Mechanism of 1,4-dehydrogenation catalyzed by a fatty acid
(1,4)-desaturase of
Calendula officinalis
Darwin W. Reed
1
, Christopher K. Savile
2
, Xiao Qiu
1
, Peter H. Buist
2
and Patrick S. Covello
1
1
Plant Biotechnology Institute, Saskatoon, SK, Canada;
2
Department of Chemistry, Carleton University, Ottawa, Ontario, Canada
The mechanism by which the fatty acid (1,4)-desaturase of
Calendula officinalis produces calendic acid from linoleic acid
has been probed through the use of kinetic isotope effect
(KIE) measurements. This was accomplished by incubating
appropriate mixtures of linoleate and regiospecifically
dideuterated isotopomers with a strain of Saccharomyces
cerevisiae expressing a functional (1,4)-desaturase. GC-MS
analysis of methyl calendate obtained in these experiments
showed that the oxidation of linoleate occurs in two discrete
steps since the cleavage of the C11-H bond is very sensitive to
isotopic substitution (k
H
/k
D

[6–8](Fig. 1B). The latter reaction is particularly noteworthy
given the current interest in conjugated fatty acids with
respect to their role in human nutrition [9] as well as
commercial applications [10].
As part of ongoing research into the structure–function
relationships of FAD2 type enzymes, a closer examination
of calendate formation is clearly warranted. Early labelling
experiments using marigold seed homogenates and labelled
linoleate precursors demonstrated that calendic acid is
produced by an apparent (1,4)-dehydrogenation process
whereby a linoleoyl substrate loses one hydrogen from C8
and C11, respectively [11]. No oxygenated intermediates
were detected. These results as well as related substrate
specificity data [6] point to a mechanism which is analogous
to that proposed for the more common (1,2)-dehydrogen-
ation reactions of fatty acid desaturases (Fig. 2). The
mechanistic model [12] for the latter process features an
initial, energetically difficult hydrogen abstraction step,
which generates a very short-lived, carbon-centered radical
intermediate, or its iron-bound equivalent (not shown). This
species collapses rapidly to give an unsaturated product by
what is formally a second hydrogen abstraction, although a
one electron oxidation/proton removal sequence cannot be
rigorously excluded at this time. The stepwise nature of this
transformation is supported by kinetic isotope effect (KIE)
studies of several membrane-bound fatty acid desaturases.
In all cases examined, one C-H cleavage was found to be
subject to a large primary deuterium kinetic isotope effect
while the second C-H bond rupture was insensitive to
isotopic substitution [12–22]. This pattern of KIEs is

Definition: The term (1,4)-desaturase denotes an enzyme that converts
an isolated carbon-carbon double bond in a fatty acid into two
conjugated double bonds by what is formally a 1,4-dehydrogenation
reaction. Such enzymes have also been termed conjugases [7].
(Received 23 July 2002, accepted 28 August 2002)
Eur. J. Biochem. 269, 5024–5029 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03209.x
EXPERIMENTAL PROCEDURES
Materials
Methyl linoleate (> 99%) was purchased from Nu-Chek-
Prep, Inc. The two regiospecifically dideuterated methyl
linoleates ([8,8
2
H
2
]-1,11,11-
2
H
2
]-1) required for the KIE
study were prepared by routes which were very similar to
those reported previously for the synthesis of the corres-
ponding chiral monodeutero analogues [27]. Thus, the
tosylate of 8-hydroxy-[8,8
2
H
2
]octanoic acid was reacted
with lithium acetylide-ethylenediamine complex to give
[8,8
2

tion of substrates was carried out by flash chromatography
(silica gel, 0.5% v/v ethyl acetate/hexane) and HPLC
fractionation as previously described [16]. GC-MS analysis
[16] of the final deuterated products revealed that each
isotopomer consisted essentially entirely of dideuterated
species (m/z 296; 294 for nondeuterated analogue).
1
Hand
13
C NMR analysis confirmed the position of the two
deuterium atoms for each isotopomer as indicated by the
presence/absence of the diagnostic bisallylic signals at d
2.77 p.p.m. (
1
H) and 25.67 p.p.m. (
13
C) for [11,11-
2
H
2
]-1,
respectively, and an approximately two-fold attenuation of
the overlapping allylic signals (C-8,C-13) at d 2.02 p.p.m.
(
1
H) and 27.25 p.p.m. (
13
C) for [8,8
2
H

