An alternative mechanism of product chain-length determination in type III geranylgeranyl diphosphate synthase Hisashi Hemmi, Motoyoshi Noike, Toru Nakayama and Tokuzo Nishino Department of Biomolecular Engineering, Graduate School of Engineering, Tohoku University, Sendai, Miyagi, Japan (All-E) prenyl diphosphate synthases catalyze the con- secutive condensation of isopentenyl diphosphates with allylic prenyl diphosphates, producing products with vari- ous chain-lengths that are unique for each enzyme. Some short-chain (all-E) prenyl diphosphate synthases, i.e. farnesyl diphosphate synthases and geranylgeranyl diphosphate synthases contain characteristic amino acid sequences around the allylic substrate binding sites, which have been shown to play a role in determining the chain- length of the product. However, among these enzymes, which are classified into several types based on the pos- sessive patterns of such characteristics, type III geranyl- geranyl diphosphate synthases, which consist of enzymes from eukaryotes (excepting plants), lack these features. In this study, we report that mutagenesis at the second position before the conserved G(Q/E) motif, which is distant from the well-studied region, affects the chain- length of the product for a type III geranylgeranyl diphosphate synthase from Saccharomyces cerevisiae.This clearly suggests that a novel mechanism is operative in the product determination for this type of enzyme. We also show herein that mutagenesis at the corresponding posi- tion of an archaeal medium-chain enzyme also alters its product specificity. These results provide valuable infor- mation on the molecular evolution of (all-E)prenyl diphosphate synthases. Keywords: prenyltransferase; geranylgeranyl diphosphate synthase; hexaprenyl diphosphate synthase; mutagenesis; molecular evolution. (All-E) prenyl diphosphate synthases catalyze consecutive condensations of isopentenyl diphosphate (IPP) in the E-type configuration with allylic primer substrates and yield products with various hydrocarbon-chain lengths that are specific to each enzyme. The products are utilized as precursors for numerous types of isoprenoid compounds such as steroids, carotenoids, respiratory quinones and prenylated proteins (Fig. 1). The enzymes have been classified into three groups based on their quaternary structure and the chain-length of the product produced, i.e. short-, medium-, and long-chain enzymes yielding C 10)25 , C 30)35 and C 40)50 products, respectively [1,2]; therefore, these designations are used in this paper. The enzymes of the three groups are thought to have similar structures and the same catalytic mechanism is involved, because their amino acid sequences have a high degree of similarity [3,4]. For example, two aspartate-rich motifs, designated as FARM (the first aspartate-rich motif) and SARM (the second aspartate-rich motif), are completely conserved among these enzymes and act as binding sites for allylic substrates and IPP, respectively. A crystallographic analysis of avian farnesyl diphosphate synthase (FPS) revealed that the homodimeric enzyme consists almost entirely of a-helices, some of which constitute a reaction cavity in the center of a subunit of the enzyme, and that FARM and SARM both exist on distinct a-helices and face each other at different sides of the rim of the cavity [5]. Short-chain (all-E) prenyl diphosphate synthases are known to have strict product specificities, and the mecha- nisms involved in product determination have so far been investigated using FPSs and geranylgeranyl diphosphate synthases (GGPSs) that yield C 15 and C 20 products, respectively. Several mutagenic studies have revealed that aromatic amino acids, frequently found at the fourth and fifth positions, before FARM of short-chain enzymes, are involved in these mechanisms [6–12]. It is thought that these aromatic amino acids act as the bottom of a reaction cavity to prevent further elongation of the prenyl chain of the final product. In addition, two amino acids inserted into FARM, which occurs in FPSs and GGPSs from bacteria and plants, are also considered to be involved in the mechanism [9]. Thus, we designated the area, including FARM and several amino acids upstream of it, as the CLD (chain-length determination) region and proposed a classification of these enzymes based on the patterns of such characteristic amino acid residues (Fig. 2). However, no such residue