Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

When kidney function is impaired, a variety of pathological changes occur that are collectively referred to as the uremic syndrome. A part of this syndrome involves elevated serum levels of a number of substances (Ringoir et al., 1987). In patients who have chronic renal failure, uremic toxins accumulate in the serum by a combination of the following four mechanisms: 1) a decrease in renal clearance (indoxyl sulfate), 2) an accumulation of abnormal metabolites (methylguanidine), 3) an increase in production (parathyroid hormone), and 4) a decreased rate of catabolism by the kidney (₂-microgloblin) (Niwa, 1996).

3-Carboxy-4-methyl-5-propyl-2-furanpropanoic acid (CMPF), a furan dicarboxylic acid derivative, was first detected in normal human urine in 1979 (Spiteller and Spiteller, 1979) and was subsequently isolated and identified in human blood (Pfordt et al., 1981). Since these initial reports, it has also been reported to accumulate in uremic plasma, reaching concentrations in the range of 60 to 370 µM (Niwa et al., 1988). In addition, a growing body of evidence suggests that it is a significant uremic toxin (Niwa, 1996).

CMPF has a high affinity for albumin (number of binding sites, n = 1; apparent association constant, K_a = 1.3 × 10⁷ M¹) and inhibits the binding of other ligands, especially those that bind to site I (Lindup et al., 1986; Mabuchi and Nakahashi, 1988b; Henderson and Lindup, 1990; Sakai et al., 1995). It is believed to be a major factor in the decreased level of drug binding in uremic plasma because of its considerable affinity for albumin and also because of increased CMPF concentrations in uremic plasma. The concentration of CMPF can approach a 1:1 molar ratio with albumin, particularly when the production of albumin is reduced during chronic renal failure.

It has also been suggested that CMPF may play a role in a variety of pathological conditions, including the anemia that occurs during chronic renal failure (Niwa et al., 1990; Costigan et al., 1995), irregularities in thyroid function (Lim et al., 1993), and neurological symptoms that may be caused by the inhibition of organic anion transport at the blood-brain barrier (Costigan et al., 1996a). This furan dicarboxylic acid also inhibits phase I (O-demethylation) and phase II (glutathione conjugation and glucuronidation) pathways of drug metabolism in rabbit liver homogenates in vitro (Walters et al., 1995). In addition, there is evidence that CMPF inhibits active tubular secretion in the kidney (Henderson and Lindup, 1992). Thus, CMPF can be classified as a uremic toxin and, as a result, a compound of pharmacological interest.

Nevertheless, the mechanism by which CMPF accumulates remains unclear. To understand the pathway for the accumulation of CMPF, we have examined the pharmacokinetic properties of CMPF after intravenous administration and compared them with those of other uremic toxins, namely, indoxyl sulfate (IS), indole acetic acid (IA), and hippuric acid (HA). The renal and biliary excretion of CMPF and its tissue distribution after intravenous administration to the anesthetized rat have also been studied. We have also investigated the uptake of CMPF by renal cortical slices in vitro to see whether an organic anion/dicarboxylate exchanger was involved in the renal excretion of CMPF.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials. CMPF was synthesized according to the method of Costigan et al. (1996b), which was based on that of Pfordt et al. (1981). IS, IA, HA, p-aminohippuric acid (PAH), glutarate, and succinate were obtained from Nacalai Tesque (Kyoto, Japan). -Ketoglutarate (-KG) was obtained from Sigma-Aldrich (St. Louis, MO). Tetraethylammonium (TEA) was obtained from Wako Pure Chemicals (Osaka, Japan). Malonate was purchased from Tokyo Kasei (Tokyo, Japan), and methylsuccinate was bought from Aldrich Chemical Co. (Milwaukee, WI). All other chemicals were of analytical grade.

