Integration of metabolomics and in vitro metabolism assays for investigating the stereoselective transformation of triadimefon in rainbow trout

Article abstract of DOI:10.1002/chir.20725

Triadimefon is a systemic agricultural fungicide of the triazole class whose major metabolite, triadimenol, also a commercial fungicide, provides the majority of the actual fungicidal activity, i.e., inhibition of steroid demethylation. Both chemicals are chiral: triadimefon has one chiral center with two enantiomers while its enzymatic reduction to triadimenol produces a second chiral center and two diastereomers with two enantiomers each. All six stereoisomers of the two fungicides were separated from each other using a chiral BGB-172 column on a GC-MS system so as to follow stereospecificity in metabolism by rainbow trout hepatic microsomes. In these microsomes the S-(+) enantiomer of triadimefon was transformed to triadimenol 27% faster than the R-(-) enantiomer, forming the four triadimenol stereoisomers at rates different from each other. The most fungi-toxic stereoisomer (1S,2R) was produced at the slowest rate; it was detectable after 8 h, but below the level of method quantitation. The triadimenol stereoisomer ratio pattern produced by the trout microsomes was very different from that of the commercial triadimenol standard, in which the most rat-toxic pair of enantiomers (known as Diastereomer A ) is about 85% of the total stereoisomer composition. The trout microsomes produced only about 4% of Diastereomer A . Complementary metabolomic studies with NMR showed that exposure of the separate triadimefon enantiomers and the racemate to rainbow trout for 48 h resulted in different metabolic profiles in the trout liver extracts, i.e., different endogenous metabolite patterns that indicated differences in effects of the two enantiomers.

Full text of DOI:10.1002/chir.20725

CHIRALITY 22:183–192 (2010)
Integration of Metabolomics and In Vitro Metabolism Assays for
Investigating the Stereoselective Transformation of Triadimefon
in Rainbow Trout
JOHN F. KENNEKE,1 DREW R. EKMAN,1 CHRIS S. MAZUR,1 BRAD J. KONWICK,2 AARON T. FISK,3
1
JIMMY K. AVANTS,4 AND A. WAYNE GARRISON
*
1U.S. Environmental Protection Agency, National Exposure Research Laboratory, Ecosystems Research Division, Athens, Georgia
2Student Services Authority Contract, U.S. EPA, Athens, Georgia
3Great Lakes Institute for Environmental Research, University of Windsor, Ontario, Canada
4Senior Service America, U.S. EPA, Athens, Georgia
ABSTRACT
Triadimefon is a systemic agricultural fungicide of the triazole class
whose major metabolite, triadimenol, also a commercial fungicide, provides the majority
of the actual fungicidal activity, i.e., inhibition of steroid demethylation. Both chemicals
are chiral: triadimefon has one chiral center with two enantiomers while its enzymatic
reduction to triadimenol produces a second chiral center and two diastereomers with
two enantiomers each. All six stereoisomers of the two fungicides were separated from
each other using a chiral BGB-172 column on a GC-MS system so as to follow stereo-
specificity in metabolism by rainbow trout hepatic microsomes. In these microsomes
the S-(1) enantiomer of triadimefon was transformed to triadimenol 27% faster than the
R-(2) enantiomer, forming the four triadimenol stereoisomers at rates different from
each other. The most fungi-toxic stereoisomer (1S,2R) was produced at the slowest
rate; it was detectable after 8 h, but below the level of method quantitation. The triadi-
menol stereoisomer ratio pattern produced by the trout microsomes was very different
from that of the commercial triadimenol standard, in which the most rat-toxic pair of
enantiomers (known as ‘‘Diastereomer A’’) is about 85% of the total stereoisomer compo-
sition. The trout microsomes produced only about 4% of ‘‘Diastereomer A’’. Complemen-
tary metabolomic studies with NMR showed that exposure of the separate triadimefon
enantiomers and the racemate to rainbow trout for 48 h resulted in different metabolic
profiles in the trout liver extracts, i.e., different endogenous metabolite patterns that
indicated differences in effects of the two enantiomers. Chirality 22:183–192,
C
VV
2010.
2009 Wiley-Liss, Inc.
KEY WORDS: triadimefon; triadimenol; triazoles; conazoles; rainbow trout; metabolism;
metabolomics; stereoisomers; enantiomers; diastereomers; stereoselectivity;
exposure
INTRODUCTION
produced from triadimefon in different relative amounts by
plants and fungi, and the proportions of each stereoisomer
may differ depending upon the species.3,4 A similar meta-
bolic pathway (i.e., formation of triadimenol from triadime-
fon) is followed in animals1; to the best of our knowledge,
however, stereoselective formation of triadimenol has not
been reported for animals.
Triadimefon and triadimenol are suspected of disrupting
steroidogenesis in animals, are neurotoxins, and have
been shown to cause tumors in rodents.5 Since triadime-
fon and triadimenol are both used as agricultural fungi-
Triadimefon (Fig. 1) is a 1,2,4-triazole fungicide used for
the control of powdery mildews and fungi on fruits, vegeta-
bles, turf grasses, and other agricultural crops. It is a sys-
temic fungicide that acts by inhibiting steroid demethyla-
tion1 and is enzymatically reduced in plants, soils, and
fungi to the more active metabolite, triadimenol,1 which is
also used as an agricultural fungicide. Triadimefon has a
single chiral center and thus exists as two enantiomers.
