Regional differences in triglyceride breakdown in human adipose tissue: Effects of catecholamines, insulin, and prostaglandin E2

Regional

Differences in Triglyceride Effects of Catecholamines, Bj@rn Richelsen,

Steen

B. Pedersen,

Breakdown in Human Adipose Insulin, and Prostaglandin E, Torben

MqHler-Pedersen,

and Jens

Tissue:

Friis Bak

Regional variation of adipose tissue triglyceride breakdown (lipolysis) has been suggested to play a role for the health consequences of some forms of obesity. Thus, in the present study we investigated the regulation of lipolysis in isolated adipocytes obtained from different fat depots in females. Intra-abdominal adipose tissue (omental) and subcutaneous abdominal adipose tissue were obtained from the same individuals undergoing abdominal surgery (n = 9); in addition, adipocytes from the subcutaneous gluteal region (n = 12) and from mammary adipose tissue (n = 5) were investigated. The lipolyticlantilipolytic properties of epinephrine (EPI), insulin, clonidine, and prostaglandin E, (PGE,) were investigated. The most prominent observation was that EPI had none or only minor lipolytic effect in adipocytes from the subcutaneous regions, but significantly enhanced lipolysis by approximately 500% in omental adipocytes (P < .OOl). In the presence of the 4-adrenergic antagonist, yohimbine, EPI had similar stimulatory effects (fourfold to fivefold) in all fat depots. The antilipolytic compounds, insulin and clonidine, had greatly reduced antilipolytic properties in omental adipocytes as compared with subcutaneous adipocytes (P < .Ol and P < .05, respectively). On the other hand, PGE, had similar antilipolytic properties in adipocytes from the various depots. In conclusion, we found great regional variation in the regulation of lipolysis. Particularly, EPI was much more lipolytic in omental adipocytes than in subcutaneous adipocytes, mainly due to an enhanced functional %-receptor activity in subcutaneous adipocytes. These in vitro data suggest that free fatty acids (FFA) are more readily mobilized from omental adipose tissue than from subcutaneous adipose tissue. Copyright 0 1991 by W.B. Saunders Company

R

ECENTLY, ies indicate

both cross-sectional that accumulation

and prospective studof central,

abdominal

tissue is closely associated with obesity-related diseases such as cardiovascular complications, non-insulindependent diabetes mellitus (NIDDM), and premature death.‘-4 On the other hand, enlargement of peripheral adipose tissue (gluteal-femoral obesity) does not show such associations.“.4 Particularly, accumulation of intra-abdominal (visceral) adipose tissue, which can be determined by computed tomography (CT) scanning,s.h seemed to be associated with the enhanced health risk of obesity.‘.“,’ The factors responsible for the association between abdominal (particularly visceral) fatness and morbidity are still unknown, but abdominal adipose tissue may predispose to diseases through its correlation with numerous metabolic aberrations.1.3,7 Moreover, from recent studies it has been suggested that differential regulation of lipolysis in various fat depots may be of pathophysiological importance for the occurrence of the abnormal metabolic profile.’ ” In humans, catecholamines and insulin are the only hormones with a marked effect on triglyceride breakdown”; however, the endogenous substances prostaglandin E, (PGE,) and adenosine also have pronounced effects, but their physiological role is unknown.“-” Furthermore, in human adipocytes catecholamines mediate their effects through interaction with both p- as well as qadrenergic receptors, which mediate opposite effects on lipolysis.‘“” adipose

From the University Clinic of Endocrinology and Metabolism, Aarhus Amtsvgehus, Tage Hansensgade. Aarhus, Denmark. Supported by grants from NOVO. Nordic Insulin Foundation, the Danish Diabetic Association, P. Carl Petersens Fond, Aarhus Unilaersity, and Danish Medical Research Council. Address reprint requests to Bjr$m Richelsen, MD, UniverrsityClinic of Endocrinology and Metabolism. Aarhus Amtssygehus, Tage Hansensgade, DK-8000 Aarhus C, Denmark. Copyright 0 1991 by W.B. Saunders Cornpuny 0026-049519114009-0019$03.00/0 990

