Unusual arrangement of catalytic domains in head-to-tail
associated homodimer of 6-hydroxymellein synthase, a
multifunctional polyketide biosynthetic enzyme
Fumiya Kurosaki *, Kousuke Togashi, Munehisa Arisawa
Faculty of Pharmaceutical Sciences,Toyama Medical and Pharmaceutical Uni6ersity,Sugitani,Toyama930-0194,Japan Received 3 March 2000; received in revised form 7 April 2000; accepted 28 April 2000
Abstract
6-Hydroxymellein synthase, a multifunctional polyketide synthetic enzyme in carrot, is organized as a homodimer, and the activity of the synthase was appreciably inhibited upon the specific alkylation of cysteine- and cysteamine-SHs at the reaction center with iodoacetoamide and chloroacetyl-CoA, respectively. Dissociation and stoichiometric recombination of the unmodified and the SH-modified enzyme subunits yielded a combination of unmodified – unmodified, unmodified – modified and modified – modified hybrid dimers that together exhibit 50% activity. In contrast, hybrid dimers obtained by reconstruction of the two modified enzymes showed essentially no catalytic activity. These results suggest that the two subunits of 6-hydroxymellein synthase are aligned in head-to-tail orientation to organize two reaction centers which are comprised of a cysteine and a complementary cysteamine SH group, belonging to and contributed from the same subunit in the homodimer structure. © 2000 Elsevier Science Ireland Ltd. All rights reserved.
Keywords:Polyketide biosynthetic enzyme; Multifunctional enzyme;Daucus carota; Reaction center
www.elsevier.com/locate/plantsci
1. Introduction
6-Hydroxymellein (6HM) synthase is a multi-functional polyketide biosynthetic enzyme induced in carrot cells upon microbial invasion [1,2]. The synthase catalyzes the condensation of 1 mol acetyl-CoA and 4 mol malonyl-CoA, and an NADPH-dependent ketoreduction of the carbonyl group of the top C2 unit takes place at the
en-zyme-bound triketide intermediate stage to form a dihydroisocoumarin skeleton (Fig. 1) [1]. We re-ported previously [3] that the active form of 6HM synthase is organized as a homodimer. However, the subunits of multifunctional property (approxi-mately 130 kDa each) are readily dissociated to monomeric form under high ionic strength condi-tions. Recently, we found [4] that the monomeric
synthase retains the ability of catalyzing the acyl condensation but is lacking in the ketoreducing activity, and triacetic acid lactone (TAL) is liber-ated from the monomer enzyme instead of 6HM (Fig. 1). These results led us to assume that two subunits of 6HM synthase are aligned in a head-to-tail direction, and the functional domain for the ketoreduction associates with that for the acyl condensation of another subunit to form two reac-tion centers in each molecule of the enzyme. It appears [4] that two SH groups at the reaction center, Cys-SH of the condensation enzyme and 4%-phosphopantetheine-SH attached to acyl carrier
protein (ACP), belong to the same subunit since the synthase does not lose the activity of the acyl condensation even in the monomeric form. It is well known [5 – 7] that fatty acid synthases (FASs) share many common properties with polyketide biosynthetic enzymes, and the overall organization of 6HM synthase is similar to multifunctional
* Corresponding author. Tel.: +81-76-4342281; fax: + 81-76-4345052.
E-mail address:[email protected] (F. Kurosaki).
FAS in animal cells. However, in sharp contrast to 6HM synthase, it has been demonstrated that Cys-SH and ACP-SH at the reaction center of animal FAS are contributed from the other sub-units of the homodimer, respectively [5 – 7]. In the present study, the two SH groups at the reaction center of 6HM synthase were selectively alkylated, and the hybrid dimers composed of these chemi-cally modified subunits were constructed. Catalytic activities of these 6HM synthase variants were determined to confirm the organization and the arrangement of catalytic domains at the reaction center of the enzyme.
