CHIRALITY 4247-251 (1992)

Enzymes in Stereoselective Pharmacokinetics of Endogenous Substances A. MARZO, G. CARDACE, AND E. ARRIGONI MARTELLI Department of Drug Metabolism and PharmacokineEics, @ma Tau S.pA., Rome, Itdy

ABSTRACT The use of enzymes to assay individual components of the L-carnitine family in pharmaceuticals, foodstuffs, and biological fluids with various forms of detection is reviewed. The most useful enzyme in the assay of compounds of the L-carnitine family is carnitine acetyl transferase (CAT),which catalyses the reversible interconversion of L-carnitine and its short-chain acyl esters. CAT can be used in one or more coupled reactions combined with U.V., or radiolabelled detection, or combined with HPLC, allowing, enantioselective, structurally specific, and, in the case of radiolabelled tracing, highly sensitive assays to be carried out. When compared with chromatographic separation of enantiomers or diastereoisomers, enantioselective enzyme mediated assays may be cheaper, more sensitive, and simpler, but they do not allow the nonpreferred isomer to be assayed. Consequently,they are appropriate for the specific assay of endogenous enantiomeric substrates of the enzyme concerned, in biological samples. The analysis of the other enantiomer in raw materials or in pharmaceuticals must be more properly approached by enantioselective chromatographic methods.

KEY WORDS: radioenzyme assay, stereospecific assay, carnitine acetyl transferase, L - m i tine family, L-carnitine, acetyl-L-carnitine,propionyl-L-carnitine INTRODUCTION Most endogenous substances that possess one or more asymmetric carbon atoms are found in the bcdy in one enantiomeric form. They are in fact synthesized and often metabolized in the body by strictly enantioselective enzyme-catalysed reactions. As many enantiomeric substances are present in foods or are given as dietary additives or administered as single isomer drugs, their evaluation in biological fluids requires enantioselective methods. With endogenous substances, enzyme-involved methods, which can conveniently solve the problem of an enantioselective assay, are often available. Although much progress has been made in stereoselective chromatography and chromatographicmethods, which are predominantly used in pharmacokinetics studies, the strict stereospecificity of some enzymes is of great interest in developing enantioselective assays for endogenous compounds. In this respect, increasing interest is being devoted to enzyme methods coupled to HPLC. In fact, frequently, chromatographicor other kind of stereoselective methods possess a specificity and sensitivity adequate for analysis of raw materials or pharmaceuticals, but not biological samples. This is particularly true for endogenous molecules, where a high sensitivity is required in order to evaluate baseline levels and appreciate variations produced by exogenous administration. The advantage, but currently a limitation of enzyme-based enantioselective methods is that only the naturally occurring isomer reacts, whereas the other is not detected. However, the unreactive isomer generally has no interest for pharmacokinet~~

~

Present at the Second International Symposium on Chiral Discrimination, May 27-31, 1991, Rome, Italy. 0

1992 Wiley-Liss, Inc.

ics. unless the drug is administered as racemate or a racemization occurs in vivi, although the latter event has a very low probability of occurrence with endogenous compounds. In a few cases, as with lactate and tartrate, it is possible to achieve separate enantioselective enzyme assays, for either the L or the D enantiomer, as some enzymes extracted from particular microorganisms are able to interact with the enantiomer not present in mammals. Sometimesa molecule such as glucose may be a substrate of different enzymes and therefore several kinds of enzyme methods may be set up, generally involving a sequence of coupled reactions, one of which at least possesses an almost absolute enzyme-substrate specificity. In other situations, as with the L-carnitine (LC) family components, one specific enzyme reaction is the basis for several analytical procedures. In this study, an example is given of the use of enzymes to assay individual components of the LC family in pharmaceuticals, foodstuffs, and biological fluids with various detections. 293

REINSPECTION OF THE L-CARNITINE FAMILY ASSAY

LC is present in the body mainly in the free form and to a lesser extent as short-, medium-, and long-chain acylcarnitine esters. LC, its endogenous acyl esters, and likely impurities in bulk chemicals and pharmaceuticals are satisfactorily evaluated as such, or after chemical derivatization by HPLC, with reversed phase, -NH2or ion exchange columns.4-6 These Received for publication June 11, 1991;accepted January 16, 1992. Address reprint requests to A. hhm, Department of Drug Metabolism and Pharmacokinetics, Sigma Tau S.p.A., Via Pentina Km 30,400,00040Pomezia, Rome, Italy.

