An unusual isopentenyl diphosphate isomerase found in the mevalonate pathway gene cluster from Streptomyces sp. strain CL190

Abstract

A gene cluster encoding five enzymes of the mevalonate pathway had been cloned from Streptomyces sp. strain CL190. This gene cluster contained an additional ORF, orfD, encoding an unknown protein that was detected in some archaebacteria and some Gram-positive bacteria including Staphylococcus aureus. The recombinant product of orfD was purified as a soluble protein and characterized. The molecular mass of the enzyme was estimated to be 37 kDa by SDS-polyacrylamide gel electrophoresis and 155 kDa by gel filtration chromatography, suggesting that the enzyme is most likely to be a tetramer. The purified enzyme contained flavin mononucleotide (FMN) with the amount per tetramer being 1.4 to 1.6 mol/mol. The enzyme catalyzed the isomerization of isopentenyl diphosphate (IPP) to produce dimethylallyl diphosphate (DMAPP) in the presence of both FMN and NADPH. The Escherichia coli plasmid expressing orfD could complement the disrupted IPP isomerase gene in E. coli. These results indicate that orfD encodes an unusual IPP isomerase showing no sequence similarity to those of IPP isomerases identified to date. Based on the difference in enzymatic properties, we classify the IPP isomerases into two types: Type 2 for FMN- and NAD(P)H-dependent enzymes, and type 1 for the others. In view of the critical role of this isomerase in S. aureus and of the different enzymatic properties of mammalian (type 1) and S. aureus (type 2) isomerases, this unusual enzyme is considered to be a suitable molecular target for the screening of antibacterial drugs specific to S. aureus.

Isoprenoids play important roles in all living organisms; they function as steroid hormones in mammals, carotenoids in plants, and ubiquinone or menaquinone in bacteria (1). All of these isoprenoids are synthesized by consecutive condensations of the five-carbon precursor isopentenyl diphosphate (IPP) to its isomer dimethylallyl diphosphate (DMAPP). Two distinct pathways for the IPP biosynthesis are known. One is the mevalonate pathway that operates in eukaryotes, archaebacteria, and the cytosols of higher plants. The other is the nonmevalonate pathway (Fig. 1), which is generally used by many eubacteria, including Escherichia coli and Bacillus subtilis, green algae, and the chloroplasts of higher plants (2, 3). The initial step of this pathway is the formation of 1-deoxy-d-xylulose 5-phosphate (DXP) by condensation of pyruvate and glyceraldehyde 3-phosphate catalyzed by DXP synthase (4–7). In the second step, DXP is converted to 2-C-methyl-d-erythritol 4-phosphate (MEP) by DXP reductoisomerase (8–11). MEP is then cytidylylated by MEP cytidylyltransferase to give 4-(cytidine 5′-diphospho)-2-C-methyl-d-erythritol (CDP-ME) (12, 13), which is phosphorylated by CDP-ME kinase to yield 2-phospho-4-(cytidine 5′-diphospho)-2-C-methyl-d-erythritol (CDP-ME2P; refs. 14, 15). Next, CDP-ME2P is converted to 2-C-methyl-d-erythritol 2,4-cyclodiphosphate (MECDP) by the action of MECDP synthase (16, 17). However, the subsequent reactions leading to the formation of IPP from MECDP remain to be elucidated. Recently, Boronat and coworkers (18) demonstrated that IPP and DMAPP were independently biosynthesized in the nonmevalonate pathway.

Figure 1.

IPP biosynthetic pathway in the E. coli transformants. The subsequent reactions leading to the formation of IPP or DMAPP from MECDP remain to be elucidated. X, unidentified biosynthetic intermediate of the nonmevalonate pathway.

IPP isomerase (EC 5.3.3.2) catalyzes an essential reaction in the biosynthesis of isoprenoids by converting IPP to DMAPP (Fig. 1; ref. 19). Many IPP isomerase genes (idi) have been cloned from various organisms such as humans (20), Saccharomyces cerevisiae (21), E. coli (22), and Rhodobacter capsulatus (23). Interestingly, however, database searches using the highly conserved amino acid sequences in IPP isomerases identified to date failed to detect IPP isomerase in archaebacteria and some eubacteria that possess the mevalonate pathway. This failure to detect IPP isomerase in these organisms has suggested the existence of an unusual type of IPP isomerase the amino acid sequence of which shares no similarity to those of IPP isomerases identified to date.

