Tải bản đầy đủ (.pdf) (16 trang)

Báo cáo Y học: Structure, mechanism and function of prenyltransferases docx

Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (888.43 KB, 16 trang )

REVIEW ARTICLE
Structure, mechanism and function of prenyltransferases
Po-Huang Liang, Tzu-Ping Ko and Andrew H J. Wang
Institute of Biological Chemistry, Academia Sinica, Taipei, Taiwan
In this review, we summarize recent progress in studying
three main classes of prenyltransferases: (a) isoprenyl pyro-
phosphate synthases (IPPSs), which catalyze chain elonga-
tion of allylic pyrophosphate substrates via consecutive
condensation reactions with isopentenyl pyrophosphate
(IPP) to generate linear polymers with defined chain lengths;
(b) protein prenyltransferases, which catalyze the transfer of
an isoprenyl pyrophosphate (e.g. farnesyl pyrophosphate) to
a protein or a peptide; (c) prenyltransferases, which catalyze
the cyclization of isoprenyl pyrophosphates. The prenyl-
transferase products are widely distributed in nature and
serve a variety of important biological functions. The cata-
lytic mechanism deduced from the 3D structure and other
biochemical studies of these prenyltransferases as well as
how the protein functions are related to their reaction
mechanism and structure are discussed. In the IPPS reaction,
we focus on the mechanism that controls product chain
length and the reaction kinetics of IPP condensation in the
cis-type and trans-type enzymes. For protein prenyltrans-
ferases, the structures of Ras farnesyltransferase and Rab
geranylgeranyltransferase are used to elucidate the reaction
mechanism of this group of enzymes. For the enzymes
involved in cyclic terpene biosynthesis, the structures and
mechanisms of squalene cyclase, 5-epi-aristolochene syn-
thase, pentalenene synthase, and trichodiene synthase are
summarized.
Keywords: chain elongation; isoprenoid; lipid carrier;


prenyltransferase; site-directed mutagenesis; 3D structure.
Isoprenoids are an extensive group of natural products with
diverse structures consisting of various numbers of five-
carbon isopentenyl pyrophosphate (IPP) units [1,2]. The
more than 23 000 isoprenoid compounds identified thus far
serve a variety of essential biological functions in Eukarya,
Bacteria and Archaea. For example, steroids are cyclic
isoprenoids, which have distinct biological functions as
hormones [3]. Carotenoids contain highly conjugated
structures for absorption of light and are the most common
accessory pigments in all green plants and many photosyn-
thetic bacteria [4]. Retinoids are involved in morphogenesis
and are the light-sensitive element in vision. Prenylated
proteins including Ras and other G-proteins are involved in
specific signal-transduction pathways [5,6].
Many linear isoprenoids, generally synthesized from C
15
farnesyl pyrophosphate (FPP) and IPP, are found in nature
[7]. The enzymes responsible for the synthesis of linear
isoprenyl pyrophosphates can be classified as cis-andtrans-
isoprenyl pyrophosphate synthase (IPPS) according to the
stereochemical outcome of their products [8]. As shown in
Fig. 1, all-trans-geranyl pyrophosphate (GPP) and FPP are
synthesized by trans-type geranyl pyrophosphate synthase
(GPPS) and farnesyl pyrophosphate synthase (FPPS),
respectively. Starting from FPP, a variety of different-
chain-length products are generated by the corresponding
synthases. The geranylgeranyl pyrophosphate synthase
(GGPPS) and farnesylgeranyl pyrophosphate synthase
(FGPPS) produce C

20
and C
25
all-trans-polyprenyl pyro-
phosphates to make C
20
–C
20
and C
20
-C
25
ether-linked lipids
in archeon [9–11]. The C
40
product of octaprenyl pyro-
phosphate synthase (OPPS) constitutes the side chain of
ubiquinone in Escherichia coli [12–14]. Several cis-isoprenyl
Correspondence to P H. Liang, Institute of Biological Chemistry, Academia Sinica, Taipei 11529, Taiwan.
Fax: +886 2 2788 9759, Tel.: +886 2 2785 5696 ext. 6070, E-mail: or A.H J. Wang, Fax: +886 2788 2043,
E-mail:
Abbreviations: FPP, farnesyl pyrophosphate; GPP, geranyl pyrophosphate; GGPP, geranylgeranyl pyrophosphate; IPP, isopentenyl pyrophos-
phate; UPP, undecaprenyl pyrophosphate; IPPS, isoprenyl pyrophosphate synthase; UPPS, undecaprenyl pyrophosphate synthase; DDPPS,
dehydrodolichyl pyrophosphate synthase; PPPS, polyprenyl pyrophosphate synthase; GPPS, geranyl pyrophosphate synthase; FPPS, farnesyl
pyrophosphate synthase; GGPPS, geranylgeranyl pyrophosphate synthase; FGPPS, farnesylgeranyl pyrophosphate synthase; HexPPS, hexa-
prenyl pyrophosphate synthase; HepPPS, heptaprenyl pyrophosphate synthase; OPPS, octaprenyl pyrophosphate synthase; SPPS, solanesyl
pyrophosphate synthase; DPPS, decaprenyl pyrophosphate synthase; FTase, farnesyltransferases; GGTase, geranylgeranyltransferase.
Enzymes: UPPS from Escherichia coli (EC 2.5.1.31); DDPPS from yeast Saccharomyces cerevisiae (EC 2.5.1.31); PPPS from Arabidopsis thaliana
(EC 2.5.1.31); FPPS from E. coli (EC 2.5.1.10); GGPPS from yeast (EC 2.5.1.29); FGPPS from Aeropyrum pernix (EC 2.5.1.33); HexPPS from
Bacillus stearothermophilus and yeast (EC 2.5.1.30); HepPPS from Mycobacterium tuberculosis (EC 2.5.1.30); OPPS from E. coli (EC 2.5.1.11);

SPPS from Rhodobacter capsulatus (EC 2.5.1.11); DDPPS from human (EC 2.5.1.31); Ras farnesyltransferase from human (EC 2.1.1.100); Rab
geranylgeranyltransferase from human (EC 2.1.1.100); squalene cyclase from Alicyclobacillus acidocaldarius (2.5.1.31); 5-epi-aristolochene synthase
from Tobacco (EC 4.2.3.6); pentalenene synthase from Streptomyces UC5319 (EC 4.2.3.7); trichodiene synthase from Fusarium sporotrichioides
(EC 4.2.3.6).
(Received 22 March 2002, accepted 17 May 2002)
Eur. J. Biochem. 269, 3339–3354 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03014.x
pyrophosphate synthases including solanesyl pyrophos-
phate synthase (SPPS) [15–17], decaprenyl pyrophosphate
synthase (DPPS) [18], heptaprenyl pyrophosphate synthase
(HepPPS) [19], and hexaprenyl pyrophosphate synthase
(HexPPS) [20,21] are responsible for making C
45
,C
50
,C
35
,
and C
30
side chains, respectively, of ubiquinone in different
species [22].
Among the cis-polyprenyl pyrophosphates, the C
55
product of the bacterial undecaprenyl pyrophosphate
synthase (UPPS) serves as a lipid carrier in cell wall
peptidoglycan biosynthesis [23,24]. Its homologous dehy-
drodolichyl pyrophosphate synthase (DDPPS) in eukaryo-
tes is responsible for making C
55
–C

100
dolichols for
glycoprotein biosynthesis, a pathway similar to that of the
bacterial peptidoglycan synthesis [25,26]. An even longer
C
120
polymer was found as the final product of an isoprenyl
pyrophosphate synthase in plant Arabidopsis thaliana with
unknown function [27]. As the cis-prenyltransferases com-
monly synthesize long-chain products, a unique short-chain
cis,trans-FPP is made by Mycobacterium tuberculosis FPPS
which utilizes C
10
GPP and IPP to produce a FPP with a cis
double bond [28] (Fig. 1). This bacterium has a decaprenyl
pyrophosphate synthase (DPPS) to produce C
50
decaprenyl
pyrophosphate as a lipid carrier, which is one IPP unit
shorter than UPP found in other bacteria.
The products of these prenyltransferases have specific
chain lengths essential for their biological functions. An
intriguing question is how do they achieve product chain-
length specificity. In theory, restriction of the size of the
enzyme active site should play a major role in determining
the chain length of the final product. With the increasing
numbers of 3D structures available for prenyltransferases,
the mechanism of product chain-length determination has
begun to be elucidated [29]. The cis-type UPPS crystal
structures from Micrococcus luteus and E. coli have been

solved by Fujihashi et al.[30]andKoet al. [31], respect-
ively, providing the first two structures of the cis-prenyl-
transferase family. The structure in conjunction with
site-directed mutagenesis studies has revealed how
cis-prenyltransferase controls its product size [31].
Protein prenyltransferases form another prenyltrans-
ferase family. This enzyme catalyzes the transfer of the
carbon moiety of FPP or GGPP to a conserved cysteine
residue in a CaaX motif of protein and peptide substrates
[32]. This postranslational modification is essential for a
number of proteins including Ras, Rab, nuclear lamins,
trimeric G-protein c subunits, protein kinases, and small
Ras-related GTP-binding proteins [33,34]. The addition of a
farnesyl group to these proteins is required to anchor them
to the cell membrane, a step required for them to function.
As oncogenic forms of Ras in nearly 30% of human cancers
were observed, inhibition of Ras farnesyltransferases
(FTases) became a new strategy for anticancer therapy
[35,36]. The crystal structure of a mammalian FTase had
been determined at 2.25 A
˚
resolution [37]. Subsequently, the
structures of the enzyme in complex with FPP substrate [38]
and a ternary complex of the enzyme with FPP and a CaaX
peptide substrate were solved [39]. Later, a Rab geranyl-
geranyltransferase (GGTase) was solved at 2.0 A
˚
resolution
[40].
FPP is considered to be a branching point in the synthesis

of different types of natural isoprenoids. A number of
enzymes catalyze cyclization of FPP to generate natural
products such as pentalenene (pentalenene synthase), 5-epi-
aristolochene (5-epi-aristolochene synthase) and trichodiene
(trichodiene synthase). Squalene synthase catalyzes the
cyclization of squalene which is synthesized by the coupling
of two FPP molecules. The crystal structures of these
enzymes have been solved and provide insights into the
catalytic mechanism of terpenoid cyclization [41–44].
This review summarizes these recent advances in the
structural and mechanistic studies of the above three
families of prenyltransferases with emphasis on IPPS, which
has essential biological functions (Table 1). A general
mechanism of product chain-length determination and the
reaction kinetics derived from a pre-steady-state kinetic
analysis for trans-IPPS and cis-IPPS are described.
CLASS I: ISOPRENYL PYROPHOSPHATE
SYNTHASES (IPPSs)
Structure and active site of
trans
-IPPS
Over the past decade, many trans-IPPSs have been purified
and their genes cloned [45,46]. The deduced amino-acid
sequences of these enzymes show amino-acid sequence
homology and two common DDxxD motifs [47], suggesting
that they evolved from the same origin (Fig. 2) [48,49].
These Asp-rich motifs were recognized from the 3D
structure [50] and site-directed mutagenesis studies [51–56]
to be involved in substrate binding and catalysis via
chelation with Mg

