Introduction

Top Abstract Introduction Materials and methods Results Discussion Appendix References

The kidney, a key organ for the maintenance of water and electrolyte balance, is capable of synthesis, metabolism, and secretion of hormones and is responsible for the excretion of drugs and synthetic compounds, particularly biotransformation products that are usually of greater polarity than the parent compounds (see recent review, Lohr et al., 1998). The kinetics of renal drug excretion are complex (Cummings et al., 1967, Garrett, 1978): the free species undergoes glomerular filtration and additional secretion results when substrate in the postglomerular circulation enters the basolateral and then the brush border (luminal) membranes into tubular urine, with the kinetic events being complicated by reabsorption at the brush border membrane. The scenario is complex because membrane transport processes at the basolateral and brush border membranes operate in both directions and are subject to local pH conditions that influence passive diffusion and saturation behaviors of carrier proteins that effect facilitative transport (Garrett, 1978; Levy, 1980; Tucker, 1981; Lohr et al., 1998).

The kinetics involving metabolite formation and excretion in the kidney are even more elaborate (Diamond and Quebbeman, 1981; Tremaine et al., 1984; de Lannoy et al., 1990). For the in vivo system in which metabolite formed by the kidney is directly excreted into urine and reenters the circulation, the apparent clearance of the formed metabolite estimated by conventional methods (urinary clearance/midpoint concentration) is not constant and exceeds that for the preformed metabolite (Wan and Riegelman, 1972; de Lannoy et al., 1989, 1990; de Lannoy and Pang, 1993; Kugler et al., 1996). In vivo, has been suggested that the handling of a metabolite formed from drug differs from that of the metabolite (preformed) administered to the kidney (Garrett, 1978), that the apparent renal clearance of the formed metabolite exceeds the renal clearance of the metabolite given per se as a preformed species, provided that the same transport processes are involved in the handling of both the formed and preformed metabolites, and given that there is lack of reabsorption (efflux) of the metabolite and absence of interaction between the parent drug and metabolite (Tucker, 1981).

Differentiation of the renal handling of formed and preformed metabolite in a recirculating system or in vivo is difficult (de Lannoy et al., 1990). For this reason, the single-pass isolated perfused rat kidney (IPK) preparation is ideal for distinguishing the fate of the preformed versus formed metabolite in the kidney, when both precursor and metabolite, labeled with different radiolabeled isotopes, are given simultaneously at tracer concentrations to the IPK (de Lannoy et al., 1989). The single-pass IPK precludes recirculation of the venous metabolite back to the kidney and allows for mass balance considerations and quantitation of the kinetics in absence of added organs. In this instance, one could postulate that because of filtration of the preformed but not the formed metabolite species which originates within the postglomerular region in situ the kidney that precludes filtration, the urinary clearance (CL_tot,K{pmi}) for a preformed metabolite that is solely excreted is greater in comparison to the apparent clearance of the metabolite formed from its precursor CL_tot,K{mi}, unless a transport barrier for entry of the metabolite exists (deLannoy et al., 1990). Renal metabolic studies on the comparative fates of the diacid metabolite [¹⁴C]enalaprilat generated from [¹⁴C]enalapril and preformed [³H]enalaprilat in the single-pass isolated perfused rat kidney preparation indeed showed that CL_tot,K{mi} > CL_tot,K{pmi} (de Lannoy et al., 1989). The observation was explained by the existence of a transport barrier at the basolateral membrane for the metabolite enalaprilat (de Lannoy et al., 1990) and was later confirmed with multiple indicator dilution studies (Schwab et al., 1992). Since renal metabolism and the resulting toxicity of xenobiotics have been well recognized (Elfarra and Anders, 1984; Spry et al., 1985; Stevens et al., 1988; Monks et al., 1990), the circumstances and the determining factors surrounding the renal clearance of the preformed metabolite that result in a different apparent renal clearance of the formed metabolite need to be clarified.