Pharmacokinetics of Uremic Toxins in the Anesthetized Rat. Male Wistar rats (250-290 g; bred and maintained in the departmental animal facility) were anesthetized with sodium pentobarbital (60 mg/kg) by intraperitoneal injection, and the left femoral vein and artery were then cannulated.

Determination of the Free (Unbound) Concentration of Uremic Toxins. The serum-free concentrations of uremic toxins were estimated by ultrafiltration as described previously (Tsutsumi et al., 1999). The free fractions of uremic toxins were determined according to the following equation:

(1)

where f represents free fraction of uremic toxin, C_f represents the free concentration, and C_t represents the total concentration of uremic toxin, respectively.

Urinary and Biliary Excretion. The urinary excretion and biliary excretion of CMPF were measured to find the major excretory pathway of CMPF. Male Wistar rats (250-290 g; bred and maintained in the departmental animal facility) underwent a surgical procedure under light anesthesia with phenobarbital where cannulae were inserted into the femoral vein and artery, using polyethylene tubing (polyethylene-50; 0.58 mm i.d.; 0.9655 mm o.d.; BD Biosciences, Parsippany, NJ). The bile duct was also cannulated with polyethylene tubing (polyethylene-10; 0.28-mm i.d.; 0.61-mm o.d.), as was the bladder (polyethylene-8; 2.33-mm o.d.; Hibiki Co., Tokyo, Japan). The body temperature of the rats was maintained by heat from a lamp. Thirty minutes before the i.v. injection of CMPF, control samples of bile and urine were collected. Bile and urine were collected at 0 to 60, 60 to120, 120 to 180, 180 to 240, and 240 to 300 min.

Tissue Distribution of CMPF. Rats were weighed and cannulated as described above. CMPF was then administered via the femoral vein, and 1 h later the rats were sacrificed by decapitation. The brain, heart, lungs, liver, spleen, kidneys, and testes were removed and weighed. A sample (0.5 g) of each tissue was homogenized in 3 ml of 1 M KH₂PO₄, and the CMPF therein was extracted using the procedure described for the preparation of plasma samples. The distribution of CMPF in each tissue is expressed as the concentration of CMPF per gram of each tissue divided by the concentration of CMPF in plasma (i.e., the C_t/C_p ratio).

Uptake by Rat Renal Cortical Slices. The uptake of CMPF and PAH by rat renal cortical slices was investigated using the procedure described by Henderson and Lindup (1992). Rats were anesthetized and the kidneys promptly removed, decapsulated, and placed in an ice-cold oxygenated rinse medium, which consisted of 97 mM NaCl, 40 mM KCl, 0.74 mM CaCl₂ · 2H₂O, and 7.5 mM sodium phosphate-chloride buffer, pH 7.4. Renal cortical slices (weight 10-20 mg/slice; about 0.5 mm in thickness) were cut freehand with Gillette valet strip blades to about 3 inches in length (Sabre International Products Ltd., Reading, UK). Two cortical slices were prepared from each half-kidney and were stored in the oxygenated rinse medium on ice for no longer than 15 min before the start of incubation. Two slices were placed in each flask and the medium (which consisted of rinse medium, 10 mM L-lactate, and 10 mM L-pyruvate) in each flask was thoroughly gassed with 100% oxygen for about 1 min both before and after the addition of the slices. A concentration of 20 µM was chosen for the substrate (CMPF) because this is the concentration normally found in healthy humans (Niwa et al., 1988). The flasks were then tightly sealed with rubber stoppers and incubated in a shaking water bath at 60 cycles/min for 60 min at 25°C. After the incubation, the flasks were placed on ice, the slices promptly removed from the flasks, gently blotted, and weighed. Tissue blanks were prepared by omission of CMPF from the incubation medium.

