Cyclic Nucleotides and Gluconeogenesis Rat Liver Cells John N. Fain, Margaret

E. M. Tolbert,

Fred R. Butcher, and Andrew Gluconeogenesis from lactate, pyruvate, fructose, alanine, and other substrates was accelerated by glucagon or epinephrine in hepatocytes isolated from rat liver. Glucagon and epinephrine also increased cyclic AMP accumulation by mt hepatocytes. lsoproterenol increased cyclic AMP but not gluconeogenesis, while phenylephrine accelerated gluconeogenesis. The activation of gluconeogenesis by epinephrine was unaffected by propmnolol but blocked by dihydroergotamine. Dibutyryl cyclic AMP added to hepatocytes stimulated gluconeogenesis al concentrations as low as 1 fl. Exogenous cyclic GMP (0.1-l PM) inhibited gluconeogenesis due to either glucagon or epinephrine without affecting basal gluconeogenesis. However, carbamylcholine did not affect gluconeogenesis by hepato-

by

Richard H. Pointer, Arnold

cytes. Basal gluconeogenesis and the increases due to all agents were inhibited by removal of extracellular calcium or the presence of A-231 87, D-600, or tetracaine. In contrast, added 0.1 PM cyclic GMP, 2 mM NH,CI, and 10 MM phenethylbiguanide inhibited glucagon- or epinephrine-stimulated gluconeogenesis without affecting basal values. Studies with hepatocytes indicate that the hormonal activation of gluconeogenesis is not limited to substrates entering prior to triose phosphate formation. Glucagon may act by increasing cyclic AMP which acts via unknown mechanisms to increase gluconeogenesis. In contrast, epinephrine acts via a cyclic AMP-independent mechanism which does not appear to involve cyclicGMP, Ca’+ fiux, or K+ flux.

C

YCLIC AMP was discovered as a result of studies by Sutherland on the activation of phosphorylase by epinephrine and glucagon in dog liver slices.‘-3 Subsequent studies have supported the hypothesis that glucagon activation of glycogenolysis is secondary to activation of adenylate cyclase and intracellular accumulation of cyclic AMP.* Those effects of catecholamines which have been classically described as betaadrenergic effects are accompanied by an increased formation of cyclic AMP and blocked by beta blockers such as propranolol.4 In contrast, the alpha effects of catecholamines have not been found to be associated with an increased accumulation of cyclic AMP and are blocked by phentolamine, phenoxybenzamine, and dihydroergotamine .4 However, fairly high concentrations of ergotamine (160 &4) blocked the activation of dog liver phosphorylase5 and adenylate cyclase b catecholamines.6 We suggest that this could be explained if ergot alkaloids at low concentrations are alpha blockers while at concentrations in the range of 0. l-l mM they are beta blockers. From the Division of Biological and Medical Sciences. Brown University, Providence, R.I. Received for publication October 15. 1974. Supported by USPHS Research Grants AM 10149 and AM 14648 from the National Institute of Arthritis, Metabolism and Digestive Diseases. R.H.P. is a Predoctoral Fellow of the National Fellowship Fund. Reprint requests should be addressed to John N. Fain, Brown University, Box G. Providence, R.I. 02912. 0 1975 by Grune & Stratton, Inc. Metabolism, Vol. 24, No. 3 (March), 1975

395

396

FAIN ET AL.

In perfused rat livers, glucagon was a marked activator of cyclic AMP accumulation and also stimulated gluconeogenesis.’ The further finding that added cyclic AMP and analogs of cyclic AMP mimicked the effects of glucagon led Exton and Park* to postulate that glucagon increased gluconeogenesis through cyclic AMP. It should be noted that the mechanism by which cyclic AMP increases gluconeogenesis remains to be elucidated. In addition, Exton and Parks found that catecholamines such as epinephrine or norepinephrine also stimulated gluconeogenesis. However, isoproterenol did not increase gluconeogenesis, and epinephrine had only a very small effect on cyclic AMP accumulation as compared to glucagon.’ The failure of isoproterenol to accelerate gluconeogenesis in perfused rat livers is disturbing as it is generally more potent than epinephrine as an activator of adenylate cyclase and cyclic AMP accumulation.4 Isoproterenol was also relatively ineffective in activating hepatic glycogenolysisp~‘O and glycogen phosphorylase in vivo” or in perfused rat livers. ‘* The activation of rat hepatic adenylate cyclase in vitro and of phosphorylase in vivo due to isoproterenol was blocked by the beat-adrenergic antagonist propranolol.” This suggests that isoproterenol acts as a beta-adrenergic agent to accelerate glycogenolysis through cyclic AMP, as originally suggested by Sutherland and Rall. However, in rat liver there appears to be another mechanism not involving cyclic AMP by which catecholamines increase glycogenolysis and gluconeogenesis. There has been a reluctance to accept this conclusion despite the large amount of data which are best explained by this hypothesis. The most convincing evidence is that propranolol blocked the epinephrine-stimulated increases in rat hepatic adenylate cyclase and cyclic AMP release by perfused rat livers’* but not the increase in hepatic phosphorylase in viva” or in perfused livers.‘2 Phenylephrine, a predominantly alpha agonist, was found to increase glycogen phosphorylase activity in perfused livers but did not stimulate cyclic AMP release or accumulation.‘2 Phentolamine, an alpha adrenergic antagonist, reduced the activation of phosphorylase by epinephrine but not cyclic AMP accumulation.‘* These data indicate that hepatic glycogenolysis due to epinephrine is not a classic beta response and probably occurs by a mechanism independent of cyclic AMP. The stimulation by catecholamines of hepatic gluconeogenesis in the rat also appears not to be mediated through cyclic AMP. Epinephrine was effective in stimulating gluconeogenesis in perfused rat livers, while isoproterenol was inactive.’ Propranolol was also unable to inhibit gluconeogenesis in perfused rat livers.‘-’ The present studies were designed to investigate the role of cyclic nucleotides in gluconeogenesis by suspensions of rat parenchymal cells. These studies support the hypothesis that gluconeogenesis is accelerated by a cyclic AMPdependent (glucagon) and -independent (catecholamine) process. MATERIALS

