Isoflavone, Glyphosate, and Aminomethylphosphonic Acid Levels in Seeds of Glyphosate-Treated, Glyphosate-Resistant Soybean

340
J. Agric. Food Chem. 2003, 51, 340?344
Isoflavone, Glyphosate, and Aminomethylphosphonic Acid
Levels in Seeds of Glyphosate-Treated, Glyphosate-Resistant
Soybean
STEPHEN O. DUKE,*,† AGNES M. RIMANDO,† PATRICK F. PACE,†
KRISHNA N. REDDY,‡ AND REID J. SMEDA§
Natural Products Utilization Research Unit, Agricultural Research Service, U.S. Department of
Agriculture, P.O. Box 8048, University, Mississippi 38677; Southern Weed Science Research Unit,
P.O. Box 350, Stoneville, Mississippi 38776; and Department of Agronomy, University of Missouri,
Columbia, Missouri 65211
The estrogenic isoflavones of soybeans and their glycosides are products of the shikimate pathway,
the target pathway of glyphosate. This study tested the hypothesis that nonphytotoxic levels of
glyphosate and other herbicides known to affect phenolic compound biosynthesis might influence
levels of these nutraceutical compounds in glyphosate-resistant soybeans. The effects of glyphosate
and other herbicides were determined on estrogenic isoflavones and shikimate in glyphosate-resistant
soybeans from identical experiments conducted on different cultivars in Mississippi and Missouri.
Four commonly used herbicide treatments were compared to a hand-weeded control. The herbicide
treatments were (1) glyphosate at 1260 g/ha at 3 weeks after planting (WAP), followed by glyphosate
at 840 g/ha at 6 WAP; (2) sulfentrazone at 168 g/ha plus chlorimuron at 34 g/ha applied preemergence
(PRE), followed by glyphosate at 1260 g/ha at 6 WAP; (3) sulfentrazone at 168 g/ha plus chlorimuron
at 34 g/ha applied PRE, followed by glyphosate at 1260 g/ha at full bloom; and (4) sulfentrazone at
168 g/ha plus chlorimuron at 34 g/ha applied PRE, followed by acifluorfen at 280 g/ha plus bentazon
at 560 g/ha plus clethodim at 140 g/ha at 6 WAP. Soybeans were harvested at maturity, and seeds
were analyzed for daidzein, daidzin, genistein, genistin, glycitin, glycitein, shikimate, glyphosate, and
the glyphosate degradation product, aminomethylphosphonic acid (AMPA). There were no remarkable
effects of any treatment on the contents of any of the biosynthetic compounds in soybean seed from
either test site, indicating that early and later season applications of glyphosate have no effects on
phytoestrogen levels in glyphosate-resistant soybeans. Glyphosate and AMPA residues were higher
in seeds from treatment 3 than from the other two treatments in which glyphosate was used earlier.
Intermediate levels were found in treatments 1 and 2. Low levels of glyphosate and AMPA were
found in treatment 4 and a hand-weeded control, apparently due to herbicide drift.
KEYWORDS: Aminomethylphosphonic acid; glyphosate; herbicide-resistant crop; isoflavone; shikimic
acid; transgenic crop
INTRODUCTION
increased since it was introduced in 1995, until approximately
The isoflavones of soybeans have a number of nutraceutical
75% of all U.S. soybeans planted in 2002 were GR (6). Before
properties, including estrogenic and hypocholesterolemic activi-
and since transgenic crops were introduced, questions have been
ties (1, 2), as well as reportedly being able to reduce the risk of
posed regarding potential subtle, pleiotrophic effects of the
cancer (3). They may have adverse health effects on certain
transgenes on food quality (7) and similar effects that might be
animals fed soybean meal (4). Thus, the levels of these
due to positional effects of the transgene in the genome (8).
compounds in soybeans are of great interest to both human and
One study indicates that GR soybean lines contain lower levels
animal nutritionists.
of estrogenic isoflavones than non-GR soybean lines (9). A more
The most successful transgenic crop in the world has been
thorough study has shown that there are no effects of the CP4
glyphosate-resistant (GR) soybeans (5). Its use has steadily
5-enolpyruvylshikimatae-3-phosphate synthase (CP4 EPSPS)
gene, which confers glyphosate resistance to all GR cultivars
sold, on isoflavone content of soybeans (10).