The metabolic transformation of triadimefon to triadimenol
involves the reduction of the carbonyl group to an alcohol,
resulting in the formation of a second chiral center2
(Fig. 1). Thus, triadimenol consists of two diastereomers:
A [enantiomers A1 (1R,2S) and A2 (1S,2R)] and B [enan-
tiomers B1 (1R,2R) and B2 (1S,2S)], for a total of four
stereoisomers. Diastereomers A and B, and the four ste-
reoisomers that make up these diastereomers, are each
C
*Correspondence to: A. Wayne Garrison, U.S. Environmental Protection
Agency, 960 College Station Road, Athens, GA.
E-mail: garrison.wayne@epa.gov
Received for publication 30 September 2008; Accepted 10 February 2009
DOI: 10.1002/chir.20725
Published online 5 May 2009 in Wiley InterScience
(www.interscience.wiley.com).
VV
2009 Wiley-Liss, Inc.

184
KENNEKE ET AL.
Fig. 1. The metabolic transformation of triadimefon to triadimenol. The reduction of a carbonyl group to an alcohol yields a second chiral center and
four stereoisomers.
cides, there is concern regarding potential human expo- to paraoxon, which acts as a more potent cholinesterase
sure, as well as concern about their residues in the envi- inhibitor.10
ronment. Triadimenol, for example, has been detected in
While metabolism studies provide information on xeno-
water samples from ditches and streams at concentrations biotic metabolites, metabolomics, a relatively recent addi-
of up to a few micrograms per liter.6 Thus, it is important tion to the array of molecular techniques used to assess
to understand the fate of these compounds in the environ- chemical exposures, uses various types of analytical instru-
ment as well as the impact of their exposure to humans mentation to detect and identify changes in endogenous
and various wildlife species. In addition, the enantioselec- metabolites in selected tissues and/or biofluids of various
tive loss of triadimefon to form triadimenol during meta- organisms after exposure.11 Changes in these endogenous
bolic transformation is of concern. Established data show metabolites can be the direct result of exposure to a xeno-
that a wide variety of chiral pesticides occur nonracemi- biotic, and can be used to assess xenobiotic exposure
cally in various environmental compartments, are trans- even after the chemical stressor is no longer present. In
formed enantioselectively in environmental microcosms, the environmental arena, metabolomics involves a compar-
or produce enantioselective toxic effects on various organ- ison of endogenous metabolite patterns in tissues or fluids
isms.7 For similar reasons, since it is known that triadime- from organisms exposed to some environmental stressor,
nol diastereomer A is 10 times more acutely toxic to rats such as a toxic chemical, with those from unexposed con-
(oral LD50) than is diastereomer B,8 the possible stereose- trols, using advanced data analysis techniques. This
lective formation of triadimenol is an important issue for approach allows graphical depiction of any significant dif-
both human health and ecological risk assessment.
ferences in the patterns, which often provides information
In vivo and in vitro metabolism studies are useful for about toxicity mechanisms, pathways, and possible bio-
understanding the fate of a xenobiotic in an organism. In markers of exposure. Among those analytical techniques
vitro experiments using microsomal material from various most commonly used for metabolomics studies, nuclear
organs (e.g., liver) can shed light on such metabolism magnetic resonance (NMR) spectroscopy provides a
events with a minimum of expense and provide valuable powerful approach for obtaining information-rich spectra
information on toxicant exposure.2,9 In certain cases the for a wide variety of tissues and biofluids.12
chemical that enters the organism is metabolized to a
Rainbow trout (Onchorhynchus mykiss) have historically
more toxic form inside the organism. For example, the been among one of the most frequently tested freshwater
organophosphorus insecticide parathion is biotransformed fish with respect to toxicology and physiology for the
Chirality DOI 10.1002/chir

STEREOSELECTIVE TRANSFORMATION OF TRIADIMEFON
185
Preparative Separation of Triadimefon Enantiomers
purpose of assessing ecological risk from exposure to spe-
cific xenobiotics.13 Here, we report results of investiga-
tions of the in vitro stereoselective metabolism of triadime-
fon to triadimenol in rainbow trout liver microsomes, as
well as results of complementary NMR metabolomic
experiments involving exposure of the separate triadime-
fon enantiomers and the racemate to rainbow trout in vivo.
To our knowledge, this is the first time that xenobiotic me-
tabolism has been used to invoke the influence of chirality
on endogenous metabolite distributions as measured
through metabolomics.
Triadimefon enantiomers were separated by Chiral
Technologies (West Chester, PA) using preparative super-
critical fluid chromatography. Racemic triadimefon (2 g)
was applied to a Chiralpak1ADH1 (Chiral Technologies)
preparative column (3.0 cm i.d. 3 25 cm long). Elution
was by CO2:MeOH/85%:15% at 258C, flow rate was 2.0 ml/
min, and detection was by UV at 275 nm. Preparative
yields were 0.89 and 0.90 g for the 1st and 2nd eluting
enantiomer, respectively, with an enantiomeric excess of
>99% for each. The optical rotation in methanol of the first
eluting enantiomer was (2) and that of the second was
(1) as determined by Chiral Technologies using a PDR
Chiral Advanced Laser Polarimeter under ambient condi-
tions.
EXPERIMENTAL
Reagents and Chemicals
Trout Microsome Preparation
Triadimefon and triadimenol were obtained in 99.4 and
96.4% purity, respectively, from the U.S. Environmental
Protection Agency National Pesticide Standard Repository
(Ft. Meade, MD). Nicotinamide adenine dinucleotide
phosphate (NADP), glucose-6-phosphate, glucose-6-phos-
phate dehydrogenase, magnesium chloride, phosphate,
and trizma buffers were purchased from Sigma Chemical
(St. Louis, MO). Methanol and acetonitrile were from
Fisher Chemicals (Fair Lawn, NJ) and were of analytical
grade. Perchloric acid (60–62%) was obtained from J.T.
