Pertussis toxin effects on adenylate cyclase activity, cyclic AMP accumulation and lipolysis in adipocytes from hypothyroid, euthyroid and hyperthyroid rats

619

~j~c~~~jcu et Biophysiea Acta 876 (1986) 619430

Ekevier

BBA 51246

Pertussis toxin effects on adenylate cyclase activity, cyclic AMP accumulation and lipolysis in adipoeytes from hy~thyroid,

euthyroid and hy~~hyroid

rats

Ira Mills a,*, J. Adolf0 Garcia-Sainz b and John N. Fain ay** *Departmeni

of Bioehemist~,

Brown ~niversjty, Providence, RI 02906, and University of Tennessee, Memphis, (U.S.A.), and b Departamento de Bivenerg&ea, Institute de Fisiologia Cehdar, Universidad National Autbnoma de Mkxico, Mexico City (Mexico) (Received

Key words:

Pertussis

toxin; Adenylate

November

TN

12&h, 1985)

cycle; cyclic AMP accumulation;

Lipolysis;

Thyroid

hormone;

(Rat adipocyte)

Adipocytes from hypothyroid rats have a decreased responsiveness to agents that activate adenylate cyciase, whereas cells from hyperthyroid rats have an increased responsiveness as compared to the controls. This is reflected in cyclic AMP accumulation as well as lipolysis. Administration of pertussis toxin to rats or its in vitro addition to adipocytes increased basal lipolysis and cyclic AMP accumulation as well as the response to norepinephrine or forskolin. The effects of thyroid status was not abolished by toxin treatment. Pertussis toxin-catalyzed ADP ribosylation of Ni was increased in adipocyte membranes from hypothyroid rats as compared to those from euthyroid rats. However, no change in sensitivity to N6-(phenylisopropyl)adenosine was observed. The data suggest that the amount of Ni might not be rate-limiting for the inhibitor action of adenosine. A consistent decrease in maximal lipolysis was observed in freshly isolated adipocytes from hypothyroid animals as compared to those from the controls. Such defective maximal lipolysis was not corrected by adenosine deaminase or in vivo administration of pertussis toxin. The relationship betvveen cyclic AMP levels and lipolysis suggests that in fat cells from hy~thyroid rats either the cyclic AMP-dependent protein kinase or the lipase activity itself may limit maximal lipolysis. There appears to be multiple effects of thyroid status on lipolysis involving factors other than those affecting adenylate cyclase activation.

Introduction It is well known that adipocytes obtained from hypothyroid rats have a reduced responsiveness to beta-adrenergic agents and all other lipolytic stimuli as compared to cells obtained from euthyroid animals [l-4]. However, the mechanism(s) involved in such reduced responsiveness are incompletely understood. * Present address: Department of Medicine, Harvard University Medical School, Boston, MA 02115, U.S.A. ** To whom correspondence should be addressed at (present address): Department of Biochemistry, College of Medicine, University of Tennessee at Memphis, 800 Madison Avenue, Memphis, TN 38163, U.S.A.

0005-2760/86/$03.50

0 1986 Elsevier Science Publishers

Reduced lipolytic responsiveness is associated with a smaller accumulation of cyclic AMP in response to hormones [l-4]. It has been suggested that defective adenylate cyclase activation is responsible for the observed defects [l-4]. Hypothyroidism induces alterations at the receptor level [4] and also at the level of the proteins that couple receptors to the catalytic subunit of adenylate cyclase, the so-called ‘N’ proteins [5]. We have carried out studies to examine the effect of thyroid status on the ability of adipocytes to increase cyclic AMP levels and lipolysis. Our studies suggest that multiple defects are present in hypothyroidism and include changes beyond the generation of cyclic AMP.

B.V. (Biomedical

Division)

620

Materials and Methods Female Sprague-Dawley rats (150-200 g) of the Charles River CD strain were used. Thyroid status was manipulated as described previously [2]. Hypothyroidism was induced by maintaining rats on an iodine-deficient diet (U.S. Biochemical Corp., No. 17700) with 6-N-propyl-2-thiouracil (0.00625%) added to the drinking water for 3 weeks. Hyperthyroidism was induced by injecting T3 (25 pg/lOO g body weight) 48 and 24 h before killing. Euthyroid and hyperthyroid rats were provided with the same diet as given to the hypothyroid rats but with a normal amount of iodine added. The weight (g) of the rats at the time of killing were as follows: hypothyroid, 224 + 49; pertussis toxin-treated hypothyroid, 225 & 5; euthyroid, 248 + 7; pertussis toxin-treated euthyroid, 232 + 10; hyperthyroid, 226 f 7; pertussis toxin-treated hyperthyroid, 220 f 7. The parametrial and omental adipose tissue weights (g) at the time of killing were reduced by treatment in vivo with pertussis toxin as follows: hypothyroid, 9.0 k 0.3; hypothyroid, pertussis toxin-treated, 7.6 + 1.0; Euthyroid, control, 7.8 + 1.2; Euthyroid, pertussis toxin-treated, 5.3 f 0.8; hyperthyroid, control, 6.9 + 1.4; hyperthyroid, pertussis toxin-treated, 5.4 + 1.3. These values are the mean + S.E. for four experiments from the studies reported in Tables I and II utilizing eight rats per group in each experiment. Parametrial and omental adipose tissue was used for the preparation of adipocytes according to the method of Rodbell [6]. Adipocytes were isolated and incubated in Krebs-Ringer phosphate buffer containing 3% albumin (fatty acid-poor bovine albumin of clinical reagent grade (CRG-7 from Armour)). Cyclic AMP accumulation [7] and glycerol release [8] were determined as previously described. All data are expressed per cell based on counting cells in an aliquot of the dilute cell suspension placed on a slide which was inverted (hanging drop procedure) on a light microscope stage. A preparation of pertussis toxin purified 1800fold was used [9]. For in vitro studies, 10 ml of adipocytes were incubated with 28 ml of KrebsRinger phosphate buffer containing 28 ~1 of per-

