APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Nov. 1997, p. 4272–4281 Vol. 63, No.

11
0099-2240/97/$04.0010
Copyright © 1997, American Society for Microbiology

Degradation of the Fluoroquinolone Enrofloxacin by the Brown Rot

Fungus Gloeophyllum striatum: Identification of Metabolites
HEINZ-GEORG WETZSTEIN,1* NORBERT SCHMEER,1 AND WOLFGANG KARL2
1 2
Animal Health Research and Central Research, Bayer AG,
D-51368 Leverkusen, Germany
Received 9 December 1996/Accepted 22 August 1997

The degradation of enrofloxacin, a fluoroquinolone antibacterial drug used in veterinary medicine, was
investigated with the brown rot fungus Gloeophyllum striatum. After 8 weeks, mycelia suspended in a defined
liquid medium had produced 27.3, 18.5, and 6.7% 14CO2 from [14C]enrofloxacin labeled either at position C-2,
at position C-4, or in the piperazinyl moiety, respectively. Enrofloxacin, applied at 10 ppm, was transformed
into metabolites already after about 1 week. The most stable intermediates present in 2-day-old supernatants
were analyzed by high-performance liquid chromatography combined with electrospray ionization mass spec-
trometry. Eight of 11 proposed molecular structures could be confirmed by 1H nuclear magnetic resonance
spectroscopy or by cochromatography with reference compounds. We identified (i) 3-, 6-, and 8-hydroxylated
congeners of enrofloxacin, which have no or only very little residual antibacterial activity; (ii) 5,6- (or 6,8-), 5,8-,
and 7,8-dihydroxylated congeners, which were prone to autoxidative transformation; (iii) an isatin-type com-
pound as well as an anthranilic acid derivative, directly demonstrating cleavage of the heterocyclic core of
enrofloxacin; and (iv) 1-ethylpiperazine, the 7-amino congener, and desethylene-enrofloxacin, representing
both elimination and degradation of the piperazinyl moiety. The pattern of metabolites implies four principle
routes of degradation which might be simultaneously employed. Each route, initiated by either oxidative
decarboxylation, defluorination, hydroxylation at C-8, or oxidation of the piperazinyl moiety, may reflect an
initial attack by hydroxyl radicals at a different site of the drug. During chemical degradation of [4-14C]en-
rofloxacin with Fenton’s reagent, five confirmatory metabolites, contained in groups i and iv, were identified.
These findings provide new evidence in support of the hypothesis that brown rot fungi may be capable of
producing hydroxyl radicals, which could be utilized to degrade wood and certain xenobiotics.

Fluoroquinolones (FQs) have found wide application in hu- (among other antibiotics) in a marine sediment was assessed.
man and veterinary medicine. They are active against a broad Its apparent depuration was attributed to leaching and redis-
spectrum of pathogenic gram-negative and gram-positive bac- tribution rather than to degradation (21).
teria (18, 49). Enrofloxacin (Fig. 1) has been developed for Recently, we have shown in vitro degradation of enrofloxa-
veterinary use to treat infections in pet animals and livestock cin by the white rot fungus Phanerochaete chrysosporium and by
(7, 44). In mammals, FQs can be metabolized by glucuronida- Gloeophyllum striatum, representing the brown rot fungi. After
tion, sulfation, N dealkylation, and oxidation of the amine 4 weeks of incubation, 6 and 16% 14CO2, respectively, were
substituent (27, 42). However, a major fraction is excreted produced from [4-14C]enrofloxacin in liquid cultures (45). Fur-
unchanged and introduced into the environment via animal thermore, four species of white rot fungi and three strains of
waste (51). Furthermore, manure from livestock is often dis- G. striatum were found to degrade [4-14C]enrofloxacin bound
posed of by being spread onto agricultural soil and pastures. to straw by up to 50% within 8 weeks. Enrofloxacin pread-
Little is known about both the biodegradability and fate of sorbed to agricultural soil could also be metabolized, although
antibiotics in the environment (17, 21, 31). Fluorinated aro- at a much lower rate, possibly reflecting decomposition of the
matic compounds, as exemplified by FQs, are xenobiotics; i.e., matrix to which the drug was bound (33).
they are man-made and have not been found to occur naturally White rot fungi like P. chrysosporium are known to degrade
(20, 36). However, FQs are tightly bound to soil and feces (13, the lignin component of woody plant cell walls as well as a wide
31, 34). In contrast to earlier views, such binding might be spectrum of pollutants such as polyaromatic hydrocarbons,
regarded as being quite a favorable characteristic of a sub- polychlorinated biphenyls, chlorinated pesticides (e.g., dichlo-
stance (23), because bound molecules should hardly be bio- rodiphenyltrichloroethane and lindane), and explosives (e.g.,
available. As a consequence, FQs might not exert a significant trinitrotoluene) (4, 6, 8–10, 15, 35). Such activities were attrib-
selection pressure in situ, e.g., in eliminating specific parts of a uted to lignin peroxidases, manganese-dependent peroxidases,
bacterial population or in selecting for resistance. However, and laccases. These extracellular enzymes are thought to cat-
strong binding can be expected to delay degradation and may alyze oxidative degradation reactions via diffusible agents like
partly explain the apparent recalcitrance of FQs (21, 31). aryloxy radicals, Mn31, and even hydroxyl radicals, with the
Biodegradation of sarafloxacin, an FQ used in poultry med- latter being produced by secondary reactions. Mechanistically,
icine, has been studied in three different types of soils. From none of these processes is fully understood (4, 14, 15, 19, 25).
[2-14C]sarafloxacin, #0.6% 14CO2 was produced after 80 days Brown rot fungi preferentially degrade the cellulose and
(31). In another investigation, the persistence of sarafloxacin hemicellulose components of plant cell walls, while lignin has
been shown to be modified primarily by hydroxylation and
demethylation and to a minor extent by depolymerization (15,
* Corresponding author. Phone: 49 2173 38 4882. Fax: 49 2173 38 32). For the decay of wood, a Fenton-type reaction mechanism
3766. was first postulated by Koenigs (30). In various model systems,

