ISOPRENOID METABOLISM IN HIGHER PLANTS

Plants contain an enormous range of isoprenoid compounds with a wide variety of structures and functions. Some are
primary metabolites, e.g., the steroids, but the majority are synthesized as secondary metabolites that are uniquely plant
products.
Isoprenoids have been known since antiquity as ingredients of perfumes, soaps, flavoring and as food colorants.

The isoprenoids are built up of C5 isoprene units and the nomenclature of the main classes reflects the number of isoprene
units present (Table 1).
Table 1. Main Classes of Isoprenoids found in Plants

Carbon atoms Name Parent Isoprenoid

10 Monoterpenoids GPP
15 Sesquiterpenoids FPP
20 Diterpenoids GGPP
25 Sesterterpenoids GFPP
30 Triterpenoids Squalene
40 Tetraterpenoids Phytoene
>40 Rubbers GGPP + (C5)n

General Pathway of Terpenoid Biosynthesis

HMG CoA glyceraldehyde-3-phosphate

+ pyruvate

mevalonic acid DXP

IPP DMAPP

Monoterpenoids GPP

Sesquiterpenoids FPP Squalene Sterols

Diterpenoids GGPP Carotenoids

Sesterterpenoids GFPP rubber

FIGURE 1.

MEVALONIC ACID PATHWAY

Biosynthesis of Mevalonic Acid

A molecule of acetyl-CoA undergoes a Claisen condensation with a molecule of malonyl-CoA (as in fatty acid biosynthesis;
see Biochemistry of Yeast Fermentation Fatty Acids) to form acetoacetyl-CoA which reacts with a third molecule of
manonyl-CoA in an Aldol Condensation to give 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA). Reduction with HMG-CoA
reductase yields mevaldic acid which is reduced again to mevalonic acid.
 HMG-CoA O HO O
O O synthase
acetoacetyl CoA
O O
thiolase
SCoA HO SCoA
O2C Claisen
SCoA SCoA malonyl-CoA
Condensation acetoacetyl-CoA HMG-CoA
acetyl-CoA malonyl-CoA Aldol
Condensation
NADPH
HMG-CoA
reductase
NADP

HMG-CoA O HO O
O HO reductase
HO H
HO OH
NADP NADPH mevaldic acid
(3R)-mevalonic acid

Conversion of Mevalonic acid to Isopentenyl pyrophosphate (3-methylbut-3-enylpyrophosphate)

Mevalonate is converted into 3-phospho-5-pyrophosphomevalonate by three consecutive phosphorylations. This labile
intermediate loses CO2 and phosphate and yields 3-isopentenyl pyrophosphate (IPP).

ATP ADP O HO ATP ADP O HO

O HO O O O
MVA kinase O O P O mevalonate O O P O P O
O OH 5-phosphate
O kinase O O
(3R)-mevalonate mevalonate-5-phosphate mevalonate-5-pyrophosphate
(MVAPP)

ATP
mevalonate-5-diphosphate
decarboxylase
ADP

2
O3PO
O O O
O O
O P O P O
3 O O P O P O
O O CO2, PO4
isopentenyl pyrophosphate (IPP) O O

DXP PATHWAY

Recently a second pathway to isopentenyl pyrophosphate was discovered and in 2002 this pathway was shown to occur in
grapes.

Biosynthesis of 1-deoxy-D-xylulose 5-phosphate (DXP)

The first step in this pathway is the formation of 1-deoxy-D-xylulose 5-phosphate (DXP) from pyruvate and glyceraldehyde
3-phosphate, both glycolysis intermediates. See Biochemistry of Yeast Fermentation; Sugars; Glycolysis.
O OH
O
H OP OP
COOH OH O OH

pyruvate 1-deoxy-D-xylulose 5-phosphate

glyceraldehyde 3-phosphate
DXP

Pyruvate reacts with thiamine diphosphate (TPP) to form active acetaldehyde (TPP-C2). See Biochemistry of Yeast
Fermentation; Sugars; Alcoholic Fermentation. Active acetaldehyde then reacts with glyceraldehyde-3-phosphate to form
DXP. See Figure 2.

Biosynthesis of 1-deoxy-D-xylulose 5-phosphate (DXP)

 H H
O pyruvate
O O

COOH HO H
O H OP
OH
OH
R N S
R glyceraldehyde 3-phosphate
R S N S
N from glycolysis

OPP

OPP OPP

thiamine diphosphate active acetaldehyde (TPP-C2)

1-deoxyxylulose 5-phosphate synthase

H OH
O
OH
OP
OP OH
R N S
O OH

1-deoxy-D-xylulose 5-phosphate
DXP

OPP
FIGURE 2

Conversion of DXP to Isopentenyl pyrophosphate (3-methylbut-3-enylpyrophosphate)

DXP undergoes a pinacol-like rearrangement to form 2C-methyl-D-erythritol-4-phosphate, which reacts with cytidine
triphosphate (CTP). After phosphorylation, an internal rearrangement takes place yielding 1,4-dihydroxy-3-methylbutan-2-
one 1-phosphate. A series of steps yet to be elucidated produces ispentenyl diphosphate. See Figure 3.

OH OH

OP OP OP OP
OP
O OH OH OH OH O OH
1-deoxyxylulose 2-C-methyl-D-erythritol 1,4-dihydroxy-3-methylbutan-2-one
isopentenyl diphosphate
5-phosphate 4-phosphate 1-phosphate
IPP
DXP MEP
 Conversion of DXP to Isopentenyl pyrophosphate
H
O O
pinacol-like OH OH
rearangement
OP OP OP
DXP OP
O OH reductoisomerase HO O OH OH OH
OH
H 2-C-methyl-D-erythritol
1-deoxy-D-xylulose 5-phosphate 4-phosphate
DXP MEP

CDP-ME
CTP
synthase
O
NH2 NH2
O P OH
OH N N
O O CDP-ME OH O O
kinase
O P O P O N O O P O P O N O
O O
OH OH OH OH OH OH OH OH

HO OH HO OH
4-(CDP)-2-C-methyl-D-erythritol 2-phosphate 4-(CDP)-2-C-methyl-D-erythritol
CDP-ME2P CDP-ME

O OH
O O O
P
HO P
OP OP OP OP
O OH
H OH O OH OH OH OH
OH
2-C-methyl-D-erythritol- steps to be determined
2,4-cyclodiphosphate

OP
OP
dimethylallyl diphosphate isopentenyl diphosphate

DMAPP IPP

FIGURE 3

ISOMERIZATION OF ISOPENTENYL PYROPHOSPHATE AND PRENYL TRANSFERASE REACTIONS

IPP is isomerized to dimethylallyl pyrophosphate (DMAPP) with IPP isomerase. The isomerization involves addition of a
proton to the sp2 carbon of isopentenyl pyrophosphate and generation of the more stable tertiary carbocation. Elimination of
a proton yields the most stable alkene (trisubstituted, Zaitevs Rule).
 B
HS HR H
HS HR PPO
PPO IPP isomerase
PPO

dimethylallyl pyrophosphate (DMAPP)

isopentenyl pyrophosphate (IPP)
most stable carbocation

Next IPP reacts with DMAPP to give geranyl pyrophosphate (GPP). In the first step of the reaction, IPP acts as a
nucleophile and displaces a pyrophosphate group from DMAPP. Pyrophosphate is an excellent leaving group. Its four OH
groups have pKa values of 0.9, 2.0, 6.6, and 9.4. Therefore three of the four groups will be primarily in their basic forms at
physiological pH (pH 7.3). A proton is removed in the next step, again yielding the most stable alkene. GPP rearranges to
give the monoterpenes. GPP reacts with a second molecule of IPP to produce farnesyl pyrophosphate (FPP). FPP is the
precursor to the sesquiterpenoids, and the steroids. FPP reacts with a third molecule of IPP to produce geranylgeranyl
pyrophosphate (GGPP). GGPP is the precursor of the carotenoids.

prenyl transferase
OPP OPP OPP
OPP
DMAPP HS
HS geranyl pyrophosphate (GPP)
IPP B
IPP
prenyl transferase

OPP

farnesyl pyrophosphate (FPP)

IPP
prenyl transferase

OPP

geranylgeranyl pyrophosphate (GGPP)

 MONOTERPENOIDS

Interest in the monoterpenes originated due to their use in perfumes and as food flavors. About forty monoterpene
compounds have been identified in grapes. Some of the monoterpene alcohols have a strong odor, especially geraniol, nerol,
citronellol, linalol, -terpineol, wine lactone, rose oxide and hotrienol, which has a floral aroma reminiscent of rose essence.
The olfactory perception thresholds of these compounds are rather low, as little as a few hundred micrograms per liter. The
most odoriferous are rose oxide and wine lactone. Furthermore, the olfactory impact of monoterpene compounds is
synergistic. Geraniol and linalol play an important role in the aromas of grapes and wines from the Muscat family (Muscat
Petits Grains, Muscat of Alexandria, Muscat of Ottonel and Muscat dAlsace), as concentrations are often well above the
olfactory perception thresholds.
These compounds also play a role in the Muscat aroma of some Alsatian and German grape varieties: Gewrztraminer (rose
oxide and wine lactone), Pinot Gris, Riesling, Auxerrois, Scheurebe, MuIler-Thurgau, etc. However, monoterpenes are only
partially responsible for the varietal aromas of these wines and do not explain all of the nuances.
Figure 4 shows the biosynthesis of the important monoterpenes skeletons found in grapes.

OH

OH OH

OH

geraniol nerol citronellol linalool

OH

O
O
OH O rose oxide
hotrienol -terpineol wine lactone

As with the anthocyanidins, most of the monoterpenes found in grape juice are present as glycosides. The main glycoside is
the glucoside (i.e. -D-glucopyranose attached to the monoterpene) but three disaccharides are also present: 6-O--L-
arabinofuranosyl--D-glucopyranose, 6-O--rhamnosyl--D-glucopyranoside and 6-O--D-glucopyranoside.

OH O HO O
O O
HO O HO O

OH HO monoterpene OH OH HO monoterpene
HO OH OH

6-O--L-Arabinofuranosyl--D-glucopyranoside 6-O--Apiofuranosyl--D-glucopyranoside

O OH
HO O
O O
HO monoterpene HO
HO
HO OH HO monoterpene
OH OH

6-O--L-Rhamnopyranosyl--D-glucopyranoside -D-glucopyranoside
 Formation of the Major Monoterpenes in Wine from Geranyl pyrophosphate

OH OH

geraniol citronellol

OPP
OPP
HS
OPP OH
PPO IPP

geranyl diphosphate neryl diphosphate nerol

DMAPP

OPP
OH

linalool
linalyl cation

OH

HO O

trans furan linalool

OH OH oxide
diendiol
-terpineol

FIGURE 4.
 CAROTENOIDS

The carotenoids are an abundant group of naturally occurring pigments, present in all green tissues, where they are
constituents of the chloroplast, as well as being responsible for most of the yellow to red colors of flowers and fruits.
Carotenoid hydrocarbons are called carotenes, whereas derivatives containing oxygen functions are the xanthophylls.

Unripe green fruit contains the same pigments as other photosynthetic tissues but upon ripening the chloroplasts
differentiate into chromoplasts (storage organelles for pigments) and there is often, but not always, de novo synthesis of
carotenoids.

In the grape, the concentrations of the carotenoids decrease steadily from bloom on, and at the same time the concentrations
of the C13-norisoprenoids steadily increase. Exposure of grapes to sunlight during ripening accelerates carotenoid
breakdown and in accompanied by an increase in the glycosylated C13-norisoprenoid derivative content (glycosides).

The most important carotenoids in grapes, in decreasing order, are lutein, -carotene, neoxanthin and lutein-5,6-epoxide.
OH

lutein

HO

-carotene

C neoxanthin
O
OH OH
HO

OH

lutein-5,6-epoxide

O
HO
The first dedicated step in the formation of carotenoids is the head-to-head condensation of two molecules of all-trans
geranylgeranyl pyrophosphate (GGPP) via the cyclopropyl carbonyl pyrophosphate, prephytoene pyrophosphate (PPPP) to
form phytoene. See Figure 5.
The Formation of 15-cis-phytoene from GGPP
PPO

OPP

GGPP GGPP

OPP

PPPP

OPP

H
H

15-cis-phytoene

FIGURE 5.
The sequence from phytoene to lycopene involves four stepwise dehydrogenations, occurring alternatively to either side of
the chromophore to form phytofluene, -carotene, neurosporene and lycopene. In addition, in higher plants, the 15-15
double bond is isomerized from the cid to the trans configuration. See Figure 6.

The Formation of Lycopene from 15-cis-Phytoene

15-cis-phytoene

phytoene desaturase

15-cis-phytofluene

all trans-phytofluene

phytoene desaturase

-carotene

-carotene desaturase

neurosporene

-carotene desaturase

lycopene

FIGURE 6.
The cyclohexane-ring end-groups on many of the carotenoids are formed independently.

The Formation of -Carotene and the Xanthophylls

Lycopene

Nu Enzyme
H Lycopene

Nu Enzyme Nu Enzyme

Nu Enzyme

-carotene

OH
zeaxanthin

HO

OH
violaxanthin O

O
HO

FIGURE 7.
-Carotene can result from the dehydration and enzyme release step taking place at one of the cyclohexane end-groups via
the removal of a proton on the other side of the carbon which has been attacked by the enzymic nucleophile.

The Formation of -Carotene and the resulting Xanthophylls

Lycopene

Nu Enzyme
H Lycopene

Nu Enzyme
Enzyme Nu H

Nu Enzyme

-carotene

OH
Lutein

HO

FIGURE 8.
 C13-NORISOPRENOIDS

A vast number of apparently carotenoid-derived substances have been identified in plant extracts, but knowledge of the
biochemistry of carotenoid catabolism is still extremely limited. The in vivo cleavage of the carotenoid chain is generally
considered to be catalyzed by dioxygenase systems. Although all the in-chain double-bonds seem to be vulnerable to
enzymatic attack, thus resulting in the formation of major fragment classes with 10, 13, 15 or 20 carbon atoms, in fruit
tissues a bio-oxidative cleavage of the 9,10 (9',10') double bond seems to be the most preferred. This last cleavage reaction
creates a large number of aroma products.
' 4'

3'
5
' ' ' '
' 14
' 12 ' 10 ' 8 '
15 13 11 2'
1 7 9 11 13 15 9 7 6'
2 6 8 10 12 14 1'
3 5
4

C20 (Retinoids)
C15 (Plant hormones)

C13 Aroma compounds

C10

The three essential steps for the formation of C13-norisoprenoid aroma compounds can therefore be considered to consist of
(i) the initial dioxygenase cleavage,
(ii) the subsequent enzymatic transformation(s) of the primary degradation product in natural tissues, and
(iii) formation of the flavorless glycoside.
During winemaking, acid-catalyzed conversion of the glycoside into an aroma compound occurs. Figure 9 shows how the
carotenoid, neoxanthin may yield two C13-norisoprenoid glycosides in this way, which during the winemaking process
produce the two powerful odorants, vitispirane and damascenone.

Glycosylation of any secondary metabolite will significantly increase its water solubility and may occur for a number of
reasons. Some of these may be for
(i) accumulation and storage,
(ii) transport,
(iii) protect the plant from any toxicity of the liphophilic compounds (e.g., the cyanogenic glycosides are hydrolyzed by -
glucosidase releasing toxic cyanide as a defense mechanism),
(iv) stability of compound (the anthocyanidins are more stable as their glycosides, the anthocyanins).
Formation of Two C13-Norisoprenoid Glycosides from the Carotenoid Neoxanthin

C
O
OH OH
HO
neoxanthin carotenoid cleavage enzymes

O
C
O
OH O
HO
OH
grasshopper ketone
epoxide ring opening
enzymatic reduction
O

OH
C OH
HO
OH OH
HO
enzymatic reduction
allenic triol
glucosylation OH

O OH
C OH HO
OH O
HO glucosylation
OH
HO HO
OH OH
HO
OH
O
winemaking OH
HO
process
O

O
winemaking
process

damascenone

vitispirane
FIGURE 9.
Pure Appl. Chem., Vol. 71, No. 12, pp. 22792284, 1999.
Printed in Great Britain.
q 1999 IUPAC

The mevalonate-independent methylerythritol

4-phosphate (MEP) pathway for isoprenoid
biosynthesis, including carotenoids*

Michel Rohmer
Universite Louis Pasteur/CNRS/Institut Universitaire de France, Institut Le Bel,
4 rue Blaise Pascal, 67070 Strasbourg Cedex, France

Abstract: A mevalonate-independent route to isopentenyl diphosphate (IPP), the universal

precursor of isoprenoids, is present in many bacteria, in some unicellular green algae and
in the plant plastids. All essential isoprenoids related to photosynthesis, including the
carotenoids, are synthesized via this alternative metabolic route. IPP is formed from pyruvate
and glyceraldehyde 3-phosphate via 1-deoxy-D-xylulose 5-phosphate and 2-C-methyl-D-
erythritol 4-phosphate. Later steps are still poorly known, although extensive data on the
origin of the hydrogen atoms of IPP are now available.

INTRODUCTION
Isoprenoids are present in all living organisms. The rst investigations on their biosynthesis were mainly
performed on sterols, with liver tissues, yeast and several plant systems [13]. The C5 isoprenic skeleton
was shown to be derived from the well-known sequence found in all textbooks (Fig. 1, Scheme A),
starting from acetyl-CoA 4 and leading to isopentenyl diphosphate 7 (IPP) via mevalonate 6 (MVA). The
committed step of this biosynthetic route is the reduction of hydroxymethylglutaryl CoA 5 into MVA
catalyzed by the HMGCoA reductase. Mevalonate 6 was unanimously accepted as the universal
isoprenoid precursor in all living organisms, despite some contradictory results mainly obtained when
working on isoprenoid biosynthesis from plant chloroplasts and from bacteria. Feeding experiments using
13
C-labeled carbon sources allowed the detection in bacteria and the partial elucidation of an alternative
route towards IPP (Fig. 1, Scheme B), starting from triose phosphate derivatives, in which 2-C-methyl-D-
erythritol 4-phosphate 10 (MEP) is the rst intermediate presenting the branched isoprenic skeleton. IPP
7 remains, however, a common intermediate to both routes and must therefore be considered as the
universal isoprenoid precursor. This brief contribution is not intended as an extensive survey of the topic.
It summarizes some of the important steps of the discovery and the elucidation of this novel metabolic
route. More details are found in the accompanying references.

ON THE ORIGIN OF THE CARBON ATOMS

The rst conclusive feeding experiments concerning an alternative metabolic route towards IPP were
performed with hopanoid producing bacteria. The C35 triterpenoids of the bacteriohopane series 11
(Fig. 2) are characterized by an unusual polyhydroxylated C5 side chain linked by a carboncarbon bond
to the triterpenic hopane skeleton [4]. In order to decipher the origin of this unusual feature, labeling
experiments were performed with [1-13C]- and [2-13C]-acetate [5]. All feeding experiments were made
using a minimal medium, only containing mineral salts and a single 13C-labeled carbon source. The
bacteria were thus forced to utilize this labeled carbon source. The metabolic pathways were usually well
known, and it was accordingly possible to deduce from the observed labeling pattern the origin of the

*Lecture presented at the 12th International Symposium on Carotenoids, Cairns, Australia, 1823 July 1999, pp.
22052302.
Correspondence: E-mail: mirohmer@chimie.u-strasbg.fr

2279
2280 M. ROHMER

Fig. 1 Incorporation of [1-13C]-glucose into isoprenoids via the MVA pathway (A) or via the MEP pathway (B).
carbon atoms by a retrobiosynthetic analysis. Such labeling conditions were later extended to most of
the labeling experiments performed, and were quite different from previous labeling studies, which
were made with bacteria grown on complex media. The additional carbon atoms of the
bacteriohopanepolyol side chain were derived from a D-pentose (according to the stereochemistry,
most probably from a D-ribose derivative) linked via its C-5 carbon to the hopane isopropyl group. The
most interesting problem was, however, found in the triterpenic moiety. The observed labeling pattern
found in the hopane isoprene units did not t with that expected from the MVA pathway. The rst
attempts to interpret these results in the frame of the MVA route [5] could not be conrmed by further
incorporations of 13C-labeled glucose isotopomers into the hopanoids 11 from the bacteria Zymomonas
mobilis, Methylobacterium fujisawaense and Alicyclobacillus acidoterrestris and the prenyl side chain
of ubiquinone 12 from Escherichia coli [6].
Isoprenic units were formed in the new metabolic route from two subunits: the rst was derived from
the C-2 and C-3 carbon atoms of pyruvate 3, and the other was D-glyceraldehyde phosphate 2 (GAP)
[6,7]. All labeling patterns resulting from the above-described labeling experiments were consistent with
the formation of pyruvate and GAP from glucose either via the glycolysis or via the EntnerDoudoroff
pathway or from acetate via the tricarboxylic acid and the glyoxylate cycles [5,6]. Incorporation of doubly

q 1999 IUPAC, Pure Appl. Chem. 71, 22792284

 Mevalonate-independent MEP pathway for isoprenoid biosynthesis 2281

Fig. 2 Key isoprenoids for the identication of the MEP pathway: hopanoids (11), ubiquinone (12), ginkgolides
(13), carotenoids (14), phytol (15) and plastoquinone (16).

labeled [4,5-13C2]-glucose into the hopanoids and ubiquinone from Methylobacterium fujisawaense
showed, in addition, from the long-range 13C/13C 2J coupling constants that a rearrangement allowed the
insertion of the C2 subunit derived from pyruvate decarboxylation between the C-2 and C-3 carbon atoms
from GAP [6]. The presence of such a rearrangement was corroborated by the incorporation of [U-13C6]-
glucose into the hopanoids of Zymomonas mobilis [7]. From these data, a hypothetical biogenetic scheme,
which proved later to be veried, was proposed. Similar experiments were performed by Schwarz and
Arigoni on the biosynthesis of diterpenoids of the ginkgolide 13 series in Ginkgo biloba embryos and led
to identical conclusions [8,9].

MEVALONATE PATHWAY VERSUS METHYLERYTHRITOL PHOSPHATE PATHWAY

Labeling experiments using 13C-labeled carbon sources allow a clear differentiation to be made between
the MVA and the MEP pathways [6,8,10]. In an organism utilizing glucose via glycolysis, one has to
determine how the rst isoprenoid precursors (acetyl-CoA 4 for the MVA route, pyruvate 3 and GAP 2 for
the MEP route) are obtained from [1-13C]-glucose 1 for instance (Fig. 1). The resulting labeling patterns
are quite distinct and characteristic for each biosynthetic pathway (Fig. 1, Schemes A and B). Many plant
isoprenoids, such as carotenoids 14, the phytyl chain 15 of chlorophylls, and mono- and diterpenes, are
usually poorly labeled from [14C]-MVA [11]. Furthermore, mevinolin, a potent inhibitor of the HMGCoA
reductase, efciently blocks sterol biosynthesis in plant systems, whereas it does not affect carotenoid
biosynthesis in chloroplasts. These results have usually been interpreted within the framework of the
MVA pathway in terms of a lack of permeability of the plastid membrane towards MVA and mevinolin
[11] and of a cytoplasm-independent IPP biosynthesis in the chloroplasts via the MVA route, even if a
completely different biosynthetic route cannot be excluded. In contrast, carotenoids for instance are easily
labeled from [14C]-pyruvate or from carbon dioxide. All these results suggest the presence of an
alternative biosynthetic route for the plastid isoprenoids. The approach utilizing 13C-labeled glucose in
the place of acetate, which proved useful for the elucidation of the biosynthesis of bacterial isoprenoids,

q1999 IUPAC, Pure Appl. Chem. 71, 22792284

2282 M. ROHMER

was applied to labeling experiments performed with phototrophic organisms. A clear dichotomy was
found in higher plants. The MVA route occurred in the cytoplasm and was responsible, as expected, for
the formation of the triterpenoids, including the sterols, whereas the MEP pathway afforded in the
chloroplasts the essential isoprenoids required for the photosynthetic apparatus (carotenoids 14, phytol 15
and plastoquinone 16) [10], as well as the whole series of plastid-related secondary metabolites (isoprene,
mono- and diterpenes) [12,13].

1-DEOXY-D-XYLULOSE 5-PHOSPHATE AND 2-C-METHYL-D-ERYTHRITOL 4-PHOSPHATE

AS IPP AND DMAPP PRECURSORS
Condensation of (hydroxyethyl)thiamine diphosphate and GAP yields 1-deoxy-D-xylulose 5-phosphate 9
(DXP) (Fig. 1, Scheme B). The rst evidence for DXP as an isoprenoid precursor was given by Broers and
Arigoni et al. [15]. Deuterium-labeled free 1-deoxy-D-xylulose (DX) was efciently incorporated into the
prenyl [14] chain of ubiquinone and menaquinone from a wild-type E. coli strain [14]. Similar
experiments were later performed with several plant systems and conrmed the role of DXP as an
isoprenoid precursor. Incorporation of [2,3,4,5-13C4]-DX into phytol and lutein from Catharanthus
roseus cell cultures [15] and of [2,3-13C2]- or [2,4-13C2]-DX into the prenyl chain of the ubiquinone from
E. coli [16] showed that DX in a plant and in a bacterium was incorporated without any previous
degradation and that the branched isoprenic skeleton resulted from an intramolecular rearrangement.
2-C-Methyl-D-erythritol (ME) or the corresponding lactone was repeatedly reported from higher plants
[11]. The cyclodiphosphate of this tetrol is found in normal growth conditions in the bacterium
Desulfovibrio desulfuricans or under oxidative stress induced by benzylviologen in several Gram-
positive bacteria [11]. Formally, ME directly results from DX via an acid-catalyzed a-ketol
rearrangement followed by a reduction [7]. It was therefore tempting to check whether ME or one of
its derivatives was involved in the alternative route for isoprenoid biosynthesis. Several 13C-labeled
glucose isotopomers were incorporated into the prenyl chain of the dihydromenaquinones from
Corynebacterium ammoniagenes and into the ME cyclodiphosphate. The C5 carbon skeletons of the
isoprenic units, as well as that of the branched tetrol, showed identical labeling patterns, indicating that
both resulted from the same reaction sequence of the MVA-independent route [17]. Deuterium-labeled
ME was incorporated, although in low yield, into the prenyl chain of ubiquinone and menaquinone from a
wild-type E. coli, indicating that ME is an isoprenoid precursor [18]. 13C-labeled DX was, in addition,
incorporated into ME by the leaves of Liriodendron tulipifera [19].
The gene of the DXP synthase was cloned from E. coli and from peppermint and overexpressed in
E. coli [2022], and that of the DXP isomero-reductase from E. coli and later from peppermint and
Arabidopsis thaliana [2325]. The latter enzyme utilizes only DXP as substrate (and not free DX). It
catalyzes the rearrangement of the pentulose phosphate and the concomitant nicotinamide adenine
dinucleotide phosphate (NADPH)-dependent reduction of the resulting methylerythrose phosphate
yielding MEP 10. The identication of these genes allowed the construction of E. coli mutants with
disrupted DXP synthase and/or isomero-reductase genes. Such mutants required free ME for their growth,
suggesting that a kinase capable of converting ME into MEP was present and proved useful to conrm the
role of MEP in isoprenoid biosynthesis [26].

ON THE ORIGIN OF THE HYDROGEN ATOMS IN ISOPRENIC UNITS

Incorporation experiments with 13C-labeled precursors allowed the identication of the rst precursors of
the MEP pathway. However, little is known about the further steps. Additional information was expected
from the knowledge of the fate of the hydrogen atoms in this pathway. Incorporation of [5,5,5-2H3]-DX
[27] and [3,5,5,5-2H4]-ME [26] into the ubiquinone prenyl side chain showed that all methyl hydrogen
atoms were preserved in the isoprenoid biosynthetic pathway in E. coli. Feeding of E. coli with [4,4-2H2]-
or [1,1,4,4-2H4]-ME [18,26] or of Zymomonas mobilis with [6,6-2H2]-glucose [28] also indicated that all
these deuterium atoms were retained in the isoprenic units of ubiquinone. Incorporation of [2-13C,4-2H]-
DX into phytol and lutein by cell cultures of Catharanthus roseus resulted in a complete loss of the
deuterium in all isoprenic units [29]. In contrast, the deuterium from [4-2H]-DX was retained in the
DMAPP 8-derived starter unit of the prenyl chain of the ubiquinone from E. coli and lost in all those

q 1999 IUPAC, Pure Appl. Chem. 71, 22792284

 Mevalonate-independent MEP pathway for isoprenoid biosynthesis 2283

derived from IPP 7 [30]. This observation was corroborated by the incorporation of [3,5,5,5-2H4]-ME
in an E. coli mutant lacking the DXP synthase gene, which solely synthesized its isoprenoid from the
exogenous ME added to the culture medium. Whereas the ubiquinone methyl groups retained their three
deuteriums of ME, the C-3 deuterium was only found in the DMAPP-derived unit and was integrally lost
in those derived from IPP [26].
The conversion of MEP 10 into IPP 7 formally requires the elimination of three molecules of water,
two reduction steps and one phosphorylation. Evidence for a rst reduction was obtained by the discovery
of the DXP isomero-reductase, which introduces a hydride from NADPH on the carbon atom
corresponding to C-4 of IPP. Further evidence for a reduction was obtained by feeding Zymomonas
mobilis with [1-2H]-glucose as the only carbon source [31]. This bacterium utilizes glucose via the
EntnerDoudoroff pathway, is unable to convert pyruvate into GAP and is a strong ethanol producer in
anaerobic growth conditions. C-1 of glucose is completely lost as CO2 by pyruvate decarboxylation and,
accordingly, is not incorporated into the isoprenic units of the hopanoids. This bacterium has no
tricarboxylic cycle. Under such growth conditions, deuterium-labeled NADPH pools are synthesized. In
the triterpenoids of the hopane series, deuterium was found in all carbon atoms derived from C-4 of IPP or
DMAPP, as expected from the DXP isomero-reductase, as well as on all carbon atoms corresponding to
C-2 of DMAPP and IPP. The latter deuterium atom was the signature of an additional reduction step
occurring at an unknown stage of the reaction sequence of the biosynthetic route. The results of this
experiment on the biosynthesis of hopanoids from Zymomonas mobilis shed light on the equivalence of
the isoprenic units, whether they were derived from IPP or from DMAPP, and are consistent with those
obtained with the Catharanthus roseus cell culture [29].
The stereochemistry of the reaction catalyzed by an IPP isomerase [32] and a prenyl transferase [33]
from E. coli has recently been determined. Both enyzmes eliminate the pro R hydrogen from IPP, like all
other known equivalent enzymes. Furthermore, it was shown that the IPP isomerase is not essential in
E. coli: its disruption does not prevent the growth of the bacterium [34]. These data suggest that, if no
other IPP isomerase is present, two distinct routes occur in E. coli, yielding separately IPP 7 and DMAPP
8 from the same intermediate derived from MEP 10.

CONCLUSION
The discovery of a second route for isoprenoid biosynthesis has opened up new elds in isoprenoid
biochemistry. The simultaneous presence of both the MVA and MEP pathways in plant cells and
the exchange of intermediates (IPP, GP, FPP) between the cytoplasm and the plastids indicate the
signicance of this cross-talking [8,15,35,36]. An interesting interpretation is the regulation of the
biosynthesis of volatile isoprenoids released in stress conditions [36].
The presence of the MEP route in bacteria, which are either opportunistic pathogens or closely related
to pathogenic species, opens up a route towards the discovery of enzyme inhibitors, which should
represent novel types of antibacterial agents [37,38]. Finally, the recent nding of the presence of the
two known genes of the MEP pathway in Plasmodium falciparum, the microorganism responsible for
malaria, as well as the growth inhibition of this parasite by fosmidomycin, an inhibitor of the DXP
isomero-reductase, allows new hopes for curing this major disease [39].

REFERENCES
1 S. L. Spurgeon, J. W. Porter. In Biosynthesis of Isoprenoid Compounds (J. W. Porter, S. L. Spurgeon, eds.), vol 1,
pp. 146. John Wiley and Sons, New York (1981).
2 N. Qureshi, J. W. Porter. In Biosynthesis of Isoprenoid Compounds (J. W. Porter, S. L. Spurgeon, eds.), vol. 1,
pp. 4794. John Wiley and Sons, New York (1981).
3 D. A. Bochar, J. A. Friesen, C. V. Stauffacher, V. W. Rodwell. In Comprehensive Natural Products Chemistry.
Isoprenoids Including Carotenoids and Steroids (D. E. Cane, ed.), vol. 2, pp. 1544. Pergamon Press, Oxford,
UK (1999).
4 M. Rohmer. Pure Appl. Chem. 65, 12931298 (1993).
5 G. Flesch, M. Rohmer. Eur. J. Biochem. 175, 405411 (1998).

q1999 IUPAC, Pure Appl. Chem. 71, 22792284

2284 M. ROHMER

6 M. Rohmer, M. Knani, P. Simonin, B. Sutter, H. Sahm. Biochem. J. 295, 517524 (1993).

7 M. Rohmer, M. Seemann, S. Horbach, S. Bringer-Meyer, H. Sahm. J. Am. Chem. Soc. 118, 25642566 (1996).
8 M. Schwarz. Ph.D. Thesis Nb 10951. Eidgenossische Technische Hochschule, Zurich, Switzerland (1994).
9 M. Schwarz, D. Arigoni. In Comprehensive Natural Products Chemistry. Isoprenoids Including Carotenoids
and Steroids, (D. E. Cane, ed.), vol. 2, pp. 367400. Pergamon Press, Oxford, UK (1999), and references cited
therein.
10 H. K. Lichtenthaler, J. Schwender, A. Disch, M. Rohmer. FEBS Lett. 400, 271274 (1997).
11 M. Rohmer. In Comprehensive Natural Products Chemistry. Isoprenoids Including Carotenoids and Steroids,
(D. E. Cane, ed.), vol. 2, pp. 4567. Pergamon Press, Oxford, UK (1999), and references cited therein.
12 W. Eisenreich, M. Schwarz, A. Cartayrade, D. Arigoni, M. H. Zenk, A. Bacher. Chem. Biol. 5, R221233
(1998), and references cited therein.
13 H. K. Lichtenthaler. Annu. Rev. Plant Physiol. Mol. Biol. 50, 4765 (1999), and references cited therein.
14 S. T. J. Broers. Ph.D. Thesis Nb. 10978. Eidgenossische Technische Hochschule, Zurich, Switzerland (1994).
15 D. Arigoni, S. Sagner, C. Latzel, W. Eisenreich, A. Bacher, M. H. Zenk. Proc. Natl. Acad. Sci. USA 94,
10 60010 605 (1997).
16 S. Rosa Putra, L. Lois, N. Campos, A. Boronat, M. Rohmer. Tetrahedron Lett. 39, 2326 (1998).
17 T. Duvold, J.-M. Bravo, C. Pale-Grosdemange, M. Rohmer. Tetrahedron Lett. 38, 47694772 (1997).
18 T. Duvold, P. Cali, J.-M. Bravo, M. Rohmer. Tetrahedron Lett. 38, 61816184 (1997).
19 S. Sagner, W. Eisenreich, M. Fellermeier, C. Latzel, A. Bacher, M. H. Zenk. Tetrahedron Lett. 39, 20912094
(1998).
20 G. A. Sprenger, U. Schorken, T. Wiegert, S. Grolle, A. A. de Graaf, S. V. Taylor, T. P. Begley, S. Bringer-
Meyer, H. Sahm. Proc. Natl. Acad. Sci. USA 94, 12 85712 862 (1997).
21 B. M. Lange, M. R. Wildung, D. McCaskill, R. Croteau. Proc. Natl. Acad. Sci. USA 95, 21002104 (1998).
22 L. M. Lois, N. Campos, S. Rosa Putra, K. Danielson, M. Rohmer, A. Boronat. Proc. Natl. Acad. Sci. USA 95,
21052110 (1998).
23 S. Takahashi, T. Kuzuyama, H. Watanabe, H. Seto. Proc. Natl. Acad. Sci. USA 95, 98799884 (1998).
24 B. M. Lange, R. Croteau. Arch. Biochem. Biophys. 365, 170174 (1999).
25 J. Schwender, C. Muller, J. Zeidler, H. K. Lichtenthaler. FEBS Lett. 455, 140144 (1991).
26 L. Charon, J.-F. Hoefer, C. Pale-Grosdemange, L. M. Lois, N. Campos, A. Boronat, M. Rohmer. Biochem. J.
346, 737742 (2000).
27 J. L. Giner, B. Jaun. Tetrahedron Lett. 39, 80218022 (1998).
28 T. Duvold. Ph.D. Thesis. Universite Louis Pasteur, Strasbourg, France (1997).
29 D. Arigoni, W. Eisenreich, C. Latzel, S. Sagner, T. Radykewicz, M. H. Zenk, A. Bacher. Proc. Natl. Acad. Sci.
USA 96, 13091314 (1999).
30 J. L. Giner, B. Jaun, D. Arigoni. Chem. Commun. 18571858 (1998).
31 L. Charon, C. Pale-Grosdemange, M. Rohmer. Tetrahedron Lett. 40, 72317234 (1999).
32 A. E. Leyes, J. A. Baker, F. M. Hahn, C. D. Poulter. Chem. Commun. 717718 (1999).
33 A. E. Leyes, J. A. Baker, C. D. Poulter. Org. Lett. 1, 10711073 (1999).
34 F. M. Hahn, A. P. Hurlburt, C. D. Poulter. J. Bacteriol. 181, 44994504 (1991).
35 K. P. Adam, J. Zapp. Phytochemistry 48, 953959 (1998).
36 J. Piel, J. Donath, K. Bandemer, W. Boland. Angew. Chem. Int. Ed. 37, 24782481 (1998).
37 S. Rosa Putra, A. Disch, J.-M. Bravo, M. Rohmer. FEMS Microbiol. Lett. 164, 169175 (1998).
38 M. Rohmer. Progr. Drug. Res. 50, 135154 (1998).
39 H. Jomaa, J. Wiesner, S. Sanderbrand, A. Altinciak, C. Weidemeyer, M. Hintz, I. Turbachova, M. Eberl, J.
Zeidler, H. K. Lichtenthaler, D. Soldati, E. Beck. Science 285, 15731576 (1999).

q 1999 IUPAC, Pure Appl. Chem. 71, 22792284

Pure Appl. Chem., Vol. 75, Nos. 23, pp. 375387, 2003.
2003 IUPAC

Mevalonate-independent methylerythritol
phosphate pathway for isoprenoid biosynthesis.
Elucidation and distribution*
Michel Rohmer

Universit Louis Pasteur/CNRS/Institut Universitaire de France, Institut Le Bel,

4 rue Blaise Pascal, 67070 Strasbourg Cedex, France

Abstract: A long-overlooked metabolic pathway for isoprenoid biosynthesis, the mevalonate-

independent methylerythritol phosphate (MEP) pathway, is present in many bacteria and in
the chloroplasts of all phototrophic organisms. It represents an alternative to the well-known
mevalonate pathway, which is present in animals, fungi, plant cytoplasm, archaebacteria, and
some eubacteria. This contribution summarizes key steps of its elucidation and the state-of-
the-art knowledge of this biosynthetic pathway, which represents a novel target for antibac-
terial and antiparasitic drugs.

INTRODUCTION: THE MEVALONATE PATHWAY FOR ISOPRENOID BIOSYNTHESIS

Isoprenoids represent the most diverse family of natural products. They are all formally derived from
the branched C5 skeleton of isoprene 1 (Fig. 1). Isopentenyl diphosphate 19 (IPP) and dimethylallyl
diphosphate 20 (DMAPP) (Fig. 2) correspond both to the biological equivalents of isoprene. The elu-
cidation of their biosynthesis resulted from the first applications of labeling with radioactive as well
as stable isotopes and ended up in the discovery of the mevalonate pathway, which starts from acetate
12 activated as acetyl-coenzyme A 13 (Fig. 2). This first work on the early steps of isoprenoid biosyn-
thesis was essentially performed with liver tissues and yeast for the biosynthesis of cholesterol and
ergosterol (9, Fig. 1), and was later confirmed for the biosynthesis of plant sterols and triterpenes.
Mevalonate 16 was thus accepted as the universal precursor of all isoprenoids in all living organisms
[1,2].
Discrepancies with this assertion were, however, reported. 14C-Labeled mevalonate was poorly
incorporated into the carotenoids (e.g., 7, Fig. 1) from plant chloroplasts [3,4]. This was interpreted in
terms of lack of permeability of chloroplasts toward mevalonate. In addition, incorporation yields of ra-
diolabelled mevalonate into plant mono- and diterpenes were usually very low, at least as compared to
those into sterols, which are synthesized in the cytoplasm [5]. Other data, which did not fit with the gen-
eral occurrence of the mevalonate pathway, were obtained with mevinolin, a potent hypocholes-
terolemic agent, inhibiting the HMGCoA reductase, a key enzyme of the mevalonate pathway reducing
hydroxymethylglutaryl-CoA 15 into mevalonic acid (MVA) 16. Mevinolin strongly inhibits sterol
biosynthesis in plant systems, but had no effect on carotenoid and phytol biosynthesis [6,7]. Again, such
negative results were interpreted in terms of impermeability of the chloroplast membrane toward the an-
tibiotic. In addition, a cytoplasm-independent IPP biosynthesis was discovered in plant chloroplasts and
chromoplasts [8]. All these contradictory results did not really question the accepted dogma of the

*Pure Appl. Chem. 75, 141419 (2003). An issue of reviews and research papers based on lectures presented at the 23rd IUPAC

International Symposium on the Chemistry of Natural Products, Florence, Italy, 28 July2 August 2002 on the theme of natural
products.
E-mail: mirohmer@chimie.u-strasbg.fr

375
376 M. ROHMER

Fig. 1 Isoprenoids investigated for the elucidation of the MEP pathway: isoprene 1, bacterial hopanoid 2,
ubiquinone 3, menaquinone 4, phytol 5, plastoquinone 6, -carotene 7, ginkgolide 8, sterol 9.

mevalonate pathway, and the identity of the biosynthetic route followed for the biosynthesis of chloro-
plast isoprenoids.

DISCOVERY OF THE MEVALONATE-INDEPENDENT METHYLERYTHRITOL

PHOSPHATE PATHWAY IN BACTERIA
As compared to plants, most bacteria are poor isoprenoid producers. All of them produce essential iso-
prene metabolites such as bactoprenol, the carbohydrate carrier for the biosynthesis of the peptidogly-
cane of the cell wall, or ubiquinone and/or menaquinone of the electron transport chains, but in rather
tiny amounts. Accordingly, very few investigations were performed on the biosynthesis of bacterial iso-
prenoids. The few studies leading to positive results pointed to the presence of the MVA pathway.

2003 IUPAC, Pure and Applied Chemistry 75, 375387

 Mevalonate-independent MEP pathway for isoprenoid biosynthesis 377

Fig. 2 MVA pathway for isoprenoid biosynthesis with labeling pattern from [1-13C]glucose metabolized via
glycolysis.

14C- and 13C-labeled MVA were respectively incorporated into a sterol of the gliding bacterium
Nannocystis exedens [9] or into the carotenoids of a Flavobacterium sp. [10]. 13C-Labeled acetate
was incorporated into the prenyl moiety of antibiotics of mixed origin from several actinomyces, and
the labeling pattern corresponded to the expected one resulting from the direct incorporation of acetate
into the MVA pathway [11]. Other incorporation attempts led to inconclusive results, which could not
be readily interpreted. Thus, the incorporation of 14C-labeled acetate into the prenyl chain of
ubiquinone from Escherichia coli resulted in a radioactivity distribution, which did not fit with the ex-
pected one from the MVA pathway, and led the authors to propose an alternative variant of this path-
way via acetolactate [12]. 13C-Labeled acetate was not incorporated into this prenyl chain of
ubiquinone, when E. coli was grown on a culture medium containing glucose and amino acid-derived
carbon sources, whereas pyruvate labeled two contiguous carbon atoms derived from C3 and C5 of
DMAPP or IPP [13]. Such results were indeed the premises for the presence of an alternative route for
the formation of isoprene units.
The discovery of a novel pathway for the early steps of isoprenoid biosynthesis was indeed a side-
product of our investigations on the biosynthesis of bacterial triterpenoids of the hopane series (e.g. 2,
Fig. 1) [14]. A C35 skeleton, in which the triterpenic hopane moiety is linked by a carbon/carbon bond
to a polyhydroxylated n-alkyl side chain, characterizes all major bacterial hopanoids [15]. All bacteria
do not produce hopanoids, but in the hopanoid-producing bacteria, such triterpenoids were shown to
play a role of membrane stabilizers, much like sterol do in eukaryotic membranes. Hopanoid concen-

2003 IUPAC, Pure and Applied Chemistry 75, 375387

378 M. ROHMER

tration is usually of the same order of magnitude as the sterol concentration in eukaryotic cells. This
makes them well suited for investigation by NMR spectroscopy after feeding of 13C-labeled precursors.
Feeding of 13C-labeled acetate to the hopanoid producing bacteria Rhodopseudomonas palustris,
Rhodopseudomonas acidophila, or Methylobacterium organophilum were performed with bacteria
grown on a mineral medium only containing 13C-labeled acetate as the only carbon and energy source
[14]. Such experimental conditions ensured the incorporation of the labeled precursor into the
hopanoids. As most biosynthetic pathways were known, a retrobiosynthetic analysis of the observed la-
beling patterns allowed reconstructing the reaction sequence involved in the formation of the hopanoids.
Indeed, the experiments performed with [1-13C]- and [2-13C]acetate showed that the C5 side chain was
derived from a D-pentose synthesized via the nonoxidative pentose phosphate pathway and linked to the
triterpenic skeleton via its C5 carbon atom. The most interesting problem was, however, found in the
isoprene moiety. The observed labeling pattern completely differed from the expected one from the
MVA pathway. Interpretation was not obvious from the sole experiments performed with labeled ac-
etate. The clues for the elucidation of this metabolic pathway were obtained after incorporation of glu-
cose isotopomers with 13C labeling either at C1, C2, C3, C5, or C6 into the hopanoids of Zymomonas
mobilis [16]. Indeed, Z. mobilis only utilizes glucose as carbon source. This glucose is metabolized via
the EntnerDoudoroff pathway. This bacterium does not possess a complete tricarboxylic acid cycle
and does not convert pyruvate into glyceraldehyde phosphate because of the lack of an active enolase.
From the observed labeling patterns, it could be deduced that isoprene units in the hopanoids were de-
rived from a C2 subunit synthesized via pyruvate decarboxylation and from a C3 subunit corresponding
to a triose phosphate derivative (Fig. 3) [16]. These results were in accordance with the formerly de-
scribed incorporations of 13C-labeled acetate. The observed labeling pattern in the hopanoids from

Fig. 3 MEP pathway for the biosynthesis of isoprenoids with labeling pattern from [1-13C]glucose metabolized via
glycolysis.

2003 IUPAC, Pure and Applied Chemistry 75, 375387

 Mevalonate-independent MEP pathway for isoprenoid biosynthesis 379

Rhodopseudomonas spp. or from M. organophilum simply resulted from the incorporation of acetate
into the glyoxylate cycle, which allows growth on acetate as sole carbon source, and into the tricar-
boxylic acid cycle, yielding finally phosphoenolpyruvate, the precursor of pyruvate and the triose phos-
phate derivatives [16]. They were also confirmed by incorporation of 13C-labeled glucose into iso-
prenoids from bacteria utilizing glucose via the EmbdenMeyerhofParnas pathway and by studies in
other isoprenoid series such as the ubiquitous ubiquinones and menaquinones [16]. Incorporation of
doubly labeled [4,5-13C2]glucose into the hopanoids and ubiquinone of Methylobacterium fujisawaense
and of uniformly labeled [U-13C6]glucose into the hopanoids from Z. mobilis afforded a definitive proof
for the formation of the isoprene units via a C2 and a C3 subunit and shed light on a rearrangement re-
action allowing the insertion of the C2 subunit derived from pyruvate 11 between two carbon atoms
from the triose phosphate-derived C3 subunit [16,17]. This triose phosphate was identified as glycer-
aldehyde phosphate 10 by feeding E. coli mutants lacking each enzymes of the triose phosphate me-
tabolism with either 13C-labeled glycerol in the presence of unlabeled pyruvate, or 13C-labeled pyru-
vate in the presence of unlabeled glycerol [17]. A hypothetical biogenetic pathway was proposed,
including as first step a reaction involving a transketolase-like, thiamin diphosphate-dependent enzyme
catalyzing the condensation of hydroxyethylthiamin derived from pyruvate decarboxylation on the car-
bonyl group of a triose phosphate derivative and a second step with an -ketol rearrangement resem-
bling the reaction catalyzed by the acetolactate isomero-reductase followed by the concomitant reduc-
tion of the resulting aldehyde and yielding methylerythritol 4-phosphate 22 (MEP) [16,17]. MEP
presents the characteristic branched C5 skeleton of all isoprenoids and has to be considered as a
hemiterpene. The pathway was named after this intermediate, which has to date no other known role
than that of an isoprenoid precursor and which most probably is the result of a committed step of the
pathway.

UBIQUITOUS PRESENCE OF THE MEP PATHWAY IN CHLOROPLASTS OF

PHOTOTROPHIC ORGANISMS
Once an alternative route for isoprenoid biosynthesis has been found in bacteria, it was tempting to
check whether such a pathway is involved in the formation of plant isoprenoids such as carotenoids,
phytol, mono- and diterpenes. Indeed, independent IPP formation was formerly pointed out in plant
chloroplasts, and the presence of an alternative route would be consistent with the discrepancies with
respect to the MVA pathway mentioned above.
Labeling experiments using 13C-labeled glucose isotopomers were performed on ginkgo embryos
by Schwarz and Arigoni, independently of our own assays with bacteria, and afforded the first proof for
the presence of the MEP pathway in plants [18,19]. They resulted in labeling patterns of ginkgolides 8
(Fig. 1), belonging to a diterpenoid series, which could be interpreted in the frame of the same bio-
genetic pathway as that proposed for the formation of bacterial isoprenoids (Fig. 3). They also shed light
on the dichotomy in the isoprenoid biosynthesis in plant cells: the mevalonate pathway is involved, as
expected, in the sterol biosynthesis in the cytoplasm, whereas isoprenoids that are most likely plastid
related are synthesized via the MEP pathway [18]. Incorporation of 13C-labeled glucose into the iso-
prenoids from several higher plant systems afforded additional proofs for the presence of the MEP path-
way in plants. The ubiquitous terpenoids of the photosynthetic apparatus (carotenoids 7, phytol 5 from
chlorophyll, and the prenyl chain of plastoquinone 6) are also derived from the MEP pathway [20].
Finally, labeling experiments, performed with 13C-labeled glucose or with deuterium-labeled deoxy-
xylulose, showed the wide distribution of the MEP pathway. It is involved in the formation of all in-
vestigated hemi- (e.g., isoprene 2, 2-methylbut-3-en-2-ol, dimethylallyl and the rest linked to aromatic
rings) [2123], mono- [23,24] and diterpenoids [2527] and some sesquiterpenoids [28].

2003 IUPAC, Pure and Applied Chemistry 75, 375387

380 M. ROHMER

TOWARD THE IDENTIFICATION OF INTERMEDIATES

1-Deoxyxylulose 5-phosphate
From the origin of the carbon atoms in the isoprene units derived from the MEP pathway, as deduced
from the 13C glucose labeling experiments, a 1-deoxy-D-xylulose (DX) derivative 21 (Fig. 3) was a
likely isoprenoid precursor. This pentulose was not an unknown natural product. It has been found in
the fermentation broth of a Streptomyces sp. [29] and was known as precursor for pyridoxol phosphate
and thiamin diphosphate in E. coli [30,31]. Efficient incorporation of deuterium-labeled DX iso-
topomers into the prenyl chain of ubiquinone and menaquinone of wild-type E. coli represented the first
proof for the role of a DX derivative as isoprenoid precursor [32]. Similar experiments performed with
other systems from bacteria or from plants confirmed this conclusion. The gene of the 1-deoxyxylulose-
5-phosphate (DXP) synthase (dxs or ispA) was rapidly identified in E. coli [33,34] and in peppermint
[35], and later in many other organisms, owing to its expected similarity with the transketolase genes.
As predicted, the enzyme is thiamin diphosphate-dependent and utilizes GAP 10 and pyruvate 11 as
substrates to yield DXP 21.

2-C-Methylerythritol 4-phosphate and its derivatives

2-C-Methyl-D-erythritol (ME) is a natural product found in several higher plants [3]. Its 2,4-cy-
clodiphosphate 25 (Fig. 3) is accumulated in large amounts by some gram-positive bacteria in oxida-
tive stress conditions induced by one-electron transfer inhibitors such as benzylviologen [3]. According
to its structure, this tetrol fits well into the biogenetic scheme. It was, therefore, tempting to check
whether ME is an isoprenoid precursor. Incorporation of 13C-labeled glucose showed that the carbon
skeleton of ME from Corynebacterium ammoniagenes is synthesized via the same reaction sequence as
the isoprene units of the menaquinones [36]. Deuterium-labeled ME was incorporated by wild-type
E. coli into the prenyl chain of ubiquinone and menaquinone with a localization of the labeling in ac-
cord with the hypothetical biogenetic scheme [37]. This allowed the search for E. coli mutants aux-
otrophic for ME. Analysis of such mutants was followed by the discovery of the DXP isomero-reduc-
tase gene (dxr or ispB) [38]. This enzyme catalyzes, as predicted, the acid-catalyzed -ketol
rearrangement of DXP as well as the reduced nicotinamide adenine dinucleotide phosphate
(NADPH)-dependent concomitant reduction of the resulting aldehyde, yielding MEP 22 as reaction
product. The enzyme requires a divalent cation (Co2+, Mg2+, or Mn2+) for catalysis.
The next three steps correspond to an activation of MEP, yielding three novel intermediates.
Incubation of 3H-labeled MEP in the presence of nucleotide triphosphate with crude cell-free systems
from E. coli resulted in the formation of a novel metabolite. The purification and identification of this
novel radiolabeled metabolite was, however, difficult. The search in databases for genes encoding en-
zymes using a polyol phosphate (such as MEP) and a nucleotide triphosphate as substrate revealed the
homology of a gene encoding an enzyme catalyzing the formation of 5-diphosphocytidyl ribitol from
cytidine 5-triphosphate and from ribitol 5-phosphate and of a gene of unknown function in E. coli
(ygbP or ispC). Cloning and overexpression of the E. coli ygbP gene resulted in the identification of an
enzyme using MEP and cytidine triphosphate as substrates, yielding 4-diphosphocytidyl ME 23 as re-
action product [39]. Whereas all other genes of the MEP pathway are scattered in the E. coli genome,
this gene forms a small operon with the two next ones. This feature facilitated their identification. The
enzyme encoded by the ychB (or ispD) gene catalyzes the adenosine 5-triphosphate (ATP)-dependent
phosphorylation of the tertiary hydroxy group of 4-diphosphocytidyl ME yielding 4-diphosphocytidyl
ME 2-phosphate 24 [40], and that encoded by ychB (or ispE) the formation of ME cyclodiphosphate 25
from the former intermediate with elimination of cytidine 5-monophosphate (CMP) [41]. The three
novel intermediates were incorporated into carotenoids by red pepper chromoplasts, suggesting that
they corresponded to intermediates of the MEP pathway [3941]. The latter three genes were also iden-
tified by another approach. In an E. coli strain capable of utilizing MVA after introduction of the genes

2003 IUPAC, Pure and Applied Chemistry 75, 375387

 Mevalonate-independent MEP pathway for isoprenoid biosynthesis 381

permitting its conversion into IPP (MVA kinase, phosphomevalonate 17 kinase, and diphosphomeval-
onate 18 decarboxylase), random mutations, which were rescued by the addition of MVA to the culture
medium allowed the identification of the ygbP, ychB, and ygbB genes and indicated that they are es-
sential genes of the MEP pathway [4244].

ORIGIN OF HYDROGEN ATOMS IN ISOPRENE UNITS DERIVED FROM MEP PATHWAY:

INDEPENDENT IPP AND DMAPP BIOSYNTHESIS
The formation of allylic alcohol diphosphates such as IPP 19 and DMAPP 20 from an ME derivative (a
tetrol) is not an obvious process. In order to get more insight into the steps downstream of ME cy-
clodiphosphate, more information was required. Knowledge on the fate of the hydrogen atoms in the
pathway was expected to afford useful clues for the further elucidation of the MEP pathway.
Labeling experiments using deuterated DX isotopomers and a wild-type E. coli or deuterated ME
isotopomers and an E. coli mutant with no functional deoxyxylulose phosphate synthase dxs gene
showed that no apparent changes occurred at the level of the carbon atoms corresponding to C1, C4,
and C5 of IPP or DMAPP. All corresponding deuteriums from DX or from ME were preserved in the
prenyl chains of ubiquinone 3 and menaquinone 4 [4547].
Feeding of wild-type E. coli with [4-2H]DX 27 yielded a surprising labeling pattern of the prenyl
chains (Fig. 4). The terminal isoprene units derived from DMAPP 20 and serving as starter for the elon-
gation of the prenyl chain by the prenyl transferase showed considerable deuterium retention (up to
70 %), whereas all those derived from IPP 19 had no trace of deuterium [48]. This was confirmed by
feeding [3,5,5,5-2H4]ME to an E. coli strain with a deleted dxs gene [47]. In such a strain, which has to
be grown in the presence of free DX or ME, all isoprene units are derived from labeled precursor added
to the culture medium. The expected retention of the three deuteriums of the ME methyl group was ob-

Fig. 4 Independent IPP and DMAPP biosynthesis in the MEP pathway. Incubation of [4-2H]-1-deoxy-D-xylulose
27 or [3-2H]-2-C-methyl-D-erythritol 28.

2003 IUPAC, Pure and Applied Chemistry 75, 375387

382 M. ROHMER

served, and their signal served as internal reference. The 3-2H was quantitatively retained only in the
DMAPP 20 derived unit, whereas it was completely lost in those derived from IPP 19 (Fig. 4). Such a
different labeling pattern in isoprene units, depending on their origin from IPP or from DMAPP, pointed
to the possible presence of two branches leading separately to the two precursors of isoprene units.
Furthermore, the activity of IPP isomerase is barely detectable in wild-type E. coli, and the deletion of
the idi gene encoding the IPP isomerase does not affect the growth of the bacterium [49]. Definitive proof
of the branching was obtained by a genetic method [50]. The deletion of the dxr gene in E. coli is res-
cued either by the addition of free ME to the culture medium or by the introduction of the genes allow-
ing the utilization of exogenous MVA. When the strain is growing on MVA, it only synthesizes IPP from
MVA and has to rely on a functional protein capable of isomerizing IPP into DMAPP. If in addition the
idi gene is deleted, the novel strain does not grow anymore on MVA, indicating that there is no other way
in E. coli for converting IPP into DMAPP than the reaction catalyzed by the protein encoded by the idi
gene. It grew, however, normally on ME. This shows that IPP and DMAPP can be independently syn-
thesized from MEP derivative by two separate branches: an IPP branch characterized by deuterium loss
from [4-2H]DX or from [3-2H]ME and a DMAPP branch characterized by deuterium retention.
The fate of the 4-H hydrogen of DX can thus be utilized for the detection of a possible branching
of the MEP pathway in plants (Fig. 4). Incubation of a Catharanthus roseus cell culture with
[1-13C, 4-2H]DX resulted in 13C labeling of all isoprene units with no trace of deuterium, suggesting
that no branching occurred, at least in the utilized culture conditions [51]. Feeding of [4-2H]DX to eu-
calyptus twiglets was followed by a slight (0.9 %) but significant deuterium incorporation into the
monoterpene cineol, but only in the DMAPP-derived moiety, with nearly no deuterium found in the iso-
prene unit derived from IPP [52]. Finally, incorporation of [4-2H]DX into the plastid isoprenoids from
a tobacco BY-2 cell culture resulted in the labeling of all isoprene units of phytoene and plastoquinone,
whatever their origin from IPP or from DMAPP was, with the same isotope abundance (ca. 15 %) [53].
These experiments revealed the presence of the branching in plant cells. In C. roseus, only the IPP
branch was detected. In eucalyptus, like in E. coli, both IPP and DMAPP branches were found. In to-
bacco, the DMAPP branch significantly contributed to the formation of all isoprene units. Definitive in-
terpretation of the labeling patterns was only possible once more insight into the reduction reactions in-
volved in the last two steps has been obtained.

LAST STEPS: FROM METHYLERYTHRITOL CYCLODIPHOSPHATE TO IPP AND

DMAPP
Two additional genes, gcpE (or ispG) and lytB (or ispH), were always found accompanying the formerly
identified genes of the MEP pathway. Deletion of these genes was lethal for E. coli, but rescued by the
insertion of the three genes allowing the utilization of exogenous mevalonate, indicating that they were
implied in this biosynthetic route [5456]. An E. coli strain with a deleted gcpE gene and capable of
utilizing mevalonate accumulated methylerythritol cyclodiphosphate 25, suggesting that this intermedi-
ate is the substrate of the corresponding GcpE enzyme [57]. Crude cell-free systems from an E. coli
strain overexpressing the biosynthesis of GcpE were capable of converting ME cyclodiphosphate 25
into (E)-4-hydroxy-2-methylbut-2-enyl diphosphate 25 (HMBPP) [58]. This protein was later shown to
possess an oxygen-sensitive [4Fe-4S] cluster involved in a reductive process corresponding to two suc-
cessive one-electron transfers [59]. Regeneration of the reducing system could be performed either by
flavodoxin/flavodoxin reductase/NADPH or by deazaflavin via the semiquinone radical [59]. The ac-
cumulation of HMBPP 26 in a lytB-deficient E. coli mutant strongly suggested that HMBPP is the sub-
strate of the LytB protein [60]. A different approach also shed light on the role of the gcpE gene. An E.
coli strain overexpressing the gene of the xylulose kinase, an enzyme capable of phosphorylating free
DX, and all genes encoding enzymes located upstream of the GcpE protein in the MEP pathway accu-
mulated ME cyclodiphosphate upon incubation with uniformly 13C-labeled DX [61]. This experimen-

2003 IUPAC, Pure and Applied Chemistry 75, 375387

 Mevalonate-independent MEP pathway for isoprenoid biosynthesis 383

tal procedure allowed the detection of intermediates by directly running the 13C NMR spectra on crude
cell-free systems. Additional overexpression of the GcpE gene was followed by the accumulation of
HMBPP 26 [61] and of LytB resulted in the simultaneous formation of IPP 19 and DMAPP 20 [62].
This represented the first indication that the LytB protein was responsible for the conversion of HMBPP
26 into IPP 19 and DMAPP 20 and that the reaction catalyzed by this protein corresponds to the branch-
ing previously described. Finally incubation of 3H-labeled HMBPP with chromoplast from Capsicum
annuum and Narcissus pseudonarcissus led to the formation of radiolabeled carotenoids, confirming the
role of HMBPP 26 as intermediate in the MEP pathway [63].

CHARACTERIZATION OF REDUCTION STEPS

Z. mobilis proved a useful organism for the elucidation of the MEP pathway. In this bacterium, which
has no complete tricarboxylic acid cycle, the reducing agent pool derived from NADPH or reduced
nicotinamide adenine dinucleotide (NADH) results from three reactions of the glucose catabolism: the
oxidation of glucose 6-phosphate into gluconolactone 6-phosphate, the oxidation of glyceraldehyde 3-
phosphate into phosphoglyceric acid and the reduction of acetaldehyde into ethanol. When [1-2H]glu-
cose is utilized as sole carbon and energy source, a pool of the two possible isotopomers of 2H-labeled
NADPH/NADH in equivalent amounts results from the conjugate action of the three former enzymes.
Accordingly, a 2H transfer should characterize the reduction steps. After growing Z. mobilis on [1-
2H]glucose, two deuterium-labeled sites were characterized in the isoprene units of the hopanoids [64].
The first one corresponded to all carbon atoms derived from C4 of IPP or DMAPP. This labeling was
introduced by the well-documented reduction step catalyzed by the NADPH-dependent DXP reducto-
isomerase. The second site corresponded to carbon atoms derived from C2 and represented the signa-
ture of an additional reduction step, which was later shown to be catalyzed by the LytB protein. As the
next two enzymes implied in isoprenoid biosynthesis, the IPP isomerase and the prenyl transferase,
eliminate both the pro-R hydrogen of IPP, it is possible to propose a nearly complete hypothetical bio-
genetic scheme for the MEP pathway, including an interpretation for the different labeling patterns ob-
tained after incubation of deuterium-labeled DX or ME (Fig. 4) [53].