tussis toxin (56 pg) or 28 pl of vehicle (0.1 M phosphate buffer plus 2 M urea, pH 7.4). Adipocytes were incubated with pertussis toxin or vehicle for 4 h but there did not appear to be any lysis of adipocytes during this procedure. The cells were washed with fresh buffer and comparable numbers of cells from each group were used in the experiments. In some of the studies, a preparation of pertussis toxin obtained from Dr. L.L. Irons of the PHLS Centre of Applied Microbiology and Research, Porton Down, U.K. was used at a concentration of 0.2 pg/ml and gave results indistinguishable from those of the toxin prepared in Mexico. Adipocyte plasma membranes were prepared by suspending one volume of free adipocytes in two volumes of homogenization medium (10 mM EDTA, 0.25 sucrose, pH 7.4, at 22°C) in a 50 ml glass tissue grinder (Arthur H. Thomas, type C) and immediately homogenized with ten up-anddown strokes of a motor-driven Fisher Dyna-Mix stirrer; (1700-1900 rm) with a Teflon pestle. The homogenates were placed on ice and all subsequent manipulations were performed at 0-4°C. The homogenates were centrifuged in a Sorvall SS-34 rotor (15 000 X g maximum) for 20 min. The congealed fat cake and infranatant were removed and the pellets were resuspended in homogenization buffer (0.5-1.5 mg of protein/ml). 1 ml aliquots of the crude microsmal preparation were frozen in a solid CO, acetone bath and then stored at -80°C in those experiments utilizing frozen-thawed membranes. Adenylate cyclase activity was measured using a modification of the method of Cooper et al. [lo] as described previously [ll] in a mixture containing 200 PM a-[ 32P]ATP (50 cpm/pmol), 30 mM Tris-HCl (pH 7.5) 1 mM MgCl,, 0.1 mM cyclic AMP, 0.1% CRG-7 albumin, 10 mM creatine phosphate, 10 U/ml creatine phosphokinase, 1.0 pg/ml adenosine deaminase, the indicated additions and membrane protein (5 -15 pg) in a final volume of 100 ~1. The assays were initiated by the addition of membranes and terminated by the addition of 1 ml of 1% SDS. All assays were conducted for 6 min at 30°C. Cyclic [32P]AMP was purified by sequential chromatography on Dowex and alumina columns as described by Salomon [12]; Cyclic [ 3H]AMP was added as a

621

recovery standard. Protein was determined by the dye-binding procedure of Bio-Rad Laboratories (Richmond, CA). Toxin-catalyzed ADP-ribosylation, electrophoresis and autoradiography were performed essentially as described by Malbon et al. [5]. Membranes were incubated for 15 min in 50 mM Tris buffer, (pH 7.4) with 2 mM MgCl,, 1 mM ATP, 10 mM thymidine, 10 PM [32P] NAD (approx. 2 Ci/mmol), 5 mM dithiothreitol, 0.2 mM GTP and activated pertussis toxin (100 pg/ml) in a total volume of 200 ~1. Samples were washed with 800 ~1 ice-cold 50 mM Tris buffer (pH 7.4) and immediately centrifuged for 5 min in a microcentrifuge (Fisher, Model 235B) at 13000 X g. Pellets were resuspended in buffer containing 0.25 M Tris-HCl (pH 6.8) 5% SDS, 1% glycerol, 0.05% bromophenol blue and 0.1% beta-mercaptoethanol. Samples were heated for 5 min at 80°C and subjected to SDS-polyacrylamide gel electrophoresis with a 5% stacking gel and 10% separating gel (1.5 mm thick), stained with Coomassie brilliant blue R-250, destained for 1 h with fast destain (50% methanol, 10% acetic acid) and several hours with low destain (7% acetic acid, 5% methanol). Gels were dried and exposed to Kodak X-OMAT film overnight at -80°C. ( - )-Norepinephrine was obtained from Sigma Chemical Co. (St. Louis, MO); N,-(phenylisopropyl)adenosine (PIA) and adenosine deaminase from Boehringer Mannheim (Indianapolis, IN); forskolin (7fi-acetoxy-8,13-epoxy-l-cx,6/?,9-cY-trihydroxy-labd-14-ene-1 l-one) from Calbiochem; crude collagenase (Clostridium histolyticum), lot 42N083, from Worthington; Dowex AG 50-WX-4 (200-400 mesh) and neutral alumina AG7 (100-200 mesh) for adenylate cyclase assay from Bio-Rad Laboratories. Results Pertussis toxin catalyzes the ADP-ribosylation of N, (the guanine nucleotide-binding regulatory protein involved in inhibition of adenylate cyclase activity) and blocks the effects of this protein [13]. It has been reported that the level of Ni in adipocytes is elevated in hypothyroidism [5]. If this accounts for the defect in cyclic AMP accumulation seen in adipocytes from hypothyroid rats,

