Reactivity of diflunisal acyl glucuronide in human and rat plasma and albumin solutions

New Perspectives on Drug-Induced Liver Injury Risk Assessment of Acyl Glucuronides

Drug-induced liver injury (DILI) remains one of the key challenges in drug development due to the mechanisms of action being multifactorial in nature. This is particularly the case for idiosyncratic DILI which occurs in a very low frequency in humans (e.g., 1:10,000). Despite perceptions that acyl glucuronide metabolites are defacto risks for DILI, scientific evidence suggests that acyl glucuronide formation alone does not pose an increased risk compared to other drug metabolites. This applies in particular to those acyl glucuronides which are not reactive and do not form covalent adducts with proteins. The goal of this paper is to provide guidance on preclinical and clinical strategies to evaluate the potential for acyl glucuronide formation to contribute to DILI. A key element of our proposed safety assessment is to investigate whether a particular acyl glucuronide is reactive or not and whether systemic exposure in humans can be demonstrated in animal toxicology studies following administration of the parent drug. While standard animal toxicology studies can identify overtly hepatotoxic compounds, these studies are not predictive for drugs that produce idiosyncratic forms of DILI. In addition, we do not recommend conducting toxicology studies of administered individual acyl glucuronides due to differences in pharmacokinetic and dispositional properties from the endogenously produced metabolites. Once a drug candidate has entered clinical trials, the focus should be on clinical safety data and emerging risk-benefit analysis.

Metabolic activation of carboxylic acids

BACKGROUND

Carboxylic acids constitute a large and heterogeneous class of both endogenous and xenobiotic compounds. A number of carboxylic acid drugs have been associated with adverse reactions, linked to the metabolic activation of the carboxylic acid moiety of the compounds, i.e., formation of acyl-glucuronides and acyl-CoA thioesters.

OBJECTIVE

The objective is to give an overview of the current knowledge on metabolic activation of carboxylic acids and how such metabolites may play a role in adverse reactions and toxicity.

METHODS

Literature concerning the formation and disposition of acyl glucuronides and acyl-CoA thioesters was searched. Also included were papers on the chemical reactivity of acyl glutathione-thioesters, and literature concerning possible links between metabolic activation of carboxylic acids and reported cellular and clinical effects.

RESULTS/CONCLUSION

This review demonstrates that metabolites of carboxylic acid drugs must be considered chemically reactive, and that the current knowledge about metabolic activation of this compound class can be a good starting-point for further studies on the consequences of chemically reactive metabolites.

Drug bioactivation, covalent binding to target proteins and toxicity relevance

A number of therapeutic drugs with different structures and mechanisms of action have been reported to undergo metabolic activation by Phase I or Phase II drug-metabolizing enzymes. The bioactivation gives rise to reactive metabolites/intermediates, which readily confer covalent binding to various target proteins by nucleophilic substitution and/or Schiff's base mechanism. These drugs include analgesics (e.g., acetaminophen), antibacterial agents (e.g., sulfonamides and macrolide antibiotics), anticancer drugs (e.g., irinotecan), antiepileptic drugs (e.g., carbamazepine), anti-HIV agents (e.g., ritonavir), antipsychotics (e.g., clozapine), cardiovascular drugs (e.g., procainamide and hydralazine), immunosupressants (e.g., cyclosporine A), inhalational anesthetics (e.g., halothane), nonsteroidal anti-inflammatory drugs (NSAIDSs) (e.g., diclofenac), and steroids and their receptor modulators (e.g., estrogens and tamoxifen). Some herbal and dietary constituents are also bioactivated to reactive metabolites capable of binding covalently and inactivating cytochrome P450s (CYPs). A number of important target proteins of drugs have been identified by mass spectrometric techniques and proteomic approaches. The covalent binding and formation of drug-protein adducts are generally considered to be related to drug toxicity, and selective protein covalent binding by drug metabolites may lead to selective organ toxicity. However, the mechanisms involved in the protein adduct-induced toxicity are largely undefined, although it has been suggested that drug-protein adducts may cause toxicity either through impairing physiological functions of the modified proteins or through immune-mediated mechanisms. In addition, mechanism-based inhibition of CYPs may result in toxic drug-drug interactions. The clinical consequences of drug bioactivation and covalent binding to proteins are unpredictable, depending on many factors that are associated with the administered drugs and patients. Further studies using proteomic and genomic approaches with high throughput capacity are needed to identify the protein targets of reactive drug metabolites, and to elucidate the structure-activity relationships of drug's covalent binding to proteins and their clinical outcomes.

