Converging catabolism of 2,4,6-trinitrophenol (picric acid) and 2,4-dinitrophenol by Nocardioides simplex FJ2-1A

Article abstract of DOI:10.1023/A:1014447700775

Initial F₄₂₀-dependent hydrogenation of 2,4,6-trinitrophenol (picric acid) generated the hydride σ-complex of picrate and finally the dihydride complex. With 2,4-dinitrophenol the hydride σ-complex of 2,4-dinitrophenol is generated. The hydride transferring enzyme system showed activity against several substituted 2,4-dinitrophenols but not with mononitrophenols. A K_m-value of 0.06 mM of the hydride transfer for picrate as substrate was found. The pH optima of the NADPH-dependent F₄₂₀ reductase and for the hydride transferase were 5.5 and 7.5, respectively. An enzymatic activity has been identified catalyzing the release of stoichometric amounts of 1 mol nitrite from 1 mol of the dihydride σ-complex of picrate. This complex was synthesized by chemical reduction of picrate and characterized by ¹H and ¹³C NMR spectroscopy. The hydride σ-complex of 2,4-dinitrophenol has been identified as the denitration product. The nitrite-eliminating activity was enriched and clearly separated from the hydride transferring enzyme system by FPLC. 2,4-Dinitrophenol has been disproven as a metabolite of picrate (Ebert et al. 1999) and a convergent catabolic pathway for picrate and 2,4-dinitrophenol with the hydride σ-complex of 2,4-dinitrophenol as the common intermediate has been demonstrated.

Full text of DOI:10.1023/A:1014447700775

Biodegradation 12: 367–376, 2001.
© 2002 Kluwer Academic Publishers. Printed in the Netherlands.
367
Converging catabolism of 2,4,6-trinitrophenol (picric acid) and
2,4-dinitrophenol by Nocardioides simplex FJ2-1A
Sybille Ebert1, Peter Fischer2 & Hans-Joachim Knackmuss1∗
1Institut für Mikrobiologie and Institut für 2Organische Chemie der Universität Stuttgart, Germany (∗ author for
correspondence)
Accepted 18 October 2001
Key words: denitrating activity, F420, hydride transfer, hydride σ-complex, hydride transferase
Abstract
Initial F420-dependent hydrogenation of 2,4,6-trinitrophenol (picric acid) generated the hydride σ-complex of
picrate and finally the dihydride complex. With 2,4-dinitrophenol the hydride σ-complex of 2,4-dinitrophenol is
generated. The hydride transferring enzyme system showed activity against several substituted 2,4-dinitrophenols
but not with mononitrophenols. A Km-value of 0.06 mM of the hydride transfer for picrate as substrate was
found. The pH optima of the NADPH-dependent F420 reductase and for the hydride transferase were 5.5 and 7.5,
respectively. An enzymatic activity has been identified catalyzing the release of stoichometric amounts of 1 mol
nitrite from 1 mol of the dihydride σ-complex of picrate. This complex was synthesized by chemical reduction
1
of picrate and characterized by H and 13C NMR spectroscopy. The hydride σ-complex of 2,4-dinitrophenol
has been identified as the denitration product. The nitrite-eliminating activity was enriched and clearly separated
from the hydride transferring enzyme system by FPLC. 2,4-Dinitrophenol has been disproven as a metabolite of
picrate (Ebert et al. 1999) and a convergent catabolic pathway for picrate and 2,4-dinitrophenol with the hydride
σ-complex of 2,4-dinitrophenol as the common intermediate has been demonstrated.
Abbreviations: 2-Cl-4,6-DNP, 2-chloro-4,6-dinitrophenol; 4,6-DNH, 4,6-dinitrohexanoate; DNOC, 2-methyl-
4,6-dinitrophenol; 2,4-DNP, 2,4-dinitrophenol; 2,6-DNP, 2,6-dinitrophenol; F420, coenzyme F420; 2H−-picrate,
dihydride complex of picrate; H−-picrate, hydride σ-complex of picrate; HTES, hydride-transferring enzyme
system; PCP, pentachlorophenol; 1,3,5-TNP, 1,3,5-trinitropentane; TNT, 2,4,6-trinitrotoluene.
Introduction
is the production of nitrobenzene from benzene (Patil
& Shinde 1989).
Because of the presence of the electron-withdrawing
nitro groups as substituents dinitrophenols and
trinitrophenols are not readily attacked by oxygenases
as generally observed with aromatic xenobiotic com-
pounds including o-nitrophenol (Zeyer & Kearney
1984) and p-nitrophenol (Hanne et al. 1993; Kadiy-
ala & Spain 1998; Munnecke & Hsieh 1974; Spain
et al. 1979). Consequently, initial biological reduction
of the aromatic system is more likely than an oxid-
ative attack (Rieger & Knackmuss 1995; Vorbeck et
al. 1998). Up to now aerobic bacterial degradation
of 2,4,6-trinitrophenol, 2,4,6-trinitrotoluene (TNT),
or 2,4-dinitrophenol (2,4-DNP) through initial mono-
Nitrophenols are important building blocks for the
synthesis of dyes, pesticides, and explosives. Because
of their ubiquitous use nitrophenols occur as contam-
inants in industrial effluents and at low concentrations
also in surface and groundwaters (Hallas & Alexan-
der 1983). 2,4,6-Trinitrophenol (picric acid) and its
ammonium or potassium salts (picrates) were used as
explosives during the first and second world war. As
a consequence they have been found as pollutants in
ground water at former production and certain mil-
itary sites (Wyman et al. 1992). The main source of
dinitrophenols and picric acid in industrial wastewater

368
or dioxygenation is unknown. An exception is 2,6-
dinitrophenol (2,6-DNP) which is subject to denitra-
tion by an initial dioxygenase (Ecker et al. 1992). Re-
duction occurs preferably at the nitro groups revealing
unproductive dead end metabolites as products. Char-
acteristically, gratuitous reduction of picric acid gener-
ates picramic acid as a dead end product (Tsukamura
1960). For picric acid, 2,4-DNP and TNT an un-
usual initial hydrogenation of the aromatic ring system
is observed during in vivo transformations. Only for
the nitrophenols was this hydrogenation identified as
part of a complete catabolic sequence. For these com-
pounds hydride σ-complexes have been identified as
the initial metabolites (Lenke & Knackmuss 1992;
Rieger et al. 1999; Vorbeck et al. 1998).
