Structure-activity relationships of trichothecene toxins in an Arabidopsis thaliana leaf assay

Article abstract of DOI:10.1021/jf0709193

Many Fusarium species produce trichothecenes, sesquiterpene epoxides that differ in patterns of oxygenation and esterification at carbon positions C-3, C-4, C-7, C-8, and C-15. For the first comprehensive and quantitative comparison of the effects of oxygenation and esterification on trichothecene phytotoxicity, we tested 24 precursors, intermediates, and end products of the trichothecene biosynthetic pathway in an Arabidopsis thaliana detached leaf assay. At 100 μM, the highest concentration tested, only the trichothecene precursor trichodiene was nontoxic. Among trichothecenes, toxicity varied more than 200-fold. Oxygenation at C-4, C-8, C-7/8, or C-15 was, on average, as likely to decrease as to increase toxicity. Esterification at C-4, C-8, or C-15 generally increased toxicity. Esterification at C-3 increased toxicity in one case and decreased toxicity in three of eight cases tested. Thus, the increase in structural complexity along the trichothecene biosynthetic pathway in Fusarium is not necessarily associated with an increase in phytotoxicity.

Full text of DOI:10.1021/jf0709193

J. Agric. Food Chem. 2007, 55, 6487−6492 6487
Structure−Activity Relationships of Trichothecene Toxins in an
Arabidopsis thaliana Leaf Assay
ANNE E. DESJARDINS,* SUSAN P. MCCORMICK, AND MICHAEL APPELL
National Center for Agricultural Utilization Research, Agricultural Research Service, U.S. Department
of Agriculture, Peoria, Illinois 61604
Many Fusarium species produce trichothecenes, sesquiterpene epoxides that differ in patterns of
oxygenation and esterification at carbon positions C-3, C-4, C-7, C-8, and C-15. For the first
comprehensive and quantitative comparison of the effects of oxygenation and esterification on
trichothecene phytotoxicity, we tested 24 precursors, intermediates, and end products of the
trichothecene biosynthetic pathway in an Arabidopsis thaliana detached leaf assay. At 100 µM, the
highest concentration tested, only the trichothecene precursor trichodiene was nontoxic. Among
trichothecenes, toxicity varied more than 200-fold. Oxygenation at C-4, C-8, C-7/8, or C-15 was, on
average, as likely to decrease as to increase toxicity. Esterification at C-4, C-8, or C-15 generally
increased toxicity. Esterification at C-3 increased toxicity in one case and decreased toxicity in three
of eight cases tested. Thus, the increase in structural complexity along the trichothecene biosynthetic
pathway in Fusarium is not necessarily associated with an increase in phytotoxicity.
KEYWORDS: Trichothecenes; Arabidopsis thaliana; Fusarium; phytotoxicity
INTRODUCTION
Many plant pathogenic species of the genus Fusarium produce
trichothecenes (1-24; Table 1 and Figure 1), a large group of
sesquiterpene epoxides that are inhibitors of eukaryotic protein
synthesis. Fusarium graminearum, Fusarium sporotrichioides,
and other trichothecene-producing Fusarium species are com-
mon pathogens of cereal grains, especially in temperate climates
in Europe, Asia, and the Americas, where trichothecene
contamination of wheat, barley, rye, and maize can be a serious
problem.ThetrichothecenebiosyntheticpathwayinF.graminearum
ultimately produces the 8-keto, 7-hydroxy trichothecenes deox-
ynivalenol 18 and nivalenol 22, and in F. sporotrichioides, it
produces the 8-O-isovalerate ester T-2 toxin 16. These com-
pounds and other trichothecenes cause feed refusal, ill thrift,
and emetic syndromes in swine and other animals and are
associated epidemiologically with chronic and fatal toxicoses
in humans (1).
The trichothecene skeleton contains a carbon position-1 (C-
1) pyran group and a C-12,13 epoxide, but trichothecenes differ
otherwise in patterns of extraskeletal oxygenation and esterifi-
cation at C-3, C-4, C-7, C-8, and C-15 and in the presence of
a keto group at C-8. Fusarium trichothecenes contain only
acetate and other short-chain esters and lack the complex
macrocylic esters found in some other trichothecene-producing
fungi. Determination and quantification of biological effects of
trichothecenes in animal systems began shortly after the
discovery of T-2 toxin 16, deoxynivalenol 18, nivalenol 22, and
other trichothecenes in the 1960s and 1970s. Among Fusarium
Figure 1. Structures of trichodiene 1 and isotrichotriol 2.
trichothecenes, animal toxicity is higher among compounds with
C-3, C-4, C-8, and C-15 oxygenations, such as T-2 toxin 16
and nivalenol 22, than among compounds with only C-3 and
C-15 oxygenations, such as calonectrin 8 (2). Relative biological
activities of the more-oxygenated trichothecenes in a wide range
of animal systems have been extensively studied and recently
reviewed (1, 3).
As compared to animal systems, assessment of the effects of
trichothecenes in plant systems has been rather neglected,
although toxicity of 4,15-diacetoxyscirpenol 11 to peas, lettuce,
and other plants was first reported more than 40 years ago (4).
Comparative studies of trichothecene phytotoxicity have been
limited, in large part, by the difficulty and expense of obtaining
trichothecenes in highly purified form in quantities sufficient
for phytotoxicity assays. Typically, one or two compounds have
been isolated in different laboratories and then tested with
different assay methods and plant species, thus making quantita-
tive comparisons very difficult. The fragmentary studies avail-
able have found differences in toxicity of trichothecenes between
plant and animal systems, which might have been expected,
but have also found differences in relative toxicity of tricho-
thecenes in different plant systems (5-8).
