Effects of glucagon on glycogenolysis and gluconeogenesis are region-specific in periportal and perivenous hepatocytes

Effects of glucagon on glycogenolysis and gluconeogenesis are region-specific in periportal and perivenous hepatocytes YOSHIHIRO IKEZAWA, KEIICHI YAMATANI, ATSUSHI OGAWA, HIROSHI OHNUMA, MASAHIKO IGARASHI, MAKOTO DAIMON, HIDEO MANAKA, and HIDEO SASAKI YAMAGATA, JAPAN

It has been established, mainly by histochemical and immunohistochemical studies, that liver cells are functionally heterogeneous, with periportal hepatocytes (PPHs) being predominantly gluconeogenic and perivenous hepatocytes (PVHs) being glycolytic. We therefore investigated the region-specific functional effects of glucagon on glycogenolysis and gluconeogenesis in isolated PPHs and PVHs prepared by the digitonin-collagenase method. BB rats, a model of insulin-dependent diabetes, were used to study the region-specific heterogeneity of gluconeogenesis in the diabetic state. Although glycogen content was not different between PVHs and PPHs in rats fed the normal diet, basal glucose release was 1.37 times greater in PVHs than in PPHs (P < .05). The increase in glucose release stimulated by 0.01 to 0.1 nmol/L glucagon was 1.52 times greater in PVHs than in PPHs (P < .05), whereas no differences were seen in response to I to 100 nmol/L glucagon. Glucose release from gluconeogenic substrates was 1.57 times greater in the PPHs than in the PVHs of fasted normal rats (P < .05), whereas the increase in gluconeogenesis produced by glucagon was not different between PPHs and PVHs. The glucagon-binding capacity, the cAMP release, and the increase in intracellular Ca 2+ stimulated by glucagon were not different between PPHs and PVHs in the fed or fasted states. Gluconeogenesis from gluconeogenic substrates was 1.52 times greater in the PPHs than in the PVHs of fasted nondiabetic BB rats (P < .05). After the development of diabetes, the gluconeogenic capacity in PVHs increased to the level observed in PPHs, but that in PPHs did not change. Thus there was no difference in gluconeogenesis between the PPHs and PVHs of diabetic BB rats. In both the PPHs and PVHs of diabetic BB rats, the 0.01 to 100 nmol/L glucagon-induced increase in gluconeogenesis was greater than that in PPHs from nondiabetic BB rats (2.30 and 3.07 times, P < .01, respectively). We conclude that PPHs and PVHs of normal rat liver express region-specific differences in their glycogenolytic and gluconeogenic responses to glucagon. In diabetic BB rats, the difference in the gluconeogenic capacity between PPHs and PVHs disappeared, whereas glucagon-induced gluconeogenesis was enhanced. (J Lab Clin Med 1998; 132:547-55)

Abbreviations: [Ca2+]i = intracellular free calcium concentration; cAMP = cyclic 3",5"-adenosine monophosphate; FBS= fetal bovine serum; PEPCK= phosphoenolpyruvate carboxykinase, PPH = periportal hepatocyte; PVH = perivenous hepatocyte

From The Third Department of Internal Medicine, Yamagata University School of Medicine. Submitted for publication January 6, 1998; revision submitted July 6, 1998; accepted August 10, 1998. Reprint requests: Yoshihiro Ikezawa, MD, The Third Department of

Internal Medicine, Yamagata University School of Medicine, 2-2-2 Iida Nishi, Yamagata 990-9585, Japan. Copyright © 1998 by Mosby, Inc. 0022-2143/98 $5.00 + 0 5/1/93812

547

548

Ikezawa et al

The liver is an important center of carbohydrate, lipid, and protein metabolism. Hepatic glucose production in the fasting state plays a key role in the maintenance of glycemia, which is necessary for survival of the organism. Under these conditions glucagon is the mostimportant hormone regulating hepatic glucose output, through stimulation of both glycogenolysis and gluconeogenesis. It has been reported that glycogenolysis and gluconeogenesis are similarly sensitive to stimulation by glucagon in vivo. 1 The concept of the hepatic acinus as the unit of microcirculation, first proposed by Novikoff, 2 is currently generally accepted. Following the bloodstream, at least 2 different zones can be discerned, the periportal zone perfused with blood rich in oxygen, substrates, and hormones; and the perivenous zone perfused with blood depleted in oxygen, substrates, and hormones but enriched in CO 2 and other products of metabolism. 3 A number of histochemical 4 and immunohistochemical 5 studies have shown that the hepatocytes along the sinusold differ in a variety of morphologic, histochemical, and biochemical characteristics. It was also reported that the activity of gluconeogenic enzymes (ie, PEPCK, fructose 1,6-bisphosphatase, and glucose 6-phosphatase) is 1.5 to 3 times higher in PPHs than in PVHs, whereas the activity of glycolytic enzymes (ie, glucokinase and pyruvate kinase) is higher in PVHs. 3 Thus it was suggested that liver cells are functionally heterogeneous, with PPHs being predominantly gluconeogenic and PVHs glycolytic. It remains to be established whether PPHs and PVHs demonstrate the same characteristics in response to hormones in vivo. The purpose of the present study was to investigate in detail the short-term regulation of glycogenolysis and gluconeogenesis by glucagon in isolated PPHs and PVHs. The glucagon-binding capacity and the increments in c A M P and intracellular Ca 2+ stimulated by glucagon were also examined in PPHs and PVHs. In addition, because hepatic glucose overproduction contributes to the hyperglycemia of diabetes, 6,7 BB rats, a model of insulin-dependent diabetes, 8 were used to investigate the region-specific heterogeneity of gluconeogenesis in the diabetic state. METHODS Animals. Male Wistar rats (body wt 180 to 220 g) were maintained on a 12-hour day/night rhythm with unlimited access to a standard laboratory diet and tap water in the Laboratory Animal Center, Yamagata University. BB rats were provided by the Animal Research Center, Tokyo Medical College, 9 and bred in the Laboratory Animal Center, Yamagata University. Only male BB rats (body wt 180 to 220 g) were used in the present study. The appearance of glucosuria, based on weekly analysis, was used to establish the onset of dia-

