Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

LC is an endogenous compound essential for the transport of long-chain fatty acids across the inner mitochondrial membrane (Bahl and Bressler, 1987; Bremer, 1983). In mammals, the body stores of LC are maintained by dietary intake and endogenous biosynthesis by numerous organs, including the kidney (Bremer, 1983; Rebouche and Engel, 1980; Rebouche and Paulson, 1986). Carnitine supplementation has also been used therapeutically for treating deficiency syndromes, and the compound may also be beneficially used for the treatment of cardiovascular conditions such as ischemic heart disease and angina pectoris (Bahl and Bressler, 1987; Goa and Brogden, 1987). PLC, a short-chain carnitine ester produced by esterification of the hydroxyl group of LC, displays an unique pattern of biochemical activity (Paulson et al., 1986; Siliprandi et al., 1987; 1991) and is currently being evaluated for the treatment of peripheral arterial diseases and other cardiovascular disorders (Bartels et al., 1992; Brevetti et al., 1992; Greco and Mingrone, 1992).

LC and short-chain carnitine esters are highly polar compounds and at physiological pH possess both cationic and anionic functional groups. As a consequence, this class of compound is believed to move across biological membranes primarily by carrier-mediated transport (Bieber, 1988; Gross and Henderson, 1984; Vary and Neely, 1983). In the kidney, transport systems ensure that the body conserves LC, and at physiological plasma concentrations more than 95% of the filtered load is reabsorbed from the renal tubule (Engel et al., 1981; Harper et al., 1988; Mancinelli et al., 1995; Rebouche et al., 1993). ALC also undergoes extensive tubular reabsorption (Mancinelli et al., 1995), and studies with rat kidney brush-border membrane vesicles (Rebouche and Mack, 1984) and the IPK (Gross and Henderson, 1984; Hokland and Bremer, 1986; Mancinelli et al., 1995) indicate that the tubular reabsorption of LC and ALC exhibit typical characteristics of carrier-mediated processes, including saturation kinetics. Interconversion of LC and ALC in the IPK has also been observed (Hokland and Bremer, 1986; Mancinelli et al., 1995), and LC and ALC as locally formed, or `generated' metabolites were found to undergo direct movement into tubular urine. These findings suggest that bidirectional transport systems exist in the kidney and that the site of interconversion is, at least in part, distal to the major sites of tubular reabsorption (Mancinelli et al., 1995).

Currently, very little information exists on the disposition of PLC in the kidney. However, based on previous findings for LC and ALC, the kidney may play an important role in regulating the body stores of endogenous PLC and the pharmacokinetics of exogenous PLC administered for therapeutic aims. Similarly, the kidney may be involved in the metabolism of PLC. In 1986, Hokland and Bremer examined the distribution of LC and several short-chain carnitine esters in kidney extracts, urine and perfusate from IPK experiments in which PLC was added to the perfusion medium as the tritiated compound at a concentration of 30 µM. The findings suggested that PLC was converted to LC and ALC by the kidney. It was claimed that `extensive' tubular reabsorption of PLC was noted in a single perfusion, but no data were provided.

With the recent use of PLC as a pharmacological agent, it is important that the processes involved in renal handling of this compound be clearly understood. Hence, in the present studies, the IPK was used to investigate the renal excretion and metabolism of PLC and to examine the fate of any renally formed metabolites. Studies were performed with deuterium-labeled PLC with analysis by HPLC/MS, which permitted discrimination between renally formed metabolites of the added substrate and the corresponding endogenous compounds released from the kidney during the perfusion experiments.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. L-Carnitine HCl, propionyl-L-carnitine HCl, acetyl-L-carnitine HCl and L-()-4-trimethylammonium-3-chlorobutyric acid (ST 1085, as internal standard) were obtained from Sigma Tau S.p.A (Pomezia, Rome, Italy). PLC-CD₃, ALC-CD₃ and LC-CD₃ were synthesized by the Academy of Wissensshaften (Berlin, Germany) and their purity was >98% as judged by HPLC/MS. ¹⁴C-Carboxyinulin (3 mCi/g) and ¹⁴C-PLC were purchased from New England Nuclear Co. (Boston, MA) and Amersham (Buckinghamshire, UK), respectively. Fraction V bovine serum albumin (Miles Inc., Kaukakee, IL) was dissolved in Krebs-Henseleit buffer and purified by dialyzing against three exchanges of protein-free perfusate buffer over 3 days at 4°C. Before each perfusion, the concentration of bovine serum albumin in the perfusion medium was adjusted to 65 g/l. D-Glucose, L-cysteine, glycine, L-glutamic acid and D-mannitol were obtained from Sigma Chemical Co. (St. Louis, MO). Acetonitrile (UV cutoff 190 nm), methanol and all other chemicals were of analytical grade and were purchased commercially.