2+
, a cofactor required for enzyme
activity.
The first crystal structure of a trans-prenyltransferase
reported is that of avian FPPS [50]. As shown in Fig. 3A,
this 2.6-A
˚
-resolution structure contains 13 a helices, 10 of
them surrounding the large central cavity. Two conserved
DDxxD sequences are located in this deep cleft, which
Fig. 1. Synthesis of linear all trans-isoprenyl
pyrophosphates (top) and trans,cis-isoprenyl
pyrophosphates (bottom) catalyzed by trans-
IPPSs and cis-IPPSs, respectively. The
stereoisomers are not specified in the nomen-
clature of the enzymes. Beside a trans-FPPS, a
cis-type FPPS from M. tuberculosis has been
showntosynthesizecis,trans-FPP.
3340 P H. Liang et al.(Eur. J. Biochem. 269) Ó FEBS 2002
forms the substrate-binding pocket. In site-directed muta-
genesis studies of yeast FPPS, substitution of Asp with Ala
in the first negatively charged DDxxD motif decreased k
cat
by 4–5 orders of magnitude [53]. In the second DDxxD
motif of yeast FPPS, the first and second Asp to Ala
mutations resulted in k
cat
values 4–5 orders of magnitude
smaller [53]. However, when the third Asp was mutated to
Ala, there was a less pronounced effect ( 100-fold

smaller k
cat
)[53].A10
4
)10
5
lower activity was observed
for the FPPS from Bacillus stearothermophilus when the first
Table 1.
7
The biological functions and three-dimensional structures of prenyltransferases presented in this review.
Prenyltransferases Biological functions 3D structure [ref]
trans-FPPS Precursor of steroids, cholesterol, sesquiterpenes, farnesylated proteins,
heme, and vitamin K12
[50,57]
trans-GGPPS Precursor of carotenoids, retinoids, diterpenes, geranylgeranylated
chlorophylls, and archaeal ether linked lipids
trans-GFPPS Archaeal ether linked lipids
trans-HexPPS Ubiquinone side chains
–DPPS
cis-UPPS Lipid carrier for peptidoglycan synthesis [30,31]
cis-DDPPS Lipid carrier for glycoprotein synthesis
Ras FTase Farnesylated Ras for signal transduction [37–39]
Rab GGTase Geranylgeranylated Rab for signal transduction [40]
Squalene cyclase Precursor of cholesterol [41]
epi-Aristolochene synthase Precursor of antifungal phytoalexin capsidol [42]
Pentalenene synthase Precursor of pentalenolactone antibiotics [43]
Trichodiene synthase Precursor of antibiotics and mycotoxins [44]
Fig. 2. Sequence alignment of trans-prenyl-
transferases. The sequence-related proteins

including FPPS from E. coli, GGPPS from
yeast, FGPPS from E. coli, HexPPS
from yeast, HepPPS from B. subtilis, OPPS
from E. coli, SPPS from M. luteus,andDPPS
from human are shown. Black and gray
outlines indicate identical and similar amino
acid residues, respectively. The two DDxxD
motifs are conserved in all proteins. The 5th
aminoacidupstreamfromthefirstDDxxDis
a large residue (Leu, Phe, or Tyr) for FPPS
and GGPPS, and is small amino acid (Ala) for
FGPPS, HexPPS, HepPPS, OPPS, SPPS, and
DPPS.
Ó FEBS 2002 Structure, mechanism and function of prenyltransferases (Eur. J. Biochem. 269) 3341
and second Asp of the second DDxxD motif were replaced
with Ala [55]. In contrast, the third Asp seems to be less
important to catalysis as its mutation to Ala only resulted in
6–16 times lower k
cat
values, but with a 10-fold increased
IPP K
m
value [55].
For rat FPPS, replacement of the second and third Asp
with Glu in the first Asp-rich motif decreased k
cat
by
 1000-fold [52]. However, no significant change in the K
m
values for IPP and GPP were observed. In the second

DDxxD motif of rat FPPS, substituting the first Asp with
Glu decreased k
cat
90-fold and increased IPP K
m
26-fold,
whereas GPP K
m
remained unchanged [51]. On the other
hand, mutation of the third Asp resulted in no change in the
kinetic parameters.
These results indicate that all the Asp residues in the two
DDxxD motifs except the last one in the second motif are
important for catalysis. In addition, the second motif is
essential for IPP binding. These results are consistent with
the cocrystal structure of avian FPP in complex with GPP
and IPP [57]. The structure of FPPS clearly shows that the
first DDxxD is bound to the allylic substrate GPP and the
second motif is the binding site of the homoallylic substrate
IPP. Site-directed mutagenesis of other amino acids around
the two DDxxD motifs was also performed to show their
effects in substrate binding and catalysis [54,56].
Amino-acid residues essential to product chain-length
determination of
trans
-prenyltransferases
In parallel with the site-directed mutagenesis studies, a
random chemical mutagenesis approach was used to select
FPPS mutants induced by NaNO
2

treatment that could
synthesize longer-chain products. FPPS was converted into
GGPPS which synthesizes a C
20
product by generating
mutations at position 81, 34 and 157 of FPPS from
B. stearothermophilus [58]. The Y81H mutant was the most
effective at increasing the production of GGPP (C
20
)
compared with FPP (C
15
) [58]. The subsequent site-directed
mutagenesis study in which Y81 was systematically replaced
with each of the other 19 amino-acid residues showed that
Y81A, Y81G and Y81S are capable of making hexaprenyl
pyrophosphate (C
30
) [59]. The final products of Y81C,
Y81H, Y81I, Y81L, Y81N, Y81T and Y81V are C
25
geranylfarnesyl pyrophosphate. On the other hand, Y81D,
Y81E, Y81F, Y81K, Y81M, Y81Q, and Y81R cannot
synthesize products larger than GGPP (C
20
). It appears that
this residue, located at the fifth position before the first
DDxxD motif, is the key residue in determining product
chain length. Predicted from the secondary structure of
FPPS, this residue is located at a distance of  12 A

˚
from
the first Asp-rich motif, which is similar to the length of the
hydrocarbon moiety of FPP [59]. The chain length of the
product catalyzed by these mutants is inversely proportional
to the accessible surface volume of the substituted amino-
acid residue in the first DDxxD. Also in archaebacterial
GGPPS, mutation of Phe77, which is upstream from the
first DDxxD, led to a change in product [60]. For instance,
replacement of this large residue with the smaller Ser
resulted in the production of C
25
rather than C
20
.
There was a similar finding in avian FPPS, in which
replacement of aromatic Phe112 and Phe113 with smaller
amino acids resulted in the product specificity shifting from
C
15
(FPP) to C
20
(GGPP)
1
(F112A), C
25
geranylfarnesyl
pyrophosphate (F113S) and longer products (F112A/
F113S double mutant) [57]. These two residues are located
in the fifth and sixth position before the first DDxxD

Fig. 3. (A) Crystal structure of trans-type FPPS and (B) schematic representation of the active site with FPP (product) and IPP bound. (A) The model
of avian FPPS is shown using a ribbon diagram. Two identical subunits are associated into a dimer by forming a four-layer helix bundle. The
N-terminal a helical hairpins in the outer layers are shown in blue and red for the two individual subunits, while the eight helices in the inner layers
are shown in cyan and magenta. Two small peripheral domains are colored green and yellow. The locations of the active sites are indicated by red
arrows, where the aspartate side chains of the two DDxxD motifs in each subunits are also shown in red. The phenylalanine side chains of F112 and
F113, which are involved in product length control, are shown in green. (B) The 5th amino acid (shown as a filled black circle) before the first
DDxxD motif in the sequence is located next to the tail of FPP. The large amino acid at this position can block further chain elongation of the
product and represents a mechanism to control the product chain length.
3342 P H. Liang et al.(Eur. J. Biochem. 269) Ó FEBS 2002
motif. The X-ray crystal structure of the enzyme reveals
that the first DDxxD is near a large hydrophobic pocket
which contains the mutated Phe112 and Phe113 residues
(Fig. 3A). It is proposed that this pocket binds the growing
hydrocarbon tail of the product, and the first DDxxD is
responsible for the binding of its pyrophosphate moiety.
An enlarged active site of the mutant enzymes is probably
related to increased product size based on the 3D
structuresoftheFPPSintheapostateandtotheallylic
substrate bound. Accordingly, an increase of 5.8 A
˚
in the
depth of the hydrophobic pocket agrees with the 5.2 A
˚
difference of an extra IPP unit between FPP (C
15
)and
GGPP (C
20
). The role of F113 in controlling the product
chain length is further supported by the fact that GGPPS

from Methanobacterium thermoautotrophicum contains
phenylalanine and serine at the positions corresponding
to Phe112 and Phe113 in avian FPPS. From the X-ray
crystal structure and the sequence alignment of a variety of
polyprenyl pyrophosphate synthases (PPPSs), which gen-
erate C
15
,C
20
and larger products, the importance of
Phe112 and Phe113 in the mechanism of product chain-
length determination is evident. In conclusion, the large
amino-acid residue located before the first DDxxD motif
provides the ÔfloorÕ to block further elongation of the
product (Fig. 3B). Once this residue is replaced with a
smaller one, elongation can continue.
In addition to the 5th and 6th amino-acid residues, the
roles of the 8th and/or the 11th positions before the first
DDxxD in the control of product specificity of archaeal
GGPPS and FPPS were also examined [61]. The single
mutant (F77S, 5th amino acid residue before the first Asp-
rich motif), double mutant (L74G/F77G) and triple mutant
(I71G/74G/F77G) of GGPPS mainly produce C
25
,C
35
and
C
40
, respectively [61]. FPPS mutants display a similar