J Lab Clin Med D e c e m b e r 1998

betes. Rats were given unlimited access to food for the study of glycogenolysis or were fasted for 24 hours in the study of gluconeogenesis. All animals were anesthetized before surgery by intraperitoneal injection of pentobarbital (100 mg/kg) between 1 and 1:30 PM. Selective hepatocyte isolation. PPHs and PVHs were isolated by the digitonin-collagenase method, 10,11 with minor modifications. 12 In brief, after a loose tie was placed around the inferior vena cava above the right kidney, the portal vein was cannulated, and perfusion (15 mL/min) was commenced with Hanks' CaZ+-free buffer containing 0.5 mmol/L ethyl-

eneglycol-bis-(~l-aminoethylether)-N,N,N',N'-tetraacetic acid. The chest cavity was then opened for cannulation of the superior vena cava, and the ligature around the inferior vena cava was fastened. For preparation of PPHs, a 1 mmol/L digitonin (Sigma Chemical Co, St Louis, MO) solution, prepared by boiling in Hanks' solution, was infused at 37 ° C through the cannula from the superior vena cava at a rate of 10 mL/min for 10 seconds, followed by Ca2+-free Hanks' buffer for 3 minutes. Collagenase (95 U/mL; Wako, Osaka, Japan) was then infused for 7 minutes at a rate of 15 mL/min through the portal cannula. The liver cells were then dispersed in Eagle minimum essential medium (Nissui, Tokyo, Japan), filtered through gauze mesh, and washed 4 times by centrifugation at 50 g for 1 minute. For preparation of PVHs, the procedure was the same except that the direction of the preperfusion, digitonin, and collagenase infusions were reversed. Viability was determined by exclusion of 0.05% Trypan blue dye, and the cell yield was determined by the hepatocrit method. Cell viability was routinely more than 95%. Glycogenolysis and gluconeogenesis. Cells were diluted to 106 cells/mL in Williams' medium E (Dainippon, Tokyo, Japan) containing 10% (vol/vol) FBS (BioWhittaker, Walkersville, MD), 1 nmol/L insulin (Novo-Nordisk, Copenhagen, Denmark), and 1 nmol/L dexamethasone (Upjohn, Tokyo, Japan). A 2-mL sample of the inocula of hepatocytes was plated in 35-ram collagen-coated dishes and placed in a CO 2 incubator at 37°C for 1 hour. After cell attachment, these monolayers were washed with Krebs-Henseleit buffer twice with 10-minute intervals. For the analysis of glycogenolysis, glucose released from PPHs or PVHs of fed rats was determined during a subsequent 30-minute incubation in Krebs-Henseleit buffer with 0 to 100 nmol/L glucagon (Novo-Nordisk). For the determination of gluconeogenesis, glucose released from PPHs or PVHs of 24-hour-fasted rats was studied during a subsequent 30-minute incubation in Krebs-Henseleit buffer containing 5 mmol/L lactate, 0.5 mmol/L pyruvate, and 5 mmol/L L-alanine with 0 to 100 nmol/L glucagon. The medium was collected and stored at -20 ° C until the assay of glucose by the glucose oxidase method (BoehringerYamanouchi, Tokyo, Japan) and the assay of cAMP by radioimmunoassay (Yamasa, Chyoshi, Japan). The cells were treated with 1% Triton X-100 (Sigma), scraped, and homogenized. The homogenate was stored at -20 ° C until protein assay by the Pyrogallol method (Boehringer-Yamanouchi) or analysis of glycogen by the amyloglucosidase method. 13

