Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

The kidney is capable of both metabolism and excretion, processes that compete for substrate removal within the eliminating organ. However, the principles of clearance have been re-examined recently, and metabolism and excretion are found to confound estimates of clearance in the presence of each other (Sirianni and Pang, 1997). The confusion lies in the traditional method with which we infer net loss of a compound in the kidney since the urinary (or excretory) clearance, expressed as the fractional excretion, FE¹ (unbound urinary clearance/glomerular filtration rate), is often used to infer the secretory potential of the kidney: FE > 1 implies net secretion, FE < 1 implies net reabsorption, and FE = 1 implies net filtration (Brenner et al., 1986). Excretion is expected to be devoid of effect on the metabolic clearance, and vice versa. However, it was argued that intrarenal metabolism would greatly mask the true secretory ability of the kidney since metabolism might reduce high values of FE to less than unity (Smith and Kugler, 1994). Intuitively, renal metabolism is expected to reduce the unchanged drug appearing in the urine, thereby lowering the FE estimate and affecting interpretation on the net movement of drug across the kidney.

This clearance principle for the kidney has been exemplified with enalapril, an angiotensin converting enzyme (ACE) inhibitor, which is metabolized to a single dicarboxylic acid metabolite, enalaprilat; both are excreted unchanged by the kidney (de Lannoy et al., 1989; Tocco et al., 1982). When viewed theoretically with simulations according to a physiologically based model (Sirianni and Pang, 1997), the suspicion that metabolism of enalapril had reduced the FE value was confirmed. The extent of the decrease, however, was highly dependent on whether intracellular or intraluminal metabolism had occurred. With intracellular metabolism of enalapril, its FE values remain less than unity regardless of whether metabolism occurs. With intraluminal metabolism, FE decreases from a value greater than one (control condition, in the absence of metabolism) to a value less than one in the presence of metabolism.

Consideration of the true secretory ability of the kidney towards enalapril therefore necessitates verification of the site of intrarenal metabolism. The discrimination between intracellular and intraluminal metabolism for this drug is not readily apparent since the solute is transported in both directions across the brush border and basolateral membranes. Methods such as microperfusion (Quamme and Dirks, 1986) or in vitro renal tubule perfusion (Burg and Knepper, 1986) would be useful in determining the location of metabolism for a compound that is not reabsorbed across the brush border membrane. Since these techniques involve direct perfusion of the renal tubule, metabolite appearing in urine would provide evidence of luminal metabolism, whereas the absence of metabolite would suggest intracellular metabolism. The nonfiltering isolated perfused rat kidney preparation (NFK), normally used for assessment of drug transport into renal tubular cells [reabsorption of filtered component across the luminal membrane or via the postglomerular circulation (Gillatt et al., 1990; Minami et al., 1992; Suzuki et al., 1984)], was used for the study of the site of intrarenal metabolism. In this preparation, glomerular filtration is obliterated by ureter ligation and by the high oncotic pressure exerted by 8% albumin in perfusate plasma. Hence, the drug is only able to gain access to the renal tubule indirectly by means of the postglomerular circulation. Since metabolism is a major component of the renal clearance of enalapril, we hypothesized that the site of intrarenal metabolism could be deduced by examining the difference in metabolite data in the nonfiltering (NFK) versus the filtering, isolated perfused rat kidney (IPK). Because of the lack of glomerular filtration and urine flow in the NFK, it is likely that drug and metabolite disposition in the two preparations will differ. Moreover, the expected behavior of the drug and metabolite could be predicted for the IPK and NFK by a physiological model that incorporates either cellular or luminal metabolism in the kidney. The correlation between the expected drug and metabolite data and observations provided the basis for model selection. In this study, we studied the handling of tracer concentrations of [³H]enalapril by the NFK. The derived data were compared with those data on tracer enalapril previously acquired for the IPK (de Lannoy et al., 1989; de Lannoy and Pang, 1993). Furthermore, the renal handling of the metabolite, enalaprilat, is also known (Schwab et al., 1992) such that a set of parameters was available for simulation (Sirianni and Pang, 1997; de Lannoy et al., 1989; de Lannoy and Pang, 1993; Schwab et al., 1992; de Lannoy et al., 1990; Spector, 1956).

