Studies on the reactivity of acyl glucuronides--VII. Salicyl acyl glucuronide reactivity in vitro and covalent binding of salicylic acid to plasma protein of humans taking aspirin

Uniform Microgels Containing Agglomerates of Silver Nanocubes for Molecular Size-Selectivity and High SERS Activity

Surface-enhanced Raman scattering (SERS) is a promising technique for molecular analysis as the molecular fingerprints (Raman spectra) are amplified to detectable levels compared with common spectroscopy. Metal nanostructures localize electromagnetic field on their surfaces, which can lead to dramatic increase of Raman intensity of molecules adsorbed. However, the metal surfaces are prone to contamination, thereby requiring pretreatment of samples to remove adhesive molecules. To avoid the pretreatment and potentially achieve point-of-care (POC) analysis, we have developed SERS-active microgels using the droplet-microfluidic system. As the microgels are composed of water-swollen network with consistent mesh size, they selectively allow diffusion of molecules smaller than the mesh, thereby excluding large adhesives. To render the microgels highly SERS-active, we destabilize silver nanocubes to form agglomerates, which are embedded in the matrix of microgels. The nanogaps in the agglomerates provide high sensitivity in Raman measurement and size-selective permeability of the microgel matrix obviates the pretreatment of samples. To validate the functions, we demonstrate the direct detection of Aspirin dissolved in whole blood without any pretreatment.

A molecular functional study on the interactions of drugs with plasma proteins

The binding of drugs to plasma proteins, such as albumin and alpha1-acid glycoprotein (AGP) is a major determinant in the disposition of drugs. A topology analysis of drug binding sites on HSA and AGP was determined using various methods, including spectroscopy, QSAR, photoaffinity labeling and site directed mutagenesis. Recombinant albumin was found to be useful for rapidly identifying drug binding sites. The binding sites on AGP are not completely separated but are partially overlapped, and Trp, Tyr, Lys and His residues in the drug binding pockets play important roles in this process. Drug displacement is somewhat complex, due to the involvement of multiple effects. The reduced binding in uremic patients may be explained by a mechanism that involves a combination of direct displacement by free fatty acids as well as cascade effects of free fatty acids and unbound uremic toxins for significant inhibition in serum binding. Albumin-containing dialysate is useful for the extracorporeal removal of endogenous toxins and in the treatment of drug overdoses. Oxidized albumin is a useful biomarker for the quantitative and qualitative evaluation of oxidative stress. Interestingly, AGP undergoes a structural transition to a unique structure that differs from the native and denatured states, when it interacts with membranes.

Role of benoxaprofen and flunoxaprofen acyl glucuronides in covalent binding to rat plasma and liver proteins in vivo

Benoxaprofen (BNX) has been implicated in rare but serious hepatotoxicity which led to its withdrawal from the world market. Flunoxaprofen (FLX), a structural analog, appears to be less toxic. It has been postulated that the nonsteroidal antiinflammatory drugs associated toxicity may be related to covalent modification of proteins by their reactive acyl glucuronides, and the extent of covalent protein binding depends on both reactivity of the acyl glucuronide and the exposure to the reactive metabolite. The disposition of BNX and FLX in rats were compared upon intravenous administration of 20 mg/kg of BNX, FLX or their metabolites. Covalent binding of BNX and FLX to plasma and liver proteins were also determined, and an immunochemical approach was used to detect their hepatic targets. Similar concentrations of plasma protein adducts for BNX and FLX were detected even though the AUC of BNX-glucuronide (BNX-G) was almost twice that of FLX-glucuronide (FLX-G). Similar concentrations of liver protein adducts for BNX and FLX were also detected at 8 h, however, the hepatobiliary exposure of BNX-G was only 1/3rd that of FLX-G indicating that BNX-G was more reactive than FLX-G, which was in agreement with in vitro data. Proteins of 110 and 70 kDa were the major liver protein targets modified by covalent attachment of BNX and FLX. In conclusion, measuring covalent binding to tissue proteins in animals in addition to plasma adducts should be considered when evaluating and comparing carboxylic acid analogs that form reactive acyl glucuronides.

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.