Thus, the balance between these antagonistic receptors seems to be responsible for the final lipolytic effect of catecholamines. Recently, it has been found that intraabdominal adipocytes may be more lipolytically responsive to catecholamines than subcutaneous adipocytes.““’ It is suggested that the resulting high levels of free fatty acids (FFA) exposed to the liver from visceral fat may have implications for the metabolic aberrations associated with abdominal adiposity such as hyperinsulinaemia, hypertriglyceridaemia, and impaired glucose tolerance.“.‘,h.7 To get more insight into the diversity of the regulation of lipolysis in adipocytes from various fat depots, we have examined human adipose tissue from the visceral depot (omental), in addition to various subcutaneous depots (gluteal, abdominal, and mammary adipose tissue) from both normal-weighing and obese subjects. The focus of the present study is the evaluation of the effects of catecholamines, insulin, and PGE, on triglyceride breakdown. MATERIAL AND METHODS Subjects

Intra-abdominal

(omental) and subcutaneous

abdominal adifrom nine women undergoing abdominal surgery. Three were morbidly obese and were undergoing operation for their obesity (jejuneal bypass); the non-obese women underwent surgery for gallstone diseases (cholecystectomy). Gluteal subcutaneous adipose tissue was obtained by an open biopsy”’ in a separate group of women (n = 12). Mammary adipose tissue was obtained from breast reduction surgery (n = 5). Body mass index (BMI) of the subjects was calculated from their weight and height (kg/m’). All subjects were fasted overnight before tissue removal. Omental and abdominal adipose tissue were removed at the beginning of the operation. In the study of sex differences, adipose tissue was taken from the gluteal region from moderately obese subjects (six women and five men). None of the subjects had any identified metabolic disorder or malignant diseases and no subjects received drugs that might influence triglyceride breakdown (P-blockers. antidepressiva. etc). The study was approved by the local ethical committee. pose tissue were obtained

concomitantly

Metabolism,Vol40. No 9 (September),

1991: pp 990-996

REGIONAL VARIATION

IN HUMAN ADIPOCYTE

991

LIPOLYSIS

Adipocyte Preparation Immediately Hanks

after

RESULTS

removal,

BS at 37°C and

Adipocytes

were isolated

tissue

transported

by collagenase

was placed

to the

digestion

as previously

described.“‘,”

nmolil.

was included

reduce

cell lysis during

isolation

of adipose

the

the adipocytes

buffer containing

Determination performed

since

isolation

5% albumin

at 37°C of 200

it has been

found

procedure.“‘,”

were

washed

and were

2.5% human

After four

finally

to the

times

in

resuspended

in

serum albumin.

of the size and the number

by measuring

tissue

at a concentration

in the buffer,

procedure, containing

incubation

Adenosine

in

laboratory.

in 10 mmol/L Hepes buffer for 45 to 60 minutes

fragments

buffer

the adipose

quickly

(microscopically)

of the adipocytes the diameter

was

of at least

per sample and from this parameter estimate the adipocyte volume and adipocyte surface were determined by conventional equations.“‘,”

200 adipocytes

Lipo!ysis Assay Adipocytes Hepes

glucose,

and electrolytes

adenosine minutes

2.5% human

as previously

at 37°C. with gentle

shaking

and responsiveness

the effect

where

of EPI

albumin,

described.”

Incubations

mined in incubations

in 250 JLL in 10

IO’ cells) were incubated

buffer containing

(200 nmol/L).“‘,”

sensitivity with

x

(20 to 50

mmol/L

5 mmol/L

in addition

were performed in duplicates.

of epinephrine

to

for 120

The lipolytic

(EPI)

were deter-

the effect of EPI alone was compared

in combination

with

the cu2-adrenergic

antagonist, yohimbine. These incubations were performed in a parallel set-up from the individual fat depots. The antilipolytic effects

of insulin,

clonidine,

lipolysis. which was stimulated isoproterenol(50

nmol/L)

and

in the presence

(ADA).‘”

Glycerol

measured

by an enzymatic-fluorometric

It has recently

PGEz

was taken been found

were

with a submaximal

determined

on

concentration

of

of adenosine

as an index

deaminase

of lipolysis

and was

method.

that the observed

great variation

in

in adipocyte suspensions could be attributable to variable amounts of endogenously produced adenosine.” Adenosine has several biological effects on adipocyte metabolism lipolytic responsiveness

where, especially, addition.

a pronounced

it has been found that adenosine

adipocyte

suspensions

is almost

cells” and is, thus, an artificial to get reproducible

results,

tion of endogenous chosen

antilipolytic

effect is prominent. accumulating

exclusively

derived

phenomenon.

it is necessary

adenosine.“‘,”