HPLC Conditions. The HPLC system consisted of an L-6200 intelligent pump (Hitachi, Tokyo, Japan) and either an F-1050 fluorescence spectrophotometer or L-4000 UV detector (Hitachi). A column of LiChrosorb RP-18 (Cica Merck, Tokyo, Japan) was used as a stationary phase. The mobile phase consisted of acetate buffer (0.2 M, pH 4.5)/acetonitrile (93:7, v/v, for HA; 65:35, v/v, for IA and IS), acetate buffer (0.2 M, pH 4.5)/acetonitrile]/acetic acid (60:40:0.5, v/v, for CMPF), and methanol/acetonitrile/tetrahydrofuran/acetate buffer (70 mM, pH 4.0), which contained 3 mM tetra-n-butylammonium bromide (6.5:2:0.2:91.3, v/v, for PAH). The flow rate was 1.0 ml/min. HA, CMPF, and PAH were detected by UV at 240, 261, and 254 nm, respectively. IA and IS were detected by means of a fluorescence monitor. The excitation/emission wavelengths were 280/375 nm, respectively, for both IA and IS. The coefficients of variation of the HPLC methods were similar (<5%).

Data and Statistical Analysis. Plasma concentration profiles were analyzed by fitting the following biexponential equation with the nonlinear least-squares method (MULTI) (Yamaoka et al., 1981).

(2)

Pharmacokinetics parameters were calculated using the following equations.

(3)

(4)

(5)

where area under the curve (AUC)_0, CL_tot, and t_1/2 represent AUC from zero to infinity, total body clearance, and half-life of the  phase, respectively.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The recovery of authentic CMPF, IS, IA, and HA from rat plasma was greater than 95% in each case. Normal rat plasma contained substances that have chromatographic characteristics that are identical to CMPF, IS, IA, and HA, and the mean (±S.E.) endogenous concentrations of these substances were 0.85 ± 0.06, 1.62 ± 0.17, 0.25 ± 0.04, and 1.25 ± 0.14 µg/ml, respectively (n = 5).

Pharmacokinetics of Uremic Toxins. Fig. 1 shows the plasma-time concentration profiles for the uremic toxins after administration of doses of 5 mg/kg to the rats. The plasma clearance of CMPF was the lowest of all the uremic toxins tested in this experiment (Table 1). The biological half-life (t_1/2) was 356 ± 18 min (Table 1), and this value seems to reflect its high affinity for albumin (the percentage bound was about 98%; Table 1). IS, however, which had a similar unbound fraction (2%) to CMPF in rat plasma under the conditions of this experiment, was eliminated from plasma much more rapidly than CMPF. Other uremic toxins, including IA and HA, had a moderate plasma clearance and a lower affinity for albumin than either CMPF or IS.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Plasma concentration-time profiles for a series of uremic toxins. All uremic toxins (, CMPF; , IA; , HA; and , IS) were administered at a dose of 5 mg/kg by rapid infusion into the femoral vein. Blood samples were taken from the femoral artery at 1, 3, 6, 15, 30, 60, 90, 120, 180, 240, and 300 min. Each point represents the mean ± S.E. of four different experiments.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Tissue distribution of CMPF after i.v. administration. CMPF was administered at a dose of 5 mg/kg via the femoral vein and 1 h later various tissues were removed. The distribution of CMPF in each tissue is expressed as the Ct/Cp ratio, i.e., the concentration of CMPF per gram of each tissue divided by the concentration of CMPF in plasma (28.5 µg/ml) at 1 h after administration.

