Biotransformation of benzydamine by microsomes and precision-cut slices prepared from cattle liver

Article abstract of DOI:10.1080/00498250110085827

1. Benzydamine (BZ), a non-steroidal anti-inflammatory drug used in human and veterinary medicine, is not licensed for use in food-producing species. Biotransformation of BZ in cattle has not been reported previously and is investigated here using liver microsomes and precision-cut liver slices. 2. BZ was metabolized by cattle liver microsomes to benzydamine N-oxide (BZ-NO) and monodesmethyl-BZ (Nor-BZ). Both reactions followed Michaelis-Menten kinetics (K_m = 76.4 ± 16.0 and 58.9 ± 6.4 μM, V_max = 6.5 ± 0.8 and 7.4 ?? 0.5 nmol mg^-1 min^-1, respectively); sensitivity to heat and pH suggested that the N-oxidation is catalysed by the flavin-containing monooxygenases. 3. BZ-NO and Nor-BZ were the most abundant products derived from liver slice incubations, and nine other BZ metabolites were found and tentatively identified by LC-MS. Desbenzylated and hydroxylated BZ-NO analogues and a hydroxylated product of BZ were detected, which have been reported in other species. Product ion mass spectra of other metabolites, which are described here for the first time, indicated the formation of a BZ N⁺-glucuronide and five hydroxylated and N⁺-glucuronidated derivatives of BZ, BZ-NO and Nor-BZ. 4. The results indicate that BZ is extensively metabolized in cattle. Clearly, differences in metabolism compared with, for example, rat and human, will need to be considered in the event of submission for marketing authorization for use in food animals.

Full text of DOI:10.1080/00498250110085827

xenobiotica, 2002, vol. 32, no. 1, 73±86
Biotransformation of benzydamine by
microsomes and precision-cut slices prepared
from cattle liver
A. SANTI , P. ANFOSSI *, N. G. COLDHAM ,
{
{
}
F. CAPOLONGO , M. J. SAUER and C. MONTESISSA
{
{
}
{ Istituto di Patologia e Igiene veterinaria, Universita di Padova, Agripolis, I-35020
Legnaro (PD), Italy
{ Dipartimento di Sanita Pubblica Veterinaria e Patologia Animale, Universita di
Bologna, Ozzano Emilia, Bologna, Italy
} Department of Risk Research, Veterinary Laboratories Agency, New Haw, Addlestone,
Surrey KT15 3NB, UK
Received 14 May 2001
1. Benzydamine (BZ), a non-steroidal anti-in¯ammatory drug used in human and
veterinary medicine, is not licensed for use in food-producing species. Biotransformation
of BZ in cattle has not been reported previously and is investigated here using liver
microsomes and precision-cut liver slices.
2. BZ was metabolized by cattle liver microsomes to benzydamine N-oxide (BZ-NO)
and monodesmethyl-BZ (Nor-BZ). Both reactions followed Michaelis±Menten kinetics
(Km ˆ 76:4 § 16:0 and 58:9 § 6:4 mm, Vmax ˆ 6:5 § 0:8 and 7:4 § 0:5 nmol mg¡1 min¡1
,
respectively); sensitivity to heat and pH suggested that the N-oxidation is catalysed by the
¯avin-containing monooxygenases.
3. BZ-NO and Nor-BZ were the most abundant products derived from liver slice
incubations, and nine other BZ metabolites were found and tentatively identi®ed by LC-
MS. Desbenzylated and hydroxylated BZ-NO analogues and a hydroxylated product of
BZ were detected, which have been reported in other species. Product ion mass spectra of
other metabolites, which are described here for the ®rst time, indicated the formation of a
BZ N‡-glucuronide and ®ve hydroxylated and N‡-glucuronidated derivatives of BZ, BZ-
NO and Nor-BZ.
4. The results indicate that BZ is extensively metabolized in cattle. Clearly,
di€erences in metabolism compared with, for example, rat and human, will need to be
considered in the event of submission for marketing authorization for use in food animals.
Introduction
Benzydamine (1-benzyl-3-[3-(dimethylamino)propoxy]-1H-indazole; BZ) is a
non-steroidal analgesic anti-in¯ammatory drug used in humans by topical applica-
tion. Although BZ has been used in veterinary medicine in cattle and pig
(intramuscular administration) for the treatment of in¯ammatory conditions of
the respiratory, gastrointestinal, genital and urinary systems and in horse by
topical application for soft tissue injuries, it is not currently licensed in the EU
for applications in food-producing species.
A wide range of BZ metabolites has been identi®ed in rat, rabbit and man
(Kataoka et al. 1971, 1973, Koppel and Tenczer 1985, Baldock et al. 1991). In rat,
benzydamine N-oxide (BZ-NO) and glucuronide metabolites were found in urine
* Author for correspondence; e-mail: anfossi@vet.unibo.it
#
Xenobiotica ISSN 0049±8254 print/ISSN 1366±5928 online
2002 Taylor & Francis Ltd
http://www.tandf.co.uk/journals
DOI: 10.1080/00498250110085827