We previously cloned a gene cluster encoding enzymes of the mevalonate pathway from Streptomyces sp. strain CL190 and demonstrated that the mevalonate pathway could function in an E. coli strain transformed with this gene cluster (24, 25). In addition to the known genes of the mevalonate pathway, this gene cluster contains an additional ORF, orfD, encoding a protein with unknown function (25). This orfD has recently been detected in some archaebacteria, some Gram-positive bacteria, including Staphylococcus aureus (26), and in the carotenoid biosynthetic gene cluster of Erwinia herbicola Eho10 (27). These facts led us to assume that orfD encodes a previously uncharacterized type of IPP isomerase.

In this paper, we demonstrate that the orfD product expressed in E. coli catalyzes the isomerization of IPP to produce DMAPP in the presence of both flavin mononucleotide (FMN) and NADPH. Based on the enzymatic properties of this enzyme, which are different from those of known IPP isomerases, we propose that this orfD product is an unusual type of IPP isomerase the sequence of which has no similarity to those of known IPP isomerases. This paper also describes the identification of this unusual IPP isomerase in archaebacteria and some Gram-positive bacteria such as S. aureus, Streptococcus pneumoniae, Streptococcus pyrogenes, and Enterococcus faecalis. This finding suggests that the IPP isomerase of S. aureus represents an effective target for the screening of anti-staphylococcal drugs.

Materials and Methods

Strains and Plasmids.

We previously cloned a gene cluster containing orfD, and genes encoding 3-hydroxy-3-methylglutaryl CoA (HMG-CoA) synthase, HMG-CoA reductase, mevalonate (MVA) kinase, phosphomevalonate (PMVA) kinase and diphosphomevalonate (DPMVA) decarboxylase from Streptomyces sp. strain CL190 (24, 25). The expression plasmid, pUMV19, contains all of these genes (25). The pUMV19-derived plasmids and E. coli strains used in this study are listed in Table 1. Because E. coli W3110 was not able to maintain the pUC118-derived plasmids, probably because of the high-level expressions of the mevalonate pathway genes from the lacZ promoter in pUC118, a pTTQ18 vector harboring lacI^q, which encodes the lac repressor, was used to introduce these genes into this strain. The pTMV20S contained orfD in addition to genes encoding the three enzymes of the mevalonate pathway, MVA kinase, PMVA kinase, and DPMVA decarboxylase, present in pTMV20SΔM.

Table 1.

Bacterial strains and plasmids used in this study

E. coli strain and plasmid	Description	Reference or source
Strains
E. coli DK223	W3110 dxr∷Kmr	This study
E. coli DK331	W3110 dxr∷Kmr and idi∷Cmr	This study
Plasmids
pTTQ18	cloning vector; lacIq, Ampr	S. Matsuyama (University of Tokyo)
pGEM-11zf(+)aad-1	contains 2.0-kb Spmr cassette	H. Ikeda (Kitasato University)
pTTQ18S	pTTQ18 with 2.0-kb Spmr cassette from pGEM-11zf(+)aad-1 inserted at the ScaI site in the Ampr gene	This study
pUMV19	pUC118 with the 6.8-kb SnaBI-SnaBI fragment (accession no. AB037666)	25
pUMV20	pUMV19 with a deletion of the 0.7-kb Sse8387I-Sse8387I fragment	This study
pUMV20ΔM	pUMV20 with a deletion of the 0.4-kb MluI-MluI fragment	This study
pTMV20S	pTTQ18S with the 6.1-kb XbaI-HindIII fragment from pUMV20	This study
pTMV20SΔM	pTTQ18S with the 5.7-kb XbaI-HindIII fragment from pUMV20ΔM	This study

Construction of the Plasmid for Overexpression of orfD in E. coli.

