Eur. J . Biochem. 170,443-445 (1987) 0 FEBS 1987

Ester composition of carnitine in the perfusate of liver and in the plasma of donor rats Attila SANDOR, Gyula KISPAL, Bela MELEGH and Istvan ALKONYI Institute of Biochemistry, University Medical School, Ptcs

(Received April 13/July 13, 1987) - EJB 87 0434

When the carnitine pool of fed rats was labelled with tritium, in non-recirculating perfusate of their liver 44% of acid-soluble 3H activity was identified as free carnitine and 47% as short-chain acylcarnitine. Of the latter component acetylcarnitine accounted for 30% and propionylcarnitine for 10% of total acid-soluble.

In plasma the contribution of short-chain acylcarnitines to total carnitine in fed, fasted and diabetic rats was 15.6%, 43.1 % and 48.0%, respectively. Recirculating perfusion of livers from the same animals revealed that livers from fed rats released short-chain acylcarnitines as much as 56.2% of total and this proportion did not increase further in the other two groups. At the same time, ketone bodies in the perfusate increased gradually in the fed, fasted and diabetic group, paralleling the plasma ketone levels. Although liver supplies the organism with carnitine the increment of plasma short-chain acylcarnitines seen in ketosis is not a result of some extra output by the liver.

During enhanced fatty acid oxidation the level of plasma short-chain acylcarnitines sharply increases. This has been established in fasting and diabetic ketosis in adult humans [I - 31, in children [4,5] and animal studies [6,7]. The assump- tion that this increased plasma acylcarnitine/carnitine ratio reflects the corresponding ratio in the liver [8], in the organ of ketone body production, led to the conception that excess plasma acylcarnitines derive from the liver too.

The observation that plasma acylcarnitines parallel plasma ketone body level and liver acylcarnitine/carnitine ratio was made on rats [6,7]. In rats the liver supplies the whole organism with carnitine [9] and, additionally, liver releases carnitine in excess relative to the demand of the organism, as we established earlier, using perfused liver [lo]. These latter observations also support the idea that liver may be the sup- plier of surplus acylcarnitines. About 50% participitation of acylcarnitines was observed earlier in the carnitine pool re- leased by isolated rat liver cells [ l I] and by perfused rat liver [lo]. However, due to cell membrane demage during isolation [ 1 I], isolated liver cells have some limitations especially when the release of carnitine is to be examined.

As it appears from the above, the rat and its liver is a proper model to study the role of liver in the carnitine metabolism. This study was undertaken (a) to characterize the ester composition of carnitine released by the liver and (b) to ascertain the contribution of liver to the plasma short-chain acylcarni tines.

MATERIALS AND METHODS

Reagents L-Carnitine and acetyl-L-carnitine were gifts from Sigma-

Tau (Roma). Propionyl-L-carnitine was a kind gift from Dr

Correspondence to A. Sandor, Biokemiai IntCzet, Ptcsi Orvo- studomanyi Egyetem, Pecs, H-7624, Hungary

Enzymes. Carnitine acetyltransferase (EC 2.3.1.7); 3-hydroxy- butyrate dehydrogenase (EC 1.1.1.30).

H. Seim (Leipzig). [l-’4C]Acetyl-CoA was purchased from Amershdm. Dowex ion-exchange resins were from Serva (Heidelberg) and converted to the desired form as rec- ommended by the Bio-Rad catalog. Carnitine acetyl- transferase and 3-hydroxybutyrate dehydrogenase were from Boehringer (Mannheim). [Me-3H]Butyrobetaine was pre- pared as described [12].

