Effects of octopamine on lipolysis, glucose transport and amine oxidation in mammalian fat cells

Effects of octopamine on lipolysis, glucose transport and amine oxidation in mammalian fat cells

Comparative Biochemistry and Physiology Part C 125 (2000) 33 – 44 www.elsevier.com/locate/cbpc Effects of octopamine on lipolysis, glucose transport...

160KB Sizes 0 Downloads 6 Views

Comparative Biochemistry and Physiology Part C 125 (2000) 33 – 44

www.elsevier.com/locate/cbpc

Effects of octopamine on lipolysis, glucose transport and amine oxidation in mammalian fat cells Emi Fontana, Nathalie Morin, Danielle Pre´vot, Christian Carpe´ne´ * Institut National de la Sante´ et de la Recherche Me´dicale (INSERM), Unite´ 317, Institut Louis Bugnard Bat L3, CHU Rangueil, 31403 Toulouse, France Received 15 April 1999; received in revised form 3 September 1999; accepted 9 September 1999

Abstract Octopamine is known to exert adrenergic effects in mammals although speficic octopamine receptors have been cloned only in invertebrates. It has been shown that octopamine can stimulate a2-adrenoceptors (ARs) in Chinese hamster ovary cells tranfected with human a2-ARs. More recently, we reported that octopamine stimulates lipolysis through b3-rather than b1-or b2-AR activation in white adipocytes from different mammalian species. The present study was thus undertaken to further characterize the adrenergic properties of octopamine. For this purpose, several biological processes known to be regulated by adrenergic stimulation were studied in response to octopamine, noradrenaline, adrenaline and tyramine in white adipocytes from different mammals. First, octopamine was fully lipolytic in garden dormouse and Siberian hamster while tyramine was ineffective. Although being around one hundred-fold less potent that noradrenaline, octopamine was slightly more potent in these hibernators known for their high sensitivity to b3-AR agonists than in rat and chiefly more active than in human adipocytes known for their limited responses to b3-AR agonists. Second, octopamine reduced insulin-dependent glucose transport in rat fat cells, a response also observed with noradrenaline and selective b3-AR agonists but not with b1-or b2-agonists. Third, human adipocytes, which endogenously express a high level of a2-ARs, exhibited a clear a2-adrenergic antilipolytic response to adrenaline but not to octopamine. Moreover, octopamine exhibited only a very weak affinity for the a2A-ARs labeled by [3H]RX 821002 in human adipocyte membranes. In Syrian hamster adipocytes, which also possess a2-ARs, octopamine induced only a weak antilipolysis. Finally, octopamine was a substrate of fat cell amine oxidases, with an apparent affinity similar to that of noradrenaline. All these results demonstrate that octopamine, tyramine noradrenaline and adrenaline can be degraded by adipocyte amine oxidases. However these biogenic amines interact differently with adipocyte adrenoceptors: tyramine is inactive, adrenaline and noradrenaline activate both b- and a2-ARs while octopamine activates only b3-ARs and is devoid of a2-adrenergic agonism. Thus, octopamine could be considered as an endogenous selective b3-AR agonist. © 2000 Elsevier Science Inc. All rights reserved. Keywords: b3-adrenoceptor; a2-adrenoceptor; Octopamine; Adipocyte; Lipolysis; Tyramine; Monoamine oxidase; Semicarbazide-sensitive amine oxidase

1. Introduction

* Corresponding author. Tel.: + 33-5-62172955; fax: +335-61173321. E-mail address: [email protected] (C. Carpe´ne´)

Octopamine, which structurally corresponds to the ring-dehydroxylated product of noradrenaline, has been shown to act as a neurohormone or a neuromodulator in invertebrates (Osborne et al.,

0742-8413/00/$ - see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 7 4 2 - 8 4 1 3 ( 9 9 ) 0 0 0 8 6 - 9

34

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

1996). At least three octopamine receptors (OAR1 to OAR3) have been cloned in crustaceans and insects (Roeder, 1995; Roeder et al., 1995) and some of their intracellular signalling pathways have been depicted (Evans et al., 1993). In insects, some of the octopamine receptors are also stimulated by tyramine (Robb et al., 1994). In mammals, octopamine was initially classified as a false neurotransmitter (Trendelenburg et al., 1972) and its physiological function is still poorly defined. Several in vivo actions of octopamine have been described in mammals such as an induction of salivary secretion in rat, with a pharmacological profile suggesting the involvement of both b- and a1-adrenoceptors (Okina et al., 1992) or a hypoprolactinemic action, probably due to dopamine release or by direct action on pituitary cell dopaminergic receptors (Becu-Villalobos et al., 1992). From these studies, it is difficult to elucidate whether octopamine acts on the receptors, either by direct binding at postsynaptic receptors or through modification of neurotransmitter release. However, in vitro studies evidenced that octopamine is able to activate a1-adrenoceptors in the rat corpus cavernosum (Tong and Cheng, 1997) or a2-adrenoceptors in a transfected cell system (CHO overexpressing human a2-ARs) (Airriess et al., 1997). In these models, octopamine acted as a full agonist but with weak potency (pD2 around 5.3 – 4.3). b3-AR agonist properties of octopamine were also reported in dog (Galitzky et al., 1993) and rat (Yen et al., 1998) adipocytes. Recently, we observed that octopamine acts in vitro as a lipolytic agent on rat, hamster and dog fat cells through direct b3-AR activation with a relatively weak potency (pD2 in the range of 5.6–4.6) while it was inactive at b1and b2-ARs in guinea pig and human adipocytes (Carpe´ne´ et al., 1999). From these studies it is therefore conceivable that octopamine can mimic some of the effects of noradrenaline and adrenaline on the ARs. Since adipose cells express many of the diverse a- and b-AR subtypes, it is widely accepted that, the relative proportion of each AR subtype influences the b-lipolytic/a-antilipolytic balance, and thus controls the magnitude of lipid mobilization in response to adrenergic system activation (Lafontan and Berlan, 1993). Indeed, the expression of adipose a- and b-ARs exhibits large differences, according to the hormonal status, the anatomical location of fat depot, or the animal