Based on the entire nucleotide sequence of orfD from Streptomyces sp. strain CL190 (accession no. AB037666), two oligonucleotide primers, 5′-GGGGATCCACCAGCGCCCAACGCAAGGACG-3′ (5′ of orfD) and 5′-GGGGATCCTCGTGTGCTTCCCGTCGTCTGG-3′ (3′ of orfD), including a BamHI restriction site (underlined) were synthesized (Amersham Pharmacia) and used together with total DNA from Streptomyces sp. CL190 to amplify orfD. By using Taq DNA polymerase (Roche Molecular Biochemicals) and the protocol recommended by the supplier, a 1.1-kb fragment was amplified. The PCR fragment was digested with BamHI and cloned into the multicloning site of the expression vector pQE30 (Qiagen, Chatsworth, CA) to give pQCLD41.

Expression and Purification of the Recombinant orfD Product.

E. coli M15 containing pREP4 [neo, lacI] (Qiagen) was used as a host for the expression of orfD. M15(pREP4, pQCLD41) was cultured at 18°C in 1 liter of LB medium containing 25 μg/ml kanamycin (Nacalai Tesque, Kyoto) and 200 μg/ml ampicillin (Sigma) for 6 h with the addition of 2 mM isopropyl β-d-thiogalactopyranoside (IPTG) when an optical density at 660 nm of 0.7 was reached. Cells were harvested by centrifugation and resuspended in Buffer A composed of 100 ml of 0.1 M Tris⋅HCl (pH 8.0) and 5 mM DTT. After sonication, the lysate was centrifuged at 10,000 × g for 20 min and the supernatant retained. The crude extract was applied on a Ni-nitrilotriacetic acid agarose column (Qiagen) previously equilibrated with Buffer A. The resin was washed with 50 mM imidazole in Buffer A and then the protein bound to the resin was eluted with 200 mM imidazole in Buffer A.

Cloning, Expression, and Purification of the Recombinant S. aureus IPP Isomerase.

On the basis of the entire nucleotide sequence of the orfD homolog, which was located in Contig111, from S. aureus (http://www.sanger.ac.uk/Projects/S aureus/), two oligonucleotide primers, 5′-GGGATCCAGTGATTTTCAAAGAGAACAGAG-3′ (5′ of the S. aureus orfD homolog) and 5′-GGGATCCTCCTCGATGTATATTCAAGTTACG-3′ (3′ of the S. aureus orfD homolog), including a BamHI restriction site (underlined) were synthesized and used together with total DNA from S. aureus ATCC25923 to amplify the IPP isomerase gene. Cloning, expression, and purification of the recombinant S. aureus IPP isomerase were done in the same way as for CL190 orfD. An expression plasmid for S. aureus IPP isomerase was named pQSAU39.

Determination of Molecular Mass.

The molecular masses of the overexpressed proteins were estimated by gel filtration using a Superdex 200 (1.6 × 60 cm) column (Amersham Pharmacia), which had been equilibrated with 20 mM sodium phosphate buffer (pH 7.1) containing 0.15 M NaCl. As the standard proteins, ferritin (440 kDa), catalase (232 kDa), aldolase (158 kDa), and BSA (66 kDa) were used.

Reconstitution of the orfD Product with FMN.

The orfD product (2 mg) in a volume of 1 ml was incubated with 100 μM FMN at 4°C for 12 h. After incubation, the protein solution was desalted to remove free FMN by gel filtration using a PD-10 column (Amersham Pharmacia). Absorption spectra of each 0.5 ml of fraction were recorded by using a UV-visible spectrophotometer (U-3210; Hitachi, Tokyo). It was confirmed that the fractions containing protein and those containing free FMN were well separated using this column.

Assay for IPP Isomerase.

