Fungicidal activity of natural and synthetic sesquiterpene lactone analogs

Article abstract of DOI:10.1016/S0031-9422(00)00008-X

Fungicidal activity of 36 natural and synthetic sesquiterpene lactones with guaianolide, trans, trans-germacranolide, cis, cis-germacranolide, melampolide, and eudesmanolide carbon skeletons was evaluated against the phytopathogenic fungi Colletotrichum acutatum, C. fragariae, C. gloeosporioides, Fusarium oxysporum, Botrytis cinerea, and Phomopsis sp. Dose-response data for the active compounds dehydrozaluzanin C, dehydrocostuslactone, 5α-hydroxydehydrocostuslacone, costunolide, and zaluzanin C are presented. A new 96-well microbioassay procedure for fast and easy evaluation of antifungal activity was used to compare these compounds with commercial fungicide standards. Some structure-activity conclusions are also presented. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0031-9422(00)00008-X

Phytochemistry 53 (2000) 747±757
www.elsevier.com/locate/phytochem
Fungicidal activity of natural and synthetic sesquiterpene lactone
analogs
D.E. Wedgea,*, J.C.G. Galindob, F.A. Macõasb
aUnited States Department of Agriculture, Agricultural Research Service, Natural Products Utilization Research Unit, P.O. Box 8048, University,
MS 38677, USA
bDepartamento de QuõÂmica OrgaÂnica, Facultad de Ciencias, Universidad de CaÂdiz, c/ RepuÂblica Saharahui s/n, Apdo. 40, 11520 Puerto Real, CaÂdiz,
Spain
Received 10 August 1999; received in revised form 5 November 1999
Abstract
Fungicidal activity of 36 natural and synthetic sesquiterpene lactones with guaianolide, trans, trans-germacranolide, cis, cis-
germacranolide, melampolide, and eudesmanolide carbon skeletons was evaluated against the phytopathogenic fungi
Colletotrichum acutatum, C. fragariae, C. gloeosporioides, Fusarium oxysporum, Botrytis cinerea, and Phomopsis sp. Dose-
response data for the active compounds dehydrozaluzanin C, dehydrocostuslactone, 5a-hydroxydehydrocostuslacone,
costunolide, and zaluzanin C are presented. A new 96-well microbioassay procedure for fast and easy evaluation of antifungal
activity was used to compare these compounds with commercial fungicide standards. Some structure-activity conclusions are
also presented. 7 2000 Published by Elsevier Science Ltd. All rights reserved.
Keywords: Fungicides; Dehydrozaluzanin C; Zaluzanin C; Dehydrocostuslactone; Costunolide; 5a-Hydroxydehydrocostuslactone; Sesquiterpene
lactones; Colletotrichum acutatum; Colletotrichum fragariae; Colletotrichum gloeosporioides; Fusarium oxysporum; Botrytis cinerea; Phomopsis sp
1. Introduction
agrochemicals for disease control. This study evaluates
several sesquiterpene lactones for their potential use as
disease control agents for pathogenic fungi.
Owing to the continuing development of microbial
resistance in medicine and agriculture, discovery of
new antimicrobial substances is an important, if not
urgent, research objective. In addition, the desire for
safer agrochemicals with less environmental and mam-
malian toxicity is a major concern. Particularly desir-
able is the discovery of novel prototype antimicrobial
agents representing new chemical classes that operate
by di€erent modes of action than existing antifungal
agents and, consequently, lack cross-resistance to
chemicals currently used (McChesney, 1993; Kirst et
al., 1992). Following natural product leads o€ers an
ecient approach to discovering and optimizing new
Filamentous fungi of the genera Colletotrichum,
Botrytis, Fusarium and Phomopsis species, all con-
sidered major plant pathogens worldwide (Farr, Bills,
Chamuris & Rossman, 1989) were selected as test
organisms. Failure to control these fungi can result in
serious economic losses to both US and worldwide
agriculture. Anthracnose (caused by Colletotrichum
sp.), Phomposis and Botrytis diseases are serious pro-
blems for strawberry fruit and plant production in
many areas of the world but are especially serious in
the southeastern US (Maas, 1998). New shortened
crop production cycles and increased growing pressure
to meet the rapidly expanding mass market of orchids
has facilitated Fusarium oxysporum in becoming a
serious fungal pathogen often uncontrolled by com-
mercial fungicides.
* Corresponding author. Tel.: +1-601-232-1134; fax: +1-601-232-
1035.
E-mail address: dwedge@olemiss.edu (D.E. Wedge).
Sesquiterpene lactones constitute a large family of
0031-9422/00/$ - see front matter 7 2000 Published by Elsevier Science Ltd. All rights reserved.
PII: S0031-9422(00)00008-X

748
D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
more than 5000 compounds (Fraga, 1996, 1997, 1998)
mainly isolated from members of the Compositae. Ses-
quiterpenes possess a wide spectrum of biological ac-
those reported for the natural guaianolide annuolide A
(Macõas, Varela, Torres & Molinillo, 1993). The main
di€erences were observed in the position of the H-3 (d
5.99, 1H, br s) and H-15 (d 4.28, 1H, br s, H-15; d
4.36, 1H, br s, H-15') signals, which were shifted
up®eld with respect to annuolide A. Analysis of 2D
1H±1H COSY and 13C-NMR spectroscopic data con-
®rmed the proposed structure. Substitution of primary
and secondary allylic hydroxyl groups by chloride ion
in reactions with chloride derivatives possessing good
leaving groups have been reported (Collington &
Meyers, 1971). Since both the C-3 epimers and the re-
arrangement product were obtained, the mechanism
must have a strong SN1 component, since the for-
mation of the intermediate carbocation is favored by
the use of aprotic and polar solvents such as pyridine.
Compound 2 was protected as the tetrahydropyranyl
ether 9 (99%) by stirring a dry THF solution of 2 with
an excess of dihydropyran (DHP) and a catalytic
amount of p-toluene sulfonic acid. The structure of 9
was con®rmed by its EIMS spectrum (m/z 330), con-
sistent with the molecular formula C20H26O4 in ad-
dition to analysis of its and 1HNMR (Table 1) and
13C-NMR (Table 2) spectral data.
tivity (Fischer, 1991; Robles, Aregullin, West
&
Rodrõguez, 1995; Marles, Pazos-Sanou, Compadre,
Pezzuto, Bloszyk & Arnason, 1995) in which they
appear to play a role in plant defense mechanisms.