74
A. Santi et al.
and bile, whereas in rabbit the major metabolites were BZ-NO, a hydroxylbenzy-
damine derivative (BZ-OH) and its acetyl-, glucuronide and sulphate conjugates.
In man, the major metabolites excreted in urine were BZ-NO, monodesmethyl-BZ
(Nor-BZ), BZ-OH, the dedimethylaminopropyl-BZ, the desbenzyl-BZ and sev-
eral glucuronides. Despite the use of BZ in food animals in the past and the
knowledge that the drug undergoes a very complex metabolic fate in laboratory
animals and man, there are no published data describing the metabolism of BZ in
food species such as cattle. Establishing the extent to which the metabolite pro®le
of the target species is representative of that evaluated in previous studies in
laboratory animals provides the basis for determining whether further toxicity
assessment is required for extension of licensing to a new food species. Further-
more, this provides an indication of the potential target residues for requisite
pharmacokinetic and residues studies, etc.
Several in vitro models, including isolated hepatocytes and microsomes, are
available for the assessment of xenobiotic metabolism. Recently, precision-cut
liver slices have been utilized as an alternative method since the approach main-
tains tissue architecture and functional heterogeneity and provides a pathway for
drug metabolism that is more consistent with in vivo metabolism than other
models (Smith et al. 1987, Barr et al. 1991, Dogterom 1993, Miller et al. 1993,
Saylers et al. 1994, George et al. 1999).
In vitro studies using rat liver and brain microsomes and hepatocytes from
several species have demonstrated that the N-oxidation of BZ to BZ-NO is
catalysed by the ¯avin-containing monooxygenases (FMO), whereas the N-de-
methylation of BZ to Nor-BZ involves cytochrome P450 monooxygenases (Kawaji
et al. 1993, 1994, Ubeaud et al. 1999).
In vitro metabolism studies can be viewed as a means of readily enabling the
generation of valuable qualitative comparative data as a basis for the initial
assessment of equivalence with other target species and with model species used
in toxicity studies, such as the rat. In addition, such preliminary data may better
enable more e€ective planning of essential in vivo studies.
The biotransformation of BZ was investigated in the present study by using
microsomes and precision-cut slices prepared from cattle liver to provide enzyme
kinetic details on the production of BZ-NO and Nor-BZ as well as preliminary
information on the likely pathways of BZ metabolism in the species. Metabolites
were analysed by high-performance liquid chromatography (HPLC) with ¯uori-
metric detection, and the structural identity of metabolites was investigated by
mass spectrometry (MS).
Material and methods
Chemicals
Benzydamine hydrochloride (purity 99.8%), benzydamine N-oxide hydrochloride (97.3%) and nor-
benzydamine hydrochloride (99.7%) were supplied by Angelini Ricerche S.p.A. (Societa per Azioni).
NADPH and bovine serum albumin (BSA) were from Boehringer Mannheim Italia S.p.A. (Milan,
Italy). Earl’s balanced salt solution (EBSS) without phenol red, RPMI 1640 w/GLUTAMAX II with
25 mm HEPES and glutamine were supplied by Gibco BRL (Life Technologies, Paisley, UK).
Methionine, hydrocortisone, insulin, foetal calf serum, gentamycin and amphoteracin B were supplied
by Sigma Chemical Co. (St Louis, MO, USA). All other reagents, salts (analytical grade) and solvents
(hplc grade) were purchased from Mallinchrodt Baker S.p.A. (Deventer, The Netherlands) and Carlo
Erba (Milan, Italy).

