Sulfation and glucuronidation are two parallel pathways for the metabolism of phenolic substrates. Gentisamide (GAM) was used as a model compound to examine the effects of parallel competing pathways on drug disappearance and metabolite formation in the once-through perfused rat liver preparation. GAM was found to form one glucuronide (GAM-5G) and two sulfate (GAM-2S and GAM-5S) conjugates. These GAM conjugates were biosynthesized in recirculating rat liver preparations, and were isolated by preparative high-performance liquid chromatography. Specific incorporation of 35S-sodium sulfate and [14C]glucose into GAM sulfate and glucuronide conjugates revealed corresponding elution patterns as labeled GAM metabolites. Their identities were characterized by enzymatic and acid hydrolyses and by NMR spectroscopy. Gentisamide-5-sulfate (GAM-5S) and gentisamide-5-glucuronide (GAM-5G) are major metabolites, and gentisamide-2-sulfate (GAM-2S) is a minor metabolite. Single-pass rat liver perfusions were used to examine the effect of stepwise increases/decreases of input GAM concentration (CIn) on the extraction ratio (E) of GAM and formation of metabolites. The E of GAM remained constant (about 0.89) at input concentrations from 0.9 to 120 microM and decreased at CIn greater than 120 microM. Metabolite patterns, however, changed with GAM CIn, even when E was constant at CIn up to 120 microM. GAM-5S was present as the major metabolite of GAM at all GAM CInS in most liver preparations but the proportions of GAM-5S and GAM-2S decreased at increasing CIn; the proportion of GAM-5G, a minor metabolite at low CIn, increased with increasing CIn. Biliary excretion rates at steady state accounted for 5.3 +/- 2.7% (mean +/- S.D.) of the input rate: GAM-5G was the predominant metabolite found. Fitting the metabolic data (sum of the rates of efflux in bile and perfusate at steady state) and the logarithmic average drug concentration at the various CIn to the Michaelis-Menten equation furnished parameter estimates for the three metabolic pathways. The estimated Km and Vmax values were quite comparable: for GAM-2S formation, 22 microM and 287 nmol/min; for GAM-5S formation, 26 microM and 978 nmol/min; for GAM-5G formation, 71 microM and 1062 nmol/min, indicating that sulfation and glucuronidation are effective competing pathways of each other. In viewing the fate of GAM over the CIn range, a changing metabolic fate of GAM over a constant E was noted at CIn less than 120 microM: the sum of the rates of metabolite formation was proportional to CIn.(ABSTRACT TRUNCATED AT 400 WORDS)