* To whom correspondence should be addressed. Tel: 309-681-6378.
Fax: 309-681-6665. E-mail: anne.desjardins@ars.usda.gov.
10.1021/jf0709193 This article not subject to U.S. Copyright. Published 2007 by the American Chemical Society
Published on Web 07/14/2007

6488 J. Agric. Food Chem., Vol. 55, No. 16, 2007
Table 1. Structures and Lethality (LD50) of Trichothecenes
Desjardins et al.
substituentsa
LD50
(µ
± SE
M)b
no.
compound
isotrichodermol
R1
R2
R3
R4
R5
3
4
5
6
7
8
9
OH
OAc
OH
OA
OH
H
H
OH
OH
OAc
H
H
H
H
H
H
OH
H
H
9.0 ± 1.5
isotrichodermin
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
16.1
>100
>100
26.6
±
5.1
3,15-dideacetylcalonectrin
15-deacetylcalonectrin
3-deacetylcalonectrin
calonectrin
±
2.9
Figure 2. Representative healthy (top row) and dead (bottom row)
OAc OAc
2.6
±
1.9
detached leaves of A. thaliana after
trichothecene.
7 days of incubation with a
scirpentriol
OH
OH
OH
OH
OAc OH
OAc OAc
>100
10 15-acetoxyscirpenol
11 4,15-diacetoxyscirpenol
12 3,4,15-triacetoxyscirpenol
13 neosolaniol
14 3-acetylneosolaniol
15 HT-2 toxin
16 T-2 toxin
17 3-acetyl-T-2 toxin
18 deoxynivalenol
19 3-acetyldeoxynivalenol
20 15-acetyldeoxynivalenol
21 3,15-diacetyldeoxynivalenol OAc OAc
22 nivalenol
23 4-acetylnivalenol
24 3,15-diacetylnivalenol
3.7
1.5
±
±
1.1
0.2
OAc OAc OAc
OH
OAc OAc OAc OH
OH
OH
27.6
13.0
14.6
±
±
±
3.5
1.4
5.6
Table 2. Toxicity (ED50) of Selected Trichothecenes and Precursors
OAc OAc OH
compounds
trichodiene
isotrichotriol
ED50
>100
>100
(±SE) (µ
M)a
OAc OH
OAc OAc Iso
Iso
1.0
0.5
4.1
± 0.8
1
2
±
0.05
0.2
OAc OAc OAc Iso
±
3
4
isotrichodermol
isotrichodermin
3.5 (
±
1.9)
2.2)
OH
OAc OH
OH OAc
OH
H
H
H
H
OH
OAc
dO
dO
dO
dO
dO
dO
dO
OH 23.0
OH 18.7
OH 4.0
OH 28.2
OH >100
OH 79.7
±
6.8
8.6 (±
±
±
±
12.4
2.8
0.4
5
6
7
9
3,15-dideacetylcalonectrin
15-deacetylcalonectrin
3-deacetylcalonectrin
scirpentriol
90.1 (
>100
±7.7)
±
7.6 (±
1.4)
4.4)
OH
OH
OH
OH
53.6 (±
13.6
12
18
19
20
21
22
23
24
3,4,15-triacetoxyscirpenol
deoxynivalenol
3-acetyldeoxynivalenol
15-acetyldeoxynivalenol
3,15-diacetyldeoxynivalenol
nivalenol
6.7 (
3.8 (
3.4 (
1.3 (
10.0 (
>100
16.1 (
±
1.7)
3.0)
2.6)
1.1)
OAc OAc OH
OH 4.3 ± 1.4
±
±
±
a Iso
)
isovaleryl ester. b LD50 was the concentration at which 50% of the A.
±
1.5)
thaliana leaves were dead following treatment.
4-acetylnivalenol
3,15-diacetylnivalenol
±
1.9)
2.2 (±0.6)
During the course of long-term research on the biosynthesis
of deoxynivalenol 18 and nivalenol 22 by F. graminearum and
of T-2 toxin 16 by F. sporotrichioides, we have used gene-
disruption mutants, precursor feeding experiments, and other
approaches to create a library of trichothecene biosynthetic
pathway precursors, intermediates, and end products. The
cruciferous weed Arabidopsis thaliana has been used for assay
of fungal phytotoxins and is sensitive to trichothecenes (9-
12). In this study, we combined our compound library and A.
thaliana for a comprehensive and quantitative analysis of
structure-activity relationships of trichothecenes biosynthesized
by Fusarium.
a ED50 is the concentration at which 50% of the A. thaliana leaves showed
symptoms of chlorosis, shriveling, or death following treatment.