CONCLUSION: DISTRIBUTION OF THE PATHWAYS FOR ISOPRENOID

BIOSYNTHESISTHE MEP PATHWAY, A TARGET FOR NEW BIOCIDES
The MEP pathway is nearly completely elucidated, and its distribution is now fairly well known
[6568]. MVA and MEP pathways can be usually readily differentiated by the labeling pattern result-
ing from the incorporation of 13C-labeled glucose (Figs. 2 and 3). The MEP metabolic route is present
in most investigated bacteria. Its presence was detected either directly by biochemical proofs obtained
after incubation of 13C-labeled carbon sources and analyzing the labeling pattern of the isoprenoids or
via computer search of the genes encoding the enzymes of the pathway. Distribution of the MEP path-
way in eubacteria does not clearly fit with the phylogeny and is interpreted in terms of lateral gene
transfer [69,70]. Several Streptomyces spp. possess both pathways [71]. In this case, they are not ex-
pressed simultaneously. The MEP pathway is implied is the biosynthesis of essential metabolites such
as the prenyl chains of the menaquinones 4 (Fig. 1), whereas the MVA route is utilized for the forma-
tion of the prenyl moiety of secondary metabolites with antibiotic activity.
The MEP pathway is also present in all phototrophic organisms. In higher plants and most algae,
there is a dichotomy in the isoprenoid biosynthesis [18,65,66]. The MVA pathway is located in the cy-
toplasm and is responsible for the formation of sterols 9, triterpenes, the prenyl chain of ubiquinone 3
and most sesquiterpenoids [3,66,72]. The presence of the MEP pathway in contrast is restricted to the
chloroplasts. This is evidenced by the specific labeling pattern of typical chloroplast isoprenoids from
the photosynthetic apparatus (phytol 5 from chlorophylls, plastoquinone 6, carotenoids 7) [20] or sec-

2003 IUPAC, Pure and Applied Chemistry 75, 375387

384 M. ROHMER

ondary metabolites, whose biosynthesis is related to plastid-like structures (isoprene 1, monoterpenes,

diterpenes) [2328]. Definitive proof for the localization of the MEP pathway in chloroplasts is given
by the presence of a nucleotide sequence corresponding to a plastid-targeting peptide in the genes en-
coding the plant enzymes of the MEP pathway [35,7377]. In unicellular green algae, however, no di-
chotomy in the isoprenoid biosynthesis was observed [78]. All isoprenoids, including the sterols from
the cytoplasm, are derived from the MEP pathway. The clear-cut separation of the two biosynthetic
routes to isoprenoids in plants is a simplified presentation. Cross-talk between the two compartments,
cytoplasm and chloroplasts, and exchange of metabolites (IPP, DMAPP, geranyl diphosphate, and/or
farnesyl diphosphate) were often observed and are probably a way for the regulation of isoprenoid
biosynthesis to be explored [18,28,51]. Finally, the MEP pathway is also present in nonphototrophic eu-
karyotes, but belonging to phyla related to phototrophic unicellular eukaryotes. This is clearly the case
for the parasite responsible for malaria, Plasmodium falciparum, and for related species, which are all
characterized by the presence of apicoplasts, representing a residual chloroplast structure [79].
The MEP pathway is present in many pathogenic bacteria, in opportunistic pathogenic bacteria
responsible for severe infections in immunodeficient patients and in the parasite responsible for malaria
[79,80]. This means that all enzymes of this pathway are targets for inhibitors, which represent a novel
class of antibacterial or antiparasitic drugs [79,81]. This novel and unexpected aspect is rather interest-
ing in a context of increasing antibiotic and drug resistance of bacteria as well as of the malaria para-
site.

NOTE ADDED IN PROOF

Additional information concerning the last two steps of the MEP pathway was recently made available
[8285]. Both GcpE and LytB proteins possess a [4Fe-4S] cluster as prosthetic group. A spectrophoto-
metric enzyme test was described for both enzymes using dithionite as reducing agent [82,83]. Purified
LytB catalyzes the reduction of HMBP 26 into IPP 19 and DMAPP 20 via two successive one-electron
transfers and protonation of the resulting allylic anion. Much like for GcpE, the reduction may be per-
formed either chemically by the 5-deazaflavin semiquinone radical, or by the flavodoxin/flavodoxin re-
ductase/NADPH system [84,85]. The active reduced form of the LytB [4Fe-4S]1+ cluster was directly
characterized by electron paramagnetic resonance after reconstitution of the prosthetic group in the pu-
rified enzyme and reduction with dithionite [85].

REFERENCES
1. K. Bloch. Steroids 57, 378383 (1992).
2. D. A. Bochar, J. A. Friesen, C. V. Stauffacher, V. W. Rodwell. In Comprehensive Natural Product
Chemistry. Isoprenoids Including Carotenoids and Isoprenoids, D. E. Cane (Ed.), Vol. 2, pp.
1544, Pergamon, Oxford, UK (1999) and references therein.
3. M. Rohmer. In Comprehensive Natural Product Chemistry. Isoprenoids Including Carotenoids
and Steroids, D. E. Cane (Ed.), Vol. 2, pp. 4567, Pergamon, Oxford, UK (1999) and references
therein.
4. K. S. Treharne, E. I. Mercer, T. W. Goodwin. Biochem. J. 99, 239245 (1966).
5. D. V. Banthorpe, B. V. Charlwood, M. J. O. Francis. Chem. Rev. 72, 115155 (1972).
6. T. J. Bach and H. K. Lichtenthaler. In Ecology and Metabolism of Plant Lipids, G. Fuller and W.
D. Nes (Eds.), Vol. 325, 109139, American Chemical Society, Washington, DC (1987).
7. T. J. Bach, A. Motel, T. Weber. Rec. Adv. Phytochem. 24, 182 (1990).
8. F. Ltke-Brinkhaus and H. Kleinig. Planta 171, 406411 (1987).
9. W. Kohl, A. Gloe, H. Reichenbach. J. Gen. Microbiol. 129, 16291635 (1983).
10. G. Britton, T. W. Goodwin, W. S. J. Lockley, A. P. Mundy, N. J. Patel. Chem. Commun. 2728
(1979).

2003 IUPAC, Pure and Applied Chemistry 75, 375387

 Mevalonate-independent MEP pathway for isoprenoid biosynthesis 385

11. N. Orihara, K. Furihata, H. Seto. J. Antibotics 50, 979981 (1997), and references therein.
12. S. Pandian, S. Saengchan, T. S. Raman. Biochem. J. 295, 675691 (1981).
13. D. Zhou and R. H. White. Biochem. J. 273, 627634 (1991).
14. G. Flesch and M. Rohmer. Eur. J. Biochem. 175, 405411 (1988).
15. M. Rohmer. Pure Appl. Chem. 65, 12931298 (1993).
16. M. Rohmer, M. Knani, P. Simonin, B. Sutter, H. Sahm. Biochem. J. 295, 517524 (1993).
17. M. Rohmer, M. Seemann, S. Horbach, S. Bringer-Meyer, H. Sahm. J. Am. Chem. Soc. 118,
25642566 (1996).
18. M. Schwarz. Ph.D. Thesis Nb 10951. Eidgenssische Technische Hochschule, Zrich,
Switzerland (1994).
19. M. Schwarz and D. Arigoni. In Comprehensive Natural Product Chemistry. Isoprenoids Including
Carotenoids and Steroids, D. E. Cane (Ed.), Vol. 2, pp. 367340, Pergamon, Oxford, UK (1999)
and references therein.
20. H. K. Lichtenthaler, J. Schwender, A. Disch, M. Rohmer. FEBS Lett. 400, 271274 (1997).
21. J. G. Zeidler, H. K. Lichtenthaler, H. U. May, F. W. Lichtenthaler. Z. Naturforsch. 52c, 1523
(1997).
22. J. G. Zeidler and H. K. Lichtenthaler. Planta 213, 323326 (2001).
23. A. J. Barlow, H. Becker, K. P. Adam. Phytochemistry 57, 714 (2001).
24. W. Eisenreich, S. Sagner, M. H. Zenk, A. Bacher. Tetrahedron Lett. 38, 38893892 (1997).
25. W. Eisenreich, B. Menhard, P. J. Hylands, M. H. Zenk, A. Bacher. Proc. Natl. Acad. Sci. USA 93,
64316436 (1996).
26. W. Knss, B. Reuter, J. Zapp. Biochem. J. 326, 449454 (1997).
27. N. Tott, L. Charon, M. Rohmer, F. Compernolle, I. Baboeuf, J. M. C. Geuns. Tetrahedron Lett.
41, 64076410 (2000).
28. K.P. Adam and J. Zapp. Phytochemistry 48, 953959 (1998).
29. H. Hoeksema and L. Baczynskyj. J. Antibiotics 29, 688690 (1976).
30. S. David, B. Estramareix, J. C. Fischer, M. Thrisod. J. Am. Chem. Soc. 103, 73417342 (1981).
31. D. E. Cane, S. C. Du, J. K. Robinson, Y. J. Hsiung, I. D. Spenser. J. Am. Chem. Soc. 121,
77227723 (1999).
32. S. T. J. Broers. Ph.D. Thesis Nb. 10978. Eidgenssische Technische Hochschule, Zrich,
Switzerland (1994).
33. G. A. Sprenger, U. Schrken, T. Wiegert, S. Grolle, A. A. de Graaf, S. V. Taylor, T. P. Begley, S.
Bringer-Meyer, H. Sahm. Proc. Natl. Acad. Sci. USA 94, 1285712862 (1997).
34. L. M. Lois, N. Campos, S. Rosa-Putra, K. Danielsen, M. Rohmer, A. Boronat. Proc. Natl. Acad.
Sci. USA 95, 21052110 (1998).
35. B. M. Lange, M. R. Wildung, D. McCaskill, R. Croteau. Proc. Natl. Acad. Sci. USA 95,
21002104 (1998).
36. T. Duvold, J. M. Bravo, C. Pale-Grosdemange, M. Rohmer. Tetrahedron Lett. 38, 47694772
(1997).
37. T. Duvold, P. Cal, J. M. Bravo, M. Rohmer. Tetrahedron Lett. 38, 61816184 (1997).
38. S. Takahashi, T. Kuzuyama, H. Watanabe, H. Seto. Proc. Natl. Acad. Sci. USA 95, 98799884
(1998).
39. F. Rohdich, J. Wungsintaweekul, W. Eisenreich, G. Richter, C. A. Schuhr, S. Hecht, M. H. Zenk,
A. Bacher. Proc. Natl. Acad. Sci. USA 97, 64516456 (2000).
40. H. Lttgen, F. Rohdich, S. Herz, J. Wungsintaweekul, S. Hecht, C. A. Schuhr, M. Fellermeier, S.
Sagner, M. H. Zenk, A. Bacher, W. Eisenreich. Proc. Natl. Acad. Sci. USA 97, 10621067 (2000).
41. S. Herz, J. Wungsintaweekul, C. A. Schuhr, S. Hecht, H. Lttgen, S. Sagner, M. Fellermeier, W.
Eisenreich, M. H. Zenk, A. Bacher, F. Rohdich. Proc. Natl. Acad. Sci. USA 97, 24862490
(2000).

2003 IUPAC, Pure and Applied Chemistry 75, 375387

386 M. ROHMER

42. T. Kuzuyama, M. Takagi, K. Kaneda, T. Dairi, H. Seto. Tetrahedron Lett. 41, 403706 (2000).
43. T. Kuzuyama, M. Takagi, K. Kaneda, H. Watanabe, T. Dairi, H. Seto. Tetrahedron Lett. 41,
29252928 (2000).
44. M. Takagi, T. Kuzuyama, K. Kaneda, H. Watanabe, T. Dairi, H. Seto. Tetrahedron Lett. 41,
33953398 (2000).
45. T. Duvold. Ph.D. Thesis. Universit Louis Pasteur, Strasbourg, France (1997).
46. J. L. Giner and B. Jaun. Tetrahedron Lett. 39, 80218022 (1998).
47. L. Charon, J. F. Hoeffler, C. Pale-Grosdemange, L. M. Lois, N. Campos, A. Boronat, M. Rohmer.
Biochem. J. 346, 737742 (2000).
48. J. L. Giner, B. Jaun, D. Arigoni. Chem. Commun. 717718 (1998).
49. F. M. Hahn, A. P. Hurlburt, C. D. Poulter. J. Bacteriol. 181, 44994504 (1991).
50. M. Rodrguez-Concepcin, N. Campos, L. M. Lois, C. Maldonaldo, J. M. Hoeffler, C.
Grosdemange-Billiard, M. Rohmer, A. Boronat. FEBS Lett. 465, 157160 (2000).
51. D. Arigoni, W. Eisenreich, C. Latzel, S. Sagner, T. Radykewicz, M. H. Zenk, A. Bacher. Proc.
Natl. Acad. Sci. USA 96, 13091314 (1999).
52. C. Rieder, B. Jaun, D. Arigoni. Helv. Chim. Acta 83, 25042513 (2000).
53. J. F. Hoeffler, A. Hemmerlin, C. Grosdemange-Billiard, T. J. Bach, M. Rohmer. Biochem. J. 366,
573583 (2002).
54. N. Campos, M. Rodrguez-Concepcin, M. Seemann, M. Rohmer, A. Boronat. FEBS Lett. 488,
170173 (2001).
55. B. Altincicek, A. K. Kollas, S. Sanderbrand, J. Wiesner, M. Hintz, E. Beck, H. Jomaa. J.
Bacteriol. 183, 24112416 (2001).
56. B. Altincicek, A. K. Kollas, M. Eberl, J. Wiesner, S. Sanderbrand, M. Hintz, E. Beck, H. Jomaa.
FEBS Lett. 499, 3740 (2001).
57. M. Seemann, N. Campos, M. Rodrguez-Concepcin, J. F. Hoeffler, C. Grosdemange-Billiard, A.
Boronat, M. Rohmer. Tetrahedron Lett. 43, 775778 (2002).
58. M. Wolff, M. Seemann, C. Grosdemange-Billiard, D. Tritsch, N. Campos, M. Rodrguez-
Concepcin, A. Boronat, M. Rohmer. Tetrahedron Lett. 43, 25552559 (2002).
59. M. Seemann, B. Tse Sum Bui, M. Wolff, D. Tritsch, N. Campos, A. Boronat, A. Marquet, M.
Rohmer. Angew. Chem. 41, 43374339 (2002).
60. M. Hintz, A. Reichenberg, B. Altincicek, U. Bahr, R. M. Gschwind, A. K. Kollas, E. Beck, J.
Wiesner, M. Eberl, H. Jomaa. FEBS Lett. 509, 317322 (2001).
61. S. Hecht, W. Eisenreich, P. Adam, S. Amslinger, K. Kis, A. Bacher, D. Arigoni, F. Rohdich. Proc.
Natl. Acad. Sci. USA 98, 1483714842 (2001).
62. F. Rohdich, S. Hecht, K. Grtner, P. Adam, C. Krieger, S. Amslinger, D. Arigoni, A. Bacher, W.
Eisenreich. Proc. Natl. Acad. Sci. USA 99, 11581163 (2002).
63. W. Gao, R. Loeser, M. Raschke, M. A. Dessoy, M. Fulhorst, H. Alpermann, L. A. Wessjohann,
M. H. Zenk. Angew. Chem. 41, 26042607 (2002).
64. L. Charon, C. Pale-Grosdemange, M. Rohmer. Tetrahedron Lett. 40, 72317234 (1999).
65. M. Rohmer. Nat. Prod. Rep. 11, 565574 (1999) and references therein.
66. H. K. Lichtenthaler. Annu. Rev. Plant Physiol. Plant Mol. Biol. 50, 4665 (1999) and references
therein.
67. F. Rohdich, K. Kis, A. Bacher, W. Eisenreich. Curr. Opinion Chem. Biol. 5, 535540 (2001) and
references therein.
68. W. Eisenreich, F. Rohdich, A. Bacher. Trends Plant Sci. 6, 7884 (2001) and references therein.
69. Y. Boucher and W. F. Doolittle. Mol. Microbiol. 37, 703716 (2000).
70. B. M. Lange, T. Rujan, W. Martin, R. Croteau. Proc. Natl. Acad. Sci. USA 97, 1317213177
(2000).
71. H. Seto, H. Watanabe, K. Furihata. Tetrahedron Lett. 37, 79797982 (1996).

2003 IUPAC, Pure and Applied Chemistry 75, 375387

 Mevalonate-independent MEP pathway for isoprenoid biosynthesis 387

72. A. Disch, A. Hemmerlin, T. J. Bach, M. Rohmer. Biochem. J. 231, 615621 (1998).

73. B. M. Lange and R. Croteau. Arch. Biochem. Biophys. 365, 170174 (1999).
74. F. Rohdich, J. Wungsintaweekul, W. Eisenreich, G. Richter, C. A. Schuhr, S. Hecht, M. H. Zenk,
A. Bacher. Proc. Natl. Acad. Sci. USA 97, 64516456 (2000).
75. F. Rohdich, J. Wungsintaweekul, H. Lttgen, M. Fischer, W. Eisenreich, C. A. Schuhr, M.
Felleremeier, N. Schramek, M. H. Zenk, A. Bacher. Proc. Natl. Acad. Sci. USA 97, 82518256
(2000).
76. J. Querol, N. Campos, S. Imperial, A. Boronat, M. Rodrguez-Concepcin. FEBS Lett. 514,
343346 (2002).
77. F. X. Cunningham Jr., T. P. Lafond, E. Gantt. J. Bacteriol. 182, 58415848 (2000).
78. A. Disch, J. Schwender, C. Mller, H. K. Lichtenthaler, M. Rohmer. Biochem. J. 333, 381388
(1998).
79. H. Jomaa, J. Wiesner, S. Sanderbrand, A. Altincicek, C. Weidemeyer, M. Hintz, I. Trbachova,
M. Eberl, J. Zeidler, H. K. Lichtenthaler, D. Soldati, E. Beck. Science 285, 15731576 (1999).
80. S. Rosa-Putra, A. Disch, J. M. Bravo, M. Rohmer. FEMS Microbiol. Lett. 164, 169175 (1998).
81. M. Rohmer. Progr. Drug Res. 50, 135154 (1998).
82. A. K. Kollas, E. C. Duin, M. Eberl, B. Altincicek, M. Hintz, A. Reichenberg, D. Henschker, A.
Henne, I. Steinbrecher, D. N. Ostrovsky, R. Hedderich, E. Beck, H. Jomaa, J. Wiesner. FEBS Lett.
532, 432436 (2002).
83. B. Altincicek, E. C. Duin, A. Reichenberg, R. Hedderich, A. K. Kollas, M. Hintz, S. Wagner, J.
Wiesner, E. Beck, H. Jomaa. FEBS Lett. 532, 437440 (2002).
84. F. Rohdich, F. Zepeck, P. Adam, S. Hecht, J. Kaiser, R. Laupitz, T. Grwert, S. Amslinger, W.
Eisenreich, A. Bacher, D. Arigoni. Proc. Natl. Acad. Sci. USA 100, 15861591 (2003).
85. M. Wolff, M. Seemann, B. Tse Sum Bui, Y. Frapart, D. Tritsch, A. Garcia-Estrabot, M.
Rodriguez-Concepcion, A. Boronat, A. Marquet, M. Rohmer. FEBS Lett. In press.

2003 IUPAC, Pure and Applied Chemistry 75, 375387

78 Review TRENDS in Plant Science Vol.6 No.2 February 2001
Deoxyxylulose phosphate pathway
to terpenoids
Wolfgang Eisenreich, Felix Rohdich and Adelbert Bacher
Recently, a mevalonate-independent pathway was discovered in bacteria and
plants that leads to the formation of isopentenyl diphosphate and dimethylallyl
diphosphate, the two basic precursors of isoprenoids. Although many details
of the widely distributed pathway are unknown, some intermediates,
mechanisms, enzymes and genes of this novel route have been identified.
Information on this pathway could provide the basis for the development of
new antibiotics, herbicides and antimalarials.
Terpenoids and isoprenoids are a structurally diverse
group of natural products. More than 25 000
representatives with a variety of biological functions
have been reported in the plant kingdom1. To give just
a few examples, carotenoids serve as light-harvesting
and light-protecting pigments, sterols play important
roles as modulators of membrane properties, the
phytol side chain of chlorophyll (the most abundant
organic pigment) is of terpenoid origin, and a wide
variety of plant terpenoids function as insect
attractants or repellents. Various terpenoids have
attracted commercial interest as pharmaceuticals or
nutraceuticals. Thus, paclitaxel (Taxol), a diterpene
from yew, has been established as a major cytostatic
agent. Lycopene and lutein were recently registered
as oncopreventive agents.
Pathways of terpenoid biosynthesis
All terpenoids are biosynthesized from just two C5
precursors: isopentenyl diphosphate (IPP) and
dimethylallyl diphosphate (DMAPP). The
biosynthesis of these universal terpene precursors via
mevalonate (Fig. 1) has been elucidated by studies of
yeast and animal cells (reviewed in Ref. 2).
Radioactivity from labelled mevalonate has been
reported to be diverted to a wide variety of plant
terpenes, albeit at incorporation rates well below 1%,
which were explained by poor uptake of mevalonate
into plant cells or by compartmentalization.
Several results inconsistent with the mevalonate
pathway led to the discovery of a non-mevalonate
pathway of terpenoid biosynthesis. For example,
Wolfgang Eisenreich*
Felix Rohdich
unexpected 13C labels were found in hopanoids from
Adelbert Bacher certain bacteria that had been grown with 13C-
Lehrstuhl fr Organische labelled acetates3. Furthermore, 13C-labelled glucose
Chemie und Biochemie, applied to E. coli and to seedlings of Ginkgo biloba
Technische Universitt
Mnchen,
caused labelling patterns in ubiquinones
Lichtenbergstrasse 4, (S.T.J. Broers, PhD thesis, ETH Zrich, 1994) and
D-85747 Garching, ginkgolides (M.K. Schwarz, PhD thesis, ETH Zrich,
Germany.
1994) that were clearly inconsistent with the
*e-mail:
wolfgang.eisenreich@ mevalonate route. These labelling patterns suggested
ch.tum.de that C-3, C-2 and C-1 of a triose could supply C-1, C-2
http://plants.trends.com 1360-1385/01/$ see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S1360-1385(00)01812-4
and C-4 of IPP, and that C-2 and C-3 of pyruvate could
supply C-3 and C-5 of IPP in the novel pathway
(reviewed in Refs 46).
A hypothetical mechanism was suggested with a
head-to-head condensation of D-glyceraldehyde 3-
phosphate and activated acetaldehyde derived from
pyruvate, resulting in the formation of 1-deoxy-D-
xylulose 5-phosphate as the first precursor of the
novel pathway (S.T.J. Broers, PhD thesis, ETH
Zrich, 1994; M.K. Schwarz, PhD thesis, ETH Zrich,
1994) (Fig. 1). This hypothesis was confirmed by
efficient incorporation of 2H-labelled 1-deoxy-D-
xylulose into the isoprenoid side chain of ubiquinone
in E. coli (S.T.J. Broers, PhD thesis, ETH Zrich,
1994). This pentulose had earlier been reported to
serve as a precursor for the biosynthesis of thiamine7,8
and pyridoxal9.
Deoxyxylulose phosphate pathway is widely
distributed in nature
Subsequent isotope incorporation studies using
bacteria, certain algae, plant cell cultures and plant
tissues demonstrated the formation of IPP and
DMAPP from exogenous 1-deoxyxylulose. These
seminal experiments have been reviewed
recently46,10 and are not covered in detail here. In
summary (Fig. 2), archaea, certain bacteria, yeasts,
fungi, some protozoa and animals appear to use the
mevalonate pathway. However, many bacteria
(including many human pathogens), green algae and
possibly the malaria parasite Plasmodium
falciparum appear to rely exclusively on the
mevalonate-independent pathway. Streptomycetes,
some algae, mosses and liverworts, marine diatoms,
and higher plants appear to use both pathways.
Finally, the capacity to biosynthesize isoprenoid
precursors appears to be absent in some parasitic
microorganisms that might be able to use terpenoids
from their host.
Crosstalk between the mevalonate and deoxyxylulose
phosphate pathways in plants
In higher plants, the mevalonate route operates in
the cytoplasm and mitochondria; sterols,
sesquiterpenes and ubiquinones are formed
predominantly via mevalonate. Isoprenoids
synthesized in the plastids (hemiterpenes,
monoterpenes, diterpenes and carotenoids) are
formed predominantly via 1-deoxyxylulose
5-phosphate (reviewed in Refs 46). The
 Review TRENDS in Plant Science Vol.6 No.2 February 2001 79

Fig. 1. Pathways of
isopentenyl diphosphate (a) (b)
(6) and dimethylallyl O
O O O O
diphosphate (7) 1 + 1
formation. The origins of + 9
SCoA SCoA 8 COO H O P O
carbon atoms from acetyl-
CoA (1), pyruvate (8) and OH O
glyceraldehyde 3-
phosphate (9) are CO2 dxs
indicated in blue, red and
green, respectively. (a) In O O OH O
the mevalonate pathway, 2 10
O
three molecules of acetyl-
SCoA O P O
CoA (1) are condensed
OH O
to 3-hydroxy-3- O
methylglutaryl-CoA (3) via 1 NADPH
dxr
acetoacetyl-CoA (2) by the SCoA
consecutive action of two
HSCoA NADP+
enzymes. Subsequent O
HO HO
reduction to mevalonate
(4), phosphorylation and 3 O P O 11
O
ATP-dependent HOOC COSCoA OH OH
decarboxylation produces
IPP (6), which can be 2 NADPH CTP
ispD NH2
converted to and from
DMAPP (7) by an 2 NADP+ HSCoA PPi
isomerase. (b) In
N
the deoxyxylulose HO HO O O
phosphate pathway, N O
D-glyceraldehyde
4 O P O P O O 12
3-phosphate (9) and HOOC CH2OH OH OH O O
pyruvate (8) are converted ATP
into 1-deoxy-D-xylulose HO OH
5-phosphate (10) in a ADP ATP ispE
decarboxylative reaction. ADP NH2
ATP
Subsequent
rearrangement and ADP O N
reduction leads to 2C-
HO O P O O
methyl-D-erythritol 4- O
N O
phosphate (11), which is
5
O O O O P O P O
converted into 2C-methyl-
O
D-erythritol 2,4-
HOOC CH2O P O P O OH OH O O

13
cyclodiphosphate (14) via O O
4-diphosphocytidyl-2C-
ispF HO OH
methyl-D-erythritol (12) ATP CMP
and 4-diphosphocytidyl- ADP + Pi CO2
2C-methyl-D-erythritol
2-phosphate (13). The O O
further reactions leading O O
O O P
to IPP (6) and DMAPP
6 P O
(7) are unknown. O P O P O O O
14
O O OH OH

O O O O
O O
7 O P O P O O P O
O P O P O O P
O O
O
O O O
6 7

TRENDS in Plant Science

compartmental separation of the two different IPP
biosynthetic pathways is not absolute because at
least one metabolite can be exchanged between the
compartments1113 (M.K. Schwarz, PhD thesis, ETH
Zrich, 1994). The extent of this crosstalk depends on
the species as well as on the presence and
concentration of exogenous precursors. Generally,
crosstalk contributions appear to be small in intact
plants under physiological conditions (<1%). Higher
values have been found in plant cell cultures in the
presence of exogenous 1-deoxyxylulose or
mevalonate. The nature of the metabolite(s)
exchanged between the compartments and the
regulation of the process remain to be established.
The crosstalk between the two terpenoid pathways
is one of the reasons for the historical failure to

http://plants.trends.com
80 Review TRENDS in Plant Science Vol.6 No.2 February 2001

recognize the deoxyxylulose route. In retrospect, the

Mevalonate Deoxyxylulose observed low incorporation rates of mevalonate and
acetate into many plant terpenoids were not because
of restricted uptake or restricted use of the labelled
Bacteria or
precursors but rather reflected the small crosstalk
Archaea contributions of mevalonate to the biosynthesis of
terpenoids that are predominantly formed from
Fungi 1-deoxyxylulose.

Algae and/or Genes and enzymes of the deoxyxylulose phosphate

pathway
Higher plants 1-Deoxy-D-xylulose 5-phosphate synthase
plastidic compartment In vivo experiments described above suggested that
cytosolic compartment the presumed intermediate 1-deoxy-D-xylulose
5-phosphate is synthesized by a transketolase-like
Protozoa decarboxylation from pyruvate and glyceraldehyde
3-phosphate (Fig. 1). Following this hypothesis,
Animals
TRENDS in Plant Science
database similarity searches with the E1 subunit
(EC 1.2.4.1) of the pyruvate dehydrogenase complex,
Fig. 2. Distribution in nature of the mevalonate and deoxyxylulose pyruvate decarboxylase (EC 4.1.1.1), and
phosphate pathway of isoprenoid biosynthesis. transketolase (EC 2.2.1.1) as query sequences
revealed a homologous unannotated gene at 9.0 min
Table 1. Recombinant enzymes of the deoxyxylulose phosphate pathway on the E. coli genome, directly downstream from the
Enzyme Property Refs ispA gene, that encodes farnesyldiphosphate
synthase14,15. The unannotated gene and ispA were
Organism KM vmax found as part of an operon, which suggested a
M mol min1 mg1 functional relationship. The unknown gene (now
1-Deoxy-D-xylulose 5-phosphate designated as dxs) was expressed in E. coli and the
synthase recombinant protein shown to catalyse the formation
Arabidopsis 18 of 1-deoxy-D-xylulose 5-phosphate from pyruvate and
Capsicum annuum 500a 750.0b 500.0 17 glyceraldehyde 3-phosphate.
Chlamydomonas 19 Further genome comparisons revealed putative
E. coli 96a 250.0b 300.0 14,15
dxs orthologues in many bacteria and plants.
Mentha piperita 16
Streptomyces sp. 65a 120.0b 370.0 20 1-Deoxy-D-xylulose 5-phosphate synthases from
Synechococcus leopoliensis 21 Mentha piperita16, Capsicum annuum17,
1-Deoxy-D-xylulose 5-phosphate Arabidopsis18, Chlamydomonas19, a Streptomyces
reductoisomerase species20 and Synechococcus leopoliensis21 have been
Arabidopsis 26 cloned and characterized in some detail. These
E. coli 250c 7.4d 11.8 23 enzymes require thiamine diphosphate and divalent
Mentha piperita 27 cations such as Mg2+ or Mn2+ for activity14,17,20.
Plasmodium falciparum 28
Catalytic properties have been reported for the
Synechocystis 25
enzymes from Streptomyces, E. coli and
4-Diphosphocytidyl-2C-methyl-D-
erythritol synthase
C. annuum14,15,17,20 (Table 1). The proteins from
Arabidopsis 500e 114f 67.0 34 M. piperita, C. annuum and Arabidopsis are in the
E. coli 131e 3.0f 23.0 31,33 nuclear genome and are preceded by putative
4-Diphosphocytidyl-2C-methyl-D- N-terminal plastid targeting sequences1618.
erythritol kinase Translocation into chloroplasts was recently shown
E. coli 34.0 35,38,39 for 1-deoxy-D-xylulose 5-phosphate synthase in
Lycopersicon esculentum 33.0 37 Arabidopsis18 and Lycopersicon esculentum22.
Mentha piperita 38
2C-Methyl-D-erythritol 2,4-
1-Deoxy-D-xylulose 5-phosphate reductoisomerase
cyclodiphosphate synthase
A rearrangement is required for the formation of the
E. coli 40,41
branched-chain isoprenoid precursors IPP and
Substrates are indicated by:
aD-glyceraldehyde
DMAPP from the linear pentulose phosphate11
3-phosphate.
bPyruvate.
(Fig. 1). The presence of a 1,2-ketol function in
c1-deoxy-D-xylulose
1-deoxy-D-xylulose 5-phosphate fulfils the
5-phosphate.
dNADPH.
requirements for a sigmatropic rearrangement
reaction as predicted by various in vivo labelling
e2C-methyl-D-erythritol 4-phosphate.
studies described above. An unannotated gene (yaeM)
fCTP.
at 4.2 min on the chromosomal map of E. coli was

http://plants.trends.com
Review TRENDS in Plant Science Vol.6 No.2 February 2001 81

Table 2. Occurrence of deoxyxylulose pathway genes in 21 completely sequenced bacteriaa

Organism Gene

dxs dxr ispD ispE ispF

Aquifex aeolicus AE000712 AE000688 AE000734 AE000713 AE000715
Bacillus subtilis D84432 Z99112 Z99104 Z99104 Z99104
Campylobacter jejuni AL139074 AL139078 AL139079 AL139077 AL139079
Chlamydia muridarum AE002329 AE002301 AE002343 AE002285 AE002340
Chlamydia pneumoniae CWLO29 AE001686 AE001618 AE001642 AE001675 AE001639
Chlamydophila pneumoniae AR39 AE002238 AE002203 AE002178 AE002249 AE002181
Chlamydia trachomatis AE001306 AE001281 AE001320 AE001352 AE001317
Deinococcus radiodurans AE001992 AE001994 AE002089 AE002089 AE001885
E. coli AF035440 AB013300 AF230736 AF216300 AF230738
Haemophilus influenzae U32822 U32763 U32750 U32834 U32750
Helicobacter pylori 26695 AE000552 AE000541 AE000610 AE000644 AE000610
Helicobacter pylori J99 AE001468 AE000001458 AE001474 AE001556 AE0001474
Mycobacterium tuberculosis Z96072 Z74024 Z92774 Z92752 Z94774
Neisseria meningitidis Z2491 AL162753 AL162752 AL162756 AL162755 AL162756
Neisseria meningitidis MC58 AE002536 AE002375 AE002501 AE002439 AE002501
Pseudomonas aeruginosa AE004821 AE004785 AE004783 AE004880 AE004782
Synechocystis sp. PCC6803 D90903 D64000 D90914 D90899 D90906
Thermotoga maritima AE001815.1 AE001754.1 AE001792 AE001791 AE001738
Treponema pallidum AE001253 AE001235 AE001227 AE001226 AE001227
Vibrio cholerae AE004173 AE004297 AE004139 AE004289 AE004139
Xylella fastidiosa AE004037 AE003942 AE003962 AE004071 AE003962
aPutative orthologs of dxs, dxr, ispD, ispE, and ispF are absent in the genomes of archaea, the parasitic bacteria Mycoplasma genitalium,
Borrelia burgdorferi and Rickettsia prowazekii, as well as in the genomes of the eukaryotes Saccharomyces cerevisiae, Caenorhabditis elegans
and Drosophila melanogaster.

identified that complemented mutants requiring
2C-methyl-D-erythritol for growth23. The gene, now
designated dxr, was expressed in E. coli and the
recombinant protein shown to catalyse the formation
of 2C-methyl-D-erythritol 4-phosphate from 1-deoxy-
D-xylulose 5-phosphate in a NADPH-dependent
reaction23 (Fig. 1).
The E. coli enzyme23,24 and its orthologues from the
bluegreen alga Synechocystis25, the plants
Arabidopsis26 and M. piperita27, and the protist
Plasmodium falciparum28 have been characterized in
some detail. The E. coli enzyme was reported to have
a molecular mass of 140 kDa with four subunits. The
enzyme requires divalent metal ions such as Mn2+,
Co2+ or Mg2+, and is strictly dependent on NADPH as
a cofactor23. The stereochemical features of the
reductoisomerase reaction have been studied with the
recombinant enzymes from E. coli and Synechocystis,
showing that HSi from C-4 of NADPH is transferred to
the HRe position at C-1 of 2C-methyl-D-erythritol
4-phosphate. Thus, 1-deoxy-D-xylulose 5-phosphate
reductoisomerases from E. coli29 and Synechocystis25
are class B dehydrogenases. The kinetic properties of
the E. coli enzyme were also studied in some detail
(Table 1). Catalytic rates of 11.8 mol min1 mg1 were
measured with Mn2+ at a pH optimum of 7.5. The
enzyme does not accept 1-deoxy-D-xylulose as a
substrate23.
Site-directed mutagenesis of the recombinant
E. coli protein revealed that Glu231 is involved in the
reaction and that His153, His209 and His257 might
be involved in the binding of 1-deoxy-D-xylulose
5-phosphate24. The N-terminal region is thought to
be the binding site for NADPH (Refs 24,26,27).
Fosmidomycin, a long-known antibiotic compound
that mimics the structure of the hypothetical
reaction intermediate 2C-methyl-D-erythrose
4-phosphate, inhibits the reaction with a Ki of
38 nM (Refs 28,30).
4-Diphosphocytidyl-2C-methyl-D-erythritol synthase
Crude enzyme fractions from wild-type E. coli
convert 2C-methyl-D-erythritol 4-phosphate into a
novel product in a CTP-dependent reaction31.
A database search for enzymes transferring a
cytidylphosphate residue to a polyol retrieved the
asc1 gene of a serotype-specific DNA region from
Haemophilus influenzae. The N-terminal domain of
the cognate enzyme had been shown to catalyse the
formation of CDPribitol from ribitol 5-phosphate
and CTP (Ref. 32). Subsequent similarity searches
starting from that gene retrieved the hitherto
unknown ygbP gene family (now designated ispD),
which is present in many bacteria as well as in
Arabidopsis. More specifically, ispD is present in
organisms with dxs and dxr genes but not in
organisms devoid of dxs and dxr genes (Table 2).
Thus, it appeared likely that ispD had a role in the
novel terpenoid pathway.
This hypothesis was confirmed by studies on
recombinant IspD protein, which was shown to
catalyse the formation of 4-diphosphocytidyl-2C-
methyl-D-erythritol from 2C-methyl-D-erythritol
4-phosphate and CTP (Refs 31,33; Fig. 1). Moreover, it
was shown that 4-diphosphocytidyl-2C-methyl-D-
erythritol is incorporated efficiently into carotenoids

http://plants.trends.com
82 Review TRENDS in Plant Science Vol.6 No.2 February 2001

Table 3. Putative ispD, ispE and ispF orthologues in incompletely sequenced organisms
Organism Gene

ispD ispE ispF

Eubacteria
Actinobacillus actinomycetes Contig320 Contig344 Contig320
Bacillus anthracis Banth_1680 Banth_1607 Banth_1680
Bacillus stearothermophilus Contig541 Contig541
Bordetella bronchisepticum Contig2244
Bordetella pertussis Contig301 Contig278 Contig301
Caulobacter crescentus C.cres._12574 C.cres._12574 C.cres._12574
Clostridium acetobutylicum C.aceto_gnl C.aceto_gnl C.aceto_gnl
Clostridium difficile Contig935 Contig985 Contig935
Corynebacterium diphteriae Contig239 Contig14 Contig239
Dehalococcoides ethenogenes Deth_1507 Deth_1545 Deth_1507
Desulfovibrio vulgaris Dvulg_gdv23 Dvulg_gdv23
Enterococcus faecalis Gef_10297 Gef_10286 Gef_10286
Geobacter sulfurreducens Gsulf_ggs98 Gsulf_ggs110 Gsulf_ggs98
Haemophilus ducreyi HTSC_730 HTSC_730 HTSC_730
Klebsiella pneumoniae Contig1719 Contig31 Contig1719
Legionella pneumophila CUGC_446 CUC_446
Mycobacterium avium M.avium_183 M.avium_24 M.avium_183
Mycobacterium bovis Contig442 Contig365 Contig353
Mycobacterium leprae Sanger_1769 Sanger_1769 Sanger_1769
Neisseria gonorrhoeae Contig9 Contig9 Contig9
Pasteurella multicoda CBCUMN_747 CBCUMN_747 CBCUMN_747
Porphyromonas gingivalis P.ging._GPG.con P.ging._GPG.con P.ging._GPG.con
Pseudomonas putida Pputida_10705 Pputida_10708 Pputida_10705
Rhodobacter capsulatus X12382 X12382
Salmonella enteritidis Contig1007
Salmonella paratyphi Spara_B_SPA.0.2916 Spara_B_SPA.0.2446 Spara_B_SPA.0.2916
Salmonella typhi S.typhi_CT18 S.typhi_CT18 S.typhi_CT18
Salmonella typhimurium Contig997 M77236 Contig997
Shewanella putrefaciens Sputr_6406 Sputr_6406 Sputr_6406
Streptomyces coelicor AL160331 AL359989 AL160331
Thiobacillus ferrooxidans T_ferro._6139 T_ferro._6154 T_ferro._6139
Treponema denticola Tdent_gtd269 Tdent_gtd51 Tdent_gtd269
Yersinia pestis Contig854 Contig855 Contig854
Zymomonas mobilis AF176314 AF088896 AF176314
Plants
Arabidopsis thaliana AF230737 AF288615 AC010852
Catharanthus roseus AF250236
Mentha piperita AF179283
Solanum esculentum AF263101
Protists
Plasmodium falciparum AF279661
indicates data not available.

of red pepper31 and that IspD is essential for IPP
biosynthesis in E. coli33. More recently, recombinant
IspD from Arabidopsis has been shown to catalyse the
same reaction34. The kinetic properties of the E. coli
and Arabidopsis enzyme are summarized in Table 1.
Both enzymes require a divalent cation, preferably
Mg2+. The E. coli protein appears to be a dimer with
25 kDa subunits31, but the Arabidopsis protein shows
a tendency for self-aggregation depending on the
buffer system34. Sequence analysis of the Arabidopsis
protein sequence revealed a putative plastid import
sequence34.
4-Diphosphocytidyl-2C-methyl-D-erythritol kinase
A bioinformatics approach has been used to identify
more genes following the distribution of the
orthologous sets of dxs, dxr and ispD genes. Whole
genome comparisons retrieved putative
deoxyxylulose pathway genes from the completely
sequenced genomes of organisms known to use the
deoxyxylulose pathway but not from the genomes of
organisms that use the mevalonate pathway35. As a
hypothetical candidate following this distribution, the
ychB gene with hitherto unknown function (now
designated ispE) was identified in the genome of
E. coli and many other organisms35. The involvement
of the orthologous gene product from tomato in
chromoplast ripening had already been proposed36.
Both known plant orthologues (tomato and
peppermint) carry putative plastid target
sequences37,38. Sequence analysis revealed putative
ATP binding sites35,37,38.

http://plants.trends.com
 Review TRENDS in Plant Science Vol.6 No.2 February 2001 83

The purified recombinant enzyme from E. coli was diphosphate (Fig. 1) is an intermediate of the
shown to catalyse the ATP-dependent deoxyxylulose phosphate pathway40.
phosphorylation of 4-diphosphocytidyl-2C-methyl-D-
erythritol into 4-diphosphocytidyl-2C-methyl-D- Further downstream reactions
erythritol 2-phosphate35,39 (Fig. 1, Table 1) at a rate35 The genes, enzymes and reactions involved in the
of 34 mol min1 mg1. The recombinant enzyme from conversion of 2C-methyl-D-erythritol 2,4-
tomato37 catalysed the same reaction at a rate of cyclodiphosphate into IPP and DMAPP are currently
33 mol min1 mg1 (Table 1). However, unknown. However, from a structural point of view, a
phosphorylation of isopentenyl monophosphate by ring-opening reaction, two dehydration steps and two
IspE from E. coli or M. piperita was reported, albeit at reduction steps are required to biosynthesize IPP and
extremely low rates at 1.0 nmol min1 mg1 and DMAPP. Moreover, there is evidence that IPP and
0.1 nmol min1 mg1, respectively38. More recent data DMAPP are biosynthesized via independent
with the recombinant enzymes from E. coli and mechanisms in the late steps of the novel pathway45,46
tomato37 gave catalytic rates below 0.002 nmol and that the gene product of lytB from E. coli is
min1 mg1. Therefore, it is hardly conceivable that involved in one of the late steps of IPP and DMAPP
phosphorylation of isopentenyl monophosphate biosynthesis47.
catalysed by IspE is metabolically relevant.
Possible applications
2C-Methyl-D-erythritol 2,4-cyclodiphosphate synthase The rapidly increasing resistance of many human
The next step in the pathway was also identified pathogens to all currently available drugs generates
using bioinformatics. Genome analyses had shown an urgent demand for novel therapeutic approaches.
that many putative orthologues of the E. coli ygbB The deoxyxylulose phosphate pathway is used in
gene (now designated ispF) were linked or fused to many pathogenic microorganisms, as well as in
putative orthologues of ispD (Refs 31,40). Moreover, P. falciparum (Tables 2,3). In addition, the pathway
the occurrence of ispF correlates with the presence of does not occur in humans and animals. This makes
the deoxyxylulose pathway genes (Tables 2,3). Based the deoxyxylulose pathway of isoprenoid biosynthesis
on these findings, the E. coli ispF gene was expressed an ideal target for the development of novel
and the recombinant protein shown to catalyse the antibiotics and antimalarial agents. P. falciparum is
conversion of 4-diphosphocytidyl-2C-methyl-D- sensitive to the inhibition of this alternative
erythritol 2-phosphate into 2C-methyl-D-erythritol terpenoid biosynthetic pathway by fosmidomycin28.
Acknowledgements 2,4-cyclodiphosphate40,41 (Fig. 1). The enzyme In plants, the inhibition of the deoxyxylulose
Our work was
supported by grants
requires Mg2+ or Mn2+ but no other cofactors. pathway might be the basis for the development of
from the Deutsche The product, 2C-methyl-D-erythritol 2,4- novel herbicides19,48. Detailed knowledge of the
Forschungsgemeinschaft cyclodiphosphate, is well known as it accumulates mechanisms and regulation of the pathway will also
(SFB 369), the Fonds der
under oxidative stress in some bacteria4244. Results benefit the biotechnological production of
Chemischen Industrie and
the Hans-Fischer- obtained from incorporation experiments with commercially interesting isoprenoids, such as
Gesellschaft. chromoplasts of C. annuum suggest that the cyclic carotenoids49.
References 8 David, S. et al. (1981) 1-Deoxy-D-threo-2- thiamin, and pyridoxol. Proc. Natl. Acad. Sci.
1 Sacchettini, J.C. and Poulter, C.D. (1997) pentulose: the precursor of the five-carbon chain U. S. A. 94, 1285712862
Creating isoprenoid diversity. Science of the thiazole of thiamine. J. Am. Chem. Soc. 15 Lois, L.M. et al. (1998) Cloning and
277, 17881789 103, 73417342 characterization of a gene from Escherichia coli
2 Bochar, D.A. et al. (1999) Biosynthesis of 9 Hill, R.E. et al. (1989) Biosynthesis of vitamin B6: encoding a transketolase-like enzyme that
mevalonic acid from acetyl-CoA. In incorporation of D-1-deoxyxylulose. J. Am. Chem. catalyses the synthesis of D-1-deoxyxylulose 5-
Comprehensive Natural Product Chemistry (Vol. Soc. 111, 19161917 phosphate, a common precursur for isoprenoid,
2) (Cane, D., ed.), pp. 1544, Pergamon 10 Boucher, Y. and Doolitle, W.F. (2000) The role of thiamin, and pyridoxol biosynthesis. Proc. Natl.
3 Rohmer, M. et al. (1993) Isoprenoid biosynthesis lateral gene transfer in the evolution of isoprenoid Acad. Sci. U. S. A. 95, 21052110
in bacteria: a novel pathway for the early steps biosynthesis pathways. Mol. Microbiol. 16 Lange, B.M. et al. (1998) A family of transketolases
leading to isopentenyl diphosphate. Biochem. J. 37, 703716 that directs isoprenoid biosynthesis via a
295, 517524 11 Arigoni, D. et al. (1997) Terpene biosynthesis from mevalonate-independent pathway. Proc. Natl.
4 Eisenreich, W. et al. (1998) The deoxyxylulose 1-deoxy-D-xylulose in higher plants by Acad. Sci. U. S. A. 95, 21002104
phosphate pathway of terpenoid biosynthesis in intramolecular skeletal rearrangement. Proc. 17 Bouvier, F. et al. (1998) Dedicated roles of plastid
plants and microorganisms. Chem. Biol. Natl. Acad. Sci. U. S. A. 94, 1060010605 transketolases during the early onset of
5, R221R233 12 Adam, K-P. and Zapp, J. (1998) Biosynthesis of isoprenoid biogenesis in pepper fruits. Plant
5 Arigoni, D. and Schwarz, M.K. (1999) Ginkgolide the isoprene units of chamomile sesqiterpenes. Physiol. 117, 14231431
biosynthesis. In Comprehensive Natural Product Phytochemistry 48, 953959 18 Araki, N. et al. (2000) Temperature-sensitive
Chemistry (Vol. 2) (Cane, D., ed.), pp. 367399, 13 Nabeta, K. et al. (1995) Biosynthesis of Arabidopsis mutant defective in 1-deoxy-D-
Pergamon heteroscyphic acid A in cell cultures of xylulose 5-phosphate synthase within the plastid
6 Rohmer, M. (1999) A mevalonate-independent Heteroscyphus planus: non-equivalent labelling of non-mevalonate pathway of isoprenoid
route to isopentenyl diphosphate. In C-5 units in diterpene biosynthesis. J. Chem. Soc., biosynthesis. Physiol. Plant. 108, 1924
Comprehensive Natural Product Chemistry (Vol. Chem. Commun. 681682 19 Lichtenthaler, H.K. et al. (2000) The non-
2) (Cane, D., ed.), pp. 4568, Pergamon 14 Sprenger, G.A. et al. (1997) Identification of a mevalonate isoprenoid biosynthesis of plants as a
7 White, R.H. (1978) Stable isotope studies on the thiamin-dependent synthase in Escherichia coli test system for new herbicides and drugs against
biosynthesis of the thiazole moiety of thiamin in required for the formation of the 1-deoxy-D- pathogenic bacteria and the malaria parasite.
Escherichia coli. Biochemistry 17, 38333840 xylulose 5-phosphate precursor to isoprenoids, Z. Naturforsch. 55c, 305313

http://plants.trends.com
84 Review
20 Kuzuyama, T. et al. (2000) Cloning and
characterization of 1-deoxy-D-xylulose 5-phosphate
synthase from Streptomyces sp. strain CL190, which
uses both the mevalonate and nonmevalonate
pathways for isopentenyl diphosphate biosynthesis.
J. Bacteriol. 182, 891897
21 Miller, B. et al. (1999) A Synechococcus
leopoliensis SAUG 1402-1 operon harboring the
1-deoxy-D-xylulose 5-phosphate gene and two
additional open reading frames is functionally
involved in the dimethylallyl diphosphate
synthesis. FEBS Lett. 460, 485490
22 Lois, L.M. et al. (2000) Carotenoid biosynthesis
during tomato fruit development: regulatory role
of 1-deoxy-D-xylulose 5 phosphate synthase. Plant
J. 22, 503513
23 Takahashi, S. et al. (1998) A 1-deoxy-D-xylulose
5-phosphate reductoisomerase catalysing the
formation of 2-C-methyl-D-erythritol 4-phosphate
in an alternative nonmevalonate pathway for
terpenoid biosynthesis. Proc. Natl. Acad. Sci.
U. S. A. 95, 98799884
24 Kuzuyama, T. et al. (2000) Characterization of 1-
deoxy-D-xylulose 5-phosphate reductoisomerase, an
enzyme involved in isopentenyl diphosphate
biosynthesis, and identification of its catalytic amino
acid residues. J. Biol. Chem. 275, 1992819932
25 Proteau, P.J. et al. (1999) Stereochemistry of the
reduction step mediated by recombinant 1-deoxy-
D-xylulose 5-phosphate isomeroreductase. Org.
Lett. 1, 921923
26 Schwender, J. et al. (1999) Cloning and
heterologous expression of a cDNA encoding 1-
deoxy-D-xylulose-5-phosphate reductoisomerase
of Arabidopsis thaliana. FEBS Lett. 455, 140144
27 Lange, B.M. and 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, 170174
28 Jomaa, H. et al. (1999) Inhibitors of the non-
mevalonate pathway of isoprenoid biosynthesis as
antimalarial drugs. Science 285, 15731576
29 Radykewicz, T. et al. (2000) Biosynthesis of
terpenoids: 1-deoxy-D-xylulose-5-phosphate
reductoisomerase from Escherichia coli is a class
B dehydrogenase. FEBS Lett. 465, 157160
30 Kuzuyama, T. et al. (1998) Fosmidomycin, a
specific inhibitor of 1-deoxy-D-xylulose
5-phosphate reductoisomerase in the
nonmevalonate pathway for terpenoid
biosynthesis. Tetrahedron Lett. 39, 79137916
31 Rohdich, F. et al. (1999) Cytidine 5-triphosphate
biosynthesis of isoprenoids: YgbP protein of
Escherichia coli catalyses the formation of
4-diphosphocytidyl-2C-methylerythritol. Proc.
Natl. Acad. Sci. U. S. A. 96, 1175811763
32 Follens, A. et al. (1999) acs1 of Haemophilus
influenzae type a capsulation locus region II
encodes a bifunctional ribulose 5-phosphate
reductase CDPribitol pyrophosphorylase.
J. Bacteriol. 181, 20012007
33 Kuzuyama, T. et al. (2000) Formation of
4-(cytidine 5-diphospho)-2-C-methyl-D-erythritol
from 2-C-methyl-D-erythritol 4-phosphate by
2-C-methyl-D-erythritol 4-phosphate
cytidyltransferase, a new enzyme in the
nonmevalonate pathway. Tetrahedron Lett. 41,
703706
34 Rohdich, F. et al. (2000) Biosynthesis of
terpenoids: 4-diphoshocytidyl-2C-methyl-D-
erythritol synthase of Arabidopsis thaliana. Proc.
http://plants.trends.com
TRENDS in Plant Science Vol.6 No.2 February 2001
Natl. Acad. Sci. U. S. A. 97, 64516456 42 Turner, D. et al. (1992) Structure determination of
35 Lttgen, H. et al. (2000) Biosynthesis of a novel cyclic phosphocompound isolated from
terpenoids: YchB protein of Escherichia coli Desulfovibrio desulfuricans. Biochem. J.
phosphorylates the 2-hydroxy group of 285, 387390
4-diphosphocytidyl-2C-methyl-D-erythritol. 43 Ostrovsky, D. et al. (1992) The ability of bacteria
Proc. Natl. Acad. Sci. U. S. A. 97, 10621067 to synthesize a new cyclodiphosphate correlates
36 Lawrence, D.L. et al. (1997) Chromoplast with their tolerance to redox-cycling drugs: on a
development in ripening tomato fruit: crossroad of chemotherapy, environmental
identification of cDNAs for chromoplast-targeted toxicology and immunobiochemical problems.
proteins and characterization of a cDNA encoding Biofactors 4, 6368
a plastid-localized low-molecular-weight heat 44 Ostrovsky, D. et al. (1998) Effect of oxidative
shock protein. Plant Mol. Biol. 33, 483492 stress on the biosynthesis of 2-C-methyl-D-
37 Rohdich, F. et al. (2000) Biosynthesis of erythritol-2,4-cyclopyrophosphate and
terpenoids: 4-diphosphocytidyl-2-C-methyl-D- isoprenoids by several bacterial strains. Arch.
erythritol kinase from tomato. Proc. Natl. Acad. Microbiol. 171, 6972
Sci. U. S. A. 97, 82518256 45 Giner, J-L. et al. (1998) Biosynthesis of
38 Lange, M. and Croteau, R. (1999) Isopentenyl isoprenoids in Escherichia coli: the fate of the 3-H
diphosphate biosynthesis via a mevalonate- and 4-H atoms of 1-deoxy-D-xylulose. J. Chem.
independent pathway: isopentenyl Soc., Chem. Commun. 18571858
monophosphate kinase catalyses the terminal 46 Rodriguez-Concepcion, M. et al. (2000) Genetic
enzymatic step. Proc. Natl. Acad. Sci. U. S. A. evidence of branching in the isoprenoid pathway
96, 1371413719 for the production of isopentenyl diphosphate and
39 Kuzuyama, T. et al. (2000) Studies on the dimethylallyl diphosphate in Escherichia coli.
nonmevalonate pathway: conversion of FEBS Lett. 473, 328332
4-(cytidine 5-diphospho)-2-C-methyl-D-erythritol 47 Cunningham, F.X., Jr et al. (2000) Evidence of a
to its 2-phospho derivative by 4-(cytidine role for LytB in the nonmevalonate pathway of
5-diphospho)-2-C-methyl-D-erythritol kinase. isoprenoid biosynthesis. J. Bacteriol. 182,
Tetrahedron Lett. 41, 29252928 58415848
40 Herz, S. et al. (2000) Biosynthesis of terpenoids: 48 Fellermeier, M. et al. (1999) Cell-free conversion of
YgbB protein converts 4-diphosphocytidyl-2C- 1-deoxy-D-xylulose 5-phosphate and 2-C-methyl-
methyl-D-erythritol 2-phosphate to 2C-methyl-D- D-erythritol 4-phosphate into -carotene in higher
erythritol 2,4-cyclodiphosphate. Proc. Natl. Acad. plants and its inhibition by fosmidomycin.
Sci. U. S. A. 97, 24862490 Tetrahedron Lett. 40, 27432746
41 Takagi, M. et al. (2000) Studies on the 49 Matthews, P.D. and Wurtzel, E.T. (2000) Metabolic
nonmevalonate pathway: formation of 2-C- engineering of carotenoid accumulation in E. coli
methyl-D-erythritol 2,4-cyclodiphosphate from by modulation of the isoprenoid precursor pool
2-phospho-4-(cytidine 5-diphospho)-2-C-methyl- with expression of deoxyxylulose phosphate
D-erythritol. Tetrahedron Lett. 41, 33953398 synthase. Appl. Microbiol. Biotechnol. 53, 396400
Trends in Plant Science online
Making the most of your personal subscription
High quality printouts (from PDF files)
Links to other articles, other journals and cited software and databases
All you have to do is:
Obtain your subscription key from the address label of your print subscription
Then go to http://www.trends.com/free_access.html
Click on the large Click Here button at the bottom of the page
You will see one of the following:
(1) A BioMedNet login screen. If you see this, please enter your
BioMedNet username and password. If you are not already a
member please click on the Join Now button and register.
Once registered you will go straight to (2) below.
(2) A box to enter a subscription key. Please enter your
subscription key here and click on the Enter button.
Once confirmed, go to http//plants.trends.com and view the full text
of Trends in Plant Science
If you get an error message please contact Customer Services (info@current-
trends.com) stating your subscription key and BioMedNet username and password.
Please note that you do not need to re-enter your subscription key for Trends in Plant
Science, BioMedNet remembers your subscription. Institutional online access is
available at a premium. If your institute is interested in subscribing to print and online
please ask them to contact ct.subs@rbi.co.uk
Chapter 24

Lipid Biosynthesis

........................
Chapter Outline
Fatty acid biosynthesis
Biosynthesis localized in cytosol: Fatty acid degradation in mitochondria
Intermediates held on acyl carrier protein (ACP): Phosphopantetheine group attached to
serine: CoA in degradation
Fatty acid synthase: Multienzyme complex
Carbons derived from acetyl units
Acetyl CoA to malonyl CoA by carboxylation
Acetyl unit added to fatty acid with decarboxylation of malonyl CoA
Carbonyl carbons of acetyl units reduced using NADPH
Source of acetyl units
Amino acids, glucose
Acetyl CoA used to produce citrate
Citrate exported to cytosol: ATP-citrate lyase forms acetyl-CoA and oxaloacetate
Source of NADPH
Oxaloacetate utilization
Oxaloacetate (from citrate) to malate: NADH dependent reaction
Malate to pyruvate: Malic enzyme: NADPH produced
Pentose phosphate pathway
Malonyl-CoA production: Acetyl-CoA carboxylase
Biotin-dependent enzyme
ATP drives carboxylation
Enzyme regulation
Filamentous polymeric form active
Citrate favors active polymer
Palmitoyl-CoA favors inactive protomer (polymers monomer or
building block molecule)
Citrate/palmitoyl-CoA effects depend on state of phosphorylation of
protein
o Unphosphorylated protein binds citrate with high affinity:
Activation
o Phosphorylated protein binds palmitoyl with high affinity:
Inactivation
Acetyl transacetylase: Acetylates acyl carrier protein (ACP): Destined to become methyl end of
fatty acid
Malonyl transacetylase: Malonylates ACP
-Ketoacyl-ACP synthase (acyl-malonyl ACP condensing enzyme): Accepts acetyl group:
Transfers acyl group to malonyl-ACP
Malonyl carboxyl group released: Decarboxylation drives synthesis
Malonyl-ACP converted to acetoacetyl-ACP
-Ketoacyl-ACP reductase
Carbonyl carbon reduced to alcohol
Chapter 24 Lipid Biosynthesis

NADPH provides electrons

-Hydroxyacyl-ACP dehydratase: Elements of water removed: Double bond created
2,3-trans-Enoyl-ACP reductase
Double bond reduced
NADPH provides electrons
Subsequent cycles: C-16: Palmitoyl-CoA
Additional modifications
Elongation
Mitochondrial-based system uses reversal of -oxidation
Endoplasmic reticulum-based system uses malonyl-CoA
Monounsaturation: One double bond
Bacteria: Oxygen-independent pathway: Chemistry performed on carbonyl
carbon
Eukaryotes: Oxygen-dependent pathway
Polyunsaturation
Plants can add double bonds between C-9 and methyl end
Animals
Add double bonds between C-9 and carboxyl end
Require essential fatty acids to have double bonds closer to methyl end
Regulation
Malonyl-CoA inhibition of carnitine-acyl transferase: Blocks fatty acid uptake
Citrate/palmitoyl regulation of acetyl-CoA carboxylase
Complex lipids
Glycerolipids: Glycerol backbone
Glycerophospholipids
Triacylglycerols
Sphingolipids: Sphingosine backbone
Phospholipids
Sphingolipids
Glycerophospholipids
Glycerolipid biosynthesis
Phosphatidic acid is precursor
Glycerokinase produces glycerol-3-P
Glycerol-3-phosphate acyltransferase acylates C-1 with saturated fatty acid:
Monoacylglycerol phosphate
Eukaryotes can produce monoacylglycerol phosphate using DHAP
Acyldihydroxyacetone phosphate reduced by NAPDH to monoacylglycerol
phosphate
Acyltransferase acylates C-2: Phosphatidic acid
Phosphatidic acid used to synthesize two precursors of complex lipids
Diacylglycerol: Precursor of
Triacylglycerol: Diacylglycerol acyltransferase
Phosphatidylethanolamine, phosphatidylcholine
o Ethanolamine phosphorylated
o CTP and phosphoethanolamine produce CDP-ethanolamine
o Transferase moves phosphoethanolamine onto diacylglycerol
o (Dietary choline: As per ethanolamine)
o (Phosphatidylethanolamine to phosphatidylcholine by
methylation)
Phosphatidylserine: Serine for ethanolamine exchange
CDP-diacylglycerol
Phosphatidate cytidylyltransferase produces CDP-diacylglycerol
CDP-diacylglycerol used to produce
o Phosphaphatidyl inositol
o Phosphaphatidyl glycerol
o Cardiolipin
Plasmalogens: ,-unsaturated ether-linked chain at C-1

385
Chapter 24 Lipid Biosynthesis

DHAP acetylated
Acyl group exchanged for alcohol
Keto group on DHAP reduced to alcohol and acylated
Head group attached
Desaturase produces double bonds
Sphingolipid biosynthesis
Serine and palmitoyl-CoA condensed with decarboxylation to produce 3-
ketosphinganine
Reduction forms sphingamine
Sphingamine acylated to form N-acyl sphingamine
Desaturase produces ceramide
Cerebrosides: Galactose or glucose added
Gangliosides: Sugar polymers: Sugars derive from UDP-monosaccharides
Eicosanoids: Derived from 20-C fatty acids: Arachidonate is precursor
Local hormones: Prostaglandins, thromboxanes, leukotrienes, hydroxyeicosanoic acids
Prostaglandins
Cyclopentanoic acid formed from arachidonate by prostaglandin
endoperoxidase synthase
Aspirin inhibits enzyme
Cholesterol
Membrane component
Precursor of important biomolecules
Bile salts
Steroid hormones
Vitamin D
Cholesterol biosynthesis: In liver
Mevalonate biosynthesis
Thiolase condenses two acetyl-CoA to produce acetoacetyl-CoA
HMG-CoA synthase produces HMG-CoA
HMG-CoA reductase produces mevalonate
Rate limiting step
Regulation
o Inactivated by cAMP-dependent protein kinase
o Short half life of enzyme when cholesterol levels high
o Gene expression regulated
Pharmacological target for blood cholesterol regulation
Isopentenyl pyrophosphate and dimethylallyl pyrophosphate from mevalonate
Squalene to lanosterol to cholesterol
Lipid transport
Fatty acids complexed to serum albumin
Phospholipids, triacylglycerol, cholesterol transported as lipoprotein complexes
Lipoprotein complex types: HDL, LDL, IDL, VLDL, Chylomicrons
Chylomicrons formed in intestine
HDL, VLDL assembled in liver
o Core of triacylglycerol
o Single layer of phospholipid
o Proteins and cholesterol inserted
o VLDL to IDL to LDL to liver for uptake and degradation
o HDL: Assembled without cholesterol but picks up cholesterol
during circulation
Bile salts
Glycocholic acid
Taurocholic acid
Steroid hormones
Cholesterol to pregnenolone
Pregnenolone to progesterone
Hormone

386
Chapter 24 Lipid Biosynthesis

Sex hormone precursor

Androgens
Estrogens
Corticosteroids precursor
Glucocorticoids
Mineralocorticoids
Chapter Objectives
Fatty Acid Biosynthesis
The steps of fatty acid biosynthesis (Figure 24.7) are similar in chemistry to the reverse of -
oxidation. Two-carbon acetyl units are used to build a fatty acid chain. The carbonyl carbon is
reduced to a methylene carbon in three steps: reduction to an alcoholic carbon, dehydration to a
carbon-carbon double bond intermediate, and reduction of the double bond. The two reduction
steps utilize NADPH as reductant. Two-carbon acetyl units are moved out of the mitochondria as
citrate and activated by carboxylation to malonyl-CoA. We have already seen similar
carboxylation reactions and should remember that biotin is involved when carbons are added at
the oxidation level of a carboxyl group. The enzyme, acetyl-CoA carboxylase, is regulated by
polymerization/depolymerization with the filamentous polymeric state being active. You should
understand the regulatory effects of citrate (favors polymer formation), palmitoyl-CoA
(depolymerizes) and covalent phosphorylation (blocks citrate binding) on acetyl-CoA carboxylase
activity.
In -oxidation, we saw that the phosphopantetheine group of coenzyme A functioned as a
molecular chauffeur for two-carbon acetyl units. In synthesis, phosphopantetheine, attached to
the acyl carrier protein, functions as a molecular chaperone by guiding the growth of fatty acid
chains.
In plants and bacteria, the steps of fatty acid biosynthesis are catalyzed by individual proteins
whereas in animals a large multifunctional protein is involved. Synthesis starts with formation of
acetyl-ACP and malonyl-ACP by specific transferases. The carboxyl group of malonyl-ACP
departs, leaving a carbanion that attacks the acetyl group of acetyl-ACP to produce a four-carbon
-ketoacyl intermediate, which is subsequently reduced by an NADPH-dependent reductase,
dehydrated, and reduced a second time by another NADPH-dependent reductase. To continue
the cycle, malonyl-ACP is reformed, decarboxylates, and attacks the acyl-ACP. The original acetyl
group is the methyl-end of the fatty acid, whereas the malonyl groups are added at the carboxyl
end. NADPH is supplied by the pentose phosphate pathway and by malic enzyme, which
converts the oxaloacetate skeleton, used to transport acetyl groups out of the mitochondria as
citrate, into pyruvate and CO2 with NADP+ reduction.
Additional elongation and introduction of double bonds can occur after synthesis of a C16 fatty
acid. Elongation can occur in the endoplasmic reticulum, where malonyl CoA is utilized, or in
the mitochondria where acetyl-CoA is used. Introduction of double bonds occurs via oxygen-
independent mechanisms in bacteria and oxygen-dependent mechanisms in eukaryotes. Be
familiar with the reaction catalyzed by stearoyl-CoA desaturase, involving stearoyl-CoA and
oxygen as substrates and oleoyl-CoA and water as products.