4272
VOL. 63, 1997 METABOLITES OF ENROFLOXACIN 4273

Mineral medium. The unbuffered mineral medium, devoid of a carbon, nitro-

gen, or phosphate source, contained 0.8 mM MgSO4, 0.2 mM CaCl2, 12 mM
H3BO3, 0.4 mM CoSO4, 0.2 mM CuSO4, 0.04 mM (NH4)6Mo7O24 (Merck,
Darmstadt, Germany), 2 mM MnSO4, and 0.4 mM ZnSO4 (Riedel-deHaën,
Seelze, Germany). This medium was sterilized at 121°C for 20 min. Prior to
inoculation, it was supplemented with 20 mM FeSO4 (Riedel-deHaën) and 1 ml
(per liter of medium) of a vitamin stock solution containing (per liter) 20 mg of
4-aminobenzoic acid sodium salt, 6 mg of D-(1)-biotin, 50 mg of DL-a-lipoic acid,
50 mg of nicotinamide, 10 mg of D-(1)-pantothenic acid sodium salt, 100 mg of
pyridoxine z HCl, 50 mg of (2)-riboflavin (Aldrich, Steinheim, Germany), 20 mg
of folic acid (Fluka, Neu-Ulm, Germany), 10 mg of thiamine z HCl (Merck), 20
mg of thiamine pyrophosphate chloride, and 50 mg of cyanocobalamin (Sigma,
FIG. 1. Molecular structure of enrofloxacin and positions of the 14C label in Deisenhofen, Germany). Sterile-filtered stock solution was stored frozen until
[2-14C]enrofloxacin (p), [4-14C]enrofloxacin (3), and [piperazine-2,3-14C]enro- use. The final pH of around 6 was maintained throughout the incubation period.
floxacin (1). Enrofloxacin and reference compounds. The chemical structure of enrofloxa-
cin z HCl [1-cyclopropyl-7-(4-ethyl-1-piperazinyl)-6-fluoro-1,4-dihydro-4-oxo-3-
quinolinecarboxylic acid hydrochloride], including the positions of 14C labeling,
is shown in Fig. 1. The unlabeled standard compound was provided by H. Rast
hydroxyl radicals are thought to be produced by a reductive (Bayer AG, Leverkusen, Germany); its chemical purity was .99.4%. Chemically
cleavage of hydrogen peroxide by Fe21 (15, 19, 25, 28, 50): synthesized reference compounds (Table 1) were kindly provided by W. Hallen-
bach (Bayer AG) (F-6, F-8, and F-11) and U. Petersen (Bayer AG) (F-4 and
Fe21 1 H2O2 3 Fe31 1 HOz 1 HO2. F-9).
Iron is abundant in decaying wood, while hydrogen peroxide Radioactively labeled enrofloxacin. 14C-labeled enrofloxacin was synthesized
could be generated by enzymes like cellobiose dehydrogenase, by R. Koch (Bayer Corp., Stilwell, Kans.) (C-2 and C-4 label) and by R. Thomas
aryl alcohol oxidase, or glyoxal oxidase. Superoxide radicals and R. Körmeling (Bayer AG, Wuppertal, Germany) (piperazine label). In the
latter laboratory, all labeled substrates were freshly purified immediately before
also have to be considered as a potential source of hydroxyl use. The specific activities of [4-14C]-, [2-14C]-, and [piperazine-2,3-14C]enro-
radicals (15, 19, 22, 24, 30, 37, 50). A partially purified glyco- floxacin z HCl were 4.98, 0.63, and 3.38 MBq/mg, respectively; the radiochemical
peptide from Tyromyces palustris was proposed to catalyze both purity was .98%, as determined by high-performance liquid chromatography
the formation of hydrogen peroxide from superoxide radicals (HPLC). The compounds were dissolved in sterile distilled water and stored
frozen until use.
and the subsequent reduction of hydrogen peroxide by bound Experimental procedures. Precultures of G. striatum were grown in malt broth,
Fe21 (22). Moreover, the formation of hydroxyl radicals during unagitated, at room temperature for 7 days. After the supernatant was discarded,
growth of Antrodia xantha on wood shavings has been demon- each mycelium (22 6 4 mg [dry weight] [mean 6 standard deviation for qua-
strated by a chemiluminescence method utilizing phthalic hy- druplicate assays]) was washed once with 30 ml of mineral medium (with vita-
drazide as radical trap (3). mins omitted) before being transferred to the culture vessels of test system A or
B. Dry weight was determined on cellulose acetate membrane filters (pore size,
For our attempts to identify some of the metabolites formed 0.45 mm; Sartorius, Göttingen, Germany) which had been dried to a constant
during the degradation of enrofloxacin, we selected the brown weight at 85°C overnight.
rot fungus G. striatum DSM 9592. This strain did not exhibit Test system A was employed to determine the degradation kinetics of [14C]en-
any peroxidase or laccase activity typical of white rot fungi, rofloxacin. The reaction vessel is described elsewhere (1). Briefly, each 100-ml
screw-capped jar contained 10 ml of sterile mineral medium and a sterile 10-ml
despite having the highest activity in degrading enrofloxacin of glass tube holding 3 ml of 1 M NaOH as a trap for CO2. To prevent volatile
all wood-rotting basidiomycetes studied so far (33, 45). 14
C-labeled metabolites from entering the NaOH, a bore hole in the tube wall
was loosely sealed with a cotton plug coated with 2% (wt/vol) paraffin in hexane.
MATERIALS AND METHODS Each flask received 2.2 6 0.4 mg (dry weight) of mycelium per ml. 14C-labeled
enrofloxacin (8 to 9 kBq) was diluted with unlabeled drug to give a final con-
Organism. G. striatum DSM 9592 degraded enrofloxacin in liquid culture (45) centration of 10 ppm. All experiments were run at room temperature in the dark
and bound to straw (33) but exhibited neither lignin peroxidase, laccase, nor to prevent photodegradation of enrofloxacin.
tyrosinase activity when tested by the methods described by Stalpers (40); ABTS Test system B was utilized for the production of metabolites. Each 250-ml
[2,29-azinobis(3-ethylbenzthiazoline-6-sulfonic acid)], which is used for the de- Erlenmeyer flask, containing 30 ml of mineral medium, was inoculated with
tection of laccase (38), also was not oxidized (43). 2.2 6 0.4 mg (dry weight) of mycelium per ml. Thereafter, 300 mg of enrofloxa-
Culture media and growth conditions. Stock cultures of G. striatum were kept cin, 14C labeled with 448.5, 229.1, or 691.9 kBq (of C-4, C-2, or piperazine label,
on malt extract agar CM 59, pH 5.4 (Oxoid) (Unipath, Wesel, Germany), con- respectively), was added. The vessels were kept at 150 rpm at room temperature
taining (per liter) 15 g of malt extract, 2.5 g of mycological peptone, and 15 g of in the dark. Produced CO2 was adsorbed to 6 g of soda lime placed in a funnel
agar. Cultures on agar plates were incubated at room temperature for at least 7 at the top of the flask. To determine produced 14CO2, the soda lime was acidified
days. To inoculate liquid cultures, small agar pieces were cut out of the agar. with 5 M HCl. Released 14CO2 was transferred into scintillation fluid via a
Precultures in liquid medium were grown in malt extract broth CM 57, pH 5.5 stream of nitrogen (see below).
(Oxoid). Five grams of dry substance was dissolved in 1 liter of deionized water In order to produce the quantity of a metabolite required for structure eluci-
and distributed in portions of 30 ml into 250-ml Erlenmeyer flasks. These were dation by 1H nuclear magnetic resonance (NMR) (approximately 20 mg), 180 ml
sealed with a cotton plug and autoclaved for 10 min at 115°C. of supernatant from 2-day-old cultures containing degraded [4-14C]enrofloxacin

TABLE 1. Nomenclature of metabolites generated by G. striatum from enrofloxacin

Designation Chemical name

F-1......................................................1-Cyclopropyl-7-(4-ethyl-1-piperazinyl)-6-fluoro-3-hydroxy-4-1H-quinolinone
F-2......................................................1-Cyclopropyl-7-(4-ethyl-1-piperazinyl)-1,4-dihydro-6-hydroxy-4-oxo-3-quinolinecarboxylic acid
F-3......................................................1-Cyclopropyl-6-(4-ethyl-1-piperazinyl)-5-fluoro-1H-indole-2,3-dione
F-4a ....................................................1-Cyclopropyl-7-{[2-(ethylamino)ethyl]amino}-6-fluoro-1,4-dihydro-4-oxo-3-quinolinecarboxylic acid
F-5......................................................1-Cyclopropyl-7-(4-ethyl-1-piperazinyl)-1,4-dihydro-6,8(5,6)-dihydroxy-4-oxo-3-quinolinecarboxylic acid
F-6a ....................................................1-Cyclopropyl-7-(4-ethyl-1-piperazinyl)-6-fluoro-1,4-dihydro-8-hydroxy-4-oxo-3-quinolinecarboxylic acid
F-7......................................................1-Cyclopropyl-7-(4-ethyl-1-piperazinyl)-6-fluoro-1,4-dihydro-5,8-dihydroxy-4-oxo-3-quinolinecarboxylic acid
F-8a ....................................................1-Cyclopropyl-6-fluoro-1,4-dihydro-7,8-dihydroxy-4-oxo-3-quinolinecarboxylic acid
F-9a ....................................................7-Amino-1-cyclopropyl-6-fluoro-1,4-dihydro-4-oxo-3-quinolinecarboxylic acid
F-10a ..................................................1-Ethylpiperazine
F-11a ..................................................2-Cyclopropylamino-4-(4-ethyl-1-piperazinyl)-5-fluorobenzoic acid
a
A chemically prepared reference compound is available.
4274 WETZSTEIN ET AL. APPL. ENVIRON. MICROBIOL.