* Corresponding author [telephone (662) 915-1036; fax (662) 915-1035;
e-mail sduke@olemiss.edu].
There is the possibility of sublethal levels of the herbicide to
† USDA-ARS Natural Products Utilization Research Unit.
‡
which the crop has been made resistant, if resistance is not
USDA-ARS Southern Weed Science Research Unit.
§ Department of Agronomy, University of Missouri.
complete. Isoflavones are products of the shikimate pathway,
10.1021/jf025908i CCC: $25.00
© 2003 American Chemical Society
Published on Web 12/05/2002

---

Glyphosate Effects on Isoflavones in Soybean
J. Agric. Food Chem., Vol. 51, No. 1, 2003
341
the target pathway of glyphosate (11). Glyphosate has been
Table 1. Herbicide Treatments Used in Glyphosate-Resistant Soybean
shown to reduce the levels of related compounds in nontrans-
Experiments in Mississippi and Missouri in 2000
genic soybeans (see, e.g., refs 12 and 13) and of the phytoestro-
genic compound genistein in Lupinus luteus L. (14). Reddy et
rateb
application
application
al. (15) found that GR soybeans are not completely resistant,
treatmenta
(g/ha)
methodc
timing
with significant inhibition of growth occurring at application
1. glyphosate
1260
POST
3 weeks after planting
rates as low as 2.24 kg/ha under certain growing conditions.
840
POST
6 weeks after planting
2. sulfentrazone +
168
PRE
at planting
Therefore, it is possible that glyphosate affects the levels of
chlorimuron fb
34
PRE
at planting
estrogenic isoflavones in seed produced from GR soybeans.
glyphosate
1260
POST
6 weeks after planting
Taylor et al. (16) addressed this question and found no effects.
3. sulfentrazone +
168
PRE
at planting
However, their study did not examine all estrogenic isoflavones
chlorimuron fb
34
PRE
at planting
glyphosate
1260
POST
at full bloom
and their glycosides. Furthermore, the glyphosate application
(8 weeks after planting)
rates used were not as high or applied as late in plant de-
4. sulfentrazone +
168
PRE
at planting
velopment as those commonly used by many farmers.
chlorimuron fb
34
PRE
at planting
Glyphosate is labeled for use in GR soybeans from emergence
acifluorfen +
280
POST
6 weeks after planting
bentazon +
560
POST
6 weeks after planting
to flowering. Two applications of glyphosate alone or one
clethodim
140
POST
6 weeks after planting
application of glyphosate following preemergence herbicide
5. hand-weeded
manual hoeing as needed to remove weeds
applications are commonly used to achieve effective weed
control. Use of preemergence herbicides at planting provides
a fb, followed by. b Rate refers to acid equivalent for glyphosate and active
the flexibility for late-season glyphosate application (17). In this
ingredient for all other herbicides. c Pre- (PRE) or post- (POST)emergence.
paper, we reinvestigate this question, comparing several com-
monly used herbicide-based weed management regimes, includ-
weeds in the experimental area included common waterhemp (Ama-
ing herbicide programs that include other herbicides that have
ranthus rudis Sauer), giant foxtail (Setaria faberi Herrm.), ivyleaf
been reported to affect the synthesis of compounds related to
morningglory [I. hederacea (L.) Jacq.], pitted morningglory, and prickly
isoflavones. Environmental conditions are known to influence
sida.
the level of resistance of glyphosate-resistant crops (see, e.g.,
Glyphosate-resistant soybeans (Asgrow 3701 RR; determinant, mid
ref 18), so we conducted these experiments with GR soybeans
III maturity group) were planted on May 8 at a population of 432000
seeds/ha in four-row plots with rows 76 cm apart and 13.7 m long.
at two locations with different environmental conditions.
The PRE and POST herbicide treatments were applied as described
below (Table 1). All herbicides were applied with a CO2-pressurized
MATERIALS AND METHODS
backpack sprayer at a carrier volume of 187 L/ha and a spray pressure
of 207 kPa using flat fan TeeJet XR8003 nozzle tips at a ground speed
Field Experimental Conditions. Mississippi Experiment. This
of 4.8 km/h. Rainfall was adequate throughout the growing season,
experiment was conducted in 2000 at the USDA-ARS Southern Weed
negating the use of irrigation.