Biotransformation of benzydamine in cattle
75
Metabolism of BZ by liver microsomes
Samples of liver tissue (caudate lobe) from four healthy steers (550±600 kg b.w.) were collected at the
slaughterhouse immediately post mortem. Samples of liver (¹50 g) were cut into small pieces and washed
three times with ice-cold 1.15% (w/v) KCl, suspended in 2 vols 0.1 m phosphate bu€er (pH 7.4) and
1.15% (w/v) KCl solution (1:1 v/v) and homogenized in a Potter apparatus with a Te¯on pestle (Heavy
Stirrer, Apparecchi per Laboratori Chimici s.r.l., Milan, Italy). The homogenate was centrifuged at
9000g at 48C for 20 min and the supernatant retained and further centrifuged at 105 000g for 60min to
obtain the microsomal fraction. The microsomes were suspended in 0.1m phosphate bu€er (pH
7.4):glycerol 80:20 (v/v) containing 0.1 m EDTA and stored under liquid N2 pending biotransformation
studies. Immediately before use, the microsomal fraction was thawed at 378C and suspended in 0.1 m
phosphate bu€er (pH 7.4) or 0.1 m Tris-HCl bu€er (pH 8.5) for evaluation of CYP- and FMO-
catalysed reactions, respectively. The protein concentration of the microsomal preparation was
determined using bovine serum albumin (BSA) as a calibration standard (Lowry et al. 1951). Total
microsomal cytochrome P450 was determined by measuring the CO spectra of dithionite reduced P450
(Omura and Sato 1964).
Biotransformation was studied in triplicate by incubating 0.2 mg ml¡1 microsomal protein, 1mm
NADPH, 2.5 mm MgCl2, and substrate (BZ; concentrations from 2.5 to 1000 mm) in a ®nal volume of
0.4 ml, with either 0.1 m phosphate bu€er (pH 7.4) or 0.1 m Tris-HCl (pH 8.5) for 10 min at 378C. Blank
samples without NADPH or BZ were used to control for non-enzymatic transformations.
The e€ects of microsomal protein concentration and of incubation time on the formation of
metabolites was evaluated as above over the range 0.1±1 mg ml¡1 microsomal protein and from 5 to
30 min of incubation. The reactions were stopped by addition of acetonitrile (0.8 ml). After vortexing
for 1 min, samples were centrifuged at 13 000g for 5 min (Microfuge E, Beckman, San Ramon, CA,
USA), and the supernatant diluted from 1:1 to 1:20 v/v with the HPLC mobile phase.
To evaluate the contribution of CYP- and FMO-catalysed reactions, further assays were performed
with substrate concentrations of 1.0 and 2.5 mm in a similar manner § heat treatment (478C for 3 min) to
inactivate FMO (Ziegler 1988).
Metabolism of BZ by precision-cut liver slices
Samples of liver tissue were obtained from three healthy calves Angus/Friesian (cross steer, 7±10
months old) slaughtered by captive bolt and exsanguination. Livers were perfused within 10 min of
slaughter with ice-cold Earl’s balanced salt solution (EBSS, previously equilibrated with 95% O2/5%
CO2) and transported to the laboratory on ice.
Cylindrical tissue cores were prepared using a 10-mm diameter motor-driven tissue-coring tool.
Tissue slices (200±300 mm thick) were cut from the cylinders with a Krumdieck tissue slicer (Alabama
Research and Development Corp., Munford, AL, USA) in oxygenated EBSS containing 25 mm glucose
at 48C. Liver slices were ¯oated into 12-well plates (one slice per well) and incubated with culture
medium (1 ml) employing a continuously submerged incubation system. The culture medium was
prepared from RPMI 1640 medium containing 5% (v/v) foetal calf serum, 1 mm insulin, 0.5 mm l-
methionine, 0.1 mm hydrocortisone-21-hemisuccinate and 50mg ml¡1 gentamycin (Coldham et al.
1995).
Culture plates were incubated at 378C in a humidi®ed atmosphere of 95% O2/5% CO2. After 30-min
incubation, treatment was commenced by replacing the culture medium with fresh medium and a range
of BZ concentrations (0, 100, 500, 1000mm) was added to the appropriate wells.
Following incubations of 0, 2, 4, 6, 8 and 24 h, the medium was removed for analysis of metabolites.
Liver slices were washed in 1.15% KCl, containing 50 mm Tris-HCl (1ml), homogenized in this
medium with a IKA-Labotechnic Ultra Turrax 25 and sonicated with an MSE Soniprep 150 Ultrasonic
Disintegrator (MSE Scienti®c Instruments, Crawley, UK).
The liver slice homogenates and samples of culture medium were stored at 7808C before analysis.
Whole homogenates of liver slices were assayed for protein content with BSA as calibration standard
(Lowry et al. 1951). Samples of culture medium were diluted and analysed by HPLC and HPLC-MS.
HPLC analysis
HPLC consisted of a Jasco Pump PU 980 equipped with an LG 980 02 Ternary Gradient Unit and a
GASSTORR GT-103 degasser, a Jasco Autosampler AS 950, and a Jasco ¯uorometric detector FP 920
(Japan Spectroscopic, Tokyo, Japan) set at 303 nm excitation and 350 nm emission.
Samples of microsomal supernatant and culture medium (20 ml) were chromatographed on a Prodigy
ODS(3) column (5 mm; 250 £ 4.6 mm i.d.) (Phenomenex, Torrance, CA, USA) using an isocratic binary
mobile phase consisting of water 0.1% tri¯uoracetic acid (TFA):acetonitrile 0.082%TFA 65:35 (v/v) at a
¯ow rate of 0.8 ml min¡1. Under these chromatographic conditions reference compounds eluted
(retention time §SD, min) in the order: Nor-BZ …11:8 § 0:3†, BZ …13:1 § 0:3† and BZ-NO …15:7 § 0:3†.
Calibration curves of BZ, BZ-NO and Nor-BZ in the range 0.01¡1 mm were prepared in the mobile
phase starting from aqueous standard solutions (1 mm). The concentration of analyte in unknown
samples was determined by interpolating from the appropriate calibration curve.

76
A. Santi et al.
LC-MS analysis
Culture medium (20 ml) from precision-cut liver slices was analysed by LC-MS using an HP1050
HPLC system (Hewlett Packard, Bracknell, UK) and an ion-trap mass spectrometer (LCQ, Finnigan,
Hemel Hempstead, UK) in series with a Shimatzu RF-535 ¯uorescence detector set at 303 nm
excitation and 350nm emission.
BZ metabolites were chromatographed on the same Prodigy RP column and eluted using a binary
mixture of (A) water 2% formic acid:(B) acetonitrile 2% formic acid. Metabolites were eluted at ¯ow rate
of 0.8 ml min¡1 with a linear gradient from 90% A to 65% A over 10 min, followed by an isocratic period
at 65% A for 15 min.
Under these chromatographic conditions, the reference compounds were resolved (retention
time § SD, min) in the order Nor-BZ …16:4 § 0:2†, BZ …17:1 § 0:3† and BZ-NO …18:2 § 0:4†.
MS was optimized by infusion of a BZ standard (1 mg ml¡1). Positive-ion electrospray provided the
greatest sensitivity. The probe settings included a heated capillary temperature of 271.58C and voltage
of 23.6 V.
Mass spectra were collected in either full scan (m/z 100±550) or product ion modes, the latter being
derived from the molecular ion ‰M ‡ HŠ‡. MS experiments on BZ and metabolites were performed with
CID energies of 30% and mass isolation width of 2 amu.
Analysis of BZ, Nor-BZ and BZ-NO fragmentation by MS was determined by direct infusion of
10 mm solutions into mobile phase at a ¯ow rate of 10 ml min¡1
.
Potential metabolites were investigated by selecting masses at m/z 236, 326, 342, 486, 488, 502 and
518 for further MS experiments (taking account of published literature and fragmentation pathways of
BZ, Nor-BZ and BZ-NO), at elution times corresponding with peaks detected by ¯uorescence.
Results
Metabolism of BZ by liver microsomes
The content of cytochrome P450 in bovine microsomes was
0 61 § 0 05 nmolmg¡1 protein (mean § SD, n ˆ 4). After incubation of BZ with
:
:
microsomes, three metabolites were observed in the supernatant (®gure 1), the ®rst
unidenti®ed product eluting at an earlier retention time (¹9.9 min) than BZ (12.8
min) and two further metabolites co-eluting at retention times corresponding with
UF
5.0E+05
BZ
12.752
4.0E+05
Nor-BZ
BZ-NO
3.0E+05
2.0E+05
1.0E+05
0.0E+00
11.517
15.358
9.926
5.00
10.00
minutes
15.00
20.00
Figure 1. HPLC chromatogram obtained from cattle liver microsomes incubated with BZ (100mm).
The HPLC conditions were: Prodigy ODS(3) column (5mm; 250 £ 4.6mm i.d.) (Phenomenex,
Torrance, CA, USA); isocratic binary mobile phase consisting of water 0.1%TFA:acetonitrile
0.082%TFA 65:35 (v/v); ¯ow rate 0.8ml min¡1; ¯uorometric detection (303 nm excitation,
350 nm emission). UF, units of ¯uorescence.

