Regulation of oxytocin-induced phosphoinositide breakdown in adipocytes by adenosine, isoproterenol and insulin

Biochimica et Biophysica Acta, 1013 (1989) 73-79

73

Elsevier BBAMCR 12516

Regulation of oxytocin-induced phosphoinositide breakdown in adipocytes by adenosine, isoproterenol and insulin Horng-Mo Lee and John N. Fain Department of Biochemistry, University of Tetmessee, Memphis, TN (U.S.A.) (Received 30 January 1989)

Key words: Phosphoinositide; Adipocyte; Adenosine; Oxytocin; Catecholamine; Insulin

In rat adipocytes, the breakdown of phosphoinositides labelled by a 3 h incubation with [3Hlinositoi resulted in the accumulation of labelled inositol mono-, bis. and trisphosphates in the presence of oxytoein, vasotocin or vasopressin. Oxytocin at a concentration of I nM markedly increased phosphoinositide breakdown. Incubation of adipocytes both during the 3 h labelling and the 10 rain breakdown period in a low adenosine medium (presence of adenosine deaminase) or high adenosine medium (presence of 0.1 /tM N6-(phenylisopropyi)adenosine) (P|A) did not affect basal or iigand-stimulated phosphoinositide breakdown. The addition of I / t M PIA only during the measurement of pbosphoinositide breakdown variably stimulated basal breakdown but significantly potentiated that due to oxytocin, lsoproterenol similarly had little effect on basal but inhibited oxytocin stimulation of phosphoino~itide breakdown. Insulin did not affect basal or ligand-stimulated phosphoinositide breakdown in the low or high adenosine medium. However, in adil~cytes incubated in the absence of added adenosine deaminase or PIA, insulin stimulated basal accumulation of inositol phosphates by about 29% and inhibited that due to oxytociv by about 20%. There was no significant effect of insulin on the stimulation by vasopressin or vasotocin of phosphoifiositide breakdown. These results indicate that, in adipocytes, pbosphoinosifide breakdown stimulated by oxytoein is enhanced by adenosine, inhibited by isoproterenol and, under some conditions is inhibited by insulin.

Introduction

Both insulin and oxytocin are able to stimulate glucose metabolism in rat adipocytes [1,2] but have opposite effects on glycogen synthase activity [3]. Oxytocin [4], like vasopressin and phenylephrine [5], is able to stimulate phosphoinositide breakdown in adipocytes but the effects of insulin on phosphoinositide breakdown are unclear. Insulin has been reported by Farese et al. [6] to activate, but others have been unable to see any stimulation by insulin of phosphoinositide breakdown in adipocytes [4,5]. All groups have been able to see an increase in incerporation of [32P]Pi into adipocyte phosphoinositides [4-8]. Furthermore, Moreno et al. [9] reported that the increase in 32p uptake into adipocyte phosphoinositides due to epinephrine was inhibited, while that due to insulin was enhanced by pertussis toxin. Abbreviations: G, guanine nucleotide bi~.ding p~otein - i, inhibitory; s, stimulatory; p, phospholipase; PIA, (R)-N6-(phenylisopropyl)aden mine. Correspondence: J.N. Fain, Department of Biochemistry, University of Tennessee, 800 Madison Avenue, Memphis, TN 38163, U.S.A.

In adipocytes, adenosine inhibits adenylate cyclase activation through the inhibitory guanine nucleotide binding protein (Gi), while isoproterenol is able ~o activate adenylate cyclase through the stimulatory guanine nucleotide binding protein (Gs) [10]. It was of interest to see if activation of Gs or Gi might affect phospholipase C activation. Presumably, epinephrine, oxytocin and vasopressin stimulate phosphoinositide breakdown through a separate guanine nucleotide binding protein, usually referred to as Gp [11]. The present results indicate that the activation of phospholipase C by agonists in adipocytes is influenced by insulin, Gi activation or Gs activation to only a limited extent. Materials and Methods The (R)-N6-(phenyli~opropyl)adenosine (PIA) was purchased frotn Boehringer-Mannheim. Calf intestinal mucosa adenosine aeaminase (Type VIII 200 U/mg), synthetic oxytocin (acetate salt), synthetic arginine vasopressin (acetate salt), synthetic vasotocin (acetate salt), synthetic human angiotensin II (acetate salt), bovine insulin (24 U/rng), isoproterenol (free base) and collagenase (Type II from Clostridium histolyticum), were