J Lab Clin M e d Volume 132,Number 6

Glucagon binding study. After isolation, PPH or PVH (105/mL) cells were incubated in 50-mL polystyrene tubes at 37 ° C for 1 hour in Williams' medium E containing 10% (vol/vol) FBS to allow the cell surface to recover from the collagenase-induced damage. The hepatocytes were then washed 3 times with Krebs-Ringer buffer containing 1% FBS and 0.35 mmol/L bacitracin (Sigma). Iodine 125-labeled glucagon was prepared with chloramine T 14 and separated by diethylaminoethyl-Sephadex A25 chromatography to a specific activity of 3 to 11 x 105 Ci/mol. PPHs or PVHs (5 x 105 cells)were incubated with [125I]labeled glucagon and unlabeled glucagon (0 to 1000 nmol/L) simultaneously (total volume 0.5 mL) at 37 ° C for 30 minutes with shaking. The hepatocytes were washed 3 times and centrifuged at 3000 rpm for 5 seconds, and the [125I]-labeled glucagon bound to hepatocytes was determined by using a ycounter (Wallac, Turku, Finland). Scatchard analysis of the binding data was carried out after subtraction of nonspecifically bound [125I]-labeled glucagon, as defined in incubations containing 10 gmol/L unlabeled glucagon. Values are means of triplicate determinations. The number of glucagon receptors and the affinity constants were calculated by the method of Rosenthal. 15 ~ntracellular calcium fluorimetry. After isolation, PPHs or PVHs (5 x 107 cells) were incubated with 5 gmol/L fura2/AM (Dojindo, Mashiki, Japan) in 3 mL of Williams' medium E containing 10% (vol/vol) FBSat 37 ° C for 1 hour. After washing, cells were placed in a temperature-controlled (37 ° C) flow-through chamber on the stage of an inverted microscope (Diaphoto-TMD; Nikon, Tokyo, Japan) equipped with a dual excitation imaging system (C3329; Hamamatsu Photonics, Hamamatsu, Japan) and were perfused with KrebsHenseleit buffer. [Ca2+]i was calculated from the ratio between the fluorescence intensities of fura-2 at 340 and 380 nm excitation. 16 The increase in [Ca2+]i was determined as the difference from basal of the average value during the last 30 seconds of a 9-minute infusion of 0 to 10 nmol/L glucagon. Statistics. Values are expressed as mean _+SEM. Analysis of variance or paired or unpaired Student's t test was used for the comparison between the 2 groups, as appropriate. P < .05 was considered significant. RESULTS Glycogenolysis in PPHs and PVHs of fed normal rats. G l y c o g e n content was not different between the PPHs and PVHs o f normal rats given free access to food (42.4 + 1.7 gg/mg protein in PPHs and 42.7 _+ 3. l gg/mg protein in PVHs 1 hour after isolation, respectively). Basal glucose release was 1.37 times greater from P V H s (89.7 _+ 8.6 n m o l / m g protein/30 minutes) than from PPHs (65.3 _+ 4.4 nmol/mg protein/30 minutes, P < .05) from fed normal rats (Fig 1, A). W h e n incubated with 0.01 n m o l / L glucagon, glucose release increased significantly, by 30.1 + 7.5 n m o l / m g protein/30 minutes (35.3% + 8.5%), from PVHs (P < .05), but insignificantly, b y 7.2 _+ 3.6 n m o l / m g protein/30 minutes (9.9% + 4.6%), from PPHs. On the other hand,

Ikezawa e t al

549

250 ._= 200-

o 150o

0. 100 "6 E ¢-

50-

°

A

o

o.51

o11

i

io

16o

Glucagon (nmol/L) ~" 140 ~= 120

~

~

__.

100 o

s0

"1 0.

60 40

o E C

20 0

S

o.ol

o11

1

lo

16o

Glucagon (nmol/L)

Fig 1. The effect of glucagon on glucose output from PPHs (B, n = 6) and PVHs (O, n = 6) of fed normal Wistar rats (A) and on glucose output with gluconeogenic substrates (5 mmol/L lactate, 0.5 mmol/L pyruvate, and 5 mmol/L L-alanine) from PPHs (BI, n = 6) and PVHs (0, n = 6) of 24-hour fasted normal Wistar rats (B). The glucose outputs without glucagon or gluconeogenic substrates were 17.2 -+ 3.7 nmol/mg proteird30 minutes in PPHs and 12.4 -+1.4 nmol/mg protein/30 minutes in PVHs.

the maximal increment in glucose release produced by 100 nmol/L glucagon was not different between PVHs and PPHs, increasing b y 112.1 -+ 19.3 n m o l / m g protein/30 minutes and 82.4 _+ 13.6 n m o l / m g protein/30 minutes (131.9% + 27.0% and 129.4% _+ 25.3%), respectively. Thus, although the increase in glucose release stimulated by 0.01 to 0.1 nmol/L glucagon was 1.52 times greater from P V H s than f r o m PPHs (P < .05), that stimulated by 1 to 100 nmol/L glucagon was not different between PVHs and PPHs. Although it not shown in Fig 1, A, the basal glucose release from hepatocytes of the whole liver, which was not treated by digitonin, was 78.4 _+ 7.2 n m o l / m g protein/30 minutes. The d o s e - r e s p o n s e curve o f g l y c o g e n o l y s i s i n c r e a s e d by 0.01 to 100 n m o l / L glucagon in hepatocytes of the whole liver and appeared intermediate between that of PPHs and PVHs.

550

J Lab Clin Med D e c e m b e r 1998

Ikezawa et al

0.07 0.06 LL

0.05-

"o

0.04

o

0.03

G3

k

~

0.02 0.01 0

0

A

~ °

100

......

200

300

400

500

600

Bound glueagon (fmol / 10Seells)

0.07 0.06.:~ u.

0.05 !

° ? %~p.

m

0.03:

0.02:

"" ....

o.o1:

~

~ . . . ~ ,....

O ~

0 B

"""-...

100

200

300

400

500

600

Bound glucagon (fmol / 106cells)

Fig 2. The Scatchard's plot that caluculated from the specific binding of [125I]-labeled glucagon to PPHs (11) and PVHs ( 0 ) of fed normal Wistar rats (A) and of 24-hour fasted normal Wistar rats (B). Values are means of triplicates.

