Role of hydroxylamine intermediates in the phytotransformation of 2,4,6-trinitrotoluene by Myriophyllum aquaticum

Article abstract of DOI:10.1021/es030010a

Phytotransformation studies of 2,4,6-trinitrotoluene (TNT) were conducted using Myriophyllum aquaticum to clarify the role of initial intermediates of TNT transformation in the complex product distributions reported previously. 2-Hydroxylamino-4,6-dinitrotoluene (2HA46DNT) and 4-hydroxylamino-2,6-dinitrotoluene (4HA26DNT) were the initial intermediates of TNT phytotransformation. 2HA46DNT and 4HA26DNT were both abiotically transformed to 4,4′,6,6′-tetranitro-2,2′-azoxytoluene (2,2′azoxy) and 2,2′,6,6′-tetranitro-4,4′-azoxytoluene (4,4′azoxy) and also phytoreduced to the related amines 2-amino-4,6-dinitrotoluene (2A46DNT) and 4-amino-2,6-dinitrotoluene (4A26DNT). To further elucidate the initial steps of this TNT phytotransformation pathway, the transformations of known intermediates (including 2HA46DNT, 4HA26DNT, 2A46DNT, 4A26DNT, 2,2′azoxy, and 4,4′azoxy) were monitored in plant systems. The transformation rates were measured, and kinetic analysis using pseudo-first-order models was used to evaluate the relative rates of competing reactions. The formation of the azoxy products was determined to be more rapid than the formation of the amine products. Both the azoxy and amine products were subject to uptake and further transformation by the plant.

Full text of DOI:10.1021/es030010a

Environ. Sci. Technol. 2003, 37, 3595-3600
to determine the intermediates and products formed and
their potential to increase or decrease toxicity.
Role of Hydroxylamine Intermediates
in the Phytotransformation of
2,4,6-Trinitrotoluene by
In studies of TNT metabolism by microorganisms, it has
been widely demonstrated that the aryl nitro groups are the
site of initial biological attack (13-15). Studies using several
plant species have consistently showed the formation of two
nitro-reduction products, 2-amino-4,6-dinitrotoluene (2A46-
DNT) and 4-amino-2,6-dinitrotoluene (4A26DNT). However,
these compounds generally account for a limited fraction
(∼15 mol %) of the TNT transformed by the plant (9, 11).
A previous study using the aquatic plant Myriophyllum
aquaticum showed a rapid disappearance of TNT that was
attributed to plant uptake and transformation (9). In this
study, the use of axenic plants and plant tissue cultures
confirmed that plant-associated bacteria were not responsible
for TNT transformation. The accumulation of low concen-
trations of 2A46DNT and 4A26DNT were produced by plant
metabolism and observed in the extracellular medium (9,
11). Mass balances of 14C showed that the TNT transformation
products were distributed nearly evenly between the plant
medium and the plant tissues. The 14C associated with the
plant were characterized as bound residues, as they could
be assayed only through the combustion of plant samples.
Metabolites other than the amines in the aqueous medium
were characterized as conjugation metabolites and oxida-
tion products (2-amino-4,6-dinitrobenzoate, 2,4-dinitro-6-
hydroxybenzyl alcohol, 2-N-acetoxyamino-4,6-dinitroben-
zaldehyde, and 2,4-dinitro-6-hydroxytoluene). The pathways
leading to bound, conjugated, or partially oxidized forms
are believed to involve initial reduction reactions, although
the mechanism(s) for their formation is not well understood
(6, 16).
Little is known about the initial reduction reactions of
TNT in plants. Ferredoxin NADP+ reductase isolated from
spinach leaves is known to reduce TNT to 4HA26DNT (17).
Burken et al. state “hydroxylamines are postulated as a key
intermediate in the transformation pathway of TNT, but
conclusive research is needed” (18). The purpose of the
studies presented was to evaluate the initial stages of the
TNT phytoremediation pathway in M. aquaticum and to
better understand how initial reactions may influence the
formation of aminated products or other known metabolites
that are not strictly the result of reduction reactions. This
involved isolating and characterizing initial transformation
intermediates, followed by studies in which these intermedi-
ates were exposed to the plants to evaluate their further
transformation.
