THE ANATOMICAL RECORD 227:321-333 (1990)
Glycogen Metabolism in Cultured Chick Hepatocytes: A Morphological Study JOAN LEE PARKES, EMMA LOU CARDELL, GERD GRIENZNGER, AND ROBERT R. CARDELL Department of Environmental Medicine, New York University Medical Center, New York, New York 10016 (J.L.P.);Department of Anatomy and Cell Biology, University of Cincinnati, College of Medicine, Cincinnati, Ohio 45267 (E.L.C.,R.R.C.); The Lindsley F. Kimball Research Institute of the New York Blood Center, New York, New York 10021 (G.G.)
ABSTRACT
Ultrastructural and autoradiographic observations of cultured chick hepatocytes under the following conditions are described: Induction of glycogen synthesis with glucose alone and glucose plus insulin, and glucagon-induced glycogen breakdown. Profiles of hepatocytes cultured in medium containing 10 mM glucose showed typical cellular organelles and occasionally a few glycogen granules. After incubation of hepatocytes with 3H-glucose, silver grains were found over these sparse glycogen granules, indicating a low level of glycogen synthesis by a few cells. After addition of 75 mM glucose for 1 h r about 3% of the profiles of cells showed glycogen, and by 24 h r half of the hepatocytes had glycogen. Addition of insulin plus glucose induced glycogen synthesis in 82% of the cells after 6 hr, and by 24 h r almost every cellular profile showed glyocgen particles. Morphologically, glycogen accumulation was similar whether the cells were stimulated by high glucose or by glucose plus insulin: glycogen granules appeared in restricted regions of the cytoplasm, which were rich in smooth endoplasmic reticulum (SER), and peroxisomes were found close to the newly deposited glycogen particles. At maximum glycogen accumulation the association of SER and peroxisomes with glycogen was less obvious. Glycogenolysis induced by incubation of glycogen-rich hepatocytes with glucagon resulted in proliferation of SER in the glycogen regions of the cells. These observations are compatible with the concept of regions in the hepatocyte cytoplasm specialized for glycogen metabolism. Possible roles for SER and peroxisomes found near glycogen particles and other organelles in hepatic glycogen metabolism are discussed.
Recently the well-established embryonic chicken hepatocyte culture system (Grieninger and Granick, 1978; Grieninger, 1983) was used to study the regulation of glycogen metabolism under precisely defined conditions. By culturing hepatocytes in a chemically defined medium (without serum) it was possible to evaluate the role of glucose and insulin in hepatic glycogenesis, to study the effects of glucagon on glycogen breakdown, and to observe the influence of insulin-like growth factors (Parkes and Grieninger, 1985;Parkes et al., 1986). These studies showed that addition of high concentrations of glucose to the culture medium caused rapid glycogen deposition, but the levels obtained never equaled maximal in vivo fed hepatic glycogen levels. It was concluded that glycogen synthesis in response to glucose is driven primarily by mass action (i.e., the effect of increasing substrate concentration). In contrast, insulin added to the culture medium with glucose causes the restoration of physiological maximal levels of glycogen. Thus insulin was proposed as the major regulator of glycogen synthesis in the hepatocyte. No major differences were observed between insulin and insulin-like growth factors with respect to glycogen deposition. The addition of glucagon to the 0 1990 WILEY-LISS,
INC.
culture medium caused a rapid breakdown of glycogen stores in these cells. In this paper we report ultrastructural and autoradiographic observations on cultured chick hepatocytes after stimulation with glucose alone and with glucose plus insulin to induce glycogen synthesis under highly regulated conditions. In addition, we show the ultrastructural changes in hepatocytes under conditions of glucagon-induced glycogen breakdown. In general, these observations confirm earlier reports (Jerome and Cardell, 1983; Michaels et al., 1984; Cardell et al., 1985; Striffler et al., 1981) on the morphology of hepatic glycogen synthesis and breakdown in vivo. Another observation described in the present work is the close proximity of peroxisomes to regions of newly deposited glycogen particles. We discuss a possible role for this organelle in glycogen metabolism.
Received June 16, 1989; accepted November 2, 1989.
322
J.L. PARKES ET AL.