Biotransformation of benzydamine in cattle
77
10
9
pH 8.5
8
7
6
5
4
3
2
1
0
0
200
400
600
800
1000
10
9
8
7
6
5
4
3
2
1
0
Conc (mM)
pH 7.4
0
200
400
600
800
1000
Conc (mM)
Figure 2. Production of Nor-BZ (*) and BZ-NO (^) as a function of substrate concentration in
cattle liver microsomes. Values are the mean § SD of data derived from samples from four
animals. A non-linear regression software (Fig.P 2.0, Biosoft, Cambridge, UK) was used to
provide curves of best ®t to data derived from the Michaelis±Menten equation.
Table 1. Kinetic parameters derived from the Michaelis±Menten
equation for Nor-BZ and BZ-NO formation from BZ by cattle liver
microsomes.
Product
Km(mm)
Vmax (nmol mg¡1 min¡1
)
Nor-BZ
pH 7.4
pH 8.5
58:9 § 6:4 (52.7±67.5)
46:7 § 4:2 (41.2±51.4)
7:4 § 0:5 (6.9±7.9)
4:8 § 0:5 (4.1±5.2)
BZ-NO
pH 7.4
pH 8.5
76:4 § 16:0 (64.6±100.0)
63:6 § 14:0 (46.8±79.6)
6:5 § 0:8 (5.6±7.5)
8:0 § 1:6 (6.4±10.3)
Values are mean § SD …n ˆ 4† (range).
authentic Nor-BZ (11.5 min) and BZ-NO (15.4 min). The production of meta-
bolites was linear up to 0.5 mg ml¡1 of microsomal protein and for incubation
times up to 15 min (data not shown).
The formation of Nor-BZ and BZ-NO followed Michaelis±Menten kinetics
either when the incubations were performed at pH 8.5 or at pH 7.4 (®gure 2). Km
and Vmax (mean§ SD and ranges) are provided in table 1. The production of Nor-
BZ and BZ-NO was greatest at pH 7.4 and 8.5, respectively, although the small
number of animals weakened the statistical analysis.

78
A. Santi et al.
The formation of Nor-BZ was not reduced by heat treatment of the micro-
somal protein, whereas that of BZ-NO decreased to 29±33% of the control
samples.
Metabolism of BZ by precision-cut liver slices
Figure 3 shows HPLC chromatograms obtained from medium after incubation
of precision-cut liver slices with BZ for 2, 6 and 24 h compared with a blank sample
(liver slices incubated for 24 h). After 2 h of incubation, the main component was
parent compound (95%), with only traces of Nor-BZ (1%) and BZ-NO (4%)
evident. The percentage abundance of BZ-NO increased between 6 and 24h
UF
Blank
2 hours
6 hours
24 hours
minutes
Figure 3. Metabolite pro®les from precision-cut liver slices. Representative HPLC chromatograms
have been selected to illustrate metabolites present in extracts of medium following incubation of
precision-cut liver slices with BZ for 2, 6 and 24 h. The HPLC conditions were: Prodigy ODS(3)
column (5 mm; 250 £ 4.6 mm i.d.) (Phenomenex); elution by binary mixture of (A) water 2%
formic acid; (B) acetonitrile 2% formic acid with a gradient from 90% A to 65% A over 10 min,
followed by an isocratic period at 65% A for 15 min; ¯ow rate of 0.8 ml min¡1; ¯uorometric
detection (303nm excitation, 350 nm emission). UF, units of ¯uorescence.