Covalent binding of 2-phenylpropionyl-S-acyl-CoA thioester to tissue proteins in vitro

In this study, we investigated the possible involvement of acyl-CoA, reactive intermediary metabolites of 2-arylpropionic acids (profens), in protein adduct formation in rat liver homogenate and in human serum albumin (HSA) in buffer. (RS)-[1-14C]-2-Phenylpropionic acid (14C-2-PPA, 1 mM) was incubated with rat liver homogenate (1.5 mg/ml) in the presence of cofactors of acyl-CoA formation (Mg2+, ATP, and CoA). Aliquots of the incubation mixture were analyzed for covalent binding and acyl-CoA formation over a 3-h period. High-performance liquid chromatographic analysis of the products from such incubations showed the presence of 2-phenylpropionyl-S-acyl-CoA (2-PPA-CoA), which was confirmed by coelution with authentic 2-PPA-CoA, as well as by mass spectrometry. In the same incubations, 2-PPA was shown to bind covalently to hepatic proteins in a time- and ATP-dependent fashion. Inhibition of 2-PPA-CoA formation by acyl-CoA synthetase inhibitors, such as palmitic acid, lauric acid, octanoic acid, and ibuprofen, markedly decreased the extent of covalent binding of 2-PPA to hepatic proteins. Results from these in vitro studies strongly suggest that acyl-CoA thioester derivatives are chemically reactive and are able to bind covalently to tissue proteins in vitro, and, therefore, may contribute significantly to covalent adduct formation of profen drugs in vivo.

Interaction of human serum albumin with furosemide glucuronide: a role of albumin in isomerization, hydrolysis, reversible binding and irreversible binding of a 1-O-acyl glucuronide metabolite

Furosemide 1-O-acyl glucuronide (Fgnd) was reversibly bound to a single class of binding sites on human serum albumin (HSA), and the binding of Fgnd decreased with increasing F concentrations, suggesting that Fgnd binds to the same warfarin binding sites on HSA as F binds. The rate of Fgnd degradation (hydrolysis and acyl migration) decreased in the presence of HSA. Although the formation of acyl migration isomers of Fgnd was slower in the presence of HSA than in its absence, hydrolysis of Fgnd to F was faster in the presence of HSA. Rapid minor irreversible binding of Fgnd to HSA within 30 min was followed by slow major irreversible binding. Slow irreversible binding of Fgnd to HSA was decreased by F, though not significantly. This suggests that major irreversible binding may proceed via reversible binding. It has been reported that acyl migration is a prerequisite for irreversible binding. Therefore, these results indicate that HSA decreases irreversible binding of Fgnd to protein by suppressing acyl migration. Furthermore, these results suggest that HSA may prevent irreversible binding of Fgnd to other proteins in the body by decreasing the concentration of reactive Fgnd in the unbound form. HSA eliminates reactive Fgnd by hydrolysis to F. Therefore, it is concluded that HSA works as a scavenger to decrease reactive compounds by reversible binding or eliminates reactive compounds by irreversible binding.

Hepatobiliary transport of diflunisal conjugates and taurocholate by the perfused rat liver: the effect of chronic exposure of rats to diflunisal

Acyl glucuronides are reactive electrophilic metabolites of carboxylate drugs which can form covalent adducts with endogenous macromolecules such as serum albumin and hepatic proteins. Such adducts have been suggested as initiating factors in certain immune and toxic responses to acidic drugs. In the present study, pretreatment of rats with high daily doses (50 mg/kg orally) of the non-steroidal anti-inflammatory drug (NSAID) diflunisal (DF) for 35 days, followed by perfusion of the isolated liver with 3 mg DF for 3 hr, resulted in appreciable concentrations of covalent adducts of DF with hepatic tissue (3.68 microg DF/g liver). Immunoblotting using a rabbit polyclonal DF antiserum showed the major DF-modified bands at about 110, 140 and 200 kDa. A vehicle-pretreated control group achieved adduct concentrations of only 0.37 microg DF/g liver, with the 200 kDa band not detectable in immunoblots. Elimination of DF from perfusate of the isolated perfused rat liver (IPRL) preparation was the same (t1/2 about 3.4 hr) in both DF- and vehicle-pretreated groups. Appearance of the sulfate (DS) conjugate, the major metabolite in perfusate, was also similar. However, higher concentrations of the acyl glucuronide (DAG) and phenolic glucuronide (DPG) conjugates were found in perfusate at later times, though a statistically significant difference in area under the concentration-time curve was found only in the case of DAG. At 3 hr, recoveries of dose as DAG and DPG were significantly higher in perfusate, but not in bile. No significant differences in uptake and biliary excretion of taurocholate were found between the two groups. The finding of higher perfusate concentrations of DAG and DPG could signal a minor compromise to biliary excretion processes for the glucuronides, though whether such a result is simply coincident with or attributable to DAG-derived covalent DF-protein adducts in liver remains indeterminate.