In the present

In preliminary studies, the influences of adenosine and ADA were examined on the EPI-induced lipolysis. These investigations were performed in abdominal adipocytes (n = 5). In the absence of ADA (ie, in the presence of 200 nmol/L adenosine), EPI had only very weak lipolytic action at very high concentrations (> 5 umol/L) (Fig 1). If adenosine was removed by ADA (0.5 U/mL), the “basal” lipolysis was enhanced by approximately 50%. In addition, ADA showed a moderate lipolytic effect of EPI in the lower concentration range (10 to 100 nmol/L). However, at higher concentrations (> 100 nmol/L), the lipolytic effect of EPI gradually diminished until it returned to initial values (Fig 1). In the presence of yohimbine (50 kmol/L), EPI was lipolytic both in the presence and absence of ADA; moreover, the maximally lipolytic effect of EPI was similar (Fig 1). However. in the presence of ADA, the lipolytic sensitivity of EPI (f yohimbine) was about IO-fold higher than in its absence (ie, in the presence of 200 nmol/L adenosine) (ED,,,, 65 2 12 nmol/L v 510 ? 81 nmol/L; P < .Ol). Regional Variation in Catecholamine-Induced Lipolysis In these comparative studies, omental and abdominal adipocytes were obtained from the same individuals. In addition, lipolytic data from gluteal adipocytes and mammary adipocytes from other groups of women were also presented (Table 1). Three morbidly obese women (jejuneal bypass operated) were included in these studies, since it was found that the regional differences (omental/ abdominal) in the effect of EPI (? yohimbine) were not significantly different from the non-obese females. However, if these three obese women were not included, the 1,5

In

in human

1

from leaking

Accordingly, to control

Influence of ADA

in order

the concentrastudy, we have

to express “basal” lipolysis as the glycerol release in the

presence of a known amount of adenosine (200 nmol/L).‘” However, in the studies of regional variations, all other incubations related to determination of lipolyticiantilipolytic responsiveness and sensitivity

are performed

in the presence

of ADA (0.5 UlmL)

in order to remove both endogenously produced added adenosine (adenosine-free conditions). lipolysis

through

endogenous

removal

adenosine.

of the tonical

inhibition

Thus, when expressing

ness) in relative terms, the glycerol release alone is taken as the reference value (=l).

and exogenously ADA stimulates mediated

by 0,o

the data (responsive-

in the presence

;

The

values

presented

ANOVA

unpaired

t tests were used.

are

for repeated

the

means

I

I

1

I

8

7

6

5

of ADA -log

Statistical Methods analysis,

A

2

measurements,

SEM. and

Regression paired

or

[Epinephrine](M)

Fig 1. The influence of ADA on the lipolytic affect of EPI. Adipocytes from subcutaneous abdominal adipose tissue (n = 5) were incubated with EPI in the absence (A.4 and presence (0.0) of ADA (0.5 U/mL}. and in the absence (A,O) and presence (A#} of yohimbine (50 pmol/L). The incubation buffer contains 200 nmol/L of adenosine (see Materials and Methods).

RICHELSEN ET AL

992

Table 1. Characterization

of the Subjects

was comparable between omental adipocytes and subcutaneous adipocytes with a fourfold to fivefold increment of lipolysis. However, the lipolytic sensitivity of EPI was still higher in omental adipocytes than in abdominal and gluteal adipocytes (P < .Ol; Table 2). ADA alone had a minor lipolytic effect in all fat depots, enhancing lipolysis by approximately 50% above basal levels (Table 2) with no significant regional differences. In adipose tissue from the same region, the maximal lipolysis induced by EPI in the presence of yohimbine was comparable to lipolysis induced by isoproterenol at maximally effective concentrations (1 umol/L) (Table 2).

Adiwse Tissue 0lllent.dAbdominal

Gluteal

9

12

No. Age (vr) BMI (kg/m2) Adipocyte size (pL)

