Studies on the nonmevalonate pathway of terpene biosynthesis
The role of 2
C
-methyl-D-erythritol 2,4-cyclodiphosphate in plants
Monika Fellermeier
1
, Maja Raschke
1
, Silvia Sagner
1
, Juraithip Wungsintaweekul
2
, Christoph A. Schuhr
2
,
Stefan Hecht
2
, Klaus Kis
2
, Tanja Radykewicz
2
, Petra Adam
2
, Felix Rohdich
2
, Wolfgang Eisenreich
2
,
Adelbert Bacher
2
, Duilio Arigoni

phate by the consecutive action of IspD, IspE, and IspF
proteins in the nonmevalonate pathway of terpenoid
biosynthesis. To complement previous work with radio-
labelled precursors, we have now demonstrated that
[U-
13
C
5
]2C-methyl-D-erythritol 4-phosphate affords
[U-
13
C
5
]2C-methyl-D-erythritol 2,4-cyclodiphosphate in
isolated chromoplasts of Capsicum annuum and Narcissus
pseudonarcissus. Moreover, chromoplasts are shown to
efficiently convert 2C-methyl-
D-erythritol 4-phosphate as
well as 2C-methyl-
D-erythritol 2,4-cyclodiphosphate into
the carotene precursor phytoene. The bulk of the kinetic data
collected in competition experiments with radiolabeled
substrates is consistent with the notion that the cyclodipho-
sphate is an obligatory intermediate in the nonmevalonate
pathway to terpenes. Studies with [2,2
0
-
13
C
2

lose, a known precursor of the vitamins thiamine [11] and
pyridoxol [12], could be incorporated into terpenoids by
E. coli [9] as well as by higher plants [7]. More specifically,
plants were shown to utilize the mevalonate pathway in the
cytoplasmic compartment and the nonmevalonate pathway
in the plastid compartment [7,10,13,14]. More recently, the
origin of a variety of plant terpenoids could be assigned to
this plastid-based nonmevalonate pathway (reviewed in [6]).
Recent studies by several research groups identified
1-deoxy-
D-xylulose 5-phosphate synthase as the first
enzyme of the alternative terpenoid pathway in certain
bacteria [15–17] and plants [18,19]. The enzyme product is
converted into the branched chain polyol, 2C-methyl-
D-erythritol 4-phosphate, by a reductoisomerase via a
skeletal rearrangement followed by an NADPH-dependent
reduction [20– 23].
We have shown that in E. coli 2C-methyl-
D-erythritol
4-phosphate can be converted into a cyclic diphosphate by the
consecutive action of 4-diphosphocytidyl-2C-methyl-
D-
erythritol synthase, 4-diphosphocytidyl-2C-methyl-
D-erythritol
kinase and 2C-methyl-
D-erythritol 2,4-cyclodiphosphate
synthase [24–26] (Fig. 1). In the meantime, some of these
results have been confirmed by other authors [27–29].
We have also shown that
14

¨
nchen, Lichtenbergstr. 4,
D-85747 Garching, Germany. Fax: þ 49 89 289 13363,
Tel.: þ 49 89 289 13043, E-mail:
(Received 5 July 2001, revised 8 October 2001, accepted
9 October 2001)
Eur. J. Biochem. 268, 6302–6310 (2001) q FEBS 2001
Fig. 1. Biosynthesis of phytoene via the nonmevalonate pathway.
q FEBS 2001 Isoprenoid biosynthesis in plants (Eur. J. Biochem. 268) 6303
[
3
H]borohydride as reducing agent [32] [2,2
0
-
13
C
2
]-,
[2-
14
C]2C-methyl-D-erythritol 2,4-cyclodiphosphate, and
[U-
13
C
5
]2C-methyl-D-erythritol 4-phosphate were prepared
as described [33,34].
Isolation of chromoplasts from
C. annuum
Chromoplasts were isolated by a slight modification of the