Myriophyllum aquaticum
C H U A N YU E W A N G , † D E L I N A Y. L YO N , †
J O S E P H B . H U G H E S , * , † A N D
G E O R G E N . B E N N E T T ‡
Department of Civil and Environmental Engineering, MS-317,
and Department of Biochemistry and Cell Biology, MS-140,
Rice University, Houston, Texas 77005-1892
Phytotransformation studies of 2,4,6-trinitrotoluene (TNT)
were conducted using Myriophyllum aquaticum to clarify the
role of initial intermediates of TNT transformation in the
complex product distributions reported previously.
2-Hydroxylamino-4,6-dinitrotoluene (2HA46DNT) and
4-hydroxylamino-2,6-dinitrotoluene (4HA26DNT) were the
initial intermediates of TNT phytotransformation. 2HA46DNT
and 4HA26DNT were both abiotically transformed to 4,4′,6,6′-
tetranitro-2,2′-azoxytoluene (2,2′azoxy) and 2,2′,6,6′-
tetranitro-4,4′-azoxytoluene (4,4′azoxy) and also phytoreduced
to the related amines 2-amino-4,6-dinitrotoluene (2A46DNT)
and 4-amino-2,6-dinitrotoluene (4A26DNT). To further
elucidate the initial steps of this TNT phytotransformation
pathway, the transformations of known intermediates
(including 2HA46DNT, 4HA26DNT, 2A46DNT, 4A26DNT,
2,2′azoxy, and 4,4′azoxy) were monitored in plant systems.
The transformation rates were measured, and kinetic
analysis using pseudo-first-order models was used to
evaluate the relative rates of competing reactions. The
formation of the azoxy products was determined to be more
rapid than the formation of the amine products. Both the
azoxy and amine products were subject to uptake and further
transformation by the plant.
Introduction
As a result of munitions production and storage, 2,4,6-
trinitrotoluene (TNT) and its byproducts are widespread and
persistent contaminants of soil and groundwater at a number
of government facilities. Since it is toxic and poses a serious
environmental risk (1-3), contaminated soil and water are
currently being remediated. During the evaluation of re-
mediation processes, phytoremediation has been identified
as a viable low-cost option for the cleanup of TNT-con-
taminated media (4, 5). Phytoremediation processes rely on
the ability of plants to take up and transform contaminants.
Investigations of TNT phytoremediation have been con-
ducted using aquatic and terrestrial plant systems (5-12).
While TNT levels decreased, none of these studies showed
significant mineralization of TNT. Still, considerable interest
exists in the evaluation of TNT phytotransformation pathways
Materials and Methods
Chem icals. Nitroaromatic compounds used were TNT (99%
purity: Chem Service, West Chester, PA); (U-ring-14C)-TNT
(Chemsyn Science, Lenexa, KS) purified (98.6%) as de-
scribed previously (13); 2HA46DNT and 4HA26DNT syn-
thesized and purified as described previously (13); 2A46DNT,
4A26DNT, and 2,4-diamino-6-nitrotoluene (24DA6NT) (Ac-
custandard Inc., New Haven, CT); and 2,2′,6,6′-tetranitro-
4,4′-azoxytoluene (4,4′azoxy) and 4,4′,6,6′-tetranitro-2,2′-azoxy-
toluene (2,2′azoxy) (Dr. Deborah Roberts, University of
Houston, Houston, TX). HPLC grade solvents included ethyl
acetate, methylene chloride, and acetonitrile (99.9%) (Fisher
Scientific, Alanta, GA). Silica gel for column chromatography
was 70-230 mesh (Fisher Scientific). NMR was conducted
in methanol-d (99.8 atom %) (Aldrich, St. Louis, MO).
Plants. As described in our previous report (9), M.
aquaticum was purchased from a local supplier. The plants
were grown to between 10 and 30 g/ L (wet wt) in outdoor
* Corresponding author e-mail: hughes@rice.edu; phone: (713)-
348-4761; fax: (713)348-5203.
† Department of Civil and Environmental Engineering.
‡ Department of Biochemistry and Cell Biology.