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Source of Materials. [³H]Enalapril (specific activity 2,654 µCi/mg), unlabeled enalapril maleate and enalaprilat were supplied by Merck Sharp and Dohme Research Laboratories (West Point, PA). The radiolabeled compounds were purified prior to use, and the radiochemical purity of enalapril was 96% pure as judged by thin layer chromatography (TLC, 1-propanol:1 M acetic acid:water (10:1:1 v/v/v with Silica Gel GF plates obtained from Analtech, Newark, DE).

Kidney Perfusion. Male Sprague-Dawley rats (Charles River, St. Constant, Quebec, Canada; 385 ± 46 g) were allowed free access to food and water. The isolated perfused rat kidney preparation was similar to that used by Johnson and Maack (1977). Isolation of the kidney was performed under pentobarbital anesthesia (50 mg/kg intraperitoneal injection) according to the technique described by Ross et al. (1972), with minor modifications. Both IPK and NFK experiments were conducted. For the IPK, mannitol (50 mg/ml) dissolved in heparin (150 U/ml) was injected into the penile vein, followed by cannulation of the right ureter (PE-50). For the NFK, stimulation of diuresis by mannitol was avoided and the ureter was ligated close to its origin as a precautionary measure to stop urinary flow. The right adrenal artery was tied near its origin at the renal artery. A perfusate-filled cannula (18-gauge stainless steel needle) was inserted into the superior mesenteric artery and guided across the aorta into the right renal artery for immediate perfusion. The right kidney was removed quickly from the rat and placed in a water-jacketed glass holder and maintained at 37°C. Surgery was usually completed within 15 min.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Schematic representation of the single pass nonfiltering isolated perfused rat kidney (NFK) preparation. Perfusate, consisting of bovine red blood cells, amino acids, dextrose, and bovine serum albumin (BSA), is filtered, pumped, and gassed with a mixture of O2:CO2 (95:5 v/v) prior to reaching the renal artery. Venous outflow is not recirculated to the reservoir. Perfusion pressure is monitored by a sphyngomanometer. The obliteration of filtration at the glomerulus is accomplished by a high BSA content (8% albumin) which increases the oncotic pressure in the arterial perfusate at the glomerulus, and by ligation of the ureter.

Protein Binding. The unbound fraction of drug in plasma perfusate containing 8% BSA was determined using equilibrium dialysis at 37°C in a rotating water bath. Preliminary studies revealed that equilibrium was attained at 8 hr. Freshly prepared plasma, identical to that used in the NFK perfusion experiments, was spiked with radiolabeled enalapril and used for equilibrium dialysis. The radioactivity in both plasma and buffer after 8 hr was quantified by dual channel liquid scintillation spectrometry (Beckman Instruments model 6800, Palo Alto, CA). The unbound fraction in plasma (f_p) was estimated by the ratio of the concentration of radiolabeled enalapril in buffer to that in plasma. The leakage of protein from the plasma side of the membrane to the buffer side was determined by the Lowry method (Lowry et al., 1951). Significant volume shifts and protein leakage across the membrane were not observed. The membranes used in the experiments were rinsed with distilled water, and then immersed in scintillation cocktail for -counting. No radioactivity was detected on the membranes. The binding experiments were repeated with 5% BSA to view binding data for previous comparable IPK studies.