pertussis toxin should normalize the response. Therefore, the effect of thyroid status on cyclic AMP accumulation and lipolysis was examined in adipocytes from normal and pertussis toxin-treated rats (Table I). The ability of norepinephrine to stimulate cyclic AMP accumulation and lipolysis was depressed in adipocytes obtained from hypothyroid rats and enhanced in adipocytes obtained from hyperthyroid rats as noted previously [l-4]. Pertussis toxin administration to rats, 3 days prior to removal of adipose tissue, increased the stimulation of cyclic AMP due to norepinephrine in adipocytes from hypothyroid, euthyroid or hyperthyroid rats (Table I). Although pertussis toxin treatment of hypothyroid rats markedly enhanced the ability of norepinephrine to stimulate adipocyte cyclic AMP accumulation, the responses in adipocytes from euthyroid rats were also enhanced by the treatment. Thus, after (in vivo) pertussis toxin treatment, the difference due to thyroid status was still evident (Table I). It should be mentioned, however, that the accumulation of cyclic AMP in response to norepinephrine was smaller in adipocytes from pertussis toxin-treated hyperthyroid rats as compared to those from toxin-treated euthyroid animals (Table I). Cyclic AMP accumulation at 60 min was similar to or less than that seen at 4 min except in cells from rats treated with pertussis toxin where the values were 2-3-fold higher at 60 min (Table I). Cyclic AMP accumulation due to norepinephrine at 60 min in adipocytes from hyperthyroid rats treated with pertussis toxin was much less than that in euthyroid rats (Table I). Administration of pertussis toxin to rats markedly enhanced the basal lipolysis of their adipocytes, in agreement with previous findings [9,14,15]. Interestingly, basal lipolysis was much smaller in adipocytes from toxin-treated hypothyroid rats and larger in cells from toxin-treated hyperthyroid rats as compared to cells from toxin-treated euthyroid animals (Table I). As little as 0.1 PM norepinephrine maximally (cells from toxin-treated euthyroid and hyperthyroid rats) or near maximally (cells from toxin-treated hypothyroid rats) activated lipolysis (Table I). However, the maximal lipolysis observed in cells from toxin-treated hypothyroid rats was significantly less than that in toxin-treated euthyroid or hyper-

622

thyroid rats (Table I). Furthermore, although the cyclic AMP accumulation induced by 10 PM norepinephrine was similar in cells from toxintreated hypothyroid and euthyroid rats, lipolysis was less in cells from toxin-treated hypothyroid rats (Table I). We used forskolin as an activator of adipocyte lipolysis because it bypasses the catecholamine receptor to directly activate adenylate cyclase and lipolysis in adipocytes [ll]. The ability of forskolin to stimulate cyclic AMP accumulation and lipolysis was markedly reduced in adipocytes from hypothyroid rats (Table II). Forskolin stimulation of cyclic AMP accumulation was enhanced in adipocytes prepared from hypothyroid rats injected with pertussis toxin, but not to the same levels as noted in adipocytes prepared from euthyroid rats treated with pertussis toxin. Adenosine deaminase (0.5 pg/ml) potentiated the stimulation by forskolin of cyclic AMP accumulation and lipolysis in adipocytes from hyperthyroid, euthyroid, or hypothyroid rats to about the same extent as prior injection of pertussis

TABLE

toxin (Table II). We observed a defect in cyclic AMP accumulation due to forskolin in adipocytes from hypothyroid rats even in the presence of adenosine deaminase but no greater stimulation of cyclic AMP accumulation due to forskolin in adipocytes from hyperthyroid as compared to euthyroid rats (Table II). However, in adipocytes from hyperthyroid rats incubated in the presence of adenosine deaminase, forskolin stimulation of glycerol release was greater than in the adipocytes from euthyroid rats (Table II). Again, we observed that under these conditions, the maximal rate of lipolysis of cells from hypothyroid rats was reduced as compared to cells from euthyroid or hyperthyroid rats. All further studies were conducted using adipocytes incubated with pertussis toxin in vitro for 4 hours. The ability of norepinephrine to stimulate cyclic AMP in cells not exposed to pertussis toxin was reduced by the 4 h incubation period (Fig. 1) as compared to the studies shown in Table I using freshly isolated adipocytes. In adipocytes from hypothyroid rats previously

I

EFFECT OF THYROID STATUS ON CYCLIC NORMAL AND PERTUSSIS-TREATED RATS

AMP

ACCUMULATION

AND

LIPOLYSIS

IN ADIPOCYTES

FROM

Adipocytes (200000 cells/ml) were incubated for 4 or 60 min with the additions indicated. Adipocytes were prepared from rats injected intraperitoneally without or with pertussis toxin (50-75 pg) 3 days prior to them being killed. The values are the means* S.E. of five experiments. Results are given as CAMP at 4 min (pmol/106 cells) (A), CAMP at 60 min (pmol/106 cells) (B) and glycerol release at 60 min (nmol/105 cells) (C). Control