Due to their toxicity, some sesquiterpene lactones with
eudesmanolide (Sabanero, Quijano, Rõos & Trejo,
1995), pseudoguaianolide, and trans,trans-germacrano-
lide (Picman, 1984) skeletons have been evaluated for
fungicidal activity. Sesquiterpene lactones have not
been reported as phytoalexins or other induced anti-
fungal metabolites. This paper describes the antifungal
characteristics of 36 natural and semi-synthetic sesqui-
terpene lactones evaluated by direct TLC-bioautogra-
phy and 96-well microtiter plate assays. Compounds
tested belong to guaianolide (1±17), trans,trans-germa-
cranolide (18±23), melampolide (24±27), cis,cis-germa-
cranolide (28), or eudesmanolide (29±36) classes.
2. Results and discussion
The three chloro derivatives 7 (42%), 8 (20%) and
16 (5%) were prepared by treatment of a dry pyridine
solution of 2 with freshly distilled mesyl chloride in
dry pyridine at 08C. Compound 7 shows an EIMS
with a molecular ion at m/z 264:266, in accordance
with the molecular formula C15H19O2Cl. The IR spec-
Acetonide derivative 11 was prepared by re¯uxing
an acetone solution of 6 for 24 h with a catalytic
amount of p-toluene sulfonic acid. The structure was
con®rmed by its EIMS spectrum (molecular ion at m/z
318, according to the molecular formula C18H22O5)
and 1H- (Table 1) and 13C-NMR (Table 2) spectral
data: d 1.49 (3H, s, Me); d 41.54 (3H, m, Me); d 113.2
(s, C-16); d 28.6 (q, C-17); d 28.3 (q, C-18).
1
trum shows an absorption band at 813 cm for the
1
(C±Cl) functional group. The H-NMR spectrum was
also very close to that of the starting product 2
(Macõas, Galindo & Massanet, 1992). Major di€er-
ences were observed for signals corresponding to the
lactonic proton and the exomethylene at C4±C15. The
Oxidation of 12 with SeO2 and tert-butyl hydroper-
oxide (Macõas et al., 1992) a€orded 14 (62%) as the
major product. Its EIMS gave a molecular ion at m/z
264, consistent with the molecular formula C15H18O4.
Comparison of the 1H- and 13C NMR spectroscopic
data of 12 and 14 showed a new signal in the 1H-
NMR spectrum at d 4.63 (1H, dd, J2a,3b ˆ 6:8 Hz,
J2b,3b ˆ 6:9 Hz) for H-3b and the anticipated deshield-
ing of C-3 (12: d 30.6, 14; d 74.4). An a-orientation
was proposed for the C-3 hydroxyl group on the basis
of previous observations with similar compounds
(Macõas et al., 1992; Kalsi, Kaur, Sharma & Talwar,
1984).
lactonic proton resonated at
J5,6 ˆ J6,7 ˆ 9:5 Hz, H-6), down®eld in comparison
with its equivalent in (d 3.83, 1H, dd,
d 4.11 (1H, dd,
2
J5,6 ˆ J6,7 ˆ 9:5 Hz, H-6); the exomethylene protons
appeared at d 5.56 (1H, dd, J3a,15 ˆ J5,15 ˆ 2:1 Hz, H-
0
0
15) and d 5.50 (1H, dd, J3a,15 ˆ J5,15 ˆ 2:1 Hz, H-15').
Both signals were shifted up®eld due to the introduc-
tion of the chlorine atom with a b orientation. A 2D
COSY experiment con®rmed the structure of 7.
Compound 8 had the same IR and EIMS spectra as
7. Major di€erences were observed in the 1H-NMR
spectrum, where the signal corresponding to H-3b in 8
appeared down®eld (d 4.89, 1H, dd, J2a,3b ˆ 6:8 Hz,
J2b,3b ˆ 2:5 Hz); the signal of H-6 was shifted to a
higher ®eld (d 3.93, 1H, dd, J5,6 ˆ J6,7 ˆ 9:2 Hz)
because of the a-oriented chlorine atom.
Compound 16 exhibited an EIMS spectrum close to
those of 7 and 8, showing a molecular ion at m/z
264:266 suggesting the molecular formula C15H19O2Cl.
The 1H-NMR spectroscopic data were very close to
Reduction of 1 with diisobutylaluminum hydride
(DIBAL) in dry toluene ( 708C, N2 atmosphere)
yielded the corresponding lactol characterized by the
lack of the g-lactone absorption band and the presence
of a strong ±OH absorption band at 3396 cm in the
1
IR spectrum. Notable changes in the H-NMR spec-
trum are the lactol proton at d 5.71 (1H, br d,
J12,13a ˆ 1:8 Hz, H-12), and the up®eld signals corre-
sponding to the lactonic exomethylene protons (d 5.26,
1H, dd, J7,13a ˆ 2:7 Hz, J12,13a ˆ 1:8 Hz, H-13a; d 5.05
1H, dd, J7,13b ˆ 3:1 Hz, J12,13b ˆ 1:3 Hz, H-13b).