Animals, perfusions and sample preparation Male Wistar rats (200-250 g) were used. Fasted animals

were deprived of food 24 h before experiment. Diabetes was induced by intraperitoneal injection of 3 5 mg streptozotocin. Plasma glucose levels in the diabetic animals averaged 19.7 1.9 mM vs 4.8 1.1 mM in controls (not shown). Rats were killed by decapitation and immediately their blood was collected into plastic heparinized tubes to obtain plasma. Livers were perfused in situ in recirculating or in outflow fashion or in combination as previously described [lo]. Unless otherwise stated, the perfusate did not contain added carnitine. In one type of experiment the carnitine pool of the animal was labelled in vivo with intraperitoneal injection of 1.5 x lo8 cpm of [Me-3H]butyrobetaine 12 h before exper- iment, by which time all activity was present in carnitine [12], then the liver was perfused in outflow mode. In order to desalt, neutralized perchloric acid extract (pH 7.4) of 1 ml effluent (Krebs-Ringer bicarbonate) was applied to the fol- lowing ‘double’ column contained by two pasteur pipettes: 0.5 cm x 7.0-cm Dowex 50x8 (200-400 mesh) in NHZ form for the upper one, 0.5 cm x 7.0-cm Dowex 1x8 (200- 400 mesh) in F- form for the bottom one. (This ‘double’ column was describer with Dowex 1 (X8) in -OH form for total carnitine [13]; the resin in F- form saves carnitine esters from hydrolysis and completely removes C1- and H2P0T .) The column was washed with 4.0 ml water and the collected effluent was lyophilized. The sample was reconstituted in 0.1 ml methanol and was used for TLC analysis.

444

Table 1. Levels of carnitines and ketone bodies in the liver, in the perfusate of the liver and in the plasma of the donor animals Livers were perfused in recirculating fashion with 30 ml perfusate for 60 min. Plasma of donor animals, livers and perfusates were then analyzed. Livers at the end of perfusion weighed ll.OkO.8 g, 7.9+ 0.7 g and 9.6k0.6 g in the fed, fasted and diabetic groups, respectively. Ketone bodies refers to the sum of acetoacetate and 3-hydroxybutyrate. SC indicates short-chain acylcarnitines. All numbers in the table and subtitle are mean k SEM for five animals

Condition of Plasma donors

Perfusate Liver tissue

carnitine ketone carnitine ketone carnitine

total free SC total free SC total free SC bodies bodies

nmol/ml % total nmol/ml nmol/ml % total

Fed 52.3 76.0 15.6 104 22.3 35.1 56.2

Fasted 45.3 43.5 43.1b 863b 15.3" 29.7 58.0

Diabetic 55.1 38.3 49.0' 907 23.1 31.5 57.5

k3.1 k1.2 k8.8 k2.1 f 5.2

k4.2 - +3.3 k55.5 k1.1 k 5.2

k 4.2 - +3.5 *lo1 - + 2.2 - + 4.8

pmol/ml nmol/g

0.88 206 +0.01 k15.1 2.30b 319b - + 0.4 k20.1 5.52b 382b k0.71 k30.1

~~

% total

50.6 39.1

43.9 37.1

40.9 39.8

& 3.8

k2.2

* 20.2

a P < 0.01 when compared with the two other groups. ' P < 0.001 when compared with the fed group. ' P < 0.01 vs fasted group and P < 0.001 vs fed group.

number of fr

0.34

1, M:O.42

h 15 20 ----

lions I4rnrn wch)

Fig. 1. Ester composition of acid-soluble carnitines in the outflowing perfusate from the liver of a fed rat. The carnitine pool of the rat was labelled in vivo, its liver was perfused in outflow fashion at a rate of 10 ml/min; 1.0 ml effluent was prepared for TLC as described under Materials and Methods. 40 pl final extract (9800 cpm) was applied and run to 145 mm. RF values were defined with standards: RF 0.22, free carnitine; RF 0.34, acetylcarnitine; RF 0.42, propionylcarnitine. All the 9000 cpm was recovered on the plate

Analyses

At termination of perfusion samples were frozen immedi- ately in liquid N2. Carnitine was assayed by enzymatic radioisotropic method in principle as described [14] and modi- fied [7]. Modifying the preocedure we introduced the Dowex 1 x 8 resin in F- form which ensures lower background than does the C1- form. Free carnitine, short-chain acylcar- nitines and long-chain esters were distinguished by the en- zymatic analysis as described 1151. TLC was performed on Whatman silica gel plate, linear K LK6D type (Pierce). The solvent was methanol/chloroform/water/ammonium formate (55:50:10: 7.5:2.5, v/v). Plasma glucose was assayed with a commercial kit GOD-POD (Galenopharm, Geneva). Ketone bodies were assayed enzymatically as described [16, 171.