AND METHODS

Hepatic parenchymal cells can be readily isolated by digestion of perfused rat livers with collagenase either in the absence or presence of hyaluronidase.14 Studies in our laboratory have indicated that this procedure for isolating hepatocytes is the best of those which have been published to date.” The technical problems involved in setting up such a procedure are numerous. However,

CYCLIC NUCLEOTIDES

397

AND GLUCONEOGENESIS

the advantages of being able to obtain a large population of cells whieh could be divided among many tubes and used for the examination of hormone action have made it worthwhile to use this procedure. Laboratories such as those of Krebs,16 Lardy,“,‘* Park,” and Haynes,20F2’ which formally used perfused livers, have all recently used isolated rat liver cells. The techniques used for isolation of rat liver cells and assay of glucose have been described.‘5,22 The sources of drugs and hormones were as folows: N602’ -dibutyryl adenosine 3’,5’ monophosphate (DcAMP), N6-monobutyryl adenosine 3’,5’ monophosphate (MBcAMP), adenosine 3’,5’ monophosphate (CAMP), guanosine 3’,5’ monophosphate (cGMP), Schwartz Bioresearch; (-) epinephrine bitartrate, glucagon, and carbamylcholine (Carbachol), Sigma Chemical Co.; tetracaine, K and K Labs Inc; A-23187, gift of Eli Lilly and Company, Indianapolis, Ind; N’-@-phenethylbiguanide (PEBG, DBI, or phenformin), gift of Ciba-Geigy; D-600, gift of Dr. Jean Marshall of Brown. RESULTS

The first problem we had in trying to link cyclic AMP with gluconeogenesis in rat liver cells was that isoproterenol was unable to stimulate gluconeogenesis in isolated liver cells.** This was not due to an inhibitory effect of isoproterenol on the pathways of gluconeogenesis, since stimulation of gluconeogenesis by epinephrine was the same in the presence or absence of isoproterenol.** However, isoproterenol was more effective than epinephrine in stimulating cyclic AMP accumulation in rat liver cells (Fig. 1). The stimulation of cyclic AMP that one can see with either catecholamine is far less than that produced by glucagon under identical conditions in isolated cells** or perfused livers.’ We looked at other catecholamines which are primarily alpha-agonists for their ability to stimulate gluconeogenesis and found that phenylephrine stimulated gluconeogenesis. ** These studies suggested that the activation of gluconeo-

1S-

GLUCOSE

P’OC’zrr

16=

14r

-2-

Fig. 1. Cyclic AMP and gluconeogenesis in response to epinephrine and isoproterenol. liver cells (6.8 x lo6 per tube) were incubated for 1 min for measurement of cyclic AMP accumulation and 2 hr for glucose production in the presence of 10 mM lactate. The values are shown as increments due to isoptoterenol or epinephrine over the basal values in the absence of catecholamines, which for glucose production was 21 flug of glucose per 10’ cells, while that of cyclic AMP accumulation was 5.6 pmoles/106 cells. (Reproduced by permission from Tolbert et alF2)

IS0p~0t~~.d

-4CYCLIC

AMP

Concentration of Catechokninbo in uM (log scale)

20

FAIN

AMP All

DLUCOSE

PRODU

ET At.

MIN.

>N OVRR 2 HOURS

Fig. 2. Inhibition by propmnolol of cyclic AMP accumulation due to epinephrine without any effect on gluconeogenesis. liver cells (7 x lo* per tube) were incubated for 1 min or 2 hr with IO mM lactate in the absence (left side of figure), or presence of 1 fl D( - ) epinephrine (right side of figure). The cells were incubated either without (open bars) or plus 5 NM propmnolol (diagonal bars), 10 CM phenoxybenzamine (solid bars), 10 CM phentolomine (dotted ban), and 0.5 pM dihydroergotamine (cross-hatched ban). The data am taken from reference 22.

