Specificity of geranylgeranyl diphosphate synthase for homoallylic substrate analogs

Article abstract of DOI:10.1016/j.molcatb.2015.07.006

The goal of this study was to determine the substrate specificity of Homo sapiens geranylgeranyl diphosphate synthase (GGPPase) for analogs of isopentenyl diphosphate (IPP) to facilitate the application to organic synthesis techniques to the study of prenyl chain elongation enzymes. For this purpose, we used the IPP analogs 2a-d, which contain different alkyl side-chains at the 3-position, as substrates of the condensation reaction with the allylic substrate geranyl diphosphate (GPP) that is catalyzed by GGPPase. GGPPase catalyzed the reaction of GPP with 3-desmethylisopentenyl diphosphate (but-3-enyl diphosphate) to yield 3-desmethylfarnesyl diphosphate (12.1%), as well as the reaction of GPP with 3-ethylbut-3-enyl diphosphate or 3-propylbut-3-enyl diphosphate to yield 3-ethylfarnesyl diphosphate (46.9%) or 3-propylfarnesyl diphosphate (22.6%), respectively. However, a reaction product was not detected when 3-butylbut-3-enyl diphosphate was used as substrate.

Full text of DOI:10.1016/j.molcatb.2015.07.006

Journal of Molecular Catalysis B: Enzymatic 120 (2015) 179–182
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Specificity of geranylgeranyl diphosphate synthase for homoallylic
substrate analogs
Norimasa Ohyaa, Takumi Ichijob, Hana Satoa, Takeshi Nakamuraa, Saki Yokotac,
Hiroshi Sagamid, Masahiko Nagakie,∗
a Department of Material and Biological Chemistry, Faculty of Science, Yamagata University, Kojirakawa-machi, Yamagata 990-8560, Japan
b Graduate School of Science and Technology, Hirosaki University, 3 Bunkyo-cho, Hirosaki, Aomori 036-8561, Japan
c Graduate School of Engineering and Resource Science Department of Applied Chemistry, Akita University, 1-1 Tegata gakuen-machi, Akita-shi, Akita
010-8502, Japan
d Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, Aoba-ku, Sendai, Miyagi 980-8577, Japan
e Department of Nursing, School of Health Sciences, Hirosaki University of Health and Welfare, 3-18-1 Sanpinai, Hirosaki, Aomori 036-8102, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 2 May 2015
Received in revised form 20 June 2015
Accepted 12 July 2015
Available online 23 July 2015
The goal of this study was to determine the substrate specificity of Homo sapiens geranylgeranyl diphos-
phate synthase (GGPPase) for analogs of isopentenyl diphosphate (IPP) to facilitate the application to
organic synthesis techniques to the study of prenyl chain elongation enzymes. For this purpose, we
used the IPP analogs 2a–d, which contain different alkyl side-chains at the 3-position, as substrates
of the condensation reaction with the allylic substrate geranyl diphosphate (GPP) that is catalyzed by
GGPPase. GGPPase catalyzed the reaction of GPP with 3-desmethylisopentenyl diphosphate (but-3-enyl
diphosphate) to yield 3-desmethylfarnesyl diphosphate (12.1%), as well as the reaction of GPP with 3-
ethylbut-3-enyl diphosphate or 3-propylbut-3-enyl diphosphate to yield 3-ethylfarnesyl diphosphate
(46.9%) or 3-propylfarnesyl diphosphate (22.6%), respectively. However, a reaction product was not
detected when 3-butylbut-3-enyl diphosphate was used as substrate.