We first used the standard assay for IPP isomerase described by Satterwhite (28). This standard assay system is based on the acid lability of DMAPP. The assay mixture consisted of 0.1 M Hepes buffer (pH 7.0), 5 mM MgCl₂, 1 mM DTT, 10 μM [1-¹⁴C]IPP (2.0 GBq/mmol; American Radiolabeled Chemicals, St. Louis) and 40 μM IPP in a final volume of 50 μl. The reaction was initiated by adding enzyme to the assay system followed by incubation at 37°C for 10 min. The reaction was terminated by adding 0.2 ml of 25% concentrated HCl in MeOH and 0.5 ml of H₂O, followed by incubation at 37°C for 10 min. This mixture was then saturated with NaCl and subjected to extraction twice with 0.5-ml aliquots of toluene. The extracts were combined and then dried by adding Na₂SO₄. The radioactivity in the 0.3-ml aliquot of the toluene solution was determined in a scintillation solution (Scintisol AL-1; Dojin, Kumamoto, Japan) by using a scintillation counter (LS 6000LL; Beckman Coulter).

However, all attempts to detect IPP isomerase activity of the orfD product were unsuccessful using the standard assay system described above. We then established a new assay system for the IPP isomerase encoded by orfD, which is also based on the acid lability of DMAPP. The new assay mixture consisted of 0.1 M Hepes buffer (pH 7.0), 5 mM MgCl₂, 1 mM DTT, 10 μM [1-¹⁴C]IPP, 40 μM IPP, 10 μM FMN, and 5 mM NADPH in a final volume of 50 μl. The reaction was initiated by adding enzyme to the assay system followed by incubation routinely at 37°C for 1 min. Further manipulations were the same as those used in the standard assay described above.

Product Analysis by ¹H Nuclear Magnetic Resonance.

The reaction mixture (1 ml) consisting of 0.1 M K₂HPO₄ (pH 7.0), 5 mM MgCl₂, 1 mM DTT, 10 μM FMN, 2.5 mM NADPH, 1 mg of orfD product, and 5 mM IPP was incubated at 37°C for 12 h. After incubation, the reaction mixture was lyophilized, and the resulting residue was resuspended in 99.9% D₂O. The reaction product was then analyzed by using a ¹H NMR spectrometer (A-500; JEOL). As a control, a reaction mixture not containing the orfD product was also analyzed.

pH-Dependent Activity of IPP Isomerase.

The IPP isomerase assay was conducted as described above. Assay solutions consisted of 0.1 M Mes at pH 5.0–6.5, 0.1 M Hepes at pH 6.5–7.5 or 0.1 M Tris⋅HCl at pH 7.0–9.5, 5 mM MgCl₂, 10 μM FMN, 5 mM NADPH, and 1 mM DTT.

Complementation of the Disrupted E. coli idi Gene.

Strain DK331 (Table 1) was separately transformed with pTMV20SΔM and pTMV20S. These transformants were cultured at 37°C in LB liquid medium containing 0.005% MVA, 12.5 μg of kanamycin/ml, 17 μg of chloramphenicol/ml and 100 μg of spectinomycin/ml. As a positive control, DK223(pTMV20SΔM) was cultured at 37°C in LB liquid medium containing 0.005% MVA, 12.5 μg of kanamycin/ml and 100 μg of spectinomycin/ml. Growth of these transformants was monitored by optical density at 660 nm.

Results

Expression and Purification of the Recombinant orfD Product.

To elucidate the in vitro function of the orfD product, orfD was overexpressed in E. coli by using the T7 promoter system and the product was purified to homogeneity as a soluble protein (Fig. 2). SDS/PAGE showed a subunit molecular mass of 37 kDa. Native PAGE gave a single protein band with a mobility corresponding to 150 kDa (Fig. 2). The apparent molecular mass of the orfD product was estimated to be 155 kDa by Superdex 200 gel filtration. These data indicate that the orfD product is most likely to be a tetramer. After purification, the protein solution was yellow, indicating that a chromophore such as flavin was present. Since it was possible that this chromophore might have been partially removed during purification, reconstitution of the protein with FMN was done. The absorption spectrum of the resulting solution (Fig. 3) demonstrates that the orfD product is a typical flavoprotein. Based on the extinction coefficient for free FMN (12.2 mM⁻¹⋅cm⁻¹ at 450 nm), the amount of FMN per tetramer was determined to be 1.4 to 1.6 mol/mol.

