Intracellular transport of asialoglycoproteins in rat hepatocytes

Experimental

Intracellular Evidence

Cell Research 161 (1985) 285-296

Transport of Asialoglycoproteins in Rat Hepatocytes for Two Subpopulations

TROND BERG,‘,* TERRY FORD’

of Lysosomes

GRETE M. KINDBERG,’ and RUNE BLOMHOFF’

‘Znsfitute for Nutrition Research, University of Oslo, 0316 Oslo 3, Norway and ‘Department of Biology, University of Essex, Wivenhoe Park, Colchester, Essex, UK

The intracellular transport and degradation of asialoorosomucoid (AOM) in isolated rat hepatocytes was studied by means of subcellular fractionation in Nycodenz gradients. The asialoglycoprotein was labelled by covalent attachment of a radioiodinated tyraminecellobiose adduct ([1251]TC) which leads to labelled degradation products being trapped intracellularly and thus serving as markers for the degradative organelles. The ligand was initially (1 min) in a slowly sedimenting (small) vesicle and subsequently in larger endosomes. Acid-soluble, radioactive degradation products were first found in a relatively light lysosome whose distribution coincided in the gradient with that of the larger endosome. Later (30 min) degradation products were found in denser lysosomes which banded in the same region of the gradient as the lysosomal enzyme, #I-acetylglucosaminidase. Colchicine, monensin and leupeptin all inhibited degradation of [‘251]tyramine-cellobiose asialoorosomucoid ([1251]TC-AOM) and reduced the formation of degradation products in both the light and the dense lysosomes. In presence of monensin and colchicine no undegraded ligand was seen in the dense lysosome, suggesting that uptake in these vesicles was inhibited. Leupeptin allowed accumulation of undegraded ligand in the dense lysosome. Therefore, transfer from light to dense lysosomes is not dependent on degradation as such. In the presence of monensin two peaks of undegraded ligand were found in the gradients. It seems possible that in the monensin-sensitive endosomes, dissociation of the ligandreceptor complex is inhibited, allowing ligand to recycle with the receptors in small vesicles. @ 15%kademif press,1~.

Liver parenchymal cells take up a variety of molecules from the circulation via receptor-mediated endocytosis. These include peptide hormones [7], haptoglobin-hemoglobin complexes [ 181, immunoglobulin A [23], chylomicron remnants 1193, epidermal growth factor (EGF) [ll, 22, 301, and asialoglycoproteins [16]. The uptake and intracellular transport of asialoglycoproteins in rat liver parenchymal cells has been studied in cells [l], in the perfused liver [9], in monolayers [37] and in suspensions [32] of isolated cells. Morphological techniques have shown that the ligand, after binding to receptors in coated pits, is transported via coated vesicles and endosomes to multivesicular bodies and secondary lysosomes [27]. The ligand is ultimately degraded in lysosomes [16], while the * To whom offprint requests should be sent. Address: Institute for Nutrition Research, University of Oslo, P.O. 1046 Blindern, 0316 Oslo 3, Norway. 19-858342

Copyri&t @ 1985 by Academic F’ress, Inc. AU rights of reproduction in any form reserved 0014-48z7/85 $03.00

286

Berg et al.

receptor is reused [29]. Early after uptake, a proportion of the ligand/receptor complexes formed may be recycled directly to the cell surface, while in the remainder the complex rapidly dissociates with the receptor being recycled to the surface and the ligand in the endosome, continuing along its path to degradation [35, 381. The morphological findings are supported by subcellular fractionation studies by means of sucrose [6], and Percoll [24] gradients. Such studies show that the ligand is internalized in a vesicle of relatively low density and later found in denser fractions which correspond to the distribution of lysosomal enzymes. Thus, subcellular fractionation has identified a prelysosomal (endosomal) and a lysosomal step in the heterophagy of asialoglycoproteins. However, there is an obvious need for subcellular fractionation techniques that could identify the steps in the intracellular transport of ligand in more detail than those tested so far. Improved fractionation techniques could be a useful alternative to electron microscopy for studying the kinetic aspects of intracellular transport and metabolism of the asialoglycoprotein ligand in liver parenchymal cells. In the present work we have studied the intracellular transport of asialoglycoproteins by centrifugation in Nycodenz gradients. By exploiting differences in the rate of sedimentation we were able to identify a small, “early” vesicle involved in the uptake process. Further, by using asialoglycoproteins (asialoorosomucoid) labelled with [‘251]tyramine-cellobiose ([‘*‘I]TC) we could also identify, in the gradient, structures involved in degradation of the ligand. Degradation of [1251]tyramine-cellobiose asialoorosomucoid ([‘*‘I]TC-AOM) leads to labelled degradation products being trapped in the lysosomes. These degradation products may therefore serve as markers for the organelles involved in their formation [25]. A preliminary account of this study has been presented elsewhere [4].

