New chemical clues for broomrape-sunflower host - Parasite interactions: Synthesis of guaianestrigolactones

Article abstract of DOI:10.1021/jf900870j

A comparative structure-activity relationship (SAR) study has been conducted with several guaianolide sesquiterpene lactones (SLs) as inducers of the germination of sunflower broomrape (Orobanche cumana) seeds. Compounds were selected and synthesized to s

Full text of DOI:10.1021/jf900870j

J. Agric. Food Chem. 2009, 57, 5853–5864 5853
DOI:10.1021/jf900870j
New Chemical Clues for Broomrape-Sunflower Host-Parasite
Interactions: Synthesis of Guaianestrigolactones
‡
´
´
´
F
RANCISCO A. MACIAS,*,† MARIA D. GARCIA-DI´ AZ,† ALEJANDRO PEꢀREZ
-
DE-LUQUE,
§
†
D
IEGO
R
UBIALES
,
AND
JUAN C. G. GALINDO
†Grupo de Alelopatı
´
a, Depto. de Quı
´
mica Organica, Facultad de Ciencias, Universidad de Cadiz, Avda.
ꢀ
ꢀ
‡
ꢀ
ꢀ
Republica Saharaui s/n, Apdo. 40, 11510, Puerto Real, Cadiz, Spain, IFAPA-CICE, CIFA “Alameda del
Obispo”, Apdo. 3092, 14080, Cordoba, Spain, and Institute of Sustainable Agriculture, IAS-CSIC,
§
ꢀ
ꢀ
14080, Cordoba, Spain
A comparative structure-activity relationship (SAR) study has been conducted with several
guaianolide sesquiterpene lactones (SLs) as inducers of the germination of sunflower broomrape
(Orobanche cumana) seeds. Compounds were selected and synthesized to study the influence of
the lactone-enol-γ-lactone moiety on the selectivity of SLs toward the stimulation of sunflower
broomrape germination. The results clearly illustrate that SLs are recognized only by O. cumana,
while the introduction of a strigol-like second lactone moiety in the guaianolide backbone results in
the loss of specificity and hence the germination of other broomrape species. We have named this
new class of compounds guaianestrigolactones (GELs).
KEYWORDS: Sesquiterpene lactones; guaianolides; guaianestrigolactones; SAR; germination
stimulants; allelopathy; parasitic weeds; broomrape; Orobanche cumana; Orobanchaceae
INTRODUCTION
Host-parasite interaction is extremely specific and depends on
chemical recognition. Several natural germination inducers have
been isolated from their corresponding hosts (16). A number of
natural products obtained from nonhosts also present germina-
tion-inducing activity (16). With the exception of sorgoleone, all
natural inducers obtained from host plants belong to the so-called
strigolactone group of compounds, which have a postulated
biogenetic origin in the carotenoids pathway (17). These materials
have been isolated from sorghum (18-20), red clover (21), and
cow pea (22). Surprisingly, strigolactones also occur in nonhost
species such as cotton (23), Menispermum dauricum, and Stepha-
nia sepharantha (24), and it has been suggested that they might
occur widely in the plant kingdom. Host recognition might be the
result of a coevolution process by which the parasite takes
advantage of and recognizes certain chemicals exuded by the
host plant for other purposes (mainly defense, excretion, or
mycorrhyzation (25, 26)). These chemicals are detected by the
parasite as an indication of a viable host nearby and trigger the
cascade of events that lead to germination. Environmental factors
such as humidity and temperature are crucial for a positive
germination response (27). During the conditioning period,
changes occur inside the seed and prepare it for the detection of
a chemical signal. This period is water- and temperature-depen-
dent (28, 29), and such restrictions are supposed to ensure
germination in the best conditions for both the parasite and the
host. In the search for healthy plants to attach to, the healthier the
host the greater the probability for the successful development of
the parasite.
Sunflower (Helianthus annuus L.) is an important crop in
several parts of the world as it is used as a source of vegetable
oil and seeds, and also as a fertilizer. Sunflowers can be para-
sitized specifically by the root holoparasitic plant Orobanche
cumana. Sunflower crops are in danger because of severe infesta-
tion by this parasite in several million hectares throughout eastern
and southern Europe and in Asian countries including China (1).
The effectiveness of conventional control methods to date is
limited because of numerous factors including the complex nature
of the parasites, which reproduce by tiny and long-lived seeds that
can remain dormant and viable for a period of almost 20 years (2),
and the difficulty in diagnosis until the crop is irreversibly
damaged. The intimate connection between host and parasite
also hinders efficient control by herbicides. Once a new episode of
broomrape has been documented in an area, this weed spreads
rapidly because of the high level of seed production per plant
(thousands of seeds) and their small size. Spreading factors, to
name the most common ones, include the wind, the soil attached
to the tires of agricultural machinery, and the residues of the
infested crops.
Some satisfactory results have been found in certain cases by
using herbicides such as glyphosate, imidazolinones, or sulfony-
lureas (1, 3-6). Other possible agricultural practices for broom-
rape control include crop rotation (7), intercropping (8), use of
fertilizers (9), breeding (10), activators of systemic acquired
resistance (11) and resistant cultivars (12-14), and synthetic
germination stimulants (15).
To our knowledge, the specific germination stimulants of
sunflower broomrape remain unknown. However, our previous
work with SLs shows that these compounds are specific inducers
*Corresponding author. Phone: +34-956-016370. Fax: +34-956-
016193. E-mail: famacias@uca.es.
© 2009 American Chemical Society
Published on Web 06/09/2009
pubs.acs.org/JAFC

´
Macıas et al.
5854 J. Agric. Food Chem., Vol. 57, No. 13, 2009
Figure 1. (A) Synthetic procedure followed to obtain strigolactone-like SLs (GELs). The numbering in the second lactone ring has been assigned according to
the latest accepted convention followed for strigolactones (34). (B) Synthetic pathway to obtain GELs.
of O. cumana and do not induce a germination response in other
Orobanche species (30, 31). The main objective of the work
described here was to determine whether sunflower broomrape
is able to recognize the lactone ring present in SLs or if this
specificity is lost when a lactone-enol-γ-lactone system similar
to that of strigolactones is introduced into the SL backbone. We
have named such compounds guaianestrigolactones. If such a
hypothesis is confirmed, then the lactone ring alone could be
identified as the specific bioactiphore of SLs for sunflower
broomrape.
stability of this compound, it is strongly recommended to be immediately
used for the nucleophilic substitution step.
Treatment of 3 with potassium hydride under an argon atmosphere in
dry THF (0 °C) yielded the corresponding alkoxide at C-13, and this was
followed by the addition of an excess of bromolactone 5 dissolved in dry
THF. The reaction mixture was allowed to react for 7 h at room
temperature, after which further evolution was not observed by TLC.
The SN2 substitution on the bromine atom yielded the mixture of
diastereomers 6(a+b) (13%) and 7(a+b) (4%), with starting material
recovered (60%) and traces of the elimination product 1 detected.
A similar procedure was followed using 4 as the starting material, and
this gave rise to the diastereomeric mixture 8(a+b) (7.5%) with 60% of the
starting material recovered. All attempts to separate the two diastereomers
present in each mixture by chromatographic means including chiral phase
HPLC columns were unsuccessful.
MATERIALS AND METHODS
General Considerations and Starting Material. The main objective
was to verify whether the introduction of an enol-γ-lactone-like moiety
into the SL would induce any change in the bioactivity and the selectivity
toward the germination of the different broomrape species. For this study,
the guaianolide backbone was selected as a case study on the basis of our
previous results with several other SL types (30, 31) and because of the
availability of dehydrocostuslactone (1) in multigram amounts to serve as
the starting material. Pure material was obtained from the commercial
root extract of Saussurea lappa (Costus Resin Oil, Pierre Chauvet, S.A.) by
column chromatography on silica gel (Hx/AcOEt 5%) and purified by
recrystallization from Hx/Et2O mixtures along with the other major
compound costunolide. The spectroscopic (1H and 13C NMR and HRMS)
and physical data (m.p.) were fully consistent with those reported in the
literature (32, 33).
Synthesis of Guaianestrigolactones (GELs). 13-Hydroxycostuslac-
tone (2) and 13-hydroxy-epi-costuslactone (3) (or their mixture) were
oxidized to the corresponding aldehydes (11) using the Dess-Martin
periodinane reagent (57% yield). The introduction of the second lactone
ring D was achieved using a procedure (dry THF, 0 °C, base 1 equiv)
similar to that used for dihydro-GELs. Thus, enolate 12 was generated
using potassium tert-butoxide (t-BuOK) as this gave better yields than
potassium hydride. Once enolate 12 was added, it smoothly underwent
SN2 displacement of the bromine atom in racemic bromolactone
5 (Figure 1) to yield a mixture of 13 and 14 in 42% global yield. C-11(S)
(2), C-11(R) (3) hydroxycostuslactone, or a mixture of the two can be used
as starting material for the synthesis of guaianestrigolactones as the
stereospecificity at C-11 is lost after the formation of the enolate. In this
case, the racemic mixture of the diastereoismers at C-20 (13 and 14) could
be separated by HPLC (silica gel, Lichrospher Column and Hx/AcOEt
4:1 as eluant), and the compounds were characterized spectroscopically.
Bioassays. Orobanche spp. Germination Bioassays. Two popu-
lations of O. cumana, denoted 12 and 23, with different degrees of viru-
lence were collected from experimental sunflower infected fields by
Synthesis of 11,13-Dihydroguaianestrigolactones (11,13-dihy-
droGEls). In order to assess the importance in the germination activity
of the second lactone ring alone, i.e., without the enol moiety, compounds
were synthesized in two steps according to the synthetic route depicted in
Figure 1. The numbering for ring D is assigned according to the latest
accepted convention for strigolactones (34).
ꢀ
Michael addition of the hydroxy group was performed according to the
methodology developed by our group (35), and this yielded the mono- and
dihydroxylated derivatives 2 (21%), 3 (30%) and 4 (6%), with some
starting material 1 remaining (30%).
2-Methyl-4-bromobut-2-ene-1,4-olide (5) was generated by treatment
of the corresponding 2-methylbut-2-ene-1,4-olide (Sigma-Aldrich, Co.)
with NBS in CCl4 solution and a catalytic amount of benzoylperoxide as
a radical initiator (36). The reaction mixture was heated under reflux at
78 °C for 1.5 h to give 5 in almost quantitative yield. The compound
obtained was the racemic mixture of epimers at C-40. Because of the low
Dr. Gonzalez Carrascosa (Semillas Cargill, Spain). Seeds of O. ramosa were
provided by Dr. D. M. Joel (Department of Weed Research, Agricultural
Research Organization, Newe-Ya’ar Research Center, Israel). The Oro-
banche germination bioassay on filter paper was performed as previously
reported (30,37). Compounds were tested at 10, 1, 0.1, and 0.01 μM aq soln
(0.1% acetone as cosolvent). Seeds were preconditioned in darkness (24 °C,
11 days, and 0.3 mM MES (2-[N-morpholino]-ethanesulfonic) acid buffer;
pH was adjusted to 6 with 0.1 M NaOH) and germinated in the presence of
the test compound (darkness, 24 °C, 6 days). A parallel treatment with 10 μM
GR-24 was set up as an internal positive control. Germination was observed

