Journal of Hepatology, 1992; 14: 335-341 @ 1992 Elsevier Science Publishers B.V. All rights reserved. 0168-8278/92/$05.00

335

HEPAT 00990

Axe1 Holstege, Heide-

aria Gengenbacher, Linda ehle and Wolfgang

Get-ok

Department of Internal Medicine, Universityof Freiburg, Freiburg, Federal Republic of Germany

(Received 22 March1991)

A new approach in the treatment of gastrointestinal tumors with 5-tluorouracil involves the infusion of high doses of uridine to improve the chemotherapeutic efficiency of the former. High amounts of uracil formed from uridine can interfere with the hepatic catabolism of 5fluorouracil and thus increase its bioavailability and toxicity. In our study, we analysed the metabolite pattern of uridine in the effluent of isolated perfused rat livers in relation to portal uridine levels. The livers were pelfused hemoglobin-free without recirculation at a constant flow. In the perfusate, uridine was changed from 0.5 to 100 PmoYl. The Icomplete degradation of [2-14C]uridine and [2-‘4C]uracil was monitored via the r&lease of labeled CO*. Radioactive catabolites of uridine including uracil and the sum of dihydrouracil andp-ureidopropionate were separated by high-performance liquid chromatography and counted using a radioactivity flow monitor. Portal uridine concentrations ‘were increased from 0.5 to 1OOymoYl and were accompanied by a rise in the relative amount of non-metabolized uridine in the effluent from 13 to 78%. At uridine concentrations above 50fimol/l, there was a constant release of uracil into the effluent, indicating saturation of uridine phosphorolysis or transport. The amount of 14C02 formed by the liver reflecting complete uridine breakdown was higher than any other uridine metabolite when uridine concentration varied from 0.5 to 15 pmol/l. Saturation of 14C02 formation was achieved at a uridine concentration of 25 pmol/l. Higher peak values of 14C02 release were observed after direct infusion of equivalent amounts of uracil into the portal vein. Extraction of uridine was most efficient at lower nucleoside concentrations indicating dose dependency of clearance by the liver. The dose of uridine in treatment regimens including 5fluorouracil should surpass the capacity of uracil catabolism by the liver counteracting 5fluorouracil degradation and toxicity.

Uridine is a physiological constituent of biood plasma (l-6)) which is used for pyrimidine nucleotide synthesis via the salvage pathway in many tissues (1,7). The pyrimidine nucleotide requirements of cells can be satisfied by the circulating uridine in blood plasma (7). The accumulating intracellular uridine Y&phosphate (UTP) which is formed leads to feedback inhibition of pyrimidine de novo synthesis (7) and antagonizes the effects of pyrimidine antimetabolites which should interfere with this pathway in cancer chemotherapy (1). Recently, the infusion of high doses of &dine have been shown to reverse the DNAand RNA-directed actions of 5-fluoropyrimidines (reviewed in Ref. 8). Uridine rescue resulted in a considerable improvement of antitumor activity and allowed

higher doses of Wluorouracil to be administered (9-11). In order to have this beneficial effect, uridine must first be converted into nucleotides to counteract the cytotoxic effects of both 5-fluoro-2’-deoxyuridine and 5-fluorouridine triphosphates (8). Finally, uridine and uracil can also interfere with the catabolism of 5-fluorouracil(12,13) which can lead to an enhanced bioavailability of the drug (14). In addition to their metabolic effects, uridine and uracil have been shown to possess different pharmacological properties. The administration of uridine favours the restoration of cardiac glycogen (15) and may increase blood pressure in anaesthetized rats (16) presumably by binding itself to a distinct class of pyridiminoreceptors (17) before or after its conversion to uridine S-monophosphate

Correspondence: Dr. Axe1 Holstege, Medizinische Winik I der Universitlt Regenstmrg. Franz-Josef-Strauss Allee, W-8400 Regensburg,

F.R.G.

