![]() |
|
|
Vol. 26, Issue 4, 294-298, April 1998
Microchemistry Laboratory, Pathobiology Department, University of Connecticut1 (T.C.B., A.J.K., M.F., D.W.H.), Department of Biomedical Sciences, Quinnipiac College (T.C.B.), and Racing Chemistry Laboratory, Iowa State University (W.H.H.)
| |
Abstract |
|---|
|
|
|---|
A urinary metabolite of flunixin in greyhound dogs was isolated and
purified by a gradient-elution solid-phase extraction technique. The
purified metabolite was shown to be hydrolyzed to free flunixin by
strong base and by
-glucuronidase, suggesting the presence of a
C1-
-glucuronide ester of flunixin. The metabolite was further
characterized by positive-ion, tandem MS with electrospray ionization.
Mass spectral data showed the presence of a protonated molecular ion
(M+1) at m/z 473, which was consistent with the molecular
weight of protonated flunixin glucuronide, and a product ion at
m/z 297, which was consistent with the molecular weight of
protonated flunixin. Collisionally induced dissociation of the
m/z 297 product ion showed a fragmentation pattern
consistent with that of standard flunixin. These data support the
contention that this metabolite of flunixin in greyhound urine is the
C1-
-glucuronide of flunixin. Acyl glucuronide metabolites of some
organic acid drugs have been shown to bind covalently to tissue
proteins in vitro, in vivo, and ex
vivo. The presence of this metabolite may, therefore, have
pharmacokinetic and pharmacodynamic implications for flunixin in
greyhound dogs, as well as in other animal species in which the acyl
glucuronide of flunixin is a metabolite.
| |
Introduction |
|---|
|
|
|---|
Flunixin
[2-(2'-methyl-3'-trifluoromethylanilino)nicotinic acid] (fig.
1) is a nonsteroidal anti-inflammatory
drug. This drug suppresses prostaglandin synthesis through its
inhibitory effect on cyclooxygenase activity. This results in analgesia
and suppression of the inflammatory response (Ciofalo et
al., 1977
). Flunixin is used in veterinary medicine as an
alternative to narcotic analgesics and as an anti-inflammatory agent.
Flunixin is used to treat colic (Vernib and Hennessey, 1977
), lameness
(Houdeshell and Hennessey, 1977
), and endotoxemia (Moore et
al., 1981
) in horses and septic peritonitis in dogs (Hardee
et al., 1983
). The pharmacokinetics of flunixin have been
investigated in mixed-breed dogs (Chay et al., 1982
) and in
horses (Hardee et al., 1985
). Those studies show that
flunixin is rapidly and almost completely absorbed after oral
administration in those animals. The drug has a half-life of 0.6-4.2
hr in horses and 2.5-4.9 hr in dogs. These are very short half-lives
for a drug that may exhibit pharmacological effects for up to 30 hr
after a single dose (Tobin, 1981
). Flunixin has been shown to be
excreted in the urine of horses as free flunixin (Johansson and Anler,
1988
), 5-hydroxyflunixin (Jaussad et al., 1987
), and a major
metabolite that releases free flunixin after treatment with strong base
(Johansson and Anler, 1988
). There have been no reports of the
identification of this base-sensitive metabolite in horses. Unpublished
work performed in our laboratory, employing a method used successfully
with horse urine (Jaussad et al., 1987
), failed to
demonstrate the presence of an hydroxylated metabolite of flunixin in
the urine of greyhounds administered the drug. Additionally, work
performed previously in our laboratory demonstrated the presence of a
quantitatively significant urinary metabolite of flunixin in greyhound
dogs that, upon treatment with strong base or
-glucuronidase, was
hydrolyzed to free flunixin (Hill et al., 1994
). This
suggests the presence of an acyl glucuronide metabolite of flunixin.
There have been no reports of the identification of a strong
base/
-glucuronidase-hydrolyzable metabolite or the identification of
any metabolite of flunixin in dogs.
|
Acyl glucuronide metabolites of some organic acid drugs have been shown
to act as chemically reactive intermediates in the formation of
covalent bonds between the parent drug and tissue proteins (Smith
et al., 1990
, 1992
; Williams et al., 1992
;
Kauffman, 1994
; McGurk et al., 1996
). This covalent binding
has, in some cases, resulted in altered pharmacodynamics and/or
pharmacokinetics of the drug in question (Faed, 1984
). Covalent binding
of drug to tissue proteins has, in some cases, resulted in drug
hypersensitivity resulting from an immunological response to the
drug-carrier complex (Pohl et al., 1988
). This study
describes the techniques used in the isolation, purification, and
structural characterization of the acyl glucuronide of flunixin, the
major urinary metabolite of flunixin in greyhound dogs.
| |
Materials and Methods |
|---|
|
|
|---|
Chemicals and Reagents.