Article
J. Agric. Food Chem., Vol. 57, No. 13, 2009 5855
Figure 2. (A)Main diagnosticNOE effectobserved isdepictedinthe onlylocal minimum that allowssuchan effect(2%ofthe totalconformer population)inthe
diastereomer(11R,20S). (B) Minimum energy conformer (84%of the total population) in thediastereomer (11R,20R);noneofthelocal minima wereableto give
any NOE effect. Conformer distributions were obtained using the Montecarlo MMFF force field (Spartan’06).
under a dissection microscope (30ꢀ) starting on the third day after the
addition of the compound. Seeds were considered to be germinated when the
radicle was at least 0.2 mm long. Germination was expressed as a percentage
of the total seeds.
Stock solutions were prepared with acetone and diluted with 0.3 mM
MES to obtain a 10 μM (0.1% acetone) solution; test solutions at 1,
0.1, and 0.01 μM were prepared by diluting with 0.3 mM MES (0.1%
acetone aq solns).
confirmed by the new signals appearing in the 1H NMR spectrum
at δ 5.11 (1H, m, H-20) and δ 1.91 (3H, dd, H-60), which
correspond to the hemiketal proton and the methyl group present
in the new lactone ring, respectively. The 13C NMR spectrum also
contains new signals corresponding to the lactone moiety (car-
bons C20 to C60; Table 2). The two signals obtained for the vinyl
proton at δ 7.07 (0.19H, m, H-30a) and δ 7.11 (0.81H, m, H-30b)
evidence the racemic nature of the mixture, with each correspond-
ing to a different enantiomer at C-20. Signal duplicity is also
observed in the 13C NMR spectrum, and each set of signals was
Molecular Modeling. Molecular calculations were performed using
Spartan’06, PCModel 7.50.00 (Serena Software), and Chem3D Ultra 7.5.
Spartan’06, PCModel 7.50.00, and Chem3D Ultra 7.5 were used to obtain
local minimum energy structures using semiempirical PM3 and AM1
calculations. Conformer distribution (Spartan’06) was obtained using the
Montecarlo program with the MMFF force field; MAXCONFS were set
to 1000, which proved to be sufficient as none of the calculations yielded
more than 100 conformers. Single point Hartree-Fock (3-21G(*)) calcu-
lations were performed in some cases to refine the values obtained for the
heat of formation (38). For transition state calculations, a first approach to
the geometry was obtained using semiempirical calculations (PM3; key-
words, transition state geometry; geometrycycle=10000), and the resulting
geometry was then refined using ab initio calculations (restricted Hartree-
Fock using the basis set 3-21G(*)). The energy values were then refined
using electron density correlation methods (B3LYP using the basis set
6-21G(*), keywords, single point energy; geometrycycle=10000).
Chem3D Ultra 7.5 was used in some cases to obtain a fast approach to
basic minimum energy structures but not for transition state calculations.
The GMMX command was then used to generate other low energy local
minima. These conformers were further minimized using PM3 calculations
(MOPAC) (39) to obtain more accurate theoretical ΔH°f searching for
the real minimum energy conformer. The local minima obtained were
compared with those obtained from Spartan’06. Usually, geometries
obtained from Chem3D and Spartan’06 were similar or identical. In cases
where a significant difference arose, refined ab initio calculations using
Spartan’06 allowed the geometry to be chosen on the basis of more
accurate enthalpies of formation. Volume, height, and energies of
the molecular orbitals were obtained from PCModel (MOPAC) and
Spartan’06.
unequivocally correlated through H-13C g-HSQC and H-13C
g-HMBC heteronuclear correlation experiments. The connection
between the guaianolide backbone and the new lactone ring is
confirmed through the 1H-1H NOESY experiment as a correla-
tion between the signals corresponding to H-11 and H-20 can be
observed, corresponding to the NOE effect between these two
protons (Figure 2). The integral ratio (9.94:2.25, 6a/6b) also
establishes relative yields for each diastereomer as 82% (6a)
and 18% (6b) (d.e.=64%) for the total mixture (Figure 2).
The absolute stereochemistry of each diastereomer was un-
equivocally established through NOESY experiments: a positive
NOE effect was observed between the proton signals of H-11 and
H-20 in the major diastereomer (6a). A conformational search on
both C-20R and C-20S diastereomers concluded that major
minimum energy conformers, accounting for 84% and 71% of
the total conformer population, were not able to present any
NOE effect because of the opposite disposition of these protons.
However, while none of the local minima obtained for the C-20R
diastereomer were able to present NOE effects between H-11 and
H-20, a local minimum with the appropriate spatial disposition
accounting for 2% of the conformer population was found in the
C-20S diastereomer. Consequently, an absolute configuration
(11R,20S) can be established for 6a, while the minor diastereomer
has the (11R,20R) configuration (Figure 2).
1
1
Not surprisingly, treatment of 3 also led to a small amount of
the mixture (7a+7b), and this had spectroscopic properties
similar to those of mixture (6a+6b) (Tables 1 and 2) but without
any diastereomeric excess, as shown by the relative intensities of
the integrals obtained for H-20 and H-30 (7a/7b, 1:1, d.e.=0) and
the value of [R]25D=+0.0. Extensive NOE and NOESY studies
RESULTS AND DISCUSSION
Spectroscopic Characterization: DihydroGELs. The mixture of
diastereomers (6a+6b) presents a molecular ion in the HRMS at
m/z 344.1601, which is consistent with the molecular formula
C20O5H24. The incorporation of the lactone moiety is further