A. HOLSTEGE

336 (UMP). Jn addition, high doses of uridine interfere with the regulation of body temperature in different species (18-20).One of its breakdown products appears to induce fever in man and rabbits (20). Continuous administration of a diet supplemented with uracil was followed by uracil-induced urolithiasis and the development of a reversible papillomatosis in the urinary bladder of F344 rats (21). Inborn dehydropyrimidine dehydrogenase deficiency led to thymine uraciluria and a non-specific clinical picture of cerebral dysfunction in man (22). These pathological conditions strongly point to the need four ongoing unimpaired pyrimidine degradation. Thi; liver has been shown to be the key site for complete degradation of uridine, uracil and their analogs {14,23-28). Earlier studies showed that the isolated perfused rat liver eliminates uridine in a single pass and releases a constant amount of the nucleoside from hepatic pools of uridine nucleotides (27,28). In our study, we used a non-recirculating perfusion system to further clarify the role of the liver in portal vein &dine degradation end we tested a wide range of natural and non-physiological nucleoside concentrations. High uridine levels can occur under in vivo conditions during combinations of high doses of uridine with Nluorouracil in cancer chemotherapy (18). Furthermore, we analyzed the metabolite pattern of uridine in the liver vein effluent; this indicated efficient extraction with most of the nucleoside completely catabolized at low physiological uridine concentrations. We present evidence for a dose-dependent clearance of uridine.

et al.

Hemoglobin-free liver perfusion

Isolation and perfusion of the liver were performed exactly as previously described (29,30). Rats were anesthetized with sodium pentobarbital (3 mg/lOO g body weight; Ceva, Bad Segeberg, F.R.G.) injected intraperitoneally. Heparin (100 U; Hoffmann-La Roche, Grenzach-Wyhlen, F.R.G) was administered into the femoral vein. The isolated livers were perfused at constant flow in a non-recirculating system using bicarbonate-buffered Krebs-Henseleit saline plus L-lactate (2.1 mmol/l) and pyruvate (0.3 mmoyl) (29,30). The buffer was equilibrated with O&O2 (95:5, v/v) at 37 “C to give a final pH of 7.4. Perfusate flow was 4 mlmin-‘.g liver-“. The perfusate entered the liver via the cannulated portal vein (influent). A cannula fixed in the superior caval vein served to drain the liver (effluent). The hepatic artery was tied off. Portal pressure, oxygen consumption and pH were continuously monitored by using, a pressure transducer, (Sachs Elektronik, Hugstetten, F.R.G.), a Clarke-type electrode (Bachofer, Reutlingen, F.R.G), and a pH electrode (Ingold, Frankfurt, F.R.G.), respectively. Pyrimidines were infused into the influent perfusate using a precision micropump (Braun Melsungen, Melsungen, F.R.G.). Pyrimidine concentrations in influent perfusate were controlled by measurement of their ultraviolet absorbance and using published absorption coefficients (31). After an equilibration period of at least 20 min effluent samples were collected every minute. Half of the samples were used for CO, measurement and the other half for metabolite analysis. Measurement of radioactive CO2

Materials and Methods Animals, chemicals and isotopes

Male Wistar rats (Tierzuchtanstalt Jautz, Hannover, F.R.G.) weighing 150 to 200 g had free access to water and a standard rat diet (Altromin 300 R, Altromin, Lage, F.R.G.). Uridine and its catabolites (uracil, dihydrouracil, /3-ureidopropionate) were from Sigma Chemicals (Deisenhofen, F.R.G.) or from Merck (Darmstadt, F.R.G.). ZPhenylethylamine was obtained from Serva Biochimica (Heidelberg, F.R.G.). Pyruvate and L(+)-lactate were bought from Boehringer Mannheim (Mannheim, F.R.G.) and from Roth (Karlsruhe, F.R.G.), respectively. All other chemicals were from E. Merck of the highest quality available. [2-‘4C]Uridine (52 Ci/mol) and [2-14C]uracil (48 Ciimol) were from Amersham Buchler (Braunschweig, F.R.G.). Purity was more than 98% for each of the isotopes as determined by reversed-phase high-performance liquid chromatography.

