obtained both in vitro and in vivo suggest that /3,-ARs are involved in the ...... (CL 316,243). A potent beta-adrenergic agonist virtually specific for beta3 receptors.
0021-972x/961$03 00/o Journal of Clinical Endocrinology and Metabobsm Copyright 0 1996 by The Endocrine Society
Vol. 81, No. 1 Printed in US A
Evidence for Numerous Brown Adipocytes Lacking Functional P,-Adrenoceptors in Fat Pads from Nonhuman Primates* N. VIGUERIE-BASCANDS?, D. RICQUIER, M. BERLAN,
A. BOUSQUET-MliLOU, AND L. CASTEILLA
J. GALITZKY,
D. LARROUY,
CNRS URA 1937, UPS-IFR Louis Bugnard, and INSERM U-317, Laboratoire de Pharmacologic Mkdicale et Clinique, Facultd de Medecine (A.B.-M., J.G., M.B.), Toulouse; and Centre de Recherche sur I’Endocrinologie Mole’culaire et le Dkveloppement, CNRS (D.R.), Meudon, France ABSTRACT Brown adipose tissue (BAT) is involved in the control of energy balance and has been demonstrated to be activated through paadrenoceptor (&-AR) occupation in rodents. The ability to specifically activate energy expenditure via this receptor is of great interest for the treatment of obesity. Nevertheless, the extent of BAT and the presence of a functional @,-AR in humans are now debated, and this situation is difficult to clarify for evident practical and ethical reasons. We investigated the occurrence of brown adipocytes in fat deposits of prepubertal baboons using antibodies raised against uncoupling protein (UCP) in Western blotting and immunocytology experiments. UCP was detected in all types of fat pads studied and was revealed in multilocular cells. Pe&ardiac and axillary adipose tissues displayed large amounts of UCP and can be assimilated to typical BAT. Most of the other pads looked like white adipose tissue, but exhibited
areas with clusters of brown adipocytes and, thus, can be assimilated to the convertible adipose tissue as previously described in rodents. The presence of pa-ARs was evaluated by both j3,-agonist-stimulated lipolysis and messenger ribonucleic acid (mRNA) expression studies. There was no significant lipolytic effect of any of the &-AR agonists tested (SR 586118, BRL 37344, CGP 12177, or CL 316243) in either white or brown tissues. PCR analysis demonstrated that &-AR mRNA expression is not related to the UCP content of fat pads and that &-AR expression is low. This study demonstrates the presence of great proportions ofbrown adipocytes in adipose tissue and the heterogeneity of the fat pads in baboons. The lack of a metabolic effect of &agonists combined with the weak expression of &-AR mRNAs raise the question of the role of pa-ARs in adipose tissues of primates. (J Clin Erzdocrinol Metab 81: 368-375, 1996)
HITE and brown adipose tissues (WAT and BAT, respectively) play a central role in body weight control, having opposite roles in the regulation of energy balance. WAT is the greatest energy buffer of the body, ensuring both energy storage via triacylglyceride synthesis and mobilization via lipolysis. BAT is able to dissipate energy as heat via uncoupled mitochondrial respiration by a mitochondrial anion carrier, uncoupling protein (UCP) (1). This function is crucial for thermoregulation in hibernating animals and mammal neonates and contributes to rodent cold- and dietinduced thermogenesis (2). The contribution of a defect in BAT activity to the determinism of obesity has been confirmed by genetic ablation of BAT inducing the development of obesity in mice (3). The quantitative importance and functionality of brown fat in healthy human subjects is important to evaluate but difficult to establish for obvious ethical and practical considerations (4). Recently, a higher frequency of a specific restriction fragment length polymorphism in the UCP gene was found in humans with higher than average
body fat gain over time (5). This link favors the physiological involvement of UCP function in body fat gain in humans. The adrenergic system plays a major role in the physiological regulation of WAT and BAT functions in mammals. In WAT, catecholamines are able to stimulate lipolysis by the activation of P-adrenoceptors (/3-ARs) and to inhibit lipolysis by the activation of ol,-adrenoceptors (a,-ARs). Concerning P-adrenergic responsiveness, white adipocytes of many mammal species are characterized by the presence of a functional &-AR (6, 7) coexisting with the &- and &-subtypes. However, there are now convincing data showing that the adrenergic control of lipolysis in human white adipocytes is strikingly different from that in the commonly used laboratory species (8). Man is the sole mammal species whose white adipocytes exhibit both a weak (if any) Pa-adrenergic responsiveness and a potent a,-adrenergic component (8). However, recent data have demonstrated that in two monkey species, the macaque and the baboon, white adipocytes present pa- and a,-adrenergic receptivities very similar to those of human white fat cells (9,lO). Regarding BAT, results obtained both in vitro and in vivo suggest that /3,-ARs are involved in the control of catecholamine-induced differentiation and activation of brown adipocytes. Indeed, the Paagonist ICI 07114 induces UCP expression in primary cultures of rodent brown adipocytes (11). In adult dogs, which lack detectable BAT and express functional &-ARs in WAT, chronic treatment with ICI D7114 restored UCP expression in various fat pads and induced a reduction in body weight
Received February 17,1995. Revision received June 8,1995. Accepted August 7, 1995. Address all correspondence and requests for reprints to: Dr. N. Viguerie-Bascands, CNRS URA 1937, IFR Louis Bugnard, CHU Rangueil, avenue Jean Poulhes, 31054 Toulouse Cedex, France. * This work was supported by grants from the Direction de la Recherche et des Etudes Doctorales and from the Fondation pour la Recherche Medicale. t Established investigator for INSERM.
