Mutation and inhibition studies of mevalonate 5-diphosphate decarboxylase

Article abstract of DOI:10.1016/j.bmcl.2007.09.033

Mevalonate 5-diphosphate decarboxylase plays an important role in regulating cholesterol biosynthesis, which was studied through incubation with various synthetic substrate analogs and characterization of mutated enzymes. The results are potentially useful for further developing inhibitors that block the mevalonate pathway which is a drug target for treating cardiovascular disease and cancer.

Full text of DOI:10.1016/j.bmcl.2007.09.033

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 17 (2007) 6164–6168
Mutation and inhibition studies of mevalonate
5-diphosphate decarboxylase
Yongge Qiu,a,b Jinbo Gao,a Fei Guo,a Yuqin Qiaoa and Ding Lia,*
aDepartment of Biology and Chemistry, City University of Hong Kong, 83 Tat Chee Avenue, Kowloon, Hong Kong SAR, PR China
bChemistry Department, Hanshan Normal University, Chaozhou City 521041, Guang Dong Province, PR China
Received 5 June 2007; revised 5 September 2007; accepted 6 September 2007
Available online 11 September 2007
Abstract—Mevalonate 5-diphosphate decarboxylase plays an important role in regulating cholesterol biosynthesis, which was stud-
ied through incubation with various synthetic substrate analogs and characterization of mutated enzymes. The results are potentially
useful for further developing inhibitors that block the mevalonate pathway which is a drug target for treating cardiovascular disease
and cancer.
Ó 2007 Elsevier Ltd. All rights reserved.
In animal cells, the cholesterol biosynthetic pathway
contains a unique series of three sequential ATP-depen-
dent enzymes that convert mevalonate to isopentenyl
diphosphate (IPP): mevalonate kinase (MVK), phos-
phomevalonate kinase (PMK), and mevalonate 5-
diphosphate decarboxylase (MDD) (Fig. 1).1 These
three enzymes catalyze consecutive steps downstream
from the HMG-CoA reductase in the mevalonate path-
way, and are responsive for cholesterol intake in ani-
mals. The pathway has been exploited in the design of
drugs treating cardiovascular disease2 and cancer.3 In
addition, inhibition of insect juvenile hormone biosyn-
thesis by fluorinated mevalonate analogs has been re-
ported as a potential method for pest control.4 Several
investigators have suggested the involvement of MDD
as important regulatory steps in the biosynthesis of
cholesterol.5
The X-ray structure of MDD from yeast has been solved
without bound ligand,6 and the enzyme belongs to
GHMP (Galactokinase, Homoserine kinase, Mevalo-
nate kinase, and Phosphomevalonate kinase) superfam-
ily of small-molecule kinases. Kinases are a ubiquitous
group of enzymes central to many biochemical processes
such as metabolism, gene regulation, and signal trans-
duction.7 A variety of substrate analogs of MDD target-
ing mevalonate 5-diphosphate binding site have been
studied,8 and some have been found to be enzyme inhib-
itors. In the present study, we report our further investi-
gation of rat MDD through characterization of mutated
enzymes. A variety of substrate analogs were synthe-
sized and incubated with rat MDD, and structure–activ-
ity relationships were obtained for the enzymatic
reaction. The study increased our understanding of
MDD and may be useful for drug discovery for the pur-
pose of treating cardiovascular disease and cancer, or
for pest control.
Abbreviations: FPP, farnesyl diphosphate; GGPP, geranylgeranyl
diphosphate; GHMP, galactokinase, homoserine kinase, mevalonate
kinase, and phosphomevalonate kinase; GPP, geranyl diphosphate;
HMG-CoA reductase, 3-hydroxy-3-methylglutaryl-CoA reductase;
IPP, isopentenyl diphosphate; MDD, mevalonate 5-diphosphate dec-
arboxylase, also known as mevalonate 5-pyrophosphate decarboxylase
or MPD; MVAPP, mevalonate 5-diphosphate; MVK, mevalonate ki-
nase; PMK, phosphomevalonate kinase; PP, pyrophosphate or diph-
osphate; TsCl, toluene sulfonyl chloride.
