Autophagic sequestration of [14C]sucrose introduced into isolated rat hepatocytes by electrical and non-electrical methods

Experimental

Cell Research 160 (1985) 449-458

Autophagic Sequestration of [‘4C]Sucrose Introduced into Isolated Rat Hepatocytes by Electrical and Non-electrical Methods PAUL B. GORDON, HELGE TOLLESHAUG

and PER 0. SEGLEN

Department of Tissue Culture, Norsk Hydra’s Institute for Cancer Research, The Norwegian Radium Hospital, Montebello, 0310 Oslo 3, Norway

Isolated rat hepatocytes were found to become permeable to [‘4C]sucrose at 0°C under three different conditions: 1. Immediately following their liberation from the collagenase-perfused liver. 2. Following a short incubation under hypoxic conditions. 3. After electropermeabilisation. All three conditions were characterised by the formation of small protuberances (blebs) indicative of localised cell surface damage, and it is possible that the stretched plasma membrane of such blebs acted as a high-permeability region. Disappearance of blebs and restoration of normal plasma membrane impermeability could be achieved by a short (15 min) incubation at 37°C. It could be shown that [‘4C]sucrose introduced into rat hepatocytes by non-electrical means was autophagically sequestered at the same rate as [‘4C]sucrose introduced electrically. In both cases the sequestration was inhibited by the specific autophagy inhibitor 3methyladenine to a similar extent. The subcellular distribution of sequestered isotope in metrizamidekucrose density gradients was found to be independent of the conditions of its introduction into cells. @ 1985 Academic Press, Inc.

Autophagy is the term given to the cellular process(es) by which endogenous proteins destined for degradation are first sequestered into closed vacuoles, autophagosomes, by a sequestering membrane of largely unknown origin and properties [I, 21 and later delivered to the lysosomes for proteolysis. In order to investigate the early, pre-lysosomal steps of autophagy by non-morphological methods, several workers have utilised various ways of introducing radioactive proteins or inert probes into the cytosol of various cell types [Ml. In this laboratory we have developed a method which allows us to introduce [‘4Clsucrose into the cytosol of rat hepatocytes at high efficiency by reversibly permeabilising the plasma membrane using an electrical technique [7]. The time-dependent transfer of radioactivity from cytosol to sedimentable vacuoles, autophagosomes and lysosomes, is easily followed and serves as an indicator of overall autophagy, since sucrose is not metabolised by hepatocytes [8]. This technique was later optimised and used in a study of the effects of amino acids and 3methyladenine on the first step in the autophagic-lysosomal pathway of protein degradation [9]. Hepatocytes rendered permeable to small molecules by electrical discharges have an unusual surface morphology which is evident under examination by light Copyright @I 1985 by Academic Press, Inc. All rights of reproduction in any form reserved 0014-4827/85 W3.00

450 Gordon, Tolleshaug and Seglen

microscopy (x200-400), i.e. their plasma membrane is distinctly rough and small blebs, cytoplasm-filled protrusions, are apparent. Indeed, according to previous reports, electrical discharges disturb the lipid bilayer structure in the plasma membrane regionally, causing the formation of areas of increased permeability, “pores” [ 10, 1l] through which small molecules are able to pass and which cease to be permeable after a short incubation at 37°C [7]. In previous work it has been noted that the initial cell suspension obtained directly after liberating cells from the collagenase-perfused liver contains hepatocytes with surface abnormalities resembling the blebs described above, probably at loci where cell-to-cell junctions have been forcibly parted [12]. These blebs do not persist through the prepurification incubation of 30 min at 37°C routinely used in our cell preparation procedure. Likewise, hepatocytes subject to short periods (10-15 min) of incubation at 37°C under hypoxic conditions display such blebs [13]. In the present work we have investigated whether permeability to sucrose might be a general property of cells which display blebs. If so, it would be possible to measure autophagic sequestration of [14C]sucroseby non-electrical methods, and thus obtain an independent verification of the validity of our electropermeabilisation technique. MATERIALS AND METHODS Chemicals [‘C]Sucrose (555 mCi/mmol; 1 uCi/pl) was purchased from Amersham International plc, Buckinghamshire, UK; 3-methyladenine (6-amino-3-methylpuriine) was from Fhtka AG, Buchs, Switr.erland; and all other biochemicals from Sigma Chemical Co, St. Louis, MO.

