Introduction

Abstract Introduction Results Discussion References

Many chemical compounds have contained carboxylic acid function in their molecules. These compounds are metabolized via various pathways such as glucuronic acid conjugation, glycine conjugation, taurine conjugation, -oxidation and, very rarely, carnitine conjugation (fig. 1). Among these various metabolic pathways, glycine conjugation is a very common route for xenobiotic carboxylic acids. For example, in humans, sodium salicylate or aspirin is conjugated with glycine to form salicyluric acid (1). Benzoic acid is eliminated exclusively via conjugation with glycine in both humans and rats (2); however, this elimination is also dependent on the dose (3). Endogenous organic acids, such as isovaleric acid which accumulates in organic acidemia (4) and hypoglycine which is the causative agent of Jamaican vomiting sickness (5), are also subject to glycine conjugation. The glycine conjugation reaction occurs in both liver and kidney; however, the liver is quantitatively the most important organ because the renal medulla has only low synthetic activity and the mass of liver considerably exceeds that of the kidneys (6).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Metabolic pathway of carboxylic acid compounds.

Carnitine conjugates are well known in the field of endogenous metabolism. The biological importance and roles of carnitine have been reported as essential cofactors in fatty acid oxidation (7) in which carnitine acts as a carrier of acyl groups to transport fatty acids into the mitochondrial inner membrane and also as a modulator of the ratio of acyl CoA/CoASH (8). However, recent reports have suggested that carnitine conjugation plays an important role in xenobiotic metabolism. Pivalic acid (PA), whose oxymethyl ester derivatives are used in drugs to improve intestinal absorption (9, 10), is metabolized by conjugation with glucuronic acid, glycine, and carnitine in laboratory animals (11, 12). In humans, 90% of PA is excreted as carnitine conjugates into urine (13). Other examples are cycloprate (14-16), valproic acid (17), and benzoic acid (18) which have been reported as metabolites of carnitine conjugation.

Taurine conjugate is also an important metabolic route for xenobiotic carboxylic acids; however, it usually represents only a small amount of the total metabolism in rats (19). Three exceptions to date are 2-naphthylacetic acid (20), trimoprostil (21), and 3,4-dichlorobenzyloxyacetic acid (19).

In this study using rat hepatocytes and kidney slices, we focused on identifying the carboxylic acid, which is the best substrate for carnitine conjugation, and compared it with glycine conjugation, which is a competitive reaction through the CoA thioester route.

The compounds we examined were the branched-chain carboxylic acids (PA, isobutyric acid (IB)), the cyclic side chain carboxylic acids (cyclopropanecarboxylic acid (CPCA), cyclobutanecarboxylic acid (CBCA), cyclopentanecarboxylic acid (CPECA), cyclohexanecarboxylic acid (CHCA)), and benzoic acid (BA). As most of these compounds are not thought to be metabolized by oxidation, they should be appropriate substrates for studying the conjugating reactions.

Materials and Methods

Chemicals. (2-¹⁴C)-Glycine (specific activity of 3.7 MBq/mg and 99% radiochemically pure) and L-(N-methyl-¹⁴C)-carnitine hydrochloride (specific activity of 3.7 MBq/mg and 99% radiochemically pure) were obtained from Du Pont Company (Wilmington, DE). PA and BA were purchased from Nacalai Tesque Inc. (Kyoto, Japan), and IB, CPCA, CBCA, CPECA, CHCA from Aldrich Chemical Company (Milwaukee, WI). (Carbonyl-¹⁴C)-PA sodium (specific activity of 758.6 KBq/mg and 98% radiochemically pure) was synthesized by T. Nagasaki et al (22). All other chemicals and solvents were of reagent grade.

Animals. Male Sprague-Dawley rats, 6 weeks old, were purchased from Japan Clea Laboratory and raised in Shionogi Laboratories until use. Rats of 8-12 weeks old were used for the experiments.