acetyltransferase Tri101- mutant) (15). Isotrichodermin 4 was isolated
from YEPD cultures of F. sporotrichioides strain A11b (a C-15
oxygenase Tri11- mutant) (16). 3,15-Didecalonectrin 5 and 15-
decalonectrin 6 were isolated from YEPD cultures of F. sporotrichioides
strain MB2972 (14). 3-Decalonectrin 7 and calonectrin 8 were isolated
from cornmeal cultures of F. graminearum strain LH1-41 (a calonectrin
oxygenase Tri1- mutant) (17). 4,15-Diacetoxyscirpenol 11 was isolated
from YEPD cultures of F. sporotrichioides strain 1716cos 9-1#1 (a
trichothecene overexpression Tri1- mutant) (18). 3,4,15-Triacetoxy-
scirpenol 12, 3-acetylneosolaniol 14, and 3-acetyl T-2 toxin 17 were
isolated from YEPD cultures of F. sporotrichioides strain NA-476 (a
C-3 esterase Tri8- mutant) (19). Neosolaniol 13 and T-2 toxin 16 were
isolated from YEPD cultures of F. sporotrichioides strain 5493cos9-
1#11 (a trichothecene overexpressing Tri4- mutant) (18). HT-toxin 15
was isolated from YEPD cultures of F. sporotrichioides strain 7-4-7
(a C-4 acetyltransferase Tri7- mutant) (20). 15-Acetyldeoxynivalenol
20 was isolated from YEPD cultures of F. graminearum strain B4-1
(a mutant with an extra copy of trichodiene synthase, TRI5, and a
regulatory protein, TRI6) (21). 3,15-Diacetyldeoxynivalenol 21 was
isolated from rice grain cultures of F. graminearum strain NA-8b01 (a
C-8 esterase Tri8- mutant) (19). 3-Acetyldeoxynivalenol 19 was
prepared by feeding deoxynivalenol 18 to liquid cultures of a Saccha-
romyces cereVisiae transformant expressing F. sporotrichioides genes
TRI101 (C-3 acetyltransferase) and TRI12 (a transporter) (15). Scir-
pentriol 9 and 15-monoacetoxyscirpenol 10 were made by hydrolysis
of 11 with 0.1 N sodium hydroxide. Deoxynivalenol 18 was made by
hydrolysis of 15-acetyldeoxynivalenol 20 with 0.1 N sodium hydroxide.
3,15-Diacetylnivalenol 24 was prepared by mixing 10 mg of 22 in 2
mL of pyridine and acetic anhydride (1:1) and identified with GC-MS
and NMR. Mass spectra were recorded with a Hewlett-Packard 5890
MATERIALS AND METHODS
Trichothecenes. Nivalenol 22 and 4-acetylnivalenol 23 were
purchased from Sigma Chemical Co. (St. Louis, MO). The remaining
compounds were isolated from wild-type or mutant strains of F.
sporotrichioides NRRL3299 or F. graminearum Z-3639 or were
produced by synthetic modification or by biotransformation. All
compounds were checked for greater than 95% purity with Gas
chromatography-mass spectrometry (GC-MS). Just prior to phytotox-
icity assays, each test compound was completely solubilized in acetone
to produce a 0.1 M master stock solution, which was serially diluted
in acetone to produce additional stock solutions. For phytotoxicity
assays, each stock solution was added to sterile distilled water to a
final concentration of 0.1% acetone by volume. Control assays contained
water with 0.1% acetone.
Trichodiene 1 was isolated from yeast extract-peptone-dextrose
liquid (YEPD) cultures of F. sporotrichioides mutant strain F15 (a
trichodiene oxygenase Tri4- mutant) (13). Isotrichotriol 2 was isolated
from rice grain cultures of F. sporotrichioides strain MB2972 (a C-15
acetyltransferase Tri3- mutant) (14). Isotrichodermol 3 was isolated
from YEPD cultures of F. sporotrichioides strain Tri101-3D (a C-3

Trichothecene Toxicity to Arabidopsis
J. Agric. Food Chem., Vol. 55, No. 16, 2007 6489
Figure 3. Percentage of dead leaves at concentrations of selected trichothecenes. Each data point is the mean of two replicate experiments. (A) T-2
series: 9 ) T-2 toxin, O ) 3-acetyl T-2 toxin, and 2 ) HT-2 toxin. (B) Scirpentriol series: 9 ) scirpentriol, O ) 15-acetoxyscirpenol, 2 )
4,15-diacetoxyscirpenol, and 3 ) triacetoxyscirpenol. (C) Deoxynivalenol series: 9 ) deoxynivalenol, O ) 3-acetyldeoxynivalenol, 2 )
15-acetyldeoxynivalenol, and 3 ) 3,15-diacetyldeoxynivalenol. (D) Nivalenol series: 9 ) nivalenol, O ) 4-acetylnivalenol, and 2 ) 3,15-diacetylnivalenol.
MSD spectrometer equipped with a GC 5890 with a DB-5MS column
(30 m × 0.25 mm film thickness). The oven temperature was
programmed from 120 °C at injection to 210 °C at 15 °C/min, held for
1 min, then heated to 260 °C at 5 °C/min, and held for 8 min with
helium as the carrier gas. NMR spectra were recorded with a Bruker
500 MHz instrument.
on sterile distilled water. Leaves were transferred from the leaf pool
into individual wells of 96 well, nonskirted polymerase chain reaction
(PCR) microplates (Fisher Scientific, Pittsburgh, PA). Each well
contained 250 µL of freshly prepared test solution in water with 0.1%
acetone by volume. The uncovered 96 well plates were returned to the
incubator in glass trays loosely covered with plastic wrap to maintain
humidity. Leaves were observed daily and scored for symptoms after
1 week. All compounds were tested twice in a dilution series of 0, 1.0,
5.0, 10, 50, and 100 µM. As necessary, highly toxic compounds were
tested twice again in a dilution series of 0, 0.1, 0.3, 1.0, and 3.0 µM.
For each of these tests, 96 leaves were treated and scored for each
dilution of each compound. For objectivity and uniformity, assays were
blind and all assays were scored by the same individual. Control leaves
remained healthy, green, and unshriveled throughout the duration of
the assay; apparently healthy control leaves (and leaves on intact plants)
occasionally developed areas of dark reddish coloration. Leaves were
individually rated for shriveling, chlorosis, and death on a scale of 1-4,
where 1 ) healthy leaves (green or reddish) with no visible shriveling
or chlorosis; 2 ) no shriveling, <50% chlorosis or necrosis; 3 )
shriveling, up to 90% chlorosis or necrosis; and 4 ) dead. For
calculating LD50 values, categories 3 and 4 were combined as “dead”
leaves. For calculating ED50 values, categories 2-4 were combined as
“symptomatic” leaves.