Direct detection of the internal acyl migration reactions of benzoic acid 1-O-acylglucuronide by 13C-labeling and nuclear magnetic resonance spectroscopy

1-O-Acyl-beta-D-glucopyranuronates can undergo irreversible binding to proteins mainly through internal acyl migration reactions, which may have toxicological significance. A new method based on the 13C-labeling and nuclear magnetic resonance (NMR) spectroscopy has been developed to study the reactivity of the 1-O-acyl-beta-D-glucopyranuronate of benzoic acid. In phosphate buffer (pH 7.4) solution at 37 degrees C, the glucuronide showed apparent first-order degradation kinetics (T1/2, 125 min), and concurrent and sequential appearance of 2-, 3- and 4-O-acyl isomers as both alpha- and beta-anomers was observed. The isomeric glucuronides were identified by two-dimensional NMR of the reaction mixture. The direct approach using 13C-labeling and NMR could also provide insights into the reactivities of other labile drug acylglucuronides and their isomeric glucuronides.

Drug-protein binding sites. New trends in analytical and experimental methodology

In the last few years, continuous progress in instrumental analytical methodology has been achieved with a substantial increase in the number of new, more specific and more flexible methods for ligand-protein assays. In general, the methods used for drug-protein binding studies can be divided into two main groups: separation methods (enabling the calculation of binding parameters, i.e. the number of binding sites and their respective affinity constants) and non-separation methods (describing predominantly qualitative parameters of the ligand-protein complex). This review will be focussed particularly on recent trends in the development of drug-protein binding methods including stereoselective and non-stereoselective aspects using chromatography, capillary electrophoresis and microdialysis as compared to the "conventional approach" using equilibrium dialysis, ultrafiltration or size exclusion chromatography. The advantages and limitations of various methods will be discussed including a focus on "optimal" experimental strategies taking into account in vitro, ex vivo and/or in vivo studies. Furthermore, the importance of some particular aspects concerning the drug binding to proteins (covalent binding of drugs and metabolites, stereoselective interactions and evaluation of binding data) will be outlined in more detail.

Disposition and covalent binding of ibuprofen and its acyl glucuronide in the elderly

Ibuprofen is an over-the-counter nonsteroidal anti-inflammatory drug with a low incidence of severe adverse reactions. It is metabolized by oxidation to carboxyibuprofen and hydroxyibuprofen and by conjugation to an acyl glucuronide. In vitro studies have indicated that ibuprofen glucuronide is labile and reactive, forming covalent adducts with proteins. To verify the formation of ibuprofen-protein adducts in vivo, the pharmacokinetics of ibuprofen glucuronide and its covalent binding to plasma proteins were studied in five elderly patients who received long-term administration of oral doses of ibuprofen. Plasma levels of ibuprofen glucuronide were low relative to those of ibuprofen; the ratio of area under the plasma concentration versus time curve for the glucuronide relative to the parent drug was only 4%. Covalent binding of ibuprofen to plasma protein was observed in all patients, correlating well with the area under the plasma concentration versus time curve of ibuprofen glucuronide (r = 0.966). Compared with reports for other nonsteroidal anti-inflammatory drugs that form acyl glucuronides, plasma levels of ibuprofen-protein adduct are low during long-term administration. The observed lower reactivity in vivo is probably attributable to the greater stability of ibuprofen glucuronide relative to other acyl glucuronides.

Covalent binding of suprofen to renal tissue of rat correlates with excretion of its acyl glucuronide

1. Dosing rat with suprofen produces suprofen equivalents that are covalently bound to plasma and tissue proteins in vivo. 2. Suprofen acyl glucuronide is reactive in vitro, resulting in suprofen equivalents covalently bound to proteins of plasma and tissues in a time-dependent manner. 3. Bile duct ligation of rat increases exposure to suprofen acyl glucuronide in vivo, which leads to enhanced covalent binding of suprofen equivalents to plasma proteins and to kidney tissue. 4. Covalent binding of suprofen equivalents to kidney tissue correlates with excretion of suprofen and suprofen glucuronide by the kidney.

Rat serum albumin modified by diflunisal acyl glucuronide is immunogenic in rats

Acyl glucuronide metabolites of carboxylic acid drugs such as the salicylate derivative diflunisal (DF) have been shown to react with proteins in vitro and in vivo to produce covalent adducts. Such attachment of foreign compounds to endogenous molecules could be associated with toxic and/or immune consequences. In this study we have injected rats with rat serum albumin (RSA) modified (a) by DF using a carbodiimide reagent (-->DF-RSA-I, 4.9 micrograms DF/mg RSA) and (b) by incubation with DF acyl glucuronide (DAG) and its rearrangement isomers (iso-DAG) (-->DF-RSA-II, 0.34 micrograms DF/mg RSA). All of the six rats injected with DF-RSA-I produced antibodies reactive with DF-modified keyhole limpet hemocyanin (KLH), the coating protein used in the ELISA. Three out of six animals injected with DF-RSA-II generated similar antibodies. Cross-reactivity with other non-steroidal anti-inflammatory drugs (NSAIDs) such as naproxen and ketoprofen (as the free drugs) was not observed. This study shows that a self protein covalently modified by incubation with DAG and iso-DAG is immunogenic in rats. The data thus support the hypothesis that covalent modification of macromolecules by acyl glucuronide metabolites of acidic drugs in vivo can lead to the production of circulating antibodies which may be involved in aberrant immune responses such as drug hypersensitivity.

