218
Biochimica et Biophysics
Acta, 1043 (1990) 218-224 Elsevier
BBALIP
53357
Effects of insulin on inositol phosphate in cultured rat hepatocytes Richard Depnrtmenf
Key words:
Insulin;
A. Pittner and John N. Fain
of Biochemistry
(Revised
Phospholipase
production
University of Tennessee, Memphis,
(Received 24 July 1989) manuscript received 6 November
C; Vasopressin;
Glucagon;
Lithium
TN (U.S.A.)
1989)
ion; Inositol
phosphate;
(Rat hepatocyte)
Addition of vasopressin (100 nM) to rat hepatocytes prelabelled with 13H]inositol stimulated the production of inositol phosphates in the presence of 20 mM Li +. Preincubation of hepatocytes with insulin (50 nM) or glucagon (10 nM) had no significant effect alone but enhanced the effects of vasopressin after a lag period of at least 1 min. The effects of insulin and glucagon appeared additive in this respect. Insulin also enhanced the norepinephrine-mediated stimulation of inositol phosphate accumulation. The enhancement by insulin of the effects of vasopressin required at least OS-5 nM insulin and did not involve changes in 13H]inositol lipid labelling or IP, phosphatase activity. The effect of insulin appeared insensitive to prior treatment of hepatocytes with pertussis toxin (200 ng/ml for 18-24 h) or cholera toxin (100 ng/ml for 3-4 h). The glucagon enhancement of the effects of vasopressin was not affected by pertussis toxin but was mimicked by cholera toxin. The response of hepatocytes to vasopressin in the absence of Li + was smaller and more transient. Under these conditions a 5 min prior incubation with insulin inhibited the stimulation by vasopressin of inositol phosphate accumulation. A similar inhibitory effect of prior insulin exposure on the transient activation by vasopressin of exogenous phosphatidylinositol 4,5-bisphosphate breakdown by hepatocyte homogenates was also seen. These data indicate that insulin, although having no effect on basal inositol phosphate accumulation, can either enhance or antagonise the effects of vasopressin in primary rat liver hepatocyte cultures depending on the experimental conditions.
Introduction The breakdown of PIP, to yield IP, and diacylglycerol can be stimulated by Ca’+-mobilising hormones such as vasopressin in hepatocytes [1,2]. Recently, we have shown that glucagon, although having no significant effect by itself, markedly enhanced the ability of vasopressin to stimulate inositol phosphate synthesis [3]. It has been reported that insulin did not affect phosphoinositide metabolism in liver [4-61. Insulin also did not appear to inhibit vasopressin stimulation of intracellular Ca2+ mobilisation [7-lo] or phosphoinositide break-
Abbreviations: PIP,, phosphatidylinositol 4,5-bisphosphate; IP,, trisphosphates; IP,, myo-inositol tetraphosphates; IP,, myo-inositol myo-inositol bisphosphates; IP,, myo-inositol monophosphates; BSA, bovine serum albumin; PMA, 4P-phorbol 12g-myristate 13~acetate. Correspondence: R.A. Pittner, Department of Biochemistry, University of Tennessee, Memphis, 800 Madison Avenue, Memphis, TN 38163, U.S.A. 0005-2760/90/$03.50
0 1990 Elsevier Science Publishers
B.V. (Biomedical
down in freshly isolated rat hepatocytes [5], whereas it inhibited the actions of phenylephrine. However, we have shown that under certain conditions insulin antagonised the vasopressin-mediated inhibition of pyruvate kinase in cultured rat hepatocytes [ll]. In this study we investigated whether the stimulation by glucagon of vasopressin-induced phosphoinositide breakdown could be reversed by insulin. We actually found that insulin mimicked the effects of glucagon on vasopressin action under some conditions and that the effects of both hormones were additive. However, insulin was also found to antagonise the early stimulation (less than 1 min) of phosphoinositide breakdown by vasopressin. Materials and Methods Preparation of hepatocytes Hepatocytes were prepared from male 150-250 g Sprague Dawley rats, [12]. Hepatocytes were attached to 35 mm Falcon multiwell tissue culture plates at a density of approx. 1.25 . lo6 cells/well for 1 h in 1.5 ml of a Division)
219 modified Liebowitz L15 medium containing 10% (v/v) newborn calf serum. They were subsequently incubated in medium containing 5 pCi/ml [ 3H]inositol for 18 h and where indicated, 1 PM PMA, or 200 ng/ml of pertussis toxin. In some experiments, a comparison was made between hepatocytes that were maintained under serum and serum-free conditions. Following the 1 h plating period, the media were replaced with fresh medium containing 10% (v/v) newborn calf serum for 6-8 h. Hepatocytes were subsequently incubated in fresh media containing 5 pCi/ml [3H]inositol and either 10% serum or 0.2% (w/v) fatty acid poor bovine serum albumin for a further 12-18 h in the absence of any added hormones. For studies on PIP, breakdown in homogenates, hepatocytes were maintained in serum-free conditions on 60 mm plates at a density of approx. 2.5 . lo6 cells per plate, without [3H]inositol. The concentrations of vasopressin (100 nM), norepinephrine (20 pm), glucagon (10 nM) insulin (50 nM), and Li+ (20 nM) used, were optimal as described previously [3,11,12,18]. Measurement of inositol phosphate production Inositol phosphate fractions were prepared as described by Pittner and Fain [3] and separated on Dowex-formate columns [13]. Results are expressed as the total amount of IP,, IP, and IP, produce/plate. However, the IP, fraction also contains any inositol tetrakisphosphates formed during the incubation. Separation of phosphoinositides by TLC Lipids in the chloroform phase of hepatocyte extracts were dried down by vacuum and resuspended in 100 ~1 of chloroform. Lipids were separated on polyester backed Whatman silica-G plates. The solvent used was [90 : 90 : chloroform/ methanol/ H,O/ (28%)NH,OH 19 : 10, v/v] as described by Schacht [14]. R, values for PI, PIP and PIP, were approx. 0.75, 0.45 and 0.1, respectively, as determined using radioactive standards. Total 3H-labelled lipids in the chloroform phase were measured by scintillation counting a small aliquot of the chloroform extract that had previously been dried down. IP3 phosphatase assay IP, phosphatase activity in hepatocyte homogenates was determined as described Shears et al. [15]. Permeabilisation of hepatocytes After the appropriate incubations, the medium was aspirated and the plates were cooled on ice. The cells were permeabilised at 4O C with 1 ml sucrose buffer containing 0.25 M sucrose, 0.5 mM DTT, 10 mM Hepes (pH 7.4) and 0.075 mg/ml digitonin for 5 min and were then washed once with 1 ml of fresh sucrose buffer without digitonin [16]. To measure phosphoinositide hydrolysis, the permeabilised hepatocytes were in-
cubated for 10 min at 37 “C with 1 ml of buffer containing; 50 mM Hepes (pH 7.4), 0.5 mM EDTA, 20 mM LiCl, 0.3 mg/ml BSA, 140 mM NaCl, 6 mM 2 mM ATP and GTPyS as indicated [16]. MgCl,, Incubations were terminated by the addition of 1 ml of methanol: 2 M HCl [9: 11. Inositol phosphates were separated as described above. Determination of phospholipase C activity Cells were scraped in 2 ml of ice-cold sucrose buffer as described above, then homogenised for six 1 s strokes using a Potter-Elvehjem homogeniser fitted with a small Teflon pestle. The homogenate was centrifuged at 105 000 x g in a 50 Ti rotor for 30 min in a Beckman ultracentrifuge. The membrane pellet was resuspended in 2 ml of sucrose buffer with a Polytron homogeniser. Phospholipase C activity was determined in the soluble and membrane fractions by published procedures [17,18]. Each assay contained in a final volume of 100 ~1: 100 mM Tris-HCl (pH 7.0), 100 mM NaCl, 2 mM sodium cholate, deoxycholate (1 mM for soluble or 2 mM for membrane enzymes) and approx. 25000 dpm (25-30 nM) of [3H]PIP2, plus 10 ~1 of homogenate containing approx. 10 pg of protein. Incubations at 37°C were terminated after 10 min by the addition of 1.25 ml chloroform/methanol (1 : 2). The phases were separated after the addition of 0.5 ml of chloroform and 0.5 ml of 0.25 M HCl, by centrifugation (2300 X g for 5 min). Breakdown of PIP, was between 2-4s of added label and analysis by HPLC showed that the main product was IP,. A small amount ( < 10%) was also seen in IP, and IP,, presumably as a product of phosphatase action in the homogenates (results not shown). Phospholipase C activity was normalised against total homogenate lactate dehydrogenase activity. Lactate dehydrogenase activity rather than protein was used to compensate for variations in cell number [20]. There was no significant effect of vasopressin or insulin on total lactate dehydrogenase activity (Table III). Materials [ 3H]PIP, ([ inositol-2,3-Hlphosphatidylinositol 4,5bisphosphate, 4 Ci/mmol) and myo-[2-3H]inositol (15 Ci/mmol) were obtained from New England Nuclear Research Products; tissue culture reagents from Flow labs; tissue culture plates from Falcon; bovine serum albumin from Armour Pharmaceutical Company; AG l-X8 Dowex resin (100-200 pm mesh, formate form) from Bio-Rad; collagenase, IP, and other reagents were from Sigma. Results and Discussion Recently, we have shown that glucagon enhanced the ability of vasopressin to stimulate inositol stimulate inositol phosphate formation in the presence of 20 mM