9
10.1021/es030010a CCC: $25.00
Published on Web 07/16/2003
2003 Am erican Chem ical Society
VOL. 37, NO. 16, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 5 9 5

pools recharged with rainwater and rooted in compost-
amended sediment and gravel (no TNT was present in the
sediments). Before use in experiments, the plants were gently
uprooted and washed thoroughly with tap water to remove
sediment and gravel.
a wet weight plant concentration of 30 g/ L after wash-
ing with tap water. Intermediates 2HA46DNT, 4HA26DNT,
2,2′ azoxy, 4,4′ azoxy, 2A46DNT, and 4A26DNT were added
individually to a concentration of 25 mg/ L. The transforma-
tion processes were monitored over various time periods
(120 h-21 d) depending on the disappearance rate of
individual compounds added. Samples were analyzed im-
mediately by HPLC. During the control test using hydroxyl-
amines, the products were recovered by extraction with
methylene chloride. The extract was dried with anhydrous
MgSO4, filtered, and evaporated to yield white solids. These
compounds were verified as azoxy compounds by HPLC
retention time, UV spectra, and 1H NMR as compared to
standards.
Analytical Methods. The Waters system (Milford, MA)
used for HPLC analysis consisted of a separation module
(model 2690) and a diode-array UV-visible detector (model
996). The system was controlled by a PC-based workstation
equipped with the Millenium Chromatography Manager
software. Spectra were acquired continuously between 200
and 400 nm, and chromatograms were extracted at 230 nm
for quantitation. Analytes were separated on a reverse-phase
Waters Nova-Pak-C18 column (2 × 150 mm) at room
temperature with gradient mobile phases consisting of water/
acetonitrile (0-5 min, 75/ 25 (v/ v); 5-12 min, 70/ 30 (v/ v);
12-20 min, 40/ 60 (v/ v); 20-25 min, 75/ 25) at 0.6 mL/ min.
The measurement of 14C in HPLC fractions and stock
solutions was performed with a Beckman LS6500 scintillation
counter after addition of samples to ReadyGel liquid scintil-
Kinetic Analysis. Data obtained from plant transformation
studies were evaluated using a second-order kinetic mech-
anism as described by Pavlostathis et al. (10) and shown in
eq 1:
dCi
) -kiCiP
(1)
1
lation cocktail (Beckman). H NMR spectra were obtained
dt
on a Bruker AC-250 spectrometer. Chemical shifts are
reported in ppm relative to methanol-d (3.31and 4.78 ppm).
TNT Transform ation Studies. Batch experiments were
performed in 3-L glass tanks at room temperature exposed
to indoor lighting. A [14C ]TNT solution (3 L at 25 ppm and
1000 dpm/ mL of 14C) was added to the tanks. For phy-
totransformation studies, M. aquaticum (90 g wet wt per
tank or 30 g/ L) were added after washing the plants with tap
water. Control systems were identical to phytotransformation
systems except that no plants were added. The transformation
process was monitored by HPLC analysis of samples of the
TNT solution. Samples were analyzed immediately by HPLC.
Samples were not diluted since the TNT concentration was
25 ppm or less. Plant systems and controls with no plants
added were set up in duplicate. One set of plant systems was
terminated at 9 d, and the other was maintained for a total
of 15 d.
Isolation of Transform ation Interm ediates. Intermedi-
ates of TNT phytotransformation were obtained at different
points in the transformation process. Transformation was
stopped after either 9 or 15 d by removing the plants from
the reaction mixture. The TNT intermediates were extracted
from the solution with methylene chloride (200 mL) and then
ethyl acetate (300 mL). The combined organic fractions were
dried with anhydrous MgSO4. After the MgSO4 was filtered,
the filtrate was evaporated, leaving a light yellow residue.
Intermediates were then isolated by silica column chroma-
tography (ethyl acetate/ hexane ) 1/ 10-1/ 2). The products
eluted off the column in the following order: 4,4′azoxy,
2,2′azoxy, 4A26DNT, 2A46DNT, 2HA46DNT, and 4HA26DNT.
These products were subjected to 1H NMR analysis and HPLC
analysis.
Synthesis of 14C Isotope Interm ediate Standards.
14C isotope intermediate standards were synthesized bio-
chemically using [14C]TNT as a starting material (13, 19).
[14C]-4HA26DNT was obtained from reduction of [14C]TNT
by Clostridium acetobutylicum (13). [14C]-2HA46DNT,
[14C]-2A46DNT, and [14C]-4A26DNT were obtained from
[14C]TNT transformation in plant systems using the same
methods as for the isolation of transformation intermediates.
[14C]-2,2′azoxy and [14C]-4,4′azoxy were obtained as de-
scribed previously from the oxidation of [14C]-2HA46DNT
and [14C]-4HA26DNT, respectively (19).
