103

Biochimica et Biophysics Acta, 1047 (1990) 103-111 Elsevier

BBALIP 53506

Effect of chronic heparin administration on serum lipolytic activity and some aspects of lipid metabolism D.M. Goldberg and T. Chajek-Shaul Department

of Internal Medicine B and Lipid Research Laboratory, Hadassah Universig

Hospital, Jerusalem (Israel)

(Received 18 December 1989) (Revised manuscript received 15 June 1990)

Key words: Lipoprotein lipase; Hepatic triacylglycerol lipase; Heparin; VLDL; Phospholipid; Adipose tissue; (Heart); (Liver); (Skeletal muscle); (Serum); (Rat)

Chronic heparin administration to rats for periods up to 8 days by i.p. implantation of mini pumps, increased serum total lipolytic activity in a dose-dependent manner up to infusion rates of 10 U/h per 100 g body weight. This augmentation was predominantly due to lipoprotein lipase (LPL). Synchronously, heart muscle demonstrated a dose-dependent reduction in LPL activity and adipose tissue showed a biphasic response, LPL activity decreasing at low doses and rising towards control levels at higher doses. Lipolytic activities of skeletal muscle and liver were unaffected. Increased serum LPL could not be attributed to altered enzyme clearance from the circulation in chronically heparinised rats, but was accompanied by a reduced response to i.v. high-dose heparin indicating reduction in the pool of endotheliaf-bound enzyme. Fasting sermn concentrations of triacylglycerol and glycerol were unaffected in chronically heparinised animals although accelerated clearance of exogenous “C-labelled VLDL was demonstrated, together with enhanced uptake of the isotope by liver and heart. Since de novo synthesis of fatty acids and triacylglycerol from 3Hz0 was not increased by heparin, we suggest that serum triacylglycerol concentrations were maintained by enhanced re-esterification of preformed fatty acids taken up by the liver. Hepatic cholesterol synthesis from 3H20 was augmented by heparhq this observation is consistent with reported increases in serum total and HDGchofesterol mediated by chronic heparin administration in man and dog.

The anticoagulant drug heparin has been used in high dose in the acute treatment of thrombo-embolic disease for half a century. More recently, interest has focused upon the long-term use of low-dose heparin in the prophylaxis of these conditions and of atherosclerosis with which they are frequently associated [l-4]. Several actions of heparin may benefit or prevent the development of atherosclerosis besides its well known anti-coagulant effect, viz. inhibition of smooth muscle cell proliferation, and displacement of lipolytic enzymes from the endothelial surface, a process that minimises damage caused by cholesterol-rich lipoproteins (see Ref. 2 for review).

Abbreviations: LPL, lipoprotein Iipase; HTGL; hepatic glycerol lipase; VLDL, very-low-density lipoprotein.

triacyl-

Correspondence (Present address): D.M. Goldberg, Department of Clinical Biochemistry, University of Toronto, 100 College Street. Banting Institute Toronto, Ontario M5G lL5, Canada. 00052760/90/$03.50

The ability of heparin to displace two endothelialbound lipolytic enzymes, lipoprotein lipase (LPL; EC 3.1.1.34) and hepatic triacylglycerol lipase (HTGL; EC 3.1.1.3) into the circulation has been known for three decades [5-71. In man, after a bolus i.v. injection of heparin in doses ranging from lo-100 U/Kg, plasma lipolytic activity reaches a peak 5-30 min later and returns to the basal level within 2-5 h [8]. Repeated injections in the same individuals yield reproducible values for peak lipolytic activities [9]. Exhaustion of tissue lipolytic enzymes has not been demonstrated by these experiments which, for practical reasons, have been conducted over relatively short time periods. During acute i.v. heparin injection, hydrolysis of triacylglycerol-rich lipoproteins (chylomicrons and VLDL) occurs, leading to reduced concentrations of triacylglycerol and increased levels of fatty acids and glycerol in the circulation [lo-121. These changes are transient and normal plasma lipid concentrations are usually restored within l-4 h of the heparin injection. In contrast to our knowledge of the enzymatic and metabolic effects of acute high-dose heparin injection,

0 1990 Elsevier Science Publishers B.V. (Biomedical Division)

104 little is known about the long-term chronic effects of low-dose heparin administration upon the distribution of lipolytic enzymes between the tissues, the endothelial surfaces, and the circulation although it has been speculated that chronic heparinisation per se or associated with renal dialysis programs may deplete tissue lipases and lead to lower endothelial-bound or circulating levels of these enzymes [13,14]. The changes in circulating triacylglycerol, and in the metabolism of VLDL, that might accompany long-term heparin administration have not been defined. Displacement of the enzymes from their physiological sites might also be expected to disturb the organism’s ability to handle endogenous and exogenous lipids. We have, therefore, used a rat model to investigate the effects of chronic heparin administration on: (a) circulating and tissue lipases; and (b) some aspects of triacylglycerol and VLDL metabolism.

milk LPL, purified according to Bengtsson-Olivecrona and Olivecrona [16] and with specific activity of 441 mU/yg protein, by injection into the femoral vein. Blood was subsequently drawn by cannula 1, 5 and 15 min later. The zero-time activity was calculated from the dilution of the administered LPL assuming a plasma volume of 4.2% of body weight, and the clearance rate was determined, subtracting the basal lipolytic activity from that of all subsequent samples. Acute heparin pulse As in the above experiment, blood was drawn through an aortic cannula before, and 1, 10 and 20 rnin after injecting 1 U of heparin per g body weight into the femoral vein. Deter~nations of total lipolytic activity and activity resistant to 1 M NaCl were performed, and baseline values were subtracted from those of subsequent samples to derive the incremental activity released by heparin at each time interval.

