In vitro metabolism of ceftiofur in bovine tissues
Abstract
The metabolism of ceftiofur in bovine kidney, liver, muscle and lung, and the effects of the presence of cystine and glutathione in the media were evaluated using S-9 and microsomal tissue fractions. Conversion of ceftiofur to desfuroylceftiofur (DFC) was catalyzed by an esterase which was most active in kidney, followed by liver. It was not very active in muscle and lung. After DFC was liberated, it rapidly bound primarily to tissue proteins (>56%), and was also conjugated to cysteine and glutathione. Production of DFC-cysteine by disulfide exchange of DFC with cystine and production of DFC-glutathione by conjugation of DFC to glutathione occurred in buffer if glutathione and cystine were present in the medium. These conjugations were also observed in incubations with tissue fractions, indicating that they were not inhibited by the tissues endogenous molecules. In addition, the metabolism of DFC-glutathione to DFC-cysteine was observed when tissue proteins were present. The metabolism of DFC-glutathione to DFC-cysteine was faster in kidney than in liver. Metabolites devoid of an intact β-lactam ring were not observed in these in vitro studies.
INTRODUCTION
In the interest of consumer safety, drug sponsors and regulatory agencies worldwide require intricate knowledge of the metabolism processes of all drugs used in food producing animals. Traditionally, whole animal models have been used for characterization of the metabolism in the target species and in the species used for pathological and toxicological evaluations. Whole animal studies indicated that ceftiofur sodium (NAXCEL®/EXCENEL Sterile Powder) is rapidly metabolized to desfuroylceftiofur (DFC) and furoic acid in the target species and in the rat (Jaglan et al., 1989; Jaglan et al., 1992; Gilbertson et al., 1995; Beconi-Barker et al., 1995). In these metabolism studies DFC, the microbiologically active metabolite of ceftiofur, was never detected free, but was always found conjugated to proteins (70–90%), and to cysteine (DFC-cysteine) in tissues and in plasma. In one overdose study in swine Gilbertson et al. (1995) reported the presence of the glutathione conjugate of DFC (DFC-glutathione) in liver. DFC binding to biological molecules is reversible (Jaglan et al., 1989), thus, disulfides and protein conjugates can serve as reservoirs of DFC in plasma and tissue. Cysteine exists freely in plasma as its reduced disulfide linked dimer, cystine, and is highly susceptible to disulfide exchanges (Creighton, 1984). Formation of DFC-cysteine could result from the direct interaction of DFC with either cysteine or cystine or be produced through the metabolism of DFC-glutathione. The extent of the metabolism of ceftiofur by different tissues remains unknown. Metabolite production and rates of ceftiofur biotransformation in different tissues can be investigated through in vitro incubation with subcellular fractions and subsequent HPLC analysis. Though the relationship between chemical behaviour in subcellular systems and in the whole animal may seem distant, examinations of subcellular fractions such as S-9 and microsome fractions can yield information on discrete steps in biotransformations that may be impossible to determine in a whole animal (Crawford & Hutson, 1979). One disadvantage of incubations with S-9 fractions, which isolate endoplasmic reticulum, and membrane portions yielding a large system of enzymes, is that results can be affected by the interaction of previously isolated proteins. Microsomal fractions can eliminate this factor by isolating specific enzyme complexes such as the cytochromes P450 (Estabrook et al., 1979) which serve as the dominant oxidative complex necessary for the biotransformation of drugs.
This study investigated the rate and extent of metabolism of ceftiofur in different bovine tissues, the dependence of ceftiofur metabolism on S-9 and microsomal enzyme complexes and the effects of the presence of cystine and glutathione.
