Chloroaniline/lignin conjugates as model system for nonextractable pesticide residues in crop plants

Article abstract of DOI:10.1021/es9801588

In vitro lignins formed by the peroxidase/H₂O₂-mediated polymerization of coniferyl alcohol in the presence of 3,4-dichloroaniline or [¹⁵N]aniline were studied by ¹H, ¹³C, and ¹⁵N NMR spectroscopy. The anilines were >95% bound to the benzylic α-position of lignin side chains. Mild acid hydrolysis under simulated stomach conditions (0.1 M HCl, 37 °C) was studied as a first estimate of animal bioavailability. The two extremes of facile or slow acid hydrolysis that are known for chloroaniline/lignin complexes could be reproduced by using low or high incorporation ratios of aniline to coniferyl alcohol (10 or 40 mol %, respectively). The case of facile acid hydrolysis and high animal bioavailability may be due to the high mole ratios used and may not be relevant for pesticidal crop plant residue levels of 3,4- dichloroaniline. The latter are typically in the parts per million range. On the basis of ¹⁵N NMR spectral fine structure, we propose that the acid- labile linkage may be due to anchimeric assistance in conformers formed at the high aniline molar ratio. The optimized methods presented here allow the use of in vitro lignin copolymers as a reference system for the structural features and the bioavailability of nonextractable pesticide residues in crop plants. In vitro lignins formed by the peroxidase/H₂O₂-mediated polymerization of coniferyl alcohol in the presence of 3,4-dichloroaniline or [¹⁵N]aniline were studied by ¹H,¹³C, and ¹⁵N NMR spectroscopy. The anilines were >95% bound to the benzylic α-position of lignin side chains. Mild acid hydrolysis under simulated stomach conditions (0.1 M HCl, 37 °C) was studied as a first estimate of animal bioavailability. The two extremes of facile or slow acid hydrolysis that are known for chloroaniline/lignin complexes could be reproduced by using low or high incorporation ratios of aniline to coniferyl alcohol (10 or 40 mol %, respectively). The case of facile acid hydrolysis and high animal bioavailability may be due to the high mole ratios used and may not be relevant for pesticidal crop plant residue levels of 3,4-dichloroaniline. The latter are typically in the parts per million range. On the basis of ¹⁵N NMR spectral fine structure, we propose that the acid-labile linkage may be due to anchimeric assistance in conformers formed at the high aniline molar ratio. The optimized methods presented here allow the use of in vitro lignin copolymers as a reference system for the structural features and the bioavailability of nonextractable pesticide residues in crop plants.

Full text of DOI:10.1021/es9801588

Environ. Sci. Technol. 1998, 32, 2113-2118
Figure 1. Pyrolysis/ mass spectrometry has indicated a
Chloroaniline/Lignin Conjugates as
Model System for Nonextractable
Pesticide Residues in Crop Plants
preferential binding of 3-chloroaniline and 3,4-dichloroa-
niline (DCA) to the R-carbon of lignin side chains (4). This
position of chloroaniline binding has also been demonstrated
1
13
by H and C NMR spectroscopy of in vitro lignins (dehy-
drogenation polymers, DHP), which were synthesized by
2 2
peroxidase/ H O -mediated polymerization of coniferyl al-
cohol in the presence of 4-chloroaniline or DCA (5).
B . M A R K U S L A N G E , † , ‡
N O R B E R T H E R T K O R N , A N D
H E I N R I C H S A N D E R M A N N , J R . * , †
§
The ecotoxicological significance of nonextractable pes-
ticide residues depends on the bioavailability to an animal
that ingests it and the potential release of low molecular
weight xenobiotic fragments. Bound pesticide residues in
plants generally have but low bioavailability in animals (6,
GSF Forschungszentrum f u¨ r Umwelt und Gesundheit GmbH,
Institut f u¨ r Biochemische Pflanzenpathologie, D-85764
Oberschleissheim, Germany, and GSF Forschungszentrum f u¨ r
Umwelt und Gesundheit GmbH, Institut f u¨ r O¨ kologische
Chemie, D-85764 Oberschleissheim, Germany
7
). Feeding of a nonextractable wheat metabolite fraction
1
4
of [ C] DCA to rats and lambs led to the excretion of 11-20%
of the bound xenobiotic in the form of soluble metabolites.
Under comparable conditions, lignin DHPs containing 14 or
44 mol % chloroanilines released about 65% of the bound
xenobiotic (8, 9), indicating that structural differences existed
between the wheat and the enzymatic chloroaniline/ lignin
conjugates. So far, no experimental evidence exists to explain
the difference in bioavailability. It has also remained open
whether the case of high bioavailability (8, 9) is relevant for
pesticidal crop plant residue levels of chlorinated anilines.