Materials and Methods Animals Male rats of Hebrew University strain, weighing 150-175 g, received a pelleted diet (Am Rod 931, Hadera, Israel) ad libitum, but in some experiments were fasted I6 h prior to sacrifice. Heparin was administered to certain rats by peritoneal impl~tation of Alzet @ mini-osmotic pumps, (Model 2001 (Alza Corporation, Palo Alto, CA, U.S.A.), Lot No. 60887, specified mean pumping rate 0.96 f 0.06 PI/h and mean fill volume 210 + 6 ~1) under light ether anesthesia. The standard solution of heparin comprised 140 mg (Sigma, Sodium salt, Grade 1 from porcine intestinal mucosa; 17X USP units/mg, Lot 34F-0742) dissolved in 1 ml of 0.15 M NaCl, giving a dosage of 140 pg or 15 units (U) per h, but other concentrations were used as indicated in experiments to assess the heparin dose-response relationship. Controls initially received unfilled pumps, but were subsequently subjected to laparotomy alone when it was established that the food intake and growth rate and changes in serum lipolytic activity in the two groups were sim.iIar. Unless otherwise indicated, blood was collected from the tail vein under light ether anesthesia, but to terminate the study animals were killed by exsanguination and various tissues were collected. Homogenates of liver, heart muscle, skeletal muscle (diaphragm) and acetoneether powders of adipose tissue (epi~dymal fat pads) were prepared as previously described flS] for assay of lipolytic enzymes. Clearance of exogenous LPL Under light ether anesthesia, a cannula was placed in the aorta and blood was taken to determine basal Iipolytic activity. Each animal received 50 pg bovine

incorporation of ‘H,U into lipid tractions These studies were based on the procedures of Kalderon et al. [17] and were performed on control rats and those receiving the standard dose of heparin 5 days after operation, commencing at 10 a.m. without prior fasting. 25 mCi of 3H,0 (Amersham, U.K., TRS Lot 76, 5 Ci/ml) was injected i.p. and the animals were killed by exsanguination 4 h later. Residual blood was flushed out with 50 ml cold isotonic saline through an aortic cannula after which various tissues were collected, blotted dry, weighed and 1 g of each as well as 1 ml of serum were homogen~ed in 20 ml of Folch reagent [IS], using the Polytron homogenizer (Kinematica, Luzern, Switzerland) with a pt lo-11 probe at maximum speed for 0.5-l min at 0 o C. Determinations of radioactivity in the chloroform phase were used to calculate the incorporation of ‘H into total lipids; the relative percentages of the total counts attributable to each lipid fraction were obtained after separation by thin-layer chromatography; the incorporation of 3H into esterified fatty acids was measured after their saponification in ethanolic KOH and extraction into heptane. A small aliquot (2-10 ~1) or a suitable dilution of native serum was counted directly and the specific activity of serum water was derived as dpm per mole assuming a serum water content of 52 mmol/ml. This was then used to calculate the mmoles of 3H,0 incorporated into the various lipid fractions of each tissue in each animal. All the above procedures were performed as previously described 1171. Hepatic triacylglycerol secretion rate This was determined in controls and rats treated with the standard dose of heparin 5 days after operation ~mmen~ing at 10 a.m. without prior fasting. 0.5 FCi of t3Hfpalrnitate complexed to bovine serum albumin and

105 1 mg Triton WR 1339 (Serva, Heidelberg, F.R.G.) per g body weight were injected into the femoral vein and 90 min later the animals were killed by exsanguination. The fractional secretion of hepatic triacylglycerol was calculated as total dpm in serum triacylglycerol (assuming total serum volume to be 4.2% of body weight) divided by the sum of total serum triacylglycerol dpm and total liver triacylglycerol dpm as previously described [ 191. Fractional hepatic triacylglycerol secretion was also determined by measuring the increment of triacylglycerol in serum which accumulated during the 90 min period; this was expressed as a percentage of the sum of total hepatic and total serum triacylglycerol.

tions were made with the Peridochrom Triglycerides GPO-PAP kits (Cat. Nos. 701882 and 877557; Boehringer Mannheim) which also allow free glycerol measurement. Results from both procedures agreed within 7.5% and were averaged for presentation. Absorbance was monitored with the Spectronic 601 (Milton Roy, Rochester, NY, U.S.A.). Radioactivity was determined with a Tricarb beta scintillation spectrometer (Minaxy, Packard) with absolute activity analyzer. The scintillation fluid was 20% Triton X-100, 0.5% POPOP and 0.4% PPO in toluene. Routine centrifugation was performed with the Model J-6B (Beckman Instruments).