Solvents used in extractions and chromatography were of HPLC grade and
were obtained from EM Sciences (Gibbstown, NJ). These included
acetonitrile, methanol, ethyl acetate, and chloroform. HPLC-grade
phosphoric acid, reagent grade glacial acetic acid, and reagent-grade
sodium hydroxide were obtained from Fisher Scientific. TEA2 was obtained from Acros
Organics, Fisher Scientific (Springfield, NJ).
-Glucuronidase (EC
3.2.1.31, 1-3 × 106 units/g) was obtained
from Sigma Chemical Co. (St. Louis, MO). The C18
bonded-phase column used for the SPE was obtained from J.T. Baker, Inc.
(Phillipsburg, NJ). The Zorbax Rx C8 HPLC column (4.6 mm × 25 cm, packed with 5-µm particles) was obtained from MAC-MOD Analytical Inc. (Chadds Ford, PA). Flunixin, as the meglumine salt, was a gift from Schering Corp. (Union, NJ). Flunixin meglumine paste (Banamine), used in the greyhound administration studies, was
obtained from Schering.
Instrumentation. HPLC analyses were performed with a Hewlett-Packard (Palo Alto, CA) HP1090 system with an HP1040 diode-array detector. Solvent A was 0.15 M phosphoric acid/0.05 M TEA in water and solvent B was 0.15 M phosphoric acid/0.05 M TEA/20% water in acetonitrile. The flow rate was 2.0 ml/min. The detector was set to monitor absorbance at 210 nm, and UV spectral data were collected from 195 to 405 nm at 640-msec intervals. In the gradient-elution program, the solvent composition was held at 100% solvent A for 2.2 min and then changed linearly to 100% solvent B in 20 min. The column was a C8 Zorbax Rx column (4.6 mm × 25 cm) packed with 5-µm spherical particles. The column temperature was maintained at 30°C. Ten microliters of sample were injected onto the column by the HP1090 autosampler.
MS analyses of standard flunixin (100 µg/ml in 0.1 N acetic acid, pH 4.5) and the isolated urinary metabolite from greyhound dogs were performed by isocratic HPLC followed by on-line, positive-ion MS with electrospray ionization, using a VG QUATTRO I tandem quadrupole mass spectrometer (Fisons Instruments, Manchester, UK). The HPLC system was a Hewlett-Packard HP1090 system with an HP1040 diode-array detector. The mobile phase was 40% (v/v) CH3CN/water containing 0.05% (v/v) trifluoroacetic acid. The flow rate was 0.85 ml/min. The column was a C8 Zorbax Rx column (3.0 mm × 25 cm) packed with 5-µm spherical particles. Four microliters of sample were injected onto the column by the HP1090 autosampler. Effluent from the column was split (7.0:1.5) between the diode-array detector and the electrospray ionization source. The electrospray source cone potential was set at 50 V, and the CID collision energy for the second quadrupole was set at 60 eV. Argon was used as the collision gas. A total ion chromatogram was collected over a 6-min period. The electrospray and CID spectra of the m/z 297 product ion were collected from standard flunixin (RT = 4.67 min) and the isolated flunixin metabolite (RT = 2.55 min). Spectra from the SPE eluate of 0-hr DFU samples were also collected. Acetic acid (0.1 N, pH 4.5) was analyzed between each test sample to detect carryover.Administration of Flunixin to Greyhounds.
Flunixin meglumine (Banamine) was administered orally to four greyhound
dogs, two with a dose of 0.55 mg/kg flunixin and two with a dose of 1.0 mg/kg flunixin. Urine samples were collected before administration (0 hr) and 2, 4, 8, 24, 48, and 72 hr after administration for all four
dogs. A 96-hr urine sample was collected for each dog treated at 1.0 mg/kg. Analysis of the urine samples by HPLC before and after base
hydrolysis showed that the 8-hr urine samples contained an appreciable
concentration of flunixin metabolite (Brady et al., 1997
).