Materials and methods
Test materials
Ceftiofur sodium, radiolabelled with 14C at the 2-position of the thiazole ring, was synthesized at New England Nuclear, Du Pont, and provided as a dry powder. The radiochemical purity was determined by high performance liquid chromatography (HPLC) to be > 94.5%. The specific activity was 17.39 mCi/mmol. 14C-Ceftiofur was diluted with cold ceftiofur hydrochloride (Issue E, Pharmacia & Upjohn, Inc., Kalamazoo, MI, control reference standard) to achieve the final specific activity used in the incubations. Desfuroylceftiofur-cysteine disulfide (DFC-cysteine), desfuroylceftiofur-glutathione disulfide (DFC-glutathione), and 3,3′-desfuroylceftiofur-disulfide (DFC-dimer) were synthesized by C.L. Gatchell at Pharmacia and Upjohn Inc. following previously described techniques (Gilbertson et al., 1995). Desfuroylceftiofur (DFC), was synthesized by S.C. Olson at Pharmacia and Upjohn Inc. as follows: Fifty-four (54.0) grams of Ceftiofur hydrochloride standard, (Issue E, Pharmacia and Upjohn, Inc. control reference standard), 0.776 mL of 14C-ceftiofur sodium solution (specific activity 0.29 mCi/mL, Dupont NEN Products, Boston, MA, 94.5% pure), and 1 g dithioerythritol (Aldrich Chemical Company, Milawaukee, WI) were dissolved in 50 mL of 0.1 m borate buffer (38 g sodium borate, 7.9 g KCl, 2 L H2O, pH 9). The solution was left at room temperature (≈25° C) for 1 h with periodic mixing. DFC was separated from the reaction mixture by preparative HPLC using solvent system 2 (Instrumental Methods). Fractions were collected, pooled, neutralized with 600 μL of 10% ammonium hydroxide, diluted with water and lyophilized. The resulting powder was resuspended in 5 mL 0.025 m NaPO4 buffer, divided into 1 mL aliquots, assayed for purity and frozen.
S-9 preparations
Ten (10) grams of male bovine tissues (kidney, liver, lung or muscle) were homogenized in 20 mL 1.15% KCl buffer. The homogenate was transferred to centrifuge tubes and centrifuged at 9000×g for 20 min at 5°C. The supernatant was divided into 2 mL aliquots and stored at −80°C.
Microsome preparations
Bovine liver and kidney microsomal fractions were prepared according to previously described procedures (Imai & Sato, 1974). Briefly, male kidney and liver tissue were homogenized in 4× (weight/volume) ice cold 0.1 m KCl/10 mm EDTA, pH 7.4 buffer. The homogenate was transferred to centrifuge tubes and centrifuged at 10 000 g for 20 min at 5°C. The lipid layer was aspirated and the supernatant transferred to ultra-centrifuge tubes and centrifuged at 227 000×g for 40 min at 5°C. The supernatant was discarded and the microsomal pellet washed and re-homogenized in 0.1 mm Na2H2P2O7/1 mm EDTA buffer. The homogenate was ultra-centrifuged again as above. The remaining microsomal pellet was resuspended in KCl/EDTA buffer, divided into 1 mL aliquots and stored at −80°C. Before incubations, stock microsome solutions were diluted in 50 mm Hepes/0.1 mm EDTA buffer as follows: 500 μL kidney microsomes (9.6 mg protein/mL) in 2565 μL buffer or 330 μL liver microsomes (14.2 mg protein/mL) in 2670 μL buffer for a final concentration of 1.50 mg protein/mL.
Protein determination of S-9 and microsome fractions
Protein determination was done following previously described methods (Smith et al., 1985). Briefly, a standard curve was con-structed by diluting Bovine Serum Albumin (Sigma, St. Louis, MO) in 0.9% saline solution at a range of known concentrations from 100 to 2000 p.p.m. Dilutions of each S-9 and microsomal fraction were prepared in duplicate from 1/2 to 1/30. Bicinchoninic acid protein assay reagent (Pierce, Rockford, IL) was added to 10 μL samples of each solution and the optical densities read at 570 nm on a THERMOmax™ microplate reader (Molecular Devices Corporation, Sunnyvale, CA). An external standard calibration curve was generated from the calibration standards using log transformation. Unknown values were determined by comparing their absorbance to that of the external calibration standards.