Clearance of labelled VLDL

Materials In addition to those already specified, bovine serum albumin and Triton X-100 were from Sigma (St. Louis, MO, U.S.A.). All others were of highest analytical grade.

Rat serum was obtained by exsanguination 45 min after i.v. injection of 75 pCi of [ 3H]palmitate complexed to bovine serum albumin. VLDL were isolated by ultracentrifugation for 20 h at 190000 X g in the SWA 41 Rotor using the L5-50B ultracentrifuge (Beckman Instruments, Fullerton, CA, U.S.A.) and collected by tube-slicing. The final preparation had a specific activity of 658 dpm/pg triacylglycerol at a concentration of 0.95 mg triacylglycerol/ml. 1 ml was injected into a femoral vein and 5 min later the animal was killed by exsanguination and residual blood was washed out as described above. Tissues were collected and 500 mg portions were homogenized in 10 ml of methanol/chloroform (1 : 1, v/v). After centrifugation at 3 000 x g for 10 min, the liquid phase was decanted into glass vials and evaporated to dryness before adding scintillation fluid and was counted to determine the uptake of VLDL or its degradation products. The radioactivity of serum was also measured to determine the residual circulating VLDL. Specific assays Lipolytic activities in serum and tissues were determined at 37°C as previously described [20] and the data are presented as micromoles of substrate hydrolysed per min (U). The glycerol tri-[9,10(n)-3H]oleate, specific activity 1 Ci/mmol, was obtained from Amersham International, U.K. Salt-resistant lipase was assayed after incubating serum in a final concentration of 1 M NaCl for 10 min at room temperature synchronously with controls at a final concentration of 0.15 M-NaCl before adding substrate to initiate the reaction. In preliminary experiments we were able to demonstrate that 1 M NaCl and goat antiserum for LPL similarly inhibited serum lipolytic activity. Triacylglycerol content of serum and liver was assayed after extraction and saponication by an enzymatic method allowing independent determination of free glycerol, using ultraviolet-test kits (Cat. No. 125032; Boehringer Marmheim, Tutzing, F.R.G.) as previously described [21]. In some experiments parallel determina-

Statistical evaluation Student’s t-test was used to assess the significance differences between groups. All data are presented mean f S.E.

of as

Results Although food consumption during the g-days between operation and sacrifice was significantly lower in rats receiving heparin at 15 U/h (530 f 19 g/kg; P < 0.01) and at 30 U/h (476 + 16 g/kg; P < 0.005) than in controls (627 f 30 g/kg), body weight was not significantly altered. Serum lipolytic activity Total lipolytic activity of the serum was higher in heparin-treated rats than in controls at all intervals after operation (Table I). Preliminary assays performed in some animals at 6, 30 and 54 h after operation indicated that peak activity occurred between the latter two times. The data in Table I show that activity with all heparin doses was highest at 48 h and declined slowly over the next 6 days. In general, the values in rats receiving 6 U/h were at least double those of rats receiving 3 U/h at all times; in like fashion, rats infused with 15 U/h increased their serum lipolytic activity at least 2.5-fold above the values in those infused with 6 U/h. Thus, from 3-15 U/h, heparin appeared to raise the serum lipolytic activity in a dose-dependent manner. A further increase in serum lipolytic activity was seen in rats infused with 30 U/h of heparin but this fell considerably short of a doubling of the values in those receiving 15 U/h, suggesting a plateau in the, response. In a small number of experiments, the serum lipolytic activities in fed and fasting rats were compared in sham-operated animals and those receiving 15 U/h of heparin. The latter animals demonstrated abolition of

106 TABLE

1

Serum lipolytic activity at variow intervals after sham-operatron

(controls) or implantation

of peritoneal pumps delivering heparrn at specified dose rues

Open data are total Epolytic activities and data in square brackets represent lipoiytic activity resistant to 1 M NaCl in the same animals as U/l at 37“C, mean& SE. Dete~ati~ns of total lipolytic activity were performed on 6-14 male rats in each group weighing 150-175 g, except for highest heparin dose where only 2 were used. For salt resistant activity. data from 3-4 rats are given. 24 h after operation, serum total lipoiytic activities were 10.9 + 1.6, 13.2 f 2.4 and 52.0 + 7.3 in controls and rats treated with 6 U/h heparin and 15 U/h, respectively. (n = 6 for all groups). Treatment

Baseline

group

Days after procedure 2

Controf

7.6 * 0.6 j2.3 f 0.31 7.6 * 0.7

Heparin 3 U/h Heparin 6 U/h Heparin 15 U/h Heparin 30 U/h

8.1 zt 1.3 7.6 + 0.6 6.1 kO.8

5 6.1kO.5

8.4k1.2 [0.9 + 0.31 21.4? 2.4 [2.4+0.8] 71.8k5.3 H3.6 + 3.5] 100.5 + 6.6 [17.0&2.7]

8” 5.98 f 0.50

3.77*

0.55

6.50k0.80 H.33rfrO.271 16.2 -fr:3.8 [3.1$-1.01 45.02 8.7

5.92+

0.60

[6.3+1.3] 77.0 * 7.6 HO.1 ct 2.81

15.4+ 3.6 [1.9& 0.61 36.0+ 7.7 [7.1 i 2.11 46.5 + 15.2

a The 8-day sample was collected between lo-12 a.m. after a 16 h fast, so that assays for serum triacylglycerol and free glycerol and for tissue lipolytic enzymes could be performed in these animals in a defined nutritional state. All other samples were at the same time of day but under non-fasting

conditions.