MATERIALS

AND METHODS

Chemicals Nycodenz was obtained from Nyegaard & Co AS, Oslo. Asialoorosomucoid

was prepared as before [33] and labelled by covalent attachment of a radioiodinated tyramine-cellobiose adduct ([‘z51]TC) according to Pittman et al. [25]. The adduct of tyramine-cellobiose was a generous gift from Ray Pittman, Department of Medicine, University of California, San Diego, Calif. Collagenase, enzyme substrates, albumin (fraction V), colchicine, and leupeptin were from Sigma Chemical Co. St. Louis, MO.). [lUI]Na was from the Radiochemical Centre, Amersham, UK.

Animals and Cells Isolated rat liver parenchymal cells (hepatocytes) were prepared essentially according to Seglen [28] as described before [6]. Male Wistar rats, about 200 g, were used. The animals were given food and water ad libitum. The cells were incubated in a medium consisting of 8.5 g NaCI, 0.4 g KCl, 0.06 g Na2HP04x2H20, 0.047 g KH2P0.,, 0.2 g MgS04x7H20, 4.76 g Hepes, 0.29 g CaC12x2Hz0, and water to 1000 ml. pH was adjusted to 7.5 by the addition of NaOH. Osmolality was 300 mOsm/l. Albumin (1 %, w/v) was included in the incubation medium, unless otherwise stated. The cell viability was always above 90% (trypan blue exclusion test). fip

Cell Res 161 (1985)

Degradation Subcellular

of asialoglycoproteins

287

fractionation

The rat hepatocytes (5 ml, 10’ cells/ml) were washed twice in an isotonic sucrose solution (0.25 M sucrose, 1 mM EDTA, 1 mM HEPES, pH 7.2) and homogenized by 20 strokes in a tight-fitting Dounce homogenizer. About 5x10’ cells in 5 ml of sucrose solution were homogenized. A postnuclear supematant was prepared by centrifugation at 2000 g for 2 min. The nuclear pellet was resuspended in 4 ml of sucrose solution, rehomogenized and centrifuged again for 2 min at 2000 g. The two post-nuclear supematants were mixed and 4 ml of the combined fractions were layered on to linear Nycodenz gradients prepared by mixing 0.25 M sucrose solution (see above) with 35 % (w/v) Nycodenz (dissolved in 1.2 mM KCl, 0.1 mM CaNa EDTA, 0.1 mM NaCl, and 0.2 mM ‘I&-HCl, pH 7.5) in 40 ml centrifuge tubes. The volume of the gradient itself was 32 ml. The gradients were usually centrifuged for 45 min at 85 000 g. Under these centrifugation conditions smaller endocytic vesicles will not reach their buoyant density and can therefore be separated from larger vesicles of similar buoyant density [20]. Following centrifugation the gradients were divided into 18x2 ml fractions by upward displacement using Maxidens as the displacement fluid.

Analytical procedures Radioactivities were measured in a Kontron gamma counter. Degradation of [iZ51]TC-AOM was followed by measuring radioactivity soluble in 10% (w/v) trichloroacetic acid (IXZA) (acid-soluble radioactivity). /?-Acetylglucosaminidase was assayed according to Barrett [2] and S’nucleotidase according to El-Aaser & Reid [12]. The density of each fraction was obtained from its refractive index using the formula: d=n.3.41-3.555 (n, refractive index). The degradation products in the gradient are not formed because of proteolysis in the gradient fractions or the gradient itself. No increase in labelled degradation products were noted in fractions stored up to 2 days at room temperature.

Design of experiments To get a relatively synchronized cellular uptake of labelled asialoglycoprotein the cells were first incubated at 4°C for 1 h in the presence of 50 nM [“‘I]TC-AOM. At this temperature the ligand is bound to the surface receptors but not internalized. The cells were then washed twice in ice-cold medium and resuspended in incubation medium at 37°C. At various times after the start of the incubation at 37°C cell aliquots were removed (usually 5 ml suspension containing 10’ cells/ml) and diluted with four parts of ice-cold medium. The cells were sedimented and washed twice by centrifugation (50 g, 1 min) before homogenization and fractionation in Nycodenz gradients, as described above.