pattern. This indicates that replacing amino acids near the
5th amino acid with smaller ones can further increase
production of longer polymers.
Conversely, replacement of a small amino acid with a
larger one shortens the chain length of the product. The
avian FPPS mutants A116W and N114W produce C
10
GPP
instead of C
15
FPP synthesized by the wild-type enzyme [62].
These residues are also located in the hydrophobic pocket of
the allylic substrate site, as revealed by docking dimethyl-
allyl pyrophosphate into the crystal structure of FPPS.
Dimethylallyl pyrophosphate is an isomer of IPP and its
condensation with IPP produces GPP. The mutations
A116W and N114W could fill in the bottom of the active-
site cavity, forcing the synthesis of shorter GPPs. Steady-
state kinetic measurements indicate that the two mutant
enzymes have larger k
cat
values with dimethylallyl pyro-
phosphate as substrate and binding of GPP is therefore
shifted to the allylic site because of the smaller internal space
oftheactivesite.
Structure and active site of
cis
-IPPS
UPPS has been partially purified from Lactobacillus
plantarum and characterized [63]. Subsequently, the UPPS

encoding gene cloned from M. luteus is the first cis-
prenyltransferase gene identified, and the deduced amino-
acid sequence shows no sequence similarity to those of
trans-prenyltransferases [64]. The sequence comparison
with UPPS allows the identification of many cis-type
IPPSs in bacteria, plant and animals as a family [65].
Several regions of conserved sequences can be identified
(Fig. 4). Broadly grouped, they are (with the amino-acid
sequence shown in parentheses): region I (20–32), region II
(42–46), region III (66–88), region IV (142–154) and region
V (190–224). Most of the fully conserved amino acids are
involved in catalysis, substrate binding, or structural
interactions as revealed later by the crystal structures and
mutagenesis studies.
Unlike the trans-type enzymes, cis-prenyltransferases
lack the DDxxD motifs, although they require Mg
2+
for
activity. Earlier site-directed mutagenesis studies examined
the conserved Asp and Glu of E. coli UPPS and revealed
the importance of Asp26, Asp150 and Glu213 in substrate
binding and catalysis [66]. Replacement of Asp26 with Ala
results in 10
3
-fold smaller k
cat
without any significant
change in FPP and IPP K
m
values. Mutagenesis of Asp150

to alanine leads to a 50 times larger IPP K
m
but no change
in FPP K
m
and k
cat
values. If Glu213 is replaced with Ala,
there is a 70-fold increase in IPP K
m
and a 100-fold
decrease in k
cat
. The subsequent 3D structure of UPPS
from M. luteus suggests that Asp29 (equivalent to Asp26
in E. coli UPPS) is located in a p-loop, a conserved motif
for the pyrophosphate-binding site in many phosphate-
binding proteins such as nucleotide triphosphate hydrolase,
phosphofructokinase, cAMP-binding domain, and sugar
phosphatase [67]. Site-directed mutagenesis studies on
M. luteus UPPS have also identified Glu216 (equivalent
to Glu213 in the E. coli enzyme) as being responsible for
IPP binding via Mg
2+
[68]. A hydrophobic cleft in the
enzyme with four positively charged Arg residues located
at the entrance of the cleft and the hydrophobic residues
covering the interior is proposed as the site for substrate
recognition [30].
The E. coli UPP structure reveals a larger hydrophobic

tunnel surrounded by two a helices (a2anda3) and four
b strands (bB, bA, bD, and bC)asshowninFig.5A.In
contrast with the symmetric structure of M. luteus UPPS,
two protein conformations (open and closed forms) are
seen in the two subunits of E. coli UPPS, implicating a
closed/open conformational change mechanism in sub-
strate binding and product release [31]. The difference
between the two conformers is mainly in the position of
the a3 helix. On the basis of site-directed mutagenesis
studies [31], the flexible loop with amino acids 72–83
connected to the a3 helix has been suggested to serve as a
hinge for the interconversion of two conformers. This
study also suggested a role for a Trp residue in the loop for
FPP binding and catalysis [69]. Fluorescent stopped-flow
technology and steady-state spectrophotometer have
recently been used to directly observe a Trp fluorescence
intensity change due to the change in protein conformation
during catalysis [70]. When Trp91, which is located in the
a3 helix, is mutated to Phe, the fluorescence quenching
upon addition of FPP is abolished, suggesting that the a3
helix moves toward the active site during substrate binding
[70]. Thus the change in UPPS conformation to a closed
form results in better interaction between the enzyme and
the substrates and intermediates. After the reaction, the
UPPS structure shifts to an open form for product release,
triggered by crowding of the prenyl chain of the product
because the large amino-acid residues seal the bottom of
the tunnel-shaped active site.
Ó FEBS 2002 Structure, mechanism and function of prenyltransferases (Eur. J. Biochem. 269) 3343
Mechanism of product chain-length determination

in
cis
-IPPS
E. coli UPPS has a hydrophobic tunnel formed by two
a helices and four b strands, which is sufficiently large to
accommodate the whole UPP (Fig. 5A). Large amino acids
including I62, L137, V105, and H103 are located on the
bottom of the tunnel [31]. Among the mutants produced by
replacing these residues with the smaller Ala, L137A
synthesizes C
70
and C
75
as major products, which are longer
than the C
55
product of the wild-type enzyme [31]. Ko et al.
[31] proposed that this residue plays an essential role in
determining the product chain length by blocking further
product elongation, analogous to the above amino-acid
residue located upstream from the first DDxxD motif in the
trans-type enzyme (Fig. 5B). This residue may represent a
common residue in the mechanism of product chain-length
determination among cis-prenyltransferases. From their
sequence homology, yeast Rer2, which synthesizes longer-
chain C
70
–C
80
products, has an Ala residue at this position.

The single UPPS mutation, L137A, has converted UPPS
into DDPPS in terms of product specificity. V105 may also
play a role in blocking chain elongation, as its mutation
increases the proportion of C
60
,C
65
and C
70
in the absence
of Triton, but it is not as critical as L137.
All cis-prenyltransferases so far identified have products
of chain length at least C
55
, except a short-chain FPP-
synthesizing enzyme recently identified from M. tuberculo-
sis. When the amino-acid sequence of UPPS is compared
with that of the short-chain FPPS, the A69 and A143 in
E. coli UPPS correspond to the large L84 and V156 in
M. tuberculosis FPPS, respectively. Substitution of Leu for
A69 in E. coli UPPS indeed leads to production of a greater
amount of C
30
intermediate which is longer lived, suggesting
that this residue interferes with the chain elongation of the
C
30
product. From the structure, it is reasonable to assume
that A69 is located midway in FPP elongation to the C
55

product. The conclusion derived from the changes in
product specificity in E. coli UPPS mutants needs to be
confirmed when the structures of the other cis-prenyltrans-
ferases are available.
Reaction mechanism and kinetics of
trans
-IPPS
and
cis
-IPPS
An ionization–condensation–elimination mechanism has
been proposed for the trans-prenyltransferase reaction. As
shown below in Fig. 9A, the carbocation resulting from the
dissection of the pyrophosphate group from GPP is
proposed as the intermediate during the FPPS-catalyzed
condensation reaction of IPP with allylic pyrophosphate
substrate [1]. This conclusion is based on the finding that the
enzyme, which normally catalyzes the addition of a GPP to
IPP, is able to catalyze the hydrolysis of GPP [71]. Hydrolysis
carried out in H
2
18
Oorwith(1S)-[1-
3
H]GPP shows that the
C–O bond is broken, and the chirality of the C1 carbon of
GPP is inverted in the process. Furthermore, the trifluoro-
methyl substituent at the C3 position or the fluoro atom at
the C2 position of the allylic substrate destabilizes the
carbocation and retards the enzyme reaction [72].

Fig. 4. Alignment of the cis-type IPPSs. These
include, in turn, E. coli UPPS, yeast DDPPS
Rer2, Yeast DDPPS Srt1, M. tuberculosis FPs
Rv1086, M. tuberculosis DPPS Rv2361c,
Arabidopsis thaliana PPPS and human
DDPPS. The numbers and secondary-struc-
tural elements shown above the sequences
arefortheUPPSfromE. coli,basedon
PROCHECK
analysis of the crystal structure.
The green arrows denote the locations of
b-strands, and the cylinders in red, magenta
andcyanarefora helices, 3
10
helices and
turns, respectively. In the aligned sequences,
several conserved regions including (I) resi-
dues 20–32, (II) 42–46, (III) 66–88, (IV)
142–154, and (V) 190–224 can be identified,
which probably also have corresponding
secondary structures in common.
3344 P H. Liang et al.(Eur. J. Biochem. 269) Ó FEBS 2002
A tertiary carbocation on the C3 carbon of the IPP part is
proposed to form during the 1¢-4 condensation. The aza
analogues with nitrogen replacing the cationic carbon to
mimic the transition state have previously been demonstra-
ted to strongly inhibit enzymes in isoprenoid biosynthesis
where the carbocation transition state is presumably formed
during catalysis [73]. An aza analogue of the transition state,
3-azageranylgeranyl diphosphate, is also a potent inhibitor

of GGPPS [74,75].
In the elimination step, a hydrogen is removed from C2
of IPP with simultaneous formation of a C¼C double
bond between C2 and C3. The formation of the trans or
cis double bond during the IPPS reaction depends on the
spatial arrangement of the IPP relative to the elongating
oligoprenyl substrate. In the trans-prenyltransferases, the
allylic substrate is added to the si face of the double bond
of the IPP, and the pro-R proton is removed from the
methylene group next to the double bond, with the
concomitant formation of a new trans double bond [72].
The amino acid involved in the hydrogen removal of IPP
in prenyltransferase has not yet been reported. However,
several important amino acids, including the nucleophilic
His309 responsible for the removal of the pro-R hydrogen,
have been proposed from analysis of the crystal structure
of pentalenene synthase, which catalyzes cyclization of
FPP initiated by the cleavage of its pyrophosphate moiety
and formation of a carbocation intermediate (see below
for details) [43]. The active-site residues Phe77 and Asn219
have been implicated in the stabilization of the carboca-
tion intermediate [43]. In UPPS, there are several
conserved hydrophilic amino acids in the vicinity of
Asp26, including Asn28, Arg30, His43, Phe70, Ser71,
Arg194 and Glu198. It is conceivable that some of them
may play roles analogous to those of His309, Phe77 and
Asn219 in pentalenene synthase. However, to form a cis
double bond in UPPS, the position of the requisite
nucleophile (possibly His43) will need to be close to the
pro-S proton of IPP.