´
Macıas et al.
5856 J. Agric. Food Chem., Vol. 57, No. 13, 2009
Table 1. 1H NMR Chemical Shifts of 6-11, 13, and 14a
H
6a/6b
7a/7b
8a/8b
2.87
1.82a
1.94a
2.50b
2.02b
2.81
9
10
11
13
14
1
2.89 ddd
2.87
1.86
1.94
2.45a
2.17am
2.83
3.95
2.89 m
1.82
2.90 ddd
1.83 dddd
1.97
2.93 m
1.87 m
1.94 m
2.50
2.89 ddd
1.87
1.87
2.89
1.87
1.87
2.51
2.51
2.78
3.97
3.03
2.54
1.40
2.13m
2.3
2R
2β
3R
3β
5
1.85 dddd
1.93 m
2.53 m
1.97dddd
2.51
2.51
2.48
2.48
2.50
2.50
2.15 m
2.50
2.82
2.80 dd
3.97 dd
2.23 m
2.89
4.15
3.14
4.23
2.77
3.98
6
4.04/3.99
2.46
2.08
4.03
7
2.41
2.15 m
1.31
2.05b
2.42b
2.56 ddd
2.18
1.44
3.28
2.72 dddd
2.17
1.32
3.01
2.50
8R
8β
9R
9β
11
13
130
14
140
15
150
20
2.26 dddd
1.31 dddd
2.15 dddd
1.50
2.54
1.42
1.40
2.12ddd
2.38ddd
2.08c
1.98c
2.02
2.51
2.02
2.04 brdd
3.94 dd (2H)
2.41 m
3.85 dd/4.07 brs
3.97
3.26d
9.88
4.05/3.79 d
3.70/3.73 d
4.89/4.87 s
4.79/4.76 s
5.18/5.15
5.06/5.07 dd
5.80/5.71
6.77
4.48
4.37
4.76
4.66
4.82
4.74
5.33
5.11
7.43
7.43
4.86 brs
4.77 brs
4.87
4.77
4.87
4.77
4.89brs
4.79dd
5.17d
4.86s
4.80s
5.25
4.86
4.80
5.25
5.04
6.09
6.90
5.16 brdd
5.04 brdd
5.18
5.20 brd
5.07 brd
5.04brd
5.89/5.83
6.85/6.80
5.05d
5.05
6.08dq
6.91dq
5.11 (a) ddq/ 5.46 (b) dq
7.11(a) /7.07 (b) dq
30
7.99brt (2H)
7.43 dd (2H)
7.57 tt
7.99
7.58
40
50
7.46
7.58
600
1.91(a) dd
1.94/1.96
1.95/1.93
7.43 dd (2H)
7.99brt (2H)
2.02
2.02
700
3
7.99
8.04 dd
7.60 dd
7.43 dd
7.60 dd
8.04 dd
400
500
600
700
a 600 and 400 MHz, CDCl3, signal of residual CHCl3 centered at δ 7.25 ppm. a,b,c Superscripts within the same column: these signals may be interchanged.
Table 2. 13C NMR Chemical Shifts of 6-11, 13, and 14a
C
6
7
8
9
10
47.1
11
13
48.0
14
1
47.1 d
30.1 t
46.9/47.0
30.2
47.04/46.7
30.1/29.7
32.5
47.1
27.4
47.1
30.2
32.6
151.3
51.7
86.1
47.9
30.3
2
27.6
32.9
30.3
32.5
3
32.5ad
151.5 s
51.8 d
85.5 d
43.2 d
32.2at
37.2 t
32.5/32.6
151.6/151.7
51.7/51.8
85.6/85.7
47.7/48.1
32.4
32.7
32.5
4
151.39/151.33
52.5
151.6
52.6
151.9
52.6
151.0
51.9
151.5
51.9
5
6
83.8/83.6
51.0/50.9
26.9/26.7
37.2/37.1
149.6/149.5
76.1/75.8 s
177.2
83.8
51.5
84.5
46.5
85.1
43.5
85.2
43.4
7
8
31.2
37.7
30.3
37.6
29.6
35.4
29.8
35.7
9
37.5/37.7
149.8
37.3
149.3
10
11
12
13
14
15
10
20
30
40
50
600
700
100
2
149.6 s
31.7 d
177.2 s
85.5 t
149.9
76.1
149.2
80.6
149.0
112.2
171.4
150.5 d
112.8
109.8
149.0
112.0
171.3
150.4
112.7
109.7
43.3/44.6
175.4
176.6
65.0
172.0
63.5
177.8
195.2
112.4
109.4
64.6 (a)/ 66.1 (b)
111.95/112.0
109.1/109.2
70.4/68.2
112.3
109.63/109.56
112.1/112.2 t
109.3 t
112.6
109.8
166.2 s
129.3 s
130.1
128.9 d
133.9 d
128.9 d
130.1 d
112.4
109.3
165.4
128.85
130.0
128.63
133.9
128.63
130.0
164.6 s
128.77 s
129.8 d
128.60 d
133.6 d
128.60 d
129.8 d
78.3 (a)/78.6(b) d
148.4 (a) /148.6 (b) d
130.5 s
101.1 (a)/102.0 (b)
142.7/142.8
134.1/134.7
101.2/100.6
142.14/142.09
135.1/134.6
171.1
100.7 d
140.9
135.9
170.2
10.8
100.6
140.7
136.1
170.2
10.8
173.8 s
10.6 q
171.7/171.8
10.60/10.64
10.6
300
400
500
600
700
a 100.6 and 50.3 MHz, CDCl3, signal of residual CHCl3 centered at δ 77.0 ppm. Degree of protonation and assignments were established by APT, HETCOR, and HSQC
experiments; multiplicities are not repeated if identical with those in the preceding column. a,b,c Superscripts within the same column: these signals may be interchanged.

Article
J. Agric. Food Chem., Vol. 57, No. 13, 2009 5857
did not show any signal of diagnostic importance to assign the
absolute configuration of each diastereomer. The formation of
this mixture is explained through base-catalyzed epimerization
at C11 of the starting material, as this has an acidic proton in
this position. The higher yield obtained for the mixture with an
(11R) configuration (6a+6b) is in good agreement with our
previous results concerning the relative stability of the two
epimers (2 and 3) when obtained by a Michael addition of a
hydroxy group to the dehdyrocostuslactone (1) (35).
from the addition of the benzoyl moiety, 9 (10%) and 10 (6%),
can be isolated. Compound 9 gave rise to a molecular ion in the
HRMS at m/z 368.1618, corresponding to the molecular formula
C22O5H24. The major diagnostic signals in the 1H NMR spectrum
were those corresponding to the aromatic ring at δ 7.43 (2H,
dd, H-40), δ 7.57 (1H, dd, H-50), and δ 7.99(2H, dd, H-30). Thiswas
further substantiated by the 13C NMR spectrum, which con-
tained signals corresponding to the carbon ester at δ 166.2 (C-10)
and the aromatic ring (one quaternary carbon, one methyne
carbon, and two equivalent methyne carbons; Table 2).
Finally, treatment of 4 under the same conditions yielded the
mixture of diastereomers 8a and 8b, again as an equimolecular
mixture. As discussed previously, the introduction of the new
lactone ring was confirmed by HRMS, which gave a molecular
ion at m/z 360.1543, consistent with a molecular formula
Compound 10 gave a molecular ion in the HRMS at m/z
472.1889, corresponding to the molecular formula C29O6H28, and
the main fragmentation products at m/z 350.1533 ([M -
PhCOOH]+) and 228.1146 ([M - 2PhCOOH]+). Evidence for
the incorporation of two benzoyl ester groups could also be
1
C20O6H24. The H NMR spectrum presented major diagnostic
1
signals at δ 6.77 (2H, m, H-30), corresponding to the vinyl proton
at ring D in both diastereomers, and differentiated signals for
the methyl group and the H-20 proton in each diastereomer
(δ 1.93 3H, dd, H-60b; δ 1.95 3H, dd, H-60a; δ 5.71 1H, dd,
H-20b; δ 5.80 1H, dd, H-20a). The integral ratio for each pair of
signals was 1:1, which established their relative proportions as 1:1.
The incorporation of the lactone ring into the guaianolide back-
bone was further substantiated in the 13C NMR spectrum by the
presence of new signals at δ 171.1 (C-50, lactone group), δ 142.1
(C-30, vinyl carbon), and δ 10.6 (C-60, methyl group). Once again,
each diastereomer presented differentiated chemical shifts for
carbon C-20 (δ 100.6, C-20a; δ 101.2, C-20b). Surprisingly,
differentiated signals can also be observed for H-13, H-130,
H-6, C-13, and C-6 (Tables 1 and 2). Signals for each diastereomer
were assigned by using homo- and heteronuclear correlation
experiments (1H-1H COSY and 1H-13C gHMQC), but the abso-
lute configuration at C-20 could not be elucidated.
observed in the H NMR spectrum as two independent sets of
aromatic signals, as clearly established through the 1H-1H COSY
experiment. As these compounds represent an interesting mod-
ification of the primary skeleton to investigate the effect of
increasing volume and steric requirements in the vicinity of the
reactive center on the germination activity, they were also used in
the bioassays.
(6R),(7S),(11R)-13-[40-Methyl-30-ene-20-oxy-γ-butyrolactone]-
4(15),10(14)-diene-guaiane-6R,12-olide (mixture of 6a and 6b;
d.e.=64%) was obtained as a colorless oil; [R]25D+5.6 (c 0.04,
CHCl3); IR (neat, KBr) νmax 2963, 2924, 2851 (C-H), 1760, 1730
(CdO, γ-lactone) cm-1 1H NMR (600 MHz): See Table 1.
;
J (Hz): 1,2R=8.1; 1,2β=5.0; 5,6=5,1=9.2; 6.7=9.3; 7,8R=
3.1; 7,8β=13.2; 8R,8β=12.1; 8R,9R=12.1; 8R,9β=9.4; 8β,9R=
11.1; 8β,9β=4.8; 11,13=11,130 =9.4; 3,15=5,15=2.3; 20(a),
30(a)=3.1; 30(a),60(a)=1.5; 20(a),60(a)=1.8; 20(b),60(b)=4.0. 13C
NMR (100 MHz): see Table 2. EIMS (70 eV) m/z (rel. int.):
344 [M]+ (0.5), 328 [M-CH4]+ (23), 216 [M-CH4 -CH3O-
C4O2H]+ (17), 159 [M-CH4-CH3O-C4O2H-CO2-CH3]+
(100); HRMS m/z 344.1606 (calcd for C20O5H24 344.1624).
(6R),(7S),(11S)-13-[40-Methyl-30-ene-20-oxy-γ-butyrolactone]-
4(15),10(14)-diene-guaiane-6R,12-olide (racemic mixture of 7a
and 7b) was obtained as a colorless oil; [R]25D +0.0 (c 0.04,
CHCl3); IR (neat, KBr) νmax 2963, 2926, 2856 (C-H), 1770, 1734
(CdO, γ-lactone) cm-1; 1H NMR (600 MHz): see Table 1. J (Hz):
1,2R=8.2; 1,2β=4.2; 1,5=9.3; 2R,3R=4.0; 2R,3β=6.0; 5,6=6,7=
9.5; 8R,8β=12.1; 8β,7=12.1; 8β,9R=11.3; 8β,9β=5.0; 9R,8R=
5.2; 9R,9β=11.3; 11,13(a)=3.6; 11,130(a)=4.4; 13(a),130(a)=10;
1,14=1,140=1.3; 3,15=5,15=30,15=2.2; 20(a),30(a)=1.5; 30(a),
60(a) = 1.3; 20(a),60(a) = 1.3; 20(b),30(b) = 1.3; 30(b),60(b) = 1.8;
20(b),60(b)=1.3. 13C NMR (100 MHz): see Table 2. EIMS (70 eV)
m/z (rel. int.): 344 [M]+ (0.5), 328 [M-CH4]+ (23), 216 [M-CH4-
CH3O-C4O2H]+ (17), 159 [M-CH4-CH3O-C4O2H-CO2-
Molecular modeling of the transition state (B3LYP using the
basic set 6-31-G*) shows a difference in energy of -0.08444
hartree (-52.986 kcal/mol) between the C20S and C20R transition
state in favor of the C20S diastereoisomer in the case of 6. Such a
difference was not observed when the calculations were carried
out with the transition state that yields 7 (see Supporting
Information). These results are in good agreement with the
diastereomeric excess obtained in the case of 6. The spatial
requirements of the methyl group in the lactone ring D could
be responsible for the observed effect asit forces the orientation of
this ring in the transition state to minimize the steric hindrance.
The R orientation of this moiety in 6 makes the steric hindrance
higher in the transition state of the C20R diastereomer, while the
β orientation causes this moiety to lie away from the rest of the
molecule in 7. Consequently, the methyl group does not influence
the orientation, and the differences in steric hindrance between
the two possible diastereomers are minimized (a more detailed
discussion on the theoretical data is carried out in the Supporting
Information section).
Elimination of the hydroxyl group under strong basic condi-
tions can compete with the generation of the alkoxyde; in
compounds 2 and 3, the acidic proton H-11 easily undergoes
elimination toward 1, while the same reaction with the tertiary
hydroxyl group at C-11 in 4 will lead to the aldehyde 11. All
attempts to increase the yield of the alkoxyde varying the type and
amount of base used resulted in the increase of 1 and 11. Only
under the conditions described was it possible to obtain the
compounds resulting from the coupling of the alkoxyde and the
bromo-lactones 6-8 without detecting the elimination products
1 and 11.
CH3]+ (100); HRMS m/z 344.1606 (calcd for C20O5H24
-
344.1624).
(6R),(7S),(11R)-11-Hydroxy-13-[30-methyl-30-ene-40-oxy-
γ-butyrolactone]-4(15),10(14)-diene-guaiane-6R,12-olide (race-
mic mixture of 8a and 8b) was obtained as a colorless oil;
[R]25D -3.4 (c 0.04, CHCl3); IR (neat, KBr) νmax 3390 (OH, st),
2925, 2853 (C-H), 1789, 1770, (CdO, γ-lactone) cm-1. 1H NMR
(600 MHz): see Table 1. J (Hz): 1,2R=4.8; 1,2β=8.1; 1,5=16.0;
5,6=6,7=9.7; 8R,8β=16.1; 7,8R=4.5; 8R,9R=8R,9β=7.8; 13,
130(a)=9.5; 13,130(b)=9.7; 3,15(a,b)=5,15=2.1; 3,150(a)=5,150
(a)=2.3; 20,30(a)=20,60(a,b)=1.4; 30,60(a)=1.6; 30,60(b)=1.4. 13C
NMR (100 MHz): see Table 2. EIMS (70 eV) m/z (rel. int.):
360 [M]+ (0.5), 342 [M-H2O]+ (18), 263 [M-C5H5O2]+ (34),
159 (100); HRMS m/z 360.1543 (calcd for C20O6H24 360.1573).
(6R),(7S),(11R)-11-Hydroxy-13-benzoyloxy-4(15),10(14)-
diene-guaiane-6R,12-olide (9) was obtained as an amor-
Benzoylperoxide can compete with the nucleophilic substitu-
tion of the alkoxide if the reaction mixture is not carefully filtered.
If such a competitive process occurs, the compounds resulting
1
phous white solid. H NMR (600 MHz): see Table 1. J (Hz):