UCP
AND
&-AR
EXPRESSION
gain (12). In addition, &-agonists have been demonstrated to increase rodent thermogenesis and induce BAT hypertrophy and mitochondrial proliferation (13). In these species, various Pa-agonists induce weight loss in obese animals and possess antidiabetic properties (2,13,14). According to these results, the use of &-agonists was proposed as a potential therapeutic approach in the treatment of obesity and related endocrine-metabolic disorders in humans. However, although brown adipocytes have been detected in autopsied adult subjects (15, 16), the effects of &agonists on energy expenditure remain controversial in humans (8, 17). Interestingly, a recent study reported that brown adipocytes of the guinea pig lack &-AR thermogenic responsiveness (18). These data point out the discrepancies among mammal species concerning the expression of a functional &-AR in BAT. The baboon has been shown to possess an adrenergic regulation of lipolysis in WAT very close to that of humans (9, lo), thus providing a good model to develop treatment of obesity. Because the &-AR messenger ribonucleic acid (mRNA) has always been detected together with the UCP mRNA in adipose tissues from humans (71, the current study investigates the presence of brown adipocytes and &ARs in healthy baboons. Firstly, the proportion of brown adipocytes was systematically assessed in several fat pads from male and female baboons. Secondly, the functionality and expression of P,-ARs were simultaneously investigated according to the UCP content of the pads. Materials Sample
and
Methods
collections
Fat pads were obtained from five female and five male baboons (Papio pupio), weighing 5-7 kg. The animals had previously been individually housed at 23 C. After overnight fasting and pentobarbital anesthesia, the animals were bled, and biopsies (l-16 g) were taken. Tissue biopsies were dissected from nerves, vessels, and lymph nodes. Then, one sample of adipose tissue was immediately fixed in alcoholic Bouin’s fluid for 24 h, dehydrated, and embedded in paraffin for histological studies; another sample was used for lipolysis measurements. The remaining adipose tissue was frozen in liquid nitrogen and stored at ~80 C until mitochondria and RNA preparations. All procedures were performed under the control of INSERM according to the National Authorization of Animal Care Investigations.
Histological
studies
Sections of 3-grn thickness were deparaffinized, dehydrated, and boiled for 5 min in citrate buffer (0.1 mol/L) in a microwave oven, then dried. Unspecific binding was blocked by 1% skimmed milk in Trisbuffered saline. The sections were incubated with purified sheep IgG raised against rat UCP (1:2000) for 1 h at room temperature, rinsed, and incubated in the presence of rabbit antisheep IgG coupled to alkaline phosphatase (1:lOO). The phosphatase was revealed using Sigma Fast Red TR/Naphthol AS-MX tablets (Sigma Chemical Co., St. Louis, MO) as substrate and counterstained with Harris’s hematoxylin, then mounted in Aquamount. Control experiments were performed using purified sheep immunoglobulin G (Sigma) and yielded no staining.
Preparation
and analysis
of the mitochondrial
fraction
Mitochondria fractions were prepared by differential centrifugation .I of a homogenate as described by dasteilla St nl. (19). For the detection of UCP, from 2-30 fig mitochondrial proteins were electrouhoresed in 10% polyacrylamide-SDS gels, transferred onto nitrocellulhse, and incubated with sheep antiserum raised against rat UCP (1:2000) and rabbit
IN BABOON
FAT PADS
369
antisheep IgG conjugated to peroxidase (1:lOOO). Unspecific binding was blocked by 2% skimmed milk in phosphate-buffered saline containing Tween 20 (0.5%). I’eroxidase was revealed using 4-chloro-1-naphthol and hydrogen peroxide as substrate. The bands on the nitrocellulose membrane were scanned using an image analysis system (ElecphorCRIS-TM) that calculates the densitometric value of each band. Then, for relative quantitation, arbitrary units were assigned as described in the figure legends.
Preparation
and analysis
of RNA
Whole frozen pads were powdered to homogeneity, and total RNA was prepared from the frozen powdered tissues, according to the single step method of Chomczynski and Sacchi (20). The RNA pellet was dried and dissolved in water containing diethyl pyrocarbonate (0.1%). RNA was quantitated by UV spectrophotometry at 260 and 280 nm. RNA preparations with an A260/A280 ratio between 1.9-2.1 were used for complementary DNA (cDNA) synthesis as follows. RNA (1 fig) was treated with 200 U Maloney murine leukemia virus RT in 20 PL RT buffer containing 5 mmol/L deoxynucleotides, 50 ng/pL of the homooligomeric deoxynucleotide pfdeoxythymidine),5, dithiothreitol (10 mmol/L), and 33 U placenta ribonuclease inhibitor. The reaction was carried out at 25 C for 10 min, then at 42 C for 50 min, and the tubes were heated to 95 C for 5 min and chilled on ice. PCR amplification was performed on 10 PL of each cDNA preparation by adding 150 pmol/L of each primer and 1.25 U Tlzermus dq&cl~~ DNA oolvmerase in a final volume of 50 tiL PCR buffer containing either 3% (v&/vol) formamide for &-AR mRNA amplification or 10% (vol/ vol) dimethylsulfoxide for p-actin mRNA amplification. Samples were denatured for 3 min at 95 C, then PCR was performed for 30 or 40 cycles (1 min at 94 C, 1 min and 30 s at 56 C, 1 min at 72 C) with an Appligene Crocodile thermocycler. To confirm the integrity of the RNAs, a housekeeping gene mRNA, P-actin mRNA, was also amplified. The PCR products for &AR or @actin were amplified from the same cDNA and analyzed on 2% agarose gels stained with ethidium bromide. The gel was photographed with Polaroid type 667 films under UV light at the same exposure and developing time. Amplification in the absence of cDNA yielded no bands other than those of the primers for &AR mRNA (data not shown). The &-AR primers were selected in the coding region (exon 1) of the cDNA from the alignment of the rat, bovine, and human sequences of the Pa-AR. The primers were 20 nucleotides in length with 50% GC composition. The sense mimer 5’-ACCAACGTGTTCCTCACTTC-3’ was defined by bases 850-870 and matched at a position corresponding to the second transmembrane segment (tm2 as described by Emorine t; al. (21)). The antisense mimer 5’-TAGATGAGCGCGTTGAAGGC-3’ was defined by bases 165’4-1674 and matched at a position corresponding to the end of the seventh transmembrane segment ftm7). The cDNA amplification product was predicted to be 823 bp in length. The sequences of sense and antisense p-actin primers were 5’.CGACGAGGCCCAGAGCAAGC3’ and 5’-CCAGGGCGACGTAGCACAGC-3’, respectively. Prdw for thy ~&XXI ,&-AX mKN4 was n.bf~Pd by.anpJ3.h~m uf 250 ng genomic DNA, as mentioned above. The PCR product was purified with the Qiaquick spin PCR purification kit (Qiagen Corp., Chatsworth, CA), cloned in PCR-Script SK vector (Stratagene, La Jolla, CA) and sequenced. I
Southern
blot analysis
To assess the specificity of the observed signals, Southern blots were performed. After amplification, 25 FL of each PCR product were electrophoresed through a 2% agarose gel. The gel was denatured, neutralized, and blotted overnight onto a standard nylon membrane (Hybond-N, Amersham Corp., Arlington Heights, IL) as described by Sambrook it al. (22) with 20 X sodium chloride-sodium citrate buffer as the transfer buffer. The DNA was fixed onto the nylon membrane under UV radiation (254 nm). The probe for baboon &AR mRNA was labeled with [321’Jdeoxy-CTP by random priming (Amersham Megaprime kit). After hybridization in stringent conditions with the radiolabeled probe (50% formamide, 42 C, overnight), the nylon was washed once with 0.1 X SSC-0.1% SDS buffer at 55 C, then twice at 65 C for 30 min each time.