The reaction mechanism of the cis-prenyltransferases is
less well understood. The hydrolysis of the allylic substrate
by cis-prenyltransferases has not been observed. Analog-
ously to the two DDxxD motifs in the trans-prenyltrans-
ferases, site-directed mutagenesis studies of UPPS indicate
that Asp and Glu play a significant role in IPP binding and
Fig. 5. (A) Two orthogonal views of the ribbon representation of an E. coli UPPS dimer and (B) the proposed active site of the E. coli UPPS located in
a tunnel-like crevice surrounded by a2, a3, bD, bB, bA, and bC. (A) The top view is perpendicular to, and the bottom view is parallel to, the molecular
dyad axis. The seven a helices and six b strands are shown in red and green, respectively, for subunit A, and in magenta and cyan for subunit B, and
they are labelled separately. The blue arrows indicate locations of the active sites, each having a substrate-binding tunnel formed by helices a2, a3
and the central b-sheet. (B) The substrate site is located on the top of the hydrophobic tunnel with D26 and D213 playing a significant role in
substrate binding and catalysis. A69 is in the midle of FPP chain elongation to the product. The large amino acid L137 on the bottom of the tunnel
is essential for determination of product chain length by blocking further elongation of UPP.
Ó FEBS 2002 Structure, mechanism and function of prenyltransferases (Eur. J. Biochem. 269) 3345
activity probably through co-ordination with Mg
2+
.The
IPP condensation kinetics for UPPS, DDPPS (cis-type) and
OPPS (trans-type)hadbeenmeasuredforcomparisonof
IPP condensation catalyzed by cis-IPPS and trans-IPPS
[76–78]. These experiments were performed under the
enzyme single-turnover condition using a rapid-quench
instrument, because the slow release of the large hydropho-
bic product limits IPPS catalysis under steady-state condi-
tions. The IPP condensation rate constant in the UPPS
reaction is similar to that of OPPS (2.5 s
)1
vs. 2.0 s
)1
)

derived from the time course data simulated by the Kinsim
program (Fig. 6). This suggests that the chemical environ-
ment of the active site in the two types of enzyme may be
similar, despite the completely different primary sequence.
Moreover, Triton could facilitate product release and
increase the UPPS steady-state rate by 190-fold by switching
the rate-limiting step from product release to IPP conden-
sation. However, OPPS activity is only slightly increased
(threefold) by Triton. Also determined from these studies,
the bond-forming or bond-breaking step during IPP
condensation must be rate limiting because the first reaction
of C
15
to C
20
, in which UPPS or OPPS is preincubated with
FPP so that the pyrophosphate dissociation and the
relocation of the intermediate are not involved in the
reaction, has the same rate as the subsequent steps [76].
Product distribution of
cis
-prenyltransferases
and
trans
-prenyltransferases
In the UPPS reaction, significant amounts of intermediates
are yielded as products under the conditions of 50 l
M
FPP
and 50 l

M
IPP, whereas the C
55
-UPP is synthesized with
5 l
M
FPP and 50 l
M
IPP. In contrast, OPPS produces no
intermediate products under these conditions [77]. Another
trans-prenyltransferase, SPPS from M. luteus, shows a
similar pattern of product distribution in which no inter-
mediate is displaced from the active site at high concentra-
tion of FPP [15]. As OPPS and UPPS have similar FPP and
IPP K
m
values, OPPS and SPPS (trans-type) apparently
have higher affinities for intermediates compared with
UPPS (cis-type). At a fixed concentration of allylic sub-
strate, the increased ratio of IPP to FPP results in the
synthesis of longer-chain products. As shown for a UPPS
from the Archaeon Sulfolobus acidocaldarius, the enzyme
mainly generates C
50
and C
55
when IPP is present in excess
of GGPP (or FPP), but decreasing the concentration of IPP
results in larger amounts of short-chain products [79]. This
could be due to the low affinity of IPP for the enzyme.

In the presence of Triton, UPPS mainly synthesizes C
55
product. In contrast, under conditions in which product
release is slow, chain elongation beyond normal can occur.
Therefore, in the absence of Triton, UPPS could produce
C
55
–C
75
polymers [76]. For the microsomal DDPPS of rat
liver, the chain length of products shifted downward from
C
90
and C
95
with increasing concentration of detergent [80].
The trans-typeOPPScouldalsogenerateC
40
–C
60
com-
pounds as the final products. Derived from pre-steady-state
kinetic studies, the rate constant for condensation of an extra
IPP with C
55
to produce C
60
is fivefold lower than that for
regular IPP condensation (e.g. C
50

to C
55
)intheUPPS
reaction (Fig. 6). On the other hand, the rate for elongation
from C
40
to C
45
catalyzed by OPPS is 100 times slower than
for elongation from C
35
to C
40
(Fig. 6). The trans enzyme
seemstohaveamorerigidactivesitethanthecis-
prenyltransferase, as shown by the higher product specificity.
The most surprising observation on the product distri-
bution of the UPPS and OPPS reactions is that, when the
Fig. 6. Comparison of the kinetic pathways of UPPS and OPPS. OPPS
catalyzes elongation of C
15
to C
40
, resulting in all trans double bonds,
and UPPS catalyzes formation of the C
55
product containing newly
formed cis double bonds. The IPP condensation steps have rate con-
stants of 2 s
)1

and 2.5 s
)1
for OPPS and UPPS, respectively. The
similar rate constants suggest closely related reaction mechanisms and
perhaps similar active-site environments. However, OPPS has higher
product specificity as judged from its much lower rate of conversion
into C
45
extra product.
3346 P H. Liang et al.(Eur. J. Biochem. 269) Ó FEBS 2002
concentration of the UPPS and FPP complex is in 10-fold
excess over that of IPP, long-chain polymers larger than C
20
(GGPP), including C
55
, are generated. Pan et al. [76]
proposed that UPPS–intermediate complexes may have
greater affinity than the UPPS–FPP complex for the limited
amount of IPP, leading to the formation of product with the
correct chain length. OPPS shows the same phenomenon in
producing C
20
–C
40
under the same conditions [77]. This
strategy for synthesizing the desired chain length seems to be
shared by both types of enzyme.
CLASS II: PROTEIN
PRENYL-TRANSFERASES
Structure and active site of protein prenyltransferases

Ras FTase is a Zn
2+
-dependent prenyltransferase contain-
ing a and b heterodimer, which catalyzes the farnesylation
on a C-terminal CaaX motif of the Ras protein. As shown
in Fig. 9B, the Zn
2+
-activated thiolate of Cys acts as a
nucelophile to attack the ionized farnesyl group. In the 3D
structure of a mammalian Ras FTase, both subunits are
largely composed of a helices (Fig. 7A) [37]. The a-2 to a-15
helices in the a subunit fold into a novel helical hairpin
structure, resulting in a crescent-shape domain that enve-
lopes part of the b subunit. On the other hand, the 12 helices
of the b subunit form an a–a barrel. Six additional helices
connect the inner core of helices and form the outside of the
helical barrel. A deep cleft surrounded by hydrophobic
amino acids in the center of the barrel is proposed as
the FPP-binding pocket. A single Zn
2+
ion is located at the
junction between the a-hydrophilic surface groove near the
subunit interface and the deep cleft in the b subunit. This
Zn
2+
ion is pentaco-ordinated by the Asp297 and Cys299
located in the N-terminal helix 11, His362 in helix 13 of the
b subunit, and a water molecule as well as a bidentate
2
ligand

Asp297b. Replacement of Cys299b with Ala results in lower
Zn
2+
affinity and abolishes enzyme activity [81]. A nine-
amino-acid portion of the adjacent b subunit in the crystal
lattice was found to bind in the positively charged pocket of
the b subunit close to the FPP site. This observation allowed
the authors to speculate that the nonapeptide mimics some
aspects of normal CaaX peptide binding. This was suppor-
ted by the fact that the first four residues of the peptide form
atype1b turn which indeed is similar to the observed
conformation for the natural CaaX peptide when bound to
FTase [82]. Furthermore, the C-terminal residue of the
nonapeptide could form hydrogen bonds with Lys164a,
Arg291b,andLys294b.
Since the report of the first FTase structure, the same
research group has published the structure of FTase in
complex with FPP [38] as well as the structure of the ternary
complex containing an inactive FPP analogue and a CaaX
peptide substrate [39]. Many features of the apo-FTase
structure are retained in the complexes. According to the
structure of FTase in complex with FPP, the highly
conserved residues Trp303b,Try251b, Trp102b,Tyr205b,
and Tyr200a form hydrophobic interactions with FPP.
Arg202b in the binary complex adopts a different confor-
mation from the structure of the apoenzyme to further
interact with the substrate. This residue is stabilized by
Asp200b and Met193b which also adopts a different
conformation. From the binary structure, two additional
amino acids, Cys254b and Gly250b, participate in the

binding of FPP. The diphosphate moiety of the FPP
substrate is hydrogen-bonded with His248b,Arg291b,and
Tyr300b. Lys164a and Lys294b are also within hydrogen-
bonding distance of the diphosphate. Mutations of His248b,
Arg291b, Lys294b, Tyr300b, and Trp303b cause 3–15 times
increased FPP K
d
compared with the wild-type FTase [83],
consistent with the crystal structure. On the other hand,
replacement of Lys164a with Asn results in a markedly
decreased k
cat
value, suggesting that this residue plays a
catalyticrole[84].Fromthebinarystructure,Lys164may
interact with the diphosphate moiety of FPP and be
involved in the transfer of Cys thiol to C1 of the substrate.
As for the binding of the 5th CaaX peptide substrate,
the structure of FTase complexed with a-hydroxyfarnesyl-
Fig. 7. (A) Ribbon diagram of Ras FTase and (B) the molecular ruler mechanism for substrate specificity of FTase and GGTase. (A) The heterodimeric
enzyme consists of two subunits, a and b, colored in cyan–blue and yellow–green, respectively. Most of the secondary structures are helices, with the
a subunit comprising seven helical hairpins that surround the more compact b subunit. The peptide substrate, colored red, binds to a cleft between
the two subunits, and so does the substrate analogue, colored magenta. The active site is located in the b subunit. It contains a zinc ion, shown in
blue, which is bound by three residues Asp297, Cys299 and His362, shown in cyan, and also makes bonds with the substrate molecules. (B) When
FPP is replaced with GGPP in the active site of FTase, the thiolate nucleophile is further away from the electrophilic carbon next to the
pyrophosphate leaving group, thereby decreasing the enzyme activity. Similarly, FPP is a poor substrate for GGTase.
Ó FEBS 2002 Structure, mechanism and function of prenyltransferases (Eur. J. Biochem. 269) 3347
phosphonic acid and acetyl-Cys-Val-Ile-selenoMet-COOH
reveals that it binds in a cleft located in the subunit interface,
in agreement with the pervious apoenzyme structure [37].
The peptide is in contact with FPP and several amino acids