In vitro lignins formed by the peroxidase/H2O2-mediated
polymerization of coniferyl alcohol in the presence of 3,4-
15
1
13
dichloroaniline or [ N]aniline were studied by H, C,
15
and N NMR spectroscopy. The anilines were >95% bound
to the benzylic R-position of lignin side chains. Mild acid
hydrolysis under simulated stomach conditions (0.1 M HCl,
37 °C) was studied as a first estimate of animal bioavail-
Here, we present an improved DHP model system that
ability. The two extremes of facile or slow acid hydrolysis
that are known for chloroaniline/lignin complexes could be
reproduced by using low or high incorporation ratios of
aniline to coniferyl alcohol (10 or 40 mol %, respectively). The
case of facile acid hydrolysis and high animal bioavailability
may be due to the high mole ratios used and may not be
relevant for pesticidal crop plant residue levels of 3,4-
dichloroaniline. The latter are typically in the parts per
15
utilizes derivatives of [ N]- and unlabeled aniline derivatives
1
13
15
and their characterization by H, C, and N NMR spec-
troscopy. Mild acid hydrolysis under simulated stomach
conditions (0.1 M HCl, 37 °C) is used to mimick the digestion
in the animal stomach. Variation of the aniline incorporation
rate allowed us to produce and characterize chloroaniline/
lignin conjugates with high or low acid sensitivity.
15
Experimental Section
million range. On the basis of N NMR spectral fine structure,
we propose that the acid-labile linkage may be due to
anchimeric assistance in conformers formed at the high
aniline molar ratio. The optimized methods presented
here allow the use of in vitro lignin copolymers as a reference
system for the structural features and the bioavailability
of nonextractable pesticide residues in crop plants.
Chem icals. Coniferyl alcohol was purchased from Fluka
(Neu-Ulm, Germany), 3,4-dichloroaniline was from Riedel
1
5
de Haen (Seelze, Germ any), [ N]aniline was from
Chemotrade (Leipzig, Germany), and horseradish peroxidase
1
4
was from Boehringer (Mannheim, Germany). [Ring U- C]-
DCA (Sigma, St. Louis, MO) was purified by HPLC to 99%
(
10). All reagents for NMR spectroscopy were obtained from
Merck (Darmstadt, Germany).
Introduction
Synthesis of Model Com pounds. erythro-1-(4-Hydroxy-
A wide range of commercially important herbicides contains
chlorinated anilines as structural components, e.g., acyla-
nilide, N-phenylurea, and carbamyl derivatives. The free
chloroanilines, which are formed as primary metabolites,
are conjugated in plants, yielding soluble as well as insoluble
3
-m ethoxyphen yl)-2-(2-m ethoxyphen oxy)propan e-1,3-
diol (2a) was synthesized according to Nakatsubo et al. (11).
-(4-Hydroxy-3-methoxyphenyl)-3-(3,4-dichlorophenylami-
3
15
no)-2-(2-methoxyphenoxy)propan-1-ol (2b) and [ N]-3-(4-
hydroxy-3-m ethoxyphenyl)-3-phenylam ino-2-(2-m ethox-
yphenoxy)-propan-1-ol (2c) were synthesized as follows. In
a 100-mL two neck round-bottom flask, 0.1 mmol of 2a in
(
nonextractable) conjugates (1, 2). In intact plants and plant
cell cultures, the operationally defined lignin fraction has
been identified as a major covalent binding site for nonex-
tractable chloroaniline residues (3-5).
Lignins are formed by an oxidative polymerization of
monolignol precursors such as coniferyl alcohol (1 in Figure
4
0 mL of dichloromethane was placed under a nitrogen
atmosphere. At room temperature, 1 mmol of trimethylsilyl
bromide was added within 1 min to the stirred solution with
a syringe through a septum, and the reaction was allowed
to proceed for 4 min to yield the R-brominated product. The
reaction mixture was washed twice with 20 mL of saturated
NaCl. The organic phase containing the quinone methide
was dried over 3 g of Na
removed by a stream of dry nitrogen at 0 °C. Chloroform (20
mL) was added to the residue, and Na SO was removed by
filtration. To the organic solution, 0.2 mmol of 3,4-dichlo-
roaniline or [ N]aniline was added during a 20-min period
under a nitrogen atmosphere for the synthesis of 2b and 2c,
respectively. The mixture was incubated for 2 h at 22 °C. An
aliquot was examined for product purity by analytical HPLC
1
). The proposed reaction mechanism for lignin biosynthesis
involves the formation of a mesomeric monolignol radical
that gives rise to achiral and polydisperse polymers with a
multitude of binding types. The main interunit linkage
present in lignins is the â-aryl ether structure shown in 2 of
2 4
SO for 30 min, and the solvent was
2
4
*
Corresponding author e-mail: sandermann@gsf.de; telephone:
+ †
49 89 3187 2285; fax: +49 89 3187 3383.
15
Institut f u¨ r Biochemische Pflanzenpathologie.
‡
Present address: Institute of Biological Chemistry, Washington
State University, Pullman, WA 99164-6340.
§
Institut f u¨ r O¨ kologische Chemie.
S0013-936X(98)00158-8 CCC: $15.00
Published on Web 05/30/1998

1998 Am erican Chem ical Society
VOL. 32, NO. 14, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2 1 1 3

FIGURE 1. Structures of the lignin precursor, coniferyl alcohol (1), of the â-aryl ether side chain structure of lignins (2), and of model
15
compounds without xenobiotic (2a) or containing 3,4-dichloroaniline (2b) or [ N]aniline (2c) in benzylamine linkages.