Effluent samples were collected in glass flasks and occluded with a rubber stopper. After addition of 3 ml perchloric acid (3 mol/l) radioactive CO2 was trapped in 1 ml 2-phenylethylamine over 16 to 24 h and counted at an efficiency of 85% in Instant Stint Gel (Packard Instrument, Downers Grove, IL). Metabolite analysis

For high-performance liquid chromatography 10 ml of each sample was lyophilized and resuspended in 1 ml of water. After filtration (0.22 pm Millipore filter units; Eschbom, F.R.G.) 30~1 HClO, (2.5 mol/l) were added to 300 ,ul of the filtrate. The supematant was neutralized with solid KHC03, centrifuged and filtrated using Millex HV 0.45 pm filter units (Millipore). Fifty ~1 of the clear sample wac injected onto two Cts-PBondapak columns connected in series with a Guard-PAK precolumn module containing a C,s-PBondapak insert (Waters, Eschbom, F.R.G.). HPLC and counting of the radioactive peaks was performed as described (30) using an isocratic Waters

URIDINE

CATABOLISM

337

HPLC equipment and a radioactivity flow monitor connected in series. The PLC eluent was continuously mixed with liquid scintillator (Rialuma; Baker Chemicals, Deventer, The Netherlands) at a flow rate of 3 ml/min using a 1 ml mixing chamber (Model LB-1 thold, Wildbad, F.R.G.). The sum of dihydrouracil an\dpureidopropionate is given since minor conversions of dihydrouracil into @reidopropionate cannot be excluded. Alkaline treatment of samples at 4 “C, however, did not reduce dihydrouracil contents indicating a higher stability of this compound as compared to its fluorinated analog (14,33). Uridine clearance (C&; ml/min) was calculated from the portal flow (V, ml/min) as well as from the influent (Ci) and effluent (c,) uridine concentration (nmol/ml) according to the following formula: Clu,a = V.Ci-C,JCi.

Changes ofuridine catabolite pattern with increasing portal vein nucleoside concentrations Five different radioactive metabolites were detected in

URIDINE Fig

1 &lease

wi& &rent

of ufiane

CONCENTRATION

IN INFLUENT

urac~ CC& and the sum

uridine Cc&n&ons

(~mol/l)

uent perfusate of isolated rat livers perfused with increasing concentrations of “4C-labeled uridine including uridine, uracil, dihydrouracil, @reidopropionate and 14C02 (Fig. 1). Total radioactivity in the liver vein effluent was 98.5 + 3.6% (mean f SD.) of the label icfused into the portal vein excluding any significant intracellular trapping of uridine in the form of uracil nucleotides. When the livers were perfused with low concentrations of uridine (0.5 to IO~umoVl), radioactive CO, was the predominant catabolite. Ilt peaked at 55 + 4% of total radioactivity in the hepatic outflow and reflected complete degradation of the nucleosides. Under these conditions, uracil and the sum of dibydrouracil and j&ureidopropionate amounted to 18.5 and 5.3%, respectively (Fig. 2). .4t an infItlent uridine concentration of 3 ymoY1, which is close tc the amount detected in blood plasma under physiological conditions (2-6), 79% of the nucleoside was converted into catabolites by the liver in a single pass. At higher portal m-idine concentrations (215 PmoVI), unchanged uridine became the predominant metabolite in liver vein effluent (Fig. 1): the relative proportion of this nucleoside increased from 31 to 78% at influent uridine

IJRlDlNE CONCENTRATION

IN INFLUENT

(pmol/l)

of dihydrouracil and @-ureidopropionate (catabolites) into the effluent of rat livers perfused

@.5 to 100 pmol/l). Isolation and perfusion of rat livers was performed as described under

Materials and M&o&. (2-‘4C]ufidine(15 Ca) and the indicated amount of nonlabeled uridine were included into the influent perfusate. Uridine and its cata-

bolites were quantitated by HPLC and a radioactivity flow monitor. 14C02was trapped in 2-phenylethylamine and counted. Symbols represent the mean +- SD. (vertical bars) least three separate perfusions. The right half of the figure indicates an enhanced view low uridine concentration range.

of at

of the

A. HOLSTEGE et al.

338 I

.

1 ’

’

’

60

I ‘0

Urd

c

URIDINE

“I.

“1.