RESULTS Intracellular transport and degradation of [‘251]TC-AOM in isolated rat hepatocytes and control conditions

Suspensions of hepatocytes ‘&ere first’“allowed to bind [‘2’I]TC-AOM to surface receptors at 4°C for 1 h. The cells were then washed and incubated in new medium at 37°C. During the incubation at 37°C aliquots of cells were removed after 1; 5; 15; 30; 60 and 90 min. The cell samples were homogenized and postnuclear fractions prepared and centrifuged in Nycodenz gradients as described in Material and Methods. After fractionation the gradient fractions were treated with 10% TCA to separate acid-soluble degradation products from undegraded ligand. The values for acid-soluble radioactivity underestimate degradation to some extent [25] but should nevertheless indicate the time-course for intracellular degradation. The acid-soluble radioactive degradation products were analyzed by gel filtration: their molecular weights were about 1000-2000 (fig. l), probably representing the TC ligand still attached to short peptides. Exp Cell Res 161 (1985)

288 Berg et al. -

'2

i

24-

void

cyanocobalamin

I

20-

1

_

0

2

4

6

8

IO

12

14

16

18

20

Fig. 1. Gel fdtration in Sephadex G-75 of soluble radioactivity in 10% TCA. The cells were incubated with 50 nM [‘*‘I]TCAOM for 2 h at 37”C, and then precipitated in the presence of 10% TCA. Aliquots,(0.2 ml) of the supematant were separated on a Sephadex G-75 column (0.9~ 10 cm). Blue Dextran (MW=2x l@) and ! cyanocobalamin l(MW= 1355) were ‘analysed in separate runs.

FRACTION NUMBER

Fig. 2 shows the distribution of acid-soluble and acid-precipitable radioactivities after centrifuging post-nuclear fractions from cells incubated as described for various time intervals at 37°C. The distribution curves for acid-precipitable radioactivity reveal that the ligand is initially in an organelle which bands at a density of 1.05 g/ml and, at later stages, in vesicles whose distribution shows a peak at 1.10 g/ml. No acid-soluble radioactivity was detected in the gradient before 15 min of incubation at 37°C. From this time point, acid-soluble radioactivity started to appear at a density of about 1.10 g/ml in the gradient. These early degradation products coincided with the acid-precipitable radioactivity and were separated from the main /I-acetylglucosaminidase activity (fig. 2). However, the lysosomal enzyme showed a well-defined shoulder in the region of the radioactivity peak. After 30 min an additional peak of radioactive degradation products appeared in the gradient at a density of about 1.13 g/ml. This latter peak was almost coincident with the lysosomal enzyme #I-acetylglucosaminidase but at slightly lower densities than the enzyme. No detectable undegraded ligand was associated with this high-density subcellular fraction containing degraded [‘?]TCAOM. The present data confirm earlier studies [4, 201, and suggest that the asialoglycoprotein is sequentially associated with (at least) five organelles or vesicles: the plasma membrane, the “small” vesicle at 1.05 g/ml, the larger endocytic vesicle (density 1.10 g/ml), and two types of lysosomes, a light and a dense lysosome. The denser endosome and the light lysosome coincide in the gradient but are functionally different organelles. EJsects of inhibitors on the subcellular distribution of [“‘I,’ TC-AOM Colchicine. Fig. 3 shows the effect of colchicine (added at a concentration of 25 uM at the start of incubation at 37°C) on the subcellular distribution of degraded and undegraded [‘251]TC-AOM. The colchicine-treated cells contained only 50% of degraded ligand, as compared to control values after 60 min, with concomitantExpCell

Res 161 (1985)