of the enzyme. The Ile-Met portion interacts with the
adjacent Tyr166a, and the carbonyl group of the main chain
forms a hydrogen bond with Arg202b. The Cys sulfur atom
of the peptide co-ordinates with a water molecule and the
Zn
2+
ion which is stabilized by interacting with Asp297a,
Cys299b,andHis362b. The N-terminal acetyl group makes
no contacts with the enzyme.
As revealed by the FTase structure in complex with
FPP, large amino-acid residues (Trp102b and Tyr205b)
are located on the bottom of the hydrophobic funnel for
FPP binding. The distance from these amino acids to the
bound Zn
2+
ion is approximately the length of FPP from
C1 to C15 (Fig. 7B). According to this model, if a C
20
GGPP is within the active site, its pyrophosphate is out of
the reach of Zn
2+
. This explains why GGPP binds
competitively with FPP to FTase but only FPP serves as
an effective substrate. On the other hand, FPP binds to
GGTase with 330-fold less affinity than GGPP [85].
However, it still serves as a moderately effective substrate
because its diphosphate moiety can be situated around the
Zn
2+
. This hypothesis has been confirmed by the solving

of the 3D structure of Rab GGTase [40]. The GGPP is
bound in the central cavity of the a–a barrel in the
b subunit with its diphosphate head group to the positively
charged cluster composed of Arg232b, Lys235b,and
L105a. The diphosphate is closed to the Zn
2+
,whichis
co-ordinated with Asp238b, Cys240 b and His290b in a
similar way to that observed in FTase, and an additional
His residue. The structure of Rab GGTase can be
superimposed on that of Ras FTase. One of the most
striking differences is that on the bottom of the GGTase
active-site cavity, Ser48b and Leu99b replace the more
bulky Trp102b and Tyr154b seen in FTase, thereby
enlarging the active site to accommodate GGPP. This is
consistent the molecular ruler mechanism (Fig. 7B) pro-
posed in FTase for substrate specificity.
Rab GGTase is unique and different from FTase and
type I GGTase in that it is able to prenylate Rab only in the
presence of Rab escort protein. It exclusively modifies
members of a single subfamily of Ras-related small GTPase,
the Rab proteins involved in the regulation of intracellular
vesicular transport in the biosynthetic secretory and exocy-
tic/endocytic pathways [86,87]. Upon binding with the
protein complex, the Rab GGTase transfers two GGPPs to
the two Cys residues of Rab C-terminal -CC, -CXC, -CCX,
or -CCXX motif [88]. As shown in the superimposition of
the GGTase and FTase structures, several residues in the
peptide-binding pockets are different, reflecting their pep-
tide substrate specificity.

Reaction mechanism and kinetics of protein:
prenyltransferase
Kinetic analysis of the enzymatic prenylation reaction
reveals a relatively fast chemical step  0.8–12 s
)1
followed
by rate-limiting product release (0.06 s
)1
) [89,90]. Substrate
binding follows an ordered sequential mechanism with the
formation of the FTase–FPP complex before binding of the
CaaX substrate [91]. When the FPP analogue with a strong
electron-withdrawing fluorine atom replacing the C3
hydrogen was used, the reaction rate was significantly
decreased, suggesting formation of a carbocation at C1
during the reaction [92,93]. This phenomenon is similar to
that previously observed for FPPS as described above, but
the fluorinated FPP had less effect in the FTase reaction
than in the FPPS reaction. The formation of the carbo-
cation with the simultaneous separation of the pyrophos-
phate of FPP represents in part the rate-determining step in
the FTase reaction as the metal required for co-ordination
with Cys of the peptide substrate is also involved in the
catalysis.ThedirectcontactofZn
2+
with Cys thiol is
evident from the crystal structure of the FTase ternary
complex and the studies using the Co
2+
-substituted FTase

[94]. The use of the more thiophilic Cd
2+
to increase the
thio affinity and lowering its pK
a
to display lower activity
suggests the direct participation of a metal ion in FTase
catalysis [95]. However, inclusion of high concentrations of
Mg
2+
in the reaction mixture increases enzyme activity
700-fold because excess Mg
2+
occupies a separate site,
facilitating departure of the diphosphate group [96].
Therefore, an associated character with partially positive
charge at C1 of FPP and partially negative charge
pyrophosphate oxygen as well as in the metal-co-ordinated
thiolate was proposed as the transition state of the FTase
reaction (see Fig. 9B).
CLASS III: TERPENOID CYCLASES
General structure and mechanism of terpenoid cyclase
Terpenoid cyclases such as squalene cyclase, pentalenene
synthase, 5-epi-aristolochene synthase, and trichodiene
synthase are responsible for the synthesis of cholesterol, a
hydrocarbon precursor of the pentalenolactone family of
antibiotics, a precursor of the antifungal phytoalexin
capsidiol, and the precursor of antibiotics and mycotoxins,
respectively. The last three enzymes catalyze the cyclization
of FPP involving: (a) ionization of FPP to an allylic cation

which acts as electrophile to react with one of the p bonds of
the substrate for cyclization; (b) relocalization of the
carbocation via hydride transfer and Wagner-Meerwein
rearrangements; (c) deprotonation or capture of an exo-
genous nucleophile such as water to eliminate the carboca-
tion. On the other hand, squalene synthase catalyzes the
cyclization of squalene, which is formed by coupling two
FPP molecules. Like trans-type IPPSs, which make linear
polymers from FPP, the four cyclases also contain con-
served Asp-rich motifs, suggesting that these enzymes have
similar strategies for activating FPP. In the structures of
these three enzymes, the similar structural feature referred to
as Ôterpenoid synthase foldÕ with 10–12 mostly antiparallel
a helices is found, as also observed in IPPS and FTase
(Fig. 8). The high structural similarity provides support for
the hypothesis that the three families of prenyltransferases
described in this review have related evolution despite their
low sequence similarity.
Pentalenene synthase
AsshowninFig. 8A,theactivesiteofbacterialpentalenene
synthase cloned from Streptomyces UC5319 [97] is located
in a hydrophobic pocket with the bottom sealed by aromatic
3348 P H. Liang et al.(Eur. J. Biochem. 269) Ó FEBS 2002
residues Phe57, Phe76, and Trp308, and aliphatic residues
Leu53, Val177, Val179, Thr182, and Val301 [43]. These
residues together create the necessary conformation for FPP
to undergo cyclization to pentalenene. The top of the pocket
is surrounded by hydrophilic amino acids, including His309,
Asn219, Arg44, Arg157, Arg173, Lys226, Arg230, and
particularly Asp80 and Asp84 with their carboxylate

protruded into the active site to complex with Mg
2+
.
Together they facilitate the FPP carbocation formation as
shown in Fig. 9C (path a). Subsequent cyclization and
carbocation rearrangement processes use a His309 catalytic
residue in the active site [43].
5-epi-Aristolochene synthase
Tobacco 5-epi-aristolochene synthase adopts entirely an
a-helical structure with short loops and turns (Fig. 8B). The
structure folds into two domains: the N-terminal domain
aligns structurally with two glycosyl hydroxylases, and its
function is not known; the C-terminal domain shares the
common fold with FPPS. The active site of epi-aristolochene
synthase is located in the C-terminal domains containing two
Mg
2+
-binding sites. The Asp301 and Asp305 found in the
conserved DDxxD motif co-ordinate with a Mg
2+
[42]. The
Asp444, Thr448, Glu452 and a water molecule co-ordinate
Fig. 8. Ribbon representation of the structures of cyclic terpenoid-synthetic enzymes pentalenene synthase (A), 5-epi-aristolochene synthase (B),
trichodiene synthase (C) and squalene synthase (D). They are shown in different colors with the N-terminus and C-terminus labeled. All of these
enzymes have the terpenoid synthase fold and contain a helices as their major secondary-structure elements. The homologous helices that constitute
the active site are emphasized with broad ribbons.
Ó FEBS 2002 Structure, mechanism and function of prenyltransferases (Eur. J. Biochem. 269) 3349
with the other Mg
2+
.ThesetwoMg

2+
ions located in the
entrance of the active-site pocket are responsible for binding
to the FPP substrate. A number of hydrophobic amino acids
around the FPP site were proposed to provide further
hydrophobic binding interactions.
The catalytic mechanism of epi-aristolochene synthase
was elucidated by its structure in complex with the substrate
analogue farnesyl hydroxyphosphonate. Accordingly, as a
loop (residues 521–534) becomes ordered, it forms a lid that
clamps down over the active-site entrance in the presence of
farnesyl hydroxyphosphonate. This brings the loop con-
taining Arg264 and Arg266 close to the C1 hydroxy group
of farnesyl hydroxyphosphonate, and an additional Mg
2+
site forms. In analogy, the released negatively charged
diphosphate group of the FPP substrate is stabilized by
interaction with Arg264, Arg441 and the three Mg
2+
ions.
The newly formed allylic carbocation is positioned near the
main-chain carbonyl oxygens of residues 401 and 402 at
the kink in helix G, and the C1 is near the p orbitals of the
C10–C11 bond. The bond forming between C1 and C10
leaves a tertiary carbocation on C11 (see Fig. 9C, path b),
which is stabilized by Tyr527. A deprotonation at C13 by
Asp525 leads to the germacrene intermediate. The forma-
tion of a C–C bond between C2 and C7 of the intermediate
is triggered by protonation of C6 with a proton donated by
Tyr520. This leaves a positive charge on C3 to form a

bicyclic eudesmane carbocation intermediate. The following
rearrangement of the carbocation is facilitated by the
electron-rich side chain of Trp273 to generate the final
product aristolochene.
Trichodiene synthase
The recently solved 3D structure of trichodiene synthase
(Fig. 8C) reveals 17 a helices, six of which (C, D, G, H, I,
and J) form the hydrophobic active-site cleft [44]. An Asp-
rich D
100
DSKD motif, which is important for catalytic
activity as suggested by the mutagenesis studies [98], is
located in the active site at the C-terminus of helix D.
Three Mg
2+
ions were found in the electron map of the
enzyme–pyrophosphate complex. Among them, two metal
ions in the active-site region co-ordinate with Asp100 (the
first residue of the conserved Asp-rich motif). The
conserved Asp residues are also found in aristolochene
synthase and FPPS. The other Mg
2+
is located in the site
of N225, S229, and E233 on helix H, which are in a
consensus sequence conserved among all known terpene
synthase sequences. The metal ions are bound to the
enzyme in the presence of pyrophosphate, and the
pyrophosphate (also probably FPP) triggers a protein
conformational change, in which D101 breaks a salt link
with R62 and forms a new salt bridge with R304 after

