structural and mutational studies of anthocyanin

Structural and Mutational Studies of AnthocyaninMalonyltransferases Establish the Features ofBAHD Enzyme Catalysis*□S

Received for publication, January 23, 2007, and in revised form, March 1, 2007 Published, JBC Papers in Press, March 23, 2007, DOI 10.1074/jbc.M700638200

Hideaki Unno‡1, Fumiko Ichimaida§1, Hirokazu Suzuki§1, Seiji Takahashi§, Yoshikazu Tanaka¶, Atsushi Saito§,Tokuzo Nishino§, Masami Kusunoki‡2, and Toru Nakayama§3

From the ‡Institute for Protein Research, Osaka University, 3-2 Yamada-oka, Suita, Osaka 565-0871, Japan, §Departmentof Biomolecular Engineering, Graduate School of Engineering, Tohoku University, 6-6-11 Aobayama, Sendai,Miyagi 980-8579, Japan, and ¶Suntory Research Center, Shimamoto-cho, Mishima-gun, Osaka 618-8503, Japan

The BAHD family is a class of acyl-CoA-dependent acyltrans-ferases that are involved inplant secondarymetabolismandshowadiverse range of specificities for acyl acceptors. Anthocyanin acyl-transferases make up an important class of the BAHD family andcatalyze theacylationofanthocyanins thatare responsible formostof the red-to-blue colors of flowers. Here, we describe crystallo-graphic andmutational studies of three similar anthocyaninmalo-nyltransferases from red chrysanthemum petals: anthocyanidin3-O-glucoside-6�-O-malonyltransferase (Dm3MaT1), anthocyani-din 3-O-glucoside-3�, 6�-O-dimalonyltransferase (Dm3MaT2),and a homolog (Dm3MaT3). Mutational analyses revealed thatseven amino acid residues in the N- and C-terminal regions areimportant for the differential acyl-acceptor specificity betweenDm3MaT1andDm3MaT2.CrystallographicstudiesofDm3MaT3provided the first structure of a BAHD member, complexed withacyl-CoA, showing the detailed interactions between the enzymeand acyl-CoAmolecules. The structure, combinedwith the resultsof mutational analyses, allowed us to identify the acyl-acceptorbinding siteof anthocyaninmalonyltransferases,which is structur-ally different from the correspondingportion of vinorine synthase,another BAHDmember, thus permitting the diversity of the acyl-acceptor specificity of BAHD family to be understood.

Plants producemore than 2� 105 types of secondarymetab-olites, many of which are of biomedical, pharmaceutical, andagricultural importance. A class of acyl-CoA-dependent acyl-

transferases plays versatile roles in the biosyntheses of thesemetabolites,making a very important contribution to the estab-lishment of the structural and functional diversities of themetabolites (1). Although these acyltransferases show only lowsequence similarities (15–30% identity) with each other, theyshare two highly conserved sequences, suggesting that theyevolved from a common ancestor. Thus, a single protein familyhas been proposed for these diversified acyltransferases of bio-logical and industrial significance and is referred to as theBAHD4 family (1–4).

Anthocyanin acyltransferases (AATs) form an importantclass of this family, catalyzing the transfer of an acyl group fromacyl-CoA to a sugar moiety of an anthocyanin (2), a class offlavonoids that is the origin of most of the orange to blue colorsof flowers and that plays important roles in plant reproductionand survival (5). The AAT-catalyzed acylation of anthocyaninsis important for the stabilization, accumulation in vacuoles, andmodulating the coloration of the pigments (2). It has been pro-posed that AAT catalysis proceeds through the formation of aternary complex consisting of acyl-CoA, an anthocyanin, andthe enzyme, where a general-base amino acid residue deproto-nates a hydroxy group of the anthocyanin substrate, therebypromoting its nucleophilic attack on the carbonyl of the thio-ester of acyl-CoA (6).The red color of the chrysanthemum (Dendranthema �

morifolium) mainly arises from cyanidin 3-O-3�,6�-O-dimalo-nylglucoside (Fig. 1, compound 3), which contains twomalonylgroups (7, 8). During the biosynthesis of the pigment the firstmalonylation takes place at the 6�-position of the cyanidin 3-O-glucoside (1) (termed 3MaT1 activity) and is catalyzed by either* This work was supported by the National Project on Protein Structure and

Functional Analysis promoted by the Ministry of Education, Culture,Sports, Science and Technology of Japan. The costs of publication of thisarticle were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.

□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Figs. 1S–3S.

The nucleotide sequence(s) reported in this paper has been submitted to theDDBJ/GenBankTM/EBI Data Bank with accession number(s) AB290338.

The atomic coordinates and structure factors (code 2E1U, 2E1T, and 2E1V)have been deposited in the Protein Data Bank, Research Collaboratory forStructural Bioinformatics, Rutgers University, New Brunswick, NJ(http://www.rcsb.org/).

1 These authors are equal contributors.2 To whom correspondence may be addressed. Tel. and Fax: 81-6-6879-8636;

E-mail: [email protected] To whom correspondence may be addressed. Tel. and Fax: 81-22-795-7270;

E-mail: [email protected].

4 The abbreviations used are: BAHD, benzyl alcohol acetyl-, anthocyanin-O-hydroxycinnamoyl-, anthranilate-N-hydroxycinnamoyl/benzoyl-, anddeacetylvindoline acetyltransferase; AAT, anthocyanin acyltransferase;Dm3MaT1, malonyl-CoA:anthocyanidin 3-O-glucoside-6�-O-malonyl-transferase of D. x morifolium; MaT, malonyltransferase; Dm3MaT2, malo-nyl-CoA:anthocyanidin 3-O-glucoside-3�,6�-O-dimalonyltransferase of D. xmorifolium; Dm3MaT3, an AAT homolog of D. x morifolium; Gt5AT,hydroxycinnamoyl-CoA:anthocyanin 3,5-O-diglucoside 6�-O-hydroxycin-namoyltransferase of G. triflora; Pf3AT, hydroxycinnamoyl-CoA:anthocya-nin 3-O-glucoside 6�-O-hydroxycinnamoyltransferase of P. frutescens;Ss5MaT1, malonyl-CoA:anthocyanin 5-O-glucoside-6�-O-malonyltrans-ferase of S. splendens; Ss5MaT2, malonyl-CoA:anthocyanin 5-O-glucoside-4�-O-malonyltransferase of S. splendens; Sc3MaT, malonyl-CoA:anthocy-anidin 3-O-glucoside-6�-O-malonyltransferase of S. cruentus; kb,kilobase(s).

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 282, NO. 21, pp. 15812–15822, May 25, 2007© 2007 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

15812 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 21 • MAY 25, 2007

by guest on February 2, 2018http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

of two distinct anthocyanin malonyltransferases, Dm3MaT1(malonyl-CoA:anthocyanidin 3-O-glucoside-6�-O-malonyl-transferase) or Dm3MaT2 (malonyl-CoA:anthocyanidin 3-O-glucoside-3�,6�-O-dimalonyltransferase) (Fig. 1) (8). The sec-ond malonylation (termed 3MaT2 activity) is exclusivelycatalyzed byDm3MaT2, which furthermalonylates the 3� posi-tion of the mono-malonylated product (cyanidin 3-O-6�-O-malonylglucoside, 2), producing cyanidin 3-O-3�,6�-O-dimalo-nylglucoside (3). Thus, Dm3MaT1 shows exclusive 3MaT1activity, whereas Dm3MaT2 shows both 3MaT1 and 3MaT2activities. It is particularly noteworthy that these two malonyl-transferases (MaTs) are 89% identical to each other in their pri-mary structures (8), serving as very good targets for studies ofstructural factors thatgovern theacyl acceptor specificityofAATs.The present study was undertaken to establish the struc-

tural basis of the acyl acceptor and donor specificities ofAATs. As the first step we carried out region swapping andsite-specific mutagenesis studies of Dm3MaT1 andDm3MaT2 to identify the residues responsible for differen-tial acceptor specificity between these two enzymes. In thesecond step we isolated a cDNA of another AAT homolog,Dm3MaT3, from red chrysanthemum petals, which yielded arecombinant protein that was suitable for x-ray crystallo-graphic analyses. Although the first crystal structure of a freeenzyme form of a BAHD family member, vinorine synthaseof the medicinal plant Rauvolfia serpentina, was reported in2005 (9), the crystal structure of a BAHD member com-plexed with acyl-CoA has not been reported.Wewere able todetermine the crystal structures of Dm3MaT3 at 1.8–2.2 Åof resolutions, providing the first crystal structure of aBAHD member complexed with acyl-CoA. The crystalstructures, along with the results of the present mutagenesisstudies, allowed us to unambiguously identify the acyl-CoAand acyl-acceptor binding sites in AATs, providing impor-tant information for general insights into the specificity andmechanism of BAHD catalysis.

