Fish Physiology and Biochemistry vol. 8 no. 4 pp 299-309 (1990) Kugler Publications, Amsterdam/Berkeley

The role of glycogen phosphorylase in the regulation of glycogenolysis by insulin and glucagon in isolated eel (Anguilla rostrata) hepatocytes Glen D. Foster and T.W. Moon Department of Biology, University of Ottawa, Ottawa, Ontario, KIN 6N5, Canada Keywords: glycogen phosphorylase, glycogenolysis, American eel, Anguilla rostrata, hepatocytes, insulin, glucagon

Abstract The efects of porcine, scombroid, and salmon insulins, and bovine and anglerfish glucagons on glycogen depletion and glycogen phosphorylase (GPase) activities were examined in freshly isolated American eel (Anguilla rostrata) hepatocytes. Eel liver GPase in crude homogenates was activated (increase in %GPase a) by phosphorylating conditions and was rapidly inactivated (less than 1 h) when a phosphatase inhibitor (fluoride) was absent. Caffeine inhibits, and AMP activates, the b form of GPase consistent with their effects on rat liver GPase. Both mammalian and fish glucagons increased glucose production in eel hepatocytes, but had more ambiguous effects on glycogen levels and GPase activities. The magnitude of bovine glucagon effects were dependent on the initial glycogen content of the cells; only at glycogen concentrations less than approximately 70 pzmoles.g - ' did glucagon significantly increase %GPase a. Anglerfish glucagon significantly increased cyclic AMP (cAMP) concentrations by 90% at 10- 7 M, but had no effects at 10- 9 M and 10-8 M. Scombroid and salmon insulins maintained hepatocyte glycogen concentrations and decreased glucose production, with these effects more pronounced at low (10- 9 to 10-8 M) rather than high (10 - 7 M) hormone concentrations. Porcine and salmon insulins decreased total GPase and %GPase a activities, and salmon insulin decreased cAMP levels, but only at 10-8 M (by 44%). Glycogen is, therefore, depleted by glucagon and maintained by insulin in freshly isolated American eel hepatocytes, and these changes are accomplished, at least in part, by changes in the activities of GPase. Changes in cAMP do not explain all of the observed hormone effects.

Introduction Mammalian liver glycogen phosphorylase (GPase) exists in two interconvertible forms, the active phosphorylase a (GPase a) and the less active phosphorylase b (GPase b). This interconversion is regulated by glucose and a number of hormones, including insulin and glucagon, and involves reversible phosphorylation (Hems and Whitton 1980). Inter-

convertible forms of GPase have not been directly demonstrated in teleost liver, although GPase activities can be altered by varying the phosphorylation conditions of crude rainbow trout liver homogenates (Vernier and Sire 1978a). The hormonal regulation of GPase has been poorly studied in teleosts. Stimulation of GPase activity by glucagon and epinephrine has been reported in the liver of trout (Salmo gairdneri: Vernier

Correspondence to: Dr. T.W. Moon, Dept. of Biology, University of Ottawa, Ottawa, Ontario, KIN 6N5, (613) 564-2336.

300 and Sire 1978), catfish (Ictalurus sp.: Ottolenghi et al. 1986, 1988), killifish (Fundulus heteroclitus: Umminger and Benziger 1975; Umminger et al. 1975), and salmon (Oncorhynchus tshawytcha, 0. kisutch: Plisetskaya et al. 1989). Epinephrine, however, did not alter GPase activities in the liver of the carp (Cyprinus carpio) (Demael-Suard and Garin 1970), a fish that contains an amylase with activities exceeding that of GPase (Murat 1976; Pikucans and Umminger 1979). A glucagon-stimulated increase in GPase activities is consistent with its recognize effects on liver glycogen content in teleosts, both in vivo (e.g. Chan and Woo 1978; Plisetskaya et al. 1989) and in vitro (e.g. Morata et al. 1982; Foster and Moon 1987, 1989). The effects of insulin on this enzyme have not been studied to our knowledge. Mammalian insulin has a generally anabolic effect on glycogen content in isolated hepatocytes of teleost species (Plisetskaya et al. 1984; Foster and Moon 1987, 1989). In addition, insulin increases C3 precursor and glucose fluxes to glycogen in both Japanese eel hepatocyte primary cultures (Hayashi and Ooshiro 1985) and freshly isolated sea raven, Hemitripterusamericanus(Foster and Moon 1987) and American eel (Foster and Moon 1989) hepatocyte suspensions. Studies of the role of teleost insulin on glycogen metabolism have not been conclusive. Plisetskaya et al. (1984) reported greater glycogen content in coho samon hepatocytes incubated in the presence of salmon insulin, while no effects on this parameter were found in sea raven hepatocytes using a scombroid insulin (Foster and Moon 1987). This study was undertaken to characterize American eel liver GPase, and to determine the effects of glucagon and insulin on these properties in vitro. Changes in the phosphorylation status of GPase were monitored in crude homogenates, and in the presence of dibutyryl cAMP, glucagon, and insulin in freshly isolated hepatocytes. The criteria of Stalman and Hers (1975) were applied to separate the GPase forms of the eel liver enzyme, and the effects of both kinase (EDTA and EGTA) and phosphatase (F-) inhibitors on these forms were assessed in order to develop a reliable and meaningful GPase assay for teleost species.