pyrophosphate binding. The purpose of the protein
conformational change is proposed to facilitate the depar-
ture of the pyrophosphate group of the substrate. These
ligand-induced conformational changes were also observed
in other terpene synthases such as 5-epi-aristolochene
synthase, suggestive of divergent evolution from a common
ancestor. On the basis of the structures and the mechanism
of trichodiene synthase, a model of the enzyme reaction
has been constructed (Fig. 9C). Path c is the initial step of
the reaction; the recapture of a-phosphate oxygen by the
C3 cation forms the intermediate, which then undergoes
dephosphorylation and C1–C6 bond formation [44]. The
resulting C7 carbocation attracts the electrons of the
C10–C11 p orbital to form the C7–C11 bond. Subse-
quently, three carbocation-driven migrations occur to
generate the final product.
Squalene cyclase
Squalene cyclase and 2,3-oxidosqualene synthase catalyze
the cyclization reactions to give steroid compounds and
have become the targets for development of anticholeste-
remic and antifungal drugs. Bacterial squalene synthase
had been crystallized in three distinct forms, and the
structure of form A solved. The enzyme is a homodimer,
and the structure shows two domains in the a subunit
Fig. 9. The transition state for reactions of cis -IPPS and trans-IPPS
(A), protein farnesyltransferase (B), and pentalenene synthase, epi-aris-
tolochene synthase, and trichodiene synthase (C). In three cases, the
carbocation intermediate at C1–C3 carbons of FPP is formed, and the
negative charge developed on pyrophosphate is stabilized by co-ordi-
nation with Mg

2+
. The deprotonated IPP or a Zn
2+
-activated thiolate
anion of a protein acts as a nucleophile to attack the carbocation at C1
resulting in condensation with departure of the pyrophosphate. In the
cyclase reaction, the nearby C¼C bond donates electrons for ring
closure. The paths a, b, and c denote the routes of electron migration
from the C¼C bond to C1 with the release of pyrophosphate catalyzed
by pentalenene synthase, epi-aristolochene synthase, and trichodiene
synthase, respectively.
3350 P H. Liang et al.(Eur. J. Biochem. 269) Ó FEBS 2002
(Fig. 8D) [41]. Domain 1 is an a6–a6 barrel of two
concentric rings of parallel a helices, which is similar to the
fold of FPPSs and glucanases [99]. Domain 2 forms an a–a
barrel and inserts into domain 1. The active site is located
in a central cavity, with two amino acids in a DxDD motif,
Asp376 and Asp377, on the top of the cavity [41].
Mutagenesis of these two Asp residues abolishes enzyme
activity, indicating their importance in catalysis [100].
Based on the structure, Asp376 is the catalytic acid to
protonate the first C¼C double bond, resulting in a
carbocation for the initiation of ring cyclization. A water
molecule acts as a base to receive a proton on the other
end of the squalene molecule, which is polarized by the
other water in a hydrogen-bonding network formed by
Glu45 and Gln262. A constriction formed by Phe166,
Val174, Phe434, and Cys435 in a mobile loop may serve as
a gate for substrate passage.
CONCLUSION

This review summarizes the 3D structures, IPP condensa-
tion kinetics and mechanism of trans-IPPS and cis-IPPS for
the production of linear polyprenyl pyrophosphates, and
the reaction mechanism and structures of protein prenyl-
transferases and isoprenyl diphosphate cyclases. As illus-
trated in Fig. 9, the three families of prenyltransferases have
similar strategies for substrate binding and catalysis. First,
they use hydrophobic interactions to bind the hydrocarbon
moiety of the allylic pyrophosphate substrate. Secondly,
pyrophosphtae leaving is facilitated by co-ordination of
active-site amino acids such as Asp, Glu and Lys with
Mg
2+
. Although the protein FTase was isolated with Zn
2+
bound, the inclusion of Mg
2+
greatly enhances enzyme
activity. The carbocation character of the allylic substrate
(FPP) exists during reactions of all these prenyltransferases.
In the IPPS reaction, IPP with a proton abstracted attacks
the carbocation of FPP. For the protein farnesyltransferase,
a thiolate anion of the substrate protein or a peptide acts as
the nucelophile. The FPP cyclase reaction is also initiated by
the formation of an FPP carbocation which attracts a pair
of electrons from the nearby C¼C double bond for the
ring closure (paths a, b, and c for pentalenene synthase,
5-epi-aristolochene synthase, and trichodiene synthase,
respectively).
As the X-ray structure of cis-type UPPS has become

available, the general mechanism for product chain-length
determination of the cis and trans forms of IPPS has
gradually been understood. As revealed by the 3D
structures, the size of the active-site cavity apparently
controls the chain length of the final product of the IPPS
reaction. Several large amino acids form a floor to seal the
bottom of the active site and block further elongation of
the product. Site-directed substitution of key large amino
acids with smaller ones extends the size of the active site,
thereby allowing the formation of a longer product and
vice versa. The substrate specificities of protein FTase and
GGTase are controlled by the size of the substrates. The
larger GGPP is more distant from the Zn
2+
catalytic site
when binding on FTase and hence produces lower activity.
The accurately tailored active sites of FPP cyclases
(pentalenene synthase, 5-epi-aristolochene synthase, and
trichodiene synthase) guarantee formation of the correct
cyclization products.
A large number of isoprenoids have significant biological
functions in nature. The structural and mechanistic infor-
mation available so far provides an initial step to under-
standing the functions of many other prenyltransferases and
a basis for designing enzyme agonists and antagonists to
treat various diseases. However, the structures and even
functions of many isoprenoid-synthetic enzymes, as well as
the proteins catalyzing and undergoing prenylation, are still
not known. Further studies on the structure, mechanism
and function of prenyltransferases are therefore required.

ACKNOWLEDGEMENTS
We are grateful to Yi-Kai Chen for the preparation of the sequence
alignment figures.
REFERENCES
1. Poulter, C.D. & Rilling, H.C. (1981)
3
Prenyl transferases and
isomerase. In Biosynthesis of Isoprenoid Compounds (Spurgeon,
S.L., ed.), Vol. 1, pp. 161–224. John Wiley & Sons, New York.
2. Poulter, C.D. (1974)
4
Model studies in terpene biosynthesis.
J. Agric. Food Chem. 22, 167–173.
3. Poulter, C.D. & Rilling, H.C. (1981) Conversion of farnesyl
pyrophosphate to squalene.
5
In Biosynthesis of Isoprenoid Com-
pounds (Spurgeon, S.L., ed.), Vol. 1, pp. 225–282. John Wiley &
Sons, New York.
4. Spurgeon, S.L. & Porter, J.W. (1981) Biosynthesis of caro-
tenoids.
6
In Biosynthesis of Isoprenoid Compounds (Spurgeon,
S.L., ed.), Vol. 2, pp. 1–122. John Wiley & Sons, New York.
5. Sinensky, M. (2000) Recent advances in the study of prenylated
proteins. Biochim. Biophys. Acta 1484, 93–106.
6. Gelb, M.H., Scholten, J.D. & Sebolt-Leopold, J.S. (1998) Protein
prenylation: from discovery to prospects for cancer treatment.
Curr. Opin. Chem. Biol. 2, 40–48.
7. Kellogg, B.A. & Poulter, C.D. (1997) Chain elongation in the

isoprenoid biosynthetic pathway. Curr. Opin. Chem. Biol. 1, 570–
578.
8. Ogura, K. & Koyama, T. (1998) Enzymatic aspects of isoprenoid
chain elongation. Chem. Rev. 98, 1263–1276.
9. Ohnuma, S I., Suzuki, M. & Nishino, T. (1994) Archaebacterial
ether-linked lipid biosynthetic gene. Expression cloning,
sequencing and characterization of geranylgeranyl-diphosphate
synthase. J. Biol. Chem. 269, 14792–14797.
10. Chen, A. & Poulter, C.D. (1994) Isolation and characterization
of idsA: the gene for the short chain isoprenyl diphosphate syn-
thase from Methanolbacterium thermoautotrophicum. Arch. Bio-
chem. Biophys. 314, 399–404.
11. Tachibana, A., Yano, Y.Y., Otani, S., Nomura, N., Sako, Y. &
Taniguchi, M. (2000) Novel prenyltransferase gene encoding
farnesylgeranyl diphosphate synthase from a hyperthermophilic
archeon, Aeropyrum pernix. Eur. J. Biochem. 267, 321–328.
12. Asai, K I., Fujisaki, S., Nishimura, Y., Nishino, T., Okada, K.,
Nakagawa, T., Kawamukai, M. & Matsuda, H. (1994) The
identification of Escherichia coli ispB (cel ) gene encoding the
octaprenyl diphosphate synthase. Biochem. Biophys. Res. Com-
mun. 202, 340–345.
13. Okada, K., Suzuki, K., Kamiya, Y., Zhu, X., Fujisaki, S.,
Nishimura, Y., Nishino, T., Nakagawa, T., Kawamukai, M. &
Matsuda, H. (1996) Polyprenyl diphosphate synthase essentially
defines the length of the side chain of ubiquinone. Biochim.
Biophys. Acta 1302, 217–223.
14. Okada, K., Minehira, M., Zhu, X., Suzuki, K., Nakagawa, T.,
Matsuda, H. & Kawamukai, M. (1997) The ispB gene encoding
octaprenyl diphosphate synthase is essential for growth of
Escherichia coli. J. Bacteriol. 179, 3058–3060.