220 TABLE
I
Insulin and glucagon enhance uasopressin-mediated
stimulation
of inositol phosphate production
Hepatocytes were incubated in medium containing 10% serum then incubated with or without vasopressin (100 nM) for 5 min and IP, from basal values due to agents alone or as % change plus glucagon for 5 min or 8 h. The results are means f S.E. for in the presence of vasopressin as compared to incubations * * P < 0.02; * * * P i 0.01; a P < 0.005; b P < 0.001. Additions
without, with glucagon (10 nM) or with insulin (50 nM) for either 5 min or 8 h and in the presence of 20 mM Li+. Results are expressed as the % change of IP,, IP,, 1s over the vasopressin response due to prior exposure to insulin, glucagon, or insulin six independent experiments. The significance of the effects of insulin and glucagon containing vasopressin alone was calculated using a paired ‘t’-test: * P i 0.05;
IPI
None
IP, and IP,
155
160
dpm 2655 % change -5* +8* -6k +12+ +46+
Glucagon for 5 min Insulin for 5 min Glucagon for 8 h Insulin for 8 h Vasopressin for 5 min
due to hormones
+ + + + + +
glucagon insulin 5 glucagon glucagon insulin 8 glucagon
5 min min and insulin 5 min 8h h and insulin 8 h
due to insulin or glucagon
-7* 3 -7+ 8 -2* 7 +11*10 +114*15 B
over that due to vasopressin 51 gb 64** 54** 30 * **
+63*30 +44*13 +98+27 +83*25 +55*10 +164*36
* *** *** *** 1 a
10% serum, insulin (50 nM), like glucagon (10 nM), had no significant effect on basal inositol phoshate accumulation in the presence of 20 mM Lif. This was true even after an 8 h exposure to insulin or glucagon. Insulin and
3500
bl 6.9 t Li+
14 13 9 6 33 ***
+82+ +59+ +184+ +168+ +71+ +297*101
+12*15 +8& 8 +20*16 +22*12 +21*12 +41*13 **
Li+, while having no significant effect alone [3]. These experiments were repeated to see whether insulin could reverse the effects of glucogen. The results in Table I show that in hepatocytes incubated in medium with
8) BSA - Li+
alone +8+ +10* +6+ +6k +183*
3 7 3 8 7 ***
% change Vasopressin Vasopressin Vasopressin Vasopressin Vasopressin Vasopressin
‘P2
/t
cl Serum - Li+
dl
Serum t Li+
t
j 3000 B t ii
l
62sOO-
b : 2000
.
::
)Iinutrr
Minutea
Winutrr
Fig. 1. Effects of insulin on the vasopressin-mediated stimulation of inositol phosphate production in the presence or absence of Li+ in hepatocytes incubated without or with serum. Hepatocytes were incubated for 18 h with [3H]inositol in the presence of 10% serum (c, d) or in serum-free medium containing 0.2% albumin (BSA) (a, b). Hepatocytes were then incubated for 5 min. in the presence (0) or absence (0) of insulin (50 nM) and in the presence (b, d) or absence (a, c) of Li+ (20 mM) and then further incubated for up to 15 min with vasopressin (100 nM). Results are expressed as the increase in dpm of total inositol phosphate production over basal values for three independent experiments. The basal values of IF’,, IP, and IP, and IP3+4 are as shown in Table I. SE. values have been omitted for the sake of clarity and are similar to those shown in Table 1.
221 the absence of Li+, the inositol phosphate response to vasopressin was maximum at 1 min and returned to near basal values by 15 min (Fig. la and c). Insulin under these conditions significantly inhibited at 15 s the actions of vasopressin (Table II), consistent with the finding that insulin could antagonise vasopressin-mediated inhibition of pyruvate kinase activity [ll]. In the presence of 20 mM Li+, a different response was noted at 5 and 15 min (Fig. lb and d). For the first minute, the increase due to vasopressin in the accumulation of inositol phosphates in the presence of Li+ was identical to that seen in the absence of Li+. However, in the presence of 20 mM Li+, there was a much greater accumulation of inositol phosphates at 5 and 15 min. Increasing the concentration of Li+ to 50 mM further increased vasopressin-stimulated inositol phosphate accumulation but the increase due to insulin was still apparent (results not shown). This suggests that insulin is not increasing Lif entry into hepatocytes but this remains a possibility. However, we know of no report suggesting insulin regulation of Lit transport. Insulin in the presence of Lif had two effects. For the first minute, insulin antagonised the effects of vasopressin in an almost identical fashion to incubations performed in the absence of Li+. At 5 or 15 min, however, insulin significantly enhanced the effects of
glucagon also did not affect the levels of PI, PIP or PIP, (results not shown). If vasopressin (100 nM) was added to hepatocytes that had previously been incubated with glucagon for 5 min or 8 h, an enhancement of the effects of vasopressin was seen, as expected [3]. Similar (although smaller) effects were also seen with insulin (Table I), in contrast to the findings of Taylor et al. [5] using freshly isolated hepatocytes. Furthermore, the enhancement seen with insulin was additive to that of glucagon (Table I). That insulin and glucagon should have similar and additive effects was unexpected. Previously we reported that insulin acutely reduced the vasopressin-mediated inhibition of pyruvate kinase in cultured hepatocytes [ll]. In those experiments hepatocytes had been maintained under serum-free conditions and in the absence of Li+. However, in the experiments shown in Table I, hepatocytes were incubated with 10% newborn calf serum and 20 mM Li+. A comparison was therefore made of different culture conditions. The results in Fig. 1 show time-courses for inositol phosphate accumulation due to vasopressin (100 nM) with or without a 5 min preincubation with insulin (50 nM) and in the presence or absence of 20 mM Li+. The responses to vasopressin of hepatocytes that had been maintained without or with serum were similar. In TABLE II Insulin inhibits vasopressin-induced response at 5 min
phosphoinositide
breakdown in the absence of Li + at IS s, while in the presence of Li +, it enhances the vasopressin
Hepatocytes were incubated for 12-18 h in the presence of 10% serum or serum-free medium containing 0.2% albumin. Hepatocytes were then incubated for 5 min without or with insulin (50 nM) or Li+ (20 mM). Vasopressin (100 nm) was added and the incubation continued for 15 s or 5 min. The basal values for the inositol monophosphates (IP,) were not affected by serum, insulin, or Li+ and were approx. 3200 dpm, while those for the inositol bisphosphates (IP,) were 215 dpm and for the inositol tris- plus tetrakisphosphates (IP, +IP,) were 200 dpm. The values for the increases due to vasopressin are the means f S.E. of five paired independent experiments and for insulin are the changes from cells incubated with vasopressin plus insulin as compared to those incubated with vasopressin alone. There was no significant effect of insulin alone on any parameter (see Table I) so those data are not shown. Significant effects of insulin, as based on paired comparison for the five independent experiments, are shown with an asterisk (P < 0.05). Additions
A due to vasopressin alone (dpm)
I Pi Albumin + vasopressin for 15 s Albumin + vasopressin for 5 min Albumin + Li+ and vasopressin for 15 s Albumin + Li+ and vasopressin for 5 min Serum + vasopressin for 15 s Serum + vasopressin for 5 min Serum + Li+ and vasopressin for 15 s Serum + Li and vasopressin for 5 min
IP2
+239*
50
+108+176
+294*
74
+1084+372
A due to insulin in the presence of vasopressin (dpm)
1PI
IP3
IP3
+69*
13
+81*16
-144*
+105*
21
+69*35
-163+118
+112*
44
+98*26
-139*
74
+349*98
+131*
34 *
+148*27
*
-30+44
92 *
+18*44
+1*12
-33*21
+2*20
+471+167
+260+
27
+88f
22
+85f
+258+
97
+108*
25
+87*29
+26Ozt
60
+96f
43
+107*19
-239*142
+305*49
+419*134
+644*223
IP2
+347*103
6
-260f124 -2185
58.