Incubation with Hepatocytes. Rat liver parenchymal cells were isolated by collagenase perfusion method according to Moldeus et al. (23). Hepatocyte viability averaged 92% (90-93%) as determined by the trypan blue exclusion test. The cells were suspended at 1 × 10⁷ cells/ml in Krebs Hensleit buffer, pH 7.4, supplemented with 0.2% bovine serum albumin, 10 mM glucose, 10 mM Hepes and benzylpenicillin (200 IU/ml). The incubation mixtures consisted of 2.5 ml of cell suspension (1 × 10⁷ cells/ml), 5 µl (10 nmol) of ¹⁴C-carnitine (2 µmol/ml of aqueous solution) or ¹⁴C-glycine (2 µmol/ml of 0.1 N HCl solution) and 100 µl (100 µg) of nonlabeled carboxylic acids (1 mg/ml of methanol solution) of the chemical structures shown in fig. 2. When ¹⁴C-PA was used as substrate, 100 µg of ¹⁴C-PA (1 mg/ml of aqueous solution) was added instead of nonlabeled substrate without ¹⁴C-labeled co-substrate. The incubation was carried out in a rotating round-bottom flask under an atmosphere of O₂:CO₂ (95:5) for 1 hr at 37°C. Under these conditions, cell viability did not fall below 88%. After incubation, 2 ml of mixture was taken, treated with sonication, mixed with 0.2 ml of 1 N HCl, and then centrifuged, and the resultant supernatant was used for analysis.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Chemical structures of carboxylic acids used as substrates.

Incubation with Kidney Slices. Rats were decapitated, and the kidneys were removed and stored in chilled Krebs-Ringer bicarbonate buffer (KRB). After the connective tissues and fat had been removed, the kidneys were cut into slices about 0.2 mm thick with a tissue slicer KN-822 (Natsume Co., Ltd., Tokyo, Japan). One gram of kidney slices was added to a 30-ml flask containing 4 ml of KRB which had been bubbled with an O₂:CO₂ (95:5) gas mixture for more than 15 min. Next, 10 nmol of ¹⁴C-carnitine or ¹⁴C-glycine and 100 µg of the nonlabeled carboxylic acids were added. When ¹⁴C-PA was used as substrate, 100 µg of ¹⁴C-PA was added instead of nonlabeled substrate without ¹⁴C-labeled co-substrate. The mixtures were finally bubbled again with the same gas mixture for 30 sec, then incubated for 1 hr at 37°C with constant shaking. After incubation, the mixtures were homogenized in a glass homogenizer. Portions of 3 ml of each homogenate were acidified by adding 0.3 ml of 1 N HCl and centrifuged, and then the resultant supernatant was used for analysis.

Analytical Procedures. The supernatant obtained from the incubation with the nonlabeled substrate was directly applied to DIAION HP-20 (Mitsubishi Kasei Co. Ltd., Tokyo, Japan) column (60 mm × 7 mm i.d.). The supernatant obtained from the incubation with ¹⁴C-PA was extracted with a solvent mixture of n-hexane/ethyl acetate (9:1), then applied to HP-20. The column was washed with 3 ml of 0.1 N HCl three times, then eluted with 6 ml of acetone. The eluate was dried in vacuo, dissolved into a small volume of 95% of tetrahydrofuran aqueous solution, and subjected to TLC using silica gel 60F254 precoated plates (0.25 mm thickness 20 × 20 cm; No. 5715; Merck, Darmstadt, Germany) with a developing solvent system of ethyl acetate/acetic acid/water (4:1:1) for the glycine conjugate, and ethyl acetate/acetic acid/water (2:1:1) for the carnitine conjugate. The TLC plate was brought into contact with X-ray film (Fuji Photo Film Co., Ltd., Tokyo, Japan) to obtain a radioautogram; then each metabolite fraction was scraped off and its radioactivity was counted with a Packard Tri-Carb 2000CA liquid scintillation spectrometer. In case of the incubating product from ¹⁴C-co-substrate, the radioactivity of each metabolite fraction was corrected by subtracting the radioactivity of the same area on the TLC plate after incubation without nonlabeled substrate.

	Results

Abstract Introduction Results Discussion References

Carnitine and Glycine Conjugation Activity in Hepatocytes and Kidney Slices. Carnitine or glycine conjugation activity in hepatocytes or kidney slices was expressed as the production ratio (%) of radioactive carnitine or glycine conjugate to the total radioactivity which was added as ¹⁴C-co-substrate into the incubation mixture as described in Methods (tables 1 and 2).