Science Research farm, Stoneville, MS (33° N latitude). The soil was
The experiment was conducted in a randomized complete block
a Dundee silt loam (fine-silty, mixed, thermic Aeric Ochraqualf) with
design with four replications. At maturity, the two center rows of each
pH 6.4, 1.6% organic matter, and soil textural fractions of 19% sand,
plot were harvested with a combine.
57% silt, and 24% clay. The experimental area was tilled in the fall of
Herbicide Treatments. Herbicide treatments at both locations
1999 and the following spring with a disk harrow, followed by a field
included glyphosate at 1260 g of ae/ha at 3 weeks after planting (WAP)
cultivator before planting. The experimental area was naturally infested
followed by (fb) glyphosate at 840 g ae/ha at 6 WAP; sulfentrazone at
with weeds. Predominant weed species in the experimental area included
168 g of ai/ha plus chlorimuron at 34 g of ai/ha applied PRE fb
barnyardgrass [Echinochloa crus-galli (L.) Beauv.], browntop millet
glyphosate at 1260 g ae/ha at 6 WAP; sulfentrazone at 168 g of ai/ha
[Brachiaria ramosa (L.) Stapf], hemp sesbania [Sesbania exaltata (Raf.)
plus chlorimuron at 34 g of ai/ha applied PRE fb glyphosate at 1260
Rydb. ex A.W. Hill], hyssop spurge (Euphorbia hyssopifolia L.), pitted
g of ae/ha at full bloom (8 WAP); sulfentrazone at 168 g of ai/ha plus
morningglory (Ipomoea lacunosa L.), prickly sida (Sida spinosa L.),
chlorimuron at 34 g of ai/ha applied PRE fb acifluorfen at 280 g of
sicklepod [Senna obtusifolia (L.) Irwin and Barneby], and yellow
ai/ha plus bentazon at 560 g of ai/ha plus clethodim at 140 g of ai/ha
nutsedge (Cyperus esculentus L.).
at 6 WAP; and a hand-weeded control (Table 1). PRE herbicides were
Glyphosate-resistant soybean cultivar DP 5806 RR (determinant, late
applied broadcast immediately after planting. POST herbicides were
V maturity group) was planted at a population of 359000 seeds/ha on
applied at several stages of soybean growth. A nonionic surfactant
May 12, 2000, in four-row plots with rows 100 cm apart and 24.3 m
(paraffinic petroleum oil concentrate) was added to all POST treatments
long. Preemergence (PRE) and postemergence (POST) herbicide
except glyphosate as suggested by the manufacturer.
treatments were applied as described below and summarized in Table
Soybean Isoflavone Analysis. Extraction. Ten-gram samples of
1. All herbicides were applied with a tractor-mounted sprayer with
oven-dried (100 °C, 15-16 h), ground (Cyclotech 1093 sample mill,
TeeJet 8004 (Spraying Systems Co., Wheaton, IL) standard flat spray
Foss Tecator, Ho¨gana¨s, Sweden) soybeans were extracted with 80%
tips delivering 187 L/ha water at 179 kPa at a ground speed of 6.8
methanol/20% water using a Dionex 200 ASE extractor (Dionex Corp.,
km/h.
Sunnyvale, CA), programmed to four cycles to ensure complete
The experiment was conducted in a randomized complete block
extraction of the isoflavones. Extracts were dried in a vacuum.
design with four replications. Rainfall during May and June was normal,
Analysis. Samples were analyzed for their content of daidzin,
but the months of July and August were extremely hot and dry.
genistein, genistin, glycitin, formononetin, and biochanin A by HPLC
Soybeans were irrigated five times during July and August. At maturity,
(Hewlett-Packard 1050, Agilent Technologies Inc., Palo Alto, CA) using
soybeans were harvested from two center rows of each plot using a
a reversed-phase C18 column, Zorbax SB-Aq, 5 µm, 4.6 × 150 mm
combine.