Figure 2.

Expression and purification of the recombinant orfD product. (A) SDS/PAGE of the enzyme. Lanes: 1, low molecular mass markers; 2, the pellet of the extract from E. coli M15 (pREP4, pQCLD41) culture after induction with isopropyl β-d-thiogalactopyranoside (IPTG); 3, the supernatant of the same extract; 4, proteins passed through an Ni-nitrilotriacetic acid agarose column; 5, the fraction eluted with 50 mM imidazole; 6, purified enzyme (1 μg) eluted with 200 mM imidazole. (B) Native PAGE of the purified enzyme. Lanes: 1, high molecular mass markers; 2, purified enzyme (0.7 μg).

Figure 3.

Absorption spectrum of the recombinant orfD product reconstituted with FMN. A fraction containing the highest concentration of protein (2.8 mg/ml) is shown.

FMN and NADPH Dependence of IPP Isomerase Encoded by orfD.

We next determined whether the reconstituted orfD product had IPP isomerase activity. Initial efforts to detect the activity were carried out by using the standard method described by Satterwhite (28). However, these attempts were totally unsuccessful. This raised the possibility that the orfD product might require some cofactors. Thus, we added various cofactors into the standard assay mixture. A high level of radioactivity was detected in the toluene extract (see Materials and Methods) when both FMN and NADPH were added to the reaction mixture. This radioactivity could have resulted from DMAPP, suggesting that the orfD product catalyzed the isomerization of IPP to produce DMAPP. As described below, structure analysis of the reaction product by ¹H NMR spectroscopy proved that the orfD product converted IPP to DMAPP in the presence of FMN and NADPH.

Table 2 shows the FMN and NAD(P)H dependence of IPP isomerase encoded by orfD. Without addition of FMN and NAD(P)H, no activity was detected, demonstrating that IPP isomerization catalyzed by the orfD product was strictly dependent on both FMN and NAD(P)H. The activity was maximal when 10 μM FMN and 5 mM NADPH were added. The activity was slightly below maximal when 10 μM FMN and 5 mM NADH were used. However, replacement of FMN with FAD decreased the reaction rate to 51% of the original rate. Replacement of FMN with riboflavin resulted in the loss of IPP isomerase activity. Replacement of NAD(P)H with NAD or NADP also resulted in the loss of IPP isomerase activity. The FMN- and NAD(P)H-dependence of the orfD product was unchanged under anaerobic conditions. The enzyme also required Mg²⁺, Mn²⁺, or Ca²⁺. The activity was maximal with Mg²⁺, but lower with Mn²⁺ and Ca²⁺. No effects of other divalent cations such as Co²⁺, Cu²⁺, Ni²⁺, and Zn²⁺ at 5 mM were detected. Addition of 5 mM EDTA resulted in the almost complete loss of enzyme activity.

Table 2.

FMN and NAD(P)H dependence of the activity of the reconstituted IPP isomerase

Additions to reaction buffer*
FMN or FAD, μM	NAD(P)H, mM	Relative activity, %
FMN	NADPH
0	0	ND
0	0.1	ND
0	0.5	3.6
0	1	6.5
0	5	19
0	10	15
0	20	11
10	0.1	36
10	1	95
10	5	100
10	10	90
10	20	96
	NADH
10	1	92
10	5	97
FAD	NADPH
10	5	51

^*The basal reaction buffer consisted of 0.1 M HEPES (pH 7.0), 2.5 mM MgCl₂, 1 mM DTT, 5 μg enzyme. ND, not detected. 

The IPP isomerase encoded by orfD was most active in Hepes buffer. The optimum activity of the enzyme occurred at approximately pH 7.0. The effect of temperature on the enzyme activity was investigated over the range of 15–55°C, with the maximum activity being observed at approximately 35–37°C.

Product Analysis by ¹H NMR.