1

D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
749
Table 1
1H-NMR spectral data for compounds 7±9, 11, 16±17, and 26 (399.95 MHz, CDCl3)a
H
7
8
9
11
14
16
17
26
1
2
2.96 dd
3.27
3.07 m
2.17 m
1.83 m
b4.68 dd
±
3.01
3.26 dd
a2.52 m
b2.52 m
b5.99 s
2.90
5.59
a2.15 ddd
b2.47 ddd
a4.71 ddd
a2.38
b2.21
b4.89
a2.65 dd
b2.06 dd
b4.51
a2.15 ddd
b1.83 ddd
b4.63
a1.72 ddd
b2.05 m
a1.96 m
b2.42 m
2.73
a1.99b
b2.15b
a1.99
b2.15
5.00 d
4.58
m
m
3
5
6
7
8
2.84 dd
4.11 dd
2.82 m
3.18
3.93
2.87
a2.27
b1.42
a2.14
b2.50
±
3.07 m
3.91
2.85
±
3.06
3.10 dd
4.06
4.01 d
3.32
3.88 dd
2.37 dddd
a2.10 m
b1.27 dddd
a2.45 ddd
b1.99 ddd
±
3.26
2.33
2.84 m
a2.27 dddd
b1.44 dddd
a2.14 dd
b2.53 m
±
2.43
a2.25 dddd
b1.47 dddd
a2.15 dddd
b2.56 dddd
±
2.07 m
2.23 m
2.52 ddd
1.40 ddd
±
a1.45 ddd
b2.16 dd
a2.54 m
b2.22 m
±
a2.36 m
b1.57
a1.97 m
b2.03
±
b1.49 m
9
12
13
14
15
4.12 d
4.02 d
a5.49 d
b6.21 d
5.01 s
a5.50
b6.23
4.96
4.82
5.63
5.51
a5.48
b6.21
a5.48
b6.20
5.36
3.96 dd
3.68 dd
4.86
a5.47 d
b6.19 d
4.93
a5.14 s
b5.10 s
4.85
a5.46 d
b6.18 d
4.13 d
3.99 d
1.84 d
4.91
4.75
4.96 s
5.56 dd
5.50 dd
5.15
5.75
4.68
5.35 dd
5.29 dd
4.89
4.36 d
4.28 d
4.74 dd
5.13 br s
4.97 br s
5.44 br s
5.25 br s
5.54 m
5.55
1.54 s
1.49 s
2'
3'
6'
3.92 dd
3.52 m
a Coupling constants (Hz): 7 1,2b = 2b,3a = 6; 2a,2b = 12; 2a,3a = 2; 2b,3a = 8; 3a,15 = 2; 5,6 = 6,7 = 10; 8a,8b = 12; 8a,9a = 8b,9a
= 8b,9b 8a,9b = 8; 9a,9b = 14; 7,13a = 3; 7,13b = 4; 5,15' = 2. 8 1,2a = 7; 1,2,b = 7; 1,5 = 10; 2a,2b = 14; 2a,3b = 7; 2b 3b = 2; 5,6
= 6,7 = 9; 7,8a = 9; 7,8b = 10; 8a,8b = 9a,9b = 13; 8a,9a = 8b,9b = 8b,9a = 8a,9b = 8; 8a,8b = 12; 9a,9b = 14; 7,13a = 7,13b = 3. 9
5,6 = 6,7 = 9; 5'6' = 8; 7,13a = 7,13b = 3. 11 2a,2b = 14; 2a,3a = 2a,3b = 7; 3b,15 = 2; 3b,15' = 2; 6,7 = 9; 7,13a = 3; 7,13b = 4. 14
1,2a = 4; 1,2b = 7; 1,5 = 2; 2a,2b = 13; 2a3b = 7; 2b,3b = 7; 5,6 = 6,7 = 9; 7,8b = 9; 8a,8b = 13; 8a,9a = 8b,9b = 4; 8a,9b = 8b,9a =
12; 9a,9b = 13; 11,13 = 3; 11,13' = 12; 15',1 = 15',3b = 2; 15,1 = 15,3b = 2. 16 1,2b = 16; 1,5 = 10; 5,6 = 9, 6,7 = 10; 7,8a = 5; 7,8b =
10; 8a,8b = 13; 8a,9a = 5; 8a,9b = 11; 8b,9a = 10; 8b,9b = 5; 9a,9b = 13; 7,13a = 7,13b = 3. 17 1,2b = 2b,3a = 10; 1,5 = 8; 2a,2b =
13; 5,6 = 6,7 = 10; 8a,8b = 15; 12,12' = 121. 26 1,2a, 1,2b = 8; 5,6 = 10; 5,15 = 1; 6,7 = 9; 7,13a = 7,13b = 3; 14,14' = 11.
b Signals in the same column may be interchanged.
Further reduction of the lactol with NaBH4 yielded 17
(80%). Its EIMS spectrum showed a molecular ion at
Table 2
13C-NMR data of compounds 9, 11, 16, 26, and lactol (50.00 MHz,
CDCl3)a
m/z 234, corresponding to the molecular formula
1
C15H22O2: A strong hydroxyl band at 3250 cm (±
OH st.) was detected in the IR spectrum of 17. The
1H-NMR spectrum shows signals corresponding to a
hydroxymethylene group at d 4.12, 4.02 (1H, both d,
9
11
14
16
Lactol
26
1
50.2 d
37.6 t
73.7 d
150.1 s
44.6 d
85.1 d
45.9 d
88.8 s
45.4
71.8
49.6 d
38.0
74.4
48.0
34.0
48.8
30.5
130.4
22.8
24.4
0
J12,12 ˆ 12:1 Hz, H-12).
2
3
131.7
140.5
43.4 d
84.1
32.1t
151.9
48.2
The chlorinated derivative 26 was obtained from 24
by treatment with an excess of tosyl chloride in dry
pyridine. Its EIMS spectrum showed a molecular ion
at m/z 266, according to the molecular formula
C15H19O2Cl. No important changes appeared in the
1H-NMR spectrum compared to 24, di€erences arising
from the 13C-NMR spectrum, where the signal corre-
sponding to C-14 was strongly shifted down®eld for 24
(24: d 66.1, t, C-14; 26; d 49.2, t, C-14). The halogen
displacement of an hydroxyl group by a halogen de-
rivative is similar to that reported for compounds 7, 8,
and 9 (Collington & Meyers, 1971). No traces of the
C1-chloro rearrangement product were found. This
could be easily explained since the reaction involved a
4
153.4
91.7 s
82.7
154.4
49.3 d
85.9
43.6b
32.4
139.6
125.3
80.3
5
6
83.0
52.5
7
39.7
30.6
46.5
30.9
45.4
25.1
8
30.7b
30.7b
t
t
32.7
32.3
9
29.4
39.7
36.5
139.0
148.5 s
38.1
10
11
12
13
14
15
138.8 s
147.8 s
169.0 s
119.8 t
112.8 t
113.5 t
140.1
147.8
169.3
120.3
116.2
115.7
149.1
149.9
153.5
97.4 d
108.2
110.9
108.1
137.6
138.2
170.2 s
119.2
48.9
43.1b
177.4
58.6
d
120.2
52.6
114.0
112.4
112.4
17.2 q
a 9: d 95.9 (s, C-2'), d 62.2 (t, C-6'), d 26.2 (t, C-3', C-5'), d 19.9 (t,
C-4'); 11: d 113.2 (s, C-1'), d 28.6 (q, C-2'), d 28.3 (q, C-3').
b Signals in the same column may be interchanged.

