Transforming growth factor alpha inhibits glycogen synthesis and counteracts the stimulation by insulin in hepatocytes

Vol. March

183,

No.

2, 1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

16, 1992

Pages

TRANSFORMING GROWTH FACTOR ALPHA AND COUNTERACTS THE STIMULATION

560-565

INHIBITS GLYCOGEN SYNTHESIS BY INSULIN IN HEPATOCYTES

M.Peak and L.Agius* Department of Medicine, University of Newcastle upon ?)rne Newcastle upon Tyne, NE2 4HH, U.K. Received

November

19,

1991

SUMMARY Rat transforming growth factor o! (TGFol) inhibits glycogen synthesis in rat and guinea pig hepatocyte cultures and counteracts the stimulation of glycogen deposition and activation of glycogen synthase caused by insulin. The EC,, for inhibition of glycogen deposition was 0.2nM. The inhibition of glycogen synthesis was also observed in the absence of extracellular Ca” and was not blocked by indomethacin, suggesting that it is not mediated by production of prostaglandins. Since TGFcr is produced by hepatocytes during liver regeneration and by macrophages during endotoxin stimulation, it may have an autocrine / paracrine effect on hepatic carbohydrate metabolism in these states, and may account for the low hepatic glycogen levels during liver regeneration and the impaired glucose tolerance associated with sepsis. 0 1992 Academic Pre.55,Inc.

Transforming growth factor alpha (TGFol) is a 50 amino acid peptide that shares homology (30-40%) with epidermal growth factor (EGF) [1,2]. Rat TGFcx shares 34% sequence identity with mouse EGF but no immunological cross reactivity [3,4], and human TGFa is 42% homologous with human EGF [2]. The TGFa! sequence, unlike the EGF sequence, is highly conserved with only 4 out of 50 amino acid substitutions between rat and human [5]. The sequencesof the precursor proteins of rat and human TGFrr (160 and I59 amino acid residues, respectively) are also highly conserved with about 90% homology [5,6]. TGFcY and EGF bind to the same 170 kDa glycoprotein membrane receptor [7] and activate the intrinsic tyrosine kinase activity, leading to receptor autophosphorylation, internalization and down regulation [8]. The sequence of the TGFa/EGF receptor is highly conserved in different species even in birds and mammals [9]. Both similarities [1,2] and differences [lo141 in the biological effects of TGFol and EGF have been reported. Both growth factors induce cell proliferation in various cell types [1,2] and inhibit acid secretion in gastric mucosal

* To

whom correspondence

should

be addressed.

Abbreviations: TGFcr, transforming growth factor alpha, EGF, epidermal factor; MEM, Minimum Essential Medium. 0006-291X/92 $1.50 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

560

growth

Vol.

183,

No.

2, 1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

cells [15] but they have different effects on collagen synthesis and bone resorption in calvaria cells [ll] and on prostacyclin production by vascular endothelial cells [14]. TGFa is expressed by virally transformed cells and tumour cell lines [16] and also by various normal cells including hepatocytes [17]. The expression of TGFcY by hepatocytes increases several-fold after partial hepatectomy [17]. Since TGFcx stimulates the proliferation of hepatocytes [17], it is proposed to have an autocrine role in the control of liver regeneration [ 171. We report in this study that TGFa potently inhibits glycogen synthesis in hepatocytes and totally counteracts the stimulation by insulin. This suggests that TGFol may also have an autocrine/paracrine role in controlling hepatic carbohydrate metabolism during liver regeneration and possibly also during sepsis. MATERIALS

and METHODS

Mufetials: Rat TGFcx (synthetic [3]), mouse EGF (extracted from submaxillary gland), insulin and indomethacin were from Sigma Chemical CO. (St. Louis, MO, U.S.A.). Sources of other materials were as described previously [ 18,191. culture: Hepatocytes were isolated by collagenase perfusion of the liver [18] of male Wistar rats (body wt. 200-280g) or male Dunkin-Hartley guinea pigs (body wt. 300400g) that were fed on standard rat or guinea pig diet ad libitum. They were inoculated in MEM containing 5% (v/v) serum as described previously [19]. After attachment the hepatocytes were cultured for 16h in serum-free MEM with lOnM-dexamethasone [18,19]. Heparocyte