Gluconeogenesis in PPHs and PVHs of fasted normal rats. Basal glucose release was 17.2 _ 3.7 nmol/mg pro-

tein/30 minutes from PPHs and 12.3 ___1.4 nmol/mg protein/30 minutes from PVHs of 24-hour-fasted rats. When incubated with gluconeogenic substrates, glucose release was 1.57 times greater from PPHs (86.7 _+ 6.9 nmol/mg protein/30 minutes) than from PVHs (56.9 + 5.7 nmol/mg protein/30 minutes, P < .05) in PPHs (Fig 1, B). When 0.01 nmol/L glucagon was added, glucose release increased significantly, by 2.4 _+ 0.8 nmol/mg proteird30 minutes (3.2% + 1.2%), from PPHs (P < .05) but insignificantly, by 0.2 _+0.9 nmol/mg protein/30 minutes (0.2% + 1.5%)~ from PVHs. The 100 nmol/L glucagon-induced increase in glucose release was not different between PPHs and PVHs, rising by 27.3 _+5.4 nmol/mg protein/30 minutes and 32.6 ___3.9 nmol/mg proteird30 minutes (33.5% _+7.7% and 61.7% + 11.6%), respectively. As a whole, glucagon-stimulated increases in glucose release were not different between PPHs and PVHs of fasted rats (Fig 1, B). In the absence of digitonin treatment, glucose release from hepatocytes of the whole liver of fasted rats was

62.3 _+ 6.8 nmol/mg protein/30 minutes with the substrates alone, and the dose-response curve of gluconeogenesis produced by 0.01 to 100 nmol/L glucagon ranged between that of PPHs and that of PVHs. Gluconeogenic rates were constant during the test period in the presence and absence of glucagon (data not shown). G l u c a g o n binding to PPHs and PVHs. Maximal binding of [125I]-labeled glucagon was 4.9% to PPHs and 5.5% to PVHs of fed rats. Assuming 2 classes of saturable binding sites, analyzed by the method of Rosenthal, 15 there were approximately 44,000 binding sites/cell with a high-affinity constant of about 1.3 x 10-9/mol and approximately 211,000 binding sites/cell with a low-affinity constant of about 0.14 x 10-9/mol in PPHs (Fig 2, A). There were approximately 39,000 binding sites/cell with a high-affinity constant of about 1.4 x 10-9/mol and approximately 333,000 binding sites with a low-affinity constant of about 0.09 x 10-9/mol in PVHs (Fig 2, A). When rats were fasted for 24 hours, maximal binding of labeled glucagon was 4.7% to PPHs and 5.8% to PVHs. There were approximately 60,000 binding sites/cell with a high-affinity constant of about 0.9 x 10-9/mol and approximately 198,000 binding sites with a low-affinity constant of about 0.13 x 10-9/mol in PPHs, whereas PVHs exhibited approximately 52,000 binding sites/cell with a high-affinity constant of about 1.2 x 10-9/mol and approximately 134,000 binding sites with a low-affinity constant of about 0.23 x 10-9/mol (Fig 2, B). Thus there was no difference in glucagon binding between PPHs and PVHs in either the fed or fasted state. cAMP released into the medium from PPHs and PVHs.

During 30-minute incubations with 10 and 100 nmol/L glucagon, cAMP released from PPHs of fed normal rats was 1.45 _+0.49 pmol/mg protein/30 minutes and 2.53 + 1.35 pmol/mg protein/30 minutes, respectively (Fig 3, A). This was not different from that observed in PVHs (1.39 __ 0.58 pmol/mg protein/30 minutes at 10 nmol/L glucagon and 1.8 _+ 0.56 pmol/mg protein/30 minutes at 100 nmol/L glucagon, respectively) (Fig 3, A). When PPHs or PVHs were incubated with 0.01 to 1 nmol/L glucagon, cAMP released into the medium was below detectable levels (< 0.3 pmol/mg protein/30 minutes). In PPHs from 24-hour fasted rats, cAMP released by treatment with 10 and 100 nmol/L glucagon was 1.11 + 0.48 pmol/mg protein/30 minutes and 1.42 _+ 0.76 pmol/mg protein/30 minutes, respectively (Fig 3, B). The cAMP release by PVHs of 24-hour fasted rats in response to 10 and 100 nmol/L glucagon was 1.65 _+ 0.71 pmol/mg protein/30 minutes and 2.93 _+ 0.89 pmol/mg protein/30 minutes, respectively. There was

J Lab Clin Med Volume 132, Number 6

Ikezawa et al

200-

~5. o.

E

O o. o~

551

150~

~=, 100

2

E

,~

O

E

0

A

0'.1

i

10

50

16o

Glucagon (nmol/L)

o.0ol

A

o.01

o11

1

1'o

Glucagon (nmol/L) 200

;-= E 4-

~o 150E "-" 100

3-

P

o

50

<~

o

6

B

o11

10

10o

Glucagon (nmol/L)

Fig 3. The

effect of glucagon on cAMP output from PPHs (11) and PVHs ( 0 ) of fed (A) or 24-hour fasted (B) normal Wistar rats. A, PPH (n = 5), PVH (n = 5). B, PPH (n = 5), PVH (n = 4). Detectable level, 0.3 pmol/mg protein/30 minutes.