2. Materials and methods
2.1. Chemicals
6HM was prepared by demethylating 6-methoxymellein isolated from fungi-infected carrot roots with BBr3 in anhydrous CH2Cl2 as reported
previously in detail [1,8]. Chloroacetyl-CoA (ClAcCoA) was synthesized from chloroacetic acid and CoA (Sigma) according to the method of Kawaguchi et al. [9], and TAL was synthesized from dehydroacetic acid (Nacalai Tesque) as de-scribed previously [1]. 2-Chloroethylphosphonic acid, acetyl-CoA, malonyl-CoA and NADPH were purchased from Sigma, while iodoacetoamide (IoAA) and dithiothreitol (DTT) were from Wako Pure Chemicals. [2-14C]Malonyl-CoA (specific
ac-tivity, 2.2 GBq/mmol) was from New England Nuclear. All other chemicals were reagent grade.
2.2. Induction, purification and assay of 6HM synthase
6HM synthase was induced in carrot cells by
treatment of the root tissues with
2-chloroethylphosphonic acid [1,2], and the synthase protein was highly purified according to the meth-ods described previously [10]. The enzyme solution was appropriately concentrated by the ultrafiltra-tion on an Amicon cell (YM-10 membrane), and protein concentrations were determined by the method of Bradford [11]. The purity of the syn-thase in the preparation was determined by
densit-ometric scan on a dual wavelength
chromatoscanner after separation by sodium do-decyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (8% gel)) according to the method of Laemmli [12], and the results were as reported previously [8,10]. The standard assay mixture of the synthase activity consisted of 10 mM K-phos-phate (pH 7.0), 100 mM acetyl-CoA, 50 mM [2-14C]malonyl-CoA (3.7 kBq), 2 – 5 pkat of the
enzyme preparation, 2 mM DTT and 1 mM NADPH in a total volume of 100 ml. The mixture
was incubated at 37°C for 30 min, and the reac-tion was terminated by the addireac-tion of 50 ml of
50% (v/v) acetic acid. The products were extracted with 200 ml ethyl acetate by blending, and 50 ml
aliquots were applied onto a silica gel thin-layer chromatography plate. After the development, ra-dioactivities co-migrated with the authentic 6HM or TAL were determined as reported previously in detail [1,8,10].
2.3. Chemical modification of 6HM synthase
Two SH groups at the reaction center of 6HM synthase were blocked by alkylation with IoAA and ClAcCoA according to the method described previously in detail [13]. In brief, DTT in the synthase preparation was removed by repeated dialysis, and Cys-SH of the condensation enzyme was alkylated by the incubation of the enzyme protein with 5 mM IoAA in 10 mM K-phosphate buffer (pH 7.0) at 37°C for 15 min. For the alkylation of ACP-SH, 1 mM ClAcCoA was em-ployed instead of IoAA. After the alkylation reac-tions, DTT (final concentration, 7 mM) was added to the mixtures to quench the excess SH inhibitors, and the sample solutions were dialyzed against the K-phosphate buffer containing 5 mM DTT (pH 7.0) to remove these reagents.
2.4. Preparation of hybrid dimers of 6HM synthase
To 2.5 ml aliquots of the unmodified, and IoAA- and ClAcCoA-modified enzyme solutions (approximately 5 pkat equivalent in 10 mM K-phosphate buffer containing 2 mM DTT, pH 7.0), an equal volume of the phosphate buffer con-taining 4 M NaCl was added, and the mixtures were incubated at 4°C for 2 h, respectively. As reported previously, the homodimeric form of 6HM synthase is quantitatively dissociated to the monomer subunits in these conditions [3,4]. These three monomer subunit solutions were mixed in a 1:1 ratio (600 ml each) in various
combin-ations (unmodified – unmodified, unmodified – IoAA modified, unmodified – ClAcCoA modified, and IoAA modified – ClAcCoA modified), and re-construction of these unmodified and modified monomers was initiated by dialysis of the mixtures against 10 mM K-phosphate buffer (pH 7.0) con-taining 2 mM DTT at 4°C [3,4]. After the recom-bination of the monomer subunits by desalting, the volume of the hybrid solutions was adjusted to 1.5 ml, and the activities of 6HM synthase in these preparations were determined.