Adipose Tissue

Serum 40

q n

Fasting Non-Fasting

Control Fig. 1. Effect

of fasting

Heparin Treated

Control

Heparin Treated

and feeding on lipolytic activity of serum and adipose tissue in control and heparin-treated commencing the experiment. Bars represent mean i S.E. (n = 3 in all groups).

rats (15 U/h)

48 h after

107 the usual increase in serum lipolytic activity in response to fasting at 48 h after commencing heparin (Fig. l), whereas in sham-operated controls the fasting values were significantly higher than those of non-fasting animals (t = 2.47; P < 0.05). At all time periods after heparin, as well as in the controls, the lipolytic activity resistant to 1 M NaCl was a rather small fraction of the total (Table I), 80-90s of which apparently comprised true LPL. This conclusion was supported by a number of studies aimed at characterizing the enzyme utilizing specific antibodies to LPL (unpublished data) and by the use of a specific assay for HTGL (unpublished data). Nevertheless, an increase in this salt-resistant activity was noted with the higher heparin dose schedules, the activities averaging 6-fold above baseline values at 48h with the 15 U/h dose and 7.7-fold at this time with the 30 U/h dose, in comparison with the respective increments of 9.5-fold and 16fold in total lipolytic activity under these conditions. Tissue lipolytic activities After 8 days, heparin treatment did not influence the lipolytic activity of diaphragm or liver from fasting animals in a consistent or significant way. Neither did the distribution of hepatic lipolytic activity between salt-resistant and salt-labile fractions alter. There was a decrease in the lipolytic activity of heart muscle with increase in the dose of heparin (Table II). This reduction was significant relative to the values in controls in animals receiving the dose of 30 U/h (t = 2.85; P < 0.05). In adipose tissue, lipolytic activity was significantly decreased at the two lower doses of heparin, but rose progressively with increasing dose.

TABLE

II

Lipolytic activities of rat heart and adipose tissue d-days after sham-operation (control) or implantation of peritoneal pumps containing heparin at different dose schedules Data (mean* SE.) as mU/g wet weight of heart and mU/g acetone powder for adipose tissue. Animals were sacrificed between lo-12 a.m. after a 16 h fast Treatment group (No. of rats)

Heart

Adipose

Control (8) Heparin (6) 3 U/h Heparin (8) 6 U/h

1032 f 205 853+208

3 137 f 218 1427f182b

845k132

tissue

1887+240

Heparin 15 U/h

(7)

652+

85

2 763 f 365

Heparin 30 U/h

(2)

446k

16a

3465k772

a Significantly b Significantly ’ Significantly

lower than control lower than control lower than control

(t = 2.85; P < 0.05). (t = 6.02; P -C0.001). (t = 3.85; P -C0.005).

=

These observations were essentially confirmed in fasted rats (n = 3, each group) sacrificed after 2 days of treatment. As before, heparin did not affect the lipolytic activity of diaphragm or liver, but reduced the activity in heart muscle. Compared with values (means + S.E.) of 1063 k 22 mU/g in the controls, the activities in heart muscle of heparin-treated rats were 780 + 48 at 6 U/h and 658 + 181 at 15 U/h; the difference between the first two groups was highly significant (t = 5.40; P < 0.01). In view of the abolition by heparin of the expected increment in serum lipolytic activity consequent upon fasting, we compared adipose tissue lipolytic activity in control rats and those receiving 15 U/h heparin under fasting and non-fasting conditions 48 h after commencing heparin (Fig. 1). In control and heparinized rats adipose tissue lipolytic activity was reduced by fasting to 14% and 26%, respectively, of the values in non-fasting animals, suggesting that heparin may blunt the sensitivity of adipose tissue lipolytic activity to changes in nutritional status. The lipolytic activity of the fasting heparin-treated rats was more than twice that of controls (t = 2.39; P < 0.05) but the differences in the non-fasting state were not significant. Clearance of exogenous LPL A potential explanation for the increase in serum lipolytic activity after heparin administration was a delayed clearance of the circulating enzyme. This possibility was tested by measuring the disappearance of purified bovine LPL given i.v. to sham-operated animals, and to two groups of rats treated with heparin (15 U/h) for 2 days and 8 days. Previous experiments demonstrated that bovine LPL behaves in vivo and in organ perfusion systems in the same way as rat adipose tissue LPL [22]. In all three groups the disappearance of exogenous LPL was rapid and the time-course of disappearance was virtually identical during the 15 min observation period, during which 90% of the injected LPL activity disappeared (Fig. 2). Thus, no evidence could be obtained to support the notion that chronic heparin administration at the doses used in this study delays the clearance of circulating LPL, and this cannot be a factor in raising the serum enzyme activities in this condition. Superimposition of acute high-dose heparin injection Since increased serum lipolytic activity during chronic heparin administration did not seem to be accompanied by delayed enzyme clearance, and the increased serum activity could only partially be attributed to intracellular depletion within the tissues examined, we considered the possibility that the enzyme may be liberated predominantly from the vascular or surface endothelial cells. If this were so, depletion of the enzyme might be expected to limit the incremental activity usually seen