3,15-Diacetylnivalenol. m/z 396 (M+, 0.5), 336 (M-60, 2.5), 305
(55), 277 (81), 189 (100). 13C NMR (Bruker 500 MHz) in CDCl3: δ7.14
(C14); 15.28 (C16); 20.58 (C3 AcMe); 20.90 (C15 AcMe); 45.82 (C13);
49.47 (C5); 52.24 (C6); 61.70 (C15); 63.88 (C12); 69.56 (C11); 72.89
(C7); 78.14 (C4), 78.34 (C2); 83.15 (C3); 136.02 (C9); 137.88 (C10);
1
170.20 (C15Ac); 171.73 (C3Ac); 198.89 (C8). H NMR (Bruker 500
MHz) in CDCl3: δ1.17 (s, C14Me); 1.92 (s, C15Me); 1.93 (s, C16Me);
2.24 (s, C3AcMe); 3.10 (d, J ) 4.2 Hz, H13a); 3.12 (d, J ) 4.2 Hz,
H13b); 3.98 (d, J ) 4.7 Hz, H2); 4.27 (d, J ) 12.2 Hz, H15a); 4.40 (d,
J ) 12.2 Hz, H15b); 4.40 (d, J ) 5.9 Hz, H11); 4.43 (d, J ) 3.1 Hz,
H4); 4.87 (s, H7); 4.96 (dd, J ) 4.7 Hz, H3); 6.59 (dd, J ) 5.9 and 1.5
Hz, H10).
Phytotoxicity Assay. Seeds of A. thaliana Columbia (Col-4) ecotype
were obtained from Lehle Seeds (Round Rock, TX). For surface
disinfestation, seeds were placed on filter paper in a Buchner funnel,
rinsed for approximately 15 s each with 0.5% sodium hypochlorite and
sterile distilled water, and dried with suction. For stratification, seeds
were sown on the surface of Petri plates of Murashige-Skoog mineral
medium (Sigma, St. Louis, MO) adjusted to pH 5.9 in 1.5% agar, sealed
with Parafilm, and incubated for 3 days in the dark at 4 °C. Plates
were unsealed, placed in glass trays lined with moist toweling,
transferred to a biological incubator (Percival Scientific Inc., Perry,
IA), and incubated for 11 days under 24 h fluorescent lights at 25 °C.
Agar blocks with uniform, healthy plants were transferred to 24 well
tissue culture-treated multiwell plates (Becton Dickson Labware,
Franklin Lakes, NJ), and the 24 well plates were returned to the
incubator without covers for one more week with additional sterile
distilled water added to the wells as needed.
Statistical Analysis. Statistical analysis of the phytotoxicity assay
results was performed using OriginPro 7.5 software (OriginLab,
Northampton, MA). Dose-response data were analyzed by the logistic
fitting function with the upper and lower asymptotes fixed at 100 and
0, respectively. LD50 and ED50 values are reported as means ( standard
errors.
RESULTS
Phytotoxicity. Twenty-two trichothecenes and two trichoth-
ecene precursors were screened for toxicity to detached leaves
of A. thaliana. Toxicity was expressed as an LD50, the
concentration at which one-half of the leaves were dead (Table
Uniform, healthy leaves with petioles (approximately 1 cm long and
0.5 cm wide) were cut from the rosettes of 3 week old plants and floated

6490 J. Agric. Food Chem., Vol. 55, No. 16, 2007
Desjardins et al.
compounds differ in hydroxylation at a single site, C-4, C-8, or
C-15. Five pairs of compounds differ in both hydroxylation at
C-7 and the presence of a keto-group at C-8. The early
intermediates isotrichodermol 3 and its derivative 4 were quite
toxic, with LD50 values of 9-16 µM, but toxicity was reduced
>10-fold by C-15 hydroxylation (3 f 5, 4 f 6). C-4
hydroxylation had diverse effects: a slight increase in toxicity
(3,15-dideacetylcalonectrin 5 f 9); >5-fold increases in toxicity
(3-deacetylcalonectrin 7f 10 and 3,15-diacetyldeoxynivalenol
21f 24); or a >5-fold decrease in toxicity (deoxynivalenol 18
f 22). C-8 hydroxylation also had diverse effects: a slight
increase in toxicity (3,4,15-triacetoxyscirpenol 12 f14) or a
>10-fold decrease in toxicity (4,15-diacetoxyscirpenol 11f 13).
Last, concurrent oxygenation at both C-7 (hydroxyl) and C-8
(keto) had diverse effects: >5-fold increases in toxicity
(calonectrin derivatives 5 f 18, 6 f 19, 7 f 20) or at least
10-fold decreases in toxicity (calonectrin 8 f 21 and scirpentriol
9 f 22).
The effect of multiple oxygenations on toxicity was examined
by comparing three series of compounds that differ only in
sequential oxygenations at C-4, C-15, and C-7/C-8. In one series
from isotrichodermol to nivalenol, 3 f 5 f 18 f 22, toxicity
first decreased 10-fold, then increased 5-fold and decreased 10-
fold, resulting in a >10-fold decrease in toxicity overall. In the
second series, 3 f 5 f 9 f 22, toxicity first decreased >10-
fold, then remained very low. In the series from calonectrin to
3,15-diacetylnivalenol 8 f 21 f 24, toxicity was unchanged
overall.