750
D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
primary hydroxyl group where the mechanism must
necessarily be of the SN2-type.
bendazole) to which many fungi have developed resist-
ance demonstrated small or no zone of inhibition with
a di€use margin. Compounds possessing strong anti-
fungal activity produced a clear zone of inhibition
bounded by a sharp margin regardless of the size of
the inhibitory zone. Active compounds in decreasing
order of activity were 10 (dehydrozaluzanin C), 1
(dehydrocostuslactone), 4 (5a-hydroxydehydrocostusla-
cone), 18 (costunolide), 23 (parthenolide), and 3 (zalu-
Antifungal pre-screening of 36 sesquiterpene lac-
tones (Figs. 1 and 2) in a matrix format was conducted
directly on TLC plates sprayed with a spore suspen-
sion. Bioautography indicated that six sesquiterpene
lactones had antifungal activity against either C. acuta-
tum, C. fragariae, Phomopsis sp. or Botrytis cinerea.
(Table 3). Commercial fungicides (vinclozolin and thia-
Fig. 1. Guaianolides 1±17 and trans, trans-germacranolides 18±23 tested for antifungal activity.

D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
751
Fig. 2. Melampolides 24±27, cis, cis-germacranolide 28, and eudesmanolides 29±36 tested for antifungal activity.
zanin C). Bioautography indicated that guaianolides
10 and 1 appeared to be the most e€ective against C.
acutatum and C. fragariae, and 10 inhibited growth of
all four fungi. None of these compounds, except
parthenolide, have previously reported fungicidal ac-
tivity, and parthenolide was previously tested only
against Phytophthora capsici (Gennari, Abbattista-
Gentile & Cugudda, 1987).
The ®ve most active antifungal compounds identi®ed
by bioautography (1, 3, 4, 10, and 18) were sub-
sequently evaluated using a 96-well microbioassay sys-
tem (Wedge & Kuhajek, 1998) for activity against B.
cinerea, C. acutatum, C. fragariae, C. gloeosporioides
(Figs. 3±6), and F. oxysporum (data not shown). The
96-well microbioassay system used a liquid broth cul-
ture system in a dose-response format to evaluate each
compound in more detail. Secondary screening of
active compounds using this microbioassay system
identi®ed one compound (10) as having signi®cant
antifungal properties, one compound (1) with moder-
ate antifungal activity, and eliminated three weakly
active compounds (3, 4, and 18). Microbioassay results
indicated that the antifungal activity of sesquiterpene
lactones (1 and 10) appear to be species-dependent and
Table 3
Fungicidal activity of sesquiterpene lactones 1±36 on the autobiogra-
phy assay by visual evaluation of the clear inhibitory antifungal
zonea
Test compound C. acutatum C. fragariae Phomopsis sp. B. cinerea
1
WA (7)
±
WA (8)
±
±
2
±
WA (4)
WA (6)
±
±
3
WA (4)
WA (7)
±
±
±
4
5±9
±
±
±
SA (8)
±
SA (11)
±
SA (7)
10
11±17
18
SA (7)
±
±
VW (6)
±
±
VW (4)
±
±
±
19±22
23
±
±
±
VW (4)
±
VW (10)
WA (4)
±
±
±
24±36
Vinclozolin
±
±
VW (3.5)
SA (11)
±
VW (7.5)
SA (15.5)
SA (16.5)
±
Chlorothalonil SA (7)
Thiabendazole
SA (10)
±
±
a a: Vinclozolin; b: Chlorothalonil; c: Thiabendazole; dashes: No
Activity; VW: Very Weak activity; WA: Weak Activity; SA; Strong
Activity. Letters indicate zone clairity and numbers indicate zone di-
ameter (mm). Mean zone diameter from two experiments is
expressed in parentheses.

752
D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
Fig. 3. Antifungal activity of compounds 1, 10, 3, 4, and 18 against Botrytis cinerea growth. Commercial standard fungicides: Chl.: chlorothalo-
nil, Thiab.: thiabendazole, Vincl.: vinclozolin. Least Square Means for percent inhibition of each sesquiterpene lactone is listed for each of three
concentrations (n = 4) by fungal species. Positive values indicate growth stimulation and negative values indicate growth inhibition relative the
untreated control (n = 32).
Fig. 4. Antifungal activity of compounds 1, 10, 3, 4, and 18 against Colletotrichum acutatum growth. Commercial standard fungicides: Chl.:
chlorothalonil, Thiab.: thiabendazole, Vincl.: vinclozolin. Least Square Means for percent inhibition of each sesquiterpene lactone is listed for
each of three concentrations (n = 4) by fungal species. Positive values indicate growth stimulation and negative values indicate growth inhibition
relative the untreated control (n = 32).

D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
753
Fig. 5. Antifungal activity of compounds 1, 10, 3, 4, and 18 against Colletotrichum fragariae growth. Commercial standard fungicides: Chl.:
chlorothalonil, Thiab.: thiabendazole, Vincl.: vinclozolin. Least Square Means for percent inhibition of each sesquiterpene lactone is listed for
each of three concentrations (n = 4) by fungal species. Positive values indicate growth stimulation and negative values indicate growth inhibition
relative the untreated control (n = 32).