Conditions

thyroid

status

Pertussis

hypo-

eu-

hyper-

33+3 30*7 29k6 28*2

28+ 5 33+ 1 59* 7 703+26

39* 5 385 8 118i41 158+50

3452 37+4 40&5 37*4

24k 3 28+ 6 48+ 4 118+32

28+ 6 32+ 5 46k 7 86k13

0+5 2+2 0+2 8&4

2+ 3 8k 3 48+16 68+22

lo* 1 34k 8 128*40 148rt45

toxin-treated

thyroid

status

l?U-

hypo-

hyper-

A. Control Norepinephrine 1.0 PM 10.0 PM

0.1 pM

26+ 35+ 595* 141Ok

3 2 195 365

48+ 170& 1665k 2160+

10 31 249 602

48+ 240+ 1350+ 1505*

8 52 245 300

B. Control Norepinephrine 1.0 gM 10.0 PM

0.1 PM

38+ 4 42+ 4 400 1745i 3380?1040

6 29i 120 475+ 540 5840+ 7550+1536

4 31* 528? 340 2825 k 885 361Ok1030

c. Control Norepinephrine 1.0 pM 10.0 PM

0.1 PM

9* 41+ 65+ 64+

4 7 13 10

51* llO+ 105* 103+

7 16 13 14

92+ 132k 126k 120*

26 35 35 30

623

ti +I +i +I 2 f 55 m

624 HYPOTHYROID

EUTHYROlD

HYPOTHYROID

HYPERTHYROID

ELITHYROID

c

r 160 1coo

2 :

1&I -

PERTUSSIS TOXIN

%

t

0

0.1

1.0

10.0

0

0.1

1.0

NOREPINEPHRINE

200

I

10.0

CONC.

0

0.1

1.0

10.0

(uhi)

Fig. 2. Effect of thyroid status on norepinephrine-stimulated lipolysis accumulation in rat adipocytes previously incubated for 4 h in the presence or absence of pertussis toxin. Lipolysis was based on a 60 min incubation of adipocytes under conditions identical to those mentioned in Fig. 1.

+

A-A-A-A 0

4

’

’

0

0.1

1.0

’ 10.0

0

0.1

NOREPlNEPHRlNE

1.0

10.0

0

CONC.

(4)

0.1

1.0

HYPOTHYROID

10.0

Fig. 1. Effect of thyroid status on norepinephrine-stimulated cyclic AMP accumulation in rat adipocytes incubated for 4 h in the presence or absence of pertussis toxin. Rat adipocytes (160 000 cells/ml) were incubated for 4 h without (A) or with (A) 2 yg/ml of pertussis toxin. The cells were washed twice and then exposed to the indicated concentratjons of norepinephrine for 4 min. The left side of the figure represents the mean&SE. of four experiments using adipocytes from hypothyroid rats, the middle portion of the figure represents the mean*S.E. of seven experiments using adipocytes from euthyroid rats and the right side of the figure represents the mean + SE. of three experiments using adipocytes from hyperthyroid rats.

exposed to pertussis toxin (2 pg/ml) for 4 h, cyclic AMP accumulation in response to norepinephrine was markedly enhanced (Fig. 1). In adipocytes from euthyroid rats, the stimulation by norepinephrine of cyclic AMP accumulation was actually less and further reduced in adipocytes from hyperthyroid rats. The effects of thyroid status on lipolysis were still seen after the 4 h incubation in medium containing albumin (Fig. 2). Bsal lipolysis in toxin-treated adipocytes from euthyroid rats was much higher than in toxintreated adipocytes from hypothyroid rats (Fig. 2). Lipolysis in the presence of 0.1 to 10 PM norepi-

140

1

0

EUTHYROID

GLYCEROL

1.0

1 RELEASE

3.3

10.0

PIA

CONC.

0

1.0

3.3

10.0

(n&l)

Fig. 3. N6-(Phenylisopropyl)adenosine inhibition of lipolysis and reversal by pertussis toxin in adipocytes obtained from hypothyroid or euthyroid rats. Rat adipocytes were incubated without (+) or with 2 pg/ml of pertussis toxin (A) for 4 h and then washed twice with fresh buffer. Adipocytes from euthyroid rats (158~ cells/ml, vehicle; 161000 cells/ml, pertussis toxin-treated) were incubated with 3.3 pM forskolin and 0.5 pg/ml adenosine dearninase. Adipocytes from hypothyroid rats (136000 cells/ml, vehicle; 142000 cells/ml, pertussis toxin-treated) were incubated with 20 pM forskolin and 0.5 pg/ml adenosine deaminase. Lipolysis was measured after 60 min. The values represent the means of four paired experiments. The mean+ S.E. of the percentage inhibition due to N6-(phenylisopropyl)adenosine (PIA) of glycerol release was 56 * 6, 74_+ 6 and 97 f 2 for 1. 3.3 and 10 nM PIA in ceils from hypothyroid rats incubated without and 3 + 8, I i 3 and 13 ct 2 in cells incubated with pertussis toxin, respectively. In cells from euthyroid rats, the respective percentages were 52_+6. 83+7 and 93&2 without and 7&4, 11+6 and I717 in cells incubated with pertussis toxin.