Effect of Esterification on Phytotoxicity. The effect of
trichothecene esterification on toxicity was examined by com-
paring 18 pairs of compounds that differ in esterification at a
single site, C-3, C-4, C-8, or C-15 (Tables 1 and 2). C-3
acetylation had diverse effects: a 10-fold increase in toxicity
(3-deactylcalonectrin 7 f 8); at least 10-fold decreases in
toxicity (4,15-diacetoxyscirpenol 11 f 12, T-2 toxin 16 f 17,
and 15-acetyldeoxynivalenol 20 f 21); or little effect on toxicity
(isotrichodermol 3 f 4, 3,15-diacetylcalonectrin 5 f 6,
neosolaniol 13 f 14, and deoxynivalenol 18 f 19). C-4
acetylation increased toxicity (15-acetoxyscirpenol 10 f 11,
4-hydroxy-T-2 toxin 15 f 16, and nivalenol 22 f 23). C-8
esterification with isovalerate increased toxicity (neosolaniol 13
f 16 and its derivative 14 f 17). Last, C-15 esterification had
no effect on toxicity (3-acetyldeoxynivalenol 19 f 21) or
increased toxicity (two calonectrin derivatives 5 f 7, 6 f 8,
scirpentriol 9 f10, and deoxynivalenol 18 f 20).
The effect of multiple esterifications on toxicity was examined
by comparing two series of compounds that differed only in
sequential esterifications at C-3, C-4, and C-15. Among scir-
pentriol and its derivatives 9 f 10 f 11f 12, toxicity first
increased >50-fold and then decreased, resulting in an 8-fold
increase overall. Among deoxynivalenol and its derivatives 18
f 19 f 21, toxicity was unchanged.
Figure 4. Stepwise changes in toxicity of T-2 toxin 16 (right side) and
related trichothecenes and of nivalenol 22 and related trichothecenes (left
side). Nodes represent compounds, nodes with numbers represent
compounds described in Table 1, and nodes without numbers represent
compounds not tested in this study. Arrows from left to right represent
single oxygenations. Arrows from top to bottom represent single esteri-
fications: Thin black arrows indicate C-3, broad white arrows indicate
C-15, and broad black arrows indicate C-4 and C-8. Plus (
+) and minus
(
−) signs on an arrow indicate the direction of the change in toxicity, as
the ratio of the substrate toxicity to product toxicity (LD50). A plus sign
indicates a two-fold or greater increase in toxicity, and a minus sign
indicates a two-fold or greater decrease in toxicity from substrate to
product. Arrows represent stepwise changes between compounds but not
necessarily the major endogenous biosynthetic pathways.
1) and (Figure 2). For less toxic compounds, toxicity was also
expressed as an ED50, the concentration at which one-half of
the leaves showed symptoms of chlorosis (Table 2). At 100
µM, the highest concentration tested, the trichothecene precursor
trichodiene 1 was essentially nontoxic, producing mild symp-
toms of chlorosis in 10% and death in 3% of the leaves.
Isotrichotriol 2, an epoxide of 1, was slightly more toxic,
producing mild symptoms of chlorosis in 10% and death in 10%
of the leaves treated at 10, 50, or 100 µM, with no dose-
response effect. In more than 50 independent tests, a mean of
4% (range ) 0-9%) of control leaves treated with water showed
mild symptoms of chlorosis and fewer than 1% of control leaves
were dead.
All 22 trichothecenes tested were toxic to A. thaliana, but
they varied more than 200-fold in toxicity as measured by LD50
and ED50 (Tables 1 and 2). The nine most toxic compounds
(LD50 < 10 µM) (3, 8, 10, 11, 15-17, 20, and 24) were
structurally diverse and included highly oxygenated compounds
such as 3,15-diacetylnivalenol 24 and T-2 toxin and its
derivatives 15-17, along with the less-oxygenated intermediates
isotrichodermol 3 and calonectrin 8. The eight moderately toxic
compounds (4, 7, 12-14, 18, 19, and 21) were similar in
toxicity, with LD50 values of 13-28 µM. The five least toxic
compounds (LD50 > 80 µM) (5, 6, 9, 22, and 23) included
highly oxygenated nivalenol 22 and its derivative 23, as well
as less-oxygenated calonectrin 5 and its derivative 6. Even at
100 µM, 6 produced symptoms in fewer than half of the leaves
and 22 produced symptoms in fewer than 10% of the leaves.
Dose-response curves for toxicity (percent dead) of four
agriculturally important trichothecenes, 4,15-diacetoxyscirpenol
11, T-2 toxin 16, deoxynivalenol 18, nivalenol 22, and their
acetylated derivatives are shown in Figure 3.
The overall effect of oxygenation and esterification on toxicity
of the 22 trichothecenes in this study is summarized in network
diagrams of T-2 toxin 16 and related trichothecenes (Figure 4,
right side) and of nivalenol 22 and related trichothecenes (Figure
4, left side). Nodes represent compounds, and arrows connecting
nodes represent stepwise oxygenations and esterifications. Two-
fold or greater changes in toxicity, as assessed by LD50 values,
are indicated by a plus sign for an increase in toxicity and a
minus sign for a decrease in toxicity from substrate to product.
DISCUSSION
Effect of Oxygenation on Phytotoxicity. The effect of
trichothecene oxygenation on toxicity was examined by com-
paring 13 pairs of compounds (Tables 1 and 2). Eight pairs of
For a comprehensive and quantitative analysis of structure-
activity relationships of trichothecenes biosynthesized by Fusar-

Trichothecene Toxicity to Arabidopsis
J. Agric. Food Chem., Vol. 55, No. 16, 2007 6491
ium, we used an A. thaliana detached leaf assay to compare
phytotoxicity of 24 biosynthetic pathway precursors, intermedi-
ates, and end products. All trichothecenes and intermediates with
an epoxide group were toxic to A. thaliana; only the nonoxy-
genated precursor trichodiene 1 was nontoxic. Three previous
studies have demonstrated toxicity of trichothecenes to the Col-0
ecotype of A. thaliana, but none of these studies included a
large number of compounds or a dose-response analysis. In
the first study (12), deoxynivalenol 18 at 50 µM inhibited seed
germination on an agar medium. In the second study (11),
lesions developed in leaves injected with 1 µM diacetoxyscir-
penol 11, HT-2 toxin 15, and T-2 toxin 16 or with 10 µM 18.