no significant difference in glucagon-induced cAMP release between PPHs and PVHs of 24-hour fasted rats. Glucagon-induced increase in [Ca2+)i in PPHs and PVHs. Basal [Ca2+]i was 74 to 120 nmol/L in PPHs and 70 to 101 nmol/L in PVHs of fed rats and was 73 to 113 nmolfL in PPHs and 90 to 109 nmol/L in PVHs of 24-hour fasted rats. When 0.1 to 10 nmol/L glucagon was perfused for 9 minutes, [Ca2+] i in hepatocytes increased gradually in a dose-dependent manner. The increase in [CaZ+]i induced by 10 nmol/L glucagon was 134 _+ 10 nmol/L in PPHs and 140 + 18 nmol/L in PVHs of fed rats (Fig 4, A) and was 84 _+24 nmol/L in PPHs and 104 + 21 nmol/L in PVHs of 24-hour fasted rats (Fig 4, B). There was no difference in the increase in [Ca2+] i stimulated by 0.1 to 10 nmol/L glucagon between PPHs and PVHs either in the fed or the fasted state. Gluconeogenesis in PPHs and PVHs of fasting diabetic and nondiabetic BB rats. Hepatic glycogen content 1 hour after the isolation of hepatocytes from fed diabetic BB rats was highly variable (3.4 to 45.5 gg/mg protein in PPHs and 6.9 to 37.4 gg/mg protein in PVHs), precluding an examination of glycogenolysis. Therefore we investigated only the effect of glucagon on intrahepatic region-specific gluconeogenesis in the dia-

-50 o.ool a

o.ol

o11

1

io

Glucagon (nmol/L)

Fig 4. The effect of glucagon on [Ca2+]i in PPHs (Ill) and PVHs ( 0 ) of fed (A) or 24-hour fasted (B) normal Wistar rats. A[Ca2+] i is the difference of the average of the last 30 seconds during 9-minute infusion of glucagon from the basal period. Each point shows the mean of 10 to 20 observations from 3 to 4 experiments.

betic state. After 24-hour fasting, the plasma glucose concentration was 5.9 + 0.2 mmol/L in nondiabetic BB rats and 9.6 + 0.9 mmol/L in diabetic BB rats (P < .01). Basal glucose release was 7.7 + 2.0 nmol/mg prorein/30 minutes from PPHs and 4.6 _+ 1.6 nmol/mg prorein/30 minutes from PVHs of 24-hour fasted nondiabetic BB rats. When incubated with gluconeogenic substrates, glucose release was 1.52 times greater from PPHs than from PVHs (50.3 _ 5.8 nmol/mg protein/30 minutes vs 33.1 _+4.2 nmol/mg protein/30 minutes, P < .05), which was further increased by glucagon in a dose-dependent manner (Fig 5). The 100 nmol/L glucagon-stimulated increment in glucose release was not different between PPHs and PVHs of fasted nondiabetic BB rats, increasing by 16.3 +_ 2.6 nmol/mg prorein/30 minutes and 10.2 + 1.1 nmol/mg protein/30 minutes (34.8% _+6.9% and 33.1% _+5.1%), respectively. In 24-hour fasted diabetic BB rats, basal glucose release was 13.6 + 3.3 nmol/mg protein/30 minutes in PPHs and 11.0 + 3.5 nmol/mg protein/30 minutes in PVHs. When incubation was done with gluconeogenic substrates, glucose release was 56.8 __+.5.7 nmol/mg prorein/30 minutes from PPHs and 51.6 _+ 5.0 nmol/mg

552

J Lab Clin Med D e c e m b e r 1998

Ikezawa et al

120 lOO E

~

60 40 E 20

6 o.01 o11

i

1'o 16o

Glucagon (nmol/L) Fig 5. The effect of glucagon on glucose output with gluconeogenic substrates (5 mmol/L lactate, 0.5 mmol/L pyruvate, and 5 mmol/L Lalanine) from PPHs (El, n = 6) and PVHs (O, n = 6) of 24-hour fasted diabetic or PPHs (11, u = 6) and PVHs ( 0 , n = 6) of nondiabetic BB rats.

proteirl/30 minutes from PVHs, and this was further increased by glucagon in a dose-dependent manner (Fig 5). The 100 nmol/L glucagon-induced increase in glucose release was not different between PPHs and PVHs of fasted diabetic BB rats, rising by 41.7 _+ 12.2 nmol/mg protein/30 minutes and 42.3 __ 5.3 nmol/mg protein/30 minutes (86.3% +_ 32.4% and 88.4% + 17.3%), respectively. There was no difference between PPHs and PVHs of diabetic BB rats. As a whole, 0.01 to 100 nmol/L glucagon stimulated an increase in gluconeogenesis that was greater in the hepatocytes from diabetic BB rats than in those from nondiabetic rats (2.30 times greater in PPHs and 3.07 times greater in PVHs, P < .01). DISCUSSION The results of the present study demonstrate the existence of region-specific differences in the metabolic responses of hepatocytes to glucagon. To study such hormone responses, cells prepared from the PPH and PVH regions of the liver must retain their normal receptor and enzyme activities. The action of digitonin is believed to reside in its ability to complex with cellmembrane cholesterol, so cells are destroyed. 17 Thus, differential perfusion of the liver with digitonin permitted selective isolation of PPHs and PVHs from whole livers. We have confirmed, using this method, that PEPCK mRNA transcript levels are 2-fold greater in PPHs than in PVHs, whereas albumin mRNA levels are not different between PPHs and PVHs. 12 This is in accord with reports using in situ hybridization to localize PEPCK mRNA in the rat liver, 18 as well as with demonstrated PEPCK enzyme activity in PPHs and