RESULTS

Kinetics of 14CO2 formation from [14C]enrofloxacin. Pre-

cultured mycelia were resuspended in a defined mineral me-
dium containing 10 ppm of enrofloxacin, which was 14C labeled
at either of three different positions (Fig. 1). After 8 weeks,
27.3% 6 1.1% and 18.5% 6 2.1% 14CO2 were produced from
[2-14C]- and [4-14C]enrofloxacin, respectively. Under identical
conditions, only 6.7% 6 0.4% 14CO2 was generated from [pi-
perazine-2,3-14C]enrofloxacin (Fig. 2). The maximum rate of
CO2 formation (observed with [2-14C]enrofloxacin) was 5.8%
6 1.5% per week, which corresponds to a specific activity of
2.6 6 0.7 nmol of CO2 per mg (dry weight) of mycelium per
week. Nonspecific binding of 14C-labeled compounds to myce-
lium was assessed by a series of balances of radioactivity em-
ploying cultures of ages varying from 2 to 42 days. Recoveries
were close to 100% throughout. For example, in 5-day-old
cultures containing [4-14C]enrofloxacin, the amounts of (i) pro-
FIG. 2. Total 14CO2 production from [14C]enrofloxacin by G. striatum. My- duced 14CO2, (ii) activity remaining in the supernatant, and
celia, pregrown in malt medium, were washed and transferred into a defined
mineral medium containing 10 ppm of enrofloxacin labeled either at C-2 (h), at (iii) activity nonspecifically bound to mycelium had reached
C-4 (E), or in the piperazinyl moiety (Ç). Values given are the means 6 standard 4.3% 6 1.7%, 90.6% 6 2.3%, and 4.2% 6 0.4% (mean 6
deviations for five replicate cultures. standard deviations for three cultures), respectively (data not
shown in detail).
The determination of the incubation time resulting in max-
imum concentrations of major metabolites is shown in Fig. 3.
was produced. Each of six replicate cultures was first checked by HPLC to verify After 5 days, enrofloxacin had a residual concentration of only
the expected metabolite pattern. After separation of the mycelia by centrifuga-
tion, combined supernatants were lyophilized, redissolved in 8.5 ml of distilled
9%, and it was hardly detectable from day 7 onwards (not
water, analyzed, and fractionated by HPLC method I (see below). shown). The HPLC elution profiles indicated five major me-
HPLC. Analytical and micropreparative HPLCs were performed on an HPLC tabolites, reaching similar maximum concentrations after 2
system HP 1050 equipped with a diode array detector (Hewlett-Packard, Wald- days. Hence, an incubation time of 2 days was used to produce
bronn, Germany), a Ramona 90 radioactivity flowthrough detector (Raytest,
Straubenhardt, Germany), and a Kromasil 100 C18 column (0.46 by 25 cm;
metabolites for structure elucidation.
particle size, 5 mm; Eka-Chemicals, Bohuds, Sweden). The elution solvent con- Profiles of metabolites generated from enrofloxacin. Cul-
sisted of 0.1 mM ammonium formate in 1% formic acid (component A) and tures of G. striatum, each containing either [2-14C]-, [4-14C]- or
acetonitrile (component B). Routinely, the following gradient was employed [piperazine-2,3-14C]enrofloxacin, were incubated under iden-
(HPLC method I): starting at 100% component A for 2 min, component A was
linearly decreased to 94% over 3 min and then to 85% over 10 min. Thereafter,
tical conditions for 2 days. Supernatants were analyzed by
component A was kept at 85% for 15 min and then decreased to 72% over 5 min HPLC with radioactivity detection. Representative elution
and finally to 0% over 10 min. For micropreparative purification of isolated profiles are shown in Fig. 4. All major metabolites, arbitrarily
metabolites, HPLC method II was used, in which component A contained, in designated F-1, F-2, and F-4 to F-7 (F indicates their fungal
addition, 1% (vol/vol) isopropanol. The flow rate was 1 ml/min.
HPLC-MS. Combined HPLC-electrospray ionization mass spectrometry
origin), were present in all three supernatants, regardless of
(HPLC-MS) was performed on an HPLC system HP 1090 (Hewlett-Packard) the applied label. Obviously, C-2, C-4, and the piperazinyl
equipped with a Kromasil 100 C18 column (0.2 by 25 cm; particle size, 5 mm;
Eka-Chemicals), linked to a TSQ 7000 mass spectrometer (Finnigan-MAT, Bre-
men, Germany), which was operated in the electrospray ionization mode. The
atmospheric pressure ionization source (Finnigan) was controlled by a DECsta-
tion computer system (Digital Equipment Corp., Maynard, Mass.). Applied
gradients were as described above. The effluent of the HPLC column (200
ml/min) was passed on to the MS interface without splitting. The spray needle
voltage was 4.5 kV, and nitrogen at 500 kPa served as the sheath gas. The
capillary was held at 210°C. The electrospray ionization ion source was used in
the positive mode, scanning the range of m/z 150 to 900 in 2.5 s. For the
identification of F-10 (see below), a scan range of m/z 30 to 600 was used.
1
H-NMR spectroscopy. NMR spectra were recorded at 500 MHz on a Bruker
NMR spectrometer AMX 500 (Bruker, Rheinstetten, Germany) in D2O (deu-
teration degree, $99.96%) containing 0.1 or 5% (vol/vol) trifluoroacetic acid-d1.
Chemical shifts are reported in parts per million (d). Acetone was used as an
internal standard (d 5 2.00 ppm).
Liquid scintillation counting. An LS 3801 liquid scintillation counter (Beck-
man, Fullerton, Calif.) was employed for determination of radioactivity at am-
bient temperature. Radioactivities in stock solutions and small-volume samples
were determined in Instant-Scint-Gel Plus (Beckman). The sodium hydroxide
solution of test system A was transferred into 7 ml of Quickszint 401 (Zinser
Analytic, Frankfurt, Germany). Remobilized CO2 was trapped in a mixture of 5.6
ml of Permafluor V and 4.4 ml of Carbosorb (Beckman).
Fenton’s reaction. The reaction mixture described previously (28) was modi-
fied and consisted of 10 ml of 1 mM sodium acetate buffer (pH 4.2) containing
5 mM FeSO4, 0.01% hydrogen peroxide, and 10 ppm of [4-14C]enrofloxacin (141
kBq). After 4 h of incubation at room temperature in the dark with stirring at 100 FIG. 3. Degradation of [4-14C]enrofloxacin and time course of formation of
rpm, the solution was frozen and lyophilized. The resulting dry matter was major metabolites by G. striatum. Mycelia were transferred into mineral medium
redissolved in 0.5 ml of distilled water and analyzed by HPLC and HPLC-MS as containing 10 ppm of enrofloxacin. The supernatant was analyzed by HPLC to
described above. Metabolites formed from enrofloxacin were characterized by determine the pattern of metabolites formed after up to 5 days. Due to the use
retention time, molecular weight, and the characteristic isotope pattern caused of a different HPLC column, the retention times here were slightly smaller than
by the 14C label. those indicated in Table 2.
VOL. 63, 1997 METABOLITES OF ENROFLOXACIN 4275

Identification of metabolites by HPLC cochromatography.