Biotransformation of benzydamine in cattle
79
Table 2. Summary of major product ions formed following mass spectral analysis
of BZ, Nor-BZ and BZ-NO ‰M ‡ HŠ‡ parent ions (m/z 310, 296 and 326,
respectively).
MSn
n
[Molecular ion], precursor ion
…m=z†
Product ions
…m=z†
Compound
BZ
2
3
4
4
2
3
2
3
[310]
[310], 265
[310], 265, 174
[310], 265, 237
[296]
[296], 265
[326]
[326], 265
265
237, 187, 174*, 91
146*
209
265*, 225
237, 209, 187, 174*
265, 209, 102*
237, 174*
Nor-BZ
BZ-NO
* Base product ion in the spectra.
from 33 to 56%. In contrast, Nor-BZ production was maximal (5%) by 6 h. At 24h
several other peaks eluting before Nor-BZ were also evident.
LC-MS analysis of metabolites
Product ions generated from the molecular ions ([M + H]‡) of BZ (m/z 310),
Nor-BZ (m/z 296) and BZ-NO (m/z 326) are summarized in table 2. A product ion
of m/z 265 was evident in the spectrum of BZ, Nor-BZ and BZ-NO, whereas those
at m/z 225 and 102 were only evident in the spectrum of Nor-BZ and BZ-NO,
respectively.
Further LC-MS experiments were performed on ions that co-chromato-
graphed with unidenti®ed ¯uorescent entities found in the medium at 6- and
24-h incubations. Figure 4 shows selected ion chromatograms corresponding to
the [M + H]‡ of eight putative metabolites (labelled M1±8 in order of elution)
obtained from a representative 24-h medium sample; retention times, molecular
and product ions for each are provided in table 3.
Two metabolites eluting at 13.3 and 17.9 min (M9 and BZ-NO respectively)
were evident from the total ion current trace for the parent ion of m/z 326. The
retention time and product ion spectrum of the latter were consistent with BZ-NO.
The former metabolite was not detected by ¯uorescence but generated a single
product ion of m/z 281, which was also evident as a product ion of other
metabolites (table 3).
Discussion
The di€erent roles of FMO and cytochrome P450 in the N-oxidation of BZ
and N-demethylation of BZ, respectively, were previously established by Kawaji
et al. (1993, 1994) using rat liver microsomes. From that model, Km and Vmax for
the formation of Nor-BZ were produced that were ¹50 times greater and 2000
times lower, respectively, than those for BZ-NO.
In the present study using liver microsomes from cattle, BZ-NO production
increased at pH 8.5 and was sensitive to heat treatment, suggesting that FMOs
may be largely responsible for catalysing this reaction. However, the closeness of

80
A. Santi et al.
M 3
M 8
M 7
M 4
M 5
1
M 2
M 6
Figure 4. Identi®cation of potential [M + H‡] molecular ions (m/z 236, 342, 486, 488, 502 and 518) of
metabolites from precision-cut liver slices by LC-MS by extraction of ions from full-scan data.
Precision-cut liver slices were incubated with BZ for 24 h and the incubation medium extracted
and subjected to analysis by full scan (m/z 200±600) HPLC-MS.
Table 3. Fragmentation of precursor ions of metabolites found in the medium
following incubation of precision-cut liver slices for 24 h.
Retention time
(min)
Precursor ions
…m=z†
MSn
n
Product ions
…m=z†
Metabolite
M1
9.1
502 …M ‡ H‡†
2
3
2
2
2
3
2
3
2
2
2
2
457*, 326, 281
281*, 191
457*, 281
175, 102*
457, 312*, 281
281*, 175
457, 326*, 281
281*, 175
457, 342*, 281
281, 102*
502, 457
M2
M3
M4
9.6
10.1
10.8
518 …M ‡ H‡‡†
236 …M ‡ H‡†
488 …M ‡ H †
488, 312
M5
10.9
502 …M ‡ H‡†
502, 457
M6
M7
M8
M9
11.4
13.8
14.7
13.3
518 …M ‡ H‡‡†
342 …M ‡ H †
486 …M ‡ H‡†
326 …M ‡ H‡†
441, 310, 265*
281
* Base product ion in the MS spectra.