Isolated perfused rat kidney. Kidneys of male Sprague-Dawley rats (300400 g; Charles River, Calco, Italy) were perfused as described previously (Mancinelli et al., 1995). Within a thermostatically controlled (37°C) cabinet, a rotary pump (Masterflex model 7518-10, Cole Palmer, Chicago, IL) was used to deliver recirculating perfusate through an 8-µm in-line filter (Millipore, Bedford, MA), a silastic tubing oxygenator, a glass bubble trap and finally a glass cannula inserted into the right renal artery of the excised kidney. Venous outflow was collected directly into the perfusate reservoir and urine was collected via a catheter inserted into the ureter. A constant renal arterial pressure of 110 to 120 mm Hg distal to the tip of the arterial cannula was maintained by adjusting the perfusion flow rate (30-40 ml/min). Urine was collected into preweighed 5-ml polypropylene tubes, and the volume was determined gravimetrically.

Analytical methods. ¹⁴C-Inulin in perfusate and urine was determined radiometrically with a 50-µl aliquot of sample diluted in 4.5 ml of liquid scintillation fluid (Optifluor, Packard, Milan, Italy), and radioactivity was measured by a Packard B1900-TR detector. An external standard method was used for quench correction. Sodium was measured in perfusate and urine by flame photometry (Perkin Elmer Atomic Spectrophotometer 1100B, Monza, Italy), and glucose was determined with use of a commercial kit (Sigma Diagnostic 510).

Protein binding of PLC. The binding of PLC to perfusate protein was determined by ultrafiltration. ¹⁴C-PLC (0.02 µCi) and unlabeled PLC were added to blank perfusate to yield final concentrations of 1, 50 and 500 µM. After incubation at 37°C for 30 min, six replicate aliquots (1.2 ml) of each perfusate sample were transferred to Centrifree tubes (MPS-1, Amicon Co., Danvers, MA) and centrifuged at 1600 × g for 15 min at room temperature. Under these conditions, approximately 300 µl of ultrafiltrate was obtained. Fraction unbound (fu) was taken to be the ratio of the concentration of radioactive ¹⁴C-PLC in ultrafiltrate to that in unfiltered perfusion medium.

Calculations and statistical analysis. The total clearance of PLC-CD₃ by the perfused kidney (CL_T) was estimated as the dose added to the perfusion medium divided by the total area under the perfusate concentration versus time curve from zero to infinite time (AUC_0-), as determined by the trapezoidal method with extrapolation. The half-life (t_1/2) of PLC-CD₃ was estimated from the slope of the terminal phase of the log plasma concentration versus time plot. Apparent renal excretory clearance (CL_R) of PLC-CD₃, ALC-CD₃ and LC-CD₃ was estimated as the rate of excretion into urine divided by midpoint concentration in perfusate. CL_Rave, the time-averaged renal clearance over the entire perfusion period, was calculated as the mean of individual time estimates. The nonexcretory (metabolic) clearance of PLC-CD₃ by the perfused kidney (CL_M) was taken to be the difference between the corresponding CL_T and CL_Rave values. The renal excretory clearance of ¹⁴C-inulin was taken to be GFR and the %TR of PLC, sodium and glucose was calculated with equation 1.

(1)

where fu·GFR represents the contribution of glomerular filtration to renal excretion. For sodium and glucose, fu was assigned a value of unity. For PLC, perfusate-binding experiments indicated that the extent of perfusate binding was negligible (see below) and so fu was again assigned a value of unity.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

In this study, a stable isotope marker was used to study the renal disposition of PLC. In this way, added substrate and its metabolites could be readily distinguished from the corresponding endogenous compounds released from the kidney during the perfusion experiments. Preliminary experiments, in which perfusate containing PLC-CD₃ was recycled through the IPK system at 37°C for 2 hr, in the absence of a kidney, resulted in no detectable LC-CD₃ and ALC-CD₃. These experiments also confirmed that PLC was not taken up by the components of the perfusion system. The binding of PLC to perfusate protein was found to be negligible; the fu determined by ultrafiltration at 1, 50 and 500 µM was 1.05 ± 0.02, 1.07 ± 0.03 and 1.03 ± 0.03, respectively. Therefore, in the analysis of data on the renal handling of PLC, fu was taken to be unity. The functional parameters for the IPK experiments were similar to those reported previously (Mancinelli et al., 1995) with GFR values of 400 to 800 µl/min.

