The hepatocellular entry of 4-methylumbelliferyl sulfate (4MUS) a highly ionized and highly bound anion capable of futile cycling, was examined in the single-pass albumin-free perfused rat liver preparation. Desulfation of 4MUS to 4-methylumbelliferone (4MU) was verified in vitro to be a low-affinity, high-capacity process (Km = 731 micromol/L; Vmax = 414 nmol min(-1) g(-1) liver). With 4MUS given to the perfused rat liver, sulfation of 4MU, the formed metabolite, was attenuated in the presence of 2,6-dichloro-4-nitrophenol (DCNP), a sulfation inhibitor, and when sulfate ion was substituted by chloride ion. 4MU sulfation, being a high-affinity system, was reduced most effectively at the lowest 4MUS concentration (15 micromol/L) used, evidenced by the increased (24%) net hepatic extraction ratio of 4MUS and reduced utilization (72%) of infused tracer 35SO4(2-) by 4MU for 4MU35S formation. Single-pass multiple indicator dilution (MID) studies were thus conducted under identical conditions (DCNP and absence of inorganic sulfate), with injection of [3H]4MUS and a set of noneliminated vascular and cellular reference indicators into the portal vein (prograde) or hepatic vein (retrograde), against varying background bulk concentrations of 4MUS (5 to 900 micromol/L). The steady-state removal rate of 4MUS and formation rates of 4MU and its glucuronide conjugate (4MUG) were not altered with perfusion flow direction, suggesting the presence of even or parallel distributions of 4MUS desulfation and 4MU glucuronidation activities. When the outflow dilution profile of [3H]4MUS was evaluated with the barrier-limited model of Goresky, a slight red cell carriage effect was found for 4MUS. The permeability surface area product for cellular entry for prograde showed a dramatic concentration-dependent decrease (from 0.13 to 0.01 mL sec(-1) g(-1), or 7.4 to 0.56 times the blood perfusate flow rate) and was resolved as saturable and nonsaturable components, while data for retrograde were more scattered, varying from 2.8 to 1 times the blood perfusate flow rate. Efflux (coefficient = 0.0096 +/- 0.0024 and 0.0088 +/- 0.0062 mL sec(-1) g(-1), respectively) was relatively insensitive to concentration and flow direction. The same was observed for the removal capacity for metabolism and excretion (sequestration coefficient: for prograde, 0.0056 +/- 0.0017 mL sec(-1) g(-1); for retrograde, 0.0056 +/- 0.003 mL sec(-1) g(-1)). The decrease in the apparent partition coefficient (ratio of 4MUS concentration estimated in tissue to unbound plasma concentration) and the increase in relative throughput component with concentration further substantiate the claim on the presence of concentrative processes at the sinusoidal membrane.