108

-

CONTROL(nd)

-.---I-

HEPAFIIN. 2 Days (n3) HEPARIN. 8 Days (n-3)

Minutes After LPL Injection Fig. 2. Clearance of bovine milk LPL from serum of control rats on day 8 of experiment and rats receiving heparin (15 U/h) on days 2 and 8 of the experiment. 50 cg LPL was injected via the femoral vein and blood was sampled before this injection and 1, 5 and 15 min after injection by means of an aortic cannula. The zero-time value was calculated as the basal activity in each animal plus the concentration of injected LPL assuming a plasma volume of 4.2% of body weight. Data are mean + S.E.

after a bolus i.v. injection of heparin. Accordingly, 1 unit of heparin per g body weight was injected into the femoral vein after drawing blood at zero-time and blood samples were obtained from an aortic cannula 1, 10 and 20 min later. At all times the activities of total lipase, LPL and HTGL were lower in the rats chronically treated with heparin (Table III). The difference in HTGL between the two sets of rats was less striking than for LPL, but in keeping with the known action of

TABLE

III

Post-heparin release of lipolytic enzymes into the ~erwn of control and heparinised rats after S-days at 15 U/h Heparin at a dose of 1 U per g body weight was injected into fasting rats via the femoral vein. Blood was sampled through an aortic cannula before this injection, and 1, 10 and 20 mm after injection. The zero-time activity for each animal was subtracted from the subsequent values to give the incremental enzyme activity released by heparin. Hepatic lipase (HTGL) was taken to be the activity resistant to 1 M-NaCl. LPL was the difference between the latter and total lipolytic activity. Data are mean + SE. Time after i.v. heparin 1

(min)

10

20

Total lipolytic activity (U/l) (n=5) 1120&135 1235f 53 14451k 63 control 996 f 215 584+130 a 891f 92 a heparin-treated (n = 4) HTGL (U/l) 947 * 139 (n=5) 651+ 67 848* 40a control 793f170 442* 94 542+111 heparin-treated (n = 4) LPL (U/l) (n=5) 470+llZa 3875 49 500+ 98” control 201* 49 166+ 48 353+ 50 heparin-treated (n = 4) a Heparin-treated

rats significantly

lower than control

( P < 0.05).

acute heparin in large dosage and in contrast with our previous characterization of serum lipolytic activity in rats chronically treated with heparin, HTGL accounted for most of the total serum lipolytic activity, ranging from 60-80% in the heparin-treated animals and 58-68% in the controls. These results suggest that chronic heparin administration increases circulating lipolytic activity by depleting endothelial surface-bound LPL (and possibly also, to a lesser extent, HTGL) without consistent effect upon intracellular tissue levels of the enzyme. This implies that less enzyme is available for release during massive acute heparin administration. Serum triacylglycerol The serum concentration of triacylglycerol was assayed in fasted rats 8 days after starting heparin treatment in doses ranging from 3 U/h to 30 U/h, or after sham-operation. Surprisingly, serum triacylglycerol concentration tended towards higher values in the heparintreated animals than in the controls, but these trends were not statistically significant. Neither did free glycerol concentrations show any particular trend between the groups, suggesting that increased intravascular lipolysis is not a response to chronic heparin treatment, or that alternatively the resynthesis and secretion of triacylglycerol and reutilization of the products released by lipolysis are sufficiently enhanced to balance the latter. In vivo lipogenesis To determine whether de novo synthesis of lipids, and in particular triacylglycerol, was affected by chronic heparin treatment, we measured the incorporation of 3H,0 into various lipid fractions 4 h after injecting 25 mCi i.p. Significantly lower incorporation of tritium into total serum lipids and esterified fatty acids was noted (Table IV), and this was attributed to a sharply reduced incorporation into triacylglycerol which averaged only about 10% of the control values in the heparin-treated rats. Quantitatively less dramatic but statistically significant reductions in the incorporation of 3H into serum mono- and diacylglycerol and phospholipids also followed heparin treatment. The livers of heparin-treated rats also seemed to show reduced ‘H incorporation into total lipids, triacylglycerol and phospholipids in the chloroform fraction but these changes were not statistically significant. However. a significant reduction in the incorporation into mono- and diacylglycerol was demonstrated in these animals (P < 0.05). A striking finding was a dramatic increase in 3H incorporation into hepatic cholesterol in the heparin-treated rats. Paradoxically, there was reduced incorporation into serum cholesterol in these animals. Heparin treatment did not affect 3H incorporation into cholesterol esters of liver or serum, nor did it modulate total lipid or triacylglycerol ‘H content of adipose tissue.