0167-4889/89/$03.50 © 1989 Elsevier Science Publisher.q B.V. (Biomedical Division)

74 purchased from Sigma. ScintiVerse was purchased from Fisher and was used for all radioactivity determinations by liquid scintillation counting. Ten ad libitum fed 125-149 g male Sprague-Dawley rats were killed and the epididymal fat pads were rapidly removed for each experiment. Approx. 5 g of adipose tissue were incubated in a 37 °C orbital shaking waterbath at 200 rpm for 40-45 rain in 3 ml of buffer per g of fat containing 1 mg/ml of collagenase plus 120 mM NaCI, 4.75 mM KCI, 1.3 mM CaCI 2, 1.2 mM K2HPO4, 1.3 mM MgSO4, 24 mM NaHCO 3, 2 mM glucose, 10 mM Hepes and 3% bovine fraction V albumin. The buffer was gassed for 30 rain with a 95% 02/5% CO2 mixture, and the pH was adjusted to 7.4. Adipocytes were then filtered through a 250 ttm nylon mesh and washed free of collagenase. The adipocytes (about 5 ml packed volume) were incubated in 5 ml of bicarbonate buffer containing 100 ~tCi of [3H]inositol without, in the presence of 1 U / m l of adenosine deaminase or in the presence of both adenosine deaminase and 100nM PIA for 3 h. The presence of adenosine deaminase results in the removal of endogenous adenosine released by isolated rat adipocytes [12]. Honnor et al. [13] suggested that the level of endogenous adenosine might account for much of the variability in the response of adipocytes to ligands. Furthermore Londos et al. [14] reported that the anti. lipolytic action of insulin was enhanced in the presence of adenosine deaminase and PIA. We compared an adenosine-free incubation condition with a controlled high adenosine environment (100 nM PlA in the pr,~sence of adenosine deaminase). This analog of adenosine is not deaminated by adenosine deandnase and at a concentration of 100 nM maximally potentiates insulin-stimulated glucose oxidation over a 3 h period

[121. The adipocytes were washed twice with the respective buffers without labelled inositol but with adenosine deaminase or the adenosine analog at the end of the 3 h incubation. The cells were divided among tubes contain-

ing buffer plus 20 mM Li + with or without 4 nM insulin and incubated for 5 min before other ligands were added. The incubation was continued for 10 or 45 min. 50/~1 of sample was taken from each tube and kept in freezer for glycerol determination [15]. The reactions were stopped by adding 2.5 ml of a 2 : 1 methanol/ chloroform mixture, and further extracted with I ml 0.2 M HCI and 1 ml chloroform. The extraction mixture was .shaken vigorously and allowed to stand at room temperature for 15 rain, then centrifuged at 2800 rpm for 10 rain. 2-ml fractions were taken from the aqueous phase and neutralized to pll 7.0 with 1.5 M NH4OH. 1 ml of Dowex AG18 resin was added to the neutralized samples and the mixture loaded onto columns after 10 min. Columns were washed with 20 ml of water, then 10 ml of 60 mM sodium formate/5 mM sodium tetraborate. The [3H]inositol phosphates were eluted with 8 ml ~f 0.2 M ammonium formate/0.1 M formic acid, the [3Hi~inositol bisphosphates with 8 ml of 0.4 M ammo:iium formate/0.1 M formic acid, and the [3H]inositol tris- and tetrakisphosphates with 8 ml of 1.0 M ammonium formate/0.1 M formic acid [16]. Each 8 ml fr~',tion was divided between two 20 ml scintillation vials and ScintiVerse (13 ml) was added; the vials were shaken vigorously and then counted for 3 mill in a liquid scintillation counter. For isolation of [3H]phosphoinositides a 1.0 ml fraction of organic fraction was also taken and dried overnight in room temperature. The dried samples are counted in 5 ml of ScintiVerse for 3 rain. The data for inositol phosphates are expressed as percentage of [3H]inositol-labelled lipid in each tube at the end of the ht~'ubation to correct for variable uptake of label from t~,b~,~ to tube and one day to the next. There were between 35 000 and 50000 cpm incorporated into lipid per tube. Results

The addition of o×ytocin, vasopressin or vasotocin to rat adipocytes stimulated the accumulation of inositol

TABLE ! Comparison of the effects of insulin, adenosine.deaminase and PIA on the phosphoinositide breakdown due to agonists