248

MARZO ET AL.

methods may be highly specific, but they are limited by poor sensitivity and lack of enantioselectivity. N.M.R. with chiral shift reagents may be used in the enantioselective assay of these compounds in pharmaceuticals,not in biological samples. The discovery and commercial availability of the enzyme carnitineacetyl transferase (CAT)has allowed the development of enantioselective and sensitive methods, some of which are useful in pharmacokinetic studies. Coupling the enzyme reaction with HPLC, U.V., or a radiolabelled isotope, an enantioselective assay is developed, as described in the following sections. As shown in Figure 1, CAT catalyses the reversible interconversion of LC and its short-chain acyl esters. Other transferases, namely, camitine octanoyl transferase (COT) and carnitine palmitoyl transferase (CPT), are involved in mediumand in long-chain L-carnitine acyl esters, respectively (Fig. 1). All of these transferases are strictly specific for the L-enantiomer of free camitine and acylcarnitine esters. CAT is particularly useful, as it catalyses in vivo a very quick interconversion between LC and its short-chain esters. Therefore, analytical procedures either for free, or short-chain, LC esters have been developed, based on the enantio-specific reaction catalysed by that enzyme. 697

CAT IN SPECTROPHOTOMETRIC AND FLUOROMETRIC ASSAY

+

the spectrophotometric or fluorometric change in NADH H concentration derived from the following coupled reactions: l1 +

ALC

CAT c* LC

+ CoASH

acetyl-S-CoA + oxalacetate

+ acetyl-S-CoA

cs -+

citrate

+ CoASH

(2

MDH c* oxalacetate + NADH+ H . These methods are sensitive only in the nanomole range and, in spite of their high specificity and enantioselectivity, some interferences,mainly from endogenoussulphydryl groups that react with DTNB, can affect the analyticalprocedure.Therefore they are not suitable for pharmacokinetic studies. malate

+ NAD

+

HPLC ASSAY USING CAT In reaction 1, LC can be assayed through the concentration of CoASH,whereas LC esters can be estimated from the concentration of the related acyl-S-CoA,separated on a reversed phase HPLC column, and detected at 254 nm.12p13This method possesses the enantioselectivityderived from the enzyme, the specificity derived from the chromatographic process, and has a sensitivity good for evaluation in urine, although not in plasma.

CAT IN RADIOENZYME ASSAY Purified CAT, used for the first time by Marquis and Fritz8 Using radiolabelledsubstrates,the above enzyme assays are in an enzyme assay developed in 1964, catalyses the following transformed into radioenzyme assays, allowing determination enantio-specificreaction: of the picomole amount of LC family components.These highly CAT sensitive, reproducible, and enantioselective methods are suitAL€ + CoASH. (1 able for measurement in all tissues and biological fluids, and, acetyl-S-CoA + LC Only the L-enantiomer is a substrate for CAT, the D-isomer therefore,they are particularly useful in pharmacokineticstudbeing inactive in this respect; indeed, at high concentrations, ies. Direct methods are available for LC and ALC. The problem the D-form inhibits the enzyme. CoASH, produced in a stoichio- of analysing short-chainLC esters (C2 - C,) is approached by metric amount, is reacted with 5,5’-dithiobis-2-nitrobemic a CAT-mediated radioexchange method. acid (DTNB), releasing the chromogenic thiophenolate ion, LCamitine which absorbs at 412 nm. Alternatively, &ASH is involved in According to reaction 1, in the presence of CAT, LC reacts coupled reactions in several spectrophotometricor fluorometric with [l 14C]-acetyl-S-CoA, producing labelled ALC, which is methods. hsed on the method of Marquis and Fritz, autoseparated from unreacted acetyl-SCoA and measured for ramated spectrophotometricassays have been developed. l4 The reverse reacdioactivity by liquid scintillation counting. In the case of acetyl-L-carnitine(ALC), the assay is based on tion is prevented by trapping the GASH released with N-ethyl maleimide (NEM) or sodium tetrathionate.The use of NEM has been demonstrated to be far superior to that of tetrathionate in assaying LC even in the presence of large amounts of ALC, l5 ACYLCARNITINES as occurs in samples from patients suffering from renal failure ACETYL-L-CARNITINE OCTANOYL-L-CARNITINE or in pharmacokinetics of intravenously administered ALC. A C c A A T h L-CARNITINE ~ A O T perchloric acid extraction allows a partition to be achieved ACYLCARNITINES between a “total acid-soluble”fraction, present in the supernatant, comprising free LC and its short-chainesters, and an “acid PROPIONYL-L-CARNITINE PALMITOYL-L-CARNITINE insoluble” fraction, present in the pellet and comprising longchain LC esters. Both fractionscan be measured as free LC after ACYLCARNlTlNES alkaline hydrolysis. Medium chain acyl-L-camitineesters have CAT CARNlTlNE ACETYLTRANSFERASE a critical solubility, so it is difficult to standardize their extracCOT - CARNlTlNE OCTANOYLTRANSFERASE tion. Several modifications of the LC radioenzyme assay have CPT CARNlTlNE PALMITOYLTRANSFERASE been published, mainly involving the extraction procedure of Fig. 1. Transferases involved in reversible interconversion of L-camitine samples and reagents used, mostly buffer, but all using the original analytical procedure. and L-carnitine esters. c-)