Baker (Phillipsburg, NJ). All reagents used for metabolo-
mic analyses were purchased from Sigma-Aldrich (St.
Louis, MO) and were of ACS reagent grade. Reagent
water for all experiments was produced by a Barnstead
Nanopure Infinity water purification system (Thermo Sci-
entific, USA). Solvents for the various experiments were
pesticide grade.
Juvenile rainbow trout of mixed gender (Lake Burton
Fish Hatchery, Clarkesville, GA) were held for a 10-day
acclimation period in 50 l fiberglass aquaria with carbon-fil-
tered, recirculating, dechlorinated tap water chilled to
128C. Trout livers were excised upon cervical dislocation,
washed, and coarsely minced in an ice-chilled 1.15% KCl
solution. The KCl solution was then drained and a wet
weight of liver material was recorded. The liver was trans-
ferred to a prechilled glass homogenizer where 4 ml vol-
umes of a 0.25 M sucrose solution were added per gram
of liver weight. The sample was homogenized on ice and
then transferred to an ice-chilled centrifuge tube. The liver
homogenate was centrifuged in a high speed centrifuge
(Beckman) at 8000g for 20 min at 48C. Next, the superna-
tant was carefully decanted and ultracentrifuged at
100,000g for 60 min at 48C. The supernatant was then dis-
carded and the microsome pellet was resuspended with a
1:1 volume per gram of original wet weight with a pH 8.0,
0.066 M trizma base buffer containing 0.25 M sucrose and
5.4 mM EDTA. Trout liver microsome total protein content
was determined using the Bradford assay and microsomal
material was stored at 2808C until use.
GC-MS Analysis of Triadimefon
and Triadimenol Enantiomers
GC-MS analysis (e.g., of the fungicide standards and the
microsome metabolism extracts) was by use of a Hewlett-
Packard 5973 mass spectrometer linked to a 6890 gas
chromatograph equipped with a BGB 172 (BGB Analytik
AG, Switzerland) chiral column. Column description: 30 m
Trout Microsome Incubation Procedure
Incubations of microsome samples with triadimefon
3 0.25 mm ID 3 0.25 lm film thickness; chiral stationary were conducted in microcentrifuge tubes placed in a cool-
phase, 20% tert-butyldimethylsilylated-B-cyclodextrin. GC ing block at 118C. Microsomal suspensions were prepared
conditions were as follows: injection, splitless at temp. at a final concentration of 0.125 mg microsome protein/ml
2758C; column temp. program, 150–2208C at 48/min, fol- in trizma buffer (100 mM) at pH 8.0. Next, triadimefon
lowed by temp. hold for 60 min; helium gas flow, 1.5 ml/ stock solutions (in acetonitrile and stored in amber vials at
min; MS inlet temp., 2758; MS source temp., 2308; and 48C) were added to achieve a final substrate concentration
fragmentation voltage, 70 eV. Sample injection volume was level of 20–40 lM while not exceeding 1% organic solvent.
1 ll. Detection was by selected ion monitoring (SIM); SIM The system was vortexed and allowed to stand for 10 min
ions were m/z 181, 208, and 210 for triadimefon and m/z in the cooling block to ensure temperature equilibration.
128 and 168 for triadimenol. Quantification was by com- In a separate vial, a NADPH-regenerating system (NRS)
parison of enantiomer peak areas to standards of similar was preincubated at 118C for 10 min with 100 mM trizma
concentration analyzed the same day; quality control buffer at pH 8.0 and a final reaction concentration contain-
included analysis of at least one standard per day as well ing 0.5 mM NADP, 7 mM glucose-6-phosphate, 1.25 mM
as analysis of a standard before and after each 10 sam- MgCl2, and 1.5 U of glucose-6-phosphate dehydrogenase.
ples. These standards were referenced to a standard Incubation was initiated by the addition of 250 ll of the
curve of triadimefon or triadimenol enantiomer peak NRS to the microsomal suspension, and the low tempera-
area versus concentration.
ture trout microsome assays were conducted for 0–48 h.
Chirality DOI 10.1002/chir

186
KENNEKE ET AL.
NMR Instrumentation and Measurements
Each assay sample, taken at selected time points, was ter-
minated with a 1:1 addition of MTBE, vortexed, and imme-
diately placed on ice. The samples were centrifuged at 48C
for 10 min at 10,600g, and the MTBE was transferred to
seal-cap vials containing a polyspring microinsert for sub-
sequent analysis of triadimefon and triadimenol.
All tissue extract NMR data were acquired at 258C on a
Varian Inova 800 NMR spectrometer (799.7 MHz, 1H)
using a cryogenic triple resonance probe. The residual
water signal was suppressed by presaturation. Acquisition
parameters for the 1D 1H experiments included a 2 sec
presaturation, 12 kHz spectral window, 16 k data points
and 64 scans. All 1D data were processed with double-size
zero-filling and 0.3 Hz line broadening. To aid in metabo-
lite resonance assignments, 2D gradient COSY spectra
were collected. Acquisition parameters for these spectra
include a 1 sec presaturation, 6 kHz spectral window, 1024
data points, 128 increments, and n-type for the indirect
dimension. All 1H chemical shifts were internally refer-
enced to TSP. Peak assignments were made according to
previously reported values.16–18
Fish Exposures and Liver Tissue Collection
for Metabolomic Analysis
Juvenile rainbow trout (n 5 160) of mixed gender,
obtained from the Lake Burton Fish Hatchery (Clarkes-
ville, GA), were randomly assigned to 50 L fiberglass aqua-
ria (20 fish per tank) containing recirculating, dechlori-
nated tap water chilled to 128C, and carbon filtered to
remove any contaminant residues in the water. Fish were
maintained on a 12 h light, 12 h dark photoperiod and fed
trout chow (Aquamax, Carolina Pet Supply) during the
acclimation period (7 days) before exposure initiation.