-51*
*
9*
-2O*
+15*18
-8f16
-42*26
-7+c28
+140*54
*
+79*28
8
*
-15*12
*
+100*19
*
222 vasopressin if hepatocytes were incubated with Li+ (Fig. 1). Similar results were obtained from hepatocytes that had been maintained in the presence or absence of serum. In the presence of Li+, the stimulatory effects of insulin were seen after a lag time of at least 1 min. With incubations of 1 min or less, insulin consistently antagonised vasopressin actions even in the presence of 20 mM Li+ which gave an optimal response (data not shown). Table II shows the accumulation of the individual inositol phosphates (IP,, IP, and IP3+4) under the same conditions described in Fig. 1, at 15 s and 5 min following the addition of vasopressin. A 5 min incubation with insulin alone did not significantly affect basal accumulation of any inositol phosphate fraction (data not shown). The pattern of accumulation of IP,, IP, and IP3+4 following a 15 s incubation with vasopressin was very similar for all conditions, in that IP, accumulated more than the other isomers (Table II). If hepatocytes were incubated with insulin for 5 min prior to the exposure to vasopressin for 15 s, the accumulation of IP, was significantly inhibited in the absence of Li+. The levels of IP, were also decreased in the absence of serum or Lif and there was no significant change in the accumulation of IP,. After a 5 min incubation with vasopressin in the absence of Li+, the levels of inositol phosphates were essentially the same as they were after 15 s. by In the presence of Li+, the 15 s stimulation vasopressin of IP, was reduced by insulin and the levels of IP, and IP, were affected to a lesser extent, but these decreases were quite variable and not statistically significant (Table II). In the presence of Lif, accumulation of all inositol phosphates was much greater at 5 min than at 15 s. If the hepatocytes were preincubated with insulin there was a further enhancement of the accumulation of all inositol phosphates (Table II). In view of the finding in Table II that insulin inhibited the 15 s response to vasopressin with regard to inositol monophosphate accumulation, the question arose whether insulin could inhibit phospholipase C activation in membranes seen after prior exposure of hepatocytes to vasopressin. We reported that exposure of hepatocytes to vasopressin for 5 s prior to homogenization resulted in an increased breakdown of added PIP, by both cytosol and membrane fractions [18]. We therefore incubated hepatocytes with or without insulin for 5 min in the absence of Lif prior to a 5 s exposure to vasopressin and then prepared soluble and membrane fractions from homogenates. The results in Table III indicate that prior exposure to insulin abolished the ability of homogenates incubated with vasopressin to hydrolyze added PIP,. These results suggest that the acute inhibition by insulin is due to inhibition of a phospholipase C enzyme activated by vasopressin which can use added PIP, as a substrate.
TABLE
111
Insulin reversal of the actwat~on by vasopressin of phospholipase activity as assayed wing exogenous PIP, in homogenates
C
Hepatocytes were incubated in serum-free medium without added Li+ for 5 min in the presence or absence of 50 nM insulin and then further incubated for 5 s without or with 100 nM vasopressin as indicated. Soluble and membrane fractions were isolated and assayed for phospholipase C activity in the presence of 2 mM cholate, 100 mM NaCl and either 1 mM or 2 mM deoxycholate for soluble and membrane fractions, respectively. Results are expressed as the percent change in phospholipase C and lactate dehydrogenase activity relative to the 5 s control incubations. Results are means* SE. from six independent experiments. The significance of difference between the groups was calculated using a paired t-test; * P < 0.005. Additions
Vasopressin Insulin Insulin + vasopressin
Phospholipase
C (% change
due to hormones)
soluble activity
membrane activity
LDH recovery
+50+11* +11+ 7 +9* 7
+45*10 * +1* 7 +13+ 7
-6k-l -3+5 -4+9
Insulin did not appear to alter the sensitivity of hepatocytes to vasopressin but rather the maximal response in the presence of 20 mM Li+. Vasopressin at 1 nM gave a response about half that seen with 100 nM vasopressin and the effect of insulin was about the same on a percentage basis (data not shown). The addition of 0.5 nM insulin gave a response about 40% of that seen with 50 nM insulin (data not shown). Insulin alone, or in combination with vasopressin or norepinephrine, did not appear to significantly affect the total 3H-label in inositol-containing phospholipids in the chloroform phase of hepatocyte extracts even after an 8 h exposure (data not shown). This is consistent with the findings of other [4-61. Insulin and vasopressin, alone or in combination, did not affect IP, phosphatase activity measured in hepatocyte homogenates in the presence or absence of Li+ (results not shown), in agreement with the findings of Shears et al. [15]. Therefore, insulin does not appear to be exerting its effects by changing the incorporation of [ 3H]inositol into lipids or affecting the breakdown of inositol 1,4,5trisphosphate. To see whether insulin and glucagon are exerting their effects at the level of the vasopressin or norepinephrine receptors or at a common site distal to the receptors, the effects of these hormones were determined on GTPyS-mediated stimulation of inositol phosphate production in permeabilised hepatocytes. Hepatocytes were incubated for 5 min with insulin or glucagon prior to being permeabilised by digitonin. Incubation of these hepatocytes with GTPyS (10 PM) in the presence of 20 mM Lit-stimulated inositol phosphate production (Fig. 2). Addition of vasopressin or norepinephrine did not further stimulate inositol phosphate production (results not shown). Presumably, the detergent properties of the
223
600:H
?Y 500. z
0 None
Insulin
qGlucagan
!#I
-
*
:
500 :Ez!
b z
r
IF .-+4
Fig. 2. Insulin and glucagon pretreatment stimulate GTPyS-induced stimulation of inositol phosphate accumulation in permeabilized hepatocytes. Hepatocytes were incubated without or with 50 nM insulin or 10 nM glucagon for 5 min in medium containing 10% serum and then permeabilized as described in the Methods section. They were then incubated for 10 min in the presence of 10 pM GTPyS without (open bars), with insulin pre-treatment (diagonal bars) or plus 10 nM glucagon (horizontal bars) in the presence of 20 mM Li+. The basal accumulation of IP, was approx. 850 dpm, for IP,, 210 dpm and for IP3+,, 170 dpm. The results are shown as the percent changes from basal values. The effects of insulin and glucagon are shown as the means* S.E. of five or four paired independent experiments, respectively, and were significant based on a paired ‘r ‘-test (* P -C 0.05 * * P < 0.02 * * P < 0.01).
digitonin
uncoupled
the
receptors
from
their
ap-
propriate ‘G,,’ proteins. However, preincubation of the hepatocytes with insulin or glucagon enhanced the inositol phosphate production mediated by GTPyS (Fig. 2). These results suggest that insulin and glucagon are acting at a post-receptor site, possibly ‘Gr’ or phospholipase C. To further investigate the nature of the site of glucagon or insulin action, experiments were performed in the presence of pertussis and cholera toxins. Pertussis toxin is known to ADP-ribosylate the cY-subunit of Gi, the G protein involved in the inhibition of adenylate cyclase by ligand-receptor complexes. In contrast, cholera toxin ADP-ribosylates the a-subunit of Gs, involved in the activation of adenylate cyclase [21]. Hepatocytes were incubated with pertussis toxin (200 ng/ml) for 18-24 h, or cholera toxin (100 ng/ml) for 3-4 h alone or in combination. These conditions have been reported to result in complete ADP-ribosylation of the respective G proteins in hepatocytes [22-241 and GH, cells [25]. Basal inositol phosphate accumulation was affected by less than 3% after toxin treatment (results not shown). Treatment of hepatocytes with pertussis toxin slightly enhanced the effects of vasopressin in stimulating inositol phosphate accumulation (32 of: 15%, n = 5, P < 0.05). However, this did not impair the ability of insulin or glucagon to further enhance the effects of vasopressin (Fig. 3). Cholera toxin appeared to mimic the effects of glucagon in enhancing the
Insulin
HGlucagon
looIPl
0 None
z 2
vasopressin
vasopressin t pertussis toxin
fasopressin lasopressin . t cnolera + pertussis toxin * cholera toxin
Fig. 3. Effects of pertussis toxin and cholera toxin pretreatment on the ability of insulin and glucagon to enhance vasopressin-mediated stimulation of inositol phosphate accumulation. Hepatocytes were incubated in medium containing 10% serum for 18-24 h with pertussis toxin (200 ng/ml) or 3-4 with cholera toxin (100 ng/ml) alone or in combination. Insulin (50 nM) or glucagon (10 nM) were added with 20 mM Li+ 5 min prior to the addition of vasopressin (100 nM) for 5 min. Results are expressed as the I increase in dpm of IP, + Is +4 over basal accumulation which was 384 f 30 dpm, and are means + S.E. from five independent experiments. The significance of the effects of insulin and glucagon in the presence of vasopressin was calculated using a paired ‘I ‘-test: * P < 0.02; **P x 0.05.