Supernatant containing degraded [4-14C]enrofloxacin was an-
alyzed by HPLC cochromatography with UV detection. Four
metabolites (F-4, F-6, F-8, and F-9) could be identified by
coelution with chemically prepared reference compounds. The
systematic names of all identified metabolites are listed in
Table 1; the corresponding chemical structures are shown in
Fig. 5. HPLC retention times of all metabolites, their pseudo-
molecular ions, and characteristics of their UV absorption
spectrum are given in Table 2. The UV spectra of F-4, F-6, F-8,
and F-9 were identical to those of the respective reference
compounds.
Structure elucidation by 1H-NMR spectroscopy. Fractions
containing major metabolites were isolated from the combined
supernatant preparation and subjected to HPLC-MS and, fi-
nally, NMR analysis.
(i) F-1. After separation of a nonradioactive matrix compo-
nent, the fraction containing F-1 (Table 1) was subjected to
FIG. 4. HPLC analysis of 2-day-old supernatants from cultures of G. striatum
containing metabolites generated from [14C]enrofloxacin. Enrofloxacin had a HPLC-MS. Its mass of 331 Da indicated a loss of 28 mass units
concentration of 10 ppm and was labeled at one of three alternative positions: (Table 2). The NMR spectrum of F-1, shown in Fig. 6, was very
either [4-14C]enrofloxacin (A), [2-14C]enrofloxacin (B), or [piperazine-2,3-14C] similar to that of enrofloxacin (Table 3). All proton signals of
enrofloxacin (C) was used. In trace C, the apparent loss of F-8 is marked by an
arrow; also note the oversized peak of F-10.
the cyclopropyl and the ethylpiperazinyl moieties could be
assigned. The aromatic protons H-5 and H-8 had almost iden-
tical chemical shifts (d) as well as H,F coupling constants
(JH,F). However, the signal of H-2 was shifted upfield (d 5 8.37
moiety all had been retained. The maximum concentrations of ppm instead of 8.65 ppm), indicating the replacement of the
such metabolites fell between 6% (F-5) and 13% (F-6) of the carboxyl group by the electron-rich hydroxyl group, which was
initially applied radioactivity. Metabolite F-8 was missing in in agreement with the determined molecular weight. The pro-
the profile generated from piperazine-labeled enrofloxacin posed structure of F-1 is shown in Fig. 5.
(Fig. 4, trace C). In addition, a broad and tailing, oversized (ii) F-2. The molecular weight of metabolite F-2 (Table 1)
peak appeared in the front of that chromatogram. This fraction was 357 (Table 2). Its NMR spectrum showed only one—but a
comprised a polar compound, designated F-10, which was decisive—difference when compared with enrofloxacin: the sig-
gradually retained by the solid-state scintillator of the detector
nals of H-5 and H-8 did not exhibit H,F coupling, indicating
cell, thereby causing such an artificially enlarged signal. When
the loss of the fluorine atom (Table 3). The molecular mass of
this fraction was collected and analyzed with an LS counter, an
appropriate amount of radioactivity (approximately 13%) was F-2, two mass units smaller than enrofloxacin, was in agree-
found. ment with a hydroxyl group replacing fluorine (Fig. 5).
F-10. HPLC-MS analysis of supernatant containing de- (iii) F-4. Metabolite F-4 (Table 1), identified primarily by
graded [piperazine-2,3-14C]enrofloxacin directly permitted the retention time and UV absorption spectrum, had a molecular
identification of F-10 (1-ethylpiperazine) by its pseudomolecu- weight of 333 (Table 2). Even after further purification, the
lar ion (M 1 H)1 at m/z 115 and by its specific ion pattern (m/z NMR spectrum indicated the presence of matrix components
115/117/119 5 10:1:2), caused by the presence of either one or superimposed on the aliphatic region. The aromatic region,
two 14C-labeled carbon atoms in the piperazinyl moiety (Table however, provided the signals of protons H-2, H-5, and H-8;
2). A corresponding pattern found in [piperazine-2,3-14C]en- these were identical in chemical shift and multiplicity to the
rofloxacin at m/z 360, 362, and 364 proved that F-10 was a signals of the reference compound (data not shown). Further
specific derivative thereof. attempts to purify F-4 failed, because of its lability.

TABLE 2. HPLC retention times, pseudomolecular ions in MS, and characteristics of the UV absorption spectra of
[4-14C]enrofloxacin and metabolites
Value for:
Characteristic
Enrofloxacin F-1 F-2 F-3 F-4 F-5 F-6 F-7 F-8 F-9 F-10c F-11
a b
Retention time (min) 27.7 18.1 19.5 — 22.4 24.6 26.7 36.7 37.2 38.5 2.5 28.4e
Pseudomolecular iond 360 332 358 318 334 374 376 392 280 263 115 308
(M 1 H)1) (m/z)
UV absorption (nm)
lmax (major) 278 283 278 —b 276 288 245 247 272 271 250e
lmax (minor) 316, 328 346 339 260 330 353 250, 327 278e, 352e
Shoulder 335 338 271 264
a
Determined by HPLC method I, as described in Materials and Methods.
b
—, not detectable by HPLC with either UV or radioactivity detection.
c
Data are derived from an experiment with [piperazine-2,3-14C]enrofloxacin.
d
(M 1 H)1 was accompanied by (M 1 2 1 H)1 resulting from the 14C label; in addition, (M 1 4 1 H)1 was present in F-10.
e
Generated from the reference substance.
4276 WETZSTEIN ET AL. APPL. ENVIRON. MICROBIOL.

FIG. 5. Proposed principle routes of degradation of enrofloxacin (routes A to D) employed by the brown rot fungus G. striatum. Monohydroxylated metabolites
(F-1, F-2, and F-3) may result from an initial attack of enrofloxacin by hydroxyl radicals at different sites of the molecule. Secondary and subsequent attacks at one of
various possible sites in each of these primary metabolites may cause branching of the routes, resulting in a network of metabolites. However, only the most stable
intermediates could be isolated and characterized.

(iv) F-5. Metabolite F-5 (Table 1) had a molecular weight of After purification, i.e., separation from enrofloxacin, F-6 was
373 (Table 2). This could have resulted from an elimination of analyzed by NMR. The spectrum contained all signals typical
fluorine, combined with a twofold hydroxylation. Due to matrix of the ethylpiperazinyl and the cyclopropyl groups (Table 3).
components, the aliphatic region of its NMR spectrum could Only two protons were detected in the aromatic region, a
not be analyzed. However, in the aromatic region two signals singlet assigned to H-2 and a doublet specifically assigned to
were detected. A singlet at 8.58 ppm could be specifically H-5 due to ortho coupling with fluorine (Table 3). Since the
assigned to H-2. Another singlet at 7.02 ppm (absence of H,F molecular mass was increased by 16 Da, F-6 was identified
coupling) indicated the loss of fluorine. Because of the absence as 8-hydroxyenrofloxacin (Fig. 5). The reference compound
of a further aromatic proton signal, and based on the molec- showed an identical NMR spectrum (not shown).
ular weight, two structural alternatives have to be envisaged: a (vi) F-7. The molecular weight of 391 determined for F-7
5,6- or a 6,8-dihydroxylated congener of enrofloxacin (Fig. 5). suggested a twofold hydroxylation (Tables 1 and 2). The NMR
F-5 was rapidly decomposed. spectrum of F-7 revealed only one singlet at 8.77 ppm, typical
(v) F-6. Identified primarily by cochromatography, metabo- of H-2. Hence, the two hydroxyl groups could have been in-
lite F-6 (Table 1) had a molecular weight of 375 (Table 2). troduced only at C-5 and C-8 (Fig. 5). Matrix compounds
VOL. 63, 1997 METABOLITES OF ENROFLOXACIN 4277