P450 activity determination in microsomes
Previously established methods (Omura & Sato, 1964; Matsubara et al., 1976) were used to determine P450 content in each microsomal portion. Briefly, each microsomal sample was diluted 1/30 in 0.1 m phosphate, 0.1 mm EDTA, pH 7.4 buffer to ≈ 1 mg/mL. The resulting solution was reduced with sodium dithionite, vortexed, divided into 1 mL portions in glass cuvettes and placed in a spectrophotometer. After bubbling carbon monoxide through the sample, the absorbance spectrum was recorded and wavelength of maximum absorbance noted. Concentration of P450 cytochromes was determined from its extinction coefficient.
Incubations
Solutions and reagents. A cofactor solution containing 0.7638 g nicotinamide, 0.6331 g MgCl2, 1.333 g Na2 HPO4, and 2.130 g KH2 PO4 in 250 mL water was prepared for use in S-9 incubations. The pH of this solution was brought to 7.2 with 50% NaOH. Within 24 h of incubation, 0.0141 g glucose-6-phosphate and 0.0018 g NADP were dissolved in 10 mL of cofactor solution. The incubation buffer for microsomal incubations was prepared from stock solutions of 1 m Hepes, pH 7.4 (23.83 g Hepes in 100 mL water) and 100 mm EDTA, pH 7.4 (3.72 g EDTA in 100 mL water). One (1) mL of Hepes solution and 20 μL EDTA solution were brought to 20 mL of final volume in water. The NADPH generating solution contained 10 mm-NADP+ Na, 50 mm DL-isocitrate Na3, and 50 mm MgCl2. 6H2O, to which ≈ 4 units/mL isocitrate dehydrogenase was added immediately before incubation.
Preliminary incubations with ceftiofur in S-9 fractions. Preliminary incubations were performed to observe the effects of ceftiofur concentration and incubation time on ceftiofur metabolism by kidney, liver, lung and muscle S-9 fractions. All incubations were carried out in duplicate at two incubation times (10 and 40 min) and using two concentrations of ceftiofur (40 μm, 644 652 total decays per minute (DPM) and 1 mm, 295 790 total DPM). Each run also included duplicate blank vials (no drug) and controls (no S-9 or cofactor solution). As the esterase which converts ceftiofur to DFC is not very active in lung and muscle (M. Beconi-Barker, personal communication), only the 40 min incubation time was conducted for lung and muscle incubations. Each incubation vial contained 1 mm14C-labelled ceftiofur (295 790 total DPM) in a final incubation volume of 0.5 mL NADPH cofactor solution. Reactions were initiated with the addition of 4.0 mg protein. All vials were incubated at 37°C until reactions were terminated with an 80:20 acetonitrile:methanol solution, iced and centrifuged. Aliquots (200 μL) of each supernatant were analysed by HPLC system 1.
Rate of ceftiofur cleavage by S-9 fractions. The rate of ceftiofur cleavage was investigated in incubations with kidney and liver S-9 fractions. Incubation vials were prepared exactly as above. In kidney, duplicate reaction vials were quenched at 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 and 15 min after S-9 addition. From the preliminary incubations it was observed that liver metabolized ceftiofur more slowly, thus, liver incubation vials were quenched at 5, 10, 15, 20, 25, 30, 35, 40, 42.5, 45, 47.5, 50, 52.5, 55, 57.5 and 60 min after addition of S-9 protein.
Rate of ceftiofur cleavage in microsomal incubations. Incubations were carried out with both kidney and liver microsomes. Kidney microsome incubations contained excess substrate (14C-ceftiofur, ≈ 700 μm). Duplicate vials were incubated for 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 and 15 min. Liver microsomes were incubated with ≈ 4 μm ceftiofur in duplicate vials for 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 and 15 min. Each run also included duplicate blank vials (no drug) and controls (no microsomes or NADPH solution). Each microsomal incubation contained 0.16 mg microsomal protein in 0.3 mL of incubation buffer. After preincubation for 2 min at 37°C, reactions were initiated with the addition of 0.1 volume (50 μL) of prewarmed NADPH and 100 μL of 5 mm14C-ceftiofur solution. This resulted in a final incubation volume of 0.5 mL and 14C-ceftiofur concentration of 1 mm. Reactions were terminated with 0.5 mL of 80:20 acetonitrile:methanol, iced and centrifuged. Aliquots (200 μL) of each vial were analys ed by HPLC system 1.