Transform ation of Observed Interm ediates. Comple-
mentary experiments using methods similar to those de-
scribed for TNT transformation studies were carried out for
observed intermediates. These batch experiments were
carried out in 1L glass containers at room temperature and
exposed to indoor light. M. aquaticum was added to obtain
where Ci is the concentration (mM) of the compound of
interest, t is time (h), ki is a second-order rate coefficient (L
g-1 t-1), and P is plant concentration (g L-1). Under the
conditions of this study, plant growth/ decay was assumed
to be negligible (i.e., plant concentration was constant during
the tests) causing the analysis to become pseudo-first-order.
To analyze the rate of abiotic transformation in controls
amended with 4HA26DNT and 2HA46DNT, disappearance
profiles were evaluated with the first-order model in eq 2:
dCi
) -kCi
(2)
dt
where k is the first-order rate coefficient (time-1) and Ci is
the concentration of the parent compound. This model is a
simplification of a more complex mechanism that would
describe the oxidation of a hydroxylamine to a nitroso
functionality followed by a reaction with a separate hy-
droxylamine to form an azoxy compound. The use of a simple
first-order relationship was predictive, however, as the rate-
controlling step is the oxidation of hydroxylamine to form
a nitroso intermediate.
Results
TNT Transform ation Studies. Six intermediates (2HA46DNT,
4HA26DNT, 2A46DNT, 4A26DNT, 2,2′azoxy, and 4,4′azoxy)
were extracted and isolated from the TNT phytotransfor-
mation solution on day 9. The identities of these compounds
were confirmed by HPLC retention time, UV spectra, and 1H
NMR as compared to standards. Figure 1 illustrates the HPLC
separation and UV spectra of the six compounds. The
chemical shifts in 1H NMR spectra of 2HA46DNT, 4HA26DNT,
2A46DNT, and 4A26DNT are summarized in Table 1 and
correspond to values published previously (13). Only three
metabolites could be isolated on the 15th day and were
confirmed as 2A46DNT, 4A26DNT, and 24DA6NT.
The results from temporal sampling and HPLC analysis
of TNT transformation in M. aquaticum are shown in Figure
2. In the negative control, TNT was stable in the water
solution. In the plant system, TNT disappeared nearly
completely after 15 d. During the transformation of TNT,
2HA46DNT and 4HA26DNT were the primary products
initially observed by HPLC analysis. The hydroxylamines
reached their maximum observed combined levels on day
9 and then dissipated to only trace levels by day 15. Following
the production of the hydroxylamine intermediates, ami-
nodinitrotoluenes (2A46DNT and 4A26DNT) and azoxy
compounds (2,2′azoxy and 4,4′azoxy) were detected. 2A46DNT
9
3 5 9 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 16, 2003

FIGURE 1. HPLC separation and UV spectra of 2HA46DNT (A), 4HA26DNT (B), 2A46DNT (C), 4A26DNT (D), 2,2′azoxy (E), 2,4′azoxy (F), and
4,4′azoxy (G).
TABLE 1. 1H NMR and 13C NMR Spectra of TNT and Isolated Intermediates
compound
H-3
H-5
H-7
C-1
C-2
C-3
C-4
C-5
C-6
C-7
TNT
8.95 (s)
7.7 (d)
7.21 (s)
7.49 (s)
6.21 (d)
2.64 (s)
2.22
2.21
2.32
2.04
134.9
126.2
112.8
116.6
106.5
153.0
150.1
153.8
153.5
149.6
123.6
117.2
113.1
112.3
101.4
147.5
147.4
149.6
152.9
148.0
15.7
13.6
13.8
14.0
12.0
2A46DNT
4A26DNT
4HA26DNT
24DA6NT
7.8 (d)
108.7
150.9
6.45 (d)
106.1
153.8
compound
H-3
H-3′
H-5
H-5′
8.80 (d)
H-7
H-7′
2,2′azoxy
4,4′azoxy
9.32 (d)
8.99 (s)
9.19 (d)
9.02 (d)
9.15 (s)
2.65
2.65
2.76
2.68
and 4A26DNT accumulated to
a
combined maximum
Figures 3 and 4. In the control test of 4HA26DNT (Figure 3A)
and 2HA46DNT (Figure 4A), an abiotic transformation of
hydroxylaminonitrotoluenes in aqueous solution occurred
at a relatively rapid rate. Initially the yellow solution turned
clear, and then a white precipitate was observed. By com-
paring the retention time and UV spectra with standards,
the white precipitate was confirmed as 2,2′azoxy and
4,4′azoxy, which were the same compounds as those observed
during TNT phytotransformation. Following the extraction
of the precipitates, the yields of 2,2′azoxy and 4,4′azoxy
products were 91% and 89%, respectively.