FIG. 6. The 500-MHz 1H-NMR spectrum (including expanded areas) of the decarboxylated congener of enrofloxacin (F-1) formed by G. striatum.

masked the aliphatic region and prevented further NMR H2O)1], which were also observed with reference compound.
assignments. F-7 was already decomposed during purifica- The NMR spectrum showed that the 1-ethylpiperazine moiety
tion. had been cleaved off (compare with Fig. 4, trace C). Only two
(vii) F-8. Metabolite F-8 (Table 1) was first identified by aromatic protons were present, H-5 (d, d 5 7.57 ppm, JH,F 5
cochromatography. Its molecular weight of 279 (Table 2) was 10.1 Hz) and H-2 (s, d 5 8.82 ppm). Two 2H multiplets (at 0.99
indicated by ions at m/z 280 and 262 [(M 1 H)1 and (M 1 H 2 and 1.10 ppm) and a 1H multiplet (at 4.18 ppm) were charac-

TABLE 3. 1H-NMR data for enrofloxacin and monohydroxylated metabolitesa

Enrofloxacin F-1 F-2 F-6
Proton(s)b d Multi- J d Multi- J d Multi- J d Multi- J
Integral Integral Integral Integral
(ppm) plicity (Hz) (ppm) plicity (Hz) (ppm) plicity (Hz) (ppm) plicity (Hz)

CH2 (cyclopropyl) 1.01 2 m 1.06 2 m 1.07 2 m 0.98 2 m

CH3 (ethyl) 1.17 3 t 7.3 1.17 3 t 7.3 1.18 3 t 7.3 1.17 3 t 7.4
CH2 (cyclopropyl) 1.21 2 m 1.27 2 m 1.26 2 m 1.08 2 m
CH2 (ethyl) 3.12 2 q 7.3 3.12 2 q 7.3 3.12 2 q 7.3 3.11 2 q 7.4
CH2 (piperazine, H-b) 3.15 4 m 3.15 4 m 3.10 4 m '3.2 4 m
CH2 (piperazine, H-a2) 3.54 2 m 3.54 2 m 3.55 2 m 3.42 2 m
CH (cyclopropyl) 3.58 1 m '3.8 1 m 3.74 1 m 4.20 1 m
CH2 (piperazine, H-a1) 3.79 2 m 3.79 2 m 3.90 2 m 3.49 2 m
H-8 7.45 1 d 7.2c 7.54 1 d 7.5c 7.54 1 s
H-5 7.66 1 d 13.0c 7.85 1 d 12.8c 7.52 1 s 7.41 1 d 11.5c
H-2 8.65 1 s 8.37 1 s 8.84 1 s 8.76 1 s
a
Spectra were determined at 500 MHz in a deuterated solvent (0.1% trifluoroacetic acid-d1 in D2O for enrofloxacin and 5% trifluoroacetic acid-d1 in D2O for F-1,
F-2, and F-6). Acetone was used as internal standard (d 5 2.00 ppm). Multiplicities are abbreviated as follows: s, singlet; d, doublet; t, triplet; q, quadruplet; m, multiplet.
b
Positioning is as shown in Fig. 1.
c
JH,F.
4278 WETZSTEIN ET AL. APPL. ENVIRON. MICROBIOL.

teristic of the cyclopropyl moiety. The structure of F-8 is shown

in Fig. 5.
Identification of minor metabolites by HPLC-MS. HPLC-
MS analysis revealed three additional minor metabolites,
which could not be unequivocally assigned in HPLC elution
profiles by either UV or radioactivity detection because of
their very low concentrations.
(i) F-3. Metabolite F-3 (Table 1), with a molecular weight of
317 (Table 2), showed in mass spectra the characteristic ion
patterns generated from supernatants containing [4-14C]enro-
floxacin (m/z 318 and 320) or [piperazine-2,3-14C]enrofloxacin
(m/z 318, 320, and 322). The specific signal at m/z 320 was
absent when [2-14C]enrofloxacin was employed as a substrate,
indicating that C-2 of enrofloxacin was eliminated in F-3.
Based on these observations, an isatin-type structure was pro-
posed for F-3 (Fig. 5).
(ii) F-9. Metabolite F-9 (Table 1), initially identified by co- FIG. 7. Comparison of HPLC elution profiles showing metabolites generated
chromatography, had a molecular weight of 262 (Table 2). In by G. striatum from [4-14C]enrofloxacin (A) and metabolites produced by chem-
ical degradation of [4-14C]enrofloxacin with Fenton’s reagent (B). The metabo-
order to further confirm its structure, the supernatant was lites in trace B were identified by retention time, molecular weight, and ion
spiked by adding approximately the same concentration of pattern in MS.
unlabeled reference compound. Analysis by HPLC-MS re-
vealed the expected isotope dilution effect: the height of the
signal at m/z 265 (caused by [4-14C]enrofloxacin) was reduced
to about 50%. The proposed structure of F-9 is given in Fig. 5. DISCUSSION
(iii) F-11. The identification of metabolite F-11 (Table 1;
Fig. 5) had to be based on an analysis of the reference com- In a previous paper, we described biodegradation of enro-
pound, which had a molecular weight of 307 (Table 2); it floxacin adsorbed to straw. Three strains of the brown rot
proved to be quite labile in aqueous solution. When HPLC-MS fungus G. striatum were found to possess a higher activity than
was done in the single-ion monitoring mode at m/z 308, a peak four species of white rot fungi, including P. chrysosporium (33).
was detected in the supernatant at the retention time of the Hence, we selected G. striatum for our attempts to identify the
metabolites generated in this process. Because G. striatum did
reference compound (Table 2). To further assess its specificity,
not exhibit peroxidase or laccase activity, which is typically
a supernatant sample was spiked with reference compound. As
found in white rot fungi (43), another degradation mechanism
expected, an enlarged peak was observed in the single-ion could be expected to be operative, possibly one of the Fenton
trace at m/z 308, indicating the presence of F-11 in the super- type as first proposed by Koenigs for the decay of wood by
natant of G. striatum. brown rot basidiomycetes (30). Experimental conditions under
Metabolites at trace concentrations. Approximately 10 ad- which the lignocellulose-depolymerizing activity of a brown rot
ditional metabolites of enrofloxacin were detected by HPLC- fungus could be induced in liquid culture have not been de-
MS. Most had retained all 14C labels (C-2, C-4, and piperazine) scribed before to our knowledge (29). However, we had ini-
and showed the specific patterns of pseudomolecular ions. Five tially observed that after the formation of biomass in malt
examples of such molecular weights and hypothetical struc- broth, mycelia of G. striatum could be transferred into a de-
tures should be mentioned (compare with Fig. 5): a molecular fined mineral medium, lacking a carbon, nitrogen, and phos-
weight of 305 would be in agreement with an oxidatively de- phorus source, to enhance the expression of the degradation
carboxylated F-4, a mass of 329 Da could result from oxida- potential for enrofloxacin, which otherwise remained undetect-
tively decarboxylated F-2, a 347-molecular-weight metabolite able, even in diluted malt medium (43).
may have been derived from F-1 by hydroxylation at C-8 or Elimination kinetics of specific C atoms of enrofloxacin. At
C-5, a 349-molecular-weight metabolite could have been gen- first, the degradation process was characterized by the forma-
erated from F-4 hydroxylated at C-8 or C-5, and a metabolite tion of 14CO2 from [14C]enrofloxacin labeled either at position
with a molecular weight of 389 might have been generated C-2, at position C-4, or in the piperazine ring. Because both the
from metabolite F-2 by twofold hydroxylation at C-5 and C-8 rate and extent of 14CO2 production from [2-14C]enrofloxacin
or, alternatively, by oxidation of F-7, giving rise to the respec- were higher than those of 14CO2 production from [4-14C]en-
tive quinone. In fact, two metabolites with a molecular weight rofloxacin (Fig. 2), C-2 was most likely eliminated prior to C-4,
which would be in accordance with degradation route A shown
of 389 were observed at distinct retention times. However, due
in Fig. 5. The high initial rates of 14CO2 formation could not be
to the extremely low concentrations of these metabolites,
maintained for incubation times exceeding 3 weeks. This was
which probably reflect their lability, no additional information due either to the lack of nutrients or to unknown factors
could be obtained. limiting G. striatum in this closed in vitro test system. It is
Fenton’s reaction. [4-14C]enrofloxacin was degraded by Fen- remarkable that from [piperazine-2,3-14C]enrofloxacin only
ton’s reaction as described in Materials and Methods. A typical about one-quarter of the amount of 14CO2 was produced un-
HPLC elution profile is shown in Fig. 7. Five metabolites were der identical conditions. Obviously, the core part of the quin-
identified by HPLC-MS, namely, three monohydroxylated con- olone molecule was the preferred target of the degradation
geners of enrofloxacin (F-1, F-2, and F-6) and two metabolites activity expressed by G. striatum. As only about 5% of the 14C
(F-4 and F-9) indicating degradation of the piperazinyl sub- label was nonspecifically bound to mycelium, the bulk of ra-
stituent. The systematic names and analytical characteristics of dioactivity, i.e., that in enrofloxacin and its metabolites, re-
identified metabolites are given in Tables 1 and 2. mained fully bioavailable in the supernatant.
VOL. 63, 1997 METABOLITES OF ENROFLOXACIN 4279