i.d. × length (Agilent Technologies Inc.) maintained at 26 °C. The
Missouri Experiment. This experiment was conducted in 2000 at
mobile phase and solvent elution were as follows (solvent A is 0.05%
the University of Missouri Bradford Research Farm near Columbia (39°
acetic acid in water; solvent B is 0.05% acetic acid in acetonitrile):
N latitude). The soil was a Mexico silt loam (fine, montmorillonitic,
0-2 min, 20% B; 2-18 min, 20-40% B; 18-23 min, 40-100% B;
Mesic, Aeric Vertic Epiaqualf) with pH 6.0, 2.3% organic matter, and
23-26 min, 100% B; 26-27 min, 100-20% B; 27-34 min, 20% B.
soil textural fractions of 6% sand, 73% silt, and 21% clay. The
The mobile phase flow rate was 0.6 mL/min. Sample volume injection
experimental area was tilled in the fall 1999 with a disk harrow and
was 10 µL. The isoflavones were detected using a photodiode array
the seedbed prepared in spring 2000 with a field cultivator. Predominant
detector while on-line monitoring was done at 260 nm.

---

342
J. Agric. Food Chem., Vol. 51, No. 1, 2003
Duke et al.
Daidzein, glycitein, and shikimic acid were analyzed by GC-MS on
Table 2. Effects of Different Herbicide Treatments on Shikimate,
an Agilent 6890 gas chromatograph coupled to a JEOL GC Mate II
Isoflavones and Their Glycosides, Glyphosate, and AMPA in Soybean
mass spectrometer (JEOL Corp., Peabody, MA). The capillary column
Seeda
used was a DB-17HT (15 m length × 0.25 mm i.d. × 0.15 µm film;
J&W Scientific, Folsom, CA). The carrier gas was helium (flow rate
µg/g for herbicide treatment
)
seed
1.0 mL/min). The injection port was maintained at 250 °C. The
constituent
1
2
3
4
5
volume of injection was 1 µL, splitless injection. The GC temperature
was programmed as follows: initial temperature, 120 °C; held for 2
Stoneville
min; then increased to 250 °C at a rate of 25 °C/min; held at this
shikimate
52
45
55
42
26
temperature for 2 min; then increased to 340 °C at a rate of 60 °C/
daidzein
1023a
634b
883ab
625b
612b
daidzin
1102
773
973
1049
888
min; and held at this temperature for 3 min. The GC interface and MS
genistein
258
150
147
107
113
ionization chamber were kept at 250 and 200 °C, respectively. For the
genistin
1136
962
1105
1202
1041
quantitation of the isoflavones, standard curves were prepared for each
glycitein
973
656
806
636
676
isoflavone using chalcone as an internal standard.
glycitin
383
441
394
477
459
Glyphosate and Aminomethylphosphonic Acid (AMPA) Deter-
glyphosate
0.181b
0.480b
2.18a
0.166c
0.103c
mination. Extraction and DeriVatization. Soybeans were dried in the
AMPA
0.602b
0.729b
7.27a
0.269b
0.263b
oven at 100 °C for 15-16 h and milled (Cyclotec 1093 sample mill,
Columbia
Ho¨ganta¨s, Sweden). Soybeans were extracted and derivatized following
shikimate
29
24
60
41
57
a published procedure (19), with minor modifications. One gram of
daidzein
805
856
967
1013
1002
ground soybeans was extracted with 5 mL of water in a sonicating
daidzin
1367
1562
1704
1398
1696
bath for 20 min and then centrifuged at 200g for 10 min. Two milliliters
genistein
250
311
389
382
294
genistin
1403
1413
1347
1385
1451
of supernatant was taken and transferred to a 20-mL vial. Extraction
glycitein
631
562
940
810
674
was repeated by adding 5 mL of water to the sample; the vial was
glycitin
583
555
556
502
502
shaken and sonicated for 20 min and then centrifuged at 200g for 10
glyphosate
0.234b
0.552b
3.08a
0.086c
0.126c
min. One milliliter of supernatant was taken and combined with the 2
AMPA
0.862b
0.492b
25.00a
0.158b
0.126b
mL obtained from the first extraction. Concentrated HCl (15 µL) was
added to this combined supernatant and shaken. A 2.5-mL portion was
a Means in the same row with different letters are significantly different (P )
pipeted into a 20-mL vial, and 2.5 mL of CH2Cl2 was added, shaken,
0.05) based on Fisher’s protected LSD. There were no significant differences
and centrifuged for 10 min at 200g. A portion (1.8 mL) of the water
between means in rows without letters.