MATERIAL AND METHODS Cell Culture
Primary monolayer cultures of liver cells from 16 day chicken embryos were prepared as described elsewhere (Parkes and Grieninger, 1985). Briefly, suspensions of hepatocytes were obtained by perfusion of the liver with buffer, mechanical disruption, and treatment with purified dissociating enzymes. Hepatocytes were washed and plated in modified Ham's F12 medium containing 25 mM glucose but no hormones or serum supplement. The modified medium contained no lactate or pyruvate and no other sugar but glucose. Hepatocyte suspensions were plated in 35- or 55-mmdiameter untreated Falcon plastic culture dishes. Culture medium was replaced with a n equal volume of fresh medium every 24 h r unless otherwise indicated. Monolayers contained approximately 2 x lo6 or 5 x 106 cells in 35 or 55 mm dishes, respectively; 3 x lo6 cells represent about 1 mg of protein (Grieninger and Granick, 1978). These nondividing cultures suffer no loss of viability for several days in the absence of hormone (Liang and Grieninger, 1981; Plant et al., 1983). Insulin, which is not mitogenic in these cultures (Liang and Grieninger, 1981; Plant et al., 1983), was introduced only after removal of unattached cells during the first medium change to avoid hormonal effects on plating efficiency. Medium varied by the addition of 10,25, or 75 mM glucose with and without 35 nM insulin. 3H-glucose was added to the medium of the 35 mm dishes for autoradiographic localization of label incorporated into glycogen. Additional details of the various experimental culture conditions are given below. Samples of cells were obtained at various times for biochemical analysis of glycogen content, light and electron microscopic evaluation, andlor autoradiography. Microscopy
Monolayers were fixed briefly (5 min) in a 2% glutaraldehyde-2% paraformaldehyde mixture in 0.1 M sodium cacodylate buffer (pH 7.3) (Cardell et al., 1985). Cells were gently scraped from the culture dishes as sheets and/or strips of cells in a small amount of fixative and transferred in aliquots which yielded a pellet 0.5-1 mm thick when centrifuged for 2 min in 0.5 ml microfuge tubes containing fresh fixative. Pellets were removed from the tubes and placed into vials of fresh fixative for a minimum of 2 h r at room temperature, washed in buffer for 1 hr, and postfixed for 2 h r in cacodylate-buffered 1% osmium tetroxide. Following dehydration in a graded series of ethanol and infiltration with propylene oxide, the samples were embedded in Epon 812 (Cardell et al., 1985). Thick (0.5-1 Fm) and thin (silver) sections were cut on a Sorvall MT-2B ultramicrotome. Thick sections were adhered to glass slides by gentle heat. Sections were evaluated immediately after staining with toluidine blue to confirm that the section was from the portion of pellet1 that 'Semithin longitudinal sections of a n entire pellet show that the pellet consists almost entirely of sheets of cells throughout, except for the most distal (widest) portion, which is a thin layer of single cells, broken cells, and debris. An optimal region of the pellet, therefore, is composed of sheets of strips of cells that have retained their cell to cell relationships. Centrifugation in the microfuge does not distort or disassociate the cells; the sheets of cells are not packed tightly against each other in the pellet.
consists of sheets of cells. Slides with unstained sections of the optimal region of the pellets were coated with Kodak NTB-2 nuclear tracking emulsion for light microscopy autoradiography. Thin sections for routine transmission electron microscopy (EM) were picked up on copper grids and stained with uranyl and lead salts; thin sections for autoradiography were transferred by wire loop and air-dried onto nitrocellulose-coated glass slides. Prior to coating with Ilford L-4 emulsion diluted to yield a monolayer of silver halide crystals, thin sections on slides were stained with uranyl acetate and lead citrate and coated with carbon. After various exposure times, autoradiograms were developed (Michaels et al., 1984), the sections were floated of" the slides, picked up on grids, and evaluated. Reagents
Bovine pancreas insulin (crystalline, Sigma) and porcine glucagon (crystalline, NOVO,Denmark) were prepared and stored in 0.01 N HC1. 3H-D-glucose (20 Cilm mole) was obtained from ICN Pharmaceuticals. Glycogen Determination
Culture dishes were treated as described, with minor modification (Parkes and Grieninger, 19851, for determination of glycogen. Monolayers were washed twice and immediately frozen in a n ethanolldry-ice bath. The cells were scraped from the dishes in water with a Teflon policeman. Glycogen was measured by digestion for 10 min at 55°C with amyloglucosidase (Sigma), followed by centrifugation, and subsequent colorimetric determination of glucose by the glucose-oxidase procedure (Roehrig and Allred, 1974) using a kit from Sigma. Glycogen Synthesis
Three hours after plating, cells were washed once and incubated for 16 h r in medium containing 10 mM glucose to deplete glycogen stores [Note: the blood glucose level of a starved chicken is approximately 15 mM (see Parkes and Grieninger, 198511. Hepatocytes were then given fresh medium containing 25 mM HEPES (pH 7.4) and either 10 or 75 mM glucose with and without insulin (35 nM). Metabolic reactions were stopped at the appropriate times by removal of medium and immediate addition of fixative for EM studies or immersion in a ethanolldry-ice bath for biochemical determination of glycogen content. For the autoradiographic experiments, confluent cells in 35 mm dishes were given medium containing 10 or 75 mM glucose with and without insulin, along with approximately 1.25 mCi of 3H-glucose. The cells were then harvested at various times (5, 15, and 30 min, 2,12 and 24 hr). Hepatocytes that were harvested after 12 or 24 h r received 3H-glucose 2 h r prior to harvesting. Glycogen Degradation
Three hours after plating, glycogen stores were enhanced by incubating cells for 48 h r in medium containing 75 mM glucose plus 35 nM insulin with one change into fresh medium after 24 hr. At the end of this pretreatment period, culture dishes were washed once and incubated in medium containing 10 mM glucose. Thirty minutes later, at time 0, the dishes were given
GLYCOGEN METABOLISM I N CHICK HEPATOCYTES
" ' J 4 40
323
[GLYCOGEN SYNTHESIS]
t
0lOmM
glucose
75mM glucose M 7 5 m M glucose + 35nM insulin
. 1
0
1
2
5
10
15
Fig 1 . Time course of glucose and insulin-induced glycogen accumulation. Three hours after plating, cells were washed once and incubated for 16 hr in medium containing 10 mM glucose. Hepatocytes were then given fresh medium (time 0 ) containing either 10 mM glu-
20
30
60
6
24
65
cose or 75 mM glucose with and without 35 nM insulin. Glycogen content was determined at the times indicated. Cells incubated for 65 hr were given fresh experimental medium every 24 hr.
