Effect of hormones on cytosolic free calcium in adipocytes

cd

cwdum (lese)

10. 581-e87

Q Longman Gmup UK LM 1989

Effect of hormones on cytosolic free calcium in adipocytes P. F. BLACKMORE’and G. AUGERT2 Department of Pharmacology,Eastern VirginiaMedical School, Norfolk, Virginia, USA 2Nestf6 Research Centre, Nestec Ltd, Lausanne, Switzerland ’

Abstract - Some studies have indicated that insulin was able to increase the level of free cytosolic calcium in adipocytes [e.g. 71. The present study was designed to examine this phenomenon. Insulin did not increase free cytosolic calcium, however oxytocin, vasopressin, c-adrenergic agonists and ATP did increase free cytosolic caIclurn in adipocytes. Cther agonists which also did not aher calcium were epidermal growth factor, angiotensin II, glucagon, and badrenergic agonists. The effect of oxytocln at increasing free cytosolic calcium was inhibited by activation of protein kinase C with phorbol 12-myristate M-acetate and by ADP ribosylation of a Gi like protein with islet activating protein. The hormones that did increase cytosolic free calcium did so by mobilizing internal calcium and by promotingcalcium influx. Even though insulin did not increase free cytosolic calcium, it was able to attenuate the cl-adrenergicmediated increase in cytosolic free calcium. The fact that certain hormones can increase the level of the second messenger calcium in adipocytes implies that it may be a key intracellular regulator of adipocyte function as it is in many other tissues. Some studies have shown that insulin transiently increases the levels of inositol phosphates and decreases the levels of phosphoiuositides in adipocytes from epididymal fat pads [l]. Insulin was also shown to increase the level of 1.2~diacylglycerol in insulin-treated fat pads and BC3H-1 myocytes [2, 31. The effects of insulin on phosphoinositide and phospholipid turnover in adipose tissue remains contradictory [l-6], with some studies not being able to demonstrate changes in inositol phosphates [5,6]. Some data on measurements of free cytosolic calcium ([Ca’+]i) [7, 81, however, support the studies of Famse et al. [l, 41. In these experiments insulin was shown to increase adipocyte [Ca2+]i

using the fluorescent indicator Fura- 7, 81. It was suggested that insulin stimulated CaI ’ influ via voltage dependent Ca2+ channels, an effect which was potentiated by glucose [7,8]. In this present study, we investigated the ability of insulin to increase [Ca2+]iin adipocytes. Since oxytocin can mimic some of the effects of insulin in adipocytes [e.g. 9-111, and since it has recently been shown to stimulate phosphoinositide metabolism in adipocytes [6], we utilized this hormone to show that in the adipocyte system that we were using, an increase in [Ca2+]i could be observed. We show here that oxytocin together with several other agonists, but not insulin, can increase [Ca2+]iin adipocytes.

561

CELLCALCIUM

562

Materialsand Methods Adipocytepreparation

Isolated adipocytes were prepared by the collagenase method [12] from rat epididymal adipose tissue of approximately 220 g male Sprague-Dawley rats. The cells were suspended in Krebs-Ringer-HEPES (5 mM) buffer (pH 7.4) containing 20 mg/ml fraction V bovine serum albumin, 1 mM glucose, and 2.5 mM CaCh. Cells (5 ml) were incubated with gentle shaking at 37°C in 25 ml polypropylene Erlenmeyer flasks before any additions were made. Measurementof [C2+]i using Fur-a-2

Adipocyte suspensions were incubated for 45 min with 4 p.M Fur&YAM essentially as described by Draznin et al. [7]. The cells were separated from the media by centrifugation (50 g for 2 min); the cells were resuspended in fresh media (13 ml), and fluorescence measurements made 10 min after incubation. Cell suspensions (2 ml) were incubated in a polypropylene test tube with stirring in a SPEX dual excitation fluorometer at 37°C. Agents (20 pl) -