The rat liver MDD with N-terminal hexa His-tag was
obtained and assayed as previously described.8c Tris
buffer was used instead of phosphate buffer for storing
the enzyme and for the enzyme assays. A sequence align-
ment analysis of MDDs from different sources indicates
that acidic residue D306 and basic residues K23, K27,
R154, R162, K208 are highly conserved, and may have
important roles in substrate binding or catalysis. Mutant
expression plasmids of rat MDD K23A, K27A, R154A,
R162A, K208A, D306A, D306E, D306N were subse-
quently constructed using site-directed mutagenesis.
Keywords: Mevalonate 5-diphosphate decarboxylase; Substrate ana-
log; Mutation; Cholesterol; Cardiovascular disease; Cancer.
*
Corresponding author. Tel.: +852 2788 7824; fax: +852 2788
7406; e-mail: bhdingli@cityu.edu.hk
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.09.033

Y. Qiu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6164–6168
6165
O
O
O
O
HO
HO
HMG-CoA
reductase
MVK
SCoA
Acetyl-CoA
HO
SCoA
HO
OH
HMG-CoA
Mevalonic acid
O
HO
O
HO
Terpenoids
MDD
PMK
Steroid Hormones
Cholesterol
HO
OPP
HO
OP
OPP
Isopentenyl-
pyrophosphate (IPP)
Mevalonate-5-diphosphate
(MVAPP)
Mevalonate-5-phosphate
Figure 1. Mevalonate pathway in animal cells.
´
˚
6
The mutated enzymes were overexpressed and purified
with a nickel metal affinity column to apparent homoge-
neity based on SDS–PAGE. Spectrophotometric activ-
ity assays demonstrated that mutants K23A, R162A,
K208A, D306A, D306E, and D306N have no detectable
activity. The low activities of mutants K27A and R154A
were detected and their kinetic parameters were ob-
tained as shown in Table 1. For K27A, the Vmax value
is decreased to 0.14 0.01 lmol/min/mg, and KM values
increased to 0.75 0.16 mM for mevalonate 5-diphos-
phate and 1.87 0.15 mM for ATP. For R154A, the
Vmax value is 0.34 0.02 lmol/min/mg, and KM values
are 0.92 0.29 mM and 1.40 0.20 mM for mevalonate
5-diphosphate and ATP, respectively.
Asp302 (2.75 A), forming an invariant acid/base pair.
The activity of the mutated enzyme K23A is below
detectable levels, indicating that the residue is critical
for catalysis or substrate binding. Krepkiy and Mizi-
orko reported a 30-fold increase in activity and 16-
fold inflation of KM for ATP for yeast MDD K18M
mutant, indicating that Lys18 influences the active site
but is not crucial for reaction chemistry.10 Our results
show that the mutation of the residue to alanine has a
larger effect on the activity of the protein. Comparing
with the case of mevalonate kinase, residue K23 in rat
MDD may function like K13 in rat MVK to interact
with c-phosphoryl group of ATP and 3-OH of meva-
lonate 5-diphosphate by salt bridge or hydrogen
bond.9b
The MDD-catalyzed reaction has been proved to be a
stepwise process, which includes phosphorylation at 3-
OH and decarboxylation.8a For mevalonate kinase,
another member of GHMP superfamily, the highly con-
served acidic residue D204 is determined to play a cru-
cial catalytic role for phosphorylation at 5-OH of the
substrate.9 It is reasonable to assume that a conserved
aspartate residue in MDD may also work as a general
base to abstract proton from 3-hydroxy of the substrate.
A sequence alignment demonstrates that D306 is located
in a highly conserved motif (306DAGPN310), and the
three mutants D306A, D306E, and D306N were all
found to have no detectable activity. This result strongly
suggests that D306 has a catalytic role in the reaction.