(
Millenium system, Waters, Eschborn, Germany); column:
Vydac (4.6 mm × 25 cm, particle size 5 µm), Separations
Group, Hesperia, CA; gradient: 0.05% H PO in water (5 min),
followed by a linear gradient from 0 to 100% acetonitrile in
difference spectra were recorded with 90° pulses to avoid
SPT effects. For semiquantitative 15N NMR spectra of DHPs,
inverse gated decoupling with a relaxation delay of 45 s was
3
4
1
5
used; N-refocused-INEPT spectra were recorded with a 90°
-
1
1
water (45 min) with a solvent flow of 1 mL min ; eluting
compounds were detected with a diode array detector
(
2
H purge pulse at the end of the t delay. Typical conditions
for 2D NMR spectra: magnitude HH-COSY: 1K data points
in F2 and 256 increments in F1; phase-sensitive (TPPI)
HH-COSY: 2K data points in F2 and up to 1K increments
in F1 corresponding to a digital resolution in F2 of 1.5 Hz;
heteronuclear correlation spectra: 4K data points in F2 and
256 increments in F1.
Mild Acid Hydrolysis. A threo/ erythro mixture of com-
pound 2b (threo/ erythro ratio: 85:15) was incubated in 100
µL of 0.1 M HCl at 37 °C. Hydrolysis was terminated by
addition of 110 µL of 0.1 M NaOH, followed by extraction
with 200 µL of ethyl acetate. The decrease of the amount of
Waters, Eschborn, Germany) between 200 and 450 nm. Two
main products were detected and separated by semi-
preparative HPLC (Millenium system, Waters, Eschborn,
Germany); column: Nucleosil C-18 (10 mm × 25 cm, particle
size 5 µm), Knauer, Berlin, Germany; gradient: 50% aceto-
nitrile in water (5 min), followed by a linear gradient to 100%
acetonitrile (25 min) with a solvent flow of 3 mL min (diode
array detection between 200 and 450 nm). The purity of the
homogeneous compounds was assessed by diode array HPLC
-1
(
(
(
t
analytical method as above. R : 2a, 19.2 min; 2b, 28.9 min
threo isomer) and 29.2 min (erythro isomer); 2c, 25.9 min
threo isomer) and 26.8 min (erythro isomer). The identity
2b [R
t
, 28.9 min (threo isomer); R
t
, 29.2 min (erythro isomer)]
and the increase of DCA (R
t
, 22.6 min) and 2 (R
t
, 19.2 min)
of the compounds was assured by UV, IR, and NMR
spectroscopy (only NMR data are shown in this report; for
details see ref 10).
Synthesis of Dehydrogenation Polym ers (DHPs). All
i
syntheses were carried out in 0.1 M sodium phosphate (NaP )
were followed by diode-array RP-HPLC analysis and used to
quantify the hydrolysis rate (internal standard: 2,4-dichlo-
t
roaniline; R , 25.8 min). Alternatively, hydrolysis was followed
1
by H NMR spectroscopy without an extraction step. DHP
3 and DHP 5 were incubated in an NMR tube in DMSO-
buffer (pH 6.5) under a nitrogen atmosphere at room
temperature (10). To a stirred solution of 2000 U of
6 2
d / D O (3:1, v/ v) acidified with HCl to pH 1.0 at 37 °C. The
1
5
protonated [ N]aniline accumulating during hydrolysis was
1
peroxidase in 100 mL of NaP
mmol in 75 mL of NaP ) and of coniferyl alcohol (2.0 mmol
in 75 mL of NaP ) were added without an additional
component to yield DHP 1. Aniline substrates (DCA; 2.0 or
.2 mmol in 75 mL of NaP to yield DHP 2 and DHP 4,
respectively) or [ N]aniline (2.0. or 0.2 mmol in 75 mL of
NaP to yield DHP 3 and DHP 5, respectively) were also
employed. All solutions were added at rates of 20 mL h
i
, separate solutions of H
2
O
2
(2.5
quantified by H signal integration (H-2, H-6: 7.34 ppm;
i
H-4: 7.40 ppm; H-3, H-5: 7.45 ppm); all signals assigned to
1
5
i
protonated [ N] aniline had higher ppm values than
copolymer signals (details see ref 10). In addition, DHP
hydrolysis was followed by 15N NMR analysis. DHP 3 and
0
i
1
5
15
DHP 5 were treated as described above. The changes in
N
i
signal intensity of the resonances at -313.7 and -317.3 ppm
-
1
15
.
(bound [ N]aniline; threo and erythro isomer, respectively)
1
4
15
A[ C]-3,4-DCA tracer was included in some of the 3,4-DCA
experiments (10). The total reaction time was 5 h. The
formed DHPs were pelleted by centrifugation (25 min,
and at -323.7 ppm (free [ N]aniline) were used to quantify
the reaction.
Results and Discussion
1
5000g); washed with 20 mL of 1% (w/ v) SDS, 20 mL of 1 M
NaCl, and 3 × 20 mL of H O; lyophylized; and stored at -20
C until further analysis. Polymer yields with regard to
2
Synthesis and Characterization of Model Com pounds. The
present investigation on the binding type of chloroanilines
in DHPs employs NMR spectra of model compounds related
to the lignin â-aryl ether subunit (2), which is the predominant
linkage found in native lignins. The anilines were attached
to the R-position by a biomimetic mechanism involving a
quinone methide intermediate resulting in the formation of
benzylamines. 2b and 2c (Figure 1) have two optical centers,
corresponding to a mixture of threo (RR/ SS) and erythro
(RS/ SR) forms. Accordingly, two main diastereomeric reac-
tion products were obtained with 3,4-dichloroaniline (DCA)
°
incorporated coniferyl alcohol were always >90%.