Complex Lipids
The glycerolipids, including glycerophospholipids and triacylglycerols, are synthesized from
glycerol, fatty acids, and head groups. Synthesis starts with the formation of phosphatidic acid
from glycerol-3-phosphate and fatty acyl-CoA. C-1 is esterified usually with a saturated fatty
acid. Phosphatidic acid may be converted to diacylglycerol and then to triacylglycerol.
Alternately, diacylglycerol can be used to synthesize phosphatidylethanolamine and
phosphatidylcholine with CDP-derivatized head groups serving as substrates. Phosphatidylserine
is produced by exchange of the ethanol head-group from PE with serine. Phosphatidylinositol,
phosphatidylglycerol, and cardiolipin (two diacylglycerols linked together by glycerol) are
synthesized using CDP-diacylglycerol as an intermediate. Plasmalogens are synthesized from
acylated DHAP. The acyl group is exchanged for a long-chain alcohol followed by reduction of the
keto carbon of DHAP, acyl group transfer from acyl-CoA to C-2, head group transfer from CDP-
ethanolamine and formation of a cis double bond between C-1 and C-2 of the long-chain alcohol.
The sphingolipids all derive from ceramide, whose synthesis starts with bond formation
between palmitic acid and the -carbon of serine (with loss of the serine carboxyl carbon as
bicarbonate). After a few steps a second fatty acid is attached to serine in amide linkage.
Subsequent sugar additions lead to cerebrosides and gangliosides.

Prostaglandins

387
Chapter 24 Lipid Biosynthesis

The prostaglandins are produced from arachidonic acid released by phospholipase A2 action
on phospholipids. Production of these local hormones is blocked by aspirin, and nonsteroid anti-
inflammatory agents such as ibuprofen and phenylbutazone.

Cholesterol
Cholesterol derives from HMG-CoA, a product we already encountered in ketone body
formation. You might recall that ketone bodies are produced from acetyl-CoA units. HMG-CoA is
a six-carbon CoA derivative produced from three acetyl units. The rate-limiting step in
cholesterol synthesis is formation of 3R-mevalonate from HMG-CoA by HMG-CoA reductase,
which catalyzes two NADPH-dependent reductions. This enzyme is carefully regulated by 1)
phosphorylation leading to inactivation, 2) degradation, and 3) gene expression. Mevalonate, a
six-carbon intermediate, is converted to isopentenyl pyrophosphate, which is used to synthesize
cholesterol. Cholesterol is the precursor of bile salts and the steroid hormones. You should
understand how lipoproteins are responsible for movement of cholesterol and other lipids in the
body.

388
Chapter 24 Lipid Biosynthesis

Figure 24.7 The pathway of palmitate synthesis from acetyl-CoA and malonyl-CoA. Acetyl and malonyl
building blocks are introduced as acyl carrier protein conjugates. Decarboxylation drives the -ketoacyl-ACP
synthase and results in the addition of two-carbon units to the growing chain. Concentrations of free fatty
acids are extremely low in most cells, and newly synthesized fatty acids exist primarily as acyl-CoA esters.

389
Chapter 24 Lipid Biosynthesis

Problems and Solutions

1. Carefully count and account for each of the atoms and charges in the equations for
the synthesis of palmitoyl-CoA, the synthesis of malonyl-CoA, and the overall reaction for
the synthesis of palmitoyl-CoA from acetyl-CoA.

Answer: Malonyl-CoA is synthesized as follows

Acetyl-CoA + HCO3- + ATP4- malonyl-CoA- + ADP3- + Pi2- + H+
The carbons in the acetyl group of acetyl-CoA derive from glucose via glycolysis or from the side
chains of various amino acids. The bicarbonate anion is produced from CO2 and H2O by carbonic
anhydrase CO2 + H2O H2CO3 H+ + HCO3-. Generation of ADP and Pi from ATP is a hydrolysis
reaction; however, water does not show up in the equation because incorporation of bicarbonate
carbon into malonyl-CoA is accompanied by release of water.

Synthesis of palmitoyl-CoA is described as follows

Acetyl-CoA + 7 malonyl-CoA- + 14NADPH + 7 H+ + 7 ATP4-
palmitoyl-CoA + 7 HCO3- + 14 NADP+ + 7 ADP3- + 7 Pi2- + 7 CoASH
For bicarbonate to show up on the right hand side of the equation, the carbon dioxide released by
reacting malonyl-CoA and acetyl-CoA must be hydrated and subsequently ionized. So, each
bicarbonate is accompanied by production of protons. This is the reason why only half as many
protons as NADPH are found in the reaction. Carbons 15 and 16 derive from acetyl-CoA directly;
the remaining carbons in palmitoyl-CoA derive from acetyl-CoA by way of malonyl-CoA.

2. Use the relationships shown in Figure 24.1 to determine which carbons of glucose will
be incorporated into palmitic acid. Consider the cases of both citrate that is immediately
exported to the cytosol following its synthesis and citrate that enters the TCA cycle.

Answer: The six carbons of glucose are converted into two molecules each of CO2 and acetyl
units of acetyl-coenzyme A. Carbons 1, 2, and 3 of glyceraldehyde derive from carbons 3 and 4, 2
and 5, and 1 and 6 of glucose respectively. Carbon 1 of glyceraldehyde is lost as CO2 in
conversion to acetyl-CoA, so we expect no label in palmitic acid from glucose labeled only at
carbons 3 and 4. The carbonyl carbon and the methyl carbon of the acetyl group of acetyl-CoA
derive from carbons 2 and 5, and carbons 1 and 6 of glucose, respectively. The methyl carbon is
incorporated into palmitoyl-CoA at every even-numbered carbon, whereas the carbonyl carbon is
incorporated at every odd-numbered carbon.
Acetyl-CoA is produced in the mitochondria and exported to the cytosol for fatty acid
biosynthesis by being converted to citrate. The cytosolic enzyme, citrate lyase, converts citrate to
acetyl-CoA and oxaloacetate. When newly synthesized citrate is immediately exported to the
cytosol, the labeling pattern described above will result. However, where citrate is instead
metabolized in the citric acid cycle, back to oxaloacetate, label derived from acetyl-CoA shows up
at carbons 1, 2, 3 and 4 of oxaloacetate. These carbons do not get incorporated into palmitoyl-
CoA.

3. Based on the information presented in the text and in Figures 24.4 and 24.5, suggest
a model for the regulation of acetyl-CoA carboxylase. Consider the possible roles of
subunit interaction, phosphorylation, and conformation changes in your model.

Answer: Acetyl-CoA carboxylase catalyzes the formation of malonyl-CoA, the committed step in
synthesis of fatty acids. This enzyme is a polymeric protein composed of protomers, or subunits,
of 230 kD. In the polymeric form, the enzyme is active whereas in the protomeric form the
enzyme is inactive. Polymerization is regulated by citrate and palmitoyl-CoA such that citrate, a
metabolic signal for excess acetyl units, favors the polymeric and, therefore, active form of the
enzyme whereas palmitoyl-CoA shifts the equilibrium to the inactive form. The activity of acetyl-
CoA carboxylase is also under hormonal regulation. Glucagon and epinephrine stimulate cyclic
AMP-dependent protein kinase that will phosphorylate a large number of sites on the enzyme.
The phosphorylated form of the enzyme binds citrate poorly and citrate binding occurs only at
high citrate levels. Citrate is a tricarboxylic acid with three negative charges and its binding site
on the enzyme is likely to be composed of positively-charged residues. Phosphorylation
introduces negative charges, which may be responsible for the decrease in citrate binding.
In the phosphorylated form, low levels of palmitoyl-CoA will inhibit the enzyme. Thus, the
enzyme is sensitive to palmitoyl-CoA binding and to depolymerization in the phosphorylated form.
If we assume that the palmitoyl-CoA binding site is located at a subunit-subunit interface, and

390
Chapter 24 Lipid Biosynthesis

that phosphorylated, and hence negatively charged subunits interact with lower affinity than do
unphosphorylated subunits, we see that it is easier for palmitoyl-CoA to bind to the enzyme.

4. Consider the role of the pantothenic acid groups in animal fatty acyl synthase and the
size of the pantothenic acid group itself, and estimate a maximal separation between the
malonyl transferase and the ketoacyl-ACP synthase active sites.

Answer: In fatty acyl synthase, pantothenic acid is attached to a serine residue as shown below.
O O CH3 O
H
HS CH2 C CH2 C C P serine
CH2 N CH2 N C O OH
CH3 OH
H H OH
The approximate distance from the pantothenic group to the -carbon of serine is calculated as
follows. For carbon-carbon single bonds the bond length is approximately 0.15 nm. The distance
between carbon atoms is calculated as follows.
109o
C
0.15 nm
35.5 o 35.5 o
C
d
d = 0.15nm cos(35.5 ) = 0.22nm
Let us use this length for carbon-carbon single bonds, carbon-oxygen bonds, oxygen-
phosphorous bonds, and carbon-nitrogen bonds exclusive of the amide bond. For the amide
bond we will use a distance of 0.132 nm. The overall length is approximately 1.85 nm from the -
carbon of serine to the sulfur. The maximal separation between malonyl transferase and
ketoacyl-ACP is about twice this distance or approximately 3.7 nm. The actual distance between
these sites is smaller than this upper limit.

5. Carefully study the reaction mechanism for the stearoyl-CoA desaturase in Figure
24.14, and account for all of the electrons flowing through the reactions shown. Also
account for all of the hydrogen and oxygen atoms involved in this reaction, and convince
yourself that the stoichiometry is correct as shown.

Answer: Stearoyl-CoA desaturase catalyzes the following reaction

Stearoyl-CoA + NADH + H+ + O2 oleoyl-CoA + 2 H2O
This reaction involves a four-electron reduction of molecular oxygen to produce two water
molecules. Two of the electrons come from the desaturation reaction directly, in which
desaturase removes two electrons and two protons from stearoyl-CoA to produce the carbon-
carbon double bond in oleoyl-CoA. The other two electrons and protons derive from NADH + H+.
Two electrons from NADH are used by another enzyme, NADH-cytochrome b5 reductase, to
reduce FAD to FADH2. Electrons are then passed one at a time to cytochrome b5, which passes
electrons to the desaturase to reduce oxygen to water. So, two electrons and two protons come
from palmitoyl-CoA and two electrons come from NADH with two protons being supplied by the
surrounding solution.

6. Write a balanced, stoichiometric reaction for the synthesis of phosphatidyl-

ethanolamine from glycerol, fatty acyl-CoA, and ethanolamine. Make an estimate of the
G' for the overall process.

Answer: The synthesis of phosphatidylethanolamine involves the convergence of two separate

pathways: A diacylglycerol backbone is synthesized from glycerol and fatty acids; ethanolamine is
phosphorylated and activated by transfer to CTP to produce CDP-ethanolamine. CDP-
ethanolamine: 1,2-diacylglycerol phosophoethanolamine transferase then catalyzes the formation
of phosphatidylethanolamine from diacyl-glycerol and CDP-ethanolamine.
Starting from glycerol, production of diacylglycerol involves the following reactions:
Glycerol + ATP4- glycerol-3-phosphate + ADP3- + H+
Glycerol-3-phosphate + fatty acyl-CoA lysophosphatidic acid + CoA-SH

391
Chapter 24 Lipid Biosynthesis

Lysophosphatidic acid + fatty acyl-CoA phosphatidic acid + CoA-SH

Phosphatidic acid + H2O diacylglycerol + Pi2-
We have:
Glycerol + 2 fatty acyl-CoA + H2O + ATP4- diacylglycerol + 2 CoA-SH+ ADP3- + Pi2-+H+
Production of CDP-ethanolamine involves the following:
Ethanolamine + ATP4- phosphoethanolamine + ADP3- + H+
Phosphoethanolamine + CTP4- CDP-ethanolamine + PPi 4-
PPi 4- + H2O 2 Pi2-
Or,
Ethanolamine + ATP4- + CTP4-+ H2O CDP-ethanolamine + ADP3- + 2 Pi2- + H+
Finally, for the reaction catalyzed by CDP-ethanolamine: 1,2-diacylglycerol
phosophoethanolamine transferase we have:
Diacylglycerol + CDP-ethanolamine phosphatidylethanolamine + CMP2- + H+
The balanced, stoichiometric reaction is:
Glycerol + ethanolamine + 2 fatty acyl-CoA + 2 ATP4- + CTP4- + 2 H2O
Phosphatidylethanolamine + 2 CoA-SH + 2 ADP3- + 2 H+ + 3 Pi2- + CMP2-

7. Write a balanced, stoichiometric reaction for the synthesis of cholesterol from acetyl-
CoA.

Answer: The immediate precursors of cholesterol are isopentenyl pyrophosphate and

dimethylallyl pyrophosphate, both of which derive from hydroxymethylglutaryl-CoA (HMG-CoA).
HMG-CoA is synthesized from acetyl-CoA by the following route:
H 2O + O
O
H3C C S-CoA
H3C C S-CoA O O OH O
-
+ H3C C CH2 C S-CoA OOC CH2 C CH2 C S-CoA
O CH3
H3C C S-CoA CoA-SH CoA-SH
The reaction is:
3 Acetyl-CoA HMG-CoA + 2 CoA-SH

HMG-CoA is anabolized into isopentenyl pyrophosphate and dimethylallyl pyrophosphate, both of

which are used to synthesize squalene, which is converted by way of lanosterol into cholesterol.
(The next question asks us to trace carbons from mevalonate to cholesterol, so it is worthwhile
now to look at these reactions in detail).
Synthesis of isopentenyl pyrophosphate from HMG-CoA is as follows:

3 ATP 3 ADP +
2 NADPH 2 NADP+ + H2O Pi + CO2
OH O OH
-OOC CH2 C CH2 C S-CoA -
OOC CH2 C CH2 CH2 OH
CH3 CoA-SH CH3
HMG-CoA mevalonate

CH3 O O CH3 O O
CH2 C CH2 CH2 O P O P OH CH3 C CH CH2 O P O P OH
OH OH OH OH

isopentenyl pyrophosphate dimethylallyl pyrophosphate

Overall, the reaction is:
HMG-CoA + 2 NAPDH + 3 ATP isopentenyl pyrophosphate (or dimethylallyl
pyrophosphate) + CoA-SH + 2 NADP+ + 3 ADP + Pi + CO2

392
Chapter 24 Lipid Biosynthesis

Production of squalene using isopentenyl pyrophosphate and dimethylallyl pyrophosphate

proceeds as follows. Two farnesyl pyrophosphates are produced from two dimethylallyl
pyrophosphates and four isopentenyl pyrophosphates. The farnesyl pyrophosphates are reacted
to produce squalene as follows:
2 farnesyl pyrophosphates

O- O-

O- P O P O
O O
+ O O

O P O P O-
O- O-

NADPH

NADP+
+ 2 PPi

squalene

Squalene is converted to lanosterol in two steps catalyzed by squalene epoxidase and squalene
oxidocyclase. Squalene + 0.5 O2 + NADPH lanosterol
The overall equation for acetyl-CoA to lanosterol is:
18 Acetyl-CoA + 13 NADPH +13 H+ + 18 ATP + 0.5 O2
Lanosterol + 18 CoA-SH + 13 NADP+ + 18 ADP + 6 Pi + 6 PPi + 6 CO2
The pathway from lanosterol to cholesterol involves the oxidation and loss of three carbons.

8. Trace each of the carbon atoms of mevalonate through the synthesis of cholesterol,
and determine the source (i.e., the position in the mevalonate structure) of each carbon in
the final structure.

Answer:

393
Chapter 24 Lipid Biosynthesis

mevalonate
OH
-
OOC 4C 3C 2C 1C OH

4C

4C O O 4C O O

4C 3C 2C 1C O P O P O- 4C 3C 2C 1C O P O P O-

O- O- O- O-

isopentenyl pyrophosphate dimethylallyl pyrophosphate

2C 4C
1C 3C
2 farnesyl pyrophosphates
2C 4C 4C
O -
O - 1C 3C

O- P O P O 2C 4C 4C
1C 3C
4C O + O
O O 4C
3C 1C
4C 4C 2C O P O P O-

3C 1C O- O-
4C 4C 2C
NADPH
3C 1C
4C 2C

farnesyl pyrophosphate 2C 4C
1C 3C
+
NADP
2C 4C 4C
+ 2 PPi
1C 3C
4C

2C 4C 4C
3C 1C
2C 1C 3C
4C 4C

4C
3C 1C squalene
4C 4C 2C

3C 1C
4C 2C

394
Chapter 24 Lipid Biosynthesis

2C 4C
1C 3C

2C 4C 4C
1C 3C
4C

2C 4C 4C
3C 1C
2C 1C 3C
4C 4C

4C
3C 1C squalene
4C 4C 2C

3C 1C
4C 2C 4C
4C
3C
4C
2C 4C
3C
1C

2C

1C 1C
1C 2C

3C 4C
4C 2C
4C
1C 3C 4C
3C
4C
4C
2C 2C
1C
3C
4C 4C 4C 2C 4C
4C
3C 1C 3C
4C
1C 4C
2C
1C 2C
4C 1C
4C 2C 3C
4C
1C 3C 3C

2C 2C 4C
HO 3C 1C

cholesterol
9. Suggest a structural or functional role for the O-linked saccharide domain in the LDL
receptor (Figure 24.40).

Answer: LDL receptors are synthesized on the rough endoplasmic reticulum and move through
the smooth endoplasmic reticulum and Golgi apparatus before being incorporated into the
plasma membrane. On the plasma membrane, LDL receptors bind LDL, aggregate into patches,
and are internalized into coated vesicles that fuse with lysosomes where LDL is degraded. The O-
linked saccharide domain functions to extend the receptor domain away from the cell surface,
above the glycocalyx coat. This allows the receptor to bind circulating lipoproteins.

10. Identify the lipid synthetic pathways that would be affected by abnormally low levels
of CTP.

Answer: Phosphatidylethanolamine and phosphatidylcholine synthesis depend on the formation

of CDP-ethanolamine and CDP-choline respectively. Phosphatidyl-inositol and
phosphatidylglycerol biosynthesis utilize CDP-diacylglycerol. The synthetic pathways of all of
these compounds may be affected if the cell experiences low levels of CTP.

395
Chapter 24 Lipid Biosynthesis

11. Determine the number of ATP equivalents needed to form palmitic acid from acetyl-
CoA. (Assume for this calculation that each NADPH is worth 3.5 ATP.)

Answer: Palmitate is synthesized with the following stoichiometry:

8 Acetyl-CoA + 7 ATP + 14 NADPH palmitate + 7 ADP + 14 NADP+ + 8 CoA-SH + 6 H2O + 7 Pi
The 14 NADPHs are equivalent to (14 3.5 =) 49 ATPs. Combining these with the 7 ATPs used to
synthesize malonyl-CoA gives a total of 56 ATPs consumed.

12. Write a reasonable mechanism for the 3-ketosphinganine synthase reaction,

remembering that it is a pyridoxal phosphate-dependent reaction.

Answer:
O
H3C (CH2)14 C S CoA
Palmitoyl-CoA enzyme
enzyme HB
:B
H
HO CH2 C- COO-
HO CH2 C COO-
+
+ H NH
H NH
C C

OH OH
2-
2-
O3PO O3PO

N+ CH3 N+ CH3

Seryl-pyridoxal phosphate

H3C
H 3C
CoASH
(CH2)14
(CH2)14
C O
C O
HO CH2 C enzyme O-
HB HO CH2 C C
+
H NH
+ NH O
C- H
C
OH
2-
O3PO OH
2-
O3PO
N+ CH3
N+ CH3

396
Chapter 24 Lipid Biosynthesis

H 3C

H 3C (CH2)14

(CH2)14 C O

H 2O HO CH2 C H
C O
NH3+
HO CH2 C H
+
H NH H O
C C

OH OH
2- 2-O PO
O3PO 3

N+ CH3 N+ CH3

13. Why is the involvement of FAD important in the conversion of stearic acid to oleic
acid?

Answer: In eukaryotes, unsaturation reactions are catalyzed by stearoyl-CoA desaturase. This

enzyme functions along with cytochrome b5 reductase and cytochrome b5 to pass electrons, one
at a time to desaturase. Desaturase reduces O2, a 4-electron reduction for which two electrons
come from the fatty acid that is desaturated and two ultimately from NADH. The two electrons
from NADH pass through cytochrome b5 reductase, an FAD-containing enzyme, which must pass
electrons one at a time to cytochrome b5. NAD cannot participate in one-electron transfers
whereas FAD can. FADH2 can lose one electron or two.

14. Write a suitable mechanism for the HMG-CoA synthase reaction. What is the
chemistry that drives this condensation reaction.

Answer: The mechanism for HMG-CoA was already discussed in problem 14 of chapter 23. The
chemistry is not unlike that of citrate synthase. HMG-CoA synthase produces
hydroxymethylglutaryl-CoA from acetyl-CoA and acetoacetyl-CoA. The reaction is accompanied
by hydrolysis of a thioester bond linking coenzyme to the acetyl group. We learned in earlier
chapters that these thioester bonds are high energy. Production of fatty acyl-CoA by acyl-CoA
synthetase, which is used to activate fatty acids for beta oxidation, uses in effect two
phosphoanhydride bonds to create the thioester bond. (As an interesting aside, synthase and
synthetase both run reactions that join two substrates to form a product. Synthetases, in
general, drive these reaction with hydrolysis of high-energy phosphoanhydride bonds. Synthases
do not.)

15. Write a suitable reaction mechanism for the -ketoacyl ACP synthase, showing how
the involvement of malonyl-CoA drives this reaction forward.

Answer: -Ketoacyl ACP synthase links an acetyl group from malonyl-ACP onto an acyl-group
(acetyl- in the first round) during fatty acid synthesis. Malonyl-CoA is, in effect, an activated
acetyl group produced at the expense of hydrolysis of ATP by acetyl-CoA carboxylase. The
mechanism of action involves decarboxylation of malonyl-CoA to produce a carbanion on the beta
carbon, which attacks the carbonyl carbon of acyl-ACP producing -ketoacyl ACP. This
mechanism is shown below.

397
Chapter 24 Lipid Biosynthesis

-O O CO2 O O O

C CH2 C S ACP -CH2 C S ACP H3 C C CH 2 C S ACP

O
H3 C C S KSase

16. Consider the synthesis of linoleic acid from palmitic acid and identify a series of
three consecutive reactions that embody chemistry similar to three reactions in the
tricarboxylic acid cycle.

Answer: Palmitic acid is a saturated 16-carbon fatty acid whereas linoleic acid shown below- is
18 carbons long with two carbon-carbon double bonds. To convert palmitic acid to linoleic acid it
must first be elongated using one cycle of fatty acid synthesis. Elongation of palmitoyl-CoA
involves a thiolase reaction using acetyl-CoA, which adds to the carboxyl end. The -keto acyl
CoA derivative is then reduced to -hydroxy, dehydrated to form a carbon-carbon double bond
and then reduced to produce stearyl-CoA. The chemistry of this series of reactions is similar to
the chemistry found in the TCA cycle but going in reverse from oxaloacetate to succinate.

To convert stearyl-CoA to linoleilyl-CoA we would have to produce two carbon-carbon double

bonds by oxidation of the saturated fatty acid. While plants can produce this polyunsaturated
fatty acid, mammals cannot.
O

C
H3C OH

17. Rewrite the equation in Section 24.1 to describe the synthesis of behenic acid (see
Table 8.1).

Answer: Behenic acid (a.k.a., docosanoic acid) is a 22-carbon long fatty acid shown below. On
page 770 we are given the equation for synthesis of palmitoyl-CoA, a 16-carbon long fatty acid.
To produce behenic acid we need to recognize that we will need to run three more cycles of fatty
acid synthesis. Each cycle will consume one acetyl-CoA, 1 ATP and 2 NADPH.

H3C COOH
The equation is:
11-Acetyl-CoA + 10 ATP4- + 20 NADPH + 10 H+ J behenoyl-CoA + 20 NADP+ + 10 CoA-SH + 10
ADP3- + 10 Pi2-. (There are only 10 H+ consumed, and not 20 (thinking that each NADPH used is
actually used with a proton), because hydrolysis of ATP releases a proton. You can see this by
counting charge in the equation.)

Questions for Self Study

1. How are acetate units moved from the mitochondria to the cytosol? What other role does the
acetate carrier play in regulation of metabolism?

2. Although acetate units are incorporated into fatty acids during synthesis, they derive from
three-carbon compounds attached to coenzyme A. What is this three-carbon coenzyme A
derivative? What enzyme is responsible for its formation? How many high-energy phosphate
bonds are cleaved to drive its synthesis?

3. Describe how, in animals, the activity of acetyl-CoA carboxylase is regulated by citrate and
palmitoyl-CoA and how this regulation is sensitive to covalent modification of the enzyme.

398
Chapter 24 Lipid Biosynthesis

4. Match an enzyme with an activity.

a. Acetyltransferase 1. Keto carbon converted to alcoholic carbon.
b. Dehydrase 2. Attaches malonyl group to fatty acid synthase.
c. Malonyltransferase 3. Attaches acetyl group to acyl carrier protein.
d. Enoyl reductase 4. Carbon-carbon double bond reduced.
e. -Ketoacyl reductase 5. Condensation of acetyl group and malonyl group.
f. -Ketoacyl synthase 6. Enoyl intermediate formed.

5. From the following list of compounds identify those that are cholesterol derivatives and
appropriately identify each cholesterol derivative as a hormone (H), bile salt (B), or vitamin (V).
a. Prostaglandin D2
b. Glycocholic acid.
c. Squalene
d. Testosterone
e. Arachidonic Acid.
f. Cortisol
g. Cholecalciferol
h. Progesterone
i. Thromboxanes
j. Taurocholic acid
k. Aldosterone

6. What is the rate-limiting step in cholesterol biosynthesis?

7. Match a lipoprotein complex with a function.

a. Chylomicrons 1. Formed from very low density lipoproteins.
b. Very low density lipoproteins 2. Formed in the intestine.
c. Low-density lipoproteins 3. Slowly accumulate cholesterol.
d. High-density lipoproteins 4. Carry lipid from the liver.

8. In eukaryotes, glycerolipids are all derived from phosphatidic acid. Draw the structure of
phosphatidic acid and outline its biosynthesis from dihydroxyacetone-phosphate and from
glycerol-3-phosphate.

9. What is the role of cytidine in lipid biosynthesis?

10. Fill in the blanks. The prostaglandins are that function locally and at very low
concentrations. They are synthesized from , a 20-carbon polyunsaturated fatty acid.
Mammals can produce this fatty acid from (18:29,12) but must acquire this polyunsaturated
fatty acid from their diet.

Answers
1. Acetate units on acetyl-CoA are used to produce citrate in the mitochondria. Citrate is
exported to the cytosol where it is converted to acetyl-CoA and oxaloacetate. Citrate inhibits
phosphofructokinase and thus serves as a regulator of glycolysis. It also stimulates fatty acid
synthesis.

2. Malonyl-CoA is produced by acetyl-CoA carboxylase at the expense of one high-energy

phosphate bond.

3. Acetyl-CoA carboxylase is active in a polymeric state. The equilibrium between active polymer
and inactive protomers is affected by citrate and palmitoyl-CoA. Citrate is an allosteric activator
of the enzyme and shifts the equilibrium to the polymer. Palmitoyl-CoA shifts the equilibrium to
the inactive, protomeric state. The enzyme is phosphorylated by a number of kinases and the
phosphorylated state has a low affinity for citrate and a high affinity for palmitate.

4. a. 3; b. 6; c. 2; d. 4; e. 1; f. 5.

5. b. B; d. H; f. H; g. V; h. H; j. B; k. H.

6. The production 3R-mevalonate from HMG-CoA catalyzed by HMG-CoA reductase.

399
Chapter 24 Lipid Biosynthesis

7. a. 2; b. 4; c. 1; d. 3.

8. Dihydroxyacetone phosphate is converted to 1-acyldihydroxyacetone-phosphate by an

acyltransferase reaction and reduced to 1-acylglycerol-3-phosphate by a reductase. This
compound can also be synthesized from glycerol-3-phosphate by acyltransferase. Transfer of a
second acyl group to C-2 produces phosphatidic acid whose structure is shown below.
O
R1 C O CH2

O
R2 C O C H O
CH2 O C O-
O-
9. The head groups of phosphatidylethanolamine and phosphatidylcholine derive from cytidine
diphosphate derivatives. CDP-diacylglycerol is a precursor of phosphatidylinositol,
phosphatidylglycerol, and cardiolipin.

10. Hormones; arachidonic acid; linoleic acid.

Additional Problems
1. What are the sources of carbons for fatty acid biosynthesis? What is the role of the citrate-
malate-pyruvate shuttle in making carbon compounds available for fatty acid biosynthesis?

2. Movement of citrate out of the mitochondria coordinates glycolysis and fatty acid biosynthesis.
Explain.

3. Name the three water soluble vitamins that are crucial to fatty acid synthesis and briefly
describe the roles they play in this process.

4. Why do mammals require certain essential fatty acids in their diet?

5. Outline the synthesis of glycerophospholipid.

6. Lovastatin lowers serum cholesterol by interfering with HMG-CoA reductase, the enzyme that
catalyzes the rate limiting step in cholesterol synthesis. The drug is administered as an inactive
lactone that is activated by hydrolysis to mevinolinic acid, a competitive inhibitor of HMG-CoA
reductase. Can you recall another lactone hydrolysis reaction encountered in an earlier chapter?

7. What is the role of high-density lipoproteins in regulation of cholesterol levels in the blood?

8. Synthesis of the steroid hormones from cholesterol starts with the reaction catalyzed by
desmolase shown below. Why is this a critical reaction for formation of steroid hormones?

400
Chapter 24 Lipid Biosynthesis

H3C

H3C

O
HO
Cholesterol C H

Isocaproic aldehyde

O
H3C

H3C

HO
Pregnenolone

Abbreviated Answers
1. The immediate source of carbons is acetyl-CoAs, which are produced from carbohydrates,
amino acids, and lipids. Acetyl-CoA is produced in the mitochondria but fatty acid biosynthesis
occurs in the cytosol. To move acetyl units out of the mitochondria, they are condensed onto
oxaloacetate to form citrate, in a citric acid cycle reaction. Citrate is then transported out of the
mitochondria to the cytosol, where ATP-citrate lyase catabolizes citrate to acetyl-CoA and
oxaloacetate. This cytosolic acetyl-CoA is used to synthesize fatty acids. So, the citrate-malate-
pyruvate shuttle is responsible for moving two-carbon units from the mitochondria to the cytosol.
However, it has another purpose: it supplies some of the NADPH needed for fatty acid synthesis.
Cytosolic oxaloacetate is reduced to malate and then oxidatively decarboxylated to CO2 and
pyruvate by malic enzyme, in an NADP+-dependent reaction. The NADPH thus formed is
consumed during the reduction steps of fatty acid biosynthesis.

2. When glycolysis was covered, it was pointed out that phosphofructokinase activity is
inhibited by citrate. In this chapter, we saw how citrate is used to move two-carbon units from
the mitochondria to the cytosol for fatty acid biosynthesis. An increase in the concentration of
citrate is a signal that the citric acid cycle is backing up, either because energy stores are
satisfactory or because there is an abundance of acetyl units. In either case, there is not reason
to continue glycolysis. Movement of citrate out of the mitochondria shifts acetyl units from
degradation via the citric acid to storage via cytosolic fatty acid synthesis and serves to turn down
glycolysis at phosphofructokinase.

3. Biotin is a component of acetyl-CoA carboxylase. Nicotinamide is found in NADPH.

Phosphopantetheine is covalently attached to acyl carrier protein. Biotin functions as an
intermediate carrier of activated carboxyl groups in malonyl-CoA biosynthesis by acetyl-CoA
carboxylase. NADPH is required at two reduction steps in each round of chain elongation in fatty
acid biosynthesis. Phosphopantotheine serves as a carrier of the growing fatty acid. This group
is covalently attached to a serine residue in acyl carrier protein and serves to carry acetyl groups,
malonyl groups, and acyl groups during various stages of fatty acid biosynthesis.

401
Chapter 24 Lipid Biosynthesis

4. Mammals cannot introduce a double bond beyond C-9 in a given fatty acid. The
prostaglandins are synthesized from linoleic acid, 9,12-octadecadienoic acid, which cannot be
produced by mammals and is therefore an essential fatty acid.

5. The components of glycerophospholipids are glycerol, phosphate, fatty acids, and an alcoholic
head group. Synthesis starts with either glycerol (via reduction of glyceraldehyde) or DHAP being
converted to glycerol-3-phosphate by glycerokinase or glycerol-3-phosphate dehydrogenase,
respectively. Glycerol-3-phosphate is converted to 1-acylglycerol-3-P and then to phosphatidic
acid (1,2-diacylglycerol-3-P) by two acyltransferase reactions. Phosphatidic acid serves as the
precursor for triacylglycerol and the glycerophospholipids phosphatidylethanolamine (PE),
phosphatidylcholine (PC), phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylglycerol
(PG), and cardiolipin (diphosphatidylglycerol). Phosphatidic acid is converted to diacylglycerol,
which is converted to PE or PC by transferases using CDP-derivatized ethanolamine or choline.
Alternatively, phosphatidic acid can be converted to its CDP derivative, CDP-diacylglycerol, which
is metabolized to PI or PG. PS is produced from PE using serine to displace ethanolamine.

6. In the pentose phosphate pathway, conversion of 6-phosphogluconolactone to 6-

phosphogluconate involves hydrolysis of a lactone.

7. HDL is assembled in the endoplasmic reticulum of liver cells and secreted into the blood.
Newly synthesized HDL contains very little cholesterol but with time it accumulates cholesterol as
both free cholesterol and as cholesterol esters. HDL then returns to the liver where cholesterol is
either stored or converted to bile salts and excreted?

8. The steroid hormones are transported in the blood to target tissues and must therefore be
slightly more soluble than cholesterol. The reaction catalyzed by desmolase removes the
hydrocarbon tail of cholesterol, making the product more soluble.

Summary
The biosynthesis of lipid molecules proceeds via mechanisms and pathways, which are
different from those of their degradation. In the synthesis of fatty acids, for example, 1)
intermediates are linked covalently to the -SH groups of acyl carrier proteins instead of coenzyme
A, 2) synthesis occurs in the cytosol instead of the mitochondria, 3) the nicotinamide coenzyme
used is NADPH instead of NADH, and 4) in eukaryotes, the enzymes of fatty acid synthesis are
associated in one large polypeptide chain instead of being separate enzymes. Fatty acids are
synthesized by the addition of two carbon acetate units, which have been activated by the
formation of malonyl-CoA, decarboxylation of which drives the reaction forward. Once the
growing fatty acid chain reaches 16 carbons in length, it dissociates from the fatty acid synthase
and is subject to the introduction of unsaturations or additional elongation. Acetyl-CoA needed
for fatty acid synthesis is provided in the cytosol by citrate that is transported across the
mitochondrial membrane and converted to acetyl-CoA and oxaloacetate by ATP-citrate lyase.
Formation of malonyl-CoA by acetyl-CoA carboxylase (ACC), a biotin-dependent enzyme, commits
acetate units to fatty acid synthesis. In animals, ACC is a multifunctional protein, which forms
long, filamentous polymers. It is allosterically activated (and polymerized) by citrate and
inhibited (and depolymerized) by palmitoyl-CoA. Affinities for both these regulators are decreased
by phosphorylation of the enzyme at up to 8 to 10 separate sites. The fatty acid synthesis
reactions involve formation of O-acetyl and O-malonyl enzyme intermediates, followed by transfer
of the acetyl group to the -SH of an acyl carrier protein (ACP) and then to the -ketoacyl-ACP
synthase. Transfer of the malonyl group to the ACP is followed by decarboxylation of the malonyl
group and condensation of the remaining two-carbon unit with the carbonyl carbon of the acetate
group on the synthase. This is followed by reduction of the -carbonyl to an alcohol, dehydration
to yield a trans-, double bond and reduction to yield a saturated bond. Introduction of
unsaturations in the nascent chain occurs by O2-dependent and O2-independent pathways and
may be followed by further chain elongation. Several mechanisms are utilized to introduce
multiple unsaturations in a fatty acid chain. Regulation of fatty acid synthesis is related to
regulation of fatty acid breakdown and the activity of the TCA cycle, because of the importance of
acetyl-CoA in all these processes. Malonyl-CoA inhibits carnitine transport, blocking fatty acid
oxidation. Citrate activates ACC and palmitoyl-CoA inhibits, both in chain-length-dependent
fashion. The enzymes of fatty acid synthesis are also under hormonal control.
Glycerolipid synthesis is built around the synthesis of phosphatidic acid from glycerol-3-
phosphate or dihydroxyacetone phosphate. Specific acyltransferases add acyl chains to these
glycerol derivatives. Other glycerolipids, such as phosphatidylcholine (PC) and

402
Chapter 24 Lipid Biosynthesis

phosphatidylethanolamine (PE) are synthesized from phosphatidic acid via CDP-diacylglycerol

and diacylglycerol. Base exchange converts phosphatidylethanolamine to phosphatidylserine.
Other phospholipids, such as phosphatidylinositol, phosphatidylglycerol and cardiolipin are
synthesized from CDP-diacylglycerol. Dihydroxyacetone phosphate is a precursor to the
plasmalogens. Platelet activating factor (PAF), an ether lipid, dilates blood vessels, reduces blood
pressure and aggregates platelets. Sphingolipids are produced via condensation of serine and
palmitoyl-CoA by 3-ketosphinganine synthase and reduction of the ketone product to form
sphingamine. Acylation followed by desaturation yields ceramide, the precursor to other
sphingolipids and cerebrosides.
Eicosanoids, derived from arachidonic acid by oxidation and cyclization, are ubiquitous local
hormones. They include the prostaglandins, thromboxanes, leukotrienes and other
hydroxyeicosanoic acids. A variety of stimuli, including histamine, epinephrine, bradykinin,
proteases and other agents associated with tissue injury and inflammation, can stimulate the
release of eicosanoids, which have short half-lives and are rapidly degraded. Aspirin acetylates
endoperoxide synthase on its cyclooxygenase subunit, irreversibly inhibiting the synthesis of
prostaglandins.
Cholesterol biosynthesis begins with mevalonic acid, which is formed from acetyl-CoA by
thiolase, HMG-CoA synthase and HMG-CoA reductase. The HMG-CoA reductase reaction is the
rate-limiting step in cholesterol biosynthesis. Inhibition of this enzyme by lovastatin blocks
cholesterol biosynthesis and can significantly lower serum cholesterol. Mevalonate is converted
to squalene via isopentenyl pyrophosphate and dimethylallyl pyrophosphate, which join to yield
farnesyl pyrophosphate and then squalene. Squalene is cyclized in two steps and converted to
lanosterol. The conversion of lanosterol to cholesterol requires another 20 steps.
Lipids circulate in the body in lipoprotein complexes, including high density lipoproteins, low
density lipoproteins, very low density lipoproteins and chylomicrons. Lipoproteins consist of a
core of mobile triacylglycerols and cholesterol esters, surrounded by a single layer of
phospholipid, into which is inserted a mixture of cholesterol and proteins. Lipoproteins are
bound to lipoprotein receptors at target sites and progressively degraded in circulation by
lipoprotein lipases. Defects of lipoprotein metabolism can lead to elevated serum cholesterol.
Steroids such as the bile acids and steroid hormones are synthesized from cholesterol via key
intermediates such as pregnenolone and progesterone. The male hormone testosterone is a
precursor to the female hormones including estradiol. Steroid hormones modulate transcription
of DNA to RNA in the cell nucleus. The corticosteroids, including glucocorticoids and
mineralocorticoids, synthesized by the adrenal glands, are important physiological regulators.

403
 16

Terpenoids and Gibberellic Acids

Interaction in Plants
Zahra Asrar
Department of Biology, Shahid Bahonar University of Kerman, Kerman
Iran

1. Introduction
Plants synthesize an astonishing diversity of isoprenoids, some of which play essential roles
in photosynthesis, respiration, and the regulation of growth and development. In spite of
economic significance of the terpenoids and their many essential functions, relatively little is
known about terpenoid metabolism and its regulation in plants (Mansouri et al, 2009).
However, two independent patheways for the biosynthesis of isoprenoid precursors coexist
within the plant cell: the cytosolic mevalonic acid (MVA) pathway and the plastidial
methyerythritol phosphate (MEP) pathway. Cannabis is a diecious species that is a source of
fiber, food, oil and medicine. Cannabinoids represent a distinctive class of compounds
belong to the chemical class of natural terpenophenols. Among others, 9-tetrahycannabinol
(THC) and cannabidiol (CBD) are the most important of these compounds. The experiments
with labeling patterns showed that the cannabinoids are derived entirely or predominantly
(98%) from the deoxyxylulose pathway (Fellemeier et al., 2001). They are produced by
glandular trichomes that occur on most aerial surfaces of the plant (Hilling 2004). Cannabis
is used in modern medicine for the treatment of emesis in chemotherapy. As well as being
useful anti-emetics, cannabinoids appear to have therapeutic value as antispasmodics,
analgesics, and appetite stimulants and have potential in the treatment of epilepsy,
glaucoma, and asthma (Agurell and Nilsson 1972; Guzman 2003; Howlett et al., 2004).
However little is known about the effects of plant hormones on the regulation of these
pathways. Thus, we investigated the effect of gibberellic acid (GA3) on changes in the amount
of many produced terpenoids and the activity of the key enzymes, 1-deoxy-D-xylulose 5-
phosphate synthase (DXS) and 3-hydroxy-3-methy glutaryl coenzyme A reductase (HMGR) in
these pathways. To understand the role of gibberellic acid (GA3) in the regulation of two
terpenoid biosynthesis pathways, we gained experience on the response of the main end
products of these pathways and cannabinoids under GA3 treatment in cannabis plants.
This chapter will provide focus on recent development of the interaction effects of
terpenoids and gibberellic acid in Cannabis sativa.

2. Biosynthesis of terpenoids
Isoprenoids (also known as terpenoides) made from assembly of isoprene units. Isoprenes are
flexible five carbon units (CH2)2C=CH-CH2- , and they comprise a diverse class of plant

www.intechopen.com
346 Advances in Selected Plant Physiology Aspects

metabolites; most of them are known as defense-related compounds, flavors or scents (Chris et
al., 2007). Terpenoids constitute the largest family of natural plant products, with over 30,000
members (Dewick, 2002). Members of this diverse group of natural products are found in all
organisms. In primary metabolism, they function as photosynthetic pigments (chlorophylls
and carotenoids), electron transport (ubiquinone and plastoquinone), plant hormones (abscisic
acid and gibberellins) and membrane fluidity such as sterols (Lang and Ghassemian, 2003;
Mcgarvey and Croteau, 1995). The plant produced isoprenoids -carotene (provitamine A)
and -tocopherol (VitE) are required for the maintenance of human health (Shintani and
Dellapenna, 1998; Hirschberg, 1999). Industrial uses of isoprenoids include products such as
colorants, fragrance, and flavorings (Lang and Croteau, 1999).
Otto Wallach who in 1910 recognized that isoprene is the basic constituent of terpens (Fig. 1)
and Leopold Ruzicha found that isoprene is the basic element for the synthesis of many
natural compounds including steroids. He postulated the biogenic isoprene rule, according
to which all terpenoids (derivatives of terpens) are synthesized via a hypothetical precursor
which he named active isoprene. This speculation was verified by Feodor Lynen in 1964,
when he identified isopentenyl pyrophosphate to be the active isoprene (Heldt, 2005).

Fig. 1. Isoprene unit and Terpens

2.1 Terpenoids have two different biosynthetic pathways

All isoprenoides are derived from the ubiquitous C5 building blocks isopentenyl
diphosphate (isopentenyl pyrophosphate) (IPP) and dimethylallyl diphosphate (DMAPP).
These precursors can be synthesized by independent pathways in higher plants and algae
(Lichtenthaler, 1999), the classical mevalonate pathway in the cytoplasm or the alternative
non-mevalonate pathway in plastids (Arigoni et al., 1997; Rohmer, 1999). The plastidial
pathway, now known as the 2-C-methy-D-erythriol-4-phosphate (MEP) pathway, has been
fully elucidated by a combination of biochemical and genomic approaches (Rodriguez and
Boronat, 2002); it provides the precursors for monoterpenes, diterpenes, carotenoids,
tocopherols, and the prenyl moiety to chlorophyll (Eisenreich et al., 2001).

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 347

2.1.1 Mevalonate pathway

Konrad Bloch in 1964 discovered that acetyl CoA is a precursor for the biosynthesis of
steroids. In fact, the mevalonate pathway (MVA pathway) has long been assumed to be the
exclusive biosynthesis route used in all organisms for the synthesis of Isopentenyl
pyrophosphate (IPP); particularly in mammalian where it is responsible for the biosynthesis
of cholesterols. Fig. 2 shows the synthesis of the intermediary product IPP: two molecules of
acetyl CoA react to produce acetoacety CoA and with another acety CoA yielding -
hydroxy- methyl glutaryl CoA (HMG CoA). In plants a single enzyme, HMG CoA
synthase catalyzes both reactions (Heldt, 2005). Several studies have emphasized the
regulatory role of 3-hydroxy 3-methy glutaryl CoA enzyme reductase (HMGR) in the
mevalonate pathway. HMG R is considered to be the most heavily regulated enzyme in
mammalian metabolism (Goldstein and Brown, 1990) and there are numerous studies
demonstrating that it is also closely regulated in plants, particularly in the case of sterol and
phytoalexine synthesis (Weissenborn, 1995; Stermer et al., 1994).
The esterified carboxyl group of HMG Co A is reduced by two molecules of NADPH to a
hydroxyl group accompanied by hydrolysis of the energy-rich thioester bond. Thus
mevalonic acid is formed. A pyrophosphate ester is formed in two successive
phosphorylation steps, catalyzed by two different kinases. Third molecule of ATP, then
involving the transitory formation of a phosphate ester, a carbon-carbon double bond is
generated and the remaining carboxyl group is removed (Heldt, 2005). Isopentenyl
pyrophosphate is the basic unit for the formation of an isoprenoid chain and the cytosolic
pathway, which starts from acetyl-CoA and proceeds through the intermediate mevalonate
(MVA), provides the precursors for sterols and ubiquinons (Laule et al., 2003).

Fig. 2. Scheme of Mevalonate (MVA) pathway from acetyl CoA to Isopenteny

phyrophosphate (IPP)

2.1.2 The DOXP/ MEP pathway

Experiments with plants led to the discovery that synthesize of isopentenyl pyrophosphate
in the plastids follow a different pathway (Fig. 3). The precursors for this pathway are

www.intechopen.com
348 Advances in Selected Plant Physiology Aspects

Fig. 3. Scheme of 1-deoxy-D-xylulose-5-phosphate/methyl-D-erythrithol (DOXP/MEP

pathway). The enzymes of both pathways are numbered. MEP/DOXP pathway for
isopentenyl diphosphate (IPP) and isoprenoid biosynthesis

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 349

pyruvate and D-glyceraldehyde-3-phosphate. Pyruvate is decarboxylated via thiamine

pyrophosphate (TPP) and then is transferred to D-glyceraldehyde-3-phosphate to yield 1-
deoxy-D-xylulose-5-phosphate (DOXP). After isomerization and reduction by NADPH, 2
C-methyl erythriol-4-phosphate (MEP) is synthesized. MEP is then activated by reacting
with CTP to yield CDP methyl erythriol. Two further reduction steps, following
dehydration and phosphorylation. Finally yield isopentenyl pyrophosphate (Heldt, 2005).
The MEP-synthase pathway for isoprenids is present in bacteria, algae, and plants, but not
in animals. A large part of plant isoprenoids, including the hemiterpene, monoterpenes like
limnone, phytohormones such as gibberellins (Crozier et al., 2000), brasinoesteroids, and
phytoalexine (Lang and Gassemian, 2003; Kanno et al., 2006).
The production and accumulation of cannabinoids in plants of Cannabis sativa follow non-
mevalonate pathway (Fellermeier et al., 2001) synthesized via the MEP pathway located in
plastids (Rohmer, 1999).
1-deoxy-D-xylulose-erythriol phosphate synthase (DXS) catalyzes the first step in the
methy-erythriol phosphate (MEP) pathway and is hypothesized to be an important control
step within MEP pathway (Lichtenthaler, 1999). The role of the DXS in plants can be seen in
Arabidopsis, the over expression of DXS led to an increase in terpenoids concentrations while
the repression of DXS decreased terpenoid concentrations (Estevez et al., 2001).

3. Biosynthesis of gibberellins
Farmers in Asia were aware of a disease of rice plants called bakanae (foolish seedling)
disease. Infected plants would grow excessively taller than normal healthy plants, and being
fall over and be unharvestable. This disease was found to be caused by a fungus known as
Gibberella fujikuroi (Yabuta, T., 1935). Thus, Work on this special disease in Japan occurred by
Japanese scientists who isolated a growth promoting substances from cultures of a fungus
that parasitizes rice plants in the 1930s. They called it gibberellin. Scientists in the 1950s
rediscovered this work and extracted a range of chemical that elicit the growth response in
rice seedlings. Since that time, more than 100 slightly different gibberellins have been
identified chemically.
Gibberellins (GAs) are diterpene plant hormones or plant growth regulator of vascular
plants, fungi, and bacteria that are biosynthesized through complex pathways and control
diverse aspects of growth and development, including seed germination, shoot elongation,
leaf expansion, pollen-tube growth, trichome, flower and fruit development, cell division,
cell elongation and floral transition (Olszewski et al., 2002; Razem et al., 2006). Among more
than a hundered GAs identified from plants (http://www.plant-hormones.info/
gibberelline-nomenclature.htm) , only a small number of them, such as GA1, GA3, GA4, and
GA7 are thought to function as bioactive (biologically active gibberellins) hormones.
Therefore, many non bioactive gibberellins exist in plants as precursors for the bioactive
forms or deactivated metabolism (Yamaguchi, 2008). The major bioactive GAs, commonly
have a hydroxyl group on C-3 , a carboxyl group on C-6 and a lactone between C-4 and C-
10 (Fig. 4). GAs has been identified frequently in a variety of plant species, implying that it
acts as a widespread bioactive hormone.

www.intechopen.com
350 Advances in Selected Plant Physiology Aspects

Fig. 4. Active Gibberellins

According to the nature of the enzymes involved, gibberellins biosynthesis pathways are
usually divided into three parts. Each of the three parts consists of several steps:
I- Biosynthesis of ent-Kaurene
Recentely, a second, MVA- independent pathway to isopentenyl pyrophosphate (IPP) via
glyceraldehydes 3-phosphate and pyruvate was discovered in a green alga (Schwender et
al., 1996). Isopentenyl pyrophosphate is further converted to gerany geranyl pyrophosphate
(GGPP) by two enzymes, IPP-isomerase and GGPP-synthase in plastids of higher plants
(Dogbo & Camara, 1987).
Gerany gerany pyrophosphate is cyclized to ent-copalyl pyrophosphate (CPP) and finally
to ent-kaurene. In higher plants these reactions are catalyzed by two enzymes, CPP
synthase formerly Kaurene synthase A and ent-Kaurene (K) synthase formerly Kaurene
synthase B, renaming was suggested by Mcmillan 1979. Both enzymes were localized in
isolated proplastids of meristematic shoot tissues, but not in mature chloroplasts of pea
and wheat (Aachet al., 1995, 1997) or of pumpkins endosperm (Simcox et al., 1975; Aach et
al., 1995).
II- From ent-Kaurene to GA12
In the second stage of GA biosynthesis (Fig. 5), the C-19 methy (CH3) group of Kaurene is
oxidized in three steps to give ent-kaurenoic acid (KA). These oxidations are catalyzed by
ent-kaurene oxidase (KO), which is multifunctional because it can catalyze all three
reactions. In Arabidopsis, KO is encoded by GA3, and mutation in this gene will again
produce severly dwarfed plant (Shinjiro, 2008).
The intermediate, second, part of the pathway is catalyzed by microsomal NADPH-
dependent cytochrome P-450 monooxygenases at the endoplasmic reticulum (Lange, 1998).
Thus, the precursor ent-kaurene must be translocated from the proplastids to the
endoplasmic reticulum, although nothing is known about the transport mechanism (Lange,
1998). Ent-kaurene is oxidized into GA12, via ent-kaurenol, ent kaurenal, entkaurenoic acid,
ent-7 -hydroxy kaurenoic acid, and GA12-aldehyde. Certain biosynthesis steps, Such as 7-
oxidation, 12 -hydroxylation, and 13-hydroxylation are catalyzed by both monooxygenases
and soluble dioxygenases, which occasionally occur together within the same species or
even within the same tissue (Lange & Graebe, 1993).

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 351

III- Gibberellins from GA12

Some initial steps of the final part of pathway might overlap with the second part. However,
the third part starts with GA12-aldehyde is oxidized by soluble 2-oxoglutarate- dependent
dioxygenases (Fig. 5). The GA dioxygenases are multifunctional with broad substrate
specificity, resulting in many side reactions and numerous gibberellines found in higher
plants (Lange & Graebe, 1993; Hedden & Kamiya, 1997).

Fig. 5. Scheme of gibberellins biosynthesis (DXS) in the leaves of cannabis plants.

4. The interaction effect of terpenoids and GA3 at vegetative stage

The influence of gibberellic acid (GA3) on plastidic and cytosolic terpenoids and on two key
enzymes, 1-deoxy-D-xylulose-5-phosphate synthase (DXS) and 3-hydroxy-3-methylglutaryl
coenzyme A reductase (HMGR), for terpenoid biosynthesis was compared in vegetative

www.intechopen.com
352 Advances in Selected Plant Physiology Aspects

cannabis plants. However, the regulation of the biosynthesis of terpenoids in plant cell is
poorly understood. We do not know how the metabolite fluxes between the primary and
secondary metabolisms are orchestrated. Plant growth regulators have essential roles in
growth and development of plants and plant resposes to environment. In order to understand
how interact with terpenoid biosynthesis; we focused on the role of GA3 in the regulation of
primary and secondary terpenoid production in Cannabis sativum at vegetative stage.
Treatment with exogenous GA3 resulted in a decrease of DXS activity in comparison with
the control plants (Fig. 6). Although, there is no report on the effect of GA3 on DXS activity,
many reports support a regulatory role of DXS for the production of MEP-derived
isoprenoids in plants (Rodriguez-concepcion, 2006). Estevez et al (2001) reported that
transgene mediated up regulation or down regulation of DXS levels in Arabidopsis were
correlated with concomitant changes in the levels of MEP derived isoprenoid end product.

Fig. 6. Effects of gibberellic acid (GA3) on 1-deoxy-D-xylulose 5-phosphate synthase (DXS)

in the leaves of cannabis plants.

The effects of the concentration of GA3 on the chlorophyll and carotenoid contents of
cannabis leaf are present in Fig. 7 & 8. Reduction in chlorophyll content after GA3
application has been reported in wheat (Misra & biswal, 1980), rice seedlings (Yim et al.,
1997) and pea (Bora & Sarma, 2006) which are consistence with our results (Mansouri et al.,
2011).The variation in the level of the cartenoids were similar to those observed for the
chlorophylls. Apparently carotenoids and chlorophylls accumulation are controlled through
a similar mechanism, because both of them are reduced by GA3. The changes in chlorophyll
and carotenoid were parallel with changes in DXS activity. It can show the limiting role of
DXS activity in chlorophyll and carotenoid synthesis.
The tocopherol is lipophilic antioxidants that are synthesized by photosynthetic
organisms, occurring mainly in leaves and seeds. Literature sources indicate that the
major tocopherol form in leaf tissues is tocopherol (Szymanska and Kruk, 2008).
Abdul Jaleel et al. (2007) reported that GA3 treatment stimulated tocopherol
accumulation in Catharanthus roseus.The tocopherolcontent of cannabis plants
decrease in 50 M GA3 treatment and increased with increasing GA3 at vegetative stage
(Fig. 9) which is consistence with other reports.

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 353

Fig. 7. Effects of gibberellic acid (GA3) on chlorophylls a, b and total in the leaves of
cannabis plants.

Fig. 8. Effects of gibberellic acid (GA3) on carotenoids in the leaves of cannabis plants.

Fig. 9. Effects of gibberellic acid (GA3) on a-tocopherol content in the leaves of cannabis plants

www.intechopen.com
354 Advances in Selected Plant Physiology Aspects

Fig. 10. Effects of gibberellic acid (GA3) on 3-hydroxy-3-methylglutaryl coenzyme A

reductase (HMGR) activities in the leaves of cannabis plants.

Exogenous GA3 caused an increasing in HMGR activity (Fig. 10). Since HMGR is an
important control point for the MVA pathway in plants (Kato-Emori et al., 2001), the
increase in HMGR activity should result in increasing the supply of phytosterols. Squalene
and phytosterol contents showed similar changes and increased concomitant with HMGR
activity in cannabis plants (Fig.11, 12). Consistent with our results, pea seedlings treated
with GA3 showed an increase in the HMGR activity (Russell & Davidson, 1982).
The amount of THC and CBD increase with increasing GA3 treatment in comparison with the
control plants (Fig. 13).The THC content was higher than those of CBD content ( Mansouri et al.,
2011). Perhaps, the increase observed in the THC and CBD content at high level of GA3 is not a
direct effect of GA3 treatment and could reflect the GA3 interaction with other plant hormones.
As has been shown that exogenous application of GA3 caused a clear increase in ACC content.
ACC oxidase activity and ethylene biosynthesis occur during the breaking of dormancy and
onset of germination in Fagus sylvatica L. seeds (Calvo et al., 2004). Furthermore, it is possible
that ethylene caused the increase observed in THC and CBD content.

Fig. 11. Effects of gibberellic acid (GA3) on squalene content in the leaves of cannabis plants.

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 355

Fig. 12. Effects of gibberellic acid (GA3) on phytosterol content in the leaves of cannabis
plants.

On the whole, these results showed that GA3 treatment had opposite effect on primary
terpenoid biosynthesis by MVA and MEP pathways. GA3 treatment caused a decrease in
DXS activity and biosynthesized primary terpenoids from MEP pathway, but this treatment
increased HMGR activity and phytosterols from MVA pathway. Whereas, secondary
terpenoids showed different response to GA3 treatment and it could be because of
interference of two biosynthetic pathways in their formation.

Fig. 13. Effects of gibberellic acid (GA3) on THC and CBD content in the leaves of cannabis
plants.

5. The effect of GA3 and terpenoids at flowering stage

In spite of economic importance of the terpenoids and their many essential functions,
relatively little is known about terpenoid metabolism and its regulation in plants. To
understand the role of gibberellic acid (GA3) in the regulation of two terpenoid biosynthesis

www.intechopen.com
356 Advances in Selected Plant Physiology Aspects

pathways, we studied the response of the main end products of these pathways and
cannabinoids under GA3 treatment in cannabis plants.
Effects of different concentrations of GA3 on levels of chlorophyll, carotenoids and
tocopherol in leaves of male and female cannabis plants were investigated. Male plants
treated with 50uM GA3 had lower chlorophyll a and total contents compared to the control
(Fig. 14A). Treatment of female leaves with 50 and 100uM GA3 significantly decreased
chlorophyll a, b and total contents in a dose-dependent pattern (Fig. 14B). The carotenoid
contents of treated plants were lower compared with the control (Fig. 15A). Low
concentrations of GA3 were more effective on decreasing carotenoid contents in male plants.
The phytyl (C20) conjugates chlorophylls and tocopheroles, and carotenoids (C40) are
produced by the MEP pathway. GA3 treatment decreased chlorophyll contents in C.sativa
plants. Reduction in chlorophyll content after GA3 application has been reported in, wheat
(Misra & Biswal 1980), peach trees (Monge et al. 1994), rice seedlings (Yim et al. 1997) and
pea (Bora & Sarma 2006). Perez et al. (1974) have shown that a mutant of tomato with
increased levels of GA3 contains less chlorophyll. Our results showed that GA3 caused a
decrease in carotenoid contents (Fig. 14). Apparently carotenoid and chlorophyll
accumulationis controlled through a similar mechanism, because both of them are reduced
by GA3. The previous studies indicate that GA3 treatments delayed the chloroplast
chromoplast conversion of colored fruit (Goldshmidt, 1998; Pfander, 1992). The
development of chromoplasts is accompanied by the accumulation of carotenoids (Vainstein
et al., 1994). Also by Rodrigo and Zacarias (2007) reported that GA3 reduced the ethylene
induced expression of early carotenoid biosynthetic genes and the accumulation of
phytoene in orange.