Studies on the reactivity of acyl glucuronides--VIII. Generation of an antiserum for the detection of diflunisal-modified proteins in diflunisal-dosed rats

Acyl glucuronide metabolites of carboxylic drugs such as the salicylate derivative diflunisal (DF) have been shown to react with proteins to produce covalent adducts. To aid in the study of the formation and distribution of these adducts in both humans and rats, we raised an antiserum against human serum albumin modified by covalent attachment of DF via an amide bond, using a carbodiimide reagent. This antiserum had wide reactivity, reacting with all types of DF-modified proteins tested and with free DF (albeit at a lower affinity). It did not cross-react with other salicylates or other non-steroidal anti-inflammatory drugs. The antiserum has been used in immunoblotting to detect proteins covalently modified by DF in the plasma and livers of rats treated with the drug for 7 days. Although some cross-reactivity was apparent on the blots, a series of DF-modified proteins was found in cytosolic, mitochondrial and mixed membrane fractions of hepatocytes, with molecular weights ranging from 28 to 130 kDa.

Studies on the reactivity of acyl glucuronides--VI. Modulation of reversible and covalent interaction of diflunisal acyl glucuronide and its isomers with human plasma protein in vitro

Acyl glucuronide conjugates are chemically reactive metabolites which can undergo hydrolysis, rearrangement (isomerization via acyl migration) and covalent binding reactions with protein. The present study was undertaken to identify factors modulating the reactivity of diflunisal acyl glucuronide (DAG) with human serum albumin (HSA) in vitro, by comprehensively evaluating the interplay of the three pathways above when DAG and a mixture of its 2-, 3- and 4-isomers (iso-DAG) were incubated with protein. Buffer, plasma, fraction V HSA, fatty acid-free HSA, globulin-free HSA and fatty acid- and globulin-free HSA were investigated at pH 7.4 and 37 degrees, each in the absence and presence of warfarin, diazepam and diflunisal (DF) as reversible binding competitors. DAG and iso-DAG were highly reversibly bound (ca. 98-99.5%) in plasma and HSA solutions. The binding was primarily at the benzodiazepine site, since displacement occurred in the presence of diazepam and fatty acids but not warfarin. DAG degradation, via rearrangement, hydrolysis and covalent adduct formation (in that order of quantitative importance), was retarded in plasma and HSA solutions compared to buffer. The protective effect of protein was afforded by the high reversible binding to the (non-catalytic) benzodiazepine site. The warfarin site appeared to be catalytic for DAG hydrolysis, whereas rearrangement appeared to be hydroxide ion-catalysed only. In contrast to DAG, iso-DAG degradation was greatly accelerated in the presence of protein, through both covalent binding and catalysis of hydrolysis. Covalent binding via DAG was increased in the presence of warfarin but decreased in the presence of diazepam, DF and fatty acids. The opposite effects were found for covalent binding via iso-DAG. The data suggest that covalent binding of DF to HSA via DAG and iso-DAG occurs by different mechanisms (presumably transacylation and glycation, respectively) at different sites (benzodiazepine and warfarin, respectively) whereas reversible binding occurs primarily at the same site (benzodiazepine).

Direct determination of ibuprofen and ibuprofen acyl glucuronide in plasma by high-performance liquid chromatography using solid-phase extraction

A method for the simultaneous determination of ibuprofen and its labile, reactive metabolite, ibuprofen acyl glucuronide, in plasma is described. Reversed-phase high-performance liquid chromatography (HPLC) employed a C18 column using methanol-10 mM trifluoroacetic acid as the mobile phase with ultraviolet detection at 225 or 214 nm. It is essential that blood is handled rapidly and plasma is acidified upon collection prior to freezing. Plasma samples first are deproteinated with acetonitrile, the supernatant is diluted with phosphate buffer, and ibuprofen, ibuprofen glucuronide, and ibufenac (internal standard) are initially isolated by solid-phase extraction on C18 cartridges. Upon elution, the residue is evaporated, dissolved and injected onto the HPLC system. Recovery is 94 +/- 8 and 70 +/- 9% for ibuprofen glucuronide and ibuprofen, respectively. The measurable concentration range is linear from 0.1 to 10 micrograms/ml for ibuprofen glucuronide and from 0.5 to 100 micrograms/ml for ibuprofen. The method is satisfactory for the analysis of ibuprofen and ibuprofen glucuronide from pharmacokinetic studies in humans. The direct determination of ibuprofen glucuronide allows accurate measurement of this conjugate at low levels relative to the parent compound, ibuprofen, a distinct advantage compared to previously employed indirect methods.