Food was withheld over the 48 h course of the exposure.
Fish were exposed to each of the triadimefon enantio-
mers or the racemate via oral gavage at two dose levels
(high dose, 720 mg/kg body weight; low dose, 144 mg/kg
body weight). Oral gavage stock solutions (1000 mg/l) for
each triadimefon enantiomer or racemate were prepared
by dissolving each in methanol and warmed gelatin.14 Fish
were lightly anesthetized in tricaine methanesulfonate,
MS-222; 20 mg/l buffered with 200 mg/l NaHCO3;
(Finquel, Argent, Redmond, WA), removed and weighed,
and the corresponding dose was administered via oral
gavage by using a syringe. Fish were allowed to recover in
clean water before being returned to their respective treat-
ment aquaria. Controls were treated in an identical manner
but without the addition of any triadimefon chemical to the
methanol/gelatin solution. Twenty-four hours after the
exposure, 10 fish from each treatment (e.g., racemate,
720 mg/kg/day) and the corresponding controls were ran-
domly removed and anesthetized in a buffered solution of
MS-222, 100 mg/l. Total body weight was measured and
livers were excised, weighed, and immediately transferred
to preweighed microcentrifuge tubes, then flash-frozen in
liquid nitrogen. After an additional 24 h (the 48-h time
Metabolomic Data Reduction and Multivariate Analyses
All NMR spectra were phased and baseline corrected
before data reduction. Reduction consisted of segmenting
the chemical shift region from 0 to 10 ppm into 500 bins
0.02 ppm in size. All spectral processing and binning was
done using the ACD/1D NMR Manager module as part of
the ACD/SpecManager software package (Advanced
Chemistry Development, Toronto, Canada). The reduced
data sets were then imported into Microsoft1 Excel. For
each binned spectrum, those bins containing water reso-
nances (d 4.7–5.0) were removed, and the remaining bins
were scaled by the total integrated area as a means for
normalization. Finally, all variables were mean-centered
and unit variance-scaled.19 To identify changes in metabo-
lite profiles among dosed and control animals, principal
components analysis (PCA) was performed using the
SIMCA-P1 software package version 10.0.4 (Umetrics,
Umea, Sweden).
RESULTS AND DISCUSSION
GC-MS Analysis of Triadimefon and Triadimenol
Stereoisomers
Figure 2 shows a typical GC-MS chromatogram (SIM)
point), the remaining fish were removed and identically of a mixture of triadimefon and triadimenol commercial
processed. Liver samples were stored at 2808C before standards at a concentration of 4 lg/ml each obtained
further analysis.
using the BGB-172 chiral column. The two triadimefon
enantiomers are almost baseline-resolved, while the four
triadimenol stereoisomers are separated from each other
by approximately one minute or more. The triadimefon
Liver Extraction and Preparation for NMR Analysis
Polar liver extracts were prepared according to the pro- peaks are nominally of equal area, indicating the expected
cedure published by Viant et al.15 Briefly, pulverized tissue racemic form of the compound. The EF of triadimefon is
samples were extracted using 5 ml/g (wet mass) of ice- 0.503 6 0.005 (n 5 6), according to eq. 1. It is known that
cold 6% perchloric acid followed by centrifugation at the absolute configurations of the two enantiomers relative
12,000g for 10 min at 48C. Supernatants were collected and to their optical rotation is R-(2) and S-(1)20; the first elut-
neutralized to pH 7.4 using 2 M K2CO3 followed by centrif- ing peak on this BGB-172 column is the (R)-(2) enan-
ugation once again to remove salt (KClO4). Finally, 500 ll tiomer.
of each supernatant was lyophilized and reconstituted in
In fungi, plants, and animals, the carbonyl group of tria-
550 ll of sodium phosphate buffered (100 mM, pH 7.4) dimefon is reduced to an alcohol. The resulting metabo-
deuterium oxide (99.96%) containing 0.25 mM TSP. Fol- lite, triadimenol, is a considerably more potent fungicide
lowing another round of centrifugation, samples were than triadimefon.21 Since triadimenol has two chiral cen-
pipetted into 5-mm tubes and analyzed by NMR.
ters, it exists as four stereoisomers; these exist as two dia-
Chirality DOI 10.1002/chir

STEREOSELECTIVE TRANSFORMATION OF TRIADIMEFON
187
Fig. 2. GC-MS chromatogram (selected ion mode) of a mixed standard of the commercial formulations of triadimefon (first two peaks; ions m/z 181,
208, 210) and triadimenol (following four peaks; ions m/z 128, 168) showing enantiomer separation on a BGB-172 chiral column.