concentration on day 14, while the azoxy compounds had
a maximum concentration on day 5 and then decreased to
trace levels by day 15. The total initial 14C in the extracellular
medium decreased to 60% by day 9. At the completion of the
study, 41% of the initial 14C in the extracellular media was
in forms that could not be confirmed using HPLC protocols
and were likely polar oxidation products previously identified
(16).
Transform ation of 2HA46DNT and 4HA26DNT. The
results from temporal sampling and HPLC analysis of
2HA46DNT and 4HA26DNT transformation are shown in
9
VOL. 37, NO. 16, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 5 9 7

FIGURE 5. Results obtained from the HPLC fractionation of azoxy
compounds phytotransformation: 4,4′azoxy (9), 2,2′azoxy (2),
4,4′azoxy control (1), and 2,2′azoxy control ([).
FIGURE 2. Results obtained from the HPLC fractionation of TNT
phytotransformation. TNT (b), 2HA46DNT and 4HA26DNT (3),
2A46DNT and 4A26DNT (O), TNT (control) ([), and water-soluble
14C in extracellular medium (2).
FIGURE 6. Results obtained from the HPLC fractionation of amino
compounds phytotransformation: 2A46DNT (9), 4A26DNT(2),
24DA6NT (1), 2A46DNT control ([), and 4A26DNT control (b).
of both amine compounds (2A46DNT and 4A26DNT) and
both azoxy compounds (2,2′azoxy and 4,4′azoxy). In contrast
to studies with TNT as the parent compound, the phy-
totransformation of hydroxylamines resulted in higher levels
of azoxy compound formation and at a faster rate.
Transform ation of 2,2′azoxy and 4,4′azoxy. The re-
sults of HPLC analysis of the 2,2′azoxy and 4,4′azoxy trans-
formations are shown in Figure 5. Both compounds were
stable in the control systems, but both were rapidly re-
moved from solution when plants were present. Both
2,2′azoxy and 4,4′azoxy were rapidly depleted within 140 h
of exposure, and no degradation products from either azoxy
transformation were observed to accumulate according to
HPLC analysis.
FIGURE 3. (A) Results obtained from the HPLC fraction of 4HA26DNT
control test: 4HA26DNT (9) and 4,4′azoxy (2). (B) Results obtained
fromthe HPLCfractionof4HA26DNTphytotransformation: 4HA26DNT-
(9), 4,4′azoxy (2), and 4A26DNT (1).
Transform ation of 2A46DNT and 4A26DNT. The results
of HPLC analysis of the 2A46DNT and 4A26DNT transfor-
mations are shown in Figure 6. While 2A46DNT and 4A26DNT
were stable in the controls, both 2A46DNT and 4A26DNT
were slowly transformed in the plant systems. After 21 d of
exposure, a small amount of the original 2A46DNT and
4A26DNT remained in the aqueous phase of plant systems,
but only 1% of the products could be identified as 24DA6NT.
The identities of the other products were unconfirmed.
Kinetic Analysis. Data taken from the plant studies and
the abiotic transformation of 4HA26DNT and 2HA46DNT
were analyzed using the integrated form of eqs 1 and 2
yielding eq 3:
C
Co
ln ) -kobs
t
(3)
FIGURE 4. (A) Results obtained from the HPLC fraction of 2HA46DNT
control test: 2HA46DNT (9) and 2,2′azoxy (2). (B) Results obtained
from the HPLC fraction of 2HA46DNT phytotransformation: 2HA46DNT
(9), 2,2′-azoxy (2), and 2A46DNT (1).
where Co is the initial concentration (mM), C is the
concentration (mM) at any time t (h), and kobs is the observed
rate coefficient determined by linear regression. Results of
this analysis are presented in Table 2. In the case of biotic
systems, it is assumed that the kobs obtained is a function of
During the phytotransformation of 4HA26DNT (Figure
3B) and 2HA46DNT (Figure 4B), most of the initial com-
pounds had disappeared within 120 h, with the production
P (g/ L). Therefore, Table 1 also presents the kbio (L h1 P1-
by dividing the kobs by P in experimental systems.