Analytical procedures
For fatty acid analysis, yeast pellets were saponified by
adding 2 mL 10% KOH/methanol and heating at 80 °Cfor
2 h. The mixture was then cooled and pre-extracted with
2 · 2 mL hexane to remove nonsaponifiable lipids. The
reaction mixture was then neutralized with 50% acetic acid
to  pH 5 and the fatty acids were extracted with 2 · 2mL
hexane. The hexane was removed under a nitrogen stream
and the mixture, including the conjugated fatty acids, was
esterifiedwith2mL1%H
2
SO
4
in methanol at 50 °Cfor1
h (This methylation method has been found to be most
suitable for conjugated fatty acid ester analysis; W. Christie,
Mylnefield Research Services Ltd., Dundee, Scotland,
personal communication.) The cooled mixture was extrac-
tedwith2· 2 mL hexane. The pooled hexane was washed
with 2 mL H
2
O and concentrated under N
2
for HPLC
purification, GC or GC-MS analysis.
GC-MS analysis of yeast lipids was performed using a
Fisons VG TRIO 2000 mass spectrometer (VG Analytical,
UK) controlled by Masslynx version 2.0 software, coupled
to a GC 8000 Series gas chromatograph as previously
described [16] except that a narrow EI

considerations affect the conclusions reached in this paper.
The use of a competitive rather than a noncompetitive
experimental design has allowed KIE determinations to be
carried out for both in vitro and in vivo desaturase systems.
The results have correlated well with KIE data obtained by
other methods [30–34]. Our methodology dictates that
interference by endogenous d
0
-substrate, if present, must be
eliminated: this has been accomplished previously through
the use of unnatural chain-shortened substrates or ana-
logues bearing a remote ÔthiaÕ- or deuterium mass label.
Such measures proved unnecessary in the case of the
linoleate-calendate reaction since the host yeast system used
for this purpose does not biosynthesize the relevant
substrate.
Optimal incubation conditions for our KIE studies were
set up in a preliminary experiment: methyl linoleate 1
(100 mgÆL
)1
) was administered to cultures (50 mL) of the
pYJ/DTY10a2 strain of S. cerevisiae incubated at 20 °Cfor
3 days to permit relatively rapid growth and then at 15 °C
for a further 3 days to reach saturation at a temperature
which has been found to give better substrate conversion
rates. The cells were harvested by centrifugation and the
lipids were isolated via a hydrolysis/methylation sequence
known to be suitable for conjugated fatty acid esters (10%
w/v KOH/CH
3

that of the trial experiment. The deuterium content of the
olefinic fatty acid methyl esters in the cellular lipid extract
was assessed by GC-MS as described in the Experimental
section. The d
2
/d
0
ratio of the linoleate isotopomers found
in the cells was essentially identical to that of the starting
material in both incubations, as is required for these types of
competitive KIE measurements [35]. No loss of label due to
reversible exchange of deuterated linoleate at C-8 or C-11
could be detected. Mass spectral analysis of the calendate
fraction revealed that in both incubations, this material
consisted entirely of a d
0
/d
1
mixture indicating a loss of one
deuterium from the d
2
-substrate as expected. Product
kinetic isotope effects (k
H
/k
D
) were calculated using the
ratio: [% d
0
(product)/% d