effects of vasopressin (129 k 37’%, n = 5, P -c 0.02). This is presumably due to the ability of cholera toxin to raise CAMP levels in hepatocytes, as has been reported by Janicot et al. [24]. Addition of glucagon to cholera toxin-treated hepatocytes did not further enhance the
-Insulin
+Insulin -PM
-Insulin
+Insulin
tPMA 5 min
-Insulin
tInsulin
+PHA 18 hrs
Fig. 4. Insulin enhancement of vasopressinand norepinephrinestimulated inositol phosphate accumulation is not altered by phorbol ester treatment. Hepatocytes were incubated in medium containing 10% serum without or with 1 pM Ph4A for 5 min or 18 h. Insulin (50 nM) was added as indicated 8 h prior to the addition of vasopressin (100 nM) or norepinephrine (20 PM) plus 20 mM Li+ for 5 min. Results are expressed as the % change of IP, +IPs+4 over basal (100%) which was 396 *44 dpm. The values are the means f S.E. from three independent experiments. The significance of the effects of insulin were calculated using a paired ‘t’-test where * signifies at least P < 0.05.
224 effects of vasopressin (Fig. 3). However, insulin under these conditions did enhance the effects of vasopressin and this effect was only slightly diminished in hepatocytes treated with both toxins. Insulin therefore, appears to be acting at a site different from that of glucagon. This would rule out the possibility of glucagon contamination of our insulin preparations. Cholera or pertussis toxin treatment also did not affect the ability of insulin to enhance the stimulation of phosphoinositide breakdown due to norepinephrine (results not shown). These results suggest that the actions of insulin are not mediated through either Gs or Gi. The actions of glucagon are not’mediated by Gi but are mimicked by activation of Gs. The data also suggest that neither Gs nor Gi is the putative Gp protein involved in vasopressin-mediated activation of phospholipase C, in agreement with the results of Uhing et al. [26]. The following experiments were performed to determine whether protein kinase C is involved in the actions of insulin. Short-term treatment of hepatocytes with PMA activates protein kinase C through translocation of the enzyme to membrane locations [1,27]. Longterm treatment down regulates protein kinase C and provides a convenient method for depleting the cell of this enzyme [27-291. Short-term treatment (5 min) of hepatocytes with PMA (1 PM) had no significant effect on vasopressin (100 nM)-mediated stimulation of IP, + accumulation but blocked the effects of norIP,+, epinephrine (Fig. 4). Long-term treatment (18 h) markedly enhanced the actions of vasopressin and norepinephrine (Fig. 4). Insulin (50 nM) treatment for 8 h enhanced the effects of vasopressin and norepinephrine (Fig. 4) after either short- or long-term treatment of hepatocytes with PMA. Similar, although smaller, effects were also seen with a 5 min exposure to insulin (results not shown). The ability of insulin to enhance the effects of vasopressin and norepinephrine therefore appear independent of the state of activation or depletion of protein kinase C. In summary, the data show that insulin, like glucagon, in the presence of Li+ enhances the vasopressinand norepinephrine-mediated stimulation of inositol phosphate production. These stimulatory effects were seen on all inositol phosphates after a lag period of at least 1 min. Since insulin also enhanced the GTPySmediated breakdown of phosphoinositides in permeabilised hepatocytes, the site for insulin stimulation of inositol phosphate accumulation probably invokes processes beyond the vasopressin or norepinephrine receptor. Acknowledgements The authors would like to thank Anita Hardeman and Doreen Enns for typing the manuscript. This re-
search was supported by a postdoctoral fellowship to R.A.P. from the Juvenile Diabetes Foundation and by U.S.P.H.S. grant (NIH DK 37004) to J.N.F.
References
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
23 24 25 26 27 28
29
Nishizuka, Y. (1986) Science 233, 305-312. Berridge, M.J. (1987) Annu. Rev. B&hem. 56, 159-193. Pittner, R.A. and Fain, J.N. (1989) Biochem. J. 257, 455-460. Creba, J.A., Downes, C.P., Hawkins, P., Brewster, G., Michell, R.H. and Kirk, C.J. (1983) Biochem. J. 212, 7333747. Taylor, D., Uhing, R.J., Blackmore, P.F., Prpic, V. and Exton, J.H. (1985) J. Biol. Chem. 260, 2011-2014. Thakker, J.K., DiMarchi, R., MacDonald, K. and Caro, J.F. (1989) J. Biol. Chem. 264, 7169-7175. Dehaye, J.P., Hughes, B.P., Blackmore, P.F. and Exton, J.H. (1981) Biochem. J. 194, 949-956. Cooper, R.H., COB, K.E. and Williamson, J.R. (1985) J. Biol. Chem. 260, 3781-3788. Corvera, S., Schwarz, K.R., Graham, R.M. and Garcia-Sainz, J.A. (1986) J. Biol. Chem. 261, 570-576. Lynch, C.J., Charest, R., Bocckino, S.B., Exton, J.H. and Blackmore, P.F. (1985) J. Biol. Chem. 260, 2844-2851. Pittner, R.A. and Fain, J.N. (1988) B&hem. J. 252, 717-721. Pittner, R.A., Fears, R. and Brindley. D.W. (1985) B&hem. J. 225. 455-462. Downes, C.P. and Michell, R.H. (1981) B&hem. J. 198, 133-140. Schacht, J. (1978) J. Lipid Res. 19, 1063-1067. Shears, S.B., Parry, J.B., Tang, E.K.Y., Irvine, R.F., Michell, R.H. and Kirk, C.J. (1987) Biochem. J. 246, 139-147. Wojcikiewicz, R.J.H. and Fain, J.N. (1988) Biochem. J. 255, 1015-1021. Bojanic, D., Wallace, M.A., Wojcikiewicz, R.J.H. and Fain, J.N. (1987) B&hem. Biophys. Res. Commun. 147, 1088-1094. Pittner, R.A. and Fain, J.N. (1989) Biochim. Biophys. Acta 1010. 227-232. Saggerson, K.D. and Greenbaum, A.L. (1969) B&hem. J. 115, 405-417. Berg, T., Boman, D. and Segler, P.O. (1972) Exp. Cell Res. 72, 517-576. Gilman, A.G. (1987) Annu. Rev. Biochem. 56, 615-649. Fujinaga, Y., Morozumi, N., Sato, K., Tokimitsu, Y., Fujinaga, K., Kondo, Y., Ui, M. and Okajima, F. (1989) FEBS Lett. 245, 117-121. Pobiner, B.F., Hewlett, E.L. and Garrison, J.C. (1985) J. Biol. Chem. 260, 16200-6209. Janicot, M., Clot, J-P. and Desbuguois, B. (1988) B&hem. J. 253, 735-743. Wojcikiewicz, R.J.H., Kent, P.A. and Fain, J.N. (1986) Biochem. Biophys. Res. Commun. 138, 1383-1389. Uhing, R.J., Prpic, V., Jiang, H. and Exton, J.H. (1986) J. Biol. Chem. 261, 2140-2146. Pittner, R.A. and Fain, J.N. (1990) Biochim. Biophys. Acta 1043, 211-217. Whetton, A.D., Monk, P.N., Consalvey, S.D., Huang, S.J., Dexter, T.M. and Downes, C.P. (1988) Proc. Natl. Acad. Sci. USA 85, 3284-3288. Stassen, F.L., Schmidt, D.B., Papadopoulos, M. and Sarau, H.M. (1989) J. Biol. Chem. 264, 4916-4923.