1 7 1 = =

249

ENZYMES IN STEREOSELECTIVE PHARMACOKINETICS

Acetyl-L- Carnitine

As described in reaction 2, acetyl-S-COA is first formed by ALC in the reaction catalysed by CAT. In the presence of an excess of - 14C]-oxalacetate,acetyl-S-CoA is then converted ted is then converted into into [ ’‘C]citrate. U ~ ~ ~ coxalacetate aspartate in a reaction catalysed by glutamic oxaloacetic transaminase (GOT), and in this form it is trapped by a cation exchangeresin, whereas the radioactivity of citrate is measured as an expression of ALC.16 Because of possible interferences from endogenous substrates such as acetyl-SCoA,oxalacetate, and citrate, which participate in the equilibrium of reactions involved in the assay, a purification step is required. High molar excess of LC does not invalidate the assay. Interferences from propionyl-S-&A, propionyl-L-camitine, and other shortchain esters of L-carnitine are possible, but only if present in very large amount compared with ALC.

Then, after separation by WLC or TLC, individual radioactive acyl-L-camitine esters are detected and quantified if the specific activity of L-carnitine in the pool is known. l7 h e to the fact that here the CAT catalysed reaction is used in the equilibrium form, many interferences arising either from samples or from impurity of reagents may affect the system. Therefore, although applied successfully to different biological samples, the method may fail in some situations such as with urine samples. As recently stated by Marm et al., l8 to monitor the chemical synthesis of PLC and its presence in pharmaceutical formulations, PLC, and more generally acylesters of L-camitine, can be chemically hydrolysed to LC, which is then evaluated using the enzyme assay, while HPLC is used to ascertain the complete hydrolysis of the acyl-L-camitineand the net crotonoylbetaine formation.

A NEW APPROACH TO ENZYME-MEDIATED L-CARNITJIVE ASSAY A new enzyme method for the assay of L-carnitine,based on Like other short-chainLC esters (other than ALC), propionylthe use of an enzyme other than CAT, has recently appeared in L-carnitine (PLC) does not have an individual radioenzyme the literature.19 In that method, L-camitine is oxidised by a assay, because no specific biochemical reactant is routinely camitine dehydrogenase extracted from Pseudomonas available. A general procedure is currently a CAT-mediated radioisotopic exchange using labelled L-carnitine with high aeruginosa according to the following reactions: specific activity. An aliquot of the “total acid soluble” fraction, carnitine + P-NAD Sdehydrocarnitine + B-NADH + H . containing a pool of free LC and short-chain acyl-L-carnitine esters, is incubated with labelled LC in the presence of GASH The enzyme is strictly specific for L-camitine as it does not and CAT. As the reaction catalysed by CAT is easily revers- catalyse reactions with D-carnitine,acylcarnitines,and choline. ible, an isotopic equilibrium is attained, accordingto the follow- Since the fluorescence of P-NADH is affected by many substances in biological samples, the p-NADH produced by the ing: above reaction is allowed to convert reSazurin to resorufin, acyl-L-carnitine acyl-S-CoA which fluoresces at 580 nm (h ex = 560 nm), in a subsequent acyl-Lamitine* + + + diaphorase-catalysed reaction. Both carnitine dehydrogenase COASH c* LC &ASH and diaphorase are immobilized on a small column connected + to a flow-injection system. The method is reported to be as LC* sensitive as the radioenzyme assay. It cannot be used for mea(added). surement of L-carnitine acylesters but only for free and total CAT in the Assay of Short-Chain Acyl Esters (C, - C,) of G Carnitine