EXPERIMENTAL PROCEDURES

Materials

Anthocyanins were generousgifts from Prof. Masa-atsu Yamagu-chi, Minami-Kyushu University. Allother chemicals used were of ana-lytical grade or sequencing grade, asappropriate.

Enzyme Assay

The standard reaction mixture(50 �l) consisted of 20 mM potas-sium phosphate, pH 7.0, 60 �Mmal-onyl-CoA, 120 �M cyanidin 3-O-glucoside, and enzyme. Themixturewithout enzyme was incubated at30 °C for 10 min. The reaction wasstarted by the addition of theenzyme. After incubation at 30 °Cfor 20min, the reactionwas stoppedby adding 50 �l of ice-cold 0.5%

(v/v) trifluoroacetic acid. Anthocyanins in the resulting mix-ture were analyzed by reversed-phase high performance liquidchromatography using a Shimadzu Prominence LC solutionsystem (Shimadzu, Kyoto, Japan) with an Asahipak ODP-50 4Ecolumn (4.6 � 250 mm; Shoko, Tokyo, Japan). Chromato-graphic conditions were as follows: solvent A, 0.1% (v/v) triflu-oroacetic acid; solvent B, 0.1% (v/v) trifluoroacetic acid and 90%(v/v) acetonitrile; column temperature, ambient; detection at520 nm; flow rate, 0.7 ml/min. The column was equilibratedwith 20% B before use. After injection, the column was initiallydeveloped with 20% B for 3 min followed by a linear gradientfrom 20 to 31%B in 15min followed by that from 31 to 56%B in1 min. The column was then washed with 56% B for 4 minfollowed by a linear gradient from 56 to 20% B in 1 min.

Enzyme Kinetics

Assays for the initial velocity for 3MaT1 and 3MaT2 activi-ties were carried out under steady-state conditions using astandard assay system (see above) with varying concentrationsof substrates. Typically, the anthocyanin and malonyl-CoAconcentrations were varied in the range of 5–54 and 5–30 �M,respectively, in the assays for Dm3MaT1, Dm3MaT2, and theirmutants. Double reciprocal plots of the activities of these AATsgenerally yielded a family of apparently parallel lines in therange of substrate concentrations employed, as has beenobserved for other AATs (6). The apparent kinetic parametersand their standard deviations were determined by fitting theinitial velocity data by a nonlinear least squares method using acomputer program (two-dimensional scientific graph-plottingtool program) to Equation 1, which best describes the observedparallel patterns of double reciprocal plots,

1/v � KmA/�kcat�E��A�� 1 � KmB/�B��/kcat�E� (Eq. 1)

where [A] and [B] denote the substrate concentrations, [E]denotes enzyme concentration, kcat denotes the catalytic rateconstant, and KmA and KmB, respectively, denote [A] and [B],

FIGURE 1. Malonyltransferase activities of Dm3MaT1, Dm3MaT2, and Dm3MaT3 from red chrysanthe-mum petals. The names of the substrates for 3MaT1 activity are pelargonidin 3-O-glucoside (R1 � R2 � -H),cyanidin 3-O-glucoside (R1 � -OH, R2 � -H), and delphinidin 3-O-glucoside (R1 � R2 � -OH) (1). Substrates for3MaT2 activity (2) and the product (3) are collectively referred to as anthocyanidin 3-O-6�-O-malonylglucosideand anthocyanidin 3-O-3�,6�-O-dimalonylglucoside, respectively. The structures of the anthocyanins areshown as their flavylium forms. Key positional numberings are labeled on 1.

Structures of a BAHD Family Enzyme

MAY 25, 2007 • VOLUME 282 • NUMBER 21 JOURNAL OF BIOLOGICAL CHEMISTRY 15813


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

which give a kcat[E]/2 value in the presence of a saturating con-centration of a counter substrate. Note that our previous prod-uct inhibition studies showed that the AAT-catalyzed malonyltransfer proceeds via a ternary complex mechanism (6), con-sistent with the results of the present crystallographic studies(see below). Under some circumstances, double reciprocalplots of ternary complex mechanisms are known to yield paral-lel patterns (10).

Region Swapping and Site-specific Mutagenesis

The expression plasmids pQE-Dm3MaT1 and pQE-Dm3MaT2 (8) share unique restriction enzyme sites (SphI,PmlI, SpeI, and SalI in the order of 533 orientation) in eachAAT sequence. Digestion of each of these plasmids with SphI-PmlI produces two DNA fragments that are 0.5 and 3.3 kb insize. The 0.5-kb fragment derived from pQE-Dm3MaT2 wereligated with the 3.3-kb fragment from pQE-Dm3MaT1 to pro-duce the plasmid pQE-Dm3MaT(211), which expresses aDm3MaT1 mutant in which region A is replaced with that ofDm3MaT2 (see Fig. 2 and supplemental Fig. 1S). Likewise, the0.5-kb fragment obtained from pQE-Dm3MaT1 was ligatedwith the 3.3-kb fragment from pQE-Dm3MaT2 to produce a

plasmid expressing Dm3MaT(122).Plasmids expressing Dm3MaT(121)and Dm3MaT(212) were preparedby exchanging the PmlI/SpeI-di-gested DNA fragments of pQE-Dm3MaT1 and pQE-Dm3MaT2,and those expressing Dm3MaT(112)and Dm3MaT(221) were preparedby exchanging the SpeI/SalI-di-gested DNA fragments of pQE-Dm3MaT1 and pQE-Dm3MaT2.Escherichia coli JM109 cells weretransformed with the resultingplasmids to overexpress eachregion-swapped mutant. The plas-mids for the expression of site-specific mutants of Dm3MaT(212)was prepared by PCR using pQE-Dm3MaT(212) as a template andmutagenesis primers (not shown).

cDNA Cloning, Protein Expression,and Purification of Dm3MaT3

cDNA Cloning and Expression ofDm3MaT3—The Dm3MaT3 cDNAwas isolated from red buds of D. �morifolium essentially as describedpreviously (8).TheDm3MaT3 cDNAwas cloned into pBluescript SK.From the resulting plasmid, thefull-length Dm3MaT3 cDNA wasamplified by PCR using the prim-ers 5-TGCGAGCATATGGCTT-CTCTTC-3 and 5-CCCGGGG-ATCCTTAAAGG-3 to introduceNdeI and BamHI restriction sites

(respective underlined sites) containing in-frame start and stopcodons into the 5- and 3-ends of the cDNA, respectively. Theamplified fragment was cloned into a pCR4Blunt-TOPOvector(Invitrogen) and sequenced to confirm the absence of PCRerrors. The fragment was then digested with NdeI and BamHIand ligated with an expression vector, pET-15b (Novagen,Madison, WI), which had previously been digested with NdeIand BamHI. The resulting plasmid, pET-Dm3MaT3, whichencodes an N-terminal in-frame fusion of Dm3MaT3 with aHis6 tag, was used to transform E. coli BL21(DE3) cells. Afterpreculturing the transformant cells at 37 °C overnight in Luria-Bertani broth containing 50 �g/ml ampicillin, 10 ml of the cul-ture was inoculated into 2000 ml of the same medium. Afterincubation at 30 °C until the A600 reached 0.5, isopropyl 1-�-D-thiogalactoside was added to the broth to a final concentrationof 0.4 mM followed by cultivation at 30 °C for 14 h.Purification of Recombinant Dm3MaT3—All subsequent

operations were conducted at 0–4 °C. The cells were harvestedby centrifugation (5000 � g, 15 min) and resuspended in bufferH (20 mM potassium phosphate, pH 7.4, containing 15 mM2-mercaptoethanol, 10 mM imidazole, and 0.5 M NaCl). Thecells were disrupted at 4 °C by 10 cycles of ultrasonication