Mammary 5

41 r 3.9

34 t- 2.7

30 r 2.8

24.5 + 1.6

23.3 ? 1.4

598 + 41

302 ? 49

316 + 77

465 + 53

28 t 5

BMI (25.1 ? 0.8) of this group (n = 6) was comparable to the BMI in the two other groups (Table 1). EPI was exclusively lipolytic in omental adipocytes; moreover, the lipolytic response was much more pronounced in omental adipocytes than in any of the subcutaneous adipocytes (Fig 2A, Table 2). At maximally effective concentrations, EPI (1 to 10 kmol/L) enhanced lipolysis almost fivefold in omental adipocytes, which was significantly more than in subcutaneous adipocytes (P < ,001). In this concentration range, EPI seemed to be antilipolytic in gluteal adipocytes in relation to ADA-induced lipolysis (inhibition by 30%) without any effect in abdominal adipocytes and had only minor lipolytic action in mammary adipocytes (stimulation with 50%). In the lower concentration range, EPI (10 to 100 nmol/L) had a small but consistently lipolytic effect (from 50% to 115% above reference levels) in subcutaneous adipocytes from all three regions (Fig 2A). In the presence of the cqadrenergic antagonist, yohimbine, a completely different picture was found (Fig 2B). Yohimbine (50 umol/L) had only minor effects on the EPI-induced lipolysis in omental adipocytes, whereas it unmasked a pronounced lipolytic response of EPI in adipocytes from all subcutaneous regions. Thus, in the presence of yohimbine, the lipolytic responsiveness of EPI

6

Regional Variation in Antilipolysis The antilipolytic effect of insulin showed pronounced regional differences (Fig 3). The responsiveness to insulin was highest in gluteal adipocytes (maximum inhibition, 73% r+ 5.9%; P < .Ol and P < .05 as compared with omental and abdominal adipocytes, respectively; Table 2). On the other hand, the lowest responsiveness to insulin was seen in omental and mammary adipocytes with a maximum inhibition of 18% to 25%. The antilipolytic effect of insulin was intermediate in abdominal adipocytes, with an inhibition of approximately 50%, which, however, was significantly higher than in omental adipocytes (P < .05). Moreover, the antilipolytic sensitivity was also highest in gluteal adipocytes (IC,,,, 133 ? 11 pmol/L v 288 ? 76 pmol/L in omental adipocytes; P < .05). In contrast to insulin, the maximally antilipolytic effect of PGE, was similar in adipocytes from various depots, with a maximum inhibition of 86% to 95% (Fig 4). However. the antilipolytic sensitivity of PGEL was higher in the gluteal adipocytes (IC,,, 1.4 ? 0.24 nmol/L) than in the other

11

A

B

+Yohimbine

1

-O+ + -Cl-

;,

;r

i

-log [Epinephrine]

fi (M)

4

4

h

Omental Abdominal Gluteal Mammary

I

I

I

I

8

7

6

5

-log [Epinephrine]

(M)

Fig 2. Regional variation in the lipolytic effect of EPI in human adipocytes. Isolated adipocytes ware incubated with EPI for 2 hours at 37°C and glycerol release was taken as an index of lipolysis. ADA (0.5 U/mL) was present in all incubations and the glycerol release in the presence of ADA alone was taken as the reference value (=l). (A) EPI-induced lipolytic response. (B) %-Receptor antagonist, yohimbine (50 pmol/L), included in the adipocyte suspensions and the incubations performed in parallel set-up with A. Omental(0) and abdominal (A) adipocytes were obtained from the same individuals (n = 9). Gluteal adipocytes (0) (n = 12) and mammary adipocytes (0) (n = 5) were from separate groups of women. The absolute values for glycerol release are shown in Table 2.

REGIONAL VARIATION IN HUMAN ADIPOCYTE LIPOLYSIS

993

Table 2. Data of Lioolvsis Adioose Tissue Omental

Abdominal

Gluteal

Mammary

Glycerol release (nmol/106 cells/2 h) Basal

203 2 26

308 + 43

272 t 29

146 r 19

ADA

313 * 49

445235

369 ? 42

265 r 51

188

EPI

1,439 +

539 2 51

323 + 26

EPI + yohimbine

1,643 k 201

1,654 2 82

1,556 -t 64

1,598 2 132

1,722 k 91

1,634 k 73

lsoproterenol % Inhibition Clonidine

393 + 38 1,099 r 151

53 k a.8

72 k 5.3

a4 5 4.9

24.5 + 1.9

49.5 + 3.6

73 + 5.9

17 2 3

a7 k 4.0

90 k 4.1

89 + 3.9

85 k 5.1

EPI (+yohimbine)*

29 k 9

82 2 23

221 + 37

43 k 18

Clonidine*

a2 k 19

32 + 8

29 + 6

Insulin PGE, Sensitivity (ED,,)

lnsulint PGE,’

288 + 76

201 + 32

2.4 + 0.4

2.6 k 0.4

133+

11

1.4 f 0.2

9.1 2 3.1

NOTE. Absolute values for the lipolytic data presented in Figs 2-5 (see Materials and Methods). In addition, basal glycerol release, which was determrned in the presence of exogenous adenosine (200 nmol/L), is shown. Thus, except for basal lipolysis, all other incubations were performed in the presence of ADA (adenosine-free conditions). ADA (0.5 U/mL), EPI (5 kmol/L), yohimbine (50 pmol/L), isoproterenol (1 kmol/L). The sensitivity of the compounds was determined as half-maximally stimulatory concentration of EPI (in the presence of yohimbine) or half-maximally inhibitory concentration (clonidine, insulin, and PGE,). *nmol/L. tpmol/L.