109 TABLE

IV

Incorporation

of tritiated water into various lipid fractions of serum, liver and adipose tissue in sham-operated

and heparin-treated

rats (j-days

at 15

U/h) 4 h after i.v. injection of the label 25 mCi of 3H20 was injected i.p. and animals were killed 4 h later. Serum, liver and adipose tissue were extracted with Folch reagent [19]. The final washed chloroform layer was taken for scintillation counting and separation of lipids by thin-layer chromatography. An aliquot of the chloroform layer was saponified with ethanolic KOH, after which the hydrolysed fatty acids were extracted into heptane for scintillation counting. The specific activity of serum water was determined (see text) and used to calculate incorporation of the label into each lipid fraction. All procedures were as described in Ref. 17 pmol tritiated water incorporated into:

Chloroform

Whole serum control (3) heparin (4) Whole liver control (3) heparin (4) Adipose

285+ 76+

Heptane

extracts

total lipid

triacylglycerol

59.3’ 5.1

1895 + 335 1510+227

155.2* 15.9+ 1036 782

69.0 b 8.1

+250 +215

phospholipid

mono/ diacylglycerol

15.7* 8.0* 232 101

2.4 = 1.5

f39.8 k21.1

g

63.8 f 12.6 d 26.9+ 6.0 464 300

+96 +50

cholesterol

19.9+ 4.5*

cholesterol ester

6.2 e 0.6

36.4+ 2.6 ’ 204.2 f 19.0

pm01 water into: fatty

8.4k1.94 5.6k1.26 47.3f9.5 52.9 f 6.0

extract

tritiated incorporated esterified acids

239 f 56.3+ 1500 1076

61.0 7.6

’

+360 +280

tissue

(per g) control (2) heparin (4)

225* 225+

14.5 42.5

191.5 f 190.2 f

15.9 36.1

_

_

_

150.0+ 161.9 f

15.9 28.5

a f = 3.51, P < 0.02. d t = 2.64, P qO.05. gt= 2.91, P ~0.05. b t = 2.01, P > 0.05. e t = 2.57, P < 0.05. h t = 8.75, P < 0.001. ’ t = 2.71, Pi 0.05. ‘t = 2.97, P < 0.05.

Clearance of labeled VLDL

hanced to provide an adequate source for resynthesis of triacylglycerol. Table V shows that VLDL clearance was indeed increased in heparin-treated rats. The percentage of injected dose remaining in the circulation after 5 mm was only 28.5% in the heparinised rats compared with 43.5% in the controls (P -C0.05). Simultaneously, there was a 230% increase in uptake of label by the liver (P < O.OOl), from 7.7% to 17.8% of the dose administered. Reduced uptake of 3H by muscle, adipose tissue

Despite higher circulating lipase levels, rats treated chronically with heparin did not manifest a reduction in serum triacylglycerol concentrations. Yet they did not seem to increase their de novo synthesis of fatty acid or of triacylglycerol utilizing these newly synthesised fatty acids. We, therefore, tested the possibility that VLDL was degraded more rapidly in heparinised animals and that peripheral uptake of fatty acid was thereby en-

TABLE

V

Clearance of “H-1abelled VLDL from serum and uptake of “H by tissues 5 min after i.v. injection in sham-operated for j-days) rats Injected

dose corresponded

to 625.103

dpm and 0.95 mg triacylglycerol Control (n=4)

Total serum dpm.10e3 Total liver dpm.10e3 Total heart dpm Adipose

tissue (per g)

dpm Muscle (per g)

272 f (43.5 f 48 f (7.7* 850 f

in 1 ml 0.15 M NaCl

Heparin-treated (n =4)

t

P

178 (28.5 111 (17.8 1272

2.73

< 0.05

6.52

< 0.001

2.97

i 0.025

f 22 f 3.5) + 5.1 f 0.8) *111

1509

f407

997

f168

NS

870

*176

664

f

NS

dpm ’ Data in parenthesis NS, not significant.

27 4.3) a 8.2 1.3) 89

as VLDL suspended

(control) and heparin-treated

are % of dose injected.

81

(15 U/h

110 and heart seemed to occur in the heparinised animals, but only in the latter tissue was the change significant (P < 0.025). Hepatic triacylglycerol secretion rate Using the Triton WR 1339 technique, the secretion of triacylglycerol by the liver was measured in control and heparin-treated rats (15 U/h for 5 days). The fractional secretion of triacylglycerol over the 90 min period calculated from incorporation of [ ‘H]palmitate into liver and serum triacylglycerol was 52.7 + 4.5% in the controls and 40.6 k 6.5% in the heparin-treated rats (n = 4, each group). Based upon the accumulation of chemically determined triacylglycerol in serum as a percentage of total hepatic triacylglycerol, values for fractional clearance were 34.9 + 1.8% and 31.5 + 2.1% respectively. In neither instance was the difference in triacylglycerol secretion rate statistically significant. Discussion Acute heparin injection is known to dramatically increase the activities and concentrations of lipolytic enzymes (LPL and HTGL) in the circulation. One mechanism is displacement of endothelial-bound enzymes by competition with heparan sulphate which is believed to be the natural ligand present at the surface of endothelial cells [5,7]. Another possibility is inhibition of enzyme clearance, which is otherwise very rapid and mediated predominantly by the liver [23]; a third is direct stabilization of the enzyme by heparin [24,25]. Enhanced secretion is a further potential mechanism and has been demonstrated in vitro for adipocytes, skeletal muscle and heart muscle [26-281. Our results with chronic heparin treatment clearly show a dose-dependent increase in circulating lipolytic activity up to a rate of 15 U/h in rats weighing 150-175 g, corresponding to 85-100 U/h per Kg body weight. The lipolytic response to chronic low-dose heparin in our rats was qualitatively different from that occurring subsequent to acute i.v. bolus injection in the rat (Table III) and in man [29-311. Under the latter circumstances, HTGL accounts for 60-80% of the lipolytic activity, but with low-dose heparin true LPL predominated according to the criteria of inhibition by 1 M NaCl and by antisera specific for LPL [15], as well as by specific catalytic assays for HTGL activity [32]. This observation is consistent with previous results when rats were treated with i.p. heparin in doses as low as 0.1-0.5 U/Kg [33]. One explanation might be that the liver is able to degrade heparin quite efficiently, but this capacity is overwhelmed at high heparin doses; only under the latter circumstance is the heparin concentration in the hepatic circulation adequate to displace hepatic lipase from its endothelial binding sites. The increased LPL activity was not due to delayed clearance of the