-

+

-

TABLE 1. Amount excreted via urine (AE)in p o l e s and renal clearance (Cl,) in L h-I before (BAS) and after 2461 pmoles‘

BAS (0-24h)

ALC

AE clr

0-2h

2-4h

M F M F

44.51 f 26.05 27.89 f 21.40 0.32 f 0.21 0.43 f 0.34

465.00 532.50 3.81 4.54

f

M F M F

180.88 f 123.10 132.72 f 90.33 0.14 f 0.10 0.18 f 0.13

618.38 602.25 4.01 4.32

f 117.76 f 183.77

M F M F

332.34 247.24 0.23 0.30

105.12

f 159.31 f

f

0.82 0.93

72.88 71.90 2.03 2.87

4-8h

f 24.82 f 18.10 f

f

0.76 0.92

33.65 f 27.06 f 1.08 f 1.59 f

&12h 17.51 f 11.21 f 0.74 f 1.22 f

6.66 8.70 0.12 0.73

12-24h

16.79 5.25 0.34 0.65

44.85 23.36 0.74 0.77

f 27.51

f 13.67 f 0.44 f 0.37

~~

LC

AE clr

TC

AE clr

f 168.54 f 130.42 f f

0.12 0.17

f f

0.89 1.13

1108.63 f 218.08 1157.00 f 243.52 3.91 f 0.83 4.28 f 0.78

f 19.53 f 24.87

0.40

60.50 64.86 0.29 0.43

42.08 75.73 0.11 0.43

96.71 96.80 0.41 0.59

f 34.11

284.13 f 83.36 250.00 f 37.10 2.10 f 0.62 2.47 f 0.70

153.00 148.84 0.64 0.90

f 32.97 f 67.25 f 0.11 f

f 103.84 f 55.96

214.88 193.80 0.77 1.04

f f f f

376.38 338.88 2.14 2.60

f f

0.59 0.74

“ALC-HC1i.v. injection in healthy male 0 and female (F) volunteers. Mean value of 8 findings f SD.

f

0.12

f 0.15

38.34 f 0.16 f 0.22 f

172.21 f 63.72 110.68 f 75.20 0.29 f 0.13 0.26 f 0.18 302.13 f 107.3 178.14 f 99.77 0.45 f 0.18 0.38 f 0.21

250

MARZO ET AL.