´
Macıas et al.
5858 J. Agric. Food Chem., Vol. 57, No. 13, 2009
1,2R=6.3; 1,2β=1.1; 2R, 2β=13.1; 2R,3R=8.5; 2R,3β=4.2; 5,6=
8.7; 6,7=10.0; 7,8R=3.1; 7,8β=13.1; 8R,8β=12.7; 8β,9R=12.7;
8β,9β=4.6; 8R,9R=9.8; 8R,9β=8.8; 13,130=11.5; 3,15=3,150=
1.8; 30,40 =8.2; 30,50 =1.5; 40,50 =7.9. 13C NMR (100 MHz): see
Table 2. EIMS (70 eV) m/z (rel. int.): 368 [M]+ (2.5), 263 [M-
PhCO]+ (2), 105 [PhCO]+ (100), 77 [C6H5]+ (34); HRMS m/z
368.1618 (calcd for C22O5H24 368.1624).
(6R),(7S),(11R)-11,13-Dibenzoyloxy-4(15),10(14)-diene-guaiane-
6R,12-olide (10) was obtained as an amorphous white solid.
1H NMR (600 MHz): see Table 1. J (Hz): 1,2R = 8.0; 1,2β =
4.0; 1,5=9.0; 2R, 2β=13.6; 2R,3R=2R,3β=6.8; 2β,3β=9.2; 5,6=
6,7=9.6; 7,8R=2.8; 7,8β=12.4; 8R,8β=12.4; 8β,9R=12.4; 8β,
9β=4.9; 8R,9R=11.2; 8R,9β=4.2; 13,130=11.0; 1,15=3,150=1.7;
1,15=3,150=1.6; 30,40=8.8; 30,50=1.6; 40,50=7.2; 300,400=8.4; 300,
500 =1.6; 400,500 =7.6. 13C NMR (100 MHz): see Table 2. EIMS
(70 eV) m/z (rel. int.): 472 [M]+ (0.5), 350 [M-PhCOOH]+ (4),
228 [M - 2PhCOOH]+ (17), 105 [PhCO]+ (100); HRMS m/z
472.1889 (calcd for C29O6H28 472.1886).
Spectroscopic Characterization: GELs. The oxidation step
was optimized with 13-hydroxy-epi-costuslactone (3) to yield
13-oxo-epi-costuslactone 11. With this aim in mind, several
oxidizing agents were tested, and periodinane gave the best yields
(57%, data not shown for other reagents). Compound 11 pre-
sented a molecular ion in the HRMS at m/z 246.12559, and this
corresponds to the molecular formula C15H18O3. The main
diagnostic changes that support oxidation to the aldehyde were
the shielding effect experienced by the signal of the H-13 protons
(δ 3.74 ppm, 1H, brd, H-130; δ 4.00 ppm, 1H, brd, H-13) and the
C-13 carbon (δ 67.8 ppm) in 3, which appeared now at δ 9.88 ppm
(1H, brs, H-13) and δ 195.2 ppm, respectively, in the new
compound (11).
formula C20O5H22 and the new signals in the 1H and 13C NMR
spectra corresponding to the H-20, H-30, and H-60 protons on the
lactone ring (Tables 1 and 2). Ring D is attached to the SL
backbone through the oxygen atom, as shown by the position
of the signals corresponding to proton H-20 (δ 6.08 ppm, dd,
J=1.3 Hz, J=1.5 Hz, 13; δ 6.09 ppm, dd, J=1.3 Hz, J=1.5 Hz,
14), which are similar to those of strigolactones (23, 40, 41) and
11,13-dihydro-GELs obtained previously; the shielding of the
H-13 proton from the typical aldehyde position (δ 9.88 ppm, 11)
to an enol position (δ 7.43 ppm in both 13 and 14); and the
g-HMBC connection established through the oxygen bridge
between the signals of H-20 and C-13.
Usually, the reaction between the enolate from a 1,3-dicarbo-
nyl system and an electrophile proceeds through the carbon of the
enolate. Examples of reaction through the oxygen are not
uncommon, but they are rare. Molecular semiempirical (AM1,
PM3) and ab initio (Hartree-Fock 3-21G(*)) calculations show
that there is a significant difference of about 0.1 e- between the
electronic density at the O-13 oxygen atom and carbon C-11 in
enolate 12 (Table 3). This difference appears in the two possible
stereochemistries (Z and E) of the enolate. Such a difference alone
would not explain why only the oxygen undergoes electrophilic
attack, but along withthe sterichindrance at C-11, this is expected
to be enough to direct the reaction pathway through the oxygen
and not through C-11.
Extensive NOE and NOESY studies on both diastereomers 13
and 14 did not provide any helpful positive NOE effect that could
serve to discriminate between a Z or E configuration at the Δ11-
13 double bond. However, the synthetic approach used here has
previously been applied to obtain strigolactones (36, 42) and its
synthetic analogues (43), and the E enolate was obtained in all
cases. The similarity between the ABC systems in guaianolides
and strigolactones, where both backbones adopt a similar spatial
arrangement, makes it reasonable to believe that the result should
be the same in GELs (Figure 3A). A similar arrangement
is obtained when the E enolate of aldehyde 11 is overlapped
with strigol. Consequently, using as a working hypothesis the
E stereoisomer at the C11-C13 double bond, an exhaustive study
of the theoretical distances between H-13a and H-8β in all
conformers in dehydrocostuslactone (1) and guaianestrigolac-
tones 13 and 14 was performed.
The results show that in the case of the double bond with a
Z disposition, the distance between H-13a and H-8β is almost the
same in the minimum energy conformer (Table 4, Figure 3B) for
all three compounds. Also, calculation of the Boltzmann dis-
tribution shows that the minimum energy conformer accounts for
more than 90% of the population in each of the three compounds
(Figure 4) and that in each local minima H-13,H-8 distances are in
the range of 2.1 to 3.2 A. The overlapping of the theoretical
minimum energy conformer in 1, (Δ11,13Z)-13 and (Δ11,13Z)-14
The incorporation of the second lactone ring is supported by
the HRMS spectrum, which presents a molecular ion at m/z
342.1483 (13) and m/z 342.1473 (14) according to the molecular
Table 3. Net Atomic Charges Calculated Using Semiempirical (PM3, AM1)
and Ab Initio (Hartree-Fock, Basis Set 3-21G(*)) Methods for the Two
Possible Stereochemical Configurations for Enolate 12a
Z enolate
Spartan
AM1 3-21G(*) PM3
E enolate
Spartan
PM3 AM1 3-21G(*)
CS Draw
PM3
CS Draw
PM3
O13 -0.6317 -0.493 -0.451 -0.689 -0.6848 -0.530 -0.487 -0.732
C11 -0.5537 -0.668 -0.560 -0.551 -0.6482 -0.658 -0.556 -0.534
C13 +0.2178 +0.381 +0.270 +0.403 +0.4265 +0.382 +0.266 +0.399
O12 -0.5283 -0.460 -0.404 -0.687 -0.6535 -0.490 -0.435 -0.717
O6 -0.3909 -0.307 -0.321 -0.654 -0.3019 -0.292 -0.310 -0.649
a Values were calculated using the MOPAC interface included in Chem3D Ultra
(v. 7.0) and Spartan’06.
Figure 3. (A) Overlapping of DHC (1) and strigol minimum energy conformers as obtained from PM3 calculations (Spartan’06). (B) Overlapping of theoretical
minimum energy conformers of 1, Δ11,13(Z )-13 and Δ11,13(Z )-14 calculated using the MMFF force field. The arrow depicts the NOE effect observed in
DHC (1).