species considered. In this work, we further studied the effects of octopamine on adipocytes from different mammalian species in order to determine whether octopamine is able to directly activate both a- and b-AR subtypes (as do the endogenous catecholamines adrenaline and noradrenaline) or can be considered as a selective agonist stimulating only a given AR subtype. Several adrenergic responses of white adipocytes, namely stimulation of lipolysis (mediated by b1-, b2- and/or b3-AR activation) (Lafontan and Berlan, 1993), inhibition of insulin action on glucose transport (mediated by b3-ARs) (Carpe´ne´ et al., 1993) and antilipolysis (mediated by a2-AR) (Lafontan and Berlan, 1993) were thus tested for octopamine and for other biogenic amines (noradrenaline, adrenaline and tyramine) in white fat cells from various species chosen in function of their previously described particularities, such as the predominence of a given AR subtype. In this context, two hibernators were studied, namely the garden dormouse or lerot (Eliomys quercinus) and the Siberian hamster (Phodopus sungorus sungorus). These two species behave differently when exposed to cold and/or short photoperiod: garden dormice readily hibernate whereas Siberian hamsters only undergo short torpor periods and their fur turns into white (Lyman, 1982). Despite these different adaptative behaviours, both models share a particularly high sensitivity of their white adipocytes to b3-adrenergic agonists (Carpe´ne´ et al., 1994; Atgie´ et al., 1998) and were classified as highly-responsive, regarding to their b3-adrenergic-dependent lipolytic responsiveness (Lafontan and Berlan, 1993; Carpe´ne´ et al., 1998). By contrast, human fat cells exhibit only weak activation of lipolysis in response to b3-agonists when compared to b1-, b2selective agonists or to the mixed b-agonist isoprenaline (Langin et al., 1991; Lafontan and Berlan, 1993; Umekawa et al., 1996; Carpe´ne´ et al., 1998, 1999). Nevertheless, human adipocytes are known to express high levels of a2-AR, the stimulation of which induces strong antilipolytic response (Lafontan and Berlan, 1993), and were therefore used to test the a2-AR agonist properties of octopamine in antilipolysis and binding experiments. Syrian hamster (Mesocricetus auratus) adipocytes, which are characterized by a substantial a2-adrenergic receptivity (Lafontan and Berlan, 1993) were also used since in rat and in the two hibernators mentionned above,

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

adipocytes exhibit only blunted a2-adrenergic antilipolytic responses (Carpe´ne´ et al., 1994; Atgie´ et al., 1998). Whatever the relative proportion of their adrenoceptor subtypes, adipocytes are able to oxidize biogenic amines via mitochondrial monoamine oxidases (MAO) and/or membranebound semicarbazide-sensitive amine oxidases (SSAO) (Raimondi et al., 1992). We therefore tested the capacity of octopamine of being a substrate of such amine oxidases. Tyramine, a good substrate of MAOs in mammals (Yu, 1986) also served as reference since it is known to be active at the octopamine/tyramine receptor in insects (Robb et al., 1994). In the following study, we report that, in different mammals, octopamine stimulates only several of the adrenergic responses of white adipocytes. Octopamine effects (increase in lipolytic activity and reduction of insulin action on glucose uptake) are mainly b3-adrenergic in nature since no activation of a-ARs was clearly evidenced. In mammals, octopamine behaves differently from tyramine which is more readily oxidized by amine oxidases and does not activate ARs.

2. Materials and methods

2.1. Origin of adipose tissue samples Samples of human subcutaneous adipose tissue were obtained from non obese women undergoing either mammary or abdominal dermolipectomy in the Department of plastic surgery of Toulouse Rangueil hospital (mean body mass index was 24.6 9 0.9 and age ranged from 22 to 55 years; n= 12). The patients gave their consent and the study was approved by the Ethical Committee of Toulouse University Hospital. Samples of human adipose tissue were transferred in less than 30 min from surgery to the incubation bath in which they were subjected to collagenase digestion. Male Wistar rats (200–250 g) were purchased from Harlan breeding center (Gannat, F) and male Syrian hamsters (Mesocricetus auratus) of 90 – 110 g were from Janvier breeding center (Le Genest, F). The garden dormice were lawfully captured in France at least 6 months before experiments and served as controls in other investigations relative to hibernation conducted by Pr. L. Ambid (UPRESA CNRS 5018, Toulouse). Mature animals

35

were of both sexes and weighed around 100 g. Siberian hamsters (Phodopus sungorus sungorus) were raised in Pr. L. Ambid breeding colony and weighed 45–50 g at sacrifice. All animals were housed individually at 22–24°C on a long-day cycle (16 h:8 h, light:dark) with free access to food and water. Samples of white adipose tissues were removed from retroperitoneal, perirenal and perigonadal locations, then pooled and subjected to collagenase digestion (1–1.5 mg/ml) in a Krebs–Ringer solution buffered with 10 mM Hepes plus 15 mM bicarbonate (pH 7.4) and containing 3.5% (w/v) bovine serum albumin and 6 mM glucose, (KRBHA). Adipocytes were then isolated and washed three times in KRBHA as previously reported (Carpe´ne´ et al., 1994).

2.2. Lipolysis and glucose uptake measurement Isolated fat cells (:20 mg lipids) were incubated in 0.5 ml KRBHA at 37°C with gentle shaking. Pharmacological agents were added at suitable dilutions in a 5 ml volume to the cell suspension just before the beginning of the assay. After 60 or 90 min of incubation, the tubes were placed in an ice bath and 200 ml aliquots of the infranatant were taken for enzymatic determination of glycerol, used as lipolysis index as previously reported (Carpe´ne´ et al., 1994). Deoxyglucose uptake was measured during 5 min after preincubation of 45 min with the indicated drugs as already described (Carpe´ne´ et al., 1993). Final concentration of 2-deoxyglucose was 0.1 mM and KRBHA buffer was supplemented with 2 mM pyruvate instead of 6 mM glucose present in lipolytic assays.