stereomers, termed A and B, each of which consists of a ments, and is representative of actual commercial triadime-
pair of enantiomers (Figs. 1 and 2). The EF of the diaster- nol. It is noteworthy, however, that different analytical
eomer A enantiomers is 0.497 6 0.005 (n 5 6), and of dia- standards of triadimenol may not contain the same ratios
stereomer B is 0.483 6 0.010 (n 5 6). (This ‘‘diaster- of diastereomers A and B as the commercial pesticide for-
eomer’’ terminology is technically incorrect and may be mulation. For example, a standard from Sigma (www.sig-
confusing: A and B are commercially called diastereomers, ma.com) contained less than 1% of diastereomer B, as
but technically a diastereomer is one isomer of a chemical measured by GC-MS, whereas the EPA repository formu-
that has two chiral centers, such as triadimenol, and each lation contains about 15%. Since the more abundant dia-
of these individual isomers has a counterpart, called an stereomer A is the most toxic, at least to rats8 and fungi,23
epimer, which is another of the four stereoisomers that dif- the Sigma standard would be overall more toxic—to the
fers in configuration at only one of the chiral centers.22)
same or similar species—since the toxicity of the EPA re-
Figure 2 shows for a commercial mixture of triadimenol pository product would be diluted by the presence of the
a pattern of chromatographic separation of diastereomers less toxic diastereomer B. Thus, it is critical to use the
A and B, each of which shows two enantiomer peaks in commercial pesticide formulation standard in fate and
equal amounts (A1 and A2; B1 and B2); i.e., each diaster- effect experiments on triadimenol.
eomer is a racemic mixture. Diastereomer A is 85% of the
total triadimenol, while diastereomer B is about 15%. As
shown in Figure 1, A1 and B1 epimers are derived from
Metabolism of Triadimefon in Trout Microsomes
Research results demonstrated that the S-(1) enan-
tiomer of triadimefon was metabolized faster than the R-
(2) enantiomer in trout liver microsomes, as shown in
Figure 3 where the enantiomer fraction (EF, eq. 1) of tria-
dimefon is shown to decrease slowly with time, from 0.513
at to to 0.495 at t48 (2889 min).
metabolism of (R)-(2)-triadimefon and A2 and B2 are from
(S)-(1)-triadimefon. Similar to the two triadimefon enan-
tiomers, the R stereoisomer elutes before the S stereoiso-
mer for each diastereomer (see Fig. 2). Little is known
about toxicity of the four individual stereoisomers, except
that most of the fungicidal activity of triadimenol can be
attributed to the A2 (1S,2R)-stereoisomer.23 This is one of
the predominant enantiomers in the commercial formula-
tion of triadimenol (see Fig. 2), and contributes to the
greater fungal toxicity of diastereomer A over that of B.
The enantiomer composition of the triadimenol formula-
½S-ðþÞ enantiomerꢀ
EF ¼
ð1Þ
½S-ðþÞ enantiomerꢀ þ ½R-ðꢁÞ enantiomerꢀ
Since the decrease in EF was relatively small, additional
tion used in the metabolism studies reported here corre- metabolism studies were conducted with the individual S-
sponds to the relative peak areas depicted in Figure 2, (1) and R-(2) enantiomers. The S-(1) enantiomer metab-
with the EF values reported earlier. This is the formulation olized 27% faster than R-(2) and would theoretically yield
supplied by the EPA repository and used in our experi- a change in EF of 0.07 after 2500 min based upon reaction
Chirality DOI 10.1002/chir

188
KENNEKE ET AL.
rates. These results compare favorably with the racemic
study, which exhibited a 22% faster rate of S-(1) metabo-
lism and an EF change of about 0.02 after 2500 min (see
Fig. 3).
Figure 4 is a GC-MS chromatogram (SIM) depicting the
distribution of triadimenol stereoisomers formed after in
vitro exposure of racemic triadimefon to trout liver micro-
somes for 480 min. Although the stereoisomer ratio pat-
tern is very different from that of the commercial triadime-
nol standard (see Fig. 2), it is obvious that the first two
eluting peaks correspond to diastereomer A (i.e., 1R,2S
and 1S,2R) and the second pair to diastereomer B (1R,2R
and 1S,2S) based upon comparison of retention times with
those of the standard. However, it was not possible to as-
certain the absolute configuration of each isomer based
upon existing information. To address this issue, metabo-
lism studies were conducted using the separate R-(2)- and
S-(1)-triadimefon enantiomers. The R-(2) enantiomer pro-
duced significant metabolites at 46.5 and 49.6 min while
the S-(1) yielded metabolites at 48.4 and 53.2 min (see
Fig. 4). Assuming that R could only produce 1R,2R and
1R,2S (see Fig. 1) and knowing that one of the first two
eluting peaks had to be either 1R,2S or 1S,2R (the compo-
sition of diastereomer A), it was concluded that the first
Fig. 3. The enantiomer fraction (EF) of triadimefon decreases with
time, indicating that the S-(1)-enantiomer metabolizes faster than the R-
(2).
Fig. 4. GC-MS chromatogram (selected ion mode; m/z 128, 168) depicting the distribution of tiradimenol enantiomers formed from racemic triadime-
fon after in vitro exposure to trout liver microsomes for 480 min.
Chirality DOI 10.1002/chir

STEREOSELECTIVE TRANSFORMATION OF TRIADIMEFON
189
Fig. 5. One-dimensional 1H NMR spectrum (800 MHz) of a typical trout liver extract used in the metabolomics study. ‘‘HDO’’ indicates the reso-
nance produced by residual water in the sample.
peak had to be 1R,2S (46.5 min) and by default the second isomers of triadimenol produced by metabolism of triadi-
peak was 1S,2R (48.4 min). Following the same logic, the mefon can vary greatly from species to species. Since it is
remaining two peaks were determined to be 1R,2R (49.6 known that triadimenol diastereomer A is 10 times more
min) and 1S,2S (53.2 min), respectively.