concentration was about 1– 2 mg·mL
21
.
Isolation of chromoplasts from
N. pseudonarcissus
The isolation followed a procedure described by Kleinig &
Beyer [37]. Inner coronae of N. pseudonarcissus (80 g)
were homogenized in 250 mL of 67 m
M Tris/HCl, pH 7.5,
containing 5 m
M MgCl
2
,1mM dithioerythritol, 1 mM
EDTA, 0.2% (w/v) polyvinylpyrrolidone K90, and 0.74 M
sucrose. The suspension was filtered (three layers of nylon
cloth, 50 mm) and centrifuged (5 min, 1990 g, GSA rotor).
The supernatant was centrifuged (20 min, 25 400 g, GSA
rotor) affording a pellet of crude chromoplasts which was
resuspended in 2 mL of 67 m
M Tris/HCl, pH 7.5, contain-
ing 5 m
M MgCl
2
,1mM dithioerythritol and 50% (w/v)
sucrose. The suspension was filtered through one layer of
nylon cloth (50 mm). Aliquots of 2 mL were transferred to
centrifuge tubes. Equal volumes of 40, 30 and 15% (w/v)
sucrose in 67 m
M Tris/HCl, pH 7.5, containing 5 mM
MgCl

D-erythritol
4-phosphate and/or 2C-methyl-
D-erythritol 2,4-cyclodiphos-
phate were added as indicated in Table 1. The mixtures were
incubated at 30 8C. The reaction was terminated by ethyl
acetate extraction. The lipid extract was dried over sodium
sulfate. In experiments with radiolabeled substrates the
residue was analyzed by scintillation counting and/or
HPLC. The aqueous phase was analyzed by reversed phase
ion pair HPLC monitored by scintillation counting.
HPLC analysis of phosphorylated metabolites
Reversed phase ion pair HPLC separations were performed
with a Luna C8 column (Phenomenex, 5 mm,
4 £ 250 mm). The column was developed with a linear
gradient of 0–42% methanol in 10 m
M tetrabutylammo-
nium sulfate, pH 6.0 (total volume, 60 mL; flow rate,
0.75 mL min
21
). The retention volumes of 2C-methyl-
D-erythritol 4-phosphate and 2C-methyl-D-erythritol
2,4-cyclodiphosphate were 10.0 and 29.0 mL, respectively.
Table 1. Competition experiments with [1-
3
H]2C-methyl-D-erythritol 4-phosphate (MEP) and [2-
14
C]2C-methyl-D-erythritol 2,4-cyclodi-
phosphate (cMEPP) in a chromoplast system from C. annuum. In each experiment, the sample volume was 150 mL. Conc., concentration; Sp.
radioact., specific radioactivity.
Proffered precursors

)
produced
(nmol)
14
C
(nmol)
3
H
(nmol)
14
C:
3
H
(%
a
)
A 0.040 20 0.01 360 80 : 20 0.1 2.2 0.3 88 : 12
B 0.040 20 0.08 45 33 : 67 0.9 2.0 2.3 46 : 54
C 0.040 20 0.68 4.40 5.5 : 94.5 6.8 1.8 8.5 17 : 83
D 0.040 20 6.68 0.41 0.6 : 99.4 13.0 0.4 8.0 4.7 : 95.3
a
Relative molar contribution of
14
C vs.
3
H pool.
6304 M. Fellermeier et al. (Eur. J. Biochem. 268) q FEBS 2001
Isolation of 2C-methyl-D-erythritol 2,4-cyclodiphosphate
The reaction mixture was centrifuged and the supernatant
was applied to a CHROMABONDw SB column (500 mg,

were applied to a Hypersil RP18 HPLC column (5 mm,
4.5 £ 250 mm, ThermoQuest Germna GmbH, Egelsbach,
Germany) that was developed with a mixture of isopropanol/
acetonitrile/water (50 : 45 : 5, v/v). The effluent was
monitored photometrically (280 nm). The retention
volumes of b-carotene, phytoene, and xanthophylls were
12, 14, and 20 mL, respectively.
NMR spectroscopy
NMR spectra were recorded with a DRX 500 spectrometer
from Bruker Instruments (Karlsruhe, Germany) equipped
with four channels and a pulsed gradient unit. Two
dimensional homocorrelation and heterocorrelation experi-
ments were performed with
XWINNMR software from Bruker
Instruments. Phytoene was measured in CDCl
3
,and
2C-methyl-
D-erythritol 2,4-cyclodiphosphate was measured
in D
2
O.
RESULTS
Isolated chromoplasts of C. annuum were incubated with
mixtures of [1-
3
H]2C-methyl-D-erythritol 4-phosphate and
[2-
14
C]2C-methyl-D-erythritol 2,4-cyclodiphosphate