fresh medium containing HEPES (25 mM, pH 7.3) plus 10 mM glucose and 350 nM glucagon, incubated for various lengths of time, and prepared for glycogen determinations and morphological studies as described above. OBSERVATIONS AND RESULTS Glycogen Synthesis
Figure 1 shows the response of the cultured chick hepatocytes to medium containing 10 mM glucose, 75 mM glucose, or 75 mM glucose plus 35 nM insulin for several minutes through 65 hr. The presence of 10 mM glucose resulted in biochemically undetectable or very low levels of glycogen over the entire time course, although the addition of 75 mM glucose stimulated measurable glycogen synthesis by 5 min with additional accumulations by 1h r in culture. Glycogen accumulation equal to physiologically maximal levels was obtained by long-term culture in medium containing 75 mM glucose plus 35 nM insulin. Although the effect of added glucose on glycogenesis is seen within minutes, the effect of insulin on glycogenesis is apparent after approximately 6 h r of incubation (Fig. 1; and Parkes and Grieninger, 1985). As noted above, hepatocytes maintained in medium with 10 mM glucose had very low levels of glycogen (Fig. 1).Morphologically, these glycogen-depleted cells had large lipid droplets, typical profiles of Golgi apparatus, elongated mitochondria, little or no smooth endoplasmic reticulum (SER), isolated cisternae of rough endoplasmic reticulum (RER), and a few peroxisomes scattered throughout the cytosome and near the ends of RER cisternae (Fig. 3). Most profiles of cells were devoid of glycogen granules, but a few cells (1% of the cell profiles. SER was present in the cytoplasmic regions displaying glycogen particles. There also appeared to be a close association of peroxisomes with these glycogen regions. Otherwise, the basic morphology of these hepatocytes after short incubation in higher levels of glucose with and without insulin was similar to that of the glycogen-depleted cells (Fig. 3). Five minutes after addition of 75 mM glucose (not shown) or 75 mM glucose plus 35 mM insulin (Fig. 5B), approximately 2% of the hepatocyte profiles contained glycogen. Again, there was a close association of SER with CY and p particles of glycogen; peroxisomes were near these regions. After 1 h r incubation with 75 mM glucose plus 35 nM insulin, the percent profiles containing glycogen was almost twice that of hepatocytes cultured for the same time with 75 mM glucose and no insulin (about 6% and 3%, respectively). After six hours of culture in the presence of 75 mM glucose plus 35 nM insulin, approximately 82% of the cell profiles contained glycogen, sometimes in large compact masses (Fig. 6A), whereas only 12% of the hepatocyte profiles contained glycogen when cultured in glucose alone (not shown). The morphology of the glycogencontaining cells was similar in both groups and showed SER near the glycogen particles (Fig. 6A). By 24 h r of culture in 75 mM glucose and 35 nM insulin, approximately 99% of the cell profiles showed glycogen in the cytosome, whereas 53% of the hepatocyte profiles contained glycogen when cultured with glucose alone. By 65 h r of culture with insulin, cells in which glycogen practically filled the cytosome (Fig. 6B) were common. At this stage of maximum deposition, elements of SER, though not abundant, were found near the glycogen masses. This observation is consistent with in vivo studies (Cardell, 1977; Striffler, et al., 19811, which showed relatively little SER in glycogen-filled hepatocytes compared to more abundant SER in hepatocytes actively metabolizing glycogen. Profiles of hepatocytes maintained for about 3 days in 10 mM glucose showed little or no glycogen (not shown), consistent with the biochemical data (Fig. 1). To determine if the glycogen observed in the hepatocytes was newly synthesized, cells were incubated for various times in medium containing 10 or 75 mM 51ucose, with and without insulin, plus -1.25 mCi of Hglucose. Labeled glucose, a precursor to glycogen, was then localized in the cells by autoradiography. Newly synthesized glycogen was identified at all times studied, and it is significant to note that hepatocytes cul‘To make estimates of cells containing glycogen, we randomly evaluated 500 profiles of cells in a single thin section with the EM and determined for each experimental group the percentage of profiles withiwithout glycogen in that plane of section. Because these hepatocytes grown to confluency have a fairly regular shape, any portion of an hepatocyte that included mitochondria and other organelles, but not necessarily the nucleus, was counted. Every profile within contiguous grid squares was counted until a total of 500 were recorded. Micropodia and other small cellular projections did not contain mitochondria, etc., and were not counted. These uniform-size hepatocytes are not amoeboid or stellate in shape, so profiles of cells closely correspond to actual cell numbers. However, it is obvious that changes in cellular size and the random distribution of glycogen within hepatocytes cause predictable errors in this estimate.