2.3. Membrane preparation and binding studies Isolated fat cells were lysed in a hypotonic medium (0.5 mM MgCl2, 2 mM Tris–HCl, pH 7.5). Crude membranes were pelleted by centrifugation (45 000×g, 15 min at 4°C) and resuspended in binding medium (0.5 mM MgCl2, 50 mM Tris–HCl, pH 7.5). The radioactive ligand retained on glass fiber filters (Whatman GF/C) was counted in liquid scintillation cocktail (Emulsifier Safe, Packard) using a Packard beta counter. Binding studies on human fat cell membranes were performed using [3H]RX 821002 for labelling a2-ARs in a final volume of 400 ml at 25°C. Non-specific binding was evaluated with

36

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

200 mM adrenaline. This non-specific binding accounted for 79 1% of total [3H]RX 821002 binding and was equivalent to that obtained with the a2-antagonist rauwolscine at 100 mM.

2.4. Amine oxidase acti6ity Crude membranes were incubated in 200 mM phosphate buffer with 0.5 mM [14C]tyramine for 30 min and the oxidation products were measured as previously described (Carpe´ne´ et al., 1995) according to the method of Yu (1986).

2.5. Drugs The adrenergic properties, the relative selectivies and the complete chemical names of the drugs used in this study were previously reported (Lafontan and Berlan, 1993; Carpe´ne´ et al., 1999). The b3-AR agonists used were the following: CL 316243 (full b3-AR agonist) obtained from Lederle´ Laboratories (American Cyanamid Company, New York); BRL 37344 (full b3-AR agonist) and Z D7114 (partial b3-AR agonist, previously named ICI D7114), respectively provided by Dr M.A. Cawthorne (Smith KlineBeecham Pharmaceuticals, Epsom, UK) and by Dr B. Holloway (ICI Ltd, Macclesfield, UK). The b-AR antagonists used were: CGP 20712A (b1AR antagonist) from Ciba Geigy (Basle, Switzerland), ICI 118551 (b2-AR antagonist) from Imperial Chemistry Industries (Alderly Park, Macclesfield, UK), (9)Bupranolol (b1/b2/b3-AR mixed antagonist) given by Schwarz Pharma (Monheim, Germany) and (9)propranolol (b1/ b2/b3-AR mixed antagonist) purchased from Sigma. The a2-AR agonists used were clonidine (Boerhinger Mannnheim) and UK 14304. Three distinct a2-AR antagonists were used: MK 912 (generously donated by Merck Research, West Point, PA); (9 )RX 821002 and yohimbine (from Sigma). Crude collagenase, bovine serum albumin (fraction V), adenosine deaminase (ADA), tyramine, (− )isoprenaline, (−)noradrenaline, (−) adrenaline, (9 )p-octopamine, and other chemicals were obtained from Sigma (St-Quentin, France). [3H]RX 821002 (43 Ci/mmol) came from Amersham, [3H]2-deoxyglucose (26 Ci/mmol) and [14C]tyramine (45 mCi/mmol) were from New England Nuclear.

2.6. Statistical analysis Values are means 9 S.E.M. Unless otherwhise stated, unpaired Student’s t-tests were used for comparisons, differences being considered significant when P was smaller than 0.05. The pD2 value represents −log(EC50), EC50 being the molar concentration of agonist inducing half of the maximal response determined by Hill plot linear regression.

3. Results

3.1. Effect of octopamine on adipocyte lipolysis in hibernators and in man Adipocyte preparations from garden dormice and Siberian hamsters exhibited similar lipolytic activity and were both very sensitive to the selective b3-AR agonist CL 316243 with respective pD2 values of 8.48 and 8.53 (Fig. 1). The maximal lipolytic response to 10 − 5 M isoprenaline was also reached with the three biogenic amines tested: noradrenaline, adrenaline and octopamine with respective pD2 values of: 7.63, 7.42 and 5.50 in garden dormouse; 7.76, 6.80 and 5.13 in Siberian hamster (Fig. 1). On the contrary, human subcutaneous adipocyte preparations hardly responded to 1 mM CL 316243 (11.2 9 3.2% of the maximal isoprenaline effect, n= 12). In these cells, octopamine was inefficient at 10 − 5 –10 − 4 M and partially stimulated lipolysis only at the highest tested concentration (10 − 3 M), reaching 53.7 9 12.5% of the maximal isoprenaline effect (n= 12). Indeed, octopamine activated less than 2-fold the spontaneous lipolytic activity of human adipocytes while isoprenaline 10 − 5 M maximally stimulated basal lipolysis by a 3.5490.64-fold factor (not shown). Noteworthy, tyramine (up to 1 mM) was not lipolytic at all in any of the species studied (not shown).

3.2. Comparison between the effects of noradrenaline and octopamine on glucose transport into rat adipocytes In rat, b3-AR agonists not only stimulate lipolysis, but also inhibit insulin-stimulated glucose transport when the adenosine endogenously produced by isolated adipocytes is removed from the incubation medium. This inhibitory effect is not

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

37

observed with b1- or b2-AR agonists (Carpe´ne´ et al., 1993). In order to test whether octopamine was able to counteract the insulin effect in a b3-adrenergic-dependent manner, 2-deoxyglucose uptake was stimulated in rat adipocytes by insulin 10 nM in the presence of adenosine deaminase (ADA, 2 IU/ml). In these conditions, the stimu-

Fig. 2. Inhibition of insulin-stimulated glucose transport by b3-AR agonists, noradrenaline and octopamine in rat adipocytes. Experiments were carried out in the presence of adenosine deaminase (ADA, 2 IU/ml) in order to remove adenosine accumulated in the incubation medium. 2-deoxyglucose uptake (2-DG) was 0.59 90.06 and 4.879 0;75 nmol/100 mg lipid/5 min in basal and insulin-stimulated conditions, respectively. b3-AR agonists (CL 316243 and Z D7114) or biogenic amines (noradrenaline and octopamine) were present at the indicated concentrations Results are expressed as percentage of stimulated transport. Values are mean 9 S.E.M. (n =4).