FIGURE 2. Amino acid sequence comparison of Dm3MaT1, Dm3MaT2, and Dm3MaT3. Arrows above theDm3MaT1 sequence show regions A, B, and C that were defined in region swapping studies. The box labeledwith motif 2 shows the motif that is specifically conserved among AATs, and those labeled with motif 1 andmotif 3 show motifs that are highly conserved among BAHD family of enzymes, including AATs. Filled circlesbelow the Dm3MaT3 sequence indicate the location of amino acid residues corresponding to those importantfor 3MaT2 activity of Dm3MaT2 as revealed by mutational studies (see also Fig. 4). Amino acid residues that aredifferent from those of Dm3MaT1 are shown in red.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

(where one cycle corresponds to 10 kHz for 10 s followed by aninterval of 50 s). Cell debris was removed by centrifugation(5000 � g, 15 min). Polyethyleneimine was slowly added to thesupernatant solution to a final concentration of 0.3% (v/v). Themixture was allowed to stand at 4 °C for 30 min followed bycentrifugation (8000 � g, 15 min) and filtration through a0.22-�m filter. The supernatant was applied to a HisTrap HPcolumn (5 ml; GE Healthcare) equilibrated with buffer H. Thecolumn was extensively washed with buffer H, and the enzymewas eluted with buffer H containing 200 mM imidazole. Theactive fractions were collected, concentrated, and equilibratedwith buffer T (20 mM Tris-HCl, pH 7.4, containing 15 mM2-mercaptoethanol) by ultrafiltration with an Amicon Ultra-15centrifugal filter device (30,000 molecular weight cut-off; Mil-lipore, Bedford, MA). The enzyme was then digested withthrombin (Novagen; 1 unit/mg Dm3MaT3 protein) at 4 °C for16 h in 20mMTris-HCl, pH 8.4, containing 0.15 MNaCl and 2.5mM CaCl2 to remove the His6 tag from the recombinantenzyme. The resulting mixture was applied to a HisTrap HPcolumn (5 ml) equilibrated with buffer H. The column wasextensively washed with buffer H, and the flow-through frac-tions, which containedDm3MaT3, were collected. The enzymesolution was concentrated and equilibrated with buffer Q (10mMpotassiumphosphate, pH 7.4, containing 15mM 2-mercap-toethanol) by ultrafiltration. The resulting enzyme solutionwasapplied to a HiTrap QHP column (5 ml; GE Healthcare) equil-ibrated with buffer Q at a flow rate of 1 ml/min using ÅKTApurifier (GE Healthcare). The column was washed with thesame buffer. The enzyme was eluted with a linear gradient of0–600mMNaCl in buffer Q. The enzyme solutionwas concen-trated and equilibrated with buffer A (0.01 MHepes-NaOH, pH7.0, containing 15 mM 2-mercaptoethanol) as described above.Proteins were quantified by the method of Bradford using a kit(Bio-Rad) with bovine serum albumin as the standard.For the production of selenomethionyl Dm3MaT3, E. coli

BL21-CodonPlus(DE3)-RIPL cells (Stratagene, La Jolla, CA)harboring pET-Dm3MaT3were grown in 10 liters of aminimalmedium (Se-Met core medium; Wako, Tokyo, Japan) supple-mented with L-selenomethionine (2.5 mg/ml, final concentra-tion) at 35 °C. The selenomethionyl Dm3MaT3 was purifiedfrom crude extracts of the recombinant cells essentially asdescribed above, except that an extra HiTrap Q chromatogra-phy was carried out before HisTrap column chromatography.

Crystallization

Dm3MaT3 crystals were grown at 20 °C over a period of 1–3weeks by the sitting-drop vapor-diffusion method. Dm3MaT3crystals complexedwithmalonyl-CoAwere obtained bymixing2 �l of protein solution including 2 mMmalonyl-CoA with 2 �lof reservoir solution (30% (w/v) polyethylene glycol 8000(PEG8000, Hampton Research), 100 mM sodium cacodylate,pH 6.5, 120 mM ammonium sulfate, and 10% glycerol). Crystalsof substrate-free native and selenomethionyl Dm3MaT3 wereobtained by mixing 2 �l of the protein solution with 2 �l ofreservoir solution (25% (w/v) polyethylene glycol 8000(PEG8000, Hampton Research), 100 mM sodium cacodylate(pH 6.5), 100 mM ammonium sulfate, 10% glycerol, and 10%KCl). Crystals were frozen at 90 K.

Data Collection, Structure Solution, and Refinement

All data sets were collected on beamline BL-5A at the photonfactory (KEK, Tsukuba, Japan). Measurements were performedat 90 K with � � 0.97909 Å (Se peak), � � 0.97928 Å (Se edge),� � 0.95000 Å high-remote), and � � 1.000 Å (native-free andnative-complex). Complete data sets were collected throughcontiguous rotation ranges at a given wavelength before pro-ceeding to the next wavelength. The data sets were processedand scaled with HKL2000 (11). All data sets belonged to spacegroup P21 with two molecules per asymmetric unit. TheDm3MaT3 structurewas solved by themultiwavelength anom-alous diffraction method, with selenomethionyl Dm3MaT3crystal.Initial phases were determined by the SOLVE program (12).

Phase improvement by density modification including non-crystallography symmetry averaging was performed by theRESOLVE program (13). A model was then obtained by auto-matic tracingwithARP/wARPprogram (14). The structurewasthen rebuilt byO (15) and refined using Refmac (16), with 5% ofthe data set aside as a free set. In subsequent refinement, theselenomethionyl Dm3MaT3 data set was replaced with native-free and native-complex data sets. Malonyl-CoA models werefitted into the substrate binding sites based on non-crystallog-raphy symmetry averaged difference electron density maps(Fig. 3C). Of 454 residues, 14 residues (N-terminal 5 residues,C-terminal 4 residues, and residues 365–369) in chains A and Bare not visible in the electron density and are probably disor-dered. All figures were produced using POVSCRIPT� (17),PyMOL (18), and RASTER3D (19). The C� atoms of native-complex of Dm3MaT3 and vinorine synthase structures weresuperimposed with the program LSQKAB (20).

RESULTS AND DISCUSSION

Region Swapping and Site-specificMutagenesis of Dm3MaT1and Dm3MaT2—To explore the amino acid residues respon-sible for the difference in the acyl-acceptor specificity ofmalonyl transfer between Dm3MaT1 and Dm3MaT2, wefirst carried out region swapping between these enzymes.Namely, regions A, B, and C of these two enzymes (see Fig. 2)were shuffled to obtain six swapped mutants: Dm3MaT(112),Dm3MaT(121), Dm3MaT(211), Dm3MaT(122), Dm3MaT(212),and Dm3MaT(221) (see also supplemental Fig. 1S). In theDm3MaT(112) mutant, for example, regions A and B comefrom Dm3MaT1, and region C comes from Dm3MaT2. Thesemutants were examined for their 3MaT1 and 3MaT2 activities(Table 1; see also Fig. 1). The results showed thatDm3MaT(122),aDm3MaT2mutant containing regionAofDm3MaT1, aswell asDm3MaT(121) and Dm3MaT(112) displayed no appreciable3MaT2 activity, although they showed 3MaT1 activities that werecomparable with that of Dm3MaT1. These results suggest thatregion A of Dm3MaT2 contains an amino acid residue(s) that isimportant for 3MaT2 activity. Dm3MaT(211) also did not show3MaT2 activity, suggesting that region B and/or C of Dm3MaT2may also contain an amino acid residue(s) that is important for3MaT2 activity. A comparison of the 3MaT2 activities ofDm3MaT(212) and Dm3MaT(221) showed that the apparentkcat/Km value for 3MaT2 activity of the formermutant was signif-