The reaction catalyzed by the orfD product was analyzed by incubating 5 mM IPP with 1 mg orfD product in phosphate buffer containing 10 μM FMN and 2.5 mM NADPH. After 12 h of incubation, conversion of IPP to DMAPP was observed as evidenced by the appearance of additional signals for DMAPP at δ 5.25 (t, J = 6.4 Hz, 1H, H-2), 4.27 (d, J_H–H = 6.4, J_H–P = 6.1, 2H, H-1), 1.57 (s, 3H, 5-methyl), and 1.53 (s, 3H, 4-methyl) ppm and the concomitant decrease in the intensities of the signals for IPP at δ 4.68 (s, 1H, H-4a), 4.65 (s, 1H, H-4b), 3.87 (dt, J_H–H = 6.4, J_H–P = 1.7, 2H, H-1), 2.21 (t, J = 6.4, 2H, H-2), and 1.58 (s, 3H, 5-methyl) ppm (Fig. 4). These ¹H NMR spectral features were in complete agreement with those of authentic IPP and DMAPP, the spectral data of which were newly collected in this study by using a 500 MHz ¹H NMR spectrometer. This result unambiguously established that the orfD product catalyzed the isomerization of IPP to produce DMAPP. Thus, orfD encodes an FMN- and NAD(P)H-dependent IPP isomerase. The isomerization reaction was reversible, because the enzyme also converted DMAPP to IPP in the presence of FMN and NADPH (data not shown).

Figure 4.

1H NMR spectra of the reaction mixtures. The reaction buffer contained 5 mM MgCl₂, 1 mM DTT, 10 μM FMN, 2.5 mM NADPH, and 5 mM IPP. The spectra were obtained after incubation for 12 h at 37°C without (B) and with (A) the orfD product (1 mg). Small signals in ≈3.9–4.8 ppm can be ascribed to NADPH.

The equilibrium ratio of IPP to DMAPP in the reaction mixture was estimated to be 1.4 by comparing the integral of the IPP H-1 signal (2H) at δ 3.87 ppm with that of the DMAPP H-2 signal (1H) at δ 5.25 ppm.

Complementation of the Disrupted E. coli IPP Isomerase Gene.

We next determined whether orfD could complement the disrupted idi gene in E. coli. Poulter and coworkers (22) reported that disruption of the idi gene was not lethal to E. coli; this phenomenon is explained by the independent formation of IPP and DMAPP in the nonmevalonate pathway without involvement of IPP isomerase (Fig. 1; ref. 18). Therefore, to determine whether orfD could complement the disrupted idi gene, it was essential to construct E. coli disruptants defective in the nonmevalonate pathway in addition to having a disruption of the idi gene. We simultaneously disrupted the idi gene and the dxr gene, which encodes DXP reductoisomerase in the nonmevalonate pathway (Fig. 1), to prepare a new disruptant named DK331 (Table 1). Despite the idi and dxr disruptions, strain DK331 grew in the presence of 2-C-methylerythritol (ME), a biological equivalent of 2-C-methylerythritol 4-phosphate that is an intermediate of the nonmevalonate pathway (Fig. 1). DK331 was then separately transformed with pTMV20SΔM and pTMV20S. The pTMV20S contained orfD in addition to genes encoding the three enzymes of the mevalonate pathway, MVA kinase, PMVA kinase, and DPMVA decarboxylase, present in pTMV20SΔM.

Having constructed the transformants DK331(pTMV20S) and DK331(pTMV20SΔM), we examined their ability to grow in the absence of ME by using DK223(pTMV20SΔM) as a control strain possessing the endogenous idi gene. If the orfD product could function as IPP isomerase in DK331, the transformant DK331(pTMV20S) should grow even in the absence of ME. As expected, DK331(pTMV20SΔM) did not grow in the absence of ME. On the other hand, DK331(pTMV20S) and DK223(pTMV20SΔM) grew in the absence of ME. The growth of DK331(pTMV20S) is explained by its ability to synthesize both IPP and DMAPP from MVA added into the growth medium. Thus, in DK331(pTMV20S), IPP was concluded to be synthesized by the actions of MVA kinase, PMVA kinase, and DPMVA decarboxylase encoded on pTMV20S, and then IPP was isomerized by the orfD product to give DMAPP. On the other hand, in DK223(pTMV20SΔM), IPP was synthesized as in DK331(pTMV20S) and then isomerized by the endogenous idi gene product to give DMAPP. These results demonstrated that the orfD product functioned as IPP isomerase in vivo and had sufficient IPP isomerase activity to complement the disrupted E. coli idi gene.