Following the OECD guidelines 2,4-DNP and
picrate proved to be not biodegradable by activated
sludge. Nevertheless, these compounds are utilized
and mineralized aerobically by Nocardioides simplex
FJ2-1A (Ebert et al. 1999) and several Rhodococ-
cus strains (Blasco et al. 1999; Lenke & Knackmuss
1992). It is suggested that extensive ring hydrogena-
tion results in a non-oxidative ring cleavage yielding
4,6-dinitrohexanoate (4,6-DNH). This metabolite was
found in the supernatant of resting cell experiments
with Rhodococcus erythropolis strains, HL 24-1 and
HL 24-2 (Lenke et al. 1992). 3-Nitroadipate has been
identified as a metabolite in the degradation of 2,4-
DNP by Rhodococcus sp. strain RB1 which is pro-
posed to be a product of an alternative pathway of
oxidative ring cleavage and concomitant elimination
of nitrite (Blasco et al. 1999).
We have described the hydride transfer to the aro-
matic ring system by a two-componentenzyme system
consisting of a NADPH-dependent F420 reductase and
a hydride transferase. Coenzyme F420 functions as
mediator for the hydride transfer and NADPH as the
source of reduction equivalents (Ebert et al. 1999).
In the present paper we further characterize the
hydride-transferring enzyme system (HTES) and its
substrate specifity. We identify the hydrogenation
products and an enzymatic denitration of the dihydride
σ-complex of picrate (2H−-picrate) to the hydride σ-
complex of 2,4-DNP (H–2,4-DNP). These findings
refute the former hypothesis that nitrite is eliminated
from the σ-complex of picrate (H−-picrate) generating
2,4-DNP (Ebert et al. 1999; Rieger et al. 1999).
Materials and methods
Resting cell experiments with nitrophenols
Cells of N. simplex were grown as described earlier
(Ebert et al. 1999) and harvested by centrifugation,
resuspended in phosphate buffer (pH 7.4; 50 mM), and
incubated on a water bath shaker at 30 ◦C. Turnover of
nitrophenols was followed by HPLC analysis.
Purification of the NADPH-dependent F420
reductase, the hydride transferase and the nitrite
eliminating activity
Preparation of cell-free extract by French press treat-
ment and purification of the enzymes of the HTES
to electrophoretic homogeneity were performed as
described previously (Ebert et al. 1999). Nitrite elim-
inating activity was concentrated by fast-protein li-
quid chromatography (FPLC). Cell extract (74.1 mg
of protein) was passed through an anion exchange
column (Q Sepharose HP 16/10, Pharmacia, Uppsala,
Sweden) equilibrated with 50 mM TRIS-HCl (pH
7.5). Nitrite eliminating activity was eluted from the
column with a linear gradient (300 ml) from 0 to 1
M NaCl in 50 mM TRIS-HCl (pH 7.5). Protein con-
centrations were estimated by the method of Bradford
(1976).
Purification of coenzyme F420
For isolation and quantification of coenzyme F420 the
strain was grown aerobically in batch culture in a 100 l
stirred tank reactor (Bioengineering, Wald, Switzer-
land) at 30 ◦C and 550 rpm in liquid culture medium
as described earlier (Ebert et al. 1999). The culture
was aerated at a rate of 50 l per min. After consump-
tion of picrate (decolorization of the medium) the cells
were fed again with 0.35 mM picrate and harvested by
centrifugation immediately after decolorization of the
medium.
Coenzyme F420 was extracted with 70% ethanol
out of 500 g wet cells of N. simplex as described by
Lin and White (1986). Purification was performed ac-
cording to Abken (1998). After the last purification
step coenzyme F420 containing fractions were pooled
and methanol was evaporated in vacuum at 40 ◦C bath
temperature with a rotation evaporator. The residual
aqueous solution of coenzyme F420 was lyophilized.
The concentration of coenzyme F420 was calculated on
the basis of a molar extinction coefficient at 420 nm of
41 400 M−1 cm−1 (pH 7.5) (Purwantini et al. 1992).

369
For recycling of the cofactor coenzyme F420 con-
taining reaction mixtures were heated to 100 C for
in 50 mM TRIS-HCl (pH 7.5). Reactions were started
by the addition of protein.
◦
15 min. The precipitate was removed by centrifuga-
tion. The supernatant was diluted 1 : 1 with 50 mM
TRIS-HCl (pH 8.5) and applied to an anion exchange
column (Resource Q (6 ml), Pharmacia, Uppsala,
Sweden) equilibrated with 50 mM TRIS-HCl (pH
8.5). The cofactor was eluted with 0.42 M NaCl.