20

40

CONCENTRATION

60

60

IN INFLUENT

1 T-

‘I 100 (pmol/l)

Fig. 2. Relative amount of uridine and its catabolites in the liver vein effhtent of rat livers perfused with different uridine concentrations. Values were calculated from the data in Fig. 1.

concentrations of 15 and lOO~moV1,respectively (Fig. 2). This concentration-dependent escape of uridine from hepatocellular degradation was inversely related to a fall in the relative amount of labeled CO*, whereas the percentage of uracil and of the sum of dihydrouracil and /3ureidopropionate remained almost constant. Steady-state concentrations of uracil and CO2 in hepatic vein effluent were reached when the amount of portal uridine was increased above 50 and 20pmoUl, respectivelyDose-dependent cleurance of uridine The clearance of different amounts of uridine by the

isolated perfused rat liver was calculated from the nucleoside concentrations in influent (CJ and effluent (c,) and the perfusate flow. Since perfusate flow was kept constant under our experimental conditions, uridine clearance was directly related to the difference between ci and c, divided by ci. With rising uridine concentrations (ci = 0.5 to 20 pmol/l) there was a rapid decline in the clearance of the nucleoside (Fig. ‘3). A further enhancement of portal uridine levels beyond 20 pmolil was associated with a less pronounced fall in uridine clearance. Accordingly, extraction of portal vein uridine was most efficient at lower extracellular nucleoside concentrations; clearance was

I’

1

c

’

3

‘0

20

URIDINE

CONCENTRATION

40

‘1

60

“1

80

IN INFLUENT

1oc

(~mol/l)

Fig. 3. Relationship between uridine clearance and increasing influent uracil nucleoside concentrations in the isolated perfused rat liver. Uridine clearance was calculated from the nucleoside concentration in influent and effluent and the perfusate flow. The mean f SD. (vertical bars) of at least three determinations is given.

maximal at the lowest uridine dose tested (0.5 PmoVl) with 24 ml/min representing 86% of total perfusate flow. Uridine clearance dropped by 50% when portal uridine was enhanced to 50pmol/l Catabolismof uracil

Similar to the degradation of uridine, t4COzformation from [2-t4C]uracil increased when higher concentrations of the pyrimidine base were infused into the portal vein (Fig. 4). Plateau values of CO, formation were reached when uracil in the perfusate was enhanced to more than 40 pmol/l. Under these conditions, the rate of uracil degradation peaked at a value of 50 nmol.min-l-g liver-‘. The maximal rate of CO1 formation after uridine infusion only amounted to 62% of the rate of portal uracil catabolism. Direct administration of uracil (50 pmol/l or 240 nmoLmin_t-g liver-‘) was followed by.a release of CO, at a rate of 50 nmol-min-t,g liver-’ (Fig. 4). The apparent rate of uracil formation from uridine (50 pmol/l) was calculated from the rate of uracil, dihydrouracil, /3ureidopropionate, and CO2 formation, the sum of which amounted to 73.2 nmol-min-l-g liver-’ (~18

URIDINE

CATABOLISM

URACIL

CONCENTRATION

339

IN INFLUENT

(,umol/l)

Fig. 4. Release of radioactive CO, from isolated rat livers after perfusion with [2-‘4C]uracil (15 @i/l; 1 to 2OOpmoUl). Each symbol represents an individual perfusion experiment. The mean of five CO, measurements f. SD. (vertical bars) is given.

pmol/l). If this amount of uracil was included into the perfusate, a similar rate of CO, release would have been measured (30 nmol~min-‘g liver-’ vs. 27.5 nmol.min-‘g liver-’ after 50 pmol/l uridine). Uridine transport and/or phosphorolysis thus appear to limit maximal CO2 formation from sinusoidal uridine.

iseussion The complete recovery of label and the pattern of radioactive uridine metabolites in the effluent of isolated perfused rat livers indicated that in the absence of any significant uridine anabolism, portal vein uridine leaves the sinusoidal space either unchanged or degraded (Fig. 1). Minimal hepatic anabolism has been confirmed by studies in the whole animal (4,14,23,25), m the perfused rat liver (27,28), or in isolated liver cells (34). In contrast to earlier observations (27), the highest percentage of uridine which was completely converted into the endproducts of the catabolic pyrimidine pathway in our study only amounted to 55% at the influent concentration of 3/~mol/l (Fig. 2). However, if incomplete breakdown products of uridine other than labeled COz were also considered, py-