625

TABLE

III

EFFECT OF THYROID STATUS ON NOREPINEPHRINE AND FORSKOLIN ACTIVITY IN MEMBRANES OBTAINED FROM RAT ADIPOCYTES

STIMULATION

OF ADENYLATE

CYCLASE

Adipocyte membranes (S-15 pg/tube) were incubated for 6 min with the additions indicated plus adenosine deaminase (1.0 pg/ml). Adipocytes were incubated without or with pertussis toxin (2 pg/mI) for 4 h prior to preparation of the membranes. The values are the means + S.E. of five experiments. Results are expressed as pmol/min per mg for cyclic AMP formation. Control

Conditions

thyroid

status

Pertussis

eu-

hypo-

hyper-

toxin-treated

thyroid

status

eu-

hypo-

hyper

- GTP Control Norepinephrine 1.0 pM 10.0 PM +GTP(l

0.1 PM

56k 81k 175k 262+

6 6 23 45

12 10 10 9

58klS 84k16 134*26 193 k 38

49+ 1 131* 22 319+ 68 488& 110 154+ 38 410* 130 906+283

48+14 78+12 183+27 276+ 14 87+17 198*40 442+80

16+ 114+ 166k 222*

37k 64+ 145* 219+

39* 59k 91+ 125k

6 8 23 28

3 6 8 12

38&- 6 61* 8 105* 13 161k 39

1 66* 238+ 36 657*116

53* 9 138k 32 385k 66 5765 78 127+ 28 376k 95 763+188

PM)

Control Norepinephrine 1.0 PM 10.0 PM

0.1 pM

Forskolin 0.1 pM 1.0 pM 10.0 PM

38k 10 16k 9 230+ 13 311* 17 93+ 27 228+ 58 628k 171

.58+ 8 201* 53 684k226 1273 f 326 180+ 49 635 + 231 1420 f 491

HYPOTHYROID

nephrine was the same in all three groups and maximally activated by this low concentration of norepinephrine (Fig. 2). These data indicate that pertussis toxin clearly had different effects after a 4 h in vitro incubation and in vivo administration but neither procedure reversed the defect induced by hypothyroidism. Thyroid status had little effect on forskolin or norepinephrine stimulation of adenylate cyclase activity, in adipocyte membranes (Table III) as compared to its effects on cyclic AMP in intact cells (Tables I and II and Fig. 1). The activation of adenylate cyclase by norepinephrine or forskolin was less in adipocyte membranes from hypothyroid rats at low concentrations of forskolin or norepinephrine. However, adipocyte membranes from hypothyroid rats previously incubated with pertussis toxin actually had an enhanced response to the two highest concentrations of norepinephrine in the absence of GTP (Table III). Adenylate cyclase activity of membranes from hyperthyroid rats had a diminished response to norepinephrine or forskolin in the presence of GTP (Table III). An increased sensitivity to adenosine has been suggested to account for the defect in catecholamine or forskolin stimulation of cyclic AMP and

1034+194 206+ 41 573*128 llllk226

EUTHYROID

HYPERTHYROID

A-$

t

F

40

2

30

fl

20

m

CONTROL a

50

/

/ 11

t t -/ 10

t

-I-

TOXIN t

+

0 1.0

I’+-+

PERTUSSIS

1 /-

+’

3.3

10.0

*-&A 1.0 PIA

3.3 CONC.

10.0

Ku_ 1.0

3.3

10.0

(nM)

Fig. 4. N6-(PhenylisopropyI)adenosine inhibition of forskolinstimulated adenylate cyclase in fresh membranes prepared from adipocytes of hypothyroid, euthyroid and hyperthyroid rats. Adipocyte membranes (5-15 gg tube) were incubated for 6 min in the presence of 10 pM forskolin, 10 PM GTP, 120 mM NaCl and 1.0 pg/mI adenosine deaminase. Membranes were prepared from rat adipocytes treated without (0) or with 2 pg/mI pertussis toxin (A) for 4 h. The values are the mean f SE. of three paired replications and expressed as percentage inhibition due to the indicated concentrations of N6(phenyIisopropyI)adenosine. Adenylate cyclase activity in the presence of forskolin alone was 680, 1089, and 512 pmoI/min per mg of protein in membranes from hypothyroid, euthyroid and hyperthyroid rats treated without, and 1032, 1256 and 646, respectively, in membranes from cells previously incubated with pertussis toxin.

626

lipolysis in rat adipocytes in hypothyroid rats [5,16-181. Therefore, we examined the effect of the adenosine analog, N6-(phenylisopropyl)adenosine (PIA) on forskolin-stimulated cyclic AMP accumulation and lipolysis in freshly isolated

6 0” 0

100

1

200

”

300

400

CYCLIC

11

500

AMP

600 AT

11

700

800

11

900

1000

4 MINUTES

Fig. 6. Comparison of cyclic AMP accumulation versus lipolysis in adipocytes from hypothyroid and euthyroid rats. The data are taken from Table IV. Adipocytes obtained from hypothyroid rats (0) and adipocytes from euthyroid rats (m) are shown.