)
9
3 5 9 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 16, 2003

but that the formation of a second amine group is not
favorable. The identification of the compounds in this study
was facilitated by the HPLC method employed, and care was
taken to avoid decomposition. Critical to the HPLC approach
is the ability to resolve hydroxylamines from corresponding
amines since they have some similar spectral properties (6).
Decomposition was avoided by minimizing storage times
during sample handling; otherwise the hydroxylamines would
have decomposed, leaving only low levels of stable amines
(21).
Previous studies that used TNT as the starting material
to identify phytotransformation intermediates have not
recovered azoxy compounds. According to the research
presented here, 2,2′azoxy and 4,4′azoxy were very stable in
plant-free controls but were taken up and transformed by
plants at rapid rates. The rapid uptake of the azoxy
compounds observed in this study explains why the con-
centration of azoxy compounds was extremely low in the
TNT phytotransformation system and perhaps why they have
not been monitored as a phytotransformation product.
Extraction of plant biomass did not recover the azoxy
compounds, implying that they are subject to further
transformation by the plant. This hypothesis is supported by
previous studies that specifically investigated the nature of
plant associated residues and found that the bulk of
metabolites in the plant biomass were in a “bound residue”
form (11).
TABLE 2. Calculated Pseudo-First- and Second-Order Rate
Constants Observed for Various Parent Compounds in Biotic
and Abiotic Systemsa
compound
kobs (h-1
)
r 2
kbio (L h-1 P1-
)
TNT
biotic
1.1 × 10-2 ( 0.1 × 10-2 0.97
3.6 × 10-4
abiotic
4HA26DNT
biotic
abiotic
2HA46DNT
biotic
abiotic
4A26DNT
biotic
abiotic
2A46DNT
biotic
abiotic
4,4′azoxy
biotic
abiotic
2,2′azoxy
biotic
nobsd
na
na
2.1 × 10-2 ( 0.1 × 10-2 0.99
0.7 × 10-3
4.5 × 10-2 ( 0.2 × 10-2 0.98
na
2.6 × 10-2 ( 0.1 × 10-2 0.98
2.6 × 10-2 ( 0.1 × 10-2 0.98
8.6 × 10-4
na
4.7 × 10-3 ( 0.3 × 10-3 0.98
1.5 × 10-4
nobsd
na
na
5.3 × 10-3 ( 0.3 × 10-3 0.97
1.7 × 10-4
nobsd
na
na
1.9 × 10-2 ( 0.2 × 10-2 0.89
6.3 × 10-4
nobsd
na
na
1.7 × 10-2 ( 0.1 × 10-2 0.97
5.7 × 10-4
abiotic
nobsd
na
na
a
nobsd, not observed; na, not available.
An important consideration that was not evaluated in
this study is the effect of plant concentration (or net activity)
on the relative proportions of azoxy compound formation as
compared to reduced aminated derivatives. On the basis of
the control studies, oxidation of hydroxylamines is inde-
pendent of the metabolic activity of the plant. Reduction of
hydroxylamines is a plant-mediated reaction, which may be
enhanced relative to abiotic transformations by increasing
the plant concentration. Differences in plant concentration
(or net activity) may explain why the formation of azoxy
compounds is not always reported and may also influence
the degree to which aminated forms accumulate.
By combining previously reported TNT phytotransfor-
mation end products and the intermediates observed in this
study, the initial steps of the TNT phytotransformation
pathway in M. aquaticum can be deduced, as shown in Figure
7. In plant systems, TNT was taken up and transformed to
Discussion
The purpose of these studies was to evaluate how initial
reduction reactions may contribute to the product distribu-
tion of TNT phytotransformation using the aquatic plant M.
aquaticum. Experimental results of TNT phytotransformation
were consistent with previous reports demonstrating the
plant’s ability to transform TNT resulting in both extracellular
and plant associated fractions (11). In this research, however,
six transformation products were observed in the extracellular
medium of the plant system, including 2HA46DNT, 4HA26-
DNT, 2A46DNT, 4A26DNT, 2,2′azoxy, and 4,4′azoxy.
