Bamberger rearrangement during TNT metabolism by Clostridium acetobutylicum

Article abstract of DOI:10.1021/es970612s

Studies were conducted to isolate and identify polar and oxygen- sensitive intermediates of 2,4,6-trinitrotoluene (TNT) transformation by Clostridium acetobutylicum. Studies conducted in anaerobic cell extracts demonstrated that a polar product formed fro

Full text of DOI:10.1021/es970612s

Environ. Sci. Technol. 1998, 32, 494-500
the reduction of one or more aryl nitro groups via nitroso
Bamberger Rearrangement during
TNT Metabolism by Clostridium
acetobutylicum
and hydroxylamino intermediates forming aryl amines (3).
Commonly cited intermediates of TNT metabolism are
aminodinitrotoluenes and diaminonitrotoluenes (3-5). Un-
der strong reducing conditions, 2,4,6-triaminotoluene (TAT)
can be formed and undergo further metabolism (6-8).
Reaction mechanisms other than reduction to the amine
that may influence the fate of TNT in anaerobic bioreme-
diation systems, include condensation forming azoxy dimers
(1, 6), substitution of aminated intermediates (9), and ring
reduction (10). Despite the numerous known products of
TNT transformation, determining the fate of TNT in biore-
mediation systems remains challenging (3, 8, 11). TNT fate
is generally evaluated by the final distribution of 14C-labeled
products (3, 12). Mineralization of TNT has been observed,
but is often only a small percent (8, 13). Many products
“bind” to soil in a poorly reversible manner (1, 14), and
undefined “polar metabolites” may reside in the aqueous
phase (12, 15, 16).
Recent studies have demonstrated the temporal ac-
cumulation of hydroxylamino intermediates during the
reduction of TNT (12, 16, 17). Because hydroxylamines are
mutagenic and exhibit chemical properties quite different
than the amine group (18), their formation and further
metabolism is of interest in TNT bioremediation systems.
The bioorganic chemistry of arylhydroxylamines has been
reviewed by Corbett and Corbett (18). Possible reactions
include reduction to the amine, nucleophilic, and/ or elec-
trophilic substitutions or rearrangement (e.g., Bamberger
rearrangement) to phenolic amines (4, 18, 19). The formation
of phenolic amines through the Bamberger rearrangement
is particularly interesting since many phenols can undergo
anaerobic ring cleavage (2). The formation of phenolic
metabolites from TNT has been proposed (20) through the
conversion of TAT to 2,4,6-trihydroxytoluene (methylphlo-
roglucinol) and further metabolism to 4-hydroxytoluene (p-
cresol). However, this processes does not involve hydrox-
ylamino intermediates, and unlike the Bamberger rearrange-
ment this reaction eliminates N-functionalities from the ring.
In previously reported studies of TNT transformation by
Clostridium acetobutylicum, we have observed the rapid
reduction of TNT to 2,4-dihydroxylamino-6-nitrotoluene and
the subsequent the formation of a polar product (15, 16, 21)
without detection of intermediate aminated forms. Attempts
to isolate this product(s) were unsuccessful due to the
complexities of purification from an organic-rich nutrient
broth and the compound’s oxygen sensitivity. This paper
presents results of studies using C. acetobutylicum cell extracts
coupled with chemical derivatization of intermediates to
identify the mechanism of 2,4-dihydroxylamino-6-nitrotolu-
ene metabolism by this organism. Studies are also presented
confirming product formation in cell cultures.
J . B . H U G H E S , * , † C . W A N G , †
K . YE S L A N D , ‡ A . R I C H A R D S O N , †
R . B H A D R A , † G . B E N N E T T , ‡ A N D
‡
F . R U D O L P H
Department of Environmental Science and Engineering and
Department of Biochemistry and Cell Biology, Rice University,
6100 Main, Houston, Texas 77005-1892
Studies were conducted to isolate and identify polar and
oxygen-sensitive intermediates of 2,4,6-trinitrotoluene
(TNT) transformation by Clostridium acetobutylicum. Studies
conducted in anaerobic cell extracts demonstrated that
a polar product formed from the transformation of 2,4-
dihydroxylamino-6-nitrotoluene by a mechanism known
as the Bamberger rearrangement. The product was stabilized
by derivatization with acetic anhydride, and the structure
1
was confirmed by mass spectroscopy, H NMR, and IR
spectroscopy techniques. The reaction occurred in the
presence of cell extract and H but did not occur in cell
2
extract-free controls. From spectroscopic data, the product
of 2,4-dihydroxylamino-6-nitrotoluene rearrangement was
identified as either 2-amino-4-hydroxylamino-5-hydroxyl-
6-nitrotoluene (4-amino-6-hydroxylamino-3-methyl-2-nitrophe-
nol) or 2-hydroxylamino-4-amino-5-hydroxyl-6-nitrotoluene
(6-amino-4-hydroxylamino-3-methyl-2-nitrophenol).