Keywords:
Prenyl chain elongating enzyme
Geranylgeranyl diphosphate synthase
Substrate specificity
Isopentenyl diphosphate analogs
Docking simulation
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
thase (FPPase) for artificial substrate analogs [7–12]. For example,
we targeted geranylgeranyl diphosphate synthase (GGPPase) [EC
Approximately 40,000 isoprenoids (Scheme 1) are widely dis-
tributed among organisms. For example, prenyl diphosphate is
a precursor in the biosynthetic pathways that produce steroids,
carotenoids, dolichols, prenyl quinones, prenyl proteins, and nat-
ural rubber. Biosynthesis of these prenyl diphosphate polymers
occurs through consecutive condensations of isopentenyl diphos-
phate (IPP) with its isomer dimethylallyl diphosphate (DMAPP)
or with geranyl diphosphate (GPP). The stereospecific “head-to-
tail” condensations proceed to form polymers of specific lengths,
and prenyltransferases catalyze these reactions [1–6]. Because
prenyltransferases exhibit substrate selectivity under mild reaction
conditions, they will likely be useful for organic synthesis.
2.5.1.29], which catalyzes the early steps in biosynthetic pathways
of the products described above as well as those of catabolic path-
ways, in concert with FPPase [13–15].
GGPPase catalyzes the sequential condensation of IPPs with
the allylic substrates dimethylallyl diphosphate (DMAPP), geranyl
diphosphate (GPP), or both to produce geranylgeranyl diphosphate
(GGPP) [16,17]. An aspartate (D) residue-rich motif within E-type
prenyl chain elongating enzymes (FPPase or GGPPase) may reside
within the active sites of the enzymes [18–20]. The predicted
primary structures of FPPase and GGPPase are similar [21–23],
and they react with the same substrate. However, the detailed
mechanism of the reaction catalyzed by GGPPase remains to be
determined.
In order to review in which position of the active site of GGPPase
the reaction product exists, we performed the docking simulations.
Here, we examined the reactions catalyzed by GGPPase using 3-
alkyl analogs of IPP as substrates and conducted enzyme-substrate
docking simulations.
To exploit prenyltransferases for this purpose, our research has
focused on the substrate specificity of farnesyl diphosphate syn-
∗
Corresponding author. Tel.: +81 0172 27 1001; fax: +81 172 27 1023.
E-mail address: nagakim@jyoto-gakuen.ac.jp (M. Nagaki).
http://dx.doi.org/10.1016/j.molcatb.2015.07.006
1381-1177/© 2015 Elsevier B.V. All rights reserved.

180
N. Ohya et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 179–182
Scheme 1. Backbone subunits and bioactive isoprenoids.
2. Experimental
reaction to yield 3-ethylbut-3-en-1-ol after deprotection. The cog-
nate alcohols were diphosphorylated [19].
2.1. Analytical procedures
The prenyl alcohols produced by treating enzyme reaction prod-
ucts with alkaline phosphatase (Grade 1 from calf intestine, Roche)
were measured using high-performance liquid chromatography
(HPLC). Relative yields of the GGPPase reactions were calculated
relative to those of GGPP produced by the standard enzymatic reac-
tion of IPP with GPP. We used a LichroCART column (Merck-Japan,
Tokyo, Japan) with a hexane:2-propanol (40:1) eluent for HPLC.
Reaction products were identified using liquid chromatography-
mass spectrometry (LC–MS) (Hitachi NanoFrontier LD, Hitachi
High Technologies, Tokyo, Japan). All HPLC analyses were per-
formed using a LaChrom Elite HPLC system (Hitachi L-7420, a
Hitachi L-6000 pump and UV–vis spectroscopy detector), an Ulvac
Chromato-DAQ II A/D converter (Ulvac Inc., Chigasaki, Japan), and
a Merck LiChroCART column (Merck KGoA, Darmstadt, Germany).
GC–MS analysis was used to identify the prenyl alcohols derived
from alkaline phosphatase treatment of the reaction products. The
analytical system comprised a JEOL Q 1000GC-Md II Mass Spec-
trometer (Japan Electron Optics Laboratory Co. Ltd., Tokyo, Japan)
with an HP-5 column (30 m × 0.32 mmϕ). A 1 ml/min helium car-
rier flow rate was used. The temperature gradient started with a
4.7-min hold at 50 ◦C, increased by 15 ◦C/min, and ended with a
2-min hold at 250 ◦C. Injector and GC-transfer-line temperatures
were 200 ◦C.