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

icantly higher than that of the latter, leading us to propose thatregionC,butnot regionB, containsanaminoacid residue(s) that isimportant for 3MaT2 activity.To identify the amino acid residues that are important for

3MaT2 activity in regions A and C, 9 site-specific mutants ofDm3MaT(212) were created in which each or a set of aminoacid residues unique to Dm3MaT2 in regions A and C ofDm3MaT(212) were replaced by those unique to Dm3MaT1(see Table 1), and their 3MaT and 3MaT2 activities were exam-ined. The results showed that the following single amino acidsubstitutions in region A resulted in a significant loss of 3MaT2activity of Dm3MaT(212) (relative kcat/Km for acyl acceptor,�20%): T35L, L37V, A38P, and Q51P. However, M8 mutant,which is an 8-fold Dm3MaT(212) mutant containing P5S,H11Q, A49R, T72A, G81S, F84Y, F86S, and Y92F in region A,displayed a considerable level of 3MaT2 activity. These resultsindicate that Thr-35, Leu-37, Ala-38, and Gln-51 of Dm3MaT2are important for 3MaT2 activity. As for region C, each of theY411N, A413T, and K419N substitutions in Dm3MaT(212)resulted in a significant loss of 3MaT2 activity (relativekcat/Km for acyl acceptor, �20%), whereas the 14-foldmutant of Dm3MaT(212), mutant CM14, where all of the

amino acid residues unique toDm3MaT2 in domain C except forTyr-411, Ala-413, and Lys-419were replaced by those that wereunique to Dm3MaT1, displayed asignificant level of 3MaT2 activity(Table 1). These results indicate theimportance of Tyr-411, Ala-413,and Lys-419 for the 3MaT2 activityof Dm3MaT2. Considering theabove findings, we concluded thatThr-35, Leu-37, Ala-38, Gln-51,Tyr-411, Ala-413, and Lys-419 areimportant for 3MaT2 activity ofDm3MaT2. This conclusion wasfurther supported by the fact thatmutant M7, 7-fold mutant ofDm3MaT1 (G35T/V37L/P38A/P51Q/N410Y/T412A/N418K), dis-played a significant level of 3MaT2activity (Table 1).Phylogenetics, Specificity, and Bio-

chemical Properties of Dm3MaT3—During the course of our attemptsto find the structural basis of theabove conclusion, we identified anovel AAT homolog, Dm3MaT3,from red chrysanthemum petals,which unlike Dm3MaT1 andDm3MaT2 yields protein crystalssuitable for x-ray crystallographicstudies (for details of the isolationand expression of Dm3MaT3cDNA, see “Experimental Proce-dures”). The Dm3MaT3 cDNAcoded for a protein of 454 amino

acids with a calculated molecular weight of 50,805. Thededuced amino acid sequence of Dm3MaT3 (DDBJ accessionnumber, AB290338; Fig. 2) showed the highest sequence simi-larity to those of Dm3MaT1 (55% identity) and Dm3MaT2(53%) as well as the malonyl-CoA:anthocyanidin 3-O-gluco-side-6�-O-malonyltransferases of other Compositae plants,such as those of Senecio cruentus (Sc3MaT) (21) (54%) andDahlia variabilis (Dv3MaT) (22) (52%). It also shows a 40%sequence identity tomalonyl-CoA:anthocyanin 5-O-glucoside-6�-O-malonyltransferase of Salvia splendens petals (Ss5MaT1)(23). The primary structure of Dm3MaT1 contains sequencesthat are absolutely conserved among all members of the BAHDfamily: His-Xaa3-Asp (motif 1) and Asp-Phe-Gly-Trp-Gly(motif 3) (1, 2) (Fig. 2). It also contains the sequence that isuniquely conserved among the AATs: Tyr-Phe-Gly-Asn-Cys(motif 2) (2). As expected, a molecular phylogenetic analysisclearly indicated that Dm3MaT3 belongs to the AAT-relatedsubfamily and is most closely related to Compositae anthocya-nin malonyltransferases (see supplemental Fig. 2S).The Dm3MaT3 protein was expressed under the control of a

T5 promoter in E. coli BL21(DE3) cells, and the AAT activitiesof the expressed product were examined. The recombinant

FIGURE 3. Structures of Dm3MaT3 complexed with malonyl-CoA. A, ribbon diagram showing the tracing ofthe main chain of Dm3MaT3 in complex with malonyl-CoA. Helices and strands are shown as ribbons in greenand yellow, respectively, and are numbered from the N terminus, whereas loops are shown as thin lines. Boundmalonyl-CoA molecule is shown as sticks. B, a stereo view of Dm3MaT3 molecule complexed with malonyl-CoA.The locations of motifs 1, 2, and 3, each of which forms a part of a loop, is marked in red. C, model andnon-crystallography symmetry averaged 2Fo Fc electron density map of malonyl-CoA. The electron map(green) is contoured at 1�. Carbon, oxygen, nitrogen, phosphorus, and sulfur atoms are colored white, red, blue,orange, and yellow, respectively.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Dm3MaT3 showed 3MaT1 activity but no 3MaT2 activity (Fig.1); it catalyzed the transfer of a malonyl group from malonyl-CoA (Km 35 11�M) to cyanidin 3-O-glucoside (Km 470 100�M) to produce cyanidin 3-O-6�-O-malonylglucoside (kcat0.061 0.013 s1). Dm3MaT3 could also act on pelargonidin3-O-glucoside (kcat 0.022 0.002 s1; Km for malonyl-CoA,35 11 �M; Km for pelargonidin 3-O-glucoside, 120 12 �M)and delphinidin 3-O-glucoside (kcat, 0.021 0.001 s1; Km formalonyl-CoA, 24 2 �M; Km for delphinidin 3-O-glucoside,170 12 �M). However, flavonol 3-O-glucoside (quercetin3-O-glucoside), cyanidin 3-O-6�-O-malonylglucoside, pelar-gonidin 3,5-O-diglucoside, pelargonidin 3-O-6�-O-malonyl-glucoside 5-O-glucoside did not serve as acyl acceptors. For acyldonors, Dm3MaT3 was highly specific for malonyl-CoA. Suc-cinyl-CoA served as a very weak substrate (relative activity,5.1% of activity for malonyl-Co-A), whereas methylmalonyl-CoA, acetyl-CoA, p-coumaroyl-CoA, and caffeoyl-CoA wereinert as acyl donors. Thus, the observed acyl-donor specificityunambiguously showed that Dm3MaT3 is a malonyltrans-ferase. It should be noted in this respect that the AAT-relatedsubfamily of the BAHD family (see supplemental Fig. 2S) can bespecific for phenolics other than anthocyanins, as exemplifiedby a malonyltransferase of Nicotiana tabacum (24), and physi-ological malonyl acceptors of Dm3MaT3 in the chrysanthe-mum plant might not be anthocyanins, although it displayedAAT activities. The molecular weight of the native Dm3MaT3was estimated to be 60,000 by gel permeation high performanceliquid chromatography on Superdex 200 (GEHealthcare), indi-cating that Dm3MaT3 is monomeric. The recombinantenzyme showed a maximum activity at pH 7.5 and was stableover the pH range 4.5–8.0 (at 20 °C for 20 h) and at tempera-tures up to 40 °C (at pH 7.0 for 20 min).