P coupling patterns are indicated.
Fig. 2. Diversion of radioactivity from [2-
14
C]2C-methyl-D-ery-
thritol 2,4-cyclodiphosphate into lipid-soluble material of chromo-
plasts from Narcissus pseudonarcissus. (A) 2C-methyl-
D-erythritol
2,4-cyclodiphosphate; (B) lipid-soluble fraction.
q FEBS 2001 Isoprenoid biosynthesis in plants (Eur. J. Biochem. 268) 6305
pool. The amount of newly formed [1-
3
H]2C-methyl-
D-erythritol 2,4-cyclodiphosphate increased with the con-
centration of the proffered [1-
3
H]2C-methyl-D-erythritol
4-phosphate; the transformation showed saturation
characteristics.
[1-
3
H]2C-methyl-D-erythritol 4-phosphate as well as
[2-
14
C]2C-methyl-D-erythritol 2,4-cyclodiphosphate were
efficiently converted into lipid-soluble material. The amount
of [1-
3
H]2C-methyl-D-erythritol 4-phosphate converted into
lipid-soluble material increased with the concentration of
the profferred substrate; saturation was reached at a

chromoplasts of C. annuum.
13
C coupling patterns are indicated.
Table 2.
13
C NMR assignments for phytoene.
Position
13
C-Chemical shift
(d, p.p.m.)
a
J
CC
b
(Hz) INADEQUATE
b
1, 1
0
131.23 42.3 17, 17
0
2, 2
0
123.97
3, 3
0
26.77
4, 4
0
39.72 42.5, 3.5
5, 5

0
14, 14
0
120.22
15, 15
0
123.35
16, 16
0
25.69 43.3
17, 17
0
17.68 42.2
18, 18
0
16.00 42.2
19, 19
0
16.04 42.2
20, 20
0
16.52 42.0
a
Referenced to external TMS;
b
from the experiment with [2,2
0
-
13
C

1, 1
0
ND 86
2, 2
0
1.1
3, 3
0
c
1.1
4, 4
0
ND 31
5, 5
0
d
ND 85
6, 6
0
1.3
7, 7
0
c
1.3
8, 8
0
ND 31
9, 9
0
d

19, 19
0
c
8.5 83
20, 20
0
8.9 86
a
Calculated as the relative
13
C abundance by comparison of
13
C NMR signal
intensities of the labeled sample with
13
C NMR signal intensities of an
unlabeled phytoene sample.
b
calculated as the fraction of the
13
C-coupled
satellites in the global
13
C NMR intensity of a given atom. c–e assignments
may be interchanged.
6306 M. Fellermeier et al. (Eur. J. Biochem. 268) q FEBS 2001
2,4-cyclodiphosphate using isolated chromoplasts from
N. pseudonarcissus (Fig. 2). The incorporation of radio-
actvity into lipid-soluble material was again checked by
solvent extraction of reaction mixtures and the consumption

C
13
C coupling. Based on chemical shift
values and coupling constants, the compound was identified
as 2C-methyl-
D-erythritol 2,4-cyclodiphosphate (see [26]
for NMR data of the authentic compound). The absence of
singlet signals for the carbon atoms 1, 2, 2-Me, 3 and 4 in the
spectrum of the isolated material demonstrates that the
proffered material had not been diluted by significant
amounts of endogenous material with natural
13
C abun-
dance. It follows that the chromoplast extract used did not
contain significant amounts of endogenous, unlabeled
2C-methyl-
D-erythritol 2,4-cyclodiphosphate.
Isolated chromoplasts from C. annuum were sub-
sequently incubated with 0.7 m
M [2,2
0
-
13
C
2
]2C-methyl-
D-erythritol 2,4-cyclodiphosphate at 30 8C for 12 h. The
suspension was extracted with ethyl acetate, and phytoene
(Fig. 1, compound 10) was isolated from the resulting
mixture of lipophilic compounds.