NMR Spectroscopy. The NMR spectra were acquired at
3
03 K with a Bruker AC 400 spectrometer operating at 400.13
MHz proton frequency. Dimethyl sulfoxide-d was used as
6
1
5
solvent and internal reference (2.49/ 39.00 ppm). N NMR
spectra were referenced to internal nitromethane (0.00 ppm),
1
13
and methanol was an internal reference for H/ C. Proton
and HH-COSY spectra were acquired with an inverse 5-mm
1
3
15
broadband probe (90°: 8 µs), while C and N (90°: 7.5 and
1
5
1
(
3.5 µs) as well as two-dimensional CH- and NH-XHCORR
or [ N]aniline as substrates as determined by HPLC analysis.
The substances were purified by semipreparative HPLC and
subjected to NMR spectroscopy. By comparison of these
data with literature data (12-15), the presence of a threo:
1
n
J: 150, 80 Hz) and COLOC [ J: 10, 7.5 (CH), 4 (NH)] spectra
1
3
were obtained with a broadband 5 mm probe (90°: C, 8 µs;
1
5
N, 13.5 µs) using Bruker standard software. Proton NOE
2
1 1 4
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 14, 1998

1
13
TABLE 1. H and C NMR Data of Model Substances for the Binding Type of Chloroanilines in Dehydrogenation Polymers
1H NMR data
chemical shift
ppm)
(threo) (erythro) (intergral) (threo) (erythro)
13C NMR data
chemical shift
(ppm)
(
splittinga
coupling const. (Hz)
1
13
assign-
ment
assign-
ment
cross-peaks in H, C COSY,
1
13
compd
2
(threo) (erythro) long-range couplings ( H, C-COLOC)
2
2
3
a
H-2
7.03
d (1 H)
2.0
C-1
133.28 J (H-2), J (H-R), J (H-5),
3
J (OH-R)
1
3
3
H-5
H-6
OCH3-3
OH-4
H-3′
H-4′
H-5′
H-6′
6.69
6.79
3.66
8.70
6.90
6.84
6.80
6.98
d (1 H)
dd (1 H)
s (3 H)
8.0
C-2
111.30 J (H-2), J (H-6), J (H-R)
3 3 3
8.0, 2.0 C-3
C-4
146.99 J (H-2), J (OCH3-3), J (H-5)
3
3
145.47 J (H-2), J (H-6)
1
2
s (1 H)
C-5
114.60 J (H-5), J (H-6)
1
2
3
3
dd (1 H)
dd (1 H)
m (1 H)
dd (1 H)
7.4, 2.1 C-6
7.4, 2.1 OCH3-3
C-1′
119.56 J (H-6), J (H-5), J (H-2), J (H-R)
1
55.47 J (m ethyl H atom s)
2
4
148.13 J (H-6′), J (H-4′)
2
3
3
7.5, 2.1 C-2′
149.81 J (H-3′), J (OCH3-2′), J (H-4′),
3
J (H-6′)
1
2
OCH3-2′
H-R
H-â
H-γ
OH-R
OH-γ
3.65
4.73
4.31
3.61
5.27
4.58
s (3 H)
dd (1 H)
m (1 H)
m (2 H)
d (1 H)
t (1 H)
C-3′
C-4′
C-5′
112.75 J (H-3′), J (H-4′)
1
2
3
4.7
4.7
121.02 J (H-4′), J (H-3′), J (H-6′)
1
2
3
120.68 J (H-5′), J (H-6′), J (H-3′)
1
2
11.6, 4.7 C-6′
4.7
4.7
116.00 J (H-6′), J (H-5′)
1
OCH3-2′
55.63 J (m ethyl H atom s)
1
2
4
C-R
C-â
C-γ
71.71 J (H-R), J (OH-R), J (H-2)
1 3 3
83.79 J (H-â), J (OH-â), J (OH-R)
1
60.15 J (H-γ)
2
2
2
bb H-2
H-5
7.28
7.00
6.79
3.70
6.95
6.86
6.80
6.93
7.26
6.99
6.79
3.71
6.94
6.85
6.78
6.91
3.72
6.86
7.19
6.61
4.83
4.72
d (1 H)
d (1 H)
2.3
8.2
2.4
8.3
C-1
C-2
138.50 138.30 J (H-2), J (H-6)
1
112.05 112.25 J (H-2)
2
H-6
OCH3-3
H-3′
H-4′
H-5′
H-6′
dd (1 H) 8.2, 2.3 8.3, 2.4 C-3
s (3 H) C-4
dd (1 H) 8.0, 1.8 7.9, 1.9 C-5
dd (1 H) 8.1, 2.0 7.9, 1.9 C-6
150.61 150.63 J (H-2)