Acid-catalyzed rearrangement of 2,4-dihydroxylamino-6-
nitrotoluene resulted in a single product, which after
derivatization, was identical to a derivatized product from
cell extracts. Acid-catalyzed Bamberger rearrangement oc-
curs with the hydroxyl addition para to the participating
hydroxylamine, indicating that the 2-amino-4-hydroxylamino-
5-hydroxyl-6-nitrotoluene (4-amino-6-hydroxylamino-3-
methyl-2-nitrophenol) was the product isolated form cell
extracts. This product was also confirmed in whole cell
systems that had been fed TNT. Following derivatization
of the culture broth, a product was isolated that was
identical to those isolated from crude cell extracts and acid
catalysis experiments.
Introduction
Materials and Methods
The development of biological processes for treatment of
2,4,6-trinitrotoluene (TNT) contaminated soils or ground-
waters has been an area of research and testing for several
decades (1). These studies have demonstrated that TNT can
be biotransformed under a range of conditions and that the
final products of metabolism are often difficult to determine
(1, 2). Generally, the initial stages of TNT metabolism involve
Chem icals. Nitroaromatics used were 2,4,6-TNT (99%;
purity; Chem Service, West Chester, PA), [U-ring-14C]TNT
(Chemsyn Science, Lenexa, KS) purified (98.6%) as described
previously (16), and 4-hydroxylamino-2,6-dinitrotoluene and
2,4-dihydroxylamino-6-nitrotoluene [synthesized and puri-
fied as described previously (16)]. Other chemicals included
acetic anhydride (Reagent ACS 97%, TITR, ACROS), hydro-
chloric acid (37.5%, Fisher scientific), and sodium acetate
(anhydrous, Fisher Scientific). HPLC grade solvents included
ethyl acetate, hexane, and methylene chloride (99.9%, Fisher
scientific). NMR was conducted in chloroform-d (99.8 atom
%, Aldrich).
* To whom correspondence should be addressed. Telephone:
(713)285-5903; fax: (713)285-5203; e-mail address: hughes@
owlnet.rice.edu.
† Department of Environmental Science and Engineering.
‡ Department of Biochemistry and Cell Biology.
9
4 9 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 4, 1998
S0013-936X(97)00612-3 CCC: $15.00
1998 Am erican Chem ical Society
Published on Web 01/08/1998

Preparation of Clostridium a cetobutylicum Crude Ex-
tract. Cells were grown as described previously (15) to mid
log phase (o.d. 595 ) 1.2) and collected by centrifugation
(12000g) for 10 min. Cells were washed in 5 mM Tris (pH
) 7.8) and again collected by centrifugation (12000g for 10
min). The cells were resuspended in 5 mM Tris (pH 7.8)
using a volume equal to twice that of the cell pellet, and
lysozyme (Sigma) was added (10 mg/ g of cells). After mixing,
the suspension was incubated at 37 °C for 30 min and then
sonicated for 12 min. The lysate was centrifuged (15000g)
for 15 min to remove cellular debris. The supernatant (i.e.,
crude extract) was frozen with liquid nitrogen and stored at
-80 °C. All steps were carried out under anaerobic condi-
tions. Transfers and the sonication of cells were conducted
in an anaerobic glovebox (90%/ 10% (v/ v) N2 and H2,
respectively). Centrifuge tubes were sealed in the glovebox
prior to removal for centrifugation. The protein concentra-
tion of crude extract aliquots was determined by the Lowery
assay before use.
anaerobic Tris buffer (5 mM, pH ) 7.8) containing cell extract
at a concentration of 0.05 mg of protein/ mL. The reaction
mixture was continuously mixed by bubbling gas (90%/ 10%
N2 and H2, respectively) through the solution using an
aquarium pump and a glass frit.
The transformation process was monitored by HPLC
analysis of samples taken from the solution. Samples were
placed in HPLC autosampler vials, sealed within the glovebox,
cooled in an ice bath, and analyzed immediately. After 72
h, the remaining cell extract system was derivatized in the
glovebox with deoxygenated reagents.
The crude cell extract suspension containing the polar
product was cooled in an ice bath to 0 °C, and hydrochloric
acid was added slowly to reach pH ) 1. With the solution
still cold, acetic anhydride (12 mL) was added drop by drop.