Semilogarithmic plots of the mean (± S.D.) concentration in perfusates of PLC-CD₃, LC-CD₃ and ALC-CD₃ versus time, at initial nominal concentrations of 10 and 200 µM, are shown in fig. 1. Perfusate PLC-CD₃ concentrations decreased in a monoexponential manner with mean half-lives of 90 ± 24 min and 94 ± 11 min for the 10 µM and 200 µM experiments, respectively (table 1). The gradual decrease in the perfusate concentration of PLC-CD₃ (40-50% during the period of perfusion) coincided with a progressive rise in the perfusate concentration of LC-CD₃ and ALC-CD₃ (fig. 1). For example, in the 200 µM experiments, the perfusate concentration ratio of LC-CD₃ to PLC-CD₃ increased from 0.093 ± 0.001 at 10 min to 0.919 ± 0.172 at 92.5 min. The corresponding change for ALC-CD₃ was from 0.014 ± 0.006 to 0.245 ± 0.047. 

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Mean (± S.D.) perfusate PLC-CD3 (open squares), LC-CD3 (open circles) and ALC-CD3 (closed circles) concentrations during perfusion with PLC-CD3 at an initial nominal concentration of 10 µM (a; n = 4) and 200 µM (b; n = 5) in the IPK.

Urinary excretion of PLC-CD₃ and both of the deuterated metabolites was observed during the perfusion experiments at both concentrations. Data on the urinary recovery of PLC-CD₃, LC-CD₃ and ALC-CD₃ are summarized in table 2. The total recovery (sum of PLC-CD₃, LC-CD₃ and ALC-CD₃) represented 3.44% (low dose) and 12.71% (high dose) of the administered dose. The percentage recovery for all three analytes was significantly higher in all cases at an initial concentration of 200 µM.

The total, average renal excretory and metabolic clearance parameters for PLC-CD₃ are summarized in table 1. The CL_Rave values represented only 2.4% (10 µM) and 5.8% (200 µM) of the respective CL_T values. Hence, the major clearance route of PLC-CD₃ by the IPK was nonexcretory. The CL_Rave values for PLC-CD₃ were substantially less than GFR and, at both concentrations, tubular reabsorption was extensive (table 1). Although CL_T and CL_M were not affected by changes in the initial concentration of PLC-CD₃, a significantly higher CL_R and a lower %TR were observed at the higher concentration (table 1).

The individual time-period estimates for the CL_R of PLC-CD₃, LC-CD₃ and ALC-CD₃ for the 200 µM experiments are depicted in figure 2. The CL_R of PLC-CD₃ remained relatively constant over time, although a significant increase (relative to the 0-15 min period) was observed between 30 to 45 and 45 to 60 min (P < .05, ANOVA). In contrast, the CL_R of the generated metabolites, LC-CD₃ and ALC-CD₃ progressively fell (by almost 10-fold) over the perfusion period (fig. 2, b and c). The CL_R values for LC-CD₃ and ALC-CD₃ were substantially higher than those for PLC-CD₃, particularly during the early urine collection intervals, where 10-fold differences were usually observed (fig. 2).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Mean (± S.D.) apparent renal clearance estimates for PLC-CD3 (a), LC-CD3 (b) and ALC-CD3 (c) during successive urine collection intervals in experiments conducted at an initial nominal perfusate PLC-CD3 concentration of 200 µM (n = 5). Dashed line represents GFR values (µl/min). ANOVA (followed by two-tailed Dunnett's tests) was used to test for changes related to the value for the initial (015 min) urine collection interval: * P < .05; ** P < .01.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The results of the present study provide a great deal of insight into the renal handling of PLC and of the time-dependent changes which can occur in the disposition of renally formed metabolites. The most notable findings can be summarized as follows: 1) PLC-CD₃ was found to undergo extensive tubular reabsorption that was concentration dependent; 2) the excretory renal clearance of PLC-CD₃ was small relative to the magnitude of nonexcretory (or metabolic) clearance; 3) PLC-CD₃ was converted to LC-CD₃ and ALC-CD₃ by the perfused rat kidney and both of these metabolites appeared in perfusate and urine; and 4) there was evidence for direct movement of renally formed LC-CD₃ and ALC-CD₃, leading to time- and concentration-dependent changes in the apparent renal excretory clearance values for these metabolites. Each of these pivotal finding are discussed in detail below.