2.5. Gel-filtration analysis of hybridized 6HM synthase
The equal amounts of IoAA- and ClAcCoA-modified monomer subunits (0.5 mg protein each)
were combined, and the mixture was applied to a Toyopearl HW-55 column (Tosoh, 1.6×93 cm) that had been previously calibrated with gel-filtra-tion standard proteins (Bio-Rad). The column was eluted with 10 mM K-phosphate buffer (pH 7.0) containing 2 M DTT and 2 mM NaCl, and frac-tions (2.5 ml each) corresponding to the dimeric and monomeric forms of 6HM synthase were combined, respectively [3]. The samples were then subjected to SDS-PAGE with standard proteins (Bio-Rad), and the distribution of the modified subunits in these two fractions was analyzed by staining with Coomassie brilliant blue. In a paral-lel experiment, hybrid dimers were prepared from the mixture of IoAA- and ClAcCoA-modified monomer subunits by desalting as already de-scribed, and the distribution of the reconstituted enzymes in the dimer and monomer fractions was similarly analyzed. In this case, however, the NaCl-omitted buffer was employed for the gel-filtration chromatography.
3. Results
3.1. Hybrid formation of chemically modified 6HM synthase
Fig. 2. Distribution of 6HM synthase subunits in dimeric and monomeric forms after the alkylation of SH groups. IoAA-and ClAcCoA-modified 6HM synthase proteins were incu-bated in the presence of 2 M NaCl, and the mixture of the modified enzymes was applied on a Toyopearl HW-55 column. Proteins involved in the fractions corresponding to the dimer (D) and the monomer (M) forms of 6HM synthase were analyzed by SDS-PAGE with standard proteins (left panel). In a parallel experiment, a NaCl-treated mixture of the modified enzyme was desalted prior to the fractionation by the gel-filtration chromatography, and proteins in the dimer (D) and monomer (M) fractions were similarly ana-lyzed by SDS-PAGE (right panel). An arrow indicates the position of the 6HM synthase subunit.
band was not detected in the dimer fraction under the present experimental conditions (Fig. 2, left panel). In contrast, when the mixture of the alky-lated 6HM synthase variants was fractionated af-ter desalting, the protein band of the subunits was found only in the dimer fractions (Fig. 2, right panel). This set of results strongly suggests that, even after the alkylation of Cys- and ACP-SHs, 6HM synthase subunits are quantitatively dissoci-ated to the monomers under high ionic strength conditions, and the modified proteins are capable of forming the hybrid dimers in response to the decrease in the salt concentration of the buffer, as is the unmodified enzyme.
3.2. Hybrids of unmodified and chemically modified 6HM synthase
The activity of 6HM synthase markedly de-creased by treatment with either IoAA or ClAc-CoA, and the residual activities immediately after the alkylation were 9 – 13% for IoAA and 7 – 11% for ClAcCoA, respectively. In our previous experi-ments, it was confirmed [13] that, even after the alkylation with IoAA or ClAcCoA, Ser-OH in the transacylase domain and the unmodified SH group, Cys-SH or ACP-SH, at the reaction center of 6HM synthase retained either the binding activ-ities toward the substrates or the channeling abili-ties of the acyl groups from the primary binding site to the SH groups. These observations suggest that ACP-SH in IoAA-treated enzyme and Cys-SH in ClAcCoA-treated enzyme do not lose the essential functions for substrate entry and for its channeling under the present experimental condi-tions. Dissociation and reconstitution of IoAA-modified 6HM synthase in the absence of other synthases yielded inactive modified homodimer en-zyme, and the activity regain was 11 – 14% of the control, reconstructed homodimer from the disso-ciated unmodified subunits, in independent experi-ments (Table 1). A similar result was obtained when the synthase was recombined with dissoci-ated ClAcCoA-modified enzyme, and the activity of the reconstructed homodimer decreased to 9 – 13% of the level of the control. These residual activities after the alkylation corresponded closely to the amounts of activity determined immediately after the chemical modifications of the dimeric 6HM synthase (Table 1), probably due to small amounts of unmodified enzyme present. When between the two polypeptides [3,4]. Therefore, we