Article
J. Agric. Food Chem., Vol. 57, No. 13, 2009 5859
also shows a similar spatial disposition for all of these species
(Figure 3B). Irradiation of the H-13a signal gave a strong
NOE effect between the signals corresponding to protons
H-13a and H-8R in the case of dehydrocostuslactone (1).
However, on irradiating either H-13 or H-8β signals no NOE
effect nor, in fact, a single correlation was detected in the
NOESY experiment in the case of guaianestrigolactones
13 and 14. Consequently, even though the absence of NOE
effects cannot usually be taken as evidence for the stereo-
chemical configuration, in this case the comparison between
the theoretical distances in DHC (1) and GELs 13 and
14 correlate well with an E stereoisomer of the C11-C13
double bond on the basis of the absence of any NOE effect.
Table 4. Theoretical Distances between H-13, H-8R, and H-8β as Calculated
from the Minimum Energy Conformers of Compounds 1, Δ11,13(Z) (13), and
Δ11,13(Z) (14) Obtained Using PM3 and MMFF Calculations (Spartan’06)
ꢀ
ꢀ
H13-H8R (A)
H13-H8β (A)
DHC (1)
(C20R), Δ11-13(Z) (13)
2.33
2.29
2.30
3.19
3.17
3.13
(C20S), Δ11-13(Z) (14)
Figure 4. ComparisonoftheH-13-H8R and H-13-H8β distancesthroughouttheconformerdistribution of1(A) andΔ11,13(Z ) (14) (B) ascalculatedusingthe
MMFF force field from a previously PM3 minimized structure. Key: 1, Boltzmann distribution (total population normalized to 1); 2, torsion and steric energy of
each conformer (kcal/mol); 3, distance between H13-H8R (A); 4, distance between H13-H8β (A).

´
Macıas et al.
5860 J. Agric. Food Chem., Vol. 57, No. 13, 2009
The third problem to be addressed with these compounds is the
issue of the configuration at C-20. Assignment of the absolute
configuration at C-20 is a crucial issue as it has been reported that
this chiral center greatly affects the germination activity on Striga
and Orobanche spp. (44, 45).
C-20 was obtained for each enantiomer (Table 5) as C-20R for
GEL 13 and C-20S for GEL 14.
(11E)-(20R)-13-[40-Methyl-30-ene-20-oxy-γ-butyrolactone]-
4(15),10(14),11(13)-triene-guaiane-6R,12-olide (13) was ob-
tained as a colorless oil; [R]25D +0.038 (c 0.04, CHCl3); IR
(neat, KBr) νmax 2927 (C-H), 1778, 1759 (CdO, γ-lactone),
1678 (CdC) cm-1. 1H NMR (600 MHz): see Table 1. J (Hz):
1,5=5,6=9.2; 6,7=8.5; 7,8R=3.3; 7,8β=12.0; 9R,9β=13.1;
8R,9β=3.1; 8R,9R=8β,9β=5.6; 8β,9R=9.7; 7,13=3.1; 3,
15=3,150 =1.8; 20,30 =1.5; 20,60 =1.3; 30,60 =1.8. 13C NMR
(100 MHz): see Table 2. EIMS (70 eV) m/z (rel. int.): 342
[M]+ (2); 245 [M-C5H5O2]+ (20); 97 [M-C15H17O3]+ (100);
HRMS m/z 342.1473 (calcd for C20O5H22 344.1479).
(11E)-(20S)-13-[40-Methyl-30-ene-20-oxy-γ-butyrolactone]-
4(15),10(14),11(13)-triene-guaiane-6R,12-olide (14) was ob-
Frischmuth et al. developed an empirical rule to assign the
absolute configuration of strigolactones based on the sign of
the circular dichroism (CD) spectra above λ=270 nm, which
corresponds to the change in the arrangement at the C-2 stereo-
center of Feringa’s lactone (46). When recorded alone, each of the
two possible C-20 enantiomers (17 and 18, Figure 5) gives opposite
CD spectra, and the sign above 270 nm correlates with the C-20S
(+) or C-20R (-) configuration at this center, this being the
empirical rule developed for these compounds. The basic assump-
tion is that the whole of the CD spectra of strigolactones can be
split in two as the sum of the ABC ring spectrum (15 and 16)
(which basically contributes to the spectra below λ=250 nm) and
that of Feringa’s lactone (17 and 18) (Figure 5) and that the two
chromophores (the enol-lactone system and the unsaturated
cyclobutanone ring) can be considered as independent. The case
of GELs can be considered to be essentially the same, as they
share with strigolactones the lactone-enol-γ-lactone moiety
(C and D rings). The CD spectra of 13 and 14 (Figure 6) are
similar to those of strigolactones (46) and show a change in the
sign of the CD above λ=270 nm. However, the CD spectrum of
11 shows that this part of the molecule contributes to the overall
spectrum mainly at wavelengths below 270 nm and that its
contribution beyond this value can be dismissed (Figure 6).
Consequently, we applied Frischmuth’s rule to GELs once it
had been proven that both skeletons behave in a similar manner
in terms of CD. The assignment of the absolute configuration at
tained as a colorless oil; [R]25 -0.037 (c 0.04, CHCl3); IR
D
(neat, KBr) νmax 2927 (C-H), 1783, 1751 (CdO, γ-lactone),
1680 (CdC) cm-1. 1H NMR (600 MHz): see Table 1. J (Hz):
1,5=9.0; 5,6=10.0; 6,7=8.7; 7,8R=3.6; 7,8β=11.8; 9R,9β=
13.1; 8R,9β=8β,9β=5.4; 7,13=3.1; 3,15=3,150=1.5; 20,30=
30,60=1.5; 20,60=1.3. 13C NMR (100 MHz): see Table 2. EIMS
(70 eV) m/z (rel. int.): 342 [M]+ (2); 245 [M-C5H5O2]+ (20);
97 [M - C15H17O3]+ (100); HRMS m/z 342.1483 (calcd for
C20O5H22 344.1479).
Bioassays. Compounds 1-4, 6-11, and 13-15 were tested as
germination inducers. Compound 15 is the parent 2-methylpent-
2-ene-lactone, that has been included in the bioassay to test the
lactone ring alone. The guaianolide backbone was selected as
a target based on our previous results, which showed that
eudesmanolidesand melampolidespresented a lowergermination
index (31). In this study, compounds were designed and synthe-
sized to check the following hypotheses: (i) the lactone ring itself is
sufficient to induce germination in O. cumana (sunflower broom-
rape); (ii) the lactone-enol-γ-lactone-like system is a more
general bioactiphore; it can be generally recognized by broom-
rapes other than sunflower-parasitizing species; and (iii) the
introduction of a second lactone ring can increase the levels of
low-germination activity compounds.
Compounds 1-4 and 6-10 were tested on two different
populations of O. cumana (IN12 and IN23) and one population
of O. ramosa, while compounds 11-14 were tested only on one
Figure 5. Feringa’s lactone and ABC system of strigolactones.
Figure 6. Comparison of CD spectra from GELs 13 (A) and 14 (B), and the aldehyde 11 (C).