Adipocyteswereincubated during the 3 h labellingperiod without, in the presenceof adenosinedeaminase(1 unit/ml) or in the presence of both adenosine deaminase and PIA (0.1 ItM). The peptide agonists were added 5 rain after the addition of insulin (4 nM) and the incubation was continued for 10 rain. The basal data are the means+ S.E. for six independent experiments, whilethe changesdue to peptides are the means=l:S.E. for the increases due to the peptides over basal. Treatment

Total inositol -- d due to phosphates as 8ngiotensin I! ~$of labelled 1/tM phosphoino.
d due to oxytocin 0.1 ~tM

d due to vasuptesAn 0.1/~M

d due ~o va~otocin 0.03/tM

Adenosinedeaminase 1 U/ml Adenosinedeaminase + PIA,0.1/tM Adenosinedeaminase + insulin4 nM AdenosiP.edeaminase+ PlA+ insulin

0.98+0.12 0.90+.0.07 1.10+.0.10 0.96+.0.06

+ 5.08:i:0.66 +5.04+.0.37 +3.'72+0.64 +4.60+0.39

+1.76+.0.39 +1.56+.0.15 +1.69+.0.35 +2.31 +0.52

+3.68+0.59 +3.73+0.35 +2.90+.0.44 +3.32+.0.53

+0.21+-0.09 +0.29-t:0.09 +0.34+.0.03 +0.41 +.0.09

si

75 .~ 2.0"

A

0.5'

"O iml

A

.._~ O

,m

D

4

0.4' O

O n_

o=

O

o

3

im

O

O

D

O

•1=

O

.e_

O

i

0.3'

.¢

1.0

_o

2

a

J~

0.2'

i

.O O

,¢ m

0

#

O

#

0.5

1

0111 C X

o

,¢

,¢ C

C

-~-

i

÷

0.0

0t/I 0.0-

•~

.~

.=

o

.0.o

,,,

x o

>~

~ o

,_

.0

_

n

o

D

o

~

o c

Fig. 1. Effect of adenosine status on ligand-stimulated accumulation of IP l, IP2 and IP.a + IP,~. Adipocytes prelabeled with [~H]innsitol without (cross-hatched bars), in the presence of 1 u n i t / m l adenosin~e deaminase (solid bars) or with adenosine deaminase + 0.1/~M PIA (stippled bars) were incubated with iigands under the same conditions for 16 min. The concentrations of the ligands were: oxytocin, 0.1 /~M; vasotocin, 0.1 /~M; vasopressin, 0.1 /~M; and angiotensin ll, 1 ~M. The values are for the accumulation of label in inositol monophosphates (IPtL inositol bisphosphates (IP2) and inositol tris- and tetrakisphosphates (IP 3 + IP4). The value are the means of six independent experiments from the same experiments shown in Table l, where total inositol phosphate accumulation is given.

phosphates over a 10 min incubation period (Table I). A 5-fold increase in phosphoinositide breakdown was seen with 100 nM oxy!ocin. Vasopressin at 100 nM gave a response about 4070, while vasotocin at 30 nM gave a response about 7570 of that seen with 100 nM oxytocin. These values compare favorably with those for activation of glucose oxidation in adipocytes, where 30 nM vasotocin gave a stimulation about 6070 and 100 nM vasopressin about 5070 of that by 100 nM oxytocin [2]. Furthermore, the maximal stimulation of glucose oxidation seen with vasopressin was less than 7070 of that seen with oxytocin [2]. Goren et al. [17] reported that adenosine deaminase markedly reduced the insulin-like effects of oxytocin on adipocytes. However, in rat adipocytes adenosine status had little effect on the stimulation of phosphoinositide breakdown by oxytocin or the other peptides (Table I and Fig. 1). In these studies the cells were labehed for 3 h in low-adenosine medium (added adenosine deaminase) or in high-adenosine medium (adenosine deaminase plus 0.I/~M PIA) and this had no effect on the extent of inositol incorporation of label into phosphoinositides (data not shown). Adenosine status also had no effect on the relative distribution of inositol phosphates between the monophosphates, bisphosphates, or tris- plus tetrakisphosphates in the presence of oxytocin, vasotocin or vasopressin (Fig. 1). We were