Analytical Procedures. It has been shown that neither enalapril nor its metabolite enalaprilat is significantly bound to red blood cells (de Lannoy et al., 1989), so that measurements can be made using plasma concentration and plasma flow rate rather than whole blood concentration and blood perfusate flow rate. The inflow and outflow plasma samples were centrifuged to provide plasma for analysis. TLC was used to separate enalapril and enalaprilat in inflow and outflow plasma samples. Initially, we attempted a protein precipitation-drying procedure prior to plating onto TLC plates. However, band widening and poor resolution resulted. Thus, plasma (100 µl) was directly applied onto the origin of the TLC plates (Silica gel GF; Analtech, Newark, DE), prespotted with unlabeled enalapril and enalaprilat standards. After development of the plates in a system of 1-propanol:1 M acetic acid:water (10:1:1 v/v/v), the entire TLC plate was scraped into 0.5-cm segments, spanning from 1 cm below the origin up to the solvent front. The R_f value (distance compound traveled from origin/distance between origin and solvent front) for authentic enalapril varied from 0.57 and 0.70 and was between 0.31 and 0.39 for enalaprilat, and the radiolabeled compounds closely followed the migration of the unlabeled compounds. After the addition of 0.5 ml of water and 5 ml of scintillation cocktail (Ready Safe, Beckman Instruments), the samples were mixed and then kept in the dark for 24 to 48 hr before -counting. To correct for recovery from TLC, separate aliquots of plasma (100 µl) were also subjected to -counting. The amounts (DPMs) associated with [³H]enalapril and [³H]enalaprilat in the samples were summed and appropriately corrected for the recovery and volume of the sample to provide the concentrations.

Calculations. The renal extraction ratio E_K is expressed as the difference between the steady-state input rate (Q_KC_In) and output rate [(Q_K  Q_u)C_Out] divided by the input rate; C_In and C_Out are the steady-state inflow and outflow plasma concentrations of drug. Q_K and Q_u are the plasma and urine flow rates, respectively. The equation becomes simplified when Q_u = 0 for the NFK.

(1)

Total renal clearance of enalapril was calculated as follows:

(2)

The fractional excretion (FE) or the unbound urinary clearance normalized to GFR, was calculated as follows:

(3)

where C_u and Q_u are the concentration of drug in urine and the urine flow rate (the product yields the urinary excretion rate), respectively; GFR is the glomerular filtration rate, and f_p is the unbound fraction of enalapril in plasma. The venous rate out of enalaprilat, normalized to the rate of input of enalapril, was as follows:

(4)

and equals the ratio of the steady-state output plasma concentration of enalaprilat to the input concentration of enalapril [C_Out{mi}/C_In] since Q_u is zero for the NFK.

Simulations. A physiological model of the kidney, which describes either intracellular or intraluminal metabolism (fig. 2) was considered. In this model, the kidney is subdivided into the vascular, tissue, and tubular lumen (or urine) compartments. The drug is first filtered by the glomerulus before reaching the peritubular cells via the postglomerular circulation, whose flow rate is given by the difference in renal plasma flow rate (Q_K) and the glomerular filtration rate (GFR). Parameters for the transport of enalapril: CL_in^b and CL_ef^b for the influx and efflux clearances at the basolateral membrane (b), and CL_in^l and CL_ef^l for the influx and efflux clearances at the luminal membrane (l), and corresponding transfer clearances for enalaprilat across the basolateral membrane (CL_in^b{mi} and CL_ef^b{mi}) and luminal membrane (CL_in^l{mi} and CL_ef^l{mi}) are described. Intrarenal esterolysis of enalapril occurs intracellularly with CL_int,K, the cellular metabolic intrinsic clearance, or intraluminally, with CL_int,u, the intraluminal metabolic intrinsic clearance.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Physiological models of the kidney depicting either (A) intracellular or (B) intraluminal metabolism of enalapril. Enalapril is eliminated by the kidney via both excretion and metabolism to enalaprilat, which is also excreted. The kidney is subdivided into three compartments: plasma, tissue, and urine. The outflow plasma can either be recirculated back to the reservoir (the recirculating preparation), or drained into a collecting vessel (single-pass preparation). Exchange of drug (D) and metabolite (mi) between renal plasma and tissue is characterized by influx (CLinb, CLinb{mi}) and efflux (CLefb, CLefb{mi}) clearances across the basolateral membrane, and the exchange between urine and tissue is characterized by influx (CLinl, CLinl{mi}) and efflux (CLefl, CLefl{mi}) clearances at the luminal membrane. The arterial plasma and urine flow rates are represented by QK and Qu, respectively. (A) Drug within the tissue is metabolized with a renal metabolic intrinsic clearance, CLint,K. (B) Drug within the urine compartment or renal tubule is metabolized with a renal metabolic intrinsic clearance, CLint,u.