(i.e. C-16 and C-17) of this moiety showed
13
C–
13
C coupling
satellites, albeit of different intensities (Table 3). The labeling
pattern of the reconstructed DMAPP unit is summarized in
Fig. 6 and the evaluation of the signal intensities indicated a
ratio of 10 : 1 for the two isotopomers a and b. The
13
C
NMR signals of the methyl groups C-18, C-19, and C-20 of
phytoene (biosynthetically equivalent to C-5 of IPP) showed
13
C-coupled satellites of high intensity (Table 3). From the
signal intensities the molar fraction of the IPP isotopomer c
can be calculated (Fig. 6).
The signals of C-12 and the coincident signals of carbon
atoms 4 and 8 showed one bond
13
C–
13
C coupling satellites
of lower intensities that were substantially broadened by
comparison with the central signal (Fig. 4). When processed
for maximum resolution, these satellites appeared as
pseudotriplets that could be due to long range coupling
involving vicinal isoprenoid moieties. Due to the line
broadening, the precision of signal integration is substan-
tially reduced. However, within the experimental limits, it

sphate by incubation with chromoplasts of C. annuum.
Fig. 6. Reconstruction of the labeling pattern of IPP (isotopomers a
and b) and DMAPP (isotopomers c and d) from the labeling pattern
of the phytoene sample obtained in the experiment with
[2,2
0
-
13
C
2
]2C-methyl-D-erythritol 2,4-cyclodiphosphate. Bold lines
denote bonds linking adjacent
13
C atoms, numbers indicate the
percentage molar fraction of the isotopomers.
q FEBS 2001 Isoprenoid biosynthesis in plants (Eur. J. Biochem. 268) 6307
into phytoene (the main labeled component of the lipid-
soluble fraction) is systematically diminished by the
addition of increasing amounts of [1-
3
H]2C-methyl-D-ery-
thritol 4-phosphate. Moreover, the data show that even at
saturating concentrations of the tritiated compound, the
relative transfer of
14
C-label from the cyclodiphosphate pool
is always in excess of the value calculated from the original
molar concentration of the two precursors. This requires that
within the nonmevalonate pathway the cyclodiphosphate is
nearer than the 4-phosphate to IPP and DMAPP, the two C5

probably because of the inadequacy of the analytical tools
employed in earlier work using a
14
C label. In all the cases
in which such a scrambling was observed it was usually
ascribed to a lack of fidelity of the isomerase that inter-
converts IPP and DMAPP. Participation of the isomerase is
of crucial importance in the mevalonate pathway, in which
formation of IPP and DMAPP take place in sequential steps;
in contrast, the available evidence indicates that within the
new pathway IPP and DMAPP are formed in independent
steps from a common and yet unidentified intermediate
[45–50], but a subsequent partial equilibration of the pre-
formed units can nevertheless occur in organisms equipped
with the isomerase, as is the case in higher plants in which
the two metabolic pathways are known to coexist.
The isomerization of IPP to DMAPP is an antarafacial
process in which a proton is added to the re-re face of the
double bond with subsequent or concomitant stereospecific
removal of the H
B
hydrogen at C-2 (Fig. 7) from the
opposite face of the molecule [51]; in the specific case of a
recombinant yeast enzyme, the catalytic groups have been
identified as Cys139, respectively, Glu207 [52]. In
refinement and extension of previous observations by
other authors [53], the Poulter group has carried out a
thorough investigation on the lack of fidelity of this
isomerase by analyzing the proton exchange that occurs
when IPP is incubated with the enzyme in D