Calculation of the Dilution Ratio of ¹⁴C-Carnitine and ¹⁴C-Glycine with Endogenous Components. To calculate the amounts of carnitine and glycine conjugates produced, we needed to know the estimated amounts of endogenous carnitine or glycine that might contribute to each conjugation. We compared the amounts produced by incubating with ¹⁴C-PA and those from each ¹⁴C-co-substrate, and calculated the dilution ratio (table 3). The dilution ratio of ¹⁴C-carnitine was 38.2 in hepatocytes and 144 in kidney slices, leading to the estimation of 38.2 × 10 nmol/2.5 × 10⁷ cells of endogenous carnitine in hepatocytes and 144 × 10 nmol/g in kidney slices. In the same manner, the estimated amounts of endogenous glycine were calculated to be 104 × 10 nmol/2.5 × 10⁷ cells in hepatocytes and 627 × 10 nmol/g in kidney slices.

Comparison of the Amounts of Carnitine and Glycine Conjugates Produced. The amounts of carnitine conjugate produced were calculated by multiplying the estimated amounts of carnitine that might contribute to the conjugation reaction (table 3) by the production ratio of carnitine conjugate to the added radioactivity of ¹⁴C-carnitine into the incubation mixture (tables 1 and 2). The values for the glycine conjugate were calculated in the same manner, and the results are shown in figs. 3 for hepatocytes and fig. 4 for kidney slices. In the hepatocytes, although CPCA was the best substrate for carnitine conjugate, it was subjected to more conjugation with glycine (ca. 90 nmol) than with carnitine (ca. 70 nmol). CBCA was conjugated 6 times more with glycine (ca. 181 nmol) than with carnitine (ca. 33 nmol). BA (ca. 548 nmol) and CHCA (ca. 371 nmol) were preferentially conjugated with glycine. In the kidney, CPCA was conjugated 8 times more with carnitine (ca. 319 nmol) than with glycine (ca. 45 nmol). CBCA was also a good substrate for carnitine conjugate (ca. 143 nmol), but it still conjugated more with glycine (ca. 273 nmol). BA was the best substrate for glycine conjugate (ca. 968 nmol), and CHCA was also good (ca. 224 nmol); they did not conjugate with carnitine at all.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Carnitine and glycine conjugation activity in rat hepatocytes. The values were calculated by multiplying the estimated amounts of carnitine (382 nmol/2.5 × 107 cells) or glycine (1040 nmol/2.5 × 107 cells) by the production ratio of carnitine or glycine conjugate in rat hepatocytes (table 1).

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Carnitine and glycine conjugation activity in rat kidney slices. The values were calculated by multiplying the estimated amounts of carnitine (1440 nmol/g) or glycine (6270 nmol/g) by the production ratio of carnitine or glycine conjugate in rat kidney slices (table 2).

	Discussion

Abstract Introduction Results Discussion References

The biological importance of carnitine is well established in the field of fatty acid metabolism (7); however, its roles in the metabolism of xenobiotics are not well known except for a few examples such as PA (11-13), cycloprate (14-16), valproic acid (17), and benzoic acid (18). This is the first report concerning the substrate specificity of carnitine conjugation examined systematically in vitro using rat tissues. In both hepatocytes and kidney slices, the best substrate for carnitine conjugate was CPCA, followed by CBCA and CHCA. On the other hand, the best substrate for glycine conjugate was BA, with conjugation also occurring with CHCA and CBCA (tables 1 and 2). These results suggest that there is some kind of rule for the selection of the co-substrate that is transferred to the CoA thioester of carboxylic acid. Namely, carnitine transferase shows substrate specificity for carboxylic acids with lesser carbon number on the cyclic side chain, while glycine transferase shows inverse specificity. However, as there were many factors that could have affected the carnitine or glycine conjugation activity in this study using a cell system, such as cell permeability, CoA thioester synthetase activity, and glycine or carnitine transferase activity, we need to conduct more precise experiments to confirm the above hypothesis.