After 15 min, sodium acetate was added to adjust the pH to
7, and the solution was removed from the glovebox. After
extraction with ethyl acetate (three times, total 500 mL), the
combined organic phases were dried over anhydrous MgSO4,
filtered, and evaporated to yield a brown residue. Two 14C-
labeled compounds were isolated by silica column chro-
matography eluted with 3:1:2 (v/ v/ v) ethyl acetate/ methylene
chloride/ hexane. These fractions were then purified using
the preparative-scale HPLC method and characterized using
1H NMR, CI mass spectroscopy, high-resolution mass
spectroscopy, and IR spectroscopy.
Cell Cultures. A cell culture (300 mL) of C. acetobutylicum
was grown as described previously (15). During the log-
growth phase, the culture was spiked with TNT and 14C-TNT
predissolved in methanol (20 mg of TNT and 3 130 000 dpm/
mL) resulting in an initial TNT concentration of 100 mg/ L
(1.56 × 103 dpm/ mL). Temporal samples (2 mL) were
removed aseptically and centrifuged (13000g for 4 min), and
the supernatant was analyzed by HPLC for TNT and its
reduced products. After 100 min of incubation, neither TNT
or its hydroxylamino derivatives were detected. The culture
was transferred to an anaerobic chamber (100% N2), cooled
in an ice/ water bath (0 °C), and acidified to pH 1 with HCl;
20 mL of acetic anhydride was added. After 30 min, the pH
was adjusted (pH ) 7) with sodium acetate, and the
suspension was removed from the glovebox and centrifuged
at 25000gfor 10 min in sealed, nitrogen-purged polypropylene
centrifuge tubes. The supernatant was recovered, gravity-
filtered through a 1-µm glass fiber filter (Gelman Sciences,
Ann Arbor, MI), and extracted three times with ethyl acetate
(each volume of ethyl acetate equal to supernatant volume).
The extracts were combined and evaporated to dryness.
Two methods were used for comparison of products from
cell extracts and cell cultures. First, the residue from cell
cultures was analyzed by HPLC using the preparative-scale
method described previously comparing retention time and
UV spectra of eluting peaks. Secondly, 1H NMR was
conducted on the 14C-labeled fraction separated by silica
column chromatography (mobile phase was 3:1:2 ethyl
acetate/ methylene chloride/ hexane). Desired fractions were
evaporated to dryness, dissolved in CDCl3, and 1H NMR
conducted.
Analytical Methods. An HPLC equipped with a diode-
array UV/ vis detector (Waters, Milford MA) was used to
quantify and fractionate TNT and its transformation products.
Monitoring of analyte concentration during transformation
studies was conducted at room temperature with a Waters
Nova-Pak-C8 column (3.9 × 150 mm) and an isocratic mobile
phase mixture of 82% water and 18% 2-propanol at 1 mL/
min. Spectra were acquired continuously between 200 and
400 nm, and chromatograms were extracted at 230 nm for
quantitation. Purification of derivatized products (deriva-
tization procedure is described in the next section) was
conducted at room temperature with a preparative-scale
Waters Nova-Pak-C8 column (7.8 × 300 mm) using an
isocratic mobile phase of 75% water and 25% 2-propanol at
2 mL/ min.
The measurement of 14C in HPLC fractions and stock
solutions was performed with a Beckman LS3801 scintillation
counter after addition of samples to ReadyGel scintillation
cocktail (Beckman). 1H NMR spectra were obtained on a
Bruker AC-250 spectrometer. Chemical shifts are reported
in ppm relative to CDCl3 (7.26 ppm). Chemical ionization
(CI) mass spectroscopy and high-resolution mass spectros-
copy were conducted with a MAT95 mass spectrometer.
Infrared (IR) spectra were obtained with a Nicolet 205 FTIR
spectrophotometer.
Derivatization Procedure. A derivatization procedure
capable of acetylating hydroxylamino derivatives of TNT was
developed and tested initially using 4-hydroxylamino-2,6-
dinitrotoluene. A solution containing 4-hydroxylamino-2,6-
dinitrotoluene (10 mg) dissolved in 1 mL of methanol and
4 mL of distilled water was chilled in an ice bath and acidified
to pH ) 1 by slowly adding 3 mL of HCl. Acetic anhydride
(3 mL) was then added dropwise while maintaining the
solution temperature at 0 °C . As the reaction proceeded
(approximately 10 min), the solution changed from yellow
to nearly colorless. The reaction was stopped by the addition
of sodium acetate (3.5 mg) dissolved in 5 mL of water. After
repeated extraction with ethyl acetate, the organic phases
were washed with water and dried over MgSO4. The solvent
was evaporated and the product isolated by TLC plate
separation (silica) with 1:1:1 (v/ v/ v) ethyl acetate/ methylene
chloride/ hexane eluant. Following purification by the
preparative-scale HPLC method, the compound was identi-
fied using 1H NMR, CI mass spectroscopy, and infrared
spectroscopy.