3
3
138.30 138.32 J (H-2), J (H-6)
1
122.39 122.47 J (H-5)
1
3
119.45 119.52 J (H-6), J (H-2)
1
m (1 H)
OCH3-3
55.49 55.69 J (m ethyl H atom s)
dd (1 H) 8.2, 2.0 8.3, 2.1 C-1′
147.19 147.37
2
OCH3-2′ 3.73
s (1 H)
d (1 H)
d (1 H)
C-2′
C-3′
C-4′
150.28 150.09 J (H-3′)
1
H-2′′
H-5′′
H-6′′
H-R
H-â
H-γ
6.88
7.20
6.64
4.83
4.66
2.7
8.8
2.7
8.9
112.69 112.81 J (H-3′)
1
124.38 124.27 J (H-4′)
1
dd (1 H) 8.8, 2.7 8.9, 2.7 C-5′
d (1 H)
m (1 H)
122.64 122.79 J (H-5′)
1
2
5.4
5.6
C-6′
120.58 120.73 J (H-6′), J (H-5′)
1
OCH3-2′
C-1′′
C-2′′
C-3′′
C-4′′
C-5′′
C-6′′
C-R
55.70 55.64 J (m ethyl H atom s)
2
3
4.05 (γ1) 4.15 (γ1) dd (1 H) 11.3, 6.0
147.86 147.94 J (H-2′′), J (H-5′′)
1
4
.16 (γ2) 4.27 (γ2) dd (1 H) 11.3, 4.1
113.86 113.69 J (H-2′′)
1
2
3
131.04 131.18 J (H-3′′) J (H-2′′), J (H-5′′)
3
3
116.95 116.61 J (H-2′′), J (H-6′′)
1
130.30 130.32 J (H-5′′)
1
3
113.24 113.41 J (H-6′′), J (H-2′′)
1
3
57.34 57.27 J (H-R), J (H-â)
1
1
(H- ), 3
, 3
3
C-â
C-γ
80.64 80.82
J
â
J (H-R) J (H-γ1), J (H-γ2)
3
63.15 63.17 J (H-γ), J (H-â)
2
c
H-2
H-5
H-6
6.98
6.63
6.79
3.78
6.97
6.89
6.78
7.01
7.07
6.64
6.80
3.79
6.97
6.89
6.81
7.01
3.72
6.53
6.98
6.36
4.59
4.41
m (1 H)
d (1 H)
d (1 H)
s (3 H)
d (1 H)
C-1
C-2
C-3
C-4
C-5
C-6
OCH3-3
C-1′
C-2′
C-3′
C-4′
C-5′
C-6′
OCH3-2′
C-1′′
131.98 130.71 (only COSY data available)
111.53 112.35 J (H-2)
1
8.0
8.0
8.1
147.29 147.14
114.97 145.39
114.97 114.76 J (H-5)
119.71 119.72 J (H-6)
55.65 55.71 J (m ethyl H atom s)
OCH3-3
H-3′
H-4′
H-5′
H-6′
OCH3-2′ 3.67
H-2′′,H-6′′ 6.55
H-3′′,H-5′′ 6.98
H-4′′
H-R
H-â
H-γ
1
8.1
1
dd (1 H) 7.6, 1.7
m (1 H)
1
d (1 H)
s (3 H)
d (2 H)
m (2 H)
t (1 H)
7.9
7.9
8.4
147.66 147.90
150.08 150.17
112.65 112.68 J (H-3′)
121.96 121.95 J (H-4′)
120.68 120.79 J (H-5′)
1
8.4
1
1
6.47
4.62
4.30
7.2
4.0
7.2
4.2
1
d (1 H)
m (1 H)
117.23 117.55 J (H-6′)
55.47 55.46 J (m ethyl H atom s)
1
3.39 (γ1) 3.45 (γ1) m (1 H)
147.92 147.61
1
1
3
.63 (γ2) 3.54 (γ2) m (1 H)
C-2′′,C-6′′ 113.10 113.89 J (H-2′′), J (H-6′′)
C-3′′,C-5′′ 128.64 128.74 J (H-3′′), J (H-5′′)
C-4′′
C-R
C-â
C-γ
1
1
1
116.00 114.70 J (H-4′′)
1
56.65 57.05 J (H-R)
1
84.32 83.55 J (H-â)
1
59.87 60.21 J (H-γ)
a
s, singlet; d, doublet; dd, doublet of doublets; m , m ultiplet; t, triplet. b Acetylated, acetyl group signals not listed.
1
1
13
1
erythro ratio of 85:15 was determined for both 2b and 2c.
The most informative techniques for the characterization of
the model compounds were two-dimensional NMR experi-
ments involving H, H and C, H correlation spectroscopy,
which allowed the assignment of all resonances, as sum-
marized in Table 1.
VOL. 32, NO. 14, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2 1 1 5

anilines by means of two-dimensional NMR techniques (DHP
: H-6, 6.52 ppm; H-2, 6.59 ppm; H-5, 7.18 ppm; DHP 3:
2
H-4, 6.48 ppm; H-2, H-6, 6.57 ppm; H-3, H-5, 6.99 ppm; see
Table 2).