370 Autoradiography development.
VIGUERIE-BASCANDS was performed
at room temperature
for 1 h before
ET AL.
JCE & M Vol81.
A nt --- w IEAT
-If-...
PR PA PC Ab AX PO
rat
-.-.
b&on
IEAT
PR PA PC Ab Ax PO
l
1996 No 1
..--
Lipolysis measurements
2
Samples of adipose tissue were dissected free from connective tissue and larger blood vessels and carefully cut with small scissors to obtain pieces of approximately 2 mm’. Portions of 50-100 mg were incubated in polyethylene tubes with 500 PL Krebs-Ringer bicarbonate buffer containing 3.5% BSA and 6 mmol/L glucose at pH 7.4. Incubations were carried out at 37 C in a water bath with gentle shaking (60 cycles/min). Pharmacological agents at the indicated concentrations were added to the medium just before the beginning of the assay. After 90 min of incubation, the tubes were placed in an ice bath, and 200~PL aliquots of the medium were taken for enzymatic determination of the glycerol released, which was used as the index of lipolysis. The pharmacological agents used to studv B-adrenergic activation of lipolys$ were the ‘;lonse&ctive eadrenoc&tor ago& isoproterenol and the B,-adrenoceutor agonists BRL 37344. CGP 12177. CL 316243. and SR 5861’1i. The eff&s an; concentrations bf the &-adienoceptor’agonists used have been previously destibed using white and brown fat cells (23-28) as well as &-adrenoceptors transfected in Chinese hamster ovary cells (29,30). The other lipolytic agent used was the CAMP analog dibutyryl CAMP.
4 6 8 IO male
female
B loo1
Statistical analysis Values are given as the mean -I- SEM. Student’s paired t tests were used for comparisons between matched pairs, and differences were considered significant at P < 0.05.
Chemicals Moloney murine leukemia virus RT was obtained from Life Technologies (Gaithersburg, MD), ribonuclease inhibitor from Pharmacia (Up$ala, Sweden), &d Thermus aquaticusDNA polymerase from ADulisene (Illkirch. France). BRL 37344 (4-11(2-hvdroxv-(3-chlorop&$)ethvi)-amin&rop&henoxvacetate) was a &nero&gift from br. I\;r. A. Hawthorne’ (S&G Kline-&echam Pharn&euticalsTEpsom, UK). CGP 12177 (4-13-t-butvlamino-2-hvdroxvuromJxv)benzimidazol2&e) was obtained-from & K. Scheibb (&a’ G&y; Basel, Switzerland), CL 316243 ((R,R~-5-[2-[~2-(3-chlorophenyl~-2-hydroxyethyllamino]propyl11,3-benzodioxole-2,2-dicarboxylate) was obtained from Drs. Claus and Danforth (Medical Research Division, Lederle Laboratories American Cyanamide Co., Pearl River, NY), and SR 58611A (N[2S)-7-carbethoxymethoxy-1,2,3,4-tetrahydronaphth-2-y11(2R)-2-hydroxy-2-chlorophenyl)ethanamine hydrochloride) was obtained from Dr. L. Manara (Sanofi-Midi, Milan, Italy). BSA (fraction V) and (--hsoproterenol hydrochloride were obtained from Sigma Chemical Co. Dibutyryl CAMP and enzymes for glycerol assays were purchased from Boehringer Mannheim Corp. (Mannheim, Germany). All other chemicals were of the highest grade of purity commercially available.