rimidine nucleoside extraction proved to be very efficient with up to 87% of the portal uridine being catabolixed. The hepatic clearance of uridine was maximal at the lowest concentration tested and rapidly declined with increasing portal nucleoside levels (Fig. 3). This pointed to a strong dose-dependent elimination of the nucleoside from portal vein blood. When radioactive uracil (Fig. 4) or its deoxytmcleoside analog, 5-fluoro-2’-deoxyuridine (35) were infused into the perfusate, a similar inverse relationship between dose and clearance was apparent. A prerequisite for the occurrence of uridine catabolism is its transport across the cell membrane. Cell entry of nucleosides can be mediated either by symmetrical, nonconcentrative facilitated diffusion or a sodium-dependent concentrative nucleoside transporter (36). Recently, we presented evidence for the existence of a sodium-dependent transport system in rat liver (37) which was saturated at a portal uridine concentration of 20 ,umol/l. The initial rapid decline in uridine clearance (Fig. 3) caused by only minor increases in influent uridine concentrations can be best explained by saturation of this high-affinity low capacity transport system (37,38). At these low substrate levels, uridine phosphorylase rapidly converted most of the sinusoidal uridine into uracil (Fig. 1) and apparently was not rate limiting. In mouse spleen cells the Michaelis constant (40 pmol/l) and maximum velocity of the Na+dependent nucleoside transporter was only l/5 of those of the facilitated nucleoside transporter (38). From similar first-order rate constants for both transport systems, it was concluded that at low uridine concentrations (much lower than the K,,, value), uridine entry into the cells occurred about equally via both routes. However, at higher uridine concentrations, facilitated uridine transport was the primary route of entry in rat liver (37) and in mouse spleen cells (38). At low exogenous nucleoside concentrations, first-order rate constants for the subsequent in situ phosphorylation of uridine in spleen cells was similar to those of the nucleoside transporters (38). This indicates a limiting role of transport for uridine metabolism under these conditions. In the liver, uridine anabolism was almost completely absent and uridine phosphorolysis was the predominant reaction (Fig. 1) (34). Moreover, the specific enzymatic activity of uridine phosphorylase was higher in the liver than in the spleen (40) strongly arguing for a limiting role of transport and not for successive uridine phosphorolysis at low circulating uridine concentrations. Maximal sinusoidal uracil concentrations were obtained when portal uridine levels were higher than 50 pmol/l (Fig. 1). The plateau values of uracil presumably reflect saturation of dihydrouracil dehydrogenase which converts uracil into dihydrouracil and is located exclu-

A. HOLSTEGE et al.

340

sivelv in the .oarenchymal cells of the liver (34). Additional non-competitive inhibition of dihydrouracil dehydrogenase by uridine (12) may contribute to a higher production rate of 14C02 during uracil administration than with equimolar doses of uridine. In isolated hepatocytes incubated in the presence of 5fluorouracil the fluorinated analog of dihydrouracil accumulated intracellularly indicating dihydropyrimidinase as the rate-limiting enzyme of 5-fluorouracil catabolism (41). In our perfusion experiments, the sum of dihydrouracil and p-ureidopropionate was low. There were gradual increases with higher extracellular uridine concentrations (Fig. 1). The relative amount never exceeded 10% of the total radioactivity in the hepatic vein effluent (Fig. 2). The possibility of an intracellular accumulation of dihydrouracil with limited transport of this catabolite across the cell membrane as has been observed for dihydrofluorouracil cannot be excluded by our experiments. Potentiation of 5-fluorouracil cytotoxicity by uridine can either be a result of interference with 5-fluorouracil catabolism or increased anabolism (8). Uridine-induced

increases in UTP lead to feedback inhibition of de novo pyrimidine synthesis (1) and increased conversion of 5fluorouracil into 5-fluoroUMP by orotate phosphoribosyl transferase. Uridine rescue can result from competition of UTP and dUMP with the respective cytotoxic analogs of 5-fluorouracil(8-10). The results of our study indicate a high extraction of circulating uridine at low, physiological concentrations and a lower clearance at higher, exogenously administered uridine levels. Uridine doses leading to the latter blood plasma concentration range of the nucleoside (220 eel/l) should saturate the catibolic capacity of the liver providing enough unchanged uridine to rescue nonneoplastic tissues from 5-fluorouracil toxicity.