initial attack catalyzed by a soluble plant D
9
desaturase [37].
However our results clearly demonstrate that calendate
production is in fact initiated at C11 as might be expected
for a process which is catalyzed by a homolog of FAD2 – an
enzyme which initiates the conversion of oleate to linoleate
at C12 [12]. Thus, the switch between 1,2 and 1,4-
dehydrogenation could conceivably be controlled by a
fairly small change in oxidant position relative to substrates
which both adopt a conformation allowing syn removal of
Fig. 4. Isotopomers of 1 used to probe the kinetic isotope effects on the
fatty acid (1,4)-desaturase reaction involved in calendate biosynthesis.
Fig. 3. Mass spectrum of biosynthetic methyl calendate. Arrow indi-
cates the molecular ion cluster used to calculate the isotopic content of
deuterated samples.
5026 D. W. Reed et al. (Eur. J. Biochem. 269) Ó FEBS 2002
two proximal hydrogens (H-H distance in both cases
 2.5 A
˚
). This model (Fig. 5) can be tested by determining
the stereochemistry of H-removal for calendate formation
using chiral monodeutero probes [27] and comparing this
result with the known pro-R enantioselectivity at C12,13
observed for D
12
-desaturation [38].
Further evidence for the close relationship between 1,2
and 1,4-dehydrogenation has been obtained recently for a
Spodoptera littoralis desaturating system which converts

ing the GC-MS analysis, Charles Martin for providing the yeast strain
DTY-10a2 and Michele Loewen and Robert Sasata for reviewing the
manuscript.
REFERENCES
1. Shanklin, J. & Cahoon, E.B. (1998) Desaturation and related
modifications of fatty acids. Annu. Rev. Plant Physiol. Plant Mol.
Biol. 49, 611–641.
Fig. 5. Mechanistic model showing the rela-
tionship between oxidant position in D
12
desat-
uration and the (1,4)-desaturase reaction
leading to the formation of a conjugated 8,10-
diene system.
Table 1. Intermolecular isotopic discrimination by the C. officinalis (1,4)-desaturase in the 1,4-dehydrogenation of [8,8
2
H
2
]-linoleate and [11,11-
2
H
2
]-
linoleate.
Isotopic Ratio
a
Substrates
b
Products
d

0
,d
1
,d
2
refer to undeuterated, monodeuterated
and dideuterated species, respectively.
b
The isotopic ratios of the substrates were determined by mass spectral examination of the linoleate
isotopomers isolated from the cellular pool. These values are similar (within experimental error) to those obtained for the two starting
mixtures prior to incubation: 1.02 (d
0
: 11d
2
); 1.92 (d
0
:8d
2
).
c
The average KIE (four incubations) ± standard deviation.
Ó FEBS 2002 Calendic acid biosynthesis (Eur. J. Biochem. 269) 5027
2. Lee, M., Lenman, M., Banas, A., Bafor, M., Singh, S., Schweizer,
M., Nilsson, R., Liljenberg, C., Dahlqvist, A., Gummeson, P O.,
Sjo
¨
dahl, S., Green, A. & Stymne, S. (1998) Identification of non-
heme diiron proteins that catalyze triple bond and epoxy group
formation. Science 280, 915–918.
3. Okuley, J., Lightner, J., Feldmann, K., Yadav, N., Lark, E. &

283–298.
10. Derksen, J.T.P., Cuperus, F.P. & Kolster, P. (1995) Paints and
coatings from renewable resources. Ind. Crops Prod. 3, 225–236.
11. Crombie, L. & Holloway, S.J. (1985) The biosynthesis of
calendic acid, octadeca-(8E,10E,12IX)-trienoic acid, by developing
marigold seeds: origins of (E,E,Z)and(Z,E,Z) conjugated
triene acids in higher plants. J. Chem. Soc. Perkin Trans. I, 2425–
2434.
12. Buist, P.H. & Behrouzian, B. (1998) Deciphering the cryptor-
egiochemistry of oleate D
12
desaturase: a kinetic isotope effect
study. J. Am. Chem. Soc. 120, 871–876.
13. Buist, P.H. & Behrouzian, B. (1996) Use of deuterium kinetic
isotope effects to probe the cryptoregiochemistry of D
9
desatura-
tion. J. Am. Chem. Soc. 118, 6295–6296.
14. Pinilla, A., Camps, F. & Fabrias, G. (1999) Cryptoregiochemistry
of the D
11
-myristoyl-CoA desaturase involved in the biosynthesis
of Spodoptera littoralis sex pheromone. Biochemistry 38, 15272–
15277.
15. Abad, J.L., Camps, F. & Fabria
`
s, G. (2000) Is hydrogen tun-
neling involved in acylCoA desaturase reactions? The case of a D
9
desaturase that transforms (E)-11-tetradecenoic acid into Z and