Metabolites generated from enrofloxacin. HPLC elution been studied in detail. However, degradation of enrofloxacin
profiles of supernatants from up-to-5-day-old cultures of G. was very sensitive to hydroxyl-radical-scavenging agents like
striatum indicated that even the most abundant metabolites ethanol and dimethyl sulfoxide (19, 43). Hence, our results
appeared only transiently. Already 2 days after the transfer may represent new evidence in support of the hypothesis that
from malt into mineral medium, the five major metabolites brown rot fungi are able to catalyze the formation of hydroxyl
were reaching maximum concentrations (Fig. 3). When all radicals in a Fenton-type reaction. Under our experimental
three differently labeled enrofloxacin molecules were simulta- conditions, this mechanism might have been employed to de-
neously employed in parallel cultures, almost identical patterns grade enrofloxacin. The proposed degradation routes, shown
of metabolites were obtained (Fig. 4). Such profiles were com- in Fig. 5, may reflect different sites of initial attack of enro-
posed of up to 25 peaks, as detected by HPLC-MS. Individual floxacin by hydroxyl radicals. Each of the primary metabolites
fractions were purified and analyzed by combining HPLC, offers various sites for a secondary attack, which would cause
HPLC-MS, and 1H-NMR spectroscopy. The proposed molec- branching of the routes, resulting in the formation of a network
ular structures of 3 of 11 metabolites, F-1, F-2, and F-6 (chem- of metabolites; this could explain the high number of metab-
ical names are given in Table 1), could be confirmed by com- olites observed by HPLC-MS. The molecular weights of some
plete 1H-NMR spectra. Four structures (those of F-4, F-5, F-7, unidentified metabolites, detectable only by HPLC-MS, would
and F-8) were ascertained by an analysis of the aromatic re- be consistent with such a network.
gions of their NMR spectra, while metabolites F-4, F-8, and Enrofloxacin was also chemically degraded by Fenton’s re-
F-9 could also be identified by cochromatography with chem- action, in which hydroxyl radicals are generated from hydrogen
ically prepared reference compounds and by their UV absorp- peroxide and Fe21 (50). The identification of all three charac-
tion spectra. The identification of metabolites F-3, F-10, and teristic monohydroxylated congeners of enrofloxacin as well as
F-11 had to be based on MS techniques. of the two typical intermediates indicating the degradation of
The identified metabolites were assigned to four groups of the piperazinyl moiety supports the proposed role of hydroxyl
compounds: (i) monohydroxylated congeners of enrofloxacin radicals in G. striatum. The metabolic scheme outlined in Fig.
(F-1, F-2, and F-6), (ii) dihydroxylated congeners of enrofloxa- 5 would be in agreement with the pattern of metabolites gen-
cin (F-5, F-7, and F-8), (iii) an isatin-type (F-3) and an anthra- erated by Fenton’s reaction, although it was not possible to
nilic acid (F-11) derivative of enrofloxacin, and (iv) desethyl- isolate dihydroxylated metabolites from such reaction mix-
ene-enrofloxacin (F-4), the 7-amino congener of enrofloxacin tures.
(F-9), and 1-ethylpiperazine (F-10). Preliminary experiments in our laboratory have shown that
These groups of metabolites led us to propose the metabolic G. striatum gives rise to a pronounced chemiluminescence sig-
scheme shown in Fig. 5. It consists of four principle degrada- nal under the assay conditions described by Backa and col-
tion routes, which, due to the similar concentrations of all leagues (3), which has been proposed to directly indicate the
major metabolites, might be simultaneously employed. Route formation of hydroxyl radicals in wood shavings. Degradation
A would be initiated by an oxidative decarboxylation. This of lignocellulose and pollutants by reactions involving free
irreversibly inactivates the drug, because the carboxyl group is radicals occurs extracellularly (4, 15, 25). However, the site of
essential for antibacterial activity of FQs (12). An isatin-type enrofloxacin degradation in G. striatum is not known at
intermediate (F-3) as well as the anthranilic acid derivative present.
(F-11) indicate cleavage of the heterocyclic core of enrofloxa- In addition to the reactions found in wood-rotting basidio-
cin as well as the successive loss of C-2 and C-3, while C-4 was mycetes, other mechanisms are involved in aerobic degrada-
retained. Isatin is a known intermediate in degradation path- tion of N-heterocyclic compounds in bacteria and fungi (2, 26,
ways of indole and tryptophan (2, 39). Liberation of 14CO2 39). Key steps in the degradation of compounds containing
from [4-14C]enrofloxacin (as shown in Fig. 2) involves F-11 and structural elements also found in enrofloxacin, e.g., 3-carboxy-
probably metabolites from the other degradation routes as pyridine (nicotinic acid), 3,4-dihydroxypyridine, quinoline, 1H-
well. The degradation of anthranilic acid is a well-documented 4-oxoquinoline, and anthranilic acid, are catalyzed mostly by
natural process (2, 39). molybdenum-containing dehydrogenases, dioxygenases, or mo-
Route B would be initiated by defluorination of enrofloxa- nooxygenases. Bauer and coworkers reported the degradation
cin. This eliminates the xenobiotic structural element and re- of 1H-3-hydroxy-4-oxoquinoline by Pseudomonas putida even
duces the antibacterial potential of metabolite F-2 (in terms of via 2,4-dioxygenation, with concomitant elimination of carbon
increased MICs) to #3% (12, 43). Route C could be initiated monoxide (5). However, FQs might not be readily accessible
by hydroxylation of enrofloxacin at position C-8. This modifi- for such enzymes. Due to the fluorine substituent and the
cation reduces the antibacterial potential of F-6 to #5% (43) carbonyl as well as the carboxyl group, the heterocyclic core of
and most likely enhances its further degradation. Dihydroxy- enrofloxacin tends to be electron deficient. In addition, the
lated congeners are included in routes B and C. Such autoxi- high degree of substitution might prevent an enzymatic attack
dizable structures are prone to undergo further oxidative deg- due to steric hindrance. Thus, the radical-based mechanisms,
radation, which might even include the cleavage of the potentially employed by wood-rotting basidiomycetes, may
homoaromatic part of enrofloxacin. Metabolites F-5, F-7, and provide the most suitable, if not the only, means to initiate
F-8 were found to be decomposed already during purification, degradation of such complex compounds. The ability of G.
even in the frozen state. striatum to cleave the core structure of enrofloxacin also con-
Route D shows an oxidative degradation of the piperazinyl tradicts the reported apparent inability of brown rot fungi to
moiety. This sequence of reactions is apparently initiated by degrade the aromatic components of lignin (15). Moreover,
the formation of a carbonyl group, as was shown for cipro- the reported growth of the brown rot fungus Lentinus lepideus
floxacin (27). Degradation of F-4 should result in the forma- on methoxylated aromatic compounds as sole sources of car-
tion of F-9, the 7-amino congener of enrofloxacin, which has a bon is physiological evidence implying ring cleavage (11).
residual antibacterial activity on the order of #3%, as com- Clearly, much more work is required to elucidate the reaction
pared to the parent drug (43). mechanism(s) utilized by brown rot fungi and to reveal their
Proposed degradation mechanism. So far, the molecular true potential in the degradation of lignocellulose and xenobi-
mechanism of enrofloxacin degradation by G. striatum has not otics.
4280 WETZSTEIN ET AL. APPL. ENVIRON. MICROBIOL.