Analytical methods
For quantification of substrates in stopped tests a
reversed-phase high performance liquid chromato-
graphy (HPLC) system with a Gromsil 120 Oc4
column (125 × 4 mm, particle size 5 µm, Grom,
Herrenberg, Germany) was used. The mobile phase
consisted of 20% acetonitrile, 80% water and 0.26%
H3PO4 (v/v/v). For determination of reduced nitroaro-
matic compounds an anion exchange column (Gromsil
TSK- Gel Q-5PW, 125 × 4 mm, particle size 20 µm,
Grom, Herrenberg, Germany) was used as a solid
phase. The mobile phase consisted of 25 mM TRIS-
HCl (pH 8). A linear gradient from 0.35 to 0.6 M NaCl
in basic buffer was applied. Purity of the dihydride
complex of picrate was examined by ion pair chroma-
tography on the reversed-phase Gomsil column with
an isocratic eluent consisting of 20% methanol-water
and 5 mM tetrabutylammonium hydrogen sulfate ad-
justed to pH 12 with sodium hydroxide. Concentra-
tions were determined at 210 nm, 240 nm, 260 nm
and 300 nm. Products were identified by UV-visible
spectra using a photodiode array detector (UVD 340S)
from Gynkotek (Germering, Germany).
Enzyme assay
The activity of the HTES towards picrate or 2,4-
DNP as substrates was routinely assayed photomet-
rically as described earlier (Ebert et al. 1999). Tests
which were followed by HPLC analysis contained a
higher NADPH concentration (1 mM) and substrate
concentrations between 0.05 and 0.15 mM 2,4-DNP
or picrate. Reactions for repeated recording of UV-
visible spectra were started by addition of hydride
transferase.
For determining the pH optimum of the NADPH-
dependent F420 reductase the assay mixture contained
buffer (citric acid – phosphate buffer for the pH range
of 3 to 7 or 50 mM TRIS-HCl for pH 7 to 8.5), 0.5 mM
NADPH, 0.044 mM coenzyme F420 and 0.3 µg per
ml NADPH-dependent F420 reductase. Activity was
followed by the decreasing absorption at 340 nm. The
test was carried out anaerobically in microtiterplates
with a microplate scanning spectrophotometer (Power
Nuclear magnetic resonance (NMR) spectra were
recorded with an ARX 500 spectrometer (Bruker,
Rheinstetten, Germany) at a nominal frequency of
500.14 MHz (1H) and 125.77 MHz (13C), respectively.
Samples were dissolved in D2O.
ꢀ
Wavex, Bio-tek Instruments, Inc., Winooski, VT,
USA) equipped with the KC4 software v. 2.5 (Bio-
ꢀ
tek Instruments, Inc., Winooski, VT, USA) under
nitrogen at 30 ◦C and started by addition of substrate.
For determining the pH optimum of the hy-
dride transferase the assay mixture contained buffer
(100 mM succinic acid – NaOH buffer for the pH
range of 3.5 to 5, 100 mM potassium phosphate buffer
for pH 5 to 7.5, 100 mM TRIS-HCl for pH 7.5 to 10,
100 mM glycine – NaOH for pH = 10 to 12), 1 mM
NADPH, 3.5 µM coenzyme F420, 0.1 mM 2,4-DNP,
0.1 µg per ml NADPH-dependent F420 reductase and
78 ng per ml hydride transferase.
Materials
Synthesis of the dihydride σ-complex of picrate as
authentic reference compound followed the procedure
described by Severin and Schmitz (1962). The hydride
σ-complex of picrate was synthesized as described
by Rieger et al. (1999). According to Behrend and
Heesche-Wagner (1999) an authentic standard of H−-
2,4-DNP was prepared by adding sodium borohydride
to an aequeous solution of 2,4-DNP. Immediately after
chemical reduction the pH value of the reaction mix
was adjusted to pH 7.5 with phosphoric acid. This H−-
2,4-DNP standard in water was rather unstable and
could be used only for up to 2 hours even if stored
on ice.
Transformation of 2H−-picrate was routinely mon-
itored spectrophotometrically by following the de-
crease of absorption at 390 nm or by repeated re-
cording of the UV-visible spectrum in the wavelength
range between 250 and 600 nm. The reaction mix-
ture contained the chemically synthesized dihydride
complex of picrate and cell extract of Nocardioides
simplex FJ2-1A or fractions of FPLC purification steps

370
Results
eously regenerated 2,4-DNP as detected by HPLC
measurements. This effect has been described as a
non-enzymatic disproportionation by Behrend (1999).
Several nitroaromatic compounds were tested as
potential substrates for the HTES. Reactions were fol-
lowed by repeated recording of UV-visible spectra and
HPLC measurements. Mononitrophenols such as o-
nitrophenol and p-nitrophenol and 2,6-DNP were not
transformed by the HTES. As observed with whole
cells DNOC was a poor substrate for the HTES
(nearly 1% of activity with picrate). But increasing
absorption during repeated recording of UV-visible
spectra between 400 and 550 nm indicated formation
of a hydride σ-complex. This is even more pro-
nounced with 2-chloro-4,6-dinitrophenol (2-Cl-4,6-
DNP) as substrate (approximately 90% of the activity
with picrate).