Biochimica et Biophysics
Acta, 1043 (1990) 218-224 Elsevier
BBALIP
53357
Effects of insulin on inositol phosphate in cultured rat hepatocytes Richard Depnrtmenf
Key words:
Insulin;
A. Pittner and John N. Fain
of Biochemistry
(Revised
Phospholipase
production
University of Tennessee, Memphis,
(Received 24 July 1989) manuscript received 6 November
C; Vasopressin;
Glucagon;
Lithium
TN (U.S.A.)
1989)
ion; Inositol
phosphate;
(Rat hepatocyte)
Addition of vasopressin (100 nM) to rat hepatocytes prelabelled with 13H]inositol stimulated the production of inositol phosphates in the presence of 20 mM Li +. Preincubation of hepatocytes with insulin (50 nM) or glucagon (10 nM) had no significant effect alone but enhanced the effects of vasopressin after a lag period of at least 1 min. The effects of insulin and glucagon appeared additive in this respect. Insulin also enhanced the norepinephrine-mediated stimulation of inositol phosphate accumulation. The enhancement by insulin of the effects of vasopressin required at least OS-5 nM insulin and did not involve changes in 13H]inositol lipid labelling or IP, phosphatase activity. The effect of insulin appeared insensitive to prior treatment of hepatocytes with pertussis toxin (200 ng/ml for 18-24 h) or cholera toxin (100 ng/ml for 3-4 h). The glucagon enhancement of the effects of vasopressin was not affected by pertussis toxin but was mimicked by cholera toxin. The response of hepatocytes to vasopressin in the absence of Li + was smaller and more transient. Under these conditions a 5 min prior incubation with insulin inhibited the stimulation by vasopressin of inositol phosphate accumulation. A similar inhibitory effect of prior insulin exposure on the transient activation by vasopressin of exogenous phosphatidylinositol 4,5-bisphosphate breakdown by hepatocyte homogenates was also seen. These data indicate that insulin, although having no effect on basal inositol phosphate accumulation, can either enhance or antagonise the effects of vasopressin in primary rat liver hepatocyte cultures depending on the experimental conditions.
Introduction The breakdown of PIP, to yield IP, and diacylglycerol can be stimulated by Ca’+-mobilising hormones such as vasopressin in hepatocytes [1,2]. Recently, we have shown that glucagon, although having no significant effect by itself, markedly enhanced the ability of vasopressin to stimulate inositol phosphate synthesis [3]. It has been reported that insulin did not affect phosphoinositide metabolism in liver [4-61. Insulin also did not appear to inhibit vasopressin stimulation of intracellular Ca2+ mobilisation [7-lo] or phosphoinositide break-
Abbreviations: PIP,, phosphatidylinositol 4,5-bisphosphate; IP,, trisphosphates; IP,, myo-inositol tetraphosphates; IP,, myo-inositol myo-inositol bisphosphates; IP,, myo-inositol monophosphates; BSA, bovine serum albumin; PMA, 4P-phorbol 12g-myristate 13~acetate. Correspondence: R.A. Pittner, Department of Biochemistry, University of Tennessee, Memphis, 800 Madison Avenue, Memphis, TN 38163, U.S.A. 0005-2760/90/$03.50
0 1990 Elsevier Science Publishers
B.V. (Biomedical
down in freshly isolated rat hepatocytes [5], whereas it inhibited the actions of phenylephrine. However, we have shown that under certain conditions insulin antagonised the vasopressin-mediated inhibition of pyruvate kinase in cultured rat hepatocytes [ll]. In this study we investigated whether the stimulation by glucagon of vasopressin-induced phosphoinositide breakdown could be reversed by insulin. We actually found that insulin mimicked the effects of glucagon on vasopressin action under some conditions and that the effects of both hormones were additive. However, insulin was also found to antagonise the early stimulation (less than 1 min) of phosphoinositide breakdown by vasopressin. Materials and Methods Preparation of hepatocytes Hepatocytes were prepared from male 150-250 g Sprague Dawley rats, [12]. Hepatocytes were attached to 35 mm Falcon multiwell tissue culture plates at a density of approx. 1.25 . lo6 cells/well for 1 h in 1.5 ml of a Division)
219 modified Liebowitz L15 medium containing 10% (v/v) newborn calf serum. They were subsequently incubated in medium containing 5 pCi/ml [ 3H]inositol for 18 h and where indicated, 1 PM PMA, or 200 ng/ml of pertussis toxin. In some experiments, a comparison was made between hepatocytes that were maintained under serum and serum-free conditions. Following the 1 h plating period, the media were replaced with fresh medium containing 10% (v/v) newborn calf serum for 6-8 h. Hepatocytes were subsequently incubated in fresh media containing 5 pCi/ml [3H]inositol and either 10% serum or 0.2% (w/v) fatty acid poor bovine serum albumin for a further 12-18 h in the absence of any added hormones. For studies on PIP, breakdown in homogenates, hepatocytes were maintained in serum-free conditions on 60 mm plates at a density of approx. 2.5 . lo6 cells per plate, without [3H]inositol. The concentrations of vasopressin (100 nM), norepinephrine (20 pm), glucagon (10 nM) insulin (50 nM), and Li+ (20 nM) used, were optimal as described previously [3,11,12,18]. Measurement of inositol phosphate production Inositol phosphate fractions were prepared as described by Pittner and Fain [3] and separated on Dowex-formate columns [13]. Results are expressed as the total amount of IP,, IP, and IP, produce/plate. However, the IP, fraction also contains any inositol tetrakisphosphates formed during the incubation. Separation of phosphoinositides by TLC Lipids in the chloroform phase of hepatocyte extracts were dried down by vacuum and resuspended in 100 ~1 of chloroform. Lipids were separated on polyester backed Whatman silica-G plates. The solvent used was [90 : 90 : chloroform/ methanol/ H,O/ (28%)NH,OH 19 : 10, v/v] as described by Schacht [14]. R, values for PI, PIP and PIP, were approx. 0.75, 0.45 and 0.1, respectively, as determined using radioactive standards. Total 3H-labelled lipids in the chloroform phase were measured by scintillation counting a small aliquot of the chloroform extract that had previously been dried down. IP3 phosphatase assay IP, phosphatase activity in hepatocyte homogenates was determined as described Shears et al. [15]. Permeabilisation of hepatocytes After the appropriate incubations, the medium was aspirated and the plates were cooled on ice. The cells were permeabilised at 4O C with 1 ml sucrose buffer containing 0.25 M sucrose, 0.5 mM DTT, 10 mM Hepes (pH 7.4) and 0.075 mg/ml digitonin for 5 min and were then washed once with 1 ml of fresh sucrose buffer without digitonin [16]. To measure phosphoinositide hydrolysis, the permeabilised hepatocytes were in-
cubated for 10 min at 37 “C with 1 ml of buffer containing; 50 mM Hepes (pH 7.4), 0.5 mM EDTA, 20 mM LiCl, 0.3 mg/ml BSA, 140 mM NaCl, 6 mM 2 mM ATP and GTPyS as indicated [16]. MgCl,, Incubations were terminated by the addition of 1 ml of methanol: 2 M HCl [9: 11. Inositol phosphates were separated as described above. Determination of phospholipase C activity Cells were scraped in 2 ml of ice-cold sucrose buffer as described above, then homogenised for six 1 s strokes using a Potter-Elvehjem homogeniser fitted with a small Teflon pestle. The homogenate was centrifuged at 105 000 x g in a 50 Ti rotor for 30 min in a Beckman ultracentrifuge. The membrane pellet was resuspended in 2 ml of sucrose buffer with a Polytron homogeniser. Phospholipase C activity was determined in the soluble and membrane fractions by published procedures [17,18]. Each assay contained in a final volume of 100 ~1: 100 mM Tris-HCl (pH 7.0), 100 mM NaCl, 2 mM sodium cholate, deoxycholate (1 mM for soluble or 2 mM for membrane enzymes) and approx. 25000 dpm (25-30 nM) of [3H]PIP2, plus 10 ~1 of homogenate containing approx. 10 pg of protein. Incubations at 37°C were terminated after 10 min by the addition of 1.25 ml chloroform/methanol (1 : 2). The phases were separated after the addition of 0.5 ml of chloroform and 0.5 ml of 0.25 M HCl, by centrifugation (2300 X g for 5 min). Breakdown of PIP, was between 2-4s of added label and analysis by HPLC showed that the main product was IP,. A small amount ( < 10%) was also seen in IP, and IP,, presumably as a product of phosphatase action in the homogenates (results not shown). Phospholipase C activity was normalised against total homogenate lactate dehydrogenase activity. Lactate dehydrogenase activity rather than protein was used to compensate for variations in cell number [20]. There was no significant effect of vasopressin or insulin on total lactate dehydrogenase activity (Table III). Materials [ 3H]PIP, ([ inositol-2,3-Hlphosphatidylinositol 4,5bisphosphate, 4 Ci/mmol) and myo-[2-3H]inositol (15 Ci/mmol) were obtained from New England Nuclear Research Products; tissue culture reagents from Flow labs; tissue culture plates from Falcon; bovine serum albumin from Armour Pharmaceutical Company; AG l-X8 Dowex resin (100-200 pm mesh, formate form) from Bio-Rad; collagenase, IP, and other reagents were from Sigma. Results and Discussion Recently, we have shown that glucagon enhanced the ability of vasopressin to stimulate inositol stimulate inositol phosphate formation in the presence of 20 mM