I

1001

OXYTOCINlOnb4

were added 30 s after data collection was started. Typically data were collected for 2 to 5 min with integration times of 0.1 s and time increments of 0.5 s. On completion of experiments, cells were lysed with 0.01% digitonin so as to obtain maximal fluorescence (Rmsx) of Fura-2, while the minimal fluorescence (Rmin) of Fura- was determined by adding 10 mM EGTA [13]. Autofluorescence of the cells was determined by adding 2 mM MnC12in the presence of 20 pM ionomycin. The level of adipocyte [Ca2+]i was calculated according to Grynkiewicz et al. [13]. Basal levels of [Ca2’]i were between 35 and 50 nM, with maximal concentrations of oxytocin causing the level of [Ca2+]ito transiently increase to approximately 100 nM. Representative traces are shown; all experiments were performed at least four times to assure reproducibility. Chemicals

Insulin (crystalline porcine) and glucagon were from Lilly. Oxytocin, insulin (bovine), vasopressin, epinephrine, phenylephrine, angiotensin II, verapamil, ATP, phorbol 12-my&ate 13-acetate, 4o-phorbol epidermal growth factor, digitonin, albumin fraction V, norepiuephrine, propranolol, and isoproterenol were from Sigma. Islet activating protein was from List Biochemicals. Ionomycin and Fura-2/AMwere from Calbiochem.

Results

The data in Figure 1 show that 10 nM oxytocin caused a rapid and transient increase in [Ca2+]i whereas insulin (10 nM) was without effect. Higher, 100 r&I, or lower, 0.1 nM, concentrations of 40(. 0 1 insulin did not increase [Ca2+]i;also increasing the 2 3 4 5 T-(=1 glucose concentration to 10 mM did not permit insulin to increase [Ca2+]i (data not shown). F&. 1 Effect of oxytocin (10 &I) and insulin (10 nM) on [Ca”]i Incubation of adipocytes with adenosine deaminase in isolated fat cells. Adipocyte suspensions (2ml), after loading for 30 min did not have any effect on insulin or with FuraZ,were incubatedat 37'Cina SPEXfluorometer. The oxytociu actions to increase [Ca2+]i. In Figure 2A, agents were added after 30 8 of data collection, indicated by the ElTOW. For comparison purposes the effects of oxytocin and the effect of various concentrations of oxytocin on insulin were superimpose& representative traces are shown. [Ca2+]iare shown. A small ~IICIWI.W in [Ca2+]i WAS Control incubations, vehicle alone, were tbe same as the insulin observed with 0.1 nM oxytocin while the maximum trace which did not change (data not shown) increase was seen with 100 nM. Since vasopressin

EFFESX OF HORMONES ON CYTOSOLIC FREE Ca2t IN ADIPOCYTES

563

3.

z 8

OXYTOCIN lj.~M

3*

80

3 8 E

40 1 0

0.5

I 2.0

1.5

1.0

60

0.5

0

TIiLE (MN)

40

0.5

0

1.5

1.0

0

2.0

0.5

Fig. 2

Dose response of oxytocin B) on [Ca2+]i.

observed

with a maximum

effect

(panel A).

In the case of vasopressin 1 nh4 and a maximal Both hormones

A) and vasopressin

of 0.1 nM OX~~OC~IIWAS

seen with

seen with (panel B).

(panel

A small effect

100 nM oxytocin

1.5

1.0

2.0

s

a5

oxYTocINloot-lb!

9

3c I

a detectable effect was first

effect observed

exhibited

2.0

TIbiE (MN)

TIME (MN)

(panel

1.5

1.0

TIME (MN)

with 100 nM

similar transient

effects on

[CZ12+]i

and oxytocin am related peptides, the effect of vasopressin on [Ca2+]i was measured (Fig. 2B). It can be seen that vasopmssin was slightly less effective than oxytocin since 1 nM vasopressin produced a detectable increase in [Ca2+]i whereas 1 nM oxytocin induced a large effect (Fig. 2A). Maximally effective doses of vasopressin and oxytocin were not additive; this suggests that both agonists may interact with the same receptor or that the two hormones work through the same mechanism to increase [Ca2+]. The effect of 3.0 mh4 EiTA on the ability of oxytocin to increase [Ca”]i is shown in Figure 3. EGTA had a minimal effect on the ability of 1 p.M oxytocin to increase [Ca2+]i (Fig. 3A) whereas it