The 8-hr urine samples were used as the source of metabolite in this
study. The 0-hr samples were used as control DFU samples.
Isolation and Purification of Flunixin Glucuronide from Greyhound
Dog Urine.
The metabolite of flunixin was isolated from greyhound urine by
gradient elution from a C18 SPE column. Five
milliliters of an 8-hr urine sample were adjusted to pH 4.5 with 1.0 N
acetic acid and exhaustively extracted with three 10-ml volumes of
chloroform to remove free flunixin from the sample. The extracted urine
sample was applied to a chromatographic column containing 0.5 g of
C18 bonded-phase packing material. The column was
step-eluted with 5-ml portions of CH3CN/0.1 N
acetic acid (pH 4.5), starting with 100% 0.1 N acetic acid and
progressing, in 5% CH3CN/0.1 N acetic acid
increments, to 15% CH3CN/0.1 N acetic acid. From
16 to 20% CH3CN/0.1 N acetic acid, the column
was eluted in 1% CH3CN/0.1 N acetic acid
increments. All eluates were diluted 1/100 with 0.01 M
phosphate-buffered saline (pH 7.4) and were assayed for flunixin
reactivity using a flunixin ELISA (Brady et al., 1997
). A
quantitatively significant flunixin ELISA-reactive substance was
detected in the 16% CH3CN/0.1 N acetic acid
eluate. The CH3CN was evaporated from this eluate
in a water bath at 65°C under nitrogen, and the remaining aqueous
solution was passed through a second C18 column.
After the column was washed with 0.1 N acetic acid (pH 4.5), it was
eluted with 1 ml of 30% CH3CN/0.1 N acetic acid.
A flunixin ELISA-reactive substance was shown to be present in this
eluate. The CH3CN was evaporated from this
eluate, and the remaining aqueous solution was analyzed by
reverse-phase HPLC/DAD. The 0-hr DFU sample (negative control) and a
0-hr DFU sample spiked with 10,000 ng/ml flunixin
(DFU10,000ng/ml, positive free flunixin control)
were subjected to the same SPE procedure as was the 8-hr urine sample.
The 0-hr DFU and DFU10,000ng/ml samples were not extracted with chloroform before SPE, as was the 8-hr urine specimen.
Identification and Characterization of the Flunixin Metabolite.
The eluate containing the ELISA-reactive metabolite was analyzed by
reverse-phase HPLC/DAD, along with the corresponding DFU eluate and the
ELISA-reactive DFU10,000ng/ml eluate. HPLC
RT values and UV spectral data for the metabolite
and free flunixin were obtained. The SPE eluate containing the
metabolite was treated, in separate experiments, with strong base and
-glucuronidase. Fifty microliters of the SPE eluate containing the
metabolite were mixed with 10 µl of 18 M NaOH and allowed to incubate
for 1 hr at room temperature. After incubation, 40 µl of 0.1 N acetic acid (pH 4.5) was added to bring the final volume to 0.1 ml. For the
-glucuronidase treatment, 50 µl of the metabolite eluate was mixed
with 10 µl of 10,000 units/ml
-glucuronidase in 0.1 N acetic acid
(pH 4.5) and was allowed to incubate for 18 hr at 37°C. After the
incubation period, 40 µl of 0.1 N acetic acid was added to bring the
total volume to 0.1 ml. For comparison, 50 µl of the metabolite
eluate sample was diluted with 50 µl of 0.1 N acetic acid to a final
volume of 0.1 ml. All three samples were analyzed by reverse-phase
HPLC/DAD.
| |
Results |
|---|
|
|
|---|
The SPE eluate from the 8-hr urine sample, from which free flunixin had been exhaustively extracted, contained an ELISA-reactive compound that eluted early in the gradient elution system, relative to flunixin. This ELISA-reactive compound eluted in the 16% (v/v) CH3CN/0.1 N acetic acid eluate (retention volume, 95 ml). Analysis by HPLC/DAD of the 16% SPE eluate from the 8-hr urine sample showed the presence of only one flunixin metabolite, compared with the 16% eluate from the 0-hr control urine sample. Flunixin in the spiked DFU sample (DFU10,000ng/ml) eluted in the 18% (v/v) CH3CN/0.1 N acetic acid eluate (retention volume, 130 ml). Eluates from the 0-hr DFU sample showed no reactivity in the ELISA.