able to see a substantial accumulation of labelled bisphosphates (about 4070 of the value for inositol monophosphates) and tris-~-tetrakisphosphates (about 1070 of the value for monophosphates) in the presence of 100 nM oxytocin at 10 min (Fig. 1). Effects of insulin on phosphoinositide breakdown in adipocytes are somewhat unclear. However, the present studies demonstrate that insulin at most increased the accumulation of total inositol phosphates by 2070 under basal conditions or in the presence of angiotensin II (Table II). These effects were only seen in cells labelled and incubated in the absence of adenosine deaminase or PIA. However, we observed a 2070 decrease with insulin in the presence of oxytocin but not vasopressin or vasotocin of inositol phosl~hate accumulation in cells incubated without adenosine deaminase (Table ll). There was no synergism between high adenosine and insulin with regard to inositol phosphate accumulation. Just the opposite was seen, as the effects of insulin were not statistically significant in cells labelled and incubated with adenosine deaminase eaher in the absence or presence of 0.1 #M PIA (Table II). Axelrod et al. [18] have reported that angiotensin II was almost twice as effective a stimulator of PGI2 production in rat adipocytes as norepinephrine or vasopressin. There are no stimulatory effects of angiotensin Ii or vasopressin on lipolysis [18] so presumably

76 TABLE !I

TABLE !II

The effects of insulin on the responses to ligands at different adenosine values

Inhibition by isoproterenoi of phosphoinositide breakdown due to oxytocin

The procedure was as described in Table I. The data are the means ± S.E. for six independent experiments. The changes due to insulin in the absence or presence of the various peptides are the means + S.E. of the paired differences. Significant effects of insulin are indicated by an asterisk (P < 0.05).

Adipocytes were incubated during the 3 h labelling period and during measurement of phosphoinositide breakdown in the presence of adenosine deaminase (1 unit/ml). Isoproterenol (1 ~M), PIA (1 pM), and peptide agonists were added 5 min after the addition of insulin (4 nM). The incubation was continued for 10 min. The data are the means± S.E. of six independent experiments. The changes due to isoproterenoi or isoproterenol plus insulin in the absence or presence of the various peptides are the means ± S.E. of the • change. Significant effects of isoproterenol or isoproterenol plus insulin are indicated by: * P < 0.05; ** P < 0.025; and t P <0.01.

Additions

Total inositol phosphates as $ of labelled phosphoinosifides

A. Adipocytes incubated in the abs~:nce of adenosine deaminase or PIA None 0.964-0.10 Oxytocin, 0.1 p M 6.50 ± 0.57 Vasotocin, 0.03/~ M 4.93 4. 0,52 Vasopressin, 0.1 ~ M 3.08 + 0,31 Angiotensin 11, 1/~M 1.63+0.17 B. Adipocytes incubated in the presence of 1 U / m l of adenosine deaminase None 0.98 +0.12 Oxytocin 5.88 ± 0.73 Vasotocin 4.49 ± 0.56 Vasopressin 2.74 4-0.37 Angiotensin I1 1.19 4-0.06

A due to insulin 4 nM

Additions +0.21 +0.07 * - 1.22 ± 0.47 * - 0.45 ± 0.38 - 0.04 4. 0.25 +0.31 +0.10 *

+ 0.13 ± 0.13 - 0.57 ± 0.21 - 0.19 ± 0.19 + 0.24 ± 0.15 + 0.25 + 0.12

C. Adipocytes incubated in the presence of I U / m l of adenosine deaminase plus 0.1 #M PIA Basal 0.~4)± 0.07 + 0.07 4-0.03 Oxytocin 5.94 4-0.33 - 0.37 ± 0.22 Vasotocin 4.614- 0.26 + 0.03 ± 0.28 Vasopressin 2.49 4-0.10 + 0.34 ± 0.15 Angiotensin II 1.19 ± 0.09 + 0.18 ± 0.15

Total inositol ~ change phosphates due to as ~ of isoproterlabelled enol phospho1 itM inositides