The present study was designed to address these issues through experimental and theoretical examinations. Experimentally, the handling of a formed versus preformed metabolite was compared with benzoate and hippurate in the single-pass IPK. Much is known about the precursormetabolite pair. Hippurate is formed via the conjugation of benzoate with glycine which occurs in the kidney (Wan and Riegelman, 1972; Kao et al., 1978; Poon and Pang, 1995), although the reaction also takes place in liver (Bridges et al., 1970; Chiba et al., 1994) and in the intestine (Strahl and Barr, 1971). Glycine conjugation is catalyzed by benzoyl-Co A synthetase and benzoyl-Co A-glycine N-acyltransferase in the matrix of mitochondria (Gatley and Sherratt, 1977). The inhibition of these two enzymes or the reduced availability of Co A will also affect the hepatic glycine conjugation (Amsel and Levy, 1969; Gregus et al., 1992, 1993). The uptake of hippurate across the basolateral membrane of the dog kidney was saturable with a high K_m (223 mM) (Knoefel and Huang, 1959; Russel et al., 1989). The inhibitory constants of hippurate on uptake of p-aminohippurate (PAH) were 4.8 and 11.6 mM in dog renal basolateral and brush border membrane vesicles, respectively (Russel and Vermeulen, 1994). The precursor, benzoate, also inhibited the uptake of PAH in dog renal basolateral and brush border membrane vesicles (Russel et al., 1991) and repressed the sodium-dependent reabsorption of lactate in the rat kidney (Ullrich et al., 1982; Ullrich and Rumich, 1988).

Although concentration-dependent removal of benzoate was found in the single-pass IPK preparation, the apparent extraction ratio of renally formed hippurate was constant (0.48) among all concentrations (Poon and Pang, 1995). Yet the fate of preformed hippurate was not directly investigated. In the present study, we compared the renal elimination of [¹⁴C]hippurate formed from tracer [¹⁴C]benzoate and of preformed [³H]hippurate when [¹⁴C]benzoate and [³H]hippurate were delivered simultaneously at tracer concentrations to the single-pass IPK preparation to obtain paired observations on the metabolites in the kidney. We explored, with the use of a physiological model, the analytical solutions for the clearances (or extraction ratio) of the drug and preformed metabolite, and the apparent clearance (or extraction ratio) of the formed metabolite species under linear conditions. With these solutions, parameters pertaining to the transport of benzoate and hippurate and the metabolite intrinsic clearances of benzoate were optimized for the observed data. The solutions were also used to divulge the interrelationships among the transport clearances and metabolic intrinsic clearance on the kinetics of formed versus preformed metabolite species.

	Materials and Methods

Top Abstract Introduction Materials and methods Results Discussion Appendix References

Source of Materials

[¹⁴C]Benzoic acid (specific activity, 16.0 mCi/mmol) was purchased from New England Nuclear Company (Boston, MA). The radiochemical purity of [¹⁴C]benzoate was >99%, as judged by high-performance liquid chromatography (HPLC). [2-³H]Glycine that was used for the synthesis of [³H]hippurate (38.8 Ci/mmol) was purchased from Dupont (Boston, MA). All reagents used were of glass-distilled HPLC grade or were of the highest purity available.

Synthesis of [³H]Hippurate

[³H]Hippurate was chemically synthesized from benzoyl chloride and [³H]glycine according to the method of Ingersoll and Babcock (1943). [³H]Glycine (0.026 µmol in 1 ml of 2% ethanolic water) was dried under nitrogen in a test tube, and unlabeled glycine (4.97 µmol) and 6.25 µmol of sodium hydroxide in 1 ml of distilled water were added and mixed. Then 10 µmol of NaOH (100 µl of 1 N NaOH) was added dropwise as the test tube was thoroughly mixed and the speed of sodium hydroxide addition affected the yield of the reaction. The reaction mixture was allowed to stand at room temperature for 60 min before termination of the reaction by placement in an ice bucket, and 400 µl of 0.1 N HCl was added. The formed [³H]hippuric acid in the aqueous phase was extracted into ether, which was transferred to a separate test tube and dried under N₂. The residue was reconstituted in water and further purified by HPLC. After purification, the specific activity of synthesized [³H]hippurate was 56 µCi/µmol, and the radiochemical purity was >98% as judged by HPLC.