Streptomyces IPP Isomerase Homologs are Widespread in Archaebacteria and Gram-Positive Bacteria.

Homology searches using the sequence of the Streptomyces IPP isomerase as a query sequence was performed by using the protein databases SWISS-PROT, of the DNA Data Bank of Japan (DDBJ), and Microbial Genomes Blast, of the National Center for Biotechnology Information (NCBI). Fig. 5 shows the multiple alignment of the amino acid sequences of the Streptomyces IPP isomerase and proteins with unknown functions from six organisms, E. herbicola, Synechocystis sp. strain PCC6803, S. aureus, E. faecalis, Methanococcus jannaschii, and Sulfolobus solfataricus. The Streptomyces IPP isomerase showed 29% to 46% identity to these six hypothetical proteins. Although only six homologs are shown in Fig. 5, this search revealed that orfD homologs also existed as proteins with unknown functions in many bacteria such as Borrelia burgdorferi, Rickettsia prowazekii, B. subtilis, S. pneumoniae, S. pyrogenes, Deinococcus radiodurans, Halobacterium sp., Methanobacterium thermoautotrophicum, Archaeoglobus fulgidus, Aeropyrum pernix, Pyrococcus abyssi, and Leishmania major.

Figure 5.

Multiple alignment of the amino acid sequences of orfD and other orfD homologs. Identical amino acids among six proteins are marked by asterisks. Dashes indicate gaps introduced for the optimization of the alignment. The thick lines represent glycine-rich motifs. orfD, Streptomyces sp. strain CL190; Erwi, E. herbicola Eho10 (accession no. Q01335); Syne, Synechocystis sp. strain PCC6803 (accession no. P74287); Saur, S. aureus ATCC25923; Enfa, E. faecalis (http://www.tigr.org/); Meta, M. jannaschii (accession no. Q58272); Sulf, S. solfataricus (accession no. P95997).

Biochemical Properties of S. aureus IPP Isomerase.

Because the database search identified the S. aureus IPP isomerase, this recombinant enzyme expressed in E. coli was also purified and characterized. SDS/PAGE showed a subunit molecular mass of 39 kDa. Native PAGE gave a single protein band with a mobility corresponding to 150 kDa. The apparent molecular mass was estimated to be 155 kDa by gel filtration. These data indicate that the S. aureus IPP isomerase is also most likely to be a tetramer. This enzyme also catalyzed the isomerization of IPP to produce DMAPP in the presence of FMN and NADPH.

Kinetic Parameters of Streptomyces and S. aureus IPP Isomerases.

The kinetic parameters of the Streptomyces and S. aureus IPP isomerases were calculated and compared with those of FMN- and NADPH-independent IPP isomerases from E. coli, S. cerevisiae, and humans (Table 3). The Streptomyces isomerase with a k_cat of 0.70 showed less activity than the human and S. cerevisiae enzymes with k_cat values of 1.8 and 8.0 s⁻¹, respectively. On the other hand, the Streptomyces enzyme showed almost the same activity as that of the E. coli isomerase. However, the K_m value of 450 μM for IPP of the Streptomyces isomerase was 10- to 50-fold higher than those of other isomerases. The S. aureus isomerase has the same catalytic efficiency (k_cat/K_m) of 6.8 × 10⁴ M⁻¹⋅s⁻¹ as those of the three FMN- and NADPH-independent IPP isomerases.

Table 3.