2.3. Purification of GGPPases
Purification of human GGPPase was carried out according
to a published method [23]. An expression plasmid obtained
from Dr. H. Sagami of Tohoku University was used to trans-
form Escherichia coli BL21 (DE3). Transformants were cultivated
at 37 ◦C to an absorbance at 600 nm of 0.6. Isopropyl 1-thio-␤-
galactopyranoside was added to a final concentration of 1 mM, and
then the cells were cultured at 37 ◦C overnight. The cultures were
centrifuged, and the pelleted cells were suspended in lysis buffer
(20 mM Tris–HCl, pH 7.9, 5 mM EDTA, and 1 mM phenylmethyl-
sulfonyl fluoride) and were disrupted by sonication. His-tagged
GGPPase was purified from the supernatant of the lysate using a
His GraviTrap Column (GE Healthcare), and the fractions were sub-
jected to SDS-PAGE. Coomassie Brilliant Blue staining revealed that
the final preparation was 90% pure.
2.4. Enzyme assays
The reaction mixture for human GGPPase contained 125 ␮mol
Tris–HCl buffer (pH 7.4), 50 ␮ mol ␤-mercaptoethanol, 25 ␮ mol
MgCl2, 1 ␮mol of each substrate analog (GPP 1 and IPP 2, 3-alkyl
analogs of IPP, 2a–e), and GGPPase (approximately 128 ␮g) in a
total volume of 1.0 ml. After incubation for 6 h at 37 ◦C, the reaction
mixture was treated with alkaline phosphatase for 12 h, extracted
with pentane, and analyzed using HPLC and GC–MS [8–10].
2.2. Synthetic reactions
The 3-alkyl analogs of IPP, but-3-enyl- (2b); 3-ethylbut-3-enyl-
(2c); 3-propylbut-3-enyl- (2d); and 3-butylbut-3-enyl diphos-
phates (2e), were synthesized according to published methods
[7–10,24,25]. 3-DesmethylIPP was obtained by diphosphorylation
using the method of Poulter et al. [26], after tosylating 3-buten-1-
ol. 3-Ethyl-, 3-propyl-, and 3-butylbut-3-enyl diphosphates were
synthesized according to a published method [9,10]. Briefly,
3-ethylbut-3-enyl diphsphate was synthesized as follows:2-
(3-chloromethylbut-3-enoyloxy)-tetrahydro-2H-pyran was gen-
erated by selenium oxidation and chlorination of isopentenyl
tetrahydropyranyl ether, which was methylated using a Grignard
2.5. Enzyme-substrate docking simulations
The crystal structure of human GGPPase (PDB code: 2Q80PDB)
was used for docking studies. Genetic Optimization of Ligand Dock-
ing (GOLD) [27] was employed to dock flexible ligands in silico to
the binding sites of the enzymes to determine the possible bind-
ing mechanisms. Enzyme-substrate docking was visualized using
PyMOL [28].

N. Ohya et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 179–182
181
Scheme 2. Reaction of geranyl diphosphate (GPP) with isopentenyl diphosphate (IPP) catalyzed by GGPPase.
Scheme 3. Reactions between allylic substrate (GPP 1) and homoallylic analogs (2a–2d) catalyzed by GGPPase.
3. Results
GGPPase catalyzes consecutive condensation reactions of GPP
to (all-E)-3-ethyl-7,11-dimethyldodeca-2,6,10-trien-1-ol (3b-OH),
considering the stereochemistry of the reaction. Relative yield was
49.6%.