Kidney Perfusion

Male Sprague-Dawley rats (Charles River, St. Constant, Quebec City, Canada; 402 ± 11 g) were fed ad libitum and allowed free access to water. The animals were housed in a room with a 12-h light/dark cycle. Isolated rat kidney perfusion was performed according to Ross (1978) and de Lannoy et al. (1989). Before surgery, the animals were anesthetized by an i.p. administration of sodium pentobarbital (50 mg/kg b.wt.). The surgical procedure and the perfusion apparatus were identical to those described previously (Poon and Pang, 1995). The perfusate, consisting of 20% washed bovine red blood cells (Ryding-Meat Packer, Toronto, Ontario, Canada), 4% bovine serum albumin (Fraction V; Sigma Chemical Co., St Louis, MO), 5 mM glucose, and a complement of 20 amino acids in Krebs-Henseleit bicarbonate solution buffered to pH 7.4 was oxygenated with 95% oxygen-5% carbon dioxide (Matheson Gas, Mississauga, Ontario, Canada). Unlabeled inulin (50 µM) was added to the perfusate in reservoir, and its clearance was used for estimation of the glomerular filtration rate (GFR). The kidney was first equilibrated under constant perfusion pressure (approximately 90 mm Hg) for 20 min and then the perfusion was continued under constant perfusate flow rate (8 ml/min) for the remainder of the experiment (50 min). During the entire experiment, urine was collected into preweighed microfuge tubes at 5-min intervals. The volume of urine was measured gravimetrically, with the assumption that the specific gravity is unity. Three samples were taken from the reservoir at 7.5, 27.5, and 47.5 min postequilibration to determine the steady-state input concentration (C_In), whereas the outflow perfusate was sampled at the midpoint of each of the 5-min intervals (2.5, 7.5, 12.5, 17.5, 22.5, 27.5, 32.5, 37.5, 42.5, and 47.5 min); the data obtained for the last five samples were averaged to provide the steady-state output plasma concentration (C_Out).

Red Blood Cell Distribution and Protein Binding of [³H]Hippurate

Blank blood perfusate (4% albumin) containing 40% of washed bovine red blood cells (RBCs) (v/v) was diluted with plasma perfusate containing [³H]hippurate to result in 20% RBC blood perfusate. [¹⁴C]Sucrose, a reference that does not enter into the red blood cell, was also included. After admixture, samples were incubated at 37°C and duplicate samples were taken at various times up to 3 h. The radioactivities of [³H]hippurate and [¹⁴C]sucrose in plasma before and after dilution were assayed using the liquid scintillation counter (model LS 5801; Beckman Instruments, Beckman Canada, Mississauga, Canada); the hematocrit of each sample was determined by a hematocrit centrifuge (Microfuge B; Beckman Instruments, Palo Alto, CA). The concentration ratio of [³H]hippurate in RBC water to the unbound [³H]hippurate in plasma water, _HA, was estimated as a ratio of the activities in RBC water and plasma water:

(1)

where c_r and c_p are the total (bound and unbound) concentrations of [³H]hippurate in RBC and plasma, respectively, f_r and f_p are the fractions of RBCs and plasma which are water (0.70 and 0.94 for 4% albumin), respectively, and f_u is the unbound fraction in plasma. Since the resulting plasma concentrations for the sucrose tracer (c^*_suc) and the tracer hippurate (c^*_p) after admixture are related to the original amounts of added sucrose and hippurate radioactivity, R^*_suc and R^*_HA, respectively, and the newly diluted volume (V) and hematocrit (Hct) these concentrations, c^*_suc and c^*_p, are expressed as:

(2)

(3)

where  is the volume ratio of RBC water to plasma water,

(4)

Division of eq. 2 by eq. 3 and further substitution of _HA (eq. 1) and  (eq. 4) result in the following (Pang et al., 1995),

(5)

Since (c^*_suc/c^*_p), , (R^*_suc/R^*_HA), and f_u are known, _HA is readily estimated from eq. 5.

The corresponding plasma samples containing the tracer [³H]hippurate were used for the determination of the unbound fraction (f_u) by ultrafiltration (Centricon Amicon, Beverly, MA; 10,000 molecular weight cutoff). The samples were centrifuged at 1000g for 20 min at room temperature. The unbound fraction in plasma was estimated as the ratio of the disintegrations per minute (dpm) of [³H]hippurate in the ultrafiltrate (unbound) to that in original plasma (total).