25

35

30

40

TIME (SEC) Fig. 3 Effect of chelation the ability of oxytocin B) to increase nifedipiue

of extracellular

[Ca2+]i and

the effects

(10 ILM) on 100 nM oxytocin

was added to adipocyte addition of oxytocin.

suspensions

of Ni” actions.

approximately

10 nM (panel (2 mM)

and

EGTA (3.0 mM) 30 s before the

effect on 10 nM oxytocin

of 3.0 mM EGTA only depressed

slightly; this indicated

with EGTA on

EGTA had a small effect on 1 @l oxytocin

(panel A) but a substantial The addition

calcium

1 @f (panel A) and oxytocin

that very little Fura-

cells during the time course of the experiment

(panel B).

the Fura-

signal

had leaked out of the

564

CBLLCALCIUM

100

4A

F

1

3

OXYTOCIN 1OnM

5

-

40

1 0

i

60

3 8 g

30

1 0.5

1.0

1.5

2.0

0

0.5

TIME @W

1.0

1.5

I 2.0

TIMEWE) Ng. 5 Effect of IAP on the ability of 10 nM oxytocin to increase [Ca2+]i. Adipocytes were incubated with IAP for 2 h before loading with Fura-2. The total time adipocytes were incubated with IAP was 2.75 h before cells were washed to remove extracellular FuraGYAM. IAP treatment bhmted the initial response induced by oxytocin but mt the latter mponse. concentrations

of oxytocin

Higher

(100 nMj were not effected by IAF’

treatmenb while lower concentrations of oxytociu (e.g. 1 nM) were

inhibited slightly more than that seen on 10 nM oxytocin (data not shown)

-_ 0

60

30

120

TIME(SEC) Ng. 4 Effect of PMA on the ability of 10 nM oxytocin (panel A) and 1 &I oxytocin (panel B). Cells were preincubated for either 2 or 5 min with 1 pM PMA before 10 nM oxytocin was added (30 s after commencement of data collection) (panel A). Them was a time dependent effect of PMA to attenuate oxytocin action on [Ca’+]i. Preincubation for 10 min did not produce any further inhibition. The effect of 1 p.M PMA preincubated for 2 min with adipocytes on 1 r&l oxytocin action is shown in panel B

had a much larger effect to inhibit 10 nM oxytocin, especially at earlier times. These data suggest that lower concentrations of oxytocin stimulate Ca2+ influx more so than higher concentrations. Similar results were obtained when Ni2+ (2 mM), and nifedipine (10 pIvI) were used to block Ca2+ influx (Fig. 3C), the effects being more pronounced immediately after hormone addition. Preincubation of adipocytes with the protein kinase C activator phorbol 12-myristate 13-acetate (PMA) 1 ,uM for 2 or 5 min partially inhibited the effects of 10 nM oxytocin to increase [Ca2’]i (Fig. 4A). Although not shown, the inactive tumor promotor 4a-phorbol (1 p&I) was without effect on

oxytocin actions. These data indicate that protein kinase C may be involved with the down regulation of the oxytocin response and .may explain why the effect of oxytocin on [Ca2+]i was transient Lower concentrations of oxytocin (1 r&I) were completely inhibited by 1 p.M PMA (Fig. 4B). Treatment of adipocytes with islet activating protein (WI?) 100 @ml for 2 h before loading with Fura- blunted the ability of 10 nM oxytocin to increase [Ca2+]i (Fig. 5). This result suggests that a Gi like protein was involved in the signal transduction mechanism of oxytocin. This 2 h treatment with IAP resulted in the ADP ribosylation of a protein with a molecular weight of 41 Kd to an extent of 95% similar to that seen in hepatocytes 1141(data not shown). This is the molecular weight of the 01subunit of Gi [1.5]. Several other agonists which increased [Ca2+]i were the a-adrenergic agonist phenylephrine and ATP, whereas angiotensin II did not increase [Ca2+]l (Fig. 6). The b agonist isoprotemnol (10 pM), epidermal growth factor (20 nM), and glucagon (10 nM) were without effect on [Ca2+]i (data not shown). Although insulin (10 nM) did not increase [Ca2+]i by itself (Fig. l), it was able to partially