Growth of Nocardioides simplex FJ2-1A and
turnover of substrates by resting cells
Nocardioides simplex FJ2-1A can grow with picrate
as a nitrogen, carbon, and energy source. However,
growth with the nitroaromatic compound in mineral
medium is very slow (td = 5d). Therefore, acetate,
yeast extract, peptone, and casamino acids were sup-
plemented to grow the strain within 26 h to an optical
density of 3.5 at 546 nm.
Content of coenzyme F420
To quantify the amount of coenzyme F420 involved in
the hydride transfer by the HTES, wet cells of N. sim-
plex were extracted with 70% ethanol (Lin & White
1986) and the supernatant purified by FPLC (Abken
1998). This revealed 139 µmol of cofactor F420 per kg
of wet cells. The content is lower than in Methanobac-
terium thermoautotrophicum (376 µmol per kg of wet
cells) but nearly 16-fold higher than in Streptomyces
griseus (8 µmol per kg of wet cells), Streptomyces
coelicolor (9 µmol per kg of wet cells), or Halobac-
terium cutirubrum (9 µmol per kg of wet cells) (Eker
et al. 1989).
Hydride transfer to picrate
Repeated scans of UV-visible spectra during picrate
transformation with the HTES revealed a fast increase
of absorption between 400 and 550 nm which is fol-
lowed by a slow decrease in the same spectral range
(Figure 1). The yellow color of picrate in the reaction
assay mix turned quickly to orange and then slowly
to pale yellow. The hydride σ-complex of picrate
(H−-picrate) was identified as the orange metabol-
ite by HPLC measurements. Therefore H−-picrate,
as described by Rieger et. al. (1999), could not be
the final product of the enzymatic reaction and fur-
ther transformation occurred. The enzymatic reaction
was stopped with phosphoric acid and the samples
were analyzed with HPLC under acidic conditions. A
sharp peak at a retention time of 7.6 min and a max-
imum absorption of 203 nm corresponded to the data
measured for a standard of 1,3,5-TNP. As described
by Severin and Adam (1963) 1,3,5-TNP is generated
quantitatively by chemical decarboxylation of 2H−-
picrate under acidic conditions. This indicates that
two hydride ions were transferred by the HTES and
2H−-picrate was the product.
Enzyme characterization
The NADPH-dependent F420 reductase and the hy-
dride transferase are cytosolic enzymes. No activity
was found in the membrane fraction. Activity of hy-
dride transfer was maximal at pH 7.5. The NADPH-
dependent F420 reductase had a maximum activity at
pH 5.5. The HTES could be stimulated by ammonium
sulfate at a concentration of 1 M. In order to function
as a hydride donor for the reaction catalyzed by the
hydride transferase coenzyme F420 must be reduced
by the NADPH-dependent F420 reductase and NADPH
(Ebert et al. 1999). FAD or FMN could not substitute
for coenzyme F420. Hence, the hydride transfer was
investigated in a coupled test with an excess of the
F
420-reductase and NADPH.
The affinity of the hydride transferase for picrate
Synthesis of the authentic 2H−-picrate and
spectroscopic characterization
was determined by the coupled enzyme test. The reac-
tion was stopped with phosphoric acid and substrate
turnover quantified by HPLC. The apparent Km for
picrate was 0.06 mM. Substrate inhibition occurred
at picrate concentrations above 0.125 mM. Turnover
of 2,4-DNP did not follow Michaelis–Menten kin-
etics. H−-2,4-DNP was very unstable and spontan-
2H−-Picrate was synthesized by reducing picrate in
alkaline solution with sodium borohydride (Severin &
Schmitz 1962) and precipitated by addition of meth-
anol. HPLC analysis under alkaline conditions gave a
single peak (retention time 6.2 min). The UV-visible

371
Figure 1. Repeated recording of UV-visible spectra at intervals of
35 seconds during an enzymatic transformation containing 0.1 mM
Figure 2. pH-Dependence of the UV-visible spectrum of
2H -picrate (0.1 mM). The heavy line depicts the spectrum at pH
7.5, which was used for spectrophotometric assays.
−
picrate, NADPH-dependent F
reductase and hydride transferase
420
(11.5 µg/ml and 0.04 µg/ml protein, respectively), coenzyme F
420
(11 µM), NADPH (1 mM), and TRIS-HCl (50 mM, pH 7.5). The
spectra shown by heavy lines indicate increasing absorbance due to
formation of the hydride σ- complex of picrate (0–105 sec) which
subsequently declined (thin lines) corresponding to the formation of
aliphatic sp3 carbon atoms, was assigned to C-3,5.
The 1J connectivity between this 30 ppm 13C and the
4.04 ppm proton resonance is established by a 13C, 1H
−
2H -picrate (105–455 sec).
1
COSY (C, H correlation). A H test spectrum, taken
after the time-consuming 13C and correlation NMR
experiments, clearly demonstrated the structural integ-
rity of the 2H−-picrate complex even after 24 hours.
Tests with the HTES unequivocally established the
identity of the biological transformation and chemical
reduction product of picrate and H−-picrate; HPLC
retention times and UV-visible spectra were identical.
Authentic 2H−-picrate was used to detect activity of
its further transformation.
spectrum in alkaline solution (pH 13) showed a max-
imum at 390 nm as described by Severin and Schmitz
(1962) which changed characteristically both as to po-
sition and intensity with decreasing pH (Figure 2). A
slight hypsochromic shift of the maximum from 390
to 385 nm and a considerable hypochromic effect was
observed on lowering the pH to 7.5 (enzymatic test
conditions). The increase of absorption around 450 nm
is due to the increasing instability of 2H−-picrate at
pH ≤ 6.5.