It is obvious from Figure 4 that there are large differen- stereoselective formation of triadimenol is an important
acutely toxic to rats (LD50) than is diastereomer B,8 the
ces in triadimenol stereoisomer product distribution after
triadimefon exposure to trout liver microsomes relative to
that of the standard (see Fig. 2). Using eq. 2, the EF meas-
ured for each of the four triadimenol stereoisomers after
metabolism for 480 min was: 0.06 (1R,2S), 0 (1S,2R), 0.38
(1R,2R), and 0.56 (1S,2S).
issue for both human health and ecological risk assess-
ment. Incidentally, the relatively high levels of tri-
adimefon exposure (up to about 40 lM or 1.2 mg/kg) to
the microsomes were necessary, as shown by the distri-
bution of triadimenol peak areas in Figure 4; e.g., the
peaks of diastereomer A are below quantitation levels.
Lower exposure levels would have required unreason-
ably long exposure times for measureable triadimenol
GC peaks to develop.
Based upon the distribution of triadimenol stereoiso-
mers observed in these liver metabolism studies, it can be
concluded that the reductive enzyme preferentially deliv-
ers the hydride from the si-face of the R-(2) enantiomer
and the re-face of the S-(1) enantiomer to yield the R
and S alcohol, respectively. Interestingly, this follows
both Prelog and anti-Prelog specificity,24 respectively,
and suggests that two separate carbonyl reducing
enzymes may be involved in the formation of triadimenol
from triadimefon.
½enantiomer xꢀ
EF x ¼
ð2Þ
½sum of 4 enantiomersꢀ
These EFs remained relatively unchanged throughout
the course of the metabolism study (48 h). The greater
abundance of the 1S triadimenol stereoisomers (1S,2S and
1S,2R; sum EF 5 0.56) formation relative to the 1R stereo-
isomers (1R,2S and 1R,2R; sum EF 5 0.44) formation at
480 min is consistent with the conclusion that the S-(1)-
triadimefon is metabolized more quickly than the R-(2)
(see Fig. 3); this assumes that 1S,2S and 1S,2R can only
arise via the metabolism of S-(1)-triadimefon. However,
these rates for triadimefon in trout microsomes are not
very different, which may relate to the fact that the toxicity
of the two enantiomers of triadimefon to fungi are also not
very different.21 The EF of zero of the (1S,2R) triadimenol
stereoisomer indicates no production of this most fungi-
toxic stereoisomer in trout microsomes, although a very
small amount—below the level of quantitation—is indi-
cated in the chromatogram of Figure 4. This may initially
appear surprising, but unpublished work by the authors
has shown that the relative proportions of the four stereo-
Metabolomics: The Effect of Chirality on Responses
to Triadimefon Exposure
As described earlier (experimental section), juvenile
rainbow trout were exposed (via gavage) to either one of
the triadimefon enantiomers or the racemate at two dose
levels, 144 and 720 mg/kg/day, and over two time dura-
tions, 24 and 48 h. Livers were then excised and extracted
using a perchloric acid-based procedure. Subsequent 1H
NMR analysis of the extracts revealed a number of metab-
olites in significant abundance (see Fig. 5), including
Chirality DOI 10.1002/chir

190
KENNEKE ET AL.
numerous amino acids, organic acids, and carbohydrates
(i.e., glycogen and glucose). This NMR spectrum is typical
of those obtained for the metabolomic analyses conducted
in this research.
After spectral processing and binning, the complete
dataset for these experiments was assessed using principal
components analysis (PCA).25 PCA is an unsupervised
multivariate modeling approach that is useful for determin-
ing where the greatest variation lies in the dataset. This
variation is captured through the creation of new variables
or components that are based on the original variables (or
bins in this case). Typically, the variation in the dataset
cannot be captured by one component alone, making mul-
ticomponent models necessary. In this case, components
are derived in such a way that the first component cap-
tures the greatest variation in the dataset with the subse-
quent components capturing progressively less. For
metabolomic studies, in practice, the first two or three
components are typically most useful.
Plotting the positions of the individual spectra (based on
the intensities of the bins) in a coordinate system defined
by these components is useful for determining similarities
and differences among the spectra (i.e., among the metab-
olite profiles of the fish livers) relative to the different ex-
posure classes (e.g., control vs. exposed). In such a plot,
called a scores plot, a group of fish that all display a partic-
ular liver metabolite profile (e.g., the control fish) will clus-
ter together in a particular area of the plot, while another
group that displays a different profile (e.g., the exposed
fish) will also cluster together, but in a separate area.
Thus, one can readily establish whether or not differences
exist between the classes to determine if the exposure has
produced a detectable response.
Using this approach with the data set generated by this
research provided a means to determine whether or not
the fish responded differently to the different enantiomers
or the racemate of triadimefon, and how these responses
changed with the duration of exposure. For example, Fig-
ure 6A shows a two-dimensional scores plot of the first
two components from a PCA model built using the binned
24 h spectral data acquired from the livers of the fish in
the control and high dose classes. Notice that while most
of the S-(1) enantiomer exposed fish are grouped close to
the controls, the fish exposed to the R-(2) enantiomer and
the racemate are grouped separately. This grouping of the
S-(1) enantiomer exposed fish with the controls indicates
a continued similarity in the metabolite profiles of these
two classes after 24 h of exposure. In addition, the separa-
Fig. 6. Two-dimensional PCA scores plots from the (A) 24 h and (B)
48 h high dose (720 mg/kg) triadimefon exposures. Key: controls (l),
racemate (~), R-(2)-enantiomer (h), S-(1)-enantiomer (n).
tion observed between these classes and the R-(2) enan- in vivo exposure of triadimefon for metabolomics studies.