Ó FEBS 2002 Structure, mechanism and function of prenyltransferases (Eur. J. Biochem. 269) 3351
15. Ohnuma, S I., Koyama, T. & Ogura, K. (1992) Chain length
distribution of the products formed in solanesyl diphosphate
synthase reaction. J. Biochem. (Tokyo) 112, 743–749.
16. Ohnuma, S I., Koyama, T. & Ogura, K. (1991) Purification of
solanesyl-diphosphate synthase from Micrococcus luteus.Anew
class of prenyltransferase. J. Biol. Chem. 266, 23706–23713.
17. Teclebrhan, H., Olsson, J., Swiezewska, E. & Dallner, G. (1993)
Biosynthesis of the side chain of ubiquinone: trans-pre-
nyltransferase in rat liver microsomes. J. Biol. Chem. 268, 23081–
23086.
18. Suzuki, K., Okada, K., Kamiya, Y., Zhu, X.F., Nakagawa, T.,
Kawamukai, M. & Matsuda, H. (1997) Analysis of the dec-
aprenyl diphosphate synthase (dps) gene in fission yeast suggests
a role of ubiquinone as an antioxidant. J. Biochem. (Tokyo) 121,
496–505.
19. Koike-Takeshita, A., Koyama, T., Obata, S. & Ogura, K. (1995)
Molecular cloning and nucleotide sequences of the genes for
two essential proteins constituting a novel enzyme system
for heptaprenyl diphosphate synthesis. J. Biol. Chem. 270, 18396–
18400.
20. Ashby, M.N. & Edwards, P.A. (1990) Elucidation of the defi-
ciency in two yeast coenzyme Q mutants. Characterization of the
structural gene encoding hexaprenyl pyrophosphate synthetase.
J. Biol. Chem. 265, 13157–13164.
21.Shimizu,N.,Koyama,T.&Ogura,K.(1998)Molecular
cloning, expression, and characterization of the genes encoding
the two essential protein components of Micrococcus luteus B-P
26 hexaprenyl diphosphate synthase. J. Bacteriol. 180, 1578–
1581.

22. Collin, M.D. & Jones, D. (1981) Distribution of isoprenoid
quinone structural types in bacteria and their taxonomic
implication. Microbiol. Rev. 45, 316–354.
23. Robyt, J. (1998) In Essentials of Carbohydrate Chemistry Chapter
10, pp. 305–318. Springer-Verlag, New York.
24. Sutherland, I.W. (1977) In Surface Carbohydrates of the Proka-
rotic Cell,p.27.AcademicPress,London.
25. Nishikawa, S. & Nakano, A. (1993) Identification of a gene
required for membrane protein retention in the early secretory
pathway. Proc. Natl. Acad. Sci. USA 90, 8179–8183.
26. Sato, M., Sato, K., Nishikawa, S I., Hirata, A., Kato, J I. &
Nakano, A. (1999) The yeast RER2 gene, identified by
endoplasmic reticulum protein localization mutations, encodes
cis-prenyltransferase, a key enzyme in dolichol synthesis. Mol.
Cell. Biol. 19, 471–483.
27. Oh, S.K., Han, K.H., Ryu, S.B. & Kang, H. (2000) Molecular
cloning, expression, and functional analysis of a cis-pre-
nyltransferase from Arabidopsis thaliana. J. Biol. Chem. 275,
18482–18488.
28. Schulbach, M.C., Brennan, P.J. & Crick, D.C. (2000) Identifi-
cation of a short (C
15
) chain Z-isoprenyl diphosphate synthase
and a homologous long (C
50
) chain isoprenyl diphosphate syn-
thase in Mycobacterium tuberculosis. J. Biol. Chem. 275, 22876–
22881.
29. Wang, K.C. & Ohnuma, S I. (2000) Isoprenyl diphosphate
synthases. Biochim. Biophys. Acta 1529, 33–48.

30. Fujihashi, M., Zhang, Y W., Higuchi, Y., Li, X Y., Koyama, T.
& Miki, K. (2001) Crystal structure of cis-prenyl chain elongating
enzyme, undecaprenyl diphosphate synthase. Proc. Natl. Acad.
Sci. USA 98, 4337–4342.
31. Ko, T.P., Chen, Y.K., Robinson, H., Tsai, P.C., Gao, Y G.,
Chen, A.P C., Wang, A.H J. & Liang, P.H. (2001) Mechanism
of product chain length determination and the role of a flexible
loop in E. coli undecaprenyl pyrophosphate synthase catalysis.
J. Biol. Chem. 276, 47474–47482.
32. Clarke, C.F. (1992) Protein isoprenylation and methylation of at
carboxyl-terminal cysteine residue. Annu. Rev. Biochem. 61,355–
386.
33. Glomset, J.A. & Farnsworth, C.C. (1994) Role of protein
modification reactions in programming interactions between
Ras-related FTPases and cell membranes. Annu. Rev. Biol. 10,
181–205.
34. Schafer, W.R. & Rine, J. (1992) Protein prenylation: genes,
enzymes, targets and function. Annu. Rev. Genet. 26, 209–237.
35. Ashar, H.R., Armstrong, L., James, L.J., Carr, D.M., Gray, K.,
Taveras, A., Doll, R.J., Bishop, W.R. & Kirschmeier, P.T. (2000)
Biological effects and mechanism of action of farnesyl transferase
inhibitor. Chem. Res. Toxicol. 13, 949–952.
36. Cohen, L.H., Pieterman, E., van Leeuwen, R.E.W., Overhand,
M.,Burn,B.E.A.,vanderMarel,G.A.&vanBoom,J.H.
(2000) Inhibitors of prenylation of Ras and other G-proteins
and their application as therapeutics. Biochem. Pharmacol. 60,
1061–1068.
37. Park,H.W.,Boduluri,S.R.,Moomaw,J.F.,Casey,P.J.&Beese,
L.S. (1997) Crystal structure of protein farnesyltransferase at 2.25
angstrom resolution. Science 275, 1800–1804.

38. Long, S.B., Casey, P.J. & Beese, L.S. (1998) Cocrystal structure
of protein farnesyltransferase complexed with a farnesyl dipho-
sphate substrate. Biochemistry 37, 9612–9618.
39. Strickland, C.L., Windsor, W.T., Syto, R., Wang, L., Bond, R.,
Wu, Z., Schwart, J., Le, H.V., Beese, L.S. & Weber, P.C. (1998)
Crystal structure of farnesyl protein transferase complexed with a
CaaX peptide and farnesyl diphosphate analogue. Biochemistry
37, 16601–16611.
40. Zhang, H., Seabra, M. & Deisenhofer, J. (2000) Crystal structure
of Rab geranylgeranyltransferase at 2.0 A
˚
resolution. Structure 8,
241–251.
41. Wendt, K.U., Poralla, K. & Schulz, G.E. (1997) Structure and
function of a squalene cyclase. Science 275, 1811–1815.
42. Starks, C.M., Back, K., Chappel, J. & Noel, J.P. (1997) Struc-
tural basis for cyclic terpene biosynthesis by tobacco 5-epi-aris-
tolochene synthase. Science 275, 1815–1820.
43. Lesburg, C.A., Zhai, G., Cane, D.E. & Christianson, D.W.
(1997) Crystal structure of pentalenene synthase: mechanistic
insights on terpenoid cyclization reactions in biology. Science
275, 1820–18224.
44. Rynkiewicz, M.J., Cane, D.E. & Christianson, D.W. (2001)
Structure of trichodiene synthase from Fusarium sporotrichioides
provides mechanistic inferences on the terpene cyclization cas-
cade. Proc. Natl. Acad. Sci. USA 98, 13543–13548.
45. Ogura, K., Toyama, T. & Sagami, H. (1997) Polyprenyl
diphopshate synthases. In Subcellular Biochemistry (Bittman, R.,
ed.), Vol. 28, pp. 57–88. Plenum, New York.
46. Wang, K. & Ohnuma, S. (1999) Enzymatic aspects of isoprenoid

chain elongation. Trends. Biochem. Sci. 24, 445–451.
47. Ashby, M.N., Kutsunai, S.Y., Ackerman, S., Tzagoloff, A. &
Edwards, P.A. (1992) COQ2 is a candidate for the structural gene
encoding para-hydroxybenzoate: polyprenyltransferase. J. Biol.
Chem. 267, 4128–4136.
48. Koyama, T., Obata, S., Osabe, M., Takeshita, A., Yokoyama,
K., Uchita, M., Nishino, T. & Ogura, K. (1993) Thermostable
farnesyl diphosphate synthase of Bacillus stearothermophilus:
molecular cloning, sequence determination, overproduction, and
purification. J. Biochem. (Tokyo) 113, 355–363.
49. 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.
50. 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.
51. Marrero, P.F., Poulter, C.D. & Edwards, P.A. (1992) Effects of
site-directed mutagenesis of the highly conserved aspartate
residues in domain II of farnesyl diphosphate synthase activity.
J. Biol. Chem. 267, 21873–21878.
3352 P H. Liang et al.(Eur. J. Biochem. 269) Ó FEBS 2002
52. Joly, A. & Edwards, P.A. (1993) Effect of site-directed muta-
genesis of conserved aspartate and arginine residues upon
farnesyl diphosphate synthase activity. J. Biol. Chem. 268, 26983–
26989.
53. Song, L. & Poulter, C.D. (1994) Yeast farnesyl diphosphate
synthase: site-directed mutagenesis of residues in highly con-

served prenyltransferase domain I and II. Proc. Natl. Acad. Sci.
U.S.A. 91, 3044–3048.
54. Koyama, T., Tajima, M., Nishino, T. & Ogura, K. (1995) Sig-
nificance of Phe220 and Gln-221 in the catalytic mechanism of
farnesyl diphosphate synthase of Bacillus stearothermophilus.
Biochem. Biophys. Res. Commun. 212, 681–686.
55. Koyama, T., Tajima, M., Sano, H., Doi, T., Koike-Takeshita,
A., Obata, S., Nishino, T. & Ogura, K. (1996) Identification of
significant residues in the substrate binding site of Bacillus
stearothermophilus farnesyl diphosphate synthase. Biochemistry
35, 9533–9538.
56. Koyama, T., Obata, S., Saito, K., Takeshita-Koike, A. & Ogura,
K. (1994) Structural and functional roles of the cysteine residues
of Bacillus stearothermophilus farnesyl pyrophosphate synthase.
Biochemistry 33, 12644–12648.
57. 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 synthase. Proc. Natl. Acad. Sci. USA 93,
15018–15023.
58. Ohnuma, S I., Narita, K., Nakazawa, T., Hemmi, H., Hallberg,
A M., Koyama, T., Ogura, K. & Nishito, T. (1996) Conversion
from farnesyl diphosphate synthase to geranylgeranyl diphos-
phate synthase by random chemical mutagenesis. J. Biol. Chem.
271, 10087–10095.
59. Ohnuma, S I., Narita, K., Nakazawa, T., Ishida, C., Takeuchi,
Y., Ohto, C. & Nishito, 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
product. J. Biol. Chem. 271, 30748–30754.
60. Ohnuma, S I., Hirooka, K., Hemmi, H., Ishida, C., Ohto, C. &