During the progress of this work, Krepkiy and Miziorko
reported their work with assignment of a crucial cata-
lytic role to Asp302 of yeast MDD,10 which is consistent
with our result.
A 47-fold decrease in activity was found for the K27
mutant. A 21-fold inflation of KM for mevalonate 5-
diphosphate and 3.5-fold inflation of KM for ATP were
found for the mutant. Though the residue K27 is also lo-
cated in the same invariant motif of MDD with K23, the
assay result indicates that K27 is less significant for the
reaction. The changes of KM suggest that K27 affects the
binding of mevalonate 5-diphosphate more than ATP.
The residues R154 and R162 are located in another
highly conserved motif of MDD
(
154RRGSGS
ACRS163). The activities of their corresponding mutated
enzymes are quite different. The activity of R154A mu-
tant was decreased by 19-fold, while no activity was
found for the R162A mutant, suggesting the residue
R162 may play an important role for the reaction by
providing a positive center. The Ser153 of yeast MDD
in the same conserved motif was proposed to bind with
Mg-ATP at the MDD active site.11 Finally, the activity
of K208A mutant is undetectable by spectrophotometric
measurement, indicating that the residue is also crucial
for the activity of the enzyme. More work is necessary
to investigate its catalytic or structural role.
Based on our assumption that positively charged resi-
dues may directly interact with negatively charged sub-
strates (mevalonate 5-diphosphate and ATP), the
highly conserved basic residues K23, K27, R154,
R162, and K208 were chosen for mutation. The residue
K23 is located at
a
highly conserved motif
The above results indicate that mutations of highly
conserved acidic residue D306 and basic residues
K23, K27, R154, and R162 in rat MDD to alanine
caused significant changes in enzymatic activities.
D306 may play a crucial role for the catalytic reaction
as a general base to abstract proton from 3-OH of
mevalonate 5-diphosphate. The mutated basic residues
may function critically in binding with negatively
charged substrates, or play a role in the second step
of the catalytic reaction: elimination of phosphate
and carbon dioxide.
(23KYWGK27). From the crystal structure of yeast
MDD, the corresponding Lys18 is very close to
Table 1. Kinetic parameters for MDD wild-type and mutated enzymes
MDD
Vmax
KM (MVAPP)
KM (ATP)
(lmol/min/mg)
(mM)
(mM)
Wild type
K27A
R154A
6.60 0.31
0.14 0.01
0.34 0.02
0.036 0.003
0.75 0.16
0.92 0.29
0.53 0.09
1.87 0.15
1.40 0.20

6166
Y. Qiu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6164–6168
Table 2. IC50 values of substrate analogs for rat MDD
O
TsCl
OH
R
RX + Mg
Compound
IC50 for rat MDD (lM)
7.5
20
pyridine
75%
HO
50%
HO
4
8
9, R1 = Methyl
10, R2 = Ethyl
11, R3 = Propyl
12, R4 = Benzyl
X = Br or I
17
18
19
20
25
60
80
120
140
50
OH
R
OH
R
(Bu4N)3HP2O7
50%
PPO
TsO
17, R1 = Methyl
18, R2 = Ethyl
19, R3 = Propyl
20, R4 = Benzyl
13, R1 = Methyl
14, R2 = Ethyl
15, R3 = Propyl
16, R4 = Benzyl
In order to further understand rat MDD catalyzed reac-
tion, we synthesized a variety of substrate analogs,
which were incubated with the enzyme. 2-Fluoromeval-
onate 5-diphosphate has been found to be an irreversible
inhibitor of rat MDD,8c and we further synthesized 2,2-
difluoromevalonate 5-diphosphate (4)12 and 3-ethyl-2-
fluoro-3,5-dihydroxy-pentanate 5-diphosphate (8)13 as
shown in Figures 2 and 3 for inhibition studies. 4-Hy-
droxy-2-butanone reacted with 4-toluene sulfonyl
chloride in the presence of pyridine to give 4-tosyloxy-
2-butanone (1), which was then reacted with fluorinated
ethyl 2-bromoacetate through Reformatsky reaction
yielding compound 2. Compound 3 was obtained from
compound 2 through a reaction with pyrophosphate,
and its subsequent hydrolysis gave 2,2-difluoromevalo-
nate 5-diphosphate (4). Compound 8 was synthesized
following same synthetic strategy as that for compound
4. Both compounds were found to be competitive inhib-
itors only, since the inhibitions were not time-depen-
dent. Both inhibitions were competitive with
mevalonate 5-diphosphate, and the IC50 values of 7.5
and 20 lM were determined for compounds 4 and 8,
respectively, as shown in Table 2. These results indicated
that an extra fluorine at carbon 2 position and the
replacement of methyl group at carbon 3 with an
Figure 4. Organic syntheses of substrate analogs (17–20).