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

Statistics. All data were expressed as mean ± SD. ANOVA was performed on the data; a p value of 0.05 was viewed as significant.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Viability of the Perfused Rat Kidney Preparations. The viability of the IPK preparation was established for the flow rate of 8 ml/min. The extents of sodium and glucose reabsorption were 91 ± 6% and 98 ± 1%, respectively, values which were relatively constant over the 40 min of perfusion under constant flow. These indicators suggest that the viability of the kidney was adequately maintained (de Lannoy et al., 1989). The arterial pressure was 121 ± 34 mm Hg at 8 ml/min, and the increase in kidney weight after perfusion ([weight of right kidneyweight of left kidney]/weight of left kidney) averaged 23 ± 10%. For the NFK, the mean arterial pressure and increase in kidney weight were 113 ± 34 mm Hg and 11 ± 13%, respectively. The percentage of inulin recovered in outflow perfusate was 97 ± 4%, and the plasma clearance, normally used as an estimate of GFR, was virtually zero for the NFK. Apart from the difference in GFR, viability parameters for the IPK and the NFK preparations were not statistically significant (p > 0.05).

Protein Binding. Equilibrium dialysis yielded an f_p value of 0.41 ± 0.01 (N = 4) and 0.32 ± 0.01 (N = 4) for perfusate plasma containing 5% and 8% (w/v) BSA, respectively. Protein leakage from the protein side across the membrane into the buffer side was negligible: leakage was 0.42 ± 0.07% and 0.27 ± 0.01% for the binding studies with 5% and 8% BSA, respectively. The volume shifts across the membrane, and nonspecific binding to the membrane were negligible for both the 5 and 8% BSA.

The Nonfiltering Kidney. The TLC procedure effectively separated labeled enalaprilat from enalapril, albeit there were fluctuations in R_f because of the high albumin content in the samples (fig. 3). The inflow samples contained only radiolabeled enalapril, confirming that esterolysis of enalapril had not occurred in the perfusion medium. The enalaprilat found in the plasma venous samples was necessarily formed by the kidney. The steady-state extraction ratio for enalapril in the NFKs (N = 4) was 0.39 ± 0.03 and was slightly higher than that (0.29) for the IPK (p < 0.05), presumably because of the lower renal plasma flow rate for the NFKs (Table 2). Values of the renal clearance (extraction ratio × plasma flow rate) for the NFK (1.99 ± 0.12 ml/min/g) and IPK (2.1 ml/min/g) (de Lannoy and Pang, 1993) were not different (p > 0.05). The enalaprilat venous output rate, expressed as a percent of input rate of enalapril, was 30 ± 3% for the NFK (Table 2); the parameter was not significantly different from that for the IPK (27 ± 4%; N = 6) (de Lannoy and Pang, 1993).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Profiles of radioactivity recovered on thin layer chromatography plates for inflow and outflow plasma samples. Inflow and outflow plasma samples were assayed for enalapril and enalaprilat by thin layer chromatography (TLC). The top and bottom panels illustrate the radioactive profiles of representative inflow and outflow plasma samples, respectively.