tiomer and racemate-exposed fish reflects a change in the
liver metabolite profiles of these latter two classes of fish
Observation of results in terms of differential metabolite
formation depends upon exposure concentration and time.
as a result of exposure. After 48 h of exposure, however, It is possible that enantioselectivity would have been
this grouping of the S-(1) enantiomer-exposed fish with
the controls no longer exists (Fig. 6B). At this time, the S-
observed at the lower concentration with longer duration
of exposure. However, higher concentrations of chemical
(1) enantiomer-exposed fish have become grouped with are often used over shorter durations, and the information
the R-(2) enantiomer and racemate-exposed fish. There
was no significant difference in trout responses between
gleaned from such exposures is often useful for determin-
ing the nature of toxicity. In this case, that approach
controls and the low dose (144 mg/kg/day) classes for ei- showed that the two enantiomers of triadimefon produce
ther enantiomer or the racemate at either sampling time.
This begs the question of what is the appropriate level of
Chirality DOI 10.1002/chir
different metabolic responses as a function of exposure
duration.

STEREOSELECTIVE TRANSFORMATION OF TRIADIMEFON
191
The metabolomic results highlight the significant use- rameters, accurate risk assessment would require mea-
fulness of this approach for relating temporal responses in surement of acute and chronic toxicity of the separate
endogenous metabolite patterns to chemical exposures. enantiomers of both triadimefon and triadimenol for each
However, if possible, it is important to anchor such organism of concern. Considerable effort and expense is
responses to the more classical metabolism measures. For required to develop such data, but it is a feasible endeavor.
example, the EF data presented in Figure 3 allow for some Toxicity measurements for the separate enantiomers of
speculation as to why this effect is seen. Specifically, the chiral pharmaceuticals is the prescribed practice, and such
delay in the response of the fish to the S-(1) enantiomer measurements have also been conducted for a few enan-
(Fig. 6A) suggests a difference in the rate of metabolism tiomers of chiral pesticides.26–28
of this enantiomer compared to that of the R-(2) enan-
ACKNOWLEDGMENTS
tiomer and the racemate. As shown in Figure 3, the nega-
tive slope of the line, corresponding to a decrease in EF of
triadimefon, indicates that the S-(1) enantiomer is con-
verted to triadimenol in trout liver microsomes more rap-
idly than the R-(2) enantiomer. As this conversion likely
provides a path for reducing the toxicity of triadimefon,
the higher rate observed for the S-(1) enantiomer might
explain why, in the metabolomic analysis, the fish exposed
to this enantiomer are still grouped with the controls after
24 h of exposure (Fig. 6A); i.e., the faster metabolism of
the S-(1) enantiomer reduces its toxicity more than the
slower metabolism of the R-(2) enantiomer, keeping the
S-(1) enantiomer-exposed fish more like the controls.
Toxicity reduction is only possible if the products of the
S-(1) transformation are less toxic than those of the S-(1)
enantiomer itself. Although adequate data is not available
to prove this unequivocally, the fact that diastereomer B is
about 10-fold less toxic to rats than diastereomer A, and
that the major product of S-(1) metabolism, the 1S2S
enantiomer, is one of the components of diastereomer B
(see Fig. 4), suggests transformation to less toxic prod-
ucts. In support of this, Figure 4 shows very little produc-
tion of the other S-(1)-triadimefon metabolism product,
the 1S2R enantiomer, which is known to be the most toxic
of the four triadimenol enantiomers, at least to fungi.
As the exposure is continued for an additional 24 h,
however, the capacity to metabolize the S-(1) enantiomer
may become overwhelmed, eventually producing an effect
similar to the R-(2) enantiomer and the racemate. As a
result, the S-(1) enantiomer is observed to group with
the R-(2) enantiomer and the racemate in Figure 6B.
While these conclusions are based on compelling ration-
alization, further research will be required to fully test
this explanation.
This paper has been reviewed in accordance with the
U.S. Environmental Protection Agency’s peer and adminis-
trative review policies and approved for publication. Men-
tion of trade names or commercial products does not con-
stitute endorsement or recommendation for use. The
authors thank Tim Collette (EPA, Athens, GA) and Ross
Highsmith (EPA, RTP, NC) for insight and review of the
manuscript, and the Lake Burton Fish Hatchery, Clarkes-
ville, GA, for supplying the rainbow trout.
LITERATURE CITED
1. Roberts TR, Hutson DH, editors. Metabolic pathways of agrochemi-
cals, Part 2: Insecticides and fungicides. Cambridge, UK: The Royal
Society of Chemistry; 1999. p 1090–1097.
2. Kenneke JF, Mazur CS, Ritger SE, Sack TJ. Mechanistic investigation
of the noncytochrome P450-mediated metabolism of triadimefon to tri-
adimenol in hepatic microsomes. Chem Res Toxicol 2008;21:1997–2004.
3. Koller W. Isomers of sterol synthesis inhibitors: fungicidal effects and
plant growth regulator activities. Pestic Sci 1987;18:129–147.
4. Fuchs A. In: Ariens EJ, van Rensen JJS, Welling W, editors. Stereose-
lectivity of pesticides: biological and chemical problems. Implications
of Stereoisomerism in Agricultural Fungicides. Amsterdam: Elsevier;
1988. p 203.
5. INCHEM. Pesticide residues in food: 1981 evaluations. Available
at: http://www.inchem.org/documents/jmpr/jmpmono/v81pr32.htm.
6. Kahle M, Buerge IJ, Hauser A, Muller MD, Poiger T. Azole fungi-
cides: occurrence and fate in wastewater and surface waters. Environ
Sci Technol 2008;42:7191–7200.