References

12 Tuchman M, Ramnaraine MLR, O’Der RF. Effects of uridine and thymidine on the degradation of S-fluorouracil, uracil and thymine by rat liver dihydropyrimidine dehydrogenase. Cancer Res 1985; 45: 5553-6. 13 Sommadossi J-P, Gewirtz DA, Cross DS, Goldman ID, Cano J-P, Diasio RB. Modulation of 5-fluorouracil catabolism in isolated rat hepatocytes with enhancement of S-fluorouracil glucuronide formation. Cancer Res 1985; 45: 116-21. 14 Holstege A, Pausch J, Gerok W. Effect of S-diazouracil on the catabolism of circulating pyrimidines in rat liver and kidneys. Cancer Res 1986; 46: 5576-81. 15 Aussedat J, Verdetti J, Grabiy S, Rossi A. Nucleotides uridyliques et glycogene cardiaques: effet de l’administration d’uridine et de ribose chez le rat. J Physioll982; 78: 331-6. 16 Macdonald G, Assef R, Guiffre A, Lo E. Vasoconstrictor effects of uridine and its nucleotides and their inhibition by adenosine. Clin Exp Pharmacol Physioll984; 11: 381-4. 17 Seifert R, Schultz G. Involvement of pyrimidinoceptors in the regulation of cell functions by uridine and by uracil nucleotides. Trends Physiol Sci 1989; 10: 365-7. 18 Leyva A, van Groeningen Cl, Kraal I, et al. Phase I and pharmacokinetic studies of high-dose uridine intended for rescue from Sfluorouracil toxicity. Cancer Res 1984; 44: 4928-5933. 19 Cradock JC, Vishnuvajjala BR, Chin TF, Hochstein HD, Ackerman SK. Uridine-induced hyperthermia in the rabbit. J Pharm Pharmacol1986; 38: 226-9. 20 Peters GJ, van Groeningen CJ, Laurensse E, et al. Effect of pyrimidine nucleosides on body temperatures of man and rabbit in relation to pharmacokinetic data. Pharma Res 1987; 4: 113-9. 21 Shirai T, Ikawa E, Fujushima S, Masui T, Ito N. Uracil-induced urolithiasis and the development of reversible papillomatosis in the urinary bladder of F344 rats. Cancer Res 1986; 46: 2062-7. 22 Berger R, Stoiker-de Vries SA, Wadman SK, et al. Diiydropyrimidine dehydrogenase deficiency leading to thymine-uraciluria. An inborn error of pyrimidine metabolism. Clii Chim Acta 1984; 141: 227-34.

1 Keppler D, Holstege A. Pyrimidine nucleotide catabolism and its compartmentation. In: Sies H, ed. Metabolic Compartmentation. London: Academic Press, 1982; 147-203. 2 Karle JM. Anderson LW, Dietrick DD, Cysyk RL. Determination of serum and plasma uridine levels in mice, rats, and humans by high-pressure liquid chroms?kvagraphy.An31 Biochem1980; 109: 41-6. 3 Karle JM, Anderson LW, Dietrick DD, Cysyk RL. Effect of inhibitors of the de novo pyrimidine biosynthetic pathway on serum uridine levels in mice. Cancer Res 1981; 41: 4952-S. 4 Holstege A, Manglitz D, Gerok W. Depletion of blood plasma cytidine due to increased hepatocellular salvage in D-galactosamine-treated rats. Eur J Biochem 1984; 141: 339-44. 5 Chan TCK, Markman M, Cleary S, Howell SB. Plasma uridine changes in cancer patients treated with the combination of dipyridamole and iV-phosphonacetyl+aspartate. Cancer Res 1986; 46: 3168-72. 6 Berlinger WG, Stene RA, Spector R, Al-Jurf A. Plasma and cerebrospinal fluid nucleosides and oxypurines in acute liver failure. J Lab Clin Med 1987; 110: 137-44. 7 Karle JM, Anderson LW, Cysyk RL. Effect of plasma concentrations of uridine oii pyrimidinebiosynthesisin culturedL1210 cells. J Biol Chem 1984; 259: 67-72. 8 Klubes P, Leyland-Jones B. Enhancement of the antitumor activity of S-fluorouracil by uridine rescue. Pharmacol Ther 1989; 41: 289-302. 9 Martin DS, Stolfi RL, Sawyer RC, Spiegelman S. Young CW. High-dose 5-fluorouracil with delayed uridine ‘rescue’ in mice. Cancer Res 1982; 42: 3964-70. 10 Khbtx P, Cema I. Use of utidine rescue to enhance the antitumor selectivity of S-fluorouracil. Cancer Res 1983; 43: 3182-6. 11 Peters GJ, van Dijk J, Laurensse E. et al. In vitro biochemical and in viva biological studies of the uridine ‘rescue’ of S-fluorouracil. Br J Cancer 1988; 57: 259-65.