After the 16h pre-culture the hepatocytes were incubated with the concentrations of TGFcY or insulin indicated, either in MEM (without dexamethasone) containing 15mM-&I-‘“Cl-glucose (O.lCi/mol) and 2mM-pyruvate (Fig. 1 and Table 1) or with Farle’s balanced salts solution, with the modifications indicated, containing either lSmM-[U“C]glucose alone (Table 3) or with additional 2mM-pyruvate (Table 2), which causes a synergistic increase in glycogen synthesis [19]. Incubations were in duplicate for each experimental condition and were terminated after 90min or 120min as indicated [19]. Glycogen synthesis was determined from the incorporation of “C-glucose into glycogen, isolated by ethanol precipitation [ 191, and glycogen deposition from the incremental increase in cell glycogen (determined enzymically) [19], during the incubation with substrates. Cell protein was determined by an automated Lowry method [20] and glycogen synthesis is expressed as nmol of “C-glucose incorporated into glycogen per mg of cell protein and glycogen deposition as the increase in cell glycogen expressed as nmol of glucosyl units per mg of cell protein. Glycogen synthase activity was determined as in [21], with either O.lmM or 4mM-glucose-6phosphate representing synthase-a and total activity, respectively. The activity is expressed as munits per mg of cell protein where lmunit is the amount converting lnmol of UDP-glucose into glycogen per min at 30°C. Results are expressed as means + SEM for the number of hepatocyte cultures (experiments) indicated. Statistical analysis was by the Student’s paired t-test. Glycogen synthesis and deposition:

RESULTS TGFa! (SnM) inhibited glycogen synthesis (determined radiochemically and enzymically) in rat and guinea pig hepatocyte cultures by 28%-41% in the absence of insulin and by 39%~64% in the presence of lnM-insulin, and thereby counteracted the stimulation by insulin (Table 1). When the effects of TGFa! were compared in the presence of 1nM and 1OOnM insulin, the degree of inhibition was similar at these insulin concentrations (results not shown) indicating that increasing the [insulin] by loo-fold excess of the concentration that 561

Vol.

183,

No.

TABLE

2, 1992

BIOCHEMICAL

1. Effects of TGFa

AND

BIOPHYSICAL

on glycogen synthesis and deposition pig hepatocytes

Hepatocytesfrom:

Rat

39.0 f 28.0 + 58.2 + 35.4 -+

COMMUNICATIONS

in rat and guinea

Guineapig

Glycogen synthesis Additions: None SnM-TGFU lnM-insulin 1(+ 5nM-TGFo

RESEARCH

Glycogen deposition

2.6 47 + 7 2.7* 31 + 6* 4.4 101 * 11 4.7** 36 t- 4**

Glycogen synthesis

Glycogen deposition

31.6 f 20.8 + 47.8 f 27.9 -+

75 +_18 44 + 17* 124 & 11 62 -+ 16**

1.0 2.4* 4.7 2.5**

Rat and guinea pig hepatocyte cultures were incubated for 2h in MEM containing 15mM [U-‘“Clglucose and 2mM pyruvate and the additions indicated. Glycogen synthesisis expressed as nmolof “C-glucoseincorporatedinto glycogen/ 2h per mg of cell protein and glycogendepositionasnmolof glucosylunitsdeposited/ 2h per mg of protein. Valuesare means+ SEM for eachof 4 hepatocytecultures. * P < 0.05, ** P < 0.005, TGFa vs respectivecontrol, without or with insulin(pairedt-test).

causes maximal stimulation of glycogen synthesis [19] does not reverse the inhibition by TGFa. The concentration of TGFo that caused half maximal inhibition was independent of the presence of insulin: - 0.02nM for the incorporation of 14C-glucose into glycogen and -0.2nM for glycogen deposition determined enzymically (Fig. 1). TGFa also inhibited glycogen synthesis in a Ca ‘+-free medium containing O.lmM-EGTA (Table 2). However, the percentage inhibition was lower than in the Ca2+-containing medium (EGTA vs Ca*+: 12% vs 21% , inhibition in the absence of insulin and 34% vs 52 % inhibition in the presence of insulin, Table 2).