Comparison of enzymatic properties of IPP isomerases

	Monomer, kDa	Km (for IPP), μM	kcat (for IPP), s−1	kcat/Km, M−1⋅s−1	Divalent cation	FMN-dependence	NAD(P)H-dependence	Reference
Streptomyces sp. CL190	41	450	0.70	1.6  × 103	Mg2+	yes	yes	This study
S. aureus	39	19	1.3	6.8  × 104	Mg2+	yes	yes	This study
E. coli	25	7.9	0.33	4.2  × 104	Mg2+	no	no	22
S. cerevisiae	26	43	8.0	1.9  × 105	Mg2+	no	no	21
Human	39	33	1.8	5.5  × 104	Mg2+	no	no	20

Discussion

We have demonstrated in this study that orfD in the mevalonate pathway gene cluster of Streptomyces sp. strain CL190 encoded an unusual type of IPP isomerase with no sequence similarity to other IPP isomerases identified to date. All of the usual IPP isomerases characterized to date have a conserved cysteine residue (boldface) in the NXXCXHP motif and a conserved glutamate residue (boldface) in the EXE motif (22). Neither of the two motifs, however, were found in the type of IPP isomerases characterized in this study. Instead of the motifs, three glycine-rich sequences, MTGG, GXGGT, and (A/G)SGG, were conserved in the new type of IPP isomerases (Fig. 5). Interestingly, these new enzymes required both FMN and NAD(P)H for enzyme reaction. Thus, orfD encodes an FMN- and NAD(P)H-dependent IPP isomerase. Therefore, the glycine-rich sequences may be associated with the binding of these cofactors. Based on the difference in enzymatic properties, we classify the IPP isomerases into two types: Type 2 for FMN- and NAD(P)H-dependent enzymes, and type 1 for the others. We renamed orfD the fni gene for FMN- and NAD(P)H-dependent IPP isomerase.

IPP isomerase encoded by fni was a flavoenzyme containing FMN, as evidenced by the UV/visible spectral feature (Fig. 3). Flavoenzymes are classified into six categories based on the nature of the electron acceptors and donors; dehydrogenases, oxidases, oxidase-decarboxylases, monooxygenases, dioxygenases, metalloflavoenzymes, and flavodoxins (29). However, type 2 IPP isomerases do not belong to any of these categories. How do the flavoenzymes such as the fni product catalyze the isomerization of IPP to produce DMAPP? In type 1 IPP isomerases, the isomerization of IPP proceeds via a carbocationic intermediate (30). In this mechanism, an electrophilic attack by a proton from water occurs on the IPP double bond to yield the carbocationic intermediate and then the C-2 pro-R hydrogen of the intermediate is stereospecifically eliminated (31). In a similar reaction mechanism, the carbocationic intermediate may be involved in the reaction catalyzed by type 2 IPP isomerases. Further investigations are needed to clarify this hypothetical reaction mechanism of type 2 IPP isomerases.

Database searches identified a type 2 IPP isomerase in S. aureus, a pathogenic bacterium that is notorious as a methicillin resistant S. aureus (MRSA). We cloned and overexpressed this S. aureus IPP isomerase in E. coli and demonstrated that the recombinant enzyme had FMN- and NAD(P)H-dependent IPP isomerase activity. The essentiality of the type 2 IPP isomerase for S. aureus growth and the difference in the reaction mechanism from that of the mammalian type 1 IPP isomerase identify the type 2 enzyme as a promising target for anti-staphylococcal drugs. S. pneumoniae, S. pyrogenes, and E. faecalis IPP isomerases are also considered to be molecular targets for the screening of antimicrobials specific to these Gram-positive bacteria.

Abbreviations

IPP

isopentenyl diphosphate

DMAPP

dimethylallyl diphosphate

DXP

1-deoxy-d-xylulose 5-phosphate

ME

2-C-methylerythritol

CDP-ME

4-(cytidine 5′-diphospho)-2-C-methyl-d-erythritol

CDP-ME2P

2-phospho-4-(cytidine 5′-diphospho)-2-C-methyl-d-erythritol

MECDP

2-C-methyl-d-erythritol 2,4-cyclodiphosphate

MVA

mevalonate

PMVA

phosphomevalonate

DPMVA

diphosphomevalonate