reconstruction was carried out with unmodified and IoAA-modified 6HM synthase, 59 – 65% ac-tivity was demonstrated for the hybrid dimers. If the enzyme is organized as the head-to-tail ori-ented homodimer, it is reasonable to assume that the recombination of unmodified and IoAA-modified enzyme yields fully active unIoAA-modified – unmodified, unmodified – IoAA modified con-taining a single viable active site, and inactive IoAA modified – IoAA-modified dimers in a 1:2:1 ratio as with animal FAS [15]. Taking account of the ‘background’ activity of the unmodified en-zyme involved in the IoAA-treated preparation, it is reasonable to conclude that approximately 50% activity was demonstrated in the hybrid dimers obtained by the recombination of the unmodified and IoAA-modified variant. A similar set of re-sults was also obtained when ClAcCoA-modified 6HM synthase was employed for the hybridiza-tion with unmodified enzyme. A 50% active mix-ture of hybrid dimers was demonstrated in the reconstructed enzymes prepared from the un-modified and ClAcCoA-un-modified subunits (Table
1). These results are consistent with our previous assumption that, as with animal FAS, two multi-functional subunits of 6HM synthase are aligned in a head-to-tail direction, and the catalytic do-mains of acyl condensation and ketoreduction be-longing to other subunits are associated with each other to form the two complete reaction centers in the homodimeric enzyme.
3.3. Hybrids of chemically modified 6HM synthase
As already described, it is likely [4] that both Cys-SH and ACP-SH at the reaction center of 6HM synthase contribute from the same subunit in the homodimer structure. To confirm this ‘un-usual’ arrangement of the catalytic domains at the functional site of the synthase, the activity of the reconstructed hybrid dimers prepared from the IoAA- and ClAcCoA-modified monomer sub-units were determined (Table 1). The reconsti-tuted 6HM synthase protein showed only very low activity, almost similar to the ‘background’ observed for IoAA- or ClAcCoA-modified ho-modimers recombined in the absence of others. The formation of TAL was also undetectable in these hybridized enzymes in the presence or ab-sence of 1 mM NADPH (data not shown). These observations strongly suggest that none of the variants obtained by the recombination of Cys-blocked and ACP-Cys-blocked 6HM synthase exhibit the catalytic activity of acyl condensation. If the arrangement of two SH groups of 6HM synthase resembles that of animal FAS, the dissociation of IoAA- and ClAcCoA-treated enzyme should re-sult in the formation of monomer subunits I and
II in Fig. 3, respectively. Reconstitution of I and
II would form the variants of 1, 2 and 3 in a 1:2:1 ratio, and, as with FAS [15], regaining of 25% activity is expected. In contrast, as in our previously proposed model [4], reconstruction of the SH-blocked 6HM synthase is expected to form 4, 5 and 6 via III and IV, and none of the variants should catalyze the 6HM-forming reac-tion. Therefore, the results obtained in the present study strongly support our previous hy-pothesis that both SH groups at the reaction
cen-ter of 6HM synthase belong to and are
contributed by the same subunit in the homod-imer structure associated in a head-to-tail orienta-tion.
Table 1
Activity of unmodified and modified 6HM synthase variants
6HM synthase variants Relative activity (%)
Experiment 1 Experiment 2
Without dissociationa
Unmodified 100 100
IoAA 9 13
11
ClAcCoA 7
Reconstructed from dissociated subunitsb
100
aCys-SH and ACP-SH at the reaction center of 6HM
synthase were alkylated with IoAA and ClAcCoA, respec-tively, and the residual activities of the modified enzymes were determined immediately after the chemical modifications. Re-sults were expressed as percentages of that of the unmodified control.
b6HM synthase variants (unmodified, IoAA modified and
Fig. 3. Scheme of the possible arrangement of SH groups and reconstructed dimer structures prepared from the modified 6HM synthase variants. In the animal FAS-like arrangement (upper part), ACP-SH locates near the ketoreducing domain, while Cys-SH of the condensation enzyme locates at another terminal in the primary structure. Treatment with IoAA and ClAcCoA should result in the formation ofIandII, respectively, and reconstitution of these variants should form1,2, and3, in the ratio 1:2:1. In contrast, in our putative model of 6HM synthase (lower part), reconstruction of IoAA- and ClAcCoA-modified enzymes should form4,5, and 6viaIIIandIVin the ratio 1:2:1.