(1) as its natural substrate with two molecules of IPP (2) via farne-
syl diphosphate (3) to yield the final product (all-E)-geranylgeranyl
diphosphate (GGPP) (4) (Scheme 2). To investigate the substrate
specificity of human GGPPase for IPP analogs with different alkyl
chains at the 3-position, we examined a series of analogs (2a–d)
in the condensation reaction with GPP as the allylic substrate
(Scheme 3).
3.3. Reaction of GPP (1) with 3-propylbut-3-enyl diphosphate
(2c) (or 3-butylbut-3-enyl diphosphate (2d))
The GGPPase-catalyzed reaction of GPP with (2c) produced alco-
hols that eluted from the HPLC column as a peak at 14.0 min.
GC–MS analysis of the HPLC-purified alcohol revealed a molec-
ular ion [M]+ at m/z 250 (0.1%), corresponding to C17H30O,
as well as other major ions at m/z 232 [M − 18]+ (6.3%), 217
[M − 18–15] + (1.6%), 203 [M − 18–15–14] + (2.9%), 163 [M − 18–69]
3.1. Reaction of GPP (1) with but-3-enyl diphosphate (2a)
catalyzed by GGPPase
+
+
+
(2.9), 148 [M − 18–69–15]
(4.9%), 134 [M − 18–69–15–14]
+
(1.3%), 69 (base peak), and 66 [M − 18–69–15–14–68]
Reaction of GPP with (2a) catalyzed by human GGPPase fol-
lowed by phosphatase treatment of the reaction mixture produced
an alcohol which was detected using HPLC as a peak that eluted at
17.8 min (GC, 13.5 min). The MS spectrum of the alcohol showed a
molecular ion at m/z 208 (relative intensity, 0.1%), corresponding
to C14H24O, together with fragments at m/z 190 [M − 18]+ (0.6), 121
[M − 18–69]+ (2.7), 69 (base peak), and 53 [M − 18–69–68]+ (3.9).
These data indicated the presence of a norfarnesyl moiety lacking a
methyl group at the 3-position. The alcohol therefore corresponds
to (all-E)-7,11-dimethyldodeca-2,6,10-trien-1-ol (norfarnesol), 3a-
OH, considering the nature of the enzymatic reaction. Accordingly,
the GGPPase-catalyzed reaction of GPP with 2a proceeded through
prenyl chain elongation and terminated after the first condensa-
tion (Scheme 3). The relative yield of 3a-OH was 12.1% of that of
FPP produced by the reaction between GPP and IPP (Table 1)
(1.4%). These data are consistent with the structure 3-propyl-
7,11-dimethyldodeca-2,6,10-trien-1-ol (3-propylfarnesol, 3c-OH).
We therefore designated the product as (all-E)-3-propyl-7,11-
dimethyldodeca-2,6,10-trien-1-ol (3c-OH) by considering the
stereochemistry of the reaction. Relative yield was 22.6%. However,
2d did not react at all as a substrate for GGPPase.
3.4. Docking simulation of the enzymatic reaction
Kinetic parameters are generally used to evaluate enzyme
catalysis, but predicting the substrate preference from these exper-
imental data is difficult. To better define the substrate affinity of
human GGPPase, the final reaction products (GGPP, desmethylFPP,
ethylFPP, or propylFPP) derived from each substrate (IPP, but-
3-enyl, ethylbut-3-enyl, or propylbut-3-enyl diphosphate) were
docked to the active sites of the enzyme using GOLD (Fig.1). The
diphosphate group of the substrate interacted with the DDXXD
motif of the first aspartic acid-rich motif in the GGPPase binding
site, which is required for catalysis.