Materials and methods Animals Immature American eels, A. rostrata (LeSueur), were obtained from the St. Lawrence River at the Saunders Hydro Electric Dam (Cornwall, Ont.) in August. The fish were held in flowing dechlorinated Ottawa tap water that followed seasonal fluctuations in temperature. Enzyme characterization was performed the following February (6 mo. following capture; water temperature = 6C) and hepatocyte experiments were carried out either in September (3 mo. following capture, temperature = 15C) or in April (8 mo. following capture, temperature = 9-10°C). Eels were not fed during captivity and were maintained on a 12L/12D photoperiod. The extent of food-deprivation is not inconsistent with that observed in nature.

Chemicals Bovine glucagon and porcine insulin were obtained from Eli Lilly Research Laboratories (Indianapolis, IN). Anglerfish glucagon was provided by Dr. P.C. Andrews (Purdue University), coho salmon insulin by Dr. E.M. Plisetskaya (University of Washington), and scombroid insulin from Novo Laboratories (Toronto, Ont.). All hormones were pure except the scombroid insulin, which is a mixture of insulins from scombroid species eluting as a broad peak on Sephadex G-25 (Fig. 1). Hormones were dissolved in 0.001 N HCI at 1 mg/ml, diluted in appropriate buffers, and then frozen in aliquots until use. All other biochemicals were purchased from either Boehringer-Mannheim (Lachine, PQ) or Sigma Chemical Co. (St. Louis, MO). Other reagents were from local suppliers and were of the highest available purity.

Characterizationof glycogen phosphorylase The effects of AMP and caffeine on GPase a and b were determined using homogenates prepared from livers rapidly excised from stunned and de-

301 50 '0

Insulin

genized. These homogenates were gel filtered (as above) and assayed for GPase activities.

GlucIgon

40

E o ._ 0

a-

0

I

0

30 20

o0

Hepatocyte incubations

0

10

32

36

40

44

48

52

56

60

64

68

Fraction Number Fig. 1. The elution of scombroid insulin (dotted line) on Sephadex G-25 (medium beads; I x 25 cm column). One mg of protein 1 was applied to the column and eluted at I ml.h- ; each fraction represents a 30 min collection period. Standards were porcine insuline and bovine glucagon (solid line) applied together each at 1 mg.

capitated eels, and homogenized (Polytron PCU-2; Brinkman Instruments) in either a complete "stopping buffer", after Stalman and Hers (1975), consisting of 10 mM HEPES, 4.5 mM 3-mercaptoethanol, 0.18 mM phenyl methyl sulphonyl fluoride (PMSF), 10 mM EDTA, 4.5 mM EGTA, and 100 mM KF, adjusted to pH 7.8, or a F--free buffer (F- inhibits phosphatase activities; Stalman and Hers 1975). The resulting homogenate was centrifuged at 12,000 x g for 10 min and the supernatant was applied to and was eluted from a Sephadex G-25 column (medium beads; 1 x 25 cm column) equilibrated with the equivalent buffer to eliminate small molecules; GPase was assayed in the initial protein-containing fractions. F--free homogenates were not assayed earlier than 1 h following tissue preparation, a time when GPase activities were found to be at a minimum. The effects of phosphorylation conditions on GPase activities were assessed in crude liver homogenates. Livers were collected as above and homogenized in complete "stopping buffer", F--free buffer, or inhibitor-free buffer (no F- or EDTA/ EGTA), and centrifuged. The inhibitor-free homogenate was incubated with or without 2 mM ATP, 2 mM MgCI 2, and 1 mM cAMP at 100 C for 20 min. Following incubation, KF and EDTA/EGTA were added to give a final concentration of 100 mM and 10/4.5 mM, respectively to stop further changes in the enzyme, and the sample re-homo-

Eels were anaesthetized in buffered MS-222, the hepatic portal vein cannulated, and the liver perfused and hepatocytes prepared by collagenase treatment according to Foster and Moon (1989). Cells were resuspended in a medium containing 110 mM NaCI, 3 mm KCI, 1.25 mm KH 2PO 4, 5 mM NaHCO 3, 0.6 mM MgSO 4, 1 mM MgC1 2, 1 mM CaC1 2, 10 mM HEPES, 1.5% fatty acid-free bovine serum albumin, and adjusted to pH 7.8 following aeration with 99.5% 02, remainder CO 2. Cell wet weight varied between experiments, ranging from 25 to 80 mg.ml- l . Cells were divided into vials containing the stated concentrations of hormones and were incubated at 10°C for 15 min (cAMP assay), 45 min (GPase assay), or 2 h (glucose/glycogen assays). Following the 15 min incubation, the cells were centrifuged, decanted, the pellet frozen in liquid N2, and held at -800 C until assayed for cAMP. The cell pellet was sonicated in 0.7% PCA containing 5 mM EDTA and was assayed for cAMP content using the Amersham 3 H-cAMP kit (Amersham Canada Ltd., Oakville, Ont.). Following the 45 min incubation, the cells were centrifuged at 1500 x g for 90 sec, the supernatant removed, and the pellet homogenized immediately in the complete "stopping buffer" and assayed for GPase a and total GPase activity. The 2 h incubations were collected and assayed for glycogen and glucose (Foster and Moon 1989).