Fig. 1. Comparative study of the lipolytic reponses induced by the agonist CL 316243, and noradrenaline, adrenaline or octopamine. White fat cells from (upper panel) or Siberian hamster (lower panel) were incubated with the indicated concentrations of CL 316243 (closed circles), noradrenaline (squares), adrenaline (triangles) or octopamine (open circles). To standardize the responses to the agonists, results are expressed in percentage of the maximal increase in glycerol release promoted by isoprenaline (10 mM). Spontaneous and isoprenaline-stimulated glycerol release were: 0.60 9 0.14 vs. 3.739 0.46 and 0.589 0.08 vs. 3.049 0.20 mmol of glycerol released per 100 mg of cell lipid per 90 min, in garden dormouse and Siberian hamster, respectively. Values are the mean9 S.E.M. of four to six different experiments.

lated glucose transport (equivalent to 11.69 1.7fold over basal uptake, n=4) was partially inhibited in a dose-dependent manner by b3-AR agonists, but also by noradrenaline and octopamine (Fig. 2). Octopamine effect was weaker than that of noradrenaline or CL 316243, but was comparable to that of a partial b3-AR agonist, Z D7114. Moreover, the effect of 100 mM octopamine was abolished by 10 mM bupranolol or propranolol (b1/b2/b3-AR mixed antagonists) and was resistant to 10 mM CGP 20712A or ICI 118551 (b1- and b2-AR selective antagonist, respectively, not shown). In keeping with the previously reported pharmacological profile of octopamine on lipolysis activation in rat adipocytes (Carpe´ne´ et al., 1999) and with the similarities observed here between biogenic amines and b3-agonists, the counter-regulatory effect of octopamine on insulin action appears to be mediated by b3-AR activation.

38

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

3.3. Comparison of the a2 -AR-mediated antilipolytic responses to adrenaline and octopamine Since a2-AR agonists induce antilipolytic responses in adipocytes, we investigated whether octopamine was able to induce such antilipolysis. This was first conducted on human adipocytes since they exhibit the strongest a2-adrenergic antilipolytic responsiveness among the species studied so far (Lafontan and Berlan, 1993). As shown in Fig. 3, the lipolysis promoted by ADA 2 IU/ml was inhibited by adrenaline, which behaved as an a2-AR agonist in the presence of 10 − 5 M propranolol (blocking the b-adrenergic component). In the same conditions, octopamine was unable to clearly reduce ADA-activated lipolysis. In complementary experiments (Carpe´ne´ et al., 1998), the a2-AR antagonist RX 821002 potentiated the lipolytic effect of adrenaline or noradrenaline by unmasking their inhibitory a2-adrenergic component in human adipocytes but did not modify the weak response to octopamine (not shown). In hamster adipocytes, which also express high levels of a2-ARs (Lafontan and Berlan, 1993;

Fig. 3. Lack of antilipolytic effect of octopamine plus propranolol in human subcutaneous adipocytes. Spontaneous lipolysis of human adipocytes (0.1490.04 mmol of glycerol released per 100 mg cell lipids during 90 min) was stimulated by adenosine deaminase (ADA 2 IU/ml) plus 10 mM propranolol (control) in order to detect a2-adrenergic antilipolytic response to the indicated concentrations of octopamine (open symbols) or adrenaline (closed symbols). Values are mean 9 S.E.M. of six experiments.

Fig. 4. Antilipolytic effects of octopamine and clonidine in Syrian hamster adipocytes. Lipolysis was stimulated by adenosine deaminase (ADA 4 IU/ml) plus 100 mM bupranolol (4.1 90.5-fold increase over basal) in order to detect antilipolytic responses to the indicated concentrations of octopamine (circles) or clonidine (squares) without (control, closed symbols) or with 10 mM yohimbine (open symbols). Results expressed as percentage of ADA-stimulated lipolysis are mean 9S.E.M. of three experiments.

Carpe´ne´ et al., 1995), 100 mM octopamine induced only 21+8% inhibition of ADA-dependent lipolysis in the presence of 100 mM bupranolol (n=3) while the a2-AR agonist clonidine counteracted quite totally the lipolytic effect of ADA in a dose-dependent manner (Fig. 4). Clonidine effect was unmodified by bupranolol (not shown) but was sensitive to a2-adrenergic blockade. When tested at higher concentrations (1 mM), octopamine further inhibited ADA-stimulated lipolysis but this effect was uncomplete and insensitive to the blockade by the a2-AR antagonist yohimbine (Fig. 4). Finally, no antilipolytic effect of octopamine could be evidenced in rat adipocytes because they exhibit poor a2-adrenergic antilipolytic effect: (1) clonidine 10 mM induced only 8.09 3.3% inhibition of the lipolysis stimulated by ADA 5 IU/ml (n= 4, NS, compare with syrian hamster in Fig. 4); (2) the presence of 10 mM RX 821002 did not modify the dose responsecurves for octopamine. Tyramine was not antilipolytic in all the species studied (not shown). Although octopamine was apparently unable to fully activate an a2-AR-mediated process, its affinity towards a2-AR was tested in binding studies.