~$change due to isoproterenol + insulin, 4 nM

Adipocytes incubated without PIA None 0.83 +8.74.17.0 +22.5±25 Oxytocin, 1 nN[ 3.39 - 2 1 . 4 ± 7.5 * - 2 7 . 2 ± 9.0 * Oxytocin, 10nM 6.48 - 1 5 . 8 ± 3.5 * * - 1 5 . 9 ± 6.3 Oxytocin, 100 nM 7.31 - 9.2 4- 5.7 - 3.5 ± 9.9 Adipocytes incubated in the presence of 1 ~M PIA Basal 0.87 + 18.8 + 18.0 + 16.8 ± 27.0 Oxytocin, 1 nM 4.30 - 31.6± 5.0 t - 32.8± 6.3 t Oxytocin, 10nM 9.20 - 2 5 . 0 + 5.5 * * - 3 1 . 0 ± 7.8 ** Oxytocin, 100 nM 9.03 - 13.5 ± 2.7 * - 20.5 ± 3.8 t

only if PIA was present. The inhibition by isoproterenol was the same in the presence as in the absence of 4 nM insulin (Table III). TABLE IV

the stimulation of PGI., formation is related to the release of arachidonic acid from adipocyte phospholipids secondary to activation of phospholipase A2 or C by angiotensin II. However, the data in Table III as compared to those in Table I clearly demonstrate that 1 nM oxytocin is at least 2.5-times more effective than 1000 nM angiotensin II as an activator of phosphoinositide breakdown in adipocytes. Oxytocin at a concentration of 10 nM gave a response 87~ and 1 nM 40~ of that seen with 100 nM oxytocin (Table III). The small stimulation of inositol phosphate accumulation due to 1 ~tM angiotensin II was enhanced by the combination of PIA plus insulin (Table I). In hepatocytes, the response to vasopressin was specifically enhanced by incubation of the cells with an active analog of cyclic AMP or glucagon which elevated cyclic AMP [19]. In adipocytes, isoproterenol, a stunulator of cyclic AMP accumulation, either in low- or high-adenosine conditions actually inhibited by 15-33~ the response to low concentrations (1 or 10 riM) oxytocin (Table If). However, at 100 nM oxytocin the inhibitory effect of isoproterenol was statistically significant

Potentiation by PIA of oxytocin-induced breakdown of phosphoinositides Adipocytes were incubated during the 3 h labelling period and during measurement of phosphoinositide breakdown in the presence of adenosine deaminase (1 unit/ml). Isoproterenol (1/~M), PIA (1 pM) and oxytocin were added 5 rain after the addition of insulin (4 nM) and the incubation was continued for 10 rain. The data are the means 4-S.E. of the percentage change due to PIA from the six paired experiments shown in Table III. Significant effects of PIA (1/~M) are indicated by: * P < 0.02 and * * P < 0.005. Additions

Total inositol phosphates as • of labelled phosphoinositides

Adipocytes incubated without isoproterenol None 0.83 Oxytocin, 1 nM 3.39 Oxytocin, 10 nM 6.48 Oxytocin, 100 nM 7.31 Adipocytes incubated in Basal Oxytocin, I nM Oxytocin, 10 nM Oxytocin, 100 nM

~ change due to PIA

+ 8.2 ± + 29.0 ± + 40.0 ± + 29.0 ±

14.0 5.0 * * 6.3 * * 8.7 *

the presence of 1 ~tM isoproterenol 0.88 + 17.5 ± 11 2.56 + 14.8 ± 10 5.42 +23.4± 6 * 6.46 + 24.0 :i: 7.6 *

77

-

[00

m

II

l

| 150[-

[

o Basel

A ~. t~ lsopPoLe,enol n~. t ~ lnpPoLercnol ' 4.n' 2n.ulln

/ N

[ J |

m t00 S a

50 .....

p.

CD

0

0

~' . . . . .

O .....

L

tO

~

t00

....

"

nH Oxytocln Fig. 2. Effect of oxytocin on basal and isoproterenol-stimulated lipolysis. Adipocytes prelabdled with [3H]inositol in the presence of 1 unit/ml adenosine deaminase were incubated with ligands ,~'or10 min in the presence of 1 pM PIA and the indicated concentrations of oxytocin without (circles), plus 1/tM isoproterenol (triangles) or 1 ~tM tsoproterenol plus 4 nM insulin (squares). The values are the means of the same six independent experiments shown in Tables III and IV, where total inositol phosphate accumulation is given. There was no significant effect of isoproterenol alone on phosphoinositide breakdown (Table III). Isoproterenol was a potent activator of lipolysis over the 10 min incubation and the lipolytic action of isoproterenol was unaffected by 100 n M oxytocin in the presence of PIA (Fig. 2). There was also no effect of any concentration of oxytocin on basal or stimulated lipolysis (Fig. 2), In the studies shown in Table iII, PiA if added only during the 10 min period used for measurement of phosphoinositide breakdown, increased this breakdown. Analysis of the paired differences revealed a 23-40% increase due to 1 # M PIA of phosphoiinositide break-