Glycogen phosphorylase assay Glycogen phosphorylase (EC 2.4.1.1) was assayed in 67 mM phosphate buffer (pH 7.1) containing 0.35 mM NADP + , 0.4 ,aM glucose-1,6-bisphosphate, 15 mM MgCI 2, 2 mg.ml - l dialyzed glycogen, and excess dialyzed phosphoglucomutase and glucose-6- phosphate dehydrogenase. Total GPase is defined as the activity in the presence of 2.5 mM

302 Table 1. The effects of AMP and caffeine on eel liver glycogen phosphorylase activities. assay condition

glycogen phosphorylase activity (mU/mg protein) complete buffer

F--free buffer

- AMP/- caffeine (a + b)

2.60 (2.45-2.75)

0.16 (0.09-0.22)

+ AMP/- caffeine (Total)

3.23 (3.13-3.30)

0.58 (0.46-0.71)

- AMP/ + caffeine (a, no b)

1.83 (1.65-1.98)

0.003 (0-0.05)

+ AMP/ + caffeine (true a)

2.03 (1.45-2.60)

0.02 (0.02-0.02)

Eel livers were homogenized in either a complete "stopping buffer" or a F--free buffer (see Materials and Methods) and assayed a minimum of 1 h following homogenization. Assays are as described in the Materials and Methods. Values are mean (range) of 3 separate preparations. AMP and caffeine concentrations were 2.5 mM and 5 mM, respectively.

AMP but without caffeine, and GPase a activity as the activity with 2.5 mM AMP and 5 mM caffeine according to Stalman and Hers (1975); GPase b is defined as the difference between these two estimates. Activities were monitored as an increase in the absorbancy of NADP + at 340 nm using either a Unicam SP 1800 or a Unicam PU 6 spectrophotometer connected to a Zipp and Konen chart recorder; cuvette temperatures were held at 12°C. Activities are presented as munits (nmoles NADPH produced/min) per mg protein at 120 C. Protein was determined according to Lowry et al. (1956).

Statistics Significant differences between means were determined using paired or unpaired Student t-tests, depending on the experimental design. Significant differences between %/oGPasea activities were determined using the Mann-Whiney U-test, and significant correlations were determined using the Pearson r-test.

Results Characterizationof GPase Preliminary experiments established that glycogen phosphorylase activities (-AMP/-caffeine) in a F--free buffer fall to 6% of those in a complete buffer within 1h. This is consistent with the conversion of the active GPase a to the inactive GPase b as reported for rat liver GPase (Stalman and Hers 1975). The residual activity in F--free buffer is stimulated by AMP, but this stimulation is completely blocked by caffeine (Table 1), consistent with this activity being GPase b (Stalman and Hers 1975). In the "complete buffer" and assayed with caffeine, AMP stimulates apparent GPase a activities by 11 %. Higher caffeine and AMP concentrations did not alter the magnitude of these effects, and all effects are consistent with the characterization of the GPase system in the rat liver by Stalman and Hers (1975). As AMP stimulates GPase a by only 11%, and caffeine inhibits GPase b, the combination of AMP and caffeine gives an estimate of %GPase a of 63 % compared to 81% in the absence of caffeine. Assaying of GPase activities in the presence of these modifiers provides a more accurate estimate of GPase a and b in this species as previously shown in rats by Stalman and Hers (1975), and all further GPase activities reported here use caffeine and AMP routinely. Glycogen phosphorylase activities are affected by pH (Fig. 2). As pH increases, total GPase and %GPase a values decline, with the latter decreasing from 74% at pH 6.8 to 55% at pH 7.5. GPase b values are low and are essentially unaffected by pH changes. When livers are rapidly removed from eels and homogenized in the complete "stopping buffer", GPase is primarily in the a, or phosphorylated, form (77%, Table 2). After only a 20 min incubation in the inhibitor-free buffer, %GPase a activities are reduced to 40%, or to activities equivalent to that after a 6 h incubation in the same buffer. Thus, the enzyme rapidly falls into a less active (dephosphorylated) state in the absence of F- and EDTA/EGTA. Mg2 +-ATP + cAMP added to the inhibitor-free incubates counteracts this inactiva-

303 J

E .

were correlated with initial hepatocyte glycogen contents in freshly isolated eel hepatocytes (Fig. 3).

_

7

,

5.0-

I

O-o

CD

\

U) 2.5rK

Glucagon effects

70

.\

Bovine glucagon increased total glucose production by freshly isolated eel hepatocytes in a dose-dependent manner, while its negative effect on glycogen content leveled off at high hormone concentrations (Fig. 4). Anglerfish glucagon had equivalent effects on these parameters, although effects on glycogen content were small. The effect of bovine glucagon on glycogen content of eel hepatocytes was greatest at low mean initial hepatocyte glycogen contents, as demonstrated most strikingly in Fig. 5A. This observation explains the larger effect of the hormone on this parameter in 1986, when the glycogen content was 6.4 compared to 38 /tmoles.g - 1 in 1988 (Fig. 4). Anglerfish glucagon effects were absent (Fig. 4) and the liver glycogen content in these experiments was 112 /imoles.g - 1. It also appears that the sensitivity to this effect is seasonally and/or glycogen-dependent (Fig. 5A). Glycogen phosphorylase activities were not significantly affected by bovine glucagon, even though mean values are higher (Table 3). However, the glucagon effect on %GPase a is significantly

cD 60 0

o

o\

I O 0

U) o

Q" T nn _ ._

A _a --

6

. _ ,m ._..

_

7

AA _ --

-JU .