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

3.4. Affinity of octopamine towards a2 -ARs in human and hamster fat cells Bmax value of [3H]RX 821002 binding to human adipocyte membranes was 8899 130 fmol/mg protein and Kd was 1.990.4 nM (n =4). For competition studies, [3H]RX 821002 was used at 3 nM and bound up to 447941 fmol mg − 1 protein in human adipocyte membranes. In these conditions, the a2-AR antagonists MK 912 and RX 821002 exhibited high affinity in competing for [3H]RX 821002 binding (Ki was 0.690.1 and 1.4 9 0.5 nM, n= 3), confirming that the a2-ARs present on human adipocytes are of the a2A-subtype, as previously evidenced by pharmacological and genetic approaches (Castan et al., 1995). In the same conditions, Ki for octopamine in displacing RX 821002 binding was 39109360 nM (n = 3). For comparison, the Ki of an a2-AR agonist of reference, UK 14304, was 17 nM and that of adrenaline was 140 nM in the same conditions (not shown). Thus, the low affinity binding of octopamine to a2A-ARs can probably explain its weak antilipolytic effect. In hamster adipocyte membranes, [3H]RX 821002 labeled 770960 fmol mg − 1 protein with a Kd of 1.0 9 0.1 nM (n = 4). The a2-AR antagonists MK 912 and RX 821002 competed for the radioligand binding with respective Ki of 1.190.1 and 1.49 0.1 nM (n= 3). While UK 14304 and adrenaline totally inhibited

39

[3H]RX 821002 binding with respective Ki of 14 9 1 and 419 14 nM, octopamine hardly inhibited 25% of the binding at 1 mM, the maximal dose tested (n=3, not shown).

3.5. Comparison between the oxidati6e deamination of adrenaline and octopamine by adipocyte amine oxidases All the weak agonist properties of octopamine could be the result of a substantial degradation of this amine by fat cells. This hypothesis was tested by measuring the capacity of adipocytes to oxidize amines. In crude membrane preparations from rat adipocytes, [14C]tyramine oxidation was saturable, reaching a plateau at 0.5 mM and characterized by a Vmax value of 12 9 1 nmol of tyramine oxidized/mg protein/min and a Km of 85 918 mM. As already reported (Marti et al., 1998), tyramine oxidation by rat fat cells is sensitive to pargyline, an inhibitor of mitochondrial monoamine oxidases (MAO) and to semicarbazide, an inhibitor of the membrane-bound semicarbazide-sensitive amine oxidase (SSAO). The effect of these two inhibitors are additive suggesting that MAO-A and MAO-B, as well as SSAO are involved in this oxidative deamination (Marti et al., 1998). The oxidation of 0.5 mM [14C]tyramine was dose-dependently inhibited by cold tyramine, octopamine, noradrenaline and adrenaline with respective IC50 values of 5219230, 20079 592,

Fig. 5. Inhibition of tyramine oxidation by octopamine, noradrenaline and adrenaline in rat adipocyte membranes. [14C]Tyramine (0.5 mM) was oxidized during 30 min with crude membrane preparations in the presence of the indicated concentrations of biogenic amines which were added 15 min prior to incubation. Crude membranes (95 mg protein/100 ml) were prepared from isolated rat adipocytes as described in Section 2. Results are expressed as the percentage of the amine oxidase activity measured in the absence of any competitor and which was equivalent to: 10.7 9 2.0 nmol of tyramine oxidized/min/mg protein. Values are mean 9 S.E.M. of three to five experiments.

40

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

1872998 and 129709 2855 mM (Fig. 5). Octopamine was oxidized both by MAO (at 10 mM, it inhibited 70% of tyramine oxidation in the presence of 1 mM semicarbazide which totally blocked SSAO) and by SSAO (10 mM octopamine inhibited 98% of tyramine oxidation in the presence of 0.5 mM pargyline in order to block MAO). In rat adipocytes, all these inhibitors did not improve noticeably the lipolytic potency of octopamine (not shown) or noradrenaline (Carpe´ne´ et al., 1995) which shared the same maximal effect. Crude membrane preparations from human adipocytes were also able to oxidize [14C]tyramine with a Vmax of 219 3 nmol of tyramine oxidized/ mg protein − 1 per min (n =3). The oxidation of 0.3 mM [14C]tyramine was dose-dependently inhibited by octopamine and noradrenaline with respective IC50 of 236 and 850 mM (not shown). Finally, we tested whether the oxidation of octopamine could explain its low efficiency in the lipolytic assays in human fat cells. The presence of 1 mM pargyline, semicarbazide, or phenelzine (an inhibitor of both MAO and SSAO) did not modify the weak lipolytic effect of octopamine 0.1 mM in human adipocytes (basal: 0.23 90.06; octopamine: 0.369 0.07; octopamine +pargyline: 0.31 9 0.07; octopamine+ semicarbazide: 0.359 0.08; octopamine+phenelzine: 0.359 0.07 mmol glycerol/100 mg lipid/90 min, n =7). Taken together, all these data demonstrate that octopamine can be considered as a good substrate for MAO or SSAO, but like other biogenic amines, it is readily oxidized at concentrations higher than 100 mM.

4. Discussion In the present study, convergent data suggest that octopamine can directly interact with mammalian fat cells: (1) Although less potent than noradrenaline, octopamine is fully lipolytic in adipocytes isolated from hibernators which are extremely sensitive to selective b3-AR agonists (Carpe´ne´ et al., 1994; Atgie´ et al., 1998); (2) octopamine inhibits insulin-dependent activation of glucose transport in rat white fat cells, as did b3-selective but not b1- or b2-selective AR agonists (Carpe´ne´ et al., 1993); (3) octopamine does not exhibit an antilipolytic a2-adrenergic component like adrenaline in species (hamster, man) which