layer was taken, and 200 µL of acidic modifier (16 g of KH2PO4, 160
mL of H2O, 40 mL of MeOH, and 13.4 mL HCl) was added. One
milliliter was transferred to a cation-exchange resin column (2-mL
implemented in that the LSD was interpreted only if the ANOVA F
packed volume; AG 50W-X8, H+; Bio-Rad Laboratories, Hercules, CA)
test for treatment effect was significant (P ) 0.05).
that had been previously washed with two 5-mL portions of water.
The sample was drained to the top of the column bed, and to the column
RESULTS AND DISCUSSION
was added 0.7 mL of CAX mobile phase (160 mL of H2O, 40 mL of
MeOH, and 2.7 mL of HCl), eluted, and discarded. Twelve milliliters
Qualitatively, the results at the Stoneville, MS, and Columbia,
of CAX mobile phase was again added to the column to elute the
MO, sites were similar (Table 2). Shikimate levels were low
analyses. The eluate was collected in a 20-mL vial and evaporated to
and not significantly affected by any treatment. In healthy plants
dryness using a Savant Speed Vac (model SVC 200, Savant Instruments,
of most species, including nontransgenic soybean, shikimate
Inc., Holbrook, NY). The dried sample was dissolved in 1.5 mL of
levels are low. By blocking EPSPS, glyphosate causes manyfold
CAX mobile phase. A 20-µL aliquot was taken and added to 640 µL
increases in shikimate levels in glyphosate-treated soybean
of a solution of 2,2,3,3,4,4,4-heptafluoro-1-butanol and trifluoroacetic
plants (21, 22). The effects on shikimate are much more
anhydride (1:2) in a chilled 4-mL vial. The mixture was allowed to
equilibrate at room temperature for 10-15 min. The vial was transferred
dramatic than glyphosate-induced decreases in levels of com-
to a heating block maintained at 90 °C for 1 h and then allowed to
pounds derived from shikimate such as anthocyanin (12, 13).
cool to room temperature. The solvent was evaporated under a stream
In fact, elevated shikimate levels are used as an early and highly
of nitrogen, and the residue was dissolved in 80 µL of ethyl acetate
sensitive indicator of glyphosate effects on plant tissues (22).
containing 0.2% citral; 50 µL was transferred to a GC vial and analyzed
In transgenic, glyphosate-resistant cotton, shikimate levels rise
by GC-MS.
when the plants are treated with enough glyphosate to cause
Analysis. GC-MS (Agilent 6890 Series GC coupled to a JEOL
sublethal effects on reproductive tissues (23). Thus, the absence
GCMateII mass spectrometer) analysis was done using a DB-5 capillary
of an effect or only slight increases in shikimate observed in
column (J&W Scientific, Inc.), 30 m length × 0.25 mm i.d. × 0.25
this study indicated that the CP4 EPSPS was either not inhibited
µm film, run under the following GC temperature program: initial, 70
°
or minimally inhibited and that the CP4 EPSPS utilized all or
C; held for 3.5 min; raised to 160 °C at 30 °C/min rate; raised to 270
°
most of the shikimate that would have accumulated from
C at 70 °C/min rate; raised to 310 °C at 35 °C/min rate; and finally
held at this temperature for 3 min. The injection port, GC interface,
inhibition of the native EPSPS. If so, one would expect no
and ionization chamber were maintained at 260, 200, and 120 °C,
effects of glyphosate on shikimate products, such as isoflavones
respectively. The carrier gas was ultrahigh-purity helium at a 1 mL/
and their glycosides. There are no reports of inhibitors of
min flow rate. The sample injection volume was 1 µL. The MS detector
acetolactate synthase (chlorimuron), protoporphyrinogen oxidase
was a magnetic sector; spectra were acquired in the positive, low-
(acifluorfen and sulfentrazone), or photosystem II (bentazon)
resolution, selected-ion monitoring mode. AMPA derivative was
on shikimate levels.