Effects of cystine and glutathione. Production of the cysteine and glutathione conjugates of ceftiofur is dependent on the presence of cystine and glutathione in the environment. Their effect was evaluated in S-9 liver and kidney incubations containing cystine and glutathione. Cystine, rather than cysteine, was chosen for addition in incubations based on its presence in tissue as a free dimer, susceptible to disulfide exchanges. Glutathione is naturally present in tissue. Liver and kidney S-9 and microsomal fractions were incubated in duplicate with either 1 mm14C-ceftiofur or ≈ 1 mm14C-DFC (specific activity: 9269 046 DPM/mg) and with additions of 1 mm cystine, 1 mm glutathione or 1 mm cystine + glutathione. All vials were incubated for 40 min at 37°C (except for kidney vials for which an additional time point of 10 min was added) and quenched with 0.5 mL 80:20 acetonitrile:methanol solution. Each run included duplicate blanks (no drug) and duplicate controls of drug with cystine, glutathione, and cystine + gluathione (no S-9 or microsomal protein, no NADPH generating system). Incubations were prepared as described in the Preliminary incubations and Rate of ceftiofur cleavage in microsomal fractions sections.
Instrumental analysis
HPLC methods. Two HPLC systems were used. System 1 consisted of a Waters 600 powerline multi solvent delivery system (Waters, Milford, MA), controller and pump equipped with a Waters 717 Plus autosampler and a Waters 996 photodiode array detector (254 nm), and a Radiomatic FLO-ONE/Beta Model A-500 radioactive flow detector (Packard, Downers Grove, IL). A Keystone Scientific BDS Hypersil C18 column (5 μm, 250×4.6 mm) (Keystone Scientific, Bellefonte, PA) was used. Mobile phases consisted of A, 0.1% trifluoroacetic acid (TFA) in water and B, 0.1% TFA in acetonitrile. Conditions were isocratic (0% B) from 0 to 5 min; the gradient increased linearly 0–35% B from 5 to 40 min, 35–50% B from 40 to 50 min, and 50–80% from 50 to 50.10 min; conditions remained isocratic (80% B) from 50.10 to 55 min. The flow was 1 mL/min. The column was equilibrated with the starting conditions for 25 min before the next injection.
System 2 consisted of a Waters Delta Prep 4000 equipped with a Waters 440 Absorbance Detector with a 254 nm filter. A Keystone Scientific BDS Hypersil C18 column (5 μm, 250×20 mm) was used. Mobile phases were the same as in System 1. Conditions were isocratic (10% B) from 0 to 5 min; the gradient increased linearly 10–40% B from 5 to 65 min. The flow was 10 mL/min. Fractions were collected at 30 second intervals beginning with gradient flow.
Gel electrophoresis. Protein portions were isolated through filtration with an Amicon® Centricon-10 concentrator (Amicon Inc., Beverly, MA) from the corresponding incubations. Protein identification was performed using a NOVEX Tricine minigel (NOVEX, San Diego, CA) with a 10–20% acrylamide gradient. Samples were prepared by diluting unknown protein portions 1:1 in NOVEX Tricine SDS Sample Buffer (3 mL 3.0 m Tris HCl, pH 8.45, 2.4 mL glycerol, 0.8 g SDS, 1.5 mL 0.1% Coomassie Blue G, 0.1% phenol red, distilled water to 10 mL) and heating for 2 min at 85°C. Samples were loaded to the gel in 20 μL aliquots. NOVEX Tricine Running Buffer (121 g Tris Base, 179 g Tricine, 10 g SDS, distilled water to 1 L) was diluted 1:9 and loaded to the upper and lower buffer chambers of the gel. The gel was run at a constant voltage of 125 V for ≈ 90 min, with a starting current of 80 mA/gel and an ending current of 40 mA/gel. An image of the gel was taken in a 48 h exposure using Biomax film (Eastman Kodak Co., Rochester, NY).