Fig. 6. Antifungal activity of compounds 1, 10, 3, 4, and 18 against Colletotrichum gloeosporioides growth. Commercial standard fungicides: Chl.:
chlorothalonil, Thiab.: thiabendazole, Vincl.: vinclozolin. Least Square Means for percent inhibition of each sesquiterpene lactone is listed for
each of three concentrations (n =4) by fungal species. Positive values indicate growth stimulation and negative values indicate growth inhibition
relative the untreated control (n =32).

754
D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
that the biological activity depends upon the type of
carbon skeleton. Furthermore, the results demon-
strated that our microbioassay test system was su-
ciently sensitive to make biological predictions about
structure-activity relationships.
play an important role (Macõas, et al., 1992; Rodrõguez
et al., 1997). In this study, we observed that several
compounds possessing the a-methylene-g-lactone ring
did not exert any inhibition of mycelial growth (2, 5±
9, 19, 21, 24±33; Table 3). Some of these compounds
also have other reactive groups like epoxides, but this
does not appear to improve activity. Comparison
between compounds 1, 3, 9, 11 shows that the intro-
duction of steric impediments in the form of bulky
groups can cause the loss of antifungal activity in com-
pounds 9 and 11 with respect to those of 1 and 3;
these results are in agreement with previous reports
(Macõas et al., 1992). Lipophilicity is another factor to
be considered (Beekman et al., 1997); compounds 5
and 6, with an increasing number of hydroxyl groups,
presented no activity in comparison with mono-hy-
droxylated compounds 3 and 4, or with the weekly
polar compounds 1 and 10. The fact that 1, 3, 4, and
10 were toxic to all fungi tested but with di€erent pro-
®les of activity may be indicative of di€erent modes of
action from those of the commercial fungicidal stan-
dards.
Compound 10 represents a special case, since the in-
hibition of mycelial growth was greater than that
shown by the rest of guaianolides and sesquiterpene
lactones tested, even with species having fungicidal re-
sistance (C. acutatum, C. fragariae Ð thiabendazole;
B. cinerea Ð thiabendazole, vinclozolin; F. oxysporum
Ð vinclozolin).
Previous results (unpublished) showed that the phy-
totoxicity of 10 was greater than that of other guaia-
nolides and sesquiterpene lactones without an
additional a,b-unsaturated carbonyl group. A di€erent
mode of action has been proposed for 10 due to this
di€erence (Galindo, Dayan, Hernandez, Macõas and
Duke, 1999). No similar e€ects have been reported in
fungi yet, except in the case of encelin, a cis-7,8-eudes-
manolide with a double a,b-unsaturated carbonyl
group, which has been reported to exert a detergent
action on the growth of fungal cells of Mucor rouxii
(Robles et al., 1995). Altered cellular permeability and
quantitative changes in protein synthesis could also be
detected in both the carbonyl containing encelin, and
in the dihydro-derivative 3-hydroxyencelin (Trejo,
Gaytan, Mendoza & Sabanero, 1996).
The most active compound in bioautography, 10
(dehydrozaluzanin), exhibited selective activity against
Colletotrichum species (Figs. 4±6). Dehydrozaluzanin
at 30 mM reduced C. fragariae growth by 90% (Fig. 5),
C. gloeosporioides by 89% (Fig. 6), and C. acutatum
by 29% (Fig. 4). Compound 10 was nearly as e€ective
as the commercial fungicide chlorothalonil against C.
fragariae and C. gloeosporioides, and was more e€ec-
tive than thiabendazole and vinclozolin. However, at
30 mM, 10 was ine€ective against B. cinerea (Fig. 3) or
F. oxysporum (data not shown). Colletotrichum acuta-
tum showed less sensitivity than C. fragariae or C.
gloeosporioides to 10. The e€ective concentration
against C. acutatum appears to be about 10-fold higher
in comparison to the other two Colletotrichum species.
Compound 1 had the second highest fungicidal ac-
tivity of the compounds tested in bioautography. At
30 mM, growth inhibition of 51% was observed in C.
fragariae (Fig. 5), 25% in C. acutatum (Fig. 4), 17% in
C. gloeosporioides (Fig. 6), and 22% in B. cinerea
(Fig. 3). There was no signi®cant inhibition of F. oxy-
sporum (data not shown) at 30 mM. Compound 1 was
the only compound that showed any signi®cant
ꢀp < 0:05† activity against B. cinerea. Compounds 4
(zaluzanin C), 18 (costunolide), and 23 (parthenolide)
all showed weak activity in the bioautography assay
but were ine€ective as antifungal agents in the liquid
broth microbioassay when compared to commercially
available fungicides. None of the sesquiterpene lac-
tones tested showed signi®cant inhibition against F.
oxysporum. Antifungal activity of several sesquiterpene
lactones against Colletotrichum species suggested that
these compounds might have a plant defense role in
maintaining leaf integrity by inhibiting foliar patho-
gens. Further research into the use of 10 and 1 as pro-
tectant fungicides is warranted. However, most of
these compounds with an exocyclic methylene function
are strongly allergenic (Type IV allergy) and may
induce dermatoses (Hausen and Schmalle, 1985;
Schae€er, Talaga, Stampf & Benezra, 1990). Ulti-
mately, it is expected that the use of these compounds
as antifungal agents will be limited by their phytotoxi-
city rather than their allergenic nature.
Though further investigation is needed, it is reason-
able to postulate that the di€erence in the activity
between 10 and the rest of the sesquiterpene lactones
tested relies on the presence of the a,b-unsaturated car-
bonyl group in the cyclopentanone ring. Nevertheless,
the group needs to be accessible since reaction through
a Michael addition mechanism is also viable (Galindo
et al., 1999).
It has been reported that the biological activity of
sesquiterpene lactones is generally due to the a-methyl-
ene-g-lactone system (Hall, Lee, Starnes, Muraoka,
Sumida & Waddell, 1980). However, this does not
appear to be the only requisite in fungi (Marles et al.,
1995) or other living systems (Rodrõguez, Enriz, Santa-
gata, Jauregui, Pestchanker & Giordano, 1997). Other
requirements such as molecular accessibility seem to
Hydroxyachillin (15) has no exomethylene group in
the lactone ring, but has an enone group with two
double bonds conjugated to the carbonyl group. How-

D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
755
ever, compound 15 had no activity, probably because
the two double bonds are hindered (b-methyl groups)
and the accessibility to allow nucleophilic additions is
extremely restricted, especially with nucleophiles such
as bulky biomolecules.