Results and Discussion
Derivatization of 4-Hydroxylam ino-2,6-dinitrotoluene. A
single product was isolated from the acetylation of 4-hy-
droxylamino-2,6-dinitrotoluene and identified as 2,6-dinitro-
4-(N-acetoxyacetamido)toluene. The structure of this com-
pound and relevant spectroscopic data are presented in Table
1. Purification yielded 12.2 mg of product (87.5% of initial).
Interestingly, this procedure was capable of completely
acetylating the hydroxylamine at both reactive centers (N-
and O-). Without excess acetic anhydride, addition at the
N- is generally favored, although O-derivatization is some-
times observed (18). No information could be located in the
TNT Transform ation and Polar Product Characteriza-
tion. A methanol stock solution was prepared containing 50
mg/ mL unlabeled TNT and 5.0 × 105 dpm/ mL [U-ring-14C]-
TNT. In a glovebox (90%/ 10% N2 and H2, respectively),
methanol stock solution (2 mL) was added to 500 mL of
9
VOL. 32, NO. 4, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 4 9 5

1
TABLE 1. Structure of Product
TABLE 2. Results from CI Mass Spectrometry, H NMR
[2,6-Dinitro-(4-(N)-acetoxyacetamido)toluene] Isolated from
the Derivatization of 4-Hydroxylamino-2,6-dinitrotoluene and
Pertinent Spectroscopic Data
Spectroscopy, and FTIR Analysis of 14C-Labeled Derivatization
Products Obtained from Crude Cell Extracts
CI mass spectroscopy
I massa
II massa
326.2
282.4,
268.2, 226.2
CH3
O2N
NO2
m olecular ion (M + 1)b
fragm ent ions (F + 1)b
368.3
326.2,
308.2, 266.2
O
C
N
CH3
1H NMR
I ppm
(relative intensity)
splitting
O
II ppm
(relative intensity)
splitting
C
O
CH3
Ar-CH3
2.20 (3)
singlet
7.86 (1)
singlet
2.39 (3)
singlet
2.24 (3)
singlet
8.07 (1)
broad singlet
2.49 (3)
singlet
2.24 (3)
singlet
2.12 (3)
singlet
7.55 (1)
singlet
2.38 (3)
singlet
2.25 (3)
singlet
7.52 (1)
broad singlet
2.25 (3)
singlet
CI Mass Spectroscopy
mass
Ar-H
m olecular ion (M + 1)b
fragm ent ions (F + 1)b
298.0
256.0, 238.0, 196.0
Ar-OCOCH3
c
R-NCOCH3
R-NHc
1H NMR
ppm (relative intensity) splitting
Ar-CH3
Ar-H
2.38 (3)
singlet
8.16 (2)
singlet
2.23 (3)
singlet
2.55 (3)
singlet
d
R-NOCOCH3
d
R-NCOCH3
not applicable
a
R-NCOCH3
d
R-NOH
not applicable
8.93 (1)
broad singlet
a
R-NOCOCH3
FTIR Spectroscopy
I wavelength
(intensity)
II wavelength
(intensity)
FTIR Spectroscopy
wavelength (intensity)
NO2
1538.6
(very strong)
1370.6
1536.4
(very strong)
1369.8
NO2
1543.6
(very strong)
1370.0
asym m etric stretch
NO2
sym m etric stretch
asym m etric stretch
NO2
sym m etric stretch
(strong)
(strong)
(strong)
a
I em pirical form ula C15H17N3O8; II em pirical form ula C13H15N3O7.
a
b
b
Groups from derivatized aryl hydroxylam ine. Molecular and
prim ary fragm ent ions reported from CI/MS include the m ass of the
ionizing proton.
Molecular and prim ary fragm ent ions reported from CI/MS include
the m ass of the ionizing proton. Groups from derivatized arylam ine.
Groups from derivatized aryl hydroxylam ine.
c
d
TNT Transform ation in Cell Extracts and Polar Product
Identification. The results from sampling and HPLC analysis
of crude cell extracts spiked with TNT are shown in Figure
1. TNT was rapidly transformed to 4-hydroxylamino-2,6-
dinitrotoluene and 2-hydroxylamino-4,6-dinitrotoluene, with
a greater extent of initial reduction at the 4-position. After
approximately 24 h, the 4-hydroxylamino-2,6-dinitrotoluene
and 2-hydroxylamino-4,6-dinitrotoluene were converted to
the 2,4-dihydroxylamino-6-nitrotoluene, which then de-
creased in concentration with a commensurate accumulation
of the polar product. After 72 h, 83% of the 14C was present
in the polar fraction. The UV spectra obtained for the polar
product during HPLC analysis is presented in Figure 2.