Article
J. Agric. Food Chem., Vol. 57, No. 13, 2009 5861
Table 5. Assignment of the Stereochemical Configuration at C-20 in GELs
Based on the Change of the Sign in the CD Spectrum (46)
Sign of CD above λ = 270 nm
(260 for GELs)
compound
C-20 absolute configuration
Feringa’s lactone 17
Feringa’s lactone 18
strigol (19)
R
S
R
S
R
S
-
+
-
+
20-epi-strigol (20)
GEL 13
GEL 14
-
+
Figure 8. Effects of guaianolides 1-4, 9-11, γ-butirolactone 15, dihy-
droGELs6-8, andGELs13-14 onthegerminationof Orobancheramosa
seeds. GR24 was tested only at 10 μM as the internal standard for
bioassay reproducibility.
Table 6. EC50 Values Obtained for O. cumana and O. ramosa with
Guaianolides 1-4, 9-11, γ-Butirolactone 15, DihydroGELs 6-8, and GELs
13-14
O. cumana
O. ramosa
EC50 (μM)
r2
EC50 (μM)
r2
1
2
2.8
0.9905
0.9524
n. a.
n. a.
n. a.
n. a.
6.7
Figure 7. Effects of guaianolides 1-4, 9-11, γ-butirolactone 15, dihy-
droGELs 6-8, and GELs 13-14 on the germination of Orobanche
cumana (IN23) seeds. GR24 was tested at the same concentrations as
the internal standard for bioassay reproducibility.
3.04
n. a.
3
4
27.5
8.4
6
0.9893
0.8970
0.9995
0.9917
0.9927
0.9799
0.9727
7
0.003
10.4
7.9
8
13.7
population of O. cumana (IN23) and one of O. ramosa. Seeds of
O. cumana IN12 showed a greater dormancy than IN23 (data not
shown), and the percentages of germination obtained with GR24
and test compounds were lower. However, no differences were
found in the activity profiles of compounds tested between the
two populations. Consequently, the discussion will be focused on
the population with the higher germination percentage (IN23).
The results are shown in Figures 7 (O. cumana) and 8 (O. ramosa)
and the EC50 values in Table 6.
9
6.4
n.a.
28.2
3.7
n. a.
n. a.
n. a.
0.006
1.03
n. a.
10
11
13
14
15
0.9989
0.9927
0.9935
0.9987
0.8163
0.8536
7.4
52.8a
a Because of the low activity obtained with 15, a classical approximation to
estimate the EC50 value was used as follows: for 100% activity, a concentration of
1 M was supposed, where as a 10-12 M concentration was estimated for 0% activity.
In general, all compounds induced germination of the two
populations of O. cumana with two exceptions: the dibenzoyl
derivative 10 and, interestingly, the 13-hydroxy-epi-costuslactone
3. However, its epimer at C-11 (hydroxy-costuslactone 2) remains
active, thus illustrating the importance of the stereochemical
configuration. It should also be noted that the γ-butyrolactone
alone (15) was able to induce germination of O. cumana, albeit to a
much lower extent than the other compounds or GR-24.
Major differences arose for 6-8, 13, and 14 when tested on
O. ramosa. None of the SLs (1-4 and 9-11) and γ-butyrolactone 5
induced germination of tobacco broomrape, a finding that is not
unexpected; in contrast, the new GELs (13 and 14) and dihydro-
GELs (6-8) proved to be active. These results are in good
agreement with previous results, which showed the high selec-
tivity of SL toward sunflower broomrape (31) and confirmed
hypothesis ii, i.e., that the incorporation of a new lactone ring to
mimic the strigolactone lactone-enol-γ-lactone system caused a
loss of selectivity and made compounds active toward O. ramosa
when they previously did not show any activity (2-4 and 11).
Regarding the molecular modeling of the compounds tested,
the results for molecular properties are summarized in Table 7, in
which the main physicochemical features linked to the discussion
of the bioassays are presented.
One major observation is that 2, 4, 9, and 11 present significant
or good germination activities without possessing the typical
lactone-enol-γ-lactone-like system of strigolactones, an obser-
vation that is in good agreement with our previous data (30, 31).
Moreover, the γ-butyrolactone ring alone (5) is also able to
induce germination, albeit to a lower extent. These data confirm
that the lactone ring is sufficient in its own right to induce O.
cumana germination, according to hypothesis i, presented in the
previous section. However, comparison between dehydrocostus-
lactone (1) and the rest of the SLs (2-4 and 9-11) shows that 1
presents a higher level of activity. Consequently, even though
the lactone ring by itself seems to be sufficient to induce
germination in O. cumana, the presence of the much more reactive
R,β-unsaturated lactone ring enhances the activity. This finding
is consistent with a molecular mechanism in which a bionucleo-
phile reacts with the lactone ring. When the double bond is present,
the reaction turns to a Michael addition pathway (Figure 9A),
which is probably less demanding in terms of energy than the direct
1,2-addition to the carbonyl group in the lactone (Figure 9B).
Besides, the presence of the additional enol-γ-lactone system
seems to enhance activity, as shown by comparison of the pairs
2 and 7(a+b), 3 and 6(a+b), 4 and 8(a+b), and 1 with 13 and 14.

´
Macıas et al.
5862 J. Agric. Food Chem., Vol. 57, No. 13, 2009
Table 7. Theroetical Molecular Properties from Selected Strigolactones and GELs
molecular volume (A3)
compound
height (A)
E (HOMO) (eV)
E (LUMO) (eV)
strigol (19)
sorgolactone
341.93
316.70
325.69
348.77
250.43
261.20
268.94
102.94
257.40
376.21
482.28
351.6
5.98
8.94
4.06
7.15
3.90
5.00
5.18
1.80
5.03
5.61
15.36
7.21
5.98
5.17
7.21
7.21
-9.82
-9.40
-9.39
-9.81
-9.78
-9.86
-9.86
-10.44
-9.86
-9.82
-9.77
-9.80
-9.90
-9.90
-9.73
-9.71
-0.88
-1.01
-0.95
-1.06
-0.15
+0.71
solanacol
sorgomol
1 (DHC)
2 (13-hydroxy)
4 (11,13-dihydroxy)
5 (γ-butyrolactone)
11 (aldehyde)
9 (benzoyloxi)
10 (dibenzoyloxi)
dihydroGEL 6 (20 R)
dihydroGEL 7 (20 R)
dihydroGEL 8 (20 R)
GEL 13 (20 R)
GEL 14 (20 S)
+0.71
-0.41
+0.32
-0.45
-0.51
-0.51
-0.38
-0.52
-0.91
-0.88
352.4
359.1
347.8
347.9
compound (9) and the lowest volume of an inactive compound
(10). However, this assumption proved to be questionable in view
of data obtained later with another set of compounds (data not
shown) in which GELs with volumes greater than that of 10
showed reasonable to excellent levels of activity. The lack of
activity seems to be related with the spatial requirements of the
11,13-dibenzoyl system rather than with the overall volume of the
molecule. The minimum energy conformer adopted by 10 to
minimize the steric hindrance caused by the two bulky aromatic
rings (Figure 9) gives a distance between the two extremes (that
could be considered as the height of the molecule) of 15.36 A, a
value far higher than those characteristic of strigol (5.98 A),
dihydroGEL 8 (7.2 A), or GEL 14 (7.21 A; total height 8.4 A). As
a consequence, it seems that the lack of activity of 10 must be due
to steric reasons rather than with the functional groups present in
ring C, in this particular case (Figure 10), as it will not fit the
spatial requirements of the receptor cavity.
Another point worth highlighting concerns 3. Previous
results with populations IN7 and IN21 of O. cumana (31)
showed this compound to be extremely active. However, it
turned now to be inactive with populations IN12 and IN23.
The bioassay was repeated three times in order to ensure the
accuracy of the results. This fact might be understood as a
signal of the high genetic variability of Orobanche that might
change specific recognition requirements from one popula-
tion to another, making it sometimes difficult to correlate
results from different populations.
Finally, the third key point in discussion addresses the question
of the bioactiphore and whether the lactone-enol-γ-lactone
system of strigolactones is a general recognition system for all
Orobanche species or whether there are other groups that can be
also recognized. Comparison of germination activities of O.
cumana and O. ramosa clearly show that SLs are specific inducers
of O. cumana germination. However, when a second lactone ring
is added to the guaianolide skeleton in a strigolactone fashion
(rings C and D, GELs), the selectivity of SL toward O. cumana
disappears, and the resulting compounds (6-8, 13, 14) now
induce germination in other Orobanche species, tobacco broom-
rape in this case. Consequently, it is reasonable to propose that
the lactone ring is the specific bioactiphore for O. cumana, and a
more general system that can be recognized by all Orobanchaceae
corresponds to the lactone-enol-γ-lactone system. As a conse-
quence, the proposed hypothesis that SLs are the specific sun-
flower allelochemicals that trigger germination in O. cumana is
strongly supported by these data. SLs are common constituents
of sunflowers, and if exuded from the roots of sunflowers, they
Figure 9. Possible molecular mechanisms of action for SLs under a
variety of chemical environments in the lactone ring.
The germination percentages at 10 μM increase in GELs, which
also induce germination at the subsequent dilutions. A clear
explanation for this behavior is not obvious, but it could be due to
the increased number of possible bioactiphores in these molecules
(two lactone rings instead of just a single one). However, further
bioassays to compare these types of pairs (lactones vs lactone-
enol-γ-lactones) are necessary to confirm this hypothesis.
An important point to discuss is whether the presence of the
lactone system is sufficient by itself to induce germination. It
seems that other factors such as the spatial volume must also be
considered. In this sense, a dramatic reduction in activity can be
observed when 9 and 10 are compared. Compound 9 presents
activity similar to that of its parent 4. However, 10 is inactive. The
main differences between the two compounds are their molecular
volumes. We previously reported (31) that natural strigolactones
present theoretical volumes between 290.5 and 435.6 A3. We
calculated several theoretical molecular descriptors for this new
set of compounds and found that benzoyl derivative 9 has a
molecular volume of 449.5 A3, while dibenzoyl derivative 10 has a
volume of 566.7 A3. GELs also present theoretical molecular
volumes within the strigolactone range (Table 7). Initially, the
lack of activity of 10 can be related to a molecular volume that is
too highand should not fit into the receptorcavity. Consequently,
a tentative volume for the bioactiphore could be proposed in the
range 450-560 A3, on the basis of the largest volume of an active