genesis might well be an alpha- effect and that the rise in cyclic AMP was not responsible for the increased gluconeogenesis. Measurements of total cyclic AMP in tissue are not as easy to interpret as it was once thought. Under appropriate conditions, concentrations of catecholamines can be found which increase lipolysis in fat cells23~24or contractile force in cardiac muscle” without any detectable change in total cyclic AMP. If only a small fraction of total cyclic AMP is physiologically important, then agents could double the size of this pool without its resulting in any measurable increase over basal cyclic AMP. Propranolol blocked the increases in cyclic AMP due to catecholamines in either tissue and the metabolic responses.25-27 If the activation of gluconeogenesis by catecholamines in rat liver cells is an alpha effect, then it should not be blocked by propranolol. The data in Fig. 2 indicate that propranolol did not inhibit gluconeogenesis due to epinephrine but did abolish the rise in cyclic AMP. This is in striking contrast to the situation with respect to fat cell lipolysis and cardiac contractility. Thus, gluconeogenesis would be classified as an alpha effect of catecholamines but is an unusual alpha effect in that it is associated with a rise rather than a fall in cyclic AMP. It has been postulated4 that all alpha effects of catecholamines are mediated by a fall in cyclic AMP. This is hardly compatible with our findings. More recently it has been suggested that alpha effects of catecholamines are secondary to a rise in cyclic GMP. 28 However, we found that added cyclic GMP at 0.1 PLM actually blocked the rise in gluconeogenesis in rat liver cells due to either glucagon or epinephrine, while 100 PM cyclic GMP inhibited basal gluconeogenesis (Fig. 3). Exogenous cyclic GMP at low concentrations often has effects opposite to those of added cyclic AMP.‘* However, in vivo hepatic gluconeogenesis was increased by the administration of cyclic GMP,29

CYCLIC NUCLEOTIDES

Fig.

3.

399

AND GLUCONEOGENESIS

Inhibition

of

glucose release by cyclic GMP. Liver cells (3.1 x 106 cells per tube) were incubated for 2 hr in the presence of 10 mM lactate. The values are the means of ten experiments in the absence (open circles)01 presence of 1 CM glucagon (stars) or 1 CM epinephrine (squares). Significant reductions in glucose production are indicated by small stars (p > 0.05 by paired comparisons). The glucose release represents gluconeogenesis, as the glycogen content was very low after 2 hr incubation

11-

I-

4-

o,

1

I

0.1

0

CYCUC

1 OMI

ADDED10

10

100 100

MEDlUMlrMI

using cells from fasted mts and was unaffected by cyclicGMP, glucagon, or epinephrine.

as was both glycogenolysis and gluconeogenesis in perfused rat liver by O.l1 mM cyclic GMP.30-34 The possibility that cholinergic stimulation increases cyclic GMP in rat liver parenchymal cells or inhibits gluconeogenesis was tested. However, as yet we have been unable to see any effect of carbamylcholine at concentrations up to 10 pm on gluconeogenesis (Fig. 4) or cyclic GMP (unpublished experiments, Butcher FR, Pointer RH). Similar results have been seen in perfused rat livers where carbamylcholine did not affect gluconeogenesis from lactate under conditions in which it accelerated glycogen deposition.35 The effect of carbamylcholine on the perfused liver to increase glycogen deposition was similar to the effect of insulin we reported in rat liver cells. Is Insulin and carbamylcholine

CAR6AMYL

30-

CHOLINE

A-23107

OO-o-~o

20lo-

o-o-

. + EPINEPHRINE ,, + GLUCAGON o BASAL

oJ’

0-l I 00.1

I

I

1

I

1

10

I II 1

I II

100

.l

I I

I

1

CON. OF ADDED DRUG(bgscohin~)

1 I

1

10

I

1

100

Fig. 4. Effects of various drugs on gluconeogenesis by epinephrine and glucagon. Liver cells (3.5-4.5 x lo6 cells per tube) were incubated for 2 hr in the presence of 10 mM lactate. The values represent the means of thme separate experiments for each drug.

400

FAIN

ET AL.

have been reported to increase cyclic GMP in rat liver slices,36 but we have been unable to confirm these findings using parenchymal cells. Our conclusion is that cyclic GMP is an unlikely candidate for the second messenger which regulates the activation of gluconeogenesis by alpha-adrenergic amines. If cyclic GMP is involved in the regulation of gluconeogenesis, the available evidence suggests that it would actually inhibit glucose release by diverting glucose to glycogen. Secretin is the only hormone which stimulated guanyl cyclase in broken-cell preparations from rat liver,” but we have been unable to observe any effect of it on gluconeogenesis by rat liver cells (M.E. M. Tolbert, unpublished experiments). If the activation of gluconeogenesis by epinephrine is an alpha effect, it does not follow the conventional classification with regard to inhibition by alphaadrenergic antagonists. We found that phentolamine and phenoxybenzamine** were relatively ineffective as antagonists of epinephrine-induced gluconeogenesis (Fig. 2). In contrast, dihydroergotamine at concentrations below 1 PM was an extremely effective inhibitor of catecholamine-induced gluconeogenesis (Fig. 2). It should be noted that these were low concentrations of dihydroergotamine. High concentrations of dihydroergotamine (100 PM) have a variety of nonspecific effects including inhibition of catecholamine activation of adenylate cyclase6s3* and the ability of dibutyryl cyclic AMP to increase glucose output by perfused rat livers.39 The effects of alpha blockers we observed with rat liver cells are similar to those seen in rats in vivo by Kennedy and Ellis” with respect to loss of hepatic glycogen. These authors found beta-adrenergic antagonists and Dibozane (an alpha blocker) to be relatively ineffective. Phentolamine (an alpha blocker) was partially effective, while dihydroergotamine was the most effective inhibitor tested of epinephrine action. In perfused rat livers, propranolol was ineffective in blocking the activation of phosphorylase by epinephrine, while phentolamine was an effective inhibitor.12 Phentolamine did not reduce gluconeogenesis from lactate by perfused rat livers in the presence of epinephrine.13 Similar results were seen in rat liver cells (Fig. 2). The available data are most compatible with the hypothesis that epinephrine and other alpha-adrenergic agonists stimulate hepatic gluconeogenesis and glycogenolysis by a mechanism not involving cyclic AMP which is exquisitively sensitive to inhibition by dihydroergotamine. We compared the effects of epinephrine and glucagon with those of nonhormonal agents which increase gluconeogenesis. Valinomycin or lysine stimulate gluconeogenesis by mechanisms which are apparently independent of those for either catecholamines or glucagon. This conclusion was based on the finding that the same increases due to saturating concentrations of glucagon, epinephrine, or dibutyryl cyclic AMP were seen in the presence and in the absence of lysine or valinomycin.40 If the mode of catecholamine action on glucoenogenesis does not involve cyclic AMP, what could be its second messenger? We do not have the answer to this as yet, but the following studies indicate some of the areas we have investigated. Calcium is currently in vogue as the latest fad in hormone action. It is difficult to study the calcium requirement for activation of gluconeogenesis in rat liver cells, for the omission of calcium from the medium resulted in com-