2879–2881.
22. Behrouzian, B., Fauconnot, L., Daligault, F., Nugier-Chauvin, C.,
Patin, H. & Buist, P.H. (2001) Mechanism of fatty acid desa-
turation in the green alga Chlorella vulgaris. Eur. J. Biochem. 268,
3545–3549.
23. Cook, G.K. & Mayer, J.M. (1994) C-H bond activation by metal-
oxo species: oxidation of cyclohexane by chromyl chloride. J. Am.
Chem. Soc. 116, 1855–1868.
24. Buist, P.H. & Marecak, D.M. (1992) Stereochemical analysis of
sulfoxides obtained by diverted desaturation. J. Am. Chem. Soc.
114, 5073–5080.
25. Buist, P.H., Alexopoulos, K.A., Behrouzian, B., Dawson, B. &
Black, B. (1997) Synthesis and desaturation of monofluorinated
fatty acids. J. Chem. Soc. Perkin Trans. 1, 2617–2624.
26. Broun, P., Shanklin, J., Whittle, E. & Somerville, C. (1998)
Catalyic plasticity of fatty acid modification enzymes
underlying chemical diversity of plant lipids. Science 282, 1315–
1317.
27. Crombie, L. & Heavers, A.D. (1992) Synthesis of linoleic acid with
chiral isotopic labelling at a flanking and a medial allylic methy-
lene: the (8R,9Z,12Z)-[8-
2
H] and (11R,9Z,12Z)-[11-
2
H]-stereo-
isomers, and (Z)-[2,2-
2
H
2
]-non-3-enal. J. Chem. Soc., Perkins

37. Fritsche,K.,Hornung,E.,Peitzsch,N.,Renz,A.&Feussner,I.
(1999) Isolation and characterization of a calendic acid producing
(8,11)-linoleoyl desaturase. FEBS Lett. 462, 249–253.
38. Morris,L.J.,Harris,R.V.,Kelly,W.&James,A.T.(1968)The
stereochemistry of desaturation of long-chain fatty acids in
Chlorella vulgaris. Biochem. J. 109, 673–678.
39. Rodriguez, S., Camps, F. & Fabrias, G. (2001) Synthesis of
gem-dideuterated tetradecanoic acids and their use in investigating
the enzymatic transformation of (Z)-11-tetradecenoic acid
5028 D. W. Reed et al. (Eur. J. Biochem. 269) Ó FEBS 2002
into (E,E)-10,12-tetradecadienoic acid. J. Org. Chem. 66, 8052–
8058.
40. Jones, J.P. (1987) The separation of the intermolecular isotope
effect for the cytochrome P-450 catalyzed hydroxylation of
n-octane into its primary and secondary components. J. Am.
Chem. Soc. 109, 2171–2173.
41. Broadwater, J.A., Whittle, E. & Shanklin, J. (2002) Desaturation
and Hydroxylation. Residues 148 and 324 of Arabidopsis FAD2,
in addition to substrate chain length, exert a major influence in
partitioning of catalytic specificity. J. Biol. Chem. 277, 15613–
15620.
Ó FEBS 2002 Calendic acid biosynthesis (Eur. J. Biochem. 269) 5029