Mechanism of Uptake of CMPF by Renal Cortical Slices. The uptake of CMPF by renal cortical slices was examined next. In the presence of oxygen, the slice-to-medium ratio was 4.6 ± 0.5 (n = 3), whereas under anaerobic conditions (in the presence of nitrogen), this value was 1.1 ± 0.2 (n = 3). These results suggest that active transport is involved in the uptake of CMPF by renal cortical slices and provides further support for the above-mentioned in vivo data. The active uptake of CMPF by renal cortical slices was linear for at least 20 min (data not shown). To estimate the K_m and V_max values for the active uptake of CMPF from the initial rate, the uptake velocity of CMPF over a range of concentrations (20-320 µM) was measured after incubation for 20 min at 25°C in three separate experiments. Nonspecific uptake measured in the absence of oxygen was subtracted from the uptake in the presence of oxygen. The mean K_m and V_max values for the active uptake of CMPF were 98.1 ± 15.2 µM and 36.7 ± 10.2 nmol/min/g kidney, respectively. These values were smaller than those reported by Henderson and Lindup (1992) (K_m = 194 µM; V_max = 55 nmol/min/g kidney) who used a longer incubation time of 90 min. In the present study, K_m and V_max values were calculated from the initial rate and therefore the affinity of the renal uptake system for CMPF may be higher than reported previously.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Dose-dependent inhibitory effect of PAH and TEA on the uptake of CMPF by renal cortical slices. The values are expressed as a percentage of 20 µM CMPF uptake by renal cortical slices at 25°C for 1 h in the absence of inhibitor. The ordinate represents the percentage of control renal uptake for CMPF, and the abscissa represents the logarithm of the concentration of PAH () and TEA () in the incubation buffer. Each point represents the mean ± S.E. of four to five separate experiments. Each slice was prepared from a different rat. P < 0.05 (a), P < 0.01(b), significantly different from corresponding control CMPF uptake.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Effect of CMPF on the uptake of PAH by renal cortical slices. The values are expressed as a percentage of 20 µM PAH uptake by renal cortical slices at 25°C for 1 h in the absence of CMPF. Each column represents the mean ± S.E. of three different experiments. P < 0.05 (a), P < 0.01 (b), significantly different from control.

Characterization of the CMPF Transport System in Renal Cortical Slices. It is well known that the active transport of PAH across the basolateral membrane occurs as the result of the coupling of two transport processes: Na⁺-dicarboxylate symport and PAH-dicarboxylate exchange (Pritchard, 1995; Orlov Yu, 1997). Because of the mutual inhibition between PAH and CMPF in the renal cortical slice experiments, the same mechanism seems to be involved in the uptake of CMPF. To confirm whether the uptake of CMPF might be mediated by the same transport system, we investigated the effect of -KG on the uptake of CMPF by rat renal cortical slices.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Stimulation of CMPF uptake in renal cortical slices by external -KG. CMPF uptake was assessed after a 30-min incubation with 20 µM CMPF and 0 to 400 µM -KG at 25°C in the modified Cross and Taggart saline. Lithium (5 mM) completely blocked the stimulation produced by 50 µM -KG. Each column represents the mean ± S.E. of four to five different experiments. a, P < 0.01 versus control (no added -KG). b, P < 0.01 versus 50 µM -KG.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The half-life of CMPF was longer than that of the other uremic toxins investigated in this study, in which IS was eliminated the most rapidly. CMPF and IS were bound to nearly to the same extent in rat plasma in these experiments (Table 1), and the data in Table 3 show that the elimination of CMPF and IS from plasma is a function of urinary excretion. We conclude that this difference in rate of elimination is based on differences in affinity for the renal active transport system.

The elimination of IS was more rapid than that of IA, which, like IS, is an indole derivative. IS contains a sulfate group, unlike IA. Sulfate conjugates, such as estrone sulfate and dehydroepiandrosterone sulfate, are high-affinity substrates for the various organic anion transporters (Kusuhara et al., 1999; Cha et al., 2000; Van Aubel et al., 2000). This points to the importance of the sulfate group in the recognition of compounds and their excretion by the transport system. The total clearance of IS was the same as renal clearance, and the renal clearance of IS divided by the free fraction (f = 0.02), which was much higher than that of the other uremic toxins, was about 41 times greater than the GFR (7.32 ml/min/kg; Trumper et al., 1998). This suggests that either the transport mechanism for IS or the affinity of IS for the organic anion transporters may differ from that for carboxylate derivatives (IA, HA, and CMPF).