Studies on the reactivity of acyl glucuronides--IV. Covalent binding of diflunisal to tissues of the rat

Acyl glucuronides have been shown to be reactive electrophilic metabolites capable of undergoing hydrolysis, rearrangement (isomerization via acyl migration) and covalent binding reactions to plasma protein. The present study was undertaken to explore the occurrence and extent of in vivo formation of covalent adducts of diflunisal (DF), a salicylate derivative which forms a reactive acyl glucuronide, with tissues and plasma protein of rats. Groups of rats were given 50 mg DF/kg i.v. twice daily for periods of up to 7 days. Steady state plasma concentrations of reversibly bound DF and its conjugates (as measured 6 hr after a dose) were achieved by the third day of dosing. T 1/2 values after cessation of dosing were about 5-10 hr. By contrast, covalent DF-tissue adducts steadily accumulated over the 7-day dosing period. Maximum concentrations, measured 6 hr after the last dose, were 4.8 (liver), 1.0 (kidney), 0.74 (plasma), 0.26 (small intestine minus contents), 0.27 (large intestine minus contents) and 0.20 (skeletal muscle) microgram DF/g tissue or/mL plasma. T 1/2 values of about 50, 67, 18, 38 and 43 hr were obtained for liver, kidney, plasma and small and large intestine (respectively) after cessation of dosing. Thus, the study of acyl glucuronide reactivity and the question of any derived toxicity or immune responses should consider the formation of long-lived adducts in tissues as well as in plasma.

Stereoselective interactions of ketoprofen glucuronides with human plasma protein and serum albumin

A clearance pathway common to many aryl alkanoic acids is the generation of renally eliminated ester glucuronides. These metabolites are susceptible to systemic hydrolysis which generates the parent aglycone. We have conducted in vitro studies with biosynthetic R- and S-ketoprofen glucuronides to elucidate the mechanism of this phenomenon. These conjugates were incubated in human plasma, various concentrations of human serum albumin (HSA) and protein-free buffer. It was apparent that albumin, rather than plasma esterases, catalysed the hydrolysis of the glucuronides. The albumin-catalysed hydrolysis of ketoprofen glucuronides was highly stereoselective. The mean (+/- SD) hydrolysis half-life of R-ketoprofen glucuronide in plasma (N = 4) at physiological pH and temperature was 1.37 (+/- 0.30) hr. The corresponding value for S-ketoprofen glucuronide, 3.46 (+/- 0.84) hr, was significantly different (P less than 0.005). In contrast, synthetic ethyl esters of R- and S-ketoprofen were hydrolysed by plasma esterases, but not by HSA, and with little stereoselectivity. The reversible protein binding of ketoprofen glucuronides was determined at physiological pH and temperature by a rapid ultra-filtration method. The binding of R- and S-ketoprofen glucuronide to human plasma protein was independent of concentration (P greater than 0.05) over the range of 1-20 micrograms/mL. The mean (+/- SD) percentage unbound in plasma (N = 4) of R-ketoprofen glucuronide was 12.6 (+/- 1.4)%. The corresponding value for S-ketoprofen glucuronide, 9.12 (+/- 0.54)%, was significantly different (P less than 0.005). S-Ketoprofen glucuronide was also more avidly protein bound in physiological concentrations of HSA. However, this stereoselectivity decreased in more dilute HSA solutions. Based on the hydrolysis and protein binding data for ketoprofen glucuronides, we propose the existence of separate binding and catalytic sites on the albumin molecule for these metabolites.

Studies on the reactivity of acyl glucuronides--III. Glucuronide-derived adducts of valproic acid and plasma protein and anti-adduct antibodies in humans

The major metabolite of the anti-epileptic agent valproic acid (VPA) is its acyl glucuronide conjugate (VPA-G), which undergoes non-enzymic, pH-dependent rearrangement via acyl migration to a mixture of beta-glucuronidase-resistant forms (collectively VPA-G-R). We have compared the reactivity of VPA-G and VPA-G-R towards covalent VPA-protein adduct formation by incubation in buffer, human serum albumin (HSA) and fresh human plasma at pH 7.4 and 37 degrees. In all three media, the predominant reaction of VPA-G over 30 hr was rearrangement to VPA-G-R (ca. 24%). Hydrolysis was quite minor (ca. 2%) and covalent adduct formation negligible (when protein was present). On the other hand, both hydrolysis (ca. 27%) and adduct formation (ca. 7%) were extensive when VPA-G-R was incubated with HSA or plasma. These data do not support a transacylation mechanism for VPA-protein adduct formation, since this pathway should be much more highly favoured by VPA-G (an acyl-substituted acetal) than VPA-G-R (simple esters). VPA-protein adducts were found in the plasma of epileptic patients taking VPA chronically (mean 0.77 +/- SD 0.63 microgram VPA equivalents/mL, N = 17). An enzyme linked immunosorbent assay was developed, using HSA modified by incubation with VPA-G-R, to test the immunoreactivity of the patients' plasma. Of 57 patients tested, nine showed measurable levels of antibodies to these adducts, but the titres were very low, with no difference in response to modified and unmodified protein detectable at plasma dilutions of 1:16 or greater. These results suggest that the VPA-protein adducts have little immunogenicity, and are in agreement with clinical observations that drug hypersensitivity responses have not been associated with VPA therapy. Thus, although the in vitro data show that VPA-G is an example of a relatively unreactive acyl glucuronide, covalent VPA-plasma protein adducts and anti-adduct antibodies are nonetheless formed in vivo, at least in some patients on chronic therapy with the drug.