enzyme (Fig. 2). It could be largely attributed to displacement of endothelial-bound LPL as evidenced by impaired response to an i.v. bolus of heparin (Table III). The tendency to reduced LPL content of heart at all heparin doses and of adipose tissue at the lower doses (Table II), probably reflects depletion of the LPL pool at the cell surface rather than reduced intracellular LPL content. Although the LPL released into the circulation by high dose i.v. heparin is able to rapidly hydrolyse triacylglycerol-rich lipoproteins in hyperlipidemic subjects or animals [lo,12214.331. the ability of circulating LPL to do so in normo-lipidemic subjects and over longer time intervals is less certain [ll]. Our results show that serum triacylglycerol and glycerol concentrations were not affected by chronic heparin administration. Repeated injections of heparin to human volunteers increased HDL-cholesterol but no other lipid fraction was altered [9]. After 6 days of treatment with 20000 U heparin per day, serum triacylglycerol concentrations actually increased [14], and were unchanged after 1 year of low-dose therapy [l]. Indeed. some authors have speculated that chronic heparin administration may lead to impaired lipolysis by depletion of endothelial-bound enzyme [13,14,34], or by reduced apo ClI/CIII ratio ]351. We were able to demonstrate enhanced uptake of exogenous VLDL by liver and heart in heparin-treated rat, synchronously with their accelerated clearance from the circulation (Table V). Yet de novo synthesis of fatty acids by the liver was unaffected (Table IV), suggesting that in these animals enhanced clearance was compensated by re-esterification of pre-formed fatty acids by the liver. We were unable to demonstrate increased hepatic triacylglycerol secretion as might have been expected to compensate for augmented VLDL clearance from the circulation. Triton WR 1339 inhibits LPL and is likely to block the supply of fatty acids available to the liver as a consequence of increased intravascular lipolysis mediated by heparin. Enhanced triacylglycerol secretion would still be demonstrable if de novo fatty acid synthesis was the major mechanism compensating this increased lipolysis, rather than reesterification of preformed fatty acids. Our results, therefore, favour the latter as the predominant compensatory mechanism in response to chronic heparin administration, and they are consistent with the observation that almost all triacylglycerol fatty acids of VLDL are derived from free fatty acids taken up from the splanchnic circulation ]361. The increased de novo synthesis of hepatic cholesterol which apparently occurred in heparin-treated animals (Table IV) was unexpected. We did not measure the serum concentration of cholesterol or its associated lipoproteins, although it has been shown that heparin increases serum total and HDL-cholesterol concentra-

111 tions [9,37]. Whether this occurs by blocking the binding or internalization of cholesterol-rich lipoproteins, or by an increase in their synthesis and secretion is not known, but our preliminary observations favour the latter. Viewing our results as a whole, it is possible that the observed effects of heparin upon lipid metabolism can be wholly attributed to increased activity of serum lipolytic enzymes. However, a direct effect of heparin upon cells (especially hepatic cells), has to be seriously entertained and warrants further investigation. Acknowledgements We thank Professors. Y. and 0. Stein for their advice and encouragement and for providing facilities for this research. We also thank Miss 0. Halimi for expert technical assistance, and Dr. J. Etienne for kindly providing goat anti-serum to rat lipoprotein lipase. DMG was supported by the Medical Research Council of Canada, and TC-S by a grant from the Hebrew University Medical Fund. The kits for triacylglycerol and glycerol determinations were a generous gift of Boehringer Mannheim (Dr. S. Hess) and Agentek (Mr. A. Kornreich). References 1 Buchwald, H., Rohde, T.D., Schneider, P.D., Varco, R.L. and Blackshear, P.J. (1980) Surgery 88, 507-516. Revs. 36, 91-110. 2 Engelberg, H. (1984) Pharmacol. G.F., Camovali, M, Rovelh, F., 3 Neri Semeri, G.G., Gensini, Pirelh, S. and Fontini, A. (1987) Lancet i, 937-942. 4 Neri Semeri, G.G., Abbate, R., Prisco, D., Camovali, M., Fazi, A., Casolo, G.C., Bone&i, F., Rogasi, P.G. and Gensini, G.F. (1988) Am. Heart J. 115, 60-67. 5 Robinson, D.S. (1963) Adv. Lipid Res. 1, 133-182. B. and Brockman, 6 Kinnunen, P.K.J. (1984) in Lipases (Borgstrom, H.L., eds.), pp. 307-328, Elsevier, Amsterdam. Lipase (Borensztajn, J., ed.), I Robinson, D.S. (1987) in Lipoprotein pp. l-33, Evener, Chicago. P. (1987) in Lipoprotein Lipase (Borensztajn, J., 8 Nilsson-Ehle, ed.), pp. 59-77, Evener, Chicago. P.D., Kantor, M.A., Cullinane, E.M., Sady, S.P., Sari9 Thompson, telli, A. and Herbert, P.N. (1986) Metabolism 35, 999-104. R., Hamosh, M., Swasubramanian, K.N., Chowdhry, 10 Dhanireddy, P., ScanJon, J.W. and Hamosh, P. (1981) J. Pediatrics 98, 617-622. R.A., Logan, R., Russell, D.C., Smith, H.J., Simpson, 11 Riemersma, J. and Oliver, M.F. (1982) Brit. Heart J. 48, 134-139.