Analytical Methods

Viability. Perfusate and urine samples were analyzed for sodium, potassium, glucose, and inulin. Sodium and potassium were determined by flame photometry (IL 943 Flame Photometer; Instrumentation Laboratory, Lexington, MA). Glucose was assayed by the oxygen rate method (Glucose Analyzer 2; Beckman Instruments, Inc., Fullerton, CA). The percentage of reabsorbed sodium and glucose were used as indices for the reabsorptive function of the isolated IPK. Inulin in urine and plasma was determined using the UV method of Heyrovsky (1956) at 550 nm.

Drug and Metabolite Assays. [¹⁴C]Benzoic acid, formed [¹⁴C]hippurate, and preformed [³H]hippurate in urine and plasma were separated by a HPLC method with UV detection of methoxybenzoic acid, the internal standard (Chiba et al., 1994). The radiolabeled compounds eluted from the HPLC system were collected at predetermined collection intervals. Quantification was performed with calibration curves constructed with varying amounts of radioactive [¹⁴C]benzoate and [³H]hippurate added to plasma and urine.

Calculations. The plasma clearance of inulin was used as an estimate of GFR, and the value was normalized to the weight of the unperfused left kidney. The percentage of reabsorption of sodium and glucose was calculated as: [(filtered load  excreted load)/filtered load] × 100%.

(6)

(7)

(8)

Physiological Modeling. The handling of benzoate and hippurate by the kidney was examined with a physiological model similar to that used previously (Hekman and van Ginneken, 1982; deLannoy et al., 1990; Sirianni and Pang, 1998). In this model, tubular reabsorption of water (from lumen into plasma of peritubular circulation) was not taken into consideration in mass transfer relationships. The kidney is subdivided into three distinct compartments: the vascular (plasma), tissue (tubular cell), and urine (tubular lumen) compartments (Fig. 1).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Physiological models for renal elimination of a substrate that is both metabolized and excreted by kidney as the only eliminating organ. The kidney is divided into three compartments: plasma, tissue, and urine. The outflow plasma does not recirculate to the reservoir in the single-pass system. The amount of drug in reservoir (AR) entering the kidney plasma (APK) is first filtered, and only that in the postglomerular circulation reaches the renal tubular cells. Exchange of drug between renal plasma and tissue is characterized by influx (CLinb) and efflux (CLefb) clearances across the renal basolateral (b) membrane, and the exchange between urine and tissue is characterized by CLinl and CLefl at the luminal (l) membrane. Drug within the kidney tissue (AK) is metabolized with a renal metabolic intrinsic clearance, CLint,K. Urinary excretion of drug is a function of urine flow (Qu) and the net transfer clearances at both basolateral and luminal membranes. See Appendix for the definition of the terms. Parameters associated with the metabolite are further qualified by {mi}, and mass transfer for the preformed metabolite was not shown. See Appendix for the definition of other parameters.

	Results

Top Abstract Introduction Materials and methods Results Discussion Appendix References

Protein Binding and RBC Distribution of [³H]Hippurate

The unbound fraction of [³H]hippurate in plasma (f_u{mi}) determined by ultrafiltration at 25°C was 0.48 and was not expected to change dramatically at 37°C due to the low extent of binding. The blood/plasma concentration ratio of [³H]hippurate was 0.84 ± 0.02 (n = 10), which was almost identical with the value of [1  hematocrit] (0.86 ± 0.01) in the same samples (p > .05); the RBC/plasma water distribution ratio (_HA) was zero. The blood/plasma ratio of [¹⁴C]sucrose in the same sample was 0.86 ± 0.02 and was not different from [1  hematocrit]. These observations suggest that hippurate was not distributed into RBCs, as reported elsewhere (Yoshimura et al., 1998).