Degradation

of asialoglycoproteins

IO 60min

289

r--f-

D

5

Fig. 2. Subcellular fractionation of hepatocytes containing 0, degraded; or 0, undegraded [iz51]TCAOM. The cells were first incubated at 4°C for 1 h in the presence of 50 nM [12SI]TC-AOM, and after removing extracellular ligand incubated at 37°C. At the indicated times cell ahquots were removed and fractionated in linear Nycodenz gradients, as described in Material and Methods. (A-E) Distribution of 0, acid-soluble; 0, acid-precipitable radioactivity; (F’) distribution of /Iacetylglucosaminidase for cells incubatedforO(A)and9Omin(A). Radioactivities are presented as % of total cell-associated radioactivity at the start of the incubation at 37°C. Total cell-associated radioactivity changed insignificantly during the incubation at 37°C. /J-Acetylglucosaminidase is presented as % of total recovered activity in the gradient. The recovery of radioactivity or enzyme activity in the gradient, as % of that layered on to the gradient before centrifugation varied between 88 and 98 %. (A) (Cells incubated 1 min at 37°C) results from one typical experiment. The other results are means +SE for six experiments.

L2?!isd R; ! 23

IC

DENSITY

1.05

1.10

1.15

(g/ml)

ly twice as much undegraded ligand. These findings are compatible with the earlier reports [21, 391 of retarded degradation rates in the presence of colchicine. In the presence or absence of colchicine, degradation products first appeared in a low-density fraction (1.10 g/ml) and later in a denser fraction, but the rate of appearance of degradation products was much lower in the presence.of colchitine. No undegraded ligand was detected in the denser (1.13 g/ml) fraction in either case. Monensin. Monensin inhibits both uptake and degradation of asialoglycoproteins in rat hepatocytes [3, 131. The reduced uptake is partly due to a reduction in the number of available surface receptors [3]. It has been proposed that degradation of internalized ligand is reduced because it enters the lysosomes at a reduced rate [3, 171. The effect of monensin added at a concentration of 25 uM at the start of incubation at 37°C on the subcellular distribution of degraded and undegraded Exp Cell

Res 161 (1985)

290 Berg et al. coNTFtoL

COLCHICINE

5

5

4

’

60min

Fig. 3. Effect of colchicine on the distribution of 0, degraded; 0, undegraded [‘251]TC-AOM after centrifitgation in Nycodenz gradients. Cells containing surface-bound [‘*‘I]TC-AOM were incubated at 37°C in the presence or absence (control) of 100 uM colchicine. After (A, C) 30; (B, 0) 60 min samples of cells were removed and fractionated by centrifugation in Nycodenz gradients as described in Material and Methods. 0, Acid-soluble; 0, acid-precipitable radioactivities were determined in the gradient fractions.

B

4

4

3

2

I

123 105

DENSITY

1.10

115

(g/ml)

[‘*‘I]TC-AOM is depicted in fig. 4. The drug practically abolished degradation; after 60 min of incubation at 37°C the formation of acid-soluble radioactivity in monensin-treated cells was reduced to 10 % of control values. The distribution of undegraded ligand after fractionating monensin-treated cells showed two peaks, at densities 1.06 and 1.10 g/ml (fig. 4). The peak at lower density was at the same position in the gradient as the small vesicle-containing ligand early (
Degradation

of asialoglycoproteins

291

MONENSIN E 3

,. 60 m m 3

c

1

i ’ 21 1.05

.I 1

DENSITY

1.10

1.15

i

Fig. 4: Effect of monensin on the distribution of 0, degraded; 0; undegraded [‘ZSI]TC-AOM after centrifugation in Nycodenz gradients. The cells were incubated for 1 h at 4°C in presence of 50 nM [‘251]TCAOM and then, after washing to remove~extracellularI&rnd incubated at 37°C in the (E, F, G) presence; or (A, B, C) absence of 25 uM monensin. Aliquots of cells, removed after 30 and 60 min, were fractionated by centrifugation in Nycodenz gradients (as detailed in Material and Methods). (C, c) Distribution of(C) /I-acetylglucosaminidase for control; (G) monensin-treated cells incubated for 1 h at 37°C.

(g/ml)

denz gradients was carried out with cells removed both before and 30 mitt after the addition of EGTA, as well as on cells incubated in the’ continued absence of EGTA. The results are depicted in fig. 6 and show that addition of EGTA resulted in release of the ligand from the vesicle banding at a density of 1.10 g/ml. The amount of radioactivity recovered at 1.06 g/ml remained approximately the same. Leupeptin. Leupeptin is a tripeptide which inhibits thiol proteinases including cathepsin B. It markedly inhibits the degradation of endocytosed asialoglycoproteins in rat hepatocytes, both in the perfused liver [lo], and in isolated cells [5, 341. Fig. 7 shows the effect of leupeptin on the subcellular distribution of degraded and undegraded [‘2’I]TC-AOM after fractionating cells in Nycodenz gradients. The cells were first allowed to bind [“‘I]TC-AOM to surface receptors at 4°C and were then washed and incubated for various time intervals at 37°C (without extracellular ligand) in the presence and absence of leupeptin (added at a concentration of 0.1 mM). The subcellular distribution of undegraded ligand in leupeptin treated cells was closely similar to the controls apart from in the lower five fractions in the gradient; in this region leupeptin treatment led to an additionExp Cell Res 161