3.2. Reaction of GPP (1) with 3-ethylbut-3-enyl diphosphate (2b)
The hydrolysate derived from the GGPPase-catalyzed reaction
of GPP with (2b) produced a species that eluted from the HPLC col-
umn at 15.8 min. GC–MS analysis detected a molecular ion [M]+
at m/z 236 (0.5%) corresponding to C16H28O and fragment ions at
m/z 218 [M − 18]+ (4.2%), 203 [M − 18–15] + (1.2%), 149[M − 18–69]
4. Discussion
+
+
(2.7%), 134 [M − 18–15–69]
(4.1), 69 (base peak), and 66
We show here that only desmethylFPP was produced by the
GGPPase-catalyzed reaction of GPP with desmethylIPP. However,
the allylic 3-desmethyl analog did not serve as a substrate. More-
over, norfarnesyl diphosphate (1b) is not an allylic substrate of
[M − 18–15–69–68] +(1.9). These data indicated that the structure
of the alcohol was 3-ethyl-7,11-dimethyldodeca-2,6,10-trien-1-
ol (3-ethylfarnesol, 3b-OH). The product therefore corresponds

182
N. Ohya et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 179–182
Table 1
Relative yields of alcohols produced by the reaction of GPP (1) with 3-alkyl analogs of IPP (2a–d) catalyzed by GGPPase.
Reaction of GPP (1) with IPP analogs (2a–d)
Product (%)
IPP (2)
FOH (3-OH) + GGOH (4-OH)
3-DesmethylFOH (3a-OH)
3-EthylFOH (3b-OH)
3-PropyFOH (3c-OH)
3-ButylFOH (3d-OH)
100
But-3-enylPP** (2a)
12.1
49.6
22.6
n.d.*
3-Ethylbut-3-enylPP** (2b)
3-Propylbut-3-enylPP** (2c)
3-Butylbut-3-enylPP** (2d)
Yields of the alcohols produced by the reactions catalyzed by GGPPase relative to those of FOH and GGOH produced by the corresponding reactions of GPP with IPP.
*
n.d.: not detected.
PP: diphosphate.
**
to compare FPP with butylFPP (Fig. 2), butyl FPP shifted greatly by
one phosphate.
5. Conclusion
The condensation reaction of GPP 1 with 3-desmethyl- 2a,
3-ethyl- 2b, or 3-propybut-3-enyl diphosphate 2c produced nor-
farnesol 3a-OH, ethylfarnesol 3b-OH, or propylfarnesol 3c-OH in a
single step, but did not proceed further. Further, 3-butylbut-3-enyl
diphosphate 2d did not detectably react.
Acknowledgment
This work was supported in part by a grant from the Cosmetol-
ogy Research Foundation (to M. N.) and partially by the President
Specification Research of Hirosaki University of Health and Welfare.
Fig. 1. The arrangement of the DDXXD motifs circumference assumed to reside
within the active site of GGPPase revealed by molecular docking using GOLD. Com-
References
pared with FPP, desmethylFPP formed a smaller gap.
[1] C.D. Poulter, H.C. Rilling, in: J.W. Porter, S.L. Spurgeon (Eds.), Biosynthesis of
Isoprenoid Compounds, Vol. 1, Wiley, New York, 1982, p. 161.
[2] K. Ogura, T. Koyama, Chem. Rev. 98 (1998) 1263.
[3] T. Koyama, K. Ogura, Comprehensive Natural Product Chemistry, Pergamon,
Oxford, 1999, pp. 69.
[4] S.C. Roberts, Production and engineering of terpenoids in plant cell culture,
Nat. Chem. Biol. 3 (2007) 387.
[5] T. Koyama, S. Obata, M. Osabe, A. Takeshita, K. Yokoyama, M. Uchida, T.
Nishino, K. Ogura, J. Biochem. 113 (1993) 355.
[6] H.V. Thulasiram, C.D. Poulter, J. Am. Chem. Soc. 128 (2006) 15819.
[7] M. Nagaki, T. Koyama, T. Nishino, K. Shimizu, Y. Maki, K. Ogura, Chem. Lett.
(1997) 497.
[8] M. Nagaki, H. Kannari, J. Ishibashi, Y. Maki, T. Nishino, K. Ogura, T. Koyama,
Bioorg. Med. Chem. Lett. 8 (1998) 2549.