CYCLIC

NUCLEOTIDES

AND

401

GLUCONEOGENESIS

plete inhibition of both basal gluconeogenesis and the response to all added hormones.4O The addition of 0.65 mM calcium to the medium under such conditions can partially restore basal gluconeogenesis and the response to hormones.40 If the calcium was increased from 0.65 to 2.5 mM in the medium, no further increase in basal gluconeogenesis was seen, but there was a slight increase in the ability of epinephrine to stimulate glucoenogesis.40 Pilkis et aLI found that both basal and glucagon-stimulated gluconeogenesis were increased by the addition of calcium and that glucagon enhanced the uptake of labeled calcium by rat hepatocytes. Previously Friedmann and Park4’ reported that glucagon or catecholamine addition to rat liver previously loaded with labeled calcium resulted in an immediate efflux of calcium. Cyclic AMP infusion also accelerated calcium efflux which was unimpaired in livers from adrenalectomized rats despite their markedly reduced rates of gluconegenesis. 41 Tetracaine (1 mM) was found to block the calcium efflux due to cyclic AMP as well as gluconeogenesis in perfused rat livers. 42 Tetracaine had little effect on basal gluconegenesis or glucagon-induced increases in cyclic AMP.42 However, in isolated rat liver hepatocytes, tetracaine at 0.1 mM reduced basal gluconeogenesis at the same time that it abolished the increase due to glucagon or catecholamines (Fig. 4). Similar results were seen with D-600 or the divalent cation ionophore A-23187 which inhibited both basal and hormone-stimulated gluconeogenesis (Fig. 4). Zahlten et al.” found that A-23187 inhibited gluconeogenesis from pyruvate. However, these workers reported that glucagon inhibited gluconeogenesis from pyruvate, while the omission of calcium had little effect on gluconeogenesis. I7 Neither finding is comparable to anything we have seen and might have resulted from the use of gelatin rather than albumin in the incubation medium. Another difference between our studies and those of Zahlten et a1.r8 was that 2 mM ammonium chloride did not affect basal gluconeogenesis from lactate or pyruvate but abolished the increases due to either glucagon or epinephrine (Table 1). Zahlten et al.” found just the opp osite effect of ammonium ion. The increases in gluconeogenesis due to either glucagon or epinephrine were unreTable 1. Effects of NH, Cl on Gluconeogenesis Glucose Production A Due to NH4CI Substrate

Glucogon

Epinephrine

(2.7 CIM)

(1 aM)

dCAMP

WM)

BC%il

10 mM

0

39

+12+4

+

9+3

lactate, 10 mM

2

36

+

-

5*10

Pyruvate,

10 mM

0

58

+10*3

+

7+2

+

5ztl

Pyruvate,

10 mM

2

59

+

2+1

+

1*2

+

5zt4

(2DrM)

pg/l06/cnlls lactate,

2+5

+13*

Lactate,

10 mM

+ Pyruvate,

1 mM

0

43

+

9+2

+10+2

+

5zt2

lactate,

10 mM

+ Pyruvote,

1 mM

2

48

-

6+2

+

-

5Lt2

Liver cells (4.6 mM NH,CI

pyruvote, we

x

lo6

cells

per

or lactote/pyruvate

represented

OS the f

tube)

were

incubated

(10: 1). The changes standard

error

for three

for 2 hr in the presence due to added experiments.

agents

1*2

of either

1

O&O

10 mM

in the presence

and

lactate,

10

absence

of

402

FAIN ET Al.

Table 2. Inhibition by phenethylbiguanide (PEBG) of Glucagon and Epinephrine-stimulated Gluconeogenesis Per Cent Phenethyl-

Per Cent

biguanide

Inhibition of

Concentration

Basal Due to

(mM)

BlXCll

PEBG

Per Cent

Inhibition of

Inhibition of

A Due to Glucagon fig

(2.7 FM)

gl”core/l06

A Due to

A Due to

A Due to Glucogon

Epinephrine

by PEBG

Epinephrine

by

PEBG

(l.OpM)

cells

0

23.6 f 4

+12.0&2.7

0.01

23.5 zk 5

3.5 f 4.3

+

5.9ztl.O

5lf

11

+

+17.5*4.0 0.6 f 2.2

51 f

11

0.02

22.8~~4

3.1 zt4.3

+

3.6ztl.0

7oi

14

+

6.0 f

66+

16

0.10

24.2 zt 4

0

+

4.4jzl.l

63zt

14

+

5.8zt1.7

67zt

14

1.0

Liver ceils (3.6 x lo6 per tube) were incubated for 2 hr in the presence of 10 mM lactate plus 2% bovine serum albumin. Values given are the means f

standard errars for seven paired experiments.