(1985)

292

Berg et al.

lEYYIiI 15

A

min

MONENSIN

SO

min

2

--

123 1.05

110

115

DENSITY

(g/ml)

Fig. 5. Effect

of adding monensin to cells which had internalized [iz51]TC-AOM for 15 min at 37°C. Cells with surface-bound ligand were incubated at 37°C for 15 min. Monensin (25 uM) was then added to one portion of the cells and the incubation in the presence and absence of monensin was continued for a further 15 min. Cell samples were removed after the first (A) 15 min after the additional incubation for 15 min in (C) presence; (B) absence of monensin. The curves show the density distribution of undegraded [‘Z51]TC-AOM in the Nycodenz gradients.

B

30min

(\ P

:/,!A

I

123 1.05

1.10

1.15

al peak in the distribution curve. This high density subcellular fraction coincided with the lysosomal band. The density distribution of acid-soluble radioactivity in gradients after fractionation leupeptin-treated cells showed that labelled degradation products accumulated in both low and high density subcellular fractions; their rate of formation was, however, much lower than in control cells.

MONENSIN

CCWROL

DENSITY

(g/ml)

Fig. 6. Distribution patterns of [‘251]TC-AOM in Nycodenz gradients after EGTA-induced release of ligand. Following binding of [‘*‘I]TC-AOM to surface receptors at 4°C the cells were incubated at 37°C in the presence and absence of monensin (25 @I) for 30 min. Portions of both control cells and monensin-treated cells were then recovered from the tlasks and incubated in presence of 5 mM EGTA (to release any ligand that was recycled back to the cell surface). After an additional 30 min the cells were fractionated by centrifugation in Nycodenz gradients, as described in Material and Methods. The curves show the distribution in the gradient of [‘251]TC-AOM (undegraded) in the n , presence; 0, absence of EGTA. The results for monensin-treated cells are shown in (B).

Exp CeilRes 161(1985)

Degradation

of asialoglycoproteins

293

LEUPEPTIN 6 30min

D

5

4 3 2

:

~~

Fig. 7. Effect of leupeptin (0.05 mg/ml) on the distribution of 0, degraded; 0, undegraded [“‘I]TCAOM after centrifugation in Nycodenz gradients. Cells with surfacebound ligand were incubated at 37°C in (0. E, F) presence; (A, B, C) absence of leupeptin. Cell samples were fractionated in Nycodenz gradients tier 30 and 60 min of incubation. Acid-soluble and acid-precipitable radioactivities as well as Bacetylglucosaminidase were measured in the gradient fractions. (C, D) Distribution of ~-acetylglucosaminidase for cells which had been incubated for 1 h in the (C) absence or (F) presence of leupeptin. DENSITY

(g/ml)

DISCUSSION By cell fractionation in Nycodenz gradients we could separate three groups of subcellular structures with which [‘251]TC-AOM was sequentially associated. During the first minute the ligand was in a relatively small, slowly sedimenting, vesicle, which under the centrifugation conditions used banded at about 1.05 g/ml. The ligand was rapidly transferred to larger, more rapidly sedimenting structures which were distributed in the middle part of the gradient around a density of 1.10 g/ml. Finally, from 30 min and onward, the labelled degradation products reached a subcellular particle which was located in the same region of the gradient as the peak of lysosomal enzymes. I5.q Cell Res 161 (1985)

294

Berg et al.