Rx?mlIts Macroscopic analysis Most of the fat pads exhibited a yellow color similar to that of human WAT, but the fat collected from axillary and pericardiac regions clearly looked like typical brown fat from rodents even after exhaustive dissection of the lymph nodes. Finely dissected axillary and pericardiac fat pads gave about 4-7 and 2-5 g of typical brown fat, respectively. Western blotting of UCP protein Using anti-UCP antibodies, we screened Western blots of mitochondrial proteins from fat pads of 10 baboons. The sheep IgG raised against rat UCP cross-reacted with a mon-
fat
pads
FIG. 1. Distribution of UCP in baboon fat tissues. A, Immunoblotting of mituchondria from adipose tiaaues. Two micrograms of mitochondrial rat IBAT proteins (control) or 10 ccgmitochondrial proteins from baboon fat were prepared and managed as described in Materials and Methods, then analyzed for UCP immunoreactivity. Each blot corresponds to one animal. B, Quantication of the Western blots by scanning densitometry. The values obtained from blots presenting from lo-30 H mitochondrial proteins were normalized to the rat interscapular BAT (IBAT) signal, which was used as a control, then the values were calculated as a percentage of the control rat IBAT. Results are the mean 2 SEM of values from 10 animals. PR, perirenal fat; PA, periaortic fat; PC, pericardiac fat; Ab, abdominal fat; Ax, axillary fat; Po, popliteal fat.
key fat mitochondrial protein (Fig. 1A). This protein had a molecular mass corresponding to the apparent molecular mass of UCP (-32 kDa). The UCP mRNA was easily detected in most of the pads by Northern blotting experiments with 20 pg total RNA (data not shown). Taken together, the immunoblottings showed that UCP was expressed in all studied deposits, although at different levels. The number of positive fat pads differed according to the individual animals, but for some of them, all adipose tissues tested were positive. Nevertheless, according to the richness in UCP, a gradient from periaortic to pericardiac deposits can be distinguished. Pericardiac and axillary adipose tissues displayed highly positive UCP signals in all animals. The mean estimated UCP content of the mitochondria from these tissues reached 74% of the control rat interscapular BAT (Fig. 1B). The values ranged from 20% (1 case) to 100% (3 cases) in both pericardiac and axillary deposits. Surprisingly, mitochondria from deposits macroscopically similar to WAT
UCP AND &AR
EXPRESSION IN BABOON FAT PADS
exhibited numerous positive UCP signals. This was the case for perirenal, popliteal, abdominal, and periaortic pads, in which, respectively, 7, 6, 5, and 4 positive identifications of 10 were detected (Fig. 1A). On the basis of the relative signal intensity, the mean UCP content of these latter depots ranged from lo-21% of the control rat interscapular BAT (Fig. 1B). Histological study of baboon adipose tissues To identify cells expressing UCP, we performed immunocytochemical experiments on previously tested adipose tissues. Figure 2A shows a section of the axillary deposit from baboon 7. All fat cells were multilocular and red stained, revealing the presence of UCP. The red color was generally located in adipocytes containing multilocular fat droplets. The intensity of staining differed according to the cells and could indicate a heterogeneous distribution of UCP among multilocular cells. Perirenal deposits showed either characteristics specific to typical WAT or exhibited areas of stained multilocular cells drowned among typical white adipocytes (Fig. 2, B and C, respectively) depending on the histological section. This feature of composite tissue appeared in nearly all samples except the pericardiac and axillary pads, which could be histologically considered as typical BAT. Further-
371
more, few white-like adipose cells resembling unilocular adipocytes with a big central droplet surrounded by a redstained cytoplasm containing some very small fat droplets were detected (Fig. 2D). Lipolysis experiments The effects of various lipolytic agents were investigated in two fat deposits exhibiting highly UCP-positive signals (axillary and pericardiac) and in two fat deposits exhibiting slightly UCP-positive signals (perirenal and periaortic). Lipolysis studies on fat pad fragments were performed as previously described (31). In a preliminary experiment, the validation of this technique compared to lipolysis on isolated white adipocytes in rats was tested (data not shown). The results obtained with fat pads from baboons are presented in Table 1. The CAMP analog dibutyryl CAMP was lipolytic in all fat deposits, indicating that the intracytoplasmic lipolytic cascade was functional in these experiments. Isoproterenol enhanced spontaneous lipolysis in all deposits, but the stimulatory effect was higher in periaortic and perirenal than in axillary and pericardiac pads. The four &AR agonists used in this study were unable to significantly increase lipolysis in axillary, pericardiac, or periaortic deposits. In perirenal fat
FIG. 2. Tmmunocytochemical localization of UCP in baboon fat tissues. Specific labeling of UCP using Fast Red TR/Naphthol Sigma Tablets as substrate. Bar = 50 Frn. A, Typical appearance of axillary adipose tissue (baboon 7). Brown adipocytes are multilocular and densely stained in red. B, Section of perirenal deposit (baboon 10). White adipocytes are typical unilocular cells. C, Section of perirenal deposit (baboon 10). Some areas exhibit multilocular stained cells. D, Detection of UCP staining in pseudounilocular cells (baboon 10).
372
VIGUERIE-BASCANDS
TABLE 1. Maximum lipolytic effect of the nonadrenergic &AR agonists on samples of adipose tissue from axillary,
JCE & M . 1996 Vol81. No 1
ET AL.
agent dibutyryl CAMP, the nonselective P-AR agonist isoproterenol, pericardiac, perirenal, and periaortic deposits in baboon
Highly UCP-positive pads Axillary (n = 8) P&cardiac (n = 7)
Slightly UCP-positive pads Perirenal (n = 8) Periaortic (n = 6)
pm01
Basal lipolysis Dibutyryl CAMP (10e3 mol/L) Isoproterenol (lo+ mol/L) P,-Adrenergic agonists BRL 37344 (10m5 mol/L) CL 316243 (10m5 mol/L) CGP 12177 (10e4 mol/L) SR 58611A (10m5 mol/L)
and various
glycerol
0.35 t 0.04 0.61 t 0.04” 0.52 t 0.05a
0.48 + 0.05 0.73 + 0.05” 0.60 t 0.06”
0.25 -’ 0.05 0.59 -c 0.05” 0.51 t 0.05=
0.12 -+ 0.02 0.57 2 0.13” 0.51 k 0.11”
0.34 0.32 0.36 0.34
0.52 0.44 0.48 0.48
0.28 0.27 0.29 0.25
0.13 0.14 0.11 0.09
t t ? t
0.04 0.03 0.04 0.05
t 0.05 IT 0.05 t 0.06 2 0.07
Samples of 50-100 mg adipose tissue were incubated in the presence of the indicated expressed as micromoles of glycerol per 100 mg wtJ90 min. a Significantly different from basal lipolysis (P < 0.05).