regions (P < .OS), whereas the antilipolytic effect of PGEz was similar in abdominal and omental adipocytes (Table 2). Interestingly, the sensitivity of PGEz was low in mammary adipocytes (IC,,,, 9.1 k 3.1 nmol/L) (Table 2). The maximally antilipolytic effect of the a,-agonist, clonidine, was similar in gluteal and abdominal adipocytes, with an inhibition of 84% t 4.9% and 72% 2 5.3%, respectively (Fig 5). However, the effect of clonidine was significantly lower in omental adipocytes (inhibition of 53% k 8%) than 100

I

I

I

I

i!I 11

10

9

8

-log

[Insulin]

in adipocytes from subcutaneous regions (P < .05). In addition, the sensitivity was higher in the subcutaneous adipocytes than in the omental adipocytes (P < .05; Table 2). The effect of clonidine was not examined in mammary adipose tissue. Correlation Studies

By multiple variance analysis, it was found that BMI correlated significantly with the adipocyte volume of both omental adipocytes (r = .86, P < .Ol) and abdominal adipocytes (r = .75, P < .05). Abdominal adipocytes were significantly larger ( _ 50%) than omental adipocytes within the same individual (P < .Ol; Table 1). However, from the

(M)

Fig 3. Regional variation in the antilipolytic effect of insulin. The antilipolytic effect of insulin was examined on lipolysis, which was stimulated by isoproterenol (50 nmol/L) in the presence of ADA (0.5 U/mL). The difference between maximum lipolysis (without antilipolytic compounds) and basal lipolysis was taken as 100%. Symbols as in Fig 2. Omental and abdominal adipocytes, n = 7; gluteal, n = 8; and mammary, n = 5 adipocytes.

b 10r

91

-)Dg

8I

WE

2

7I

61

1 (MI

Fig 4. Regional variation of the antilipolytic affect of PGE,. (See legend to Fig 3.)

RICHELSEN ET AL

Table 3. Characterization

of Subjects and Lipolytic Data in the Sex

Study (Gluteal Adipose Tissue) Women

Mf?n

6

No. Age W BMI (kg/m’) Adipocyte size (pL)

5

31 * 3.1

33 2 3.9

28.2 + 1.2

27.5 2 1.7

763 + 59

561 * 70

Glycerol release (nmol/106 cells) 581 + 61

Basal

402 + 63

EPI +yohimbine

2,451 -t 137

982 i 135

lsoproterenol

2,602 2 135

1,172 t 141

NOTE. EPI (5 kmol/L);

yohimbine

(50 FmoliL); isoproterenol

(I

pmol/L). ADA was not included in this study.

-log

[Clonidine]

(M)

Fig 5. Regional variation of the antilipolytic effect of the aradrenoceptor agonist, clonidine. (See legend to Fig 3.)

6). In the presence of yohimbine, a clearly lipolytic effect of EPI was observed both in women and men; however, the lipolytic effect of EPI was significantly higher in gluteal adipocytes from women than in men (P < .Ol at concentrations of EPI > 0.1 kmol/L, Fig 6). DISCUSSION

regression line (not shown), it is clear that omental adipocytes have a steeper increment in size in relation to enhanced BMI than abdominal adipocytes. A positive correlation was also found between BMI and the adipocyte size in the gluteal region (r = .65; P < .05; n = 12). Adipocyte size was positively correlated to basal lipolysis, ADA-induced lipolysis, and maximal EPI (+ yohimbine)induced lipolysis in all regions, as well as in the regions taken together (r = 57, P < .Ol; r = .64, P < .OOl; and r = .75, P < .OOl; n = 30). However, when EPI responsiveness was expressed in relative terms (in relation to glycerol release in the presence of ADA), no significant correlations were found in relation to adipocyte size, neither in the individual depots nor in all groups taken together (r = .08). The maximal antilipolytic effect of insulin (in relative terms) was not correlated to adipocyte size in the individual groups; however, a tendency to a positive correlation was shown in gluteal adipocytes (r = 51, P = .l), but when all groups were taken together a significant correlation between adipocyte size and insulin responsiveness was present (r = .63, P < .OOl; n = 30). Finally, the antilipolytic responsiveness of clonidine was negatively correlated to fat cell size in gluteal adipocytes (r = -.59. P < .Ol; n = 12); however, correlations were not found in either of the other depots or if all depots were taken together. Sex Influences on Lipolysis in Gluteal Adipocytes We have previously demonstrated sex differences in the lipolytic effect of EPI and isoproterenol in gluteal adipocytes.” Accordingly, we examined the influence of yohimbine on the EPI effect between sexes. The study was performed in moderately obese subjects (Table 3) and ADA was not added to the adipocytes in these experiments. In the absence of yohimbine, EPI had only minor lipolytic effects at very high concentrations in gluteal adipocytes (Table 3). EPI (10 kmol/L) enhanced the glycerol release by 55% ? 15% in women and by 41% * 12% in men (Fig