ethyl group will both affect the nucleophilic attack
at carbon 2 by an enzyme nucleophile, which was pro-
posed for inhibition of MDD by 2-fluoromevalonate
5-diphosphate.
Based on our results from the site directed mutagenesis
study, several basic residues play important role in enzy-
matic reaction, which can probably interact with nega-
tively charged substrate. Therefore, it is interesting to
know the kinetic properties of the substrate analogs with
carboxylate group replaced with various other func-
tional groups. Since the charged group usually makes
the molecule difficult to penetrate the cell membrane,
it would be better if it can be replaced with a neutral
group. A variety of substrate analogs (17–20, 25) were
synthesized14 for this purpose as shown in Figures 4
and 5. Compounds 17–20 were synthesized using a
Zn
OH
O
O
TsCl
CO2C2H5
BrCF2CO2C2H5
50%
TsO
HO
TsO
pyridine
75%
F
F
1
2
OH
OH
-
(Bu4N)3HP2O7
CO2
0.5 N LiOH
CO2C2H5
PPO
PPO
F
F
F
F
4
3
yield from 2 to 4: 40%
Figure 2. Organic synthesis of 2,2-difluoromevaloate 5-diphosphate (4).
O
O
Zn
OH
TsCl
CO2C2H5
TsO
BrCHFCO2C2H5
50%
HO
TsO
pyridine
75%
5
6
F
OH
OH
-
(Bu4N)3HP2O7
CO2
0.5 N LiOH
CO2C2H5
PPO
PPO
F
8
F
7
yield from 6 to 8: 40%
Figure 3. Organic synthesis of 3-ethyl-2-fluoro-3,5-dihydroxy-pentanate 5-diphosphate (8).

Y. Qiu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6164–6168
6167
TsCl
OH
OH
Bu4NF
OTs
THF
pyridine
75%
O
O
O
O
OH
F
O
O
21
23
22
(1) H+, THF / CH3OH
OH
OH
TsO
F
(2) TsCl, pyridine, CH2Cl2
24
OH
(Bu4N)3HP2O7
CH3CN
PPO
F
25
Figure 5. Organic synthesis of substrate analog (25).
5. (a) Iglesias, J.; Gonzalezpacanowska, D.; Marco, C.;
Garciaperegrin, E. Biochem. J. 1989, 260, 333; (b) Gon-
zalezpacanowska, D.; Marco, C.; Garciamartinez, J.;
Garciaperegrin, E. Biochim. Biophys. Acta 1985, 833,
449; (c) Jabalquinto, A. M.; Cardemil, E. Arch. Biochem.
Biophys. 1981, 210, 132; (d) Gonzalezpacanowska, D.;
Marco, C.; Garciamartinez, J.; Linares, A.; Garciapere-
grin, E. Nutr. Rep. Int. 1985, 31, 121; (e) Jabalquinto, A.
M.; Cardemil, E. Lipids 1980, 15, 196; (f) Gonza-
lezpacanowska, D.; Marco, C.; Garciamartinez, J.; Gar-
ciaperegrin, E. Biochim. Biophys. Acta 1986, 875, 605.