concentrations
t k k +
0.07 0.07 0.06” 0.03
t 0.01 5 0.03 t 0.02 -c 0.03
of each drug. Values are the mean t
SEM,
TABLE 2. Detection of &-AR mRNA in highly UCP-positive depots of the baboon using RT-PCR Frequency of &-AR expression Axillary Pericardiac 2l6 317 4l6 417
30 cycles 40 cycles
l33-AR
Actin
nAx7 PC2 a-F 3oc. 4Oc. 2
aAx PC2 a-z 30 c. 4oc. r--
Highly UCP-positive samples were selected from depots in which the mitochondria exhibited a UCP content greater than to 50% of the level measured in rat interscapular brown adipose tissue. RT-PCR was performed as described in Materials and Methods, and the samples were analyzed on 2% agarose gels.
pads, CGP 12177 exhibited a significant, but very weak, lipolytic effect of 15-16% of isoproterenol-induced lipolysis. RT-PCR
of &-AR
mRNA
in baboon
adipose
tissue
PCR performed on 250 ng baboon DNA led to a single product with the mol wt we expected (-820 bp; 823 bp when predicted from the human sequence). After sequencing, the PCR product displayed 81.4% homology with the rat &-AR and 95.5% with the human &-AR (data not shown). Using the same conditions that permit detection of &-AR mRNA in RNA from rat fat pads (32), the presence of this messenger was assessed in adipose tissue from baboon. Despite the amount of total RNA (30 PLg)and many modifications in blot washing stringency, the detection of &-AR mRNA failed (data not shown). Thus, the expression of &-AR mRNAs was investigated by RT-PCR. Because it has been demonstrated that &-ARs are mainly expressed in brown adipocytes, the fat pads in which UCP contents were the highest (>50% of the UCP content of rat interscapular BAT) were tested. This represented 7 and 5 samples from axillary and pericardiac deposits, respectively. Using 30 cycles of PCR, there were only 3 and 2 positive signals for the presence of &-AR mRNA in axillary and pericardiac pads, respectively (Table 2). Four samples from both deposits became positive when 40 cycles were performed (Table 2). Figure 3 shows typical results, from agarose gel electrophoresis and Southern blot hybridization, of amplification products after 30 cycles of PCR of &-AR mRNA from both negative and positive samples exhibiting similar UCP signals. A single band of the predicted size was obtained (823 bp) in the axillary sample, but none was found in the pericardiac sample. Increasing the number
S FIG. 3. Detection of &-AR mRNA in brown fat pads of the baboon. Ethidium bromide-stained gel of &AR cDNA fragments amplified using RT-PCR (expected PCR product size = 823 bp) and autoradiogram of corresponding Southern blot. Control experiments were performed using 250 ng genomic DNA (lane DNA), nonreverse transcribed RNA amplified with &-AR primers CRT-), or reverse transcribed RNA amplified with /3-actin primers (expected PCR product size = 500 bp).
of PCR cycles allowed the detection of &-AR mRNA in the negative sample. Using the baboon &-AR probe Southern blots (Fig. 31, we demonstrated the specificity of the products and confirmed their identity. It is to be noted that some samples from pads exhibiting a weak UCP content showed &AR mRNA expression (data not shown). Taken together and without restraining the investigations to the UCPenriched pads, the results indicate that the proportion of &-AR mRNA-positive pads did not rise to more than 30%. There was no link between the detection of &-AR mRNAs and the UCP content of fat deposits.
Discussion
The main result of this study is that fat pads of healthy male and female baboons contain high UCP and brown adi-
UCP AND
&-AR
EXPRESSION
pocyte levels, associated with a lack of &-AR mediated lipolytic response. The physiological functions of adipose tissues have been extensively studied in laboratory rodents, but many data have pointed out the discrepancies between these species and man, raising the question of the choice of animal model for the study of human adipose tissues. Among primate species, baboons have been shown to develop obesity and endocrine-metabolic disorders (hyperinsulinemia and hypertension) with aging, under natural or laboratory housing conditions (33). Moreover, the development (34) and the adrenergic regulation of lipolysis (9, 10) of baboon adipose tissue are very similar to those in humans. Taken together, these considerations clearly demonstrate that the baboon is a valuable model for the study of human adipose tissues. BAT is abundant in neonates, in which it contributes to the maintenance of body temperature through nonshivering thermogenesis, but its presence in adults depends on the species. In rodents, BAT persists in the adult, but in larger mammals, especially in humans, this tissue seems to be converted into WAT-like deposits, so that all fat pads appear as WAT in adulthood (2). Until now, no studies have attempted to systematically investigate the distribution of BAT in nonhuman primates; thus, we focussed the present study on the determination of the amount and distribution of brown adipocytes in the baboon. Based on the semiquantitative data obtained from Western blots, we observed that the amount of UCP in pads located in the axillary and pericardiac regions from prepubertal baboons was equivalent to that found in the typical BAT of the rat. The presence of large amounts of UCP in adipose tissues of baboons can be connected to the results of a previous study in man showing BAT distribution in autopsy samples, where UCP was detected whatever the age of the donor (16). However, this study was performed postmortem on unhealthy subjects. Furthermore, the presence of multilocular cells at stages other than neonates has only been described in cases of high levels of circulating catecholamines occurring in pheochromocytoma, in workers exposed to cold environments, and in heavy alcohol consumers (15, 16, 35, 36). The present study is the first report concerning the presence of UCP in numerous fat deposits from healthy nonneonatal primates. The baboons were in a prepubertal stage (-2 yr old ZJS.3-4 yr old for puberty), indicating that the transition from typical BAT to WAT-like fat could be slower in primates than in larger mammals such as calves (37) and delayed at least until puberty. Because the development of adipose tissue in baboons parallels that of human body fat (341, the present data obtained on BAT distribution in this species could be extended to humans. However, it may also be possible that breeding conditions (single caging and room temperature) put the baboons in an acclimation temperature below their thermoneutral zone, thus inducing UCP synthesis in their fat deposits. White and brown adipocytes were considered to be located in distinct deposits, but a recent study has described the presence of brown adipocytes in rat WAT-like pads (38). The present study extends this observation to healthy primates. These data stress the difficulties of determining the exact quantitative importance of brown adipocytes in the fat of the whole body and suggest that it could be generally underes-