Nishito, T. (1996) Conversion of product specificity of archae-
bacterial geranylgeranyl diphosphate synthase. J. Biol. Chem.
271, 18831–18837.
61. Ohnuma, S I., Hirooka, K., Tsuruoka, N., Yano, M., Ohto, C.,
Nakane, H. & Nishito, T. (1998) A pathway where polyprenyl
diphosphate elongates in prenyltransferase. J. Biol. Chem. 273,
26705–26713.
62. Fernandez, S.M.S., Kellogg, B.A. & Poulter, C.D. (2000) Far-
nesyl diphosphate synthase. Altering the catalytic site to select for
geranyl diphosphate activity. Biochemistry 39, 15316–15321.
63. Allen, C.M. (1985) Purification and characterization of undeca-
prenyl pyrophosphate synthase. Methods Enzymol. 110, 281–299.
64. Shimizu, N., Koyama, T. & Ogura, K. (1998) Molecular cloning,
expression, and purification of undecaprenyl diphosphate syn-
thase. J. Biol. Chem. 273, 19476–19481.
65. Apfel, C.M., Takacs, B., Fountoulakis, M., Stieger, M. & Keck,
W. (1999) Use of genomics to identify bacterial undecaprenyl
pyrophosphate synthetase: cloning, expression and character-
ization of the essential uppS gene. J. Bacteriol. 181, 483–492.
66. Pan, J.J., Yang, L.W. & Liang, P.H. (2000) Effect of site-directed
mutagenesis of the conserved aspartate and glutamate on E. coli
undecaprenyl pyrophosphate synthase catalysis. Biochemistry 39,
13856–13861.
67. Kinoshita, K., Sadanami, K., Kidera, A. & Go, N. (1999)
Structural motif of phosphate-binding site common to various
protein superfamilies: all-against-all structural comparison of
protein–mononucleotide complexes. Protein Eng. 12, 11–14.
68. Kharel,Y.,Zhang,Y W.,Fujihashi,M.,Miki,K.&Koyama,
T. (2001) Identification of significant residues for homoallylic
substrate binding of Micrococcus luteus B-P 26 undecaprenyl

diphosphate synthase. J. Biol. Chem. 276, 28459–28464.
69. Fujikura, K., Zhang, Y W., Yoshizki, T. & Koyama, T. (2000)
Significance of Asn-77 and Trp-78 in the catalytic function of
undecaprenyl diphosphate synthase of Micrococcus luteus B-P
26. J. Biochem. (Tokyo) 128, 917–922.
70. Chen, Y.H., Chen, A.P C., Chen, C.T., Wang, A.H J. & Liang,
P.H. (2002) Probing the conformational change of Escherichia
coli undecaprenyl pyrophosphate synthase during catalysis using
an inhibitor and tryptophan mutants. J. Biol. Chem. 277, 7369–
7376.
71. Poulter, C.D. & Rilling, H.C. (1976) Prenyltransferase: the
mechanism of the reaction. Biochemistry 15, 1079–1083.
72. Poulter, C.D. & Rilling, H.C. (1978) The prenyl transfer reaction.
Enzymatic and mechanistic studies of the 1¢-4 coupling reaction
in the terpene biosynthetic pathway. Acc. Chem. Res. 11, 307–
311.
73. Rahier, A., Taton, M. & Benveniste, P. (1990) Inhibition of sterol
biosynthesis enzymes in vitro by analogues of high-energy car-
bocationic intermediates. Biochem. Soc. Trans. 18, 48–52.
74. Steiger, A., Pyun, H J. & Coates, R.M. (1992) Synthesis and
characterization of aza analog inhibitors of squalene and
geranylgeranyl diphosphate synthases. J. Org. Chem. 57, 3444–
3449.
75. Sagami, H., Korenaga, T., Ogura, K., Steiger, A., Pyun, H.J. &
Coates, R.M. (1992) Studies on geranylgeranyl diphosphate
synthase from rat liver: specific inhibition by 3-azageranylgeranyl
diphosphate. Arch. Biochem. Biophys. 297, 314–320.
76. Pan, J.J., Chiou, S.T. & Liang, P.H. (2000) Product distribution
and pre-steady-state kinetic analysis of Escherichia coli
undecaprenyl pyrophosphate synthase reaction. Biochemistry 39,

10936–10942.
77. Pan, J.J., Kuo, T.H., Chen, Y.K., Yang, L.W. & Liang, P.H.
(2002) Insight into the activation mechanism of E. coli octaprenyl
pyrophosphate synthase derived from pre-steady-state kinetic
analysis. Biochim. Biophys. Acta 1594, 64–73.
78. Chang, S.Y., Tsai, P.C., Cheng, C.S. & Liang, P.H. (2001)
Refolding and characterization of a yeast dehydrodolichyl
pyrophosphate synthase overexpressed in E. coli. Protein Exp.
Purif. 23, 432–439.
79. Hemmi, H., Yamashita, S., Shimoyama, T., Nakayama, T. &
Nishino, T. (2001) Cloning, expression, and characterization of
cis-polyprenyl diphosphate synthase from Archaeon Sulfolobus
acidocadarius. J. Bacteriol. 183, 401–404.
80. Matsuoka, S., Sagami, H., Kurisaki, A. & Ogura, K. (1991) Vari-
able product specificity of microsomal dehydrodolichyl diphos-
phate synthase from rat liver. J. Biol. Chem. 266, 3464–3468.
81. Fu, H.W., Moomaw, J.F., Moomaw, C.R. & Casey, P.J. (1996)
Identification of a cysteine residue essential for activity of protein
farnesyltransferase. Cys299 is exposed only upon removal of zinc
from the enzyme. J. Biol. Chem. 271, 28541–28548.
82. Stradley, S.J., Rizo, J. & Gierasch, L.M. (1993) Conformation of
a heptapeptide substrate bound to protein farnesyltransferase.
Biochemistry 32, 12586–12590.
83. Kral, A.M., Diehl, R.E., deSolms, S.J., Williams, T.M., Kohl,
N.E. & Omer, C.A. (1997) Mutational analysis of conserved
residues of the b-Subunit of human farnesyl: protein transferase.
J. Biol. Chem. 272, 27319–27322.
84. Andres, D.A., Goldstein, J.L., Ho, Y.K. & Brown, M.S. (1993)
Mutational analysis of alpha-subunit of protein farnesyl-
transferase. Evidence for a catalytic role. J. Biol. Chem. 268,

1383–1390.
85. Yokoyama, K., Zimmerman, K., Scholten, J. & Gelb, M.H.
(1997) Differential prenyl pyrophosphate binding to mammalian
protein geranylgeranyltransferase-I and protein farnesyl-
transferase and its consequence on the specificity of protein
Prenylation. J. Biol. Chem. 272, 3944–3952.
86. Pfeffer, S.R. (1994) Rab GTPases: master regulators of mem-
brane trafficking. Curr. Opin. Cell Biol. 6, 522–526.
Ó FEBS 2002 Structure, mechanism and function of prenyltransferases (Eur. J. Biochem. 269) 3353
87. Novick, P. & Zerial, M. (1997) The diversity of Rab proteins in
vesicle transport. Curr. Opin. Cell Biol. 9, 496–504.
88. Farnsworth, C.C., Seabra, M.C., Ericsson, L.H., Gelb, M.H. &
Glomset, J.A. (1994) Rab geranylgeranyl transferase catalyzes
the geranylgeranylation of adjacent cysteines in the samll
GTPases Rab1A, Rab3A, and Rab5A. Proc. Natl. Acad. Sci.
USA 91, 11963–11967.
89. Furfine, E., Leban, J., Landavazo, A., Moomaw, J. & Casey, P.
(1995) Protein farnesyltransferase: kinetics of farnesyl
pyrophosphate binding and product release. Biochemistry 34,
6857–6862.
90. Dolence, J.M., Cassidy, P.B., Mathis, J.R. & Poulter, C.D. (1995)
Yeast protein farnesyltransferase: steady-state kinetic studies of
substrate binding. Biochemistry 34, 16687–16694.
91. Pompliano, D., Schaber, M., Mosser, S., Omer, C., Shafer, J. &
Gibbs, J. (1993) Isoprenoid diphosphate utilization by
recombinant human farnesyl: protein transferase: interactive
binding between substrates and a preferred kinetic pathway.
Biochemistry 32, 8341–8347.
92. Huang, C.C., Hightower, K.E. & Fierke, C.A. (2000) Mechan-
istic studies of rat protein farnesyltransferase indicate an associ-

ate transition state. Biochemistry 39, 2593–2602.
93. Dolence, J. & Poulter, C.D. (1995) A mechanism for post-
translational modifications of proteins by yeast protein farne-
syltransferase. Proc. Natl. Acad. Sci. USA 92, 5008–5011.
94. Huang, C.C., Casey, P.J. & Fierke, C.A. (1997) Evidence for a
catalytic role of zinc in protein farnesyltransferase. Spectroscopy
of Co
2+
-farnesyltransferase indicates metal coordination of the
substrate thiolate. J. Biol. Chem. 272, 20–23.
95. Hightower, K.E., Huang, C.C., Casey, P.J. & Fierke, C.A. (1998)
H-Ras peptide and protein substrates bind protein farnesyl-
transferase as an ionized thiolate. Biochemistry 37, 15555–15562.
96. Saderholm, M.J., Hightower, K.E. & Fierke, C.A. (2000) Role of
metals in the reaction catalyzed by protein farnesyltransferase.
Biochemistry 39, 12398–12405.
97. Cane, D.E., Sohng, J.K., Lamberson, C.R., Rudnicki, S.M., Wu,
Z.,Lloyd,M.D.,Oliver,J.S.&Hubbard,B.R.(1994)Pentale-
nene synthase. Purification, molecular cloning, sequencing, and
high-level expression in Escherichia coli of a terpenoid cyclase
from Streptomyces UC5319. Biochemistry 33, 5846–5857.
98. Cane, D.E., Xue, Q. & Fitzsimons, B.C. (1996) Trichodiene
synthase. Probing the role of the highly conserved aspartate-rich
region by site-directed mutagenesis. Biochemistry 35, 12369–
12376.
99. Alzari, P.M., Souchon, H. & Dominguez, R. (1996) The crystal
structure of endoglucanase CelA, a family 8 glycosyl hydrolase
from Clostridium thermocellum. Structure 4, 265–275.
100. Feil, C., Sussmuth, R., Jung, G. & Poralla, K. (1996) Site-
directed mutagenesis of putative active-site residues in squalene-

hopene cyclase. Eur. J. Biochem. 242, 51–55.
3354 P H. Liang et al.(Eur. J. Biochem. 269) Ó FEBS 2002

×