Article
J. Agric. Food Chem., Vol. 57, No. 13, 2009 5863
reactions such as those depicted in entry c are possible. Thus, the
previously reported order of reactivity parthenolide > reduced
parthenolide (31) is explained.
The accepted ecological role of SLs in sunflowers (and in other
plant species where they are present) is chemical defense against
insects and aphids, fungi, and competing plants. In order to act as
chemical clues for parasitic weeds, SLs should be exuded through
the roots or leached from the leaves by rainintothe soil. However,
such studies have not been carried out with sunflower exudates to
date nor has the presence of SLs outside the plant been confirmed.
In the absence of such data, our preliminary results strongly
support the hypothesis that SLs exuded through the roots of
sunflowers are the specific inducers of the germination of sun-
flower broomrape. Broomrape might coevolve with sunflowers in
order to take advantage of this defense mechanism, changing it
into a clear signal that a host is available in the vicinity of the seed.
Supporting Information Available: Discussion on the stereo-
selectivity of the SN2 displacement of bromine to yield 6a+6b,
7a+7b, and 8a+8b based on theoretical calculations, molecular
mechanics, semi-empirical (AM1 and PM3), ab initio (Hartree-
Fock), and denstity functional (B3LYP); molecular calculations
supporting the stereochemical configuration proposed at the
Figure 10. Distance between the bottom and roof of (A) 11,13-dibenz-
oyloxy-DHC (10); (B) strigol; (C) 11-hydroxy-13-benzoyloxy-DHC (9);
(D) 20S-dihydroGEL (7); and (E) 20S-GEL (14) as obtained from the
minimum energy conformer calculated using semiempirical PM3 calcula-
tions (Spartan’06).
1
double bond C11-C13 in GELs 13 and 14; and H and 13C
NMR spectra of all new compounds6a+6b, 7a+7b, and 8a+8b,
9-14. This material is available free of charge via the Internet at
http://pubs.acs.org.
could be the specific host signals that induce germination of
sunflower broomrape seeds.
LITERATURE CITED
(1) Joel, D. M.; Hershenhorn, Y.; Eizenberg, H.; Aly, R.; Ejeta, G.;
Rich, P. J.; Ransom, J. K.; Sauerborn, J.; Rubiales, D. Biology and
management of weedy root parasites. Hort. Rev. 2007, 33, 267–349.
(2) Cubero, J. I. Parasitic Diseases in Vicia faba L. with Special
Reference to Broomrape (Orobanche crenata Forsk.). In The Faba
Bean; Hebblethwaite, P. D., Ed.; Butterworths: Oxford, U.K., 1982;
pp 496-521.
Regarding the molecular recognition mechanism, a nucleophilic
Michael addition of the nucleophile present in the receptor’s cavity
followed by a retro-Michael reaction has been proposed (47). In the
case of SLs, several mechanisms could be proposed to explain
activity, even when the exocyclic double bond conjugated with the
lactone system is not present (Figure 9) (48).
(3) Nandula, V. K.; Westwood, J. H.; Foster, J. G.; Foy, C. L. Influence
of glyphosate on amino acid composition of Egyptian broomrape
[Orobanche aegyptiaca (Pers.)] and selected hosts. J. Agric. Food
Chem. 2001, 49, 1524–1528.
Entry a represents the most general situation. The bionucleo-
phile reacts in a Michael fashion and becomes bound irreversibly
to the SL. The end product is similar to that proposed by
Zwanenburg for strigolactones. Entry b represents those cases
where the exocyclic double bond is not present, and a Michael
addition of the bionucleophile is not possible. In these cases, the
lactone ring itself may undergo a nucleophilic addition that leads
to the opening of the lactone ring. The steric constraints are
higher, and the reactivity of the lactone ring is lower in this case,
thus explaining the lower activities obtained with the reduced SLs
in comparison withtheir unsaturatedanalogues. Accordingto this
model, the nucleophilic addition involves the reaction of the
HOMO from the bionucleophile with the LUMO of the guaia-
nolide, strigolactones, or GEL. When the double bond is not
present in the lactone ring (2-4 and 11), the energy of the LUMO
rises up as a result of the lower reactivity of these compounds
(Table 7, compounds 2, 4, and 11), which can be also correlated
with their lower germination indexes. This is not the case of 9,
which lacks the double bond at the lactone ring but presents a high
germination index and energy of the LUMO similar to dihydro-
GELs (6-9). The presence of the aromatic benzoyl group lowers
the energy LUMO and may change the reactivity of the molecule
as it is a good leaving group that can undergo SN2 displacements.
This fact, plus their location in the reactive bioactiphore zone of
the SL might serve to explain its high bioactivity.
ꢀ
~
~
(4) Castejon-Munoz, M.; Romero-Munoz, F.; Garcıa-Torres, L. Con-
´
trol of broomrape (Orobanche cernua) in sunflower (Helianthus
annuus L.) with glyphosate. Crop Prot. 1990, 9, 332–336.
(5) Qasem, J. R. Chemical control of branched broomrape (O. ramosa)
in glasshouse grown tomato. Crop Prot. 1998, 17, 625–630.
(6) Eizenberg, H.; Hershenhorn, J.; Graph, S.; Manor, H. Orobanche
aegyptiaca control in tomato with sulfonylurea herbicides. Acta
Hort. 2003, 613, 205-208.
(7) Schnell, H.; Kunisch, M.; Saxena, M. C.; Sauerborn, J. Simulation of
the seed bank dynamics of Orobanche crenata Forsk. in some crop
rotations common in Northern Syria. Exp. Agric. 1996, 32, 395–403.
ꢀ
(8) Fernandez-Aparicio, M.; Sillero, J. C.; Rubiales, D. Intercropping
with cereals reduces infection by Orobanche crenata in legumes. Crop
Prot. 2007, 26, 1166–1172.
(9) Jain, R.; Foy, C. L. Nutrient effect on parasitism and germination of
egyptian broomrape (Orobanche aegyptiaca). Weed Technol. 1992, 6,
269–275.
(10) Roman, B.; Torres, A. M.; Rubiales, D.; Cubero, J. I.; Satovic, Z.
Mapping of quantitative trait loci controlling broomrape
(Orobanche crenata Forsk.) resistance in faba bean (Vicia faba L.).
Genome 2002, 45, 1057–1063.
ꢀ
(11) Perez-de-Luque, A.; Jorrin, J. V.; Rubiales, D. Crenate broomrape
control in pea by foliar application of benzothiadiazole (BTH).
Phytoparasitica 2004, 32, 21–29.
Entry c represents a third possibility in which other nucleo-
philic reaction centers are present in the molecule. Germacranes
readily react under acidic conditions to give rise to eudesmano-
lides (49, 50). Physiological pH values are mildly acidic, and
~ ꢀ
(12) Echevarrıa-Zomeno, S.; Perez-de-Luque, A.; Jorrın, J.; Maldonado,
´ ´
A. M. Pre-haustorial resistance to broomrape (Orobanche cumana)
in sunflower (Helianthus annuus): cytochemical studies. J. Exp. Bot.
2006, 57, 4189–4200.