After derivatization, two 14C-labeled products were
isolated by silica column chromatography. These com-
pounds were purified using the preparative scale HPLC
techniqueswith
a com bined yield of nearly 80% of
initialsand subjected to spectroscopic analysis. The 1H NMR
of each compound indicated that only one aryl proton
remained on the molecule and IR spectra for each yielded
characteristic absorption of the nitro group. These results
indicate that the metabolism of 2,4-dihydroxylamino-6-
nitrotoluene involved a substitution at either the 3- or
5-position and not a continued reduction of the remaining
nitro group to trihydroxylaminotoluene.
FIGURE 1. Temporal transformation of TNT (9) to 4-hydroxylamino-
2,6-dinitrotoluene (2), 2-hydroxylamino-4,6-dinitrotoluene (b), 2,4-
dihydroxylamino-6-nitrotoluene ((), and polar product (0) in crude
cell extracts.
literature to determine the 1H NMR chemical shift corre-
sponding to an aryl N-acetoxyacetamido group. On the basis
of substituent effects for aryl nitro (∆ortho ) 0.94, ∆para ) 0.39)
and methyl groups (∆meta ) -0.09) (22), the calculated ∆ortho
for this group was -0.35.
CI mass spectroscopy yielded molecular ion peaks (M +
1) of 368.3 and 326.2, respectively. The compound with a
molecular ion of 368.3 was obtained in a greater yield
9
4 9 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 4, 1998

FIGURE 2. UV/vis spectra of polar product obtained during HPLC analysis of TNT transformation in cell extracts.
(approximately 3:1) and subjected to high-resolution mass
spectroscopy confirm ing its elem ental com position
(C15H17N3O8, MW ) 367.32, error ) 1.2 ppm). The difference
in mass between the two products (∆mass ) 42) corresponds
to the absence of an acetyl group (C13H15N3O7;MW ) 325.28),
which would occur if derivatization was not complete. The
1H NMR of this smaller compound contained a broad
downfield proton (8.93 ppm) that is consistent with an (R)2-
N-O-H proton remaining from only the N-acetylation of a
hydroxylamine group. Spectroscopic information for both
derivatized products (labeled I and II) are presented in Table
2. These results are consistent with a toluene ring substituted
with the following groups: nitro-, acetoxy- (formed from the
acetylation of a hydroxyl group), acetamide- (from the
acetylation of an amino group), and either a N-acetoxyac-
etamide or N-acetamidohydroxylamine (depending on the
degree of acetylation of a hydroxylamine group).
(∂ ) 7.86) indicates that the hydroxyl group addition is at the
5-position and the parent compound is either structure A or
B. Similar rearrangement mechanisms have been reported
in studies of the microbial metabolism of nitrobenzene by
Pseudomonas pseudoalcaligenes (19) and Ralstonia eutropha
(23) but not for TNT. In P. pseudoalcaligenes, the addition
of the hydroxyl group occurred exclusively ortho to the
intermediate hydroxylaminobenzene forming 2-aminophe-
nol. Rearrangement catalyzed by R. eutropha resulted in
the formation of 2-aminophenol (33%) and 4-aminophenol
(63%).
From the spectroscopic results alone, the assignment of
groups at the 2- or 4-positions (e.g., 2-amino-4-hydroxy-
lamino or 2-hydroxylamino-4-amino) is difficult; however,
experimental results indicate that structure A is more likely.
When derivitizing 2,4-dihydroxylamino-6-nitrotoluene using
the method employed for the acetylation of 4-hydroxylamino-
2,6-dinitrotoluene, a rapid transformation was observed in
the cold acid solution (pH ) 1) (characterized by a rapid
dissipation of the yellow color of the 2,4-dihydroxylamino-
6-nitrotoluene solution) prior to the addition of acetic
anhydride. After derivatization, the product had an HPLC
retention time (14.3 min using the preparative-scale tech-
nique), UV spectra, and 1H NMR [peaks obtained with relative
areas in parentheses were 2.24 (6), 2.39 (3), 2.49 (3), 7.86 (1),
and 8.07 (1)] identical to I from crude cell extract studies.
This result is consistent with the acid-catalyzed Bamberger
rearrangement in which the para rearrangement is strongly
favored (24). Only the 2-hydroxylamine group is capable of
para rearrangement with this molecule since the aryl methyl
group is para to the 4-hydroxylamine group. The ortho
rearrangement of the 4-hydroxylamine group is unlikely
without enzymatic catalysis, and under identical conditions,
rearrangement of the 4-hydroxylamino-2,6-dinitrotoluene
was not observed.