two competing deprotonation paths is illustrated in Fig. 8 for
the predominant ES complex A of a sample of IPP carrying a
13
C label in its methyl group; a similar scheme involving a
less stable ES complex B is necessary to account for the
observed very slow exchange of the H
A
-hydrogen of IPP. In
both cases, scrambling of the label takes place within the IPP
pool and the error is then transcribed into the DMAPP pool by
the normal action of the isomerase. The validity of the
proposed scheme is rewardingly supported by the observation
that the enzyme is capable to convert the IPP homolog X into
its isomer Y (see Fig. 9) in a process which bypasses the
formation of allylic isomers [53].
ACKNOWLEDGEMENTS
This work was supported by grants from the Fonds der Chemischen
Industrie and the Deutsche Forschungsgemeinschaft (SFB369) to A. B.,
W. E. and M. H. Z. and a fellowship from the Hans-Fischer-
Gesellschaft to T. R. We thank Katrin Ga
¨
rtner for skillfull assistence
and Prof B. Camara, Strasbourg, for a sample of phytoene. Financial
support by Novartis International AG Basel (to D. A.) is gratefully
acknowledged.
REFERENCES
1. Qureshi, N. & Porter, J.W. (1981) Biosynthesis of mevalonic acid
from acetyl-CoA. In Biosynthesis of Isoprenoid Compounds
(Porter, J.W. & Spurgeon, S.L., eds), Vol. 1, pp. 47– 94. John
Wiley, New York, USA.

Switzerland.
10. Schwarz, M.K. (1994) Terpen-Biosynthese in Ginkgo biloba: Eine
u
¨
berraschende Geschichte. PhD Thesis, ETH Zu
¨
rich, Switzerland.
11. Spenser, I.D. & White, R.L. (1997) Biosynthesis of vitamin B
1
(thiamin): an instance of biochemical diversity. Angew. Chem. Int.
Ed. 36, 1032–1046.
12. Hill, R.E., Himmeldirk, K., Kennedy, I.A., Panloski, R.M., Sayer,
B.G., Wolf, E. & Spenser, I.D. (1996) The biogenetic anatomy of
vitamin B
6
.A
13
C NMR investigation of the biosynthesis of
pyridoxol in Escherichia coli. J. Biol. Chem. 271, 30426–30435.
13. Arigoni, D., Sagner, S., Latzel, C., Eisenreich, W., Bacher, A. &
Zenk, M.H. (1997) Terpene biosynthesis from 1-deoxy-
D-xylulose
in higher plants by intramolecular skeletal rearrangement. Proc.
Natl Acad. Sci. USA 94, 10600–10605.
14. Lichtenthaler, H.K. (1999) The 1-deoxy-
D-xylulose 5-phosphate
pathway of isoprenoid biosynthesis in plants. Annu. Rev. Plant
Phys. Plant Mol. Biol. 50, 47–65.
15. Sprenger, G.A., Scho
¨

D-xylulose 5-phosphate reductoisomerase catalyzing the
formation of 2-C-methyl-
D-erythritol 4-phosphate in an alternative
nonmevalonate pathway for terpenoid biosynthesis. Proc. Natl
Acad. Sci. USA 95, 9879–9884.
21. Lange, B.M. & Croteau, R. (1999) Isoprenoid biosynthesis via a
mevalonate-independent pathway in plants: cloning and
heterologous expression of 1-deoxy-
D-xylulose-5-phosphate
reductoisomerase from peppermint. Arch. Biochem. Biophys. 365,
170– 174.
22. Schwender, J., Mu
¨
ller, C., Zeidler, J. & Lichtenthaler, H.K. (1999)
Cloning and heterologous expression of a cDNA encoding
1-deoxy-
D-xylulose-5-phosphate reductoisomerase of Arabidopsis
thaliana. FEBS Lett. 455, 140–144.
23. Jomaa, H., Wiesner, J., Sanderbrand, S., Altinicicek, B.,
Weidemeyer, C., Hintz, M., Tu
¨
rbachova, I., Eberl, M., Zeidler, J.,
Lichtenthaler, H.K., Soldati, D. & Beck, E. (1999) Inhibitors of the
non-mevalonate pathway of isoprenoid biosynthesis as antimalarial
drugs. Science 285, 1573– 1576.
24. Rohdich, F., Wungsintaweekul, J., Fellermeier, M., Sagner, S., Herz,
S., Kis, K., Eisenreich, W., Bacher, A. & Zenk, M.H. (1999) Cytidine
5
0
-triphosphate biosynthesis of isoprenoids: YgbP protein of