B

0 I

0

0.005

0.05

0.5

5

I

I

50

0

0.005

1. Inhibition

of glycogen deposition

0.5

5

TGFo hMI

TGFo fnMJ

FIGURE

0.05

and synthesis by TGF~Y .

Rat hepatocytecultureswere incubatedfor 2h in MEM containing15mM [LJ-‘JC]-glucose and 2mM pyruvate in the absence(0) or presence(0) of 1nM insulin and the concentrationsof TGFo indicated. Glycogendeposition(A) is expressedasnmol of glucosylunitsdeposited/ 2h per mg of cell proteinandglycogensynthesis(B) asnmolof “C-glucoseincorporatedinto glycogen/ 2h per mgof cell protein. Valuesaremeans+ SEM of 4 hepatocytepreparations. Statistical analysis(paired t-test) is relative to the correspondingcontrol or insulin values without TGFa: *, P < 0.05; ** P < 0.005. 562

50

Vol.

183,

No.

2, 1992

TABLE

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

2. Effects of extracellular Cat+ on the inhibition synthesis by TGFol

COMMUNICATIONS

of &Wen

Glycogensynthesis Media: Additions: None

SnM-TGFa lnM-Insulin ” + + SnM-TGFa

Ca” 41.2 32.7 69.7 33.2

EGTA ri: + & -t

1.1 0.8 2.5 1.8

(100) (79)** (100) (48)**

63.4 55.8 99.8 65.6

+ f &f

Ca’+IEGTA 0.9 (100) 1.9 (88)* 3.1(100) 1.1(66)*

(0.65) (0.59) (0.70) (0.51)

Rat hepatocyte cultureswere incubatedfor 2h with the additionsindicatedin Earle’s balancedsaltscontaining15mM [U-“CJglucose and 2mM pyruvate and either 2.5mM CaC12 or O.lmM EGTA. Glycogen synthesisis expressedas nmol of “C-glucose incorporated into glycogen /2h per mg of protein. Valuesare meansk SEM for 3 hepatocyte cultures in each group. Values in parentheses indicate values in the presence of TGF~Yasa percentageof the corresponding control or insulinvalues: * P < 0.05, P

< 0.005, TGFa vs respectivecontrol.

A Na+-free medium (in which 130mM-NaCl was replaced by 260mM-sucrose and the other Naf salts by their K+ salts) abolished the stimulation by lnM-insulin,

but not the

inhibition by SnM-TGFa (control medium: 25.6 + 1.5; +insulin, 35.0 + 1.0; +insulin + TGFa, 6.4 + 0.8: Naf-free medium, 18.0 + 1.2; +insulin, 16.7 f 0.4; + insulin + TGFa, 8.8 + 0.6, nmol / 90min per mg, means f S.E.M. n=4, P < 0.005, TGFa + insulin vs insulin). This indicates that Na+ entry, via Na+/H+ exchange or other mechanismsmay be essentialfor the stimulation by insulin but not for the inhibition by TGFa. Some effects of TGFa, for example stimulation of bone resorption [l l] and prcduction of prostacyclins by endothelial cells [14] have been suggestedto be mediated by production of prostaglandins because they are blocked by indomethacin an inhibitor of prostaglandin synthesis [ll, 141. Since several glycogenolytic

peptides exert their effects through the

production of prostaglandins [22], we determined the effects of indomethacin on the inhibition of glycogen synthesisby TGFa.

At a concentration of luM, indomethacin causeda small but

not significant increase in glycogen synthesis, while at a concentration of 1OuM it decreased the stimulation by insulin. Indomethacin (1 & 10uM) did not block the inhibition of glycogen synthesisby TGFa! (Table 3). The activation of glycogen synthase by insulin (100nM) was inhibited (100nM) and by EGF (100nM).

by TGF~Y

Synthase a: Control, 1.19 + 0.19; insulin, 1.30 i 0.19;