Acknowledgement This study was supported by a grant from the Deutsche Forschungsgemeinschaft, SFB 154, ‘Klinische und experimentelle Hepatologie’.

URIDINE CATABOLISM

23 Moyer JD, Oliver JT, Handschumacher RE. Salvage of circulating pyrimidine nucleosides in the rat. Cancer Res 1981; 41: 3010-7. 24 Cooper GM, Dunning WF, Greer S. Role of catabolism in pyrimidine utilization for nucleic acid synthesis in vivo. Cancer Res 1972; 32: 390-7. 25 Dahnke H-G, k~osebach K-G. In-vivo-Untersuchungen zur Metabolisierung der Pyrimidinnucleoside. Hoppe-Seyler’s Z Physiol Chem 1975; 356: 1565-74. 26 Heidelberger C, Leibman KC, Harbers E, Bhargava PM. The comparative utilization of uracil-2-C14 by liver, intestinal mucosa, and Flexner-Jobling carcinoma in the rat. Cancer Res 1957; 17: 399-404. 27 Gasser T, Moyer JD, Handschumacher RE. Novel single-pass exchange of circulating uridine in rat liver. Science 1981; 213: 777 -8. 28 Monks A, Cysyk RI_. Uridine regulation by the isolated rat liver: perfusion with an artificial oxygen carrier. Am J Physiol 1982; 242: R465-70. 29 Hlussinger D. Isolated perfused rat liver: an experimental model for studies on ammonium and amino acid metabolism. Infusionstherapie 1987; 14: 174-8. 30 Sies H. The use of perfusion of liver and other organs for the study of microsomal electron transport and cytochrome P-450 systems. Methods Enzymoll977; 52: 48-59. 31 Sober HA. Purines, pyrimidines, nucleotides, oligonucleotides. In: Handbook of Biochemistry. Cleveland: The chemical Rubber Co, 1970; 61-238. 32 Leser H-G, Holstege A, Gerok W. The role of nonparenchymal and parenchymal liver cells in the catabolism of extracellular pu-

341 tines. Hepatology 1989; 10: 66-71. 33 Van den Bosch N, Driessen 0, van der Velde EA, Erkelens C. The stability of 5,6-dihydrofluorouracil in plasma and the consequences for its analysis. Ther Drug Monit 1987; 9: 443-7. 34 Holstege A, Leser H-G, Pausch J, Gerok W. Uridine catabolism in Kupffer cells. endothelial cells, and hepatocytes. Eur J Biothem 1985; 149: 169-73. 35 Csaky KG, LaCreta FP, Warren BS, Williams. 5-Fluoro-2’-deoxyuridine elimination by the isolated-perfused rat liver. Cancer Rcs 1988: 48: 3561-5. 36 Plagemann PGW, ‘Wohlhueter RM, Woffendin C. Nucleosides aild nucleobase transport in animal cells. Biochim Biophys Acta 1388; 947: 405-43. 37 Holstege A, Gengenbacher H-M, Jehle L, Hoppmann J. Facilitatcd diffusion and sodium-dependent transport of purine and pyrimidine nucleosides in rat liver. Hepatology 1991; 14: 373-80. 38 Plagemann PGW, Woffendin C. Na+-dependent and -independent transport of uridine and its phosphorylation in mouse spleen cells. Biochim Biophys Acta 1989; 981: 315-25. 39 Plagemann PGW, Wohlhueter RM, Erbe J. Facilitated transport of inosine and uridine in cultured mammalian cells is independent of nucleoside phosphorylase. Biochim Biophys Acta 1981; 640: 448-62. 40 Holstege A. Verwertung extrazellullrer Nukleoside in Tumorzellen und Leber. Habilitationsschrift. Freiburg, 1986. 41 Sommadossi J-P, Gewirtz DA, Diasio RB, Aubert C, Cano J-P, Goldman ID. Rapid catabolism of 5-fluorouracil in freshly isolated rat hepatocytes as analyzed by high performance liquid chromatography. J Biol Chem 1982; 257: 8171-6.