Values are means of four replications standard deviation (SD).

Fig. 14. Effects of gibberellic acid (GA3) on chlorophyll a, b and total in leaves of male (A)
and female (B) cannabis plants.

Fig. 15B shows the effect of different concentrations of GA3 on tocopherol content all
concentrations of GA3 sprayed on leaves caused an increase in tocopherol content in
both sexes. The highest increase was detected with 50uM GA3 in male plants. Apparently,

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 357

the pattern of changes in tocopherol content was an invert of that in chlorophyll and
carotenoid contents.

Values are means of four replications standard deviation (SD).

Fig. 15. Effects of gibberellic acid (GA3) on (A) carotenoids and (B) -tocopherol in leaves of
female and male cannabis plants.

Tocopherols ( , , , and tocopherol) are lipophilic antioxidants that collectively

constitute vitamin E (Crowell et al, 2008). Therefore, we measured the changes in the
amounts of tocopherol in response to GA3 treatment. Our results showed that GA3
increased tocopherol content in cannabis plants (Mansouri et al., 2009). A stimulatory
effect of applied GA3 on tocopherol content was also absorbed in Catharanthus roseus
(Abdul Jaleel et al., 2007). On the other hand, we found a reversed relationship between
photosynthetic pigments (chlorophyll and carotenoids) and tocopherol contents in
treated plants with GA3. Both tocopherol and chlorophyll contain a phytol moiety as
part of their molecule. The substrate used for their biosynthesis may be derived from a
common pool. It is possible that the decrease in chlorophyll and carotenoid biosynthesis
caused an increase in the substrate accumulation and therefore biosynthesis. Also Rise et al.
(1989) reported a transient increase in tocopherol accompanied with a decrease in
chlorophyll during the course of senescence in several plant species. Their results indicated
that the phytol released by chlorophylase during the initial stages of chlorophyll
breakdown, perhaps used for the biosynthesis of tocopherol during senescence.
The effects of GA3 on key enzyme activity of terpenoid biosynthetic pathways (DXS and
HMGR) were investigated. As shown in Fig. 16 male plants had more DXS activity than
female plants and a significant decrease was observed over DXS activity in treated plants
(male and female) by GA3 which was liner with increasing GA3 concentration. GA3
treatment had a stronger effect on decreasing DXS activity in male plants.
GA3 treatment caused a decrease in the 1-deoxy-D-xylulose5-phosphate synthase (DXS)
activity in male and female of Cannabis sativa plants (Mansouri et al., 2009).This is the first
report about the effect of GA3 on DXS activity. Many reports support a regulatory role of
DXS for the production of MEP- derived isoprenoids in plants (Rodriguez-Concepcion
2006). Estevez et al. (2001) reported that transgene-mediated up regulation or down

www.intechopen.com
358 Advances in Selected Plant Physiology Aspects

regulation of DXS levels in Arabidopsis were correlated with concomitant changes in the
levels of MEP-derived isoprenoid end products. However, in our investigation, the decrease
in the entire MEP pathway end product. The results showed an increase in tocopherol
content and a decrease in chlorophyll and carotenids. These data support this hypothesis
that several enzymes share control over the metabolic flux through the MEP pathway. The
MEP pathway might be regulated at several control points to compensate for fluctuations in
precursor/product equilibrium and redistribute the balance of control within the pathway
(Enfissi et al., 2005). Another possibility for this uncoordinated activity between the DXS
and the plastidic isoprenoids is the exchange or precursors between the cytosol and the
plastid. Strong biochemical evidence for such exchange of precursors was reported by
Kasahara et al., (2002), Nagata et al. (2002) and Hemmerlin et al. (2003).

(HMGR) activities in the leaves of female and male cannabis plants.

Values are means of four replications standard deviation (SD).
Fig. 16. Effects of gibberellic acid (GA3) on 1-deoxy-D-xylulose 5-phosphate synthase (DXS)
and 3-hydroxy-3-methylglutaryl coenzyme A reductase.

Gibberellic acid treatment increased HMGR activity in the treated plants (Fig. 16). However
some differences are obvious between the two groups of plants of different sexes (Fig.16). The
male individuals had a greater HMGR activity at lower GA3 concentrations than female plants.
Activity of 3-hydroxy-3-methyglutaryl coenzyme A reductase (HMGR) in cannabis plants
demonstrated that HMGR activity was greater in plants treated with GA3. Consistent with
our result, Russell and Davidson (1982) reported that GA3 increased HMGR activity in pea
seedlings. It was shown that higher levels of HMGR activity were usually associated with
rapidly growing parts of plants (Brooker and Russell 1975). On the othe words, Ga is widely
regarded as a growth-promoting compound. Thus, it can be a candidate for stimulating of
the HMGR activity. The lower HMGR activity was seen in the male plants treated with 100
M GA3. The decrease observed in the HMGR activity could reflect the GA3 interaction in
high concentration with other plant hormones. As it has been shown that exogenous
application of GA3 caused a clear increase in ACC content, ACC oxidase activity and
ethylene biosynthesis occur during the breaking of dormancy and onset of germination in
Fagus sylvatica seeds (Calvo et al. 2004). In addition, it is possible that ethylene caused the
decrease observed in HMGR activity.

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 359

The mature leaves of male and female cannabis plants were used to determine the effect of
GA3 on squalene (biosynthetic precursor to all steroids), campestrol, stigmasterol, and
sitosterol (the most representative phytosteroids of the MVA pathway). A significant
increase in squalene (Fig. 17), stigmasterol and sitosterol accumulation occurred in female
and male plants treated with gibberellic acid (Fig. 18). The changes in the amount of
squalene and sitosterol were coordinate with the changes in HMGR activity. In addition,
GA3 had a stimulatory effect on campestrol accumulation in male plants.

(HMGR) activities in the leaves of female and male cannabis plants.

Values are means of four replications standard deviation (SD).
Fig. 17. Effects of gibberellic acid (GA3) on squalene content in leaves of female and male
cannabis plants.

The data in our study showed that GA3 treatment increased squalene and phytostrol
contents in a pattern similar to the changes in the HMGR acitivity. Since HMGR is an
important control point for the MVA pathway in plants (Kato-Emori et al. 2001), the increase
in HMGR activity should result in increasing the supply of phytostrols. Free sterols are
found predominantly in cell membranes and are thought to contribute to the proper
functioning of membranes by controlling the fluidity characteristics of the membrane
(Devarenne et al. 2002). Douglas and paleg (1974) demonstrated that the inhibitors of GAs
biosynthesis caused a decrease in sterol accumulation. Huttly and Philips (1995) suggested
that GA3 causes an increase in cell number and size to produce a significant effect. Since
these processes need the production of cell membrane. It can be assumed that GA3 should
induce the phytostrol biosynthesis to influence its effects on growth in plants.
A comparison between male and female plants showed that females had higher amounts of
THC; especially in the flowers (Fig. 19). THC content of the leaves was slightly lower than
that of flowers. The application of GA3 to cannabis plants resulted in a decrease in THC
content. Gibberellic acid treatment had a stronger effect in decreasing THC content in the
flowers of male plants in comparison with that of female plants. However, the leaves of the
two sexes indicated similar responses to GA3 treatment.
9-tetrahydrocannabinol (THC) is the cannabinoid responsible for the main psychoactive
effects of most Cannabis drug preparations (Mecoulam 1970). Factors that control

www.intechopen.com
360 Advances in Selected Plant Physiology Aspects

biosynthesis and distribution of cannabinoids within the plant are unknown. We

investigated the impact of altered GA3 levels on this secondary metabolite in C.sativa, and
the results indicated that apart from the influence of GA3, female plants had more THC
than male plants in leaves and flowers. The treated plants with GA3 had lower THC
content in comparison with that in control plants. It is demonstrated that cannabinoids are
synthesized from the DXP pathway (Fellemeier et al. 2001). Furthermore, our results
showed that GA3 decreased THC content by decreasing the DXS activity and necessary
precursors for THC biosynthesis.

(HMGR) activities in the leaves of female and male cannabis plants.

Values are means of four replications standard deviation (SD).
Fig. 18. Effects of gibberellic acid (GA3) on campesterol, stigmasterol and sitosterol content
in leaves of (A) male and (B) female cannabis plants.

(HMGR) activities in the leaves of female and male cannabis plants.

Values are means of four replications standard deviation (SD).
Fig. 19. Effect of gibberellic acid (GA3) on _9-tetrahydrocannabinol (THC) content in (A)
leaves and (B) flowers of the male and female cannabis plants.

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 361

6. Conclusion
In conclusion, the pattern of changes in the amounts of primary terpenoids (chlorophyll,
carotenoids, and phytosterols) in Cannabis sativa suggest that GA3 have opposite effects on
the primary terpenoid biosynthesis of MEP and MVA pathways. In addition, the appearance
of the direct relationship between DXS and HMGR activity, and their main end products
confirm an important role for these enzymes in the MEP and MVA pathways regulation.
However, to understand the role of GA3 in terpenoid biosynthesis regulation, we need
further investigation.

7. Acknowledgement
The autor wishes to thank Mr. Hossein Mozafari phD student of Shahid Bahonar university
of Kerman for his critical comments and help of the manuscript.

8. References
Aach, H.; Bode, H.; Robinson, DG. & Graebe, JE. (1997). ent-Kaurene synthase is located in
proplastids of meristematic shoot tissues, Planta, Vol.202, pp.211-219.
Aach, H.; Bose, G. & Graebe, JE. (1995).ent-Kaurene biosynthesis in a cell-free system from
wheat (Triticum aestivum l.) seedlings and the localization of ent-Kaurene synthase
in plastids of three species, Planta, Vol.197, pp. 333-342
Abdul Jaleel, c.; Gopi, R.; Manivannan, P.; Sankar, B.; Kishorekumar, A. & Panneerselvam,
R. (2007). Antioxidant potentials and ajmalicine accumulation in Catharanthus roseus
after treatment with gibberellic acid. Colloids Surf B Biointerf. Vol. 60, pp.195-200.
Agurell, S. & Nilsson, LG. (1972). The chemistry and biological activity of cannabis. Bullet.
Narcotic. pp 35-37.
Arigoni, D.; Sagner, S.; Latzel, C.; Eisenrich, W.; Bacher , A. & Zenk, MH. (1997).
Terpenoidbiosynthesis from 1-deoxy-D-xylulose in higher plants by intermolecular
skeletal rearrangement, Proc Natl Acad Sci, Vol. 94, pp.10600-10605.
Bora, RK. & Sarma, CM. (2006). Effect of gibberellic acid and Cycoel on growth , yield and
protein content of pea. Asi J plant Sci. Vol.5, pp.324-330.
Brooker, JD. Russell, DW. (1975). Properties of microsomal 3-hydroxy-3-methylglutaryl
coenzyme A reductase from Pisum sativum seedings, Arch. Biochem. Biophys. Vol.
167, pp723-729
Calvo, Ap.; Nicolas, C.; Lorenzo, O.; Nicolas, G. & Rodriguez, D. (2004). Evidence for
positive regulation by gibberellins and ethylene of ACC oxidase expression and
activity during transition from dormancy to germination in Fagus sylvatica L.
seeds,J. Plant Growth Regul. Vol. 23, pp.44-53.
Crowell, EF., McGrath, JM.& Douches, DS,. (2008). Accomulation of vitamin E in potato
(Solanum toberosum) tubers, Transgenic Res. Vol.17, pp.205-217.
Crowell, EF.; Mcgrath, Jm.& Douches, DS. (2008). Accumulation of vitE in potato ( Salanum
tuberosum) tubers, Transgenic Res. Vol. 17 pp.205-217.
Crozier, A.; Kamiya, Y.; Bishop, G. & Yokota, T. (2000). Biosynthesis of hormones and
elicitor molecules. In: Buchanan BB, Gruissem W, Jones RL (eds) Biochemistry and
molecular biology of plants, Amer Soc Plant Physio, Roclville, pp.850-929.

www.intechopen.com
362 Advances in Selected Plant Physiology Aspects

Devarenne, TP.; Ghosh, A. & Chappell, J. (2002). Regulation of squalene synthase, a key
enzyme of sterol biosynthesis, in tobacco. Plant Physiol. Vol 129, pp 181-222.
Dewick, PM. (2002). The biosynthesis of C5-C25 terpenoid compounds, Nat Prod Rep, Vol.19,
pp.181-222.
Dogbo O. & Camara, B. (1987). Purification of isopentenyl pyrophosphate isomerase and
geranylgeranyl pyrophosphate synthase from Capsicum chrooplasts by affinity
chromatography. Biochim Biophys Acta. Vol. 920, pp.140-148.
Douglas, TG.; Paleg, LG. (1974). Plant Physiol. Vol 54, pp 238-245.
Eisenreich, W.; Rojdich, F. & Bacher, A. (2001). Deoxyxylulose phosphate pathway to
terpenoids, Trends Plant Sci., Vol.6, pp.78-84.
Enfissi, MA.; Fraser, PD.; Lois, LM.; Boronat, A.; Schuch, W. & Bramley, PM. (2005).
Metabolic engineering of the mevalonate and non-mevalonate isopentenyl
diphosphate-forming pathways for the production of health-promoting
isoprenoids in tomato, Plant Biotech J, Vol.3, pp.17-27.
Estevez, JM.; Cantero, A.; Reindl, A.; Reichler, S. & Leon, P. (2001) 1-deoxy-D-Xylulose 5-
phosphate synthase, limiting enzyme for plastidic isoprenoid biosynthesis in
plants, J Bio Chem Vol. 276, pp. 22901-22909.
Fellermeier, M.; Eisenreich,W.; Bacher, A. & Zenk, MH. (2001). Biosynthesis of cannabinoids:
incorporation experiment with 13C-labeled glucoses, Eur . J. Biochem. Vol. 268,
pp1596-1604.
Fellermeier, M.; Eisereich, W.; Bacher, A. & Zenk, MH. (2001). Biosynthsis of cannabinoids:
incorporation experiments with 13C- labeled glucose. Eur J Biochem Vol.268,
pp.1596-1604.
Goldschmidt, EE. (1988). Regulatory aspects of chloro-chromoplast interconversion in
seeding citrus fruit peel, Isr. J. Bot. Vol. 37, pp.23-130.
Gusman, M. (2003). Cannabinoids: potential anticancer agents, Nat. Rev. Vol. 3, pp. 745-755.
Hedden, P. & Kamiya, Y. (1997). Gibberellin biosynthesis enzymes, genes and their
regulation, AnnRev Plant Physiol Plant Mol Biol, Vol.48, pp.431-460.
Heldt, HW., (2005). A large diversity of isoprenoids has multiple functions in plant
metabolism, Plant Biochemistry. 3rd Edition ISBN: 0-12-088391-0
Hemmerlin, A.; Hoeffler, JF.; Meyer, O.; Tritsch, D.; Kagan, IA. & Grosdemanage-Billiard C.
(2003). Cross talk between the cytosolic mevalonate and the plastidial
methlerythritol phosphate pathways in tobacco Bright Yellow-2 cells, J. BIol. Chem.
Vol.278, pp.26666-26676
Hilling, KW. (2004). A chemotaxonomic analysis of terpenoid variation in cannabis, Biochem.
Systems Ecol. Vol. 32, pp. 875-891
Hischberg, J. (1999). Production of high value compounds: carotenoids and vitamin E, Curr.
Opin. Biotechnol, Vol. 20, pp.186-191.
Howlett, AC.; Breivogel, CS.; Cholders, SR.; Deadwyler, SA. Hampson, Re. & Porrino LY
(2004). d physiology and pharmacology: 30 year progress. Neuropharmacol. Vol. 47,
pp. 345-358.
Huttly, AK. & Phillips, AL. (1995). Gibberellin regulated plant genes. Physol. Plant. Vol. 95,
pp. 310-317.
Kanno, Y.; Otomo, K.; Kenmoku, H.; Mitsuhashi, W.; Yamane, H.; Oikawa, H.; Toshima, H.;
Matsuoka, M.; Sassa, T. & Toyomasu, T. (2006). Characterization of a rice gene

www.intechopen.com
Terpenoids and Gibberellic Acids Interaction in Plants 363

family encoding type A diterpene cyclases, Biosci Biotechnol Biochem, Vol. 70,
pp.1702-1710.
Kasahara, H, Hanada A, Kuzuyama T, Takagi M, Kamiya Y, Yamaguchi S.
(2002).Contribution of the mevalonate and methylerythritol phosphate pathways to
the biosynthesis of gibberellin in Arabidopsis, J. Biol. Chem. Vol. 277, pp. 45188-
45194.
Kato-Emori, S, Higashi K, Hosoga K, Kobayashi T, Ezura H (2001). Cloning and
characterizing of the gene encoding 3-hydroxy-3-methylglutaryl coenzyme A
reductase in melon (Cucumic melo L. reticulatus), Mol. Genet. Genomics Vol. 265,
pp. 135-142.
Lang, Bm. & Ghassemian, M. (2003). Genome organization in Arabidopsis thaliana : a survey
for gene involved in isoprenoid and chlorophyll metabolism. Plant Mol Biol, vol. 51
pp. 925-948.
Lang, H. (1998). Molecular biology of gibberelline synthesis, Plant, Vol.204 pp. 409-419.
Lang, T. & Graebe, JE. (1993). Enzymes of gibberellins synthesis. In: Lea PJ (ed) Method in
plant biochemistry, Vol.9, Academic Press, London, pp.403-430.
Lange, BM. & Croteau, R. (1999). Genetic engineering of essential oil production in mint,
Curr Opin Pant Biol, Vol.2, pp.139-144.
Laule, O.; Futholz, A.; Chang, HS.; Zhu, T.; Wang, X. & Heifetz, PB. (2003). Cross talk
between cytosolic and plastidial pathways of isoprenoid biosynthesis
inArabidopsis thaliana, Plant Biol, Vol.100, pp.6866-6871.
Lichtenthaler, HK. (1999). The 1-deoxy-D-xylulose-5-phosphate pathway of isoprenoid
biosynthesis in plants. Annual Review of plant physiology and plant molecular biology,
Vol. 50, pp.47-65.
Lichtenthaler, HK.; Rohmer, M. & Schwender, J. (1997) Two independent biochemical
pathways for isopentenyl diphosphate and isoprenoid biosynthesis in higher
plants. Pysiol plant, vol. 101 pp. 643-652.
MaGarvey, DJ. & Croteau, R. (1995). Terpenoid metabolism. The plant cells Vol. 7, pp. 1015-
1026.
Mansouri, H.; Asrar, Z. & Amarowicz, R. (2011). The respose of terpenoids to exogenous
gibberellic acid in Cannabis sativa at vegetative stage, Acta physiol plant, Vol.33,
pp.1085-1091.
Mansouri, H.; Asrar, Z. & Mehrabani, M. (2009) Effect of gibberellic acid on primary
terpenoids and Tetrahydrocannabinol in Cannabis sativa at flowerin stage. Journal of
integrative plant Biology Vol. 51, pp. 553-561.
McCaskill, D. & Croteau, R. (1998). Some caveats for bioengineering terpenoid metabolism
in plants, TIBITECH, Vol.16, pp.349-355.
Mechoulam, R. (1970). Marijuana chemistry. Science, Vol. 168, pp.1159-1166.
Misra, AN. Biswal, UC. (1980). Effect of photohormons on chlorophyll degradation during
again of chloroplasts in vivo and in vitro, Protoplasma Vol. 105, pp. 1-8.
Monge, E.; Aguirre. R. & Blanco, A. (1994). Application of paclobutrazol and GA3 to adult
peach trees: effects on nutritional status and photosynthetic pigments, J. Plant
Growth Regul. Vol. 13, pp. 15-19.
Negata, N.; Suzuki, M.; Yoshida, S. & Muranaka, T. (2002). Mevalonic acid partially restres
chloroplast and etioplast development in Arabidopsis lacing the non-mevalonate
pathway, Planta, Vol.216, pp.345-350

www.intechopen.com
364 Advances in Selected Plant Physiology Aspects

Olszewski, N.; Sun, T. & Gubler, T. (2002). Gibberellin signaling: Biosynthesis, catabolism,
and response pathways, Plan Cell, Vol.14, pp.61-80.
Perez, AT.; Marsh, HV. & Lachman, WH. (1974). Physiology of the the yellow green 6 gene
in tomato, Plant Physiol, Vol.53, pp.192-197.
Pfander, H. (1992). Carotenoids: an overview, Meth Enzymol, Vol.213, pp.3-13.
Razem, FA.; Baron, K. & Hill, RD. (2006). Turning on gibberellins and abscisic acid
signaling, Curr Opin Plant Bio, Vol.9, pp.454-459.
Rise, M.; Conjocaru, M.; Gottlieb, HE. & Goldschmidt, EE. (1989). Accumulation of
tocopherol in sensscing organs as related to chlorophyll degradation, Plant Physiol,
Vol.89, pp.1028-1030.
Rodrigo, MJ. & Zacarias, L. (2007). Effect of postharvested ethylene treatment on carotenoid
accumulation and the expression of carotenoid biosynthetic genes in the flavedo of
orange (Citrus sinensis L.), Postharvest Biol Technol, Vol.43, pp.14-22.
Rodriguez-concecion, M. (2006). Early steps in isoprenoid biosynthesis: Multilevel
regulation of the supply of common precursors in plant cells, Phytochem Rev,
Vol.5, pp.1-15.
Rodriguez-concepcion, M. & Boronat, A. (2002). Elucidation of the methylerythritol
phosphate pathwayfor isoprenoids biosynthesis in bacteria and plastids, A
metabolic milestone achieved through genomics, Plant Physiol, Vol. 130, pp.1079-
1089.
Russell, DW. & Davidson, H. (1982). Regulationof cytosolic HMG-CoA reductase activity in
pea seedlings: contrasting responses to different hormones, and hormone product
interaction, suggest hormonal modulation of activity, Biochem. Biophys. Res.
Commun. Vol. 104pp. 1537-1543.
Shinjiro, Y. (2008). Gibberellin Metaboilism and its regulation, Anne. Rev. plant Biol. Vol. 59,
pp. 225-251.
Shintani, D. & Della Penna, D. (1998). Elevation the Vit E content of plants through
metabolic engineering, Science, Vol.282, pp.2098-2100.
Szymanska, R. & Kruk, J. (2008). -tocopherol content and isomers composition in selected
plant species, Plant Physiol Biochem, Vol.46, pp.29-33.
Yabuta, T. (1935). Biochemistry of the Bakabaefungus of rice. Agr. Hort. Vol. 10, pp. 17-22.
Yim, KD.; Kwon, YW. & Bayer, DE. (1997). Growth response and allocation of assimilates of
rice seedlings by paclobutrazol and gibberellins treatment, J Plant Growth Regul,
Vol.16, pp.35-41.

www.intechopen.com
 Advances in Selected Plant Physiology Aspects
Edited by Dr. Giuseppe Montanaro

ISBN 978-953-51-0557-2
Hard cover, 388 pages
Publisher InTech
Published online 25, April, 2012
Published in print edition April, 2012

The book provides general principles and new insights of some plant physiology aspects covering abiotic
stress, plant water relations, mineral nutrition and reproduction. Plant response to reduced water availability
and other abiotic stress (e.g. metals) have been analysed through changes in water absorption and transport
mechanisms, as well as by molecular and genetic approach. A relatively new aspects of fruit nutrition are
presented in order to provide the basis for the improvement of some fruit quality traits. The involvement of
hormones, nutritional and proteomic plant profiles together with some structure/function of sexual components
have also been addressed. Written by leading scientists from around the world it may serve as source of
methods, theories, ideas and tools for students, researchers and experts in that areas of plant physiology.

How to reference
In order to correctly reference this scholarly work, feel free to copy and paste the following:

Zahra Asrar (2012). Terpenoids and Gibberellic Acids Interaction in Plants, Advances in Selected Plant
Physiology Aspects, Dr. Giuseppe Montanaro (Ed.), ISBN: 978-953-51-0557-2, InTech, Available from:
http://www.intechopen.com/books/advances-in-selected-plant-physiology-aspects/the-interaction-effect-of-
some-terpenoids-and-gibberellic-acid-in-cannabis-sativa-at-vegetative

InTech Europe InTech China

University Campus STeP Ri Unit 405, Office Block, Hotel Equatorial Shanghai
Slavka Krautzeka 83/A No.65, Yan An Road (West), Shanghai, 200040, China
51000 Rijeka, Croatia
Phone: +385 (51) 770 447 Phone: +86-21-62489820
Fax: +385 (51) 686 166 Fax: +86-21-62489821
www.intechopen.com
Natural products
Connections
51
Building on: Arriving at: Looking forward to:
Stereochemistry ch16 Natural products are made by Organic synthesis ch53
secondary metabolism
Conformational analysis ch18
Enolate chemistry and synthesis Natural products come in enormous
ch24ch30 variety, but fall mainly into four types:
alkaloids, polyketides, terpenes, and
Pericyclic reactions ch35ch36
steroids
Rearrangement and fragmentation
Alkaloids are amines made from amino
ch37ch38 acids
Radicals ch39
Pyrrolidine alkaloids from ornithine;
Chemistry of life ch49 benzylisoquinoline alkaloids from
Mechanisms in biological chemistry tyrosine
ch50 Morphine alkaloids are made by radical
cyclizations
Fatty acids are built up from acetyl
CoA and malonyl CoA subunits
Polyketides are unreduced variants of
fatty acids
Terpenes are made from mevalonic
acid
Steroids are tetracyclic terpene
derivatives
Biomimetic synthesis: learning from
Nature

Introduction
By natural products, we mean the molecules of nature. Of course, all life is made of molecules, and OH
we will not be discussing in great detail the major biological molecules, such as proteins and nucleic HO NHMe
acids, which we looked at in Chapters 49 and 50. In this chapter we shall talk much more about mol-
ecules such as adrenaline (epinephrine). Adrenaline is a human hormone. It is produced in
moments of stress and increases our blood pressure and heart rate ready for fight or flight. Youve HO
adrenaline
got to sit an exam tomorrowsurge of adrenaline. To an organic chemist adrenaline is intensely
interesting because of its remarkable biological activitybut it is also a molecule whose chemical
reactions can be studied, whose NMR spectrum can be analysed, which can be synthesized, and coniine
which can be imitated in the search for new medicines. an alkaloid
N
By the end of this chapter we hope you will be able to recognize some basic classes of natural H
products and know a bit about their chemistry. We will meet alkaloids such as coniine, the molecule
in hemlock that killed Socrates, and terpenes such as thujone, which was probably the toxin in
absinthe that killed the nineteenth-century artists in Paris. O
Then there are the ambiguous natural products such as the steroid cholesterol, which may cause thujone
a terpene
innumerable deaths through heart disease but which is a vital component of cell walls, and the
polyketide thromboxane, one drop of which would instantaneously clot all the blood in your body
but without which you would bleed to death if you cut yourself.
1414 51 . Natural products

H CO2H
O
 H H O
Before moving on, just pause to
admire brevetoxin, a wonderful HO OH
and deadly molecule. Look at the cholesterola steroid thromboxane A2a polyketide
alternating oxygen atoms on the
top and bottom faces of alternate We will look at the structural variety within these four important classes and beyond, from
rings. Look at the rings perhaps the smallest natural product, nitric oxide, NO (which controls penile erections in
themselvessix-, seven-, and
eight-membered but each with
men), to something approaching the largestthe polyketide brevetoxin, the algal product in
one and no more than one oxygen red tides, which appear in coastal waters from time to time and kill fish and those who eat the
atom. Trace the continuous fish.
carbon chain running from the
O
lactone carbonyl group in the HO
bottom left-hand corner to the
Me H
aldehyde carbonyl in the top right.
There is no break in this chain brevetoxina toxic polyketide O O H
H
and, other than the methyl groups,
no branch. With 22 stereogenic H
H
centres, this is a beautiful piece of Me Me H
Me O O H
molecular architecture. If you want H Me H O
H
to read more about brevetoxin, O O
read the last chapter in Nicolaou H
and Sorensens Classics in total O O
synthesis, VCH, 1996. H H H Me
O O O
H H H
H H H Many natural products are the source of important life-saving drugsconsider the millions of
R N S lives saved by penicillin, a family of amino acid metabolites.
O N
O
CO2H
Natural products come from secondary metabolism
penicillin The chemical reactions common to all living things involve the primary metabolism of the big four
e.g. penicillin G; R = PhCH2
we met in Chapter 49nucleic acids, proteins, carbohydrates, and lipids. Now we must look at
chemical reactions that are more restricted. They occur perhaps in just one species, though more
commonly in several. They are obviously, then, not essential for life, though they usually help sur-
vival. These are the products of secondary metabolism.
The exploration of the compounds produced by the secondary metabolism of plants, microor-
ganisms, fungi, insects, mammals, and every other type of living thing has hardly begun. Even so, the
variety and richness of the structures are overwhelming. Without some kind of classification the task
of description would be hopeless. We are going to use a biosynthetic classification, grouping sub-
stances not by species but by methods of biological synthesis. Though every species is different, the
basic chemical reactions are shared by all. The chart on p. 1415 relates closely to the chart of primary
metabolism in the previous chapter.

Alkaloids are basic compounds from amino acid metabolism

Alkaloids were known in ancient times because they are easy to extract from plants and some of
them have powerful and deadly effects. Any plant contains millions of chemical compounds, but
some plants, like the deadly nightshade, can be mashed up and extracted with aqueous acid to
give a few compounds soluble in that medium, which precipitate on neutralization. These com-
pounds were seen to be like alkali and Meissner, the apothecary from Halle, in 1819 named them
alkaloids. Lucrezia Borgia already knew all about this and put the deadly nightshade extract
atropine in her eyes (to make her look beautiful: atropine dilates the pupils) and in the drinks of her
 Alkaloids are basic compounds from amino acid metabolism 1415

secondary
metabolism
CO2
HO
O
photo- in O P H
synthesis plants
O
HO O
HO
O OH
HO alkaloids
HO erythrose-3-phosphate nitrogen compounds
OH OH
glucose

CO2H

O OH
P
aromatic compounds
O O amino acids, pigments, etc.
HO OH
OH
OH

O shikimic acid

phosphoenolpyruvate
O O O
polyketides

O R SCoA
n
OH linear polyketide chain
fatty acids

O
pyruvic acid O O

O SCoA
O
malonyl coenzyme A
SCoA
acetyl
coenzyme A OH terpenes
HO2C
OH
mevalonic acid steroids

citric
acid R CO2H
cycle
H NH2 alkaloids
aliphatic
amino acids nitrogen compounds
HO2C CO2H
HO CO2H
citric acid

chemical reaction in the usual sense: the starting material is incorporated into the product
1416 51 . Natural products

political adversaries to avoid any trouble in the future. Now, we would simply say that they are basic
because they are amines. Here is a selection with the basic amino groups marked in black.
HO
Me N
O
O
N O OH
H NMe
Me H H
N
HO
nicotine morphine atropine

Natural products are often named by a combination of the name of the organism from which they are
isolated and a chemical part name. These compounds are all amines so all their names end in -ine. They
appear very diverse in structure but all are made in nature from amino acids, and we will look at three types.

Solanaceae alkaloids
The Solanaceae family includes not only deadly
nightshade (Atropa belladonnahence atropine) plants
but also potatoes and tomatoes. Parts of these plants
also contain toxic alkaloids: for example, you should not
eat green potatoes because they contain the toxic
alkaloid solanine.
H
H
Atropine is a racemic compound but the (S)-enantiomer
occurs in henbane (Hyoscyamus niger ) and was given a
different name, hyoscyamine, before the structures were
known. In fact, hyoscyamine racemizes very easily just on
heating in water or on treatment with weak base. This is
probably what happens in the deadly nightshade plant.

N
H
H
H H the very toxic alkaloid
solanine, a mixture of
RO glycosides with R =
glucose, mannose, etc.

Pyrrolidine alkaloids are made from the amino acid ornithine

Pyrrolidine is the simple five-membered cyclic amine and pyrrolidine alkaloids contain this ring
O somewhere in their structure. Both nicotine and atropine contain a pyrrolidine ring as do hygrine and
tropinone. All are made in nature from ornithine. Ornithine is an amino acid not usually found in
N N
proteins but most organisms use it, often in the excretion of toxic substances. If birds are fed benzoic
H acid (PhCO2H) they excrete dibenzoyl ornithine. When dead animals decay, the decarboxylation of
Me ornithine leads to putrescine which, as its name suggest, smells revolting. It is the smell of death.
pyrrolidine hygrine
O
Me N
CO2H PhCO2H CO2H decay
Ph N H2N H2N
O H in birds
H HN Ph H NH2 NH2
dibenzoyl ornithine ornithine putrescine
tropinone
O

pyrrolidine
alkaloids

Biosynthetic pathways are usually

worked out by isotopic labelling of O
CO2H plant
potential precursors and we shall mark H2N
the label with a coloured blob. If N
H NH2
ornithine is labelled with 14C and fed to 14C-ornithine Me
the plant, labelled hygrine is isolated. hygrine
 Alkaloids are basic compounds from amino acid metabolism 1417

If each amino group in ornithine is O

labelled in turn with 15N, the amino CO2H plant
H2N
group is lost but the amino group is N
H NH2
retained. Me
ornithine
hygrine 

Further labelling experiments along these lines showed that the CO2H group as well as the Both reagents SAM and acetyl CoA
were discussed in Chapter 50. We will
amino group was lost from ornithine and that the rest of the molecule makes the pyrrolidine ring. not be able to repeat at length the
details of the chemistry of these and
The three-carbon side-chain in hygrine comes from acetate, or rather from acetyl CoA, and the N- other common biochemical reagents
methyl group comes from SAM. We can now work through the biosynthesis. already discussed there. In general, in
this chapter we will give only the
The first step is a pyridoxal-catalysed decarboxylation of ornithine, which follows the normal distinctive or interesting steps and
sequence up to a point. leave you to consult Chapter 50 if you
O need more help.

H2N O H2N H2N

N N N
pyridoxal H
CO2H phosphate
H2N OH OH OH
PO PO PO
H NH2

N Me N Me N Me
H H H
Now the terminal amino group is methylated by SAM and the

secondary amine cyclizes on to the pyridoxal imine to give an ami-
Notice that the methylation step means that the two carbon atoms that
nal. Decomposition of the aminal the other way round expels eventually become joined to nitrogen in the five-membered ring remain
pyridoxamine and releases the salt of an electrophilic imine. different throughout the sequence. If, say, putrescine had been an
Me intermediate, they would not now be distinguishable.
N H
N NH H2N
H N H Enz
Me
SAM OH
OH
OH PO PO
PO N +

N Me Me N Me
N Me
H H
H
The rest of the biosynthesis does not need pyridoxal, but it does need two molecules of acetyl
CoA. In Chapter 50 we noted that this thiol ester is a good electrophile and also enolizes easily. We
need both reactivities now in a Claisen ester condensation of acetyl CoA.
Enz H B Enz OH
O O O OH O
Claisen ester
H condensation
O
CoAS CoAS CoAS CoAS
enolization
CoAS acetoacetyl CoA stable delocalized enol

The new keto-ester is very like the acetoacetates we used in Chapter 27 to make stable enolates
and the CoA thiol ester will exist mainly as its enol, stabilized by conjugation.
This enol reacts with the imine salt we have previously made and it will be easier to see this reaction
if we redraw the enol in a different O O
conformation. The imine salt does not
have to wait around for acetoacetyl N N
CoA to be made. The cell has a good Me Me
stock of acetyl CoA and its condensa- O SCoA O SCoA
tion product. H B Enz
1418 51 . Natural products

All that remains to form hygrine is the hydrolysis of the CoA thiol ester and decarboxylation of
the keto-acid. This is standard chemistry, but you should ensure that you can draw the mechanisms
for these steps.
O O O

N N N
thiol ester -keto-acid
Me hydrolysis Me decarboxylation
Me
O SCoA O OH hygrine

Tropinone is made from hygrine and it is clear what is needed. The methyl ketone must
enolize and it must attack another imine salt resembling the first but on the other side of the ring.
Such salts can be made chemically by oxidation with Hg(II) and biologically with an oxidizing
enzyme and, say, NAD+. The symbol [O] represents an undefined oxidizing agent, chemical or
biological.

O O OH Me N
[O] H
N N N O
H oxidation enolization cyclization
Me Me Me
tropinone

This complex route to tropinone was imitated as long ago as 1917 in one of the most celebrated
The cyclization step looks
dreadful when drawn on a flat reactions of all time, Robinsons tropinone synthesis. Robinson argued on purely chemical grounds
molecule, but it looks much that the sequence of imine salts and enols, which later (1970) turned out to be Natures route, could
better in the conformation of be produced under natural conditions (aqueous solution at pH 7) from a C4 dialdehyde, MeNH2
tropinone shown below.
and acetone dicarboxylic acid. It worked and the intermediates must be very similar to those in the
Me biosynthesis.
N CO2H
Me N
OH CHO
pH 7
+ Me NH2 + O
O
CHO water
CO2H tropinone

Other pyrrolidine alkaloids

There are many pyrrolidine alkaloids derived from decarboxylation and cyclization initiated by pyridoxal. We
ornithine and another large family of piperidine alkaloids will not discuss these compounds in detail.
derived from lysine by similar pathways involving
H 2N CO2H pyridoxal
phosphate
H NH2 pyridoxal
N N
lysine H H
Me
O O N
O
+ O
N CoAS
N
acetoacetyl CoA H
H pelletierine pseudo-pelletierine

Benzyl isoquinoline alkaloids are made from tyrosine

We switch to a completely different kind of alkaloid made from a different kind of amino acid. The
benzyl isoquinoline alkaloids have a benzyl group attached to position 2 of an isoquinoline ring.
Usually the alkaloids are oxygenated on the benzene ring and many are found in opium poppies
(Papaver somniferum). For all these reasons papaverine is an ideal example.
 Alkaloids are basic compounds from amino acid metabolism 1419

MeO

N N N
MeO
isoquinoline OMe

OMe
benzyl isoquinoline papaverine

Labelling shows that these alkaloids come from two molecules of tyrosine. One must lose CO2
and the other NH3. We can easily see how to divide the molecule in half, but the details will have to
wait a moment.
13C label
MeO

CO2H Papaver
somniferum N
MeO
H NH
2 OMe
HO
Tyr
tyrosine papaverine
OMe
The question of when the extra OH groups are added was also solved by labelling and it was
found that dihydroxyphenyl pyruvate was incorporated into both halves but the dihydroxyphenyl-
alanine (an important metabolite usually called dopa) was incorporated only into the isoquinoline
half.
HO CO2H Papaver
somniferum MeO
H NH
2 N
HO
dopa MeO
OMe
HO CO2H Papaver
somniferum
O OMe
HO
dihydroxyphenylpyruvate

The amino acid and the keto-acid are, of course, related by a pyridoxal-mediated transaminase
and the hydroxylation must occur right at the start. Both of these reactions are discussed in Chapter
50.
CO2H CO2H
pyridoxal

H NH2 transaminase O
HO HO
Tyr
tyrosine hydroxylase

HO CO2H pyridoxal HO CO2H

H NH2 transaminase O
HO HO
dopa dihydroxyphenylpyruvate
1420 51 . Natural products

Catecholamines
Dopa and dopamine are important compounds because they are the precursors oxidase (Chapter 50) hydroxylates stereospecifically at the benzylic position to
to adrenaline in humans. Decarboxylation of dopa gives dopamine, which an give noradrenaline (norepinephrine).

H H OH OH
HO NH2 benzylic HO NH2 HO NHMe
hydroxylation SAM

HO HO HO
dopamine noradrenaline adrenaline
(norepinephrine) (epinephrine)
HO
The family of hormones that includes adrenaline and noradrenaline is control the breakdown of stored sugars to release glucose and they
often called the catecholamines (catechol is 1,2-dihydroxybenzene). have a direct effect on blood pressure, heart rate, and breathing. The
The hormones are produced in the adrenal gland around the kidneys relative proportion of noradrenaline and its N-methylated analogue,
and regulate several important aspects of metabolism: they help to adrenaline, controls these things. HO
catechol

Pyridoxal-mediated decarboxylation of dopa gives dopamine and this reacts with the keto-acid to
form an imine salt. This is an open-chain imine salt unlike the cyclic ones we saw in the pyrrolidine
alkaloids, but it will prove to have similar reactivity.
HO
HO
dopamine
pyridoxal NH
NH2 HO
HO
HO2C
HO2C OH OH
CO2

DHPP O
OH OH
The imine salt is perfectly placed for an intramolecular electrophilic aromatic substitution
by the electron-rich dihydroxyphenyl ring. This closes the isoquinoline ring in a Mannich-
like process (Chapter 27) with the phenol replacing the enol in the pyrrolidine alkaloid bio-
synthesis.

 HO HO
Even in biological electrophilic aromatic
substitutions, it is still important to H
remember to write in the hydrogen
NH NH
HO HO
atom at the place of substitution
(Chapter 22)! HO2C HO2C
OH OH

OH OH
HO HO
CO2
NH NH
HO HO
HO2C pyridoxal
OH OH

OH OH

The cyclization product is still an amino acid and it can be decarboxylated by pyridoxal. Now we
have something quite like papaverine but it lacks the methyl groups and the aromatic heterocyclic
ring. Methylation needs SAM and is done in two stages for a reason we will discover soon. The final
oxidation should again remind you of the closing stages of the tropinone route.
 Alkaloids are basic compounds from amino acid metabolism 1421

MeO MeO

NH N
4 SAM MeO [O] MeO
OMe OMe

papaverine
OMe OMe
The reaction to make the isoquinoline ring can be carried out chemically under very mild condi-

tions providing that we use an aldehyde as the carbonyl component. Then it works very well with
The reaction also works with an
rather similar compounds. aryl pyruvic acid, but the
HO HO decarboxylation is more difficult
to organize without pyridoxal.
NH2 pH 6 NH
HO HO
CHO
25C
water O
O

O
O
The mechanism is straightforwardthe imine is formed and will be protonated at pH 6, ready for
the CC bond formation, which is both a Mannich reaction and an electrophilic aromatic substitution.

HO HO
Notice that it was not necessary
to protect the OH groupsthe
NH NH acetal on the lower ring is not for
HO HO
H protection, and this group
O O (methylenedioxy or dioxolan) is
present in many benzyl
isoquinoline alkaloids. It is
O O formed in nature by oxidation of
an MeO group ortho to an OH
Complex benzyl isoquinoline alkaloids are formed by radical coupling group on a benzene ring.
A more interesting series of alkaloids arises when benzyl isoquinoline alkaloids cyclize by radical
reactions. Phenols easily form radicals when treated with oxidizing agents such as Fe(III), and benzyl
isoquinoline alkaloids with free phenolic hydroxyl groups undergo radical reactions in an intra- 
molecular fashion through a similar mechanism. Here are the details of some methylations of a class See Chapter 39.
of alkaloids closely related to papaverine.
HO MeO

NH NH
HO HO The names of the alkaloids
H 2 SAM H should not, of course, be learned,
OH OH but they are a convenient handle
for quick reference. The prefix
nor means without a methyl
norlaudanosoline OH norreticuline OMe group, in this case the N-Me
group, as you can see with
2 SAM norreticuline and reticuline.
1 SAM

MeO MeO

NH NMe
MeO HO
H H
OMe OH

norlaudanosine OMe reticuline OMe

1422 51 . Natural products

Methylating only one phenol on each ring of norreticuline leaves the other one free for radical
coupling. Reticuline is oxidized in the plant to isoboldine by a radical cyclization with the formation
of a new CC bond.
MeO MeO
orthopara coupling
NMe NMe
HO [O] HO
H H
OH

reticuline OMe MeO isoboldine

OH
The new CC bond is marked in black and the free phenolic OHs in green. Notice the relationship
between them. The new bond is between a carbon atom ortho to one OH group and a carbon atom
para to the other. We shall see in all these phenolic couplings that the ortho and para positions are the
only activated ones (ortho/ortho, ortho/para, and para/para couplings are all possible). Oxidation
occurs at the phenolic hydroxyl groups, and the resulting oxygen radicals couple.
MeO
MeO

NMe
O NMe
H Enz H O
H H
H
isoboldine
ortho/
para
MeO radical
coupling MeO
O
O H Enz
Phenol coupling occurs chemically under oxidation with Fe(III). The most famous example is the
coupling of 2-naphthol to give binaphtholan ortho/ortho coupling. The stereochemistry of
binaphthyls like this was discussed in Chapter 45.

OH
Fe(III) O etc. OH

O OH

Similar phenol couplings have been attempted in the laboratory with compounds in the benzyl
isoquinoline series but the nitrogen atom interferes if it is at all basic. When it has a carbonyl sub-
stituent the reactions do work reasonably well, but the yields are poor. Nature is still much better at
this reaction than we are.
MeO nonbasic N MeO

N OEt N OEt
HO Fe(III) HO
O O
[K3Fe(CN)6]

MeO MeO
OH OH
Reticuline is also the source of the morphine alkaloids by ortho/para radical coupling. The roles of
the two rings are reversed this time and it is quite difficult to see at first how the structures are related.
 Alkaloids are basic compounds from amino acid metabolism 1423

MeO
MeO
ortho/para coupling

NMe [O]
HO O
H
OH NMe

MeO intermediate
reticuline OMe for morphine
O alkaloids

A great deal has happened in this reaction, but the new CC bond (black) is ortho to the green oxy-
gen atom in the top ring and para to the green oxygen atom in the bottom ring, so ortho/para coupling
has occurred. To draw the reaction mechanism we need to draw reticuline in the right conformation.

MeO
OH
rotate over to this
side of molecule OMe
NMe O
[O]
NMe
H
MeO MeO
OH reticuline O
MeO MeO

Enz H O HO
H

oxidative
NMe aromatize
NMe
coupling
MeO MeO
O O
One of the two rings can re-aromatize but the other has a quaternary carbon atom so no proton
can be lost from this site. Instead, the OH group in the top ring adds in conjugate fashion to the
enone in the bottom ring.
MeO MeO MeO

HO
O O

NMe NMe NMe

MeO MeO MeO

O H Enz OH O

This intermediate gives rise to the important alkaloids codeine and morphine, which differ only
by a methyl group. Nature can remove methyl groups as well as add them.
MeO MeO HO

O O O
NMe NMe NMe
H H
MeO HO HO
O codeine morphine
1424 51 . Natural products

These alkaloids have MeO HO

plenty of stereochemistry.
Indeed, if we compare the NMe
HO
structures of reticuline and H O
morphine, we can see that OH
the one stereogenic centre NMe
H
in reticuline (marked in OMe
reticuline HO
green) is still there in mor- morphine
phine (it hasnt been invertedthat part of the molecule has just been turned over) and that four
new stereogenic centres marked in black have been added. These centres all result from the original
twisting of reticuline to allow phenol coupling except for the one bearing an OH group, which comes
from a stereoselective reduction.

Boldine, an isomer of isoboldine, is formed by rearrangement

We mentioned isoboldine a while back, so there must be a boldine as well. This alkaloid is also
formed from norlaudanosoline by a different methylation sequence and oxidative radical coupling.
Looking at the structure of boldine you may see what appears to be a mistake on someones part.
HO HO HO

NH NMe NMe
HO MeO MeO
H H [O] H

HO MeO MeO

OH OH OH
boldine
norlaudanosoline

The coupling is correctly para in the bottom ring but is meta in the top ring. But there is no mis-
take (neither by the authors nor by Nature!)this structure is correct and it has been made by
para/para coupling.
O O O

NMe NMe NMe

MeO MeO MeO
H H H H

para/para
coupling
MeO MeO MeO
O O H Enz OH

One of the rings has aromatized, but the other cannotthis should remind you of the morphine
biosynthesis. However, there is no nucleophilic OH group here capable of conjugate addition to the
enone so a rearrangement occurs instead. The new bond to the lower ring migrates across the top ring.
You might even say that the lower ring does an intramolecular conjugate addition on the upper ring.
O HO
HO
Enz H
NMe NMe
NMe MeO
MeO MeO
H H
H H

MeO
MeO MeO
OH boldine
OH OH
 Fatty acids and other polyketides are made from acetyl CoA 1425

After the rearrangement there is a proton available to be lost and the cation can aromatize. The para
relationship in the original coupling product has become a meta relationship by rearrangement. You
should be able to recognize this rearrangement from Chapter 37: it is a dienonephenol rearrangement.
In rearrangements like these with cationic intermediates, the group that can best support a posi-
tive charge usually prefers to migrate. The reasons for this are discussed in Chapter 37. Here is a
purely chemical example of the same reaction, giving 82% yield in acidic solution. The bond that
migrates is marked in black.
MeO MeO

NMe NMe
HO HO
H2SO4
MeO MeO

HO
O
OMe MeO multifloramine

Fatty acids and other polyketides are made from acetyl CoA
The sections that remain in this chapter show how Nature can take a very simple moleculeacetyl
CoAand build it up into an amazing variety of structures. There are two main pathways from
acetyl CoA and each gives rise to two important series of natural products.

SCoA

O O HO
HO2C OH
HO SCoA
malonyl CoA mevalonic acid

fatty poly- terpenes steroids

acids ketides

We shall discuss these four types of compounds in the order shown so that we start with the sim-
plest, the fatty acids. You met these compounds in Chapter 49 as their glyceryl esters, but you now
need to learn about the acids in more detail and outline their biosynthesis. Compare the structures of
the typical fatty acids in the chart overleaf.
These are just a few of the fatty acids that exist, but all are present in our diet and youll find many
referred to on the labels of processed foods. You should notice a number of features.
They have straight chains with no branching
They have even numbers of carbon atoms
They may be saturated with no double bonds in the chain, or
They may have one or more C=C double bonds in the chain, in which case they are usually cis (Z)
alkenes. If there is more than one C=C double bond, they are not conjugated (either with the
CO2H group or with each other)there is normally one saturated carbon atom between them.
1426 51 . Natural products

12
saturated fatty acids 1
CO2H lauric acid
16 1
CO2H palmitic acid

18
1
CO2H stearic acid

mono-unsaturated fatty acids

18 9 1
CO2H oleic acid

poly-unsaturated fatty acids

18 12 9 1
CO2H linoleic acid

18 15 12 9 1
CO2H linolenic acid

20
14 11 8 5 1
CO2H arachidonic acid

Palmitic acid (C16 saturated) is the most common fatty acid in living things. Oleic acid (C18
mono-unsaturated) is the major fatty acid in olive oil. Arachidonic acid (C20 tetra-unsaturated) is a
rare fatty acid, which is the precursor of the very important prostaglandins, thromboxanes, and
leukotrienes, of which more later.
The prevalence of fatty acids with even numbers of carbon atoms suggests a two-carbon building
block, the most obvious being acetate. If labelled acetate is fed to plants, the fatty acids emerge with
labels on alternate carbons like this.
OH biosynthesis OH

O
O
The green blob might represent deuterium (as a CD3 group) and the black blob 13C. In fact, the
reactions are more complex than this suggests as CO2 is also needed as well as CoA and it turns out
that only the first two-carbon unit is put in as acetyl CoA. The remainder are added as malonyl CoA.
If labelled malonyl CoA is fed, the starter unit, as it is called, is not labelled.
CoAS OH biosynthesis OH

O O
no label O

Malonyl CoA is made from acetyl CoA and CO2 carried, as usual, on a molecule of biotin
(Chapter 50). The first stage in the fatty acid biosynthesis proper is a condensation between
acetyl CoA (the starter unit) and malonyl CoA with the loss of CO2. This reaction could be drawn
like this.

O O CO2
SR
SR
SR
condensing O O
enzyme
O O
 Fatty acids and other polyketides are made from acetyl CoA 1427

Notice that CO2 is lost as the new CC bond is formed. When chemists use malonates, we like to
make the stable enol using both carbonyl groups, condense, and only afterwards release CO2 (Chapter
26). Nature does this in making
acetoacetyl CoA during alkaloid SR NADPH SR
biosynthesis, but here she works differ-
ently. O O -ketoacyl-ACP OH O
reductase
The next step is reduction of the
ketone group.
This NADPH reaction is typically SR SR
stereo- and chemoselective, though the
3-hydroxyacyl-ACP
stereochemistry is rather wasted here as OH O dehydratase O
the next step is a dehydration, typical of
what is now an aldol product, and
occurring by an enzyme-catalysed E1cB SR NADPH SR
mechanism.
The elimination is known to be a cis enoyl-ACP
O reductase O
removal of H and OH and the double
bond is exclusively trans (E). Only later in the nonconjugated unsaturated fatty acids do we get Z-
alkenes. Finally, in this cycle, the double bond is reduced using another molecule of NADPH to give
the saturated side-chain.

Now the whole cycle can start again using this newly made C4 fatty acid as the starter unit and
building a C6 fatty acid and so on. Each time the cycle turns, two carbon atoms are added to the acyl
end of the growing chain.

Fatty acid synthesis uses a multienzyme complex

We have not told you the whole truth so far. Did you notice that SCoA in the structures had been
replaced by SR and that a mysterious ACP had crept into the enzyme names? That was because
these reactions actually happen while the growing molecule is attached as a thiol ester to a long side-
chain on an acyl carrier protein (ACP). The long side-chain is closely related to CoA and is attached
through a phosphate to a serine residue of the ACP.
O O
S O O Ser-ACP
N N P
H H acyl
O OH O O
carrier
pantothenic acid phosphate protein
growing chain
of fatty acid as also found in coenzyme A

All of the enzymes needed for one cycle are clumped together to form two large proteins 
(ACP, the acyl carrier protein, and CE, the condensing enzyme) which associate in a stable dimer. You saw a smaller multienzyme
complex in Chapter 50 (p. 1395), but
The long side-chain passes the substrate from enzyme to enzyme so that synthesis can be continuous this one is much more complex. More
until the chain is finished and only then is the thiol ester hydrolysed. The chart on p. 1428 illustrates are being discovered all the time
Nature invented the production line well
this. before Henry Ford.

There are three ways of making unsaturated fatty acids

Conjugated unsaturated fatty acids are made simply by stopping the acylation cycle at that stage and
hydrolysing the thiol ester linkage between the unsaturated acyl chain and ACP. They always have
the E (trans) configuration and are the starting points for other biosynthetic pathways.
1428 51 . Natural products

fatty acid biosynthesis: schematic diagram of the multienzyme dimer

HO SCoA O O

O O SH O S long flexible
pantothenic acid
side-chain on the
SCoA O acyl carrier protein
(ACP)
O SH S

cysteine residue
on the condensing
enzyme (CE) CE ACP ACP
CE

ketoacyl
reductase

SH S O SH S O

CE ACP CE ACP
Me Me

hydratase
enoyl
reductase

growing chain
O transferred to O O
cysteine residue
on CE
Me S O S

Me O

SH
S

HO SCoA
CE ACP
CE ACP
O O

multienzyme complex
is ready to start the
next cycle of acylation
 Fatty acids and other polyketides are made from acetyl CoA 1429

O O
CO2
R SACP
R SCE SACP NADPH
+
condensing O O -ketoacyl-ACP
O O enzyme reductase

R SACP R SACP hydrolysis R OH

3-hydroxyacyl-ACP
OH O dehydratase O O

The second method makes Z-3,4-unsaturated acids by deconjugation from the E-2,3-unsaturated
acids catalysed by an isomerase while the acyl chain is still attached to ACP. This is an anaerobic
route as no oxidation is required (the double bond is already thereit just has to be moved) and is
used by prokaryotes such as bacteria.

R R H Enz R H H

H SACP SACP SACP

H
O O O
Enz B H Enz H B Enz

extended enol intermediate

Removal of a proton from C4 forms an extended enol, which can be protonated at C2 

or C4. Protonation at C4 is thermodynamically favoured as it leads to the conjugated alkene. For more on this, read a specialized
book, such as Ian Flemings, Frontier
But protonation at C2 is kinetically favoured, and this leads to the nonconjugated alkene. The orbitals and organic chemical
geometry of the new alkene depends on the conformation of the chain when the first (de- reactions, Wiley, Chichester, 1976.
Similar regioselectivity is evident in the
protonation) step occurs. It is thought that this is the best conformation for the previous reac- protonation of the Birch reduction
tion, the dehydration step, and that no rotation of the chain occurs before the isomerase gets to products on p. 628.

work.
OH H Enz H B Enz
R
R R
SACP
O SACP SACP
O HO
H
H Enz
Enz B

You may think this a rather unlikely reaction, but the same thing can be done in the labora-
tory. If a simple unsaturated ester is converted into its lithium enolate and then reprotonated
with water, the major product is the ester of the Z-3,4-enoic acid. Yields and steroselectivities are
excellent.

R R H OH R H H

H OEt OEt OEt

H
O O O
Li
i-Pr2N Li extended lithium enolate 98%; R = Me

One explanation suggests that control is exercised by a favourable conformation in which 1,3- A1,3 strain (1,3-allylic strain) was
discussed in Chapter 34, p. 896.
allylic strain is preferred to 1,2-strain. It looks as though Nature has again seized on a natural chemi-
cal preference and made it even better.
1430 51 . Natural products

H H H R H R
OEt H OEt OEt
R
H
H O O OLi
i-Pr2N Li
1,2-strain
disfavoured 1,3-(allylic) strain

The third method is a concerted stereospecific removal of two adjacent hydrogen atoms from the
chain of a fatty acid after synthesis. This is an aerobic route as oxidation is required and is used by
mammals such as ourselves. The stereochemistry of the reaction is known from labelling studies to
be cis elimination.
H H O H O
H
H
H 9 1 SCoA 9 1 SCoA
R R
This oxidation involves a chain of reagents including molecular oxygen, Fe(III), FAD, and NAD+.
A hydroxylation followed by a dehydration or a sulfur-promoted dehydrogenation has been suggest-
ed for the removal of the hydrogen atoms. The chemical reaction corresponding to the biological
reaction has not yet been discovered.

What is so important about unsaturated fatty acids?

Mammals can insert a cis-alkene into the chain, providing that it is no further away from the car-
bonyl group than C9. We cannot synthesize linoleic or linolenic acids (see chart a few pages back)
directly as they have alkenes at C12 and C15. These acids must be present in our diet. And why are we
so keen to have them? They are needed for the synthesis of arachidonic acid, a C20 tetraenoic acid
that is the precursor for some very interesting and important compounds. Here is the biosynthesis of
arachidonic acid.
synthesis of unsaturated fatty acids

18 9 1
CO2H oleic acid

cis double bond

inserted MAMMALS CANNOT DO THIS

18 12 9 1
CO2H linoleic acid

cis double bond

inserted

18 1
12 9 6 CO2H -linolenic acid

one acylation cycle

20 1
14 11 8 CO2H eicosa-8,11,14-
trienoic acid

cis double bond

inserted

20
14 11 8 5 1
CO2H arachidonic acid
 Fatty acids and other polyketides are made from acetyl CoA 1431

The final product of this chain of eventsarachidonic acidis one of the eicosanoids,
so called because eicosa is Greek for twenty, and the systematic names for these com-
pounds contain eicosanoic acid in some form. The leukotrienes resemble arachidonic acid
most closely, the prostaglandins have a closed chain forming a five-membered ring, and the
thromboxanes resemble the prostaglandins but have a broken chain. All are C20 compounds
with the sites of the alkenes (C5, C8, C11, and C14) marked by functionality or some other structur-
al feature.
compounds synthesized from arachidonic acid

20
14 11 8 5 1
CO2H

HO
11 O 1
8 8 5 1
CO2H
5 CO2H

20 11 14 20
14
HO OH
leukotriene LTA4
prostaglandin F2

8 5 1
CO2H
O
11 thromboxane A2
O 14 20

OH
These compounds are all unstable and all are involved in transient events such as inflam-
mation, blood clotting, fertilization, and immune responses. They are produced locally and decay
quickly and are implicated in autoimmune diseases like asthma and arthritis. They are made by
oxidation of arachidonic acidyou can see this best if you redraw the molecule in a different con-
formation.
8 5 1 8 5 1
CO2H cyclo-
O CO2H
oxygenase

14 O 20
20 11 14
11
PGG2
O
arachidonic acid OH
The first step is a radical abstraction of a hydrogen atom from an allylic position by oxygen (per-
haps carried on an iron atom in a haem). The atom removed is between two alkenes so that the
resulting radical is doubly allylic.
O O
R R
H H H

20 20
11 14 11 14
arachidonic acid delocalized radical

This allylic radical captures a molecule of oxygen at C11 to form a new oxyradical. The reaction
occurs at one end of the delocalized radical so that the product is a conjugated diene and the new
alkene is trans (E).
1432 51 . Natural products

R R
H

O O
11 14 O 11 14
O
conjugated E,Z-diene
Now we need to resume the full structure of the intermediate because the oxyradical does an elaborate
addition to the C8 alkene and then to the newly formed diene to form a new stable allylic radical.
8 5 1
8 5 1
CO2H O CO2H
O 14 O
20 11 20
O 11 14

Three new stereogenic centres are created in this cyclization, at C8, C9, and C12, and all are under
full control both from the centre already present and from the way in which the molecule folds up
under the guidance of the enzyme. Now the allylic radical reacts with oxygen to give the unstable
hydroperoxide PGG2.
8 5 1 8 5 1
O CO2H O CO2H

O O 20
20 11 14
11 14
PGG2
O
O OH OH
This unstable prostaglandin has been isolated from sheep but, as it has a half-life of only 5 min-
utes, this is no trivial matter. Both weak OO bonds are now reduced enzymatically to give the first
reasonably stable compound, PGF2 (PG just means prostaglandin).
HO
8 5 1 8 5 1
O CO2H
CO2H
O 20
11 14 11 20
14
O HO
PGG2
OH OH
[H] prostaglandin F2

The best evidence for this pathway comes from labelled oxygen molecules. If a mixture of
16O16O (ordinary oxygen) and 18O18O is supplied to an organism making PGF2, the product
has either both black OHs as 16O or both as 18O but no molecules are formed with one 16O and one
18O. These isotopes are easily measured by mass spectrometry. Both black OHs then come from one

and the same molecule of oxygennot an obvious conclusion when you inspect the molecule of
PGF2, and thus good evidence for this pathway.

How aspirin works

CO2H
The enzyme that catalyses these remarkable reactions, drug. PGs also control acid
cyclooxygenase, is an important target for medicinal secretion in the stomach O
chemists. Inhibiting PG synthesis can bring about a and aspirin inhibits their aspirin
reduction of inflammation and pain. In fact, this is how synthesis there too so O
aspirin works. It was not, of course, designed to work that stomach ulceration can
way and its mode of action was discovered decades after result.
its use began. There is a price to pay for such a useful

Each of the other families of eicosanoidsthromboxanes and leukotrieneshas interesting

biosynthetic pathways too, but we will mention only one small detail. A completely different oxida-
tion enzyme, lipoxygenase, initiates a separate pathway leading to the leukotrienes, but the first steps
are very similar. They just occur elsewhere in the arachidonic acid molecule.
 Aromatic polyketides come in great variety 1433

11 O 1
8 CO2H
oxygen removes a hydrogen atom at C7
5
to form a stable conjugated radical
lipoxygenase
20
8 5 1 14
leukotriene LTA4
CO2H

14
20 8 5 1
11 O CO2H
arachidonic acid cyclo-
oxygenase O
14 20
oxygen removes a hydrogen atom at C13 11
to form a stable conjugated radical O
PGG2
OH
The initially formed radical is stabilized by two double bonds in the same way as that we have just
seen and reacts with oxygen in the same way again to give a trans-alkene and a new hydroperoxide.
O OH O OH
8 1 8 1
5 CO2H 5 CO2H
lipoxygenase

14 14
20 20
11 11
arachidonic acid

The next step is something quite new. No new CC bond is formed: instead, the diene attacks the
hydroperoxide to give an epoxide and a fully conjugated triene. The new double bond is cis this time,
which is what we should expect from the conformation we have been using. This is LTA4 and all the
other leukotrienes are made from this compound.
H Enz
O OH
8 1
5 CO2H
Enz B H
leukotriene LTA4
H 14
20
11 
The relatively recent discovery of these unstable molecules of incredibly powerful biological activ- Prostaglandins and leukotrienes have
appeared several times before in this
ity means that we by no means know all about them yet. They are very important to our well-being book, and you can read about aspects
and important medical advances are bound to follow from a better understanding. of their laboratory synthesis on pp.
686, 1229, 1268, and 883.