Fig. 5. Autoradiogram of pertussis toxin-catalyzed ADP ribosylation of adipocyte membranes from hypothyroid and euthyroid rats. Membranes were subjected to ADP ribosylation with [ ‘* PINAD as described in Materials and Methods. Membranes were solubilized and analysed by gel electrophoresis and autoradiography of the fixed, stained gel. The results of four separate experiments are shown. The odd-numbered lanes represent membrane from control, and the even numbered lanes. those exposed to pertussis toxin. Lanes 1, 2, 5, 6, 9, 10. 13 and 14 are from hypothyroid and lanes 3, 4, 7, 8, 11, 12, and 15, 16 are from euthyroid animals. Each experiment (4 lanes per experiment) had the same amount of protein added to the gels and compared membranes prepared from two hypothyroid with those from two euthyroid rats.

adipocytes. Adipocytes from hypothyroid rats were not more sensitive to N6-(phenylisopropyl)adenosine than those obtained from euthyroid rats (Table IV). The experiments were repeated using adipocytes previously incubated for 4 h and again 1 nM PIA gave the same inhibition of lipolysis in adipocytes from hypothyroid rats as in those from euthyroid rats (Fig. 3). In adipocytes previously exposed to 2 pg/ml of pertussis toxin for 4 h the antilipolytic action of PIA was abolished in adipocytes from hypothyroid and in those from euthyroid animals except at 10 nM PIA. However, the inhibition by 10 nM PIA was not different in adipocytes from hypothyroid rats as compared to those from euthyroid rats. This indicates that exposure of adipocytes to 2 pg/ml of pertussis toxin for 4 h was as effective in cells from hypothyroid as in those from euthyroid rats. Adenylate cyclase activity in membranes obtained from adipocytes of euthyroid rats was actually inhibited by phenylisopropyladenosine to a greater extent than that in membranes from hypothyroid rats (Fig. 4). Incubation of adipocytes for 4 h with 2 pg/ml of pertussis toxin reversed the inhibition by the adenosine analog of forskolinstimulated adenylate cyclase in membranes from euthyroid rat adipocytes (Fig. 4) but did not completely do so in membranes from hypothyroid or hyperthyroid rats (Fig. 4).

627

TABLE

IV

CYCLIC AMP ACCUMULATION AND EUTHYROID RATS EXPOSED TO PIA

GLYCEROL

RELEASE

IN

ADIPOCYTES

FROM

HYPOTHYROID

AND

Rat adipocytes (159000, hypothyroid; 136000, euthyroid) were incubated in the presence of adenosine deaminase (0.5 pg/ml). The data represent the meanfS.E. of four separate experiments. Basal glycerol release (60 min. nmol/105 cells (B)) averaged 23 nmol/lO’ cells and basal cyclic AMP accumulation (4 min, pmol/106 cells (A)) 36 pmol/106 cells. Forskolin Cont. (PM)

PIA cont. (nM) 0.33

1.0

27 73 13

135+ 11 223+ 62 457+131

86k 146+ 316*

318k 106 685 f 320 1060*400

206+ 78 629 rt 284 942 f 334

0

10.0

3.3

A. Hypothyroid 3.3 10.0 20.0

123k 265k 620*

8 43 90

81+ 112* 158k

12 13 26

57+20 55*17 57*25

117* 38 438 k 174 715 * 310

123+ 41 204k112 408+217

4Ok18 60+15 74k20

Euthyroid 3.3 10.0 20.0 B. Hypothyroid 3.3 10.0 20.0

lllk 1905 219k

56 52 49

107* 179? 217*

56 67 50

82* 143+ 200?

46 59 62

68+ 101* 168k

33 59 58

13*11 62+35 59*30

253* 345+ 344*

23 37 50

216+ 315k 399k

36 26 62

167* 260* 333*

61 13 35

134* 254+ 361+

60 22 42

43*31 135+36 248 * 27

Euthyroid

In order to obtain an index of the amount of the guanine nucleotide-binding regulatory proteins present in adipocytes from euthyroid and hypothyroid rats, toxin-catalyzed ADP ribosylation of membranes was performed. It was observed that in three out of four membrane batches, pertussis toxin-catalyzed ADP ribosylation of Ni was enhanced in membranes from hypothyroid rats as compared to that in membranes from euthyroid animals (Fig. 5). The complexity of thyroid hormone effects on lipolysis is emphasized by Fig. 6. These data indicate that for a given concentration of cyclic AMP there is about 40% more lipolysis in adipocytes from euthyroid as compared to hypothyroid rats at cyclic AMP levels over 200 pmol/106 cells. This indicates that a major unrecognized defect in activation of lipolysis is present which involves some step beyond cyclic AMP formation.

Discussion The decreased sensitivity of adipocytes from hypothyroid rats to lipolytic stimuli is well documented [l-4]. Prior research has focused on the search for a single defect responsible for this refractoriness. However, current evidence that hypothyroidism leads to alterations at several steps in the process that links receptor activation to modulation of lypolysis. Thus, there is evidence that the fat cell /3-adrenergic receptor has an impaired interaction with the stimulatory guanine nucleotide-binding regulatory protein of adenylate cyclase (N,) [4]. The defective responsiveness of adipocytes from hypothyroid rats is not exclusive for P-adrenergic agonists but is seen with all agents that activate adenylate cyclase [l]. The amount and interaction of N, with the catalytic subunit of adenylate cyclase appears to be unaffected in fat cells by hypothyroidism [2-41.