The presence of endogenous carnitine and glycine in the tissues affected the radio-specific activity of ¹⁴C-carnitine and ¹⁴C-glycine added to the incubation mixtures. Therefore, we could not compare the amounts of carnitine and glycine conjugates produced using the expression as the production ratio (%) of the radioactive conjugate to the radioactive co-substrate added to the incubation mixtures (tables 1 and 2). We tried to calculate the dilution ratio of carnitine or glycine by comparing the amounts of carnitine or glycine conjugate produced after incubation with ¹⁴C-PA with those after using each ¹⁴C-co-substrate. The calculated amounts of endogenous carnitine and glycine in hepatocytes were 382 nmol/2.5 × 10⁷ cells and 1040 nmol/2.5 × 10⁷cells, respectively, and those in kidney slices were 1440 nmol/g tissue and 6270 nmol/g tissue, respectively (table 3). These values were not very different from ones in the literature (24, 25). We then compared the activity of carnitine conjugate with that of glycine conjugate in hepatocytes or kidney slices by multiplying the estimated amounts of endogenous carnitine or glycine by the production ratio of the conjugate to the added radioactivity. To compare the conjugating activity of hepatocytes and kidney slices, we hypothesized that 2.5 × 10⁷ cells of hepatocytes have enzyme activity nearly equal to 1 g of liver slices and then compared the hepatocyte data with those for the kidney slices. Our hypothesis was confirmed by the result that similar conjugation activity was observed between hepatocytes (2.5 × 10⁷ cells) and liver slices (1 g) when ¹⁴C-PA was used as substrate (data not shown), and also by the report of similar metabolic activity of biphenyl obtained between hepatocytes (approximately 3 × 10⁷ cells) and 1 g of liver slices (26).

The enzyme activity per unit tissue weight tends to be higher in the kidney for both conjugates in comparison with hepatocytes (figs. 3 and 4). CPCA, which is the best substrate for carnitine conjugates, is also conjugated with glycine as well as in hepatocytes. However, in the kidney CPCA was preferentially conjugated with carnitine rather than glycine; its carnitine conjugating activity was about 5 times higher than that in hepatocytes. These results suggest that one of the main metabolites might be the carnitine conjugate of CPCA in rat in vivo. However, there is no report concerning the rat in vivo metabolism of CPCA, except for cycloprate (14), for which the major metabolite is the glycine conjugate of CPCA with the carnitine conjugate being a minor one. As this discrepancy may have originated from the chemical structure between cycloprate (hexadecyl CPCA) and CPCA, we are interested in the metabolism of CPCA in vivo and are now examining it. To our knowledge, this is the first report of the kidney having a high carnitine conjugating activity for xenobiotics. Cycloprate metabolism in dogs is also interesting from the viewpoint of species differences; the major metabolite is carnitine conjugate of CPCA which is excreted into urine and shows a tendency to accumulate in muscle for a long time (15).

BA was the best substrate for glycine in both hepatocytes and kidney slices, and CHCA was also a good substrate for glycine; however, these were negligibly conjugated with carnitine. It is well known that BA is metabolized mainly into the glycine conjugate in rat in vivo (2, 25); this is a good example of the in vitro metabolism coinciding with the in vivo one. On the other hand, as far as we know there has been no report concerning the in vivo metabolism of CHCA or describing it as a good substrate for the glycine conjugate. CBCA was a moderately good substrate for both carnitine and glycine conjugates in hepatocytes and kidney slices among those examined in this study. CPECA was the least effective for both conjugates in the cyclic side chain carboxylic acid series, perhaps because CPECA responds poorly to esterification with CoA, especially in hepatocytes.

The branched-side chain carboxylic acids, such as PA and IB, were also poor substrates for carnitine and glycine conjugates in rat hepatocytes and kidney slices. These findings coincide with those from the study by Diep et al. in which they suggest that the heart and the brown fat, but not the liver, play important roles in pivaloylcarnitine formation in rat (27). They also reported (28) the pivaloylcarnitine conjugating activty in isolated heart cells (5 nmol/2.5 × 10⁷ cells/hr); it was 5 times higher than that of liver (1 nmol/2.5 × 10⁷cells/hr) but lower than that of kidney (43 nmol/g/hr) when compared with our data. We therefore think that the contribution of the heart to carnitine conjugation of PA might not be larger than that of the kidney.

At present, we are very interested in the species difference of the substrate specificity of these model carboxylic acids and are examining the in vitro differences among dogs, rabbits, and monkeys.