Ecological perspectives. In practice, animals undergoing an- 11. Collett, O. 1992. Aromatic compounds as growth substrates for isolates of
tibacterial therapy will excrete a substantial fraction of an the brown-rot fungus Lentinus lepideus (Fr. ex. Fr.) Fr. Mater. Org. 27:67–77.
12. Domagala, J. M. 1994. Structure-activity and structure-side-effect relation-
applied FQ as intact drug bound to feces; glucuronidation and ship for the quinolone antibacterials. J. Antimicrob. Chemother. 33:685–706.
sulfation of FQs cause only transient inactivation. Metabolites, 13. Edlund, C., L. Lindqvist, and C. E. Nord. 1988. Norfloxacin binds to human
characterized by an oxidized piperazine moiety, have been fecal material. Antimicrob. Agents Chemother. 32:1869–1874.
detected at small concentrations in urine and feces (27, 42). 14. Eggert, C., U. Temp, J. F. D. Dean, and K.-E. L. Eriksson. 1996. A fungal
metabolite mediates degradation of non-phenolic lignin structures and syn-
Because nonspecific binding restricts the mobility of FQs in the thetic lignin by laccase. FEBS Lett. 391:144–148.
environment (13, 31, 34), manure (e.g., cattle dung) could 15. Evans, C. S., M. V. Dutton, F. Guillen, and R. G. Veness. 1994. Enzymes and
serve as a relevant model ecosystem to study the fate of drugs, small molecular mass agents involved with lignocellulose degradation. FEMS
the more so because at least some of the known coprophilous Microbiol. Rev. 13:235–240.
16. Freer, S. N., and R. W. Detroy. 1982. Biological delignification of 14C-labeled
basidiomycetes might also be able to degrade enrofloxacin. lignocelluloses by basidiomycetes: degradation and solubilization of the lig-
Two such basidiomycetes were isolated from aged cattle nin and cellulose components. Mycologia 74:943–951.
dung by Wicklow and colleagues 2 decades ago (47, 48). These 17. Frost, A. J. 1991. Antibiotics and animal production, p. 181–194. In J. B.
fungi appeared late on cattle dung, after various ascomycetes Woolcock (ed.), Microbiology of animals and animal products. Elsevier,
Amsterdam, The Netherlands.
had consumed the more easily utilizable substrates. Both iso- 18. Greene, C. E., and S. C. Budsberg. 1993. Veterinary use of quinolones, p.
lates, strain NRRL 6464 and a strain identified as Cyathus 473–488. In D. C. Hooper and J. S. Wolfson (ed.), Quinolone antimicrobial
stercoreus, physiologically resembled wood-rotting fungi in agents, 2nd ed. American Society for Microbiology, Washington, D.C.
showing a high activity in the degradation of lignocellulose in 19. Halliwell, B., and J. M. Gutterridge. 1989. Free radicals in biology and
medicine, 2nd ed. Oxford University Press, Oxford, United Kingdom.
vitro (16, 48). Preliminary results from our laboratory indicate 20. Harper, D. B., and D. O’Hagan. 1994. The fluorinated natural products. Nat.
that C. stercoreus is able to degrade enrofloxacin (43). There- Products Rep. 11:123–133.
fore, under field conditions, the degradation of enrofloxacin 21. Hektoen, H., J. A. Berge, V. Hormazabal, and M. Yndestad. 1995. Persis-
might, in principle, proceed at the rate of decomposition of tence of antibacterial agents in marine sediments. Aquaculture 133:175–184.
22. Hirano, T., H. Tanaka, and A. Enoki. 1995. Extracellular substance from the
undigested plant polymers excreted by livestock animals or at brown-rot basidiomycete Tyromyces palustris that reduces molecular oxygen
that for the surrounding humus matrix. to hydroxyl radicals and ferric iron to ferrous iron. Mokuzai Gakkaishi
Wood-rotting fungi, in degrading lignocellulose and other 41:334–341.
substances associated with such matrices, play an essential role 23. Holzman, D. 1995. Refractory chemical residues in soils may be safe. ASM
News 61:445–446.
in the global carbon cycle (6, 32), but very little is known about 24. Hyde, S. M., and P. M. Wood. 1996. Kinetic and antigenic similarities for
homologous taxa in agricultural soils and plant litter (41). So cellobiose dehydrogenase from the brown rot fungus Coniophora puteana
far, G. striatum has served as a model organism in proving that and the white rot fungus Phanerochaete chrysosporium. FEMS Microbiol.
enrofloxacin (and, proven only very recently, ciprofloxacin Lett. 145:439–444.
25. Joseleau, J.-P., S. Gharibian, J. Comtat, A. Lefebvre, and K. Ruel. 1994.
[46]) is biodegradable. However, other, perhaps more appro- Indirect involvement of ligninolytic enzyme systems in cell wall degradation.
priate, fungal species naturally present in dung and other rel- FEMS Microbiol. Rev. 13:255–264.
evant agricultural sites may have to be included to fully de- 26. Kaiser, J.-P., Y. Feng, and J.-M. Bollag. 1996. Microbial metabolism of
scribe the natural degradation potential for FQs. pyridine, quinoline, acridine, and their derivatives under aerobic and anaer-
obic conditions. Microbiol. Rev. 60:483–498.
27. Karabalut, N., and G. L. Drusano. 1993. Pharmacokinetics of the quinolone
ACKNOWLEDGMENTS antimicrobial agents, p. 195–223. In D. C. Hooper and J. S. Wolfson (ed.),
Quinolone antimicrobial agents, 2nd ed. American Society for Microbiology,
We thank our colleagues at various chemistry departments at Bayer, Washington, D.C.
namely, W. Hallenbach, R. Koch, R. Körmeling, U. Petersen, H. Rast, 28. Kirk, T. K., R. Ibach, M. D. Mozuch, A. H. Conner, and T. L. Highley. 1991.
and R. Thomas, for providing the labeled and unlabeled standard Characteristics of cotton cellulose depolymerized by a brown-rot fungus, by
compounds. Furthermore, we thank S. Ochtrop, A. Gerhardt, and acid, or by chemical oxidants. Holzforschung 45:239–244.
J. Schneider, for excellent technical assistance. 29. Kleman-Leyer, K., E. Agosin, A. H. Conner, and T. K. Kirk. 1992. Changes
in molecular size distribution of cellulose during attack by white rot and
brown rot fungi. Appl. Environ. Microbiol. 58:1266–1270.
REFERENCES 30. Koenigs, J. W. 1974. Hydrogen peroxide and iron: a proposed system for
1. Alef, K. 1991. Methodenhandbuch Bodenmikrobiologie, p. 89. Ecomed Ver- decomposition of wood by brown-rot basidiomycetes. Wood Fiber 6:66–80.
lagsgesellschaft mbH, Landsberg, Germany. 31. Marengo, J. R., R. A. Kok, K. O’Brien, R. R. Velagaletti, and J. M. Stamm.
2. Andreoni, V., G. Baggi, and S. Bernasconi. 1995. Microbial degradation of 1997. Aerobic biodegradation of (14C)-sarafloxacin hydrochloride in soil.
nitrogenous xenobiotics of environmental concern. Prog. Ind. Microbiol. Environ. Toxicol. Chem. 16:462–471.
32:1–36. 32. Markham, P., and M. J. Bazin. 1991. Decomposition of cellulose by fungi,
3. Backa, S., J. Gierer, T. Reitberger, and T. Nilsson. 1992. Hydroxyl radical p. 379–424. In D. K. Arora, B. Rai, K. G. Mukerji, and G. R. Knudsen (ed.),
activity in brown-rot fungi studied by a new chemiluminescence method. Handbook of applied mycology, vol. 1. Marcel Dekker, Inc., New York, N.Y.
Holzforschung 46:61–67. 33. Martens, R., H.-G. Wetzstein, F. Zadrazil, M. Capelari, P. Hoffmann, and N.
4. Barr, D. P., and S. D. Aust. 1994. Pollutant degradation by white rot fungi, Schmeer. 1996. Degradation of the fluoroquinolone enrofloxacin by wood-
p. 49–72. In G. W. Ware (ed.), Reviews of environmental contamination and rotting fungi. Appl. Environ. Microbiol. 62:4206–4209.
toxicology. Springer Verlag, New York, N.Y. 34. Nowara, A., J. Burhenne, and M. Spiteller. 1997. Binding of fluoroquinolone
5. Bauer, I., A. de Beyer, B. Tshisuaka, S. Fetzner, and F. Lingens. 1994. A carboxylic acid derivatives to clay. J. Agric. Food Chem. 45:1459–1463.
novel type of oxygenolytic ring cleavage: 2,4-oxygenation and decarbonyla- 35. Orth, A. B., and M. Tien. 1995. Biotechnology of lignin degradation, p.
tion of 1H-3-hydroxy-4-oxoquinaldine and 1H-3-hydroxy-4-oxoquinoline. 287–302. In K. Esser and P. A. Lemke (ed.), The mycota, vol. II. Genetics
FEMS Microbiol. Lett. 117:299–304. and biotechnology. Springer Verlag, Berlin, Germany.
6. Boominathan, K., and C. A. Reddy. 1991. Fungal degradation of lignin: 36. Ribbons, D. W., A. E. G. Cass, J. T. Rossiter, S. J. C. Taylor, M. P. Wood-
biotechnological applications, p. 763–822. In D. K. Arora, B. Rai, K. G. land, D. A. Widdowson, S. R. Williams, P. B. Baker, and R. E. Martin. 1987.
Mukerji, and G. R. Knudsen (ed.), Handbook of applied mycology, vol. 4. Biotransformation of fluoroaromatic compounds. J. Fluorine Chem. 37:299–
Marcel Dekker, Inc., New York, N.Y. 326.
7. Booth, D. M. 1994. Enrofloxacin revisited. Vet. Med. 89:744–753. 37. Ritschkoff, A.-C., M. Rättö, J. Buchert, and L. Viikari. 1995. Effect of carbon
8. Bumpus, J. A. 1993. White rot fungi and their potential use in soil bioreme- source on the production of oxalic acid and hydrogen peroxide by brown-rot
diation processes. Soil Biochem. 8:65–100. fungus Poria placenta. J. Biotechnol. 40:179–186.
9. Bumpus, J. A., M. Tien, D. Wright, and S. D. Aust. 1985. Oxidation of 38. Schlosser, D., R. Grey, and W. Fritsche. 1997. Patterns of ligninolytic en-
persistent environmental pollutants by a white rot fungus. Science 228:1434– zymes in Trametes versicolor. Distribution of extra- and intracellular enzyme
1436. activities during cultivation on glucose, wheat straw and beech wood. Appl.
10. Buswell, J. A. 1991. Fungal degradation of lignin, p. 425–480. In D. K. Arora, Microbiol. Biotechnol. 47:412–418.
B. Rai, K. G. Mukerji, and G. R. Knudsen (ed.), Handbook of applied 39. Schwarz, G., and F. Lingens. 1994. Bacterial degradation of N-heterocyclic
mycology, vol. 1. Marcel Dekker, Inc., New York, N.Y. compounds, p. 459–486. In C. Ratledge (ed.), Biochemistry of microbial
VOL. 63, 1997 METABOLITES OF ENROFLOXACIN 4281