RESULTS AND DISCUSSION

To examine the composition of released carnitine during non-recirculating perfusion requires the previous radioactive labelling of the carnitine pool [12]. The sensitivity of this method enables us determine the proportion of major individ- ual acylcarnitines, too. Fig. 1 shows the TLC analysis of acid- soluble carnitines of such an effluent. Of the radioactivity recovered on the plate, 44% was found in free carnitine, 30% in acetylcarnitine and 10% in propyonylcarnitine. Unidenti- fied esters accounted for 6.1%. All together, 44% free vs 47% short-chain acylcarnitines were recovered, which is fairly consistant with that found in recirculating perfusions (see below). The absolute amount released by the liver was quantified earlier [lo] and was found to be 6.8 nmol . min-' per 100 g body weight.

The results in Table 1 prove that livers perfused in re- circulating fashion maintain their in vivo feature with respect to ketogenesis: they gradually produce more ketone bodies in the fed, fasted and diabetic group, paralleling the plasma ketone levels. In marked contrast, short-chain acylcarnitines in the perfusate do not follow their increasing plasma concen- tration. Instead, the percentage of short-chain acylcarnitines in the perfusate, even in the fed group, reaches 56.2% and this relatively high proportion of acylcarnitines did not (could not) increase further in the other two groups.

The amount of released carnitine was significantly lower in the perfusate of 'fasted' liver (Table 1). When calculated on the basis of carnitine content of whole liver, this decrease is more pronounced [lo] and also valid for the diabetic group. (Carnitine content of the whole liver can be calculated from data of Table 1. The increase in the diabetic group is signifi- cant, r < 0.001.) This alteration of hepatic carnitine transport in fasting state was observed earlier and reported in details [lo]. The proportion of short-chain acylcarnitines, as a per- centage of the total, did not increase in the liver tissue (Table l) , while the acylcarnitine/free carnitine ratio increases in agreement with other reports [7, 11, 161. The contribution of long-chain carnitine esters can be calculated from the Table 1 ; it shows an about parallel increase in each source.

445

Table 2. Outflow perfusion of rat livers loaded with L-carnitine Livers were loaded by recirculating perfusion for 30 min with initial 0.5 mM, 1 .O mM, 5 mM and 10 mM L-carnitine in perfusate in groups I, 11, I11 and IV, respectively. Then outflow perfusion was initiated at 10 ml/min and the effluent after the first 100 ml was analyzed. Results were corrected for extracellular space as previously described [lo]. Numbers are averages of three perfusions

Group Carnitine in

liver perfusate

total short chain total short chain

nmol/g Ya total nmol/min Ya total I 278 22.3 18.8 44.7 TI 562 20.9 36.3 49.5 111 1344 25.7 70.1 42.8 IV 1989 21.3 82.7 45.9

When livers were loaded with carnitine to different extents the release of carnitine increased as a function of tissue level (Table 2) . The increasing tissue levels can saturate the rate of release as we reported with respect to total carnitine [lo]. Table 2 shows that the short-chain esters maintained a con- stant relative amount while the total amount increased 4.3- fold in the effluent. This suggests that free carnitine and short- chain carnitine esters share a common protein to promote the release.