Hepatocyte Preparation Hepatocytes were isolated from the liver of 18 h-starved male Wistar rats, 250-300 g, by the method of collagenase perfusion [12]. Following puri&ation, cells were suspended in suspension buffer [12], fortified with 20 mM pyruvate and 2 mh4 MgC12, and maintained at 0°C. In some cases, cells were liberated from the collagenase-perfused liver directly into fortified suspension buffer and maintained at 0°C without further purification before being loaded with [“Clsucrose (cf below).

[‘4C]Sucrose Loading of Cells and Incubation of Ceil Suspensions (a) EIecrricul permeubilisation. 10-12 ml of purified cell suspension at a cell concentration of 15&250 mg wet wt/ml was warmed to 37°C and electropermeabilised in a square-bottomed (2x2 cm) electrode chamber using five electrical discharges at 2 kV/cm and 1.2 pP [7]. The cell suspension was rapidly cooled to 0°C and after the addition of [“Clsucrose (5-10 uCi/ml) maintained at that temperature for 1 h with intermittent, gentle shaking. The suspension was then incubated for 30 min (for resealing) in two 10 cm 0 Petri dishes on a tilting platform (10 rpm). After three or four washes in 40 ml of ice-cold wash buffer [12], the cells were resuspended in fortified suspension buffer to a cell concentration of 75-125 mg wet wt/ml. (b) Preparation-induced permeubilisation. Following liberation of the cells from the enzymeperfused liver into ice-cold fortified suspension buffer, 10-12 ml of cell suspension was incubated at 0°C with [“Clsucrose; either 4-5 @i/ml for 2 !4 h or 15 uCi/ml for 1 h. (c) Hypoxic permeabilisation. 10-12 ml of purified cell suspension was warmed to 37T and incubated in a fiat-bottomed plastic beaker for 10 or 15 min under hypoxic conditions (i.e., without agitation), after which it was cooled to 0°C and incubated in the presence of [“~]SUC~OSC (65 fip Cell Res MO (1985)

Non-electrical permeabilisation of hepatocytes

45 1

@i/ml) for 1 h at 0°C. Following the incubation period at WC, cell suspensions in (b) and (c) were treated, resealed and washed as was the electropermeabilised cell suspension. During experimental incubations, cells were incubated as 0.4 ml aliquots (normally 0.3 ml cell suspension + 0.1 ml additives) in 15 ml glass centrifuge tubes shaking at a frequency of 215 rpm for up to 120 min at 37°C.

Density Gradient Centrifugation Me&amide/sucrose

gradients were prepared and centrifuged essentially as described before [14].

Scanning Electron Microscopy (SEM) Cells were fixed in a solution of 1% glutaraldehyde, 0.1 M cacodylate and 0.1 M sucrose for several hours at room temperature. Specimens were dehydrated in graded ethanol and critical point-dried in a Balzer critical-point dryer. Thereafter, the cells were gold-sputtered and scanning electron micrographs were taken using a Jeol-1200 EX microscope operating at 50 kV and an angle of 20“.

Miscellaneous Electrodisruption of the cells, isolation of the sedimentable cell fraction, determination of radioactivity, measurement of acid phosphatase and calculation of sequestration rates were all as previously described in detail [9].

RESULTS EfSects of Electropermeabilisation Surface Morphology

and Incubation at’37T on Hepatocytic

Cells freshly liberated from the collagenase-perfused liver are somewhat irregular in shape and display a number of plasma membrane lesions (presumably points at which undigested cell-to-cell junctions were tom apart during the liberation procedure), visible as blebs (fig. 1A). After a preincubation for 30 min at 37”C, routinely included in our preparation procedure [12], the blebs have largely disappeared and the surface of the (now rounded) cells can be seen to be covered with a smooth layer of microvilli (fig. 1B). Cells subjected to electrical discharges (fig. 1 C) display characteristic plasma membrane lesions visible as numerous small blebs on the cell surface. These electro-induced lesions largely disappeared during the 30-min resealing-incubation period following loading with [‘4C]sucrose, although the cell surface retained a somewhat rough appearance (fig. 1D). Kinetics of [‘4C]Sucrose Uptake in Cells Permeabilised Electrically and Non-electrically Purified cells permeabilised by our electrical method as well as freshly isolated cells rendered permeable by the cell preparation procedure remained permeable to, and continuously accumulated [‘4C]sucrose when maintained at 0°C (fig. 2). When the incubation temperature was raised from 0 to 37°C (fig. 2, at 60 min), in both cases the cells accumulated [‘4C]sucrose very much more rapidly for a period of approx. 5 min. Normal impermeability to sucrose was then restored, as 30-858340