Biotransformation of benzydamine in cattle
81
Km and Vmax of Nor-BZ and BZ-NO formation indicated the involvement of
either the same enzyme or di€erent enzymes with similar substrate anity.
In addition to BZ, Nor-BZ and BZ-NO, only one further minor peak was
evident in the HPLC chromatograms of microsomal incubations (®gure 1). In
contrast, chromatograms from the media of liver slices gave rise to several peaks
that eluted before reference standards for BZ, Nor-BZ and BZ-NO (®gure 3);
these were present in sucient quantity to enable direct identi®cation by LC-MS
analysis.
The concentration of BZ-NO in liver slice culture medium exhibited a striking
increase from 2 to 24 h compared with the other metabolites, and this feature of BZ
biotransformation could modify BZ residue kinetics. Back reduction to the parent
compound is a mechanism already described for N-oxides and hydroxylamines in
biological ¯uids (Bickel 1969, Kaibaf et al. 1992, Blake et al. 1995, Montesissa et
al. 1998). Previous studies have demonstrated that BZ-NO can also form a
reservoir of the parent compound, since it can be N-deoxygenated back to BZ,
both in vitro by rat liver microsomes and in vivo after oral administration of BZ to
rat (Kataoka et al. 1979a, b). In vivo, the reduction of N-oxidated metabolites is
often related to the inability to detect these labile derivatives in biological ¯uids
and to the increased half-life of their parent compounds.
The previously undetected BZ metabolites produced in cattle liver slices were
tentatively identi®ed as follows (®gure 5): M1 (m/z 502), 1-benzyl-3-[3-(dimethy-
lamino)propoxy]-1H-hydroxy-indazole N‡-glucuronide; M2 (m/z 518), 1-benzyl-
3-[3-(dimethylamino)propoxy]-1H-hydroxy-indazole N-oxide N‡-glucuronide;
M3 (m/z 236), 3-[3-(dimethylamino)propoxy]-1H-indazole N-oxide; M4 (m/z
488), 1-hydroxy-benzyl-3-[3-(methylamino)propoxy]-1H-indazole N‡-glucuro-
nide; M5 (m/z 502), 1-hydroxy-benzyl-3-[3-(dimethylamino)propoxy]-1H-
indazole N‡-glucuronide; M6 (m/z 518), 1-hydroxy-benzyl-3-[3-(dimethylami-
no)propoxy]-1H-indazole N-oxide N‡-glucuronide; M7 (m/z 342), hydroxylated
analogue of BZ-NO; M8 (m/z 486), benzydamine N‡-glucuronide; and M9 (m/z
326), hydroxylated analogue of BZ. These products are presented in a proposed
hepatic biotransformation pathway in ®gure 6.
Only three of the previously unidenti®ed metabolites (M3, M7 and M9)
(®gures 5a and 6) found in the incubation medium of cattle liver slices were
unconjugated compounds. The metabolite M3 (m/z 236) could be produced by
either the loss of the benzyl group from BZ-NO or by debenzylation of BZ
followed by an oxidation. Desbenzyl-BZ was ®rst identi®ed in rabbit urine by
Kataoka et al. (1971) and subsequently in human urine by Koppel and Tenczer
(1985).
The exact position of the hydroxyl group for the metabolites M7 (m/z 342)
(®gure 5a) and M9 (m/z 326) (®gure 6) can not be assigned from the product ion
spectra. However, Kataoka et al. (1971) described a hydroxy-benzyl derivative in
rabbit urine with the site of hydroxylation at the 4-position of the benzyl ring,
whereas Koppel and Tenczer (1985) suggested the 5-position of the indazole ring
for a hydroxylated metabolite found in human urine.
The fragmentation patterns of the other metabolites (M1, M2, M4±6 and M8)
were consistent with fragments formed from N‡-glucuronides. The peak for M8
(m/z 486) (table 3, ®gure 5b) gave rise to a product ion at m/z 441, which
correspond to the sum of the major fragment of BZ (m/z 265, table 2) and
glucuronic acid (m/z 176). This suggests that M8 is a novel N‡-glucuronide of

82
A. Santi et al.
A
M7
{M+H}+
342
O
M3
O - C3H6 - N(CH3
)
2
{M+H}+
236
O
NH
+
O - C3H6 - N(CH3
)
2
N
NH
+
CH2
NH
H2C=CH-CH2N+(CH3
)
2
OH
OH
C5H12NO
+ 102
{M+H}
O - CH2 - CH=CH2
NH
+
O - CH2 - CH=CH2
R2
N
NH
+
CH2
N
H
C17H17N2O2
C10H11N2
+ 175
{M+H}
O
{M+H}+
281
OH
B
O - C3H6 - R1
R2
O - C3H6 - R1
COOH
+
N
R2
O
N
NH
+
OH
CH2
N
OH
OH
C19H24N3
{M+H}
O
CH2
+ 310
C19H24N3O2
+ 326
{M+H}
R3
R3
M1 {M+H}+
M5 {M+H}+
:
502 R =N(CH ) ; R =OH; R =H
1
3
2
2
3
502 R =N(CH ) ; R =H; R =OH
:
1
3
2
2
3
M2 {M+H}+
M6 {M+H}+
:
518 R =NO(CH ) ; R =OH; R =H
1
3
2
2
3
518 R =NO(CH ) ; R =H; R =OH
:
1
3
2
2
3
O - CH2 - CH=CH2
R2
M4 {M+H}+
M8 {M+H}+
:
488 R =NHCH ; R =H; R =OH
1
3
2
3
COOH
+
N
C23H25N2O7
+ 441
O
486 R =N(CH ) ; R = R =H
:
1
3
2
2
3
N
{M+H}
C23H25N2O8
{M+H}
OH
CH2
OH
OH
+ 457
O - CH2 - CH=CH2
R2
NH
+
N
R3
CH2
C17H17N2
O
{M+H}+
265
C17H17N2O2
281
{M+H}+
O - CH2 - CH=CH2
R2
C10H11N2
{M+H}
O
R3
NH
+
+ 175
N
H
C10H11N2O2
+ 191
{M+H}
Figure 5. Proposed structures of precursor and product ions of BZ metabolites. Part A shows the
unconjugated derivatives M3 (m/z 236, debenzyl BZ-NO) and M7 (m/z 342, hydroxylated BZ-
NO). Part B shows the N‡-glucuronated derivatives M1 and M5 (m/z 502, hydroxylated BZ N‡-
glucuronide), M2 and M6 (m/z 518, hydroxylated BZ-NO N‡-glucuronide), M4 (m/z 488,
hydroxylated Nor-benzydamine N‡-glucuronide) and M8 (m/z 486, benzydamine N‡-glucur-
onide). The position of the hydroxyl group is suggested in every metabolite, but cannot be
de®nitely stated based on product ion spectra.