220 TABLE
I
Insulin and glucagon enhance uasopressin-mediated
stimulation
of inositol phosphate production
Hepatocytes were incubated in medium containing 10% serum then incubated with or without vasopressin (100 nM) for 5 min and IP, from basal values due to agents alone or as % change plus glucagon for 5 min or 8 h. The results are means f S.E. for in the presence of vasopressin as compared to incubations * * P < 0.02; * * * P i 0.01; a P < 0.005; b P < 0.001. Additions
without, with glucagon (10 nM) or with insulin (50 nM) for either 5 min or 8 h and in the presence of 20 mM Li+. Results are expressed as the % change of IP,, IP,, 1s over the vasopressin response due to prior exposure to insulin, glucagon, or insulin six independent experiments. The significance of the effects of insulin and glucagon containing vasopressin alone was calculated using a paired ‘t’-test: * P i 0.05;
IPI
None
IP, and IP,
155
160
dpm 2655 % change -5* +8* -6k +12+ +46+
Glucagon for 5 min Insulin for 5 min Glucagon for 8 h Insulin for 8 h Vasopressin for 5 min
due to hormones
+ + + + + +
glucagon insulin 5 glucagon glucagon insulin 8 glucagon
5 min min and insulin 5 min 8h h and insulin 8 h
due to insulin or glucagon
-7* 3 -7+ 8 -2* 7 +11*10 +114*15 B
over that due to vasopressin 51 gb 64** 54** 30 * **
+63*30 +44*13 +98+27 +83*25 +55*10 +164*36
* *** *** *** 1 a
10% serum, insulin (50 nM), like glucagon (10 nM), had no significant effect on basal inositol phoshate accumulation in the presence of 20 mM Lif. This was true even after an 8 h exposure to insulin or glucagon. Insulin and
3500
bl 6.9 t Li+
14 13 9 6 33 ***
+82+ +59+ +184+ +168+ +71+ +297*101
+12*15 +8& 8 +20*16 +22*12 +21*12 +41*13 **
Li+, while having no significant effect alone [3]. These experiments were repeated to see whether insulin could reverse the effects of glucogen. The results in Table I show that in hepatocytes incubated in medium with
8) BSA - Li+
alone +8+ +10* +6+ +6k +183*
3 7 3 8 7 ***
% change Vasopressin Vasopressin Vasopressin Vasopressin Vasopressin Vasopressin
‘P2
/t
cl Serum - Li+
dl
Serum t Li+
t
j 3000 B t ii
l
62sOO-
b : 2000
.
::
)Iinutrr
Minutea
Winutrr
Fig. 1. Effects of insulin on the vasopressin-mediated stimulation of inositol phosphate production in the presence or absence of Li+ in hepatocytes incubated without or with serum. Hepatocytes were incubated for 18 h with [3H]inositol in the presence of 10% serum (c, d) or in serum-free medium containing 0.2% albumin (BSA) (a, b). Hepatocytes were then incubated for 5 min. in the presence (0) or absence (0) of insulin (50 nM) and in the presence (b, d) or absence (a, c) of Li+ (20 mM) and then further incubated for up to 15 min with vasopressin (100 nM). Results are expressed as the increase in dpm of total inositol phosphate production over basal values for three independent experiments. The basal values of IF’,, IP, and IP, and IP3+4 are as shown in Table I. SE. values have been omitted for the sake of clarity and are similar to those shown in Table 1.
221 the absence of Li+, the inositol phosphate response to vasopressin was maximum at 1 min and returned to near basal values by 15 min (Fig. la and c). Insulin under these conditions significantly inhibited at 15 s the actions of vasopressin (Table II), consistent with the finding that insulin could antagonise vasopressin-mediated inhibition of pyruvate kinase activity [ll]. In the presence of 20 mM Li+, a different response was noted at 5 and 15 min (Fig. lb and d). For the first minute, the increase due to vasopressin in the accumulation of inositol phosphates in the presence of Li+ was identical to that seen in the absence of Li+. However, in the presence of 20 mM Li+, there was a much greater accumulation of inositol phosphates at 5 and 15 min. Increasing the concentration of Li+ to 50 mM further increased vasopressin-stimulated inositol phosphate accumulation but the increase due to insulin was still apparent (results not shown). This suggests that insulin is not increasing Lif entry into hepatocytes but this remains a possibility. However, we know of no report suggesting insulin regulation of Lit transport. Insulin in the presence of Lif had two effects. For the first minute, insulin antagonised the effects of vasopressin in an almost identical fashion to incubations performed in the absence of Li+. At 5 or 15 min, however, insulin significantly enhanced the effects of
glucagon also did not affect the levels of PI, PIP or PIP, (results not shown). If vasopressin (100 nM) was added to hepatocytes that had previously been incubated with glucagon for 5 min or 8 h, an enhancement of the effects of vasopressin was seen, as expected [3]. Similar (although smaller) effects were also seen with insulin (Table I), in contrast to the findings of Taylor et al. [5] using freshly isolated hepatocytes. Furthermore, the enhancement seen with insulin was additive to that of glucagon (Table I). That insulin and glucagon should have similar and additive effects was unexpected. Previously we reported that insulin acutely reduced the vasopressin-mediated inhibition of pyruvate kinase in cultured hepatocytes [ll]. In those experiments hepatocytes had been maintained under serum-free conditions and in the absence of Li+. However, in the experiments shown in Table I, hepatocytes were incubated with 10% newborn calf serum and 20 mM Li+. A comparison was therefore made of different culture conditions. The results in Fig. 1 show time-courses for inositol phosphate accumulation due to vasopressin (100 nM) with or without a 5 min preincubation with insulin (50 nM) and in the presence or absence of 20 mM Li+. The responses to vasopressin of hepatocytes that had been maintained without or with serum were similar. In TABLE II Insulin inhibits vasopressin-induced response at 5 min
phosphoinositide
breakdown in the absence of Li + at IS s, while in the presence of Li +, it enhances the vasopressin
Hepatocytes were incubated for 12-18 h in the presence of 10% serum or serum-free medium containing 0.2% albumin. Hepatocytes were then incubated for 5 min without or with insulin (50 nM) or Li+ (20 mM). Vasopressin (100 nm) was added and the incubation continued for 15 s or 5 min. The basal values for the inositol monophosphates (IP,) were not affected by serum, insulin, or Li+ and were approx. 3200 dpm, while those for the inositol bisphosphates (IP,) were 215 dpm and for the inositol tris- plus tetrakisphosphates (IP, +IP,) were 200 dpm. The values for the increases due to vasopressin are the means f S.E. of five paired independent experiments and for insulin are the changes from cells incubated with vasopressin plus insulin as compared to those incubated with vasopressin alone. There was no significant effect of insulin alone on any parameter (see Table I) so those data are not shown. Significant effects of insulin, as based on paired comparison for the five independent experiments, are shown with an asterisk (P < 0.05). Additions
A due to vasopressin alone (dpm)
I Pi Albumin + vasopressin for 15 s Albumin + vasopressin for 5 min Albumin + Li+ and vasopressin for 15 s Albumin + Li+ and vasopressin for 5 min Serum + vasopressin for 15 s Serum + vasopressin for 5 min Serum + Li+ and vasopressin for 15 s Serum + Li and vasopressin for 5 min
IP2
+239*
50
+108+176
+294*
74
+1084+372
A due to insulin in the presence of vasopressin (dpm)
1PI
IP3
IP3
+69*
13
+81*16
-144*
+105*
21
+69*35
-163+118
+112*
44
+98*26
-139*
74
+349*98
+131*
34 *
+148*27
*
-30+44
92 *
+18*44
+1*12
-33*21
+2*20
+471+167
+260+
27
+88f
22
+85f
+258+
97
+108*
25
+87*29
+26Ozt
60
+96f
43
+107*19
-239*142
+305*49
+419*134
+644*223
IP2
+347*103
6
-260f124 -2185
58.