is found in enzymes classified as type III GGPSs, the group consisting of GGPSs from eukaryotes (excepting plants). The typical Correspondence to Tokuzo Nishino, Department of Biomolecular Engineering, Graduate School of Engineering, Tohoku University, Aoba-yama 07, Sendai, Miyagi 980-8579, Japan. Fax/Tel.: + 81 22 217 7270, E-mail: nishino@mail.cc.tohoku.ac.jp Abbreviations: IPP, isopentenyl diphosphate; GPP, geranyl diphosphate; FPP, farnesyl diphosphate; GGPP, geranylgeranyl diphosphate; FPS, farnesyl diphosphate synthase; GGPS, geranylgeranyl diphosphate synthase; HPS, hexaprenyl diphosphate synthase; FARM, the first aspartate-rich motif; CLD, chain-length determination. Enzymes: geranylgeranyl diphosphate synthase (EC 2.5.1.29), hexaprenyl diphosphate synthase (EC 2.5.1.30), farnesyl diphosphate synthase (EC 2.5.1.10). (Received 24 January 2003, revised 10 March 2003, accepted 18 March 2003) Eur. J. Biochem. 270, 2186–2194 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03583.x sequence of the CLD region of type III GGPSs resembles that of medium- and long-chain (all-E) prenyl diphosphate synthases: no bulky aromatic amino acids exist at the fourth and fifth positions before FARM, and the insertion of two amino acids is not observed. This fact strongly suggests that some characteristic amino acids that are located outside the CLD region are conserved in type III GGPSs and play an important role in the mechanism of chain-length determin- ationintheseenzymes. In our previous report in 1996, which was the first reference to the mechanism of product determination in (all-E) prenyl diphosphate synthase, mutants of type II FPS from Bacillus stearothermophilus were produced by random mutagenesis, and those yielding longer products such as geranylgeranyl diphosphate (GGPP) were selected [6]. Two of four selected mutants shared the same amino acid substitution at the fifth position before FARM (Y81H), which led us to subsequently discover the importance of the Fig. 1. Reactions of (all-E) prenyl diphosphate synthases in isoprenoid biosynthesis. OPP, OP 2 O 5 3– . Fig. 2. Alignment of amino acid sequences around FARM and the G(Q/E) motif of various (all-E) prenyl diphosphate synthases. The partial amino acid sequences of the enzymes classified into two types of FPSs, three types of GGPSs and longer-chain enzymes are aligned. Sce FPS, S. cerevisiae FPS; Gga FPS, Gallus gallus (avian) FPS; Eco FPS, E. coli FPS; Bst FPS, B. stearothermophilus FPS; Sac GGPS, S. acidocaldarius GGPS; Mth GGPS, Methanobacterium thermoautotrophicum GGPS; Pan GGPS, Pantoea ananatis (Erwinia uredovora) GGPS; Ath GGPS, Arabidopsis thaliana GGPS; SceGGPS, S. cerevisiae GGPS; Hsa GGPS, Homo sapiens GGPS; SsoHPS, S. solfataricus HPS (hexaprenyl diphosphate synthase); Eco OPS, E. coli octaprenyl diphosphate (C 40 ) synthase. The amino acid residues that were suggested to concern product determination are shaded. Arrowheads indicate the positions into which substitutive mutations were introduced in this work. Ó FEBS 2003 Another evolutionary route of prenyltransferase (Eur. J. Biochem. 270) 2187 CLD region in product determination. However, the rest of the mutants had substitutive mutations at different posi- tions. One mutant, designated as mutant 3, possessed mutations at two positions (V157A/H182Y), and further mutational experiments revealed that a V157A mutation was sufficient to change product specificity. The mutated site is located at the second position before the G(Q/E) motif that is highly conserved among all (all-E)prenyl diphosphate synthases. This fact led us to the hypothesis that the position might play an important role in product determination in type III GGPSs because they do not possess the usual characteristic amino acid residues in the CLD region. This idea is supported by the fact that histidine, a relatively large amino acid, is conserved at this position in type III GGPSs, while smaller ones, such as valine, alanine, serine, cysteine or glutamic acid, are conserved in longer-chain enzymes and in the short-chain enzymes of the other groups (Fig. 2). In this study, substitutive mutations were introduced at the second position before the G(Q/E) motif of a type III GGPS from Saccharomyces cerevisiae. The mutations were clearly shown to affect the product determination of the enzyme. Mutational studies using an archaeal medium- chain enzyme, hexaprenyl diphosphate synthase (HPS) from