In vitro reactivity of carboxylic acid-CoA thioesters with glutathione

The chemical reactivity of acyl-CoA thioesters toward nucleophiles has been demonstrated in several recent studies. Thus, intracellularly formed acyl-CoAs of xenobiotic carboxylic acids may react covalently with endogenous proteins and potentially lead to adverse effects. The purpose of this study was to investigate whether a correlation could be found between the structure of acyl-CoA thioesters and their reactivities toward the tripeptide, glutathione (gamma-Glu-Cys-Gly). The acyl-CoA thioesters of eight carboxylic acids (ibuprofen, clofibric acid, indomethacin, fenbufen, tolmetin, salicylic acid, 2-phenoxypropionic acid, and (4-chloro-2-methyl-phenoxy)acetic acid (MCPA)) were synthesized, and each acyl-CoA (0.5 mM) was incubated with glutathione (5.0 mM) in 0.1 M potassium phosphate (pH 7.4, 37 degrees C). All of the acyl-CoAs reacted with glutathione to form the respective acyl-S-glutathione products, with MCPA-CoA having the highest rate of conjugate formation (120 +/- 10 microM/min) and ibuprofen-CoA having the lowest (1.0 +/- 0.1 microM/min). The relative reactivities of the acyl-CoAs were dependent on the substitution at the carbon atom alpha to the acyl carbon and on the presence of an oxygen atom in a position beta to the acyl carbon and were as follows: phenoxyacetic acid > o-hydroxybenzoic acid--phenoxypropionic acid > arylacetic acid derivatives > 2-methyl-2-phenoxypropionic acid--2-phenylpropionic acid. For each acyl-CoA thioester, the overall hydrolysis rate was determined as the time-dependent formation of parent compound. A linear trend was observed when comparing the reactivities of the acyl-CoAs with glutathione with the corresponding overall hydrolysis rates. Thus, the most reactive compound (MCPA-CoA) was also the compound with the highest rate of hydrolysis and the least reactive compounds (ibuprofen-CoA, clofibryl-CoA) were also the compounds least susceptible to hydrolysis.

Acyl glucuronide drug metabolites: toxicological and analytical implications

Although glucuronidation is generally considered a detoxification route of drug metabolism, the chemical reactivity of acyl glucuronides has been linked with the toxic properties of drugs that contain carboxylic acid moieties. It is now well documented that such metabolites can reach appreciable concentrations in blood. Furthermore, they are labile, undergo hydrolysis and pH-dependent intramolecular acyl migration to isomeric conjugates of glucuronic acid, and may react irreversibly with plasma proteins, tissue proteins, and with nucleic acids. This stable binding causes chemical alterations that are thought to contribute to drug toxicity either through changes in the functional properties of the modified molecules or through antigen formation with subsequent hypersensitivity and other immune reactions. Whereas in vitro data on the toxicity of acyl glucuronides have steadily accumulated, direct evidence for their toxicity in vivo is scarce. Acyl glucuronides display limited stability, which is dependent on pH, temperature, nature of the aglycon, and so on. Therefore, careful sample collection, handling, and storage procedures are critical to ensure generation of reliable pharmacologic and toxicologic data during clinical studies. Acyl glucuronides can be directly quantified in biologic specimens using chromatographic procedures. Their adducts with plasma or cell proteins can be determined after electrophoretic separation, followed by blotting. ELISA techniques have been used to assess the presence of antibodies against acyl glucuronide-protein adducts. This review summarizes the most recent evidence concerning biologic and toxicologic effects of acyl glucuronide metabolites of various drugs and discusses their relevance for drug monitoring. A critical evaluation of the available methodology is included.

Chemical reactivity of the naproxen acyl glucuronide and the naproxen coenzyme A thioester towards bionucleophiles

Drugs may be metabolised to reactive electrophilic species that spontaneously react with proteins. The presence of such drug-protein adducts has been associated with drug toxicity. In this study, the reactivity of the major metabolite of naproxen--the 1-beta-O-glucuronide (Nap-GlcU)--was compared to the corresponding naproxen coenzyme A (Nap-CoA) thioester. The reactivity of the two metabolites was assessed in vitro in a phosphate buffer (pH 7.4; 0.1 M) at 37 degrees C towards the model bionucleophiles glutathione and human serum albumin (HSA). The reaction between the electrophilic species (Nap-GlcU and Nap-CoA) and glutathione forming the Nap-glutathione conjugate was monitored using LC-MS-MS and LC-UV, respectively. It was shown that Nap-CoA resulted in an approximate 100-fold higher formation of Nap-glutathione conjugate than Nap-GlcU. The presence of Nap-CoA also resulted in acylated HSA with a rate and a yield that was significantly higher than reported for Nap-GlcU. In summary, the data suggest that CoA metabolites may be more reactive species than acyl glucuronides that previously have been associated with severe drug related side effects in vivo.