327
tured for a short time with either 10 mM glucose (Fig. 5A) or 75 mM glucose (Fig. 7B, C) showed silver grains over small glycogen regions, indicating incorporation of glucose into glycogen. The region shown in Figure 7A corresponds to a n “early SERGE,” described by Cardell et al. (1985) as a restricted area of cytoplasm in which smooth endoplasmic reticulum, glycogen (or precursors), and related enzymes are localized. Very little glycogen is visible here, although the presence of silver grains indicates that 3H-glucose has been incorporated into newly synthesized glycogen. Since glucose is utilized by hepatocytes for activities other than glycogen synthesis (Neutra and Leblond, 1966a,b), it is not surprising that several organelles of the cell, primarily the Golgi apparatus, sometimes were labeled (not shown). Glycogen Degradation
Fig. 8 shows the extensive deposits of glycogen that accumulate in the cells incubated in 75 mM glucose and 35 nM insulin for 48 hr. The large glycogen masses are compact with limited associated SER. To study glycogen degradation, such cells were given fresh insulinfree medium containing 10 mM glucose plus 350 nM glucagon and incubated for various times up to 3 hr. Figure 2 shows that glucagon causes a rapid breakdown of glycogen, which is nearly complete by 180 min (see also Parkes and Grieninger, 1985). The embryonic chick hepatocyte culture system is a n excellent model for the morphological study of glycogen depletion under highly regulated conditions. Ten minutes after addition of glucagon to the cultures, most cells showed SER infiltrations within the glycogen masses (Fig. 9A). However, other hepatocytes failed to show this early response to hormone. After longer exposure to glucagon additional cells showed decreased amounts of glycogen, and by 60 min many hepatocytes were practically depleted of glycogen (Fig. 9B). SER was observed near the remaining glycogen particles during late stages of glycogenolysis. By 180 min, virtually all cells were depleted of glycogen and SER was sparse in the cytosome. DISCUSSION
Primary cultures of chicken embryo hepatocytes were used to study morphological aspects of hepatocellular glycogen metabolism. This is a n excellent system for such studies because the cultured hepatocytes do not require hormones or other macromolecular media supplements for their viability and expression of hepatic functions (Parkes and Grieninger, 1985). Thus it was possible to study morphological aspects of hepatic glycogen metabolism under precisely controlled conditions. Hepatocytes obtained from 16-day-old chicken embryos and cultured in modified Ham’s F12 medium with low levels of glucose are healthy and do not lose viability for several days in the absence of hormones (Parkes and Grieninger, 1985), even though under these conditions the cells contain almost no biochemically detectable glycogen. Addition of increasing concentrations of glucose to glycogen-depleted cells causes the hepatocytes to synthesize glycogen; however, insulin is required to restore fed in vivo levels of glycogen in the cultured cells (Parkes and Grieninger, 1985).
328
J.L. PARKES ET AL
Fig. 6. A Sizeable glycogen accumulations are found in hepatocytes after 6 hr culture with 75 mM glucose and 35 nM insulin. Arrowheads, SER arrows, glycogen. B: Long term culture (65 hr) in medium
supplemented with glucose and insulin results in glycogen deposits, which occupy most of the cytosome. Arrowheads, SER arrows, glycogen; P, peroxisome.
Culture of hepatocytes in hormone-free medium supplemented with 10 mM glucose resulted in very little glycogen synthesis over 65 hr. Five minutes after ad-
dition of 75 mM glucose to the culture medium, measurable glycogen begins to accumulate in the hepatocytes, and by 1 hr significant quantities of glycogen are
GLYCOGEN METABOLISM IN CHICK HEPATOCYTES
Fig. 7. A After 30 min incubation with 3H-glucose, hepatocytes that have been maintained in medium containing 10 mM glucose show small glycogen deposits and associated ribbon-like silver grains indicating incorporation of the isotope into the newly formed glycogen. P, peroxisomes; arrows, glycogen; arrowheads, SER. B,C: Cells that
329
have been incubated with 75 mM glucose and labeled glucose for 30 min show distinct glycogen patches. Silver grains indicate incorporation of the isotope into glycogen synthesized after addition of the isotope. P, peroxisomes; arrows, glycogen; arrowheads, SER.
330
J.L. PARKES ET AL.
Fig. 8. Hepatocytes can be induced to deposit large quantities of glycogen by 48 h r incubation with 75 mM glucose plus 35 nM insulin added to the medium. The cytosome is practically filled with glycogen and such cultures served as baselines for studies of glucagon action. G1, glycogen.
GLYCOGEN METABOLISM IN CHICK HEPATOCYTES
Fig. 9. A Ten minutes after addition of 350 nM glucagon to fresh insulin-free medium, the glycogen areas in some hepatocytes contain many tubules of smooth endoplasmic reticulum (arrowheads). Arrows, glycogen; P, peroxisome. B: After 60 min incubation with glu-
331
cagon, cells almost devoid of glycogen are found. Notice that SER (arrowheads) is close to the remaining glycogen particles (arrows). P, peroxisomes.
332
J.L. PARKES ET AL.
synthesized. Addition of 35 nM insulin to the culture system caused a sustained synthesis of glycogen, which reached levels equal to the fed in vivo situation by 65 hr. The ultrastructure of hepatocytes cultured i n medium containing 10 mM glucose showed the usual cellular organelles and, in a n occasional cell, a few glycogen granules. The location of silver grains over these sparse glycogen granules in the autoradiographic preparations indicated that a very low rate of glycogen synthesis occurred during culture with 10 mM glucose; therefore, these infrequently observed glycogen granules were not residual glycogen in the cell. After addition of 75 mM glucose to the culture medium a small percentage of the hepatocytes responded rapidly by synthesizing glycogen. This newly synthesized glycogen appeared in small (
Glycogen Metabolism in Cultured Chick Hepatocytes: A Morphological Study JOAN LEE PARKES, EMMA LOU CARDELL, GERD GRIENZNGER, AND ROBERT R. CARDELL Department of Environmental Medicine, New York University Medical Center, New York, New York 10016 (J.L.P.);Department of Anatomy and Cell Biology, University of Cincinnati, College of Medicine, Cincinnati, Ohio 45267 (E.L.C.,R.R.C.); The Lindsley F. Kimball Research Institute of the New York Blood Center, New York, New York 10021 (G.G.)