Aromatic polyketides come in great variety

The fatty acid pathway or, as we should call it now, the acyl polymalonate pathway, also gives rise to
an inexhaustible variety of aromatic and other compounds belonging to the family of the poly-
ketides. You saw in Chapter 50 how the shikimic acid pathway makes aromatic compounds but the
compounds below are from the polyketide route.
Me O O OH O
HO
CO2H O OH

HO
HO OH
OMe O OH OR
orsellinic acid
Me OH daunomycin (R = sugar)
alternariol

You might immediately be struck by the extent of oxygenation in these compounds. The shikimic
acid route produced ArC3 compounds with at most one OH group in the para position and others
1434 51 . Natural products

added ortho to that first OH group. Here we have multiple oxygenation with a predominant 1,3 pat-
tern. If we try to arrange an acyl polymalonate product to make orsellinic acid, this is what we shall
need.
Me Me Me
enol to ketone carbonyl O
CO2H tautomerism CO2H condensation CO2H

HO OH O O O O
orsellinic acid

Merely by writing ketones instead of phenols and doing one disconnection corresponding to a
simple carbonyl condensation, we have reached a possible starting material which is a typical acyl
polymalonate product without any reductions. This is what polyketides are. The fatty acids are
assembled with full reduction at each stage. Polyketides are assembled from the same process but
without full reduction; indeed, as the name polyketide suggests, many are made without any reduc-
tion at all. This is the biosynthesis of orsellinic acid.
CO2
O a tetraketide = 4 acetate units
O
O2C
SCoA O O O
rotate
SCoA CO2H
3 malonyl CoA
acetate =
starter unit acyl polymalonate intermediate

Me Me Me
O
CO2H cyclization CO2H enolization
CO2H

O O O O HO OH
orsellinic acid

This route has been demonstrated by feeding 13C-labelled malonyl CoA to a microorganism. The
orsellinic acid produced has three 13C atoms only, seen by an M + 3 peak in the mass spectrum. The
location of the labels can be proved by NMR. The starter unit, acetate, is not labelled.
Me
O
orsellinic acid synthase CO2H
O2C
SCoA
Penicillium species
HO OH
= 13C
orsellinic acid

As the polyketide chain is built up, any of the reductions or eliminations from fatty acid biosyn-
thesis can occur at any stage. The simple metabolite 6-methyl salicylic acid (6-MSA) is made in the
microorganism Penicillium patulum, and it could come from the same intermediate as orsellinic acid
with one reduction.
Me Me
O Me
O
CO2H CO2H
reduction cyclization CO2H

O O HO O
linear tetraketide 6-methyl salicylic acid

Reduction to the alcohol or to the unsaturated acid or ketone would give the right oxidation level
and could occur as the chain is built, after it is completed, or after cyclization. In fact, reduction to
 Aromatic polyketides come in great variety 1435

the conjugated unsaturated triketide occurs as the third acetate unit is added, just as the fatty acid
route would lead us to expect.
O O O O OH O O O
NADPH

SR SR SR
linear triketide

This intermediate cannot cyclize as it has a trans double bond and the ends cannot reach each
other. First, the double bond is moved out of conjugation with the COSR group, again as in the fatty
acids, except that here the new Z double bond moves into conjugation with the remaining keto
group.
H
O H O O
H
SR SR SR

H
O O O
E -alkene conjugated extended enol Z -alkene conjugated
with acid group with ketone group

Now the last chain extension occurs and the completed Z-tetraketide cyclizes to 6-methyl salicylic
acid. Chemically, we would prefer not to carry the unstable Z-enone through several steps, but
Nature controls these reactions very precisely.
O O O
chain redraw in new
extension conformation
SR SR
malonyl
O CoA O

Me O Me O Me O
O
aldol and
SR dehydration SR enolization SR

O O OH
This precise sequence was discovered only through very careful double labelling experiments and H
after the discovery of specific inhibitors for the enzyme. Since polyketides can be made from the acyl Me
HO
polymalonate pathway with or without reduction and elimination at any step, the number of possi- O
ble structures is vast. With more reduction, no aromatic ring can be formed: macrolide antibiotics H O
OH
such as brefeldin A come from this route. brefeldin A
If you examine this structure, you should be able to find a continuous carbon chain made from an
acetate starter unit and seven malonyl CoA units with full or partial reduction occurring after many
acylation steps.

Other starter units

So far we have started the chain with acetate, but many other starter units are used. Some
important groups of compounds use shikimic acid metabolites such as cinnamic acid (Chapter
50) as starter units. They include the widespread plant flavones and the anthocyanidin flower
pigments. O
CO2H
SCoA

HO HO
4-hydroxy cinnamic acid
from shikimic acid activated for polymalonate addition
1436 51 . Natural products

The most common sequence uses three malonyl CoA acylations followed by cyclization to a new
aromatic ring. The simplest type is exemplified by resveratrole, the compound in red wine that helps
to prevent heart disease. Each step in this sequence is a simple reaction that you have met before.
O O O O
starter unit
3 malonyl CoA redraw
SCoA
3 malonyl CoA
HO

O OH
aldol cyclization
dehydration
enolization
O decarboxylation
O OH

COSCoA
HO HO
resveratrole

A different cyclization leads to the flavones and anthocyanidins. Reaction of the stable enol from a
1,3-diketone with the thiol ester as electrophile results in acylation at carbon in the manner of the
Claisen ester condensation (Chapter 28) with loss of CoASH and the formation of a trihydroxyben-
zene ring.
H
O O O OH
starter unit cyclization
enolization

HO CoAS O HO HO OH
O

This cyclization is followed by a conjugate addition of an ortho phenolic OH group on to the

enone system. The product is a flavanone structure, which is always drawn a different way up to the
molecules we have just been discussing. Redrawing the last product shows the cyclization.
OH OH
H
HO O conjugate HO O
addition

OH O H OH O a flavanone

Aromatization of the central oxygen heterocycle by oxidation leads to the flavones, which are yel-
low or orange depending on their substituents. Dehydration leads to the red or blue anthocyanidins,
pigments of flowers and fruit. This important group of molecules also includes plant growth hor-
mones and defence compounds.
flavones OH anthocyanidins OH

RO O HO O

OH
OH O OH
R = H; naringenin, R = glucose; naringin pelargonidin, pigment of raspberries,
a bitter substance from grapefruit peel geraniums, and red grape skins
 Terpenes are volatile constituents of plant resins and essential oils 1437

Terpenes are volatile constituents of plant resins and essential

oils
Terpenes were originally named after turpentine, the volatile oil from pine trees used in oil painting,
whose major constituent is -pinene. The term was rather vaguely used for all the volatile oily com-
pounds, insoluble in water and usually with resiny smells from plants. The oils distilled from plants,
which often contain perfumery or flavouring materials, are called essential oils and these too contain
O
terpenes. Examples include camphor -pinene
camphor
from the camphor tree, used to pre-
serve clothes from moths, humulene OH
from hops, which helps to give beer phytol
its flavour, and phytol, found in
many plants.
You will notice that they are all aliphatic
compounds with a scattering of double bonds humulene
and rings, few functional groups, and an abun-
dance of methyl groups. A better definition
?
(that is, a biosynthetically based definition)
arose when it was noticed that all these com-
pounds have 5n carbon atoms. Pinene and cam-
phor are C10 compounds, humulene is C15, and isoprene
3 isoprene units humulene
phytol is C20. It seemed obvious that terpenes
were made from a C5 precursor and the favourite candidate was isoprene (2-methylbuta-1,3-diene)
as all these structures can be drawn by joining together 2-, 3-, or 4-isoprene skeletons end to end.
Humulene illustrates this idea.
In fact, this is not correct. Isoprene is not an intermediate, and the discovery of the true path-
way started when acetate was, rather surprisingly, found to be the original precursor for all terpenes.
The key intermediate is mevalonic acid, formed from three acetate units and usually isolated as its
lactone.
OH
O OH
3 HO2C
SCoA OH
O O
acetyl CoA mevalonic acid mevalonolactone

The first step is the Claisen ester condensation of two molecules of acetyl CoA, one acting as an
enol and the other as an electrophilic acylating agent to give acetoacetyl CoA. We saw the same reac-
tion in the biosynthesis of the pyrrolidine alkaloids earlier in this chapter.
Enz H B Enz OH
O O O
Claisen ester
H condensation
O
CoAS CoAS CoAS
enolization
CoAS acetoacetyl CoA

The third molecule of acetyl CoA also functions as a nucleophilic enol and attacks the keto group
of acetoacetyl CoA. This is not a Claisen ester condensationit is an aldol reaction between the enol
of a thiol ester and an electrophilic ketone.
OH
O O O OH O

CoAS
SCoA aldol CoAS SCoA
acetoacetyl CoA
1438 51 . Natural products

We have drawn the product with stereochemistry even though it is not chiral. This is because one
of the two enantiotopic thiol esters is hydrolysed while this intermediate is still bound to the enzyme,
so a single enantiomer of the half-acid/half-thiol ester results.
O OH O H2O O OH O

enantioselective
CoAS SCoA hydrolysis HO SCoA
enantiotopic thiol esters HMG-CoA
3(S)-3-hydroxy-3-methylglutaryl CoA

The remaining thiol ester is more electrophilic than the acid and can be reduced by the nucle-
ophilic hydride from NADPH. Just as in LiBH4 reductions of esters (Chapter 24), the reaction does
not stop at the aldehyde level, and two molecules of NADPH are used to make the alcohol. This is
mevalonic acid. OH
O OH O O OH O O OH
NADPH NADPH

HO SCoA HMG-CoA HO H HMG-CoA HO OH

reductase reductase O O
HMG-CoA
3(S)-3-hydroxy-3-methylglutaryl CoA mevalonic acid mevalonolactone

Different pathways; different reactivity

Acetyl CoA (as an enol) and malonyl CoA are both acylated acetoacetyl CoA. Malonyl CoA is acylated while acetyl CoA
by acetyl CoA as an electrophile, but the behaviour of the does the aldol reaction. This could be enzymatic control.
two nucleophiles is different when they react with

OH O O
O acylation
O
SCoA SR
SCoA
SCoA
O

terpene and fatty acid

alkaloid O O and polyketide
biosynthesis biosynthesis
OH
O O
SR
CoAS
SCoA
O

aldol acetoacetyl CoA acylation

O OH O O O O

CoAS SCoA SR
mevalonic acid presursor linear triketide

Mevalonic acid is indeed the true precursor of the terpenes but it is a C6 compound and so it must lose
a carbon atom to give the C5 precursor. The spare carbon atom becomes CO2 by an elimination reaction.
First, the primary alcohol is pyrophosphorylated with ATP (Chapter 49); then the CO2H group and the
tertiary alcohol are lost in a concerted elimination. We know it is concerted because labelling the
diastereotopic hydrogen atoms on the CH2CO2H group reveals that the elimination is stereospecific.
O OH O OH
 ATP elimination
HB
PP indicates the pyrophosphate HO OH HO OPP OPP
group transferred from ATP. mevalonic acid HA
H B HA
isopentenyl pyrophosphate
 Terpenes are volatile constituents of plant resins and essential oils 1439

So is isopentenyl pyrophosphate the C5 intermediate at last? Well, yes and no. There are actually
two closely related C5 intermediates, each of which has a specific and appropriate role in terpene
biosynthesis. Isopentenyl pyrophosphate is in equilibrium with dimethylallyl pyrophosphate by a OPP
isopentenyl pyrophosphate
simple allylic proton transfer.
This is again a concerted reaction and again we know that by proton labelling. One of the two
enantiotopic protons (HS in the diagram) is lost from the bottom face of the allylic CH2 group while
the new proton is added to the top face of the alkene. This is an anti rearrangement overall.
OPP
H+ added to top
face of bond H dimethylallyl pyrophosphate
H
HB HB
OPP OPP

HA H
S HR H A
HR

The stereochemical details are interesting in establishing the mechanism but not important to
remember. What is important is that the origin of the two methyl groups in dimethylallyl pyrophos-
phate is quite distinct and can easily be traced if you always draw the intermediates in the way we
have drawn them. We will now switch to 13C labelling to make the point.
O OH O OH
ATP H
HO OH HO OPP OPP
OPP
mevalonic acid H H
isopentenyl pyrophosphate dimethylallyl pyrophosphate

The two C5 intermediates now react with each other. The dimethylallyl pyrophosphate is
the better electrophile because it is allylic, and allylic compounds are good at both SN1 and SN2
reactions (Chapter 17). Isopentenyl pyrophosphate is the better nucleophile because it can react
through an unhindered primary carbon atom to produce a tertiary cation. This is what we have in
mind.
OPP

OPP
OPP
better (allylic) better (unhindered)
electrophile nucleophile stable tertiary cation

Though this idea reveals the thinking behind the reaction, in fact it does not go quite like this. The
product is one particular positional and geometrical isomer of an alkene and the cation is not an
intermediate. Indeed, the reaction is also stereospecific (discovered again by proton labelling, but we
will not give the rather complex details) and this too suggests a concerted process.
OPP

Though terpenes are made from
OPP OPP C5 units, they are classified in
H geranyl pyrophosphate C10 units. The monoterpenes are
the C10 compounds, the
Geranyl pyrophosphate is the starting point for all the monoterpenes. It is still an allylic
sesquiterpenes (sesqui is Latin
pyrophosphate and repeating the alkylation with another molecule of isopentenyl pyrophosphate for one-and-a-half) are the C15
gives farnesyl pyrophosphate, the starting point for the sesquiterpenes, and so on. compounds, the diterpenes are
the C20 compounds, and so on.
OPP

C15 compounds
OPP OPP
(sesquiterpenes)
H farnesyl pyrophosphate
As soon as we start to make typical cyclic monoterpenes from geranyl pyrophosphate we run into
a snag. We cannot cyclize geranyl pyrophosphate because it has a trans double bond! We could
cyclize the cis compound (neryl pyrophosphate), and it used to be thought that this was formed from
the trans compound as an intermediate.
1440 51 . Natural products

many of these names are derived from OPP

fragrant plant oils:
cyclization cyclization
impossible possible

OH
X
OPP
geraniol from geraniums

geranyl pyrophosphate neryl pyrophosphate

It is now known that Nature gets round this problem without making neryl pyrophosphate.
nerol from neroli oil
An allylic rearrangement occurs to move the pyrophosphate group to the tertiary centre. This is an
OH
unfavourable rearrangement thermodynamically and probably occurs via the allyl cation and cata-
lysed by Mg(II). There is no longer any geometry about the alkene. The molecule can
now rotate freely about a single bond and cyclization can occur. Even if only a small
OH amount of the rearranged allylic pyrophosphate is present, that can rearrange and more
farnesol, also present in neroli oil
can isomerize.
OPP
OPP OPP OPP
rotate cyclization
about bond possible


The product here is limonenea
terpene of the peel of citrus fruits. One
enantiomer occurs in lemon peelthe
other in orange peel. See Chapter 45.
limonene
H
More interesting compounds come from the cyclization of the first formed cation. The remaining
alkene can attack the cation to form what looks at first to be a very unstable compound but which is
actually a tertiary carbocation with the pinene skeleton.

H
=
H
-pinene

The camphor skeleton looks as though it might be formed by cyclization of the wrong end of the
alkene on to the cation. This would certainly give the right skeleton but the intermediate secondary
cation is rather unlikely.

?
=

camphor O

There is a better route. The more likely cation formed on the way to pinene could rearrange to the
camphor cation. This is a known chemical reaction and is a simple 1,2-shift of the kind discussed
in Chapter 37. However the new cation is formed, addition of water and oxidation would give
camphor.
1. hydration
2. oxidation
=

camphor O
 Steroids are metabolites of terpene origin 1441

In the sesquiterpene series, similar cyclizations lead to an amazing variety of products. After the
initial unfavourable allylic rearrangement of the pyrophosphate group, farnesyl pyrophosphate can
give a six-membered ring cation known as the bisabolyl cation.
OPP redraw
farnesyl pyrophosphate

OPP the bisabolyl cation

This cation does many things but it takes its name from the three fairly random proton losses that
lead to the -, -, and -bisabolenes.
loss of green
proton
H
loss of brown
H proton
-bisabolene

H -bisabolene loss of black

proton

-bisabolene

Many other reactions give even larger and more complex terpenes with a variety of functionaliza-
tion but we will treat only one group in detail. These compounds are so important to us that they are
given a different name.

Steroids are metabolites of terpene origin

Two types of human hormone are steroidalthe sex hormones such as oestradiol and testosterone
and the adrenal hormones such as cortisone. Cholesterol is a steroid too, as is vitamin D, derived
from ergosterol.
OH OH 
For reference, here is the numbering of the
H H steroid nucleus, not because we want you to
learn it, but because it is often used without
explanation in books and it is not obvious.
H H H H 21 22 24 26
O HO 18 20
testosterone 23 25
oestradiol 12 H
11 27
19 13 17
HO 1 H
9 C 14 D
16
O
2 15
10 8 H
O OH H A B H
HO 3 5 7 cholesterol
H 4 6

H H H H
O HO
cortisone ergosterol

All share the skeleton of four fused rings, three six-membered and one five-membered and con-
ventionally lettered AD. Beyond the ring stereochemistry and some common oxygenation patterns
they share little else. Some (such as the female sex hormones) have an aromatic A ring; some have
side-chains on the five-membered ring.
1442 51 . Natural products

At first glance, it is not at all clear that steroids are terpenoid in origin. The 5n numbers are absent
cholesterol is a C27 compound while the others variously have 20, 21, or 23 carbon atoms. Studies with
labelled mevalonic acid showed that cholesterol is terpenoid, and that it is formed from two molecules
of farnesyl pyrophosphate (2 C15 = C30 so three carbon atoms must be lost). Labelling of one or other
of the methyl groups (two experiments combined in one diagram!) showed that two of the green car-
bon atoms and one of the black carbon atoms were lost during the biosynthesis.
OH
HO2C
OH OPP OPP
mevalonic acid

2 farnesyl
OPP pyrophosphate
farnesyl pyrophosphate

H H
HO cholesterol

It is not obvious how the two farnesyl pyrophospate molecules could be combined to make the
steroid skeleton, and the chemistry involved is extraordinary and very interesting. The first clues came
from the discovery of the intermediates squalene and lanosterol. Squalene is obviously the farnesyl
pyrophosphate dimer we have been looking for while lanosterol looks like cholesterol but still has all 30
carbon atoms.

2 farnesyl
NADPH
pyrophosphate
squalene

squalene cyclization
H

H
HO lanosterol

The three carbon atoms that are lost from lanosterol (C30) in its conversion to cholesterol (C27)
are marked with brown arrows. Now at least we know which carbon atoms are lost. But many ques-
tions remain to be answered.
How does farnesyl pyrophosphate dimerize so that two electrophilic carbon atoms (CH2OPP)
join together?
Why does the formation of squalene require the reducing agent NADPH?
How does squalene cyclize to lanosterol so that the very odd labelling pattern can be achieved?
Where do the three lost carbon atoms go?
How is the sterochemistry controlled?
Before we tell you the answers, be warned: prepare for some surprises, and be ready to hold back out-
right disbelief!
 Steroids are metabolites of terpene origin 1443

The formation of squalene from farnesyl pyrophosphate

If the reducing agent NADPH is omitted from the cell preparation, squalene is not formed. Instead,
another farnesyl pyrophosphate dimer accumulatespresqualene pyrophosphatewhich has a
three-membered ring and in which we can see that the two molecules of farnesyl pyrophosphate are
joined in a slightly more rational way.
OPP OPP

farnesyl pyrophosphate farnesyl pyrophosphate

presqualene pyrophosphate H
OPP
Maybe its not so obvious that this is more rational! The first CC bond formation is quite straight-
forward. The alkene in the red molecule attacks the allylic pyrophosphate in the black molecule in a
simple SN2 reaction. The product is a stable carbocation. Only one CC bond remains to be formed to
close the three-membered ring and this occurs by the loss of a proton from the black molecule.
OPP
H H R

R R R
R We will abbreviate the long terpene
R side-chain to R from now on.
H
PPO OPP
OPP
This is a very remarkable reaction. Such reactions do not occur chemically: this biological one
occurs only because the molecule is held in the right shape by the enzyme and because the new ring is
three-membered. Three-membered rings are very easily formed but also very easily openedand
that is what happens to this ring. In the presence of NADPH, a series of rearrangements gives a series
of carbocations, the last of which is trapped by reduction.
The first step is the migration of one of the bonds (shown in green) of the three-membered ring to
displace the pyrophosphate leaving group, expand the ring to four-membered, and release some
strain. Now the cyclobutyl cation breaks down to give an open-chain allylic cation stabilized by one
of the alkenes. This is the cation that is reduced by NADPH.

H H NADP

R R
R R R
R R R
OPP squalene

If you follow this sequence backwards, you will see that the originally formed rational bond
(shown in green) is the one that migrated and is retained in squalene, while the second bond is
cleaved in the last step.
This may all seem far-fetched, but it happens in laboratory reactions too! Treatment of the sim-
plest cyclopropyl alcohol with HBr gives cyclobutyl bromide by a similar rearrangement.

HBr Br Br
H H
H H
OH OH2

In fact, cyclopropylmethyl compounds, cyclobutyl compounds, and homoallyl compounds are

all in equilibrium in acid solution and mixtures of products are often formed. The delocalized cation
1444 51 . Natural products

shown has been suggested as an intermediate. Make sure that you can draw mechanisms for each
starting material to give the intermediate cation and from the cation to each product.
cyclopropyl
methanol OH X
a

b
cyclobutanol OH X
b
a
c c
but-3-en-1-ol OH X
X

Squalene to lanosterol
The next step is simplethe epoxidation of one of the terminal double bondsbut it leads to two of the
most remarkable reactions in all of biological chemistry. Squalene is not chiral, but enzymatic epoxidation
of one of the enantiotopic alkenes gives a single enantiomer of the epoxide with just one stereogenic centre.

squalene

O2, NADPH epoxidase

O
squalene oxide

We will start now to draw squalene in a coiled up way as the next step is the polycyclization of the
epoxide. The basic reaction is best seen first in the flat, though we will draw the stereochemistry
immediately. The first alkene cyclizes on to the epoxide and then each remaining alkene cyclizes on
to the next to give a stable tertiary cation.

H
H

O HO
H
H
By analogy with what has gone before, you might now expect a tame hydration or reduction of
this cation. Nothing of the sort! A rearrangement occurs in which five consecutive 1,2-shifts are fol-
lowed by an elimination. Since this reaction organizes the backbone of the steroids, it is often called
the steroid backbone rearrangement.
H

H Me
H H
Me

H Me Me
HO HO
H H lanosterol
 Steroids are metabolites of terpene origin 1445

Finally, we have reached lanosterol. Now we will go back over these two steps and discuss them a
bit more. Consider first the regiochemistry of the cyclization. The epoxide opens in the way we
would expect to give positive charge at the more substituted carbon atom and then all the alkenes
attack through their less substituted end (again as we would expect to give positive charge at the
more substituted carbon atom)all except one. The third alkene cyclizes the wrong waythis is
presumably a result of the way the molecule is folded.
We learn much more about the folding by examining the stereochemistry of the product cation.
First, all of the stereochemistry of each alkene is faithfully reproduced in the product: the cyclization
is stereospecific. This is emphasized in colour in the diagram. The green stereochemistry arises
because the green Me and H were trans in the first alkene of squalene, the black Me and H trans in the
second, and the brown trans in the third. But what about the relationship between the green methyl
and the black H? Or between the black and brown methyls? These were determined by the folding
and the key observation is that all the relationships are trans except that between the green Me and
the black H. Now we can draw a conformation for the cyclization.

H chair
chair
H
Me H
Me Me Me
O H
H Me
R
HO H Me H
H Me
boat
When the transition state for a ring closure forms a chair then a trans relationship results. This is
the case for the black Me and brown Me. When a boat is formed a cis relationship results. This is the
case for the green Me and black H. Squalene folds up in a chairboatchair conformation and that
leads to the observed stereochemistry.
Next, we need to look at the stereochemistry of the rearrangement step. If we draw the product
cation as nearly as possible in the conformation of folded squalene, we will see which substitutents
are axial and which equatorial. H
cis relationship between Me R
and H stops chain of migrations Me
H H
Me Me
H Me
all anti migrations
HO
Me Me
H R HO
Me
H H lanosterol

Each group that migrates (black) is axial and is anti-periplanar to the one before so that each
migrating group does an SN2 reaction on the migration terminus with inversion. The chain stops
because of the cis relationship between the green Me and H in ring B and an elimination of the green
H is all that can happen.
The remainder of the biosynthesis of cholesterol requires various redox reactions and is a bit of an
anticlimax: the details are summarized in the scheme below.
H
lanosterol R cholesterol H
reduce alkene R

introduce alkene H
H
H

Me
H H
HO H HO
Me Me CO2
1446 51 . Natural products

Biomimetic synthesis: learning from Nature

When new and academic-looking reactions are discovered in the laboratory, it often seems only a
short time before they are found in nature as well. However, the development of polyolefin cycliza-
tion reactions in synthesis occurred by the reverse philosophyit was inspiration from Nature that
led W. S. Johnson to use the reactions in synthesis, including steroid synthesis. This is biomimetic
synthesis, a strategy that is bound to work provided we can just master the practical details.
There are quite a lot of differences between the chemical and the biochemical versions so farthe
chemical ones are less complex and less sophisticated but more versatile. The reactions are just
cyclizations without the backbone rearrangements. The most important points of difference are:
The cyclization is usually begun with a cation from treatment of a cyclic tertiary alcohol rather
than an epoxide
The cyclization sequence is terminated with an alkyne or an allyl silane rather then with simple alkene
The substituents are placed in the correct positions in the starting material as no rearrangement
follows cyclizations
The cyclizations are all stereospecific as in nature but the rings coil up in an all-chair fashion
rather than in a chairboatchair fashion as there is no enzyme to shape the molecule
The product cation is quenched by addition of water rather than loss of a proton
Here is one of Johnsons best examples which leads eventually to a biomimetic synthesis of the
human hormone progesterone. The cyclization occurs just on treatment of the tertiary alcohol with acid.

H
H
H H
HO

The first step is the formation of a symmetrical allyl cation, which then initiates the cyclization. The
next double bond is disubstituted so that it has no built-in regioselectivity but prefers to form a six-
membered rather than a five-membered ring B. The next double bond is trisubstituted and directs the
formation of a six-membered ring C. The alkyne, being linear, can reach only through its inner end and
so a five-membered ring D is formed. The resulting linear vinyl cation picks up a molecule of water to
give the ketone via its enol.

H2O

OH
H H
H2O
H H H H

H
 Problems 1447

The five-membered ring A is there to ensure efficient initiation of the cyclization by the symmet-
rical allylic cation. It can easily be opened with ozone and the product cyclized to progesterone.

O
O

H
H 1. O3
H H
H H 2. KOH
O
progesterone

The conformation of the molecule in the moment of cyclization can be seen easily by working
backwards from the product. The green dashed lines show new bonds that are being formed. All the
six-membered rings in the transition state are chairs and all the ring junctions trans. This is an
impressive result as there is no enzyme to help the molecule fold up in this way.
O
H

H H

By studying the chemistry that Nature uses in living things we can learn new reactions as well as new
ways in which to carry out known reactions. Many of the reactions in this chapter would be laughed at
by worldly wise chemists if they appeared in a research proposal, but they have been evolved over mil-
lions of years to do precise jobs under mild conditions. Humans have been doing complex organic
chemistry for only about a hundred years so that learning from Nature is one of the most important
ways in which organic chemistry is advancing at the beginning of the twenty-first century.

Problems
1. Assign each of these natural products to a general class (such as
amino acid metabolite, terpene, polyketide) explaining what
makes you choose that class. Then assign them to a more specific
part of the general class (for example, tetraketide, sesquiterpene).

2. Some compounds can arise from different sources in different

organisms. 2,5-Dihydroxybenzoic acid comes from shikimic acid
(Chapter 50) in Primula acaulis but from acetate in Penicillium
species. Outline details.
CO2H HO CO2H

Primula acaulis
OH
HO OH
Penicillium
OH
shikimic acid CH3CO2H
1448 51 . Natural products

3. The piperidine alkaloid pelletierine was mentioned in the HO HO

chapter but full details of its biosynthesis were not given. There
follows an outline of the intermediates and reagents used. Fill in NH NH
HO HO
the details. Pyridoxal chemistry is discussed in Chapter 50. CO2H
HO HO
H2N CO2H pyridoxal phosphate

lysine H NH2 [RNH2 is pyridoxamine]

HO HO
O O
6. Concentrate now on the biosynthesis of scytalone in the first
+
R CoAS problem. You should have identified it as a pentaketide. Now
N N N
consider how many different ways the pentaketide chain might be
H H acetoacetyl-CoA
H folded to give scytalone.
O O

N N
H H
CO2H pelletierine

4. The rather similar alkaloids anabasine and anatabine come

from different biosynthetic pathways. Labelling experiments out-
lined below show the origin of one carbon atom from lysine and
others from nicotinic acid. Suggest detailed pathways. (Hint. 7. This question concerns the biosynthesis of stephanine, another
Nicotinic acid and the intermediate you have been using in compound mentioned in Problem 1. You should have deduced
Problem 3 in the biosynthesis of the piperidine alkaloid are both that it is a benzylisoquinoline alkaloid. Now suggest a biosynthesis
electrophilic at position 2. You also need an intermediate derived from orientaline.
from nicotinic acid which is nucleophilic at position 3. The bio-
synthesis involves reduction.) MeO MeO

CO2H H2N CO2H NMe NMe

HO HO
H ? H
H NH2 MeO
N
nicotinic acid lysine

HO OMe
orientaline stephanine

N N 8. Suggest a biosynthesis of olivetol.

H H
OH
N N
anatabine anabasine
olivetol
5. The three steps in the biosynthesis of papaverine set out
below involve pyridoxal (or pyridoxamine). Write detailed HO
mechanisms.
CO2H CO2H

NH2 O
HO HO
tyrosine

HO CO2H HO CO2H

NH2 O
HO HO
DOPA
 Problems 1449

9. Tetrahydrocannabinol, the major psychoactive compound in
marijuana, is derived in the Cannabis plant from olivetol and
geranyl pyrophosphate. Details of the pathway are unknown.
Make some suggestions and outline a labelling experiment to
establish whether your suggestions are correct.
olivetol
12. Borneol, camphene, and -pinene are made in nature from
geranyl pyrophosphate. The biosynthesis of -pinene and the
related camphor is described in the chapter. In the laboratory
bornyl chloride and camphene can be made from -pinene by the
reactions described below. Give mechanisms for these reactions
and say whether you consider them to be biomimetic.

OPP Cannabis sativa

HCl KOH
geranyl pyrophosphate

ArOH
OH
H Cl
-pinene bornyl chloride camphene
tetrahydro-
cannabinol
H 13. Suggest a biosynthetic route to the monoterpene chrys-
O anthemic acid that uses a reaction similar to the formation of
squalene in steroid biosynthesis.
10. Both humulene, mentioned in the chapter, and caryophyllene Pyrethrum plants
are made in nature from farnesyl pyrophosphate in different plants.
OPP
Suggest detailed pathways. How do the enzymes control which CO2H
product is formed? chrysanthemic acid
How could the same route also lead to the natural products
H yomogi alcohol and artemisia ketone?
O

HO
H
yomogi alcohol artemisia ketone
humulene caryophyllene
14. In the chapter we suggested that you could detect an acetate
starter unit and seven malonate additional units in the skeleton of
11. Abietic acid is formed in nature from mevalonate via the
brefeldin. Give the mechanism of the addition of the first malonyl
intermediates shown. Give some more details of the cyclization
CoA unit to acetate. Draw out the structure of the complete acyl
and rearrangement steps and compare this route with the bio-
polymalonate chain and state clearly what must happen to each
synthesis of the steroids.
section of it (reduction, elimination, etc.) to get brefeldin A.
OPP OPP H

HO Me
O
H O
OH
brefeldin A

H 15. This chemical experiment aims to imitate the biosynthesis of

terpenes. A mixture of products results. Draw a mechanism for the
H reaction. To what extent is it biomimetic, and what can the natural
system do better?
LiClO4
+
OAc OAc HOAc
OAc OAc
H H
+

H H OAc
HO2C abietic acid
+
D.E. Vance and J.E. Vance (Ed~,.) Biochemistl3' (~['Lipid.~, Lil~Olnoteins aml Memhl~mes (4th E~hl.)
~-~ 20(12 Elsevier Science B.V. All rights reserved

CHAPTER 4

Lipid metabolism in plants

Katherine M. S c h m i d i and John B. O h l r o g g e 2
/ Department ()fBiologieal Sciences, Butler UniversiO; Indianapolis, IN 46208-3485, USA,
Tel.: +1 (317) 940-9956: Fax: +1 (317) 940-9519; E-mail: kschmid@butlel:edu
-"Department of Plant Biolog); Michigan State Univelwio; East Lansing, MI 48824, USA,
Tel.: + l (517) 353-0611; Fax: +1 (517) 353-1926: E-mail: ohlrogge@pilot.msu.edu

1. Introduction

Plants produce the majority of the world's lipids, and most animals, including humans,
depend on these lipids as a major source of calories and essential fatty acids. Like other
eukaryotes, plants require lipids for membrane biogenesis, as signal molecules, and as a
form of stored carbon and energy. In addition, soft tissues and bark each have distinctive
protective lipids that help prevent desiccation and infection. To what extent does the
biochemistry of plant lipid metabolism resemble that in other organisms? This chapter
mentions a number of similarities, but emphasizes aspects unique to plants. Major
differences between lipid metabolism in plants and other organisms are summarized in
Table 1.
The presence of chloroplasts and related organelles in plants has a profound effect
on both gross lipid composition and the flow of lipid within the cell. Fatty acid

Table 1
Comparison of plant, mammalian, and bacterial lipid metabolism

Higher plants Mammals E. colt

FatO, acid synthase
Structure Type II Type I Type 11
(multicomponent) (multifunctional) (multicomponent)
Location Plastids Cytosol Cytosol
Acetyl-CoA Multisubunit and Multifunctional Multisubunit
carboxylase(s) multifunctional
Primary desaturase substrates
A9 18 : 0-ACP 18 : 1-CoA none
co-6 18 : 1 on glycerolipids none none
(I)-3 18 : 2 on glycerolipids none nolle
Primary substrate(s) acyl-ACP and acyl-CoA acyl-CoA acyI-ACP
for phosphatidic acid
synthesis
Prominent bilayer Galactolipid > Phospholipid Phospholipid
lipids phospholipid
Main 13-oxidation Provides acetyl-CoA tot Provides acetyl-CoA Provides acetyl-CoA
function glyoxylate cycle for TCA cycle for TCA cycle
94

synthesis occurs not in the cytosol as in animals and fungi, but in the chloroplast
and other plastids. Acyl groups must then be distributed to multiple compartments,
and the complex interactions between competing pathways are a major focus of plant
lipid biochemists. It is also significant that the lipid bilayers of chloroplasts are largely
composed of galactolipids rather than phospholipids. As a result, galactolipids are the
predominant acyl lipids in green tissues and probably on earth.
Plant lipids also have a substantial impact on the world economy and human nutrition.
More than three-quarters of the edible and industrial oils marketed annually are derived
from seed and fruit triacylglycerols. These figures are particularly impressive given that,
on a whole organism basis, plants store more carbon as carbohydrate than as lipid. Since
plants are not mobile, and since photosynthesis provides fixed carbon on a regular basis,
plant requirements for storage lipid as an efficient, light weight energy reserve are less
acute than those of animals.
Finally, hundreds of genes required for plant lipid biosynthesis, utilization and
turnover have now been cloned. In addition to providing valuable information on
enzyme structure and function, these genes are being exploited to design new, more
valuable plant oils. The coordination of lipid metabolic genes with each other and with
their potential regulators may also become better understood, as DNA microarray and
other genomic technologies mature.

2. Plant lipid geography

2.1. Plastids
Although all eukaryotic cells have much in common, the ultrastructure of a plant
cell differs from that of the typical mammalian cell in three major ways. The plasma
membrane of plant cells is shielded by the cellulosic cell wall, preventing lysis in the
naturally hypotonic environment but making preparation of cell fractions more difficult.
The nucleus, cytosol and organelles are pressed against the cell wall by the tonoplast,
the membrane of the large, central vacuole that can occupy 80% or more of the cell's
volume. Finally, all living plant cells contain one or more types of plastid.
The plastids are a family of organelles containing the same genetic material, a
circular chromosome present in multiple copies. Young or undifferentiated cells contain
tiny proplastids that, depending on the tissue, may differentiate into photosynthetic
chloroplasts, carotenoid-rich chromoplasts, or any of several varieties of colorless
leucoplasts, including plastids specialized for starch storage [1]. These different types
of plastids, which may be interconverted in vivo, have varying amounts of internal
membrane but invariably are bounded by two membranes. The internal structure of
chloroplasts is dominated by the flattened green membrane sacks known as thylakoids.
The thylakoid membranes contain chlorophyll and are the site of the light reactions of
photosynthesis.
As noted above, chloroplasts and other plastids are enriched in galactolipids (Fig. 1).
They also contain a unique sulfolipid, sulfoquinovosyldiacylglycerol, whose head
group is a modified galactose. The phospholipid components of plastids are less
 95

Monogalactosyldiacylglycerol Pea Chloroplasts %

Thylakoids 56
Inner membrane 64
CH2OH
Outer membrane 1
Daffodil chromoplasts 63
Cauliflower proplastids 31
Potato leucoplasts 14
II
0
Pea Chloroplasts %
Digalactosyldiacylglycerol Thylakoids 32
CH2OH Inner membrane 31
Outer membrane 40
I Y O,r-O-CHz Daffodil chromoplasts 18
Cauliflower proplastids 29
Potato leucoplasts 45

L.o o

Pea Chloroplasts %
Sulfoc]uinorosyldiacylglycerol Thylakoids 4
c H2SO~ Inner membrane 4
Outer membrane 4
0 0 t Daffodil chromoplasts 5
Cauliflower proplastids 6
Potato leucoplasts 5

o
Fig. I. Composition of plastid membranes. Figures given are percentages (as % of total lipid) of the pictured
lipid in the membranes specified. Data from Harwood, J.L. (1980) Plant acyl lipids: structure, distribution
and analysis. In: EK. Stumpf (Ed.) The Biochemistry of Plants, Vol. 4, Academic Press, pp. 2-56 and
Sparace, S.A., Kleppinger-Sparace, K.F. (1993) Metabolism in non-photosynthetic, non-oilseed tissues. In:
T.S. Moore Jr. (Ed.) Lipid Metabolism in Plants, Boca Raton, FL, CRC Press, pp. 569-589.

abundant. Phosphatidylglycerol is the most prominent phospholipid contributor to the

thylakoid membrane system of chloroplasts (but < 10% of the glycerolipids), whereas
most of the limited phosphatidylcholine of chloroplasts is associated with their outer
membrane.
The synthetic capabilities of plastids and other plant organelles are summarized
in Table 2. Representatives of each type of plastid have been isolated and found to
incorporate acetate into long-chain fatty acids, to desaturate 1 8 : 0 to 1 8 : 1 , and to
assemble phosphatidic acid and galactolipids. Chloroplasts have also been shown to
synthesize phosphatidylglycerol, including molecular species containing the unusual
trans-3-hexadecenoic acid at the 2-position. In addition to the components normally
retained within the plastids, large quantities of fatty acids, particularly 18 : l and 16 : 0.
96

._=

~._>,

"'o "o "o

V

o 0

~ ~ -~ ~.~
,.~ t-

o c

8
=~,;g

~ ~ ~.~= ~- ~g

e-
o
N o
;~ _ =%
g

e~
g ~ 2
~.. r..)
c~
~ g d ~5 a P

o
z

o '~ ~ z
 97

..= ~

2~
2 2 ~

.E E

2~

e-

e~

.-=

~2
? o
2~ E
. ~

~
~'~
o
z

aD

.& !
98

are produced for export to the rest of the cell. An acyl-CoA synthetase identified
on the outer membrane of plastids is thought to facilitate release of acyl groups
into the cytosol. It should also be noted that, although net lipid traffic is from the
plastids, this organelle can likewise be on the receiving end. In addition to small
quantities of plastidial phospholipids whose head groups are not known to arise in
that compartment, there may be considerable flow of extraplastidially constructed
diacylglycerol backbones into the galactolipid synthesis pathway. The quantitative
significance of this backflow depends on the plant species, as will be discussed in
Section 5.3.

2.2. Endoplasmic reticulum and lipid bodies

The endoplasmic reticulum has traditionally been viewed as the primary source of
phospholipids in plant cells. With the exception of cardiolipin, all of the common
phospholipids can be produced by microsomal fractions. The endoplasmic reticulum
also serves as the major site of fatty acid diversification. Although plastids do have the
ability to synthesize polyunsaturated fatty acids, they are formed on acyl lipid substrates
and are not typically exported. Thus, the endoplasmic reticulum desaturation pathways
are of particular importance for developing seeds that store large quantities of 18 : 2
and 18:3. Pathways for the production of unusual fatty acids found primarily in seed
oils have likewise been described in microsomes. Not surprisingly, the endoplasmic
reticulum also appears to be instrumental in the formation of the triacylglycerols
themselves and the lipid bodies in which they are stored (Section 7).

2.3. Mitochondria

Next to plastids and the endoplasmic reticulum, the plant mitochondrion is probably the
organelle investigated the most thoroughly with respect to lipid metabolism. Its ability to
synthesize phosphatidylglycerol and cardiolipin is well established. Although most fatty
acids for mitochondrial membranes are imported from the plastids or the ER, recently
mitochondria have been shown to synthesize low levels of fatty acids from malonate.
Octanoate is a major product of this pathway and serves as a precursor for the lipoic
acid cofactor needed by glycine decarboxylase and pyruvate dehydrogenase [2].

2.4. Glyoxysomes and peroxisomes

A discussion of the compartmentation of lipids and their metabolism would be in-

complete without reference to the organelles responsible for fatty acid oxidation. As
in mammals, there is evidence for both mitochondrial and peroxisomal [3-oxidation
systems. In plants, the peroxisomal system appears to be the more significant [3].
Unlike mammals, plants can use the peroxisomal enzymes to catabolize long-chain fatty
acids all the way to acetyl-CoA. Under certain conditions, such as oilseed germination,
plants also differentiate specialized peroxisomes called glyoxysomes. In addition to
the 13-oxidation pathway, glyoxysomes contain the enzymes of the glyoxylate cycle, a
pathway absent from animals. Plants are able to use the glyoxylate cycle to feed the
 99

acetyl-CoA produced by [3-oxidation into carbohydrate synthesis. Since plants cannot

transport fatty acids over long distances, only this conversion of acetate to sucrose,
which can be transported by the plant's vascular system, makes lipid a practical carbon
reserve for the growing shoots and roots of seedlings.

3. Acyl-ACP synthesis in plants"

Fatty acid synthases may be classified into two groups. 'Type I' fatty acid synthases
are characterized by the large, multifunctional proteins typical of yeast and mammals
(Chapter 6), while 'Type II' synthases of most prokaryotes are dissociable into compo-
nents that catalyze individual reactions (Chapter 3). Plants, while certainly themselves
eukaryotic, appear to have inherited a Type II fatty acid synthase from the photosynthetic
prokaryotes from which plastids originated.
The ground-breaking studies of Overath and Stumpf in 1964 [4] established not only
that the constituents of the avocado fatty acid synthesis system could be dissociated
and reconstituted, but also that the heat stable fraction from E. coli we now know
as acyl carrier protein (ACP) could replace the corresponding fraction from avocado.
Plant ACPs share both extensive sequence homology and significant elements of three-
dimensional structure with their bacterial counterparts. In plants, this small, acidic
protein not only holds the growing acyl chain during fatty acid synthesis, but also is
required for synthesis of monounsaturated fatty acids and plastidial glycerolipids.

3.1. Components of plant fato~ acid synthase

Fatty acid synthase is generally defined as including all polypeptides required for
the conversion of acetyl and malonyl-CoA to the corresponding ACP derivatives, the
acyl-ACP elongation cycle diagrammed in Chapter 3, and the cleavage of ACP from
completed fatty acids by enzymes termed thioesterases or acyl-ACP hydrolases [5]. All
components of fatty acid synthase occur in plastids, although they are encoded in the
nuclear genome and synthesized on cytosolic ribosomes. Most of the 8-10 enzymes of
the pathway are soluble when isolated from homogenates. Nevertheless, some evidence
suggests that at least ACP and some subunits of acetyl-CoA carboxylase may be
associated with the plastid membranes.
Despite the presence of acetyl-CoA : ACP acyltransferase activity in plant fatty acid
synthase preparations, acetyl-ACP does not appear to play a major role in plant fatty
acid synthesis (Jaworski, 1992). Instead, the first condensation takes place between
acetyl-CoA and malonyl-ACE This reaction is catalyzed by 13-ketoacyl-ACP synthase
III, one of three ketoacyl synthases in plant systems (Fig. 2). The acetoacetyl-ACP
product then undergoes the standard reduction-dehydration-reduction sequence to
produce 4 : 0-ACE the initial substrate of ketoacyl-ACP synthase I. KAS I is responsible
for the condensations in each elongation cycle up through that producing 16:0-ACE
The third ketoacyl synthase, KAS II, is dedicated to the final plastidial elongation, that
of 16 : 0-ACP to 18 : 0-ACE
100

acetyI-CoA 6malonyI-ACP malonyI-ACP

~ 4:O-ACP~ 16:O-ACP~ 18:0-ACP
maI~yI'ACP{~.j) (102 ~6CO2 ~.jJ CO2
malonyI-CoA 6cycles
Fig. 2. Contribution of the three ketoacyl synthases (KASI, II and llI) to fatty acid elongation. Each circle
represents one round of the elongation cycle catalyzed by ketoacyl-ACP synthase, enoyl-ACP reductase,
hydroxyacyl-ACPdehydrase, and acyl-ACPreductase.

3.2. Desaturation of acyl-ACPs

The major components of the long-chain acyl-ACP pool in most plant tissues are 16:0-
ACE 18 : 0-ACP and 18 : l-ACE This finding highlights the importance of stearoyl-ACP
desaturase, the plastidial enzyme responsible for Ag-desaturation in plants. In contrast to
the desaturation system of Escherichia coli (Chapter 3), the plant enzyme introduces the
double bond directly to the AV-position. Unlike yeast and mammalian A%desaturases,
it is a soluble enzyme and is specific for acyl-ACPs rather than acyl-CoAs (McKeon,
1982). In recent years, work on stearoyl-ACP desaturase has progressed rapidly and
is now providing a more detailed understanding of the fundamental mechanisms of
oxygenic fatty acid desaturation. Genes for the enzyme have been cloned from a number
of species, and the structure of the castor bean AV-desaturase has been determined to
2.4 ,~ resolution. A combination of the crystal structure and spectroscopic methods
has revealed two identical monomers, each with an active site containing a diiron-oxo
cluster. Reduction of the iron by ferredoxin leads to its binding of molecular oxygen.
The resulting complex ultimately removes electrons at the AV-position, resulting in
double bond formation [6].
Although the most common unsaturated fatty acids in plants are derived from oleic
acid, a wide range of unusual fatty acids are found in the seed oils of different species. Di-
vergent plastid acyl-ACP desaturases have been shown to account for some of this diver-
sity. For example, Coriandrum sativum achieves seed oils rich in A t'- 18 : 1 (petroselinic
acid) by desaturation of 16:0 at the A4-position followed by elongation, while Thun-
bergia alata attains a similar oil by direct A%desaturation of 18:0. Several structure-
function relationships suggested by unusual desaturases have been tested in the acyl-ACP
desaturase system. Shortening the acyl binding pocket of the A 9-18:0 desaturase by al-
tering a single amino acid as in Doxantha unguis-cati shifts substrate specificity in favor
of 16 : 0-ACP (Cahoon, 1998). In addition, a set of five specific amino acids suggested by
the Thunbergia gene transforms the castor bean AV-desaturase to a A%desaturase, while
enzymes with certain subsets of the five amino acids can desaturate at either position [6].

3.3. AcyI-ACP thioesterases

Among prokaryotes, all acyl groups exiting the dissociable fatty acid synthase are
transferred directly from ACP to polar lipids. However, plants must also release
sufficient fatty acid from ACP to supply the extraplastidial compartments. Since the
 101

typical chloroplast exports primarily 18:1 and 16:0, the same fatty acids that comprise
the greatest fraction of long-chain acyl-ACPs, it might be assumed that a relatively
non-specific thioesterase releases 16- and 18-carbon fatty acids from ACE However,
molecular and biochemical analyses of cloned plant thioesterases suggest that plants
possess individual thioesterases with specificity either for 18 : 1 or for one or more
saturated fatty acids [7]. The most prominent thioesterase in most plants has a strong
preference for 18: 1-ACE making 18 : 1 the fatty acid most available for extraplastidial
glycerolipid synthesis. In contrast, mangosteen, a plant with seed oil particularly high in
18 : 0, contains an 18 : 0-ACP thioesterase gene that has been used to engineer rapeseed
with high 18:0 content (Hawkins, 1998).
Plants that synthesize certain unusual fatty acids have additional or modified
thioesterases. For example, several plant species that produce storage oils contain-
ing large amounts of 8- to 14-carbon acyl chains contain thioesterases specific for
those chain lengths. By removing acyl groups from ACP prematurely, the medium-
chain thioesterases simultaneously prevent their further elongation and release them for
triacylglycerol synthesis outside the plastids. In addition, both the standard A9-18:1
thioesterase and a A6-18 : 1 thioesterase have been purified from the A6-18 : 1 accumu-
lating coriander plant. Thus plants, by regulating expression of different thioesterases,
can both fine tune and radically modify the exported fatty acid pool.

4. Acetyl-CoA carboxylase and control o f fatty acid synthesis

4.1. Two)~brms of aceo'l-CoA carboxylase

The malonyl-CoA that supplies all but two carbons per fatty acid is produced from acetyl-
CoA and carbon dioxide by acetyl-CoA carboxylase (ACC). In plants, malonyl-CoA for
fatty acid synthesis is apparently provided by a plastid ACC, while a cytosolic ACC
contributes malonyl units for fatty acid elongation as well as synthesis of flavonoids,
polyketides, and other metabolites. As with fatty acid synthase, ACC forms may be cate-
gorized either as 'eukaryotic' enzymes, which are dimers of a multifunctional polypeptide
(Chapter 6), or 'prokaryotic' enzymes, which are heteromers of biotin carboxyl carrier
protein, biotin carboxylase, and two subunits of carboxyltransferase (Chapter 3). In the
grass family, both plastids and cytosol house 'eukaryotic' enzymes. However, dicots and
monocots other than grasses appear to have both forms, with the 'eukaryotic' form lim-
ited primarily to the cytosol, and 'prokaryotic' enzymes dominating in the plastids [8].
Assembly of the 'prokaryotic' form requires participation of both the nuclear genome,
which encodes biotin carboxyl carrier protein, biotin carboxylase, and the c~-subunit of
carboxyltransferase, and the plastid genome, which has retained the gene for the car-
boxyltransferase 13-subunit, perhaps due to a requirement for RNA editing (Sasaki, 2000).

4.2. Acetvl-CoA carboxylase as control point

In other kingdoms, ACC is a major control point for fatty acid biosynthesis. Although
the mechanisms acting in plants are incompletely characterized, there is evidence that
102

plant ACCs are also tightly regulated [9]. For example, both redox regulation via
thioredoxin and phosphorylation of the carboxyltransferase have been implicated in
up-regulation of the 'prokaryotic' ACC by light. Conversely, feedback inhibition is
observed at the level of ACC when tobacco cell cultures are given exogenous fatty acids.
Due to its impact on the rate of fatty acid synthesis, ACC is considered a promising
target in oilseed improvement programs, and some increases in oil content have been
obtained by engineering a cytosolic ACC gene to be expressed in rapeseed plastids.

5. Phosphatidic acid synthesis: 'prokaryotic' and 'eukaryotic'

acyltransferases

Since phosphatidic acid serves as a precursor of phospholipids, galactolipids and

triacylglycerols, it is not surprising that its own synthesis has been reported in four
plant compartments: plastids, endoplasmic reticulum, mitochondria, and Golgi bodies.
In each case, esterification of the first acyl group to the sn-1 position of glycerol-3-
phosphate is catalyzed by glycerol-3-phosphate acyltransferase. Lysophosphatidic acid
acyltransferase then completes the synthesis by acylating the sn-2 position. However,
plastidial and extraplastidial acyltransferases show distinct differences in structure and
specificity.

5.1. Plastidial acyltran,sferases

In the plastids, acyltransferases provide a direct route for acyl groups from ACP to enter
membrane lipids. Since this is the standard pathway in E. coli and cyanobacteria, both
the enzymes of phosphatidic acid synthesis in plastids and the glycerolipid backbones
they produce are termed 'prokaryotic.' In both chloroplasts and non-green plastids,
the glycerol-3-phosphate acyltransferase is a soluble enzyme that, unlike the E. coli
enzyme, shows preference for 18 : 1-ACP over 16 : 0-ACE The lysophosphatidic acid
acyltransferase, which is a component of the inner envelope of plastids, is extremely
selective for 16: 0-ACE The presence of a 16-carbon fatty acid at the 2-position is
therefore considered diagnostic for lipids synthesized in the plastids.

5.2. Extraplastidial acyhransferases

At least superficially, the mitochondrial and Golgi acyltransferase activities resemble
those of the better studied endoplasmic reticulum system. All three compartments
have membrane-bound glycerol-3-phosphate and lysophosphatidic acid acyltransferases
that utilize acyl-CoA substrates. In the endoplasmic reticulum, which is quantitatively
the most significant of the extraplastidial sites for phosphatidic acid synthesis, saturated
fatty acids are almost entirely excluded from the sn-2 position. The glycerol-3-phosphate
acyltransferase is less selective, but, due to substrate availability, more often fills
the sn-1 position with 18:1 than with 16:0. It is therefore possible to judge the
relative contributions of the prokaryotic and eukaryotic pathways by comparing the
proportions of eukaryotic 18/18 or 16/18 glycerolipids with prokaryotic 18/16 or
 103

16/16 glycerolipids [5]. Mitochondrial lysophosphatidic acid acyltransferase, which

shows little selectivity for chain length or unsaturation, can usually be ignored in
discussions of lipid flow.

5.3. 16: 3 and 18 : 3 plants

Relative fluxes through the prokaryotic and eukaryotic pathways vary between organ-
isms and among tissues. Plastids have the potential to use phosphatidic acid from the
prokaryotic pathway for all of their glycerolipid syntheses. However, not all plants do
so; in some cases, the prokaryotic acyl chain arrangement is found only in plastidial
phosphatidylglycerol, whereas galactolipids are derived from diacylglycerol imported
to the plastids from the ER. As indicated above, the eukaryotic acyltransferases of
the endoplasmic reticulum produce substantially more 18/18 than 16/18 lipids, and
it is chiefly the 18/18 units that are assembled into galactolipids by phmts with a
minor prokaryotic pathway. Because galactolipids become highly unsaturated, plants
that import diacylglycerol for galactolipids are rich in 18 : 3 and are called 18 : 3 plants.
Species in which most galactolipid is derived from the prokaryotic 18/16 or 16/16
diacylglycerol contain substantial 16 : 3 and are known as 16 : 3 plants.
Kunst et al. [10] have demonstrated that a 16 : 3 plant, Arabidopsis thaliana, may be
converted to a de facto 18 : 3 plant by a single mutation in plastidial glycerol-3-phosphate
acyltransferase. Under these conditions, 16 : 3 content is reduced dramatically, and when
isolated chloroplasts are labeled with glycerol-3-phosphate, only phosphatidylglycerol
is labeled. Nevertheless, the percentage of galactolipids in mutant plants is practically
identical to that in wild-type plants, emphasizing the ability of plants to compensate
for reduction of the prokaryotic pathway. Other studies in mutants have confirmed that
plants have an amazing capacity to adapt to many, but not all, perturbations of lipid
metabolism (Section 11).

6. Glycerolipid synthesis p a t h w a y s

In plants, glycerolipid biosynthesis involves a complex web of reactions distributed

among multiple compartments [11,12]. As in mammals, the synthesis of individual
glycerolipids is initiated either by the formation of CDP-diacylglycerol from phospha-
tidic acid and CTP, or by cleavage of phosphate from phosphatidic acid to produce
diacylglycerol.
CTP : phosphatidate cytidylyltransferase has been observed in plastids, mitochondria
and endoplasmic reticulum. In all three compartments, the CDP-diacylglycerol derived
from phosphatidic acid is used in the synthesis of phosphatidylglycerol; in mitochondria,
the reaction of phosphatidylglycerol with a second CDP-diacylglycerol then produces
cardiolipin. The endoplasmic reticulum can also incorporate CDP-diacylglycerol into
phosphatidylinositol and phosphatidylserine.
Phosphatidic acid phosphatase is present in the same three compartments. In the
endoplasmic reticulum and mitochondria, diacylglycerol combines with CDP-ethanol-
amine or CDP-choline to produce phosphatidylethanolamine or phosphatidylcholine,
104

respectively. Although separate enzymes catalyze ethanolamine and choline transfer

in animals and yeast, there are indications that a single aminoalcoholphosphotrans-
ferase may be responsible in plants. Flux into phosphatidylcholine is at least partially
determined by regulation of the phosphocholine cytidylyltransferase that generates
CDP-choline. The production of CDP-ethanolamine from phosphoethanolamine is less
well studied, but is also considered a probable regulatory step. In addition, there
is clear evidence that phosphothanolamine can be methylated to monomethylphos-
phoethanolamine, dimethylphosphoethanolamine, and phosphocholine, and that this
pathway is inhibited by exogenous choline at the initial methylation step. The methyla-
tion of phosphoethanolamine in plants is frequently contrasted with the methylation of
phosphatidylethanolamine to phosphatidylcholine in animals and yeast. In general, no
significant methylation of phosphatidylethanolamine itself occurs in plants. However,
phosphatidylmonomethylethanolamine is synthesized and converted to phosphatidyl-
choline [ 11 ].
The diacylglycerol released in plastids reacts either with UDP-galactose or with
UDP-sulfoquinovose to generate sulfolipid or monogalactosyldiacylglycerol. Synthesis
of digalactosyldiacylglycerol from the latter may use either UDP-galactose or a second
monogalactosyldiacylglycerol as the galactose donor [12] (Kelly, 2002). There is
also evidence for some extraplastidial galactolipid synthesis in plants suffering from
phospholipid deficiency (Hartel, 2000).

6.1. Glycerolipids as substrates for desaturation

In addition to the soluble acyl-ACP desaturases, plants contain a number of membrane-

bound enzymes that desaturate fatty acids while they are esterified within glycerolipids
[5,6]. The recent cloning and characterization of these desaturases is of great interest to
the scientific community because the products of the membrane-bound systems include
A<),12_18 : 2 and A '~12'i5-18 : 3, both of which are essential to the human diet, and thought
to play a major role in human health and disease.
Once again, separate pathways occur in plastids and endoplasmic reticulum, although,
as should he evident from the discussion above, fatty acids from the endoplasmic
reticulum may make their way back to the plastids. Clarification of the number of
desaturases involved in plant lipid metabolism and isolation of their genes has been
greatly assisted by the isolation of a large number of mutants in A. thaliana, a
small weed of the mustard family used as a model organism by plant geneticists and
molecular biologists. Briefly, three membrane-bound desaturation sequences are evident
in Arabidopsis [5,6].
(1) In chloroplasts, 16:0 at the 2-position of phosphatidylglycerol is desaturated to
trans-3-16 : 1. This desaturase is most likely encoded by the FAD4 gene.
(2) Plastids are able to convert 18 : 1 to 18 : 3 and 16:0 to 16 : 3 using a combination of
three membrane-bound desaturases. One of them, encoded by FAD5, is relatively
specific for the conversion of 16:0 on monogalactosyldiacylglycerol to A7-16: 1.
This 16 : 1 and A'J-18 : 1 may then be given a second and third double bond by
the FAD6 and the FAD7 or FAD8 gene products, respectively. The latter two
desaturases are less selective in their choice of glycerolipid substrate, and will
 10 5

accept appropriate fatty acids on phosphatidylglycerol, sulfolipid, or either of the

major galactolipids.
(3) In the endoplasmic reticulum, 18 : 1 esterified to phosphatidylcholine or occasion-
ally phosphatidylethanolamine may be desaturated to 18:2 by FAD2 and to 18:3
by FAD3.
It should be noted that fatty acids entering one of the multistep desaturation pathways
listed above are not necessarily committed to completing that set of reactions. It is par-
ticularly common for 18 : 2 to be an end product of endoplasmic reticulum desaturation.
This 18 : 2 may remain in phospholipid, be incorporated into triacylglycerol, or enter the
galactolipid pathway and receive a third double bond in the chloroplast.

7. Lipid storage in plants

A plant stores reserve material in its seeds in order to allow seedling growth of the next
generation until photosynthetic capacity can be established. The three major storage
materials are oil, protein and carbohydrate, and almost all seeds contain some of each.
However, their proportions vary greatly. For example, the amount of oil in different
species may range from as little as 1-2% in grasses such as wheat, to as much as 60%
of the total dry weight of the castor seed. With the exception of the jojoba plant, which
accumulates wax esters in seeds, plants store oil as triacylglycerol (TAG).

7.1. Lipid body structure and biogenesis

In the mature seed, TAG is stored in densely packed lipid bodies, which are roughly
spherical in shape with an average diameter of 1 g m (Fig. 3) [13]. This size does
not change during seed development, and accumulation of oil is accompanied by an
increase in the number of lipid bodies. The very large number of lipid bodies in an
oilseed cell (often >1000) contrasts strikingly with animal adipose tissue where oil
droplets produced in the cytosol can coalesce into a few or only one droplet. The
plant lipid bodies appear to be surrounded by a lipid monolayer in which the polar
headgroups face the cytosol, while the non-polar acyl groups are associated with the
non-polar TAG within. The membranes of isolated lipid bodies, which comprise less
than 5% of a lipid body's weight, contain both phospholipids and characteristic proteins
known as caleosins and oleosins. The recently discovered caleosins are calcium binding
proteins whose function is still unknown. Oleosins are small (15-26 kDa) proteins
that are believed to preserve individual lipid bodies as discrete entities. Desiccated
seeds lacking oleosins undergo lipid body fusion and cell disruption when rehydrated
(Leprince, 1998). The cDNAs encoding many oleosins have been cloned and each has a
sequence encoding a totally hydrophobic domain of 68-74 amino acids which is likely
to be the longest hydrophobic sequence found in any organism. Structurally, oleosins are
roughly analogous to the animal apolipoproteins which coat the surface of lipid droplets
during their transport between tissues.
When a seed germinates, the TAG stored in the lipid bodies becomes the substrate
for lipases. In at least some cases, peroxidation of polyunsaturated fatty acids by a lipid
106

Fig. 3. Thin-sectional view of cells in a cotyledon of a developing cotton embryo harvested 42 days after
anthesis. The cells are densely packed with lipid bodies and several large storage protein bodies (dark).
Magnification x9000. Photo courtesy of Richard Trelease, Arizona State University.

body lipoxygenase precedes the release of fatty acids from TAG [14]. Typically lipases
and lipid body lipoxygenase are active only after germination is triggered by imbibition
and other environmental signals. Fatty acids released by the lipid bodies are further
metabolized through the 13-oxidation pathway and glyoxylate cycle in the glyoxysomes
(Section 2.4).