The present study was conducted to investigate the regulation of triglyceride breakdown in various fat depots and to provide information about the biochemical background for the observed differences. The existence of pronounced regional differences in triglyceride mobilization was confirmed in the present in vitro study performed with isolated adipocytes. EPI alone had only minor lipolytic effects in subcutaneous adipocytes, whereas a pronounced lipolytic action of EPI was found in omental adipocytes. These findings are in accordance with a previous study by &tman et al.“’ They found that norepinephrine (mixed B/a>-agonist) and isoproterenol (pure @agonist) had similar lipolytic effects in 5-

4-

3-

2-

l-

Ob

I

I

I

I

8

7

6

5

-log

[Epinephrine]

(M)

Sex influences on the lipolytic effect of EPI. Gluteal adipocytes from six women (0.0) and five men (0.W) were incubated with EPI in the absence (0.0) and presence (0.W) of yohimbine (50 pmol/L). ADA was not present in these experiments. The lipolytic effect of EPI was expressed in relation to nonstimulated values. The absolute values are shown in Table 3.

REGIONAL VARIATION

IN HUMAN ADIPOCYTE

LIPOLYSIS

omental adipocytes, whereas norepinephrine was less lipolytic in subcutaneous epigastric adipocytes. However, addition of the cuz-antagonist, yohimbine, in the present study showed a pronounced lipolytic effect of EPI in subcutaneous adipocytes, which nearly equals the effect in omental adipocytes. These findings suggest that amelioration or blocking of the lipolytic effect of EPI in subcutaneous adipocytes is mediated by the u2-adrenergic receptor, which is coupled to inhibition of lipolysis. Theoretically, enhanced cxz-receptor function could be due to an increased receptor number or enhanced activity of some postreceptor pathways. Recently, Mauriege et al’ have demonstrated an enhanced number of cw,-receptors as compared with B-receptors in subcutaneous human adipocytes ( _ 3:1), whereas in omental adipocytes B-sites are at least as numerous as a,-sites (1:l). Since we recently have been able to find positive correlations between the number of cy,-receptor sites in human adipocytes and the biological effects induced by cYz-agonists,‘x.‘4 the findings by Mauriege et al’ indicate that the present differences in catecholamine-induced lipolysis might partly be explained by the regional differences in the balance between B- and cY,-adrenergic receptors. Since the cYz-agonist, clonidine, also had significantly less antilipolytic actions in omental adipose tissue, it may suggest that the function of the cx,-receptors may also be affected in omental adipocytes, eg, involving the coupling to postreceptor effecters, etc. However, differences in the lipolytic sensitivity of EPI were still observed even in the presence of yohimbine, indicating that some properties of the B-receptor or B-receptor-mediated pathways may also exhibite regional variations. However, since the antilipolytic effect of PGE, was comparable at least in abdominal and omental adipocytes, there appears to be no general antilipolytic defect in omental adipocytes. To control the influence of variable amounts of endogenously produced adenosine,” the present regional differences in lipolytic/antilipolytic effects are obtained in the presence of ADA (adenosine-free conditions). Whether this procedure is optimal from a physiological point of view, in relation to the concentration of adenosine that is actually present in the adipose tissue in vivo, is unknown. It could be suggested that these manipulations might affect or obscure some important events taking place in the adipocytes, but it seems unlikely in light of the fact that endogenous adenosine is predominantly artificially produced in adipocyte suspensions through cell lysis where extracellular degradation of adenine nucleotides is the main source for adenosine production.” In the present study, we determined the nonstimulated lipolysis both in the presence of a known amount of adenosine (200 nmol/L) and under adenosine-free conditions (+ADA) (Table 2). From correlation studies, it was found that the regional differences in nonstimulated hpolysis in both situations could mainly be related to differences in adipocyte size. The effect of ADA was similar in the various fat depots stimulating the basal lipolysis by approximately 50% (Table 2), which is in accordance with previous studies in human adipocytes.“‘~“~” In addition, we found