down in the presence of 10 nM or greater oxytocin (Table iV). A comparison of the effects of 0.01, 0.1 vs. 1.0 #M PIA on lipolysis and phosphoinositide breakdown over 45 min is shown in Table V. The longer incubation period was used so that lipolysis could be more accurately measured in order to compare the dose-response curves for inhibition of lipolysis versus stimulation of oxytocin-induced phosphoinositide turnover. A lower concentration of isoproterenol (0.1 #M) was also used, since the anti-lipolytic action of PIA is reversed at high concentration of lipolytic agents [20]. The results in Table V indicate that near-maximal effect of PIA on lipolysis were seen with 0.1 #M PIA and no greater effect of 1 # M PIA was noted. A similar dose-response curve was seen for stimulation of phosphoinositide breakdown by PIA in the presence of oxytocin (Table V). In these studies there was a lesser effect of PIA on potentiation of the oxytocin-induced PI turnover than noted in Table IV Out a greater effect on turnover in the presence of isoproterenol plus oxytocin (Table V). Furthermore, the data in Table V demonstrate that 10 nM oxytocin did not stimulate or inhibit lipolysis during the 45 min incubation period. The anti-lipolytic effect of insulin, is reversed at high concentrations of lipolyt!c agents [21,22]. In the studies shown in Fig. 2 there was no antilipolytic effect of insulin in the presence of 1 # M isoproterenol. We therefore examined the antilipolytic action of insulin in the presence of 10 nM oxytocin and 0.1 # M isoproterenol in 10-min incubations using the same conditions as in Table V. Basal lipolysis in the presence of 10 nM oxytocin and adenosine deaminase was 68 nmo'., or glycerol. This was increased 136 + 3% (mean + S.E. of four experiments) by 0.1 # M isoproterenol in the absence and 59 + 6% in the presence of 4 nM insulin.

TABLE V Effect of valying PIA concentrations on lipolysis anc~phosphoinositide breakdown

Adipocytes were incubated during the 3 h labelling period and during measurement of phosphoinositide breakdown in the presence of adenosine deaminase (1 unit/ml). Agonists were added at the start of the 45 rain incubation. The data are the means+ S.E. of the 70 changes in five paired experiments for the % change due to PIA. Significant effects of PIA are indicated by: * P < 0.05 and * * P < 0.01. Additions

70change due to PIA basal 0.01 #M

0.1 #M

1.0 gM

None Oxytocin 10 nM Oxytocin 10 nM+ isoproterenol 0.1 #M

1'oral inositol phosphates as 70 of labelled phosphoinositides at 45 min +19.3+_9.7 +25.64__ 5.5 ** 27.55:13.1 1.90 +8.4+8.8 + 15.95:12.0 +27.05:il.0 * 12.58 +20.25:4.9 ** +43.2+11.0 ** +41.7+i6.4 * 8.43

None Oxytocin 10 nM Oxytocin 10 nM+ isoproterenol 0.1 #M

Glycerol accumabtion o~er 45 rain (nmol) 184 -40.2=1=11.6 * -44-t-10.3 ** 155 -28.6+10.2 * -36+. 7.9 ** 602 -22.6-1-10.8 -345: 9.3*