Insulin + TGFa, 1.12 f 0.08; insulin + EGF, 1.17 f 0.10. Total activity: control, 3.73 + 0.13; insulin, 3.73 f 0.20; insulin + TGF, 3.58 f 0.14; insulin + EGF, 3.71 f 0.23 munits / mg of protein, means + SEM, n=3). DISCUSSION The inhibition by TGFa of 14C-glucoseincorporation into glycogen is associatedwith inhibition of cellular glycogen accumulation, indicating that it reflects true inhibition of glycogen synthesisand deposition rather than an effect on 14C-labelrecycling. The similarities 563

Vol. 183, No. 2, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

TABLE 3. Effects of indomethacin on the inhibition of glycogen synthesisby TGFa

Glycogen synthesis Additions:

None

1pM Indomethacin 10pM Indomethacin

None lnM-Insulin ” + SnM-TGFcu

17.4 f 1.3 30.2 Ifr 5.0 15.5 &- 2.9 *

19.7 &- 2.8 32.9 f 3.1 14.8 k 2.9 *

16.6 f 3.3 23.4 & 4.1 13.0 & 2.3 *

Rat hepatocyteswere incubated for 90min in Earle’s balanced salts containing 15mM [U-14C]glucose and the additions indicated. Glycogen synthesisis expressedas nmol of 14C-glucoseincorporated into glycogen /90min per mg of protein. Values are means If: SEM for 4 cultures. * P < 0.01 TGFa + insulin vs insulin. in the effects of rat TGFcv on glycogen synthesis in rat and guinea pig hepatocyte cultures are consistent with the conservation of the sequence of both TGFa and its receptor in different species [5,9]. Since TGFLY inhibits glycogen synthesis in the absence of insulin as well as in its presence, this indicates an effect on glycogen metabolism (synthesis and / or degradation) rather than on the insulin-signalling sequence per se. The inhibition of glycogen synthesis by TGFcY was slightly diminished but not abolished in a Ca2+-free medium, indicating that uptake of extracellular Ca2+ might at most account for only a small component of the inhibition. EGF inhibits the stimulation of glycogen synthesis by insulin [18] and this inhibition is not associated with an increase in CAMP [18]. EGF increases the cytosolic [Ca2+] in various cells, including hepatocytes [23-251, by an initial peak increase derived from intracellular stores followed by a prolonged plateau dependent on uptake of extracellular Ca2+ [25]. Since the intial peak is much higher in magnitude than the sustained plateau [25], it is possible that the initial peak derived from intracellular stores may be more important for inhibition of glycogen synthesis. If TGFu also increases cytosolic Ca *+ b y release of intracellular stores the inhibition of glycogen synthesis may be due to the initial rise in Ca2+. Liver regeneration after partial hepatectomy is associated with increased lipogenesis but with low hepatic glycogen stores [26,27] and with increased expression of TGFu [17]. The concentration of TGFa causing half-maximal inhibition of glycogen synthesis (Fig. 1) was in the picomolar range and maximum inhibition was within the range that is effective at stimulating hepatocyte proliferation in culture [ 171. This high sensitivity of glycogen synthesis to TGFa suggests, that if TGFcw has an autocrine role in controlling hepatocyte proliferation during liver regeneration [17] it may also be responsible for the low hepatic glycogen stores. The induction of metallothionein production in liver cells during endotoxaemia (sepsis) or inflammation [28] may be mediated by the increased production of TGFa by macrophages [28]. Sepsis is associated with impaired glucose tolerance despite increased concentrations of plasma insulin [29]. Since endotoxins induce the production of TGFa by macrophages 1281, impaired glucose tolerance may be in part due to TGFcx counteracting the glycogenic effect of insulin in the liver. The metabolic effects of TGFa on the liver in pathological states such as sepsis and hepatocellular carcinoma, which is associated with increased urinary excretion of TGF~Y [30], warrant further study. 564

Vol.

183,

No.

2, 1992

Acknowledgment:

BIOCHEMICAL

We thank NOVO/Nordisk

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

for financial support.