7.2. Seed triacylglycerols often contain unusual fatty acids

The structural glycerolipids of all plant membranes contain predominantly 5 fatty

acids (18 : 1, 18 : 2, 18 : 3, 16:0, and in some species, 16: 3). However, the fatty acid
composition of storage oils varies far more than in membrane glycerolipids. Altogether
more than 300 different fatty acids are known to occur in seed TAG [15]. Chain
length may range from less than 8 to over 22 carbons. The position and number of
double bonds may also be unusual, and hydroxy, epoxy or other functional groups can
modify the acyl chain. Many of the different fatty acid structures, including hydroxy,
epoxy, acetylenic and conjugated varieties, are now known to originate from minor
modifications in the amino acid sequence of the oleate desaturase. For example, only
 107
Table 3
Some unusual fatty acids produced in plant seeds

Fatty acid type Example Major sources Major or potential Approx. US

USeS market size,
10~
Medium chain Lauric acid (12:0) Coconut, palm kernal Soaps, detergents, 350
surfactants
Long chain Erucic acid (22 : 1) Rapeseed Lubricants, anti-slip 100
agents
Epoxy Vernolic acid Epoxidized soybean Plasticizers 70
18:IA'~epoxyl2,13 oil, Vernonia
Hydroxy Ricinoleic acid Castor bean Coatings, lubricants 80
18:1A~,12OH
Acetylenic Crepenynic acid Crepisjoetida Polymers -
18: 2A'),12yne
Cyclopropene Sterculic acid 19:1 Sterculia foetida Lubricants, polymers -
Conjugated Parinaric hnpatiens balsamina Coatings -
18 : 4A';c 1 ltl 3t15c
Trienoic Linolenic acid Flax Paints, varnishes, 45
(c~18 : 3) coatings
Wax esters Jojoba oil Jojoba Lubricants, cosmetics 10

4 amino acid changes have been shown to convert a desaturase into a hydroxylase
[16].
The reason for the great diversity in plant storage oils is unknown. However, the
special physical or chemical properties of the 'unusual' plant fatty acids have been
exploited for centuries. In fact, approximately one-third of all vegetable oil is used for
non-food purposes (Table 3). Reading the ingredients of a soap or shampoo container
reveals one of the major end uses of high lauric acid specialty plant oils. Other major
applications include the use of erucic acid (22 : 1) derivatives to provide lubricants and
as a coating for plastic films. Hydroxy fatty acids from the castor bean have over 100
industrial applications including plastic and lubricant manufacture. As discussed further
below, the ability of genetic engineering to transfer genes for some unusual fatty acid
production from exotic wild species to high yielding oil crops is now providing the
ability to produce new renewable agricultural products and to replace feedstocks derived
from petroleum.

7.3. The pathway of triao, lglycerol biosynthesis

As in animal tissues, it has been suggested that TAGs are produced by a relatively simple
four reaction pathway. According to this model, phosphatidic acid is synthesized by the
extraplastidial pathway (Section 5) and dephosphorylated to diacylglycerol. A third fatty
acid is then transferred from CoA to the vacant third hydroxyl of the diacylglycerol,
producing TAG. This last and single committed step is catalyzed by diacylglycerol
acyltransferase (Fig. 4, reaction 6). Although plants possess all of the enzymes for the
reactions above, the assembly of three fatty acids onto a glycerol backbone is not always
108

CDP-
18:1 choline CMP
~ 18:1

x_Z
3 E
18:1
18:1
acyl group I- 18:1
modification |lTZ~ ~

(~- choline 4 >
^ /CMP~
..... \
W~1" chline %CDP- \
2t
r /
El::
~-~ ""ch~ine18:1F
~" E~18:1 17

E2
it
OH
Fig. 4. Pathway depicting how flux through phosphatidylcholine (product of reaction 3) can promote acyl
group diversity in plant triacylglycerols. Production of18:2 (boxed) at the sn-2 position and its transfer
to TAG is used as a sample modification. Other fatty acid alterations may be substituted. Enzymes: I
= glycerol-3-phosphate:acyl-CoA acyltransferase: 2 = lysophosphatidic acid:acyl-CoA acyltransferase;
3 -- CTP:phosphatidate cytidylyltransferase; 4 = T6 18:l-desaturase or other fatty acid modifying
enzyme: 5 = phospholipid:diacylglycerol acyltransferase; 6 = diacylglycerol acyhransferase; 7 = acyl-
CoA : phosphatidylcholine acyltransferase or phospholipase plus acyl-CoA synthetase.

as straightforward as suggested by the above pathway. In many oilseeds, pulse-chase

labeling has revealed that fatty acids reach TAG only after passing through phos-
phatidylcholine, (or phosphatidylethanolamine to a lesser extent). Given the range of
desaturation and other modification reactions that can take place on phosphatidylcholine,
transit through this phospholipid helps to explain some of the fatty acid diversity in TAG.
Fatty acids from phosphatidylcholine may become available for TAG synthesis in
several ways [17]. In some plants, a phospholipid :diacylglycerol acyltransferase pro-
duces TAG by direct transfer of a fatty acid from the 2-position of phospholipid to
diacylglycerol (Fig. 4, reaction 5) [18]. The second mechanism by which phosphatidyl-
choline can participate in TAG synthesis is by donation of its entire diacylglycerol
unit. In many plants, the synthesis of phosphatidylcholine from diacylglycerol and
CDP-choline appears to be rapidly reversible. As shown in Fig. 4 (reaction 3), this
activity of the choline phosphotransferase can allow diacylglycerol moieties modified
on phosphatidylcholine to be incorporated into TAG. Finally, a fatty acid removed
from phosphatidylcholine may subsequently be used for TAG synthesis. Such an 'acyl
exchange' may provide acyl-CoA either by the combined reverse and forward re-
actions of an acyl-CoA :phosphatidylcholine acyltransferase, or by a combination of
phospholipase and acyl-CoA synthase.

7.4. Challenges in understanding triacylglycerol synthesis

Although the basic reactions of TAG biosynthesis have been determined, several
fundamental and potentially related questions persist. As highlighted above, TAG and
membrane lipids frequently have radically different fatty acid compositions. How do
 109

plants control which fatty acids are stored in TAG as opposed to which fatty acids are
restricted to membranes? Are unusual fatty acids excluded from membranes because
their physical and chemical idiosyncracies would perturb membrane fluidity or other
physical characteristics? Is TAG synthesis spatially distinct from the synthesis of
membrane lipids, or do enzyme specificities dictate the partitioning of fatty acid species
among glycerolipids? Although all of these factors may be significant, selectivity by
enzymes such as phospholipid:diacylglycerol acyltransferase for unusual fatty acids,
and editing of unusual fatty acids from phospholipids, are currently the best documented
[17,18].

8. Protective lipids
In plants, tissues are protected against desiccation and pathogens by both a cuticle and
epicuticular wax. The cuticle itself contains some wax, but is anchored to the plant
cell wall by cutin, a complex polyester of fatty acid derivatives with a wide range of
oxygen-containing functional groups (Fig. 5). 16- and 18-carbon dicarboxylic acids with
one or more hydroxyl groups are particularly common [19]. Arabidopsis transformed
with cutinase from a fungal plant pathogen not only develops a leaky cuticle, but also
suffers from fusions between leaves and flower parts (Sieber, 2000).
Surface waxes are complex mixtures including a range of very long-chain alkanes,
aldehydes and ketones as well as wax esters and their building blocks. Although only
one mutation in cutin formation has been identified (Wellesen, 2001), visual screening
of plant surfaces has allowed isolation of several wax mutants blocked in malonyl-CoA-
dependent elongation of fatty acids, decarbonylation and reduction (Post-Beittanmiller,
1996). The cDNA of an elongase required for extension of wax acyl units beyond 24
carbons has been cloned, and resembles the condensing enzymes involved in synthesis of
erucic acid in seed oil and wax ester precursors in jojoba seeds (Millar, 1999). A cDNA
encoding the wax synthase of jojoba seeds has been cloned and successfully expressed
in Arabidopsis (Lardizabal, 2000), and may provide clues to the corresponding genes
for surface waxes.
Bark, wound callus, and specialized tissues such as the endodermis that controls
entry into the root vascular system, have walls lined with suberin. Suberin, like cutin,
is a polyester incorporating fatty acids enriched in carboxyl and hydroxyl groups. In
addition to placement on the inner surface of cell walls rather than outside, the tough,
waterproof suberin differs from cutin in its preference for longer fatty acids and in its
incorporation of large amounts of phenylpropanoids [19].

9. Sterol, isoprenoid and sphingolipid biosynthesis

In the plant kingdom, isoprenoids represent the most diverse range of natural products
with over 25,000 lipophilic structures known, ranging from small, volatile compounds
to rubber. Quantitatively, the photosynthetic apparatus is probably the primary consumer
of isoprenoids, since carotenoids, plastoquinone and the phytol tail of chlorophyll
110
-

! l O/~oC~Nc15H31

~
/ 0 ~ c%o
O ' # ' C ~ O~C~(CH2)8~C/(CH2)s~O~

~H

n
Fig. 5. Model showing some of the linkages in cuzin, after Kolattukudy [19]. Note the cross-linking made
possible by mono- and dihydroxy-fattyacids.

all belong to this group. Given that vital plant hormones such as gibberellin and
abscisic acid, plus many defensive compounds, are isoprenoids, the early steps of
this pathway have been studied intensely. However, surprisingly, it was not until
the late 1990s that researchers realized that plants have two very different pathways
for production of isopentenyl pyrophosphate, the five-carbon central precursor of all
isoprenoids [20]. For several decades it was known that, as in other organisms, plants
join three molecules of acetyl-CoA to form hydroxymethylglutaryl-CoA followed by
the highly regulated reduction of that compound to mevalonic acid. Furthermore,
plants contain multiple well-studied hydroxymethylglutaryl-CoA reductase genes that
 111

are differentially expressed during development and in response to such stimuli as light,
wounding and infection. It was incorrectly suspected that this 'mevalonate' pathway
was localized in both cytosol and plastids and produced all classes of isoprenoids. The
story has now been clarified [20] with the discovery that plastids produce isopentenyl
pyrophosphate by a 'non-mevalonate' pathway that begins with the condensation
of pyruvate with glyceraldehyde-3 phosphate to produce 1-deoxy-D-xylulose-5-R At
least three additional enzymes are required to produce isopentenyl pyrophosphate in
the plastids. In parallel to work in plants, this pathway has also been demonstrated in
bacteria and algae. The non-mevalonate pathway in plastids is responsible for production
of the classic plant photosynthetic isoprenoids such as phytols and carotenoids, as well
as mono- and diterpenes.
The mevalonate pathway in the cytosol is responsible for biosynthesis of sterols,
sesquiterpenes and triterpenoids. After conversion of mevalonic acid to isopentenyl
pyrophosphate, three C5 units can be joined head to tail to produce a C15 compound,
farnesyl pyrophosphate. Two farnesyl pyrophosphates are then united head to head to
form squalene, the progenitor of the C30 isoprenoids from which sterols are derived. The
plant squalene synthetase, like its mammalian homologue, is found in the endoplasmic
reticulum and the reaction proceeds via a presqualene pyrophosphate intermediate. In
the last step prior to cyclization, squalene is converted to squalene 2,3-epoxide.
It is also in the cyclization step that photosynthetic and non-photosynthetic organisms
diverge. Whereas animals and fungi produce lanosterol, organisms with a photosynthetic
heritage produce cycloartenol. Despite the differences in the cyclization product, there
is substantial conservation between the enzymes responsible, with 34% identity between
an Arabidopsis cycloartenol synthase and lanosterol synthase.
A complex series of reactions including opening of the cyclopropane ring, double
bond formation and isomerization, demethylation of ring carbons, and methylation of
the side chain result in formation of a number of different plant sterols. Sitosterol
is the most common plant sterol (Fig. 6); however, plants normally contain mixtures
of sterols whose proportions differ from tissue to tissue. In addition, sterol esters,
sterol glycosides, and acylated sterol glycosides are common plant constituents whose
physiological significance is under scrutiny. Both cold adaptation and pathogenesis
drastically alter free and derivatized sterol pools. Plants also produce a steroid hormone,
brassinolide, required for both light-induced development and fertility. Interestingly, the
gene for a 5c~-reductase in the brassinolide pathway can complement the corresponding
reductase in the testosterone pathway [21 ].
Sphingolipids are usually considered minor constituents of plant lipids, accounting
for 5% or less of most lipid extracts. This fact, and the more complex methods needed
for their identification and characterization have resulted in a comparative lack of
information on plant sphingolipid biosynthesis and function. Nevertheless, sphingo-
lipids make up a substantial proportion (25% or more) of the composition of plasma
and tonoplast membranes, with the glucosylceramides constituting the largest fraction.
In addition, difficulties in extraction and analysis may have led to underestimates of
sphingolipid contents. As in animals (Chapter 14), sphingolipid biosynthesis begins
with condensation of palmitoyl-CoA with serine to form 3-keto-sphinganine, with the
enzyme from A. thaliana showing 68% similarity to the human homologue (Tamura,
112

C22 C24

Fig. 6. Sitosterol (24c~-ethylcholesterol),shown here, is the most common plant sterol, but plants generally
contain complex mixtures of sterols. Other prominent phytosterols differ from sitosterol as follows.
Campesterol, 24c~-methyl;stigmasterol, C22 double bond; dihydrospinasterol, move double bond from C5
to C7; spinasterol, move C5 double bond to C7, add C22 double bond; dihydrobrassicasterol, 24~-methyl;
brassicasterol, 2413-methyl,add C22 double bond.

2000). Reduction of 3-keto-sphinganine by an NADPH-dependent reaction yields

sphinganine, and cDNAs encoding enzymes for the subsequent C4-hydroxylation and
A~-desaturation of sphinganine have recently been cloned [22]. Less information is
available on the synthesis of ceramides and glucosylceramides, although there is
evidence that either sterol glucoside or UDP-glucose could serve as the glucose donor
for the latter [23]. The finding that sphingolipids serve as signal molecules in plants,
particularly in the defensive apoptosis induced by some plant pathogens, has begun to
stimulate interest in this relatively neglected area of plant lipid biochemistry.

10. Oxylipins as plant hormones

Jasmonate is one of several lipid-derived plant growth regulators referred to as oxylipins

[24]. The structure and biosynthesis of jasmonate have intrigued plant biologists because
of the parallels to some eicosanoids (Chapter 13), which are central to inflammatory
responses and other physiological processes in mammals, in plants, jasmonate derives
from c~-linolenic, presumably released from membrane lipids by a phospholipase A.
The linolenic acid is oxidized by lipoxygenase and the resulting products, 9-hydro-
peroxylinolenic acid or 13-hydroperoxylinolenic acid may be further metabolized by
one of three routes to produce a wide variety of oxylipin (Fig. 7). The pathways
by which 13-hydroperoxylinoleic acid may be metabolized include hydroperoxide
lyase catalyzed scission of the trans-ll,12-double bond to produce a C6 aldehyde,
cis-3-hexenal, and the C12 compound, 12-oxo-cis-9-dodecenoic acid. The acid is
subsequently metabolized to 12-oxo-trans-lO-dodecenoic acid, the wound hormone
traumatin. The enzyme hydroperoxide dehydratase (allene oxide synthase) catalyzes
the dehydration of hydroperoxides to unstable allene oxides that readily decompose to
form a 9,12-ketol or a 12,13-ketol. The allene oxide of 13-hydroperoxylinolenic acid
may also be converted by allene oxide cyclase to 12-oxo-phytodienoic acid which can
 113

OHC'~/'~v/~v.~COOH
~~x~,C/~OOH traumatin

,8:3 l
OHC~~~/COOH
t
OOH
12-oxo-cis-dodecenoic acid

~COOH cis-3-hexenal~ C H O

13-hydroperoxy-18:3
HO O

OH 12,13-ketol
O OH
12,13-epoxy-18:3
~'~V~~ ~~jCOOH
O
10 9,12-ketol

~ C O O H
12-oxo-phytodienoicacid

jasmonicacid
Fig. 7. Metabolism of 18:3 to oxylipins, l, lipoxygenase; 2, allenoxide synthase; 3, allene oxide cyclase; 4,
12-oxo-phytodienoicacid reductase, [3-oxidation;5, hydroperoxide lyase.

be further metabolized to 7-iso-jasmonic acid. In the last few years, it has become
clear that jasmonate is a key component of a wound-signaling pathway that allows
plants to protect themselves against insect attack. When experimentally applied to
114

plants at low concentrations, jasmonate leads to the induction of protease inhibitors

and other defense compounds. Furthermore, mutants of tomato and Arabidopsis that
are deficient in jasmonate synthesis are much more susceptible to insect damage. In
addition to jasmonate, a number of the other oxylipins have been reported to function
as signal molecules. In particular, the oxylipin traumatin has been suggested to trigger
cell division at the site of wounds, leading to the development of a protective callus.
The lipoxygenase product 13-hydroxylinolenic acid triggers phytoalexin production.
Similarly, C6-C 10 alkenals act as volatile elicitors of a defense response in cotton.

11. Progress in plant lipid research: the value o f mutants

Biochemical approaches toward understanding plant lipid biosynthesis and function

provided much of the information summarized above. However, in recent years, the
isolation of mutants in plant lipid metabolism has been extremely fruitful in providing
new insights and new methods for gene isolation. Much of the progress in the genetic
dissection of plant lipid metabolism has come from the extensive studies of Somerville,
Browse and coworkers with A. thaliana, which has one of the smallest genomes (113
megabases) of higher plants. By using gas chromatography to screen several thousand
randomly selected plants from a mutagenized population, Somerville and Browse were
able to obtain an extensive collection of mutants showing altered leaf or seed fatty
acid compositions. As described above, these mutants included isolates instrumental
in confirming the relationships between prokaryotic and eukaryotic phosphatidic acid
synthesis and in the analysis and cloning of membrane-bound desaturases.

11.1. Mutants in lipid metabolism have helped link lipid structure and function

Two other major benefits have been derived from the Arabidopsis lipid mutants. First,
the physiological effects of the mutations have provided the opportunity to evaluate the
relationships between lipid structure and function. There has been a long-term assump-
tion, based on the strong association of high levels of polyunsaturated fatty acids with
photosynthetic membranes and the conservation of this property among higher and lower
plant species, that these fatty acids must be essential to photosynthesis. However, many
attempts to understand the relationships between membrane fatty acid composition and
cell physiology or photosynthesis have led to equivocal results. The isolation of mutants
totally lacking certain unsaturated fatty acids has now provided much more convincing
evaluations of their function and indeed, the results have forced re-evaluation of several
previous hypotheses. For example, A3-trans-hexadecenoate is an unusual plant fatty acid
which is associated with phosphatidylglycerol of chloroplast membranes, is evolutionar-
ily conserved, and is synthesized in coordination with the assembly of the photosynthetic
apparatus. These observations led to the suggestion that A3-trans-hexadecenoate is a
highly essential component of photosynthesis. However, mutants which contain no
detectable A~-trans-hexadecenoate grow as well as wild type plants, and all photo-
synthetic parameters examined appear normal (Browse, 1985). A minor difference in
stability of some components of the photosystem can be detected by polyacrylamide
 115

gel electrophoresis. It has been concluded from such analyses that, although A~-trans -
hexadecenoate may facilitate assembly of the light harvesting complex into thylakoids, a
more obvious phenotype could be restricted to certain unusual environmental conditions.
As mentioned above, a number of mutants blocked in the production of polyunsat-
urated fatty acid biosynthesis have also been isolated (Table 4). Because leaves have
desaturases both in chloroplasts and in the endoplasmic reticulum, single mutations
lead only to partial reduction of polyunsaturated fatty acid levels. Again, these mutants
grow normally under most conditions and have normal photosynthetic parameters. How-
ever, several alterations in physiology are observed including changes in chloroplast
ultrastructure, a reduction in the cross-sectional area of chloroplasts, and increased sta-
bility to thermal disruption of photosynthesis. Moreover, whereas wild-type Arabidopsis
plants are chilling resistant and can reproduce normally at temperatures as low as 6C,
the mutants blocked in plastidial A 7 (f~ld5) and co6 ~ d 6 ) desaturation become chlorotic
at 6C and show a 20-30% reduction in growth rate relative to the wild-type plant.
The fad2 mutants, in which the endoplasmic reticulum o)-6 desaturase is blocked, are
even more sensitive to 6C and die if left at this temperature for several days. These
results demonstrate that polyunsaturated fatty acids are essential fbr maintaining cellular
function and plant viability at low temperatures.
While most mutants which are reduced only in polyunsaturated fatty acid synthesis
grow and develop normally at 22C, a high-stearate mutant with 14% 18:0 is strikingly
abnormal (Fig. 8) [26]. Many cell types fail to expand, resulting in mutant plants
growing to less than one-tenth the size of wild-type. At higher growth temperatures
(36C), the effects are less dramatic, suggesting that the physical properties or fluidity
of highly saturated membranes are less impaired under these conditions.
Other large scale alterations in membrane fatty acid composition and phenotypes
have been obtained by creation of multiple-mutant lines. When mutants defective in
endoplasmic reticulum co6-desaturase were crossed with plants defective in the plastid
co6-desaturase, double mutants could be recovered only on sucrose-supplemented media.
The sucrose grown plants, which contained less than 6% polyunsaturated fatty acids,
were chlorotic and unable to carry out photosynthesis but otherwise remarkably normal.
These results, while confirming the significance of polyunsaturated fatty acids to
photosynthesis, indicate that the vast majority of membrane functions can proceed
despite drastically reduced levels of polyunsaturates [27].
Triunsaturated fatty acids normally dominate chloroplast membranes and thus are the
most abundant fatty acids on the planet. By constructing a triple mutant of fad3, fad7
and fad8, it has been possible to eliminate triunsaturated fatty acids from Arabidopsis
without affecting 16:2 and 18 : 2 production (McConn, 1996). Surprisingly, these plants
are able to grow, photosynthesize, and even flower. However, they are male sterile and
therefore cannot produce seeds. This observation led to the discovery of a very different
role for jasmonate. This mutant cannot synthesize jasmonate because it lacks the 18 : 3
precursor, and the plants are male-sterile because pollen does not mature properly and is
not released from the anthers. Application of jasmonate or linolenic acid to the anthers
restores fertility, demonstrating that jasmonate is a key signal in pollen development.
This result is a dramatic example of a change in fatty acid composition having a very
specific effect on an essential developmental and tissue specific reproductive process.
116

.~'~ _, .~~8= a

" o~ rj

o ~

-t

.&
-..!

.b. r=

~ < .2
~.-~

--> ~
o
E

t-,r.,
8

"Fo ~- ~ _~ ~ ~:
o

- - v v ;~ :~ ,,--, ~o

"6

e-,

.=_
,-j

-8

- ."2. ."2_ ~ .2..2_ -~ ."2, ~.

oo

F~
 117

Fig. 8. Increased stearic acid causes severe dwarfing of Arabidopsis. A wild-type Arabidopsis plant (left) is
compared to the.lab2 mutant (right) in which leaf stearic acid content has increased from 1 to 14%. Photo
courtesy of John Browse, Washington State University.

11.2. Arabidopsis mutants have allowed cloning of desaturases and elongases

A. thaliana mutants have also provided a means of cloning genes difficult to isolate by
other methods. As in other kingdoms, the membrane bound enzymes of plants have been
notoriously difficult to purify and characterize. However, cDNAs or genes encoding a
number of these enzymes have now been isolated using molecular genetic strategies
based on mutations. Several approaches have been successful. A cDNA encoding the
co-3 desaturase which converts linoleic to linolenic acid was isolated in 1992 by Arondel
et al. [28] after a mutation leading to the loss of function was genetically mapped and
a yeast artificial clone corresponding to this region was selected. The genomic region
was then used to screen a cDNA library, and some of the clones which hybridized to
the yeast artificial chromosome had sequence similarity to cyanobacterial desaturases.
These clones subsequently were shown to complement the loss of 18 : 3.
Gene 'tagging' strategies have also proved enormously valuable in identifying
clones of membrane bound enzymes. Arabidopsis genes can be 'tagged' by insertional
118

inactivation when T-DNA from Agrobacterium tumefaciens inserts randomly in the

genome. When a promising phenotype is observed, the inactivated gene can be identified
by probing with the T-DNA sequence. This method was used to identify the o)-6
desaturase required for the 18 : 1 to 18:2 conversion in endoplasmic reticulum [25]. In
addition, transposon tagging led to the cloning of a gene which controls elongation of
oleic acid to 20 : 1 and 22 : 1 in developing Arabidopsis seeds [29]. Since no membrane-
bound fatty acid elongase had ever been completely purified or cloned from a eukaryotic
organism, the finding that this gene encodes a 60-kDa condensing enzyme provided the
first direct evidence that membrane fatty acid elongation is not catalyzed by a type I
multifunctional fatty acid synthase.

12. Design of new plant oils

In recent years, tremendous progress has occurred not only in the isolation of many
plant lipid biosynthetic genes, but also in the use of these genes to manipulate plant oil
composition. As shown in Table 5, both substantial changes in seed oil composition and
introduction of unusual fatty acids to heterologous species have been achieved. Progress
in this area has been accelerated by several industrial laboratories whose goal has been
the production of higher value oilseeds.

Table 5
Some examples of genetic engineering of plant lipid metabolism

Modification achieved Enzyme engineered Source of gene Plant transformed

Lauric acid production AcyI-ACP thioesterase California Bay Rapeseed
Increased stearic acid Antisense of stearoyl-ACP Rapeseed Rapeseed
desaturase
Reduced saturated fatty Stearoyl-CoA desaturase Rat, yeast Tobacco
acids
AcyI-ACP thioesterase Soybean Soybean
Reduced 18 : 3 0.)-3 Desaturase Soybean, canola Soybean, rapeseed
Altered lauric acid 1-Acyl-glycerol-3-phosphate Coconut Rapeseed
distribution in TAG acyltransferase
Altered cold tolerance Acyl-ACP : glycerol-3-phosphate E. coil, squash, Tobacco,
acyltransferase Arabidopsis Arabidopsis
Petroselinic acid production AcyI-ACP desaturase Coriander Tobacco
Increased linolenic acid o)-3 Desaturase Arabidopsis Arabidopsis
Cyclopropane fatty acid Cyclopropane synthase E. c'oli. Tnbacco
production Sterculia
y-Linolenic acid production Linolenic A~-desaturase Synechocystis Tobacco
Increased long-chain fatty Fatty acid elongase Jojoba Rapeseed
acids
Wax ester synthesis Wax ester synthase, fatty acid Jojoba Arabidopsis
reductase, fatty acid elongase
 119

Soybean

Corn sz D
Sunflower

Canola

Olive

Palm

Beef tallow

0 10 20 30 40 50 60 70 80 90 100
Fig. 9. Fatty acid composition of dietary vegetableoils and beef tallow. The values shown represent typical
compositions of varieties grown commercially. Lines modified substantially through breeding or genetic
engineering are available for soybean, canola, corn and sunflower.

12.1. Design of new edible oils

12.1.1. Improvements in nutritional value and stability of vegetable oils

Vegetable oils have gradually replaced animal fats as the major source of lipids in
human diets and now constitute 15-25% of total caloric intake by industrialized nations.
As shown in Fig. 9, vegetable oils display a wide range in the relative proportions
of saturated and unsaturated fatty acid acids although in the United States, up to
80% of vegetable oil consumed is soybean oil making it the single largest source of
calories in American diets. Most nutritionists recommend a reduction in saturated fat
content in diets, and genetic engineering of plant oils can substantially help achieve this
goal. Most of the saturated fatty acid in common plant oils is palmitic acid, and its
occurrence is largely related to the action of a palmitoyl-ACP thioesterase. Reduction
of the expression of this activity in transgenic soybean using co-suppression has led to
decreases of the palmitic acid content from 15 to 6%.
A very high stability liquid soybean oil with low saturated fatty acids was also
produced in soybean by suppression of the oleoyl-desaturase (Kinney, 1996). Oleic acid
content was increased up to 89%, 18:2 content reduced from 57% to less than 3%, and
saturated fatty acids reduced to less than 7%. This oil has been produced commercially
and is extremely stable for high-temperature frying applications. In addition, its stability
matches that of mineral oil derived lubricants, and therefore such non-food uses as
biodegradable lubricants are underway. One added consumer benefit to wide future use
of the engineered high-oleic acid oils may be reduction in the pathologies associated
with high o)-6 consumption.
Besides reducing the amounts of trans and co-6 unsaturated fatty acids in vegetable
oils, efforts directed at increasing the amount of health beneficial fatty acids in vegetable
oils appear to be promising. Two potential health-promoting fatty acids are stearidonic,
(18:4-A6'9t2'15), an co-3 fatty acid precursor of the co-3 structures found in fish oils,
120

and 7-1inolenic acid (18 : 3-A6'~'12), which is implicated in relieving arthritis and other
conditions [30]. y-Linolenic acid is produced in some fungi and the seeds of some
plants by the desaturation of linoleic acid to 7-1inolenic acid by an endoplasmic
reticulum-localized A6-desaturase. Identification of cDNAs encoding this desaturase
from borage and the fungus Mortierella alpina has led to their heterologous expression
in plants. Expression of the Mortierella gene in Brassica napus seed resulted in 47%
7-1inolenic acid production (Ursin, 2000). Expression of the fungal A6-desaturase in a
low e~-linolenic acid canola line further enhanced 7-1inolenic acid production to 68%.
When canola lines with high c~-linolenic content were crossed with high 7-1inolenic
acid lines, up to 17% stearidonic acid was produced. These examples illustrate that the
level of desired end-product can sometimes be substantially increased, although often
unpredictably, by the choice of host and/or enzyme source.

12.1.2. Alternatives to hydrogenated margarines and shortenings

Since most vegetable oils are liquids at room temperature, the production of margarines
and shortenings from such oils requires alteration of their physical properties. This
is most frequently achieved by catalytic hydrogenation of the oil, a process which
reduces the double bonds and thereby increases the melting point of the oil. However,
hydrogenation substantially increases the saturated fat content of the oil. An additional
side reaction which occurs during hydrogenation is the conversion of much of the
naturally occurring cis double bonds to the trans configuration. Typically, hydrogenated
oils used in margarine or shortening contain up to 25 40% saturated or trans fatty
acids. Although convincing evidence for a deleterious effect of trans-isomers in the
diet is lacking, some nutritionists consider reduction of trans-double bonds in the diet
advantageous. An added disadvantage to hydrogenation is the 2-3 cents per pound cost
which it adds to the price of the oil. Thus, for several reasons, an alternative to vegetable
oil hydrogenation is desirable for manufacture of margarines and shortenings.
Further progress toward reducing the need for hydrogenation has been made using a
molecular genetic approach to increase the melting point of rapeseed oil. As described in
Section 3.2, the introduction of the first double bond into plant fatty acids occurs by the
action of stearoyl-ACP desaturase. An obvious route to alter the activity of this enzyme
in oilseeds is the use of antisense RNA. This objective was achieved in Brassica napus
and other plants, by suppression of stearoyl-ACP desaturase messenger RNA, which
reduced enzyme activity and desaturase protein to barely detectable levels (Knutzon,
1992). As a result, the content of stearic acid in the seed oil was increased from 2 to
40%. Due to its high saturated fatty acid content, the oil from these plants has a high
melting point and may be directly suitable for margarine or shortening manufacture
without added hydrogenation.
An alternative fatty acid modification that might simultaneously increase unsaturation
in diets and reduce the need for hydrogenation is the production of petroselinic acid
rich vegetable oils. Petroselinic acid is an isomer of oleic acid which has a cis double
bond at the 6th carbon from the carboxyl end of the molecule rather than the 9th
carbon. This minor modification of the structure alters the melting point of the fatty
acid such that petroselinic acid melts at 33C whereas oleic acid melts at 5C. Thus
petroselinic acid might provide the means to produce an unsaturated vegetable oil which
 121

is also a solid at room temperatures and therefore ideal for margarine and shortening
manufacture. Although petroselinic acid is a major component of seed oils in species
such as coriander and carrot, these crops have a low yield of oil per hectare. It is
therefore hoped that an oilseed crop can be engineered to become a commercially viable
source of petroselinic acid. When a cDNA encoding the acyl-ACP desaturase involved
in petroselinic acid synthesis was cloned from coriander and used to transform other
plants, petroselinic acid was produced but only at low levels. Therefore, to achieve
economic production of petroselinic acid, it may be necessary to introduce additional
genes. It is now known that coriander and its relatives possess several proteins devoted
to production of this special fatty acid, including ACP isoforms, a modified condensing
enzyme (3-ketoacyl-ACP synthase) specific for the elongation of cis-4-16: 1-ACP to
cis-6-18 : I-ACP and a novel acyl-ACP thioesterase specific for petroselinoyl-ACR

12.2. Design of new industrial oils

As described above, plants have evolved the ability to produce a diverse range of
fatty acid structures. A number of specialty fatty acids have already been extensively
exploited for industrial uses such as lubricants, plasticizers and surfactants (Table 3). In
fact, approximately one-third of all vegetable oil is now used for non-food products, and
this figure is expected to increase as petroleum reserves are depleted. Thus, in addition
to providing food, oilseed crops can be considered efficient, low polluting chemical
factories which are able to harness energy from sunlight and transform it into a variety
of valuable chemical structures with a multitude of non-food uses.

12.2.1. High laurate and caprate oils

One of the major non-food uses of vegetable oils, consuming approximately 500 million
pounds of oil per annum in the US, is the production of soaps, detergents and other
surfactants. The solubility and other physical properties of medium-chain fatty acids
and their derivatives make them especially suited for surfactant manufacture. Coconut
and palm kernel oils, which contain 40-60% lauric acid (12:0), are the current major
feedstocks for the surfactant industry. The mechanism for synthesis of lauric and other
medium-chain fatty acids in plants involves the action of a medium-chain acyl-ACP
thioesterase which terminates fatty acid synthesis after a 10 or 12 carbon chain has been
assembled (Pollard, 1991). A cDNA encoding such a thioesterase was isolated from
seeds of the California Bay tree and transformed into rapeseed. As shown in Fig. 10,
the introduction of this specialized thioesterase resulted in transgenic seeds that produce
up to 60 mol% lauric acid. The plants grow normally and oil yields are very similar to
those of the untransformed cultivars. Commercial production of high lauric rapeseed oil
began in 1995, and this crop has the potential to provide a new, non-tropical source of
lauric oils for the surfactant industry.
Since current oil crops do not produce significant amounts of caprate (10:0) and
caprylate (8:0), the cloning of thioesterases from plants accumulating these species
raised hopes for development of commercial sources of 8 : 0 and 10 : 0. Surprisingly,
medium-chain thioesterase from elm, a 10:0 accumulator, recognized either 10:0 or
16:0, and a thioesterase from a Cs 10-rich Cuphea species, although showing strong
122

Pathway
4:0-ACP
Lauric Acid
12:0 12:0 I 4MCTE 12:0-ACP

14:0 14:0-ACP

"o
'~ 16:0 16:0-ACP
<
18:0-~ACP
~ 18:1
u.
18:1-~ACP

18:2
Wildtype 1
[] Transgenic J
18:3
I J
10 20 30 40 50 60
Mole %
Fig. 10. Genetic engineering of rapeseed oil to produce laurie acid. tool% of major fatty acids in a typical
canola cultivar are compared to the composition achieved through genetic engineering using a California
Bay medium chain acyl-ACP thioesterase (MCTE) under control of a napin promoter.

selectivity for medium-chain fatty acids, gave disappointing results in transformed

plants. Part of this discrepancy was resolved with the discovery that medium-chain
production in Cuphea also involves a specific keto-acyl synthase. Cotransformation with
the medium-chain thioesterase and the keto-acyl synthase IV give plants with substantial
caprate and caprylate [31], and further improvements may be possible with addition of
acyltransferases better able to introduce 8 : 0 and 10 : 0 to underrepresented positions in
triacylglycerol [32].

12.2.2. Production of waxes

Long-chain wax esters were once harvested from sperm whales and were a major
ingredient of industrial lubricants and transmission fluids. Banning of whale harvests
led to searches for alternative biological sources of such structures. Jojoba, a desert
shrub found in the American southwest, is the only plant species known to accumulate
waxes (up to 60% dry weight) rather than TAG as a seed storage reserve. These waxes
are mostly derived from C20-C24 monounsaturated fatty acids and alcohols and are
synthesized by the elongation of oleate followed by reduction to alcohols by a fatty
acid reductase. The wax storage lipid is formed by a fatty acyl-CoA: fatty alcohol
acyltransferase, also referred to as wax synthase. The reductase and acyltransferase
were purified from jojoba and the cognate cDNAs cloned. Coordinated expression of
three genes, a Lunaria annua long-chain acyl-CoA elongase, and the jojoba reductase
and acyltransferase in Arabidopsis, resulted in wax production at up to 70% of the oil
present in mature seeds (Lardizabal, 2000). The high levels of accumulation indicated
that all the genes necessary for this trait were identified. If this trait can be successfully
transferred to commercial crops this would represent a large potential source of waxes
used for a variety of applications including cosmetics and industrial lubricants.
 123

12.2.3. Other industrial oils

Many of the unusual fatty acids produced by plants would have substantial value as
industrial feedstocks if they were available in sufficient quantity at low prices. Examples
in this category include fatty acids with hydroxy, epoxy, cyclopropane or branched
chains. These specialty fatty acids are often produced in wild species which have not
been optimized for high agronomic and oil yields, and therefore such specialty oils are
expensive to produce. An alternative to the long-term effort required for domestication
of such plants is the introduction of genes relevant to unusual fatty acid production into
existing high-yielding oil crops. A step in this direction was made by the isolation of
a cDNA for a fatty acid hydroxylase from the castor oil plant. When this gene was
introduced into transgenic plants, approximately 20% hydroxy fatty acids were produced
(Broun, 1997). Genes for specialty fatty acid production need not be isolated only from
plants. As mentioned above, the stearoyl-CoA desaturases from animals and yeast are
active in plants. Furthermore, the cyclopropane synthase of E. coli (Schmid, 1995) and
desaturases from cyanobacteria and fungi have been successfully expressed in transgenic
plants. Thus, in principle there are no fundamental barriers to producing a wide range
of oil compositions using genes borrowed from diverse organisms. Furthermore, protein
engineering offers even more possibilities to tailor the substrates and products of plant
enzymes to produce 'designer oil crops' for specific end uses [6].

13. Future prospects

It is now possible to say that the biosynthetic pathways have been determined for all
major plant lipids and most of the genes identified for enzymes in these pathways have
been cloned. Clearly there has been great progress, although the biosynthetic pathways
of sphingolipids, trans-A'~-16 : 1 and some details of several pathways remain elusive.
In addition, the enzymes involved in the production of many unusual fatty acids found
in seed oils are largely unexplored.
One area of expanding interest is the production of lipid hormones and signal
molecules. Several lipids including phosphatidylinositol phosphates, diacylglycerol and
N-acylphosphatidylethanolamine have been implicated in signal transduction in plants.
Another intriguing similarity between plants and animals is their use of oxygenated
fatty acids in response to wounding. As mentioned in Section 10, jasmonate, a plant
growth regulator derived from 18:3, is able at femtomolar concentrations to induce
proteinase inhibitors and other plant defense genes. Like leukotriene synthesis in
animals, jasmonate biosynthesis begins with the generation of a hydroperoxide by
lipoxygenase. Jasmonate itself contains a cyclopentane ring comparable to those of
prostaglandins. The common roles and origins of oxygenated fatty acids in plants and
animals suggest a very early common ancestor for these pathways.
Application of molecular genetics and genomics to problems in lipid biochemistry
should continue to expand. Particularly stimulating has been the complete sequenc-
ing of the Arabidopsis genome and the availability of over 1 million expressed
sequence tag (EST) sequences from a variety of plants. A recent survey of Ara-
bidopsis genes involved in plant acyl lipid metabolism identified over 450 genes
124
Table 6
Intemet resources related to plant lipid metabolism
Organization Web site content URL
National Plant Lipid Directory of scientists http://www.msu.eduluser/ohhogge/
Consortium (NPLC) involved in plant lipid
research.
E-mail newsgroup tier
int~rmation on plant lipids.
Abstracts of NPLC meetings
Michigan State University Survey and catalog of genes http://www.canr.msu.edu/lgc/index.html
for plant lipid metabolism. http://www.plantbiology.msu.edu/
gene_survey/fron| page.htm
Gene expression profiles http://www.bppmsu.edu/Seed/SeedArray.htm
based on plant lipid ESTs
and microarrays
Kathy Schmid, Butler Links to many oilseed http:lYolue.butler.edu/~kschmidllipids.html
University research laboratories and
web sites
Benning and Ohlrogge Database and analysis of http://benningnt.bchmsu.edu
Laboratories > 10,000 ESTs from
developing ArabidopsL~
seeds
USDA Oilseed Database Data on fatty acid www.ncaur.usda.gov/nc/ncdb/search.html-ssi
composition of seeds of
thousands of species

(http://www.plantbiology.msu.edu/gene_survey/front_page). Although most of these

could be assigned a tentative function based on sequence similarity to previously
characterized genes, only a handful have been examined individually at an experimental
level, and therefore the precise function of hundreds of genes awaits further work. For
example, there are 8 genes which are similar to acyl-CoA desaturases whose function
has not yet been identified and there are 19 genes for lipid transfer proteins and 51
additional lipid transfer protein-like genes. The physiological reasons underlying the
existence of large gene families for lipid transfer proteins, plastid ACPs, and stearoyl
desaturases, but only one gene for keto acyl-ACP synthase iII and for most fatty acid
synthase and ACC subunits, etc. remain to be elucidated. The availability of large
collections of T-DNA insertion mutants means that the impact of gene disruptions can
be tested. However, because 60% of Arabidopsis genes are present as duplicates, such
gene disruptions must be supplemented by other strategies of functional genomics.
Some websites related to these efforts in the plant lipid field are presented in Table 6.
Much of past lipid research has focused on a reductionist approach in which cells
are taken apart and their pieces analyzed. The overall success of this approach and
the wealth of new clones and sequence information have given us an unprecedented
knowledge of the pieces of the puzzle which represent lipid metabolism. However, as
in any puzzle, it is not just complete knowledge of the pieces, but an understanding
 125

of how (and when) they fit together that defines the challenge. Microarrays that permit
simultaneous monitoring of expression of many genes have begun to provide a more
global overview of how genes work together to control seed metabolism (Girke, 2000).
Together with the ability to rapidly over- and under-express genes in transgenic plants
and the strengths of classical biochemistry, such recent advances in analytical techniques
should allow us to enter a new stage of lipid research emphasizing the interplay between
metabolic compartments and the control of lipid synthesis during the plant life cycle.

Abbreviations

ACC acetyl-CoA carboxylase

ACP acyl carrier protein
EST expressed sequence tag
CoA coenzyme A
TAG triacylglycerol

References
I. Pyke, K.A. (1999) Plastid division and development. Plant Cell 11,549-556,
2. Gueguen, V., Macherel, D., Jaquinod, M., Douce, R. and Bourguignon, J. (2000) Fatty acid and lipoic
acid biosynthesis in higher plant mitochondria. J. Biol. Chem. 275, 5016-5025.
3. Gerhardt, B. (1993) Catabolism of fatty acids. In: T.S. Moore (Ed.), Lipid Metabolism in Plants. CRC
Press, Boca Raton, FL, pp. 527-565.
4. Overath, P. and Stumpf, E (1964) Fat metabolism in higher plants XXIII. Properties of a soluble fatty
acid synthetase from avocado mesocarp. J. Biol. Chem. 239, 4103-4110.
5. Harwood, J.L. (1996) Recent advances in the biosynthesis of plant fatty acids. Biochim. Biophys. Acta
1301, 7-56.
6. Shanklin, J. and Cahoon, E.B. (1998) Desaturation and related modifications of fatty acids. Annu. Rev.
Plant Physiol. Plant Mol. Biol. 49, 611-641.
7. Jones, A., Davies, H.M. and Voelker, T.A. (1995) Palmitoyl-acyl carrier protein (ACP) thioesterase and
the evolutionary origin of plant acyl-ACP thioesterases. Plant Cell 7, 359-371.
8. Konishi, T., Shinohara, K., Yamada, K. and Sasaki, Y. (1996) Acetyl-CoA carboxytase in higher plants:
most plants other than Gramineae have both the prokaryotic and the eukaryotic t\wms of this enzyme.
Plant Cell Physiol. 37, 117-122.
9. Ohlrogge, J.B. and Jaworski, J.G. (1997) Regulation of fatty acid synthesis. Annu. Rev. Plant Physiol.
Plant Mol. Biol. 48, 109-136.
10. Kunst, L., Browse, J. and Somerville, C. (1988) Altered regulation of lipid biosynthesis in a mutant of
Arabidopsis deficient in chloroplast glycerol phosphate acyltransferase activity. Proc. Natl. Acad. Sci.
USA 85, 4143-4147.
I I. Kinney, A.J. (1993) Phospholipid headgroups. In: T.S. Moore Jr. (Ed.), Lipid Metabolism in Plants.
CRC Press, Boca Raton, FL, pp. 259 284.
12. Froehlich, J.E., Benning, C. and Dormann, E (2001) The digalactosyldiacylglycerol (DGDG) synthase
DGDI is inserted into the outer envelope membrane of chloroplasts in a manner independent of the
general import pathway and does not depend on direct interaction with monogalactosyldiacylglycerol
synthase for DGDG biosynthesis. J. Biol. Chem. 276, 31806-31812.
13. Frandsen, G.I., Mundy, J. and Tzen, J.T.C. (2001) Oil bodies and their associated proteins, oleosin and
caleosin. Physiol. Plant. 112, 301-307.
126

14. Feussner, I., K~ihn, H. and Wasternack, C. (2001) Lipoxygenase-dependent degradation of storage
lipids. Trends Plant Sci. 6, 268-273.
15. van de Loo, EJ., Fox, B.G. and Somerville, C. (1993) Unusual fatty acids. In: T.S. Moore Jr. (Ed.),
Lipid Metabolism in Plants. CRC Press, Boca Raton, FL, pp. 167-194.
16. Broun, E, Shanklin, J., Whittle, E. and Somerville, C. (1998) Catalytic plasticity of fatty acid
modification enzymes underlying chemical diversity of plant lipids. Science 282, 1315-1317.
17. Voelker, T. and Kinney, A.J. (2001) Variations in the biosynthesis of seed-storage lipids. Annu. Rev.
Plant Physiol. 52, 335-361.
18. Dahlqvist, A., Stahl, U., Lenman, M., Banas, A., Lee, M., Sandager, L., Ronne, H. and Stymne,
S. (2000) Phospholipid:diacylglycerol acyltransferase: an enzyme that catalyzes the acyl-CoA-
independent formation of triacylglycerol in yeast and plants. Proc. Natl. Acad. Sci+ USA 97, 6487-
6492.
19+ Kolattukudy, EE. (2001) Polyesters in higher plants. Adv. Biochem. Eng./Biotechnol. 71, 1-49.
20. Lichtenthaler, H.K. (2000) Non-mevalonate isoprenoid biosynthesis: enzymes, genes and inhibitors.
Biochem. Soc. Trans. 28, 785-789.
21. Li, J., Biswas, M.G., Chao, A., Russell, D.W. and Chory, J. (1997) Conservation of function between
mammalian and plant steroid 5c~-reductases. Proc. Natl. Acad. Sci. USA 94, 3554-3559.
22. Sperling, P., Ternes, P., Moll, H., Franke, S., Zahringer, U. and Heinz, E. (2001) Functional char-
acterization of sphingolipid C4-hydroxylase genes fi'om Arabidopsis thaliana. FEBS Lett. 494, 90-
94.
23. Lynch, D.V., Criss, A.K., Lehoczky, J.L. and Bui, V.T. (1997) Ceramide glucosylation in bean
hypocotyl microsomes: evidence that steryl glucoside serves as glucose donor. Arch. Biochem.
Biophys. 340, 311-316.
24. Schaller, E (2001) Enzymes of the biosynthesis of octadecanoid-derived signalling molecules. J. Exp.
Bot. 52, 11-23.
25. Okuley, J., Lightner, J., Feldmann, K., Yadav, N., Lark, E. and Browse, J. (1994) The Arabidopsis
FAD2 gene encodes the enzyme that is essential for polyunsaturated lipid synthesis. Plant Cell 6,
147-158.
26. Lightner, J., Wu, J. and Browse, J. (1994) A mutant of Arabidopsis with increased levels of stearic
acid. Plant Physiol. 106, 1443-1451.
27. McConn, M. and Browse, J. (1998) Polyunsaturated membranes are required for photosynthetic
competence in a mutant of Arabidopsis. Plant J. 15, 521-530.
28. Arondel, V., Lemieux, B., Hwang, I., Gibson, S., Goodman, H.M. and Somerville, C.R. (1992)
Map-based cloning of a gene controlling omega-3 fatty acid desaturation in Arabidopsis. Science 258,
1353-1355.
29. James, D.W., Lim, E., Keller, J., Plooy, 1., Ralston, E. and Dooner, H.K. (1995) Directed tagging ol+
the Arabidopsis fatty acid elongation I (FAE1) gene with the maize transposon activator. Plant Cell 7,
309-319.
30. Broun, P., Gettner, S. and Somerville, C. (1999) Genetic engineering of plant lipids. Annu. Rev. Nutr.
19+ 197-216.
31. Dehesh, K., Edwards, E+ Fillatti, J., Slabaugh, M. and Byrne, J. (1998) KAS IV: a 3-ketoacyl-ACP
synthase from Cuphea sp. is a medium chain specific condensing enzyme. Plant J. 15, 383-390.
32. Knutzon, D.S., Hayes, T.R., Wyrick, A., Xiong, H., Davies, H. and Voelker, T.A. (1999) Lysophos-
phatidic acid acyltransferase from coconut endosperm mediates the insertion of laurate at the sn-2
position of triacylglycerols in lauric rapeseed oil and can increase total laurate levels. Plant Physiol.
120, 739-746.
154
Isoprenoid biosynthesis as a novel target for antibacterial and antiparasitic
drugs
Michel Rohmer*, Catherine Grosdemange-Billiard, Myriam Seemann & Denis
Tritsch
Address
Universit Louis Pasteur
CNRS
Institut Universitaire de France
Institut Le Bel
4 rue Blaise Pascal
67070 Strasbourg Cedex
France
Email: mirohmer@chimie.u-strasbg.fr
*To whom correspondence should be addressed
Current Opinion in Investigational Drugs 2004 5(2):154-162
Thomson Scientific ISSN 1472-4472
The mevalonate-independent methylerythritol phosphate pathway is
a long overlooked metabolic pathway for isoprenoid biosynthesis. It
is present in most bacteria, including pathogens and opportunistic
pathogens, in some unicellular eukaryotes, including the parasite
responsible for malaria, and in the chloroplasts of all phototrophic
organisms. It represents an alternative to the mevalonate pathway,
which is only present in animals, fungi, the plant cytoplasm,
archaebacteria and some eubacteria. This biosynthetic pathway is
thus a potential target for antibacterial and antiparasitic drugs. An
isopentenyl diphosphate isomerase that differs from the previously
known isopentenyl diphosphate isomerase found in all other
organisms, including animals, was discovered in several Gram-
positive bacteria possessing the mevalonate pathway, adding another
target related to isoprenoid biosynthesis.
Keywords Gram-positive, isopentenyl diphosphate,
isoprenoid biosynthesis, mevalonate-independent
methylerythritol phosphate pathway
Introduction: The discovery of the
methylerythritol phosphate pathway, a novel
route for isoprenoid biosynthesis
Isoprenoids are natural products formally derived from the
branched C(5)-skeleton of isoprene. They include essential
metabolites, as well as secondary metabolites, of less
obvious physiological significance. Isopentenyl diphosphate
(IPP) (9; Figure 1) and dimethylallyl diphosphate (DMAPP)
(10; Figure 1) represent the biological equivalents of isoprene
and are the ubiquitous precursors of isoprenoids in all living
organisms. The elucidation of isoprenoid biosynthesis in
liver tissue and yeast lead to the discovery of the mevalonate
(MVA) pathway [1,2]. Additional studies confirmed the
broad distribution of this metabolic route, including its
involvement in the biosynthesis of plant sterols and
triterpenes, as well as many sesquiterpenes. This metabolic
pathway was thus accepted as the universal biosynthetic
route of all isoprenoids in all living organisms, despite some
contradictory results gathered on the formation of
chloroplast isoprenoids (eg, carotenoids), plant mono- and
diterpenes and some bacterial isoprenoids [3,4]. There have
been few investigations involving the early steps of bacterial
isoprenoids and this has been mainly due to the low
concentration usually encountered in bacteria in comparison
with those usually found in other living organisms. Bacteria
synthesize essential isoprenoids such as undecaprenol
(bactoprenol) diphosphate, serving as a carbohydrate carrier
for the biosynthesis of the peptidoglycan of the cell wall or
the prenyl chains of the ubiquinones and menaquinones of
the electron transport chains. Additional isoprenoids of
more restricted distribution include: carotenoids, from
phototrophic and some heterotrophic bacteria, or
triterpenoids of the hopane series acting as membrane
stabilizers and modulating membrane rigidity and fluidity
[5]. Owing to their high intracellular concentration (0.5 to 30
mg/g), hopanoid biosynthesis could be investigated after
13 13
administration of [ C]-labeled precursors using C-NMR
spectroscopy. These experiments were designed to
investigate the origin of the additional C(5)-side chain linked
to the triterpene skeleton. They demonstrated that the side
chain was derived from a pentose and, in addition, disclosed
an unexpected and completely overlooked novel pathway
for the formation of isoprene units [6]. Extensive labeling
13
experiments, mainly performed with [ C]-labeled glucose
isotopomers, determined that the isoprene units were
derived from D-glyceraldehyde phosphate (2; Figure 1) and
from a C(2)-unit resulting from pyruvate (1; Figure 1)
decarboxylation, and that a rearrangement reaction was
responsible for the formation of the branched C(5)-skeleton
from a straight chain [7,8] (Figure 1). These findings
were extended to many bacteria species, including bacteria
that do not produce hopanoids (eg, Escherichia coli) and to
other isoprenoid series such as the prenyl side chains of
ubiquinone and menaquinone. Schwarz and Arigoni [9,10]
discovered the same pathway in a higher plant system.
13
According to the results from incorporation of [ C]-labeled
glucose isotopomers, the diterpenoids of the ginkgolide
series were synthesized in Gingko biloba embryos via the
same pathway, whereas sitosterol was derived from the
MVA pathway.
Elucidation of the methylerythritol phosphate
pathway
Several reviews have been published on the methylerythritol
phosphate (MEP) pathway [4,11,12], therefore, only the most
striking features concerning this novel metabolic route
(Figure 1) will be outlined here. The labeling experiments
discussed above determined the origin of the carbon atoms
of the isoprene units derived from the MEP pathway. They
also proposed 1-deoxy-D-xylulose (DX) and 2-C-methyl-D-
erythritol derivatives (ME) as intermediates. Incorporation
of [2H]-labeled DX [13] or ME [14] into the prenyl chains of
ubiquinone and menaquinone from E coli provided proof of
their intervention in the MEP pathway and allowed the
rapid identification of the first two enzymes. The 1-deoxy-D-
xylulose 5-phosphate (DXP) (3; Figure 1) synthase was
identified owing to the homology of the nucleotide sequence
of the corresponding gene with those of transketolases
[15-17]. Curiously, Rhodobacter capsulatus and Streptomyces
coelicolor possess two functional DXS [18,19], and up to three
or five isogenes with strong dxs gene homologies were found
 Isoprenoid biosynthesis as a novel target for antibacterial and antiparasitic drugs Rohmer et al 155

Figure 1. Methylerythritol phosphate pathway for isoprenoid biosynthesis.