The 13C NMR spectra of DHPs 1-3 are shown in Figure
3
(
. Lignin resonances of DHP 1 (Table 2) were readily assigned
10) by reference to the extensive 13C NMR data in the
literature (12-15). Spectral features related to the incor-
poration of anilines were of special interest. Therefore, only
resonances related to the covalent binding of anilines are
discussed here. In comparison to the 13C NMR spectra of
DHP 1, a remarkable sharpening of nearly all resonances
was observed with DHP 2 and DHP 3, and a number of new
signals appeared. On the basis of the model compound
spectra (Table 1) and 2D NMR spectroscopy [HH-COSY,
CH-COSY, CH-COLOC; NH-COLOC], resonances of DCA
in DHP 2 [113.3 ppm (C-6), 113.8 ppm (C-2), 116.5 ppm (C-
4
[
4
), 130.9 ppm (C-5), 131.0 ppm (C-3), 150.0 ppm (C-1)], and
N]aniline in DHP 3 [113.1 ppm (C-2, C-6), 115.0 ppm (C-
), 128.7 ppm (C-3, C-5), 148.0 ppm (C-1)] were assigned
1
5
(
Table 2). The most striking new spectral features of the
aniline/ lignine copolymers concerned the side chain signals.
Concomitant with a decrease of the signal intensity of C-R
(71.2 ppm) in the copolymer spectra, distinctive new
FIGURE 2. 1H NMR spectra of dehydrogenation polymers (DHPs) in
dimethyl sulfoxide-d at 303 K; A, reference lignin (DHP 1); B, DCA/
lignin copolymer (DHP 2); C, [ N]aniline/lignin copolymer (DHP 3).
1
13
resonances were detected at 56.7 and 84.3 ppm. The H, C-
6
15
1
COSY as well as the dublet splitting at 56.7 ppm ( J(NC) ) 10
Hz) clearly indicated that these signals were caused by C-R
and C-â, respectively. Comparison of these chemical shifts
with the model substance data (Table 1) provided unam-
biguous evidence for the benzylamine binding of the anilines
Synthesis and General Characterization of Chloroa-
niline/Lignin Conjugates. The in vitro polymerization of
coniferyl alcohol by horseradish peroxidase/ H O in the
2 2
presence or absence of aniline cosubstrates was carried out
by modification of the previously described procedure (5).
A reference DHP (DHP 1) was synthesized from coniferyl
alcohol 1 alone. Aniline/ lignin conjugates were prepared
from 1 plus DCA (DHP 2, employed molar ratio of 1 to DCA
(
Table 2).
The 15N NMR spectra of DHP 3 and DHP 5 (Figure 4)
revealed two signal groups of exactly the same chemical shift
values as observed with model compound 2c corresponding
to threo (-317.3 ppm) and erythro (-313.7 ppm) isomers in
the ratio of 85:15. These peaks exhibited a fine structure
that depended on the aniline molar ratio (Figure 4C). We
propose preferred conformations of the substituents at the
benzylic carbon in DHPs at higher proportions of incorpo-
rated aniline and deduce the existence of different stable
conformations from this fine structure of the 15N NMR
)
1:1); DHP 4 (employed molar ratio of 1 to DCA ) 10:1)
15
or 1 plus [ N]aniline (DHP 3, employed molar ratio of 1
1
5
to [ N]aniline ) 1:1); DHP 5 (employed molar ratio of 1
1
5
to [ N]aniline ) 10:1). The insoluble reaction products
were washed with strong detergent and repeatedly with
water.
resonances in the spectra of DHP 3 and DHP 5 (cf. Figure
High-perform ance size-exclusion chrom atography
HPSEC) was performed in dimethyl formamide using
1
1
). The signal pattern of INVGATE and J(NH)-INEPT and
(
1
15
J(NH)-DEPT-90 N NMR spectra of DHP 3 and DHP 5 are
identical, and this indicates that all resonances originate from
singly protonated nitrogen atoms. Considering the unifor-
mity of influences of the molecular environment on the
magnetic shielding constant in 13C and N NMR of unpro-
tonated aromatic amines, one would expect a corresponding
fine structure in the 13C NMR spectra of DHP 3 and DHP 5
if the fine structure is caused by long-range effects originating
from differences in the major interunit linkages within the
DHPs. However, this was not observed in any of the 1D and
polystyrene molecular weight standards (10). The DHP
copolymers were found to be polydisperse with a broad
maximum at 8-11 kDa. Absorbance at 280 nm and
incorporated radioactivity showed coelution at a constant
ratio over the entire molecular weight range (10).
15
Incorporation rates of the various anilines were deter-
mined through radioactivity content of a DHP synthesized
1
4
1
with a [ C]DCA tracer, integration of H NMR spectra, and
elemental analysis of DHPs 2-5. The molar ratios of coniferyl
alcohol to aniline found were 2.6 ( 0.1:1 for DHP 2 and DHP
2
D proton and carbon NMR spectra.
3
and 10.5 ( 0.5:1 for DHP 4 and DHP 5. These values
corresponded to 40 mol % aniline in DHP 2 and DHP 3 and
At lower molar aniline ratios, the 15N NMR signal splitting
was much less marked (Figure 4B). The 1H, N-COLOC
spectrum of DHP 3 showed cross-peaks of -317.3 to 4.63
ppm (threo) and of -313.7 to 4.61 ppm (erythro), indicating
a geminal NH-R coupling (data not shown). The proposed
binding of chloroanilines to aromatic guaiacyl rings (9) was
not observed. An azomethine binding of DCA or [ N]aniline
to C-γ of cinnamyl aldehyde side chains was likewise not
detected by any of the NMR techniques applied. Azomethine
model compounds were highly unstable in aqueous buffer
(data not shown). Thus, the benzylamine attachment was
the only binding type of anilines detected.