With respect to germacranolide type sesquiterpene
lactones, only 18 and its 1,10-epoxyderivative 23 pro-
duced weak mycelial growth inhibition. Parthenolide
(19), which has some fungicidal activities (Picman,
1984), did not show any activity with the fungi in our
assays. Other germacranolides did not demonstrate
any antifungal activity. Although several fungicidal ac-
tivities were reported for eudesmanolides (Marles et
al., 1995; Trejo et al., 1996), none of the eudesmano-
lides tested (compounds 29±36) showed any activity.
reaction mixt. was then evapd. under reduced pressure
and puri®ed by CC [hexane±AcOEt (1:4)] to give 11
(99% yield). Compounds 12 and 13 were synthesized
from 1 by re¯uxing with hexamethylphosphoric tria-
mide (HMPA) and 20% aqueous CaCO3 solution for
7 days according to the published method (Macõas,
Galindo, Massanet & Gomez-Madero, 2000a). Oxi-
dation of 12 with SeO2 and tert-butyl hydroperoxide
(Macõas et al., 1992) a€orded 14 as the major product
(62% yield). A dry toluene solution of 1 was stirred at
708 under N2 with a DIBAL/toluene solution (1:1.1)
for 1.5 h, following which MeOH and H2O were
sequentially added. The resulting gelatinous crude was
homogenized and extracted several times with EtOAc
and H2O. The two-phase mixt. was then separated and
the mother liquors extracted with EtOAc (Â3). All or-
ganic phases were dried over Na2SO4 and evapd.
under reduced pressure to yield 61% of the lactol de-
rivative of 1. Further reduction of this lactol derivative
with a methanolic solution of NaBH4 at 08C for 1 h
produced 17 (80% yield).
3. Experimental
3.1. Starting materials
Dehydrocostuslactone (1) and costunolide (18) were
obtained from crude Costus Resin Oil (Pierre Chauvet
S. A.) by CC puri®cation and recrystallization from
hexane:EtOAC mixt. Parthenolide (19) was obtained
from Magnolia grandi¯ora L. leaves, and was puri®ed
by CC and recrystallization from hexane:EtOAc mixt.
Hydroxyachillin (15) was obtained from Artemisia
lanata Wild. extract and puri®ed by CC. Structures
were veri®ed by comparison of spectroscopic data
(MS, IR, 1H and 13C-NMR) with those reported in
the literature. An: Refs. a-Santonin (34) was purchased
from Sigma Co (st. Louis, MO), and the commercial
fungicides vinclozolin, chlorothalonil and thiabenda-
zole from Chem. Service, West Chester PA.
3.3. Synthesis of germacranolides
Germacranolides 20±23 and melampolides 24, 25,
and 27 were obtained according to the published
methods (Rodrõgues, Garcõa & Rabi, 1978; Fischer,
Weidenhamer, Riopel, Quijano & Menelaou, 1990;
Umbrier & Sharpless, 1977). Cis,cis-germacranolide 28
was synthesized according to (Macõas et al., 1992).
Compound 26 was synthesized from 24 by treatment
with excess of tosyl chloride in dry pyridine, and stir-
ring for 24 h. Reaction mixt. was then ®ltered and pyr-
idine was evapd. under reduced pressure using
cyclohexane. After puri®cation using CC (hexane:E-
tOAc, 1:4), compound 26 was obtained in a yield of
80%.
3.2. Synthesis of guaianolides
Compounds 2±6 were synthesized according to the
previously described method (Macõas et al., 1992).
Chlorinated derivatives 7, 8 and 16 were obtained by
treatment of 2 with freshly distilled mesyl chloride in
dry pyridine at 08C during 24 h. Reaction mixt. was
then ®ltered and pyridine was evapd. under reduced
pressure using cyclohexane. HPLC puri®cation (hex-
ane±AcOEt, 1:9) yielded 7 (42%), 8 (20%), and 16
(5%). Treatment of 2 with anhydrous, freshly distilled
dihydropyran and p-toluene sulfonic acid as catalyst in
dry THF for 16 h, followed by neutralization with
K2CO3, ®ltration and evapn. under reduced pressure,
yielded a racemic mixture of 3a-O-tetrahydropyranyl
derivatives 9 (95%). Dehydrozaluzanin C 10 was syn-
thesized as previously described (Macõas, Galindo,
Molinillo & Castellano, 2000b).
3.4. Synthesis of eudesmanolides
Eudesmanolides 29±31 were synthesized as pre-
viously described (Rodrõgues et al., 1978) while 35±36
were obtained from 34 according to (Macõas, Aguilar,
Molinillo, Massanet & Fronczek, 1994).
3.5. Pathogen production
Isolates of Colletotrichum acutatum Simmonds, Col-
letotrichum fragariae Brooks, Colletotrichum gloeospor-
ioides (Penz.) Penz. & Sacc. in Penz., and a Phomopsis
sp. were obtained from B. J. Smith, USDA, ARS,
Small Fruit Research Station, Poplarville, MS. The
three Colletotrichum species were isolated from straw-
berry (Fragaria x ananassa Duchesne) and the Pho-
mopsis sp. was isolated from muscadine grape (Vitis
rotundifolia Michx.). Botrytis cinerea Pers.:Fr was iso-
Acetonide 11 was prepd. from 6 by re¯uxing in
Me2CO with p-toluene sulfonic acid for 24 h. The

756
D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
lated from commercial grape (Vitis vinifera ) and
Fusarium oxysporum Schlechtend:Fr from orchid
(Cynoches sp.). Colletotrichum, Phomopsis and Botrytis
species were used in the bioautography as a prescreen
for active antifungal compounds. Colletotrichum,
Botrytis and Fusarium species were used in the 96-well
microbioassay because of suitable in vitro growth and
economic importance. Commercial fungicides vinclozo-
lin, chlorothalonil, and thiabendazole (Chem, Service,
West Chester, PA) were used as validation controls in
all microbioassay tests.
sensitivity of B. cinerea, C. acutatum, C. fragariae, C.
gloeosporioides, and F. oxysporum to the various ses-
quiterpene lactones in comparison with known fungici-
dal standards (Wedge & Kuhajek, 1998). Vinclozolin,
chlorothalonil, and thiabendazole, which represent
three di€erent modes of action, were used as standards
in this experiment. Each fungus was challenged in a
dose-response format using test compounds where the
®nal treatment concentrations were 0.3, 3.0 and 30.0
mM. Microtiter plates (Nunc MicroWell (untreated),
Roskilde, Denmark) were covered with a plastic lid
and incubated in a growth chamber as described pre-
viously for fungal growth. Growth was then evaluated
by measuring absorbance of each well at 620 nm using
a microplate photometer (Packard Spectra Count,
Packard Instrument, Downers Grove, IL).