In the third study (9), lesions developed in leaves injected with
75 µM 18. Our dose-response assay also found that 11, 15,
and 16 were similar in toxicity (LD50 ) 0.5-1.5 µM) and that
all were more toxic than 18 (LD50 ) 23 µM).
Trichothecene biosynthesis is characterized by a generally
ordered sequence of oxygenations catalyzed by cytochrome
P450s followed by esterifications. First, trichodiene 1 is
oxygenated, and then, isotrichotriol 2 cyclizes to isotrichodermol
3, which has a C-3 hydroxyl group. After acetylation of 3 to
isotrichodermin 4, C-15 oxygenation and esterification form
calonectrin and its derivatives 5-8. After calonectrin, the
biosynthetic pathway diverges. One pathway leads to nivalenol
and its derivatives 22-24 via C-4, C-7, and C-8 oxygenations
and esterifications. Another pathway leads to T-2 toxin and its
derivatives 15-17 via C-4 and C-8 oxygenations and esterifi-
cations (1). Our compound library has made possible a survey
of the trichothecene biosynthetic pathways from trichodiene 1
to T-2 toxin 16 and nivalenol 22. During trichothecene biosyn-
thesis, there is a considerable increase in structural complexity.
In our A. thaliana assay, however, this increased structural
complexity was not necessarily associated with an increase in
phytotoxicity. Investigation into the conformational preferences
of selected trichothecenes from this study is in progress.
ynivalenol 19, was highly toxic to A. thaliana. Conversely, 18
and 19 were more toxic than 11 to wheat.
Trichothecene toxicity to the unicellular green alga Chlamy-
domonas reinhardtii was previously assessed by measuring the
change in cell density at a single dose of 80 µM (5). In that
study, the effect of C-3 acetylation on toxicity was examined
by comparing seven pairs of compounds that differ only at
C-3: 3-8, 11-14, and 16-19. In C. reinhardtii, C-3 acetylation
decreased toxicity of five of seven compounds tested as
follows: 3 f 4, 7 f 8, 11 f 12, 16 f 17, and 18 f 19. In
contrast, in A. thaliana, C-3 acetylation decreased toxicity of
only two of these compounds, 4,15-diacetoxyscirpenol 11 f
12 and T-2 toxin 16 f 17. Both 11 and 16 were highly toxic
to C. reinhardtii, A. thaliana, and wheat, but deoxynivalenol
18 was less toxic to C. reinhardtii and A. thaliana than to wheat.
The relationship between phytotoxicity and chemical structure
is complicated by the fact that toxicity in plant systems can be
affected by complex interactions, including uptake and metabo-
lism, as well as intrinsic activity. The extent of oxygenation
and esterification may not only influence the recognition and
stability of particular trichothecenes but also have an effect on
their absorption, distribution, and possible metabolism in A.
thaliana. Documented metabolism of trichothecenes by plants
includes the 3-glycosylation of 18 by maize callus cultures, the
deacetylation of 11 by potato tuber tissue, and the deacylation
and hydroxylation of 16 by Baccharis species (1). Initial
observations of the reduced phytotoxicity of C-3 acetylated
trichothecenes in C. reinhardtii and yeast systems initiated plant
biotechnology programs to transfer TRI101, the gene for
trichothecene 3-O-acetyltransferase, from Fusarium into cereal
crop plants (15, 25). The rationale for this approach was that
reduction of trichothecene phytotoxicity might lead to reduction
of disease symptoms in plant-Fusarium interactions in which
trichothecenes function as virulence factors (1). Expression of
TRI101 in tobacco and rice conferred some resistance to
phytotoxicity of 11 and 18, and expression of TRI101 conferred
some protection against infection by F. graminearum in wheat
but not in barley (26-29). Furthermore, 3-O-glycosylation of
18 has been strongly associated with a quantitative trait locus
that confers resistance to the spread of F. graminearum in wheat
(30). In addition, constitutive overexpression of a trichothecene
C-3 glucosyltransferase from A. thaliana decreased sensitivity
of the plant to 18 (12).
Although T-2 toxin 16, a structurally complex, pathway end
product, was the most toxic compound, isotrichodermol 3, the
first and simplest trichothecene in the pathway, also was quite
toxic. In fact, 3 was more toxic than 67% of the 21 more-highly
oxygenated or esterified compounds that occur later in the
biosynthetic pathway. To our knowledge, this is the first study
of the toxicity of 3 or isotrichodermin 4 in any higher plant or
animal system. Isotrichodermol 3 appears to contain the minimal
structural features necessary for phytotoxicity of Fusarium
trichothecenes, at least to A. thaliana. In this regard, it should
be noted that only two trichothecene biosynthetic pathway
structural genes should be needed for the synthesis of 3. TRI5
encodes trichodiene synthase, which cyclizes farnesyl pyro-
phosphate to produce trichodiene 1, and TRI4 encodes a
cytochrome P450 that oxygenates 1 to produce 2, which cyclizes
to 3 (1, 22). TRI4 and TRI5 flank the regulatory gene TRI6 at
the center of the trichothecene biosynthetic gene cluster in F.
graminearum and F. sporotrichioides, possibly comprising an
evolutionarily ancient module of trichothecene biosynthetic
genes.
Expression of trichothecene detoxification genes in cereal
crops has the potential to reduce levels of Fusarium head blight
and to decrease trichothecene contamination of cereal grains.