Exp Cell Res 160 (198S)

452 Gordon, Tolleshaug and Seglen

I. SEM appearance of (A) freshly isolated; (B) preincubated; (C) electropermeabilised; (D) reseated hepatocytes. (A) Cells fixed immediately after collagenase-dissociation of the liver. Notice several large surface blebs. (B) Cells preincubated for 30 min at 37°C and purified by differential centrifugation. The cell surface is smoothly covered with microvilli. (C) Cells puritled as in (B), then electropermeabilised with five pulses. Notice numerous surface blebs. (D) Cells resealed by a 30-min incubation at 37°C following electropermeabilisation. The blebs have largely disappeared. x 1500.

Fig.

evidenced by there being no further accumulation of sucrose between 70 and 80 min, i.e., all permeable lesions on the cell surfaces were apparently restituted. From the results shown in fig. 2 it can be seen that hepatocytes permeabilised electrically (with five pulses) are, under the conditions used, about twice as permeable to [r4C]sucrose as are those freshly liberated from the enzymeperfused liver. Accordingly, to obtain a similar loading efficiency of sucrose into cells not subjected to electrical discharges, either the 0°C loading period or the amount of isotope should be doubled. Sequestration of Electroinjected [‘4C]Sucrose is Independent of Pulse Number We have shown in earlier work that the extent to which hepatocytes become permeable to [‘4C]sucrose is a direct function of the number of electrical discharges (pulses) to which they are subjected. This occurs only up to a critical point, past which the plasma membrane progressively loses its ability to reseal during the repair incubation [9]. The results displayed in fig. 3A show that [14C[sucrose is sequestered as effectively in cells given up to 15 electrical pulses as in those given only two pulses. Thus, the sequestration of electro-injected [14C]sucrose would appear to be unaffected by the electrical treatment per se. Our method of calculating sequestration as the difference in fraction sequesExp Cell Res 160 (1985)

Non-electrical

,o

20

30

lNCUBATlON

40

50 TIME

60 Iman)

70

60

permeabilisation

20

453

of hepatocytes

‘0 lNC"6.4TION

60

80

100

120

T,HE ,.,n,

Fig. 2. Kinetics of [r4C]sucrose uptake into cells permeabilised by electrical and non-electrical means. 0, Freshly prepared; 0, electropermeabilised (five pulses, 2 kV/cm, 1.2 uF) hepatocytes were incubated in the presence of [t4C]sucrose (-10 @i/ml) for up to 60 min at 0°C and thereafter for up to 20 min at 37°C. The cells were rapidly sampled by centrifugation (at O’C) through a 3 ml cushion of 8 % metrizamide + 8 % sucrose. Each value is the mean of three cell samples. Fig. 4. Sequestration of [14C]sucrose in hepatocytes loaded electrically or non-electrically: effect of 3-methyladenine. A, A, Freshly isolated or 0, 0, electropermeabilised hepatocytes were loaded with [‘4C]sucrose as in fig. 3 (5 pulses). The cells were incubated for the length of time indicated, either in the 0, A, absence or 0, A, presence of 10 mM 3-methyladenine. The net increase in sedimentable radioactivity in electrodisrupted cells was determined and expressed as a percentage of total cellular radioactivity at the time of measurement. Each value is the mean of three cell samples.

tered at two different time points [9] automatically corrects for losses of total radioactivity due to leakage from dead or dying cells. However, in order to maintain cellular functions, including protein synthesis and degradation (cf [9]), as normal as possible, we routinely limit the electrical treatment to a maximum of five pulses. Sequestration

of [ 14C]Sucrose in Cells Loaded Non-electrically

[ 14C]Sucrose was sequestered continuously throughout a 2 h incubation period (fig. 3 B), irrespective of whether it was introduced into cells immediately following their liberation from the collagenase-perfused liver (upper curve) or into purified hepatocytes subjected to hypoxia (lower curve). The rates of [t4C]sucrose sequestration (7.5 and 7.3 %/h, respectively, between 30 and 90 min in the experiment shown) compared well with the value (7.7%/h) for cells into which [‘4C]sucrose was introduced electrically (fig. 3A). The rate of sequestration gradually declined during the incubation period. As discussed in earlier work, this may be due partly to the continual fall in the amount of sucrose available for sequestration (as more and more is packaged into vesicles), partly to an increasing inhibition of autophagy by proteolysis-derived amino acids [9]. The overall amount of sucrose sequestered in hypoxically loaded cells was Exp Cell Res 160 (1985)

