer K., Chapple C., Douglas C.J. (1997). Antisense suppression of 4-coumarate:coenzyme A ligase activity in Arabidopsis leads to altered lignin subunit composition. Plant Cell 9: 1985–1998 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Li L., Zhou Y.H., Cheng X.F., Sun J.Y., Marita J.M., Ralph J., Chiang V.L. (2003). Combinatorial modification of multiple lignin traits in trees through multigene cotransformation. Proc. Natl. Acad. Sci. USA 100: 4939–4944 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lindermayr C., Fliegmann J., Ebel J. (2003). Deletion of a single amino acid residue from different 4-coumarate:CoA ligases from soybean results in the generation of new substrate specificities. J. Biol. Chem. 278: 2781–2786 [DOI] [PubMed] [Google Scholar]
  26. Lindermayr C., Möllers B., Fliegmann J., Uhlmann A., Lottspeich F., Meimberg H., Ebel J. (2002). Divergent members of a soybean (Glycine max L.) 4-coumarate:coenzyme A ligase gene family. Eur. J. Biochem. 269: 1304–1315 [DOI] [PubMed] [Google Scholar]
  27. Marahiel M.A., Stachelhaus T., Mootz H.D. (1997). Modular peptide synthetases involved in nonribosomal peptide synthesis. Chem. Rev. 97: 2651–2674 [DOI] [PubMed] [Google Scholar]
  28. May J.J., Kessler N., Marahiel M.A., Stubbs M.T. (2002). Crystal structure of DhbE, an archetype for aryl acid activating domains of modular nonribosomal peptide synthetases. Proc. Natl. Acad. Sci. USA 99: 12120–12125 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. McCoy A.J., Grosse-Kunstleve R.W., Storoni L.C., Read R.J. (2005). Likelihood-enhanced fast translation functions. Acta Crystallogr. D Biol. Crystallogr. 61: 458–464 [DOI] [PubMed] [Google Scholar]
  30. McPherson A. (1999). Crystallization of Biological Macromolecules. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; ). [Google Scholar]
  31. Nakatsu T., Ichiyama S., Hiratake J., Saldanha A., Kobashi N., Sakata K., Kato H. (2006). Structural basis for the spectral difference in luciferase bioluminescence. Nature 440: 372–376 [DOI] [PubMed] [Google Scholar]
  32. Ragauskas A.J., et al. (2006). The path forward for biofuels and biomaterials. Science 311: 484–489 [DOI] [PubMed] [Google Scholar]
  33. Reger A.S., Wu R., Dunaway-Mariano D., Gulick A.M. (2008). Structural characterization of a 140 degrees domain movement in the two-step reaction catalyzed by 4-chlorobenzoate:CoA ligase. Biochemistry 47: 8016–8025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Schneider K., Hovel K., Witzel K., Hamberger B., Schomburg D., Kombrink E., Stuible H.P. (2003). The substrate specificity-determining amino acid code of 4-coumarate:CoA ligase. Proc. Natl. Acad. Sci. USA 100: 8601–8606 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Silber M.V., Meimberg H., Ebel J. (2008). Identification of a 4-coumarate:CoA ligase gene family in the moss, Physcomitrella patens. Phytochemistry 69: 2449–2456 [DOI] [PubMed] [Google Scholar]
  36. Stöckigt J., Zenk M.H. (1975). Chemical syntheses and properties of hydroxycinnamoyl-coenzyme A derivatives. Z. Naturforsch., C, Biosci. 30: 352–358 [DOI] [PubMed] [Google Scholar]
  37. Starai V.J., Celic I., Cole R.N., Boeke J.D., Escalante-Semerena J.C. (2002). Sir2-dependent activation of acetyl-CoA synthetase by deacetylation of active lysine. Science 298: 2390–2392 [DOI] [PubMed] [Google Scholar]
  38. Stuible H.P., Kombrink E. (2001). Identification of the substrate specificity-conferring amino acid residues of 4-coumarate:coenzyme A ligase allows the rational design of mutant enzymes with new catalytic properties. J. Biol. Chem. 276: 26893–26897 [DOI] [PubMed] [Google Scholar]
  39. Terwilliger T.C. (2000). Maximum-likelihood density modification. Acta Crystallogr. D Biol. Crystallogr. 56: 965–972 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Terwilliger T.C., Berendzen J. (1999). Automated MAD and MIR structure solution. Acta Crystallogr. D Biol. Crystallogr. 55: 849–861 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Voo K.S., Whetten R.W., O’Malley D.M., Sederoff R.R. (1995). 4-Coumarate:coenzyme a ligase from loblolly pine xylem. Isolation, characterization, and complementary DNA cloning. Plant Physiol. 108: 85–97 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wang R., Lu Y., Fang X., Wang S. (2004). An extensive test of 14 scoring functions using the PDBbind refined set of 800 protein-ligand complexes. J. Chem. Inf. Comput. Sci. 44: 2114–2125 [DOI] [PubMed] [Google Scholar]
  43. Zhang Y., Brown G., Whetten R., Loopstra C.A., Neale D., Kieliszewski M.J., Sederoff R.R. (2003). An arabinogalactan protein associated with secondary cell wall formation in differentiating xylem of loblolly pine. Plant Mol. Biol. 52: 91–102 [DOI] [PubMed] [Google Scholar]

Articles from The Plant Cell are provided here courtesy of Oxford University Press

RESOURCES