8

pH Fig. 2. The effect of pH on total GPase activity ( o , with AMP), GPase b activity (A), %oGPase a activity (), and GPase b activity with AMP (a). Assays were performed in liver homogenates prepared in either a complete "stopping buffer" (GPase a and total estimates) or an inhibitor-free buffer (GPase b estimates) followed by gel-filtration as described in the Materials and Methods.

tion, and their addition in an EDTA/EGTA-free buffer (with F-) results in activation of the enzyme above the initial state (89%) and to activities equivalent to total GPase activities (Table 2). No correlation was noted between initial glycogen content and GPase activities, even though both glucose production and glycogen depletion rates

Table 2. Total glycogen phosphorylase (GPase) and GPase a activities (mU/mg protein (SE)) in crude liver homogenates from 6 month fasted eels incubated under the stated conditions. 20 min incubation

6h phosphorylating conditions

complete buffer

inhibitor free

inhibitor free

with F-

inhibitor free

n total GPase

5 3.40 (0.33)

3 1.781 (0.18)

3 2.53 (0.28)

3 3.5323 (0.28)

2 0.85

GPase a

2.70 (0.35)

0.751 (0.40)

1.45 (0.48)

3.1523 (0.23)

0.35

%7oGPase a

77.3

39.91

67.22

88.9123

38.6

Phosphorylating conditions refer to incubating the liver homogenate with 1 mM cAMP and 2 mM Mg-ATP. The 20 min and 6 h incubations were performed at 10°C. Incubations were terminated by the addition of missing buffer component(s). Significant differences (p < 0.05) determined by Student t-tests (total GPase and GPase a) and the Mann-Whitney U-test (%GPase a). i significantly different from complete buffer; 2 significantly different from inhibitor-free (no cAMP); 3 significantly different from inhibitor-free (phosphorylating conditions).

304

an .

o,

A

z

60April

0 o

00

50

30 0 I

0-J r =0.72 P 0.001 O

0

0 80

100

so

so

200

-

00

0 -9 l0 k 10 -8 M 10 -7 k

5.

April~~~~~~; =07

w-I,~~ ·

70

o September

60

*April

[

0 0

U, 0 DC,

40 30 20

2.4

0 0

100

bovine glucagon 1986

10

M 10-

bovine glucogon 1988

M

10

M

M AF g ucagon 1988

C

50 -

30

20

Z

2

10O

6,

B

40

O

40

.5 a

A

0

0 September

70-

E

'3

50

200

00

100 200 300 INITIAL GLYCOGEN CONTENT (moles/g)

40 400

Fig. 3. The dependence of glycogen depletion (A) and glucose production (B) on initial glycogen content of freshly isolated eel hepatocytes in April (6 mo. fasted, ) and September (3 week fasted, o). Parameters were measured following a 2 h incubation and significance was determined by the Pearson r-test.

Fig. 4. The effects of bovine and anglerfish glucagons on glucose production (A) and end glycogen content (B) in freshly isolated eel hepatocytes. Control (no hormone) glucose production rates were 22±4 (bovine glucagon, 1986), 11 ± 4 (bovine glucagon, 1988), and 22±3 zmoles.h-l.g- 1 (anglerfish glucagon, 1988). Control glycogen contents (after incubation) were 6.4 ± 1.4 (bovine glucagon, 1986), 38 ± 7 (bovine glucagon, 1988) and - I 112 35 /umolesh.h- I .g (anglerfish glucagon). *Significant differences from control (p < 0.05) using the Mann-Whitney U-test.

cn

o

O September * April

A 80

a,

correlated with initial glycogen content (Fig. 5B); glucagon is without effect when the initial glycogen content is greater than approx. 70 tmoles.g-l. This is consistent with the correlation between initial glycogen content and the glucagon effect on glycogen content (Fig. 5A), and implicates this enzyme as an important component of the mechanism whereby glucagon alters liver glycogen content. In fact, a significant correlation is observed between the magnitude of the glucagon effect on GPase a and the glucagon effects on both glycogen levels (p < 0.001) and glucose production (p < 0.005) (data not shown) as predicted from the relationships noted in Fig. 5. Anglerfish glucagon was without effect on GPase activities (Table 3), consistent with the high glycogen content of the cells (Fig. 4). Correlations between the various parameters and anglerfish glucagon were not possible as the data were not paired. Anglerfish glucagon significantly increased cAMP

0

60

9,

40

Q

aC

63

20

I

0

\

Y -0.69

y -0.16x +70 ,.-0.74

\P (0.05 p

p 0.02

-20

0

aw----·

-20

100

200

300

400

0 a

30

s

25 o.

y =- -0.30x 2 r = 0.77

20 C)

15

t 22.8

p (0.005

a

a o 0

CL

0 . -5 0

20

40

60

80

100

INITIAL GLYCOGEN CONTENT (moles/g)

Fig. 5. The correlation between initial glycogen content and the glucagon effects (as %ochange from control) on glycogen deple3 tion (A) and %oGPase a (B) in eel hepatocytes isolated in September ( o, n = 6) and April ( , n = 8). Levels of significance determined using the Pearson r-test.