6. Bonanno, J. B.; Edo, C.; Eswar, N.; Pieper, U.; Roma-
nowski, M. J.; Ilyin, V.; Gerchman, S. E.; Kycia, H.;
Studier, F. W.; Sali, A.; Burley, S. K. Proc. Natl. Acad.
Sci. U.S.A. 2001, 98, 12896.
similar strategy to that for the synthesis of compound 4.
Difference is that we used the Grignard reaction here in-
stead of the Reformatsky reaction, since the reagents do
not have carboxylate ester functional group. Two equiv-
alents of Grignard reagent were used, with the first
equivalent reagent consumed by reacting with hydroxyl
group, which saved one protection and one deprotection
steps. The synthesis of compound 25 also followed a
similar strategy, which includes two stepwise SN2 reac-
tions on two sides of the molecule with some protection
and deprotection steps. These compounds were found to
be mild inhibitors of rat MDD as shown in Table 2, and
all are competitive with respect to mevalonate 5-diphos-
phate, suggesting that carboxyl group is relatively
important for the binding and reaction of the substrate.
7. Matte, A.; Tari, L. W.; Delbaere, L. T. J. Struct. Fold.
Des. 1998, 6, 413.
8. (a) Dhe-Paganon, S.; Magrath, J.; Abeles, R. H. Biochem-
istry 1994, 33, 13355; (b) Vlattas, I.; Dellureficio, J.; Ku,
E.; Bohacek, R.; Zhang, X. Bioorg. Med. Chem. Lett.
1996, 6, 2091; (c) Qiu, Y.; Li, D. Biochim. Biophys. Acta
2006, 1760, 1080; (d) Qiu, Y.; Li, D. Org. Lett. 2006, 8,
1013.
9. (a) Potter, D.; Miziorko, H. M. J. Biol. Chem. 1997, 272,
25449; (b) Fu, Z.; Wang, M.; Potter, D.; Miziorko, H. M.;
Kim, J. J. J. Biol. Chem. 2002, 277, 18134.
In summary, MDD is an essential enzyme in mevalonate
pathway regulating cholesterol biosynthesis. The en-
zyme was studied through incubation with various syn-
thetic substrate analogs and characterization of several
mutated enzymes. The results increased our understand-
ing of MDD and are potentially useful for developing
inhibitors that target the mevalonate pathway for treat-
ment of cardiovascular disease and cancer.
10. Krepkiy, D.; Miziorko, H. M. Protein Sci. 2004, 13, 1875.
11. Krepkiy, D.; Miziorko, H. M. Biochemistry 2005, 44,
2671.
Acknowledgment
12. The spectra data of 2,2-difluoromevaloate 5-diphosphate
(4) are shown as following: 1H NMR (300 MHz, D2O,
TMS) d 1.34 (s, 3H, CH3), 1.92–2.11 (m, 2H), 4.07–4.13
(m, 2H, POCH2); 31P NMR (121 MHz, D2O) d À1.03 (m).
13. The spectra data of 3-ethyl-2-fluoro-3,5-dihydroxy-pen-
tanate 5-diphosphate (8) are shown as following: 1H NMR
(300 MHz, D2O, TMS) d 0.95 (t, J = 7.5 Hz, 3H), 1.67 (m,
2H), 2.02 (m, 2H), 4.10 (q, 2H), 4.78 (m, 1H, partially
mixed with solvent signals); 31P NMR (121 MHz, D2O). d
À7.71 (d, J = 18.3 Hz), À3.54 (d, J = 19.0 Hz).
The work described in this paper was substantially sup-
ported by a grant from the City University of Hong
Kong (Strategic Research Grant, Project No. 7001793).