Sulfolobus solfataricus, also provided support for the importance of this specific position and, moreover, suggest the existence of a novel evolutional pathway of (all-E) prenyl diphosphate synthases. Materials and methods Materials Precoated reversed-phase thin-layer chromatography plates, LKC-18F were purchased from Whatman Chemical Separation, Inc. (All-E)-GGPP (all-E)-farnesyl diphosphate (FPP), and geranyl diphosphate (GPP) were kindly donated by K. Ogura and T. Koyama, Tohoku University. [1- 14 C]IPP was purchased from Amersham Bioscience Inc. All other chemicals were of analytical grade. General procedures Restriction enzyme digestions, transformations, and other standard molecular biological techniques were carried out as described by Sambrook et al.[13]. Site-directed mutagenesis A DNA fragment including the S. cerevisiae GGPS gene was amplified by means of the polymerase chain reaction utilizing the S. cerevisiae genome as a template, KOD DNA polymerase (Toyobo), and the primers indicated below: ScGGPS-Fw, 5¢-TAGACGGTACCAAGCTT CATATG GAGGCCAAGATAGATGAGC-3¢;ScGGPS-Rv,5¢-GT CTAGGTACCAAGCTT GGATCCTCACAATTCGGA TAAGTGGTC-3¢. The sequences corresponding to the NdeIandBamHI sites used later are underlined. The amplified fragment was cleaved with these two restriction enzymes and then inserted into the NdeI/BamHI-treated pET-15b vector (Novagen) to construct pET-HisScGG, a plasmid for the recombinant expression of His 6 -tagged S. cerevisiae GGPS. On the other hand, pET-HisHPS, constructed in our previous study [14], was employed in the recombinant expression of S. solfataricus HPS. Substitutive mutations were specifically introduced into these plasmids utilizing a QuickChange Mutagenesis Kit (Stratagene) according to the protocol provided by the manufacturer. Pairs of oligonucleotide primers used for each mutagenesis are indicated in Table 1. In addition, plasmids for the expression of mutants possessing additional mutations with mutant H139A, i.e. pET-HisScGG-H139A- 3A, pET-HisScGG-H139A-4A and pET-HisScGG- H139A-3A4A, were constructed from the plasmid pET-HisScGG-H139A. Expression and purification of the wild-type and mutated enzymes Escherichia coli BL21(DE3) transformed with each of the plasmids shown above or in Table 1 was cultivated in 50 mL of M9YG broth supplemented with ampicillin (50 mgÆL )1 ). When the D 600 of the culture reached 0.5, the transformed bacteria were induced with 1.0 m M isopropyl thio-b- D -galactoside. After an additional overnight cultiva- tion, the cells were harvested. The disruption of the cells and the purification of the tagged enzymes were conducted utilizing a MagExtractor His-tag Kit (Toyobo) according to the protocol recommended by the manufacturer. The level of purification was determined by 15% SDS/PAGE (data not shown). Measurement of prenyltransferase activity The assay mixture for wild-type and mutant S. cerevisiae GGPSs contained, in a final volume of 200 lL, 0.5 nmol [1– 14 C]IPP (2 GBqÆmmol )1 ), 0.5 nmol of an allylic primer (GGPP, FPP or GPP), 1 lmol MgCl 2 ,2lmol Tris/HCl buffer (pH 7.5), 2 lmol KF, and a suitable amount of each enzyme. This mixture was incubated at 30 °C for 1.5 h, and the reaction was stopped by chilling the mixture in an ice bath. The mixture was shaken with 600 lL of 1-butanol saturated with H 2 O. The butanol layer was washed with water saturated with NaCl, and the radioactivity in 100 lL of the butanol layer was measured with a TRI-CARB 1600 liquid scintilation counter (Packard). The residual butanol layer was used for product analysis. S. solfataricus HPS activity was assayed using the same method, except that the reaction mixture for the wild-type and mutants of S. solfataricus HPS contained 0.5 nmol [1– 14 C]IPP (2 GBqÆmmol )1 ), 0.5 or 25 nmol GPP, 1 lmol MgCl 2 ,2lmol of a phosphate buffer (pH 6.3), 0.1% (v/v) Triton X-100, and a suitable amount of enzyme in a final volume of 200 lL. It was incubated at 55 °C for 30 min and then processed as described above. Product analysis Prenyl diphosphates in the residual 1-butanol layer were treated with acid phosphatase according to the method of Fujii et al. [15]. The hydrolysates were extracted with pentane and analyzed by reversed-phase thin-layer chro- matography using a precoated plate, LKC-18F, developed with acetone/H 2 O (9 : 1, v/v). Authentic, standard alcohols 2188 H. Hemmi et al.(Eur. J. Biochem. 