Overall Structure of Dm3MaT3—We obtained three types ofDm3MaT3 crystals (native Dm3MaT3 (termed “native-free”),native Dm3MaT3 complexed with malonyl-CoA (termed“native-complex”), and selenomethionyl Dm3MaT3). Thestructure of Dm3MaT3 was solved by the multiwavelengthanomalous diffraction method using selenomethionylDm3MaT3 and was refined to a resolution of 2.2 Å (native-freeDm3MaT3), 2.1 Å (native-complex Dm3MaT3), and 1.8 Å (sel-enomethionyl Dm3MaT3). The models were refined to finalcrystallographic R/Rfree values of 19.4/24.8% (native-freeDm3MaT3), 19.6/23.4% (native-complex Dm3MaT3), and19.6/23.4% (selenomethionyl Dm3MaT3). The refinement sta-tistics and model quality parameters are listed in Table 2. Theasymmetric unit contains twomolecules in all types of the crys-tals. Although several hydrogen-bonding and hydrophobicinteractions between these two molecules in the asymmetricunit are observed, Dm3MaT3 ismost likely to bemonomeric insolution, judging from the gel filtration results (see above),which are consistent with other BAHD members that havebeen characterized (1).The crystal structures of Dm3MaT3 contain 17 �-strands

and 17�-helices, eachnumbered from theN terminus as shownin Fig. 3A (see also supplemental Fig. 3S). The �-strands formthree groups of �-sheets; i.e. group A (�1, �6, �9, �10, �3, and�15), groupB (�13,�12,�14,�16,�17,�11, and�7), and groupC (�2, �5, �4, end of �9, and �8). In the native-complex struc-ture of Dm3MaT3, the electron densities of malonyl-CoAwereidentified on one of two faces of the Dm3MaT3 molecule (Fig.3, A and C), and this face is referred to as the “front face.”Another face of the enzyme molecule, which is opposite to themalonyl-CoA-bound face, is referred to as the “back face” (Fig.4). Most of the �-sheets mentioned above are sandwiched

TABLE 13MaT1 and 3MaT2 activities of mutantsThe apparent kinetic parameters and their standard deviations were determined as described under Experimental Procedures. Values in parentheses indicate the relativepercentage of the kcat/Km value for 3MaT2 activity, with that of Dm3MaT(212) taken to be 100%.

Enzyme3MaT1 activity 3MaT2 activity

kcatkcat/Kmforacyl donor

kcat/Km foracyl acceptor kcat

kcat/Km foracyl donor

kcat/Km foracyl acceptor

s1 s1�M1 s1�M1 s1 s1mM1 s1mM1

Dm3MaT2 4.1 0.2 2.0 0.4 0.63 0.15 0.27 0.03 43 11 57 29 (150)aDm3MaT(221) 4.4 0.1 1.1 0.1 0.43 0.05 0.012 0.002 5.1 0.4 9.0 0.7 (24)Dm3MaT(212) 4.3 0.1 6.0 0.8 1.6 0.1 0.093 0.010 23 4 38 5 (100)Dm3MaT(122) 14 � 103 2 � 103 5.2 � 103 0.7 � 103 2.2 � 103 0.4 � 103 �0.001Dm3MaT(112) 3.8 0.1 1.8 0.2 1.2 0.2 �0.001Dm3MaT(121) 3.1 0.1 0.11 0.04 1.8 0.6 �0.001Dm3MaT(211) 2.7 0.1 0.83 0.08 0.40 0.05 �0.001Dm3MaT1 11 0.2 4.0 0.1 2.7 0.1 �0.001Site-specific mutants of Dm3MaT1(212)T35G 1.5 0.1 0.42 0.05 0.31 0.03 0.010 0.001 0.70 0.15 2.8 0.5 (7)L37V 3.9 0.1 1.1 0.3 1.9 0.1 0.027 0.003 5.2 1.3 4.7 0.9 (12)A38P 3.4 0.1 3.2 0.9 2.4 0.9 0.015 0.001 1.5 0.3 4.2 0.8 (11)Q51P 4.2 0.1 13 4 2.5 0.3 0.013 0.001 3.2 0.5 4.8 0.7 (12)M8a 3.2 0.1 3.6 0.4 1.5 0.4 0.041 0.001 7.1 0.6 15 1 (40)Y411N 1.2 0.1 0.70 0.09 0.67 0.16 �0.001A413T 2.3 0.1 3.1 0.8 2.9 1.1 0.015 0.002 22 6 2.6 0.4 (7)K419N 6.4 0.1 6.8 0.8 2.8 0.4 0.020 0.003 30 9 6.5 1.0 (17)CM14b 3.2 0.1 6.2 1.2 2.1 0.4 0.015 0.001 20 3 16 7 (41)

Seven-fold mutant of Dm3MaT1M7c 4.2 0.1 1.9 0.1 3.4 0.6 0.02 0.002 3.8 0.4 9.1 1.6 (24)

a Eight-fold mutant in region A; i.e. P5S/H11Q/A49R/T72A/G81S/F84Y/F86S/Y92F mutant of Dm3MaT1(212).b Fourteen-fold mutant in region B; i.e. I308C/S309T/V310A/N317D/L325F/I333V/T351A/S356G/E378D/G379A/P440S/K445E /F455L/V459I mutant of Dm3MaT(212).c G35T/V37L/P38A/P51Q/N410Y/T412A/N418K mutant of Dm3MaT1.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

between the faces. The front and back faces of the Dm3MaT3molecule are mainly composed of �-helices (�1, �2, �5, �6, �9,�10, �11, �12, �13, �15, �16, and �17 on the front face; �3, �4,�7, �8, and �14 on the back face) and several loop structures.

The overall structure of Dm3MaT3 is most similar to thatof vinorine synthase, a BAHD family enzyme, with a root

mean square deviation value of 3.6Å for 413 C� atoms with a 20%sequence identity (9). Structuraldifferences between Dm3MaT3and vinorine synthase are describedbelow in detail. The other structur-ally related proteins retrieved bymeans of the Dali server includepolyketide synthetase-associatedprotein 5 from Mycobacteriumtuberculosis with a root meansquare deviation value of 4.1 Å for388 C� atoms with a 14% sequenceidentity (25).Identification of Acyl-CoA Bind-

ing Site andActive-site Channel—Inthe native-complex structure ofDm3MaT3, well defined electrondensities of malonyl-CoA wereidentified in a pocket surrounded by�10, �11, �15, �12, and �14 on thefront face of the Dm3MaT3 mole-cule (Figs. 3,B andC, and 4A), whichforms the malonyl-CoA bindingsite. This acyl-CoA binding site isconnected to a channel that liesbetween �-sheet groups A and Band penetrates through theDm3MaT3 molecule. His-170 of

Dm3MaT3, which corresponds to an invariant residue in motif1 of the BAHD sequences, was identified in the middle of thechannel in close proximity to the thioester carbonyl carbon ofthe bound malonyl-CoA (Fig. 5).Our previous alanine-scanning mutagenesis studies of

Ss5MaT1 showed that the replacement of each of His and Asp

FIGURE 4. Surface diagram showing the front face (A) and the back face (B) of Dm3MaT3. The coordinatesshown as sticks are for bound malonyl-CoA. This figure was created with PyMOL program (18). The location inDm3MaT3 of amino acid residues that correspond to those important for the 3MaT2 activity of Dm3MaT2revealed by mutational studies are shown in green. C, a close-up view of the acyl-acceptor binding pocket.

TABLE 2Data collection and refinement statisticsNumbers in parentheses are for the highest shell.

Crystal type Native-free Native complexSeMet-Dm3MaT3

Peak Edge RemoteData collection and processing statisticsSpace group P21 P21 P21Unit cell dimension(Å)a (Å) 52.653 52.646 52.777b (Å) 122.980 122.922 122.908c (Å) 70.268 69.848 70.332� (°) 94.212 94.158 94.381

Wavelength (Å) 1.0000 1.0000 0.97909 0.97928 0.95000Resolution (Å) 50.00-2.20 (2.28-2.20) 35.49-2.10 (2.18-2.10) 70.18-1.80 (1.86-1.80) 70.18-1.90 (1.97-1.90) 70.18-1.90 (1.97-1.90)Measured 158,660 202,428 640,126 292,976 293,469Mean I/�I 18.5 (3.6) 19.7 (3.9) 31.0 (5.4) 26.8 (5.8) 27.4 (4.9)Redundancy 3.9 (3.0) 4.1 (3.0) 7.9 (6.8) 4.2 (4.1) 4.2 (4.0)Completeness (%) 90.2 (59) 95.8 (74.1) 98.6 (93.0) 99.0 (97.3) 99.0 (96.6)Rmerge

a (%) 7.3 (22) 7.0 (21) 6.4 (27.6) 5.0 (20.3) 4.9 (22.5)Refinement statisticsResolution 33.88-2.20 50.00-2.10 30.46-1.80Protein atoms 6932 6932 6932Ligand atoms 108Water molecule 697 463 946Rwork/Rfree(%) 19.4/24.8 19.6/23.4 19.6/23.4