305 Table 3. Glycogen phosphorylase activities in freshly isolated eel hepatocytes and the effects of dibutyryl cAMP (db-cAMP), bovine glucagon, and anglerfish glucagon. glucagon

db-cAMP

anglerfish

bovine 10- 4 M

OM n

6

6

OM

10- 8 M

16

13

OM 8

10-8 M 8

total GPase

3.23 (0.36)

3.701 (0.47)

2.84 (0.26)

3.03 (0.50)

3.02 (0.45)

3.06 (0.43)

GPase a

2.82 (0.43)

3.611 (0.18)

2.41 (0.24)

2.72 (0.42)

2.65 (0.39)

2.74 (0.38)

0oGPase a

86.7

95.11

84.5

89.8

88.9

91.2

Incubations proceeded for 45 min and were terminated by homogenization in the "stopping buffer" (see Materials and Methods). Bovine glucagon values are means of assays performed in 1986 and 1988. Values represent mean activities (m-units/mg protein) (SE). Significant differences (p < 0.05) were determined using the paired or unpaired (bovine glucagon) Student t-test (total and a) or the Mann Whitney U-test (o a). t significantly different from no addition.

B

A -J

0o z

10

Do U

0 0

o-

-1 V-

_o _

10 Minsulin 10 M I porcineM

m porcine

insulin

10 10

M M

-, 10insulin

E~ Scombroid insulin

_7 _R -, l- insulin salmon A

_

so~elmon insulin

Fig. 6. The effects of porcine, scombroid and salmon insulins on glucose production (A) and end glycogen content (B) in freshly isolated eel hepatocytes. Control (no hormone) glucose production rates were 34 ± 10 (porcine insulin, 1986), and 22±+ 3imoles.h- l .g - l (salmon insulin and scombroid, 1988). Control glycogen contents were 9.7 + 3.1 (porcine insulin, 1986), and 112 + 35 /Amoles.g-l (salmon and scombroid insulin, 1988).

levels from 0.1 1 ± 0.06 to 0.17 ± 0.06 nmoles.g- (p < 0.05, n = 6) in hepatocytes incubated for 15 min at 10- 7 M, but did not affect its concentration at either 10 - 9 or 10-8 M. Dibutyryl cAMP significantly activated GPase activities (Table 3.). Insulin effects Salmon and scombroid insulins prevented glycogen loss from freshly isolated eel hepatocytes, although

the effect was reduced and was reversed at high hormone concentrations (Fig. 6B). Porcine insulin did not significantly alter glycogen content at the three concentrations tested. Total glucose production rates were generally decreased by incubating hepatocytes with insulin, although the effect was dependent upon both the concentrations and type of insulin used (Fig. 6A). Both salmon and porcine insulins decreased %0GPase a in freshly isolated eel hepatocytes (Table 4) and the porcine hormone significantly depressed the high %GPase a values noted with glucagon (c.f. Tables 3 and 4). Salmon insulin decreased cAMP levels in the hepatocytes at 15 min from 0.11 ± 0.06 to 0.05 + 0.05 nmoles.g- ' (p < 0.05, n = 6) at 10-8 M, but was without effect at 10-9 and 10 -7 M (April experiments). Discussion Incubation of eel liver homogenates in the absence of the inhibitors F- and EDTA/EGTA results in a rapid inactivation of GPase. The activity of the less active form is virtually arrested by caffeine, and is stimulated to 20% of the maximal rate (i.e. complete buffer with AMP) by AMP. Caffeine also blocks the AMP stimulation of GPase b. The omission of caffeine results in elevated estimates of

306 Table 4. Glycogen phosphorylase activities in freshly isolated eel hepatocytes and the effects of porcine insuline and salmon insulin, and porcine insulin + bovine glucagon. salmon insulin

porcine insulin

n= total GPase GPase a W7oGPase a

insulin + glucagon

OM

10-8 M

OM

8 10- M

OM

10- 8 M

4 2.35 (0.53) 1.95 (0.49) 81.8

4 1.961 (0.53) 1.441 (0.43) 72.31

9 2.95 (0.40) 2.59 (0.35) 89.3

9 2.65 (0.37) 2.20 (0.35) 83.61

5 3.16 (1.37) 2.71 (1.13) 88.2

5 2.61 (0.85) 1.92 (0.60) 74.5 L

Incubations were 45 min at 10 0C. Conditions and statistics as in Table 3. i significantly different from no addition.

%GPase a. These findings indicate that for eel liver, and probably other teleost species, GPase a activity must be estimated by using the criteria of Stalman and Hers (1975). In addition, these results directly implicate an interconvertible phosphorylation-dephosphorylation system in the regulation of eel liver GPase. This is the first study of a teleost liver GPase system which has used caffeine and AMP in the assay medium to separate a and b forms of the enzyme. This may indicate that in other studies (e.g. Umminger and Benziger 1975; Vernier and Sire 1978a, 1978b; Ottolenghi et al. 1986) GPase a activities had been overestimated. Both agents should be included for accurate estimates of this enzyme. Also, desalting with Sephadex is necessary to remove any low molecular weight modifiers, such as AMP, that could affect GPase activity estimates. The amount of the enzyme in the active form or GPase a is high, ranging from 77-90% under control conditions (Tables 2, 3 and 4). Even though the animals used in this study are 6-8 months starved (a condition experienced in nature by eels) these values are consistent with other liver GPase studies in catfish (Umminger and Benziger 1975; Ottolenghi et al. 1986, 1988), goldfish (Storey 1988) and rat (e.g. Schwartz and Rall 1973; Stalman and Hers 1975; Katz et al. 1979). Eels do not exhibit rapid depletion of liver glycogen in vivo under conditions including fasting (Renaud and Moon 1980), even though %GPase a activities are high. One explanation for these very high %GPase a values is that rapid phosphorylation of this enzyme occurs during