Isolated Rat Kidney Perfusion (IPK)

Viability of the IPK. Glucose reabsorption of the IPK during the steady-state (35-50 min) single-pass perfusion was 95 ± 4% (mean ± S.D. of five preparations), whereas that for sodium reabsorption was lower (73 ± 11%). The GFR and the urinary flow rate were 0.34 ± 0.09 and 0.11 ± 0.05 ml/min/g, respectively (Table 1). The increase in weight of the perfused right kidney (1.85 ± 0.08 g) to that of the unperfused left kidney (1.54 ± 0.1 g) was only 20 ± 5%. These indices on viability were generally similar to those reported previously for the single-pass rat kidney perfused under constant flow (deLannoy et al., 1989; Poon and Pang, 1995).

Elimination of [¹⁴C]Benzoate, [¹⁴C]Hippurate, and [³H]Hippurate. Steady-state was readily attained in the rat IPK perfused in a single-pass fashion with tracer [¹⁴C]benzoate (30,200 ± 4,100 dpm/ml or 0.86 ± 0.12 µM), shown by the constancy in the output rates of [¹⁴C]benzoate and [¹⁴C]hippurate in venous plasma and urine (Fig. 2A). The steady-state extraction ratio of [¹⁴C]benzoic acid was 0.26 ± 0.04 (mean ± S.D., n = 5, Table 1). However, the urinary clearance of unchanged benzoic acid was very small and variable (0.011 ± 0.01 ml/min/g) with an FE value of 0.27 ± 0.19, and 99% of the renal clearance (1.13 ± 0.17 ml/min/g) was attributed to the metabolic clearance (1.12 ± 0.17 ml/min/g) for glycine conjugation in the formation of [¹⁴C]hippurate. These values agreed with those previously obtained for benzoate given at tracer concentrations to the rat IPK (Poon and Pang, 1995). Almost half of the formed [¹⁴C]hippurate was excreted immediately into urine, and the apparent extraction ratio or E_K{mi} was 0.39 ± 0.09. The ratio of the apparent renal clearance of [¹⁴C]hippurate (E_K{mi}Q_K) to f_u{mi} GFR was 11.3 ± 3.3. 

View larger version (19K): [in this window] [in a new window]   Fig. 2.   The observed output rates of [14C]benzoate (BA) and [14C]hippurate (HA) in venous plasma and urine, after single-pass delivery of [14C]benzoate to the constant flow IPK (8 ml/min) (A), and those for the simultaneously perfused [3H]hippurate (B) were plotted versus the postequilibration perfusion time. The lines were simulated data based on predictions afforded by the optimized parameters shown in Table 3 and the rate equations shown in the Appendix.

Physiological Modeling

The mass transfer rate equations pertaining to drug and metabolite and to the preformed hippurate are summarized in the Appendix. Inversion of the resulting matrices for benzoate and formed hippurate and for preformed hippurate yielded analytical solutions for the steady-state output concentrations of drug and formed/preformed metabolite in venous plasma and urine. These in turn were substituted into eqs. 6 to 8 to provide the total renal and urinary clearance of benzoate, and the extraction ratios of the formed and preformed hippurate (Table 2).

The analytical solution of E_K{mi}, solved for the first time, provided the needed comparison with E_K{pmi} (Table 2). It was found that GFR and parameters associated with the precursor were absent in the solution for the formed metabolite, E_K{mi}, i.e., the extent of excretion of formed hippurate was dependent only on its unbound fraction in plasma and influx and efflux clearances across the renal basolateral and brush border membranes, its unbound fraction, and the plasma and urinary flow rates. In comparison, the extraction ratio for preformed hippurate, E_K{pmi}, was additionally dependent on the GFR. The difference between E_K{pmi} and E_K{mi} became zero when CL_ef^b{mi}/CL_ef^l{mi} = (Q_K  f_u{mi}GFR)/(f_u{mi}GFR) (see Table 2) or 26.089 after substitution of the experimentally observed values of Q_K, f_u{mi}, and GFR into the relationship. At CL_ef^b{mi}/CL_ef^l{mi} > 26.089, E_K{pmi} exceeds E_K{mi}, whereas when CL_ef^b{mi}/CL_ef^l{mi} < 26.089, E_K{mi} exceeds E_K{pmi}. The ratio of E_K{pmi}/E_K{mi} was described by a relation in which the drug transport and metabolic parameters and the reabsorption clearance of mi (CL_in^l{mi}) were absent (Table 2).