This report shows that carnitine released by liver is not esterified to a higher degree when the liver derives from the ketonic rat. Of course, these livers can not constantly release a higher total amount, because it would lead to depletion of the tissue. On the contrary, in starving ketosis, liver releases less of the total carnitine for the body, as follows from the reduced urinary excretion and constant total body content [7]. The presented data prove that ketotic livers do not release an extra amount of short-chain acylcarnitines into the blood stream under the experimental condition. Earlier we estab- lished an excessive (futile) release/re-uptake cycle of carnitine transport across the liver cell membrane [lo, 181. The resultant of the two opposite transport processes is a net release accord- ing to the carnitine-exporting function of the liver. One might suppose that ketotic livers take up preferably free carnitine. This shift in the efflux/influx processes in favour of releasing carnitine esters is not likely to take place since it should have been found in the recirculating experiments also. However, the ratio of short-chain acylcarnitines in the recirculating perfusate of livers from ketotic groups was not higher than that in fed controls.

Lastly, the question arises of how the increased plasma level of short-chain acylcarnitines develops in the ketotic state.

We propose two possibilities. First, a decreased utilization by the extrahepatic tissues. Perhaps in muscle, fatty acids (and ketone bodies) can compete with acylcarnitines for oxidation. (A decreased loss by urinary excretion does not change the ratio because in starvation free and acylcarnitines were found to be conserved proportionally [7].) Secondly, extra short- chain acylcarnitines may derive from an extrahepatic source. In muscle the proportion and absolute amount of acylcarnitines do not increase in ketosis [7, 12, 191. However, the proportion of acetylcarnitine markedly increases in the kidney in starvation [12, 191, which finding makes the kidney a possible source of extra plasma acylcarnitines.

We thank Maria Miklosvari and Zsuzsa Hillebrand for their skilled technical assistance.

REFERENCES 1. 2.

3.

4.

5.

6. 7.

8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Genuth, S. M. & Hoppel, C. L. (1979) Diabetes 28,1083-1087. Hoppel, C. L. & Genuth, S. M. (1980) Am. J . Physiol. 238, E409-

Frochlich, J., Seccombe, D. W., Hanh, P., Dodek, P. & Hynie, I.

Cederblad, G., Hermansson, G. & Ludvigsson, J. (1982) Clin.

Soltesz, G., Melegh, B. & Sandor, A. (1983) Acta Ped. Scand. 72,

Brass, E. P. & Hoppel, C. L. (1980) Biochem. J . 190,495-504. Brass, E. P. & Hoppel, C. L. (1978) J. Biol. Chem. 253, 2688-

Bremer, J. (1983) Physiol. Rev. 63, 1428. Bshmer, T. (1974) Biochim. Biophys. Acta 343,551 - 557. Sandor, A,, Kispal, Gy., Melegh, B. & Alkonyi, I. (1985) Biochim.

Christiansen, R. Z. & Bremer, J. (1976) Biochim. Biophys. Acta

Bshmer, T., Norum, K. R. & Bremer, J . (1966) Biochim. Biophys.

E415.

(1978) Metabolism 27, 555-561.

Chim. Acta 125,207-217.

51 1 - 513.

2693.

Biophys. Acta 835,83 - 91.

448,562 - 577.

Acta 125. 244-251. Davis, A., Ingalls, S. T. & Hoppel, C. L. (1984) J . Chromatogr.

306,79 - 87. Cederblad, G., Lindstedt, S. & Lundholm, K. (1974) Clin. Chim.

Acta 53, 311-321. Pearson, D. J., Tubbs, P. K. & Chase, J. F. A. (1974) in Methods

ofenzymatic analysis (Bergmayer, H. U., ed.), pp. 1758- 1771. Verlag Chemie, Weinheim.

Williamson, D. H. & Mellanby, J. (1974) in Methods of enzymatic analysis (Bergmayer, U. H., ed.) pp. 1836-1839, Verlag Chemie, Weinheim.

Mellanby, J. &Williamson, D. H. (1974) in Methods of enzymatic analysis (Bergmayer, H. U., ed.), pp. 1840-1843, Verlag Chemie, Weinheim.

Kispal, Gy., Melegh, B., Alkonyi, I. & Sandor A. (1987) Biochim. Biophys. Acta 896,96 - 102.

Pearson, D. J. & Tubbs, P. K. (1967) Biochem. J . 105,953 -963.