270) Ó FEBS 2003 were visualized with iodine vapor, and the distribution of radioactivity was detected by Molecular imager (Bio-Rad). Results Mutational studies using S. cerevisiae GGPS To confirm the hypothesis that the amino acid at the second position before the G(Q/E) motif plays an important role in the product determination of type III GGPSs, a system for the recombinant expression of S. cerevisiae GGPS, which is classified as a type III GGPS, was constructed. Mutational experiments were conducted using the recombinant enzymes. The wild-type and mutant enzymes expressed by E. coli were purified utilizing a (His) 6 -tag fused at the N-terminus of each enzyme. We initially constructed mutants of S. cerevisiae GGPSinwhichhistidine139, which is located at the second position before the G(Q/E) motif, was replaced with smaller amino acids such as alanine, glycine or serine (Fig. 3A). These mutants retained high enzyme activities that are comparable with that of wild- type GGPS when GPP or FPP was used as the allylic substrate (Fig. 3B). One of the mutants, H139A, showed considerable activity when GGPP was used, while the wild- type GGPS barely reacted with the allylic substrate. Analyses demonstrated that the H139A mutant yielded a longer C 25 final product (Fig. 3C). Moreover, the mutant produced a small amount of C 30 prenyl diphosphate when GGPP was used as the allylic substrate. The H139G and H139S mutants also yielded C 25 prenyl diphosphate, but in very small amounts. This strongly suggests that the mechanism involved in the product determination in S. cerevisiae GGPS is largely dependent on histidine 139, a residue that is completely different from those which play major roles in the mechanisms of many short-chain enzymes. Based on the above data, we hypothesize that S. cerevis- iae GGPS utilizes a mechanism that is similar to those of many short-chain enzymes, in which bulky amino acids before FARM block the elongation of the prenyl chain at the bottom of the cavity in the enzyme. Thus, additional mutations were introduced into mutant H139A to deter- mine whether the cavity of the enzyme can be expanded directly below position 139. It has been reported that Sulfolobus acidocaldarius GGPS mutants that possess double mutations, i.e. the replacement of the amino acids at the fifth and eighth positions before FARM with smaller ones, are not able to control the chain-length of their final products and yield products longer than C 100 [10]. Similar mutational experiments using B. stearothermophilus FPS provided the same result. These mutated positions are thought to exist at the same side of the a-helix that contains FARM, which binds the diphosphate moiety of prenyl diphosphate. The introduction of smaller amino acids at these positions would be expected to give rise to a path through which the prenyl chain could elongate along the a-helix. According to the three-dimensional structure of avian FPS (Protein Data Bank accession no. 1FPS), which is the only (all-E) prenyl diphosphate synthase whose crystal structure has been determined, the a-helix containing the G(Q/E) motif is in the same orientation as that containing FARM. These two a-helices adjoin each other and comprise a portion of the reaction cavity. Therefore, we replaced L135 and/or I136 of mutant H139A with alanine because the distance between H139 and these residues generally corresponds to a pitch of an a-helical coil (Fig. 4A). All of the mutants L135A/H139A, I136A/H139A and Table 1. Mutagenic primers used in this work. Primers Sequences a Generated plasmids/mutants For the construction of S. cerevisiae GGPS mutants ScGGPS-H139A-Fw 5¢-GAATTGATCAATCTAgc ccgcGGACAAGGCTTGG-3¢ pET-HisScGG-H139A/ ScGGPS-H139A-Rv 5¢-CCAAGCCTTGT CCgcgggcTAGATTGATCAATTC-3¢ mutant H139A ScGGPS-H139G-Fw 5¢-GAATTGATCAATCTAgg ccgcGGACAAGGCTTGG-3¢ pET-HisScGG-H139G/ ScGGPS-H139G-Rv 5¢-CCAAGCCTTGT CCgcggccTAGATTGATCAATTC-3¢ mutant H139G ScGGPS-H139S-Fw 5¢-GAATTGATCAATCTAag ccgcGGACAAGGCTTGG-3¢ pET-HisScGG-H139S/ ScGGPS-H139S-Rv 5¢-CCAAGCCTTGTCCgcggctTAGATTGATCAATTC-3¢ mutant H139S ScGGPS-H139A-3A-Fw 5¢-CGATTTTCAAC GAgGAgTTGgcCAATCTAGCCCGCGG-3¢ pET-HisScGG-H139A-3A/ ScGGPS-H139A-3A-Rv 5¢-CCGCGGGCTAGATTGgcCAA cTCcTCGTTGAAAATCG-3¢ mutant I136A/H139A ScGGPS-H139A-4A-Fw 5¢-CGATTTTCAAC GAgGAggcGATCAATCTAGCCCG-3¢ pET-HisScGG-H139A-4A/ ScGGPS-H139A-4A-Rv 5¢-CGGGCTAGATTGATCgc cTCcTCGTTGAAAATCG-3¢ mutant L135A/H139A ScGGPS-H139A-3A4A-Fw 5¢-CGATTTTCAAC GAgGAggcGgcCAATCTAGCCCGCGG-3¢ pET-HisScGG-H139A-3A4A/ ScGGPS-H139A-3A4A-Rv 5¢-CCGCGGGCTAGATTGgcCgc cTCcTCGTTGAAAATCG-3¢ mutant L135A/I136A/H139A For the construction of S. solfataricus HPS mutants SsHPS-S140H-Fw 5¢-GTTATGGAAAGACACCcatGTGG GaGCTCTAAGGGATATG-3¢ pET-HisHPS-S140H/ SsHPS-S140H-Rv 5¢-CATATCCCTTAGAGCtCCCACatgGGTGTCTTTCCATAAC-3¢ mutant S140H SsHPS-S140V-Fw 5¢-GTTATGGAAAGACACCgtAGTGG GaGCTCTAAGGGATATG-3¢ pET-HisHPS-S140V/ SsHPS-S140V-Rv 5¢-CATATCCCTTA GAGCtCCCACTacGGTGTCTTTCCATAAC-3¢ mutant S140V SsHPS-S140F-Fw 5¢-GTTATGGAAAGACACCTttGTGG GaGCTCTAAGGGATATG-3¢ pET-HisHPS-S140F/ SsHPS-S140F-Rv 5¢-CATATCCCTTA GAGCtCCCACaaAGGTGTCTTTCCATAAC-3¢ mutant S140F SsHPS-S140Y-Fw 5¢-GTTATGGAAAGACACCTatGTGG GaGCTCTAAGGGATATG-3¢ pET-HisHPS-S140Y/ SsHPS-S140Y-Rv 5¢-CATATCCCTTAGAGCtCCCACatAGGTGTCTTTCCATAAC-3¢ mutant S140Y a Mismatched sequences are lower-cased, and newly introduced restriction sites to confirm mutagenesis are underlined. Ó FEBS 2003 Another evolutionary route of prenyltransferase (Eur. J. Biochem. 270) 2189 L135A/I136A/H139A yielded longer products than H139A when GGPP was used as the allylic substrate (Fig. 4B). The longest products of these mutants appeared to be C 40 or more. This result strongly suggests that the prenyl chain of the product elongates along the a-helix containing the G(Q/E) motif in the mutants and that residue H139 acts as the bottom of the reaction cavity in the wild-type S. cere- visiae GGPS. Fig. 3. Replacement of the amino acid residue at the second position before the G(Q/E) motif in S. cerevisiae GGPS. (A) Partial amino acid sequences around the G(Q/E) motif of wild type and mutated enzymes are aligned. The substituted amino acid residues are shaded. (B) Specific activities of the wild-type and mutated enzymes. Enzyme reactions were carried as described in Materials and methods, and activities were determined by the amount of radioactivity extracted with 1-butanol. (C) TLC autoradiochromatograms of the reaction products of the wild-type and mutated enzymes. The products were analyzed as described in Materials and methods. The allylic substrate used was indicated at the top of each autoradiochromatogram. Lane 1, mutant H139A; lane 2, mutant H139G; lane 3, mutant H139S; lane 4, wild type GGPS. Under all assay conditions, less than 30% of each substrate reacted. Ori., origin; S.F., solvent front. 2190 H. Hemmi et al.(Eur. J. Biochem. 270) Ó FEBS 2003 Mutational studies using S. solfataricus HPS In our previous phylogenetic study, S. solfataricus HPS, a medium-chain enzyme that was recently cloned and char- acterized, was shown to exist at an anomalous position in the phylogenetic tree of (all-E) prenyl diphosphate synthases [16]. The enzyme was included in a branch consisting of eukaryotic short-chain enzymes, i.e. type I FPSs and type III GGPSs, but not in that of other medium- and long-chain enzymes. Its phylogenetic position was particularly close to enzymes classified as type III GGPSs, which strongly suggests a close relationship between the archaeal medium- chain enzyme and type III GGPSs in the evolution of (all-E) prenyl diphosphate synthases. However, the issue of what difference between their amino acid sequences gave rise to the difference in their final products remains unclear. It must be noted that our attempt to change the product specificity of the medium-chain enzyme by replacing the CLD region with that of S. cereviseae GGPSfailed[14].Asaresult, S. solfataricus HPS was used for mutational analyses to demonstrate the importance of the amino acid residue corresponding to H139 of S. cerevisiae GGPS and to confirm the evolutionary relationship between the archaeal HPS and the eukaryotic GGPS. S. solfataricus HPS contains a serine residue at position 140, the second position before the G(Q/E) motif (GA in S. solfataricus HPS represents an exception); thus, S140 was replaced with a larger amino acid to determine whether this substitutive mutation could be the origin of the change in product specificity of the medium-chain enzyme to those of short-chain enzymes. We constructed two mutants, S140H and S140V, that mimic S. cerevisiae GGPS and B. stearo- thermophilus FPS, respectively, because the product speci- ficity of both of these short-chain enzymes changes as the result of a point mutation at their corresponding positions (Fig. 5A). In addition, we also constructed two mutants, S140F and S140Y, containing an aromatic amino acid residue at this position. When the same