Simulations. The values of the volumes of renal plasma, renal tissue, and tubular urine used for simulation were similar to those previously reported (Table 3), whereas the flow rates and unbound fractions were experimentally determined; CL_int,K was set as 40 ml/min/g, a value similar to the ratio V_max/K_m or 195 nmol/min/g/5.3 µM reported for the glycination of benzoate in the IPK (Poon and Pang, 1995). The solutions in Table 2 were used to provide optimized parameter of CL_in^b, CL_ef^b, CL_in^l, CL_ef^l, CL_int,K, CL_in^b{mi}, CL_ef^b{mi}, CL_in^l{mi}, or CL_ef^l{mi}. Formulations of the clearances of drug and formed and preformed metabolites (Table 2) were placed in mathematical worksheets (Excel, version 5; Microsoft, Seattle, WA), and each parameter was varied individually to examine the range of values that would yield predicted values that closely matched the observations. The optimized parameters are summarized in Table 3. The appropriateness of these parameters were shown by the ability of the parameters to provide closely matching urinary venous plasma output rates of benzoate and formed hippurate (Fig. 2A) and of preformed hippurate (Fig. 2B), and clearances and extraction ratios values for benzoate and formed and preformed hippurate similar to those observed (Table 1). The ratio of CL_ef^b{mi}/CL_ef^l{mi} was 4:38 or 0.11, a value much less than 26.089, and was consistent with the view that E_K{pmi} < E_K{mi}.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Simulated ratios of EK{pmi}/EK{mi}, based on analytical solutions and parameter values outlined in Tables 2 and 3, are plotted against ratios of CLefb{mi}/CLefl{mi} and the influx clearance at the basolateral membrane CLinb{mi} (A) or the reabsorption clearance at the luminal membrane, CLinl{mi} (B). The value of CLefl{mi} was fixed (38 ml/min/g) whereas that for CLefb{mi} was varied. The parameters varied are shown on the x- and y-axes.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion Appendix References

The single-pass IPK preparation is an ideal model to examine the dynamic interplay of filtration, metabolism, reabsorption, and secretion of the kidney, and to obtain comparative data on formed metabolite and preformed metabolite kinetics with differentially labeled precursor and preformed metabolite. In parallel, examination with the physiological model provided the theoretical basis on the kinetic behaviors expected of the observed data. The premise of the comparative kinetics and the analytical solutions was based on the lack of interaction between benzoate and preformed hippurate, as well as between the formed and preformed hippurate, and preservation of linearity of the system. The total, metabolic, and urinary clearances and FE value of benzoate (Table 1) were similar to those observed previously when the low benzoate concentration was given alone (Poon and Pang, 1995). The apparent renal clearance of formed hippurate that coexisted with tracer preformed [³H]hippurate in the present study (11.0 ± 3.3) was not different (p > .05) from the apparent renal clearance of hippurate formed from benzoate (9.4 ± 3.6) in previous IPK studies (Poon and Pang, 1995). With these indices for benzoate and hippurate remaining constant, we concluded that the experimental condition of using only tracer concentrations of [³H]hippurate and [¹⁴C]benzoate (micromolar range) is devoid of interactions, thus preserving linearity of the system and justifying use of linear algebra for arriving at the analytical solutions. The view is likely correct since the K_m values (millimolar) for transport are high.