ABSTRACT
Ultrastructural and autoradiographic observations of cultured chick hepatocytes under the following conditions are described: Induction of glycogen synthesis with glucose alone and glucose plus insulin, and glucagon-induced glycogen breakdown. Profiles of hepatocytes cultured in medium containing 10 mM glucose showed typical cellular organelles and occasionally a few glycogen granules. After incubation of hepatocytes with 3H-glucose, silver grains were found over these sparse glycogen granules, indicating a low level of glycogen synthesis by a few cells. After addition of 75 mM glucose for 1 h r about 3% of the profiles of cells showed glycogen, and by 24 h r half of the hepatocytes had glycogen. Addition of insulin plus glucose induced glycogen synthesis in 82% of the cells after 6 hr, and by 24 h r almost every cellular profile showed glyocgen particles. Morphologically, glycogen accumulation was similar whether the cells were stimulated by high glucose or by glucose plus insulin: glycogen granules appeared in restricted regions of the cytoplasm, which were rich in smooth endoplasmic reticulum (SER), and peroxisomes were found close to the newly deposited glycogen particles. At maximum glycogen accumulation the association of SER and peroxisomes with glycogen was less obvious. Glycogenolysis induced by incubation of glycogen-rich hepatocytes with glucagon resulted in proliferation of SER in the glycogen regions of the cells. These observations are compatible with the concept of regions in the hepatocyte cytoplasm specialized for glycogen metabolism. Possible roles for SER and peroxisomes found near glycogen particles and other organelles in hepatic glycogen metabolism are discussed.
Recently the well-established embryonic chicken hepatocyte culture system (Grieninger and Granick, 1978; Grieninger, 1983) was used to study the regulation of glycogen metabolism under precisely defined conditions. By culturing hepatocytes in a chemically defined medium (without serum) it was possible to evaluate the role of glucose and insulin in hepatic glycogenesis, to study the effects of glucagon on glycogen breakdown, and to observe the influence of insulin-like growth factors (Parkes and Grieninger, 1985;Parkes et al., 1986). These studies showed that addition of high concentrations of glucose to the culture medium caused rapid glycogen deposition, but the levels obtained never equaled maximal in vivo fed hepatic glycogen levels. It was concluded that glycogen synthesis in response to glucose is driven primarily by mass action (i.e., the effect of increasing substrate concentration). In contrast, insulin added to the culture medium with glucose causes the restoration of physiological maximal levels of glycogen. Thus insulin was proposed as the major regulator of glycogen synthesis in the hepatocyte. No major differences were observed between insulin and insulin-like growth factors with respect to glycogen deposition. The addition of glucagon to the 0 1990 WILEY-LISS,
INC.
culture medium caused a rapid breakdown of glycogen stores in these cells. In this paper we report ultrastructural and autoradiographic observations on cultured chick hepatocytes after stimulation with glucose alone and with glucose plus insulin to induce glycogen synthesis under highly regulated conditions. In addition, we show the ultrastructural changes in hepatocytes under conditions of glucagon-induced glycogen breakdown. In general, these observations confirm earlier reports (Jerome and Cardell, 1983; Michaels et al., 1984; Cardell et al., 1985; Striffler et al., 1981) on the morphology of hepatic glycogen synthesis and breakdown in vivo. Another observation described in the present work is the close proximity of peroxisomes to regions of newly deposited glycogen particles. We discuss a possible role for this organelle in glycogen metabolism.
Received June 16, 1989; accepted November 2, 1989.
322
J.L. PARKES ET AL.