Simulations. Four cases were considered: cellular vs. luminal metabolism for both the IPK and the NFK. According to eqs. 5 and 7, the optimized parameters (table 1) for the cellular metabolism model yielded renal clearances of 2.1 and 1.24 ml/min/g for the IPK and NFK, respectively, whereas the luminal metabolism model predicted renal clearances of 2.1 and 1.26 ml/min/g for the IPK and NFK, respectively (table 2). Moreover, venous plasma output rate of enalapril or enalaprilat, expressed in relation to the input rate of enalapril, were simulated according to eqs. 9 and 12 and presented in fig. 4A for the cellular metabolism model and fig. 4B for the luminal metabolism model. Values of the plasma clearance of enalapril were predicted to be the same for the cellular and luminal models for the IPK (2.10 ml/min/g), and these were greater than those predicted for the NFK (1.24 to 1.26 ml/min/g); the lower renal clearance of the NFK is a result of the lower unbound fraction of drug in the NFK. The enalaprilat plasma output patterns were expected to be different for the IPK (24 vs. 9% input rate of enalapril) with cellular and intraluminal metabolism, respectively. By contrast, values for the accumulation of the metabolite, enalaprilat, in plasma were similar for the NFK for both the cellular and luminal models (24 vs. 26% of input rate of enalapril). Viewing these alternatively, there was no expected difference in metabolite outflow levels between the IPK and NFK with cellular metabolism, but a large difference was expected between the IPK and NFK for luminal metabolism. The venous output of enalaprilat for the NFK was anticipated to be almost 3-fold (289%) that of the IPK (table 2; fig. 4). The experimental data for the renal clearance of enalapril decreased only by 5%, and virtually no difference was observed for enalaprilat plasma output rate for the NFK in comparison with data for the IPK. The predicted/observed output rate of enalaprilat was indeed a sensitive indicator. The correlation was better with the cellular metabolism model.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Simulated data on the output rates of enalapril or enalaprilat as a fraction of the inflow rate of enalapril. (A) Simulations based on the intracellular metabolism of enalapril predicted similar values for the plasma output rates of enalapril and enalaprilat/rate in for the NFK vs. the IPK. (B) Simulations based on the intraluminal model of enalapril metabolism again predicted similar ratios of enalapril rate out/rate in for both models. However, a higher enalaprilat rate out was expected for the NFK in relation to the IPK.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Recently it has been proposed that renal metabolism would mask the true drug secretory capacity of the kidney (Smith and Kugler, 1994). It is envisioned that for a compound that undergoes both metabolism and excretion within the kidney, the processes will compete with one another in the removal of the substrate. Undoubtedly, the excretory and metabolic clearance estimates will be affected by the presence of the competing route of elimination (Sirianni and Pang, 1997; Smith and Kugler, 1994). Since the secretory ability of the kidney is based on the excretion of unchanged drug in the urine (i.e. the FE value), the presence of metabolism within the kidney will decrease the amount of drug present in the urine and therefore mask the true secretory ability of the kidney for that particular compound (Sirianni and Pang, 1997).

Enalapril undergoes esterolysis within the kidney to form its dicarboxylic acid metabolite enalaprilat. The question that arises is whether enalapril is metabolized intracellularly or intraluminally within the rat kidney. Indeed, previous simulations based on renal physiological models for enalapril had shown that metabolism within the renal tubule (by ecto-enzymes along the brush border membrane) could decrease the FE value from above one (net secretion) in the absence of metabolism to a value below one (net reabsorption) when metabolism occurs (Sirianni and Pang, 1997). Metabolism would affect the interpretation of the secretory ability of the kidney. If metabolism of enalapril occurred intracellularly, the interpretation on net reabsorptive flux of enalapril would remain tenable since the FE values were all below unity (Sirianni and Pang, 1997). The converse, that is, when metabolism occurred intraluminally, would not hold true; the large change in FE induced by metabolism would have resulted in erroneous interpretation that there was net reabsorption. To address this question, we performed a set of NFK experiments, compared the results with those of the IPK (de Lannoy et al., 1989; de Lannoy and Pang, 1993), and further examined how these data correlated to the simulations based on physiological models for either cellular or luminal metabolism. A necessary first step experimentally was the establishment of the viability of perfused kidneys (both IPK and NFK). Since the number of viability indices is reduced in the NFK because of lack of urine, viability of the IPK preparation was first established; sodium and glucose reabsorption as well as kidney weight changes and perfusion pressure were used as markers of viability. Comparable weight changes and perfusion pressures were observed for the IPK and NFK. The second necessary step was the examination of protein binding for enalapril at 8% albumin for the NFK studies; that for enalaprilat did not enter into the formulation (Geng and Pang, unpublished equation). The unbound fractions were found to be 0.32 for 8% albumin and 0.41 at 5% albumin in the present study, and f_p for 5% albumin was lower than that (0.55) observed by de Lannoy et al. (1989). The difference might have been caused by the batch of albumin used. Thus, the respective f_p value was used for the simulations. A notable observation was the lower plasma flow rate for the NFK preparations in relation to that for the IPKs (table 2). The unbound renal metabolic clearance for the IPK (3.53 ml/min/g) previously observed by de Lannoy and Pang (1993) was also lower than that for the NFK (4.91 ml/min/g). The difference could not be attributed to the difference in flow rates. However, a higher metabolic clearance value was anticipated for the NFK because of the absence of excretion (Sirianni and Pang, 1997), a new paradigm to be reckoned with in clearance concepts. It could further be surmised that the performance of the erythrocyte-perfused kidney preparations was improved at the lower flow rate (8 vs. 10 ml/min), as also shown by the improved viability of the present IPKs over previous ones (de Lannoy et al., 1989).