Determination of the rate of ceftiofur cleavage/desfuroylceftiofur formation
A general linear model of the form:
analyte concentration (μg/mL) = β0+β1×time (min)
where the parameters β0 and β1 were estimated using least squares, was used to fit the data from ceftiofur incubations over time in microsomal fractions. A lack of fit test failed to reject the hypothesis that the model was appropriate (P > 0.65 for kidney and P > 0.96 for liver). All calculations were done using the SAS statistical software package (SAS, 1989).
Results
S-9 incubations with ceftiofur
Kidney. Chromatograms of kidney S-9 fractions incubated with low (40 μm) concentrations of ceftiofur indicated the presence of two peaks (Fig. 1A) after separation by high performance liquid chromatography coupled with radioactive monitoring (HPLC-RAM). Peak A, which accounted for 73–74% of the 14C activity eluting from the HPLC, had an early retention time of 3–4 min. Approximately 14–16% of the 14C activity eluting from the column was found in peak B at ≈ 30 min, which corresponded by retention time to DFC-cysteine.
. HPLC-RAM chromatogram after incubation of ceftiofur in kidney S-9 fractions: (A) 40 μm initial concentration of ceftiofur, incubated for 40 min, and (B) 1 mm initial concentration of ceftiofur, incubated for 10 min.
After incubations with higher ceftiofur concentrations (1 mm) two additional peaks were observed (Fig. 1B, peaks C and D at 36 and 41 min, respectively) which corresponded by retention times to DFC and DFC-dimer.
Peak A was isolated by filtering a pooled source of samples through an Amicon® centricon unit with a molecular cut-off weight of 10 000. Analysis of the filtrate portion of the sample by HPLC indicated that it did not contain the early eluting peak. As peak A was retained by the filter, it had a molecular weight > 10 000. Gel electrophoresis indicated that it corresponded to radiolabelled material bound to unidentified high molecular weight species (Fig. 2A). This peak will be referred to as DFC-macromolecule. Further analysis of the gel indicated the presence of different macromolecule fractions in kidney and liver. In kidney, a band was seen between ≈ 6000–17 000 kDa and 17 000–22 000 kDa, while in liver, two bands were found, one between 6000 and 17 000 kDa and the second one in the region of 60 000 kDa. It is speculated that the high molecular weight species is formed by DFC conjugated to tissue proteins/peptides not precipitated by the acetonitrile:methanol solution, and/or by polymerized DFC (previously identified as degradation product, Pharmacia & Upjohn Inc, unpublished results). Analysis of the macromolecule fraction obtained from incubations in which the proteins were not precipitated (no quenching with methanol:acetonitrile was performed), indicated that the radio-labelled material was bound to all the protein bands observed after gel-electrophoresis and Coomasie Blue staining in kidney and liver fractions (Fig. 2B). The intensity of the 14C-label led bands was in agreement with the intensity observed in the protein bands of the Coomassie Blue stained gel. Concentrations of metabolites resulting from ceftiofur incubations in S-9 incubations are summarized in Table 1.
. (A) One-dimensional gel electrophoresis of the early eluting peak in the HPLC-RAM chromatograms, or ‘protein’ fraction that did not precipitate with acetonitrile:methanol. (B) One dimensional gel electrophoresis of the total protein fraction present in the S-9 fractions of bovine kidney and liver after incubation with 14C-ceftiofur.