Biotransformation of benzydamine in cattle
83
Figure 6. Proposed pathways for hepatic BZ metabolism in cattle.
BZ which to date has not been described in the literature. M4 (m/z 488), M1 and
M5 (both m/z 502), M2 and M6 (both m/z 518) (table 3, ®gure 5b) generated a
common fragment at m/z 457 which is an oxidated (+16 amu) fragment analogous
to the one at m/z 441 originated from M8. The mass for the putative [M + H]‡ ion
of M8 is consistent with benzydamine N‡-glucuronide; thus it is reasonable to
propose analogous N‡-glucuronide structures for metabolites M1, M2 and M4±6
(®gures 5b and 6). Maggs et al. (1997) reported that the in vivo glucuronidation of
OH-carbamazepine (the main derivative of carbamazepine, an analgesic and
anticonvulsant agent) gave rise to both O- and N-glucuronides. Their mass spectra
have been compared and the product ion found to di€er: the O-glucuronides (m/z
429) yielded only the aglycone fragment (m/z 253), whereas the N-glucuronide
conjugate produced a fragment still bearing the glucuronic acid moiety (m/z 411).
Thus, the ion at m/z 457 in the spectra of BZ metabolites M1, M2 and M4±6 is
evidence of the N-conjugated moiety still bearing the glucuronic acid. This is
con®rmed in their MS spectra by the presence of a main fragment at m/z 281
corresponding to the loss of the glucuronic group. Further evidence for the N‡-
glucuronide structural assignment was provided by metabolites such as M1/5 and
M2/M6, which have identical masses but not HPLC retention times due to
di€erent molecular sites of hydroxylation and their polarities.
The unsaturated nitrogen of the imidazole moiety of BZ and BZ-OH is the
most likely site for conjugation with glucuronic acid. In ring systems of aromatic
tertiary amines, such as imidazole (where one ring nitrogen is substituted),
pyridine, 1,2,4-triazine and polycyclic heterocycles, where N‡-glucuronidation

84
A. Santi et al.
can occur at more than one site, conjugation has been observed to occur at the
nitrogen with the less bulky substituent(s) (Hawes 1998).
The exact structures of the metabolites of BZ produced by precision-cut slices
from cattle liver cannot be de®nitely stated without other analysis by NMR or
synthesis of reference compounds. Nevertheless, the assignment of N‡-glucur-
onides at both the parent compound and its hydroxylated derivatives is strongly
supported by the mass spectra and chromatographic data.
N‡-glucuronidation has been widely observed in man as a pathway in the
Phase II metabolism of tricyclic drugs with an aliphatic tertiary amine group.
Examples include imipramine, clomipramine, clozapine, chlorpromazine, trazo-
done, loxapine and trimipramine (Luo et al. 1995, Dain et al. 1997), but there are
no reports in the literature describing N‡-glucuronidated products from hydro-
xylated BZ or reports on the ability of cattle to perform N‡-glucuronidation. Until
now, only the rabbit, unlike the other laboratory species (dog, rat, cat) is reported
to share this metabolic pathway with humans. In both species, N‡-glucuronides of
ketotifen and cyproheptadine were produced during in vivo and in vitro studies,
and N‡-glucuronides of tioconazole and croconazole were found in both rabbit
and human urine (Hawes 1998).
As expected from published studies carried out in other animals and in man,
the ®ndings of the present study indicate that the metabolism of BZ is very
complex in cattle. In the case of drugs adopted both for human and veterinary
therapy, liver slices are an appropriate in vitro model to elucidate the metabolites to
which humans can be re-exposed following consumption of animal tissues. To
support the process of licensing of BZ in the bovine, and in food species in general,
it is essential to de®ne the composition of tissue residues after administration of the
compound and to know whether the metabolites found in the target species are
representative of those `tested’ in laboratory animals (as part of toxicity testing)
and those found in humans. Glucuronidation is generally regarded as a means of
detoxi®cation and the clinical importance of N‡-glucuronides has not been
thoroughly investigated. Nevertheless, a few studies described their e€ects on
histamine release, as well as the competition for glucuronic substrate between
tertiary amines and testosterone (Chiu and Huskey 1998).
The results of this study suggest that BZ metabolites in cattle are qualitatively
di€erent from those described in humans and laboratory animals, and further
toxicity tests on metabolites in laboratory species should be required to enable
marketing authorization of BZ for food species.
In view of the understanding of the mass balance of BZ in cattle and to
highlight a pathway of detoxi®cation that has never been reported in the bovine,
the metabolism of BZ in cattle is worthy of further investigation.
Acknowledgements
The authors are grateful to Giancarlo Biancotto (Istituto Zoopro®lattico
Sperimentale delle Venezie di Padova) for his scienti®c contribution to the
discussion of metabolite structure and to Professor Luigi Morganti (Director of
the Dipartimento di Sanita Pubblica Veterinaria e Patologia Animale) for his
support of the study. The work was supported by a MURST (Ministero
dell’Universita e della Ricerca Scienti®ca) grant (ex-40% 1998).