Root mean square deviationsBond lengths (Å) 0.012 0.011 0.009Bond angles (°) 1.348 1.401 1.104

aRmerge � 100��I �I��/�I, where I is the observed intensity, and �I� is the average intensity of multiple observations of symmetry-related reflections.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

residues of motifs 1 and 3, respectively, causes the nearly com-plete loss of catalytic activity, suggesting that one of these resi-dues acts as a catalytic base in the general acid/basemechanismof AAT catalysis (6). The crystal structures of Dm3MaT3 showthat the Asp residue in motif 3 (Asp-396) is too far from thebound substrate to interact with it (Fig. 3B). Thus, the His res-idue of motif 1 most likely fulfills the role of the general basecatalyst, and the channel forms the active site of the enzyme.Interaction of Dm3MaT3 with Bound Malonyl-CoA—The

native-complex structure of Dm3MaT3 provides the first crys-tal structure of a BAHD member complexed with acyl-CoA,permitting detailed information relative to interactionsbetween the enzyme and acyl-donor substrate to be elucidated.In the native-complex structure, the malonyl-CoA moleculeinteracts with His-170, Ala-175, Arg-178, Lys-260, Glu-271,Tyr-272, Val-273, Ser-274, Ser-275, Phe-276, Pro-301, Ile-302,Asp-303, Arg-307, Gly-386, Thr-387, and Lys-389 ofDm3MaT3 (Fig. 5), where the underlined residues are thosethat interact with the malonyl group, with the others interact-ing with the CoA moiety of the acyl-donor molecule.Fourteen hydrogen-bonding interactions between the

Dm3MaT3 protein and the CoA portion of malonyl-CoA wereidentified, five of which are mediated by main-chain atoms ofDm3MaT3 (Fig. 5). A sequence comparison (see supplementalFig. 3S) revealed that the amino acid residues involved in CoAbinding are not strictly conserved among members of theBAHD family, even among AATs, although the folds of theacyl-CoA binding site of the BAHD enzymes appear to be sim-

ilar to each other (see below).Therefore, the ability of BAHDenzymes to bind an acyl-CoA moi-etymight have evolvedwith the spe-cific spatial arrangement of polargroups being conserved, as theresult of the specific folding of thepolypeptide chain (see below)rather than from the conservationof specific amino acid residues.Concerning interactions with the

malonyl portion of the boundmalo-nyl-CoA, the thioester carbonyloxygen of malonyl-CoA is hydro-gen-bonded to His-170 of motif 1,which in turn is hydrogen-bondedwith the side chain of Asn-319 ofmotif 2 (not shown). The side chainof Arg-178 and the main-chain Natom of Gly-386 are involved inhydrogen-bonding interactions withthe carboxyl oxygen atom of a mal-onyl group (Fig. 5), thereby fixingthe terminal carboxyl group of mal-onyl-CoA in the complex. Arg-178and Gly-386 are also conserved inDm3MaT1, Dm3MaT2, and otheranthocyanin malonyltransferasesthat are specific for anthocyanidin3-O-glucosides. However, Arg-178

is not conserved in anthocyanin malonyltransferases that arespecific for anthocyanin 5-O-glucosides such as those from S.splendens (Ss5MaT1 (23) and Ss5MaT2 (26)) (see supplementalFig. 3S). It is noteworthy here that the molecular phylogenetictree of AATs (supplemental Fig. 2S) implies that the acyl-donorspecificities of AATs are diversified after the divergence ofplant species. Dm3MaT1, Dm3MaT2, and Dm3MaT3 are allenzymes of Compositae plants, whereas Ss5MaT1 andSs5MaT2 are produced by labiate plants. Thus, the conserva-tion of Arg-178 among the three Dm3MaT enzymes may beexplained in phylogenetic terms rather than functional terms.Thus, the specificity of BAHD enzymes for the acyl portion ofacyl-CoAmay also be due to the specific spatial arrangement offunctional groups rather than the conservation of specificamino acid residues, similar to the case for the specificity for theCoA portion of acyl-CoA (see above).Identification of Acyl-acceptor Binding Site—The malonyl-

CoAbinding site connects via the active-site channel to anotherpocket (Figs. 3B and 4B), which is located on the back face of theDm3MaT3molecule. This pocket is composed of Phe-35, Trp-36, Leu-37, Pro-40, Ile-42, Asn-44, Arg-178, Pro-301, Trp-383,Ser-385, Thr-407, Ser-409, His-412, and Asn-413 and appearsto be sufficiently wide to accommodate an acyl-acceptor(anthocyanin) molecule. A sequence comparison (Fig. 2) showsthat the amino acid residues (Thr-35, Leu-37, Ala-38, Gln-51,Tyr-411, Ala-413, and Lys-419 of Dm3MaT2) that are impor-tant for the ability of Dm3MaT2 to act on cyanidin 3-O-6�-O-malonylglucoside correspond to Phe-35, Leu-37, Arg-38, Pro-

FIGURE 5. Close-up view of the malonyl-CoA binding pocket in the native-complex structure ofDm3MaT3. Bound malonyl-CoA is shown as thick sticks, whereas the amino acid residues of Dm3MaT3 thatinteract with bound malonyl-CoA are shown as thin sticks. Carbon, oxygen, nitrogen, phosphorus, and sulfuratoms are colored gray, red, blue, salmon, and yellow, respectively. Dotted lines indicate hydrogen bonds. Aminoacid residues without dotted lines form hydrophobic interactions with bound malonyl-CoA. Amino acid resi-dues whose main chain atoms form hydrogen bonds with the malonyl-CoA molecule are underlined.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

51, Leu-405, Thr-407, and Asn-413 of Dm3MaT3, all of which,except for Pro-51, are among the amino acid residues that formthe pocket at the back face of the Dm3MaT3 molecule (Fig. 4).The inner part of the pocket is accessible to themalonyl moietyof the boundmalonyl-CoA through the active-site channel (Fig.4, B and C). All of these observations led to the conclusion thatthe pocket on the back face of theDm3MaT3molecule serves asthe acyl-acceptor binding site of Dm3MaT3, unambiguouslyestablishing the spatial arrangement of acyl-CoA and acyl-ac-ceptor binding sites in the enzyme molecule, both of which areconnected by an active-site channel. Such a spatial arrange-ment of substrate binding sites and active site was previouslyproposed for the vinorine synthase structure (9) and is similarto those of several other acyl-CoA-dependent acyltransferases,such as Azotobacter vinelandii dihydrolipoyl transacetylase(Ref. 27; Protein Data Bank (PDB) codes 1EAD and 1EAB) andmouse carnitine acetyltransferase (Ref. 28; PDB codes 1NDBand 1NDI), which show essentially no sequence similarity toDm3MaT3.Comparison with Vinorine Synthase Structure—The �-sheet

groups C and B of Dm3MaT3 consist of five and seven�-strands, respectively, whereas those of vinorine synthasecontains only two and six short �-strands, respectively. Thecrystal structures of Dm3MaT3 (even the 1.8-Å resolutionstructure of selenomethionyl Dm3MaT3) contain three disor-dered regions (N-terminal residues Met-1–Pro-5, residuesAsp-365–Phe-369, andC-terminal residues Lys-451–Leu-454),whereas that of vinorine synthase contains two disorderedregions (N-terminal residues Met-1–Pro-3 and residuesSer-235–Glu-239).The superposition of crystal structures of Dm3MaT3 and

vinorine synthase reveals that, whereas the structures of thefront faces of these enzymes are similar to each other, the struc-tures of the back faces differ significantly (Fig. 6). Namely, thearrangements of secondary structures on the front face of bothenzymes can be well superposed, showing similar structuresaround the acyl-CoA binding pockets in both enzymes (Fig.6A). By contrast, the spatial arrangement of secondary struc-tures (�-helices and loops) on the back face of Dm3MaT3 can-