express numerous adipocyte a2-ARs (Lafontan and Berlan, 1993; Carpe´ne´ et al., 1995); (4) octopamine can be oxidized by MAOs and SSAOs present in fat cells (Marti et al., 1998). The effects of octopamine on adipocytes are different from those of noradrenaline and adrenaline (which activate all the b- and a-AR subtypes), or tyramine (which is not lipolytic but readily oxidized). This singular profile makes octopamine an endogenous amine particulary able, when present at 1–100 mM, to activate b3-ARs without noticeable stimulation of b1-, b2- or a2-ARs or degradation by amine oxidases. The garden dormouse and the Siberian hamster differ strongly in their adaptation to cold, only the former being classified as a deep hibernator (Lyman, 1982). Nevertheless, their white adipocytes share a good responsiveness to b3-ARmediated activation of lipolysis, previously evidenced by the use of the atypical b-AR agonist BRL 37344 (Carpe´ne´ et al., 1994; Atgie´ et al., 1998), and confirmed here with CL 316243, a more selective b3-AR agonist (Bloom et al., 1992). Octopamine was clearly less potent than noradrenaline but appeared to be more potent in these two hibernators than in rat, for which we previously reported a pD2 value of 4.60 and demonstrated that octopamine effect was not mediated through b1- or b2-AR activation, but essentially resulted from b3-AR activation (Carpe´ne´ et al., 1999). These findings are in agreement with previous studies reporting pharmacological properties of octopamine close to those of the b3-AR agonists in dog (Galitzky et al., 1993), Syrian hamster (Carpe´ne´ et al., 1999) and rat fat cells (Yen et al., 1998). In these species, classified as hyper- and normo-responsive, regarding to the b3-adrenergic sensitivity of their adipocytes, the lipolytic effect of octopamine was relatively resistant to b1- or b2-AR selective antagonists (Galitzky et al., 1993; Carpe´ne´ et al., 1999). Another biological event which resists to the b1- or b2-AR blockade is the inhibition of insulin-dependent glucose transport. This b3-adrenergic inhibitory effect on glucose uptake has been already described on rat fat cells with the b3-AR agonist BRL 37344, but also with noradrenaline (Carpe´ne´ et al., 1993), octopamine (Yen et al., 1998) and dopamine (Lee et al., 1998). In the present work, we report that octopamine inhibits insulin-dependent glucose transport with a weaker efficiency than CL 316243, one of the most powerful b3-AR agonists, but in a manner

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

comparable to that of ZD 7114, a partial b3-AR agonist. In human adipocytes, b3-AR activation is far from being considered predominant in the lipolytic action of catecholamines and the human b3ARs have been evidenced only by detection of low levels of their mRNAs (Krief et al., 1993) while, in rodents, adipocyte b3-ARs have been evidenced by relatively large amounts of mRNAs and direct binding experiments (Tavernier et al., 1995). In keeping with their low b3-adrenergic responsiveness, human adipocytes which substantially respond to b1-and b2-AR agonists (Carpe´ne´ et al., 1999), exhibited weak lipolytic response to octopamine. This blunted octopamine effect contrasts with the 5- to 6-fold activation observed in the two hibernators but confirms the very poor efficacy of octopamine and b3-AR agonists we already reported in human subcutaneous adipocytes and guinea pig white adipocytes (Carpe´ne´ et al., 1999). From these data, the b-adrenergic lipolytic effect of octopamine can be considered as being predominantly due to agonism at b3-ARs, but not at b1-and b2-ARs. However, the low lipolytic efficiency of octopamine in human fat cells could have been explained by a putative dual action on both b-and a2-ARs, like for adrenaline in certain obesities (Lafontan and Berlan, 1993). The description of a direct action of octopamine isomers on cells overexpressing human a2A-ARs (Airriess et al., 1997) led us to verify whether octopamine was able to stimulate a2-ARs in human adipocytes. However, our experiments, carried out on subcutaneous human adipocytes, which endogenously express a2-ARs and do not possess the disadvantages of transfected cell systems (extra-physiological overexpression, alterations in coupling efficiency,...) did not reveal antilipolytic response to (9)p-octopamine: (1) there was no clear antilipolytic response to octopamine on ADA-stimulated lipolysis, even in the presence of a b-AR antagonist and; (2) there was no potentiation of the small lipolytic effect of octopamine by a2-AR antagonist while the antilipolytic a2-adrenergic effect of adrenaline was, as already reported, unmasked by the presence of b-AR antagonists (Lafontan and Berlan, 1993) and blocked by a2-AR antagonists (Carpe´ne´ et al., 1998). The lack of antilipolytic effect of octopamine in human adipocytes can be explained in part by its poor binding affinity for a2A-ARs which is 30-fold lower than that of adrenaline.

41

This contrats with the octopamine affinity towards b3-ARs which was only 2-fold lower than that of noradrenaline in CHO cells expressing b3-ARs (Carpe´ne´ et al., 1999). Alternatively, as reported on CHO cells overexpressing human a2AARs, octopamine could promote a peculiar activation state of the a2A-AR, leading to transduction signals and subsequent cellular responses slightly different from those observed with catecholamines (Airriess et al., 1997). Another example of biogenic amine selective for a subset of ARs is agmatine: this putative endogenous ligand of the imidazoline I2-sites (Li et al., 1994) is able to recognize a2-ARs only but is unable to activate or inhibit them (Pinthong et al., 1995). In hamster adipocytes, the antilipolytic effect of octopamine was insensitive to yohimbine and was probably due to a nonadrenergic signal, to an atypical interaction with a2-ARs, or to another unknown mechanism. Whatever the molecular mechanisms involved in the interaction between octopamine and a2-ARs, it is important to note that (9 )p-octopamine at 100 mM is able to induce clear b3-AR activation (stimulation of lipolysis, inhibition of glucose transport), while unable to promote a2-AR-mediated responses (antilipolysis). In this work, we did not studied the action of octopamine at a1-ARs but the fact that, in rat adipocytes, noradrenaline can stimulate glucose transport by an a1-AR-mediated mechanism (Faintrenie et al., 1998) whilst octopamine alone does not stimulate basal glucose uptake in our studies (not shown) suggests that octopamine is not able to efficiently activate adipocyte a1-ARs. Octopamine not only mimics the catecholamine effects on several, but not all, AR subtpes, it also shares similar pathways of degradation. We report here that octopamine is readily oxidized by all the amine oxidases present on adipocytes: MAO and SSAO (Raimondi et al., 1992; Marti et al., 1998). Tyramine is also a substrate for amine oxidation but, in turn, is not lipolytic at all, even at 1 mM, in the adipocyte models studied so far. Thus, agonism at octopamine/tyramine receptors in invertebrates is not predictive for agonism at b-ARs in mammals. Indeed, both tyramine and octopamine were found to be widely distributed in invertebrates and vertebrates and should therefore be considered as biogenic amines (Saavedra et al., 1989). In invertebrates, it is now accepted that octopamine plays a role of its own as a neuromodulator or neurotransmitter. Several well-char-