Practical aspects of the ligand-binding and enzymatic properties of human serum albumin

Recent work with approaches like recombinant mutants and X-ray crystallography has given much new information about the ligand-binding properties of human serum albumin (HSA). The information increases the understanding of this unique transport and depot protein and could give a structural basis for the possible construction of therapeutic agents with altered HSA-binding properties. A tabulation of high-affinity binding sites for both endogenous and exogenous compounds has been made; it could be useful for the above-mentioned purpose, but it could also be of value when trying to predict potential drug interactions at the protein-binding level. Drug displacement is not always a complication to therapy; it can be used to increase the biological effect of a drug. However, due to rebinding at other sites, the increase in the free concentration of a displaced ligand can be less than expected. Drugs and drug metabolites can also interact covalently with HSA; thiol-containing drugs often bind to the single free cysteine residue of HSA, and glucuronidated drugs react irreversibly with other residues of the protein. Reversible binding of ligands is often stereospecific, and therefore immobilized HSA can be used to separate drug isomers. Albumin-containing dialysates are useful for extracorporeal removal of endogenous toxins and in the treatment of drug overdoses. HSA has different types of hydrolytic activities, which also can be stereospecific. The esterase-like property seems especially useful in converting prodrugs to active drugs in plasma.

Stereospecific pH-dependent degradation kinetics of R- and S-naproxen-beta-l-O-acyl-glucuronide

The hydrolysis and acyl migration of biosynthetic S-naproxen-beta-l-O-acyl glucuronide (I) and R-naproxen-beta-l-O-acyl glucuronide (II) was followed by HPLC. Nine first-order kinetic rate constants for the hydrolysis and acyl migration between the beta-l-O-acyl glucuronide, its alpha/beta-2, alpha/beta-3-, alpha/beta-4-, and alpha-1-O-acyl isomers and naproxen aglycone were determined for I and II at pH 7.00, 7.40 and 8.00 at 37 degrees C by kinetic simulation. For I the 3-O-acyl isomer was the most stable isomer as the pseudo-equilibrium ratio for the major acyl-migrated isomers was 1:1.5:0.9 (2-O-acyl isomer:3-O-acyl isomer:4-O-acyl isomer). The 3- and 4-O-acyl isomers of II were equally stable as the pseudo-equilibrium ratio for the major acyl-migrated isomers was 1:1.4:1.4 (2-O-acyl isomer:3-O-acyl isomer:4-O-acyl isomer). For both I and II, the pseudo-equilibrium ratio between the major 2-O-acyl isomer and the minor alpha-l-O-acyl isomer was 10:1 (2-O-acyl isomer:alpha-l-O-acyl isomer). The pseudo-equilibrium found for the major acyl-migrated isomers of I and II in the present study corresponds with the pattern previously published for R- and S-ketoprofen-beta-l-O-acyl glucuronide acyl-migrated isomers, suggesting that these findings may be general for acyl-migrated beta-l-O-acyl glucuronides of enantiomeric 2-arylpropionic acids.

In vitro stereoselective degradation of carprofen glucuronide by human serum albumin. Characterization of sites and reactive amino acids

Acyl glucuronides formed from carboxylic acids can undergo hydrolysis, acyl migration, and covalent binding to proteins. In buffers at physiological pH, the degradation of acylglucuronide of a chiral NSAID, carprofen, consisted mainly of acyl migration. Acidic pH reduced hydrolysis and acyl migration, thus stabilizing the carprofen acyl glucuronides. Addition of human serum albumin (HSA) led to an increased hydrolysis of the conjugates of both enantiomers. This protein protected R-carprofen glucuronide from migration and therefore improved its overall stability. Hydrolysis was stereoselective in favor of the S conjugate. The protein domains and the amino acid residues likely to be responsible for the hydrolytic activity of HSA were deduced from the results of various investigations: competition with probes specific of binding sites, effects of pH and of chemical modifications of albumin. Dansylsarcosine (DS), a specific ligand of site II of HSA, impaired the hydrolysis, whereas dansylamide (DNSA) and digoxin, which are specific ligands of sites I and III, respectively, had no effect. The extent of hydrolysis by HSA strongly increased with pH, indicating the participation of basic amino acids in this process. The results obtained with chemically modified HSA suggest the major involvement of Tyr and Lys residues in the hydrolysis of glucuronide of S-carprofen, and of other Lys residues for that of its diastereoisomer.