thritol 4-phosphate cytidyltransferase, a new enzyme in the
nonmevalonate pathway. Tetrahedron Lett. 41, 703–706.
28. Kuzuyama, T., Takagi, M., Kaneda, K., Watanabe, H., Dairi, T. &
Seto, H. (2000) Studies on the nonmevalonate pathway: conversion
of 4-(cytidine 5
0
-diphospho)-2-C-methyl-D-erythritol to its 2-phos-
pho derivative by 4-(cytidine 5
0
-diphospho)-2-C-methyl-D-erythri-
tol kinase. Tetrahedron Lett. 41, 2925– 2928.
29. Takagi, M., Kuzuyama, T., Kaneda, K., Watanabe, H., Dairi, T. &
Seto, H. (2000) Studies on the nonmevalonate pathway: formation
of 2-C-methyl-
D-erythritol 2,4-cyclodiphosphate from 2-phospho-
4-(cytidine 5
0
-diphospho)-2-C-methyl-D-erythritol. Tetrahedron
Lett. 41, 3395–3398.
30. Turner, D., Santos, H., Fareleira, P., Pacheco, I., LeGall, Y. &
Xavier, A.V. (1992) Structure determination of a novel cyclic
phosphocompound isolated from Desulfovibrio desulfuricans.
Biochem. J. 285, 387 –390.
31. Ostrovsky, D., Kharatian, E., Dubrovsky, T., Ogrel, O., Shipanova,
I. & Sibeldina, L. (1992) The ability of bacteria to synthesize a new
cyclodiphosphate correlates with their tolerance to redox-cycling
drugs: on a crossroad of chemotherapy, environmental toxicology
and immunobiochemical problems. Biofactors 4, 63– 68.
32. Kis, K., Wungsintaweekul, J., Eisenreich, W., Zenk, M.H. & Bacher,
A. (2000) An efficient preparation of 2-C-methyl-

Schuhr, C.A., Hecht, S., Zenk, M.H. & Bacher, A. (2000)
Biosynthesis of terpenoids: 4-diphosphocytidyl-2C-methyl-
D-ery-
thritol synthase of Arabidopsis thaliana. Proc. Natl Acad. Sci. USA
97, 6451–6456.
40. Rohdich, F., Wungsintaweekul, J., Lu
¨
ttgen, H., Fischer, M.,
Eisenreich, W., Schuhr, C.A., Fellermeier, M., Schramek, N., Zenk,
M.H. & Bacher, A. (2000) Biosynthesis of terpenoids: 4-dipho-
sphocytidyl-2-C-methyl-
D-erythritol kinase from tomato. Proc.
Natl Acad. Sci. USA 97, 8251–8251.
41. Fehr, T., Acklin, W. & Arigoni, D. (1966) The role of the
chanoclavines in the biosynthesis of ergot alkaloids. J. Chem. Soc.
Chem. Commun. 801–802
42. Pachlatko, P., Tabacik, C., Acklin, W. & Arigoni, D. (1975) Natural
and unnatural precursors in the biosynthesis of ergot alkaloids.
Chimia 29, 526.
43. Shibuya, M., Chou, H M., Fountoulakis, M., Hassam, S., Kim,
S U., Kobayashi, K., Otsuka, H., Rogalska, E., Cassady, J.M. &
Floss, H.G. (1990) Stereochemistry of the isoprenylation of
tryptophan catalyzed by 4-(g,g-dimethylallyl) tryptophan synthase
from Claviceps, the first pathway-specific enzyme in ergot alkaloid
biosynthesis. J. Am. Chem. Soc. 112, 297 –304.
44. Croteau, R. & Loomis, W.D. (1972) Biosynthesis of mono- and
sesquiterpenes in peppermint from mevalonate 2-
14
C. Phytochem-
istry 11, 1055–1066.

Identification of Cys139 and Glu207 as catalytically important
groups in the active site of isopentenyl diphosphate: dimethylallyl
diphosphate isomerase. Biochemistry 33, 4212–4217.
53. Koyama, T., Ogura, K. & Seto, S. (1973) Studies on isopentenyl
pyrophosphate isomerase with artificial substrates. J. Biol. Chem.
248, 8043–8051.
54. Street, I.P., Christiansen, D.J. & Poulter, C.D. (1990) Hydrogen
exchange during the enzyme-catalyzed isomerization of isopente-
nyl diphosphate and dimethylallyl diphosphate. J. Am. Chem. Soc.
112, 8577–8578.
6310 M. Fellermeier et al. (Eur. J. Biochem. 268) q FEBS 2001