-51*
*
9*
-2O*
+15*18
-8f16
-42*26
-7+c28
+140*54
*
+79*28
8
*
-15*12
*
+100*19
*
222 vasopressin if hepatocytes were incubated with Li+ (Fig. 1). Similar results were obtained from hepatocytes that had been maintained in the presence or absence of serum. In the presence of Li+, the stimulatory effects of insulin were seen after a lag time of at least 1 min. With incubations of 1 min or less, insulin consistently antagonised vasopressin actions even in the presence of 20 mM Li+ which gave an optimal response (data not shown). Table II shows the accumulation of the individual inositol phosphates (IP,, IP, and IP3+4) under the same conditions described in Fig. 1, at 15 s and 5 min following the addition of vasopressin. A 5 min incubation with insulin alone did not significantly affect basal accumulation of any inositol phosphate fraction (data not shown). The pattern of accumulation of IP,, IP, and IP3+4 following a 15 s incubation with vasopressin was very similar for all conditions, in that IP, accumulated more than the other isomers (Table II). If hepatocytes were incubated with insulin for 5 min prior to the exposure to vasopressin for 15 s, the accumulation of IP, was significantly inhibited in the absence of Li+. The levels of IP, were also decreased in the absence of serum or Lif and there was no significant change in the accumulation of IP,. After a 5 min incubation with vasopressin in the absence of Li+, the levels of inositol phosphates were essentially the same as they were after 15 s. by In the presence of Li+, the 15 s stimulation vasopressin of IP, was reduced by insulin and the levels of IP, and IP, were affected to a lesser extent, but these decreases were quite variable and not statistically significant (Table II). In the presence of Lif, accumulation of all inositol phosphates was much greater at 5 min than at 15 s. If the hepatocytes were preincubated with insulin there was a further enhancement of the accumulation of all inositol phosphates (Table II). In view of the finding in Table II that insulin inhibited the 15 s response to vasopressin with regard to inositol monophosphate accumulation, the question arose whether insulin could inhibit phospholipase C activation in membranes seen after prior exposure of hepatocytes to vasopressin. We reported that exposure of hepatocytes to vasopressin for 5 s prior to homogenization resulted in an increased breakdown of added PIP, by both cytosol and membrane fractions [18]. We therefore incubated hepatocytes with or without insulin for 5 min in the absence of Lif prior to a 5 s exposure to vasopressin and then prepared soluble and membrane fractions from homogenates. The results in Table III indicate that prior exposure to insulin abolished the ability of homogenates incubated with vasopressin to hydrolyze added PIP,. These results suggest that the acute inhibition by insulin is due to inhibition of a phospholipase C enzyme activated by vasopressin which can use added PIP, as a substrate.
TABLE
111
Insulin reversal of the actwat~on by vasopressin of phospholipase activity as assayed wing exogenous PIP, in homogenates
C
Hepatocytes were incubated in serum-free medium without added Li+ for 5 min in the presence or absence of 50 nM insulin and then further incubated for 5 s without or with 100 nM vasopressin as indicated. Soluble and membrane fractions were isolated and assayed for phospholipase C activity in the presence of 2 mM cholate, 100 mM NaCl and either 1 mM or 2 mM deoxycholate for soluble and membrane fractions, respectively. Results are expressed as the percent change in phospholipase C and lactate dehydrogenase activity relative to the 5 s control incubations. Results are means* SE. from six independent experiments. The significance of difference between the groups was calculated using a paired t-test; * P < 0.005. Additions
Vasopressin Insulin Insulin + vasopressin
Phospholipase
C (% change
due to hormones)
soluble activity
membrane activity
LDH recovery
+50+11* +11+ 7 +9* 7
+45*10 * +1* 7 +13+ 7
-6k-l -3+5 -4+9
Insulin did not appear to alter the sensitivity of hepatocytes to vasopressin but rather the maximal response in the presence of 20 mM Li+. Vasopressin at 1 nM gave a response about half that seen with 100 nM vasopressin and the effect of insulin was about the same on a percentage basis (data not shown). The addition of 0.5 nM insulin gave a response about 40% of that seen with 50 nM insulin (data not shown). Insulin alone, or in combination with vasopressin or norepinephrine, did not appear to significantly affect the total 3H-label in inositol-containing phospholipids in the chloroform phase of hepatocyte extracts even after an 8 h exposure (data not shown). This is consistent with the findings of other [4-61. Insulin and vasopressin, alone or in combination, did not affect IP, phosphatase activity measured in hepatocyte homogenates in the presence or absence of Li+ (results not shown), in agreement with the findings of Shears et al. [15]. Therefore, insulin does not appear to be exerting its effects by changing the incorporation of [ 3H]inositol into lipids or affecting the breakdown of inositol 1,4,5trisphosphate. To see whether insulin and glucagon are exerting their effects at the level of the vasopressin or norepinephrine receptors or at a common site distal to the receptors, the effects of these hormones were determined on GTPyS-mediated stimulation of inositol phosphate production in permeabilised hepatocytes. Hepatocytes were incubated for 5 min with insulin or glucagon prior to being permeabilised by digitonin. Incubation of these hepatocytes with GTPyS (10 PM) in the presence of 20 mM Lit-stimulated inositol phosphate production (Fig. 2). Addition of vasopressin or norepinephrine did not further stimulate inositol phosphate production (results not shown). Presumably, the detergent properties of the
223
600:H
?Y 500. z
0 None
Insulin
qGlucagan
!#I
-
*
:
500 :Ez!
b z
r
IF .-+4
Fig. 2. Insulin and glucagon pretreatment stimulate GTPyS-induced stimulation of inositol phosphate accumulation in permeabilized hepatocytes. Hepatocytes were incubated without or with 50 nM insulin or 10 nM glucagon for 5 min in medium containing 10% serum and then permeabilized as described in the Methods section. They were then incubated for 10 min in the presence of 10 pM GTPyS without (open bars), with insulin pre-treatment (diagonal bars) or plus 10 nM glucagon (horizontal bars) in the presence of 20 mM Li+. The basal accumulation of IP, was approx. 850 dpm, for IP,, 210 dpm and for IP3+,, 170 dpm. The results are shown as the percent changes from basal values. The effects of insulin and glucagon are shown as the means* S.E. of five or four paired independent experiments, respectively, and were significant based on a paired ‘r ‘-test (* P -C 0.05 * * P < 0.02 * * P < 0.01).
digitonin
uncoupled
the
receptors
from
their
ap-
propriate ‘G,,’ proteins. However, preincubation of the hepatocytes with insulin or glucagon enhanced the inositol phosphate production mediated by GTPyS (Fig. 2). These results suggest that insulin and glucagon are acting at a post-receptor site, possibly ‘Gr’ or phospholipase C. To further investigate the nature of the site of glucagon or insulin action, experiments were performed in the presence of pertussis and cholera toxins. Pertussis toxin is known to ADP-ribosylate the cY-subunit of Gi, the G protein involved in the inhibition of adenylate cyclase by ligand-receptor complexes. In contrast, cholera toxin ADP-ribosylates the a-subunit of Gs, involved in the activation of adenylate cyclase [21]. Hepatocytes were incubated with pertussis toxin (200 ng/ml) for 18-24 h, or cholera toxin (100 ng/ml) for 3-4 h alone or in combination. These conditions have been reported to result in complete ADP-ribosylation of the respective G proteins in hepatocytes [22-241 and GH, cells [25]. Basal inositol phosphate accumulation was affected by less than 3% after toxin treatment (results not shown). Treatment of hepatocytes with pertussis toxin slightly enhanced the effects of vasopressin in stimulating inositol phosphate accumulation (32 of: 15%, n = 5, P < 0.05). However, this did not impair the ability of insulin or glucagon to further enhance the effects of vasopressin (Fig. 3). Cholera toxin appeared to mimic the effects of glucagon in enhancing the
Insulin
HGlucagon
looIPl
0 None
z 2
vasopressin
vasopressin t pertussis toxin
fasopressin lasopressin . t cnolera + pertussis toxin * cholera toxin
Fig. 3. Effects of pertussis toxin and cholera toxin pretreatment on the ability of insulin and glucagon to enhance vasopressin-mediated stimulation of inositol phosphate accumulation. Hepatocytes were incubated in medium containing 10% serum for 18-24 h with pertussis toxin (200 ng/ml) or 3-4 with cholera toxin (100 ng/ml) alone or in combination. Insulin (50 nM) or glucagon (10 nM) were added with 20 mM Li+ 5 min prior to the addition of vasopressin (100 nM) for 5 min. Results are expressed as the I increase in dpm of IP, + Is +4 over basal accumulation which was 384 f 30 dpm, and are means + S.E. from five independent experiments. The significance of the effects of insulin and glucagon in the presence of vasopressin was calculated using a paired ‘I ‘-test: * P < 0.02; **P x 0.05.