The role of leukocyte-generated reactive metabolites in the pathogenesis of idiosyncratic drug reactions

Evidence strongly suggests that many adverse drug reactions, including idiosyncratic drug reactions, involve reactive metabolites. Furthermore, certain functional groups, which are readily oxidized to reactive metabolites, are associated with a high incidence of adverse reactions. Most drugs can probably form reactive metabolites, but a simple comparison of covalent binding in vitro is unlikely to provide an accurate indication of the relative risk of a drug causing an idiosyncratic reaction because it does not provide an indication of how efficiently the metabolite is detoxified in vivo. In addition, the incidence and nature of adverse reactions associated with a given drug is probably determined in large measure by the location of reactive metabolite formation, as well as the chemical reactivity of the reactive metabolite. Such factors will determine which macromolecules the metabolites will bind to, and it is known that covalent binding to some proteins, such as those in the leukocyte membrane, is much more likely to lead to an immune-mediated reaction or other type of toxicity. Some reactive metabolites, such as acyl glucuronides, circulate freely and could lead to adverse reactions in almost any organ; however, most reactive metabolites have a short biological half-life, and although small amounts may escape the organ where they are formed, these metabolites are unlikely to reach sufficient concentrations to cause toxicity in other organs. Many idiosyncratic drug reactions involve leukocytes, especially agranulocytosis and drug-induced lupus. We and others have demonstrated that drugs can be metabolized by activated neutrophils and monocytes to reactive metabolites. The major reaction appears to be reaction with leukocyte-generated hypochlorous acid. Hypochlorous acid is quite reactive, and therefore it is likely that many other drugs will be found that are metabolized by activated leukocytes. Some neutrophil precursors contain myeloperoxidase and the NADPH oxidase system, and it is likely that these cells can also oxidize drugs. Therefore, although there is no direct evidence, it is reasonable to speculate that reactive metabolites generated by activated leukocytes, or neutrophil precursors in the bone marrow, could be responsible for drug-induced agranulocytosis and aplastic anemia. This could involve direct toxicity or an immune-mediated reaction. These mechanisms are not mutually exclusive, and it may be that both mechanisms contribute to the toxicity, even in the same patient. In the case of drug-induced lupus, a prevalent hypothesis for lupus involves modification of class II MHC antigens.(ABSTRACT TRUNCATED AT 400 WORDS)

Studies on the reactivity of acyl glucuronides--I. Phenolic glucuronidation of isomers of diflunisal acyl glucuronide in the rat

Diflunisal (DF) is metabolized primarily to its acyl glucuronide (DAG), phenolic glucuronide (DPG) and sulphate (DS) conjugates. Whereas DPG and DS are stable at physiological pH, DAG is unstable, undergoing hydrolysis (regeneration of DF) and rearrangement (intramolecular acyl migration to the 2-, 3- and 4-O-acyl-positional isomers). We have compared the in vivo disposition of DAG with that of an equimolar mixture of its three isomers after i.v. administration at 10 mg DF equivalents/kg to conscious, bile-exteriorized rats. After dosing with DAG, excretion in urine and bile (46% as DAG), hydrolysis (as assessed by recovery of 9% DPG and 8% DS resulting from reconjugation of liberated DF) and rearrangement (17% recovery as isomers of DAG) were important pathways. Highly polar metabolites excreted almost exclusively in bile and accounting for 13% of the dose were identified as an approximate 4:1 mixture of the 2- and 3-O-isomers of DAG which had been glucuronidated at the phenolic function of the salicylate ring i.e. "diglucuronides" of DF. Evidence for trace quantities only of the phenolic glucuronides of the 4-O-isomer of DAG, and of DAG itself, was found. After dosing rats with an equimolar mixture of the isomers, 52% was recovered (as the isomers) in urine and bile in 6 hr. Hydrolysis was less important--less than 3% (total) of the dose was recovered as DPG and DS. The phenolic glucuronides of the 2- and 3-O-isomers (ratio ca. 3:7) accounted for 37%. Evidence for appreciable formation of the phenolic glucuronide of the 4-O-isomer was not found. In one rat dosed with DPG, there was no evidence for further glucuronidation of the salicylate ring at its carboxy function. The data suggest that the 2- and 3-O-isomers of DAG, but not the 4-O-isomer, DAG itself or DPG, are good substrates for further glucuronidation.