3.6. Inoculum preparation
Conidial suspensions were prepared according to
published procedures (Wedge & Kuhajek, 1998). Coni-
dial concentrations were determined spectro photo-
metrically (Espinel-Ingrof & Kerkerin, 1991; Wedge &
Kuhajek, 1998) from a standard curve, and suspen-
3.9. Microbioassay experimental design
sions were then adjusted with sterile distilled water to
1
Each fungus was challenged using a 96-well plate
microbioassay format in duplicate. Each chemical was
evaluated in duplicate in a dose-response format where
the ®nal assay concentrations were 0.3, 3.0 and 30.0
mM. Eight wells containing broth and inoculum served
as positive controls; eight wells containing solvent at
the appropriate concentration and broth without
inoculum were used as negative controls. Mean absor-
bance values and standard errors were used to evaluate
fungal growth at 48 h. Analysis of variance of Least
Square Means for percent inhibition of each fungus at
each dose of test compound (n = 4) relative the
untreated controls (n = 32) were used to evaluate fun-
gal growth inhibition. The experiment used a Repeated
Measures Design and was repeated once. Data pre-
sented were pooled from two independent experiments.
a concentration of 1.0 Â 106 conidia ml
.
3.7. Bioautography
Inhibition of fungal growth on chromatographic
plates was evaluated by modi®cations of thin layer
chromatography
(TLC)
bioautographic
assays
(Homans & Fuchs, 1970; Osborne, Chase, Lunness,
Scott & Daniels, 1994). Sesquiterpene lactones were
dissolved in Me2CO and commercial fungicide stan-
dards in 95% EtOH. Each test compound was placed
on the TLC plate from stock solutions to achieve a
®nal amount of 2 mg using a disposable glass micropip-
ette for each sample. To detect biological activity
directly on the TLC plate, silica gel plates with a ¯uor-
escent indicator (250 mg, Silica Gel GF Uniplate, Ana-
ltech, Newark, DE) were sprayed with
a spore
suspension. Aliquots of 25±50 ml of inoculum spray
solution (ca. 3 Â 105) were prepared for each test fun-
gus with liquid potato dextrose broth (PDB) contain-
ing 12 g/500 ml (PDB), 0.1% bacto agar, and 0.1%
Tween-80. Using a 50 ml chromatographic sprayer,
each plate was sprayed lightly (to a damp appearance)
three times with conidial suspension. Inoculated plates
were placed in a 30 Â 13 Â 7.5 cm moisture chamber
(398-C, Pioneer Plastics, Dixon, KY) and incubated in
a growth chamber at 24 2 18C and 12 h photoperiod
under 60 2 5 mmol light. Inhibition of fungal growth
was measured 4 days after treatment. Sensitivity of
each fungal species to test compounds and a fungicide
standard at 2 mg was determined by comparing size of
inhibitory zones.
Acknowledgements
This work is supported, in part, by foundation from
the University of Cadiz, and the Department of Edu-
cation of the Andalusian Government. Technical as-
sistance of J.L. Robertson and J.M. Kuhajek and
statistical support provided by D.L. Boykin, USDA
ARS, Stoneville MS in these studies are appreciated.
References
Beekman, A. C., Woerdenbag, H. J., Uden, W. V., Pras, N.,
Konings, A. W. T., Wikstroem, H. V., & Schmidt, T. J. (1997).
Structure-cytotoxicity relationships of some helenanolide-type ses-
quiterpene lactones. Journal of Natural Products, 60, 252±257.
Collington, F. W., & Meyers, A. I. (1971). A facile and speci®c con-
version of allylic alcohols to allylic chlorides without rearrange-
ment. Journal of Organic Chemistry, 36, 3044±3045.
3.8. Microtiter assay
A 96-well microtiter assay was used to determine
Espinel-Ingrof, A., & Kerkering, T. M. (1991). Spectrophotometric

D.E. Wedge et al. / Phytochemistry 53 (2000) 747±757
757
method of inoculum preparation for the in vitro susceptibility
testing of ®lamentous fungi. Journal of Clinical Microbiology, 29,
393±394.
Macõas, F. A., Galindo, J. C. G., Massanet, G. M. & Gomez-
Madero, J. F. (2000a). Sythesis of 13-hydroxylated and 11,13-
dihydroxylated sesquiterpene lactones: Michael addition of
nucleophilic hydroxyl group. Synthesis, submitted.
a
Farr, D. F., Bills, G. F., Chamuris, G. P., & Rossman, A. Y. (1989).
Fungi on plant and plant products in the United States. St. Paul,
MN, USA: American Phytopathological Society Press.
Fischer, N. H. (1991). Sesquiterpene lactones. Methods in Plant
Biochemistry, 7, 187±211.
Macõas, F. A., Galindo, J. C. G., Molinillo, J. M. G. & Castellano,
D. (2000b). Dehydrozaluzanin C, a potent plant growth regulator
with potential use as a natural herbicide. Phytochemistry, in
press.
Fischer, N. H., Weidenhamer, J. D., Riopel, J. L., Quijano, L., &
Menelaou, M. A. (1990). Stimulation of witchweed germination
Macõas, F. A., Varela, R. M., Torres, A., & Molinillo, J. M. G.