Identification of the product of the hydride transfer to
2,4-DNP
1
The H and 13C NMR spectra (D2O, 300 K) un-
equivocally confirmed the structure of the dihydride
1
σ-complex of picrate. The H spectrum showed one
Intensive hydrogenation of 2,4-DNP yielding 2,4-
dinitrocyclohexanone has been discussed in literature
(Lenke et al. 1992). Therefore we analyzed whether
one or two hydride ions were transferred to 2,4-
DNP by the HTES of N. simplex FJ2-1A. During
transformation of 2,4- DNP an increasing absorption
between 400 and 550 nm and an isosbestic point at
405 nm was observed. H−-2,4-DNP as described by
Behrend and Heesche-Wagner (1999) was formed.
The UV-visible spectrum showed the characteristic
absorption spectrum of the H−-2,4-DNP which was
identical with that of an authentic standard (Behrend
& Heesche-Wagner 1999). The reduction was not
complete as analyzed in HPLC measurements. Since
H−-2,4-DNP was highly unstable and spontaneously
singlet only, at δ = 4.04 ppm, for the four chemic-
ally equivalent protons in positions 3 and 5 of the
cyclohexanoid ring system.
1
From the 13C satellites a J(C, H) coupling con-
stant of 135 Mz is extracted which is characteristic for
an sp3 carbon atom with electronegative substituents.
The 13C NMR spectrum showed two signals at δ =
120.98 and 120.15 ppm with a relative intensity of
2 : 1 for the sp2 carbon atoms bearing the nitro sub-
stituents (C-2, 6 and C-4). The quasi-carbonyl C-1
resonance was located at δ = 175.65 ppm, a value
typical for the ipso carbon of a phenolate or the C-
1 carbon of an enolate anion. An intensive singlet at
δ = 30.41 ppm, i.e., in the region characteristic for

372
disproportionated to 2,4-DNP, aminonitrophenol and
minor unidentified products (Behrend 1999) a com-
plete mass balance of 2,4-DNP hydrogenation could
not be established. HPLC analysis using an anion ex-
changer column and UV-visible spectra recorded dur-
ing an HPLC run verified the identity of the chemically
synthesized H−-2,4-DNP and the enzymatic product.
Hydride transfer and subsequent nitrite elimination
H−-Picrate has been identified as the initial metabolite
of picrate hydrogenation by its UV-visible spectrum
during HPLC measurements. Repeated recording of
UV-visible spectra clearly showed that the first hydride
transfer is faster than the second one. Transformation
of 2H−-picrate with concomitant nitrite elimination
was catalyzed by crude extract as demonstrated by
successive UV-visible spectral traces and the detection
of nitrite in the reaction mix. Therefore the 2H−-
picrate transforming activity was called denitrating
activity.
Figure 3. Successive UV-visible spectral traces (5 min intervals)
during turnover of 2H -picrate (heavy line) by enriched nitrite
eliminating activity (0.024 mg protein in 50 mM TRIS-HCl (pH
7.5)).
−
TRIS-HCl (pH 7.5) and phosphate buffer did not influ-
ence activity. No positive effects were observed when
metal ions such as Mg2+, Ca2+, Mn2n, Cu2+, Co2+
,
All fractions obtained by FPLC separation of the
crude extract were tested for activity of nitrite elim-
ination from the chemically synthesized H−-picrate
(Rieger et al. 1999) or from 2H−-picrate. As observed
with picrate no H−-picrate transforming activity was
detected in any single fraction. Only upon combin-
ation of fractions, containing the components of the
HTES, H−-picrate was transformed with concomit-
ant NADPH consumption but without nitrite release.
Both σ-complexes H−- and 2H−-picrate are stable
under physiological conditions and nitrite was re-
leased neither spontaneously nor unspecifically by
bovine serum albumin or denaturated enzyme. In con-
trast, nitrite-eliminating activity was observed in one
protein fraction after the first enrichment step with
2H−-picrate as substrate, but not with H−-picrate.
Ni2+, Zn2+, Zn2+, K+, or Na+ were added to the
reaction mix. Neither the presence of picrate nor 2,4-
DNP affected the nitrite-eliminating activity. Addition
of 0.1 M dithiothreitol, 1 M ammonium sulfate, or 1 M
cyanide proved inhibitory.
The enzymatic reaction with 2H−-picrate was fol-
lowed by repeated recording of UV-visible spectra
(Figure 3). Isosbestic points indicated a linear correl-
ation of changes in concentration of reacting species
in the reaction mix and the final UV-visible spec-
trum suggested the formation of H−-2,4-DNP. The
product of the reaction was unequivocally identified
as H−-2,4-DNP by HPLC and its UV-visible spec-
trum measured in situ. A freshly synthesized authentic
preparation of H−-2,4-DNP served as reference (Fig-
ure 4). Using an anion exchanger column H−-2,4-
DNP eluted with a retention time of 8.2 min at a salt
concentration of 0.471 M in a gradient of 0.35 to 0.6 M
sodium chloride.
Quantification of nitrite and the product of nitrite
elimination
Like HTES the nitrite-eliminating activity was loc-
ated in the cytosolic fraction of N. simplex cells and
could be enriched by FPLC using an anion exchanger
column. The activity eluted at a ionic strength of
0.28 M sodium chloride. This purification step al-
lowed a separation from the components of the HTES
which eluted at higher ionic strength (Ebert et al.