degradation. Kluwer Academic Publishers, Dordrecht, The Netherlands. 46. Wetzstein, H.-G., N. Schmeer, A. Dalhoff, and W. Karl. 1997. Degradation of
40. Stalpers, J. A. 1978. Identification of wood-inhabiting fungi in pure culture. the fluoroquinolone ciprofloxacin by the brown rot fungus Gloeophyllum
Studies in mycology, no. 16. Centraalbureau voor Schimmelcultures, Baarn, striatum DSM 9592, p. 519, abstr. Q-388. In Abstracts of the 97th General
The Netherlands. Meeting of the American Society for Microbiology 1997. American Society
41. Thorn, G. R., C. A. Reddy, D. Harris, and E. A. Paul. 1996. Isolation of for Microbiology, Washington, D.C.
saprophytic basidiomycetes from soil. Appl. Environ. Microbiol. 62:4288– 47. Wicklow, D. T. 1992. The coprophilous fungal community: an experimental
4292. system, p. 715–728. In G. C. Carrol and D. T. Wicklow (ed.), The fungal
42. Vancutsem, P. M., J. G. Babish, and W. S. Schwark. 1990. The fluoroquin- community. Its organisation and role in the ecosystem, 2nd ed. Marcel
olone antimicrobials: structure, antimicrobial activity, pharmacokinetics, Dekker, Inc., New York, N.Y.
clinical use in domestic animals and toxicity. Cornell Vet. 80:173–186. 48. Wicklow, D. T., R. W. Detroy, and B. A. Jessee. 1980. Decomposition of
43. Wetzstein, H.-G. Unpublished data. lignocellulose by Cyathus stercoreus (Schw.) de Toni NRRL 6473, a “white
44. Wetzstein, H.-G., and A. de Jong. 1996. In vitro bactericidal activity and rot” fungus from cattle dung. Appl. Environ. Microbiol. 40:169–170.
postantibiotic effect of fluoroquinolones used in veterinary medicine. Com- 49. Wolfson, J. S., and D. C. Hooper. 1989. Fluoroquinolone antimicrobial
pend. Contin. Educ. Pract. Vet. 18(Suppl.):S22–S29. agents. Clin. Microbiol. Rev. 2:378–424.
45. Wetzstein, H.-G., H.-G. Rast, J. P. E. Anderson, J. Köster, and N. Schmeer. 50. Wood, P. M. 1994. Pathways for production of Fenton’s reagent by wood-
1996. In vitro degradation of the fluoroquinolone enrofloxacin by two species rotting fungi. FEMS Microbiol. Rev. 13:313–320.
of wood-rot fungi, p. 414, abstr. Q-168. In Abstracts of the 96th General 51. World Health Organization. 1995. Evaluation of certain veterinary drug
Meeting of the American Society for Microbiology 1996. American Society residues in food. Forty-third report of the Joint FAO/WHO Expert Com-
for Microbiology, Washington, D.C. mittee on Food Additives. WHO Tech. Rep. Ser. 855:17–24.