MATERIAL AND METHODS Cell Culture
Primary monolayer cultures of liver cells from 16 day chicken embryos were prepared as described elsewhere (Parkes and Grieninger, 1985). Briefly, suspensions of hepatocytes were obtained by perfusion of the liver with buffer, mechanical disruption, and treatment with purified dissociating enzymes. Hepatocytes were washed and plated in modified Ham's F12 medium containing 25 mM glucose but no hormones or serum supplement. The modified medium contained no lactate or pyruvate and no other sugar but glucose. Hepatocyte suspensions were plated in 35- or 55-mmdiameter untreated Falcon plastic culture dishes. Culture medium was replaced with a n equal volume of fresh medium every 24 h r unless otherwise indicated. Monolayers contained approximately 2 x lo6 or 5 x 106 cells in 35 or 55 mm dishes, respectively; 3 x lo6 cells represent about 1 mg of protein (Grieninger and Granick, 1978). These nondividing cultures suffer no loss of viability for several days in the absence of hormone (Liang and Grieninger, 1981; Plant et al., 1983). Insulin, which is not mitogenic in these cultures (Liang and Grieninger, 1981; Plant et al., 1983), was introduced only after removal of unattached cells during the first medium change to avoid hormonal effects on plating efficiency. Medium varied by the addition of 10,25, or 75 mM glucose with and without 35 nM insulin. 3H-glucose was added to the medium of the 35 mm dishes for autoradiographic localization of label incorporated into glycogen. Additional details of the various experimental culture conditions are given below. Samples of cells were obtained at various times for biochemical analysis of glycogen content, light and electron microscopic evaluation, andlor autoradiography. Microscopy
Monolayers were fixed briefly (5 min) in a 2% glutaraldehyde-2% paraformaldehyde mixture in 0.1 M sodium cacodylate buffer (pH 7.3) (Cardell et al., 1985). Cells were gently scraped from the culture dishes as sheets and/or strips of cells in a small amount of fixative and transferred in aliquots which yielded a pellet 0.5-1 mm thick when centrifuged for 2 min in 0.5 ml microfuge tubes containing fresh fixative. Pellets were removed from the tubes and placed into vials of fresh fixative for a minimum of 2 h r at room temperature, washed in buffer for 1 hr, and postfixed for 2 h r in cacodylate-buffered 1% osmium tetroxide. Following dehydration in a graded series of ethanol and infiltration with propylene oxide, the samples were embedded in Epon 812 (Cardell et al., 1985). Thick (0.5-1 Fm) and thin (silver) sections were cut on a Sorvall MT-2B ultramicrotome. Thick sections were adhered to glass slides by gentle heat. Sections were evaluated immediately after staining with toluidine blue to confirm that the section was from the portion of pellet1 that 'Semithin longitudinal sections of a n entire pellet show that the pellet consists almost entirely of sheets of cells throughout, except for the most distal (widest) portion, which is a thin layer of single cells, broken cells, and debris. An optimal region of the pellet, therefore, is composed of sheets of strips of cells that have retained their cell to cell relationships. Centrifugation in the microfuge does not distort or disassociate the cells; the sheets of cells are not packed tightly against each other in the pellet.
consists of sheets of cells. Slides with unstained sections of the optimal region of the pellets were coated with Kodak NTB-2 nuclear tracking emulsion for light microscopy autoradiography. Thin sections for routine transmission electron microscopy (EM) were picked up on copper grids and stained with uranyl and lead salts; thin sections for autoradiography were transferred by wire loop and air-dried onto nitrocellulose-coated glass slides. Prior to coating with Ilford L-4 emulsion diluted to yield a monolayer of silver halide crystals, thin sections on slides were stained with uranyl acetate and lead citrate and coated with carbon. After various exposure times, autoradiograms were developed (Michaels et al., 1984), the sections were floated of" the slides, picked up on grids, and evaluated. Reagents
Bovine pancreas insulin (crystalline, Sigma) and porcine glucagon (crystalline, NOVO,Denmark) were prepared and stored in 0.01 N HC1. 3H-D-glucose (20 Cilm mole) was obtained from ICN Pharmaceuticals. Glycogen Determination
Culture dishes were treated as described, with minor modification (Parkes and Grieninger, 19851, for determination of glycogen. Monolayers were washed twice and immediately frozen in a n ethanolldry-ice bath. The cells were scraped from the dishes in water with a Teflon policeman. Glycogen was measured by digestion for 10 min at 55°C with amyloglucosidase (Sigma), followed by centrifugation, and subsequent colorimetric determination of glucose by the glucose-oxidase procedure (Roehrig and Allred, 1974) using a kit from Sigma. Glycogen Synthesis
Three hours after plating, cells were washed once and incubated for 16 h r in medium containing 10 mM glucose to deplete glycogen stores [Note: the blood glucose level of a starved chicken is approximately 15 mM (see Parkes and Grieninger, 198511. Hepatocytes were then given fresh medium containing 25 mM HEPES (pH 7.4) and either 10 or 75 mM glucose with and without insulin (35 nM). Metabolic reactions were stopped at the appropriate times by removal of medium and immediate addition of fixative for EM studies or immersion in a ethanolldry-ice bath for biochemical determination of glycogen content. For the autoradiographic experiments, confluent cells in 35 mm dishes were given medium containing 10 or 75 mM glucose with and without insulin, along with approximately 1.25 mCi of 3H-glucose. The cells were then harvested at various times (5, 15, and 30 min, 2,12 and 24 hr). Hepatocytes that were harvested after 12 or 24 h r received 3H-glucose 2 h r prior to harvesting. Glycogen Degradation
Three hours after plating, glycogen stores were enhanced by incubating cells for 48 h r in medium containing 75 mM glucose plus 35 nM insulin with one change into fresh medium after 24 hr. At the end of this pretreatment period, culture dishes were washed once and incubated in medium containing 10 mM glucose. Thirty minutes later, at time 0, the dishes were given
GLYCOGEN METABOLISM I N CHICK HEPATOCYTES
" ' J 4 40
323
[GLYCOGEN SYNTHESIS]
t
0lOmM
glucose
75mM glucose M 7 5 m M glucose + 35nM insulin
. 1
0
1
2
5
10
15
Fig 1 . Time course of glucose and insulin-induced glycogen accumulation. Three hours after plating, cells were washed once and incubated for 16 hr in medium containing 10 mM glucose. Hepatocytes were then given fresh medium (time 0 ) containing either 10 mM glu-
20
30
60
6
24
65
cose or 75 mM glucose with and without 35 nM insulin. Glycogen content was determined at the times indicated. Cells incubated for 65 hr were given fresh experimental medium every 24 hr.