In vitro regioselective stability of beta-1-O- and 2-O-acyl glucuronides of naproxen and their covalent binding to human serum albumin

beta-1-O- (NAG) and 2-O-glucuronides (2-isomer) of (S)-naproxen (NA) were prepared to determine which positional isomer(s) of the acyl glucuronide of NA is responsible for forming covalent adducts with human serum albumin (HSA). Their comparative stability and covalent binding adduct formation with HSA were investigated at pH 7.4 and at 37 degreesC. NA and its acyl glucuronides were simultaneously determined by HPLC. Three positional isomers were formed successively after incubation of NAG in the buffer only. However, when NAG was incubated with HSA (30 mg/mL), isomers other than the 2-isomer were formed in little or negligible quantities. In HSA solution, NAG (kd = 2.08 +/- 0.08 h-1) was four times less stable than 2-isomer (kd = 0.51 +/- 0.02 h-1). NAG was degraded by hydrolysis (khyd = 1.01 +/- 0.10 h-1) and isomerization (kiso = 1.07 +/- 0.07 h-1) to the same extent; however, hydrolysis was predominant for the 2-isomer (kd = 0.51 +/- 0.02 h-1). The incubation of both NAG and 2-isomer with HSA led to the formation of a covalent adduct; however, the adduct formation from the 2-isomer proceeded more slowly than that from NAG. The present results suggest that the covalent binding of NA to HSA via its acyl glucuronides proceeds through both transacylation (direct nucleophilic displacement) and glycation mechanisms; NAG rapidly forms an adduct that may be unstable, and the protein adduct from the 2-O-acyl glucuronide is as important for the covalent binding as those from the 1-O-acyl glucuronides.

Structural determination of the conjugate of human serum albumin with a mitomycin C derivative, KW-2149, by matrix-assisted laser desorption/ionization mass spectrometry

A new mitomycin C derivative, KW-2149, is known to form a covalent conjugate with human serum albumin (HSA). This conjugate exhibits 1/20 of the anticellular activity of unconjugated KW-2149. Structural studies of this conjugate were carried out using a combination of enzymatic digestion, high-performance liquid chromatography (HPLC), and matrix-assisted laser desorption/ionization (MALDI) mass spectrometry. The tryptic peptide T5 (residues 21-41) was the only peptide found to be modified by KW-2149 moieties, the [(gamma-L-glutamylamino)ethyl]thio group or the (2-aminoethyl)thio group, through a disulfide bond. Although the latter peptide lost its mitomycin C moiety in the course of tryptic digestion, these data strongly suggest that KW-2149 was bound to Cys-34, the only free cysteine on HSA.

Development of a simple liquid chromatographic method for the separation of mixtures of positional isomers and anomers of synthetic 2-, 3- and 4-fluorobenzoic acid glucuronides formed via acyl migration reactions

Many drugs containing carboxylate groups form beta-1-O-acyl glucuronides as their major phase II metabolites in vivo. These ester glucuronides are potentially reactive due to the susceptibility of the acyl group to nucleophilic reactions resulting in hydrolysis, acyl migration or covalent adduct formation. In the present study, a number of synthetic fluorobenzoic acid glucuronide conjugates were chosen as models for chromatographic studies. A high-performance liquid chromatography method is presented for the simultaneous determination of the 1-, 2-, 3- and 4-positional isomers of the acyl glucuronides, and their alpha- and beta-anomers for the 2-, 3- and 4-fluorobenzoic acids as well as each aglycone formed as a result of hydrolysis. The same elution order was found for the acyl migrated glucuronide isomers of the three fluorobenzoic acids in their equilibrium mixtures. The alpha-4-O-acyl isomer eluted first followed by the beta-4-O-acyl isomer, then the beta-1-O-acyl, the beta-3-O-acyl, the alpha-3-O-acyl, the alpha-2-O-acyl and finally the beta-2-O-acyl isomer eluted. The method was used to determine the overall degradation rates, the acyl migration rates and the hydrolysis rates of 1-O-(2-fluorobenzoyl)-beta-D-glucopyranuronic acid 1-O-(3-fluorobenzoyl)-beta-D-glucopyranuronic acid and 1-O-(4-flurobenzoyl)-beta-D-glucopyranuronic acid in a buffer system pH 7.4 at 25 degrees C. It was found that the order of beta-1-glucuronide acyl migration rates was 2-fluorobenzoyl > 3-fluorobenzoyl > 4-fluorobenzoyl. Both the acyl migration rates and the elution order were interpreted in terms of electronic effect of the fluorine substituent on the carbonyl carbon.