EFFECl. OF HORMONES

ON CYTOSOLIC

FREE! Ca2+ IN ADIPOCYTES

565

105

s

.5 x 5

85

z4 s a 0

65

E k! t 45 0

1.0

0.5

1.5

0

2.0

0.5

1.0

TIME (MN) Elg. 6 Effect of oxytocin

(0.1 mM) and angiotensin

(100 nh4), phenylephrine

(10 @I), AW

II (100 nM) on [Ca2+]i. See legend to

Figure 1 for other experimental

1.5

2.0

TIME @UN) cJ

45

7B 1

OXYTOCIN InY

L!

details

inhibit the effect of the a-adrenergic agonist phenylephrine (Fig. 7A). A similar effect of insulin to inhibit ol-admnergic effects in parenchymal cells was previously observed [e.g. 161. Epidermal growth factor which shams some common effects with insulin in liver [17] did not alter the effect of phenylephrine on [Ca*+]i (data not shown). Insulin (10 nM) did not modify the ability of oxytocin to increase [Ca’+]i (Fig. 7B) or vasopressin (data not shown).

I

25 1 0

0.5

Effect of insulin on the ability of 10 pM phenylephrine A) and

Adipocytes

1 &I

oxytocin

were preincubated

before the addition

(panel

B) to increase

for 5 min with

of either phenylephrine

was without effect on all doses of oxytociu

The present study shows that insulin was without effect on [Ca*+]i in freshly isolated adipocytes apart from its ability to inhibit a-adrenergic effects on [Ca*+]i. The reason why there are differences between our study and those of Draznin et al. [7, 81 is not obvious. We attempted to reproduce the Fura- measurements of Draznin et al. [7, 81 as closely as possible, however, we failed to observe any insulin effect to increase [Ca*+]i. We were able to observe an effect of insulin to stimulate glucose transport in our adipocyte preparation [6] (data not shown) which indicated that our cells were sensitive to insulin. Another piece of evidence to show that our preparation was insulin responsive was the data shown in Figure 6. These data showed that insulin pretreatment of adipocytes for 5 min could attenuate the efkcts of an a-adrenergic stimulus to increase

2.0

1.6

TIME (MIN) Fig. 7

(panel

Discussion

1.0

whereas

it inhibited

[C&.

10 nh4 insulin

or oxytocin.

Iusulin

tested (1 to 100 nM)

all doses of phenylephrine

tested with the

largest effect being seen on 10 p&f phenylephrine

[Ca*+]i but not of vasopressin. Insulin was previously shown to inhibit cl-adrenergic effects in liver but not those of angiotensin II and vasopressin [16]. This suggests that the effect of insulin to attenuate a-adrenergic responses is probably at the level of the a-admnergic receptor. Perhaps the receptor becomes phosphorylated by the tyrosine kinase activity of the insulin receptor which then attenuates binding of a-adrenergic agonists or prevents coupling of the receptor to the putative GP which activates the phosphatidyl inositol 4,Sbisphosphate specific phospholipase C [ 181. Apart from the negative findings of insulin to increase [Ca2+]i several other new observations were made in the present study. First, oxytocin was

CELL CALCIUM

566

shown to transiently increase [Ca2+]l in a dose dependent manner. This is consistent with the data of Augert and Exton [6] in which oxytocin was shown to cause the breakdown of inositol containing phospholipids and to cause inositol phosphate Other agonists which produced an generation. increase in [Ca2+JI were ATP, phenylephrine and vasopressin. These data support the notion that adipocytes contain P2 purinergic receptors and a-adrenergic receptors. Purinergic receptors have previously been identified in adipocytes [19] and ATP has been shown to increase the levels of inositol phosphates and [Ca’+]i in many cells including liver [e.g. 201. ATP (0.1 mlvl) was not as effective as oxytocin at increasing [Ca2+]l, the half concentration was maximum effective approximately 20 pM (data not shown). The effect of 10 @I phenyfephrine to increase [Ca2+]i was completely blocked by the selective a-adrenergic antagonist prazosin (1. pM) and unaffected by the fl-adrenergic antagonist propranolol (10 pM) (data not shown). The effects of norepinephrine and epinephrine to increase [Ca’+]i were also mediated by a-admnergic receptors since the increase in [Ca2+]i was completely inhibited by prazosin (data not shown). The b-adrenergic agonist isoproterenol (10 ,uM) was without effect on [Ca2*]i (data not shown). Thus only a-admnergic but not fi-adrenergic stimulation increases [Ca2+]i in adipocytes. The mechanism by which the various agonists increased [Ca2+]i was briefly examined in the present study. Chelation of extracellular calcium with EGTA blunted the increase in [Ca2+]i induced by oxytocin slightly (Fig. 3). This suggests that oxytocin can both mobilize intracellular calcium and promote calcium influx. The increase in [Ca2+]i induced by oxytocin in the presence of EGTA most likely reflects the increase in inositol phosphates previously observed [6]. The effect of EGTA to inhibit the increase in [Ca2+]i was more pronounced when lower concentrations of oxytocin were used, this phenomenon was also observedin liver cells stimulated with vasopressin [21]. Since lower concentrations of agonists were more effective at increasing [Ca2+]i via a calcium influx pathway, this suggests that the hormone receptor may couple directly to this process, and that the influx