O O O O OH O O H3C OH O O
dxs dxr ygbP
O P H3C P HO P
H3C + H O OH O OH O OH
O OH O OH NADPH NADP+ OH CTP PPi

1 2 3 4

NH2 NH2
O
N OH N O O
O P
H3C O P O O
H3C OH O O O O O N ychB H3C O O O O O O N ygbB
HO P
HO P P O HO P P O O O
O O O O O O - CDP
H H OH
OH ATP ADP OH
HO OH HO OH
5 6 7

CH3 O O O O
P P
H2C O O OH

CH3 O O O O idi
gcpE lytB
HO P P
O O OH

8 CH3 O O O O
P P
H3C O O OH

10

Text in italics represents genes.

in Streptomyces hygroscopicus and S spheroides, respectively containing the three known genes of the MEP pathway (dxs,
[20]. DXP is not only an isoprenoid precursor, but was dxr and ygbP, Figure 1). They encode the enzymes
previously known as a precursor for thiamine diphosphate responsible for the further activation of ME:
and pyridoxol phosphate in E coli. The search for E coli phosphorylation of the C(2)-hydroxy group of 4-
mutants auxotrophic to ME led to the identification of the diphosphocytidyl ME (ychB) [26] and the conversion of the
DXP reductoisomerase gene and to the characterization of former intermediate, 4-diphosphocytidyl ME 2-phosphate
the corresponding enzyme [21,22]. This enzyme requires a (6; Figure 1) into ME cyclodiphosphate (7; Figure 1) (yfgB)
2+ 2+ 2+
divalent cation (Mg , Mn or Co ). It catalyzes the [27]. Deletion of the ygbP, ychB or ygbB gene in E coli was
rearrangement of DXP into methylerythrose phosphate and rescued by the insertion of the three genes encoding the
the concomitant NADPH-dependent reduction of the enzymes permitting the utilization of exogenous MVA
aldehyde into 2-C-methyl-D-erythritol 4-phosphate (MEP) (4; (MVA kinase, phospho-MVA kinase and diphospho-MVA
Figure 1) [23,24]. MEP shows the typical branched C(5)- decarboxylase). This confirmed that they are essential for
isoprene skeleton. It can be considered as the first isoprenoid biosynthesis [28-30].
hemiterpene of the pathway. As no other role is known for
2 2
MEP, the DXP reductoisomerase most probably catalyzes Incorporation of [4- H]DX or [3- H]ME into the prenyl
the first committed step of the pathway, which is named chains of ubiquinone and menaquinone of E coli resulted in
3
after the substrate of this enzyme. Incubation of [ H]-labeled a different labeling pattern in IPP and DMAPP; deuterium
MEP with crude cell-free extracts from E coli in the presence retention in the isoprene units derived from DMAPP and
of nucleotide triphosphates resulted in the formation of an complete loss in those derived from IPP [31,32]. In
adduct. A small gene cluster encoding the genes of the next addition, the IPP isomerase activity is barely detectable in
three enzymes was found when searching for a gene this bacterium, and the deletion of the IPP isomerase gene,
encoding an enzyme using a polyol phosphate (like MEP) the enzyme ensuring the interconversion of IPP (9; Figure 1)
and a nucleotide triphosphate. This resulted in the finding of and DMAPP (10; Figure 1), is not lethal for this bacterium
the homologies between the diphosphocytidine ribityl [33]. This suggested the presence of a branching in the
synthase and the unannotated ygbP gene from E coli [25]. MEP pathway, allowing a separate synthesis of IPP and
Cloning of this gene and overexpression of the DMAPP. Definitive proof for such a branching was
corresponding protein allowed the identification of the 4- identified by genetic methods [34]. An engineered E coli
diphosphocytidyl ME (5; Figure 1) synthase. The next two strain lacking the DXP reductoisomerase and capable of
genes of this cluster were always found in genomes already utilizing MVA grew normally either in the presence of ME
156 Current Opinion in Investigational Drugs 2004 Vol 5 No 2

or MVA. In growth conditions on MVA, only IPP is Table 1. Distribution of the MEP pathway.
synthesized, and cells have to rely on an IPP isomerase for Pathway Reference
the synthesis of DMAPP. After additional deletion of the MEP
IPP isomerase idi gene, MVA no longer supported growth, Acinetobacter calcoaceticus [90]
Actinobacillus actinomycetemcomitans [48]
indicating that there was no other way to interconvert IPP
Actinobacillus pleuropneumoniae [50]
and DMAPP than via the dimethylallyl diphosphate Actinoplanes sp A40644 [48]
isomerases (IDI) enzyme. However, the cells grew Alicyclobacillus acidoterrestris [7]
normally on free ME, demonstrating that IPP and DMAPP Anaplasma phagocytophilum] [50
are independently synthesized in the MEP pathway from Aquifex aeolicus [48]
an ME derivative. Bacillus anthracis [50]
Bacillus cereus [50]
A bioinformatic search in fully sequenced bacterial genomes Bacillus subtilis [48]
for genes accompanying the known genes of the MEP Bacteroides fragilis [50]
pathway led to the identification of two additional genes, Bacteroides thetaiotaomicron [50]
Bifidobacterium longum [50]
gcpE [35,36] and lytB [37,38]. These genes are essential in the
Bordetella bronchiseptica [50]
MEP pathway. Their deletion in E coli was rescued by the Bordetella pertussis [48]
addition of mevalonate to the culture medium after Brucella melitensis [50]
introducing the genes, allowing the utilization of this Brucella suis [50]
unnatural substrate for E coli [35,36,38]. The Burkholderia caryophylli [90]
hydroxymethylbutenyl 4-diphosphate synthase (GcpE) Burkholderia cepacea [50]
enzyme converts ME cyclodiphosphate into (E)-4-hydroxy- Burkholderia gladioli [90]
3-methylbut-2-enyl diphosphate (HMBPP) (8; Figure 1) Burkholderia mallei [50]
[39,40] by two successive one-electron transfers induced via Burkholderia pseudomallei] [50]
a [4Fe-4S] cluster [41,42]. The LytB enzyme catalyzes the Campylobacter jejuni] [48]
last step of the pathway, converting HMBPP into IPP or Caulobacter crescentus [48]
Chlamydia muridarum [49]
DMAPP and being responsible for the branching in the
Chlamydia pneumoniae [48]
pathway [43-45]. This last step involves, like the GcpE-
Chlamydia trachomatis [48]
catalyzed reaction, two one-electron transfers via a [4Fe-4S] Chlamydophila pneumoniae] [50
cluster [46,47], which has been directly characterized by Chlorobium tepidum [48]
electron paramagnetic resonance spectroscopy [47]. Citrobacter freundii [90]
Clostridium acetobutylicum [48]
Most steps of the MEP pathway described above for E coli Clostridium botulinum [50]
were also characterized in plants [11], where this pathway Clostridium difficile [48]
is localized in chloroplasts as shown by the additional Clostridium perfringens [50]
Clostridium tetani [50]
putative plastid-targeted N-terminal sequence found in all
Corynebacterium ammoniagenes [48]
enzymes. Corynebacterium diphtheriae [48]
Deinococcus radiodurans [48]
Distribution of the MEP and MVA pathways: Ehrlichia chaffeensis [50]
The MEP pathway, a target for antibacterial Erwinia carotovora [90]
Escherichia coli [7,48]
and antiparasitic drugs Eubacterium sp [50]
First insight into the distribution of the MEP pathway has
Francisella tularensis [50]
been demonstrated biochemically, usually after incubation Fusobacterium nucleatum [50]
13
of [ C]-labeled substrates (eg, glucose) and analyzing the Gardnerella vaginalis [50]
labeling pattern of the isoprenoids by NMR spectroscopy. Haemophilus ducreyi [50]
This approach was first applied to bacteria, and later Haemophilus influenzae [48]
extended to unicellular algae, liverworts and plants [3,4,11]. Helicobacter jejuni [48]
In unicellular green algae, it is the only pathway for Helicobacter pylori [50]
isoprenoid biosynthesis. In other investigated algae phyla Klebsiella pneumoniae [50]
and in plants, its presence is restricted to the chloroplast for Leptospira interogans [50]
the synthesis of the essential isoprenoids of the Listeria monocytogenes [50]
photosynthetic apparatus (phytol from chlorophylls, Mannheimia haemolytica [50]
Methylobacterium fujisawaense [7,48]
carotenoids and plastoquinone), of mono- and diterpenoids
Methylobacterium organophilum [6]
and isoprene. Sterols and triterpenes, most sesquiterpenes Moraxella catarrhalis [50]
and the prenyl chain of ubiquinones are synthesized via the Mycobacterium avium [48]
MVA pathway in the cytoplasm. Exchanges between the Mycobacterium bovis [48]
two cellular compartments have, however, been regularly Mycobacterium leprae [48]
reported and cross-talk between the two compartments is Mycobacterium phlei [90]
probably a way for the regulation of isoprenoid biosynthesis Mycobacterium smegmatis [90]
in plant cells [11]. Once the genes of the pathway were Mycobacterium tuberculosis [48]
known, a bioinformatic analysis of known genomes pointed Mycoplasma penetrans [50]
out the broad distribution of the genes of the MEP pathway Neisseria gonorrhoeae [48]
(Table 1). Neisseria meningitidis [48]
 Isoprenoid biosynthesis as a novel target for antibacterial and antiparasitic drugs Rohmer et al 157

Table 1. Distribution of the MEP pathway (continued). Although the MVA pathway has been found in some
Pathway Reference bacteria, the MEP pathway represents probably the most
MEP widespread pathway for isoprenoid biosynthesis in
Nocardia brasiliensis [91] [91] eubacteria (Table 1). The known distribution of the two
Neorickettsia sennetsu [50] [50]
Pasteurella multocida [50] [50]
isoprenoid pathways merely reflects lateral gene transfer
Peptostreptococcus sp [50] rather than phylogenetic relationship [48,49,50]. The
Porphyromonas gingivalis [48] simultaneous presence of both pathways has been only
Prevotella intermedia [50] related in some Streptomyces species [51]. The MEP pathway
Proteus mirabilis [50] is utilized during the exponential growth phase for the
Providencia stuartii [50] biosynthesis of essential metabolites, such as the prenyl
Pseudomonas aeruginosa [90]
chains of the menaquinones, whereas the MVA route is
Pseudomonas fluorescens [90]
Pseudomonas putida [50] involved in the synthesis of the prenyl moiety of antibiotics
Psychrobacter sp [50] of mixed origin during the stationary phase. The MEP
Ralstonia pickettii [90] pathway is found in many pathogenic bacteria as well as in
Rhodobacter capsulatus [48] opportunistic pathogens (Table 1). It is also present in the
Rhodopseudomonas acidophila [6] parasitic Plasmodium species [52].
Rhodopseudomonas palustris [6]
Salmonella enterica [50]
In contrast, the MEP pathway is absent in animals, yeasts
Salmonella enteritidis [50]
Salmonella typhi [48]
and fungi. There is neither biochemical proof from labeling
Salmonella typhimurium [90] experiments for its presence, nor reports of the
Serratia marcescens [50] corresponding genes in the genomes of these organisms.
Shewanella putrefaciens [48] This makes the MEP metabolic route an interesting target for
Shigella flexneri [50] the design of novel antibacterial and antiparasitic drugs,
Shigella dysenteriae [50] with minimal side effects for the patient. Even if most
Sphingobacterium multivorum [92]
bacterial species do not produce large amounts of
Streptomyces aeriouvifer CL190* [48]
Streptomyces blastmyceticum
isoprenoids, some of these metabolites are essential for all
[48]
Streptomyces exfoliatus [48]
bacteria (eg, bactoprenol diphosphate, the carbohydrate
Streptomyces ghanaensis [48] carrier for the biosynthesis of peptidoglycan of the cell wall)
Streptomyces griseolosporeus* [48] or in most of them (prenyl chain of ubiquinones and
Streptomyces niveus [48] menaquinones from electron transport chains). Any
Streptomyces spheroides [48] inhibition of the biosynthesis of these essential terpenoids is
Streptomyces sp QC45B [93] lethal.
Streptomyces sp UC 5319 [94]
Streptomyces sp JP95 [95]
Synechocystis sp PCC 6803
Enzymes of the MEP pathway: New targets for
[48]
Tannerella forsythensis [50] antibacterial and antiparasitic drugs
Thermotoga maritima [48] Due to the relatively recent discovery of the MEP pathway,
Thiobacillus ferrooxidans [48] few investigations have been published on the potential role
Treponema denticola [50] of this metabolic route for the development of new
Tropheryma whipplei [50] antibacterial drugs.
Treponema pallidum [48]
Vibrio cholerae [48] The discovery of the genes implied in the MEP pathway
Vibrio vulnificus [50] made recombinant enzymes available, thus making the
Wolbachia sp [50] characterization of their catalytic properties possible. Kinetic
Xylella fastidiosa [49]
data have been obtained for the DXP synthase of Rhodobacter
Yersinia pestis [48]
capsulatus [18] and Streptomyces coelicolor [19], the DXP
Yersinia enterocolitica [50]
reductoisomerase from E coli [23,24] and Streptomyces
Zymomonas mobilis [7]
coelicolor [19], the 4-diphosphocytidyl-2-C-methylerythritol
MVA
Borrelia burgdorferi [48] synthase from S coelicolor [19] and from the higher plant
Chloroflexus aurantiacus [48] Arabidopsis thaliana [53], and the 2-C-methylerythritol 2,4-
Enterococcus faecalis [48] cyclodiphosphate synthase from Plasmodium falciparum [54].
Flavobacterium sp [96] Amino acid residues involved in the catalytic activity have
Lactobacillus helveticus [97] been characterized by directed mutagenesis for the DXP
Lactobacillus plantarum [48] synthase [55] and for the DXR [56] from E coli. X-ray
Myxococcus fulvus [48,90] structures are available for the DXP reductoisomerase
Nannocystis exedens [98] [57,58], the 4-diphosphocytidyl-2-C-methylerythritol
Paracoccus sp [99] synthase [59,60], the 4-cytidine 5'-diphospho-2-C-methyl-D-
Staphylococcus aureus [48]
erythritol kinase [61,62] and the 2-C-methylerythritol 2,4-
Staphylococcus carnosus [48]
cyclodiphosphate synthase [63-66] from E coli.
Streptococcus mutans [48]
Streptococcus pneumoniae [48]
Streptococcus pyogenes
Fosmidomycin (13; Figure 2) is an antibiotic, which was
[48]
discovered in the Fujisawa Research Laboratories [67]. It has a
*Bacteria possessing the MEP and the MVA pathways. potent antibacterial activity against many Gram-negative and
some Gram-positive bacteria, has a low toxicity for animals
158 Current Opinion in Investigational Drugs 2004 Vol 5 No 2
and is efficient against experimental infections in mice [68]. A
correlation was found in bacteria between fosmidomycin
insusceptibility and the ability to incorporate MVA into
isoprenoids. MVA incorporating bacteria such as Lactobacillus
spp, Streptococcus spp and Arthrobacter spp were all
fosmidomycin resistant, whereas Micrococcus spp, Bacillus spp
and the tested Gram-negative bacteria (E coli, Enterococcus
cloacae and Pseudomonas aeruginosa) were sensitive to the drug
[69]. In addition, fosmidomycin treatment was followed by a
decrease of the ubiquinone and menaquinone content in E
coli, and of carotenoids in Micrococcus luteus. These data
suggested that fosmidomycin had an effect on isoprenoid
biosynthesis and, according to the above mentioned list of
tested bacteria, on an enzyme of the mevalonate-independent
MEP pathway. This was verified once the enzymes of the
MEP pathway were available. Fosmidomycin was reported as
a mixed type inhibitor of the DXP reductoisomerase from E
coli (Ki = 38 nM) [70] and as a competitive inhibitor of the
Zymomonas mobilis enzyme (Ki = 0.6 M) [71]. It also inhibits
the enzyme from higher plants [72,73]. In addition, the
recombinant DXP reductoisomerase from Plasmodium
falciparum is inhibited by fosmidomycin and the related
antibiotic FR-900098 (14; Figure 2) in a dose-dependent
manner [52]. An additional antibiotic related to
fosmidomycin, FR-33289 (16; Figure 2) is known, but nothing
has been reported on its activity [70]. The presence of the
MEP pathway in Plasmodium spp is related to the apicoplasts,
which are plastid-like organelles acquired by the members of
the phylum Apicomplexa by endosymbiosis of an algal
ancestor. Fosmidomycin and FR-900098 are potent
antimalarial agents. For example, mice infected with
Plasmodium vinckei were treated intraperitoneally with
fosmidomycin (> 10 mg/kg) or FR-900098 (5 mg/kg) and
were apparently free of parasites, whereas untreated control
animals died from infection after 7 days. After oral
administration of either fosmidomycin or FR-900098 (50 or
100 mg/kg), mice were seemingly free of parasites, and
Figure 2. Inhibitors of the methylerythritol phosphate pathway.
H3C
O OH
H3C
F O
N OH
O
O
F F O O
11
OH O O
H3C N P H3C
OH
O O
14 FR-900098
OH F OH
F OH F OH
O OH O OH
17 18
parasitemia was lower than 1% at concentrations as low as 20
mg/kg of either drug [52]. The diaryl esters (15; Figure 2),
prodrugs of FR-900098, demonstrated superior activity
against P vinckei in mice after oral administration. One of
these prodrugs even demonstrated a comparable activity by
oral or intraperitoneal administration [74]. Density Functional
Theory (DFT) ab initio calculations were performed on DXP,
the substrate of the DXP reductoisomerase, as well as on its
inhibitors fosmidomycin and FR-900098, in order to
determine their structures and charge distributions. This was
intended for the further development of biologically active
inhibitors as well as parameters for docking experiments [75].
The crystal structure of the DXR complexing manganese and
fosmidomycin showed that the inhibitor behaves like a DXP
(3; Figure 1) analog. The oxygen atoms of the (N-formyl-N-
hydroxy)amino group provide two ligands for the
coordination of manganese, and the phosphonate group
linked through a three-methylene spacer is recognized by
hydrogen binding in a specific pocket [76]. Resistance to
fosmidomycin was investigated in E coli [77]. Adenylate
cyclase or glycerol 3-phosphate transport deficient mutants
were found to be resistant to fosmidomycin, indicating that
the antibiotic is transported via the glycerol 3-phosphate
transporter, the gene of which, like other genes involved in
glycerol metabolism, is highly dependent on the presence of
cyclic AMP.
Little additional information is available on the inhibition of
other enzymes of the MEP pathway. Fluoropyruvate (11;
Figure 2) inhibits DXP synthase, as expected for a thiamine
diphosphate dependent enzyme, with IC50 values of 400 and
80 M for the Pseudomonas aeruginosa and E coli enzymes,
respectively [78]. 5-Ketoclomazone (12; Figure 2), a
breakdown product of the herbicide clomazone, inhibits, to
some extent, the DXP synthase of the green alga
Chlamydomonas (IC50 ~ 0.1 mM), whereas clomazone itself
has no effect [79]. Free DX is readily incorporated into the
O
O O
H N P
Cl
12 13
OH O O Ar OH OH O O
N P Ar H3C N P
O OH
O
15 16 FR-33289
OH O O O O
H3C P H3C P
O OH O OH
O O OH
19 20
 Isoprenoid biosynthesis as a novel target for antibacterial and antiparasitic drugs Rohmer et al 159
isoprenoids from bacteria and plants. It is phosphorylated in
E coli by a xylulokinase to yield DXP [80]. DX analogs are
therefore supposed to be readily converted into the
corresponding 5-phosphates. 1-Fluoro- (17; Figure 2) and
1,1-difluoro-DX (18; Figure 2) were thus synthesized as
putative antibacterial agents, but showed no activity on E
coli cultures [80,81]. The two DXP analogs, lacking the
hydroxy group at C(4) (19, Figure 2; Ki = 120 M) or at C(3)
(20, Figure 2; Ki = 800 M), represented reversible mixed
type inhibitors of the DXP reductoisomerase [24].
Many Gram-positive bacteria (Corynebacterium ammoniagenes,
Mycobacterium smegmatis, Micrococcus luteus/'lysodeikticus') and
Gram-negative bacteria (Xanthomonas campestris, X maltophilia
and Pseudomonas putida) respond to an oxidative stress induced
by drugs such as benzyl- and methylviologen or nitrofurans
(furacilin, furadonin) by accumulating 2-C-methyl-D-erythritol
2,4-cyclodiphosphate [82-84]. Such compounds interfere with
one-electron transfer radical reactions, and one of their targets
may be the Fe/S cluster of the GcpE and LytB proteins
[41,42,44].
The IPP isomerase of the mevalonate pathway
in Staphylococcus and Streptococcus spp:
Another target?
Even in some bacteria possessing the MVA pathway,
isoprenoid biosynthesis may be a target for antibacterial
drugs. A novel type II IPP isomerase has been characterized
in archaea and in some Gram-positive bacteria (eg,
Streptomyces sp CL190, Staphylococcus aureus, Streptococcus
pneumoniae, Streptococcus pyogenes and Enterococcus faecalis)
[85]. The enzyme is a flavoprotein that requires the
simultaneous presence of both flavin mononucleotide and
NADP to produce DMAPP from IPP and is characterized by
completely different catalytic properties and kinetic
parameters than those of all other known IPP isomerase,
including the mammalian enzymes.
Conclusion
The MEP pathway for isoprenoid biosynthesis is
widespread in bacteria. It has been characterized by
biochemical or by genetic methods in many pathogenic
bacteria (eg, Mycobacterium tuberculosis) as well as in
opportunistic pathogens (eg, Pseudomonas spp, Burkholderia
spp, Acinetobacter calcoaceticus and enterobacteria) or in
phylogenetically related harmless bacteria. This pathway is
absent in animals and humans, and the corresponding
enzymes have no equivalents in these organisms. They
therefore represent potential targets for the design of a novel
and unexplored type of antibacterial agents [86]. Even if
fosmidomycin resistance has been characterized for E coli via
a protein facilitating the efflux of the drug [87], this aspect
is particularly relevant in times where nosocomial diseases
and resistance to antibiotics represent a severe problem. The
first results on the efficacy of fosmidomycin against malaria
or urinary tract infections in adults indicate the strong and
promising potentiality of the MEP pathway as a drug target
[88,89]. Due to the recent elucidation of this metabolic route
and to the availability of the already existing natural
antibiotics fosmidomycin and FR-900098, only one enzyme,
the DXP isomeroreductase, has been studied in detail. It is
therefore expected that in the near future, the other enzymes
will be investigated.
References
1. Bloch K: Sterol molecule: Structure, biosynthesis and function.
Steroids (1992) 57(8):378-383.
2. Bochar DA, Friesen JA, Stauffacher CV, Rodwell VW: Biosynthesis of
mevalonic acid from acetyl-CoA. In: Comprehensive Natural Product
Chemistry. Isoprenoids including Carotenoids and Steroids. Cane DE
(Ed), Pergamon Press, Oxford, UK (1999) 2:15-44.
3. Rohmer M: A mevalonate-independent route to isopentenyl
diphosphate. In: Comprehensive Natural Product Chemistry.
Isoprenoids including Carotenoids and Steroids. Cane DE (Ed),
Pergamon Press, Oxford, UK (1999) 2:45-67.
The discovery of the MEP pathway is reported with reference to earlier
investigations on the isoprenoid biosynthesis in bacteria and plants which
were not in accordance with the MVA pathway.
4. Rohmer M: The discovery of the mevalonate-independent pathway
for isoprenoid biosynthesis in bacteria, algae and higher plants.
Nat Prod Rep (1999) 16(5):565-574.
5. Rohmer M: The biosynthesis of triterpenoids of the hopane series
in eubacteria: A mine of new enzyme reactions. Pure Appl Chem
(1993) 65:1293-1298.
6. Flesch G, Rohmer M: Prokaryotic hopanoids: The biosynthesis of the
bacteriohopane skeleton. Formation of isoprenic units from two
distinct acetate pools and a novel type of carbon/carbon linkage
between a triterpene and D-ribose. Eur J Biochem (1988) 175(2):405-411.
7. Rohmer M, Knani M, Simonin P, Sutter B, Sahm H: Isoprenoid
biosynthesis in bacteria: A novel pathway for the early steps leading
to isopentenyl diphosphate. Biochem J (1993) 295(2):517-524.
This is the first report of the existence of an alternative mevalonate-
independent pathway for isoprenoid biosynthesis in bacteria. The origin of the
carbon atoms of isoprene units was determined, and the key rearrangement
step was characterized. This paper represents, with reference [9], the
starting point of all further research on the novel metabolic pathway.
8. Rohmer M, Seemann M, Horbach S, Bringer-Meyer S, Sahm H:
Glyceraldehyde 3-phosphate and pyruvate as precursors of
isoprenic units in an alternative non-mevalonate pathway for
terpenoid biosynthesis. J Am Chem Soc (1996) 118(11):2564-2566.
D-Glyceraldehyde phosphate and pyruvate are identified as the first
precursors of the MEP pathway.
9. Schwarz M: Terpen-biosynthese in Gingko biloba: Eine
berraschende geschichte. PhD Thesis Nb 10951, Eidgenssische
Technische Hochschule, Zrich, Switzerland (1994).
In addition to reference [7], this thesis presents the first report on the
occurrence of the mevalonate-independent route in higher plants. This
research was performed independently of the research on bacterial
isoprenoids. See also reference [10].
10. Schwarz M, Arigoni D: Gingkolide biosynthesis. In: Comprehensive
Natural Product Chemistry. Isoprenoids including Carotenoids and
Steroids. Cane DE (Ed), Pergamon Press, Oxford, UK (1999) 2:367-400.
11. Lichtenthaler HK: The 1-deoxy-D-xylulose 5-phosphate pathway of
isoprenoid biosynthesis in plants. Annu Rev Plant Physiol Plant Mol
Biol (1999) 50:47-65.
12. Rohdich F, Kis K, Bacher A, Eisenreich W: The non-mevalonate
pathway of isoprenoids: Genes, enzymes and intermediates. Curr
Opin Chem Biol (2001) 5(5):535-540.
13. Broers STJ: I. ber die frhen stufen der biosynthese von
isoprenoiden in Escherichia coli. II. Beitrag zur aufklrung der
biosynthese von vitamin B12 in Propionibacterium shermanii. PhD
Thesis Nb 10978, Eidgenssische Technische Hochschule, Zrich,
Switzerland (1994).
This thesis presents the first direct evidence that a 1-deoxy-D-xylulose
derivative is the precursor of isoprene units in Escherichia coli. Incorporation
experiments using labeled deoxyxylulose were later a key tool for the
elucidation of the MEP pathway.
14. Duvold T, Cal P, Bravo J-M, Rohmer M: Incorporation of 2-C-methyl-
D-erythritol, a putative isoprenoid precursor in the mevalonate-
independent pathway, into ubiquinone and menaquinone of
Escherichia coli. Tetrahedron Lett (1997) 38(35):6181-6184.
This is the first glimpse of the role of a 2-C-methyl-D-erythritol derivative in
isoprenoid biosynthesis in Escherichia coli. This information triggered the
discovery of DXP reductoisomerase, followed by the characterization of the
role of fosmidomycin.
160 Current Opinion in Investigational Drugs 2004 Vol 5 No 2
15. Sprenger GA, Schrken U, Wiegert T, Grolle S, de Graaf AA, Taylor SV,
Begley TP, Bringer-Meyer S, Sahm H: Identification of a thiamin-
dependent synthase in Escherichia coli required for the formation
of 1-deoxy-D-xylulose 5-phosphate precursor to isoprenoids,
thiamin and pyridoxol. Proc Natl Acad Sci USA (1997) 94(24):12857-
12862.
16. Lois LM, Campos N, Putra SR, Danielsen K, Rohmer M, Boronat A:
Cloning and characterization of a gene from Escherichia coli
encoding a transketolase-like enzyme that catalyzes the synthesis
of D-1-deoxyxylulose 5-phosphate, a common precursor for
isoprenoid, thiamin and pyridoxol biosynthesis. Proc Natl Acad Sci
USA (1998) 95(5):2105-2110.
17. Lange BM, Wildung MR, McCaskill D, Croteau R: A family of
transketolases that directs isoprenoid biosynthesis via a
mevalonate-independent pathway. Proc Natl Acad Sci USA (1998)
95(5):2100-2104.
18. Hahn FM, Eubanks LM, Testa CA, Blagg BSJ, Baker JA, Poulter CD: 1-
Deoxy-D-xylulose 5-phosphate synthase, the gene product of open
reading frame (ORF) 2816 and ORF 2895 in Rhodobacter
capsulatus. J Bacteriol (2001) 183(1):1-11.
19. Cane DE, Chow C, Lillo A, Kang I: Molecular cloning, expression and
characterization of the first three genes in the mevalonate-
independent isoprenoid pathway in Streptomyces coelicolor.
Bioorg Med Chem (2001) 9(6):1467-1477.
20. Li S-M, Wemakor E, Heide L: PCR amplification of 1-deoxy-D-
xylulose 5-phosphate synthase (dxs) gene from different
Streptomyces species: Evidence for the existence of two dxs gene
families. Mol Biol Today (2001) 2(5):61-66.
21. Kuzuyama T, Takahashi S, Watanabe H, Seto H: Direct formation of
2-C-methyl-D-erythritol 4-phosphate from 1-deoxy-D-xylulose 5-
phosphate by 1-deoxy-D-xylulose 5-phosphate reductoisomerase, a
new enzyme in the non-mevalonate pathway to isopentenyl
diphosphate. Tetrahedron Lett (1998) 39(25):4509-4512.
22. Takahashi S, Kuzuyama T, Watanabe H, Seto H: A 1-deoxy-D-xylulose
5-phosphate reducto-isomerase catalyzes the formation of 2-C-
methyl-D-erythritol 4-phosphate in an alternative nonmevalonate
pathway for terpenoid biosynthesis. Proc Natl Acad Sci USA (1998)
95(17):9879-9884.
Identification of the dxr gene and characterization of the DXP
reductoisomerase opened the way for the identification of fosmidomycin.
23. Koppisch AT, Fox DT, Blagg BSJ, Poulter CD: E coli MEP synthase:
Steady-state kinetic analysis and substrate binding. Biochemistry
(2002) 41(1):236-243.
24. Hoeffler JF, Tritsch D, Grosdemange-Billiard C, Rohmer M: Isoprenoid
biosynthesis via the methylerythritol phosphate pathway.
Mechanistic investigations of the 1-deoxy-D-xylulose 5-phosphate
reductoisomerase. Eur J Biochem (2002) 269(18):4446-4457.
25. Rohdich F, Wungsintaweekul J, Fellermeier M, Sagner S, Herz S, Kis K,
Eisenreich W, Bacher A, Zenk MH: Cytidine 5'-triphosphate-
dependent biosynthesis of isoprenoids: YgbP protein catalyzes the
formation of 4-diphosphocytidyl-2-C-methylerythritol. Proc Natl
Acad Sci USA (1999) 96(21):11758-11763.
Further metabolism of MEP is described, leading to the identification of
three additional steps downstream of the formation of MEP. See also
references [26] and [27].
26. Lttgen H, Rohdich F, Herz S, Wungsintaweekul J, Hecht S, Schuhr
CA, Fellermeier M, Sagner S, Zenk MH, Bacher A, Eisenreich W:
Biosynthesis of terpenoids: YchB protein of Escherichia coli
phosphorylates the 2-hydroxy group of 4-diphosphocytidyl-2C-
methyl-D-erythritol. Proc Natl Acad Sci USA (2000) 97(3):1062-
1067.
Further metabolism of MEP is described, leading to the identification of
three additional steps downstream of the formation of MEP. See also
references [25] and [27].
27. Herz S, Wungsintaweekul J, Schuhr CA, Hecht S, Lttgen H, Sagner S,
Fellermeier M, Eisenreich W, Zenk MH, Bacher A, Rohdich F:
Biosynthesis of terpenoids: YgbB protein converts 4-
diphosphocytidyl-2-C-methyl-D-erythritol 2-phosphate to 2C-
methyl-D-erythritol 2,4-cyclodiphosphate. Proc Natl Acad Sci USA
(2000) 97(6):2486-2490.
Further metabolism of MEP is described, leading to the identification of
three additional steps downstream of the formation of MEP. See also
references [25] and [26]. the pathway.
28. Kuzuyama T, Takagi M, Kaneda K, Dairi T, Seto H: Formation of 4-
(cytidine 5'-diphospho)-2-C-methyl-D-erythritol 4-phosphate by 2-C-
methyl-D-erythritol 4-phosphate cytidyl transferase, a new enzyme in
the nonmevalonate pathway. Tetrahedron Lett (2000) 41(5):703-706.
29. Kuzuyama T, Takagi M, Kaneda K, Watanabe H, Dairi T, Seto H:
Studies on the non-mevalonate pathway: Conversion of 4-(cytidine
5'-diphospho)-2C-methyl-D-erythritol 4-phosphate to its 2-phospho
derivative by 4-(cytidine 5'-diphospho)-2-C-methyl-D-erythritol
kinase. Tetrahedron Lett (2000) 41(16):2925-2928.
30. Takagi M, Kuzuyama T, Kaneda K, Watanabe H, Dairi T, Seto H:
Studies on the nonmevalonate pathway: Formation of 2-C-methyl-
D-erythritol 2,4-cyclodiphosphate from 2-phospho-4-(cytidine 5'-
diphospho)-2-C-methyl-D-erythritol. Tetrahedron Lett (2000)
41(18):3395-3398.
31. Giner J-L, Jaun B, Arigoni D: Biosynthesis of isoprenoids in
Escherichia coli: The fate of the 3-H and 4-H atoms of 1-deoxy-D-
xylulose. J Chem Soc Chem Commun (1998):1857-1858.
The different labeling patterns found in IPP and DMAPP after incorporation
2
of [4- H]DX highlighted for the first time a puzzling problem, which was later
solved by the identification of the separate formation of IPP and DMAPP from
a common intermediate (E)-4-hydroxy-3-methylbut-2-enyl diphosphate.
32. Charon L, Hoeffler JF, Pale-Grosdemange C, Lois LM, Campos N,
Boronat A, Rohmer M: Deuterium labelled isotopomers of 2-C-
methyl-D-erythritol as tools for the elucidation of the 2-C-methyl-D-
erythritol 4-phosphate pathway for isoprenoid biosynthesis.
Biochem J (2000) 346(3):737-742.
33. Hahn FM, Hurlburt AP, Poulter CD: Escherichia coli open reading
frame 696 is idi, a nonessential gene encoding isopentenyl
diphosphate isomerase. J Bacteriol (1999) 181(15):4499-4504.
The rather unexpected non-essentiality of the IPP isomerase in Escherichia coli
pointed toward the independent synthesis of IPP and DMAPP in the MEP pathway.
34. Rodriguez-Concepcin M, Campos M, Lois LM, Maldonado C, Hoeffler
JF, Grosdemange-Billiard C, Rohmer M, Boronat A: Genetic evidence
of branching in the isoprenoid pathway for the production of
isopentenyl diphosphate and dimethylallyl diphosphate in
Escherichia coli. FEBS Lett (2000) 473(3):328-332.
Genetic studies provided definite proof that IPP and DMAPP are
independently synthesized from a common precursor of the MEP pathway.
This was the solution to the problems discussed in references [31,32,33].
35. Campos N, Rodriguez-Concepcin M, Seemann M, Rohmer M, Boronat
A: Identification of gcpE as a novel gene of the 2-C-methyl-D-
erythritol 4-phosphate pathway for isoprenoid biosynthesis in
Escherichia coli. FEBS Lett (2001) 488(3):170-173.
36. Altincicek B, Kollas AK, Sanderbrand S, Wiesner J, Hintz M, Beck E,
Jomaa H: GcpE is involved in the 2-C-methyl-D-erythritol 4-
phosphate pathway for isoprenoid biosynthesis in Escherichia coli.
J Bacteriol (2001) 183(8):2411-2416.
37. Cunningham FX Jr, Lafond TP, Gantt E: Evidence for a role for LytB
in the nonmevalonate pathway of isoprenoid biosynthesis. J
Bacteriol (2000) 182(20):5841-5848.
38. Altincicek B, Kollas AK, Eberl M, Wiesner J, Sanderbrand S, Hintz M,
Beck E, Jomaa H: LytB, a novel gene of the 2-C-methyl-D-erythritol
4-phosphate pathway of isoprenoid biosynthesis in Escherichia
coli. FEBS Lett (2001) 499(1-2):37-40.
39. Hecht S, Eisenreich W, Adam P, Amslinger S, Kis K, Bacher A, Arigoni
D, Rohdich F: Studies on the nonmevalonate pathway to terpenes:
The role of the GcpE (IspG) protein. Proc Natl Acad Sci USA (2001)
98(26):14837-14842.
40. Wolff M, Seemann M, Grosdemange-Billiard C, Tritsch D, Campos N,
Rodriguez-Concepcin M, Boronat A, Rohmer M: Isoprenoid biosynthesis
via the methylerythritol phosphate pathway. (E)-4-Hydroxy-3-methylbut-
2-enyl diphosphate: Chemical synthesis and formation from
methylerythritol cyclodiphosphate by a cell-free system from
Escherichia coli. Tetrahedron Lett (2002) 43(14):2555-2559.
41. Seemann M, Bui BT, Wolf M, Tritsch D, Campos N, Boronat A, Marquet A,
Rohmer M: Isoprenoid biosynthesis through the methylerythritol
phosphate pathway: The (E)-4-hydroxy-3-methylbut-2-enyl diphosphate
synthase (GcpE) is a [4Fe-4S] protein. Angew Chem Int Ed (2002)
41(22):4337-4339.
GcpE is a metallo-enzyme with a [4Fe-4S] prosthetic group. The dioxygen
sensitivity of the protein represented a bottleneck in the further elucidation of
 Isoprenoid biosynthesis as a novel target for antibacterial and antiparasitic drugs Rohmer et al 161
42. Kollas AK, Duin EC, Eberl M, Altincicek B, Hintz M, Reichenberg A,
Henschker D, Henne A, Steinbrecher I, Ostrovsky DN, Hedderich R et
al: Functional characterization of GcpE, an essential enzyme of the
non-mevalonate pathway of isoprenoid biosynthesis. FEBS Lett
(2002) 532(3):432-436.
43. Rohdich F, Hecht S, Grtner K, Adam P, Krieger C, Amslinger S,
Arigoni D, Bacher A, Eisenreich W: Studies on the nonmevalonate
terpene biosynthetic pathway: Metabolic role of IspH (LytB)
protein. Proc Natl Acad Sci USA (2002) 99(3):1158-1163.
44. Adam P, Hecht S, Eisenreich W, Kaiser J, Grwert T, Arigoni D, Bacher
A, Rohdich F: Biosynthesis of terpenes: Studies on 1-hydroxy-2-
methyl-2-(E)-butenyl 4-diphosphate reductase. Proc Natl Acad Sci
USA (2002) 99(19):12108-12119.
45. Altincicek B, Duin EC, Reichenberg A, Hedderich R, Kollas AC, Hintz M,
Wagner S, Wiesner J, Beck E, Jomaa H: LytB protein catalyzes the
terminal step of the 2-C-methyl-D-erythritol-4-phosphate pathway
for isoprenoid biosynthesis. FEBS Lett (2002) 532(3):437-440.
46. Rohdich F, Zepeck F, Adam P, Hecht S, Kaiser J, Laupitz R, Grwert T,
Amslinger S, Eisenreich W, Bacher A, Arigoni D: The deoxyxylulose
phosphate pathway of isoprenoid biosynthesis: Studies on the
mechanisms of the reaction catalyzed by IspG and IspH proteins.
Proc Natl Acad Sci USA (2003) 100(4):1586-1591.
47. Wolff M, Seemann M, Tse Sum Bui B, Frapart Y, Tritsch D, Estrabot
AG, Rodriguez-Concepcin M, Boronat A, Marquet A, Rohmer M:
Isoprenoid biosynthesis via the methylerythritol phosphate
pathway: The (E)-4-hydroxy-3-methylbut-2-enyl diphosphate
reductase (LytB/IspH) from Escherichia coli is a [4Fe-4S] protein.
FEBS Lett (2003) 541(1-3):115-120.
48. Boucher Y, Doolittle WF: The role of lateral gene transfer in the
evolution of isoprenoid biosynthesis pathways. Mol Microbiol (2000)
37(4):703-716.
Biochemical proof and genome analyses for the presence of the MEP
pathway are presented in this paper. They show the broad distribution of the
MEP pathway in bacteria.
49. Lange BM, Rujan T, Martin W, Croteau R: Isoprenoid biosynthesis:
The evolution of two ancient and distinct pathways across
genomes. Proc Natl Acad Sci USA (2000) 97(24):13172-13177.
50. Eberl M, Hintz M, Reichenberg A, Kollas AK, Wiesner J, Jomaa H:
Microbial isoprenoid biosynthesis and human T cell activation.
FEBS Lett (2003) 544(1-3):4-10.
51. Seto H, Watanabe H, Furihata K: Simultaneous operation of the
mevalonate and non-mevalonate pathways in the biosynthesis of
isopentenyl diphosphate in Streptomyces aeriouvifer. Tetrahedron
Lett (1996) 37(44):7979-7982.
52. Jomaa H, Wiesner J, Sanderbrand S, Altincicek B, Weidemeyer C, Hintz
M, Trbachova I, Eberl M, Zeidler M, Lichtenthaler HK, Soldati D et al:
Inhibitors of the nonmevalonate pathway of isoprenoid biosynthesis
as antimalarial drugs. Science (1999) 285(5433):1573-1576.
Fosmidomycin inhibits the growth of Plasmodium spp and its target is the
DXP reductoisomerase of the parasite. This paper presents the first evidence
for the MEP pathway as a target for antiparasitic drugs.
53. Rohdich F, Wungsintaweekul J, Eisenreich W, Richter G, Schuhr CA,
Hecht S, Zenk MH, Bacher A: Biosynthesis of terpenoids: 4-
Diphosphocytidyl-2C-methyl-D-erythritol synthase of Arabidopsis
thaliana. Proc Natl Acad Sci USA (2000) 97(12):6451-6456.
54. Rohdich F, Eisenreich W, Wungsintaweekul J, Hecht S, Schuhr CA,
Bacher A: Biosynthesis of terpenoids. 2C-Methyl-D-erythritol 2,4-
cyclodiphosphate synthase (IspF) from Plasmodium falciparum.
Eur J Biochem (2001) 268(11):3190-3197.
55. Querol J, Rodriguez-Concepcion M, Boronat A, Imperial S: Essential role of
residue H49 for activity of Escherichia coli 1-deoxy-D-xylulose 5-
phosphate synthase, the enzyme catalyzing the first step of the 2-C-
methyl-D-erythritol 4-phosphate pathway for isoprenoid synthesis.
Biochem Biophys Res Commun (2001) 289(1):155-160.
56. Kuzuyama T, Takahashi S, Takagi M, Seto H: Characterization of 1-
deoxy-D-xylulose 5-phosphate reductoisomerase, an enzyme
involved in isopentenyl diphosphate biosynthesis, and
identification of its catalytic amino acid residues. J Biol Chem
(2000) 275(26):19928-19932.
57. Yajima S, Nonaka T, Kuzuyama T, Seto H, Ohsawa K: Crystal structure
of 1-deoxy-D-xylulose 5-phosphate reductoisomerase complexed
with cofactors: Implications of a flexible loop movement upon
substrate binding. J Biochem (2002) 131(3):313-317.
This paper, and references [54-56,58-60], present the three dimensional
structures of MEP pathway enzymes. They provide first insights into the
design of inhibitors.
58. Reuter K, Sanderbrand S, Jomaa H, Wiesner J, Steinbrecher I, Beck E,
Hintz M, Klebe G, Stubbs MT: Crystal structure of 1-deoxy-D-
xylulose-5-phosphate reductoisomerase, a crucial enzyme in the
non-mevalonate pathway for isoprenoid biosynthesis. J Biol Chem
(2002) 277(7):5378-5384.
59. Richard SB, Bowman ME, Kwiatkowski W, Kang I, Chow C, Lillo AM,
Cane DE, Noel JP: Structure of 4-diphosphocytidyl-2-C-
methylerythritol synthetase involved in mevalonate-independent
isoprenoid biosynthesis. Nat Struct Biol (2001) 8(7):641-648.
60. Kemp LE, Bond CS, Hunter WN: Crystallization and preliminary x-ray
diffraction studies of recombinant Escherichia coli 4-
diphosphocytidyl-2-C-methyl-D-erythritol synthetase. Acta
Crystallogr D Biol Crystallogr (2001) 57(8):1189-1191.
61. Miallau L, Alphey MS, Kemp LE, Leonard GA, McSweeney SM, Hecht
S, Bacher A, Eisenreich W, Rohdich F, Hunter WN: Biosynthesis of
isoprenoids: Crystal structure of 4-diphosphocytidyl-2C-methyl-D-
erythritol kinase. Proc Natl Acad Sci USA (2003) 100(16):9173-9178.
62. Wada T, Kuzuyama T, Satoh S, Kuramitsu S, Yokoyama S, Unzai S,
Tame JR, Park SY: Crystal structure of 4-(cytidine 5'-diphospho)-2-
C-methyl-D-erythritol kinase, an enzyme in the non-mevalonate
pathway for isoprenoid biosynthesis. J Biol Chem (2003)
278(32):30022-30027.
63. Richard SB, Ferrer JL, Bowman ME, Lillo AM, Tetzlaff CN, Cane DE,
Noel JP: Structure and mechanism of 2-C-methyl-D-erythritol 2,4-
cyclodiphosphate synthase. An enzyme in the mevalonate-
independent isoprenoid biosynthetic pathway. J Biol Chem (2002)
277(10):8667-8672.
64. Steinbacher S, Kaiser J, Wungsintaweekul J, Hecht S, Eisenreich W,
Gerhardt S, Bacher A, Rohdich F: Structure of 2C-methyl-D-erythritol-
2,4-cyclodiphosphate synthase involved in mevalonate-independent
biosynthesis of isoprenoids. J Mol Biol (2002) 316(1):79-88.
65. Kemp LE, Bond CS, Hunter WN: Structure of 2C-methyl-D-erythritol
2,4-cyclodiphosphate synthase: An essential enzyme for
isoprenoid biosynthesis and target for antimicrobial drug
development. Proc Natl Acad Sci USA (2002) 99(10):6591-6596.
66. Kishida H, Unzai S, Kuzuyama T, Takagi M, Terada T, Shirouzu M,
Yokoyama S, Tame JRH, Park SY: Structure and catalytic
mechanisms of 2-C-methyl-D-erythritol cyclodiphosphate (MECDP)
synthase, an enzyme in the non-mevalonate pathway for
isoprenoid biosynthesis. Acta Cryst (2003) D59:23-31.
67. Okuhara M, Kuroda Y, Goto T, Okamoto M, Terano H, Kohsaka M, Aoki
H, Imanaka H: Studies on new phosphonic acid antibiotics. III.
Isolation and characterization of FR-31564, FR-32863 and FR-
33289. J Antibiot (1980) 33(1):24-28.
68. Mine Y, Kamimura M, Nonoyama S, Nishida M, Goto S, Kuwahara S: In
vitro and in vivo antibacterial activities of FR-31564, a new
phosphonic acid antibiotic. J Antibiot (1980) 33(1):36-43.
69. Shigi Y: Inhibition of bacterial isoprenoid synthesis by
fosmidomycin, a phosphonic acid-containing antibiotic. J
Antimicrob Chemother (1989) 24(2):131-145.
First evidence that fosmidomycin may interfere with isoprenoid biosynthesis
in bacteria that do not use MVA.
70. Kuzuyama T, Shimizu T, Takahashi S, Seto H: Fosmidomycin, a specific
inhibitor of 1-deoxy-D-xylulose 5-phosphate reductoisomerase in the
nonmevalonate pathway for terpenoid biosynthesis. Tetrahedron Lett
(1998) 39(43):7913-7916.
Fosmidomycin is identified as an inhibitor of the DXP reductoisomerase,
indicating that isoprenoid biosynthesis via the MEP pathway is a potential
target for antibacterial drugs.
71. Grolle S, Bringer-Meyer S, Sahm H: Isolation of the dxr gene of
Zymomonas mobilis and characterization of the 1-deoxy-D-xylulose
5-phosphate reductoisomerase. FEMS Microbiol Lett (2000)
191(1):131-137.
162 Current Opinion in Investigational Drugs 2004 Vol 5 No 2
72. Zeidler JG, Schwender J, Mller C, Wiesner J, Weidemeyer C, Beck E,
Jomaa H, Lichtenthaler HK: Inhibition of the non-mevalonate 1-
deoxy-D-xylulose-5-phosphate pathway of plant isoprenoid
biosynthesis by fosmidomycin. Z Naturforsch (1998) 53C:980-986.
73. Fellermeier M, Kis K, Sagner S, Maier U, Bacher A, Zenk MH: Cell-free
conversion of 1-deoxy-D-xylulose 5-phosphate and 2-C-methyl-D-
erythritol 4-phosphate into -carotene in higher plants and its
inhibition by fosmidomycin. Tetrahedron Lett (1999) 40(14):2743-2746.
74. Reichenberg A, Wiesner J, Weidemeyer C, Dreiseidler E, Sanderbrand
S, Altincicek B, Beck E, Schlitzer M, Jomaa H: Diaryl ester prodrugs of
FR90008 with improved in vivo antimalarial activity. Bioorg Med
Chem Lett (2001) 11(6):833-835.
75. Zubrzycki IZ, Blatch GL: DFT study of a substrate and inhibitors of 1-
deoxy-2-xylulose-5-phosphate reductoisomerase - the potential
novel target molecule for anti-malaria drug development. J Mol
Model (2001) 7(10):378-383.
76. Steinbacher S, Kaiser J, Eisenreich W, Huber R, Bacher A, Rohdich F:
Structural basis of fosmidomycin action revealed by the complex
with 2-C-methyl-D-erythritol synthase (IspC). Implications for the
catalytic mechanism and anti-malaria drug development. J Biol
Chem (2003) 278(20):18401-18407.
Crystal structure of the fosmidomycin/DXR complex points out the features
of the enzyme active site.
77. Sakamoto Y, Furukawa S, Ogihara H, Yamasaki M: Fosmidimycin
resistance in adenylate cyclase (cya) mutants of Escherichia coli.
Biosci Biotechnol Biochem (2003) 67(9):2030-2033.
78. Altincicek B, Hintz M, Sanderbrand S, Wiesner J, Beck E, Jomaa H:
Tools for discovery of the 1-deoxy-D-xylulose 5-phosphate (DXP)
synthase and DXP reductoisomerase: An approach with enzymes
from the pathogenic bacterium Pseudomonas aeruginosa. FEMS
Microbiol Lett (2000) 190(2):329-333.
79. Mueller C, Schwender J, Zeidler J, Lichtenthaler HK: Properties and
inhibition of the first two enzymes of the non-mevalonate pathway
of isoprenoid biosynthesis. Biochem Soc Trans (2000) 28(6):792-793.
80. Wungsintaweekul J, Herz S, Hecht S, Eisenreich W, Feicht R, Rohdich F,
Bacher A, Zenk MH: Phosphorylation of 1-deoxy-D-xylulose by D-
xylulokinase of Escherichia coli. Eur J Biochem (2001) 268(2):310-316.
81. Bouvet D, O'Hagan D: The synthesis of 1-fluoro- and 1,1-difluoro-
analogues of 1-deoxy-D-xylulose. Tetrahedron (1999) 55(34):10481-
10486.
82. Ostrovsky D, Kharatian E, Dubrovsky T, Ogrel O, Shipanova I, Sibeldina
L: The ability of bacteria to synthesize a new cyclopyrophosphate
correlates with their tolerance to redox-cycling drugs: On a
crossroad of chemotherapy, environmental toxicology and
immunobiochemical problems. Biofactors (1992) 4(1):63-68.
83. Ostrovsky D, Kharatian E, Malarova I, Shipanova I, Sibeldina L,
Shashkov A, Tantsirev G: Synthesis of a new organic pyrophosphate
in large quantities is induced in some bacteria by oxidative stress.
Biofactors (1992) 3(4):261-264.
84. Ostrovsky D, Diomina G, Lysak E, Matveeva E, Ogrel O, Trutko S:
Effect of oxidative stress on the biosynthesis of 2-C-methyl-D-
erythritol-2,4-cyclopyrophosphate and isoprenoids by several
bacterial strains. Arch Microbiol (1998) 171(1):69-72.
85. Kaneda K, Kuzuyama T, Takagi M, Hayakawa Y, Seto H: An unusual
isopentenyl diphosphate isomerase found in the mevalonate
pathway gene cluster from Streptomyces sp strain CL190. Proc Natl
Acad Sci USA (2001) 98(3):932-937.
Unexpected discovery of a novel type of bacterial IPP isomerase that may
represent an additional target for antibacterial drugs.
86. Rohmer M: Isoprenoid biosynthesis via the mevalonate-
independent route, a novel target for antibacterial drugs? Prog Drug
Res (1998) 50:134-154.
87. Fujisaki S, Ohnuma SI, Horiuchi T, Takahashi I, Tsukui S, Nishimura Y,
Nishino T, Kitabatake M, Inokuchi H: Cloning of a gene from
Escherichia coli that confers resistance to fosmidomycin as a
consequence of amplification. Gene (1996) 175(1-2):83-87.
This is the first report for the resistance to fosmidomycin, an antibiotic acting
on the MEP pathway in bacteria.
88. Missinou MA, Borrmann S, Schindler A, Issifou S, Adegnika AA,
Matsiegui PB, Binder R, Lell B, Wiesner J, Baranek T, Jomaa H et al:
Fosmidomycin for malaria. Lancet (2002) 360(9349):1941-1942
89. Kuemmerle HP, Murakawa T, Sakamoto H, Sato N, Konishi T, De
Santis F: Fosmidomycin, a new phosphonic acid antibiotic. Part II:
1. Human pharmacokinetics. 2. Preliminary early phase IIa clinical
studies. Int J Clin Pharmacol Ther Toxicol (1985) 23(10):521-528.
90. Rosa-Putra S, Disch A, Bravo JM, Rohmer M: Distribution of
mevalonate and glyceraldehyde 3-phosphate/pyruvate routes for
isoprenoid biosynthesis in some Gram-negative bacteria and
mycobacteria. FEMS Microbiol Lett (1998) 164(1):169-175.
91. Shigemori H, Komaki H, Yazawa K, Mikami Y, Nemoto A, Tanaka Y,
Kobayashi J: Biosynthesis of diterpenoid moiety of brasilicardin A
via non-mevalonate pathway in Nocardia brasiliensis. Tetrahedron
Lett (1999) 40(23):4353-4354.
92. Rosa-Putra S, Hemmerlin A, Epperson J, Bach TJ, Guerra LH, Rohmer
M: Zeaxanthin and menaquinone-7 biosynthesis in
Sphingobacterium multivorum via the methylerythritol phosphate
pathway. FEMS Microbiol Lett (2001) 204(2):347-353.
93. Fontana A, Kelly MT, Prasad JD, Andersen RJ: Evidence for the
biosynthesis of squalene via the methylerythritol phosphate
pathway in a Streptomyces sp obtained from a marine sediment. J
Org Chem (2001) 66(18):6202-6206.
94. Cane DE, Rossi T, Tillmann M, Pachlatko JP: Stereochemical studies of
isoprenoid biosynthesis of pentalenolactone from [U-13C6]glucose
and [6-2H2]glucose. J Am Chem Soc (1981) 103(7):1838-1843.
95. Spiteller D, Jux A, Piel J, Boland W: Feeding of [5,5-2H2]-1-desoxy-D-
xylulose and [4,4,6,6,6-2H5]-mevalolactone to a geosmin-producing
Streptomyces sp and Fossombronia pusilla. Phytochemistry (2002)
61(7):827-834.
96. Britton G, Goodwin TW, Lockley WJS, Mundy AP, Patel NJ, Englert G:
Stereochemistry of cyclization in carotenoid biosynthesis; use of
13
C-labelling to elucidate the stereochemical behaviour of the C-1
methyl substituents during zeaxanthin biosynthesis in a
Flavobacterium. J Chem Soc Chem Commun (1979):27-28.
97. Smeds A, Joutsjoki T, Patva A: Identification of a gene cluster for the
mevalonate pathway in Lactobacillus helveticus. DNA Sequence
(2001) 12(3):187-190.
98. Kohl W, Gloe A, Reichenbach H: Steroids from the myxobacterium
Nannocystis exedens. J Gen Microbiol (1983) 129:1629-1635.
99. Eisenreich W, Bacher A, Berry A, Bretzel W, Hmbelin M, Lopez-Ulibarri
R, Mayer AF, Yeliseev A: Biosynthesis of zeaxanthin via mevalonate
in Paracoccus species strain PTA-3335. A product-based
retrobiosynthetic study. J Org Chem (2002) 67(3):871-875.
Update on Biochemistry of Plant Volatiles
Biochemistry of Plant Volatiles1
Natalia Dudareva*, Eran Pichersky, and Jonathan Gershenzon
Department of Horticulture and Landscape Architecture, Purdue University, West Lafayette,
Indiana 47907 (N.D.); Department of Molecular, Cellular, and Developmental Biology,
University of Michigan, Ann Arbor, Michigan 48109 (E.P.); and Max Planck Institute
for Chemical Ecology, Beutenberg Campus, D007745 Jena, Germany (J.G.)
Plants have a penchant for perfuming the atmo-
sphere around them. Since antiquity it has been
known that both floral and vegetative parts of many
species emit substances with distinctive smells. The
discovery of the gaseous hormone ethylene 70 years
ago brought the realization that at least some of the
compounds emitted may have physiological signifi-
cance without any distinctive smell to humans. At
present, more than 1,000 low Mr organic compounds
have been reported to be emitted from plants, al-
though a comprehensive list is available only for floral
volatiles (Knudsen et al., 1993).
Our knowledge of the occurrence and distribution of
plant volatiles has been significantly extended in the
last 15 years thanks to the adoption of simple, sensitive
methods for headspace sampling and the availability of
relatively inexpensive bench-top instruments for gas
chromatography-mass spectrometry. The substances
reported are largely lipophilic products with molecular
masses under 300. Most can be assigned to the follow-
ing classes (in order of decreasing size): terpenoids,
fatty acid derivatives including lipoxygenase path-
way products, benzenoids and phenylpropanoids,
C5-branched compounds, and various nitrogen and
sulfur containing compounds. Nearly all of these classes
are emitted from vegetative parts as well as flowers
(Knudsen et al., 1993), and some are even emitted from
roots (Steeghs et al., 2004). A major discovery of the last
decade is that plants commonly emit much greater
amounts and varieties of volatiles after herbivore damage,
and not just from the site of injury (Pare and Tumlinson,
1999).
1
This work was supported by the U.S. National Science Founda-
tion (grant nos. MCB0212802 to N.D., MCB0312466 and IBN
0211697 to E.P.), by the U.S. Department of Agriculture (grant no.
20033531813619 to N.D.), by the U.S. Israel Binational Agriculture
Research and Development funds (grant nos. US343703 to N.D.
and IS333202 to E.P.), by the Fred Gloeckner Foundation, Inc. (to
N.D.), by the German National Science Foundation (DE837/21 to
J.G.), by the European Commission (QLK3CT200201930 and
MRTNCT2003504720 to J.G.), and by the Max Planck Society (to
J.G.). This paper is contribution Number 17438 from Purdue Uni-
versity Agricultural Experimental Station.
* Corresponding author; e-mail dudareva@hort.purdue.edu; fax
17654940391.
www.plantphysiol.org/cgi/doi/10.1104/pp.104.049981.
Plant Physiology, August 2004, Vol. 135, pp. 18931902, www.plantphysiol.org 2004 American Society of Plant Biologists
Major progress in plant volatile research, as in other
areas of plant biology, has come from the use of
molecular and biochemical techniques. A large number
of genes encoding enzymes of volatile biosynthesis
have recently been reported. In vitro characterization of
the heterologously expressed enzymes, especially de-
termination of their substrate and product specificity,
has helped clarify the pathways of volatile formation.
In addition, investigation of the spatial and temporal
patterns of gene expression has provided new infor-
mation on the factors regulating the emission of plant
volatile compounds. In this update, we survey the
latest advances on the biosynthesis and regulation of
plant volatiles, beginning with a brief review of the
function of these substances.
FUNCTION OF PLANT VOLATILES
Perhaps the greatest mysteries surrounding vola-
tiles concern their function in the life of the plant.
While it is generally assumed that compounds emitted
from flowers serve to attract and guide pollinators
(Reinhard et al., 2004), only scattered attempts have
been made to demonstrate the ability of individual
substances to attract specific pollinators. Many floral
volatiles have anti-microbial or anti-herbivore activity
(DeMoraes et al., 2001; Friedman et al., 2002; Hammer
et al., 2003), and so could also act to protect valuable
reproductive parts of plants from enemies.
Among vegetative volatiles, the most intensively
studied substance is isoprene, a simple five-carbon
terpene emitted from the foliage of many woody spe-
cies (Sharkey and Yeh, 2001b). The function of isoprene
is still controversial, and this compound may act to
increase the tolerance of photosynthesis to high tem-
peratures by stabilizing the thylakoid membranes
(Sharkey et al., 2001a) or by quenching reactive oxygen
species (Loreto and Velikova, 2001). The release of
volatiles from vegetative organs following herbivore
damage seems to be a general property of plant spe-
cies. Contributions to this special issue cover herbivore-
induced volatiles from cabbage (Brassica oleracea;
Vuorinen et al., 2004b), cucumber (Cucumis sativus;
Mercke et al., 2004), Lotus japonicus (Arimura et al.,
2004b), and maize (Zea mays; Degen et al., 2004). These
substances have been demonstrated to serve as indirect
plant defenses. That is, they attract arthropods that
1893
Dudareva et al.
prey upon or parasitize herbivores, thus minimizing
further damage to plant tissue (Pare and Tumlinson,
1999; Dicke and Van Loon, 2000). In some cases,
herbivore-induced volatiles may also act as direct
defenses, repelling (DeMoraes et al., 2001; Kessler and
Baldwin, 2001) or intoxicating (Vancanneyt et al., 2001)
herbivores and pathogens (Andersen et al., 1994). The
possibility that these substances also act in plant-plant
communication has been discussed (Arimura et al.,
2000; Dicke and Bruin, 2001; Engelberth et al., 2004).
Herbivore-induced volatiles could additionally
have physiological roles within the plant, with their
release being a consequence of their volatility and
membrane solubility. Like isoprene, some herbivore-
induced monoterpenes and sesquiterpenes have the
potential to combine with various reactive oxygen
species (Hoffmann et al., 1997; Bonn and Moortgat,
2003), and so could protect against internal oxidative
damage (Delfine et al., 2000; Loreto et al., 2004b). In
fact, ozone fumigation has recently been reported to
promote the emission of herbivore-induced volatiles
(Vuorinen et al., 2004a). Yet, it is still unclear why
oxidative stress is likely to be significantly higher after
herbivore damage. Further studies are needed to help
elucidate the roles of these and other plant volatiles.
The growing number of reports on genes involved in
volatile formation, as described in the following sec-
tions, should enable investigators to manipulate vola-
tile emission and test its function in plants.
BIOSYNTHESIS OF VOLATILE TERPENES
Terpenes, as the largest class of plant secondary
metabolites, have many volatile representatives. The
majority of hemiterpenes (C5), monoterpenes (C10),
sesquiterpenes (C15), and even some diterpenes (C20)
have high enough vapor pressures at normal atmo-
spheric conditions to allow significant release into the
air. The basic pathway of volatile terpenoid biosynthe-
sis is conveniently treated in three phases: (1) formation
of the basic C5 units, (2) condensation of two or three C5
units to form C10, C15, or C20 prenyl diphosphates, and
(3) conversion of the resulting prenyl diphosphates to
end products.
The formation of basic C5 units, isopentenyl diphos-
phate (IPP) and dimethylallyl diphosphate (DMAPP)
proceeds via two alternative pathways: the long known
mevalonate pathway from acetyl-CoA and the meth-
ylerythritol phosphate pathway from pyruvate and
glyceraldehyde-3-phosphate, discovered only in the
last 10 years (for review, see Rodriguez-Concepcion
and Boronat, 2002). The methylerythritol phosphate
pathway, localized in the plastids, is thought to provide
IPP and DMAPP for hemiterpene, monoterpene, and
diterpene biosynthesis, while the cytosol-localized me-
valonate pathway provides C5 units for sesquiterpene
biosynthesis. However, metabolic cross-talk be- tion, acylation, and other reaction types. These are
tween the two pathways is prevalent (Schuhr et al.,
1894
2003), particularly in the direction from plastids to
cytosol (Laule et al., 2003). This issue contains two
contributions concerning the regulation of the basic
pathways in relation to isoprene formation. Although
produced largely by the plastidial pathway, isoprene
also seems to arise from extra-plastidial sources, but
there is apparently no cross-talk between the two
pathways in its formation (Loreto et al., 2004a). The
plastidial pathway is controlled by tight feedback
regulation on its first step, deoxyxylulose-5-phosphate
synthase (Wolfertz et al., 2004).
In the second phase of terpene biosynthesis, IPP and
DMAPP condense to form geranyl diphosphate (GPP),
farnesyl diphosphate (FPP), and geranylgeranyl di-
phosphate, the precursors of monoterpenes, sesquiter-
penes, and diterpenes, respectively. These reactions are
catalyzed by short-chain prenyltransferases (Koyama
and Ogura, 1999; Liang et al., 2002). FPP is synthesized
by a large family of homodimeric prenyltransferases
called FPP synthases. However, the situation regarding
GPP formation is more complex. While the GPP syn-
thases of Arabidopsis (Bouvier et al., 2000) and grand
fir (Abies grandis; Burke and Croteau, 2002) are homo-
dimers, like other short-chain prenyltransferases, those
reported from peppermint (Mentha 3 piperita) leaves
(Burke et al., 1999) and the flowers of snapdragon
(Antirrhinum majus) and Clarkia breweri (Tholl et al.,
2004) are unusual heterodimeric enzymes, with each
subunit being a member of the prenyltransferase pro-
tein family.
The third phase of terpene volatile biosynthesis
involves the conversion of the various prenyl diphos-
phates, DMAPP (C5), GPP (C10), FPP (C15), and gera-
nylgeranyl diphosphate (C20), to hemiterpenes
(isoprene and 2-methyl-3-buten-2-ol), monoterpenes,
sesquiterpenes, and diterpenes, respectively. These
reactions, carried out by a large family of enzymes
known as terpene synthases (Cane, 1999; Wise and
Croteau, 1999), produce the primary representatives of
each skeletal type. The investigation of terpene syn-
thases is a very active area of plant volatile research and
this issue contains four contributions describing the
isolation of genes of this type from Norway spruce
(Picea abies; Martin et al., 2004), Arabidopsis (Chen et al.,
2004), cucumber (Mercke et al., 2004), and L. japonicus
(Arimura et al., 2004b). These gene sequences give new
insights into the evolutionary origin and genetic regu-
lation of terpene synthases. One of the most outstand-
ing properties of these enzymes is their proclivity for
making multiple products from a single substrate.
Hence, there has been much curiosity about the carbo-
cationic reaction mechanism. The elucidation of the
first crystal structures of plant terpene synthases
(Starks et al., 1997; Whittington et al., 2002) now puts
this work on a much stronger experimental footing.
Many terpene volatiles are direct products of terpene
synthases, but others are formed through transforma-
tion of the initial products by oxidation, dehydrogena-
discussed in the following section.
Plant Physiol. Vol. 135, 2004
 Update on Biochemistry of Plant Volatiles

MODIFICATION REACTIONS THAT ENHANCE

THE VOLATILITY OF COMPOUNDS
The terpene pathways are essentially biosynthetic,
building up a carbon skeleton, and the immediate
products formed by the large family of terpene syn-
thases discussed above are mostly hydrocarbons, al-
though sometimes they contain a hydroxyl group (e.g.
linalool synthase produces linalool, a tertiary alcohol).
Such compounds are already fairly volatile. In contrast,
most other volatile compounds are produced through
the shortening of a carbon skeleton, often followed by
further modification, or simply by modification of the
existing carbon skeleton. Compounds that are already
somewhat volatile may also be modified, resulting in
enhanced volatility or changed olfactory properties.
The majority of these modifications involve the reduc-
tion or removal of carboxyl groups, the addition of
hydroxyl groups, and the formation of esters and
ethers. Each type of modification is catalyzed by
a group (or several groups) of related enzymes consti-
tuting protein families. Some of these protein families
had been previously recognized from biochemical re-
search into nonvolatile compounds but some were only
recently identified as part of the research into the
biosynthesis of plant volatiles. Modifications for which
enzymatic reactions and enzymes have been identified
in plants are described below.