The early, small vesicle could be a coated vesicle as the transit time for the ligand in the vesicle corresponds fairly well with that found by morphological methods for asialoorosomucoid in coated vesicles [36]. In agreement with our data Debanne et al. [8] recently reported that asialoorosomucoid early after uptake was in a slowly sedimenting vesicle. The radioactivity banding in the gradient around a density of 1.10 g/ml is probably contained in at least two different types of vesicles. Acid-soluble degradation products first appeared in this region of the gradient indicating that low density lysosomes must be present in addition to endosomes. The effect of monensin of the density distribution of ligand may suggest that a proportion of the endosomes is insensitive to the drug. Monensin directed part of the ligand into a smaller vesicle. However, some of the ligand still remained in the endosomes banding at 1.10 g/ml. It is tempting to ‘relate this finding to the observation that substantial fractions of asialoglycoprotein ligands endocytosed by hepatocytes can subsequently be returned to the cell surface still bound to the receptor 135, 381. Since the ligand dissociates from its receptor at low pH it is conceivable that the recycling ligand is passed through a non-acidic intracellular compartment. This compartment could be the “monensin-insensitive” endosome. This is in agreement with the finding that the addition of EGTA to monensin-treated cells released ligand from the endosome banding at 1.10 g/ml. The vesicle containing EGTA-releasable ligand may resemble the “mildly acidic para-Golgi compartment” described by Yamashiro et al. [40]. This compartment is involved in the recycling of intemalized transferrin in CHO cells. The origin of the small ligand-containing vesicle seen after monensin treatment is not clear. Monensin leads to rebinding of ligand to receptor [17], and this may introduce ligand into the tubular portions of the endosome (the “CURL”) or the organelle normally carrying the receptor back to the plasma membrane. Both these structures will probably be slowly sedimenting. The lysosomal nature of the degradation taking place in the light organelle was indicated by the very marked inhibitory effect of leupeptin and monensin. Degradation in the low density lysosomes was also markedly inhibited by colchicine. This drug does not inhibit degradation as such but rather the formation of the light lysosomes. The early degradation products in the gradients coincided with the undegraded ligand, and it is reasonable to assume that degradation is initiated by the uptake of acid hydrolases by endosomes. This is in agreement with the observations of Geuze et al. [14, 151 who suggest that the free ligand collects within the lumen of the endosomes and that these vesicles fuse with primary lysosomes. Wall et al. [36] investigated the internalization of asialoglycoproteins in rat hepatocytes by means of electron microscopic tracers and found that 15 min after injection ligand was contained in vesicles in the Golgi-lysosome area which also contained aryl sulfatase, indicating fusion with lysosomes. On the other hand, ligand was not at this time seen in the residual bodies. These observations may also suggest that lkp Cell Res 161 (1985)

Degradation

of asialoglycoproteins

295

degradation starts in lysosomes which are different from those containing the bulk of the lysosomal enzymes. However, the rapid and continuous degradation of ligand is incompatible with the formation of secondary lysosomes exclusively by fusion between endosomes and primary lysosomes. De novo formation of lysosomal enzymes simply does not take place at a sufficiently high rate. The contents of the light lysosome may be transferred to the dense lysosome, e.g., after fusion between the two organelles. The transfer may not be dependent on the degradation itself. In other words, transfer may occur even though degradation is slowed down. In cells treated with leupeptin, degradation of internalized [“‘I]TC-AOM is nearly blocked; still, undegraded ligand accumulated in dense lysosomes (albeit at a reduced rate [5]). The transfer is probably dependent on intact microtubuli, since accumulation of degraded or undegraded ligand is strongly reduced in colchicine-treated cells. Both monensin and ammonium ions blocked uptake of ligand, degraded or undegraded, in the dense lysosomes. This is in accordance with our previous studies [6] and with data from other laboratories [17]. Maybe proton gradients across the membranes is a prerequisite for fusion/interaction between primary lysosomes, endosomes and secondary lysosomes. Evidence for intermediate lysosomal steps in endocytic pathways is also provided by Storrie et al. [31], Pertoft et al. [24], and Rome et al. [26]. All these reports describe endocytic processes in which a ligand [24, 261 or a fluid phase marker [31];‘is first found in a low density lysosomal population and later in a higher density lysosomal population. Interestingly, Seglen and coworkers have recently described a similar sequence of low density and high density lysosomes in the autophagic processing of [i4C]sucrose in rat hepatocytes. The authors are grateful to MS Ba Gorbitz for her skilled assistance with the manuscript. The work was financially supported by the Norwegian Council for the Humanities, the Nansen Foundation, and the Norwegian Council on Cardiovascular Diseases.