concentrations of IPP and GPP were used in the assays, the enzyme activity of the S140H and S140Y mutants was decreased to 54 and 37% of that of the wild-type HPS, respectively, while the other mutants showed an activity comparable with that of the wild-type enzyme. Product analyses showed that their final products continued to be the hexaprenyl diphosphate, the C 30 product (Fig. 5B). However, except for S140V, the mutants accumulated large amounts of FPP as an inter- mediate. On the other hand, the S140V mutant showed a slightly increased production of GGPP, in comparison with the wild-type enzyme. When the concentration of GPP was increased to 50 times that of IPP to enhance the production of intermediates, all of the mutants showed an enzyme activity comparable to the wild-type enzyme. However, all mutants produced FPP as the main product, while the wild- type HPS yielded the C 30 product (Fig. 5C). Only the S140V mutant continued to produce a considerable amount of hexaprenyl diphosphate. These results indicate that position 140 plays a significant role in the product determination in these mutants, probably by blocking the elongation of the prenyl chain of the products, but in an incomplete manner. We hypothesize that type III GGPSs might have evolved from some longer-chain enzyme, such as S. solfataricus HPS, based on the result of a phylogenetic analysis of (all-E) prenyl diphosphate synthases. The fact that the product specificity of the archaeal medium-chain enzyme partially changed into those of short-chain enzymes by the acquisi- tion of a characteristic amino acid residue at the second position before the G(Q/E) motif supports this hypothesis. Discussion The importance of the second position before the G(Q/E) motif in the mechanism of product determination in S. cerevisiae GGPS was examined by mutagenic studies. The results from the characterization of the mutants strongly suggests that the amino acid residue at the position of S. cerevisiae GGPS might function in a manner similar to those of the aromatic amino acids at the fourth and fifth positions before FARM in some other short-chain enzymes, Fig. 4. Introduction of additional mutations into mutant H139A. (A) Partial amino acid sequences around the G(Q/E) motif of mutant H139A and mutated enzymes are aligned. The substituted amino acid residues are shaded. (B) TLC autoradiochromatograms of the reaction products of the wild-type and mutated enzymes. The products were analyzed as described in Materials and methods using GGPP as the allylic substrate. Lane 1, mutant H139A; lane 2, mutant I136A/ H139A; lane 3, mutant L135A/H139A; lane 4, mutant L135A/I136A/ H139A; lane 5, wild type GGPS. Under all assay conditions, less than 30% of each substrate reacted. Ori., origin; S.F., solvent front. Ó FEBS 2003 Another evolutionary route of prenyltransferase (Eur. J. Biochem. 270) 2191 i.e. the formation of the bottom of a reaction cavity (Fig. 6). However, this hypothesis appears to be incompatible with the fact that the H139G mutant did not yield products longer than those produced by H139A although the substitution with the smallest glycine residue would be expected to have the largest effect on the expansion of the cavity. On the other hand, the size of the side-chain of the amino acid residue used to substitute for serine 140 of S. solfataricus HPS seemed to correlate negatively with the product chain-length. This is similar to results obtained from an experiment in which the fifth position before FARM of B. stearothermophilus FPS was replaced [8]. The inconsistent result observed with the mutants of S. cerevis- iae GGPS might arise from a structural change in the cavity of the H139G mutant, which could arise from the introduction of a flexible glycine residue. The second position before the G(Q/E) motif would be expected to play a major role in the mechanism of product determination in type III GGPSs because these enzymes do not contain characteristic amino acids such as aromatic residues in the CLD region. It is known that a relatively large amino acid residue, i.e. histidine, is conserved at the position of the thus-far-characterized type III GGPSs, while smaller amino acid residues such as valine, alanine, serine, cysteine or glutamic acid are conserved in the longer-chain (all-E) prenyl diphosphate synthases and in the short-chain enzymes of other groups. This fact strongly suggests that all type III GGPSs share