Liver . HPLC-RAM chromatograms of liver S-9 incubations at low (40 μm) ceftiofur concentrations showed results similar to those for kidney (Fig. 3A). Peak A accounted for 56–68% of 14C activity eluting from the column. Peak B, corresponding by retention time to DFC-cysteine, represented 10–11% of 14C activity. Incubations with higher concentrations (1 mm) of ceftiofur indicated the presence of five peaks at all incubation times (Fig. 3B). Peak A, which accounted for 26–30% of the radiolabelled activity eluting from the HPLC, corresponded to ‘macromolecules’. Peak B (≈ 5%) corresponded by retention time to DFC-cysteine. Approximately 7% of 14C activity eluting from the HPLC column was found in peak C, corresponding to DFC. Peaks D and E corresponded by retention time to DFC-dimer and unmetabolized parent CF and accounted for 12–19% and 23–24% of the eluting 14C activity. Trace amounts of DFC-glutathione were observed in some incubations. The DFC-glutathione peak can be better resolved in this chromatographic system if the quenching solution (80:20 acetonitrile:methanol) is dried under nitrogen before chromatography. As only trace amounts of DFC-glutathione were observed, and drying and resuspending the sample can degrade the labile ceftiofur metabolites, an attempt to precisely quantify DFC-glutathione was not made. The presence of a large proportion of remaining parent 14C-ceftiofur up to 40 min of incubation in liver S-9 fractions and their absence in kidney incubations suggested that cleavage of ceftiofur to DFC in liver occurred at a slower rate than kidney.
. HPLC-RAM chromatogram after incubation of ceftiofur in liver S-9 fractions: (A) 40 μm initial concentration of ceftiofur, incubated for 40 min, and (B) 1 mm initial concentration of ceftiofur, incubated for 10 min.
Lung and Muscle. HPLC-RAM chromatograms of ceftiofur incubations with lung and muscle S-9 fractions revealed little evidence of ceftiofur cleavage after 40 min. The dominant radioactive peak present in HPLC-RAM chromatograms obtained from lung and muscle incubations corresponded to parent CF and represented 75–87% of 14C activity eluting from the HPLC column. The presence of a small ‘macromolecule’ peak eluting at an early retention time was indicated in both lung and muscle. Muscle incubations also indicated the presence of a peak corresponding by retention time to DFC, accounting for ≈ 6% of 14C activity eluting from the column.
Ceftiofur cleavage in S-9 fractions
Kidney. Ceftiofur concentration decreased steadily over time until ≈ 7–8 min, after that time, parent ceftiofur concentration became so low that it was no longer quantifiable (Fig. 4). The concentration of DFC increased steadily during the first few minutes of incubation. After ≈ 5 min, the concentration of DFC started to decrease, consistent with its subsequent conjugation to DFC-cysteine and dimerization. The larger proportion of DFC-dimer formed during these incubations can be explained by the fact that there was a limited supply of cystine necessary for the production of DFC-cysteine. The incubation medium was not enriched with either cystine or glutathione, only endogenous cystine and glutathione remaining in S-9 fractions after preparation was present. Once DFC is formed, it rapidly binds to other molecules containing disulfide bonds or sulfhydrile groups, such as cystine and glutathione or to itself. DFC-glutathione in kidney is rapidly metabolized to DFC-cysteine (see Effects of cystineandglutathione below), and thus, probably not detected in these incubations. The amount of DFC-cysteine that can be formed will be determined by the amount of cystine and glutathione present in the S-9 fractions After those sources have been depleted, DFC will only be able to bind to itself. A similar pattern was observed in liver, where the concentration of ceftiofur decreased steadily with time after incubation with liver S-9 fractions (Fig. 5). The concentration of DFC increased steadily during the earlier incubation times. After ≈ 45 min the concentration of DFC showed a steady decrease with time, consistent with its subsequent conjugation to DFC-cysteine, DFC-glutathione and dimerization. The concentration of DFC-glutathione increased up to ≈ 40–45 min, and then started to decrease, consistent with its metabolization to DFC-cysteine (see Effects of cystineandglutathione below).