preparation. However, the time to remove the liver from the eel (including capture from the tank) and its homogenization in the "stopping buffer" is less than 2 min. We find it unlikely that these animals could so rapidly activate this enzyme system. We are forced to conclude that these high %GPase a values reflect in vivo levels of this enzyme in the eel liver. Similar high activities without rapid glycogen depletion have been reported in mammals (Hems and Whitton 1980). To explain this paradox, Hems and Whitton (1980) suggested that the enzyme must be inhibited in vivo by tissue modulators that may include ATP, fructose-l-P, and inorganic phosphate. A similar condition may exist in the eel. However, if the inhibitory agents are the rate determining components of the glycogen depletion system, it is intriguing that a significant correlation exists between assayable GPase activity and whole liver glycogen content in the rat (Hems and Whitton 1980), the eel (this study), and the perch, Perca flavescens (Foster and Moon, unpublished data). Glucagon stimulates liver GPase activities in catfish (Umminger and Benziger 1975; Ottolenghi et al. 1988), trout (Vernier and Sire 1978), and killifish (Umminger et al. 1975), and this is associated with a decrease in glycogen content in catfish liver pieces or slices (Umminger and Benziger 1975; Ottolenghi et al. 1988). Glycogen content is also decreased by bovine glucagon in hepatocytes of the sea raven (Foster and Moon 1987) and the American eel (Foster and Moon 1989; this study). These studies clearly demonstrate a glycogenolytic effect of glucagon. In the present study, however, while db-

307 cAMP increased GPase activities, neither bovine nor anglerfish glucagons significantly altered the already high GPase activities (Table 3), and significant effects on glycogen content were only noted with bovine glucagon at a single concentration (10 - 8 M; Fig. 4). It may be that glucagon is affecting the cell content of inhibitors that block GPase activity (Hems and Whitton 1980). Further analysis of the data found a correlation between glycogen content and the capacity of glucagon to increase %GPase a such that glucagon was only effective at a glycogen content below approx. 70 ltmoles.g - 1 (Fig. 5B). A general characteristic of freshly isolated teleost hepatocytes is a negative glycogen balance resulting in the continuous production of glucose (Fig. 3, Moon and Mommsen 1986). The production of glucose can interfere with variety of biochemical parameters, including gluconeogenesis (Mommsen 1986), and the concentration of fructose 2,6-bisphosphate, and activator of glycolysis (Foster et al. 1989). Glucose may also be important in the hormonal regulation of glycogen metabolism, interfering directly with hormone action, such as the well known inhibitory effect of glucose on GPase in the rat liver (Hers 1976). The greater amount of glucose liberated by the cells containing more glycogen (Fig. 3) may be antagonizing the action of glucagon, thereby resulting in no change in GPase activities. This dependence of the glucagon effect on the initial glycogen content is also reflected in the correlation of initial glycogen content and the extent of the glucagon effect on glycogen depletion (Fig. 5A) and the poor glucagon response when mean liver glycogen content is high (Fig. 4). Anglerfish glucagon increased cAMP levels in the eel hepatocytes at 10- 7 M. This is an interesting result given that glucagon isolated from A. rostrata contains a glutamine instead of an acidic residue at position 3 (Conlon et al. 1988). Except for alligator gar glucagon, all other teleost glucagons examined to date, including anglerfish glucagon (Andrews et al. 1986), contain an acidic residue at this position and are without effects on cAMP levels in livers of mammals, whose glucagon contains glutamine at this position (see Conlon et al. 1988). The exact significance of these different structure-function and

species comparisons must await further reports on cAMP content after applications of homologous teleost peptide hormones. No definitive statement can be made at this time. Wakelam et al. (1986) proposed a dual mechanism for glucagon action in rat hepatocytes involving the two second messengers, inositol phosphate and cAMP, operating in a reciprocal manner. The inositol-P mechanism in the rat liver takes precedence at low hormone concentrations (10-10 to 10-8 M). The contrasting effects of glucagon found in this study compared to previous studies where significant effects of glucagon on GPase were noted (Umminger and Benziger 1975; Vernier and Sire 1978; Ottolenghi et al. 1988) may be related to the lower concentration of hormone employed in our study (10-8 M), where inositol-P may be the second messenger. In fact, we found that anglerfish glucagon did not increase hepatocyte cAMP content at concentrations below 10- 7 M, whereas significant effects of the hormone on glucose production were apparent at hormone concentrations as low as 10 - 9 M. Similar concentration-dependent effects on glycogenolysis have been found in goldfish, Carassius auratus, hepatocytes (Birnbaum et al. 1976). Dibutyryl cAMP did significantly increase GPase activities in eel hepatocytes (Table 3), so obviously a cAMP-dependant GPase mechanism exists in eel hepatocytes. One could speculate further that inositol-P could alter cellular modulators of GPase, allowing its already high activities to deplete glycogen; this would be similar to the suggestion of Hems and Whitton (1980) for the rat liver. Alternatively, cAMP may respond differently at different hormone concentrations, or 15 min is not the time period of maximum cAMP concentration. Mommsen and Moon (1989), however, have shown that cAMP levels in eel hepatocytes peak within 10-20 min of hormone addition (10-8 M glucagon and epinephrine). Previous studies have found differential effects of various insulins in sea raven (Foster and Moon 1987), salmon (Plisetskaya et al. 1984) and eel (Foster and Moon 1989) hepatocytes. The present study finds that both mammalian and teleost (salmon and scombroid) insulins enhance, or at least maintain, hepatocyte glycogen content (Fig. 6) and