IN BABOON
FAT PADS
373
timated. Furthermore, using an antibody raised against UCP, the cytoplasm of some unilocular adipocytes of baboons was positive. Such cells could indicate residual dormant brown adipocytes persisting despite the morphological disappearance of BAT and could be qualified as convertible adipocytes, as previously described (38,391. The nonuniform staining of brown adipocytes in a given site was also described in humans by Kortelainen et al. (16) and could reflect different stages of cell development in the same deposit. The possible contribution of BAT to energy metabolism has been evaluated in adult humans; 20-40 g active BAT could account for 10% of the daily energy expenditure (4). In baboons, the weight of axillary and pericardiac pads (-8 g), which could be assimilated to typical BAT, combined with the richness of brown adipocytes in the other fat pads, suggest that whole brown adipocytes could play a major role in energy metabolism in these animals. The effect of catecholamines on lipolysis in WAT results from the activation of both /3-ARs and cw,-ARs; their effect on BAT thermogenesis is mainly mediated by P-ARs. Since the discovery of the atypical P-AR subtype in adipose tissues, the &-AR, the use of selective &-AR agonists has been proposed in the pharmacological treatment of obesity and diabetes in man (23). Indeed, in species possessing functional Pa-ARs in both BAT and WAT, &-AR agonists exhibit antiobesity and antidiabetic properties without the side-effects produced by the /3,- and P,-ARs. However, until now, the efficiency of treatments with &-agonists in man remains controversial (8). Furthermore, the effects on weight loss (40) or energy expenditure (41) are often associated with side-effects (tremor), suggesting the stimulation of P,-ARs. There is a large diversity in fat cell &AR-mediated lipolytic responses among mammalian species. Lafontan and Berlan (8) depicted three major groups of species exhibiting clear-cut differences in &-agonist-induced lipolysis. The first group corresponds to hyperresponders, such as the rat, golden hamster, and garden dormouse, in which the lipolytic response of the white adipocytes exhibits a great /3,-adrenergic component. The second group is composed of larger mammals, such as the rabbit and dog, and corresponds to hyporesponders, in which the /3,-adrenergic pathway exists but does not predominate over the /3,- and P,-adrenergic components. The third group includes the guinea pig, baboon, macaque, and human, whose SC white fat cells exhibit a very weak (if any) response to BRL37344 or CGP12177. Concerning humans, controversial studies suggested the lack (9, 31, 42) or the presence (43) of a functional &AR in white adipocytes. Moreover, there are also discrepancies about the detection of the &AR mRNA in human WAT (44, 45). Thus, the presence of Pa-ARs in adipose tissues of adult humans remains controversial. Although UCP mRNA has always been detected together with &AR mRNA in man (71, no studies have attempted to systematically investigate the presence of &-AR in human BAT, probably because of the low amounts of BAT assumed to occur in humans, except in newborns. As, on the one hand, the baboon was recently demonstrated to be an appropriate model for the study of the P-adrenergic control of WAT lipolysis in humans (10) and, on the other hand, some baboon fat pads could be considered as typical BAT (present data),
374
VIGUERIE-BASCANDS
we searched for the presence of P,-ARs through both biological effects and mRNA expression in brown fat. The stimulation of mitochondrial respiration is the most significant physiological effect of catecholamines on brown adipocytes, but requires isolation of intact and functional adipose cells, which is very difficult in BAT, especially in primates. Nevertheless, it is known that lipolysis is a good index of brown adipocyte activation, and this allowed us to compare our results to those of other studies in this field. The search for functional /3,-ARs in human adipose tissue has been performed mainly in isolated white adipocytes, except for the study of Rosenbaum et al. (31), in which &AR agonists were tested on fragments of SC adipose tissue. The present study demonstrates that baboon adipose tissues contain brown adipocytes. In preliminary experiments we observed that the methodology commonly used to isolate adipocytes from WAT-like deposits (successive washes of density-isolated packed cells) resulted in a progressive decrease in the UCP level (detected by Western blot) of the packed cells, leaving most of the brown adipocytes in the stroma vascular fraction and leading to the formation of white adipocyte-enriched fractions (data not shown). Thus, the use of small pieces of adipose tissue to perform lipolysis measurements allowed us to take into account the heterogeneity of fat deposits in the baboon. The results obtained in this study are similar to those described for isolated baboon white adipocytes (9), which also showed a very weak lipolytic effect with CGP 12177 on perirenal fat pads. Even though lipolysis assay is a less sensitive method than measurement of oxygen consumption in BAT, it has been previously shown that the P-AR agonist isoproterenol stimulates the release of glycerol in BAT to the same extent as that observed in the present study (46). Therefore, if a /3,-adrenergic component is functional in the brown adipocytes of baboons, we should observe a lipolytic effect of &-AR agonists in our experiments. Taken together, these results indicate a lack of functional P,-ARs in fat deposits of baboons with either high or low amounts of brown adipocytes, suggesting the absence of efficient /3,-ARs in both white and brown fat cells. The lack of &-agonist-induced lipolysis combined with the negative Northern blots suggest that &-AR is expressed in such small amounts that no lipolytic functionality of this receptor is detectable, even in distinctive brown fat pads. The need for 40 cycles of RT-PCR for sufficient detection of the &-AR mRNA, even in typical brown fat pads such as the pericardiac and axillary deposits, suggests that this receptor could be expressed either in a small population of adipocytes within the pads or faintly in every adipose cell. Nevertheless, it cannot be ruled out that the &-AR protein is so stable that there is no need for high levels of the corresponding mRNA. Although the &-AR is the predominant P-adrenergic subtype in rat BAT (47) and the copresence of UCP and &-AR has been described in human fat tissues (71, the present study suggests that BAT of nonneonatal baboons does not express the &AR. It is noteworthy that a recent study of HimmsHagen et al. (18) shows that guinea pig BAT lacks any thermogenic response to &agonists. Surprisingly, a recent study on molecular evolution claimed that the guinea pig is closer to primates than it is to other rodents (48). In conclusion, the present study shows that there are nu-
ET AL.