On the basis of the results of spectroscopic analysis, four
possible structures for the parent compound(s) prior to
derivatization are shown in Figure 3. All are consistent with
a Bamberger rearrangement of the 2,4-dihydroxylamino-6-
nitrotoluene. Structure A would result from the addition of
a
hydroxyl group at the 5-position with participation
fromsand subsequent reduction ofsthe p-hydroxylamine.
Structure B would result from the addition at the 5-position
with participation of the o-hydroxylamine. Structures C and
D correspond to addition at the 3-position with participation
of either o-hydroxylamine. Calculation of 1H NMR chemical
shift for the aryl proton using literature values (22) and the
shift obtained of the completely acetylated product are
consistent with only two of the possible four isomers. After
complete acetylation, the predicted ∂ for a proton remaining
in the 3-position (structures A and B) is 7.83 ppm, while a
proton remaining in the 5-position would be either 8.22
(structure C) or 7.97 ppm (structure D). The observed value
9
VOL. 32, NO. 4, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 4 9 7

the 13C NMR without isotope enrichment did not allow for
positive identification. In theory, compound A is a more
likely product of the rearrangement since the intermediate
structures for rearrangement of B would favor hydroxyl
addition at the 3-position and could result in mixed
productssneither of which were observed.
The initial stages of TNT transformation in C. acetobu-
tylicum crude cell extracts are presented in Figure 4.
Reduction of the aryl nitro group yields aryl hydroxylamines
and eventually an aryl amine via the Bamberger rearrange-
ment. Since the Bamberger rearrangement can be a strict
chemical reaction or can be catalyzed by enzymes, an
experiment was conducted to evaluate which may be
occurring in crude cell extract systems. Two small-scale (20
mL) test systems (in triplicate) were spiked with 2,4-
dihydroxylamino-6-nitrotoluene to a concentration of 0.5
mM. One contained deoxygenated 5 mM Tris (pH ) 7.8);
the other contained deoxygenated 5 mM Tris (pH ) 7.8) and
0.05 mg of protein/ mL. After 24 h of incubation in an
anaerobic glovebox (90% N2/ 10% H2), no less than 95% of
the original 2,4-dihydroxylamino-6-nitrotoluene was ob-
served in extract-free systems (i.e., loss in any one system
never exceed 5%). In the systems containing cell extract,
transformation always exceeded 82%. These results clearly
demonstrated that the rapid rearrangement observed in these
systems is catalyzed by biological components of the crude
cell extract, similar to studies of hydroxylaminobenzene
rearrangement to aminophenols.
This result is particularly interesting, since C. acetobu-
tylicum was not grown in the presence of any nitroaromatic
compound, yet the enzyme that catalyzes the rearrangement
is present without a nitroaromatic inducer. In studies
mentioned previously (19, 23) where the Bamberger rear-
rangement of hydroxylaminobenzene was observed, cell
extracts were prepared from cultures actively growing on
nitroaromatics as their carbon and energy source. Conditions
might favor expression of nitroaromatic catabolic gene. Such
conditions would not be expected in C. acetobutylicum
growth on glucose.
FIGURE 3. Possible structures resulting from the Bamberger
rearrangement of 2,4-dihydroxylamino-6-nitrotoluene.
The ability of 2,4-dihydroxylamino-6-nitrotoluene to
undergo rearrangement during the derivatization procedure
implies that some of the products isolated may have resulted
from the small amount of 2,4-dihydroxylamino-6-nitrotolu-
ene remaining in the crude cell extract (7.5 mol %). The
yield after purification approached 80%, thus the product
isolated can not be a result of this potential artifact.
Attempts to further confirm structures with 13C spec-
troscopy were not successful (data not shown). The similarity
of chemical shifts resulting from derivatized amino or
derivatized hydroxylamino groups and the low response of
Any discussion of the enzymes directly involved in the
TNT transformation process by C. acetobutylicum is specu-
lative; however, activity without the presence of a nitroaro-
matic-inducing substrate may imply that the enzyme(s)
responsible are constitutive. Others studying the transfor-
mation of TNT by Clostridia in whole cell systems have also
observed that inducing substrates are not required for TNT
transformation to aminated derivitives (5). Previously, we
reported that the rate of TNT transformation in cell cultures
appears to be related to acidogenic hydrogenase activity (15).
FIGURE 4. Proposed pathway of transformation observed in C. acetobutylicum crude cell extracts. Brackets around the 6-amino-4-
hydroxylamino-3-methyl-2-nitrophenol Bamberger rearrangement product denote that this structure is possible but less likely than the
4-amino-6-hydroxylamino-3-methyl-2-nitrophenol product.