Biotransformation of benzydamine in cattle
85
References
Baldock, G. A., Brodie, R. R., Chasseaud, L. F., Taylor, T., Walmsley, L. M. and Catanese, B.,
1991, Pharmacokinetics of benzydamine after intravenous, oral, and topical doses to human
subjects. Biopharmaceutics and Drug Disposition, 12, 481±492.
Barr, J., Weir, A. J., Brendel, K. and Sipes, I. G., 1991, Liver slices in dynamic organ culture, an
alternative in vitro technique for the study of rat hepatic drug metabolism. Xenobiotica, 21, 331±
339.
Bickel, M. H., 1969, The pharmacology and biochemistry of N-oxides. Pharmacological Review, 21,
325±355.
Blake B. L., Rose, L. R., Mailman, R. B., Levi, P. E. and Hodgson, E., 1995, Metabolism of
thioridazine by microsomal monooxygenases: relative roles of P450 and ¯avin-countering
monooxygenases. Xenobiotica, 25, 377±393.
Chiu, S. L. and Huskey, S. W., 1998, Species di€erences in N-glucuronidation. Drug Metabolism and
Disposition, 26, 838±847.
Coldham, N., Moore, A. S., Dave, M., Graham, P. J., Sivapathasundaram , S., Lake, B. G. and Sauer,
M. J., 1995, Imidocarb residues in edible bovine tissues and in vitro assessment of imidocarb
metabolism and cytotoxicity. Drug Metabolism and Disposition, 23, 501±505.
Dain, J. G., Nicoletti, J. and Ballard, F., 1997, Biotransformation of clozapine in humans. Drug
Metabolism and Disposition, 25, 603±609.
Dogterom, P., 1993, Development of a simple incubation system for metabolism studies with precision-
cut liver slices. Drug Metabolism and Disposition, 21, 699±704.
George, E., Murdock, J., Aylott, M. and Westmoreland, C., 1999, Comparison of hepatocyte
cultures and liver slices in vitro toxicity testing. Alternatives to Laboratory Animals, 27, 769±781.
Hawes, E. M., 1998, N‡-glucuronidation, a common pathway in human metabolism of drugs with a
tertiary amine group. Drug Metabolism and Disposition, 26, 830±837.
Kaibaf, M., Sepai, O. and Lamb, J. H., 1992, Identi®cation of metabolites of 4,40-
diaminodiphenylmethane (methylene dianiline) using liquid chromatographic and mass
spectrometric techniques. Journal of Chromatography Biomedical Application, 583, 63±76.
Kataoka, S., Naito, T., Nishimura, K. and Taira, K., 1979b, E€ect of carrageenin treatment on the
metabolism of benzydamine and its N-oxide in rat organ preparations. Chemical and
Pharmaceutical Bulletin, 27, 2904±2912.
Kataoka, S., Nishimura, K. and Naito T., 1979a, In vivo metabolism and anti-in¯ammatory activity of
benzydamine hydrochloride in rats treated with carrageenin. Chemical and Pharmaceutical
Bulletin, 27, 2890±2903.
Kataoka, S., Taira, K., Ariyoshi, T. and Takabatake, E., 1973, Metabolism of benzydamine
hydrochloride: species di€erences and the identi®cation of unconjugated metabolites in rabbit
urine. Chemical and Pharmaceutical Bulletin, 21, 358±365.
Kataoka, S., Taira, K. and Takabatake, E., 1971, Metabolism of benzydamine hydrochloride.
Chemical and Pharmaceutical Bulletin, 19, 1511±1513.
Kawaji, A., Ohara, K. and Takabatake, E., 1993, An assay of ¯avin-containing monooxygenase activity
with benzydamine N-oxidation. Analytical Biochemistry, 214, 409±412.
Kawaji, A., Ohara, K. and Takabatake, E., 1994, Determination of ¯avin-containing monooxygenase
activity in rat brain microsomes with benzydamine N-oxidation. Biological and Pharmaceutical
Bulletin, 17, 603±606.
Koppel, C. and Tenczer, J., 1985, Metabolism of benzydamine. Arzneimittel-Forschung/Drug Research,
35, 634±635.
Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J., 1951, Protein measurement with the
Folin phenol reagent. Journal of Biological Chemistry, 293, 265±275.
Luo, H., Hawes, E. M., McKay, G., Korchinski, E. D. and Midha, K. K., 1995, N‡ glucuronidation
of aliphatic tertiary amines in human: antidepressant versus antipsychotic drugs. Xenobiotica, 25,
291±301.
Maggs, J. L., Pirmohamed, M., Kitteringham, N. R. and Park, B. K., 1997, Characterization of the
metabolites of carbamazepine in patient urine by liquid chromatography/mass spectrometry.
Drug Metabolism and Disposition, 25, 275±280.
Miller, M. G., Beyer, J., Hall, G. L., Degraffenried, L. and Adams, P. E., 1993, Predictive value of
liver slices for metabolism and toxicity in vivo: use of acetaminophen as a model hepatotoxicant.
Toxicology and Applied Pharmacology, 122, 108±116.
Montesissa, C., Anfossi, P., Biancotto, G. and Angeletti, R., 1998, In vitro metabolism of
clenbuterol and bromobuterol by pig liver microsomes. Xenobiotica, 28, 1049±1060.
Omura, T. and Sato, R., 1964, The carbon monoxide binding pigment of liver microsomes. Journal of
Biological Chemistry, 239, 2370±2378.
Saylers, K. L., Barr, J. and Sipes, I. G., 1994, In vitro metabolism of theophylline by rat and human
liver tissue. Xenobiotica, 24, 389±399.

86
Biotransformation of benzydamine in cattle
Smith, P. F., Fisher, R., Shubat, P. J., Gandol, V. J., Krumdieck, C. L. and Brendel, K., 1987, In
vitro cytotoxicity of allyl alcohol and bromobenzenes in novel organ culture system. Toxicology
and Applied Pharmacology, 87, 509±522.
Ubeaud, G., Schiller, C. L., Hurbin, F., Jaeck, D. and Coassolo, P., 1999, Estimation of ¯avin-
containing monooxygenase activity in intact hepatocyte monolayers of rat, hamster, rabbit, dog
and human by using N-oxidation of benzydamine. European Journal of Pharmaceutical Sciences,
8, 255±260.
Ziegler, D. M., 1988, Flavin-containing monooxygenases: catalytic mechanism and substrate
speci®cities. Drug Metabolism Review, 19, 1±32.