However, our finding of the high phytotoxicity of isotricho-
dermol 3 and isotrichodermin 4 indicates that early intermediates
of the trichothecene biosynthetic pathway are as biologically
active as many late pathway intermediates, at least in the
dicotyledenous plant A. thaliana. The trichothecene early
intermediates 1-8 remain to be tested for phytotoxicity to wheat
and other monocotyledenous cereal plants. A range of studies
indicates that modifications of extraskeletal hydroxyl and ester
groups are likely to be reversible in vivo, in plants and in
animals, and to result in only partial detoxification and
incomplete resistance (31). For complete detoxification of
trichothecenes, discovery of genes that confer the ability to
remove the 12,13-epoxide should be a high priority. Such de-
epoxidation activity has been demonstrated in bacteria but
apparently has not yet been successfully transferred to plants (31).
To our knowledge, trichothecene toxicities have been com-
pared in dose-response assays in only two plant species, wheat
and the present study of A. thaliana (6-8, 23, 24). For three of
the six compounds (11, 16, 18-20, and 22) tested in both
systems, relative toxicity to A. thaliana was similar to relative
toxicity to wheat. For example, toxicity of T-2 toxin 16 and
15-acetyldeoxynivalenol 20 was high and toxicity of nivalenol
22 was low in both systems. On the other hand, 4,15-
diacetoxyscirpenol 11, but not deoxynivalenol 18 or 3-acetyldeox-
In summary, we have shown that A. thaliana provides a
reproducible and quantitative assay for the characterization of
trichothecene structure-activity relationships. The phytotoxicity

6492 J. Agric. Food Chem., Vol. 55, No. 16, 2007
Desjardins et al.
(15) McCormick, S. P.; Alexander, N. J.; Trapp, S. E.; Hohn, T. M.
Disruption of TRI101, the gene encoding trichothecene 3-O-
acetyltransferase, from Fusarium sporotrichioides. Appl. EnViron.
Microbiol. 1999, 65, 5252-5256.
(16) McCormick, S. P.; Hohn, T. M. Accumulation of trichothecenes
in liquid cultures of a Fusarium sporotrichioides mutant lacking
a functional trichothecene C-15 hydroxylase. Appl. EnViron.
Microbiol. 1997, 63, 1685-1688.
(17) McCormick, S. P.; Harris, L. J.; Alexander, N. J.; Ouellet, T.;
Saparno, A.; Allard, S.; Desjardins, A. E. Tri1 in Fusarium
graminearum encodes a P450 oxygenase. Appl. EnViron. Mi-
crobiol. 2004, 70, 2044-2051.
(18) Hohn, T. M.; McCormick, S. P.; Desjardins, A. E. Evidence for
a gene cluster involving trichothecene pathway biosynthetic genes
in Fusarium sporotrichioides. Curr. Genet. 1993, 24, 291-295.
(19) McCormick, S. P.; Alexander, N. J. Fusarium Tri8 encodes a
trichothecene C-3 esterase. Appl. EnViron. Microbiol. 2002, 68,
2959-2964.
assay could be completed within 1 week, and the detached leaves
remained healthy for the duration of the experiment. A. thaliana
also provides a genetically tractable plant system for gene
discovery and a large collection of ecotypes, some of which
have already been shown to differ in susceptibility to F.
graminearum and to fungal toxins in detached leaf assays
(9, 10). Thus, A. thaliana provides a good model system for
the identification of novel plant traits and genes for trichothecene
detoxification and resistance.
ACKNOWLEDGMENT
We thank Stephanie Folmar, Kim MacDonald, and Nathaniel
Puent for excellent technical assistance and Steve Prather and
Don Fraser for scientific illustration.
(20) Brown, D. W.; McCormick, S. P.; Alexander, N. J.; Proctor, R.
H.; Desjardins, A. E. A genetic and biochemical approach to
study trichothecene diversity in Fusarium sporotrichioides and
Fusarium graminearum. Fungal Genet. Biol. 2001, 32, 121-
133.
(21) Chen, L.; McCormick, S. P.; Hohn, T. M. Altered regulation of
15-acetyldeoxynivalenol production in Fusarium graminearum.
Appl. EnViron. Microbiol. 2000, 66, 2062-2065.
(22) McCormick, S. P.; Alexander, N. J.; Proctor, R. H. Fusarium
sporotrichioides Tri4 encodes a multifunctional oxygenase
required for trichothecene biosynthesis. Can. J. Microbiol. 2006,
52, 636-642.
(23) Bruins, M. B. M.; Karsai, I.; Schepers, J.; Snijders, C. H. A.
Phytotoxicity of deoxynivalenol to wheat tissue with regard to
in vitro selection for Fusarium head blight resistance. Plant Sci.
1993, 94, 195-206.
(24) Shimada, T.; Otani, M. Effects of Fusarium mycotoxins on the
growth of shoots and roots at germination in some Japanese
wheat cultivars. Cereal Res. Commun. 1990, 18, 229-231.
(25) Kimura, M.; Kaneko, I.; Komiyama, M.; Takatsuki, A.; Koshina,
H.; Yoneyama, K.; Yamaguchi, I. Trichothecene 3-O-acetyl-
transferase protects both the producing organism and transformed
yeast from related proteins. J. Biol. Chem. 1998, 273, 1654-
1661.
(26) Okubara, P. A.; Blechl, A. E.; McCormick, S. P.; Alexander,
N. J.; Dill-Macky, R.; Hohn, T. M. Engineering deoxynivalenol
metabolism in wheat through the expression of a fungal
trichothecene acetyltransferase gene. Theor. Appl. Genet. 2002,
106, 74-83.
(27) Muhitch, M. J.; McCormick, S. P.; Alexander, N. J.; Hohn, T.