9
4 9 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 4, 1998

In crude cell extracts, TNT transformation to hydroxylamino
intermediates and eventually the Bamberger rearrangement
product occurred only when H2 was present as a component
of glovebox gassadding further evidence that hydrogenases
active during acidogenic metabolism by this organism are
involved, at least indirectly, in TNT transformation.
experimental systems. Since the product of 2,4-dihydroxy-
lamino-6-nitrotoluene rearrangement was extremely hydro-
philic, partitioning to soil organic matter would probably be
minimal in remediation systems. At this time, no experi-
ments have been conducted to assess the extent of either ion
exchange or partitioning soil systems.
Cell Cultures. As observed previously (16), TNT trans-
formation in cell cultures resulted in the rapid accumulation
and subsequent transformation of 2,4-dihydroxylamino-6-
nitrotoluene (data not shown). Following the derivatization,
samples were analyzed by HPLC comparing elution time
and UV spectra of peaks separated from the culture super-
natant to purified products obtained from cell extract studies.
One large peak was resolved from the culture broth that had
an identical retention time (14.3 min) and UV spectra when
compared to I from crude cell extract studies. This peak was
In summary, studies have demonstrated the ability of C.
acetobutylicum to reduce TNT to 2,4-dihydroxylamino-6-
nitrotoluene and then to a phenolic product(s) via the
Bamberger rearrangement. The product of transformation
is hydrophilic, highly reactive with oxygen, and decomposes
in aqueous solution in a matter of days under strict anaerobic
conditions. No inducing substances were required for C.
acetobutylicum to catalyze the transformation reactions, and
further evidence for the involvement of acidogenic hydro-
genases was obtained. Further study will be required to assess
both the propensity of other fermentative organisms to utilize
similar pathways for the transformation of nitroaromatics in
anaerobic environments and the long-term fate of Bamberger
rearrangement products in soil systems.
1
isolated and subjected to H NMR. Agreement between the
spectra was excellent [peaks obtained with relative areas in
parentheses were 2.24 (6), 2.39 (3), 2.49(3), 7.86 (1), and 8.07
(1)], confirming the that a Bamberger rearrangement product
of 2,4-dihydroxylamino-6-nitrotoluene was produced in
whole cell systems identical to that found in cell extracts.
The ability of bacteria to convert nitroaromatics, including
TNT, to aminated phenols may represent an important fate
process for these compounds in anaerobic environments.
Reduction of aryl nitro groups to the corresponding amine
has long been recognized as a reaction that occurs in natural
and engineered systems. The formation of aminophenols
through rearrangement of hydroxylamines may represent a
“competing” reaction that can alter the physical distribution
of contaminants, their chemical reactivity, and metabolic
pathways required for contaminant destruction. In this case,
the presence of a nitroaromatic inducer was not required for
activity to be expressed.
To what extent this finding may influence our under-
standing of TNT bioremediation systems is still unclear.
Certainly, anaerobic slurry processes fed sugars are ideal
systems for fermentative bacteria, including Clostridia. To
our knowledge, the analysis of samples for phenolic TNT
derivatives has not been conducted in field systems. Both
2,4-dihydroxylamino-6-nitrotoluene and “phenolic hydroly-
sis products of TAT” were observed in a study of TNT
transformation conducted with an anaerobic mixed labora-
tory enrichment culture (17). TAT was also a product in this
system, again demonstrating that several TNT transformation
processes may result in mixed culture systems.
Acknowledgments
Funds for this work were provided by the Hazardous
Substances Research Center South and Southwest. We would
like to thank Dr. Terry Marriot of the Rice University
Department of Chemistry for assistance with mass spec-
troscopy analysis and Dr. Addison Ault of the Cornell College
Department of Chemistry for insights into derivatization
procedures.
Literature Cited
(1) Rieger, P.-G.; Knackmuss, H.-J. In Biodegradation of Nitroaro-
matic Compounds; Spain, J. C., Ed.; Plenum Press: New York,
1995; Vol. 49, pp 1-18.
(2) Preuss, A.; Rieger, P.-G. In Biodegradation of Nitroaromatic
Compounds; Spain, J. C., Ed.; Plenum Press: New York, 1995;
Vol. 49, pp 69-86.
(3) Lewis, T. A.; Ederer, M. M.; Crawford, R. L.; Crawford, D. L. J.
Ind. Microbiol. Biotechnol. 1997, 18, 89-96.
(4) Preuss, A.; Fimpel, J.; Diekert, G. Arch. Microbiol. 1993, 159,
345-353.