that ADA (adenosine-free condition) did not affect the responsiveness of EPI, but significantly enhanced the sensitivity of EPI when compared with parallel incubations in the presence of 200 nmol/L adenosine (Fig 1). Thus, variable concentrations of endogenous adenosine may be able to reduce the lipolytic sensitivity of EPI without affecting the responsiveness in human adipocytes. In accordance with the study of Berlan and Lafontan,” we found that in the absence of ADA the biphasic effect of EPI in subcutaneous adipocytes disappeared (Fig 1). The insulin-induced antilipolysis also exhibited regional variations, with the most pronounced effect in gluteal adipocytes and only minor effects in omental adipocytes, which is in accordance with a recent study by Bolinder et a1.26In this latter study, the reduced antilipolytic effect of insulin was found partly to be due to reduced insulin receptor affinity in omental adipocytes.‘” However, in the present study, the insulin responsiveness was positively correlated to adipocyte size (examining all regions together), indicating that the observed regional differences in insulin action on lipolysis could partly be explained by differences in adipocyte size. The triglyceride breakdown in adipocytes from the mammary adipose tissue seems to have some other properties than in adipocytes from other subcutaneous depots (abdominal and gluteal). Particularly, the antilipolytic effects of insulin and PGE, were quite different than from the two other sites (Table 2). However, the adipose tissue from mammae was removed because of hypertrophy of this depot and may, thus, have other properties than in adipocytes from normal mammary adipose tissue. EPI had limited but similar effects on hpolysis in gluteal adipocytes from women and men (Fig 6). In the presence of yohimbine, EPI was found to be more lipolytic in women. This latter finding is in accordance with our previous studyZ4 where the pure B-agonist, isoproterenol, was found to be more lipolytic in gluteal adipocytes from women compared with men. However, Leibel and Hirsch” found no significant differences in the lipolytic response of norepinephrine in gluteal adipose tissue between sexes. The reason for this discrepancy is presently unknown. It should be emphasized that the present data are obtained with isolated adipocytes in vitro and there may be differences in adipocyte metabolism in the in vitro and the in vivo situation. For example, substances other than adenosine may accumulate in adipocyte suspensions, such as prostaglandins and FFA, which may affect the lipolytic response. N” Thus, these in vitro data should be considered with caution when extrapolated to the in vivo situation in human adipose tissue. However, if the present data are valid for the in vivo situation, the observed regional differences in lipolytic responsiveness seem to be in accordance with the idea that the intra-abdominal adipose tissue is a readily mobilizable energy depot, whereas the gluteofemorale depot is primarily a storage depot.“.” In conclusion, we found that omental adipocytes in vitro have a considerably higher hpolytic response to EPI than the subcutaneous localized adipocytes, and this may predom-

RICHELSEN ET AL

996

inantly be the result of regional differences in the (Yereceptor number and function. Furthermore, the antilipolytic effect of insulin was more pronounced in subcutaneous adipocytes than in omental adipocytes. Thus, the action of the important regulators of lipolysis in humans (catecholamines and insulin) greatly favor triglyceride breakdown (and FFA release) from the omental adipose tissue, which

is suggested to be an important link between (intra-abdominal) adiposity and the development related complications.3,4,7

abdominal of obesity-

ACKNOWLEDGMENT We are grateful to J. S@holt and P. Sonne

for their skillful

technical assistance.

REFERENCES 1. Larsson B, Svtirdsudd K. Welin L, et al: Abdominal adipose tissue distribution, obesity, and risk of cardiovaskular disease and death: 13 year follow up of participants in the study of men born in 1913. Br Med J 288:1401-1404.1984 2. Lapidus L, Bengtsson C, Larsson B, et al: Distribution of adipose tissue and risk of cardiovascular disease and death: A 12 year follow up of participants in the population study of women in Gothenburg, Sweden. Br Med J 289:1257-1261.1984 3. Bjiirntorp P: The associations between obesity, adipose distribution and disease. Acta Med Stand Suppl723:121-134,