The elimination of CMPF, a dicarboxylic acid, was slower than that of IA and HA, which are monocarboxylic acid derivatives. Among these carboxylic acid derivatives, the rank order of total renal clearance was the same as that of the free fraction, and therefore renal clearance of the free fraction of these carboxylic acid derivatives was the same (Tables 1 and 3). This suggests that plasma protein binding affects the rate of glomerular filtration and also the concentration of the carboxylic acid derivative that can access the transporter at the basolateral membrane, i.e., the higher the protein binding, the lower is the glomerular filtration rate and total renal clearance. The present data indicated that CMPF is almost totally excreted via the kidney (Table 2), mostly unchanged. CMPF can inhibit phase I (O-demethylation) and phase II (glutathione conjugation and glucuronidation) pathways of drug metabolism in the liver (Mabuchi and Nakahashi, 1988a; Walters et al., 1995). It seems that CMPF may act as an inhibitor of drug metabolism but not be metabolized itself. The renal transport system may therefore play an important role in the elimination of CMPF.

Because CMPF has a high affinity for albumin and is effectively excreted in urine, it is likely that an active transport system, such as an organic anion transporter is involved in its elimination from the kidney. It is reasonable to assume that CMPF is mainly eliminated by this mechanism because CMPF is an organic acid that structurally possesses the required hydrophobic area and two negative partial charges (Ullrich and Rumrich, 1988; Ullrich, 1997). Costigan and Lindup (1996) reported that the plasma clearance of CMPF is decreased as a result of the coadministration of PAH and probenecid. In addition, mutual inhibition between PAH and CMPF was observed in slice uptake experiments (Figs. 3 and 4), suggesting that the same transport system is shared with both PAH and CMPF.

The first step in active secretion is the extraction of organic anions from the peritubular blood plasma by the proximal tubule cells through the basolateral membrane. This basolateral uptake of organic anions has been extensively investigated with PAH as the test substrate. According to the proposed model (Pritchard, 1995; Orlov Yu, 1997), the first stage is the formation of a Na⁺ gradient as the result of the hydrolysis of ATP by Na⁺,K⁺-ATPase. This Na⁺ gradient is a driving force for the Na⁺-dicarboxylate symport, which is sensitive to lithium (Pritchard, 1990; Pajor, 1999). As a result, an anion gradient (dicarboxylate) develops, which is directed from the cell. During the last step, the dicarboxylate anion is exchanged for PAH by the anion exchange transporter. Our observations also suggest that CMPF is recognized by an organic anion/dicarboxylate exchanger because the renal uptake of CMPF was markedly stimulated by 50 µM -KG, and this stimulation was prevented by not only the presence of 400 µM -KG but also the addition of 5 mM lithium (Fig. 5). Based on these observations, we propose that an organic anion/dicarboxylate exchanger is involved in the uptake of CMPF, in the same manner as for PAH.

To obtain further evidence about whether the organic anion/dicarboxylate exchanger was responsible for the uptake of CMPF in the kidney, the ionic dependence of CMPF uptake was investigated (Table 5). Replacement of Na⁺ with an equimolar concentration of cations such as K⁺, choline, and lithium inhibited the uptake of CMPF by 38.9 ± 2.3, 43.5 ± 7.2, and 31.2 ± 8.7%, respectively. These results support the view that the organic anion/dicarboxylate exchanger could be involved in the transport of CMPF in the kidney, because the organic anion/dicarboxylate exchanger is coupled with Na⁺-dicarboxylate symport. Furthermore, CMPF uptake by renal cortical slices was strongly inhibited by -KG and glutarate, which are substrates and/or inhibitors of the organic anion transport system and Na⁺-dicarboxylate symport (Sekine et al., 1997; Uwai et al., 1998; Kekuda et al., 1999) (Table 6). These results suggest that the organic anion transport system could play an important part in the uptake process for CMPF in the kidney.