´
Macıas et al.
5864 J. Agric. Food Chem., Vol. 57, No. 13, 2009
(13) Perez-Vich, B.; Akhtouch, B.; Knapp, S. J.; Leon, A. J.; Velasco, L.;
Fernandez-Martinez, J. M.; Berry, S. T. Quantitative trait loci for
broomrape (Orobanche cumana Wallr.) resistance in sunflower.
Theor. Appl. Genet. 2004, 109, 92–102.
(33) Da Silva, A. J. R.; Garcı
´
a, M.; Baker, P. M.; Rabi, J. A. Carbon-13
NMR spectra of natural products. 1. Guaianolides. Org. Magn.
Reson. 1981, 16, 230–233.
(34) Humprehy, A. J.; Gaslter, A. M.; Beale, M. H. Strigolactones in
chemical ecology: waste products or vital allelochemicals? Nat. Prod.
Rep. 2006, 23, 592–614.
(14) Rubiales, D. Parasitic plants, wild relatives and the nature of
resistance. New Phytol. 2003, 160, 459–461.
ꢀ
(35) Macıas, F. A.; Garcıa-Dıaz, M. D.; Massanet, G. M.; Gomez-
´ ´ ´
(15) Johnson, A. W.; Gowda, G.; Hassanali, A.; Knox, J.; Monaco, S.;
Razavi, Z.; Rosebery, G. The preparation of synthetic analogs of
strigol. J. Chem. Soc., Perkin Trans.1 1981, 6, 1634–1743.
Madero, J. F.; Fronczek, F.; Galindo, J. C. G. An easy access to
bioactive 13-hydroxylated and 11,13-dihydroxylated sesquiterpene
lactones (SL) through Michael addition of a nucleophilic hydroxyl
group. Tetrahedron 2008, 64, 10996–11006.
(16) Galindo, J. C. G.; Macı
Chemistry of Host-Parasite Interactions. In Allelopathy. Chemistry
and Mode of Action of Allelochemicals; Macıas, F. A., Galindo, J. C.
as, F. A.; Garcıa-Dıaz, M. D.; Jorrın, J.
´ ´ ´ ´
(36) McAlpine, G. A.; Raphael, R. S.; Shaw, A.; Taylor, A. W.; Wild, H.
J. Synthesis of the germination stimulant (()-strigol. J. Chem. Soc.,
Perkin Trans. 1 1976, 410–416.
´
G., Molinillo, J. M. G., Cutler, H. G., Eds.; CRC Press: Boca Raton,
FL, 2004; pp 125-148.
(17) Rani, K.; Zwanenburg, B.; Sugimoto, Y.; Yoneyama, K.; Bouw-
meester, H. J. Biosynthetic considerations could assist the structure
elucidation of host plant produced rhizosphere signaling compounds
(strigolactones) for arbuscular mycorrhizal fungi and parasitic
plants. Plant Physiol. Biochem. 2008, 46, 617–626.
ꢀ
~
(37) Serghini, K.; Perez , A.; Castejon-Munoz, M.; Garcı
´
a-Torres, L.;
Jorrın, J. Sunflower response to broomrape parasitism: induced
´
synthesis and excretion of 7-hydroxylated simple coumarins. J. Exp.
Bot. 2001, 52, 2227–2234.
(38) Spartan’06; Wavefunction Inc.: Irvine, CA.
(39) Stewart, J. J. P. MOPAC2007, version 7.276w; Stewart Computa-
tional Chemistry: Colorado Springs, CO; http://OpenMOPAC.net.
(18) Siame, B. A.; Weerasuriya, Y.; Wood, K.; Ejeta, G.; Butler, L. G.
Isolation of strigol, a germination stimulant for Striga asiatica, from
(accessed Nov, 2008).
(40) Hauck, C.; Muller, S.; Schildknecht, H. A germination stimulant for
host plants. J. Agric. Food Chem. 1993, 41, 1486–1491.
::
::
(19) Hauck, C.; Muller, S.; Schildknecht, H. A germination stimulant for
parasitic flowering plants from Sorghum bicolor, a genuine host
plant. J. Plant Physiol. 1992, 139, 474–478.
parasitic flowering plants from Sorghum bicolor, a genuine host
plant. J. Plant Physiol. 1992, 139, 474–478.
(41) Yokota, T.; Sakai, H.; Okuno, K.; Yoneyama, K.; Takeuchi, Y.
Alectrol and orobanchol, germination stimulants for Orobanche
minor, from its host red clover. Phytochemistry 1998, 49, 1967–1973.
(42) Mori, K.; Matsui, J.; Bando, M.; Kido, M.; Takeuchi, Y. Synthesis
and biological evaluation of the four racemic stereoisomers of the
structure proposed for sorgolactone, the germination stimulant of
Sorghum bicolor. Tetrahedron Lett. 1997, 38, 2507–2510.
(43) Mangnus, E. M.; Jan Dommerholt, F.; de Jong, R. L. P.; Zwanenburg,
B. Improved synthesis of strigol analogue GR24 and evaluation of the
biological activity of its diastereomers. J. Agric. Food Chem. 1992, 40,
1230–1235.
(20) Yoneyama, K.; Sato, D.; Takeuchi, Y.; Sekimoto, H.; Yokota, T.
Search for Germination Stimulants and Inhibitors for Root Parasitic
Plants. Abstracts of Papers, 227th National Meeting of the American
Chemical Society, Anaheim, CA; American Chemical Society:
Washington, DC, 2004.
(21) Yokota, T.; Sakai, H.; Okuno, K.; Yoneyama, K.; Takeuchi, Y.
Alectrol and Orobanchol, germination stimulants for Orobanche
minor, from its host red clover. Phytochemistry 1998, 49, 1967–1973.
::
(22) Muller, S.; Hauck, C.; Schildknecht, H. Germination stimulants
produced by Vigna unguiculata Walp cv. Saunders. J. Plant Growth
Reg. 1992, 11, 77–84.
(44) Bergmann, C.; Wegmann, K.; Frischmuth, K.; Samson, E.; Kranz,
A.; Weigelt, D.; Koll, P.; Welzel, P. Stimulation of Orobanche
crenata seed germination by (+)-strigol and structural analogues.
Dependence on constitution and configuration of the germination
stimulants. J. Plant Physiol. 1993, 142, 338–342.
(45) Mangnus, E. M.; Zwanenburg, B. Synthesis, structural character-
ization, and biological evaluation of all four enantiomers of strigol
analogue GR7. J. Agric. Food Chem. 1992, 40, 697–700.
(46) Frischmuth, K.; Wagner, U.; Samson, E.; Weigelt, D.; Koll, P.;
Meuer, H.; Sheldrick, W. S.; Welzel, P. Configurational assignment
at C-20 of some strigol analogues. Tetrahedron: Asymmetry 1993, 4,
351–360.
(23) Cook, C. E.; Whichard, L. P.; Thurner, B.; Wall, M. E.; Egley, G. H.
Germination of witchweed (Striga lutea Lour.): isolation and prop-
erties of a potent stimulant. Science 1966, 154, 1189–1190.
(24) Yasuda, N.; Sugimoto, Y.; Kato, M.; Inanaga, S.; Yoneyama, K.
(+)-Strigol, a witchweed germination stimulant from Menispermum
dauricum root culture. Phytochemistry 2003, 62, 1115–1119.
(25) Akiyama, K.; Matsuzaki, K.-I.; Hayashi, H. Plant sesquiterpenes
induce hyphal branching in arbuscular mycorrhizal fungi. Nature
2005, 435, 824–827.
(26) Akiyama, K.; Hayashi, H. Strigolactones: chemical signals for
fungal symbionts and parasitic weeds in plant roots. Ann. Bot.
2006, 97, 925–931.
(27) Joel, D. M.; Steffens, J. C.; Matthews, D. E. Germination of Weedy
Root Parasites. In Seed Development and Germination; Kigel, J.,
Galili, G., Eds.; Marcel Dekker: New York, 1995; pp 567-597.
(28) Bar Nun, N.; Mayer, A. M. Preconditioning and germination of
Orobanche seeds: respiration and protein synthesis. Phytochemistry
1993, 34, 39–45.
(47) Mangnus, E. M.; Zwanenburg, B. Tentative molecular mechanism of
the germination stimulation of Striga and Orobanche seeds by strigol
and its synthetic analogues. J. Agric. Food Chem. 1992, 40, 1066–
1070.
ꢀ
(48) Galindo, J. C. G.; Hernandez, A.; Dayan, F. E.; Tellez, M. R.;
Macıas, F. A.; Paul, R. N.; Duke, S. O. Dehydrozaluzanin C, a
´
natural sesquiterpenolide, causes rapid plasma membrane leakage.
Phytochemistry 1999, 52, 805–813.
(49) Jain, T. C.; McCloskey, J. E. Carbocyclization in natural products. I.
Amberlite IR-210 catalyzed cyclization of costunolide. Tetrahedron
1975, 31, 2211–2214.
(50) Azarken, R.; Guerra, F. M.; Moreno-Dorado, F. J.; Jorge, Z. D.;
Massanet, G. M. Substituent effects in the transannular cyclization
of germacranes. Synthesis of 6-epi-costunolide and five steiractino-
lides. Tetrahedron 2008, 64, 10896–10905.
(29) Bar Nun, N.; Plakhine, D.; Joel, D. M.; Mayer, A. M. Changes in the
activity of the alternative oxidase in Orobanche seeds during con-
ditioning and their possible physiological function. Phytochemistry
2003, 64, 235–241.
ꢀ
(30) Perez de Luque, A.; Galindo, J. C. G.; Macıas, F. A.; Jorrın, J.
´ ´
Sunflower sesquiterpene lactone models induce Orobanche cumana
seed germination. Phytochemistry 2000, 53, 45–50.
´ ´
(31) Galindo, J. C. G.; de Luque, A. P.; Jorrın, J.; Macıas, F. A. SAR
studies of sesquiterpene lactones as Orobanche cumana seed germi-
nation stimulants. J. Agric. Food Chem. 2002, 50, 1911–1917.
Received March 13, 2009. Revised manuscript received May 13, 2009.
Accepted May 13, 2009. This research was supported by Ministerio de
Ciencia y Tecnologıa, Spain (MCYT; Project No. AGL2005-05190).
(32) Mathur, S. B.; Hiremath, S. V.; Kulkarni, G. H.; Kelkar, G. R.;
Bhattacharyya, S. C.; Simonovic, D.; Rao, A. S. Terpenoids. LXX.
Structure of dehydrocostus lactone. Tetrahedron 1965, 21, 3575–3590.