4. Discussion
In the present studies, it has been clearly demonstrated that hybrid dimers obtained by dis-sociation and recombination of unmodified and SH-modified 6HM synthase regain 50% activity. These observations appear to support our previ-ously proposed hypothetical model of the organi-zation of the enzyme structure [3,4]. As in animal FAS, two multifunctional subunits of 6HM syn-thase are aligned in a head-to-tail direction, and the catalytic domains of acyl condensation and ketoreduction belonging to the other subunits are associated with each other to form the two com-plete reaction centers in the homodimeric struc-ture. Recovery of 50% of the activity after the reconstruction also suggests that the two func-tional sites in the dimeric form of the synthase are independently active.
In our previous work, we hypothesized [4] that, unlike in FAS, both of the two SH groups at the reaction center belong to the same subunit. How-ever, it was demonstrated [16] that chalcone syn-thase, a well-characterized polyketide synthetic enzyme distributed in a wide range of higher plants, catalyzes the acyl condensation reaction even though the enzyme is lacking in ACP-SH in its structure. Therefore, although the occurrence of ACP-SH in 6HM synthase is evident [2], the
possibility could not be excluded that the monomeric subunit of 6HM synthase would liber-ate TAL only with Cys-SH of the condensation enzyme, even in the absence of the functional participation of ACP-SH. If, as in FAS, the two SH groups are contributed from the other subunits of the homodimer, the reconstitution of the par-tially blocked monomer subunits should generate unblocked reaction centers in an appropriate com-bination of the variants (Fig. 3). Demonstration of the complete loss of the activity in the hybridiza-tion experiments with IoAA- and ClAcCoA-modified 6HM synthase subunits strongly suggests that the contribution of Cys- and ACP-SHs, and the arrangement of the catalytic domains in the reaction center of 6HM synthase, are quite differ-ent from those of animal FAS. Together with our previous report [4], it is very likely that Cys-SH and ACP-SH are contributed by the same subunit in the homodimer structure.
enzymes responsible for biosynthesis of aglycon portion of the antibiotic erythromycin are located inSaccharopolyspora erythraea, and are organized in six repeated units resembling the coding region of animal FAS gene [19] (for example, acyltrans-ferase – ACP – ketoacyl synthase – acyltransacyltrans-ferase – ketoreductase – ACP – ketoacyl
synthase – acyltransferase – ketoreductase – ACP). If gene fusion occurred from sixth ACP to ninth ketoreductase during the evolution from individual monofunctional polypeptides to multifunctional protein, this might result in the formation of a protein similar to III and IV shown in Fig. 3. Therefore, it is possible that differences in gene fusion sites in the clusters responsible for monofunc-tional polyketide synthases generate the variants with unusual alignment of catalytic domains of this class of multifunctional proteins. Alternatively, it is also possible that consolidation of the individual genes occurred by a variety of chromosomal rear-rangement mechanisms, just as the movement of transposable elements results in the formation of the variants of multifunctional proteins with unusual arrangements.
In the present study, it remains to be clarified that, in the primary structure of 6HM synthase, the Cys-SH group of the condensation enzyme is lo-cated at the ‘core’ side of the polypeptide, while ACP-SH is at the terminal side, or vice versa. Elucidation of the alignment of the catalytic do-mains in 6HM-synthase structure is in progress in our laboratory by the combination of peptide mapping via limited proteolysis and gene cloning techniques.
Acknowledgements
This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan.