454 Gordon, Tolleshaug and Seglen

Fig. 3. Time course of [“Clsucrose sequestration in hepatocytes loaded with (A) [‘4C]sucrose electrically; or (B) non-electrically. (A) 10 ml portions of a hepatocyte suspension were electropermeabilised by 0, 2; 0, 5; A, 10; or A, 15 electrical pulses (all at 2 kV/cm, 1.2 uF), loaded with [“C]sucrose (5-10 @i/ml) for 1 h at 0°C resealed for 30 min at 37”C, washed 3x with ice-cold wash buffer, then incubated in fortified suspension buffer. (B) (Upper curw) 0, Freshly isolated hepatocytes loaded with [t4C]sucrose (10 uCi/ml) for 2% h at 0°C; (lower curve) hepatocytes permeabiised by incubation at 37°C under hypoxic conditions for q , 5; or n , 10 mitt, loaded with [“Clsucrose (10 uCi/ml) for 1 h at WC, resealed, washed and incubated as above. The net increase in sedimentable radioactivity in electrodistupted cells was determined and expressed as a percentage of the total cellular radioactivity at the time of measurement. Each value is the mean of three cell samples.

somewhat lower then that in the other cell preparations investigated (fig. 3B). This was probably due to the cells needing a longer period to recover from the hypoxic treatment than allowed by the 30 min resealing incubation. Furthermore, incubation under hypoxic conditions was clearly too harsh a procedure to be suitable for routine permeabilisation. This was evident from the observation that the loss of total cell-associated radioactivity during the experimental incubation period, which largely reflects cell death, was several-fold greater in hypoxiatreated cells than in the other types of permeabilised cells (data not shown). Effect of 344ethyladenine on the Sequestration of [14C]Sucrose Introduced Electrically and Non-electrically The rate of sequestration of [14C]sucrose was, within experimental limits, identical in cells into which it had been introduced by either electrical or nonelectrical (preparation-dependent) means (fig. 4, upper curves). Moreover, fig. 4 (lower curves) shows that 3-methyladenine, an established inhibitor of autophagic-lysosomal proteolysis which exerts its effect at the sequestration step [15], inhibits the sequestration of [‘4C]sucrose to an equal extent in both cases (-65 % inhibition between 30 and !I0 min). During the incubation period depicted in fig. 4, cells treated in either manner lost total cell-associated radioactivity at the same low rate (6%/h between 30 and 90 min), indicating that the cells were in equally good condition whether they had been subjected to the electrical discharge treatment or not. Erp Cell Res 160 (1985)

Non-electrical permeabilisation of hepatocytes

455

5. Density gradient distribution of sequestered [“Clsucrose in hepatocytes loaded (A) electrically or (I?) non-electrically. Cell corpse homogenates were prepared either from purified electropermeabilised cells loaded with [“Clsucrose for 1 h at 0°C (A) or from freshly isolated cells loaded with [“Clsucrose for 2 h at WC (B). Both cell types had been incubated for 2 h at 37°C after loading/resealing. The homogenates were centrifuged on metrizamidelsucrose gradients, and the fractions analysed for 0, [“Clsucrose; 0, acid phosphatase; protein.

Fig.

A,

Distribution of Sedimentable [‘4ClSucrose in Density Gradients During a 2 h incubation (at 37°C) subsequent to [‘4C]sucrose loading/resealing about l&15% of the total radioactivity associated with the cells was transferred from a non-sedimentable compartment (cytosol) to sedimentable structures. Following the 2 h incubation, samples of cells were washed in 10% w/v sucrose, electrodisrupted, separated from soluble cellular components, homogenised, and fractionated in isotonic metrizamide density gradients as described in detail elsewhere [8, 141. [14C]Sucrose was distributed between two major peaks; a sharp, dense peak at 1.15 g/ml and a broader, light (buoyant) peak around 1.08-1.10 g/ml (fig. 5). The dense radioactivity peak coincided with the major protein peak which represents, mainly, mitochondria [8]. Uptake of radioactive sucrose into mitochondria has been discussed in earlier work [2, 81. The light peak coincided with the major peak of acid phosphatase, and is thought to represent label in lysosomes and autophagosomes [8]. In fig. 5A, B is shown the distribution of sedimentable [‘4C]sucrose in isotonic metrizamide density gradients for cells loaded by electrical and non-electrical means respectively. The distribution of [‘4C]sucrose in these gradients was clearly identical, irrespective of whether it was introduced into cells electrically or non-electrically. Exp Cell Res MO (1985)