JCE & M . 1996 Vol81*Nol
merous brown adipocytes in fat pads of healthy nonneonatal baboons. This suggests that the amount of brown adipocytes has been underestimated in humans and that adipose tissues from adult primates have some potential for energy expenditure through thermogenesis. We also demonstrated that the fat deposits of the baboon display the same pattern of heterogeneity as previously described in rodents, suggesting the presence of potential convertible adipose tissues in primates. Finally, the lack of P,-ARs in typical BAT of nonneonatal baboons suggests that the expression of &ARs in brown adipocytes could be questioned in primate species (including man), but leaves opens the question of the potential involvement of the &-AR in BAT development.
Acknowledgments We are grateful to Drs. B. Verschure and G. Le Fur (Sanofi) for giving access to the baboon fat collection. We thank Mrs. M. Andre for her excellent technical assistance, Dr. C. At@ for careful review of this paper, and Dr. P. Winterton for revising the English version.
References Ricquier protein.
4. 5. 6. 7. 8. 9.
10.
11.
12.
13.
14.
15. 16.
17. 18.
19.
D, Casteilla
L, Bouillaud
F. 1991 Molecular
studies
of the uncoupling
FASEB J. 92237-2242.
Himms-Hagen J. 1990 Brown adipose tissue thermogenesis: interdisciplinary studies. FASEB J. 4:2890-2898. Lowell BB, S-Suzulic V, Hamann A, et al. 1993 Development of obesity in transgenic mice after genetic ablation of brown adipose tissue. Nature. 366: 740-742. Lean MEJ, James PT. 1986 Brown adipose tissue in man. In: Trayhurn I’, Nicholls D, eds. Brown adipose tissue. London: Arnold; 339-365. Oppert JM, Vohl MC, Chagnon M, et al. 1994 DNA polymorphism in the uncoupling protein (UCP) gene and human body fat. Int J Obesity. 18:526-531. Arch JRS, Kauman AJ. 1993 p3 and atypical p-adrenoceptors. Med Res Rev. 13663-729. Krief S, Liinnqvist F, Raimbault S, et al. 1993 Tissue distribution of p3adrenergic receptor mRNA in man. J Clin Invest. 91:344-349. Lafontan M, Berlan M. 1993 Fat cell adrenergic receptors and the control of white and brown fat cell function. J Lipid Res. 34:1057-1091. Bousquet-M&u A, Galitzky J, Carp&n& C, Lafontan M, Berlan M. 1994 P-Adrenergic control of lipolysis in primate white fat cells: a comparative study with nonprimate mammals. Am J Physiol. 267:R115-R123. Bousquet-M&u A, Galitzky J, Lafontan M, Be&n M. 1995 Control of lipolysis in intra-abdominal fat cells of non-human primates: comparison with humans. J Lipid Res. 36:451-461. Klaus S, Cassard-Doulcier AM, Ricquier D. 1991 Development of Pltodupus sun~~na brown preadipocytes in primary cell culture: effect of an atyplcal beta-adrenergic agonist, insulin, and triiodothyronin on differentiation, mitochondrial development, and expression of the uncoupling protein UCP. J Cell Biol. 115:1783-1790. Champigny 0, Ricquier D, Blonde1 0, Mayers RM, Briscoe MG, Holloway BR. 1991 P3-Adreneigic receptor stimulation restores message and expression of brown-fat mitochondrial uncoupling protein in adult dogs. Proc Nat1 Acad Sci USA. 88:10774-10777. Himms-Hagen J, Cui J, Danforth E, et al. 1994 Effect of CL 316,243, a thermogenic P,-adrenergic agonist, on energy balance and brown and white adipose tissues in rats. Am J Physiol. 266:R1371-R1382. Arch JRS, Cawthome MA, Coney KA, et al. 1991 P-Adrenoceptor-mediated control of thermogenesis, body composition and glucose homeostasis. In: Rothwell NJ, Stock MJ, eds. Obesity and cachexia. New York: Wiley and Sons; 241-268. Garutti G, Ricquier D. 1992 Analysis of uncoupling protein and its mRNA in adipose tissue deposits of adult man. Int J Obesity. 16:383-390. Kortelainen ML, Pelletier G, Ricquier D, Bukowiecki J. 1993 Immunohistochemical detection of human brown adipose tissue uncoupling protein in an autopsy series. J Histochem Cytochem. 41:759-764. Granneman JG. 1995 Why do adipocytes make the p3 adrenergic receptor? Cell Signal. 79-15. Himms-Hagen J, Triandafillou J, B&gin-Heick N, Ghorbani M, Kates AL. 1995 Apparent lack of /33-adrenoceptors and of insulin regulation of glucose transport in brown adipose tissue of guinea pigs. Am J Physiol. 268:R98-R104. Casteilla L, Forest C, Robelin J, Ricquier D, Lombet A, Ailhaud G. 1987
UCP AND
20.
21. 22. 23. 24.
25.
26.
27.
28.
29. 30.
31.
32.
33.