Oxidation by Cytochrome P450 Enzymes

The P450 cytochrome oxidases have been well
characterized from a multitude of plant and animal
species, and are involved in numerous metabolic
pathways (Schuler, 1996). Not surprisingly, these en-
zymes have been found to be involved in many of the
reactions of volatile biosynthesis. The basic skeleton of
the monoterpenes and sesquiterpenes, discussed
above, is often modified by hydroxylation. For exam-
ple, 3-hydroxylation of limonene by a P450 enzyme is
the first step in the biosynthesis of menthol (Fig. 1A;
Lupien et al., 1999), a volatile flavor compound found
in mint, whereas a 6-hydroxylation of limonene by
another P450 enzyme is the first step in the biosynthe-
sis of another volatile spice, carvone, in the caraway
(Carum carvi) fruit (Bouwmeester et al., 1999). A P450
enzyme is also responsible for the conversion of the
sesquiterpene 5-epi-aristolochene to capsidiol, a dihy-
droxylated volatile compound (Ralston et al., 2001).
Some homoterepene compounds, for example the C11
compound 4,8-dimethyl-1,3,7-nonatriene which is of-
ten emitted from injured tissues, are believed to be
derived from terpenes by cleavage catalyzed by P450
enzymes (Fig. 1A), but such enzymes have not yet
been identified conclusively (Boland and Gabler, 1989;
Degenhardt and Gershenzon, 2000).
Cytochrome P450 enzymes are also very important in
the biosynethsis of volatile phenylpropenes such as
eugenol and the benzenoid vanillin. Both of these
compounds, found in a wide variety of species both
Plant Physiol. Vol. 135, 2004
Figure 1. Representative modification reactions leading to the
biosynthesis of compounds with enhanced or changed volatility
and olfactory properties. SAM, S-adenosyl-L-Met; CVOMT, chavicol
O-methytransferase; SAMT, S-adenosyl-L-Met:salicylic acid carboxyl
methytransferase; and BEAT, acetyl-coenzyme A:benzyl alcohol acetyl-
transferase.
in flowers and in leaves, are derived from Phe and share
with nonvolatile phenylpropanoids the earlier steps of
4-hydroxylation of cinnamate by 4CH, a P450 enzyme
(Frank et al., 1996), and 3-hydroxylation by the newly
discovered P450 enzyme that utilizes the shikimic or
quinic ester of coumarate rather than coumaric acid or
coumaroyl-CoA (Schoch et al., 2001; Gang et al., 2002a).
Cytochrome P450 enzymes are crucial in the bio-
synthesis of volatiles derived from fatty acids, and in
particular, in the octodecanoic pathway. Two different
P450 enzymes, 9-LOX and 13-LOX, can introduce
a peroxide into linoleic acid (18:3) at the respective
positions (Howe and Schilmiller, 2002). Subsequent
cleavage of the hydrocarbon chain by hydroperoxide
lyases produces nonadienal and 3-cis-hexenal, respec-
tively. The later is an important component of green
1895
Dudareva et al.
leaf volatiles, the mixtures of compounds that are
emitted when the leaf is damaged (Pichersky and
Gershenzon, 2002). In addition to the C6 aldehyde,
cleavage of linoleic acid at the 12 to 13 bond also
produces a C12 compound that can subsequently be
converted to jasmonic acid (Howe and Schilmiller,
2002). While jasmonic acid is not by itself volatile, its
methylester is (see below).
Oxidation by Dehydrogenases
NADP/NAD-dependent oxidoreductases are an-
other large and well-studied family of proteins with
representatives found to be involved in the biosynthe-
sis of volatiles. Such enzymes have been implicated in
the interconversion of volatile alcohols and aldehydes
(Fig. 1B). For example, apparently nonspecific alcohol
dehydrogenases can convert short-chain aldehydes
such as hexanal and 3-cis-hexenal to hexenol and 3-cis-
hexenol, alcohols that are also found in damaged
leaves (Bate et al., 1998). This lack of tight substrate
specificity was used to alter the aroma profile of ripe
tomato (Lycopersicon esculentum) fruit by genetic engi-
neering (Prestage et al., 1999). Some terpene alcohols
such as geraniol and carveol are converted to alde-
hydes by similarly nonspecific dehydrogenases (Hal-
lahan et al., 1995; Bouwmeester et al., 1998). Geranial
and neral (which are coproduced by the oxidation of
geraniol; the mixture is termed citral), have lemony
aroma and are found in many plants, and carvone
gives caraway its distinct flavor. And benzyl alcohol,
a major floral scent component in many Nicotianeae
species (Raguso et al., 2003) and elsewhere, is also
likely derived from benzaldehyde in a reversible re-
action catalyzed by a member of the NADP/NAD-
dependent oxidoreductases family (Fig. 1B; Boatright
et al., 2004).
Methylation of Hydroxyl Groups
A large portion of plant volatiles contain a methyl-
ated hydroxyl group (i.e. a methoxyl group). The
methyl group is usually added in a reaction catalyzed
by a methyltransferase (MT) in which S-adenosyl-L-
methionnine (SAM) serves as the methyl donor. All
plant methyltransferases appear to share a similar
SAM-binding domain; however, they fall into distinct
families that share little primary sequence similarity
elsewhere (Noel et al., 2003). A large family of meth-
yltransferases with members involved in the synthesis
of both volatile and nonvolatile small molecules (as been reported in numerous floral scents (Knudsen
opposed to proteins or nucleic acids) has been iden-
tified and designated as the Type I methyltransferase
family (Noel et al., 2003). Members of this MT family
have been shown to catalyze the 4-hydroxyl methyl-
ation of eugenol to form methyleugenol in flowers of
C. breweri and in the glands of basil (Ocimum basilicum),
and also the methylation of chavicol to methylchavicol
in the basil glands (Fig. 1C; Wang et al., 1997; Lew-
insohn et al., 2000; Gang et al., 2002b).
1896
3,5-Dimethoxytoluene, a major scent compound in
many hybrid roses, is produced from orcinol (3,5-
dihydroxytoluene) in two successive methylation re-
actions catalyzed by two very similar MTs, orcinol
OMTs (OOMT1 and OOMT2; Lavid et al., 2002; Scalliet
et al., 2002). Both enzymes can carry out both reac-
tions; however, OOMT1 is more catalytically efficient
with orcinol while OOMT2 is more catalytically effi-
cient with 3-methoxy,5-hydroxytoluene (Lavid et al.
2002). Chinese rose (Rosa chinensis) flowers make
a similar compound with three methoxyl groups,
1,3,5-trimethoxybenzene, which is synthesized from
1,3,5-trihydroxybenzene. OOMT1 and OOMT2 can
catalyze the methylation of the second and third
intermediates (1-methoxy,3,5-dihydroxybenzene and
1,3-dimethoxy,5-hydroxybenzene) but not the methyl-
ation of 1,3,5-trihydroxybenzene, also known as phlor-
oglucinol (Lavid et al., 2002; Scalliet et al., 2002). The
enzyme that methylates this compound, phlorogluci-
nol OMT (POMT), also belongs to the Type I MT
family, but is only distantly related to OOMT1 and
OOMT2 (Wu et al., 2004).
An important strawberry (Fragaria 3 ananassa)
aroma compound, 2,5-dimethyl-4-methoxy-3(2H)-
furanone, was shown to be produced by the action of
another Type I MT. This MT appears to be able to
methylate a wide range of substrates, including inter-
mediates of the lignin biosynthetic pathway such as
coniferal aldehyde and coniferal alcohol (Wein et al.,
2002).
Eugenol, mentioned above as the substrate of a eu-
genol MT and an important volatile spice on its own,
and vanillin, another important aroma compound,
also contain a 3-methoxyl group on their benzene
ring. The synthetic pathways of these compounds
share the first few steps with the lignin pathway. Gang
et al. (2002a) demonstrated that eugenol is derived
from a lignin intermediate past the para and meta
hydroxylation reactions and the 3-hydroxyl methyla-
tion, which is catalyzed by CCOMT (caffeoyl-CoA
OMT). This is also likely to be the case for vanillin,
although this has not yet been demonstrated conclu-
sively. CCOMT is a member of the Type II MT family of
plants (Noel et al., 2003).
Methylation of Carboxyl Groups
Some methyl esters are extremely wide spread in the
plant kingdom. For example, methylsalicylate has
and Tollsten, 1993), and it is also commonly emitted
from vegetative tissues under attack by insects or
parasites (Van Poecke et al., 2001; Chen et al., 2003). An
enzyme capable of methylating salicylic acid (SA),
salicylic acid carboxyl methyltransferase (SAMT) was
first reported from C. breweri flowers (Fig. 1D; Ross
et al., 1999). It has since been identified from several
other plant species (Fukami et al., 2002; Negre et al.,
2002, 2003; Pott et al., 2002, 2004; Chen et al., 2003).
Plant Physiol. Vol. 135, 2004

This enzyme, which uses SAM as the methyl donor,
defines a new type of plant MT known as Type III or
SABATH MT (after the first two letters of the names of
the first three enzymes identified in this family). Some
SAMT enzymes have been shown to be able to meth-
ylate also benzoic acid (BA), a compound identical to
SA except for lacking the 2-hydroxyl group present in
SA (for example, Pott et al., 2004). On the other hand,
benzoic acid carboxyl methyltransferase (BAMT), the
enzyme responsible for the snapdragon floral volatile
methylbenzoate, cannot methylate SA (Dudareva et al.,
2000; Murfitt et al., 2000).
There are 24 SABATH MTs encoded by the Arabi-
dopsis genome, including one enzyme that methylates
both SA and BA (Chen et al., 2003). It is not known if all
of these MTs are involved in the biosynthesis of volatile
compounds, but MTs belonging to the SABATH fam-
ily have been shown to be responsible for the three
consecutive metylation reactions in the biosynthesis of
caffeine, a nonvolatile compound, in Coffea arabica
(Uefuji et al., 2003). However, another Arabidopsis
SABATH MT was shown to methylate jasmonic acid to
form methyljasmonate (Seo et al., 2001). While this
molecule may act as an internal signal molecule in
Arabidopsis and other plant species, it is also emitted
from injured plants (Howe and Schilmiller, 2002), and
has also been reported in the floral scent of several plant
species (Knudsen et al., 1993) where it is likely to be
formed by similar enzymes.
Acylation
Acylation, most often with an acetyl moiety but also
with larger acyls such butanoyl or benzoyl acyls, to
make volatile compounds is also common. In all known
examples, such plant volatile esters are synthesized by
a recently discovered family of plant acyltransferases
called BAHD, after the first letter of the first four
enzymes identified (St-Pierre and De Luca, 2000). The
basic reaction catalyzed by these enzymes is the trans-
fer of an acyl group from an acyl-CoA intermediate to
the hydroxyl group of an alcohol (Fig. 1E). Many BAHD
enzymes are involved in the synthesis of nonvolatile
compounds such as acylated alkaloids or taxol deriv-
atives (St-Pierre et al., 1998; Walker and Croteau, 2000),
or in early steps of pathways that may lead to the
synthesis of volatiles such as eugenol (Gang et al.,
2002a). BAHD acyltransferases directly involved in
volatile synthesis include benzyl alcohol acetyl-CoA
transferase from C. breweri flowers, which produced
benzyl acetate (Fig. 1E; Dudareva et al., 1998); benzyl
alcohol benzoyl-CoA transferase, which produces ben-
zylbenzoate in flowers of Clarkia (DAuria et al., 2002)
and petunia (Petunia hybrida; Boatright et al., 2004), and
in tobacco mosaic virus-infected leaves of tobacco
(Nicotiana tabacum; DAuria et al., 2002); and 3-cis-
hexen-1-ol acetyl CoA transferase, which produces
3-cis-hexenyl acetate (a green leaf volatile) and is in-
duced in damaged leaves of Arabidopsis (DAuria et al.,
2002).
Plant Physiol. Vol. 135, 2004
Update on Biochemistry of Plant Volatiles
The BAHD enzymes often show wide substrate
specificity for both the acyl moiety and the alcohol
moiety. For example, the petunia benzyl alcohol
benzoyl-CoA transferase enzyme can also transfer an
acetyl moiety to the alcohol phenylethanol, producing
phenylethylacetate (Boatright et al., 2004). Similarly,
a BAHD enzyme from ripening strawberry (Fragaria
spp) fruit, can use a series of acyl moieties such as
acetyl, butanyl, and hexanyl, and transfer them to
various alcohols such as heptanol, octanol, and gera-
niol (Aharoni et al., 2000; Beekwilder et al., 2004). An
acyltransferase from banana (Musa sapientum) has
similarly wide substrate specificity (Beekwilder et al.,
2004), and a rose (Rose hybrida) flower BAHD enzyme
can acetylate both geraniol and citronellol (Shalit et al.,
2003). In such cases, the type of volatile formed in
a given tissue depends more on the internal concen-
trations of the substrates than on the Km and Kcat
values of the enzyme for these substrates (Beekwilder
et al., 2004; Boatright et al., 2004).
The Production of the C6-C1 Benzenoids from
C6-C3 Phenylpropanoids
The shortening by two carbons of the three-carbon
chain attached to the phenyl ring of phenylpropanoids
leads to the formation of benzenoid compounds. The
mechanism by which this is achieved is not fully
understood. In vivo stable isotope labeling and com-
puter-assisted metabolic flux analysis, described in
this issue, revealed that both the CoA-dependent-
b-oxidative and CoA-independent-non-b-oxidative
pathways are involved in the formation of benzenoid
compounds in petunia (Boatright et al., 2004). How-
ever, a recent discovery also indicates that in the case
of 2-hydroxybenzoic acid (i.e. salicylic acid), a third
pathway, via the isochorismate pathway, may also
operate in plants, as it does in bacteria (Wildermuth
et al., 2001).
REGULATION OF EMISSION OF
VOLATILE COMPOUNDS
Emission of a particular volatile compound into the
atmosphere depends on both the rate of its biosynthe-
sis and the rate of its release. Substantial progress in
the last decade in the isolation and characterization of
genes responsible for the formation of volatile com-
pounds has facilitated the investigation of the regula-
tion of the biosynthesis of plant volatiles. It has been
found that volatiles are synthesized de novo in the
tissues from which they are emitted. Biosynthesis
normally occurs in the epidermal cells of plant tissues
from which they can escape into the atmosphere or
rhizosphere after being synthesized (Dudareva et al.,
1996; Dudareva and Pichersky, 2000; Kolosova et al.,
2001b; Chen et al., 2004) or in the secretory structures
or glandular trichomes as, for instance, was found in
peppermint, Artemisia annua, and sweet basil (Ocimum
1897
Dudareva et al.
basilicum; McCaskill et al., 1992; Gang et al., 2001; Lu
et al., 2002).
Formation of volatile compounds is spatially regu-
lated. Of the plant organs in scented species, flow-
ers produce the most diverse and the highest amount
of volatile compounds, which peak when the flowers
are ready for pollination. Vegetative tissue also re-
leases small quantities of volatile organic compounds,
which could be induced by mechanical damage or by
herbivore- or pathogen infection (Loughrin et al., 1994;
Pare and Tumlinson, 1997; Arimura et al., 2004a). In
herbs, such as peppermint, significant amounts of vola-
tile compounds accumulate in the leaf glandular
trichomes and the emitted volatiles represent only a
small fraction of the total pool produced (Gershenzon
et al., 2000).
Production and emission of volatile compounds is
also a developmentally regulated process. Volatile emis-
sion in flowers and accumulation in leaves and fruits
follow similar developmental patterns, increasing dur-
ing the early stages of organ development (when leaves
are young and not fully expended, fruit is not yet
mature, or when flowers are ready for pollination) and
then either remaining relatively constant or decreasing
over the organs lifespan (Bouwmeester et al., 1998;
Dudareva and Pichersky, 2000; Gershenzon et al., 2000).
The concurrent temporal changes in activities of en-
zymes responsible for the final steps of volatile forma-
tion, enzyme protein levels, and the expression of
corresponding structural genes suggest that the devel-
opmental biosynthesis of volatiles is regulated largely at
the level of gene expression (Dudareva et al., 1996;
McConkey et al., 2000). It is still unclear to what extent
transcriptional, posttranscriptional, translational, post-
translational, and other events contribute to this process.
In general, more than one biochemical pathway is
responsible for a blend of volatile compounds released
from different plant tissues. A comparative analysis of
the regulation of benzenoid and monoterpene emis-
sion in snapdragon flowers revealed that the orches-
trated emission of phenylpropanoid and isoprenoid
compounds is regulated upstream of individual met-
abolic pathways and includes the coordinated expres-
sion of genes that encode enzymes involved in the
final steps of scent biosynthesis (Dudareva et al., 2000,
2003). However, transcription factors that regulate
multiple biosynthetic pathways leading to the forma-
tion of odor bouquet have not yet been discovered.
The level of the enzyme responsible for the final step
of the biosynthesis of a particular volatile is not the only
limiting factor. The target for the regulation of devel-
opmental production of volatile compounds also in-
cludes the level of supplied substrate in the cell
(Dudareva et al., 2000). In the case of enzymes that are
able to use several similar substrates, such as SAMTand
acyltransferases, the level of supplied precursor con-
trols the type of produced product (Negre et al., 2003;
Boatright et al., 2004; Pott et al., 2004). The role of
substrate in the regulation of the biosynthesis of volatile
compounds was also recently confirmed by metabolic
1898
engineering. When the linalool synthase gene was
introduced under the control of the cauliflower mosaic
virus 35S constitutive promoter into petunia W115, the
differences between organs in the amount of the syn-
thesized linalool or its glycoside depended more on
the availability of the substrate GPP in the tissue
than on the expression of the linalool synthase gene
(Lucker et al., 2001). In peppermint the up-regulation of
1-deoxy-D-xylulose-5-phosphate reductase, which cat-
alyzes the conversion of 1-deoxy-D-xylulose-5-phos-
phate to methylerythritol phosphate, increased the flux
to GPP and led to about a 50% increase in the essential
oil production (Mahmoud and Croteau, 2001). Addi-
tional regulation of GPP formation can occur at the level
of GPP synthase, as was shown in snapdragon where
the small subunit can play a key role in GPP biosyn-
thesis (Tholl et al., 2004). Feedback regulation of GPP
synthase by product and substrate inhibition could also
contribute to the regulatory control of the flux to GPP
and subsequently to monoterpene production (Tholl
et al., 2004). When enzymes competed for the same
substrate as in the case of three monoterpene synthases
(g-terpinene cyclase, [1]-limonene cyclase, and [2]-b-
pinene cyclase) introduced into tobacco plants, the
magnitude of monoterpene emission in leaves was
close to that predicted based on the Km values of the
enzymes for GPP, while the emission levels in flowers
were comparable suggesting that the GPP pool did not
limit monoterpene production (Lucker et al., 2004).
These results show that while the investigation of the
regulation of the final steps of volatile biosynthesis is an
important starting point, a detailed understanding of
the regulation of the flux through the entire biochemical
pathway is essential for the complete understanding of
production and emission of secondary volatile com-
pounds.
Emission of volatile compounds from flowers and
leaves of some plant species, as well as herbivore-
induced volatiles, varies remarkably throughout the
photoperiod. The release of floral volatiles in these
species displays a rhythmic pattern with maximum
emission during the day or night, which generally
coincides with the foraging activities of potential
pollinators, and is controlled by a circadian clock or
regulated by light (Jakobsen and Olsen, 1994; Helsper
et al., 1998; Kolosova et al., 2001a). Isoprene, as well as
volatiles emitted from undamaged and herbivore-
attacked leaves, exhibit a distinct diurnal emission
pattern (Loughrin et al., 1994; Loreto et al., 1996; De
Moraes et al., 2001; Lerdau and Gray, 2003; Martin
et al., 2003; Arimura et al., 2004a) with some leaf
volatiles emission also controlled by a circadian clock,
for example (2)-b-pinene in Artemisia annua (Lu et al.,
2002). In flowers the rhythmic production and emis-
sion of some volatile compounds, such as the volatile
ester methylbenzoate, is regulated primarily by the
level of substrate availability (BA), which in turn could
be regulated at the level of expression of genes en-
coding the key enzymes of its biosynthesis (Kolosova
et al., 2001a). While regulation of isoprene emission
Plant Physiol. Vol. 135, 2004

is well understood (Wolfertz et al., 2003), little is
known to date about the molecular mechanisms
responsible for diurnal emission of inducible vegeta-
tive volatiles. Often, there is a delay of several hours
between the beginning of herbivore damage and the
release of induced volatile compounds, which could
be explained, in part, by the up-regulation of the ex-
pression of genes responsible for their biosynthesis
(Arimura et al., 2004a). When volatile compounds
are released immediately after herbivore damage,
they arise from stored pools (Pare and Tumlinson,
1997).
Environmental factors such as light, temperature,
and moisture status can greatly influence the emission
of volatiles and the yield and composition of essential
oils (Staudt and Bertin, 1998; Gershenzon et al., 2000).
In addition, pollination induces the decrease in emis-
sion of floral volatiles which begins after pollen tubes
reach the ovary (Negre et al., 2003).
To date very little is known about the release of
synthesized volatile compounds from plant tissues. In
general, the rate of release is a function of the physical
properties of the compound itself (its volatility) and the
properties of cellular and intracellular membranes (in
case of monoterpenes which are synthesized in the
chloroplast) through which the compound has to
diffuse. Comparative analysis of volatile compounds
emitted and present within the plant tissue revealed
that the emission of volatiles is not merely a function of
their differential volatility but could also involve a cy-
tologically organized excretory process (Altenburger
and Matile, 1990; Gershenzon et al., 2000). The mem-
branes of the storage compartment (where it exists) or
epidermal cell wall might be selectively more perme-
able to some volatile compounds or the emitted sub-
stances may be associated with an entirely different
secretory compartment than the stored volatiles. Vir-
tually nothing is presently known about metabolite
trafficking between various subcellular compart-
ments, the mechanism of the release process, and how
these processes contribute to the regulation of volatile
emission.
CONCLUSIONS
Plants produce a plethora of volatile compounds for
both general and specialized functions. Recent advances
in instrumentation, coupled with our present ability to
isolate and characterize genes and the enzymes they
encode from many diverse plant species, have greatly
enhanced our understanding of how plants synthezise
such compounds and regulate their production. The
next level of understanding should surely come from
studying the significance of these volatiles for plant
physiology and their impact on the ecological interac-
tions of plants with their environment.
Received July 15, 2004; returned for revision July 19, 2004; accepted July 19,
2004.
Plant Physiol. Vol. 135, 2004
Update on Biochemistry of Plant Volatiles
LITERATURE CITED
Aharoni A, Keizer LCP, Bouwmeester HJ, Sun ZK, Alvarez-Huerta M,
Verhoeven HA, Blaas J, van Houwelingen A, De Vos RCH, van der
Voet H, et al (2000) Identification of the SAAT gene involved in
strawberry flavor biogenesis by use of DNA microarrays. Plant Cell
12: 647661
Altenburger R, Matile P (1990) Further observations on rhythmic emission
of fragrance in flowers. Planta 180: 194197
Andersen RA, Hamilton-Kemp TR, Hildebrand DF, McCracken CT Jr,
Collins RW, Fleming PD (1994) Structure-antifungal activity relation-
ships among volatile C6 and C9 aliphatic aldehydes, ketones and
alcohols. J Agric Food Chem 42: 15631568
Arimura G, Huber DPW, Bohlmann J (2004a) Forest tent caterpillars
(Malacosoma disstria) induce local and systemic diurnal emissions of
terpenoid volatiles in hybrid poplar (Populus trichocarpa x deltoides):
cDNA cloning, functional characterization, and patterns of gene ex-
pression of (2)-germacrene D synthase, PtdTPS1. Plant J 37: 603616
Arimura G, Ozawa R, Kugimiya S, Takabayashi J, Bohlmann J (2004b)
Herbivore-induced defense response in a model legume: Two-spotted
spider mites, Tetranychus urticae, induce emission of (E)-b-ocimene and
transcript accumulation of (E)-b-ocimene synthase in Lotus japonicus.
Plant Physiol 135: 19761983
Arimura G, Ozawa R, Shomoda T, Nishioka T, Boland W, Takabayashi J
(2000) Herbivory-induced volatiles elicit defense genes in lima bean
leaves. Nature 406: 512515
Bate NJ, Riley JCM, Thompson JE, Rothstein SJ (1998) Quantitative and
qualitative differences in C-6-volatile production from the lipoxygenase
pathway in an alcohol dehydrogenase mutant of Arabidopsis thaliana.
Physiol Plant 104: 97104
Beekwilder J, Alvarez-Huerta M, Neef E, Verstappen FWA, Bouwmeester
HJ, Aharoni A (2004) Substrate usage by recombinant alcohol acyl-
transferases from various fruit species. Plant Physiol 135: 18651878
Boatright J, Negre F, Chen X, Kish CM, Wood B, Peel G, Orlova I, Gang D,
Rhodes D, Dudareva N (2004) Understanding in vivo benzenoid
metabolism in petunia petal tissue. Plant Physiol 135: 19932011
Boland W, Gabler A (1989) Biosynthesis of homoterpenes in higher plants.
Helv Chim Acta 72: 247253
Bonn B, Moortgat GK (2003) Sesquiterpene ozonolysis: origin of atmo-
spheric new particle formation from biogenic hydrocarbons. Geophys
Res Lett 30: 1585
Bouvier F, Suire C, dHarlingue A, Backhaus RA, Camara B (2000)
Molecular cloning of geranyl diphosphate synthase and compartmen-
tation of monoterpene synthesis in plant cells. Plant J 24: 241252
Bouwmeester HJ, Gershenzon J, Konings MCJM, Croteau R (1998) Bio-
synthesis of the monoterpenes limonene and carvone in the fruit of
caraway. I. Demonstration of enzyme activities and their changes with
development. Plant Physiol 117: 901912
Bouwmeester HJ, Konings MCJM, Gershenzon J, Karp F, Croteau R
(1999) Cytochrome P-450 dependent (1)-limonene-6-hydroxylation in
fruits of caraway (Carum carvi). Phytochemistry 50: 243248
Burke CC, Croteau R (2002) Geranyl diphosphate synthase from Abies
grandis: cDNA isolation, functional expression, and characterization.
Arch Biochem Biophys 405: 130136
Burke CC, Wildung MR, Croteau R (1999) Geranyl diphosphate synthase:
cloning, expression, and characterization of this prenyltransferase as
a heterodimer. Proc Natl Acad Sci USA 96: 1306213067
Cane DE (1999) Sesquiterpene biosynthesis: cyclization mechanisms. In DE
Cane, ed, Comprehensive Natural Products Chemistry. Isoprenoids
Including Carotenoids and Steroids, Vol 2. Pergamon Press, Oxford, pp
155200
Chen F, DAuria JC, Tholl D, Ross JR, Gershenzon J, Noel JP, Pichersky E
(2003) An Arabidopsis thaliana gene for methylsalicylate biosynthesis,
identified by a biochemical genomics approach, has a role in defense.
Plant J 36: 577588
Chen F, Ro D-K, Petri J, Gershenzon J, Bohlmann J, Pichersky E, Tholl D
(2004) Characterization of root-specific Arabidopsis terpene synthase
responsible for the formation of the volatile monoterpene 1,8-cineole.
Plant Physiol 135: 19561966
DAuria JC, Chen F, Pichersky E (2002) Characterization of an acyltransfer-
ase capable of synthesizing benzylbenzoate and other volatile esters in
flowers and damaged leaves of Clarkia breweri. Plant Physiol 130:
466476
1899
Dudareva et al.
Degen T, Dillmann C, Marion-Poll F, Turlings TCJ (2004) High genetic
variability of herbivore-induced volatile emission within a broad range
of maize inbred lines. Plant Physiol 135: 19281938
Degenhardt J, Gershenzon J (2000) Demonstration and characterization of
(E)-nerolidol synthase from maize: a herbivore-inducible terpene syn-
thase participating in (3E)-4,8-dimethyl-1,3,7-nonatriene biosynthesis.
Planta 210: 815822
Delfine S, Csiky O, Seufert G, Loreto F (2000) Fumigation with exogenous
monoterpenes of a non-isoprenoid-emitting oak (Quercus suber): mono-
terpene acquisition, translocation, and effect on the photosynthetic
properties at high temperatures. New Phytol 146: 2736
De Moraes CM, Mescheer MC, Tumlinson JH (2001) Caterpillar-induced
nocturnal plant volatiles repel nonspecific females. Nature 410: 577580
Dicke M, Bruin J (2001) Chemical information transfer between plans: back
to the future. Biochem Syst Ecol 29: 981994
Dicke M, Van Loon JJA (2000) Multitrophic effects of herbivore-induced
plant volatiles in an evolutionary context. Entomol Exp Appl 97: 237249
Dudareva N, Cseke L, Blanc VM, Pichersky E (1996) Evolution of floral
scent in Clarkia: novel patterns of S-linalool synthase gene expression in
the C. breweri flower. Plant Cell 8: 11371148
Dudareva N, DAuria JC, Nam KH, Raguso RA, Pichersky E (1998) Acetyl-
CoA:benzyl alcohol acetyltransferase: an enzyme involved in floral
scent production in Clarkia breweri. Plant J 14: 297304
Dudareva N, Martin D, Kish CM, Kolosova N, Gorenstein N, Faldt J,
Miller B, Bohlmann J (2003) (E)-b-Ocimene and myrcene synthase
genes of floral scent biosynthesis in snapdragon: function and expres-
sion of three terpene synthase genes of a new TPS-subfamily. Plant Cell
15: 12271241
Dudareva N, Murfitt LM, Mann CJ, Gorenstein N, Kolosova N, Kish CM,
Bonham C, Wood K (2000) Developmental regulation of methylben-
zoate biosynthesis and emission in snapdragon flowers. Plant Cell 12:
949961
Dudareva N, Pichersky E (2000) Biochemical and molecular genetic
aspects of floral scents. Plant Physiol 122: 627633
Engelberth J, Alborn HT, Schmelz EA, Tumlinson JH (2004) Airborne
signals prime plants against insect herbivore attack. Proc Natl Acad Sci
USA 101: 17811785
Frank MR, Deyneka JM, Schuler MA (1996) Cloning of wound-induced
cytochrome P450 monooxygenases expressed in pea. Plant Physiol 110:
10351046
Friedman M, Henika PR, Mandrell RE (2002) Bactericidal activities of
plant essential oils and some of their isolated constituents against
Campylobacter jejuni, Escherichia coli, Listeria monocytogenes, and Salmo-
nella enterica. J Food Prot 65: 15451560
Fukami H, Asakura T, Hirano H, Abe K, Shimomura K, Yamakawa T
(2002) Salicylic acid carboxyl methyltransferase induced in hairy root
cultures of Atropa belladonna after treatment with exogeneously added
salicylic acid. Plant Cell Physiol 43: 10541058
Gang DR, Beuerle T, Ullmann P, Werck-Reichhart D, Pichersky E (2002a)
Differential production of meta hydroxylated phenylpropanoids in
sweet basil (Ocimum basilicum L.) peltate glandular trichomes and
leaves is controlled by the activities of specific acyltransferases and
hydroxylases. Plant Physiol 130: 15361544
Gang DR, Lavid N, Zubieta C, Chen F, Beuerle T, Lewinsohn E, Noel
JP, Pichersky E (2002b) Characterization of phenylpropene O-meth-
yltransferases from sweet basil: facile change of substrate specificity
and convergent evolution within a plant OMT family. Plant Cell 14:
505519
Gang DR, Wang JH, Dudareva N, Nam KH, Simon JE, Lewinsohn E,
Pichersky E (2001) An investigation of the storage and biosynthesis of
phenylpropenes in sweet basil. Plant Physiol 125: 539555
Gershenzon J, McConkey ME, Croteau RB (2000) Regulation of mono-
terpene accumulation in leaves of peppermint. Plant Physiol 122: 205213
Hallahan DL, West JM, Wallsgrove RM, Smiley DWM, Dawson GW,
Pickett JA, Hamilton JGC (1995) Purification and characterization of an
acyclic monoterpene primary alcohol:NADP1 oxidoreductase from
catmint (Nepeta racemosa). Arch Biochem Biophys 318: 105112
Hammer KA, Carson CF, Riley TV (2003) Antifungal activity of the
components of Melaleuca alternifolia (tea tree) oil. J Appl Microbiol 95:
853860
Helsper JPFG, Davies JA, Bouwmeester HJ, Krol AF, van Kampen MH
(1998) Circadian rhythmicity in emission of volatile compounds by
flowers of Rosa hybrida L. cv. Honesty. Planta 207: 8895
1900
Hoffmann T, Odum JR, Bowman F, Collins D, Klockow D, Flagan RC,
Seinfeld JH (1997) Formation of organic aerosols from the oxidation of
biogenic hydrocarbons. J Atmos Chem 26: 189212
Howe GA, Schilmiller AL (2002) Oxylipin metabolism in response to
stress. Curr Opin Plant Biol 5: 230236
Jakobsen HB, Olsen CE (1994) Influence of climatic factors on emission of
flower volatiles in-situ. Planta 192: 365371
Kessler A, Baldwin IT (2001) Defensive function of herbivore-induced
plant volatile emissions in nature. Science 291: 21412144
Knudsen JT, Tollsten L (1993) Trends in floral scent chemistry in pollina-
tion syndromesfloral scent composition in moth-pollinated taxa. Bot J
Linn Soc 113: 263284
Knudsen JT, Tollsten L, Bergstrom G (1993) Floral scents a checklist of
volatile compounds isolated by head-space techniques. Phytochemistry
33: 253280
Kolosova N, Gorenstein N, Kish CM, Dudareva N (2001a) Regulation of
circadian methylbenzoate emission in diurnally and nocturnally emit-
ting plants. Plant Cell 13: 23332347
Kolosova N, Sherman D, Karlson D, Dudareva N (2001b) Cellular and
subcellular localization of S-adenosyl-L-methionine:benzoic acid car-
boxyl methyltransferase, the enzyme responsible for biosynthesis of the
volatile ester methylbenzoate in snapdragon flowers. Plant Physiol 126:
956964
Koyama T, Ogura K (1999) Isopentenyl diphosphate isomerase and
prenyltransferases. In DE Cane, ed, Comprehensive Natural Product
Chemistry. Isoprenoids Including Carotenoids and Steroids, Vol 2.
Pergamon Press, Oxford, pp 6996
Laule O, Furholz A, Chang H-S, Zhu T, Wang X, Heifetz PB, Gruissem W,
Lange M (2003) Crosstalk between cytosolic and plastidial pathways of
isoprenoid biosynthesis in Arabidopsis thaliana. Proc Natl Acad Sci USA
100: 68666871
Lavid N, Wang J, Shalit M, Gutterman I, Bar E, Beuerle T, Weiss D,
Vainstein A, Pichersky E, Lewinsohn E (2002) O-methyltransferases
involved in the biosynthesis of volatile phenolic derivatives in rose
petals. Plant Physiol 129: 18991907
Lerdau M, Gray D (2003) Ecology and evolution of light-dependent and
light-independent phytogenic volatile organic carbon. New Phytol 157:
199211
Lewinsohn E, Ziv-Raz I, Dudai N, Tadmor Y, Lastochkin E, Larkov O,
Chaimovitsh D, Ravid U, Putievsky E, Pichersky E, et al (2000)
Biosynthesis of estragole and methyl-eugenol in sweet basil (Ocimum
basilicum L). Developmental and chemotypic association of allylphenyl
O-methyltransferase activities. Plant Sci 160: 2735
Liang P-H, Ko T-P, Wang AH-J (2002) Structure, mechanism and function
of prenyltransferases. Eur J Biochem 269: 33393354
Loreto F, Ciccioli P, Brancaleoni E, Cecinato A, Frattoni M, Sharkey TD
(1996) Different sources of reduced carbon contribute to form three
classes of terpenoid emitted by Quercus ilex L leaves. Proc Natl Acad Sci
USA 93: 99669969
Loreto F, Pinelli P, Brancaleoni E, Ciccioli P (2004a) 13C labeling reveals
chloroplastic and extra-chloroplastic pools of dimethylallyl pyrophos-
phate and their contribution to isoprene formation. Plant Physiol 135:
19031907
Loreto F, Pinelli P, Manes F, Kollist H (2004b) Impact of ozone on
monoterpene emissions and evidence for an isoprene-like antioxidant
action of monoterpenes emitted by Quercus ilex leaves. Tree Physiol 24:
361367
Loreto F, Velikova V (2001) Isoprene produced by leaves protects the
photosynthetic apparatus against ozone damage, quenches ozone prod-
ucts, and reduces lipid peroxidation of cellular membranes. Plant
Physiol 127: 17811787
Loughrin JH, Manukian A, Heath RR, Turlings TCJ, Tumlinson JH
(1994) Diurnal cycle of emission of induced volatile terpenoids
herbivore-injured cotton plants. Proc Natl Acad Sci USA 91:
1183611840
Lu S, Xu R, Jia JW, Pang JH, Matsuda SPT, Chen XY (2002) Cloning and
functional characterization of a beta-pinene synthase from Artemisia
annua that shows a circadian pattern of expression. Plant Physiol 130:
13351348
Lucker J, Bouwmeester HJ, Schwab W, Blaas J, Van der Plas LHW,
Verhoeven HA (2001) Expression of Clarkia S-linalool synthase in
transgenic petunia plants results in the accumulation of S-linalyl-beta-
D-glucopyranosid. Plant J 27: 315324
Plant Physiol. Vol. 135, 2004

Lucker J, Schwab W, Van Hautum B, Blaas J, Van der Plas LHW,
Bouwmeester HJ, Verhoeven HA (2004) Increased and altered
fragrance of tobacco plants after metabolic engineering using three
monoterpene synthases from lemon. Plant Physiol 134: 510519
Lupien S, Karp F, Wildung M, Croteau R (1999) Regiospecific cytochrome
P450 limonene hydroxylases from mint (Mentha) species: cDNA iso-
lation, characterization, and functional expression of (-)-4S-limonene-3-
hydroxylase and (-)-4S-limonene-6-hydroxylase. Arch Biochem Biophys
368: 181192
Mahmoud SS, Croteau RB (2001) Metabolic engineering of essential oil
yield and composition in mint by altering expression of deoxyxylulose
phosphate reductoisomerase and menthofuran synthase. Proc Natl
Acad Sci USA 98: 89158920
Martin D, Faldt J, Bohlmann J (2004) Functional characterization of nine
Norway spruce TPS genes and evolution of gymnosperm terpene
synthases of the TPS-d subfamily. Plant Physiol 135: 19081927
Martin DM, Gershenzon J, Bohlmann J (2003) Induction of volatile
terpene biosynthesis and diurnal emission by methyl jasmonate in
foliage of Norway spruce. Plant Physiol 132: 15861599
McCaskill D, Gershenzon J, Croteau R (1992) Morphology and mono-
terpene biosynthetic capabilities of secretory-cell clusters isolated
from glandular trichomes of peppermint (Mentha-piperita L.). Planta
187: 445454
McConkey ME, Gershenzon J, Croteau RB (2000) Developmental regula-
tion of monoterpene biosynthesis in the glandular trichomes of pep-
permint. Plant Physiol 122: 215223
Mercke P, Kappers IF, Verstappen FWA, Vorst O, Dicke M, Bouwmeester
HJ (2004) Combined transcript and metabolite analysis reveals genes
involved in spider mite induced volatile formation in cucumber plants.
Plant Physiol 135: 20122024
Murfitt LM, Kolosova N, Mann CJ, Dudareva N (2000) Purification and
characterization of S-adenosyl-L-methionine:benzoic acid carboxyl
methyltransferase, the enzyme responsible for biosynthesis of the
volatile ester methylbenzoate in flowers of Antirrhinum majus. Arch
Biochem Biophys 382: 145151
Negre F, Kish CM, Boatright J, Underwood B, Shibuya K, Wagner C, Clark
DG, Dudareva N (2003) Regulation of methylbenzoate emission after
pollination in snapdragon and petunia flowers. Plant Cell 15: 29923006
Negre F, Kolosova N, Knoll J, Kish CM, Dudareva N (2002) Novel
S-adenosyl-L-methionine:salicylic acid carboxyl methyltransferase, an
enzyme responsible for biosynthesis of methylsalicylate and methyl-
benzoate, is not involved in floral scent production in snapdragon
flowers. Arch Biochem Biophys 406: 261270
Noel JP, Dixon RA, Pichersky E, Zubieta C, Ferrer JL (2003) Structural,
functional, and evolutionary basis for methylation of plant small
molecules. In JT Romeo, ed, Recent Advances in Phytochemistry, Vol
37. Elsevier Science, Oxford, pp 3758
Pare PW, Tumlinson JH (1997) De novo biosynthesis of volatiles induced by
insect herbivory in cotton plants. Plant Physiol 114: 11611167
Pare PW, Tumlinson JH (1999) Plant volatiles as a defense against insect
herbivores. Plant Physiol 121: 325331
Pichersky E, Gershenzon J (2002) The formation and function of plant
volatiles: perfumes for pollinator attraction and defense. Curr Opin
Plant Biol 5: 237243
Pott MB, Hippauf F, Saschenbrecker S, Chen F, Ross J, Kiefer I,
Slusarenko A, Noel JP, Pichersky E, Effmert U, et al (2004) Biochemical
and structural characterization of benzenoid carboxyl methyltrans-
ferases involved in floral scent production in Stephanotis floribunda
and Nicotiana suaveolens. Plant Physiol 135: 19461955
Pott MB, Pichersky E, Piechulla B (2002) Evening-specific oscillation of
scent emission, SAMT enzyme activity, and mRNA in flowers of
Stephanotis floribunda. J Plant Physiol 159: 925934
Prestage S, Linforth RST, Taylor AJ, Lee E, Speirs J, Schuch W (1999)
Volatile production in tomato fruit with modified alcohol dehydroge-
nase activity. J Sci Food Agric 79: 131136
Raguso RA, Levin RA, Foose SE, Holmberg MW, McDade LA (2003)
Fragrance chemistry, nocturnal rhythms and pollination syndromes
in Nicotiana. Phytochemistry 63: 265284
Ralston L, Kwon ST, Schoenbeck M, Ralston J, Schenk DJ, Coates RM,
Chappell J (2001) Cloning, heterologous expression, and functional
characterization of 5-epi-aristolochene-1,3-dihydroxylase from tobacco
(Nicotiana tabacum). Arch Biochem Biophys 393: 222235
Plant Physiol. Vol. 135, 2004
Update on Biochemistry of Plant Volatiles
Reinhard J, Srivivasan MV, Zhang S (2004) Scent-triggered navigation in
honeybees. Nature 427: 411
Rodriguez-Concepcion M, Boronat A (2002) Elucidation of the methyl-
erythritol phosphate pathway for isoprenoid biosynthesis in bacteria
and plastids. A metabolic milestone achieved through genomics. Plant
Physiol 130: 10791089
Ross JR, Nam KH, DAuria JC, Pichersky E (1999) S-adenosyl-L-
methionine:salicylic acid carboxyl methyltransferase, an enzyme in-
volved in floral scent production and plant defense, represents a new
class of plant methyltransferases. Arch Biochem Biophys 367: 916
Scalliet G, Journot N, Jullien F, Baudino S, Magnard JL, Channeliere S,
Vergne P, Dumas C, Bendahmane M, Cock JM, et al (2002) Biosynthesis
of the major scent components 3,5-dimethoxytoluene and 1,3,5-trimeth-
oxybenzene by novel rose O-methyltransferases. FEBS Lett 523:
113118
Schoch G, Goepfert S, Morant M, Hehn A, Meyer D, Ullmann P, Werck-
Reichhart D (2001) CYP98A3 from Arabidopsis thaliana is a 3#-hydrox-
ylase of phenolic esters, a missing link in the phenylpropanoid pathway.
J Biol Chem 276: 3656636574
Schuhr CA, Radykewicz T, Sagner S, Latzel C, Zenk MH, Arigoni D,
Bacher A, Rohdich F, Eisenreich W (2003) Quantitative assessment of
crosstalk between the two isoprenoid biosynthesis pathways in plants
by NMR spectroscopy. Phytochem Rev 2: 316
Schuler MA (1996) Plant cytochrome P450 monooxygenases. Crit Rev Plant
Sci 15: 235284
Seo HS, Song JT, Cheong JJ, Lee YH, Lee YW, Hwang I, Lee JS, Choi YD
(2001) Jasmonic acid carboxyl methyltransferase: a key enzyme for
jasmonate-regulated plant responses. Proc Natl Acad Sci USA 98:
47884793
Shalit M, Guterman I, Volpin H, Bar E, Tamari T, Menda N, Adam Z,
Zamir D, Vainstein A, Weiss D, et al (2003) Volatile ester formation in
roses: identification of an acetyl-CoA:geraniol acetyltransferase in de-
veloping rose petals. Plant Physiol 131: 18681876
Sharkey TD, Chen XY, Yeh S (2001a) Isoprene increases thermotolerance of
fosmidomycin-fed leaves. Plant Physiol 125: 20012006
Sharkey TD, Yeh SS (2001b) Isoprene emission from plants. Annu Rev
Plant Physiol 52: 407436
Starks CM, Back K, Chappell J, Noel JP (1997) Structural basis for cyclic
terpene biosynthesis by tobacco 5-epi-aristolochene synthase. Science
277: 18151820
Staudt M, Bertin N (1998) Light and temperature dependence of the
emission of cyclic and acyclic monoterpenes from holm oak (Quercus ilex
L.) leaves. Plant Cell Environ 21: 385395
Steeghs M, Bais HP, de Gouw J, Goldan P, Kuster W, Northway M, Fall
R, Vivanco JM (2004) Proton-transfer-reaction mass spectrometry
(PTR-MS) as a new tool for real time analysis of root-secreted volatile
organic compounds (VOCs) in Arabidopsis thaliana. Plant Physiol 135:
4758
St-Pierre B, De Luca V (2000) Evolution of acyltransferase genes: origin
and diversification of the BAHD superfamily of acyltransferases in-
volved in secondary metabolism. In JT Romeo, R Ibrahim, L Varin, V De
Luca, eds, Recent Advances in Phytochemistry Evolution of Metabolic
Pathways, Vol 34. Elsevier Science, Oxford, pp 285315
St-Pierre B, Laflamme P, Alarco AM, De Luca V (1998) The terminal
O-acetyltransferase involved in vindoline biosynthesis defines a new
class of proteins responsible for coenzyme A-dependent acyl transfer.
Plant J 14: 703713
Tholl D, Kish CM, Orlova I, Sherman D, Gershenzon J, Pichersky E,
Dudareva N (2004) Formation of monoterpenes in Antirrhinum majus
and Clarkia breweri flowers involves heterodimeric geranyl diphosphate
synthases. Plant Cell 16: 977992
Uefuji H, Ogita S, Yamaguchi Y, Koizumi N, Sano H (2003) Molecular
cloning and functional characterization of three distinct N-methyltrans-
ferases involved in the caffeine biosynthetic pathway in coffee plants.
Plant Physiol 132: 372380
Vancanneyt G, Sanz C, Farmaki T, Paneque M, Ortego F, Castanera P,
Sanchez-Serrano JJ (2001) Hydroperoxide lyase depletion in transgenic
potato plants leads to an increase in aphid performance. Proc Natl Acad
Sci USA 98: 81398144
Van Poecke RMP, Posthumus MA, Dicke M (2001) Herbivore-induced
volatile production by Arabidopsis thaliana leads to attraction of the
parasitoid Cotesia rubecula: chemical, behavioral, and gene-expression
analysis. J Chem Ecol 27: 19111928
1901
Dudareva et al.
Vuorinen T, Nerg A-M, Holopainen JK (2004a) Ozone exposure triggers
the emission of herbivore-induced plant volatiles, but does not disturb
tritrophic signalling. Environ Pollut 131: 305311
Vuorinen T, Nerg A-M, Ibrahim MA, Reddy GVP, Holopainen JK (2004b)
Emission of Plutella xylostella-induced compounds from cabbages
grown at elevated CO2 and orientation behaviour of the natural
enemies. Plant Physiol 135: 19841992
Walker K, Croteau R (2000) Molecular cloning of a 10-deacetylbaccatin III-
10-O-acetyl transferase cDNA from Taxus and functional expression in
Escherichia coli. Proc Natl Acad Sci USA 97: 583587
Wang J, Dudareva N, Bhakta S, Raguso RA, Pichersky E (1997) Floral
scent production in Clarkia breweri (Onagraceae). II. Localization
and developmental modulation of the enzyme SAM:(Iso)Eugenol
O-methyltransferase and phenylpropanoid emission. Plant Physiol
114: 213221
Wein M, Lavid N, Lunkenbein S, Lewinsohn E, Schwab W, Kaldenhoff R
(2002) Isolation, cloning and expression of a multifunctional O-methyl-
transferase capable of forming 2,5-dimethyl-4-methoxy-3(2H)-fura-
none, one of the key aroma compounds in strawberry fruits. Plant J
31: 755765
1902
Whittington DA, Wise ML, Urbansky M, Coates RM, Croteau RB,
Christianson DW (2002) Bornyl diphosphate synthase: structure and
strategy for carbocation manipulation by a terpenoid cyclases. Proc Natl
Acad Sci USA 99: 1537515380
Wildermuth MC, Dewdney J, Wu G, Ausubel FM (2001) Isochorismate
synthase is required to synthesize salicylic acid for plant defense.
Nature 414: 562565
Wise ML, Croteau R (1999) Monoterpene biosynthesis. In DE Cane, ed,
Comprehensive Natural Products Chemistry. Isoprenoids Including
Carotenoids and Steroids, Vol 2. Pergamon Press, Oxford, pp 97153
Wolfertz M, Sharkey TD, Boland W, Kuhnemann F (2004) Rapid regula-
tion of the methylerythritol 4-phosphate pathway during isoprene
synthesis. Plant Physiol 135: 19391945
Wolfertz M, Sharkey TD, Boland W, Kuhnemann F, Yeh S, Weise SE
(2003) Biochemical regulation of isoprene emission. Plant Cell Environ
26: 13571364
Wu SQ, Watanabe N, Mita S, Dohra H, Ueda Y, Shibuya M, Ebizuka Y
(2004) The key role of phloroglucinol O-methyltransferase in the bio-
synthesis of Rosa chinensis volatile 1,3,5-trimethoxybenzene. Plant
Physiol 135: 95102
Plant Physiol. Vol. 135, 2004
Molecular Plant Volume 5 Number 5 Pages 964967 September 2012
New Insights into Plant Isoprenoid Metabolism
PabloPulido, CatalinaPerello and ManuelRodriguez-Concepcion1
Department of Molecular Genetics, Centre for Research in Agricultural Genomics (CRAG) CSIC-IRTA-UAB-UB, Campus UAB Bellaterra, 08193 Barcelona,
Spain
Isoprenoids are a hugely diverse family of compounds derived
from the C5 precursors isopentenyl diphosphate (IPP) and
dimethylallyl diphosphate (DMAPP). Although all free-living
organisms synthesize isoprenoids, they are particularly abun-
dant and diverse in plants, with tens of thousands structures
known to date. The highest variety of plant isoprenoids are
specialized (secondary) metabolites that participate in the
interaction of plants with their environment. These include
pigments, volatiles, and defense compounds, some of which
have applications in industry and agriculture. For example,
isoprenoid drugs are used against cancer (taxol) or mal-
aria (artemisin). But plants also synthesize isoprenoids with
essential (primary) functions in respiration (ubiquinone),
photosynthesis (carotenoids, chlorophylls, tocopherols, phyl-
loquinones, plastoquinone), membrane architecture (sterols),
and growth regulation (brassinosteroids, cytokinins, gibber-
ellins, abscisic acid, strigolactones). Despite their economic
importance and biological relevance, our knowledge of the
core pathways for the production of the universal isoprenoid
precursors IPP and DMAPP in plant cells remained incomplete
until the mid-1990s. Impressive progress in the last decade
has resulted in the complete elucidation of several isoprenoid
pathways, the identification of regulatory mechanisms, the
discovery of new functions and properties of specific isopre-
noids, and the successful manipulation of isoprenoid biosyn-
thesis in a number of metabolic engineering approaches. In
this update article, we will discuss some of the most recent
advances in the plant isoprenoid field, focusing on the path-
ways supplying the C5 precursors in plant cells and the novel
insights into regulatory matters.
BIOSYNTHESIS OF THE UNIVERSAL
ISOPRENOID PRECURSORS
Unlike most organisms, plants use two unrelated pathways for
the production of IPP and DMAPP in different cell compart-
ments (Vranova etal., 2012). The methylerythritol 4-phosphate
(MEP) pathway simultaneously produces both IPP and DMAPP
from pyruvate and glyceraldehyde 3-phosphate, whereas the
mevalonic acid (MVA) pathway synthesizes IPP from acetyl-
CoA. IPP is then converted into DMAPP by the activity of IPP
isomerases (Figure1). The subcellular localization of the MEP
and MVA pathway enzymes in Arabidopsis thaliana is sum-
marized in Table1 and schematically represented in Figure1.
RESEARCH HIGHLIGHT
It is now well established that all the MEP pathway enzymes
are encoded by the nuclear genome and imported into plas-
tids. Recent proteomics approaches have detected all the
Arabidopsis MEP pathway enzymes in the stroma and have
Downloaded from http://mplant.oxfordjournals.org/ at North South University on November 5, 2012
provided much-needed evidence on the subplastidial local-
ization of enzymes involved in the production of specific
types of isoprenoids, including chrorophylls, carotenoids, and
prenylquinones such as plastoquinone, phylloquinone, and
tocopherol (Joyard etal., 2009). By contrast, the MVA path-
way enzymes are distributed in different subcellular compart-
ments (Table 1). The main rate-determining enzyme of the
pathway, hydroxymethylglutaryl CoA reductase (HMGR), is
anchored to the endoplasmic reticulum exposing the cata-
lytic domain of protein towards the cytosol, whereas other
enzymes of the pathway have been found in the cytosol
and peroxisomes (Sapir-Mir etal., 2008; Simkin etal., 2011).
Despite the compartmentalization of IPP and DMAPP synthe-
sis in plant cells, multiple studies using inhibitors, mutants,
and labeled precursors in feeding experiments have shown
that an exchange of isoprenoid precursors takes place among
different subcellular locations (Flores-Perez et al., 2010;
Paetzold et al., 2010). However, the genetic block of either
the MVA pathway or the MEP pathway in null mutants or the
complete inhibition of single pathway enzymes in wild-type
plants treated with specific inhibitors results in a developmen-
tal block and a seedling-lethal phenotype, indicating that the
loss of one of the two pathways cannot be compensated by
the remaining pathway. The characterization of Arabidopsis
mutants able to develop in the presence of pathway-specific
isoprenoid inhibitors has led to a better understanding of the
regulatory pathways that impact isoprenoid metabolism. For
example, this approach has shown that light and sugar sig-
naling down-regulate the MVA pathway but up-regulate the
MEP pathway (Flores-Perez etal., 2010; Vranova etal., 2012).
Recent results have also advanced our knowledge of how the
1
To whom correspondence should be addressed. E-mail manuel.rodriguez@
cragenomica.es, tel. +34 935636600, ext. 3222.
The Author 2012. Published by the Molecular Plant Shanghai Editorial
Office in association with Oxford University Press on behalf of CSPB and
IPPE, SIBS, CAS.
doi:10.1093/mp/sss088
Received 25 June 2012; accepted 30 July 2012
 Pulido et al. New Insights into Plant Isoprenoid Metabolism 965

Downloaded from http://mplant.oxfordjournals.org/ at North South University on November 5, 2012

Figure1. Isoprenoid biosynthetic pathways in the plant cell. The MVA pathway enzymes are shown in blue and the MEP pathway enzymes in green
(acronyms correspond to those in Table1). The enzymes suggested to have the highest degree of control over the metabolic flux through the MEP
pathway (DXS) or the MVA pathway (HMGR) are underlined. Recently reported positive and negative regulators of both pathways at the post-
transcriptional level are boxed in green and red, respectively. Dashed arrows mark multiple steps and open arrows represent transport of metabo-
lites between cell compartments. Isoprenoids with regulatory roles are underlined in purple (hormone-like compounds) and orange. CDP-ME,
4-(cytidine 5-diphospho)-2-C-methyl-D-erythritol; CDP-MEP, CDP-ME 2-phosphate; DMAPP, dimethylallyl diphosphate; DXP, 1-deoxy-D-xylulose
5-phosphate; FPP, farnesyl diphosphate; GGPP, geranylgeranyl diphosphate; GPP, geranyl diphosphate; HBMPP, 1-hydroxy-2-methyl-2-butenyl
4-diphosphate; HMG-CoA, 3-hydroxy-3-methylglutaryl CoA; IPP, isopentenyl diphosphate; MEcPP, 2-C-methyl-D-erythritol 2,4-cyclodiphosphate;
MEP, 2-C- methyl-D-erythritol 4-phosphate; MVA, mevalonic acid; MVP, 5-phosphomevalonate; MVPP, 5-diphosphomevalonate.

activity of rate-determining enzymes of both pathways are
controlled by different posttranslational mechanisms involv-
ing the PP2A phosphatase (for HMGR) and, in the case of MEP
pathway enzymes, the stromal ClpPR proteolytic complex
(Flores-Perez etal., 2008, 2010; Leivar etal., 2011; Vranova
et al., 2012). All these mechanisms (Figure 1) are likely to
provide a fine control of isoprenoid biosynthesis, allowing
an appropriate supply of the products needed for essential
processes such as respiration or photosynthesis but also for
regulatory processes.
REGULATORY ROLES OF ISOPRENOIDS
As described above, some isoprenoid metabolites are used
by plants as hormones to regulate multiple aspects of
their physiology. In addition to the well-known isoprenoid
hormones (brassinosteroids, cytokinins, gibberellins, and
abscisic acid), it was recently discovered that strigolactones
(a group of carotenoids-derived metabolites known to
be exuded from roots) and strigolactone-like compounds
such as carlactone can also regulate plant development
(Gomez-Roldan etal., 2008; Umehara etal., 2008; Alder et al.,
2012). Most interestingly, other isoprenoid metabolites have
been found to modulate plant responses in new, unsuspected
ways. Recent work has confirmed that methylerythritol
cyclodiphoshate (MEcPP), an intermediate of the MEP
pathway (Figure 1), acts as a signal released from the
plastids to modulate the expression of stress-related genes
(Xiao et al., 2012). Some isoprenoids can regulate protein
function by direct binding. Proteins can be modified by the
966 Pulido et al. New Insights into Plant Isoprenoid Metabolism

Table1. Subcellular localization of the Arabidopsis enzymes involved in the production of isoprenoid precursors. Cyt, cytosol; ER-Cyt,
anchored to the endoplasmic reticulum; Perox, peroxisomes; Plast, plastids; Mit, mitochondria. Data from Arabidopsis Isoprenoid
Pathway Database (www.atipd.ethz.ch/index.php), MASCP Gator Database (http://gator.masc-proteomics.org/), and the indicated
references: 1 Simkin etal., 2011; 2 Joyard etal., 2009; 3 Sapir-Mir etal., 2008.

Enzyme Accession Acronym Localization

acetoacetyl-CoA thiolase 1 At5g47720 AACT1 Cyt

acetoacetyl-CoA thiolase 2 At5g48230 AACT2 Cyt/Perox
3-hydroxy-3-methylglutaryl CoA synthase At4g11820 HMGS Cyt
3-hydroxy-3-methylglutaryl CoA reductase 1 At1g76490 HMGR1 ER-Cyt
MVAPathway 3-hydroxy-3-methylglutaryl CoA reductase 2 At2g17370 HMGR2 ER-Cyt
mevalonate kinase At5g27450 MVK Cyt1
5-phosphomevalonate kinase At1g31910 PMK Perox1

Downloaded from http://mplant.oxfordjournals.org/ at North South University on November 5, 2012

5-diphosphomevalonate decarboxylase 1 At2g38700 MVD1 Perox1
5-diphosphomevalonate decarboxylase 2 At3g54250 MVD2 Perox1

1-deoxy-D-xylulose 5-phosphate synthase At4g15560 DXS Plast2

1-deoxy-D-xylulose 5-phosphate reductoisomerase At5g62790 DXR Plast2
2-C-methyl-D-erythritol 4-phosphate cytidyltransferase At2g02500 MCT Plast2
MEPPathway 4-(cytidine 5-diphospho)-2-C-methyl-D-erythritol kinase At2g26930 CMK Plast2
2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase At1g63970 MDS Plast2
1-hydroxy-2-methyl-2-butenyl 4-diphosphate synthase At5g60600 HDS Plast2
1-hydroxy-2-methyl-2-butenyl 4-diphosphate reductase At4g34350 HDR Plast2

isopentenyl diphosphate isomerase 1 At5g16440 IDI1 Perox / Plast / (Mit)3

IPP isomerise isopentenyl diphosphate isomerase 2 At3g02780 IDI2 Perox / Mit / (Plast)3

covalent attachment of C15 (farnesyl diphosphate, FPP) or
C20 (geranylgeranyl diphosphate, GGPP) isoprenyl groups
to C-terminal cysteine residues, in a process known as
protein prenylation. It was well established that prenylation
modulates the interaction of proteins with membranes or
other proteins and hence influences their biological role.
Moreover, recent results have shown that the accumulation
of farnesylcisteine residues released after degradation of
farnesylated proteins can also impact plant responses such
as abscisic acid sensitivity (Huizinga et al., 2010). Besides
prenylation, a large-scale analysis of the proteinmetabolite
interactome in yeast reported the binding of different sterols
to many proteins, including regulatory kinases (Li etal., 2012).
Although the interaction between sterols and plant proteins
remains to be fully explored, it is becoming clearer that it
might have important regulatory consequences. For example,
sterols appear to facilitate membrane association and hence
activity of ARGONAUTE 1, an Arabidopsis protein required
for microRNA function (Brodersen etal., 2012). These results
together suggest that isoprenoids may play a more general
role as regulators of plant growth and development than
previously thought.
UNDERSTANDING THE COMPLEXITY
OF ISOPRENOID METABOLISM
The picture arising from the above described data shows
an extremely complex, highly networked system linking
isoprenoid metabolism with a variety of other processes.
Understanding how plants regulate the production of their
isoprenoids as a whole (i.e., as a system) remains a major
challenge during the coming years. The topology of the
plant isoprenoid pathway network and its dynamics at the
gene expression level in response to diverse stimuli have
been recently reviewed (Vranova etal., 2012). However, it is
becoming more and more evident that post-transcriptional
mechanisms are also essential to precisely control isoprenoid
metabolism in plant cells (Figure1) (Flores-Perez et al., 2008,
2010; Huizinga et al., 2010; Paetzold et al., 2010; Leivar et
al., 2011; Brodersen et al., 2012). Arecent proteomics study

focusing on isoprenoid biosynthesis in plastids has shown
a diverse, sometimes puzzling, distribution of biosynthetic
enzymes in a variety of subplastidial compartments, including
envelopes, thylakoids, stroma, and plastoglobuli (Joyard
etal., 2009). The wide array of Arabidopsis proteomic data
and resources available in public databases should also allow
determination of the localization of all the pathways that
produce isoprenoids in the cytosol, endoplasmic reticulum,
peroxisomes, or mitochondria (Vranova et al., 2012). This
information will help in understanding the connections
between the MEP and MVA pathways with the downstream
routes that lead to the variety of isoprenoids produced in
plant cells, including those that have a regulatory role
(Figure 1). But understanding these dynamic connections
is a major challenge that would require additional
quantitative technologies and systems biology approaches.
Promising attempts in this direction have been recently
reported, including the reconstruction of Arabidopsis
genome-wide metabolic network models that take into
account the distribution of enzymes and metabolites in
different tissues and subcellular compartments (Mintz-Oron
et al., 2012). In particular, these models were successfully
applied to computationally predict target enzymes for
genetic manipulations aimed to increase the production of
tocopherols, which are potent antioxidants and the main
source of vitamin E in the human diet. It is expected that
similar new, systems-oriented approaches will serve to
better understand isoprenoid metabolism in the context
of the whole plant, eventually allowing a more rational
design of metabolic engineering approaches for an
improved production of isoprenoids of interest as drugs,
phytonutrients, or biofuels in plant biofactories.
Funding
Work in our lab is supported by grants from the Spanish
Direccin General de Investigacin (BIO2011-23680 and
PIM2010IPO-00660), Consejo Superior de Investigaciones
Cientficas (2010CL0039), Generalitat de Catalunya
(2009SGR-26), Programa Iberoamericano de Ciencia y
Tecnologia para el Desarrollo (IBERCAROT), and European
Union FP7 (TiMet, contract 245143).
References
Alder, A., etal. (2012). The path from beta-carotene to carlactone,
a strigolactone-like plant hormone. Science. 335,13481351.
Pulido et al. New Insights into Plant Isoprenoid Metabolism 967
Brodersen, P., et al. (2012). Isoprenoid biosynthesis is required
for miRNA function and affects membrane association of
ARGONAUTE 1 in Arabidopsis. Proc. Natl. Acad. Sci. U S A.
109,17781783.
Flores-Perez, F., et al. (2010). PLEIOTROPIC REGULATORY LOCUS
1 (PRL1) Integrates the Regulation of Sugar Responses with
Isoprenoid Metabolism in Arabidopsis. Mol. Plant. 3,101112.
Flores-Perez, U., Sauret-Geto, S., Gas, E., Jarvis, P., and
Rodrguez-Concepcin, M. (2008). A Mutant Impaired in
the Production of Plastome-Encoded Proteins Uncovers a
Mechanism for the Homeostasis of Isoprenoid Biosynthetic
Enzymes in Arabidopsis Plastids. Plant. Cell. 20,13031315.
Gomez-Roldan, V., etal. (2008). Strigolactone inhibition of shoot
branching. Nature. 455,189194.
Huizinga, D.H., etal. (2010). Farnesylcysteine lyase is involved in
Downloaded from http://mplant.oxfordjournals.org/ at North South University on November 5, 2012
negative regulation of abscisic acid signaling in Arabidopsis.
Mol. Plant. 3,143155.
Joyard, J., etal. (2009). Chloroplast proteomics and the compart-
mentation of plastidial isoprenoid biosynthetic pathways. Mol.
Plant. 2,11541180.
Leivar, P., et al. (2011). Multilevel control of Arabidopsis
3-hydroxy-3-methylglutaryl coenzyme A reductase by protein
phosphatase 2A. Plant. Cell. 23,14941511.
Li, X., Gianoulis, T.A., Yip, K.Y., Gerstein, M., and Snyder, M. (2012).
Extensive in vivo metabolite-protein interactions revealed by
large-scale systematic analyses. Cell. 143,639650.
Mintz-Oron, S., Meir, S., Malitsky, S., Ruppin, E., Aharoni, A., and
Shlomi, T. (2012). Reconstruction of Arabidopsis metabolic net-
work models accounting for subcellular compartmentalization
and tissue-specificity. Proc. Natl. Acad. Sci. U S A. 109,339344.
Paetzold, H., et al. (2010). The isogene 1-deoxy-D-xylulose
5-phosphate synthase 2 controls isoprenoid profiles, precur-
sor pathway allocation, and density of tomato trichomes. Mol.
Plant. 3,904916.
Sapir-Mir, M., etal. (2008). Peroxisomal localization of Arabidopsis
isopentenyl diphosphate isomerases suggests that part of the
plant isoprenoid mevalonic acid pathway is compartmentalized
to peroxisomes. Plant. Physiol. 148,12191228.
Simkin, A.J., et al. (2011). Peroxisomal localisation of the final
steps of the mevalonic acid pathway in planta. Planta.
234,903914.
Umehara, M., etal. (2008). Inhibition of shoot branching by new
terpenoid plant hormones. Nature. 455,195200.
Vranova, E., Coman, D., and Gruissem, W. (2012). Structure and
dynamics of the isoprenoid pathway network. Mol. Plant.
5,318333.
Xiao, Y., etal. (2012) Retrograde signaling by the plastidial metab-
olite MEcPP regulates expression of nuclear stress-response
genes. Cell. 149,15251535.