456

Gordon,

Tolleshaug

and Seglen

DISCUSSION In a recent report we have described in detail how electroinjected [‘4C]sucrose can be used as a probe in the study of the early step in the autophagiclysosomal pathway of degradation, i.e., the sequestering step, and how this step is sensitive to inhibitors of proteolysis such as 3-methyladenine and amino acids 191.Electropermeabilisation of the plasma membrane is both a convenient and efficient method of introducing [14C]sucrose into cells, although alternative methods have been described by others [5, 161. The rate of [14C]sucrose sequestration is almost three times higher than the estimated rate of autophagic-lysosomal protein degradation under similar conditions [9]. Part of this discrepancy can be explained by mitochondrial [‘4C]sucrose uptake, which accounts for about one-third of the total, and which is unrelated to protein degradation [8]. However, the remaining, presumably autophagic, sequestration is still twice as high as expected. Adsorption of radioactivity to sedimentable structures does not occur to any detectable extent [9], and direct penetration of sugar into lysosomes seems extremely unlikely in view of the very strong inhibition (essentially a complete shut-down of non-mitochondrial sequestration) by the specific autophagy inhibitor, 3-methyladenine. Although autophagic sequestration appears to be a largely non-selective process according to electron microscopic studies [17, 181,we cannot presently exclude the possibility of either some recycling of protein (but not of [r4C]sucrose, cf ref. [9]) or of a selective sequestration of fluid vs protein, resulting in a rate of “fluid-phase autophagy” which is higher than the overall autophagic rate. The observation that electropermeabilised hepatocytes display surface lesions which resemble the surface blebs visible on freshly isolated hepatocytes led us to the finding that sucrose could in fact be introduced into hepatocytes without the application of any electrical disturbance. Moreover, this observation strengthens the hypothesis that small molecules are able to diffuse into cells through the plasma membrane only at points where its structure is definitely atypical [lo]. The identification of blebs as the functional “pores” postulated by the electromechanical permeabilisation hypothesis [lo] transcends that hypothesis in the sense that the electrical treatment does not seem to affect plasma membrane permeability directly, but rather provides the mechanical basis (non-specific surface damage) for the protrusion of cytoplasmic processes with an abnormally permeable (probably stretched) membrane. It is of interest that isolated surface blebs, which can be prepared by various methods [19-211, exhibit the expected permeability towards sucrose while keeping cytosolic enzymes in containment 1211. As the results shown in the present report indicate, the behaviour of [‘4C]sucrose introduced into hepatocytes is quite independent of the method by which it is introduced. Therefore, the possibility of the high rate of [14C]sucrose sequestration being an artefact associated with electropermeabilisation can be excluded, Exp Cell Res 160 (1985)

Non-electrical permeabilisation of hepatocytes

457

a conclusion strongly supported by the fact that the rate of sequestration is also independent of the number of electrical discharges employed. While [‘4C]sucrose can obviously be introduced into hepatocytes by nonelectrical means, we routinely prefer to use our electropermeabilisation procedure (five pulses only) since it enhances the loading efficiency and standardises the experimental conditions. Although the surface morphology of electropermeabilised/resealed hepatocytes is slightly different from that of cells which have been rendered permeable only as a result of the usual cell preparation procedure, we are confident that the identity in sequestration rates, inhibitor sensitivities, subcellular distribution of [i4C]sucrose, etc. show that the two types of preparation are equivalent in terms of normal autophagic-lysosomal function. If electrical treatment for some reason is considered undesirable, the preparation-induced permeabilisation of course offers a fully acceptable methodological alternative. The absence of a need for specialised electrical equipment makes the preparationdependent method particularly attractive. However, it should be emphasised that in addition to the electropermeabilisation step, the electrical apparatus is indispensable in performing a later step in the experimental procedure, i.e., the electrodisruption of the plasma membrane needed for separation of soluble radioactivity from sequestered radioactivity [7]. Some important implications of the preparation-induced permeabilisation ought to be pointed out. Our routine procedure for cell preparation [ 121includes a 30-min preincubation at 37°C during which the cells are allowed to reseal. If this preincubation period is omitted or significantly shortened, it is conceivable that some permeability may persist, resulting, e.g., in cell swelling during low-temperature storage [22] or an inability to retain ions and metabolites intracellularly upon initiation of experiments [23]. With the preincubation included, the final cell preparation shows optimal functional ability from the very start of experiments 112, 241, and the cells can be preserved in excellent condition (in the presence of pyruvate) for at least 24 h at 0°C (our unpublished observation). Experimental results can also be compromised by hyperpermeability associated with surface bleb formation during hepatocyte incubation, an almost inevitable consequence of mechanical shaking procedures 1121,as well as of treatment with cytochalasin and some other drugs [ 18,25,26]. For example, the observation that hepatocytic accumulation of [‘4C]sucrose is three times faster than that of the high-molecular weight compound [‘*‘I]PVP [27] may tentatively be ascribed to some of the sucrose entering’by direct permeation through blebs rather than by pinocytosis. Blebbing-associated hyperpermeability may be relevant to cell types other than hepatocytes and should probably be taken into consideration in experiments where small-molecular weight compounds are being used for transport studies or as membrane-impermeant probes. This project is generously supported by The Norwegian Cancer Society. We wish to thank Barbara Schiiler for preparing the scanning electron micrographs. Exp Cell Res 160 (1985)