observed at 7:23 min (m/z 571, 502, 446, 372), and glyphosate
There was no effect of glyphosate or any other herbicide
derivative was observed at 7:59 min (m/z 611, 584, 460). Glyphosate
treatment on isoflavone levels at the Columbia site (Table 2).
and AMPA in the samples were quantified from a calibration curve of
At the Stoneville location, glyphosate used alone (treatment 1)
derivatized standards of glyphosate and AMPA.
elevated daidzein levels. At the 5% level of confidence, at least
Statistical Analysis. Values from HPLC and GC-MS quantification
were statistically analyzed through analysis of variance (ANOVA) using
one significant difference would be expected in this number of
the GLM procedure of SAS software (SAS Institute Inc., Cary, NC).
treatments, even if there were no effects. Only trace amounts
Treatment means were separated using the least significant difference
of formononetin and biochanin A were found in all samples
(LSD) test in the GLM procedure. Fisher’s protected LSD (20) was
from both sites (data not shown). Acetolactate synthase inhibi-

---

Glyphosate Effects on Isoflavones in Soybean
J. Agric. Food Chem., Vol. 51, No. 1, 2003
343
tors, such as chlorimuron, and protoporphyrinogen oxidase
(3) Zhou, Y.; Lee, A. S. Mechanism for suppression of the
inhibitors, such as sulfentrazone and acifluorfen, can cause
mammalian stress response by genistein, an anticancer phyto-
elevated levels of products of the shikimate pathway, such as
estrogen from soy. J. Natl. Cancer Inst. 1998, 90, 381-388.
isoflavones (see, e.g., refs 24 and 25). There was no evidence
(4) Setchell, K. D. R.; Gosselin, S. J.; Welsh, M. B.; Johnston, J.
of such an effect in this study.
O.; Balistreri, W. F.; Kramer, L. W.; Dresser, B. L.; Tarr, M. J.
Glyphosate and its metabolite, AMPA, were found at the
Dietary estrogenssa probable cause of infertility and liver disease
in captive cheetahs. Gastroenterology 1987, 93, 225-233.
highest levels when treated at the latest date with glyphosate at
(5) Duke, S. O.; Scheffler, B. E.; Dayan, F. E.; Dyer, W. E. Genetic
both locations (treatment 3). The highest level of glyphosate
engineering crops for improved weed management traits. ACS
found (treatment 3, Columbia; 3.08 µg/g) was below the EPA
Symp. Ser. 2002, No. 829, 52-66.
tolerance level of 5 µg/g (26). The high levels of AMPA in
(6) Anonymous. U.S. GM crops edge higher to set new record.
treatment 3 (7 and 25 µg/g in Stoneville and Columbia,
Agrow World Crop Protection News 2002, July 12 (No. 404),
respectively) were a surprise, because glyphosate has not been
15.
reported to be readily degraded in soybean plants. There are no
(7) Mallory-Smith, C.; Eberlein, C. V. Possible pleiotropic effects
EPA tolerance levels for AMPA in soybean. We were also
in herbicide-resistant crops. In Herbicide-Resistant Crops,
surprised to find low levels of glyphosate and AMPA in
Agricultural, EnVironmental, Economic, Regulatory, and Techni-
treatments 4 and 5, which had not included glyphosate. These
cal Aspects; Duke, S. O., Ed.; CRC Press: Boca Raton, FL, 1996;
findings were not due to contaminated chromatograpy columns,
pp 201-210.
as we found similar values when using fresh columns that had
(8) Dyer, W. E.; Hess, F. D.; Holt, J. S.; Duke, S. O. Potential
not had a glyphosate treatment sample passed through it. The
benefits and risks of herbicide-resistant crops produced by
most likely explanation is that there was herbicide drift at both
biotechnology. Hortic. ReV. 1993, 15, 367-408.
locations, both from the glyphosate treatments in this study and,
(9) Lappe´, M. A.; Bailey, E. B.; Childress, C.; Setchell, K. D. R.