&-AR
EXPRESSION
Characterization of mitochondrial-uncoupling protein in bovine fetus and newborn calf. Am J Physiol. 252:E627-E636. Chomczynski P, Sacchi N. 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 762: 156-159. Emorine L, Blin N, Strosberg AD. The human P3-adrenoceptor: the search for a physiological function. TIBS. 15%7. Sambrook J, Fritsch EF, Maniatis T. 1989 Molecular cloning-a laboratory manual, 2nd ed. Cold Spring Harbor: Cold Spring Harbor Laboratory. Arch JRS, Ainsworth AT, Cawthome MA, et al. 1984 Atypical P-adrenoceptor on brown adipocytes as target for anti-obesity drugs. Nature. 309:163-165. Mohell N, Dicker A. 1989 The /3-adrenergic radioligand [3H1CGP-12177, generally classified as an antagonist, is a thermogenic agonist in brown adipose tissue. Biochem J. 261:401-405. Bloom JD, Dutia MD, Johnson BD, et al. 1992 Disodium (R,R)-5-[Z-[[2-(3chlorophenyl)-2 hydroxyethyl]-aminoJ-1,3-benzodioxole-2,2-dicarboxylate (CL 316,243). A potent beta-adrenergic agonist virtually specific for beta3 receptors. A promising antidiabetic and antiobesity agent. J Med Chem. 35: 3081-3084. Yoshida T, Sakane N, Wakabayashi Y, Umekawa T, Kondo M. 1994 Antiobesity and anti-diabetic effects of CL 316,243, a highly specific @3-adrenoceptor agonist, in yellow KK mice. Life Sci. 54:491-498. Langin D, Portillo M, Saulnier-Blache JS, Lafontan M. 1991 Coexistence of three p-adrenoceptor subtypes in white fat cells of various mammalian species. Eur J Pharmacol. 199:291-301. Galitzky J, Reverte M, Portillo M, Carp&16 C, Lafontan M, Berlan M. 1993 Coexistence of betal, beta2 and b&3-adrenoceptors on dog fat cells and their differential activation by catecholamines. Am J l’hysiol. 264:E403-E412. Emorine LJ, Man1110 S, Briend-Sutren MM, et al. 1989 Molecular characterization of the human p3-adrenergic receptor. Science. 245:1118-1121. Liggett SB. 1992 Functional properties of the rat and human P3-adrenergic receptors: differential agonist activation of recombinant receptors in Chinese hamster ovary cells. Mol Pharmacol. 42:634-637. Rosenbaum M, Malbon CC, Hirsch J, Leibel RL. 1992 Lack of P3-adrenergic effect of lipolysis in human subcutaneous adipose tissue. J Clin Endocrinol Metab. 77352-355. Casteilla L, Muzzin P, Revelli Jl’, Ricquier D, Giacobino JP. 1994 Expression of pl- and p2- adrenergic-receptor messages and adenylate cyclase P-adrenergic response in bovine perirenal adipose tissue during its transformation from brown into white fat. Biochem J. 297~93-97. Hansen BC. 1992 Obesity and diabetes in monkeys. In: Bjijrntorp P, Brodoff BD, eds. Obesity. Philadelphia: Lippincott; 256-265.
IN BABOON
FAT PADS
375
34. Lewis DS, Soderstrom PG. 1993 In vim and in vitro development of visceral adipose tissue in a nonhuman primate (Pnpio species). Metabolism. 42:12771283. 35. Huttunen P, Hirvonen J, Kinnula V. 1981 The occurence of adipose tissue in outdoor workers. Eur J Appl Physiol. 46~339-345. 36. Huttunen P, Kortelainen ML. 1990 Long term alcohol consumption and brown adipose tissue in man. Eur J Appl Physiol. 60~418-424. 37. Casteilla L, Champigny 0, Bouillaud F, Rob&n J, Ricquier D. 1989 Sequential changes in the expression of mitochondrial protein mRNA during the development of brown adipose tissue in bovine and ovine species. Biochem J. 257~665 671. 38. Cousin B, Cinti S, Morroni M, et al. 1992 Occurence of brown adipocytes in rat white adipose tissue: molecular and morphological characterization. J Cell Sci. 103:931-942. 39. Loncar D. 1991 Convertible adipose tissue in mice. Cell Tissue Res. 266:149161. 40. Connacher AA, Jung RT, Mitchell PEG. 1988 Weight loss in obese subjects on a restricted diet given BRL 26830A, a new atypical p adrenoceptor agonist. Br Med J. 296:1217-1220. 41. Wheeldon NM, McDevitt DG, McFarlane LC, Lipworth BJ. 1994 P-Adrenoceptor subtypes mediating the metabolic effects of BRL 35135 in man. Clin Sci. 86331-337. 42. Van Liefde I, Van-Ermen A, Vauquelin G. 1994 No functional atypical betaadrenergic receptors in human omental adipocytes. Life Sci. 54:209-214. 43. LSnnqvist F, Krief S, Strosberg D, Nyberg B, Emorine LJ, Amer P. 1933 Evidence for a functional P3-adrenoceptor in man. Br J Pharmacol. 110:929936. 44. Granneman JG, Lahners KN, Rao DD. 1992 Rodent and human P3-adrenergic receptor genes contain an intro” within the protein coding block. Mel Pharmacol.42:964-970. 45. Thomas RF, Liggett SB. 1992 Lack of /33-adrenergic receptor mRNA expression in adipose and other metabolic tissues in the adult human. Mol Pharmacol. 43:343-348. 46. Bukowiecky LJ, Follea N, Lupien J, Paradis A. 1981 Metabolic relationship between lipolysis and respiration in rat brown adipocytes. J Biol Chem. 256: 12840-12848. 47. Muzzin P, Revelli JP, Kuhne, et al. 1991 An adipose tissue specific P-adrenergic receptor. J Biol Chem. 266:24053-24058. 48. Li WH, Hide WA, Zharkikh A, Ma DP, Graur D. 1992 The molecular taxonomy and evolution of the guinea pig. J Hered. 83:174-181.