628

An increased sensitivity to adenosine has been suggested as a reason for the refractoriness of adipocytes from hypothyroid rats to lipolytic stimuli [5,16,18]. We did not see a change in the EC,, for phenyhsopropyl adenosine or restoration of responsiveness to lipolytic stimuli by adenosine deaminase or pertussis toxin treatment. We did see in three of four experiments an increase in pertussis toxin-catalyzed ADP ribosylation of Ni, as previously observed in fat cell membrane from hypothyroid rats [5], possibly reflecting an increased amount of Ni. Our data suggest that Ni is not rate limiting for adenosine action and that the observed apparent increase in Ni might be of little physiological relevance. The data concerning adenosine receptors in fat cells from hypothyroid rats is in conflict, i.e., decreases [18] and no change [.5] (as compared to the controls) in number and affinity have been reported. Previously, Fain and Malbon [19] found that adenosine release was impaired in adipocytes from hypothyroid rats. These findings, combined with the lack of evidence for an increase in adenosine receptors in hypothyroidism, suggests that an increased activation of Ni by adenosine is unlikely to be responsible for the defective activation of adenylate cyclase by hormones in adipocytes from hypothyroid rats. An apparent increase in adenosine responsiveness of adipocytes from hypothyroid rats was observed when cyclic AMP accumulation was maximally activated to 11 nmol/106 cells by 50 FM forskolin in the presence of adenosine deaminase [5]. Our highest concentration of forskolin was 33 PM and in the presence of adenosine deaminase gave a cyclic AMP value of 0.9 nmol/lO’ in cells from hypothyroid rats and 2.6 nmol/106 cells in cells from euthyroid rats (Table II). Ohisalo and Stouffer [16] incubated adipocytes from euthyroid rats with 0.5 PM epinephr~ne plus 1 pg/mI of adenosine deaminase and found very little inhibition of lipolysis by 10 nM PIA, while in adipocytes from hypothyroid rats almost 100% inhibition of lipolysis was noted. Under very similar conditions (i.e., adenosine deaminase + 0.5 PM epinephrine) Chohan et al. [18] found no antilipolytic effect of PIA even at a concentration of 100 PM in adipocytes from fed euthyroid rats, while 0.1 nM PIA gave a 60% inhibition of lipoly-

sis in adipocytes from hypothyroid rats. In contrast, in our experiments, 1 nM PIA inhibited cyclic AMP formation in the presence of 3.3 FM forskolin and adenosine deaminase by 64% in adipocytes from euthyroid rats and 40% in adipocytes from hypothyroid rats. The inhibition of lipolysis due to PIA under these circumstances was approximately 30%, regardless of the thyroid status. The present results agree with those of Garcia-Sainz et al. [20] who found no effect of thyroid status in hamsters on ru-2-adrenergic or adenosine effects. The effects of both agents were abolished in hamsters by pertussis toxin [9]. The deleterious effect of a 4 h incubation with pertussis toxin on the subsequent responsiveness of adipocytes from hyperthyroid rats, as reflected by cyclic AMP levels, might be related to ATP depletion. Under conditions of maximal cyclic AMP accumulation a significant drop in the ATP levels can take place in adipocytes from euthyroid but not in those from hypothyroid rats [19]. We previously observed a similar effect in adipocytes from hyperthyroid hamsters where the maximal cyclic AMP accumulation in response to hormones was decreased as compared to that in the controls [20]. It is also clear from our experiments that the 4 h incubation of adipocytes in the absence of pertussis toxin reduced the subsequent ability of norepinephrine to elevate cyclic AMP while actually enhancing its activation of lipolysis (Table I vs. Figs. 1 and 2). Van Inwegen et al. [21] and Correze et al. [22] found that hypothyroidism in immature rats elevated cyclic AMP phosphodiesterase in adipocytes. This is another defect which would tend to lower cyclic AMP levels but should be reversible at high concentrations of lipolytic agents. One of the major findings of our study was that freshly isolated cells from hypothyroid rats have a decreased maximal lipolysis for a given level of cyclic AMP as compared to controls (Fig. 5). Such a decreased maximal lipolysis was still present after pertussis toxin treatment but cyclic AMP accumulation was greatly enhanced by the toxin. These data suggest that in adipocytes from hypothyroid rats there is a defective propagation of the intracellular signal. It is not clear whether this involves alterations at the level of the cyclic AMP-dependent protein kinase(s), the lipase itself or some other process.