effects of vasopressin (129 k 37’%, n = 5, P -c 0.02). This is presumably due to the ability of cholera toxin to raise CAMP levels in hepatocytes, as has been reported by Janicot et al. [24]. Addition of glucagon to cholera toxin-treated hepatocytes did not further enhance the
-Insulin
+Insulin -PM
-Insulin
+Insulin
tPMA 5 min
-Insulin
tInsulin
+PHA 18 hrs
Fig. 4. Insulin enhancement of vasopressinand norepinephrinestimulated inositol phosphate accumulation is not altered by phorbol ester treatment. Hepatocytes were incubated in medium containing 10% serum without or with 1 pM Ph4A for 5 min or 18 h. Insulin (50 nM) was added as indicated 8 h prior to the addition of vasopressin (100 nM) or norepinephrine (20 PM) plus 20 mM Li+ for 5 min. Results are expressed as the % change of IP, +IPs+4 over basal (100%) which was 396 *44 dpm. The values are the means f S.E. from three independent experiments. The significance of the effects of insulin were calculated using a paired ‘t’-test where * signifies at least P < 0.05.
224 effects of vasopressin (Fig. 3). However, insulin under these conditions did enhance the effects of vasopressin and this effect was only slightly diminished in hepatocytes treated with both toxins. Insulin therefore, appears to be acting at a site different from that of glucagon. This would rule out the possibility of glucagon contamination of our insulin preparations. Cholera or pertussis toxin treatment also did not affect the ability of insulin to enhance the stimulation of phosphoinositide breakdown due to norepinephrine (results not shown). These results suggest that the actions of insulin are not mediated through either Gs or Gi. The actions of glucagon are not’mediated by Gi but are mimicked by activation of Gs. The data also suggest that neither Gs nor Gi is the putative Gp protein involved in vasopressin-mediated activation of phospholipase C, in agreement with the results of Uhing et al. [26]. The following experiments were performed to determine whether protein kinase C is involved in the actions of insulin. Short-term treatment of hepatocytes with PMA activates protein kinase C through translocation of the enzyme to membrane locations [1,27]. Longterm treatment down regulates protein kinase C and provides a convenient method for depleting the cell of this enzyme [27-291. Short-term treatment (5 min) of hepatocytes with PMA (1 PM) had no significant effect on vasopressin (100 nM)-mediated stimulation of IP, + accumulation but blocked the effects of norIP,+, epinephrine (Fig. 4). Long-term treatment (18 h) markedly enhanced the actions of vasopressin and norepinephrine (Fig. 4). Insulin (50 nM) treatment for 8 h enhanced the effects of vasopressin and norepinephrine (Fig. 4) after either short- or long-term treatment of hepatocytes with PMA. Similar, although smaller, effects were also seen with a 5 min exposure to insulin (results not shown). The ability of insulin to enhance the effects of vasopressin and norepinephrine therefore appear independent of the state of activation or depletion of protein kinase C. In summary, the data show that insulin, like glucagon, in the presence of Li+ enhances the vasopressinand norepinephrine-mediated stimulation of inositol phosphate production. These stimulatory effects were seen on all inositol phosphates after a lag period of at least 1 min. Since insulin also enhanced the GTPySmediated breakdown of phosphoinositides in permeabilised hepatocytes, the site for insulin stimulation of inositol phosphate accumulation probably invokes processes beyond the vasopressin or norepinephrine receptor. Acknowledgements The authors would like to thank Anita Hardeman and Doreen Enns for typing the manuscript. This re-
search was supported by a postdoctoral fellowship to R.A.P. from the Juvenile Diabetes Foundation and by U.S.P.H.S. grant (NIH DK 37004) to J.N.F.
References
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
23 24 25 26 27 28
29
Nishizuka, Y. (1986) Science 233, 305-312. Berridge, M.J. (1987) Annu. Rev. B&hem. 56, 159-193. Pittner, R.A. and Fain, J.N. (1989) Biochem. J. 257, 455-460. Creba, J.A., Downes, C.P., Hawkins, P., Brewster, G., Michell, R.H. and Kirk, C.J. (1983) Biochem. J. 212, 7333747. Taylor, D., Uhing, R.J., Blackmore, P.F., Prpic, V. and Exton, J.H. (1985) J. Biol. Chem. 260, 2011-2014. Thakker, J.K., DiMarchi, R., MacDonald, K. and Caro, J.F. (1989) J. Biol. Chem. 264, 7169-7175. Dehaye, J.P., Hughes, B.P., Blackmore, P.F. and Exton, J.H. (1981) Biochem. J. 194, 949-956. Cooper, R.H., COB, K.E. and Williamson, J.R. (1985) J. Biol. Chem. 260, 3781-3788. Corvera, S., Schwarz, K.R., Graham, R.M. and Garcia-Sainz, J.A. (1986) J. Biol. Chem. 261, 570-576. Lynch, C.J., Charest, R., Bocckino, S.B., Exton, J.H. and Blackmore, P.F. (1985) J. Biol. Chem. 260, 2844-2851. Pittner, R.A. and Fain, J.N. (1988) B&hem. J. 252, 717-721. Pittner, R.A., Fears, R. and Brindley. D.W. (1985) B&hem. J. 225. 455-462. Downes, C.P. and Michell, R.H. (1981) B&hem. J. 198, 133-140. Schacht, J. (1978) J. Lipid Res. 19, 1063-1067. Shears, S.B., Parry, J.B., Tang, E.K.Y., Irvine, R.F., Michell, R.H. and Kirk, C.J. (1987) Biochem. J. 246, 139-147. Wojcikiewicz, R.J.H. and Fain, J.N. (1988) Biochem. J. 255, 1015-1021. Bojanic, D., Wallace, M.A., Wojcikiewicz, R.J.H. and Fain, J.N. (1987) B&hem. Biophys. Res. Commun. 147, 1088-1094. Pittner, R.A. and Fain, J.N. (1989) Biochim. Biophys. Acta 1010. 227-232. Saggerson, K.D. and Greenbaum, A.L. (1969) B&hem. J. 115, 405-417. Berg, T., Boman, D. and Segler, P.O. (1972) Exp. Cell Res. 72, 517-576. Gilman, A.G. (1987) Annu. Rev. Biochem. 56, 615-649. Fujinaga, Y., Morozumi, N., Sato, K., Tokimitsu, Y., Fujinaga, K., Kondo, Y., Ui, M. and Okajima, F. (1989) FEBS Lett. 245, 117-121. Pobiner, B.F., Hewlett, E.L. and Garrison, J.C. (1985) J. Biol. Chem. 260, 16200-6209. Janicot, M., Clot, J-P. and Desbuguois, B. (1988) B&hem. J. 253, 735-743. Wojcikiewicz, R.J.H., Kent, P.A. and Fain, J.N. (1986) Biochem. Biophys. Res. Commun. 138, 1383-1389. Uhing, R.J., Prpic, V., Jiang, H. and Exton, J.H. (1986) J. Biol. Chem. 261, 2140-2146. Pittner, R.A. and Fain, J.N. (1990) Biochim. Biophys. Acta 1043, 211-217. Whetton, A.D., Monk, P.N., Consalvey, S.D., Huang, S.J., Dexter, T.M. and Downes, C.P. (1988) Proc. Natl. Acad. Sci. USA 85, 3284-3288. Stassen, F.L., Schmidt, D.B., Papadopoulos, M. and Sarau, H.M. (1989) J. Biol. Chem. 264, 4916-4923.