7. Garrison AW. Probing the enantioselectivity of chiral pesticides. Envi-
ron Sci Technol 2006;40:16–23.
8. INCHEM. Pesticide residues in food: 1989 evaluations Part II
Toxicology. Available at: http://www.inchem.org/documents/jmpr/
jmpmono/v89pr14.htm.
9. Mazur CS, Kenneke JF. Cross-species comparison of conazole fungi-
cide metabolites using rat and rainbow trout (Onchorhynchus mykiss)
hepatic microsomes and purified human CYP3A4. Environ Sci Tech-
nol 2008;42:947–954.
10. Sinclair CJ, Boxall ABA. Assessing the ecotoxicity of pesticide trans-
formation products. Environ Sci Technol 2003;37:4617–4625.
IMPLICATIONS
Preliminary work in the authors’ laboratory shows that
the stereochemistry of conversion of triadimefon to triadi-
menol varies with metabolism of the exposed organism;
this research is continuing. Since it is known that the ste-
reoisomers of triadimenol differ in their toxicities to
fungi23 and to rats,8 the biological activity of the triadime-
nol that is formed probably depends on the relative abun-
dance of its particular stereoisomers. Interspecies compari-
son of the production of metabolite enantiomers by chiral
analysis should be useful in probing the metabolism of
pesticide-exposed organisms as well as improving cross-
species extrapolation for risk assessment. Finally, in addi-
tion to knowledge about the stereochemical exposure pa-
11. Cherney DP, Ekman DR, Dix DJ, Collette TW. Raman spectroscopy-
based metabolomics for differentiating exposures to triazole fungi-
cides using rat urine. Anal Chempossible
12. Ekman DR, Teng Q, Villeneuve DL, Kahl MD, Jensen KM, Durham
EJ, Ankley GT, Collette TW. Profiling lipid metabolites yields unique
information on gender- and time-dependent responses of fathead min-
now (Pimephales promelas) exposed to 17a-ethynylestradiol. Metabolo-
mics 2009;5:22–32.
13. Konwick BJ, Garrison AW, Avants JK, Fisk AT. Bioaccumulation and
biotransformation of chiral triazole fungicides in rainbow trout (Onco-
rhynchus mykiss). Aquat Toxicol 2006;80:372–381.
14. Brown SB, Evans RE, Vandenbyllardt L, Finnson KW, Palace VP,
Kane AS, Yarechewski AY, Muir DCG. Altered thyroid status in lake
trout (Salvelinus namaycush) exposed to co-planar 3,30,4,40,5-penta-
chlorobiphenyl. Aquat Toxicol 2004;67:75–85.
Chirality DOI 10.1002/chir

192
KENNEKE ET AL.
15. Viant MR, Rosenblaum ES, Tjeerdema RS. NMR-based metabolomics:
a powerful approach for characterizing the effects of environmental
stressors on organism health. Environ Sci Technol 2003;37:4982–
4989.
22. Eliel EL, Wilen SH, Mander LN. Stereochemistry of organic com-
pounds. New York, NY: Wiley; 1994. p 1196.
23. Fruh T, Spindler F. Chiral acylanilides and chiral triazole-related fun-
gicides. In: Kurihara N, Miyamoto J, editors. Chirality in agrochemi-
cals. Chichester, England: Wiley; 1998. p 158–159.
16. Nicholson JK, Foxall PJD. 750MHz 1H and 1H-13C NMR spectroscopy
of human blood plasma. Anal Chem 1995;67:793–811.
24. Faber K. Biotransformations of non-natural compounds: state of the
art and future development. Pure Appl Chem 1997;69:1613–1632.
17. Fan TWM. Metabolite profiling by one- and two-dimensional NMR
analysis of complex mixtures. Prog NMR Spectrosc 1996;28:161–219.
25. Jackson JE. A user’s guide to principal components. Hoboken, NJ:
Wiley; 1991.
18. Lindon JC, Nicholson JK, Everett JR. NMR spectroscopy of biofluids.
Ann Rep NMR Spectrosc 1999;38:1–88.
26. Konwick BJ, Fisk AT, Garrison AW, Avants JK, Black MC. Acute
enantioselective toxicity of fipronil and its desulfinyl photoproduct to
Ceriodaphnia dubia. Environ Toxicol Chem 2005;24:2350–2355.
19. Wold S, Antti H, Lindgren F, Ohman J. Orthogonal signal correction
of near-infrared spectra. Chemom Intell Lab Syst 1998;44:175–185.
20. Kurihara N, Miyamoto J, Paulson GD, Zeeh B, Skidmore MW, Hol-
lingworth RM, Kuiper HA. Chirality in synthetic agrochemicals: bioac-
tivity and safety considerations. Pure Appl Chem 1997;69:1335–1348.
27. Overmyer JP, Rouse DR, Avants JK, Garrison AW, DeLorenzo ME,
Chung KW, Key PB, Wilson WA, Black MC. Toxicity of fipronil and
its enantiomers to marine and freshwater non-targets. J Environ Sci
Health Part B 2007;42:471–480.
21. Kramer W, Buchel KH, Draber W. Structure-activity correlation in the
azoles, In: Miyamoto J, Karney PC, editors. Proceedings of 5th Inter-
national Congress on Pesticide Chemistry, Kyoto, 1982, vol. 1. Oxford:
Pergamon; 1983. p 223–232.
28. Liu W, Gan J, Schlenk D, Jury WA. Enantioselectivity in environmen-
tal safety of current chiral insecticides. Proc Natl Acad Sci USA
2005;102:701–706.
Chirality DOI 10.1002/chir