The theoretical examination had advanced our understanding of the difference in kinetics between the formed and preformed metabolite in the kidney (Table 2). The theory is based only on the linearity of the system and is not predicated on the lack of efflux of metabolite back to the circulation or lack of reabsorption of the metabolite, as previously suggested (Garrett, 1978; Tucker, 1981). With these formulations on hand, optimized parameters for benzoate and hippurate transport are readily found (Table 3). These represent reasonably high influx clearances across the basolateral membrane for benzoate and hippurate (13 and 14 ml/min/g), and lower but comparable reabsorptive clearance (CL_in^l or CL_in^l{mi}) (3 and 4 ml/min/g) and efflux clearance (CL_ef^b or CL_ef^b{mi}) from the cell to the circulation (2.5 and 4 ml/min/g). However, at the brush border membrane, a higher secretory clearance (CL_ef^l{mi}) was found for hippurate (38 ml/min/g) than for benzoate (CL_ef^l or 10 ml/min/g). One could further surmise the roles of putative transporters. The recently cloned OAT-K1 transporter for methotrexate (Saito et al., 1996), however, does not seem to be involved. At the basolateral membrane, benzoate and hippurate may share the transporter for PAH (Russel et al., 1990), and this affinity may increase with increasing hydrophobicity and pKa (Ullrich and Rumrich, 1988). A sodium-independent organic anion transporter, ROAT1, which is capable of transporting PAH and urate (Sekine et al., 1997) and that is susceptible to inhibition by probenecid, glutarate, and -ketoglutarate (Sweet et al., 1997), has recently been cloned, but it is unknown whether there is overlapping specificity for benzoate or hippurate with this transport system. By contrast, the involvement of the sodium-dependent lactate transporter for benzoate reabsorption has been suggested (Ullrich et al., 1982; Barbarat and Podevin, 1987). However, PAH (Kullack-Ublick et al., 1994), benzoate, and hippurate are not transported by oatp1, the organic anion transport polypeptide (Pang et al., 1998), which has been immunologically localized to exist on the apical (luminal) membrane of the rat kidney (Bergmann et al., 1996).

The optimized parameters for benzoate and hippurate transport at both the basolateral and brush border membranes and for metabolism of benzoate (Table 3) yielded close approximations of the observed urinary and venous output data (Fig. 2) and renal clearances of benzoate and extraction ratios of the formed and preformed hippurate (Table 1). The solutions further predicted that E_K{pmi} would be more prone to changes in GFR, whereas the condition would not exist for E_K{mi} (Table 2). This was indeed observed for the kinetics of the preformed [³H]hippurate during the early times of perfusion when GFR fluctuated markedly, but the effect was absent for formed hippurate which was excreted to a greater extent than preformed [³H]hippurate (Fig. 2).

Another remarkable observation was that the extraction ratio of formed [¹⁴C]hippurate, E_K{mi} (0.39 ± 0.05) was higher than the renal extraction ratio of preformed [³H]hippurate (E_K{pmi} = 0.24 ± 0.05) (p < .05). As a first impression, a ratio of E_K{pmi}/E_K{mi} of less than unity might be erroneously attributed to low influx of hippurate at the basolateral membrane. In fact, the difference between E_K{pmi} and E_K{mi} for given GFR, f_u{mi} and Q_K values was reliant on the ratio of CL_ef^b{mi} and CL_ef^l{mi}. That E_K{mi} could be greater than E_K{pmi} is therefore not solely dependent on whether barrier-limited influx existed for the metabolite as previously envisioned (deLannoy et al., 1990).

Another interesting observation was the low FE value of benzoate [0.27 ± 0.19, or CL_u,K/(f_uGFR)] would ordinarily suggest net reabsorption of [¹⁴C]benzoate. However, recent examination of renal clearance concepts suggest that extensive intracellular metabolism could mask our ability to detect the net secretory flux of substrates across the basolateral and brush border membranes (Smith and Kugler, 1994; Sirianni and Pang, 1997). This might occur when renal metabolism reduces the FE from values above unity when metabolism is absent to values below unity in the presence of metabolism. Interestingly, when the metabolic intrinsic clearance of benzoate (40 ml/min/g) was set to zero, the FE value of benzoate was increased markedly (from 0.27 to 4.7). From the theoretical projection, the extensive intracellular metabolism of benzoate to hippurate had indeed masked our ability to detect the high secretory potential of the kidney for benzoate. By contrast, for the nonmetabolized substrate, [³H]hippurate, its FE value greatly exceeds unity, and the net handling mechanism is unmistakably due to secretion.

In summary, the renal elimination of [¹⁴C]benzoate and [³H]hippurate at tracer concentrations was studied simultaneously in the single pass IPK preparation. The renal extraction ratio of preformed [³H]hippuric acid was lower than the apparent extraction ratio of the renally formed [¹⁴C]hippurate. Based on analytical solutions provided upon solving mass balance rate equations obtained with a physiological model, the kinetic difference between formed and preformed hippurate was found to be due to the ratio of CL_ef^b{mi}/CL_ef^l{mi} for hippurate than on the influx clearance (CL_in^b{mi}) of the metabolite.