[9] M. Nagaki, S. Sato, Y. Maki, T. Nishino, T. Koyama, J. Mol. Catal. B: Enzym. 9
(2000) 33.
[10] M. Nagaki, K. Kimura, H. Kimura, Y. Maki, E. Goto, T. Nishino, T. Koyama,
Bioorg. Med. Chem. Lett. 11 (2001) 2157.
[11] M. Nagaki, H. Kanno, T. Musashi, R. Shimizu, Y. Maki, H. Sagami, T. Koyama, J.
Mol. Catal. B: Enzym. 59 (2009) 225.
[12] M. Nagaki, M. Nakada, T. Musashi, J. Kawakami, T. Endo, Y. Maki, T. Koyama, J.
Mol. Catal.B: Enzym. 60 (2009) 186.
[13] B.C. Reed, H.C. Rilling, Biochemistry 14 (1975) 50–54.
[14] H. Sagami, T. Korenaga, K. Ogura, J. Biochem. 114 (1) (1993) 118–121.
[15] M. Ishimoto, S. Minakami, S. Mizushima, T. Oshima, H. Wada, Metabolic Maps
Kyoritu, Publishing Company, Ltd, Tokyo, 1971, pp 68–69.
[16] I. Takahashi, K. Ogura, J. Biochem. 92 (1982) 1527–1537.
[17] H. Sagami, K. Ogura, Methods Enzymol. 110 (1985) 188–192.
[18] T. Kuzuguchi, Y. Morita, I. Sagami, H. Sagami, K. Ogura, J. Biol. Chem. 274
(1999) 5888.
Fig. 2. Comparison of FPP with butylFPP. One of the phosphates of butylFPP pro-
trudes through a large gap.
GGPPase, which is consistent with the substrate specificities of
other prenyl elongating enzymes. Thus, the methyl group (or alkyl
group) at the position-3 is indispensable for catalytic activity [7,10].
Further, we did not expect to find that the reactions did not proceed
after the first stage of condensation of GPP with 3-ethylbut-3-enyl
diphosphate, 3-propylbut-3-enyl diphosphate or IPP. FPP analogs
such as ethylFPP or propylFPP were obtained, and GGPP analogs
such as diethylGGPP (4b) or dipropylGGPP (4c) were not detected.
As shown in Table 1, the relative yields decreased as the bulk
of the alkyl group of the 3rd position of IPP increased, indicating
a significant effect of steric hindrance. Moreover, we reported that
3-butenyl diphosphate is not a substrate for FPP synthase [8–10] or
GGPPase. In the present study, when we used docking simulations
[19] Y. Ye, M. Fujii, A. Hirata, M. Kawamukai, C. Shimoda, T. Nakamura, Mol. Biol.
Cell. 18 (2007) 3568–3581.
[20] D. Tholl, S. Lee, The Arabidopsis Book, 9, e0143 doi: 10.1043/tab.0143 (2011).
[21] H. Sagami, Y. Morita, K. Ogura, J. Biol. Chem. 269 (1994) 20561.
[22] S. Ohnuma, K. Hirooka, C. Ohto, T. Nishino, J. Biol. Chem. 272 (1997) 5192.
[23] Y. Miyagi, Y. Matsumura, H. Sagami, J. Biochem. 142 (2008) 377.
[24] K. Ogura, T. Koyama, S. Seto, J. Chem. Soc. Commun. (1972) 881.
[25] T. Koyama, K. Ogura, S. Seto, J. Biol. Chem. 248 (1973) 8043.
[26] V.J. Daviesson, A.B. Woodside, C.D. Poulter, Methods Enzymol. 110 (1895) 153.
[27] T. Musashi, H. Kanno, J. Kawakami, S. Yokota, N. Ohya, M. Nagaki, Trans.
Mater. Res. Soc. Jpn. 35 (2010) 227.
[28] http://www.pymol.org