References
[1] F. Kurosaki, Y. Kizawa, A. Nishi, Derailment product in NADPH-dependent synthesis of dihydroisocoumarin 6-hydroxymellein by elicitor-treated carrot cell extracts, Eur. J. Biochem. 185 (1989) 85 – 89.
[2] F. Kurosaki, M. Itoh, M. Yamada, A. Nishi, 6-Hydrox-ymellein synthetase as a multifunctional enzyme complex
in elicitor-treated carrot root extract, FEBS Lett. 288 (1991) 219 – 221.
[3] F. Kurosaki, Dimer as the active form of a polyketide biosynthetic enzyme 6- hydroxymellein synthase, Phyto-chemistry 39 (1995) 515 – 519.
[4] F. Kurosaki, Dissociation of dimeric 6-hydroxymellein synthase, a polyketide biosynthetic enzyme in carrot cell extracts, with loss of ketoreducing activity, Arch. Biochem. Biophys. 321 (1995) 239 – 244.
[5] S.J. Wakil, J.K. Stoops, Structure and mechanism of fatty acid synthetase, in: P. Boyer (Ed.), The Enzymes, vol. 16, Academic Press, New York, 1983, pp. 3 – 61.
[6] S.J. Wakil, Fatty acid synthase, a proficient multifunc-tional enzyme, Biochemistry 28 (1989) 4523 – 4530. [7] S.J. Wakil, J.K. Stoops, V.C. Joshi, Fatty acid synthesis
and its regulation, Annu. Rev. Biochem. 52 (1983) 537 – 579.
[8] F. Kurosaki, Effect of NADPH-associated keto-reducing domain on substrate entry into 6- hydroxymellein syn-thase, a multifunctional polyketide synthetic enzyme in-volved in phytoalexin biosynthesis in carrot, Arch. Biochem. Biophys. 328 (1996) 170 – 172.
[9] A. Kawaguchi, T. Yoshimura, S. Okuda, A new method for the preparation of acyl-CoA thioesters, J. Biochem. 89 (1981) 337 – 339.
[10] F. Kurosaki, M. Itoh, Y. Kizawa, A. Nishi, Partial purification and characterization of a polyketide biosyn-thetic enzyme 6-hydroxymellein synthase in elicitor-treated carrot cell extracts, Arch. Biochem. Biophys. 300 (1993) 157 – 163.
[11] M.M. Bradford, A rapid and sensitive method for quan-titation of microgram quantities of protein utilizing the principle of protein – dye binding, Anal. Biochem. 72 (1976) 248 – 254.
[12] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680 – 685.
[13] F. Kurosaki, Transacylase-like structure and its role in substrate channeling of 6- hydroxymellein synthase, a multifunctional polyketide biosynthetic enzyme in carrot cell extracts, FEBS Lett. 379 (1996) 97 – 102.
[14] J.K. Stoops, S.J. Wakil, Animal fatty acid synthetase. A novel arrangement of the beta-ketoacyl synthetase sites comprising domains of the two subunits, J. Biol. Chem. 256 (1981) 5128 – 5133.
[15] Y.S. Wang, T.W. Tian, R.Y. Hsu, Hybridization studies of chicken liver fatty acid synthetase, J. Biol. Chem. 259 (1984) 13644 – 13647.
[16] S. Tropf, B. Kaercher, G. Schroeder, J. Schroeder, Reac-tion mechanisms of homodimeric plant polyketide syn-thases (stilbene and chalcone synthase), J. Biol. Chem. 270 (1995) 7922 – 7928.
[17] S. Smith, The animal fatty acid synthase: one gene, one polypeptide, seven enzymes, FASEB J. 8 (1994) 1248 – 1259.
[18] D.G. Hardie, J.R. Coggins, Multidomain Proteins: Struc-ture and Evolution, Elsevier, Amsterdam, 1986. [19] S. Donadio, M.J. Staver, J.B. McAlpine, S.J. Swanson, L.
Katz, Modular organization of genes required for com-plex polyketide biosynthesis, Science 252 (1991) 675 – 679.