mechanism is not regulated by a diffusible second This is messenger such as inositol phosphate. because high concentrations of agonist are needed to observe detectable increases in inositol phosphates [e.g. 221. Studies in platelets using ATP as a stimulus suggest that the agonist induces calcium influx before it mobilizes inttacellular calcium [23]. A similar phenomenon appears to occur in adipocytes (this study, Fig. 3) and hepatocytes [21]. The mechanism by which hormones control calcium entry in adipocytes is not known. One study has suggested that calcium influx can occur via volta e $+ gatcd calcium channels and/or by the Nat0 antiport [24]. Whether agonists regulate these processes is not known. Relatively high concentrations (i.e. 10 pM and above) of potential dependent calcium channel antagonists (e.g. verapamil, diltiazem and nifedipine) could slightly attenuate the actions of oxytocin and vasopressin to increase [ Ca2+]. This suggests that in the adipocyte, agoniits such as oxytocin most likely do not promote calcium influx via voltage gated channels since only very high concentrations of potential dependent cakium channel antagonists could attenuate the response slightly. In conclusion our study shows that insulin does not increase [Ca2+]i in adipocytes from epididymal fat pads, which contrasts to several other shtdies [7, 81. However, several other agonists such as oxytocin, vasopressin, ATP, and a-adrenergic agonists do increase [Ca2+]i in a transient fashion. The increase in [Ca2+]i is sensitive to protein kinase C activation since PMA attenuates the response. Also a G protein appears to be involved since the increase in [Ca2+]i was attenuated when adipocytes were treated with IAP, the ADP ribosylating activity of pertussis toxin. The mechanism by which agonists regulate calcium entry into adipocytes and the nature of the process await to be elucidated.

Acknowledgements This research was supported by a grant from the Juvenile Diabetes Foundation Ink-national and the Eastern Virginia Medical School Foundation. The skilled technical assistance of Annette Ross and Patricia Loose is gratefully acknowledged.