(5) Ederer, M. M.; Lewis, T. A.; Crawford, R. L. J. Ind. Microbiol.
Biotechnol. 1997, 18, 82-88.
(6) McCormick, N. G.; Feeherry, F. E.; Levinson, H. S. Appl. Environ.
Microbiol. 1976, 31, 949-958.
(7) Boopathy, R.; Kulpa, C. F.; Wilson, M. Appl. Microbiol. Bio-
technol. 1993, 39, 270-275.
Interestingly, the fate of Bamberger rearrangement prod-
ucts in anaerobic bioremediation systems may be similar to
that of TAT. In our hands, the phenolic product was
extremely oxygen sensitive. Even under anaerobic condi-
tions, the compound would decompose to a reddish-brown
residue after several days in solution. TAT is also oxygen
sensitive and under oxidizing conditions becomes irreversibly
“bound” to soils (1). The association of TAT with soils is
thought to result from ion exchange processes of one or more
of its free amino groups. Phenolic products from the
Bamberger rearrangement will have a free amine and
hydroxyl group that could participate in ion-exchange process
as well.
Attempts to recover phenolic product(s) in a variety of
nonpolar (e.g., benzene, hexane, etc.) and moderately polar
solvents (e.g., ethyl acetate, dichloromethane, etc.) resulted
in essentially zero recovery. In comparison, essentially 100%
recovery of 2,4-dihydroxylamino-6-nitrotoluene was possible
with ethyl acetate extraction of aqueous samples. Estimated
pKa values of phenolic groups are 9.1 and 8.2 (A and B,
respectively) and 3.4 and 3.5 (A and B, respectively) for amino
groups (25, 26). On the basis of these calculations, neither
proton donating/ accepting groups of structures A and B
(Figure 2) would be extensively ionized in the pH values of
(8) Funk, S.; Roberts, D. J.; Crawford, D. L.; Crawford, R. L. Appl.
Environ. Microbiol. 1993, 59, 2171-2177.
(9) Gilcrease, P. C.; Murphy, V. G. Appl. Environ. Microbiol. 1995,
61, 4209-4214.
(10) Vorbeck, C.; Lenke, H.; Fischer, P.; Knackmuss, H.-J. J. Bacteriol.
1994, 176, 932-934.
(11) Biodegradation of nitroaromatic compounds, 1st ed.; Spain, J.
C., Ed.; Plenum Press: New York, 1995; Vol. 49, p 232.
(12) Fiorella, P. D.; Spain, J. C. Appl. Environ. Microbiol. 1997, 63,
207-2015.
(13) Boopathy, R.; Wilson, M.; Montemagno, C. D.; Manning, J. F.,
Jr.; Kulpa, C. F. Bioresour. Technol. 1994, 47, 19-24.
(14) Selim, H. M.; Xue, S. K.; Iskandar, I. K. Soil Sci. 1995, 160, 328-
339.
(15) Khan, T. A.; Bhadra, R.; Hughes, J. J. Ind. Microbiol. Biotechnol.
1997, 18, 198-203.
(16) Hughes, J. B.; Wang, C.; Bhadra, R.; Richardson, A.; Bennett, G.;
Rudolph, F. Environ. Toxicol. Chem. In press.
(17) Lewis, T. A.; Goszczynski, S.; Crawford, R. L.; Korus, R. A.;
Admassu, W. Appl. Environ. Microbiol. 1996, 62, 4669-4674.
(18) 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.
(19) Nishino, S. F.; Spain, J. C. Appl. Environ. Microbiol. 1993, 59,
2520-2525.
(20) Crawford, R. In Biodegradation of Nitroaromatic Compounds;
Spain, J. C., Ed.; Plenum Press: New York, 1995; pp 87-98.
9
VOL. 32, NO. 4, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 4 9 9

(21) Khan, T. A. Thesis, Rice University, 1996.
(22) Dean, J. A. Analytical Chemistry Handbook; McGraw-Hill: New
York, 1995.
(26) Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M.
Environmental Organic Chemistry; Wiley & Sons: New York,
1993.
(23) Schenzle, A.; Lenke, H.; Fischer, P.; Williams, P.; Knackmuss, H.
Appl. Environ. Microbiol. 1997, 63, 1421-1427.
(24) March, J. Advanced organic chemistry: reactions, mechanisms,
structure, 4th ed.; John Wiley and Sons: New York, 1992.
(25) Jaffe, H. H. Chem. Rev. 1953, 53, 191-261.
Received for review July 11, 1997. Revised manuscript re-
ceived October 31, 1997. Accepted November 12, 1997.
ES970612S
9
5 0 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 4, 1998