From this study, 2HA46DNT and 4HA26DNT are estab-
lished as the initial intermediates, in contrast to previous
studies using diverse plant systems that have generally cited
aminodinitrotoluenes as the primary intermediates in TNT
phytotransformation (6, 9, 11, 16). Low concentrations of
2A46DNT and 4A26DNT were observed in M. aquaticum and
Catharanthus roseus hairy root culture (10). Other primary
products that have been reported include conjugated com-
pounds (6) and oxidation compounds (16). To the best of
our knowledge, 2HA46DNT from TNT phytotransformation
has not previously been identified, although the role of
hydroxylamines as important intermediates was proposed
earlier (11).
The hydroxylamines can be further reduced to amino-
dinitrotoluenes (2A46DNT and 4A26DNT) or undergo oxida-
tion to nitroso intermediates that then form azoxy com-
pounds (2,2′azoxy and 4,4′azoxy). The abiotic experimental
results of 2HA46DNT and 4HA26DNT transformations in
aqueous solution demonstrate that hydroxylamines are
unstable in the presence of oxygen and form azoxy com-
pounds at a rapid rate with near stoichiometric yields (19).
The oxidation reaction was more rapid than the reduction
reaction, but aminodinitrotoluenes were observed in the
extracellular medium. The formation of aminodinitrotoluenes
and azoxy dimers requires partial reduction via hydroxy-
lamine intermediates (20). Experimental results of the
transformation of 2A46DNT and 4A26DNT indicated that
plants take up and transform aminodinitrotoluenes at a
slower rate than other reactions. 24DA6NT was a minor
product after 2A46DNT and 4A26DNT were completely
transformed, indicating that further reduction is possible
4HA26DNT and 2HA46DNT through
a plant-mediated
reduction. Then two different reactions compete for the
hydroxylamines. 4HA26DNT and 2HA46DNT can be abioti-
cally oxidized to reactive nitroso derivatives that undergo
bimolecular nucleation to form azoxy compounds, or they
can undergo a slower, plant-mediated reduction to form
aminodinitrotoluenes (4A26DNT and 2A46DNT). Under the
experimental conditions used in these studies, it appears
that the further reduction of the hydroxylamines to form
mono- and diaminonitrotoluenes was unfavorable, although
they were produced to varying degrees. Under these condi-
tions, it appears that the oxidation pathway would constitute
the majority of the metabolite flux. However, the rapid uptake
of azoxy compounds by the plants results in their low levels
in solution when TNT is fed directly to the plants. Clearly,
both the azoxy and aminated products need to be investigated
in order to account for further transformation products.
Certain similarities can be identified when comparing
the results of TNT phytotransformation by plants to the
hydroxylamine pathways of TNT transformation by anaerobic
microorganisms that we have previously reported (13, 22).
Both pathways begin with the formation of hydroxylami-
nodinitrotoluene intermediates. However, while anaerobic
fermentor pathways lead to the accumulation of high
concentrations of primary monohydroxylamine intermedi-
ates, the hydroxylaminodinitrotoluenes in plant systems
9
VOL. 37, NO. 16, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 5 9 9

minimize the extent to which the washout of aminodini-
trotoluenes may occur from phytoremediation systems and
promote further metabolism by plants. Also, the oxidation
process is facilitated by lower levels of plant activity and/ or
density. Further studies are required to characterize the
transformation process of azoxy dimers and to determine
their role in the formation of bound and water soluble
transformation endpoints.
Acknowledgments
Work performed was supported by a grant from DSWA Project
01-97-1-0020.
Literature Cited
(1) Won, W. D.; DiSalvo, L. H.; Ng, J. Appl. Environ. Microbiol. 1976,
31, 576-580.
(2) Smock, L. A.; Stoneburger, D. L.; Clark, J. R. Water Res. 1976,
10, 537-543.
(3) Simini, M.; Wenstel, R. S.; Checkai, R. T.; Phillips, C. T.; Chester,
N. A.; Major, M. A.; Amos, J. C. Environ. Toxicol. Chem. 1995,
14, 623-630.
(4) Schnoor, J. L.; Licht, L. A.; McCutcheon, S. C.; Wolfe, N. L.;
Carreira, L. H. Environ. Sci. Technol. 1995, 29, 318-323.
(5) Palazzo, A. J.; Leggett, D. C. J. Environ. Qual. 1986, 15, 49-52.
(6) Bhadra, R.; Wayment, D.; Hughes, J. B.; Shanks, J. V. Environ.
Sci. Technol. 1999, 33, 446-452.
(7) Larson, S. L. Ann. N.Y. Acad. Sci. 1997, 829, 195-201.