the same mechanism of product determination, in which the histidine residue is involved. However, previous studies have proposed that the second position before the G(Q/E) motif might also be involved in product determination in short-chain enzymes other than type III GGPSs. For example, we previously reported that a point mutation at this position in B. stearothermophilus FPS, which is classified as a type II FPS, results in the elongation of the final product [6]. Kawasaki et al. recently reported that the product specificities for FPS from Streptomyces argenteolus (DNA Data Bank of Japan accession no. AB083108) and GGPS from Streptomyces greseolosporeus (AB037907), both of which anomalously contain CLD regions similar to those found in type I GGPSs, are interchangeable by simply exchanging their amino acids at the second position before the G(Q/E) motif [17]. In the three-dimensional structure of avian FPS, E182, the residue at the second position before the G(Q/E) motif, is in spatial proximity to F112 and F113, the residues at the fourth and fifth positions before FARM, respectively. The side chain of F112 is in direct contact with that of E182. This suggests that, in some short-chain enzymes, there might be Fig. 5. Introduction of substitutive mutations into the second position before the G(Q/E) motif in S. solfataricus HPS. (A) Partial amino acid sequences around the G(Q/E) motif of wild type and mutated enzymes are aligned. The substituted amino acid residues are shaded. The reaction products of the enzymes were analyzed as described in Materials and methods using 0.5 nmol (B) and 25 nmol (C) of GPP as the allylic substrate. Lane 1, mutant S140H; lane 2, mutant S140V; lane 3, mutant S140F; lane 4, mutant S140 Y; lane 5, wild type HPS. Under all assay conditions, less than 40% of each substrate reacted. Ori., origin; S.F., solvent front. 2192 H. Hemmi et al.(Eur. J. Biochem. 270) Ó FEBS 2003 some cooperation among these residues, as they relate to product determination. It should be noted that Stanley Fernandez et al. attempted to alter the product specificity of avian FPS into that of geranyl diphosphate synthase by introducing point mutations at position 181, the third position before the G(Q/E) motif, although the mutations resulted in a significant reduction in enzyme activity [18]. Moreover, as indicated herein, the replacement of serine at the second position before the G(Q/E) motif with valine to mimic type II GGPSs also affected the product specificity of S. solfataricus HPS. Our future interest is to reveal the importance of this position in the product determination mechanisms of short-chain enzymes other than type III GGPSs. The mutational experiment utilizing S. solfataricus HPS successfully demonstrated that the medium-chain enzyme could be altered to yield short-chain products as the result of a point mutation at this position. The S140H mutant constructed in this experiment and type III GGPSs share similar features at regions that are known to be involved in product determination. Neither contains characteristic amino acid residues in their CLD regions, such as bulky amino acids at the fourth or fifth position before FARM or the insertion of two amino acids into FARM, and both contain a histidine residue at the second position before the G(Q/E) motif. Based on the information obtained from the phylogenetic tree of (all-E) prenyl diphosphate synthases, in which S. solfataricus HPS shows a particularly close relationship with type III GGPSs, we conclude that type III GGPSs might have evolved from an ancestor resembling S. solfataricus HPS [16]. It is conceivable that the introduc- tion of the S140H mutation into S. solfataricus HPS simulates the evolution of type III GGPSs from the ancestral enzyme. At least, the change in product specificity as the result of a mutagenesis at the second position before the G(Q/E) motif represents a possible pathway of mole- cular evolution between type III GGPSs and the longer-chain enzymes, and the evolutionary pathway would be totally independent of known ones that have arisen from mutations in the CLD regions (Fig. 6). Acknowledgements This work was supported by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. We are grateful to Drs Kyozo Ogura and Tanetoshi Koyama, Tohoku University, for providing prenyl diphosphates. We thank Dr Tohru Dairi, Toyama Prefectural University, for his participation in helpful discussions. References 1. Ogura, K. & Koyama, T. (1998) Enzymatic aspects of isoprenoid chain elongation. Chem. Rev. 98, 1263–1276. 