HPLC RT data for free flunixin and the
ELISA-reactive metabolite before and after treatment with sodium
hydroxide or
-glucuronidase (fig. 2)
showed that the metabolite eluted several minutes earlier (RT = 13.4 min) than flunixin
(RT = 15.2 min). The UV absorption spectra were
similar for the two compounds, with the metabolite showing a shift of
the 256-nm band of flunixin to 260 nm, with a more pronounced band at
274 nm. Additionally, the spectrum for the metabolite exhibited a red
shift of the 326-nm band of flunixin to 340 nm. Treatment of the
metabolite with sodium hydroxide and
-glucuronidase, in separate
experiments, resulted in the disappearance of the metabolite peak, with
the concomitant emergence of a peak with the RT
and UV spectral characteristics of flunixin.
|
The positive-ion, electrospray mass spectrum of the flunixin metabolite is shown in fig. 3A. This spectrum showed the presence of an apparent molecular ion (M+1) at m/z 473 and a product ion at m/z 297. CID MS analysis of the m/z 297 fragment ion (fig. 3) showed a fragmentation pattern similar to that of standard flunixin (fig. 3, C and D) analyzed under identical conditions. MS analysis of the SPE eluate from the 0-hr urine sample, under identical conditions, showed an absence of the m/z 473 and m/z 297 ions.
|
| |
Discussion |
|---|
|
|
|---|
A metabolite of flunixin was isolated from greyhound dog urine by
SPE. This metabolite was shown to be hydrolyzed to free flunixin when
treated with strong base or when incubated with
-glucuronidase.
The results of these experiments provided preliminary evidence that the
isolated compound was the C1-
-glucuronide of flunixin.
Configurations of glucuronide esters at positions other than the
C1-
-position have been reported to be resistant to hydrolysis by
-glucuronidase (Dutton, 1966
; La Du et al., 1971
). Also,
N-linked glucuronides are stable and are very resistant to
hydrolysis by
-glucuronidase under the stated conditions (La Du
et al., 1971
).
Positive-ion, electrospray MS of the isolated flunixin metabolite
showed a protonated molecular ion at m/z 473, which is
consistent with the molecular weight of 472 for flunixin glucuronide.
The electrospray spectrum of the flunixin metabolite also demonstrated a protonated product ion at m/z 297, which is consistent
with the molecular weight of protonated flunixin. CID MS analysis of the m/z 297 product ion produced a fragmentation spectrum
consistent with that of flunixin, indicating that flunixin is a
component of the m/z 473 molecular ion. These data
collectively indicate that the metabolite of flunixin isolated from
greyhound dog urine was the C1-
-glucuronide of flunixin (fig.
1B).
The possibility exists that flunixin metabolites in greyhound urine
other than the C1-
-glucuronide reacted in the ELISA detection system. It is also possible that metabolites of flunixin were nonreactive in the ELISA and, therefore, not detected.
The SPE eluate containing the isolated flunixin metabolite was treated
with strong base and
-glucuronidase, in separate experiments, and
analyzed by HPLC/DAD (fig. 2). The conversion from metabolite to free
flunixin was complete and quantitative, indicating the presence of a
single metabolite.
These hydrolysis experiments were also performed with
postadministration greyhound urine samples that had been exhaustively extracted with chloroform to remove all traces of free flunixin. The
results again showed that hydrolysis resulted in complete and
quantitative conversion to free flunixin (Brady et al.,
1997
), with no trace of another metabolite, compared with the 0-hr
urine sample. The chloroform fraction from the initial exhaustive
extraction contained only free flunixin, as determined by
RT and UV spectral data from HPLC/DAD. These
experiments provide evidence that the major metabolite of flunixin in
the urine of greyhound dogs is the acyl glucuronide, and they suggest
that other metabolites, if present, are minor components.
| |
Footnotes |
|---|
Received April 9, 1997; accepted December 16, 1997.
1 Laboratory of origin for this work.
Send reprint requests to: Dr. Dennis W. Hill, Microchemistry Laboratory, U-193, University of Connecticut, 3113 Horsebarn Road, Storrs, CT 06269-4193.
| |
Abbreviations |
|---|
Abbreviations used are: TEA, triethylamine; DAD, diode-array detection; CID, collision-induced dissociation; DFU, drug-free urine; SPE, solid-phase extraction; ELISA, enzyme-linked immunosorbent assay.
| |
References |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||