lated to the redox state of the added substrate, for similar increases were seen with either 10 mM pyruvate, 10 mM lactate, or 10 mM lactate in the presence of 1 mM pyruvate (Table 1). The sole drug we have found which blocks the increase in gluconeogenesis due to epinephrine but not that due to glucagon was dihydroergotamine.” Besides 2 mM ammonium chloride, the only other agent which inhibited the increases in gluconeogenesis due to epinephrine and glucagon but not basal gluconeogenesis was phenethylbiguanide (Table 2). The striking finding was that near-maximal inhibition was seen with only 0.01 mM phenethylbiguanide (Table 2). We could see a reduction in basal gluconeogenesis only if 1 mM phenethylbiguanide was present (data not shown). Relatively high concentrations of phenethylbiguanide (0.5-l mM) were required to inhibit basal gluconeogenesis by perfused rat livers.43,” The inhibition by phenethylbiguanide was attributed to some indirect effect on mitochondrial metabolism.43,” The action of phenethylbiguanide was not secondary to any drop in total ATP, but the oxidation reduction state of both cytosol and mitochondria was more reduced. 43*44Similar conclusions were derived from studies on perfused guinea pig livers.45 The mechanism by which biguanides such as phenethylbiguanide lower blood Our opinion is that biguanides glucose clinically is not clearly established.46 primarily affect mitochondrial metabolism by interfering with electron transport processes. Davidoff4’ recently found that 1 mM phenthylbiguanide blocked calcium uptake and release by liver mitochondria. He postulated that inhibition of calcium release by mitochondria might explain the reduction in gluconeogenesis due to biguanides. The divalent cation ionophore A-23187 in the presence of extracellular calcium inhibited gluconeogenesis and presumably increased influx of calcium into the cytosol (Fig. 4). It is possible that A-23187 reduced gluconeogenesis indirectly rather than as a result of any change in calcium. The increased flux of divalent cations might result in a depletion of critical energy stores needed for gluconeogenesis. We found that A-23 187 and glucagon accelerated glycogenolysis in liver cells from fed rats. Glucagon, epinephrine, dibutyryl cyclic AMP, and 10 1LM

CYCLIC NUCLEOTIDES

AND GLUCONEOGENESIS

403

A-23 187 all accelerated glycogenolysis, and the increase due to A-23 187 was not additive to that of the other agents (R. H. Pointer and J. N. Fain, unpublished experiments). These results suggest that an increased flux of calcium due to glucagon or epinephrine might explain their glycogenolytic action. Pilkis, Claus, and Parki have reported that there was an increased uptake of radioactive calcium by rat hepatocytes incubated with glucagon. Whether this is the result or the cause of the increased glycogenolysis remains to be established. Increased flux of calcium appears less likely an explanation for the increase in gluconeogenesis due to either glucagon or epinephrine, since A-23187 actually inhibited gluconeogenesis. An interesting hypothesis to explain glucagon-stimulated gluconeogenesis is that the hormone acts to increase pyruvate uptake by mitochondria.48 This hypothesis resulted from studies which showed that hepatic mitochondria isolated from rats given epinephrine or glucagon utilized pyruvate at increased rates.48 Subsequently, Haynes et al. 21found that the addition of 10 nM valinomycin to isolated mitochondria also resulted in increased pyruvate utilization by mitochondria. Valinomycin markedly increases K+ flux by mitochondria which facilitates the entry of phosphate and organic anions. It is attractive to speculate that hormones might increase triose uptake by mitochondria for carboxylation to four-carbon precursors of gluconeogenesis either directly or secondarily to increased K+ flux. We40 and Haynes et al. 21found that low concentrations of valinomycin stimulated gluconeogenesis by isolated rat liver cells. Higher concentrations were inhibitory, which may result from an apparent uncoupling action due to dissipation of large amounts of energy for cyclic K+ flux. The increase due to valinomycin was additive to that of glucagon or epinephrine, which suggests that neither hormone works like valinomycin.40 Both basal and hormonestimulated gluconeogenesis were reduced by the absence of K+ or the presence of 1 mM ouabain.40 The inhibition by ouabainqO was similar to that noted with tetracaine. Friedmann found that tetracaine also blocked the flux of K+ seen after addition of glucagon or cyclic AMP. Tetracaine is a local anesthetic and thus may impede movement of all ions across the plasma membrane, while ouabain may be relatively specific for K +. We suggest that K+ is required in the extracellular medium for optimal gluconeogenesis but that the hormonal regulation of gluconeogenesis is not mediated through regulation of K+ flux. Friedmann and Dambachso found that glucagon, cyclic AMP, or cyclic GMP produced hyperpolarization and increased the fluxes of sodium and potassium in perfused livers from both normal and adrenalectomized rats. Basal, as well as epinephrine- or glucagon-stimulated, gluconeogenesis was reduced in livers from adrenalectomized rats, but the increases in cyclic AMP were unaffected.51,52There was also no evidence that variations in plasma calcium were involved in the effects of adrenalectomy on gluconeogenesis.sl These results suggest that the effects of glucocorticoids on gluconeogenesis are unrelated to cyclic AMP accumulation or cyclic fluxes of K +, Na+ or Ca2+. The strongest evidence that an elevation of cyclic AMP can trigger increased gluconeogenesis is the stimulation by added cyclic AMP. Exton and Park8 found that 50 PM cyclic AMP would increase gluconeogenesis by perfused rat