fresh medium containing HEPES (25 mM, pH 7.3) plus 10 mM glucose and 350 nM glucagon, incubated for various lengths of time, and prepared for glycogen determinations and morphological studies as described above. OBSERVATIONS AND RESULTS Glycogen Synthesis
Figure 1 shows the response of the cultured chick hepatocytes to medium containing 10 mM glucose, 75 mM glucose, or 75 mM glucose plus 35 nM insulin for several minutes through 65 hr. The presence of 10 mM glucose resulted in biochemically undetectable or very low levels of glycogen over the entire time course, although the addition of 75 mM glucose stimulated measurable glycogen synthesis by 5 min with additional accumulations by 1h r in culture. Glycogen accumulation equal to physiologically maximal levels was obtained by long-term culture in medium containing 75 mM glucose plus 35 nM insulin. Although the effect of added glucose on glycogenesis is seen within minutes, the effect of insulin on glycogenesis is apparent after approximately 6 h r of incubation (Fig. 1; and Parkes and Grieninger, 1985). As noted above, hepatocytes maintained in medium with 10 mM glucose had very low levels of glycogen (Fig. 1).Morphologically, these glycogen-depleted cells had large lipid droplets, typical profiles of Golgi apparatus, elongated mitochondria, little or no smooth endoplasmic reticulum (SER), isolated cisternae of rough endoplasmic reticulum (RER), and a few peroxisomes scattered throughout the cytosome and near the ends of RER cisternae (Fig. 3). Most profiles of cells were devoid of glycogen granules, but a few cells (1% of the cell profiles. SER was present in the cytoplasmic regions displaying glycogen particles. There also appeared to be a close association of peroxisomes with these glycogen regions. Otherwise, the basic morphology of these hepatocytes after short incubation in higher levels of glucose with and without insulin was similar to that of the glycogen-depleted cells (Fig. 3). Five minutes after addition of 75 mM glucose (not shown) or 75 mM glucose plus 35 mM insulin (Fig. 5B), approximately 2% of the hepatocyte profiles contained glycogen. Again, there was a close association of SER with CY and p particles of glycogen; peroxisomes were near these regions. After 1 h r incubation with 75 mM glucose plus 35 nM insulin, the percent profiles containing glycogen was almost twice that of hepatocytes cultured for the same time with 75 mM glucose and no insulin (about 6% and 3%, respectively). After six hours of culture in the presence of 75 mM glucose plus 35 nM insulin, approximately 82% of the cell profiles contained glycogen, sometimes in large compact masses (Fig. 6A), whereas only 12% of the hepatocyte profiles contained glycogen when cultured in glucose alone (not shown). The morphology of the glycogencontaining cells was similar in both groups and showed SER near the glycogen particles (Fig. 6A). By 24 h r of culture in 75 mM glucose and 35 nM insulin, approximately 99% of the cell profiles showed glycogen in the cytosome, whereas 53% of the hepatocyte profiles contained glycogen when cultured with glucose alone. By 65 h r of culture with insulin, cells in which glycogen practically filled the cytosome (Fig. 6B) were common. At this stage of maximum deposition, elements of SER, though not abundant, were found near the glycogen masses. This observation is consistent with in vivo studies (Cardell, 1977; Striffler, et al., 19811, which showed relatively little SER in glycogen-filled hepatocytes compared to more abundant SER in hepatocytes actively metabolizing glycogen. Profiles of hepatocytes maintained for about 3 days in 10 mM glucose showed little or no glycogen (not shown), consistent with the biochemical data (Fig. 1). To determine if the glycogen observed in the hepatocytes was newly synthesized, cells were incubated for various times in medium containing 10 or 75 mM 51ucose, with and without insulin, plus -1.25 mCi of Hglucose. Labeled glucose, a precursor to glycogen, was then localized in the cells by autoradiography. Newly synthesized glycogen was identified at all times studied, and it is significant to note that hepatocytes cul‘To make estimates of cells containing glycogen, we randomly evaluated 500 profiles of cells in a single thin section with the EM and determined for each experimental group the percentage of profiles withiwithout glycogen in that plane of section. Because these hepatocytes grown to confluency have a fairly regular shape, any portion of an hepatocyte that included mitochondria and other organelles, but not necessarily the nucleus, was counted. Every profile within contiguous grid squares was counted until a total of 500 were recorded. Micropodia and other small cellular projections did not contain mitochondria, etc., and were not counted. These uniform-size hepatocytes are not amoeboid or stellate in shape, so profiles of cells closely correspond to actual cell numbers. However, it is obvious that changes in cellular size and the random distribution of glycogen within hepatocytes cause predictable errors in this estimate.