L-carnitine,and, at present, it has been proposed for diagnosis entiated for LC and its esters; this mechanism allows each of L-camitine deficiency. component of the endogenous pool of L-carnitine family to be excreted with a renal clearance directly related to its plasma PHARMACOKINETICS OF L-CARNITINE FAMILY concentration. The above two mechanisms work synchroPharmacokinetics of LC, ALC,and PLC were investigated in nously and allow the homeostatic equilibrium of L-carnitine rats, dogs, and humans with the above radioenzyme methods, family to be restored even after a high i.v. dose of one compowhich allow for the best sensitivity.m*21 nent. In human volunteers (8 males and 8 females) treated with As recently noticed for most endogenous substances, a nonALC .HCl, baseline concentrations in males 0 and females(F) compartmental pharmacokinetic analysis proved to be more proved on average as follows: appropriate than the standard compartmental models in processing pharmacokinetic data of ALC and PLC.23s" TC 55 (M) 37 (F) LC 48 0 33 (F) PERSPECTIVES ALC 5 0 3Q. In line with the modern view that a chiral drug requires an After i.v. administration of ALC . HC1(2,461 pmoles) in bolus, enantioselectiveassay, enzymeshave to be regarded as an easy the plasma concentration of ALC increased immediately, and and useful way to solve some pharmacokinetic problems, then decreased, restoring baseline concentration in about 8 h. mainly those involved in endogenous substances and some of Also, LC increased and decreased according to a more sus- their derivatives. Endogenous substances found in foodstuffs tained behaviour. and, therapeutic treatments, as well as those involved in diagTable 1 shows urinary excretion (AE) and renal clearance nostic procedures, constitute a very large group. From a phar(CL,) of ALC, TC, and LC before and after i.v. injection. AE of macokinetic point of view, they usually provide several probALC, LC, and TC increased showing the highest value in the lems, mainly concerning baseline endogenous levels, which 0-2-h interval and then decreased to restore baseline situation must be taken into account in order to define the pharmacokiafter 24 h. Cl,, evaluated as the AE/AUC ratio, increased then netic profile and thereforeclearly distinguish net variations due decreased,paralleling AE behaviour. These data are consistent to exogenous administration. Uneven systemic distribution, with a previous pilot investigation carried out on 3 male and 3 controlled absorption rate, and a reversible metabolic pool usufemale healthy subjects. ally result in nonlinear plasma concentration-timebehaviour of After oral administration the same but less intensive the endogenous substance administered, when these submodificationsin levels were observed, exhibitinga statistically stances are cleared via urine, their cumulative urinary excresignificantdose proportionality,which was more evident in the tion often is a useful mirror of bioavailability, as occurs with AE than in plasma concentration and AUC. potassium.%In the case of the L-carnitinefamily, thanks to the Studiesof ALC*HClin rats have shown results close to those CAT enzyme, it was possible to carry out pharmacokinetic encountered in humans, whereas the dog proved to behave studies focusing on the main components of the endogenous differently, as the plasma concentration of ALC and LC after pool of L-carnitineinvolved in the interconversionmechanism. i.v. injection showed a second delayed peak, and the net cumuAs observed with the L-carnitine family, enzymes can be lative urinary excretion of TC averaged close to 50% vs values used in one or more coupled reactions, combined with spectroof 74% in humans and 100% in rats. photometric, HPLC, or radiochemical evaluation, and thus they When PLC*HCl(1.18 and 2.36 mmoles) was injected i.v. in can be utilized in different ways, allowing, in all cases, enantibolus in healthy volunteers (3 males and 3 females),a quick oselective, specific, and, in certain cases, highly sensitive asincreae of PLC and a lower but more sustained increase of LC says to be carried out. were observed, the former being restored in 6 h, and the latter When compared to chromatographicseparation of enantiomsomewhat later. ALC did not show any modification in plasma ers or diasteroisomers, enzyme mediated assays may be concentration,whereas its AE and CL, increased.21 The behav- cheaper, more sensitive, and simpler, but they only allow the iour after oral administration, and the species differences en- quantification of the naturally occurring isomer. Consequently countered with PLC, parallel the situation described above with they are appropriate for the assay of the endogenousenantiomALC. In all the species investigated, TC, ALC,PLC, and LC eric materials in biological samples. The problem of the presshowed higher plasma concentration and urinary excretion in ence of the other enantiomer in raw materials or in pharmales than in females. maceuticals is better addressed by the use of enantioselective According to previously published reports, the LC family chromatographic or spectroscopic methods. components are unevenly distributed in the body, 98% being REFERENCFS present in skeletal and cardiac muscle, 0.6% in plasma and extracellular fluids, and 1.4% in the remaining tissues.22 1. Campbell, D.B., Stereoselectivity in clinical pharmacokinetics and drug development. Eur. J. Drug. Met. Pharmacokin. 15:1C%125, 1990. Results obtained led to the following conclusions. When a component of the endogenous pool of L-carnitine is exoge- 2. Noll, F., L-(+ )-Lactate. In: Methods of Enzymatic Analysis, Val. 6. Bergmeyer, H.U., ed. Weinheim: Verlag Chemie GmBH, 1984582-588. nously administered, the body reacts in order to preserve or to restore its own homeostatic equilibrium by two mechanisms. 3. Gawehn, K., D-(-)-Lactate. In: Methods of Enzymatic Analysis, Val. 6. Bergmeyer, H.U., ed. Weinheim: Verlag Chemie GmBH, 19&1:5&3%592. The first is the CAT catalysed interconversion of LC into its 4. Minkler, P.E., Ingalls, S.T., Kormos, L.S., Weir, D.E. Hoppel, C.L. Detenninaesters (Fig. 1).This was carefully observed in various species tion of camitine, butyrobetaine and betaine as I'-bromo phenacyl ester after administeringALC and PLC. The second mechanism is a derivatives by high-perfonnance liquid chromatography. J. Chromatopr. 336271-283,1984. saturable tubular reabsorption process with a threshold differ-