1999). Further purification efforts including hydro-
phobic chromatography, chromatography on a hy-
droxylapatite column, and strong anionic exchange
column failed because of loss of activity. The buffers
Nitrite release from picrate, H−-picrate, or 2H−-
picrate as substrate was compared and quantified col-
orimetrically. Only with 2H−-picrate as substrate the
nitrite eliminating protein could release nitrite which
was formed in stoichiometric amount. A combina-
tion of the HTES and the nitrite-eliminating activity
transformed both picrate and the H−-picrate to H−-
2,4-DNP and nitrite as the only products. No nitrite
elimination was observed when DNOC or 2-Cl-4,6-
DNP was transformed by the combination of the
HTES and the denitrating activity.

373
The second hydrogenation step, reducing H−-
picrate to 2H−-picrate, was hitherto considered as
a reaction that misroutes the xenobiotic into an un-
productive pathway. Such metabolic misrouting has
been addressed by Vorbeck et al. (1998) as one of
the catabolic reactions that prevent complete bio-
degradation of TNT. Formation of hydride and di-
hydride σ-complexes of TNT has also been observed
with the picrate-utilizing Rhodococcus erythropolis
strain HL PM-1, the 4-nitrotoluene-utilizing Myco-
bacterium sp. strain HL 4-NT-1 (Vorbeck et al.
1998), and the pentaerythritol tetranitrate or glycerol
trinitrate-utilizing Enterobacter cloacae PB2 (Willi-
ams & Bruce 2000). Except for the latter organism
these hydrogenations of TNT are gratuitous reactions
that actually give rise to hydride complexes as dead
end products. Slow growth of the E. cloacae strain
on TNT as a nitrogen source under release of nitrite
(Williams & Bruce 2000), however, indicates that hy-
dride complex formation may allow partial utilization
of TNT.
Figure 4. UV-visible spectra recorded during HPLC run of authen-
−
tic H -2,4-DNP (right y axis, bold line)and biologically formed
−
H
-2,4-DNP during the reaction catalyzed by the nitrite eliminating
−
activity with 2H -picrate as substrate (left y-axis, thin line).
Discussion
The formerly described NADPH-dependent F420
-
reductase and the hydride transferase in N. simplex
FJ2-1A were further characterized. They were con-
stitutively expressed which is in contrast to Rhodococ-
cus erythropolis HL PM-1, another picrate degrading
strain (Russ et al. 2000). Correspondingly, the con-
centration of cofactor F420 in N. simplex was constant
irrespective of the presence of picrate in culture me-
dium. Nevertheless, the enzymes of the two organisms
are isofunctional and very similar. This is shown by
the N-terminal sequence of the NADPH-dependent
F420 reductase being almost identical and both en-
zymes are specific for NADPH as a hydride donor for
coenzyme F420. For other F420-dependent NADPH re-
ductases from actinomycetes or archaea the pH optima
for the reduction of coenzyme F420 are in the same
range of pH 5.5.
The hydride transferase of N. simplex also revealed
high similarity to the corresponding enzyme, named
F420-dependent dehydrogenase (dh2), from Rhodo-
coccus erythropolis HL PM-1 (Russ et al. 2000). Both
enzymes catalyse the transfer of hydride to picrate
or 2,4-DNP, the latter generating the same character-
istic UV-visible spectrum of H−-2,4-DNP described
by Behrend and Heesche-Wagner (1999). The color
change from the bright yellow of picrate to the or-
ange red of H−-picrate and finally to a pale yellow of
2H−-picrate can be rationalized in terms of a success-
ive confinement of the delocalized π-electron system.
The double hydride transfer into positions 3 and 5
of the carbon backbone creates two tetragonal centres
which interrupt the original planar aromatic π-system
of picrate.
Enzymatic release of nitrite from H−-picrate and
concomitant rearomatisation to 2,4-DNP were con-
sidered as the initial key reactions for productive
breakdown of picrate (Ebert et al. 1999; Rieger et al.
1999). The present results, however, clearly indicate
that cell free extract of N. simplex eliminated nitrite
from 2H−-picrate. H−-2,4-DNP and stoichiometric
amounts of nitrite were found by enzymatic trans-
formation containing the enriched nitrite-eliminating
activity and authentic 2H−-picrate as substrate. No
such activity was found with H−-picrate as substrate.
Hydride transfer to picrate, H−-picrate, or 2,4-
DNP takes place at position 3 of the aromatic
ring. Picrate, 2,4-DNP, and other structurally related
dinitrophenols, e.g., DNOC, 2-Cl-4,6-DNP, and 2,6-
DNP, were tested as substrates for the HTES and the
denitrating activity. Compared to whole cells specific
activities of the partially purified HTES with picrate or
2,4-DNP as substrates (Ebert et al. 1999) were 2000
or 1000 fold higher, respectively. Purified HTES re-
vealed specific activities of 45 U per mg of protein
with picrate and 24 U per mg of protein with 2,4-DNP
as substrate. Since 2,4-DNP is generated by dispro-
portionation of H−-2,4-DNP an apparent lower value
of activity with 2,4-DNP was observed. This is in con-
trast to the partially purified HTES which contained
H−-2,4-DNP transforming activities.
DNOC is transformed only very slowly by HTES.