327
tured for a short time with either 10 mM glucose (Fig. 5A) or 75 mM glucose (Fig. 7B, C) showed silver grains over small glycogen regions, indicating incorporation of glucose into glycogen. The region shown in Figure 7A corresponds to a n “early SERGE,” described by Cardell et al. (1985) as a restricted area of cytoplasm in which smooth endoplasmic reticulum, glycogen (or precursors), and related enzymes are localized. Very little glycogen is visible here, although the presence of silver grains indicates that 3H-glucose has been incorporated into newly synthesized glycogen. Since glucose is utilized by hepatocytes for activities other than glycogen synthesis (Neutra and Leblond, 1966a,b), it is not surprising that several organelles of the cell, primarily the Golgi apparatus, sometimes were labeled (not shown). Glycogen Degradation
Fig. 8 shows the extensive deposits of glycogen that accumulate in the cells incubated in 75 mM glucose and 35 nM insulin for 48 hr. The large glycogen masses are compact with limited associated SER. To study glycogen degradation, such cells were given fresh insulinfree medium containing 10 mM glucose plus 350 nM glucagon and incubated for various times up to 3 hr. Figure 2 shows that glucagon causes a rapid breakdown of glycogen, which is nearly complete by 180 min (see also Parkes and Grieninger, 1985). The embryonic chick hepatocyte culture system is a n excellent model for the morphological study of glycogen depletion under highly regulated conditions. Ten minutes after addition of glucagon to the cultures, most cells showed SER infiltrations within the glycogen masses (Fig. 9A). However, other hepatocytes failed to show this early response to hormone. After longer exposure to glucagon additional cells showed decreased amounts of glycogen, and by 60 min many hepatocytes were practically depleted of glycogen (Fig. 9B). SER was observed near the remaining glycogen particles during late stages of glycogenolysis. By 180 min, virtually all cells were depleted of glycogen and SER was sparse in the cytosome. DISCUSSION
Primary cultures of chicken embryo hepatocytes were used to study morphological aspects of hepatocellular glycogen metabolism. This is a n excellent system for such studies because the cultured hepatocytes do not require hormones or other macromolecular media supplements for their viability and expression of hepatic functions (Parkes and Grieninger, 1985). Thus it was possible to study morphological aspects of hepatic glycogen metabolism under precisely controlled conditions. Hepatocytes obtained from 16-day-old chicken embryos and cultured in modified Ham’s F12 medium with low levels of glucose are healthy and do not lose viability for several days in the absence of hormones (Parkes and Grieninger, 1985), even though under these conditions the cells contain almost no biochemically detectable glycogen. Addition of increasing concentrations of glucose to glycogen-depleted cells causes the hepatocytes to synthesize glycogen; however, insulin is required to restore fed in vivo levels of glycogen in the cultured cells (Parkes and Grieninger, 1985).
328
J.L. PARKES ET AL
Fig. 6. A Sizeable glycogen accumulations are found in hepatocytes after 6 hr culture with 75 mM glucose and 35 nM insulin. Arrowheads, SER arrows, glycogen. B: Long term culture (65 hr) in medium
supplemented with glucose and insulin results in glycogen deposits, which occupy most of the cytosome. Arrowheads, SER arrows, glycogen; P, peroxisome.
Culture of hepatocytes in hormone-free medium supplemented with 10 mM glucose resulted in very little glycogen synthesis over 65 hr. Five minutes after ad-
dition of 75 mM glucose to the culture medium, measurable glycogen begins to accumulate in the hepatocytes, and by 1 hr significant quantities of glycogen are
GLYCOGEN METABOLISM IN CHICK HEPATOCYTES
Fig. 7. A After 30 min incubation with 3H-glucose, hepatocytes that have been maintained in medium containing 10 mM glucose show small glycogen deposits and associated ribbon-like silver grains indicating incorporation of the isotope into the newly formed glycogen. P, peroxisomes; arrows, glycogen; arrowheads, SER. B,C: Cells that
329
have been incubated with 75 mM glucose and labeled glucose for 30 min show distinct glycogen patches. Silver grains indicate incorporation of the isotope into glycogen synthesized after addition of the isotope. P, peroxisomes; arrows, glycogen; arrowheads, SER.
330
J.L. PARKES ET AL.
Fig. 8. Hepatocytes can be induced to deposit large quantities of glycogen by 48 h r incubation with 75 mM glucose plus 35 nM insulin added to the medium. The cytosome is practically filled with glycogen and such cultures served as baselines for studies of glucagon action. G1, glycogen.
GLYCOGEN METABOLISM IN CHICK HEPATOCYTES
Fig. 9. A Ten minutes after addition of 350 nM glucagon to fresh insulin-free medium, the glycogen areas in some hepatocytes contain many tubules of smooth endoplasmic reticulum (arrowheads). Arrows, glycogen; P, peroxisome. B: After 60 min incubation with glu-
331
cagon, cells almost devoid of glycogen are found. Notice that SER (arrowheads) is close to the remaining glycogen particles (arrows). P, peroxisomes.
332
J.L. PARKES ET AL.
synthesized. Addition of 35 nM insulin to the culture system caused a sustained synthesis of glycogen, which reached levels equal to the fed in vivo situation by 65 hr. The ultrastructure of hepatocytes cultured i n medium containing 10 mM glucose showed the usual cellular organelles and, in a n occasional cell, a few glycogen granules. The location of silver grains over these sparse glycogen granules in the autoradiographic preparations indicated that a very low rate of glycogen synthesis occurred during culture with 10 mM glucose; therefore, these infrequently observed glycogen granules were not residual glycogen in the cell. After addition of 75 mM glucose to the culture medium a small percentage of the hepatocytes responded rapidly by synthesizing glycogen. This newly synthesized glycogen appeared in small (