15
1
0 mol % aniline in DHP 4 and DHP 5, respectively.
Spectral Assignm ents. The H NMR spectra of DHPs 1-3
1
are shown in Figure 2. A broad multiplet assigned to H-γ
appeared at 3.61-3.80 ppm. A further common feature was
the presence of olefinic propenyl side chains with typical
signals at 6.46 ppm (H-R, d, J ) 16 Hz), 6.20 ppm (H-â, m),
and 4.12 ppm (H-γ). The signal at 4.12 ppm was significantly
decreased in DHP 2 and DHP 3. The H-R and H-â resonances
in DHP 1 appeared at 4.70 and 4.32 ppm, respectively. DHP
1
5
2
and DHP 3 showed an additional signal at 4.63 ppm, which
was assigned to H-Rin a benzylamine bonding by comparison
1
13
with the model substances. The H- and C NMR spectra
of DHP 2 and DHP 3 showed a better resolution of signals
than DHP 1, allowing the assignment of the signals in the
lignin conjugate spectra attributed to the incorporated
Hydrolytic Stability. The structural analyses were comple-
mented by reactivity studies. The hydrolysis kinetics of the
benzylamine bond at pH 1.0 and 37 °C (simulated stomach
conditions;9) was investigated by three independent methods
2
1 1 6
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 14, 1998

TABLE 2. Assignment of NMR Signals in In Vitro Synthesized Insoluble DCA/Lignin Conjugates Using Model Substancesa
assignment
δ in 2a
δ in 2b
δ in 2c
δ in DHP 1
δ in DHP 2
δ in DHP 3
H-R
4.73
4.31
71.7
84.8
4.63
4.31
57.0
80.6
4.61
4.30
56.7 (d) (9.9 Hz)
83.6
4.70
4.10
71.2
84.9
4.73; 4.63
4.32 (m )
71.3; 56.7
84.9; 84.1
4.73; 4.63
4.30 (m )
71.0; 56.7 (10 Hz)
84.9; 84.2
-317.3 (threo)
-313.7 (threo)
H-â
1
C-R [ J (NC)]
C-â
1
5N[aniline]
-317.2 (threo)
-313.6 (erythro)
a
The chem ical shift values of additionally appearing signals in DHP 2 and DHP 3, when com pared to DHP 1, are italic (d, dublet; m , m ultiplet).
All spectra were determ ined in DMSO-d Shift values are given in ppm units.
6
.
FIGURE 3. 13C NMR spectra of dehydrogenation polymers (DHPs)
in dimethyl sulfoxide-d at 303 K; A, reference lignin (DHP 1); B,
DCA/lignin copolymer (DHP 2); C, [ N]aniline/lignin copolymer (DHP
).
6
15
3
1
5
for the following materials: DCA in 2b and [ N]aniline in
DHP 3 and DHP 5. RP-HPLC analysis of the hydrolysis of
2
b resulted in the accumulation of DCA and 2a, which were
quantified by diode array detection with 2,4-dichloroaniline
as an internal standard. 2b hydrolyzed linearly with a rate
of 7% per day (Figure 5). For a H NMR analysis of hydrolysis,
DHP 3 and DHP 5 were incubated in DMSO-d / D O (3:1) in
6 2
the NMR spectrometer (pH 1.0, 37 °C), and the liberated
1
FIGURE 4. 15N NMR spectra of (A) model compound 2c of Figure
1, (B) DHP 5, and (C) DHP 3. The smaller peak to the left corresponds
to the erythro configuration, and the main peak to the right
corresponds to the threo configuration.
protonated anilines were quantified by signal integration.
1
5
The hydrolysis kinetics of DHP 3 (40 mol % [ N]anilines)
displayed the characteristics of a nonlinear first-order
mined as 30% from DHP 3 and 15% from DHP 5 in accordance
with the hydrolysis characteristics obtained by 1H NMR
spectroscopy.
1
5
reaction with a rapid release of [ N]aniline within the first
0 min (16% initial release) followed by a significantly slower
hydrolysis (to 20% DCA after 3 h and 25% DCA after 15 h).