perhaps, from surrounding fields. With widespread adoption of
Alterations in clinically important phytotestrogens in genetically
modified, herbicide-tolerant soybeans. J. Med. Foods 1999, 1,
GR soybean and cotton, it is difficult to find an experimental
241-245.
site free from glyphosate drift from neighboring fields. Others
(10) Padgette, S. R.; Taylor, N. B.; Nida, D. L.; Bailey, M. R.;
have explained glyphosate contamination of seeds of untreated
MacDonald, J.; Holden, L. R.; Fuchs, R. L. The composition of
wheat and canola by herbicide drift (27, 28). In these previous
glyphosate-tolerant soybean seeds is equivalent to that of
papers, unsuccessful efforts were made to shield untreated plants
conventional soybeans. J. Nutr. 1996, 126, 702-716.
from glyphosate drift. Drift of glyphosate to nontarget crops
(11) Duke, S. O. Glyphosate. In Herbicides: Chemistry, Degradation,
and areas has been a significant problem in glyphosate-resistant
and Mode of Action; Kearney, P. C., Kaufman, D. D., Eds.;
crops (29).
Dekker: New York, 1988; Vol. 3, pp 1-70.
Little is known of the degradation of glyphosate to AMPA
(12) Hoagland, R. E.; Duke, S. O. Relationships between phenyl-
in plants. No plant-derived enzyme has been shown to make
alanine ammonia-lyase activity and physiological responses of
this conversion. Soybean cell cultures degrade glyphosate to
soybean (Glycine max) seedlings to herbicides. Weed Sci. 1983,
AMPA more efficiently that those of wheat or maize (30).
31, 845-852.
AMPA is mildly phytotoxic to soybean, and its mode of action
(13) Holliday, M. J.; Keen, N. T. The role of phytoalexins in the
is apparently different from that of glyphosate (31). AMPA can
resistance of soybean leaves to bacteria: effect of glyphosate
cause anthocyanin levels to increase in soybean seedlings (31).
on glyceollin accumulation. Phytopathology 1982, 72, 1470-
AMPA levels were apparently insufficient to affect isoflavone
1474.
(14) Kneer, R.; Alexander, A.; Olesinski; Raskin, I. Characterization
levels in soybean seed in our study (Table 2).
of the eliciter-induced biosynthesis and secretion of genistein
In a recent study (32) with another legume (field pea, Pisum
from roots of Lupinus luteus L. J. Exp. Bot. 1999, 50, 1553-
satiVum L.), much higher levels of glyphosate and much lower
1559.
levels of AMPA were found in the seed of plants treated with
(15) Reddy, K. N.; Hoagland, R. E.; Zablotowicz, R. M. Effect of
1.7 kg of glyphosate at an early seed maturation stage of
glyphosate on growth, chlrophyll, and nodulation in glyphosate-
development. In this case, the crop was nontransgenic, so the
resistant and susceptible soybean (Glycine max) varieties. J. New
lack of metabolism of glyphosate to AMPA could have been
Seeds 2000, 2, 37-52.
due to the high degree of toxicity of the glyphosate treatment.
(16) Taylor, N. B.; Fuchs, R. L.; MacDonald, J., Shariff, A. R.;
In summary, even at higher application rates and with later
Padgette, S. R. Compositional analysis of glyphosate-tolerant
applications than used by Taylor et al. (16), glyphosate had little
soybeans treated with glyphosate. J. Agric. Food Chem. 1999,
or no effect on shikimate or isoflavones in GR soybeans. These
47, 4469-4473.
results confirm that there should be no concern about effects of
(17) Reddy, K. N. Glyphosate-resistant soybean as a weed manage-
glyphosate on this aspect of the nutritional and nutraceutical
ment tool: opportunities and challenges. Weed Biol. Manage.
properties of GR soybeans when used at the times and rates
2001, 1, 193-202.
used in our study.
(18) Pline, W. A.; Wu, J.; Hatzios, K. K. Effects of temperature and
chemical additives on the response of transgenic herbicide-
resistant soybeans to glufosinate and glyphosate applications.
ACKNOWLEDGMENT
Pestic. Biochem. Physiol. 1999, 65, 119-131.