404

FAIN ET Al.

livers. Corm and Kipnis30 found that as little as 1 pM dibutyryl cyclic AMP would increase gluconeogesis by perfused rat livers. Garrison and Haynes2’ found that the concentration of dibutyryl cyclic AMP which produced halfmaximal increases in glucose.production by rat liver cells was 0.7 pM, while that of cyclic AMP was 80 pM. We found similar results except that the concentrations required were somewhat higher (Fig. 5). The N6-monobutyryl cyclic AMP was equivalent to dibutyryl cyclic AMP and both were 100 times more potent than cyclic AMP. The ability of relatively low concentration of mono- or dibutyryl cyclic AMP to increase gluconeogenesis is striking, as generally higher concentrations are required in other cells. The only piece of evidence we have to support the view that glucagon acts by the same mechanisms as dibutyryl cyclic AMP is that gluconeogenesis due to glucagon was not additive to that of dibutyryl cyclic AMP.40 Exton and Parks3 proposed that cyclic AMP, glucagon, or epinephrine stimulated gluconeogenesis between pyruvate and phosphopyruvate. This conclusion was primarily based on the finding that glucaton did not affect gluconeogenesis from fructose or dihydroxyacetone in perfused rat livers.s3 However, Veneziale54 found that glucagon increased glucose formation from fructose, and Blair et al.ss described similar results using xylitol. In isolated rat liver cells, all published reports have shown a stimulatory effect of glucagon on glucose synthesis from fructose.‘7*20,40 These results indicate that the regulation of gluconeogenesis by glucagon is not solely the result of effects on enzymatic reactions in the sequence from pyruvate to phosphopyruvate. Taunton et a1.56 reported that 4 min after the injection of glucagon to rats, the activityinliver homogenates of fructose diphosphatase was increased, while phosphofructokinase and pyruvate kinase activities were decreased. Insulin injection had the opposite effect but did not alter the levels of hepatic cyclic AMP, while glucagon markedly elevated cyclic AMP.56 In a preliminary report Veneziale and Swenson57 found that, in rat hepatocytes exposed to monoGLUCOSE FORMATION OVER 2 HIS Fig. 5. Stimulation of gluconeogenesis by cyclic odenine nucleoi N6-MBCAMP tides. Liver cells (2.8 x 40 1O6 cells Per tube) were DlCAMP incubated for 2 hr in CAMP the presence of 10 mM ,U 30lactate. The values are 2 the means + standard “$ errors of three paired 2 20/’ experiments and ore shown as the increlo/p ments due to added nucleotides in glucose pro, p 1 1 1 I duction over the basal o ?; IO 20 40 100 160 320 5 640 value of 70 rg of gluNUCLEOTIDE CONCENTRAllON (uMl LOG SCALE case formed during 2 hr. N6-0”-dibutyryl cyclic AMP (D&) is shown by circles, N’-monobutyryl cyclic AMP ( N6-M&) by stars, and cyclic AMP by squares. Cyclic GMP at 80 or 160 pM in the same experiments decreased gluconeogenesis by 18 and 32 cg of glucose, rospoctively.

CYCLIC NUCLEOTIDES

AND GLUCONEOGENESIS

405

butyryl cyclic AMP for 10 min, there was a decrease in phosphofructokinase activity but no change in fructose diphosphatase. We can anticipate much progress during the next few years in elucidating effects of hormones on these enzymatic steps. But at the moment it is not clear either where or how glucagon, epinephrine, or cyclic AMP increase glucose formation from fructose or other substrates. Our conclusion is that cyclic AMP is probably a second messenger for the regulation of gluconeogenesis. However, it does not appear to be the messenger for epinephrine and may not even be that for glucagon. The results shown in this report illustrate what is increasingly becoming apparent-namely, that regulation of metabolism is very complex. Cyclic nucleotides should be considered as a great advance in our understanding of the mode of hormone action rather than the final solution. REFERENCES 1. Rall TW, Sutherland EW, Wosilait WD: The relationship of epinephrine and glucagon to liver phosphorylase. III. Reactivation of liver phosphorylase in slices and in extracts. J Biol Chem 218:483-495, 1956 2. Rail TW, Sutherland EW, Berthet J: The relationship of epinephrine and glucagon to liver phosphorylase. IV. Effect of epinephrine and glucagon on the reactivation of phosphorylase in liver homogenates. J Biol Chem 224~463-415, 1957 3. Sutherland EW, Rail TW: Fractionation and characterization of a cyclic adenine ribonucleotide formed by tissue particles. J Biol Chem 232:1077-1091, 1958. 4. Robison A, Butcher RW, Sutherland EW: Cyclic AMP. New York, Academic Press, 1971 5. Berthet J, Sutherland EW, Rail TW: The assay of glucagon and epinephrine with use of liver homogenates. J Biol Chem 229:351-361, 1957 6. Murad F, Chi Y-M, Rail TW, Sutherland EW: Adenyl cyclase. III. The effect of catecholamines and choline esters on the formation of adenosine 3’,5’-phosphate by preparations from cardiac muscle and liver. J Biol Chem 237: 1233-1238, 1962 7. Exton JH, Robison GA, Sutherland EW, Park CR: Studies on the role of adenosine 3’,5’-monophosphate in the hepatic actions of glucagon and catecholamines. J Biol Chem 246: 6166-6177, 1971 8. Exton JH, Park CR: Control of gluconeogenesis in liver. Effects of glucagon, catecholamines, and adenosine 3’,5’-monophosphate on gluconeogenesis in the perfused rat liver. J Biol Chem 243:4189-4196, 1968 9. Arnold A, McAuliff JP: Positive correla-