tissue 1988

4. Kissebah AH, Peiris AN: Biology of regional body fat distribution: Relationship to non-insulin-dependent diabetes mellitus. Diabetes Metab Rev 5:83-109, 1989 5. Ashwell M. Cole TJ, Dixon AK: New insight into the anthropometric classification of fat distribution shown by computer tomography. Br Med J 290:1692-1694,1985 6. Fujioka S, Matsuzawa Y, Tokunaga K, et al: Contribution of intra-abdominal fat accumulation to the impairment of glucose and lipid metabolism in human obesity. Metabolism 36:54-59, 1987 7. Peiris AN, Sothman MS, Hoffmann RG, et al: Adiposity, fat distribution, and cardiovascular risk. Ann Intern Med 110:867-872, 1989 8. Rebuff&Scrive M: Regional adipose tissue metabolism in women during and after reproductive life and in men, in Berry EM, Blondheim SH, Eliahou HE, et al (eds): Recent Advances in Obesity, ~015. London, England, Libbey, 1987, pp 82-91 9. Mauriege P, Galitzky J, Berlan M, et al: Heterogeneous distribution of beta and alpha-2 adrenoceptor binding sites in human fat cells from various fat deposits: Functional consequences. Eur J Clin Invest 17:156-165,1987 10. Wahrenberg H, finnquist F, Arner P: Mechanisms underlying regional differences in lipolysis in human adipose tissue. J Clin Invest 84:458-467. 1989 11. Smith U, Hammersten J, Bjiirntorp P. et al: Regional differences and effect of weight reduction on human fat cell metabolism. Eur J Clin Invest 9:327-332, 1979 12. Hales CN, Luzio JP, Siddle K: Hormonal control tissue lipolysis. Biochem Sot Symp 43:97-135, 1978 13. Fain JN, Malbon CC: Regulation adenosine. Mol Cell Biochem 25:143-169,

of adenylate 1979

of adiposecyclase

by

14. Fredholm BB: Local regulation of lipolysis in adipose tissue by fatty acids, prostaglandins and adenosine. Med Biol56:249-261, 1978 15. Richelsen

B: Interaction

of indomethacin

with the antilipoly-

tic effect of prostaglandin Ez in rat adipocytes. Mol Cell Endocrinol 58:85-91, 1988 16. Burns TW, Langley PE. Terry BE, et al: Pharmacological characterization of adrenergic receptors in human adipocytes. J Clin Invest 67:467-475, 1981 17. Lafontan M, Berlan M. Villeneuve A: Preponderance of alpha,- over beta-adrenergic receptor sites in human fat cells is not predictive of the lipolytic effect of physiological catecholamines. J Lipid Res 24:429-440, 1983 18. Richelsen B, Pedersen 0: Alpha,-adrenergic binding and action in human adipocytes: Comparison between binding to plasma membrane preparations and to intact adipocytes. Eur J Pharmacol119:101-112.1985 19. &man J, Arner P, Engfeldt P. et al: Regional differences in the control of lipolysis in human adipose tissue. Metabolism 28:1198-1205.1979 20. Richelsen B: Prostaglandins E2 binding and action in human adipocytes: Effects of sex, age and obesity. Metabolism 37:268-275, 1988 21. Richelsen B. Eriksen EF, Beck-Nielsen, et al: Prostaglandin E2 receptor binding and action in human fat cells. J Clin Endocrinol Metab 59:7-12, 1984 22. Honnor RC, Dhillon GS, Londos C: CAMP-dependent protein kinase and lipolysis in rat adipocytes. I. Cell preparation, manipulation, and predictability in behavior. J Biol Chem 260: 1512215129,1985 23. Kather H: Purine accumulation in human fat cell suspensions. J Biol Chem 263:8803-8809, 1988 24. Richelsen B: Increased alpha,- but similar beta-adrenergic receptor activities in subcutaneous gluteal adipocytes from females compared with males. Eur J Clin Invest 16:302-309,1986 25. Berlan M, Lafontan M: Evidence that epinephrine acts preferentially as an antilipolytic agent in abdominal human subcutaneous fat cells: Assessment by analysis of beta and alpha, adrenoceptor properties. Eur J Clin Invest 15:341-348, 1985 26. Bolinder J, Kager L, &tman J, et al: Differences at the receptor and postreceptor levels between human omental and subcutaneous adipose tissue in the action of insulin on lipolysis. Diabetes 32:117-123, 1983 27. Leibel RL, Hirsch J: Site- and sex-related differences in adrenoreceptor status of human adipose tissue. J Clin Endocrinol Metab 64:1205-1210, 1987 28. Rebuff&Scrive M, Enk L, Crona N, et al: Fat cell metabolism in different regions in women. Effect of menstrual cycle. pregnancy and lactation. J Clin Invest 75:1973-1976. 1985