Tubular reabsorption of PLC. The tubular reabsorption of PLC-CD₃ was greater than 90% when the initial perfusate concentration was 10 µM (table 1). This extensive tubular reabsorption, which led to a low urinary recovery, was similar to that observed previously for LC and ALC in the IPK (Mancinelli et al., 1995). At an initial concentration of 200 µM, the %TR of PLC-CD₃ was significantly lower than at the lower concentration (table 1). The magnitude of the change in the overall reabsorption process can be better appreciated by considering that the fraction of filtered PLC-CD₃ escaping tubular reabsorption increased from 10% at 10 µM to nearly 20% at 200 µM. Hence, the tubular reabsorption of PLC is concentration dependent, with partial saturation between 10 and 200 µM. Although these concentrations are substantially higher than the base-line plasma concentrations of endogenous PLC in humans (0.6 µM, Millington et al., 1989; Minkler and Hoppel, 1993), they are similar to those observed in patients with metabolic disorders (Hoppel, 1991) and in individuals receiving oral or intravenous treatment with PLC (Longo and Pace, unpublished observations). Although we also conducted some experiments in which the initial perfusate concentration of PLC-CD₃ was near the normal endogenous level of plasma PLC, extensive tubular reabsorption (>95%) resulted in urinary levels of PLC-CD₃ which were unquantifiable in many cases.

Metabolic and excretory clearance of PLC. During the IPK experiments, a substantial decrease with time in the perfusate concentration of PLC-CD₃ coincided with an increase in perfusate LC-CD₃ and ALC-CD₃ concentrations, as well as urinary excretion of these metabolites. Because no degradation of PLC-CD₃ was observed in the absence of a kidney, the enrichment of perfusate in LC-CD₃ and ALC-CD₃ must have been caused by the metabolic activity of the perfused kidney.

Disposition of the metabolites of PLC-CD₃. When renal drug metabolism occurs, the traditional approach of estimating urinary clearance for the metabolite (e.g., the ratio of urinary excretion rate to the circulating metabolite concentration) may be misleading, because the observed excretion rate of the metabolite is the sum of the excretion rate of the circulating metabolite and that of the metabolite generated intrarenally and subsequently recovered in urine without entering the circulation. Because of the latter component, the measured renal clearance for the metabolite will often result in an overestimation of its true renal excretory clearance (Mancinelli et al., 1995). In addition, after administration of the parent drug, the renal clearance of a metabolite formed in the kidney and excreted into urine, will generally exceed that observed upon administration of the `pre-formed' metabolite (Bekersky et al., 1984; Smith and Kugler, 1994; Von Lehmann et al., 1973; Wan and Riegelman, 1972a, b). The CL<sub>R</sub> results for LC-CD<sub>3</sub> and ALC-CD<sub>3</sub> (fig. 2) are in keeping with these concepts. Thus, considering that the concentration of the recirculating LC-CD<sub>3</sub> and ALC-CD<sub>3</sub> never exceeded 60 µM (fig. 1b), our results provide estimates of the `renal clearance' of LC-CD₃ and ALC-CD₃ (fig. 2,b and c) which are 20 to 50 times greater than previously reported values applicable to the two compounds when added to IPKs in their preformed state (Mancinelli et al., 1995).

Tubular leakage of renally formed LC and ALC. An interesting finding from the present study is the decrease in the CL_R of LC-CD₃ and ALC-CD₃ when the initial nominal concentration of PLC-CD₃ was 200 µM (fig. 2, b and c). It should be noted that a similar finding was observed in the 10 µM experiments, but because the perfusate levels of the metabolites were, in some cases, below the limit of quantification, the data set is not complete at this lower concentration. The decreased CL_R of the metabolites with time can be explained by a change with time in the relative contributions of `generated' and `preformed' (circulating) metabolites to the renal clearance estimates. During the initial time periods, the concentration of the circulating metabolite in perfusate is low relative to that of the precursor. Hence, during this period, the renal clearance estimate essentially provides information on the leakage of the renally generated metabolite, because most of the metabolite appearing in urine is derived from a renally formed source rather than from the perfusion medium. With time, there is an increase in the concentration in perfusate of the metabolite and, as a consequence, there is an increase in the contribution of the `preformed' species to observed renal clearance estimates. Because the reabsorption of the preformed metabolite is very high, the overall CL<sub>R</sub> estimate decreases (fig. 2, b and c). It should be noted that the observation for both metabolites could also be supported by a decrease with time in the perfusate concentration of the precursor (PLC-CD<sub>3</sub>), and proportionally less generation of intrarenal metabolites which are secreted into urine. In reality, it is likely that both factors (i.e., the increase in the perfusate concentration of the metabolite and a reduction in that of the precursor) contribute to the unique pattern observed during our experiments. Certainly, however, the observation has a sound pharmacokinetic basis. Again, this finding may have important implications for the interpretation of the pharmacokinetics of short-chain carnitine esters and their metabolites in humans. Thus, a system could be operating under linear and time-independent conditions with respect to the renal handling of preformed and generated metabolites, but changes with time in the ratio of the precursor to the metabolite in plasma could lead to an apparent `time-dependent' renal clearance.