308 decrease GPase activities (Table 4). While seasonal changes in hormone sensitivity may be a factor (Mommsen et al. 1987; Foster and Moon 1987, 1989), differences in cell glycogen content or incubate glucose concentrations could also modify the responsiveness to insulin. Foster and Moon (1989) found larger effects in the summer, when hepatocyte glycogen content was high, and therefore glycogen depletion and glucose production rates were higher, as compared to the winter and spring. As suggested by Freig et al. (1985), glycogen itself may be modifying the responsiveness of hepatic glycogenesis to insulin. They report an inverse correlatin between hepatocyte glycogen content and insulin effects on glucose incorporation into glycogen, consistent with our findings of greater insulin effects on glycogen content at higher glycogen concentrations (e.g. summer vs. winter and spring). Alternatively, glucose may be necessary for insulin action on glycogen metabolism in the rat liver (Ottolenghi et al. 1984), and the higher glucose concentrations resulting from the greater glycogen depletion rates may be potentiating insulin action. Further studies using cell systems where glucose levels and glycogen depletion rates can be controlled are required to determine the interactions of glucose, glycogen, and hormones in the regulation of glycogen metabolism. There has been some controversy as to whether insulin has an effect on glycogen metabolism beyond that of antagonizing the glucagon effect (Hems and Whitton 1980). The present study demonstrates an insulin effect on glycogen content (Fig. 6) and GPase activities by insulin alone as well as in the presence of glucagon (Table 4). These data suggest, therefore, that insulin is capable of both a direct (i.e. tyrosine-kinase phosphorylation) and possibly an indirect (i.e. second messenger) effect on glycogen metabolism, with the latter operating through a cAMP-dependent process (Goldfine 1987). Insulin alone in salmon hepatocytes also conserve glycogen content (Plisetskaya et al. 1984). Studies using long-term cell cultures where hormonal conditions and the physiological state of the cells can be more readily manipulated are required to confirm the role of insulin on basal carbohydrate metabolism.

This study shows that glycogen content and glycogen phosphorylase are regulated by both insulin and glucagon in freshly isolated eel hepatocytes. Both mammalian and teleost insulins reduce glycogen depletion possible by decreasing GPase activities, while glucagon stimulates glycogen depletion and GPase activities only when cell glycogen contents are low. The relationship between glucagon effects and glycogen content suggests an involvement of glucose and/or some other modulator(s) in the observed effects. Acknowledgements The authors thank Dr. E.M. Plisetskaya (University of Washington) for providing the coho salmon insulin, Novo Laboratories for the scombroid insulin, Dr. P.C. Andrews (Purdue University) for the gift of anglerfish glucagon, and Dr. E.M. Root of Eli Lilly Research Laboratories for the bovine glucagon and porcine insulin. Thanks are also extended to the Ontario Ministry of Natural Resources for providing the eels and Ms. Jill Jensen for some technical assistance. This study was supported by a NSERC of Canada operating grant to TWM. References cited Andrews, P.C., Hawke, D.H., Lee, T.D., Legasse, K., Noe, B.D. and Shively, J.E. 1986. Isolation and structure of the principal products of preglucagon processing including an amidated glucagon-like peptide. J. Biol. Chem. 261: 81258133. Birnbaum, M.J., Schultz, J. and Fain, T.N. 1976. Hormonestimulated glycogenolysis in isolated goldfish hepatocytes. Am. J. Physiol. 231: 191-197. Chan, D.K.O. and Woo, N.Y.S. 1978. Effects of glucagon on the metabolism of the Japanese eel. Gen. Comp. Endocrinol. 35: 216-225. Conlon, J.M., Dacon, C.F., Hazon, N., Henserson, I.W. and Thim, L. 1988. Somatostatin-related and glucagon-related peptides with unusual structural features from the European eel (Anguilla anguilla). Gen. Comp. Endocrinol. 72: 181189. Demael-Suard, A. et Garin, D. 1970. L'interaction entre l'adrenaline et l'insuline dans la regulation du metabolisme glucidique de la tanche. C.R. Soc. Biol. 164: 1505-1509. Foster, G.D. and Moon, T.W. 1986. Cortisol and liver

