Covalent binding of acidic drugs via reactive intermediates: detection of benoxaprofen and flunoxaprofen protein adducts in biological material.

The purpose of the study was the direct detection of intact protein adducts-resulting from in-vitro incubations of flunoxaprofen- and benoxaprofen glucuronides in biological materials or originating from in vivo studies-by polyacrylamide gel electrophoresis (SDS-PAGE) followed by blotting and fluorescent scan, presumably yielding better specificity for the macromolecular binding partner and avoiding alkaline cleavage to release the aglycone. Glucuronides were isolated from urine samples or generated by incubation of aglycone with cofactors and rat liver microsomes. Following dialysis against BSA solution, SDS-PAGE and subsequent electrotransblotting were performed. Apparently, albumin represents the major binding protein for the covalent binding of these acyl glucuronides in plasma following incubation with blank plasma. In microsomal proteins two fluorescent peaks (appr. 39 and 62 KD) were identified for flunoxaprofen and benoxaprofen incubations. In vivo covalent binding was detected for both flunoxaprofen and benoxaprofen in plasma samples. For the racemically administered benoxaprofen a slight preponderance in adduct concentrations was found for the S-enantiomer. The pharmacokinetic analysis of in vivo data obtained for R/S-benoxaprofen (dose: 600 mg racemate) and S-flunoxaprofen (dose: 100 mg racemate), both of which have been withdrawn from the market, (employing a stereospecific HPLC method when analyzing volunteers' and patients' samples collected in the last 14 years, yet not stored longer than 3-4 years) demonstrated that significant amounts of glucuronides occur for both drugs (n = 2 for each compound; average Cmax values of the glucuronides: S-flunoxaprofen: 395 ng/ml; S-benoxaprofen: 775 ng/ml; R-benoxaprofen: 563 ng/ml). Presumably because of stereoinversion in humans, aglycone and glucuronide concentrations were higher for S- than for R-benoxaprofen. In vivo aglycone/glucuronide ratios were smaller for S- than for R-benoxaprofen, although in vitro incubation with human liver microsomes resulted in preferential glucuronidation of the R-enantiomer of benoxaprofen. Plasma concentration-time curves of the glucuronides paralleled those of the respective aglycones in their terminal phase. S-Benoxaprofen adduct concentrations were higher than R-benoxaprofen adduct concentrations (S: 28 ng/ml; R: 18 ng/ml covalently bound) and S-flunoxaprofen adduct concentrations with 29 ng/ml in the same range as S-beoxaprofen adducts, although for the latter the dose range as well as the respective glucuronide concentrations were higher. This indicates a higher reactivity of S-flunoxaprofen as opposed to S-benoxaprofen glucuronides.