629

Omri et al. [23] found that the ability of 1 mM N,,O,-dibutyryl adenosine cyclic monophosphate to increase protein phosphorylation was unimpaired in adipocytes from hypothyroid rats, while that due to 10 PM isoproterenol was reduced. These results suggest that a normal phosphorylation pattern (especially of the 84 kDa region where one would expect the lipase to migrate) can be seen if levels of cyclic AMP are very high. However their results do not explain why a given level of cyclic AMP in our experiments gave a smaller activation of lipolysis in cells from hypothyroid as compared to those from euthyroid rats (Fig. 5). Furthermore Caldwell and Fain [24] reported rather different results than did Van Inwegen et al. [21] in that the lipolytic responsiveness was markedly reduced in adipocytes from hypothyroid rats incubated with N,,O,-dibutyryl adenosine cyclic monophosphate. The decreased lipolytic response of adipocytes from hypothyroid rats and increased response of those from hyperthyroid rats is a consistent finding. Studies on the mechanisms involved have been inconsistent. Caldwell and Fain [24] originally found no effect of hyperthyroidism on adenylate cyclase or cyclic AMP phosphodiesterase activity in adipocytes. However, while lipolytic responsiveness and respiration were increased in adipocytes from hyperthyroid rats, the total ATP content was depressed by lipolytic agents under conditions where the agents had no effect on ATP content in adipocytes from euthyroid controls [25]. Subsequently, hypothyroidism was shown to increase cyclic AMP phosphodiesterase [21], decrease hormonal activation of adenylate cyclase [2] as well as P-adrenoceptor binding and interaction with N, proteins [26] in adipocytes from hypothyroid rats. There appear to be multiple sites of action for thyroid hormones in white fat as previously reported in brown fat by Sundin et al. [27]. For a given concentration of cyclic AMP a reduced lipolytic response was seen in brown adipocytes [27] as noted in the present studies using white adipocytes (Fig. 6). A reduced respiratory response to added fatty acids was also seen in brown adipocytes from hypothyroid rats [27]. Thus, in both brown and white fat, there appear to be multiple loci for thyroid hormone action rather than a single locus.

Acknowledgements This research was supported by United States Public Health Service Grants AM 10149 and AM 21470 from The National Institutes of Health and a grant from Conacyt. J.A. G.-S. is a 1985 Guggenheim Fellow. The authors thank Ms. Anita Hardeman for typing the manuscript. References 1 Fain, J.N. (1981) Life Sci. 28, 1745-1754 2 Malbon, C.C., Moreno, F.J., Cabelli, R.J. and Fain, J.N. (1978) J. Biol. Chem. 253, 671-678 3 Malbon, C.C. and Gill, D.M. (1979) B&him. Biophys. Acta 568, 518-527 4 Malbon, C.C., Graziano, M.P. and Johnson, G.L. (1984) J. Biol. Chem. 259, 3254-3260 5 Malbon, C.C., Rapiejko. P.J. and Mangano, T.J. (1985) J. Biol. Chem. 260, 2558-2564 6 Rodbell, M. (1964) J. Biol. Chem. 239, 375-380 7 Gilman, A.G. (1970) Proc. Natl. Acad. Sci. USA 67, 305-312 8 Fain, J.N., Czech, M.P. and Saperstein, R. (1973) in Methods in Investigative and Diagnostic Endocrinology, Vo.. 2A: Peptide Hormones (Berson, S.A. and Yalow, R.S., eds.), pp. 267-273, EIsevier/North-Holland, Amsterdam 9 Martinez-Olmedo, M.A. and Garcia-Sainz. J.A. (1983) Biochim. Biophys. Acta 760, 215-220 10 Cooper, D.M.F., Schlegel, W., Lin, M.C. and Rodbell, M. (1979) J. Biol. Chem. 254, 8927-8931 11 Litosch, I., Hudson, T.H., Mills, I., Li. S.-Y. and Fain, J.N. (1982) Mol. Pharmacol. 22, 109-115 12 Salomon, Y. (1979) Adv. Cyclic Nucleotide Res. 10, 35-55 13 Murayama, T. and Ui, M. (1983) J. Biol. Chem. 258, 3319-3325 14 Garcia-Sainz, J.A. (1981) FEBS Lett. 126, 306-308 15 Moreno, F.J., Mills, I., Garcia-Sainz, J.A. and Fain, J.N. (1983) J. Biol. Chem. 258, 10938-10943 16 Ohisalo, J.J. and Stouffer, J.E. (1979) B&hem. J. 178. 249-251 17 Malbon, C.C. and Graziano, M.P. (1983) FEBS Lett. 155, 35-38 18 Chohan, P., Carpenter, C. and Saggerson, E.D. (1984) B&hem. J. 223, 53-59 19 Fain, J.N. and Malbon, C.C. (1979) Mol. Cell B&hem. 25. 143-169 20 Garcia-Sainz, J.A., Litosch, I., Hoffman, B.B., Lefkowitz, R.J. and Fain, J.N. (1981) Biochim. Biophys. Acta 678, 334-341 21 Van Inwegen, R.G., Robison, G.A., Thompson, W.J., Armstrong, K.J. and Stouffer, J.E. (1975) J. Biol. Chem. 250, 2452-2456 22 Correze, C., Auclair, R. and Nunez, J. (1976) Mol. Cell. Endocrinol. 5, 67-73 23 Omri, B., Gavaret, J.M., Correze, C. and Nunez, J. (1984) Mol. Cell. Endocrinol. 38, 205-213

630 24 Caldwell,

A.B. and

Fain,

1195-1204 25 Fain, J.N. and Rosenthal, 1205-1211

J.N.

(1971)

Endocrinology

89,

J.W. (1971) Endocrinology

89,

26 Malbon, CC. (1980) Mol. Pharmacol. 18, 193-198 27 Sundin, U., Mills, I. and Fain, J.N. (1984) Metabolism 1028-1033

33,