not be superimposed over that ofvinorine synthase; hence, the struc-tures around the acyl-acceptorbinding sites of both enzymes aredifferent. These observationspermitted the unique structuralcharacteristics of the BAHD familyconcerning the specificities for theacyl-donor and acceptor substratesto be rationalized. The commonability of BAHD members to utilizeacyl-CoAs appears to bemaintainedthrough the structural conservationof the front face of the enzyme mol-ecules. However, the fact that theBAHD family displays versatilitywith respect to acyl-acceptor sub-strates should arise from the differ-ential architectures of the acyl-ac-

ceptor binding sites, which are constructed with differentialspatial arrangements of secondary structures on the back faceof the enzymemolecules. BAHD family enzymes with differentbiochemical functions show a only low sequence similarity toeach other (�20% identity), ensuring such differential architec-tures for the acyl-acceptor binding site.Induced Fit of Dm3MaT3 upon Malonyl-CoA Binding—A

comparison of native-free and native-complex structuresreveals that the overall structures of malonyl-CoA-bound and-unbound forms of Dm3MaT3 are essentially identical witheach other, showing a rootmean square deviation value of 0.327Å for 440 C� atoms. However, these structures show somemovements of the backbone (C� atoms of Pro-301–Ile-302 andGly-386–Lys-389) and side-chain positions (Arg-307 and Lys-389) around the malonyl-CoA binding site upon malonyl-CoAbinding. The backbone of Pro-301–Ile-302 and Gly-386–Lys-389 is shifted outward by more than 0.5 Å, with a maximumshift of the C� atom of Thr-387 by 1.18 Å, thus making it pos-sible to accommodate a malonyl-CoA molecule in the pocket.Moreover, upon malonyl-CoA binding, the side chains of Arg-307 and Lys-389 rotate and form hydrogen bonds with the CoAportion of malonyl-CoA. For comparison, the two types ofligand-free crystal structures (i.e. structures of native-free andselenomethionyl Dm3MaT3s) could be regarded as essentiallyidentical with each other (root mean square deviation of 0.178Å for 440 C� atoms), where the backbone shifts around resi-dues Pro-301—Ile-302 and Gly-386–Lys-389 are only less than0.3 Å. The observed deviations around malonyl-CoA bindingsite upon binding of malonyl-CoA can be consistentlyexplained in terms of an induced fit of theDm3MaT3molecule.Previous kinetic studies of malonyl transfer catalyzed by

Ss5MaT1 suggested that the kinetic mechanism could be mostconsistently described in terms of a ternary-complex mecha-nism, where malonyl-CoA was the first substrate to bind to theenzyme, and CoA-SH was the last product to dissociate fromthe enzyme-product complex (6). This predicted order of sub-strate binding and product release prompted us to examinewhether the binding of malonyl-CoA entails any conforma-tional changes in the acyl-acceptor binding site in the

FIGURE 6. Superposition of Dm3MaT3 and vinorine synthase (ribbon diagram). A, front face of enzymemolecules showing the acyl-CoA binding site. B, back face of enzyme molecules showing the acyl-acceptorbinding site. The two structures are superimposed as described under “Experimental Procedures.” The mainchain tracing of Dm3MaT3 and vinorine synthase is shown in green and magenta, respectively. Malonyl-CoAmolecule that is bound to Dm3MaT3 is shown as sticks. Dm3MaT3 structures shown in A and B of this figurecorrespond to those shown in A and B, respectively, of Fig. 4.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Dm3MaT3 molecule to allow for the subsequent binding of anacyl acceptor (anthocyanin). However, the main-chain andside-chain conformations of the acyl-acceptor binding sites ofthe native-complex structure were essentially identical to thoseof the native-free structure (shift of all atoms of amino acidresidues forming acyl-acceptor binding sites, less than 0.3 Å).Thus, sequential binding for the malonyl-CoA and acyl accep-tor appears to be less likely in the case of Dm3MaT3 catalysis.Specificity and Mechanism in AAT Catalysis—The results

obtained in the present studies provide a structural basis forgeneral insights into the specificity and mechanism of BAHDcatalysis. In the discussion below the amino acid residues ofBAHD enzymes are numbered so as to correspond to the num-bering of Dm3MaT3 based on the alignment shown in Fig. 2(see also supplemental Fig. 3S), and the positional numbering,according to this notation, is indicated by an asterisk.Thus far AATs have been strictly classified into two distinct

categories on the basis of their acyl-donor specificity, i.e., ali-phatic AATs (e.g., Dm3MaT1, Dm3MaT2, Dm3MaT3, andSs5MaT1) and aromatic AATs (e.g. hydroxycinnamoyl-CoA:anthocyanin 3,5-O-diglucoside 6�-O-hydroxycinnamoyltrans-ferase, Gt5AT, which is involved in the biosynthesis of gentio-delphin in the blue flowers of gentian (Gentiana triflora)(29), and hydroxycinnamoyl-CoA:anthocyanin 3-O-glucoside6�-O-hydroxycinnamoyltransferase, Pf3AT, which is involvedin the biosynthesis of shisonin in the red forms of Perilla frute-scens (30)). Aliphatic acyltransferases specifically utilize ali-phatic acyl-CoAs (such as malonyl-CoA) but do not act onaromatic acyl-CoAs (such as p-coumaroyl-CoA and caffeoyl-CoA); likewise, aromatic AATs utilize aromatic acyl-CoAsexclusively but do not act on aliphatic acyl-CoAs. The native-complex structure of Dm3MaT3 revealed that Arg-178 andGly-386 are involved in binding the malonyl group of malonyl-

CoA. A sequence comparison (seesupplemental Fig. 3S) revealed thatArg-178 ofDm3MaT3 is specificallyreplaced by Phe in both Gt5AT andPf3AT sequences. Phe-178* (i.e.Phe-182 of Gt5AT and Phe-170 ofPf3AT) of these aromatic AATswould be expected to interact withthe aromatic ring moiety of a boundhydroxycinnamoyl-CoA, and noother amino acid residues near acyl-CoA binding site would be pre-dicted to be capable of entering intoaromatic interactions with a boundacyl-CoA. Thus, Phe178* is likely tobe a key residue that governs thepreference of Gt5AT and Pf3AT foraromatic acyl-CoA.In Dm3MaT1 and Dm3MaT2

catalysis, the acyl-acceptor bindingsite, containing a set of Dm3MaT1-type amino acid residues (Gly-35*,Val-37*, Pro-38*, Pro-51*, Asn-405*,Thr-407*, and Asn-413*), wouldonly allow the binding of cyanidin

3-O-glucoside (Fig. 7), whereas that containing a set ofDm3MaT2-type residues (Thr-35*, Leu-37*, Ala-38*, Tyr-405*,Ala-407*, and Lys-413*) would allow the productive bindingof the mono-malonylated product (cyanidin 3-O-6�-O-ma-lonylglucoside) in addition to that of the non-malonylatedsubstrate (cyanidin 3-O-glucoside) (Fig. 7). Although detailsof the nature of anthocyanin binding in the pocket remains tobe clarified, Thr-407* of Dm3MaT1 should be located near thedead end of the anthocyanin binding pocket (Fig. 4C), and thereplacement of Thr-407* by alanine likely provides a widerspace in the pocket, making some contribution to the produc-tive binding of the mono-malonylated product in Dm3MaT2(Fig. 7). Gln-51* of Dm3MaT2 is remote from the pocket (Fig.4A), so that it is unlikely that this residue is directly involved ininteractions with a bound anthocyanin. Rather, this residuemay play a structural role in maintaining the conformation ofthe enzyme in such a way that affects the shape of the acyl-acceptor binding site. After the formation of a ternary complexconsisting of malonyl-CoA, anthocyanin, and enzyme, thehydroxy group of anthocyanin to be activated would be placedin the active-site channel near the thioester carbonyl carbon ofthe bound malonyl-CoA and His-170*, which then abstracts aproton from the hydroxy group, triggering an acyl transfer reac-tion (6). It should be noted that Trp-36 of Dm3MaT3 is locatedat the anthocyanin binding pocket and is conserved in all otherAATs (see supplemental Fig. 3S), implying that this aromaticresidue may be crucial for the acyl acceptor specificity of theanthocyanins. This residue would be a good target in amutagenesis study to alter the acyl acceptor specificity of aBAHD family enzyme, and such work is currently under way.In addition to His-170*, several amino acid residues have

been shown to be functionally important for AAT catalysis asthe result of alanine-scanningmutagenesis studies of Ss5MaT1