PVHs obtained by the microdissection method. 19 In addition, the effect of glucagon on glucose output from hepatocytes of the whole liver, which was not treated by digitonin, was intermediate between that of PPHs and PVHs in cells from normal rats. These results indicate that, in the present study, the influence of digitonin is minimal, and the isolation of PPHs and PVHs is reliable in terms of both viability and selectivity. The present study demonstrated that basal and glucagon-induced glycogenolysis was greater in PVHs than in PPHs. The increase in glycogenolysis induced by 0.01 to 0.1 nmol/L glucagon (equivalent to concentrations found in the portal vein) was greater in PVHs than in PPHs (P < .05). In contrast, gluconeogenesis from lactate, pyruvate, and alanine was greater in PPHs than in PVHs. The addition of glucagon further increased gluconeogenesis; however, the maximum increase produced by 100 nmol/L glucagon was similar between PPHs and PVHs. Such differences in glycogenolysis between PPHs and PVHs in selectively isolated preparations have not been previously reported, except for a study by Tosh and Agius 2° indicating that treatment with 100 nmol/L glucagon depleted glycogen content almost completely in both PPHs and PVHs after 17 hours in culture. Regarding gluconeogenesis, Quistorff11 reported that gluconeogenesis from lactate and pyruvate was 2-fold higher in PPHs than that in PVHs. Tosh et al21 also demonstrated that gluconeogenesis from pyruvate, palmitate, and carnitine was greater in PPHs than that in PVHs during 2-hour cultures, but the increase in glucose release produced by 100 nmol/L glucagon was not different between the 2 groups. Furthermore, Shiota et a122 reported that the increase in gluconeogenesis from lactate and pyruvate produced by 0.01 to 1000 nmol/L glucagon was similar between perfused PPHs and PVHs. These reports are in accord with the present results. Therefore, it can be said that glycogenolysis is greater and more sensitive to glucagon in PVHs than it is in PPHs and that although gluconeogenesis was greater in PPHs than it wasin PVHs, it was similarly sensitive to glucagon in both groups of cells. Although previous reports have indicated that both glycogenolysis and gluconeogenesis are similarly sensitive to glucagon, 1 it must be noted that these findings arose from the summation of PPH and PVH responses in vivo. The different sensitivity to glucagon between PPHs and PVHs may be due to differences in glucagon receptor number or post-receptor events. The density of glucagon receptors was reported to be 3-fold higher in PVHs than in PPHs. 23 However, we were unable to detect any differences between PPHs and PVHs in glucagon-binding, in either the fed or fasted state. The

J Lab Clin Med Volume 132, Number 6

numbers of high- and low-affinity glucagon receptors in the present study were also similar to findings in previous reports with whole hepatocytes. 24 These results therefore indicate that the glucagon receptor itself was not responsible for the observed differences in glucagon responses. It has been reported that 700 nmol/L glucagon-stimulated adenylate cyclase 25 and cAMPphosphodiesterase 26 do not show a zonal activity gradient during normal feeding; yet after a 48-hour fast, the cyclase activity was increased and the diesterase activity was decreased to a greater extent in PVHs. 25 In the present study, the release of cAMP into the medium, which reflects intracellular cAMP content, 27 was increased when 10 to 100 nmol/L g l u c a g o n was added 27 but was not different statistically between PPHs and PVHs of fed or fasted rats. The increase in [Ca2+]i stimulated by glucagon was also similar between PPHs and PVHs in the fed or fasted state. Therefore the differential sensitivity to glucagon appears to occur consequent to intracellular events downstream from these second messengers. On the other hand, the glucose release from PPHs increased significantly more than that from PVHs when gtuconeogenic substrates were added. This may be a reflection of the distribution of the enzymes of gluconeogenesis, such as PEPCK or fructose 1,6-bisphosphatase, which are found in higher concentrations in PPHs than in PVHs.4, 5 However, regardless of lower activities of gluconeogenic enzymes, the maximal increase in gluconeogenesis produced by glucagon in PVHs was similar to that in PPHs. In addition, although 0.01 to 0.1 nmol/L glucagon-induced glycogenolysis was greater in PVHs than in PPHs, the glycogen content and the maximal increase in glycogenolysis induced by glucagon were not different between PPHs and PVHs. Keppens and De Wulf 28 reported that the total glycogen phosphorylase content and the sensitivity to 0.05 to 5.0 nmol/L glucagon were very similar for both PPHs and PVHs, whereas Aggarwal et at 29 reported that 10 to 100 nmol/L glucagon phosphorylated glycogen phosphorylase more in PPHs than in PVHs. Glucose-6-phosphatase was reported to be predominant in PPHs. 5 These findings show that a simple summation of enzyme activities does not reflect on the rate of glycogenolysis as reported previously. 3° Thus, although the basal activity of glycogen phosphorylase and the efficiency of the cascade of second messengers on it produced by 0.01 to 0.1 nmol/L glucagon might be predominant in PVHs, the increase in gtycogenolysis stimulated by 100 nmol/L glucagon was similar between PPHs and PVHs. It was of similar interest to us to determine whether the observed differences in glycogenolysis and gluconeogenesis between PPHs and PVHs were preserved in