ENZYMES IN STEREOSELECTIVE PHARMACOKIh'ETICS 5. Minkler, P.E., Ingalls, S.T., Hoppel, C.L. Determination of total camitine in human urine by high-performance liquid chromatography.J. Chromatogr. 420:&393, 1987. 6. Marm, A., Cardace, G., Monti, N., Muck, S., Arrigoni Martelli, E. Review: Chromatographic and non-chromatographic assay of L-camitine family components. J. Chromatogr. 527 247-258,1990. 7. Voeffray, R., Perlberger, J.C., Tenud, L. Lcamitine: Novel synthesis and determination of the optical purity. Helv. Chim. Ada 70:205&%2%4,1987. 8. Marquis, N.R., Fritz, I.B. Enzymological determination of free camitine concentrations in rat tissues. J. Lipid Res. 1M184-187, 1964. 9. Seccombe, D.W., Dodek. P., Frohlich, J., Hahn,P., Skela, J.P., Campbell, D.J. Automated method for L-camitinedetermination. Clin. Chem. 22 158%1592, 1976. 10. Cederblad, G., Harper, P., and Lindgren, K. Spctrophotometry of camitine in biological fluids and tissue with a Cobas Bio centrifugal analyzer. Clin. Chem., 3234-346,1986. 11. Pearson, D.J., Chase, J.F.A., Tubbs, P.K. The assay of (- )-mitine and its 0-acyl derivatives. Methods Enzymol. 14612-622, 1969. 12. Takeyama, N.. Takagi, D., Adachi, K., Tanaka, T. Measurement of free and esterified camitine in tissue-extractsby high performance liquid-chromatography. Anal. Biochem. 158:346-354, 1986. 13. Dugan, R.E., Schmidt, M.J., Hoganson, G.E., Steele,J., Gilles, B.A., Shug, A.L. High perfomance liquid chromatography of coenzyme A esters formed by transesterificationof short-chain acylcarnitines: diagnosis of acidemias by urinary analysis. Anal. Biochem. 160:27S280, 1987. 14. Cederblad, C., Lindstedt, S. A method for the determination of camitine in the picomole range. Clin. Chim. Acta 37235243, 1972. 15. Parvin, R., Pande, S.V. Microdetermination of (- ) - m i t i n e and camitine acetyl transferase activity. Anal. Biochem. 79 190-201, 1977.

251

16. Pande, S.V., Caramancion, M.N. A simple radioisotopic assay of acetylcamitine and acetyl-CoA at picomolar levels. Anal. Biochem. 11230-38,1981. 17. Kemer,J., Bieber, L.L. A radioisotopic-exchange method for quantitation of short-chain (acid-soluble)acylcamitines. Anal. Biochem. 134:45M,l983. 18. Marm, A., Monti, N., Ripamonti, M, Arrigoni Martelli, E. Application of high-performance liquid chromatography to the analysis of propionyl-Lcamitine by a stereospecific enzyme assay. J. Chromatogr. 459:313-317, 1988. 19. Matsumoto, K., Yamada, Y., Takahashi, M., Todoroki, T., Mizoguchi, K., Misaki, H., Yuki, H. Fluoromehic determination of camitine in serum with immobilized camitine dehydrogenaseand diaphorase. Clin. Chem. 36 20722076,1990. 20. Marm, A,, Arrigoni Martelli, E., Urso, R., Rmhetti, M., Rizza, V., Kelly,J.G. Metabolism and disposition of intravenously administered acetyl-Lamitine in healthy volunteers. Eur. J. Clin. Pharmacol. 37 59-63, 1989. 21. Marm, A., Cardace, G., Corbelletta, C., Bassani, E., Solbiati, M., Foroni, B., Mancinelli, A., Arrigoni Martelli, E. Stereoselective pharmacokinetics of acetyl- and propionyl-L-camitine in humans dogs and rats after an i.v. bolus. Chirality, 1991 (in press). 22. Carnitine deficiency. Editorial. Lancet 335 631433,1990. 23. Marm, A,, Rescigno, A., Arrigoni Martelli, E. Role of the homeostatic equilibrium of endogenous substances in their pharmacokinetic behaviour. Quat r i h e s J o u r n h Mediterranhnes de Pharmacocinetique. Le C o r n Montpellier 28-30 Mars 1991, Abstract C-19. 24. Rescigno, A,, Marm, A. Area under the curve, bioavailability and clearance. J. Pharmacokin. Biopharm., 1947M80, 1991. 25. Melikian, A.P., Cheng, L.K., Wright, GJ., Cohen, A,, Bruce, R.B. Bioavailability of potassium from three dosage forms: Suspension, capsule and solution. J. Clin. Pharmacol. 28:1C4&1050, 1988.