EFFECT OF HORMONES

ON CYTOSOLIC

561

FREE Cs2+ IN ADIPOCYTES

References 1. Farese FV. Kuo JY. Babischkin JS. Davis JS. (1986) Insulin provokes a transient activation of phospholipase C in the rat epididymal fat pad. J. Biol. Chem., 261,8589-8592. 2. Famse RV. Barnes DE. Davis JS. Standaert ML. Pollet RJ. (1984) Effects of insulin and protein synthesis inhibitors on phospholipid metabolism, diacylglycerol levels, and pyruvate dehydrogenase activity in BC3H-1 cultured myocytes. J. Biol. Chem., 259,7094-7100. 3. Farese RV. Davis JS. Barnes DE. et al. (1985) The de novo phospholipid effect of insulin is associated with increases in diacylglycerol, but not inositol phosphates or cytosolic Ca2+. Biochem. J., 231,269-278. 4. Famse RV. Sabir MA. Larson RE. Trudeau WL. III (1983) Further observations on the increases in inositide phospholipids after stimulation by ACTH, CAMP and insulin, and on discrepancies in phosphatidylinositol mass and 3ZPG4-labelling during inhibition of hormonal effects by cycloheximide. Cell Calcium, 4, 195218. 5. Pennington SR. Martin BR. (1985) Insulin-stimulated phosphoinositide metabolism in isolated fat cells. J. Biol. Cbem., 260,11039-l 1045. 6. Augert G. Exton JH. (1988) Insulin and oxytocin effects on phosphoinositide metabolism in adipocytes. J. Biol. Chem., 263,3600-3609. 7. Draznin B. Kao M. Sussman KE. (1987) Insulin and glyburide increase cytosolic free-Ca” concentration in isolated rat adipocytes. Diabetes, 36, 174-178. 8. Dmznin B. Sussman KE. Kao M. Lewis D. Sherman N. (1987) The existence of an optimal range of cytosolic free calcium for insulin-stimulated glucose transport in rat adipocytes. J. Biol. Chem., 262, 14385-14388. 9. Muchmore DB. Little SA. deHaen C. (1981) A dual mechanism of action of oxytocin in rat epididymal fat cells. J. Biol. Chem., 256, 365-372. 10. Hanif K. Goren HJ. Hollenberg MD. Lederis K. (1982) Mechanism for insulin-like activity in isolated rat adipocytes. Mol. Phsrmacol., 22, 381-388. 11 Bonne D. Cohen P. (1975) Characterization of oxytocin receptors on isolated rat fat cells. Eur. J. Biochem., 56, 295303. 12. Rodbell M. (1964) Metabolism of isolated fat cells. J. Biol. Chem., 239,375-380. 13. Grynkiewicz G. Poenie M. Tsien RY. (1985) A new generation of Ca2+ indicators with greatly improved fluorescence properties. J. Biol. Chem., 260, 3440-3450.

14. Lynch CT. Prpic V. Blackmore PF. Exton JH. (1986) Effect of islet-activating pertussis toxin on the binding characteristics of Ca2’-mobilizing hormones and on agonist activation of phosphorylase in hepatocytes. Mol. Phamtacol., 29,196-203. 15. Ui M. (1984) Islet-activating pm&in, pertussis toxin a probe for functions of the inhibitory guanine nucleotide regulatory component of adenylate cyclase. Trends Pharmacol. Sci., 5,277-279. 16. Dehaye JP. Hughes BP. Blackmom PF. Exton JH. (1981) Insulin inhibition of alpha-adrenergic actions in liver. B&hem. J., 194,949-956. 17. Bosch F. Bouscaml B. Slaton J. Blackmore PF. Exton JH. ( 1986) Epidermal growth factor mimics insulin effects in mt hepatocytes. B&hem. J., 239.523-530. 18. Berridge MJ. (1987) lnositol trisphosphate and diacylglyceml: Two interacting second messengers. Annu. Rev. Biochem., 56, 159-193. 19. Lawrence. JC. Lamer J. (1977) Evidence for alpha-adrenergic activation of phosphorylase and inactivation of glycogen synthase in rat adipocytcs. Mol. Pharmacol., 13,1060-1075. 20. Chamst R. Blackmom PF. Exton JH. (1985) Characterization of responses of isolated rat hepatocytes to ATP and ADP. J. Biol. Chem., 260, 15789-15794. 21. Blackmom PF. (1988) Hormonal stimulation of Ca2’ influx in hepatocytcs by a process not involving inositol lipid breakdown: possible direct involvement of a G protein FASEB J., 2, A1343. 22. Lynch CJ. Blackmore PF. Charest R. Exton JH. (1985) The relationships between receptor binding capacity for norepinephrine, angiotensin II, and vasopressin and release of inositol trisphosphate Ca2’ mobilization and phosphorylase activation in rat liver. Mol. Pharmacol., 28, 93-99. 23. Sage SO. Rink TJ. (1987) The kinetics of changes in intracellular calcium concentration in Fura- loaded human platelets. J. Biol. Cbem., 262, 1636416369. 24. Pershadsingh HA. Lee LY. Snowdowne KW. (1989) Evidence for a sodium/calcium exchanger and voltage-dependent calcium channels in adipocytes. FEBS Lett., 244, 89-92. Please send reprint requests tc : Dr P. F. Blackmore, Department of Pharmacology, Eastern Virginia Medical School, P.O. Box 1980, Norfolk, VA 23501, USA Received Accepted

: 17 July 1989 : 28 July 1989