458

Gordon,

Tolleshaug

and Seglen

REFERENCES 1. Pfeifer, LJ, J cell biol 78 (1978) 152. 2. Seglen, P 0, Lysosomes. Their role in protein breakdown (ed H Glaumann & F J Ballard) p. 30. Academic Press, New York (1985). 3. Backer, J M, Bourret, L & Dice, J F, Proc natl acad sci US 80 (1983) 2166. 4. Freikopf-Cassel, A & Kulka, R G, FEBS lett 128 (1981) 63. 5. Hendil, K B, Exp cell res 135 (1981) 157. 6. Okada, C Y & Rechsteiner, M, Cell 29 (1982) 33. 7. Gordon, P B & Seglen, P 0, Exp cell res 142 (1982) 1. 8. Tolleshaug, H, Gordon, P B, Solheim, A E & Seglen, P 0, Biochem biophys res commun 119 (1984) 955. 9. Seglen, P 0 & Gordon, P B, J cell bio199 (1984) 435. 10. Zimmermann, U, Scheurich, P, Pilwat, G & Benz, R, Angew them 93 (1981) 332. 11. Riemann, F, Zimmermann, U & Pilwat, G, Biochim biophys acta 394 (1975) 449. 12. Seglen, P 0, Meth cell biol 13 (1976) 29. 13. Solheim, A E & Seglen, P 0, Biochem j 210 (1983) 929. 14. - Biochim biophys acta 763 (1983) 254. 15. Seglen, P 0 & Gordon, P B, Proc natl acad sci US 79 (1982) 1889. 16. Dean, R T, Acta biol med germ 36 (1977) 1815. 17. Glaumann, H, Ericsson, J L E & Marzella, L, Int rev cytol 73 (1981) 149. 18. Pfeifer, U, Verh dtsch ges path01 60 (1976) 28. 19. Seglen, P 0, Biological separations in iodinated density-gradient media (ed D Rickwood) p. 107. Information Retrieval, London (1976). 20. Nagelkerke, J F, Barto, K P & van Berkel, T J, Exp cell res 138 (1982) 183. 21. Berry, M N, Gannon, B J, Grivell, A R, Henderson, D W, Henly, D C, Norton, S L & Wallace, P G, Isolation, characterization and use of hepatocytes (ed R A Harris & N W Cornell) p. 277. Elsevier, New York (1983). 22. Krebs, H A, Cornell, N W, Lund P & Hems, R, Regulation of hepatic metabolism (ed F Lundquist & N Tygstrup) p. 726. Munksgaard, Copenhagen (1974). 23. Ayala, E C Canonico, P G, Proc sot exp biol med 149 (1975) 1019. 24. Seglen, P 0, Biochim biophys acta 338 (1974) 317. 25. Weiss, R, Sterz, I, Frimmer, M & Kroker, R, Beitr path01 anat allg path01 150 (1973) 345. 26. Faulstich, H, ‘Bischmann, H & Mayer, D, Exp cell res 144 (1983) 73. 27. Ose, L, Ose, T, Reinertsen, R & Berg, T, Exp cell res 126 (1980) 109. Received March 13, 1985 Revised version received May 22, 1985

hip Cell Res 160 (1985)

tinted

in Swden