tion of responsiveness to catecholamines of the rat liver glycogenolytic receptor with other o-receptor responses. Experientia 24:674-675, 1968 10. Kennedy BL, Ellis S: Interactions of catecholamines and adrenergic blocking agents at receptor sites mediating glycogenolysis in the rat. Arch Int Pharmacodyn Ther 177:390406, 1969 11. Newton NE, Hornbrook KR: Effects of adrenergic agents on carbohydrate metabolism of rat liver: Activities of adenyl cyclase and glycogen phosphorylase. J Pharmacol Exp Ther 181:479-488, 1972 12. Sherline P, Lynch A, Glinsmann WH: Cyclic AMP and adrenergic receptor control of rat liver glycogen metabolism. Endocrinology 91:680-690, 1972 13. Hagino Y, Nakashima M: Adrenergic receptors in rat liver. I. Effects of epinephrine and adrenergic blocking agents on gluconeogenesis in perfused liver. lap J Pharmacol 23: 543-551,1973 14. Berry MN, Friend DS: High-yield aration of isolated rat liver parenchymal A biochemical and fine structural study. Biol43:506-520, 1969

prepcells. J Cell

15. Johnson MEM, Das NM, Butcher FR, Fain JN: The regulation of gluconeogenesis in isolated rat liver cells by glucagon, insulin, dibutyryl cyclic adenosine monophosphate, and fatty acids. J Biol Chem 247:3229-3235, 1972 16. Cornell NW, Lund P, Hems R, Krebs HA: Acceleration of gluconeogenesis from lactate by lysine. Biochem J 134:67 l-672, 1973 17. Zahlten RN, Stratman FW, Lardy HA: Regulation of glucose synthesis in hormone-

406

sensitive isolated rat hepatocytes. Proc Nat1 Acad Sci USA 70:3213-3218, 1973 18. Zahlten RN, Kneer NM, Stratman FW, Lardy HA: The influence of ammonium and calcium ions on gluconeogenesis in isolated rat hepatocytes and their response to glucagon and epinephrine. Arch Biochem Biophys 161: 5288535, 1974 19. Pilkis SJ, Claus TH, Park CR: Control of gluconeogenesis and CAMP levels in isolated rat hepatocytes. Second International Conference on Cyclic AMP, 1974, p 34 20. Garrison JC, Haynes RC Jr: Hormonal control of glycogenolysis and gluconeogenesis in isolated rat liver cells. J Biol Chem 248:53335343, 1973 21. Haynes RC Jr, Garrison JC, Yamazaki RK: Comparison of effects of glucagon and valinomycin on rat liver mitochondria and cells. Mol Pharmacol l&381-388, 1974 22. Tolbert MEM, Butcher FR, Fain JN: Lack of correlation between catecholamine effects on cyclic adenosine 3’,5’-monophosphate and gluconeogenesis in isolated rat liver cells. J Biol Chem 248:5686-5792, 1973 23. Fain JN, Wieser PB: Effects of adenosine deaminase on cyclic AMP accumulation, lipolysis and glucose metabolism of fat cells. J Biol Chem (in press) 24. Fain JN: Inhibition of adenosine cyclic 3’,5’-monophosphate accumulation in fat cells by adenosine, N6-(phenylisopropyl) adenosine, and related compounds. Mol Pharmacol 9:595604, 1973 25. Oye I, Langslet A: The role of cyclic AMP in the inotropic response to isoprenaline and glucagon, in Advances in Cyclic Nucleotide Research, vol. I. New York, Raven Press. 1972, pp 291-300 26. Fain JN, Galton DJ, Kovacev VP: Effect of drugs on the lipolytic action of hormones in isolated fat cells. Mol Pharmacol 2:237-247, 1966 27, Fain JN: Biochemical aspects of drug and hormone action on adipose tissue. Pharmacol Rev 25:67-l 18, 1973 28. Goldberg ND, O’Dea RF, Haddox MK: Cyclic GMP, in Greengard P, Robison GA (eds): Advances in Cyclic Nucleotide Research, vol. 3 New York, Raven Press, 1973. pp 155223 29. Bron Th, Rous S: Comparason de quelques effets du 3’,5’-GMP cyclique et du 3’,5’AMP cyclique sur le metabolisme glucolipidique de la souris vivante. Biochim Biophys Acta 237:156-166, 1971

FAIN ET AL.