Ikezawa et al

553

diabetes mellitus, a condition of hepatic glucose overproduction. In the present study we used genetically diabetes-prone BB rats, because the streptozocininduced diabetic rat has liver injury31, 32 and therefore is not a suitable model for examining the function of hepatocytes. BB rats develop autoimmunologic insulitis between 60 and 120 days of age, 8,9 leading to insulin deficiency. When not treated with insulin, diabetic BB rats are also emaciated, and in the present study, they exhibited extremely varied glycogen content in the liver. Accordingly it was difficult to observe glycogenolysis, and therefore we examined gluconeogenesis only. When incubation was carried out with 0.01 to 100 nmol/L glucagon and gluconeogenic substrales for 30 minutes, glucose release from the PPHs of nondiabetic BB rats was greater than that from PVHs (P < .05), as observed in normal Wistar rats. In contrast, when BB rats were diabetic, glucose output stimulated by gluconeogenic substrates alone was not different between PPHs and PVHs. Appel et a133 have demonstrated increased activities of gluconeogenic enzymes in diabetic BB rats. The present study suggest~ that the increase in gluconeogenic enzymes in diabetic BB rats occm's mainly in PVHs. Loss of differences in gluconeogenic capacity between PPHs and PVHs were also observed in rats after cold exposure for 5 days 22 and partial hepatectomy. 34 Under these circumstances, PEPCK in PVHs increased to the levels seen in PPHs. On the other band, although a marked extension of the distribution of PEPCK to PVHs has been observed, the gradient of PEPCK between PPHs and PVHs persists in streptozocin-induced diabetic rats. 34-36 These reports are discordant with the present study, suggesting a difference between genetically diabetes-prone BB rats and rats with chemically induced diabetes with liver injury. Glucose output stimulated by 0.01 to 100 nmol/L glucagon together with gluconeogenic substrates was not different between PPHs and PVHs from diabetic BB rats and was greater than that of PPHs from nondiabetic BB rats. In eur preliminary study, glucagon-induced cAMP response was not enhanced in the isolated BB rat liver, suggesting that events after the production of second messengers are responsible for the increased glucagon effect in the diabetic BB rat liver. Thus characteristics of hepatic glucose overproduction in diabetic BB rats were revealed whereby the difference in the gluconeogenic capacity between PPHs and PVHs disappeared while glucagon-induced gluconeogenesis was enhanced. The results of the present study indicate that glycogenolysis is greater and more sensitive to glucagon in PVHs than in PPHs. It has been proposed that metabolic heterogeneity in hepatic acinus may

554

Ikezawa et al

result from concentration gradients of oxygen and glucagon/insulin ratio across the portal/venous vasculature. 3 Because these characteristics of PVHs disappeared in insulin-deficient diabetic BB rats, in which the g l u c a g o n / i n s u l i n ratio is very high, i n s u l i n must play a pivotal role in m a i n t a i n i n g PVH function. This is in accord with a report that whole hepatocytes developed PVH-like characteristics in response to treatment with insulin in primary c u l t u r e s Hepatocytes arising from the perivenous zone therefore seem to be more important for glucose homeostasis than those from the periportal zone. REFERENCES

1. Stevenson RW, Steiner KE, Davis MA, Hendrick GK, Williams PE, Lacy WW, et al. Similar dose responsiveness of hepatic glycogenolysis and gluconeogenesis to glucagon in vivo. Diabetes 1978;36:382-9. 2. Novikoff AB. Cell heterogeneity within the hepatic lobule of the rat (staining reactions). J Histochem Cytochem 1959;7:240-4. 3. Jungermann K, .Katz N. Functional specialization of different hepatocyte populations. Physiol Rev 1989;69:708-64. 4. Katz N, Teusch HF, Sasse D, Jungermann K. Heterogeneous distribution of glucose-6-phosphatase in rnicrodissected periportal and perivenous rat liver tissue. FEBS Lett 1977;76:226-30. 5. Lawrence GM, Jepson MA, Trayer IR Walker DG. The compartmentation of glycolytic and gluconeogenic enzymes in rat kidney and liver and its significance to renal and hepatic metabolism. Histochem J 1986;18:45-53. 6. DeFronzo RA, Simonson D, Ferrannini E. Hepatic and peripheral insulin resistance: a common feature of type 2 (non-insulin-dependent) and type 1 (insulin-dependent) diabetes mellitus. Diabetologia 1982;23:313-9. 7. Fery F. Role of hepatic glucose production and glucose uptake in the pathogenesis of fasting hyperglycemia in type 2 diabetes: normalization of glucose kinetics by short-term fasting. J Clin Endocrinol Metab 1994;78:536-42. 8. Nakhooda AF, Like AA, Chappel CI, Wei CN, Marliss EB. The spontaneously diabetic Wistar rat (the "BB" rat). Diabetologia 1978;14:199-207. 9: Yanagisawa M, Hara Y, Satoh T, Tanikawa T, Sakatsume Y, Katayama S, et al. Spontaneous autoimmune thyroiditis in Bio Breeding/Worcester (BB/W) rat. Endocrinol Japon 1986;33:851-61. 10. Lindros KO, Penttila KE. Digitonin-collagenaseperfusion for efficient separation of periportal or perivenous hepatocytes. Biochem J 1985;228:757-60. 11. Quistorff B. Gluconeogenesis in periportal and perivenous hepatocytes of rat liver, isolated by a new high-yield digitonin/collagenase perfusion technique. Biochem J 1985;22:221-6. 12. Ogawa A, Kurita K, Ikezawa Y, Igarashi M, Kuzumaki T, Daimon M, et al. Functional localization of glucose transporter 2 in rat liver. J Histochem Cytochem 1996;44:12316. 13. Keppler D, Decker K. Glycogen determination with amyloglucosidase. In: Bergmeyer HU, editor. Methods of enzymatic analysis. New York: Academic Press; 1974. p. 127-31. 14. Freychet R Roth J, Neville DM Jr. Monoiodoinsulin: demonstration of its biological activity and binding to fat cells and