M. Transgenic expression of the TRI101 or PDR5 gene increases
resistance of tobacco to the phytotoxic effects of the trichothecene
4,15-diacetoxyscirpenol. Plant Sci. 2000, 157, 201-207.
(28) Ohsato, S.; Ochiai-Fukuda, T.; Nishiuchi, T.; Takahashi-Ando,
N.; Koizumi, S.; Hamamoto, H.; Kudo, T.; Yamaguchi, I.;
Kimura, M. Transgenic rice plants expressing trichothecene 3-O-
acetyltransferase show resistance to the Fusarium phytotoxin
deoxynivalenol. Plant Cell Rep. 2006, 26, 531-538.
(29) Manoharan, M.; Dahleen, L. S.; Hohn, T. M.; Neate, S. M.; Yu,
X.; Alexander, N. J.; McCormick, S. P.; Bregitzer, P.; Schwarz,
P. B.; Horsley, R. D. Expression of 3-OH trichothecene acetyl-
transferase in barley (Hordeum Vulgare L.) and effects on
deoxynivalenol. Plant Sci. 2006, 171, 699-796.
LITERATURE CITED
(1) Desjardins, A. E. Fusarium Mycotoxins, Chemistry, Genetics,
and Biology; American Phytopathological Society Press: St.
Paul, MN, 2006.
(2) Bondy, G. S.; McCormick, S. P.; Beremand, M. N.; Pestka, J.
J. Murine lymphocyte proliferation impaired by substituted
neosolaniols and calonectrinssFusarium metabolites associated
with trichothecene biosynthesis. Toxicon 1991, 29, 1107-
1113.
(3) Rocha, O.; Ansari, K.; Doohan, F. M. Effects of trichothecene
mycotoxins on eukaryotic cells: A review. Food Addit. Contam.
2005, 22, 369-378.
(4) Brian, P. W.; Dawkins, A. W.; Grove, J. F.; Hemming, H. G.;
Lowe, R.; Morris, G. L. F. Phytotoxic compounds produced by
Fusarium equiseti. J. Exp. Bot. 1961, 12, 1-12.
(5) Alexander, N. J.; McCormick, S. P.; Zeigenhorn, S. L. Phyto-
toxicity of selected trichothecenes using Chlamydomonas rein-
hardtii as a model system. Nat. Toxins 1999, 7, 265-269.
(6) Eudes, F.; Comeau, A.; Rioux, S.; Collin, J. Phytotoxicite´ de
huit mycotoxines associe´es a` la fusariose de l’e´pi chez de ble´.
Can. J. Plant Pathol. 2000, 22, 286-292.
(7) Wakulinski, W. Phytotoxicity of the secondary metabolites of
fungi causing wheat head fusariosis (head blight). Acta Physiol.
Plant 1989, 11, 301-306.
(8) Wang, Y. Z.; Miller, J. D. Effects of Fusarium graminearum
metabolites on wheat tissue in relation to Fusarium head blight
resistance. J. Phytopathol. 1988, 122, 118-125.
(9) Chen, X.; Steed, A.; Harden, C.; Nicholson, P. Characterization
of Arabidopsis thaliana-Fusarium graminearum interactions and
identification of variation in resistance among ecotypes. Mol.
Plant Pathol. 2006, 7, 391-403.
(10) Lorang, J. M.; Carkaci-Salli, N.; Wolpert, T. J. Identi-
fication and characterization of victorin sensitivity in Arab-
idopsis thaliana. Mol. Plant-Microbe Interact. 2004, 17, 577-
582.
(11) Nishiuchi, T.; Masuda, D.; Nakashita, H.; Ichimura, K.; Shi-
nozaki, K.; Yoshida, S.; Kimura, M.; Yamaguchi, I.; Yamaguchi,
K. Fusarium phytotoxin trichothecenes have an elicitor-like
activity in Arabidopsis thaliana, but the activity differed
significantly among their molecular species. Mol. Plant-Microbe
Interact. 2006, 19, 512-520.
(12) Poppenberger, B.; Berthiller, F.; Lucyshyn, D.; Sieberer, T.;
Schuhmacher, R.; Krska, R.; Kuchler, K.; Glo¨ssl, J.; Luschnig,
C.; Adam, G. Detoxification of the Fusarium mycotoxin deox-
ynivalenol by a UDP-glucosyltransferase from Arabidopsis
thaliana. J. Biol. Chem. 2003, 278, 47905-47914.
(13) Hohn, T. M.; Desjardins, A. E.; McCormick, S. P. The Tri4 of
Fusarium sporotrichioides encodes a cytochrome P450 mo-
nooxygenase involved in trichothecene biosynthesis. Mol. Gen.
Genet. 1995, 248, 95-102.
(30) Lemmens, M.; Scholz, U.; Berthiller, F.; Dall’Asta, C.; Koutnik,
A.; Schumacher, R.; Adam, G.; Buerstmayr, H.; Mesterhazy,
A.; Krska, R.; Ruckenbauer, P. The ability to detoxify the
mycotoxin deoxynivalenol colocalizes with a major quantitative
trait locus for Fusarium head blight resistance in wheat. Mol.
Plant-Microbe Interact. 2005, 18, 1318-1324.
(31) Karlovsky, P. Biological detoxification of fungal toxins and its
use in plant breeding. Nat. Toxins 1999, 7, 1-23.
(14) McCormick, S. P.; Taylor, S. L.; Plattner, R. D.; Beremand, M.
N. New modified trichothecenes accumulated in solid culture
by mutant strains of Fusarium sporotrichioides. Appl. EnViron.
Microbiol. 1989, 55, 2195-2199.
Received for review March 29, 2007. Revised manuscript received May
29, 2007. Accepted May 29, 2007.
JF0709193