. Concentration of ceftiofur, DFC, DFC-cysteine and DFC-dimer over time after ceftiofur incubations in bovine kidney S-9 fractions. (The arrow indicates the time at which parent ceftiofur became limiting)
. Concentration of ceftiofur, DFC, DFC-cysteine, DFC-glutathione and DFC-dimer over time after ceftiofur incubations in bovine liver S-9 fractions.
Rate of ceftiofur cleavage in microsomes
As glutathione and cystine exist naturally in S-9 fractions, an environment free of cystine and glutathione was needed for examination of the rate of ceftiofur cleavage/desfuroylceftiofur formation. This environment was provided by microsomal fractions, which contain limited or no cystine or glutathione but contain the esterase responsible for the cleavage of ceftiofur. The rates of ceftiofur cleavage and desfuroylceftiofur formation in incubations with microsomal proteins showed a linear dependence on time in the presence of excess ceftiofur. When incubated in the presence of kidney microsomal proteins, ceftiofur was rapidly cleaved to DFC (Fig. 6). After 15 min, the concentration of ceftiofur became rate limiting. The rate of cleavage was approximately −0.54 μmol CE/mL min. Conversely, the rate of DFC formation was ≈ 0.43 μmol CE/mL min. The numerical difference observed between these rates was probably due to experimental error as the 95% confidence interval of these difference included the zero value. When incubated in the presence of liver microsomal proteins, the rate of cleavage of ceftiofur to DFC was approximately −2.6×10−3μmol CE/mL min (Fig. 7). Conversely, the rate of DFC formation was ≈ 1.4×10−3μmol CE/mL min. Under these conditions, the liver microsomal system degraded before ceftiofur became rate limiting. These rates were ≈ 300 times slower than those obser ved in kidney. The numerical difference observed between these rates was probably due to experimental error as the 95% confidence interval of the difference included the zero value.
. Concentration of ceftiofur (A) and desfuroylceftiofur (B) over time after ceftiofur incubations in bovine kidney microsomal fractions.
. Concentration of ceftiofur (A) and desfuroylceftiofur (B) over time after ceftiofur incubations in bovine liver microsomal fractions.
The conversion from ceftiofur to DFC was independent of the presence of NADPH and, thus, of P450 enzymes, and was catalyzed by an esterase. As the activity of this esterase in kidney was so much higher than in liver, it is thought that this esterase was present in the microsomal fractions and was not due to contamination from esterases present in the plasma.
Effects of cystine and glutathione
Incubations in the absence of proteins . After incubating DFC for 40 min in the absence of proteins (no S-9 or microsomes), most of the 14C-material was found in the form of unreacted DFC. Small percentages of DFC-dimer were also formed (15–16%). When the medium was enriched with glutathione, DFC-glutathione (peak A, 23%) and DFC-dimer (peak B, 7%) were the only DFC-metabolites present (Fig. 8). DFC-cysteine was not observed in these incubations. When the medium was enriched with cystine, the formation of DFC-cysteine (Peak B, 14–15%), DFC-dimer (Peak C, trace), and a ‘macromolecule’ component (Peak A, 50–51%) were observed (Fig. 9). The origin of the high molecular weight material was attributed to possible polymerized DFC and/or DFC binding to impurities in the added L-cystine or cystine polymers. Upon enrichment with a mixture of both cystine and glutathione, the formation of DFC-glutathione, DFC-cysteine, DFC-dimer and the ‘macromolecule’ component were observed.
. HPLC-RAM chromatogram after incubation of desfuroylceftiofur in buffer enriched with glutathione (1 mm) in the absence of tissue proteins.
. HPLC-RAM chromatogram after incubation of desfuroylceftiofur in buffer enriched with cystine (1 mm) in the absence of tissue proteins.
Incubations with S-9 protein. HPLC-RAM chromatographs of kidney and liver incubations of ceftiofur with glutathione in S-9 fractions indicated the formation of the DFC-glutathione conjugate. In kidney, DFC-glutathione was observed only in the earlier incubation times (10 min). Its absence at longer incubation times (40 min), indicated that it had already been metabolized to DFC-cysteine. In liver, formation of DFC-glutathione and DFC-cysteine were observed at both incubation times, with the addition of glutathione. On the average, more DFC-cysteine was observed when the medium was enriched with glutathione than in the controls indicating that DFC-cystine was a product from the metabolism of DFC-glutathione (Table 2).