J Lab Clin M e d D e c e m b e r 1998

liver membranes. Biochem Biophys Res Commun 1971;43:4008. 15. Rosenthal HE. A graphic method for the determination and presentation of binding parameters in a complex system. Anal Biochem 1967;20:525-32. 16. Shibuya I, Matsuyama K, Tanaka K, Doi K. A microfluorometric method for simultaneous measurement of changes in cytosolic free calcium concentration and pH in single cardiac myocytes. Jpn J Physiol 1991;41:341-50. 17. Chen KS, Katz J. Zonation of glycogen and glucose syntheses, but not glycolysis, in rat liver. Biochem J 1988;225:99104. 18. Bartels H, LinnemannH, Jungermann K. Predominant localization of phosphoenolpyruvate carboxykinase mRNA in the periportal zone of rat liver parenchyma demonstrated by in situ hybridization. FEBS Lett 1989;248:188-94. 19. Guder WG, Schmidt U. Liver cell heterogeneity: the distribution of pyruvate kinase and phosphoenolpyruvate carboxykinase (GTP) in the liver lobule of fed and starved rats. HoppeSeyler's Z Physiol Chem 1976;357:1793-800. 20. Tosh D, Agius L. Glycogen degradation by adrenergic agonists and glucagon in periportal and perivenous rat hepatocyte cultures. Biochim Biophys Acta 1994;1221:238-47. 21. Tosh D, Alberti KGMM, Agius L. Glucagon regulation of gluconeogenesis and ketogenesis in periportal and perivenons rat hepatocytes. Biochem J 1988;256:197-204. 22. Shiota M, Inagami M, Fujimoto Y, Moriyama M, Kimura K, Sugano T. Cold acclimation induces zonal heterogeneity in gluconeogenic responses to glucagon in rat liver lobule. Am J Physiol 1995;268:El184-91. 23. Berthoud VM, Iwanij V, Garcia AM, Saez JC. Connexins and glucagon receptors during development of rat hepatic acinus. Am J Physiol 1992;263:G650-8. 24. Sonne O, Berg T, Christoffersen T. Binding of 125I-labeled glucagon and glucagon stimulated accumulation of adenosine 325"-monophosphate in isolated intact rat hepatocytes. J Biol Chem 1978;253:3203-10. 25. Zierz S, Jungermann K. Alteration with the dietary state of the activity and zonal distribution of the glucagon-, fluoride, and forskolin-stimulated adenylate cyclase in microdissected rat liver tissue. Eur J Biochem 1984;145:499-504. 26. Runge D, Jungermann K. Distribution of cyclic AMP phosphodiesterase in microdissected periportal and perivenous rat liver tissue with different dietary states. Histochemistry 1991 ;96:87-92. 27. Saito K, Yamatani K, Manaka H, Takahashi K, Tominaga M, Sasaki H. Role of Ca2+ on vasoactive intestinal peptideinduced glucose and adenosine 3",5"-monophosphate production in the isolated perfused rat liver. Endocrinology 1992;130:2267-73. 28. Keppens S, De Wulf H. Periportal and perivenous hepatocytes respond equally to glycogenolytic agonist. FEBS Lett 1988;233:47-50. 29. Aggarwal SR, Lindros KO, Palmer TN. Glucagon stimulates phosphorylation of different peptides in isolated periporatal and perivenous hepatocytes. FEBS Lett 1995;377:439-43. 30. Thurman RG, Kauffman FC. Subcellar compartmentation of pharmacologic events (SCOPE): measurement of metabolic fluxes in periportal and pericentral regions of the liver lobule with microfluorometric and micropolarographic techniques. Hepatology 1986;5:144-51. 31. Laguens RP, Candela S, Hernandez RE, Galiardino JJ. Streptozotocin-induced liver damage in mice. Horm Metab Res 1980;12:197-201. 32. Yamatani K, Marubashi S, Wakasugi K, Saito K, Sato N,

J Lab Clin M e d Volume 132, Number 6

Takahashi K, et al. Catecholamine-induced cAMP response in streptozotocin-induced diabetic rat liver. Tohoku J Exp Med 1994;173:311-20. 33. Appel MC, Like AA, Rossini AA, Carp DB, Miller TB. Hepatic carbohydrate metabolism in the spontaneously diabetic Bio-Breeding Worcester rat. Am J Physiol 1981;240:E83-7. 34. Jungermann K. Dynamics of zonal hepatocyte heterogeneity. Perinatal development and adaptive alternations during regeneration after partial hepatectomy, starvation and diabetes. Acta Histochem 1986;32(suppl):89-98.

Ikezawa et al

,555

35. Miethke H, Wittig B, Nath A, Jungermann K. Gluconeogenic-glycolytic capacities and metabolic zonation in liver of rats with streptozotocin, non-ketotic as compared to alloxan, ketotic diabetes. Histochemistry 1986;85:483-9. 36. Wimmer M. Phosphoenolpyruvate carboxykinase activity patterns in the liver acinus of diabetic and estrogen treated ra~s. Histochemistry 1989;93:49-53. 37. Probst I, Schwartz P, Jungermann K. Induction in primary culture of gluconeogenic and glycolytic hepatocytes resembling periportal and perivenous cells. Eur J Biochem 1982;126:271-8.