REFERENCES

[ll I21 [31 [41

[51

Derynck, R. (1986)J. Cell. Biochem. 32, 293-304 Derynck, R. (1988)Cell, 54, 593-595 Marquardt, H. Hunkapiller,M.W., Hood, L.E. andTodaro,G.J. (1984)Science,223, 1079-1082 De Larco, J.E. andTodaro, G.J. (1978)Proc. Natl. Acad. Sci. 75, 4001-4005 Derynck, R., Roberts,A.B., Winkler, M.E., Chen, E.Y. andGoeddel,D.V. (1984) Cell 38. 287-297

Lee, D.S., Rose,T.M., Webb,N.R. andTodaro, G.J. (1985)Nature (London)313, 489-491 Pike, L.J., Marquardt, H., Todaro, G.J., Gallis, B., Casnellie,J.E., Bornstein,P. and 171 Krebs, E.G. (1982) J. Biol. Chem.257, 14628-14631 Massague,J. (1983)J. Biol. Chem.258, 13614-13620 PI Lax, I., Johnson,A., Howk, R., Sap,J., Bellot, F., Winkler, M., Ullrich, A., PI Vennstrom,B., Schlessinger, J. andGivol, D. (1988)Mol. Cell. Biol. 8, 1970-1978 Stern, P.H., Kreiger, N.S., Nissenson,R.A., Williams, R.D., Winkler, M.E., Derynck, [lOI R. and Strewler, G.J. (1985)J. Clin. Invest. 76, 2016-2019 Ibbotson,K.J., Harrod, J., Gowen, M., D’Souza, S., Smith, D.D., Winkler, M.E., 1111 Derynck, R. andMundy, G.R. (1986)Proc. Natl. Acad. Sci. 83, 2228-2232 iI21 Schreiber,A.B., Winkler, M.E. andDerynck, R. (1986)Science,232, 1250-1253 Cutry, A.F., KinniburghA.J., Twardzik, D.R. andWenner,C.E. (1988)Biochem. [I31 Biophys. Res. Commun. 152, 216-222 Ristimaki, A. (1989)Biochem.Biophys. Res.Commun.160, 1100-1105 [I41 Rhodes,J.A., Tam, J.P., Finke, U., Saunders,M., Bernanke,J., Silen, W. and WI Murphy, R.A. (1986) Proc. Natl. Acad. Sci. 83, 3844-3846 Madtes, D.K., Raines,E.W., Sakariassen, K.S., Assoian,R.K., Sporn, M.B., Bell, G.I. [I61 andRoss,R. (1988)Cell, 53, 285-293 Mead, J.E. andFausto,N. (1989)Proc. Natl. Acad. Sci. 86, 1558-1562 iI71 U81 Chowdhury, M.H. andAgius, L. (1987)Biochem.J. 247, 307-314 [191 Agius, L., Peak, M. andAlberti, K.G.M.M. (1990)Biochem.J. 266,99-102 Clifton, P.M., Chang,L. andMcKinnon, A.M. (1988)Anal. Biochem. 172, 165-168 WI 1211 Dopere,F., Vanstapel,F. andStalmans,W. (1980)Eur. J. Biochem. 104, 137-146 WI Altin, J.G. andBygrave, EL. (1988)Mol. Cell. Biochem.83, 3-14 Johnson,R.M., Connely,P.A., Sisk, R.B., Pobiner,B.F., Hewlett, E.L. andGarrison, v31 J.C. (1986)Proc. Natl. Acad. Sci. U.S.A. 83, 2032-2036 Bosch,F., Bouscarel,B., Slaton,J., Blackmore,P.F. and Exton, J.H. (1986) Biochem. [241 J. 239, 523-530 Magni,M., Meldolesi, J. andPandiella,J. (1991)J. Biol. Chem.266,6329-6335 L51 Gove, CD. and Hems,D.A. (1978)Biochem.J. 170, l-8 WI Schofield, P.S., Sugden,M.C;, Corstorphine,C.G. and Zammit, V.A. (1987) Biochem. 1271 J. 241. 469-474 I281 Iijima,‘Y., Fukushima,T. andKosaka,F. (1989)Biochem.Biophys. Res.Commun. 164, 114-118 De Vasconcelos,P.R.L., Kettlewell, M.G.W., Gibbons,G.F. and Williamson,D.H. P91 (1989) Clin. Sci. 76, 205-211 Yeh, Y.C., Tsai, J.F., Chuang,L.Y., Yeh, H.W., Tsai, J.H., Florine, D.L. andTam, [301 J.P. (1987)CancerRes.47, 896-901

56.5