(1993). Potential allelopathic guaianolides from cultivar sun¯ower
leaves, var. SH-221. Phytochemistry, 34, 669±674.
by
Phytochemistry, 29, 2479±2483.
sesquiterpene
lactones:
a
structure-activity
study.
Marles, R. J., Pazos-Sanou, L., Compadre, C. M., Pezzuto, J. M.,
Bloszyk, E., & Arnason, J. (1995). Sesquiterpene lactones re-
visited: recent developments in the assessment of biological activi-
ties and structure relationships. Recent Advances in
Phytochemistry, 29, 333±356.
Fraga, B. (1996). Natural Sesquiterpenoids. Natural Product Reports,
13, 307±326.
Fraga, B. (1997). Natural Sesquiterpenoids. Natural Product Reports,
14, 107±162.
Fraga, B. (1998). Natural Sesquiterpenoids. Natural Product Reports,
15, 73±92.
McChesney, J. D. (1993). Biological and chemical diversity and the
search for new pharmaceuticals and other bioactive natural pro-
ducts. In A. D. Kinghorn, & M. F. Balandrin, Human medicinal
agents from plants (pp. 38±47). In ACS Symposium Series 534.
Washington, DC: American Chemical Society.
Galindo, J. C. G., Dayan, F. E., Hernandez, A., Macõas, F. A., &
Duke, S. O. (1999). Dehydrozaluzanin C, a natural sesquiterpeno-
lide, causes rapid plasma membrane leakage. Phytochemistry, 52,
805±813.
Osborne, A. E., Chase, B. R., Lunness, P., Scott, P. R., & Daniels,
M. J. (1994). An oat species lacking avenacin is susceptible to
infection by Gaeumannomyces graminis var. tritici. Physiological
and Molecular Plant Pathology, 45, 457±467.
Gennari, M., Abbattista-Gentile, I., & Cugudda, L. (1987). Activity
of a sesquiterpene-g-lactone against phytopathogenic fungi.
P¯anzenschutz, 94, 68±73.
Hall, I. H., Lee, K. H., Starnes, C. O., Muraoka, O., Sumida, Y., &
Waddell, T. G. (1980). Antihyperlipidemic activity of sesquiter-
pene lactones and related compounds. Journal of Pharmacological
Science, 69(6), 694±697.
Picman, A. K. (1984). Antifungal activity of sesquiterpene lactones.
Biochemical Systematic Ecology, 12, 13±18.
Robles, M., Aregullin, M., West, J., & Rodrõguez, E. (1995). Recent
studies on the zoopharmacognosy, pharmacology, and neurotoxi-
cology of sesquiterpene lactones. Planta Medica, 61, 199±203.
Rodrõgues, A. A. S., Garcõa, M., & Rabi, J. A. (1978). Facile biomi-
metic synthesis of costunolide-1,10-epoxide, santamarin and rey-
nosin. Phytochemistry, 17, 953±954.
Hausen, B. M., & Schmalle, H. W. (1985). Structure-activity aspects
of 4 allergenic sesquiterpene lactones lacking the exocyclic alpha-
methylene at the lactone ring. Contact Dermatitis, 13, 329±332.
Homans, A. L., & Fuchs, A. (1970). Direct bioautography on thin
layer chromatography as a method for detecting fungitoxic sub-
stances. Journal of Chromatography, 51, 327±329.
Rodrõguez, A. M., Enriz, R. D., Santagata, L. N., Jauregui, E. A.,
Pestchanker, M. J., & Giordano, O. S. (1997). Structure cytopro-
tective activity relationship of simple molecules containing a,b-
unsatured carbonyl system. Journal of Medicinal Chemistry, 40,
1827±1834.
Kalsi, P. S., Kaur, G., Sharma, S., & Talwar, K. K. (1984).
Dehydrocostuslactone and plant growth activity of derived guaia-
nolides. Phytochemistry, 23, 2855±2861.
Kirst, H. A., Michel, K. H., Mynderase, J. S., Chio, E. H., Yao, R.
C., Nakasukasa, W. M., Boeck, L. D., Occlowitz, J. L., Paschal,
J. W., Deeter, J. B., & Thompson, G. D. (1992). Discovery, iso-
lation and structure elucidation of a family of the structurally
unique, fermentation derived tetracyclic macrolides. Synthesis and
Chemistry of Agrochemicals, III, 214±225.
Sabanero, M., Quijano, L., Rõos, T., & Trejo, R. (1995). Encelin: a
fungal growth inhibitor. Planta Medica, 61, 185±186.
Schae€er, M., Talaga, P., Stampf, J. L., & Benezra, C. (1990).
Cross-reaction in allergic contact dermatitis from alpha-methylen-
gamma-butyrolactones: importance of the cis or trans ring junc-
tion. Contact Dermatitis, 22, 32±36.
Maas, J. L. (1998). In Compendium of strawberry diseases (pp. 1±98).
MN, USA: American Phytopathological Society St. Paul, APS
Press.
Trejo, R., Gaytan, A., Mendoza, D., & Sabanero, M. (1996). 3-
Hydroxyencelin: synthesis and in vitro activity. Microbios, 88, 97±
104.
Macõas, F. A., Aguilar, J. M. G., Molinillo, J. M. G., Massanet, G.
M., & Fronczek, F. R. (1994). Studies on the stereostructure of
eudesmanolides from Umbelliferae; synthesis of 11b-angeloyloxy-
a-santonin. Tetrahedron, 18, 5439±5450.
Macõas, F. A., Galindo, J. C. G., & Massanet, G. M. (1992).
Natural products as allelochemicals. I. Potential allelopathic ac-
tivity of several sesquiterpene lactone models. Phytochemistry, 31,
1969±1977.
Umbreit, M. A., & Sharpless, K. B. (1977). Allylic oxidation of ole-
®ns by catalytic and stoichiometric selenium dioxide with tert-
butyl hydroperoxide. Journal of the American Chemical Society,
99, 5526±5528.
Wedge, D. E., & Kuhajek, J. M. (1998). A microbioassay for fungi-
cide discovery. SAAS Bulletin of Biochemistry and Biotechnology,
11, 1±7.