2. Koyama, T. (1999) Molecular analysis of prenyl chain elongating enzymes. Biosci. Biotechnol. Biochem. 63, 1671–1676. 3. Koyama, T., Obata, S., Osabe, M., Takeshita, A., Yokoyama, K., Uchida, M., Nishino, T. & Ogura, K. (1993) Thermostable far- nesyl diphosphate synthase of Bacillus stearothermophilus:mole- cular cloning, sequence determination, overproduction, and purification. J. Biochem. 113, 355–363. 4. Chen, A., Kroon, P.A. & Poulter, C.D. (1994) Isoprenyl diphos- phate synthases: protein sequence comparisons, a phylogenetic tree, and predictions of secondary structure. Protein Sci. 3, 600–607. 5. Tarshis, L.C., Yan, M., Poulter, C.D. & Sacchettini, J.C. (1994) Crystal structure of recombinant farnesyl diphosphate synthase at 2.6-A ˚ resolution. Biochemistry 33, 10871–10877. 6. Ohnuma, S i., Nakazawa, T., Hemmi, H., Hallberg, A.M., Koyama, T., Ogura, K. & Nishino, T. (1996) Conversion from farnesyl diphosphate synthase to geranylgeranyl diphosphate synthase by random chemical mutagenesis. J. Biol. Chem. 271, 10087–10095. 7. Ohnuma,S i.,Hirooka,K.,Hemmi,H.,Ishida,C.,Ohto,C.& Nishino, T. (1996) Conversion of product specificity of archae- bacterial geranylgeranyl-diphosphate synthase. Identification of Fig. 6. Proposed mechanism of product determination in type III GGPSs. Ó FEBS 2003 Another evolutionary route of prenyltransferase (Eur. J. Biochem. 270) 2193 essential amino acid residues for chain length determination of prenyltransferase reaction. J. Biol. Chem. 271, 18831–18837. 8. Ohnuma, S i., Narita, K., Nakazawa, T., Ishida, C., Takeuchi, Y., Ohto, C. & Nishino, T. (1996) A role of the amino acid residue located on the fifth position before the first aspartate-rich motif of farnesyl diphosphate synthase on determination of the final pro- duct. J. Biol. Chem. 271, 30748–30754. 9. Ohnuma, S i., Hirooka, K., Ohto, C. & Nishino, T. (1997) Conversion from archaeal geranylgeranyl diphosphate synthase to farnesyl diphosphate synthase. Two amino acids before the first aspartate-rich motif solely determine eukaryotic farnesyl diphos- phate synthase activity. J. Biol. Chem. 272, 5192–5198. 10. Ohnuma, S i., Hirooka, K., Tsuruoka, N., Yano, M., Ohto, C., Nakane, H. & Nishino, T. (1998) A pathway where polyprenyl diphosphate elongates in prenyltransferase. Insight into a common mechanism of chain length determination of prenyltransferases. J. Biol. Chem. 273, 26705–26713. 11. Narita, K., Ohnuma, S. & Nishino, T. (1999) Protein design of geranyl diphosphate synthase. Structural features that define the product specificities of prenyltransferases. J. Biochem. 126, 566–571. 12. Tarshis, L.C., Proteau, P.J., Kellogg, B.A., Sacchettini, J.C. & Poulter, C.D. (1996) Regulation of product chain length by iso- prenyl diphosphate synthases. Proc. Natl Acad. Sci. USA 93, 15018–15023. 13. Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring. Harbor Laboratory Press, Cold Spring Harbor, NY, USA. 14. Hemmi, H., Noike, M., Nakayama, T. & Nishino, T. (2002) Change of product specificity of hexaprenyl diphosphate synthase from Sulfolobus solfataricus by introducing mimetic mutations. Biochem. Biophys. Res. Commun. 297, 1096–1101. 15. Fujii, H., Koyama, T. & Ogura, K. (1982) Efficient enzymatic hydrolysis of polyprenyl pyrophosphates. Biochim. Biophys. Acta 712, 716–718. 16. Hemmi, H., Ikejiri, S., Yamashita, S. & Nishino, T. (2002) Novel medium-chain prenyl diphosphate synthase from the thermo- acidophilic archaeon Sulfolobus solfataricus. J. Bacteriol. 184, 615–620. 17. Kawasaki,T.,Hamano,Y.,Kuzuyama,T.,Itoh,N.,Seto,H.& Dairi, T. (2003) Interconversion of product specificity of type I eubacterial farnesyl diphosphate synthase and geranylgeranyl diphosphate synthase by one amino acid substitution. J. Biochem. 133, 83–91. 18. Stanley Fernandez, S.M., Kellogg, B.A. & Poulter, C.D. (2000) Farnesyl diphosphate synthase. Altering the catalytic site to select for geranyl diphosphate activity. Biochemistry 39, 15316– 15321. 2194 H. Hemmi et al.(Eur. J. Biochem. 270) Ó FEBS 2003 . An alternative mechanism of product chain-length determination in type III geranylgeranyl diphosphate synthase Hisashi Hemmi, Motoyoshi. conserved in type III GGPSs and play an important role in the mechanism of chain-length determin- ationintheseenzymes. In our previous report in 1996, which