12 Benderly, A., Rosenthal, E., Levi, J. and Brook, G. (1983) J. Parent. Ent. Nutr. 7, 37-39. 13 Chan, M.K., Varghese, Z., Persaud, J.W., Baillod, R.A. and Moorhead, J.F. (1980) Clin. Chim. Acta 108, 95-101. 14 Persson, E., Nordenstrom, J. and Vinnars, E. (1987) Acta Anesthesiol. Stand. 31, 189-192. 15 Knobler, H., Chajek-Shaul, T., Stein, 0.. Etiemte, J. and Stein, Y. (1984) B&him. Biophys. Acta 795, 363-371. 16 Bengtsson-Olivecrona, G. and Olivecrona, T. (1985) Biochem. J. 226, 409-413. 17 Kalderon, B. Adler, J.H., Levy, E. and Gutman, A. (1983) Am. J. Physiol. 244, E480-E486. 18 Folch, J.H., Lees, M. and Sloane-Stanley, G.H. (1957) J. Biol. Chem. 226,497-510. 19 Chajek-Sham. T., Friedman, G., Stein, O., Shiloni, E., Etienne, J. and Stein, Y. (1989) Biochim. Biophys. Acta 1001, 316-324. 20 Friedman, G., Gallily, R., Chajek-Sham, T, Stein, O., Shiloni, E., Etienne, J. and Stein, Y. (1988) B&him. Biophys. Acta 960, 220-228. 21 Goldberg, D.M., Roomi, M.W., Yu, A. and Roncari, D.A.K. (1980) B&hem. J. 192, 165-175. 22 Chajek-Sham, T., Friedman, G., Ziv, E., Bar-On, H. and Bengtsson-Olivecrona, G. (1988) B&him. Biophys. Acta 963, 183-191. 23 Ohvecrona, T. and Bengtsson-Olivecrona, G. (1987) in Lipoprotein Lipase (Borensztajn, J., ed.), pp. 15-58, Evener, Chicago. 24 Bengtsson-Ohvecrona, G. and Olivecrona, T. (1985) Biochem. J. 226, 409-413. 25 Jackson, R.L., Socorro, L., Fletcher, G.M. and Cardin, A.D. (1985) FEBS Lett. 190,297-300. 26 Chajek, T., Stein, 0. and Stein, Y. (1978) Biochim. Biophys. Acta 528, 456-465. 21 Vannier, C.. Jansen, H., Negrel, R. and Ailhaud, G. (1982) J. Biol. Chem. 257, 12387-12393. 28 Vannier, C. and Ailbaud, G. (1986) Biochim. Biophys. Acta 875, 324-333. 29 Rovamo, L.. Nikkila, E.A., Taskinen, M.R. and Ravio, K.O. (1984) Pediatr. Res. 18, 1104-1107. 30 Zaidan, H., Dhanireddy, R., Hamosh, M., Bengtsson-Olivecrona, G. and Hamosh, P. (1985) Pediatr. Res. 19, 23-25. 31 Eckel, R.H., Goldberg, I.J., Steiner, L., Yost, T.J. and Patemiti, J.R., Jr. (1988) Diabetes 37, 610-615. 32 Hemell, O., Egehud, T. and Olivecrona, T. (1975) B&him. Biophys. Acta 381, 233-241. 33 Zaidan, H., Gutman, A., Berkow, S., Hamosh, P. and Hamosh, M. (1984) Pediatr. Res. 18, 1321-1324. 34 Bagdade, J.D., Porte, D., Jr. and Bierman, E.L. (1968) New Engl. J. Med. 279, 181-185. 35 Havel, R.J., Fielding, C.J., Olivecrona, T., Shore, V.G., Fielding, P.E. and Egelrud, T. (1973) Biochemistry 12, 1828-1833. 36 Have], R.J., Kane, J.P., Balasse, E.O., Segel, N. and Basso, L.V. (1970) J. Clin. Invest. 49, 2017-2035. 31 Blackshear, P.J., Rohde, T.D., Varco, R.L. and Buchwald, H. (1977) Atherosclerosis 26, 23-27.