42

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

acterized octopamine systems have been described in invertebrates on the basis of physiological, pharmacological and molecular approaches (Evans et al., 1993). Octopamine/tyramine receptors appear to be present primarily in invertebrates and many of octopamine’s effects are mediated through an increase in intracellular calcium levels or adenylyl cyclase regulation (Roeder, 1995; Roeder et al., 1995). In vertebrates, octopamine is partially located in sympathetic nerves and is present in the brain; suggesting that it could play a role in sympathetic neurotransmission (Saavedra et al., 1989). Octopamine, taken up and stored in adrenergic vesicles and released upon nerve stimulation (Trendelenburg et al., 1972), was found to be two orders of magnitude less potent than noradrenaline in various organs but no specific receptor and no specific pharmacological agonists or blockers for octopamine have been described so far in mammalian systems (Saavedra et al., 1989). The endogenous concentration of octopamine in both the peripheral tissues and the brain of mammals is low but its turnover rate (which is a more significant index of the importance of pharmacologically active compounds) is considerably faster than that reported for noradrenaline (Saavedra et al., 1989). Octopamine synthesis from tyramine can be enhanced when MAO is inhibited while dopamine b-hydroxylase is still active. Increased octopamine levels are also observed in phenylketonuria and hepatic encephalopathy (Saavedra et al., 1989). Whether the neurological symptoms of these diseases may result from octopamine accumulation in central and peripheral sympathetic nerves and its subsequent action on b3-ARs remains to be demonstrated. In our opinion, another physiological situation may deserve more interest for the search of octopamine role in mammals: the foetal life. Studies on embryogenesis in rodents have shown that at day 16–17 of development a peak of phenylethanolamine and octopamine appears, at a time where noradrenaline is still undetectable. Then, there is a progressive decrease of the octopamine/noradrenaline ratio in brain until birth. The decrease in octopamine levels observed in vivo coincides with the appearance of tyrosine hydroxylase and monoamine oxidase activities (David et al., 1984; Saavedra et al., 1989). No data are available concerning this aspect in peripheral tissues of vertebrates but it is well known

that during the foetal and perinatal periods, mammals do not possess white adipose tissue. However, in all species, brown adipose tissue is present very early in embryos and participates in thermoregulatory processes in newborns. Noteworthy, brown adipose tissue is one of the richest tissues in b3-ARs (Krief et al., 1993; Lafontan and Berlan, 1993). Since we recently reported that rat brown adipocytes respond to octopamine by increased oxygen comsumption and lipolysis (Carpe´ne´ et al., 1999), we can assume that octopamine acts as the main neurotransmitter able to stimulate b3-ARs in foetal brown adipose tissue, before the production of the enzymes leading to the synthesis of noradrenaline has matured. Further studies are necessary to test such an hypothesis and assess a special role for octopamine in vertebrate ontogeny, corresponding to an evolutionnary remnant from remotely phylogenic ancestral species. In conclusion, octopamine, an amine that naturally occurs in central and peripheral sympathetic nervous system, acts as a specific agonist for the b3-ARs without noticeable agonism at b1-, b2- or a2-ARs, at least in white adipose tissues of various mammals. The physiological conditions in which mammalian b3-ARs can be selectively stimulated by octopamine, an endogenous amine which is, like adrenaline and noradrenaline, released upon nerve stimulation and sensitive to the scavenger action of amine oxidases, remain to be explored.

Acknowledgements We are grateful to F. Bousebaı¨ne, S. Vila, F. Leger and M. Jousseaume for their assistance, to L. Ambid for the access to hibernator adipose tissues, to J. Galitzky and M. Berlan for helpful discussions. Contribution of E. Fontana was partly supported by ‘Accords INSERM-CSIC’, ‘Actions inte´gre´es PICASSO’ and ‘Communaute´ de Travail des Pyre´ne´es’.

References Airriess, C.N., Rudling, J.E., Midgley, J.M., Evans, P.D., 1997. Selective inhibition of adenylyl cyclase by octopamine via human cloned a2A-adrenoceptor. Br. J. Pharmacol. 122, 191 – 198.