-40.6+10.5 ** -35.65:7.8 ** -37.2-t- 8.1 **

78 These results indicate that the adipocytes are responsive to insulin under appropriate conditions. Discussion

These results indicate that oxytocin stimulates phosphoinositide breakdown in adipocytes. Oxytocin at 1 nM gave a response 40~ of that seen with a maximal concentration of oxytocin (100 riM). We also observed significant accumulation of inositol bis- and trisphosphates in the presence of 20 mM Li +, in contrast to Augert and Exton [4]. Pennington and Martin [5] first reported stimulation of inositol phosphate accumulation by 100 nM vasopressin that was equivalent to that seen with 10 ~M phenylephrine. Our results provide further evidence that the effects of insulin on phosphoinositide breakdown in adipocytes are minimal in agreement with the results of Augert and Exton [4] and Pennington and Martin [5]. We were able to observe a small stimulation by insulin of phosphoinositide turnover that was less than 5~ that due to oxytocin and was seen under only one experimental condition (Table II). We were unable to see any transient stimulation by insulin of poly(phosphoinositide) breakdown (unpublished experiments) as reported by Farese et al. [6]. Apparently, the consistent stimulation by insulin of :2p uptake into phosphoinositides of adipocytes [4-9] is unrelated to phosphoinositide breakdown. Oxytocin is an insulin-like agonist in adipocytes that stimulates glucose oxidation [1,2]. With regard to lipolysis, the effects of oxytocin are inconsistent with both lipolytic and anti-lipolytic effects reported in adipocytes [1,23]. We found no effect of oxytocin on lipolysis under our experimental conditions (Fig. 2 and Table V). In fact, the anti-lipolytic action of insulin was seen in the presence of oxytocin. The present results indicate that insulin does not mimic oxytocin stimulation of phosphoinositide breakdown in adipocytes. Similar differences are seen with respect to glycogen synthase, sir~ce oxytocin inhibits while insulin stimulates this enzyme [3]. Hormones that stimulate phosphoinositide breakdown in adipocytes, such as a-1 catecholamine agonists also inhibit glycogen synthase [24], and this is presumably related to elevation of intracellular Ca 2+ and diacylglycerol. In contrast, insulin has both glucose-dependent and -independent stimulatory effects on adipocyte glycogen synthase [24]. There are no clearly established links between elevated intracellular concentrations of Ca 2+ or d[e-cylglycerol in adipocytes and glucose metabolism or lipolysis [11]. However, Stralfors [25] has postulated that insulin stimulation of glucose uptake is mediated by diacylglycerol in adipocytes. He believes that diacylglycerol is elevated by insulin in adipocytes secondary to action of a phospholipase C that acts on a phospho-

inositide glycan. This putative pathway for insulin action has been proposed by Saltiel [26,27]. In fact, before this hypothesis was presented one group reported that insulin increased the activity of a phospholipase C enzyme in adipocytes that degraded phosphatidylinositol [28]. However, il~ this ~aboratory we have been unable to confirm those findings in adipocytes (Etindi, R. and Fain, J., unpublished data). We have had equally poor success, as shown by the present results, in seeing consistent or major stimulation by insulin of the accumulation of any inositol containing compound with a sufficient turnover for its precursor to be labelled during a 3 h incubation of adipocytes with myo-inosit¢l. We could not determine by HPLC separation or Dowex chromatography what inosi~ol-containing product was stimulated by insulin in the studies shown in Table II, since the effect of insulin was so small. PIA has also been reported to increase 32p uptake into phosphoinositides to the same extent as insulin or oxytocin [7]. However, like insulin, this was not accompanied by any major stimulation of basal phosphoinositide turnover in our hands (Tables IV and V). We do see a reproducible potentiation of oxytocin-induced phosphoinositide turnover by adenosine. What this involves is unclear, but it is quite different from what is seen in other cells such as rat pituitary tumor cells where PIA, through receptors of the A 1 type, inhibits agonist-stimulated phosphoinositide breakdown [29,30]. Adenosine receptors of the A1 type are linked to inhibition of adenylate cyclase and are the adenosine receptors present in adipocytes. In contrast, A2 adenosine receptors are associated with activation of adenylate cyc!ase and are found in brain, where histamine-stimulated phosphoinositide breakdown is potentiated by adenosine [31,32]. The stimulation by oxytocin of phosphoipositide breakdown is in agreement with direct measurements of adipocyte Ca 2+ recently reported by Blackmore and Augert [33]. They found no effect of insulin alone, but an inhibition of a-1 adrenergic stimulation of intracellular Ca 2+ accumulation just as we saw inhibition under one condition of oxytocin-induced phosphoinositide turnover (Table II). The present results indicate that Gs activation by the fl-catecholamine receptor inhibits oxytocin-induced phosphoinositide turnover. Whether tiffs is a direct effect mediated through Gs or indirect through elevations in cyclic AMP, fatty acids released during lipolysis, or both, is_jmol~,-_.._T~rnechanism by which concurrent activation of Gi by adenosine is able to enhance phosphoinositide turnover inducer,~ by oxytocin is also unclear. In conclusion, these results indicate that oxytocin stimulates phosphoinositi~e breakdown in adipocytes and its effects are opposi'l.e to those of insulin on this parameter. It is not knowa how the oxytocin receptor is