Hydrogenation stops at the level of a hydride σ-
complex of DNOC as indicated by characteristic

374
Figure 5. Convergent degradation of picrate and 2,4-DNP by N. simplex.
changes of the UV-visible spectra. Obviously, hy-
dride is transferred only to dinitrophenols with nitro
groups in 2 and 4 position. No nitrite was elimin-
ated when enriched nitrite-eliminating activity was
added to the reaction mixture. Under the same con-
ditions 2-Cl-4,6-DNP was readily transformed by the
HTES. UV-visible spectroscopy again indicated hy-
dride σ-complex formation. However, nitrite was not
eliminated when denitrating activity was added. Our
observations with different substituted nitrophenols
clearly identify two structural elements that are ne-
cessary for ring hydrogenation and subsequent nitrite
elimination. (i) At least two nitro groups in 2- and
4-position must be present for hydride transfer. (ii)
Nitrite elimination requires a double hydrogenation
of the 2,4,6-trinitroaromatic ring. By conversion ex-
periments and enzyme tests shown here it is evident
that the electron withdrawing substituents of 2,4,6-
substituted dinitrophenols increase the activity and
electron donating groups such as methyl in DNOC
decrease the activity of hydrogenation.
Accumulation of non-stoichiometric amounts of
4,6-DNH in the culture fluid is characterized by a
fast increase and a slow decrease in concentration.
Its occurrence and the formation of stoichiomet-
ric amounts of 2-methyl-4,6-dinitrohexanoate from
DNOC by picrate grown cells of Rhodococcus eryth-
ropolis HL PM-1 (Lenke & Knackmuss 1996) indicate
further hydrogenation of H−-2,4-DNP to a dinitro-
cyclohexanone and subsequent hydrolytic ring cleav-
age. Actually, H−-2,4-DNP is further converted by
crude extract and NADPH stimulates this reaction
(Behrend 1999). Because of the chemical instability
of 4,6-DNH under physiological conditions (t1/2 ap-
proximately 55 min at pH 7.2) its occurrence in culture
fluid does not prove its role as a metabolite in the
degradation pathway (Ebert et al. 1999). A NADPH-
dependent monooxygenation in para-position has also
been suggested for the degradation of H−-2,4-DNP

375
by Behrend and Heesche-Wagner (1999). Zablotow-
icz et al. (1999) proposed an initial attack at 2,4-DNP
by a PCP 4-monooxygenase (pentachlorophenol 4-
monooxygenase) causing the release of nitrite and
concomitant hydroxylation in 4-position by a 2,4-DNP
degrading Sphingomonas sp. UG 30 strain. To elu-
cidate the lower pathway of 2,4-DNP and picrate de-
gradation further investigations are necessary. Figure 5
depicts the upper pathway of picrate and 2,4-DNP
catabolism.
Ecker S, Widmann T, Lenke H, Dickel O, Fischer P, Bruhn C &
Knackmuss H-J (1992) Catabolism of 2,6-dinitrophenol by Alc-
aligenes eutrophus JMP 134 and JMP 222. Arch. Microbiol. 158:
149-154
Eker APM, Hessels JKC & Meerwaldt R (1989) Characterization
of an 8-hydroxy-5-deazaflavin: NADPH oxidoreductase from
Streptomyces griseus. Biochim. Biophys. Acta 990: 80–86
Hallas LE & Alexander M (1983) Microbial transformation of
nitroaromatic compounds in sewage effluent. Appl. Environ.
Microbiol. 45(4): 1234–1241
Hanne LF, Kirk LL, Appel SM, Narayan AD & Bains KK (1993)
Degradation and induction specifity in Actinomycetes that de-
grade p-nitrophenol. Appl. Environ. Microbiol. 59(10): 3503–
3508
Kadiyala V & Spain JC (1998) A two-component monooxygenase
catalyzes both the hydroxylation of p-nitrophenol and the oxidat-
ive release of nitrite from 4- nitrocatechol in Bacillus sphaericus
JS905. Appl. Environ. Microbiol. 64(7): 2479–2484
Lenke H & Knackmuss HJ (1992) Initial hydrogenation during cata-
bolism of picric acid by Rhodococcus erythropolis HL 24-2.
Appl. Environ. Microbiol. 58(9): 2933–2937
The present data clearly identify 2H−-picrate as a
key metabolite of picrate catabolism. Nitrite is elim-
inated enzymatically from this complex but not from
H−-picrate. Therefore, 2,4-DNP cannot be a meta-
bolite in the degradation of picrate by N. simplex as
assumed for the degradation of picrate by Rhodo-
coccus erythropolis HL PM-1 (Behrend & Heesche-
Wagner 1999; Ebert et al. 1999; Lenke & Knackmuss
1992; Rieger et al. 1999). Consequently, 2,4-DNP and
picrate are degraded via a convergent pathway with
H−-2,4-DNP as a common metabolite.
Lenke H & Knackmuss H-J (1996) Initial hydrogenation and extens-
ive reduction of substituted 2,4-dinitrophenol. Appl. Environ.
Microbiol. 62(3): 784–790
Lenke H, Pieper DH, Bruhn C & Knackmuss H-J (1992) De-
gradation of 2,4-dinitrophenol by two Rhodococcus erythropolis
strains, HL 24-1 and HL-24-2. Appl. Environ. Microbiol. 58(9):
2928–2932
Lin X-L & White RH (1986) Occurence of coenzyme F
and its
420
γ -monoglutamyl derivative in nonmethanogenic archaebacteria.