On the other hand, it is possible that CMPF is transported by Na⁺-dicarboxylate symport because CMPF is a dicarboxylic acid. Transport systems for dicarboxylates have been studied in the vesicles from brush-border membranes and basolateral membranes (Wright, 1985; Wright and Wunz, 1987), mostly with succinate as a test substrate. Succinate is a high-affinity substrate for the Na⁺-dependent dicarboxylate transporter (NaDC3; K_t of succinate = 2 µM), which is present in the rat kidney (Kekuda et al., 1999). However, succinate exhibited only a moderate inhibitory effect on the uptake of CMPF by renal slices (Table 6). Furthermore, no inhibitory effect of 1 mM CMPF on the uptake of [¹⁴C]succinate by renal slices was observed (data not shown). These data suggest that CMPF and succinate would be transported by different carriers, and the contribution of Na⁺-dicarboxylate symport to the uptake of CMPF in the kidney would be very low. Considering the relatively little inhibition produced either by the replacement of Na⁺ (Table 5) or by succinate (Table 6), the effects on the uptake of CMPF may be secondary to the inhibition of Na⁺-dicarboxylate symport. This points to the importance of the organic anion transport system for CMPF excretion in the kidney. However, further work is needed to identify the CMPF transporter, and experiments with expression systems are needed to assess the contribution of individual transporters.

Recently, Sekine et al. (1997) reported the functional expression cloning of an organic anion/dicarboxylate exchanger (OAT1) and its characteristics as a multispecific organic anion transporter. They demonstrated that OAT1 plays an essential role in the elimination of numerous organic anions and that an outwardly directed dicarboxylate gradient is essential for expressing transport activity (Sekine et al., 2000). The necessity of the dicarboxylate gradient for the transport of CMPF suggests that OAT1 plays an important role in the elimination of CMPF. Moreover, other isoforms, named OAT2 and OAT3, are also expressed in the kidney and are sensitive to PAH (Sekine et al., 1998; Kusuhara et al., 1999; Kojima et al., 2002). In particular, CMPF may be a substrate of OAT3, because CMPF is a potent inhibitor of OAT3 expressed in oocytes (IC₅₀ of CMPF = 4 µM) (Deguchi et al., 2002). These transporters may be responsible for the renal uptake of CMPF, because there was mutual inhibition of slice uptake between CMPF and PAH in the uptake experiments (Figs. 3 and 4; Henderson and Lindup, 1992).

In uremia, the accumulation of CMPF has various effects, i.e., inhibition of protein binding (Sakai et al., 1995), inhibition of drug metabolism (Walters et al., 1995), and inhibition of renal excretion (Henderson and Lindup, 1992; Walters et al., 1995). CMPF may therefore cause other uremic metabolites and/or drugs to remain in circulation and elevate their concentrations to toxic levels. Interestingly, however, the free fraction of CMPF was low, but its concentration in the kidney was almost the same as that of plasma (Fig. 2). Based upon this observation, CMPF may affect pharmacokinetics of drugs by inhibition of metabolism and renal excretion. This leads us to the suggestion that organic anion transport systems play an important role not only the elimination of CMPF via the kidney but also in mediating the uremic toxicity of CMPF. Furthermore, it has been reported that an organic anion/dicarboxylate exchanger also exists in the brain (Adkison and Shen, 1996; Sekine et al., 2000) as well as the kidney. These transporters also recognize drugs and endogenous compounds, and so a study of the effects of CMPF on such transporters will lead to a better understanding of not only drug-uremic toxin interactions but also the relationship between the uremic syndrome and uremic toxins.

In conclusion, we have found that CMPF is cleared more slowly from rat plasma than indoxyl sulfate, hippuric acid, or indole acetic acid and that the highest tissue concentration was found in the kidney. Furthermore, using renal cortical slices, we conclude that the uptake of CMPF would be mediated by an anion/dicarboxylate exchanger that is similar to that for PAH.