79

able to turn on at least two separate types of response in adipocytes, since a-1 adrenergic activation in adipocytes stimulates phosphoinositide breakdown but not glucose oxidation [11]. In contrast, insulin is a potent stimulator of glucose transport and oxidation, but not of phosphoinositide breakdown. Acknowledgments The authors wish to thank Doreen Enns for excellent secretarial assistance. This work was supported by a USPHS grant (NIH DK 37004). References 1 Muchmore, D.B., Little, S.A. and De Haen, C. (1981) J. Biol. Chem. 256, 365--372. 2 Hanif, K., (]area, H.J., Hollenberg, M.D. and Lederis, K. (1982) Mol. Pharmacol. 22, 381-388. Stephenson, F.A. and Rogol, A.D. (1984) Arch Biochem. Biophys. 234, 230-234. 4 Augert, G. and Exton, J.H. (1988) J. Biol. Chem. 263, 3600-3609. 5 Pennington, S.R. and Martin, B.R. (1985) J. Biol. Chem. 260, 11039-11045. 6 Farese, R.V., Kuo, J.Y., Babischkin, J.S. and Davis, J.S. (1986) J. Biol. Chem. 261, 8589-8592. 7 Honeyman, T.W., Strohnsnitter, W., Scheid, C.R. and Schimmel, R.J. (1983) Biochem. J. 212, 489-498. 8 Garcia-Sainz, J.A. and Fain, J.N. (1980) Biochem. J. 186, 781-789. 9 Moreno, F.J., Mills, I., Garci~t-Sainz, J.A. and Fain, J.N. (1983) J. Biol. Chem. 258, 10938-10943. 10 Fain, J.N. and Garcia-Sainz, J.A. (1983) J. Lipid Res. 24, 945-966. 11 Fain, J.N., Wallace, M.A. and Wojcikiewicz, R.J.H. (1988) FASEB J. 2, 2569-2574.

12 Fain, J.N. and Wieser, P.B. (1974) J. Biol. Chem. 250, 1027-1034. 13 Honnor, R.C., Dhillon, G.S. and Londos, C. 0985) J. Biol. Chem. 260, 15122-15129. 14 Londos, C., Honnor, R.C. and Dhillon, G.S. (1985) J. Biol. Chem. 260, 15~39 ~.5145. 15 Vaughan, M. (1962) J. Biol. Chem. 237, 3354-335g. 16 Downes, C.P. and Michell, R.H. (1981) Biochem. J. 198, 133-140. 17 Goren, H.F., Hanif, K., Dudley, R., Hollenberg, M.D. and Lederis, K. (1986) Reg. Peptides 16, 125-134. 18 Axelrod, L., Minnich, A.K. and Ryan, C.A. (1985) Endocrinology 116, 2548-2553. 19 Pittner, R.A. and Fain, J.N. (1989) Biochem. J. 257, 455-460. 20 Fain, J.N., Pointer, R.H. and Ward, W.F. (1972) J. Biol. Chem. 247, 6866-6872. 21 Fain, J.N., Kovacev, V.P. and Scow, R.O. (1966) Endocrinology 78, 773-778. 22 Fain, J.N. and Rosenberg, L. (1972) Diabetes 21, Suppl. 2, 414-425. 23 Goren, H.J., Hanif, K., Hollenberg, M.D. and Lederis, K. (1982) Ann. N.Y. Acad. Sci. 394, 625-629. 24 Garcia-Sainz, J.A. and Fain, J.N. (1980) Mol. Pharmacol. 18, 72-77. 25 Stralfors, P. (1988) Nature 335, 554-556. 26 Saltiel, A.R., Fox, J.A., Sherline, P. and Cuatrecasas, P. (1986) Science 233, 967-972. 27 Fox, J.A., Soliz, N.M. and Saltiel, A.R. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 2663-2667. 28 Koepfer-Hobelsberger, B. and Wieland, O.H. 098,2.) Mol. Cellular Endocrinol. 36, 123-129. 29 Delahunty, T.M., Cronin, M.J. and Linden, J. (1988) Biochem. J. (1988) 255, 69-77. 30 Linden, J. and Delahunty, T.M. (1989) Trends Pharmacol. Sci. 10, 114-119. 31 Hollingsworth, E.B., De La Cruz, R.A. and Daly, J.W. (1980) Eur. J. Pharmacol. 122, 45-50. 32 Hill, S.J. and Kendall, D.H. (i987) Br. J. Pharmacol. 91,661-669. 33 Blackmore, P. and Augert, G. (1988) J. Cell. Biology 107, 499a (abstract).