FIGURE 7. Schematic representation of the substrate binding sites and active-site channel in ternarycomplexes of Dm3MaT1 (A) and Dm3MaT2 (B). Enzyme molecules are shown in gray. The structure of thebound anthocyanin is shown as a quinonoidal base structure.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

(6), i.e. Asp-396*, Tyr-48*, Asp-174*, Arg-305*, and Asn-319*.Asp-396* (in motif 3) is too far from the bound substrate tointeract with it (Fig. 3), so that the Asp residue in motif 3 couldplay a structural role in enzyme catalysis. Asn-319* (in motif 2)is hydrogen-bondedwithHis-170* and is thereby likely to affectthe role of the catalytic His-170*. Tyr-48* is distant from theactive-site channel and substrate binding sites and probablyplays an important role in maintaining the catalytically activestructure of the enzyme. Asp-174* (in motif 1) forms a hydro-gen bond with Arg-305* that occurs near the malonyl-CoAbinding pocket (not shown), so that this interaction may beimportant for the efficient binding of malonyl-CoA.Conclusion—The crystal structures of Dm3MaT3 along with

the results of mutagenesis studies, intended to explore key res-idues responsible for differential acyl-acceptor specificitybetween Dm3MaT1 and Dm3MaT2, allowed us to identify theacyl-CoA and acyl-acceptor binding sites in AATs. This findingprovides concrete evidence for the spatial arrangement of acyl-donor and acyl-acceptor binding sites connected by an active-site channel in BAHD enzymes. The front faces of BAHDenzyme molecules, which contain the acyl-CoA binding site,are structurally similar to each other, rationalizing the commonability of BAHD enzymes to utilize acyl-CoA. By contrast, thestructures around the acyl-acceptor binding site on the backface of BAHD enzymes can be different from each other, thusexplaining the diversity of acyl-acceptor specificities of theBAHD family. These findings establish the structural basis forgeneral insights into the specificity and mechanism of BAHDcatalysis.

Acknowledgments—We thankDrs. N. Igarashi andN.Matsugaki andthe technical staff at synchrotron beamline BL-5A of the Photon Fac-tory, High Energy Accelerator Research Organization, Japan, forassistance with the data collection.

REFERENCES1. D’Auria, J. C. (2006) Curr. Opin. Plant Biol. 9, 331–3402. Nakayama, T., Suzuki, H., and Nishino, T. (2003) J. Mol. Catal. B Enzym.

23, 117–1323. St-Pierre, B., and De Luca, V. (2000) in Evolution of Metabolic Pathways,

Recent Advances in Phytochemistry, (Romeo, J. T., Ibrahim, R., Varin, L.,and De Luca, V., eds) Vol. 34. pp. 285–315, Elsevier Science Ltd., Oxford

4. St-Pierre, B., Laflamme, P., Alarco, A.-M., and De Luca, V. (1998) Plant J.14, 703–713

5. Strack, D., and Wray, V. (1994) in The Flavonoids (Harbone, J. B., ed) pp.1–22, Chapman & Hall, Washington, D. C.

6. Suzuki, H., Nakayama, T., and Nishino, T. (2003) Biochemistry 42,1764–1771

7. Ino, I., Nishiyama, H., and Yamaguchi, M.-A. (1993) Phytochemistry 32,1425–1426

8. Suzuki, H., Nakayama, T., Yamaguchi, M., and Nishino, T. (2004) PlantSci. 166, 89–96

9. Ma, X., Koepke, J., Panjikar, S., Fritzsch, G., and Stockigt, J. (2005) J. Biol.Chem. 280, 13576–13583

10. Segel, I. H. (1975) Enzyme Kinetics: Behavior, and Analysis of Rapid Equi-librium and Steady-state Enzyme Systems, pp. 623–625, John Wiley &Sons, Inc., New York

11. Otwinowski, Z., and Minor, W. (1997)Methods Enzymol. 276, 307–32612. Terwilliger, T. C., and Berendzen, J. (1999) Acta Crystallogr. Sect. D. 55,

849–86113. Terwilliger, T. C. (2003) Acta Crystallogr. Sect. D. 59, 38–4414. Perrakis, A., Morris, R., and Lamzin, V. S. (1999) Nat. Struct. Biol. 6,

458–46315. Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta

Crystallogr. Sect. A. 47, 110–11916. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997)Acta Crystallogr.

Sect. D. 53, 240–25517. Fenn, T. D., Ringe, D., and Petsko, G. A. (2003) J. Appl. Crystallogr. 36,

944–94718. DeLano, W. L. (2002) The PyMOL Molecular Graphics Systems, DeLano

Scientific, San Carlos, CA19. Merritt, E. A., and Murphy, M. E. (1994) Acta Crystallogr. Sect. D Biol.

Crystallogr. 50, 869–87320. Kabsch, W. (1976) Acta Crystallogr. Sect. A. 32, 922–92321. Suzuki, H., Sawada, S., Yonekura-Sakakibara, K., Nakayama, T., Yamagu-

chi, M.-A., and Nishino, T. (2003) Plant Biotech. 20, 229–23422. Suzuki, H., Nakayama, T., Yonekura-Sakakibara, K., Fukui, Y., Nakamura,

N., Yamaguchi, M.-A., Tanaka, Y., Kusumi, T., and Nishino, T. (2002)Plant Physiol. 130, 2142–2151

23. Suzuki, H., Nakayama, T., Yonekura-Sakakibara, K., Fukui, Y., Nakamura,N., Nakao,M., Tanaka, Y., Yamaguchi,M.-A., Kusumi, T., andNishino, T.(2001) J. Biol. Chem. 276, 49013–49019

24. Taguchi, G., Shitchi, Y., Shirasawa, S., Yamamoto, H., and Hayashida, N.(2005) Plant J. 42, 481–491

25. Buglino, J., Onwueme, K. C., Ferreras, J. A., Quadri, L. E., and Lima, C. D.(2004) J. Biol. Chem. 279, 30634–30642

26. Suzuki, H., Sawada, S., Watanabe, K., Nagae, S., Yamaguchi, M. A., Na-kayama, T., and Nishino, T. (2004) Plant J. 38, 994–1003

27. Mattevi, A., Obmolova, G., Schulze, E., Kalk, K. H., Westphal, A. H., deKok, A., and Hol, W. G. (1992) Science 255, 1544–1550

28. Jogl, G., and Tong, L. (2003) Cell 112, 113–12229. Fujiwara, H., Tanaka, Y., Yonekura-Sakakibara, K., Fukuchi-Mizutani, M.,

Nakao, M., Fukui, Y., Yamaguchi, M., Ashikari, T., and Kusumi, T. (1998)Plant J. 16, 421–431

30. Yonekura-Sakakibara, K., Tanaka, Y., Fukuchi-Mizutani, M., Fujiwara, H.,Fukui, Y., Ashikari, T., Murakami, Y., Yamaguchi, M.-A., and Kusumi, T.(2000) Plant Cell Physiol. 41, 495–502




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Tanaka, Atsushi Saito, Tokuzo Nishino, Masami Kusunoki and Toru NakayamaHideaki Unno, Fumiko Ichimaida, Hirokazu Suzuki, Seiji Takahashi, Yoshikazu

the Features of BAHD Enzyme CatalysisStructural and Mutational Studies of Anthocyanin Malonyltransferases Establish

doi: 10.1074/jbc.M700638200 originally published online March 23, 20072007, 282:15812-15822.J. Biol. Chem.

10.1074/jbc.M700638200Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

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

Supplemental material:

http://www.jbc.org/content/suppl/2007/03/27/M700638200.DC1

http://www.jbc.org/content/282/21/15812.full.html#ref-list-1

This article cites 27 references, 5 of which can be accessed free at


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M700638200

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;282/21/15812&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/282/21/15812

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=282/21/15812&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/282/21/15812

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/suppl/2007/03/27/M700638200.DC1

http://www.jbc.org/content/282/21/15812.full.html#ref-list-1

http://www.jbc.org/

structural and mutational studies of anthocyanin

Documents