(8) Thompson, P. L.; Ramer, L.; Schnoor, J. L. Environ. Sci. Technol.
1998, 32, 975-980.
(9) Vanderford, M.; Shanks, J. V.; Hughes, J. G. Biotechnol. Lett.
1997, 19, 277-280.
(10) Pavlostathis, S. G.; Comstock, K. G.; Jacobson, M. E.; Saunders,
F. M. Environ. Toxicol. Chem. 1998, 17, 2266-2273.
(11) Hughes, J. B.; Shanks, J. V.; Vanderford, M.; Lauritzen, J.; Bhadra,
R. Environ. Sci. Technol. 1997, 31, 266-271.
(12) Harvey, S. D.; Fellows, R. J.; Cataldo, D. A.; Bean, R. M. J.
Chromatogr. 1990, 518, 361-374.
FIGURE 7. Initial steps of theTNT phytotransformation pathway in
M. aquaticum and previously identified end products.
(13) Hughes, J.; Wang, C. Y.; Bharda, R.; Richardson, A.; Bennett, G.;
Rudolph, F. Environ. Toxicol. Chem. 1998, 17, 343-348.
(14) Ederer, M. M.; Lewis, T. A.; Crawford, R. L. J. Ind. Microbiol.
Biotechnol. 1997, 18, 82-88.
(15) Lewis, T. A.; Goszczynski, S.; Crawford, R. L.; Korus, R. A.;
Admassu, W. Appl. Environ. Microbiol. 1996, 62, 4669-4674.
(16) Bhadra, R.; Wayment, D.; Hughes, J. B.; Shanks, J. V. Environ.
Sci. Technol. 1999, 33, 3354-3361.
(17) Goheen, S. C.; Campbell, J. A.; Roach, S. K.; Shi, Y.; Shah, M. M.
Presented at the Second International Symposium on Biodeg-
radation of Nitroaromatic Compounds and Explosives, 1999.
(18) Burken, J. G.; Shanks, J. V.; Thompson, P. L. In Biodegradation
of Nitroaromatic Compounds and Explosives; Spain, J. C.,
Hughes, J. B., Knackmuss, H. J., Eds.; Lewis Publishers: Wash-
ington, DC, 2000; pp 239-275.
(19) Wang, C. Y.; Zheng, D.; Hughes, J. B. Biotechnol. Lett. 2000, 22,
15-19.
(20) Corbett, M. D.; Corbett, B. R. In Biodegradation of Nitroaromatic
Compounds; Spain, J. C., Ed.; Plenum Press: New York, 1995;
Vol. 49, pp 151-182.
(21) Hughes, J. B;, Wang, C. Y.; Zhang, C. Environ. Sci. Technol. 1999,
33, 1065-1070.
(22) Hughes, J.; Wang, C.; Yesland, K.; Richardson, A.; Bhadra, R.;
Bennett, G.; Rudolph, F. Environ. Sci. Technol. 1998, 32, 494-
500.
undergo further reduction to the central metabolite 2,4-
dihydroxylamino-6-nitrotoluene and other products. An
obvious difference between the TNT phytotransformation
and anaerobic microbial TNT transformation is the presence
of oxygen in the plant system. The presence of oxygen and
the slow reduction rate of the hydroxylamine group make
oxidation of the hydroxylamine to a nitroso group favorable
and results in the scavenging of other hydroxylamines to
form azoxy metabolites. These azoxy compounds can be
rapidly taken up by the plant and apparently transformed.
To what extent the transformation of azoxy compounds may
contribute to the formation of bound residues, conjugated
intermediates, or oxidation products is not clear at this time,
and the characterization of that process requires further
study.
The pathway to diverse products of TNT metabolism (i.e.,
conjugated compounds, bound residues, and oxidation
products) carried out by plants becomes complex even in
the first several steps. Additionally, the competing rates of
early metabolite formation may differ based on experimental
conditions, resulting in varying results from study to study.
On the basis of the slow transformation rates of 2A46DNT
and 4A26DNT as compared to the relatively fast transforma-
tion rates of azoxy compounds observed in this work, it may
be advantageous to consider operational conditions that favor
hydroxylamine oxidation via abiotic mechanisms. This would
Received for review January 15, 2003. Revised manuscript
received April 22, 2003. Accepted May 23, 2003.
ES030010A
9
3 6 0 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 16, 2003