1
The faster hydrolysis of bound anilines in DHP 2 or DHP
3 (40 mol % anilines) as compared to DHP 4 or DHP 5 (10
mol % anilines) appeared to be related to anchimeric
assistance. Neighboring phenyl ether groups could be
responsible because the only significant spectral difference
between the two DHP types was the multiplet structure of
Figure 4C. Anchimeric effects are frequently utilized for the
depolymerization of lignins in wood pulping (16). In
conclusion, we present here the first unequivocal data on
the binding type of anilines in DHPs. Acid lability, which is
associated with high animal bioavailability (8, 9), was only
observed at 40 mol % aniline. This value is close to the mole
ratios of 14% and 44% employed in the animal experiments
(8) but far above the parts per million level that is typical for
pesticidal crop plant residues of chlorinated anilines. There-
fore, the previous experiments showing high animal bio-
1
5
15
[
N]Aniline release from DHP 5 (10 mol % [ N]aniline) was
much slower (6-7% after 3 h, 15% after 15 h) and reproduced
quite well the hydrolysis kinetics of nonextractable DCA
conjugates in wheat cell walls (9). As a third method, 15
N
NMR spectroscopy of DHP 3 and DHP 5 was utilized to
1
5
directly observe the hydrolytic liberation of [ N]aniline in
the NMR tube (same conditions as described above). The
integrals of the signals at -313.7 and -317.3 ppm (bound
1
5
15
[
N]aniline) and at -323.7 ppm (free [ N]aniline) were used
15
to quantify the reaction. The low sensitivity of N NMR
spectroscopy allowed integration only every 20 min with DHP
3
and every 4 h with DHP 5 (corresponding to 30 and 320
accumulated spectra, respectively). Increased broadening
of the NMR signals also impeded an exact quantification.
Still, the aniline released after 3 h was reproducibly deter-
VOL. 32, NO. 14, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2 1 1 7

would like to thank Dr. W. Heller for valuable discussion and
E. Schindlbeck for technical assistance.
Literature Cited
(
1) Still, G. G.; Herrett, R. A. In Herbicides. Chemistry, Degradation
and Mode of Action; Kearney, P. C., Kaufman, D. D., Eds.;
Dekker: New York, 1976; Vol. 2, pp 609-664.
(
2) Komossa, D.; Langebartels, C.; Sandermann, H. In Plant
Contamination. Modeling and Simulation of Organic Chemical
Processes; Trapp, S., McFarlane, J. C., Eds.; Lewis Publishers:
Boca Raton, 1995; pp 69-103.
(
(
(
(
(
3) Yih, R. Y.; McRae, D. M.; Wilson, H. F. Science 1968, 161, 376-
3
77.
4) Still, G. G.; Balba, H. M.; Mansager, E. R. J. Agric. Food Chem.
981, 29, 739-746.
1
5) von der Trenck, K. T.; Hunkler, D.; Sandermann, H. Z. Natur-
forsch. 1981, 36, 714-720.
6) Sandermann, H.; Scheel, D.; von der Trenck, T. J. Appl. Polym.
Sci., Appl. Polym. Symp. 1983, 37, 407-420.
7) Pillmoor, J. B.; Roberts, T. R. In Progress in Pesticide Biochemistry
and Toxicology; Hutson, D. H., Roberts, T. R., Eds.; John Wiley
&
Sons: New York, 1985; Vol 4, pp 85-101.
FIGURE 5. Mild acid hydrolysis (0.1 M HCl, 37 °C) of enzymatically
prepared aniline/lignin conjugates (DHPs) and of model substance
(8) Sandermann, H.; Arjmand, M.; Gennity, I.; Winkler, R.; Struble,
C. B.; Aschbacher, P. W. J. Agric. Food Chem. 1990, 38, 1877-
1
880.
2
b of Figure 1. Hydrolysis (% of initial) is plotted versus time (h).
(
9) Sandermann, H.; Musick, T. J.; Aschbacher, P. W. J. Agric. Food
Chem. 1992, 40, 2001-2007.
The release of DCA from 2b (O) was quantified by HPLC analysis.
15
The release of DCA or [ N]aniline from DHP 2 and DHP 3 (b) and
1
(10) Lange, M. Struktur und Biosynthese von StressligninensPilzlicher
Elicitor und 3,4-Dichloranilin als Stressoren. Ph.D. Dissertation,
Tectum Verlag, Marburg, 1995, ISBN 3-89608-322-8.
from DHP 4 and DHP 5 (1) was determined by quantitative H NMR
spectroscopy.
(
11) Nakatsubo, F.; Sato, K.; Higuchi, T. Holzforschung 1975, 29,
165-168.
availability (8, 9) may not be relevant for actual pesticidal
field residue levels.
(12) Ralph, J.; Helm, R. F. J. Agric. Food Chem. 1991, 39, 705-709.
In conjunction with previous studies using a native
nonextractable wheat cell wall residue (9), the present study
shows that defined DHPs can serve as simple model systems
for nonextractable pesticide residues of crop plants. The
methods presented here may help to reduce the number of
experiments on animals required in registration procedures
for new pesticides and other industrial chemicals.
(13) Lapierre, C.; Monties, B.; Guittet E.; Lallemand, J.-Y. Holzfor-
schung 1987, 41, 51-58.
(14) Brunow, G.; Sipil a¨ , J.; M a¨ kel a¨ , T. Holzforschung 1989, 43, 55-
5
9.
(
(
15) Robert D. In Methods in Lignin Chemistry; Lin, S. Y., Dence, C.
W., Eds.; Springer-Verlag: Berlin, 1993; pp 250-273.
16) Gierer, J. Holzforschung 1982, 36, 43-51.
Received for review February 18, 1998. Revised manuscript
received April 8, 1998. Accepted April 8, 1998.
Acknowledgments
This work was supported in part by Limagrain (Chappes,
France) and by Fonds der Chemischen Industrie. The authors
ES9801588
2
1 1 8
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 14, 1998