30. Corm HO, Kipnis DM: The effect of various 3’5’ cyclic nucleotides on gluconeogenesis and glycogenolysis in the perfused rat liver. Biochem Biophys Res Commun 37:319326, 1969 31. Exton JH, Hardman JG, Williams TF, Sutherland EW, Park CR: Effects of Guanosine 3’,5’-monophosphate on the perfused rat liver. J Biol Chem 246:2658-2664, 1971 32. Helderman JH, Wilson DE, Levine RA: Comparative effects of cyclic guanosine 3’,5’monophosphate and its dibutyryl derivative on hepatic glycogenolysis. Arch Int Pharmacodyn Ther 199:389-393, 1972 33. Conn HO, Karl IS, Steiner A, Kipnis DM: Studies of the mechanism of action of 3’,5’-cyclic nucleotides on hepatic glucose production. Biochem Biophys Res Commun 45: 436-443, 197 1 34. Glinsmann WH, Hern EP, Linarelli LG. Farese RV: Similarities between effects of adenosine 3’.5’-monophosphate and guanosine 3’.5’-monophosphate on liver and adrenal metabolism. Endocrinology 85:71 l-719, 1969 35. Barnabei 0, Tomasi V, Trevisani A: Action of autonomic nerve intermediates on liver plasma membrane. Adv Enzyme Regul 9:129-143, 1971 36. Illiano G, Tell GPE, Siegel MI, Cuatrecasas P: Guawosine 3’:5’-cyclic monophosphate and the action of insulin and acetylcholine. Proc Nat1 Acad Sci USA 70:2443-2447, 1973 37. Thompson WJ, Williams RH, Little SA: Activation of guanyl cyclase and adenyl cyclase by secretin. Biochim Biophys Acta 302:329-337, 1973 38. Ward WF, Fain JN: The effects of dihydroergotamine upon adenyl cyclase and phosphodiesterase of rat fat cells. Biochim Biophys Acta 237:387-390, 1971 39. Northrop G: Effects of adrenergic blocking agents on epinephrineand 3’,5’-AMPinduced responses in the perfused rat liver. J Pharmacol Exp Ther 159:22-28, 1968 40. Tolbert MEM, Fain JN: Studies on the regulation of gluconeogenesis in isolated rat liver cells by epinephrine and glucagon. J Biol Chem 249: 1162- 1166, 1974 41. Friedmann N, Park CR: Early effects of 3’,5’-adenosine monophosphate on the fluxes of calcium and potassium in the perfused liver of normal and adrenalectomized rats. Proc Natl Acad Sci USA 61:504-508, 1968 42. Friedmann N, Rasmussen H: Calcium gluconeogenesis. manganese and hepatic Biochim Biophys Acta 222:41--52, 1970

CYCLIC NUCLEOTIDES

AND GLUCONEOGENESIS

43. Cook DE, Blair JB, Lardy HA: Mode of action of hypoglycemic agents. V. Studies with phenethylbiguanide in isolated perfused rat liver. J Biol Chem 248:5272-5277, 1973 44. Toews CJ, Kyner JL, Connon JJ, Cahill GF Jr: The effect of phenformin on gluconeogenesis in isolated perfused rat liver. Diabetes 19:(Suppl 1) 368, 1970 45. Haeckel R, Haeckel H: Inhibition of gluconeogenesis from lactate by phenylethylbiguanide in the perfused guinea pig liver. Diabetologia 8:117-124, 1972 46. Davidoff F: Guanidine derivatives in medicine. N Engl J Med 46:141-146, 1973 47. Davidoff F: Phenethylbiguanide inhibits mitochondrial Ca2+ fluxes: A mechanism for inhibition of gluconeogenesis. Diabetes 23:369, 1974 48. Adam PAJ, Haynes RC Jr: Control of hepatic mitochondrial CO, fixation by glucagon, epinephrine, and cortisol. J Biol Chem 244:64446450, 1969 49. Friedmann, N: Effects of glucagon and cyclic AMP on ion fluxes in the perfused liver. Biochim Biophys Acta 274:214-225, 1972 50. Friedmann N, Dambach G: Effects of glucagon, 3’,5’-AMP and 3’,5’-GMP on ion fluxes and transmembrane potential in perfused livers of normal and adrenalectomized rats. Biochim Biophys Acta 307:399-403, 1973 51. Exton JH, Friedmann N, Wong EH-A, Brineaux JP, Corbin JD, Park CR: Interaction

407

of glucocorticoids with glucagon and epinephrine in the control of gluconeogenesis and glycogenolysis in liver and of lipolysis in adipose tissue. J Biol Chem 247:3579-3588, 1972 52. Friedmann N, Exton JH, Park CR: Interaction of adrenal steroids and glucagon on gluconeogenesis in perfused rat liver. Biochem Biophys Res Commun 29: 113-l 19, 1967 53. Exton JH, Park CR: Control of gluconeogenesis in liver. III. Effects of l-lactate, pyruvate, fructose, glucagon, epinephrine, and adenosine 3’,5’-monophosphate on gluconeogenic intermediates in the perfused rat liver. J Biol Chem 244:1424-1433, 1969 54. Veneziale CM: Gluconeogenesis from fructose in isolated rat liver. Stimulation by glucagon. Biochemistry 10:3443-3447, 1971 55. Blair JB, Cook DE, Lardy HA: Influence of glucagon on the metabolism of xylitol and dihydroxyacetone in the isolated perfused rat liver. J Biol Chem 248:3601-3607, 1973 56. Taunton OD, Stifel FB, Greene HL, Herman RH: Rapid reciprocal changes of rat hepatic glycolytic enzymes and fructose-1,6diphosphatase following glucagon and insulin injection in vivo. Biochem Biophys Res Commun 48:1663-1670, 1972. 57. Veneziale CM, Swenson RP: Effects of cyclic purine nucleotides on phosphofructokinase (PFK) activity in isolated rat hepatocytes. Diabetes 23:384, 1974