The simulations showed that the E_K and total renal clearance of enalapril (CL_tot,K) were poor indices for the discrimination of cellular vs. luminal metabolism (Table 2). The ratio of prediction/observation varied from 0.6 to 1 in total renal clearance or renal metabolic clearance predicted for the NFK and IPK. However, the models were discriminated based on metabolite data: the enalaprilat rate out in plasma, normalized to the input rate of enalapril (fig. 4; table 2); the parameter was independent of the unbound fraction of enalaprilat. The predicted/observed ratios were between 0.79 and 0.89, with the exception of 0.34 for the IPK with the luminal model. The ratio is an aberrant value. Enalaprilat, if formed luminally in the filtering kidney, is expected to be carried down the renal tubule for rapid excretion out into urine rather than being reabsorbed. The poor correlation suggests that enalapril delivered to the kidney is unlikely to be metabolized by enzymes on the luminal membrane. The observation appears to be consistent with the distribution of enzymes within the kidney, since only peptidases are known to be enriched on the luminal membrane (Carone et al., 1982; Kugler et al., 1985). Yet the presence of carboxylesterases on the luminal membrane has not, to our knowledge, been reported. With the notion of intracellular metabolism of enalapril, the interpretation on the net reabsorption of enalapril by the kidney is justified (de Lannoy et al., 1989).

The above treatment was based on the assumption that transport of enalapril across the basolateral and brush border membranes was not altered between the IPK and NFK. On the contrary, Kim et al. (1992) had studied the renal handling of the model anion, para-aminohippurate, PAH, in both the IPK and NFK and showed a decrease in luminal efflux of PAH in the NFK to 1/3 the value observed in the IPK. If the situation had prevailed for the present NFK studies, decreases would have resulted for the NFK. The decrease would only affect the output rate of enalaprilat for the luminal model (enalaprilat output rate would be halved) and not the cellular model. Since there was no decrease observed for enalaprilat outflow, reduction in secretory clearance of enalapril had not occurred for the NFK.

In summary, single pass studies were performed with tracer [³H]enalapril with the nonfiltering kidney, which provided data (extraction ratio, renal and metabolic clearances, and enalaprilat output rate) similar to those obtained previously for the IPK (de Lannoy et al., 1989). The "luminal metabolism model" which predicted an increase of the output rate of enalaprilat for the NFK vs. the IPK, did not agree well with the observations. The composite observations for the IPK and NFK correlated better with the expected changes with the "cellular metabolism model." The data suggest that the metabolism of enalapril is intracellular rather than intraluminal. The previous assumption on net reabsorption of enalapril by the kidney (de Lannoy et al., 1989) seems justified; namely, the metabolism of enalapril has not altered the interpretation on the mechanism of excretion of enalapril (Sirianni and Pang, 1997). Future studies with renal membrane vesicles will allow for the direct assessment of the metabolic capability of brush border membrane enzymes for enalapril.