J. Bacteriol. 168(1): 444–448
Acknowledgements
Munnecke DM & Hsieh DPH (1974) Microbial degradation of
parathion and p-nitrophenol. Appl. Microbiol. 28(2): 212–217
Patil SS & Shinde VM (1989) Gas chromatographic studies on the
biodegradation of nitrobenzene and 2,4-dinitrophenol. Environ.
Pollut. 57: 235–250
We thank P.-G. Rieger for providing the authentic H−-
picrate, DuPont de Nemours Company for supplying
us with Nocardioides simplex FJ2-1A and J. Rebell for
recording the NMR spectra.
Purwantini E, Mukhopadhyay B, Spencer RW & Daniels L (1992)
Effect of temperature on the spectral properties of coenzyme
F
and related compounds. Anal. Biochem. 205: 342–350
420
Rajan J, Valli K, Perkins RE, Sariaslani FS, Barns SM, Reysenbach
A-L, Rehm S & Ehringer M (1996) Mineralization of 2,4,6-
trinitrophenol (picric acid): characterization and phylogenetic
identification of microbial strains. J. Ind. Microbiol. 16: 319–324
Rieger P-G & Knackmuss H-J (1995). Basic knowledge and per-
spectives on biodegradation of 2,4,6-trinitrotoluene and related
nitroaromatic compounds in contaminated soil. In Spain JC (ed)
Biodegradation of nitroaromatic compounds. Plenum Press, New
York
Rieger P-G, Sinnwell V, Preuß A, Francke W & Knackmuss H-J
(1999) Hydride-Meisenheimer complex formation and protona-
tion as key reactions of 2,4,6-trinitrophenol biodegradation by
Rhodococcus erythropolis. J. Bacteriol. 181(4): 1189–1195
Russ R, Walters DM, Knackmuss H-J & Rouvière P (2000). Iden-
tification of genes involved in picric acid and 2,4-dinitrophenol
degradation by mRNA differential display. In Spain JC, Hughes
JB & Knackmuss H-J (ed) Biodegradation of nitroaromatic
compounds and explosives. Lewis Publishers, Boca Raton
Severin T & Adam M (1963) Umsetzung mit Natriumborhydrid, II.
Chem. Ber. 96: 448–452
References
Abken H-J (1998) Reinigung und Chrakterisierung der F H2-
420
Dehydrogenase aus Methanosarcina mazei Gö 1 und Identifizier-
ung von Methanophenazin als neuartigen methanogenen Cofak-
tor. Mathematisch- naturwissentschaftliche Fakultät. Göttingen,
Universität von Göttingen
Behrend C (1999) Pikrinsäuremetabolismus in Nocardioides sp.
CB-22. Institut für Biochemie. Witten/Herdecke, Privatuni-
versität Witten/Herdecke
Behrend C & Heesche-Wagner K (1999) Formation of hydride-
Meisenheimer complexes of picric acid (2,4,6-trinitrophenol)
and 2,4-dinitrophenol during mineralization of picric acid by
Nocardioides sp. strain CB 22-2. Appl. Environ. Microbiol.
65(4): 1372–1377 Blasco R, Moore E, Wray V, Pieper D, Timmis
K & Castillo F (1999) 3-Nitroadipate, a metabolic intermedi-
ate for mineralization of 2,4-dinitrophenol by a new strain of
Rhodococcus Species. J. Bacteriol. 181(1): 149–152
Bradford MM (1976) A rapid and sensitive method for the quanti-
fication of microgram quantities of protein utilising the principle
of protein-dye binding. Anal. Biochem. 72: 248–252
Severin T & Schmitz R (1962) Umsetzung von Nitroaromaten mit
Natriumborhydrid. Chem. Ber. 95: 1417–1419
Spain JC, Wyss O & Gibson DT (1979) Enzymatic oxidation of p-
nitrophenol. Biochem. Biophys. Res. Commun. 88(2): 634–641.
Ebert S, Rieger P-G & Knackmuss H-J (1999) Function of coen-
zyme F
in aerobic catabolism of 2,4,6-trinitrophenol and
420
2,4-dinitrophenol by Nocardioides simplex FJ2-1A. J. Bacteriol.
181(9): 2669–2674

376
Tsukamura M (1960) Enzymatic reduction of picric acid. J. Bio-
chem. 48(5): 662–671
Wyman JF, Serve MP, Hobson DW, Lee LH & Uddin D
(1992) Acute toxicity, distribution, and metabolism of 2,4,6-
trinitrophenol (picric acid) in Fischer 344 rats. J. Toxicol.
Environ. Health 37: 313–327
Zablotowicz RM, Leung KT, Alber T, Cassidy MB, Trevors JT, Lee
H, Veldhuis L & Hall JC (1999) Degradation of 2,4-dinitrophenol
and selected nitroaromatic compounds by Sphingomonas sp.
UG30. Can. J. Microbiol. 45: 840–848
Zeyer J & Kearney PC (1984) Degradation of o-nitrophenol and m-
nitrophenol by a Pseudomonas putida. J. Agric. Food Chem. 32:
238–242
Vorbeck C, Lenke H, Fischer P, Spain JC & Knackmuss H-J (1998)
Initial reductive reactions in aerobic microbial metabolism of
2,4,6-trinitrotoluene. Appl. Environ. Microbiol. 64(1): 246–252
Williams RE & Bruce NC (2000). The role of nitrate ester re-
ductase enzymes in the biodegradation of explosives. In Spain
JC, Hughes JB & Knackmuss H-J (ed) Biodegradation of ni-
troaromatic compounds and explosives. Lewis Publisher, Boca
Raton