Studies on the reactivity of acyl glucuronides--II. Interaction of diflunisal acyl glucuronide and its isomers with human serum albumin in vitro

A major metabolite of diflunisal (DF) is its reactive acyl glucuronide conjugate (DAG) which can undergo hydrolysis (regeneration of DF), intramolecular rearrangement (isomerization via acyl migration) and intermolecular reactions with nucleophiles. We have compared the fate of DAG and its individual 2-, 3- and 4-O-acyl positional isomers (at ca. 55 micrograms DF equivalents/mL) after incubation with human serum albumin (HSA, 40 mg/mL) at pH 7.4 and 37 degrees. Initial half-lives (T1/2) for DAG and its 2-, 3- and 4-isomers were 53, 75, 61 and 26 min, respectively. DAG was more labile to hydrolysis than any of its isomers but the latter, in particular the 4-isomer, were much better substrates for formation of covalent DF-HSA adducts. After a 2-hr incubation, 2.4, 8.2, 13.7 and 36.6% of substrate DAG and its 2-, 3- and 4-isomers (respectively) were present as DF-HSA adducts. With long term incubation, the concentrations of adducts so generated in situ declined in a biphasic manner, with apparent terminal T1/2 values of ca. 28 days. DAG was much more labile to transacylation with methanol (i.e. formation of DF methyl ester) than an equimolar mixture of its isomers after incubation in a 1:1 methanol:pH 7.4 buffer solution at 37 degrees (T1/2 values of 5 and 70 min, respectively). The data do not support direct transacylation with nucleophilic groups on protein as the predominant mechanism of formation of covalent DF-HSA adducts in vitro.

Absence of phenolic glucuronidation and enhanced hydroxylation of diflunisal in the homozygous Gunn rat

1. The disposition of diflunisal (DF) was investigated in bile-exteriorized and intact homozygous Gunn rats given 10 and 50 mg/kg doses i.v. and in Wistar rats given 10 mg/kg doses i.v. 2. In Gunn rats, DF sulphate, DF acyl glucuronide, and a hitherto unidentified metabolite of DF, a conjugate of 3-hydroxy-DF, were identified as the major metabolites, accounting for approximately 37%, 16% and 11% respectively of 10 mg/kg doses and 35%, 24% and 15% respectively of 50 mg/kg doses in bile-exteriorized animals. There was no evidence for formation of DF phenolic glucuronide. 3. Total plasma clearance of DF and formation clearances of DF to DF sulphate and 3-hydroxy-DF were little affected by increase of dose from 10 to 50 mg DF/kg, whereas formation clearance of DF to DF acyl glucuronide was increased, but not significantly. 4. In Gunn rats with undisturbed bile flow into the gut, recoveries of DF sulphate and total 3-hydroxy-DF in urine increased to approximately 48% and 25% dose respectively at the expense of DF acyl glucuronide through enterohepatic recirculation. 5. In bile-exteriorized Wistar rats, DF phenolic glucuronide, DF acyl glucuronide, DF sulphate and 3-hydroxy-DF accounted for 16%, 27%, 14% and 2%, respectively, of 10 mg/kg doses. In intact Wistar rats, urinary recoveries of the metabolites were 15%, 13%, 23% and 5%, respectively. 6. Thus in comparison to Wistar rats, phenolic glucuronidation of DF was absent or negligible in homozygous Gunn rats, acyl glucuronidation was significantly decreased, sulphation was unchanged, and the 3-hydroxylation of DF was significantly enhanced.

In vivo covalent binding of clofibric acid to human plasma proteins and rat liver proteins

Recent studies have shown that acyl-glucuronide conjugates are chemically reactive electrophilic metabolites that can undergo transacylation reactions resulting in intra-molecular rearrangement, hydrolysis and covalent binding of aglycone to albumin both in vitro and in vivo. The hypolipidaemic agent clofibrate is eliminated almost entirely as clofibric acid glucuronide in humans and rats. The formation of clofibric acid-protein adducts was investigated in 14 patients receiving 0.5-2.0 g/day of clofibrate for hypercholesterolaemia, and in liver homogenates from 20 rats administered 280 mg/kg/day of clofibric acid for up to 21 days. Total clofibric acid concentrations in the patients ranged from 0 to 114 mg/L. Covalently bound clofibric acid-protein adducts were detected in all patients, even in one subject in whom there was no measurable plasma clofibric acid. Concentrations ranged from 2.2 to 53.4 ng/mg protein and, in eight patients receiving 1.0 g/day of clofibrate, were correlated (P less than 0.05) with renal function as assessed by creatinine clearance. Clofibric acid-protein adducts were also present in rat liver homogenates, and increased with increasing duration of treatment (P less than 0.0001), from a mean (SE) of 10.1 (0.7) to 32.3 (1.6) ng/mg protein. The covalent binding of drugs to tissue macromolecules has traditionally been associated with toxicity. Further research is required to elucidate the role of acyl-glucuronide conjugates in the formation of drug-protein adducts and their biological consequences.

Diflunisal and its conjugates in patients with renal failure

Six patients with renal failure were given a single oral dose (250 mg) of diflunisal. In contrast to the acyl glucuronide, the phenolic glucuronide and sulphate conjugates showed the capacity to accumulate in plasma, suggesting that systemic instability of the acyl glucuronide contributes, via hydrolysis, to plasma concentrations of diflunisal itself. Although earlier studies in renal failure patients have almost certainly underestimated diflunisal clearance (by overestimation of plasma diflunisal concentrations through unrecognized acidic hydrolysis of diflunisal sulphate during analysis), the present results suggest that the reported decrease in clearance was not attributable only to this analytical artifact.