The lack of peaks which characterize the presence of DFC-glutathione in the 40 min incubations with kidney S-9 indicated that DFC-glutathione was converted to DFC-cysteine faster in the kidney than in the liver. Production of DFC-glutathione, DFC and DFC-bound to ‘macromolecule’ occurred also in lung and muscle S-9 fractions when glutathione was added to the media. Most of the radiolabelled activity remained as unmetabolized parent ceftiofur. Metabolism of DFC-glutathione to DFC-cysteine was not observed.
HPLC-RAM chromatograms of kidney and liver incubations with ceftiofur and cystine in S-9 fractions indicated the formation of the DFC-cysteine conjugate. On the average, more DFC-cysteine was formed when cystine and/or glutathione was added to the incubation than in incubations with ceftiofur alone (Tables 1 and 2) (3.48 μg/mL vs. 0.48 μg/mL for kidney and 1.66 μg/mL vs. 0.78 μg/mL for liver). Trace amounts of DFC-cysteine were formed in incubations with lung and muscle S-9 fractions. This data indicated that in tissue extracts, DFC-cysteine is not only formed by metabolism of DFC-glutathione but also by disulfide exchange of DFC with cystine.
HPLC-RAM chromatograms of kidney and liver incubations of ceftiofur with cystine plus glutathione in S-9 fractions revealed the presence of a ‘macromolecule’ peak, accounting for 37–61% of 14C-activity eluting from the HPLC column. This ‘macromolecule’ peak was larger on the average than the ‘macromolecule’ peak observed when either cystine or glutathione were added. The presence of peaks corresponding to both DFC-glutathione and DFC-cysteine were observed in most incubations. Formation of a large ‘macromolecule’ peak was also observed in incubations with lung and muscle S-9; the presence of DFC-cysteine and DFC-glutathione was also indicated. The concentration of metabolites produced after incubating 1 μm concentrations of ceftiofur in S-9 fractions in the presence of cystine and/or glutathione is presented in Table 2.
Discussion
Incubations with microsome and S-9 subcellular fractions obtained from various bovine tissues demonstrated that the conversion of ceftiofur to DFC occurs independently of P450 enzyme complexes. Ceftiofur conversion to DFC in incubations without the presence of enzyme-enabling NADPH yielded evidence that this conversion is esterase catalyzed. The esterase responsible for the biotransformation of ceftiofur to DFC is most active in kidney, followed by liver. It is not very active in muscle and lung. After DFC is liberated, it rapidly binds primarily to tissue proteins (>56%). These results are in agreement with previous observations in vivo (Beconi-Barker et al., 1995; Gilbertson et al., 1995) where over 65% of the 14C-activity was associated with tissue proteins. The nature of the binding of DFC to proteins is being characterized. In addition, DFC was found dimerized or conjugated to cysteine or glutathione.
Production of DFC-cysteine by disulfide exchange of DFC with cystine and production of DFC-glutathione by conjugation of DFC to glutathione occurred in buffer if glutathione and cystine were present in the medium. These conjugations were also observed in incubations with tissue fractions, indicating that they were not inhibited by the tissues endogenous molecules. In addition, the metabolism of DFC-glutathione to DFC-cysteine was observed when tissue proteins were present. The metabolism of DFC-glutathione to DFC-cysteine was faster in kidney than in liver.
Polar metabolites previously reported in urine of animals treated with ceftiofur (Jaglan et al., 1989; Beconi-Barker et al., 1995; Gilbertson et al., 1995) were not observed in these in vitro studies. Polar metabolites are probably DFC degradation products formed by microorganisms in the non sterile collected urine. However, the possibility that polar metabolites are a product of the intricate mammalian metabolism cannot be excluded.