We thank Albert Tidwell, Cassia Mizuno, and Gloria Hervey
(19) Alferness, P. L.; Wiebe, L. A. Determination of glyphosate and
for technical assistance.
aminomethylphosphonic acid in crops by capillary gas chroma-
tography with mass-selective detection: Collaborative study. J.
LITERATURE CITED
AOAC Int. 2001, 84, 823-846.
(20) Steel, R. G. D.; Torrie, J. H. Principles and Procedures of
(1) Potter, S. M. Soy protein and serum lipids. Curr. Opin. Lipidol.
Statistics: A Biometrical Approach, 2nd ed.; McGraw-Hill: New
1996, 7, 260-264.
York, 1980; p 176.
(2) Murphy, P. S.; Farmakalidis, E.; Johnson, L. D. Isoflavone
(21) Lydon, J.; Duke, S. O. Glyphosate induction of elevated levels
content of soya-based laboratory animal diets. Food Chem.
of hydroxybenzoic acids in higher plants. J. Agric. Food Chem.
Toxicol. 1982, 20, 315-317.
1988, 36, 813-818.

---

344
J. Agric. Food Chem., Vol. 51, No. 1, 2003
Duke et al.
(22) Harring, T.; Streibig, J. C.; Husted, S. Accumulation of shikimic
(28) Cessna, A. J.; Darwent, A. L.; Townley-Smith, L.; Harker, K.
acid: A technique for screening glyphosate efficiency. J. Agric.
N.; Kirkland, K. J. Residues of glyphosate and its metabolite
Food Chem. 1998, 46, 4406-4412.
AMPA in canola seed following preharvest applications. Can.
(23) Pline, W. A.; Wilcut, J. W.; Duke, S. O.; Edmisten, K. L.; Wells,
J. Plant Sci. 2000, 80, 425-431.
R. Accumulation of shikimic acid in response to glyphosate
(29) Baldwin, F. L. Transgenic crops: a view from the U.S. Extension
applications in glyphosate-resistant and conventional cotton
Service. Pest Manage. Sci. 2000, 56, 584-585.
(Gossypium hirsutum L.). J. Agric. Food Chem. 2002, 50, 506-
(30) Komossa, D.; Gennity, I.; Sandermann, H. Plant metabolism of
512.
herbicides with C-P bonds: glyphosate. Pestic. Biochem.
(24) Komives, T.; Casida, J. E. Acifluorfen increases the leaf content
Physiol. 1992, 43, 85-94.
of phytoalexins and stress metabolites in several crops. J. Agric.
(31) Hoagland, R. E. Effects of glyphosate on metabolism of phenolic
compounds: VI. Effects of glyphosine and glyphosate metabo-
Food Chem. 1983, 31 (4), 751-755.
lites on phenylalanine ammonia-lyase activity, growth, and
(25) Suttle, J. C.; Schreiner, D. R. Effects of DPX-4189 (2-chloro-
protein, chlorophyll, and anthocyanin levels in soybean (Glycine
N-((4-methoxy-6-methyl-1,3,5-triazin-2-yl)aminocarbonyl)ben-
max) seedlings. Weed Sci. 1980, 28, 393-400.
zenesulfoxamide) on anthocyanin synthesis, phenylalanine am-
(32) Cessna, A. J.; Darwent, A. L.; Townley-Smith, L.; Harker, K.
monia lyase activity, and ethylene product in soybean hypocotyls.
N.; Kirkland, K. J. Residues of glyphosate and its metabolite
Can. J. Bot. 1982, 60, 741-745.
AMPA in field pea, barley and flax seed following preharvest
(26) Anonymous. Pesticide tolerances for glyphosate. Fed. Regist.
applications. Can. J. Plant Sci. 2002, 82, 485-489.
1996, 61 (67, April 5), 15192-15196 (Rules and Regulations).
(27) Cessna, A. J.; Darwent, A. L.; Kirkland, K. J.; Townley-Smith,
Received for review August 21, 2002. Revised manuscript received
L.; Harker, K. N.; Lefkovitch, L. P. Residues of glyphosate and
October 31, 2002. Accepted November 1, 2002.
its metabolite AMPA in wheat seed and foliage following
preharvest applications. Can. J. Plant Sci. 1994, 74, 653-661.
JF025908I

---