REFERENCES 1. Ashwell, G & Morell, A G, Adv enzymol41 (1974) 99. 2. Barrett, A T, Lysosomes, a laboratory handbook (ed J T Dingle) p. 46. North-Holland, Amsterdam (1972). 3. Berg T, Blomhoff, R, Naess, L, Tolleshaug, H & Drevon, C A, Exp cell res 148 (1983) 319. 4. Berg, T, Ford, T, Kindberg, G, Blomhoff, R & Drevon, C A, Exp cell res 156 (1985) 570. 5. Berg, T & Tolleshaug, H, Acta biol med germ 40 (1981) 1599. 6. - Biochem pharmacol29 (1980) 917. 7. Carpentier, J L, Gorden, P, Barrazone, P, Freychet, P, LeCam, A & Orci, L, Proc natl acad sci US 76 (1979) 2803. 8. Debanne, M T, Bolyos, M, Gauldie, J & Regoeczi, E, Proc natl acad sci US 81 (1984) 2995. 9. DUM, W A & Aronson, N N, Acta biol med germ 36 (1977) 1917. 10. Dunn, W A, LaBadie, J H & Aronson, N N, J biol them 254 (1979) 4191. 11. Dunn, W A & Hubbard, A L, J cell bio198 (1984) 2148. Exp Cell Res 161 (1985)

2% 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

Berg et al.

El-Aaser, A A & Reid, E, Histochem j 1 (1%9) 417. Fiete, D, Brownell, M D & Baenziger, J U, J biol them 258 (1983) 817. Geuze, H J, Slot, J W, Strous, G J A M, Lodish, H & Schwartz, A L, Cell 32 (1983) 277. Geuze, H J, Slot, J W, Strous, G J A M & Schwartz, A L, Eur j cell biol 32 (1983) 38. Haiford, J & Ashwell, G, The glycoconjugates (ed M I Horowitz) vol. 4, p. 27. Academic Press, New York (1982). Harford, J, Wolkoff, A W, Ashwell, G & Klausner, R D, J cell biol % (1983) 1824. Higa, Y, Oshiro, S, Kino, K, Tsunoo, H & Nakajima, H, J biol them 256 (1981) 12322. Jones, A L, Hradek, G T, Homick, C, Renaud, G, Windier, E E T & Havel, R J, J lipid res 25 (1984) 1151. Kindberg, G, Ford, T, Blomhoff, R, Rickwood, D & Berg, T, Anal biochem 142 (1984) 455. Kolset, S 0, Tolleshaug, H & Berg, T, Exp cell res 122 (1979) 159. Moriarity, D M & Savage, C R, Arch biochem biophys 203 (1980) 506. Mullock, B M & Hinton, R H, Trends biochem sci 6 (1981) 188. Pertoft, H, W&rmeg&rd, B & Hook, M, Biochem j 174 (1978) 309. Pittman, R C, Carew, T E, Glass, C K, Green, S R, Taylor, C A & Attie, A D, Biochem j 212 (1983) 791. Rome, L H, Garvin, A J, Allietta, M M & Neufeld, E F, Cell 17 (1979) 143. Schwartz, A L, CRC crit rev biochem 16 (1984) 207. Seglen, P 0, Methods in cell biology (ed D M Prescott) vol. 13, p. 29. Academic Press, New York (1976). Steer, C J & Ashwell, G, J biol them 255 (1980) 3008. St Hilaire, R J & Jones, A L, Hepatology 2 (1982) 601. Storrie, B, Pool, R R, Sachdeva, M, Maurey, K M & Oliver, C, J cell bio198 (1984) 108. Tolleshaug, H & Berg, T, Acta biol med germ 36 (1977) 1753. - Hoppe-Seyler’s z physiol them 361 (1980) 1155. - Exp cell res 134 (1981) 207. Townsend, R R, Wall, D A, Hubbard, A & Lee, Y C, Proc natl acad sci US 81 (1984) 466. Wall, D A, Wilson, G & Hubbard, A L, Cell 21 (1980) 79. Warren, R & Doyle, D, J biol them 256 (1981) 1346. Weigel, P & Oka, J A, J biol them 259 (1984) 1150. Wolkoff, A W, Klausner, R D, Ashwell, G & Harford, J, J cell bio198 (1984) 375. Yamashiro, D J, Tycko, B, Fluss, J R & Maxfield, F R, Cell 37 (1984) 789.

Received March 24, 1985 Revised version received July 1, 1985

Exp Cell Res 161 (1985)

Printed in

Sweden