The sulphate conjugate of diflunisal: its synthesis and systemic stability in the rat

1. Diflunisal (DF) is metabolized in rats and humans primarily to its acyl glucuronide, phenolic glucuronide and sulphate conjugates. 2. DF sulphate was synthesized and administered i.v. to conscious bile-exteriorized rats at 10 mg DF equiv./kg. 3. DF sulphate was excreted almost exclusively in urine, and its systemic hydrolysis, monitored by glucuronidation of liberated DF, was quantitatively negligible. 4. Systemic stability of DF sulphate was therefore similar to DF phenolic glucuronide, and contrasted with the systemic instability of DF acyl glucuronide.

Contrasting systemic stabilities of the acyl and phenolic glucuronides of diflunisal in the rat

1. Diflunisal (DF) is metabolized in humans and rats primarily to its acyl glucuronide, phenolic glucuronide and sulphate conjugates. 2. After i.v. administration of DF acyl glucuronide to pentobarbitone-anaesthetized rats, DF and its phenolic glucuronide and sulphate conjugates appeared rapidly in plasma, indicating ready systemic hydrolysis of the acyl glucuronide and subsequent biotransformation of liberated DF. 3. Approximately 72% of the acyl glucuronide dose was recovered in bile and urine over 6 h: 52% as acyl glucuronide, 6% as phenolic glucuronide, 5% as sulphate, and 8% as isomers of the acyl glucuronide arising from intramolecular acyl migration. 4. Blockage of excretion routes by ligation of the ureters, bile duct, and both ureters and bile duct, decreased plasma clearance of the acyl glucuronide from 7.8 ml/min per kg to 6.0, 3.2 and 2.2 ml/min per kg respectively, and increased the apparent terminal plasma half-life of DF from 2.1 h to 2.6, 3.4 and 6.3 h, respectively. 5. By contrast, DF phenolic glucuronide was quite stable after i.v. administration at the same dose. 6. This study shows that systemic cycling between DF and its acyl glucuronide exists in the rat in vivo, with portions of each cycle of unstable acyl glucuronide through DF yielding stable phenolic glucuronide and (presumptively stable) sulphate conjugate.

Rapid high-performance liquid chromatographic assay for the simultaneous determination of probenecid and its glucuronide in urine. Irreversible binding of probenecid to serum albumin

A reversed-phase high-performance liquid chromatographic (HPLC) assay for the simultaneous determination of probenecid and its glucuronide in urine has been developed. The genuine glucuronide conjugate was isolated from urine by the use of solid-phase extraction on Amberlite XAD-2 and finally purified by the use of preparative HPLC on a Sepharon Hema 1000 RP-18 column. The purity of the product obtained was 88.9%. The isolated glucuronide was used as a standard sample. Of a p.o. dose of 500 mg to two volunteers, 26 and 29% were excreted as the ester glucuronide, while 1.0 and 2.7% were excreted unmetabolized. The stability of the ester glucuronide was investigated in aqueous buffers, buffered urine and human serum albumin solutions. The glucuronide was unstable in neutral and mildly alkaline solutions, and special precautions have to be taken during sampling and sample treatment in order to preserve the genuine glucuronide. The presence of human serum albumin in the solution stabilized the glucuronide against isomerization/rearrangements but catalysed the hydrolysis of the glucuronide. When incubating human serum albumin with the ester glucuronide, probenecid was shown to be covalently bound to the protein probably via a transacylation reaction.

Effects of blockage of urine and/or bile flow on diflunisal conjugation and disposition in rats

1. The effects of surgical blockage of either or both of the urinary and biliary excretion routes on the elimination of diflunisal (DF) and its conjugates were investigated in pentobarbitone-anaesthetized rats given DF at 10 mg/kg i.v. 2. In control animals the acyl glucuronide and phenolic glucuronide conjugates were excreted predominantly in bile, whereas the sulphate conjugate was eliminated almost exclusively in urine. 3. Bilateral ureter ligation had little effect on DF elimination, except for accumulation of the sulphate conjugate in plasma. Compensatory biliary excretion did not occur. 4. Total plasma clearance of DF decreased from 1.01 to 0.68 ml/min per kg following bile duct ligation. Plasma concentrations and urinary excretion of the glucuronides were elevated. 5. In rats with blockage of both urinary and biliary excretion routes, total plasma clearance of DF decreased to 0.59 ml/min per kg. Both the sulphate and phenolic glucuronide conjugates accumulated in plasma, whereas the acyl glucuronide peaked at 30 min and then declined in parallel with DF. The latter result indicates systemic instability of DF acyl glucuronide with hydrolytic regeneration of DF as the likely major consequence.