References and notes
14. (a) The spectra data of 3-hydroxy-3-methyl-butanyl pyro-
phosphate (17) are shown as following: 1H NMR
(300 MHz, D2O, TMS) d 1.28 (s, 6H, 2CH3), 1.92 (t,
J = 7.2 Hz, 2H, CH2C), 4.04–4.11 (m, 2 H, OCH2); 31P
NMR (121 MHz, D2O) d À6.30 (d, J = 17.0 Hz), À5.90
(d, 17.5 Hz). (b) The spectra data of 3-hydroxy-3-methyl-
1. (a) Goldstein, J. L.; Brown, M. S. Nature 1990, 343, 425;
(b) Houten, S. M.; Schneiders, M. S.; Wanders, R. J. A.;
Waterham, H. R. J. Biol. Chem. 2003, 278, 5736; (c) Kaur,
M.; Kaul, D. FASEB J. 1997, 11, A1217.
2. (a) Brown, M. S.; Goldstein, J. L. Science 1986, 232, 34;
(b) Ma, P. T. S.; Gil, G.; Sudhof, T. C.; Bilheimer, D. W.;
Goldstein, J. L.; Brown, M. S. Proc. Natl. Acad. Sci.
U.S.A. 1986, 83, 8370.
3. Schafer, W. R.; Kim, R.; Sterne, R.; Thorner, J.; Kim, S.
H.; Rine, J. Science 1989, 245, 379.
4. Quistad, G. B.; Cerf, D. C.; Schooley, D. A.; Staal, G. B.
Nature 1981, 289, 176.
1
pentanyl pyrophosphate (18) are shown as following: H
NMR (300 MHz, D2O) d 0.83 (t, 7.2 Hz, 3H, CH2CH3),
1.14 (s, 3H, CH3), 1.48 (q, J = 7.5 Hz, 2H, CH2CH3), 1.82
(t, J = 6.8 Hz, 2H, CH2C), 3.98 (q, J = 7.2 Hz, 2H,
OCH2); 13C NMR (75 MHz, D2O) d 8.1, 25.8, 34.2,
40.6, 63.1, 73.3; 31P NMR (121 MHz, D2O) d À6.10 (d,

6168
Y. Qiu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6164–6168
J = 20.2 Hz), À3.65 (d, J = 19.6 Hz). (c) The spectra
4.11–4.18 (m, 2H, OCH2), 7.15–7.26 (m, 5 H, ArH); 31P
NMR (121 MHz, D2O) d À5.97 (d, J = 18.9 Hz), À4.68
(d, J = 19.6 Hz); 13C NMR (75 MHz, D2O) d 26.1, 41.6,
48.2, 63.4, 73.1, 126.0, 128.9, 131.6, 138.4. (e) The
spectra data of 5-fluoro-3-hydroxy-3-methyl-pentanyl
diphosphate (25) are shown as following: 1H NMR
data of 3-hydroxy-3-methyl-hexyl pyrophosphate (19)
are shown as following: 1H NMR (300 MHz, D2O) d
0.88 (t, J = 6.9 Hz, 3H, CH2CH3), 1.19 (s, 3H, CH3),
1.28–1.48 (m, 4H, CH2CH2CH3), 1.86 (t, J = 7.2 Hz,
2H, CH2C), 3.99–4.06 (q, 2H, OCH2); 31P NMR
(121 MHz, D2O) d À6.12 (d, J = 19.5 Hz), À3.82 (d,
J = 19.5 Hz). (d) The spectra data of 3-hydroxy-3-
phenyl-butanyl pyrophosphate (20) are shown as fol-
(300 MHz, D2O)
d 1.20 (s, 3H, CH3), 1.85 (t,
J = 6.6 Hz, 2H, CH2), 1.95 (t, J = 6.0 Hz, 2H, CH2),
3.99 (q, 2H, OCH2), 4.62 (dt, JHF = 46.2 Hz,
J2 = 5.9 Hz, 2H, CH2F); 31P NMR (121 MHz, D2O) d
À6.67 (d, J = 22.0 Hz), À6.45 (d, J = 23.2 Hz).
1
lowing: H NMR (300 MHz, D2O) d 1.22 (s, 3H, CH3),
1.93 (t, J = 7.2 Hz, 2H, CH2C), 2.88 (s, 2H, CH2Ph),