309 metabolism of immature American eels, Anguilla rostrata (LeSueur). Fish Physiol. Biochem. 1: 113-124. Foster, G.D. and Moon, T.W. 1987. Metabolism in sea raven (Hemitripterusamericanus)hepatocytes: the effects of Insulin and glucagon. Gen. Comp. Endocrinol. 66: 102-115. Foster, G.D. and Moon, T.W. 1989. Insulin and the regulation of glycogen metabolism and gluconeogenesis in American eel (Anguilla rostrata)hepatocytes. Gen. Comp. Endocrinol. 73: 374-381. Foster, G.D., Storey, K.B. and Moon, T.W. 1989. The regulation of 6-phosphofructo-l-kinase by insulin and glucagon in isolated hepatocytes of the American eel (Anguilla rostrata). Gen. Comp. Endocrinol. 73: 382-389. Freig, W.E., Nother-Fleig, G., Steudter, S., Enderle, D. and Ditschuneit, H. 1985. Regulation of insulin binding and glycogenesis by insulin and dexamethasone in cultured rat hepatocytes. Biochim. Biophys. Acta 847: 325-331. Goldfine, I.D. 1987. The insulin receptor: molecular biology and transmembrane signalling. Endocr. Res. 8: 235-255. Hayashi, S. and Ooshiro, Z. 1985. Effects of glucagon, insulin, and the eel serum in the eel liver cells in primary culture. Bull. Jap. Soc. Sci. Fish. 51: 1123-1127. Hems, D.A. and Whitton, P.D. 1980. Control of hepatic glycogenolysis. Physiol. Rev. 60: 1-50. Hers, H.G. 1976. The control of glycogen metabolism in the liver. Am. Rev. Biochem. 45; 167-189. Katz, J., Golden, S. and Wals, P.A. 1979. Glycogen synthesis by rat hepatocytes. Biochem. J. 180: 389-402. Mommsen, T.P. 1986. Comparative gluconeogenesis in hepatocytes from salmonid fishes. Can. J. Zool. 64: 1110-1115. Mommsen, T.P. and Moon, T.W. 1989. Metabolic action of glucagon-family hormones in liver. Fish Physiol. Biochem. 7: 279-288. Mommsen, T.P., Andrews, P.C. and Plisetskaya, E.M. 1987. Glucagon-like peptides activate hepatic gluconeogenesis. FEBS Lett. 219: 227-232. Morata, P., Vargas, A.M., Pita, M.L. and Sanchez-Medina, F. 1982. Involvement of gluconeogenesis in the hyperglycemia induced by glucagon, adrenaline, and cyclic AMP in rainbow trout (Salmo gairdneri). Comp. Biochem. Physiol. 73A: 379-381. Murat, J.C. 1976. Studies on glycogenolysis in carp liver: evidence for an amylase pathway for glycogen breakdown. Comp. Biochem. Physiol. 55B: 461-465. Ottolenghi, C., Puviani, A.C., Baruffaldi, A. and Brighenti, L. 1984. Effect of insulin on glycogen metabolism in isolated catfish hepatocytes. Comp. Biochem. Physiol. 78A: 705-710. Ottolenghi, C., Puviani, A.C., Gavioli, M.E. and Brighenti, L. 1986. Epinephrine effect on glycogen phosphorylase activity

in catfish liver and muscles. Gen. Comp. Endocrinol. 61: 469-475. Ottolenghi, C., Puviani, A.C., Baruffaldi, A., Gavioli, M.E. and Brighenti, L. 1988. Glucagon control of glycogenolysis in catfish tissues. Comp. Biochem. Physiol. 90B: 285-290. Picukans, I. and Umminger, B.L. 1979. Comparative activities of glycogen phosphorylase and T-amylase in livers of carp (Cyprinus carpio) and goldfish (Carassius auratus). Comp. Biochem. Physiol. 62B: 455-457. Plisetskaya, E.M., Bhattucharya, S., Dickoff, W.W. and Gorbman, A. 1984. The effect of insulin on amino acid metabolism and glycogen content in isolated liver cells of juvenile coho salmon Oncorhynchus kisutsch. Comp. Biochem. Physiol. 78A: 773-778. Plisetskaya, E.M., Ottolenghi, C., Sheridan, M.A. Jr., Mommsen, T.P. and Gorbman, A. 1989. Metabolic effects of salmon glucagon and glucagon-like peptide in Coho and Chinook salmon. Gen. Comp. Endocrinol. 73: 205-216. Renaud, J.M. and Moon, T.W. 1980. Characterization of gluconeogenesis in hepatocytes isolated from the American eel Anguilla rostrata (LeSueur). J. Comp. Physiol. 135: 115-125. Schwartz, A.L. and Rall, T.W. 1975. Hormonal regulation of incorporation of U-14 C-alanine into glucose in human fetal liver explants. Effect of dibutyryl cAMP, glucagon, insulin, and triamcinolone. Diabetes 24: 650-657. Stalman, W. and Hers, H.G. 1975. The stimulation of liver phosphorylase b by AMP, fluoride, and sulphate. A technical note on the specific determination of the a and b forms of liver glycogen phosphorylase. Eur. J. Biochem. 54: 341-350. Storey, K.B. 1988. Tissue-specific controls on carbohydrate catabolism during anoxia in goldfish. Physiol. Zool. 60: 601-607. Umminger, B.L. and Benzinger, D. 1975. In vitro stimulation of hepatic glycogen phosphorylase activity by epinephrine and glucagon in the brown bullhead. Gen. Comp. Endocrinol. 25: 96-104. Umminger, B.L., Benziger, D. and Levy, S. 1975. In vitro stimulation of hepatic glycogen phosphorylase activity by epinephrine and glucagon in the killifish Fundulus heteroclitus. Comp. Biochem. Physiol. 51C: 111-115. Vernier, J.M. and Sire, M.F. 1978. In vitro study of hepatic glycogen phosphorylase in rainbow trout: its control by glucose, corticoids, adrenaline, and glucagon. Gen. Comp. Endocrinol. 34: 360-369. Wakelam, J.O., Murphy, G.J. Hruby, V.J. and Houslay, M.D. 1986. Activation of two signal transduction systems in hepatocytes by glucagon. Nature, Lond. 323: 68-71.

The role of glycogen phosphorylase in the regulation of glycogenolysis by insulin and glucagon in isolated eel (Anguilla rostrata) hepatocytes.

The effects of porcine, scombroid, and salmon insulins, and bovine and anglerfish glucagons on glycogen depletion and glycogen phosphorylase (GPase) a...
838KB Sizes 0 Downloads 0 Views