Direct analysis of salicylic acid, salicyl acyl glucuronide, salicyluric acid and gentisic acid in human plasma and urine by high-performance liquid chromatography

A method for the simultaneous direct determination of salicylate (SA), its labile, reactive metabolite, salicyl acyl glucuronide (SAG), and two other major metabolites, salicyluric acid and gentisic acid in plasma and urine is described. Isocratic reversed-phase high performance liquid chromatography (HPLC) employed a 15-cm C18 column using methanol-acetonitrile-25 mM acetic acid as the mobile phase, resulting in HPLC analysis time of less than 20 min. Ultraviolet detection at 310 nm permitted analysis of SAG in plasma, but did not provide sensitivity for measurement of salicyl phenol glucuronide. Plasma or urine samples are stabilized immediately upon collection by adjustment of pH to 3-4 to prevent degradation of the labile acyl glucuronide metabolite. Plasma is then deproteinated with acetonitrile, dried and reconstituted for injection, whereas urine samples are simply diluted prior to injection on HPLC. m-Hydroxybenzoic acid served as the internal standard. Recoveries from plasma were greater than 85% for all four compounds over a range of 0.2-20 micrograms/ml and linearity was observed from 0.1-200 micrograms/ml and 5-2000 micrograms/ml for SA in plasma and urine, respectively. The method was validated to 0.2 microgram/ml, thus allowing accurate measurement of SA, and three major metabolites in plasma and urine of subjects and small animals administered salicylates. The method is unique by allowing quantitation of reactive SAG in plasma at levels well below 1% that of the parent compound, SA, as is observed in patients administered salicylates.

Complete reversion and prevention of rectal adenomas in colectomized patients with familial adenomatous polyposis by rectal low-dose sulindac maintenance treatment. Advantages of a low-dose nonsteroidal anti-inflammatory drug regimen in reversing adenomas exceeding 33 months

PURPOSE

This nonrandomized, controlled Phase II pilot study aims at the lowest effective dose of rectally applied sulindac to achieve and maintain adenoma reversion in colectomized patients with familial adenomatous polyposis (FAP).

METHODS

The study group (n = 15) underwent proctoscopic and laboratory follow-up for polyp reversion every 6 to 12 weeks. Polyp reversion was followed by dose reduction in predefined steps. Proliferating cell nuclear antigen/cyclin (PCNA) and KI-67 proliferation indices (PI) were performed by point counting. Prostaglandin (PG)E2 and PGF2 alpha were quantified by time-resolved competitive fluorescence immunoassay.

RESULTS

All patients responded to therapy within 6 to 24 weeks. Sixty and 87 percent of patients achieved complete adenoma reversion after 48 weeks at 53 and 67 mg of sulindac per day per patient on average, respectively. Reversion was evident compared with the control group. Dose reduction by one-sixth to one-eighth of the usual oral dose was significant (Mann's trend test, P < 0.05). PCNA and KI-67 PIs of adenomatous and flat mucosa were significantly reduced (Wilcoxon's test, P < 0.05). Correlation of PCNA and KI-67 PIs indicate similar reaction of different tissue structures (Spearman's rank correlation test, P < 0.01). Nonsteroidal anti-inflammatory drug-induced redifferentiation from high-grade to low-grade dysplasia occurred in all but two patients. Tissue-PGE2 levels were greatly reduced. Unwanted, curable side effects were rare (gastritis, n = 2), and laboratory controls are within detection limits.

CONCLUSIONS

Low-dose rectal sulindac maintenance therapy is highly effective in achieving complete adenoma reversion without relapse in 87 percent of patients after 33 months. Rectal FAP phenotype should be crucial for the surgical decision. Colectomy with ileorectal anastomosis and regular chemoprevention might proceed to be a promising alternative to pouch procedures. Chemoprevention with lower incidence of FAP-related tumors via dysplasia reversion may be possible in the future.

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.

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.