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

Atgie´, C., Le Gouic, S., Marti, L., Hanoun, N., Casteilla, L., Pe´nicaud, L., Ambid, L., Carpe´ne´, C., 1998. Lipolytic and antilipolytic responses of the siberian hamster (Phodopus sungorus sungorus) white adipocytes after weight loss induced by short photoperiod exposure. Comp. Biochem. Physiol. 119A, 503–510. Becu-Villalobos, D., Thyssen, S.M., Rey, E.B., LuxLantos, V., Libertun, C., 1992. Octopamine and phenylephrine inhibit prolactin secretion both in vivo and vitro. Proc. Soc. Exp. Biol. Med. 199, 230–235. Bloom, J.D., Dutia, M.D., Johnson, B.D., Wissner, A., Burns, M.G., Largis, E.E., Dolan, J.A., Claus, T.H., 1992. Disodium (R,R)-5-[2-[[2-(3-chlorophenyl)-2hydroxyethyl]-amino]propyl]-1,3 benzo dioxole-2,2dicarboxylate (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. Carpe´ne´, C., Chalaux, E., Lizarbe, M., Estrada, A., Mora, C., Palacin, M., Zorzano, A., Lafontan, M., Testar, X., 1993. b3-adrenergic receptors are responsible for the adrenergic inhibition of insulin-stimulated glucose transport in rat adipocytes. Biochem. J. 296, 99–105. Carpe´ne´, C., Ambid, L., Lafontan, M., 1994. Predominence of b3-adrenergic component in catecholamine activation of lipolysis in garden dormouse adipocytes. Am. J. Physiol. 266, R896–R904. Carpe´ne´, C., Marti, L., Hudson, A., Lafontan, M., 1995. Nonadrenergic imidazoline binding sites and amine oxidase activitis in fat cells. Ann. N.Y. Acad. Sci. 763, 380–397. Carpe´ne´, C., Bousquet-Me´lou, A., Galitzky, J., Berlan, M., Lafontan, M., 1998. Lipolytic effect of b3-adrenergic agonists in white adipose tissue of mammals. Ann. N.Y. Acad. Sci. 839, 186–189. Carpe´ne´, C., Galitzky, J., Fontana, E., Atgie´, C., Lafontan, M., Berlan, M., 1999. Selective activation of b3-adrenoceptors by octopamine: comparative studies in mammalian fat cells. Naunyn Schmiedeberg’s Arch. Pharmacol. 359, 310–321. Castan, I., Devedjian, J.C., Valet, P., Paris, H., Lafontan, M., 1995. Human adipocytes express a2adrenergic receptor of the a2A-subtype only: pharmacological and genetic evidence. Fund. Clin. Pharmacol. 9, 569–575. David, J.C., 1984. Relationship between phenolamines and catecholamines during rat brain embryogenic development in vivo and in vitro. J. Neurochem. 43, 668–674. Evans, P.D., Robb, S., 1993. Octopamine receptor subtypes and their mode of action. Neurochem. Res. 18, 869–874.

43

Faintrenie, G., Geloe¨n, A., 1998. Alpha1-adrenergic stimulation of glucose uptake in rat white adipocytes. J. Pharmacol. Exp. Ther. 286, 607 – 610. Galitzky, J., Carpe´ne´, C., Lafontan, M., Berlan, M., 1993. Stimulation spe´cifique des re´cepteurs b3-adre´nergiques du tissu adipeux par l’octopamine. C.R. Acad. Sci. 316, 519 – 523. Krief, S., Lo¨nnqvist, F., Raimbault, S., Baude, B., VanSpronsen, A., Arner, P., Strosberg, A.D., Ricquier, D., Emorine, L., 1993. Tissue distribution of b3-adrenergic 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. Langin, D., Portillo, M., Saulnier-Blache, J.S., Lafontan, M., 1991. Coexistence of three b-adrenoceptor subtypes in white fat cells of various mammalian species. Eur. J. Pharmacol. 199, 291 – 301. Lee, T.L., Hsu, C.T., Yen, S.T., Lai, C.W., Cheng, J.T., 1998. Activation of beta3-adrenoceptors by exogenous dopamine to lower glucose uptake into rat adipocytes. J. Auton. Nerv. Syst. 74, 86 – 90. Li, G., Regunathan, S., Barrow, C.J., Eshraghi, J., Cooper, R., Reis, D.J., 1994. Agmatine: an endogenous clonidine-displacing substance in the brain. Science 263, 966 – 969. Lyman, C.P., 1982. Who is who among the hibernators. In: Lyman, C.P., Willis, J.S., Malan, A., Wang, L.C.H. (Eds.), Hibernation and Torpor in Mammals and Birds, pp. 12 – 36. Marti, L., Morin, N., Enrique-Tarancon, G., Pre´vot, D., Lafontan, M., Testar, X., Zorzano, A., Carpe´ne´, C., 1998. Tyramine and vanadate synergistically stimulate glucose transport in rat adipocytes by amine oxidase-dependent generation of hydrogen peroxide. J. Pharmacol. Exp. Ther. 285, 342 – 349. Okina, A., Abe, K., Inuzuka, T., Yano, T., Okina, T., Nakashima, T., Nishiura, T., 1992. The effects of m-octopamine on salivary flow rates and protein secretion by rat submandibular glands. Comp. Biochem. Physiol. 103C, 469 – 476. Osborne, R.H., 1996. Insect neurotransmission: neurotransmitters and their receptors. Pharmacol. Ther. 69, 117 – 142. Pinthong, D., Wright, I.K., Hanmer, C., Millns, P., Mason, R., Kendall, D.A., Wilson, V.G., 1995. Agmatine recognizes alpha2-adrenoceptor binding sites but neither activates nor inhibits alpha2-adrenoceptors. Naunyn-Schmiedeberg’s Arch. Pharmacol. 351, 10 – 16. Raimondi, L., Pirisino, R., Banchelli, G., Ignesti, G., Conforti, L., Romanelli, E., Buffoni, F., 1992. Further studies on semicarbazide-sensitive amine oxidase activities (SSAO) of white adipose tissue. Comp. Biochem. Physiol. 102B, 953 – 960.

44

E. Fontana et al. / Comparati6e Biochemistry and Physiology, Part C 125 (2000) 33–44

Robb, S., Cheek, T.R., Hannan, F.L., Hall, L.M., Midgley, J.M., Evans, P.D., 1994. Agonist-specific coupling of a cloned Drosophila octopamine/tyramine receptor to multiple second messenger systems. EMBO J. 13, 1325–1330. Roeder, T., Degen, J., Dyczkowski, C., Gewecke, M, 1995. Pharmacology and molecular biology of octopamine receptors from different insect species. Prog. Brain Res. 106, 249–258. Roeder, T., 1995. Pharmacology of the octopamine receptor from locust central nervous tissue (OAR3). Br. J. Pharmacol. 210–216, 114. Saavedra, J.M., 1989. b-phenylethylamine, phenylethanolamine, tyramine and octopamine. In: Trendelenburg, U., Weiner, N. (Eds.), Catecholamines II, pp. 181–210. Tavernier, G., Galitzky, J., Valet, P., Remaury, A., Bouloumie´, A., Lafontan, M., Langin, D., 1995. Molecular mechanisms underlying regional variations of catecholamine-induced lipolysis in rat adipocytes. Am. J. Physiol. 268, E1135–E1142.

.

Tong, Y.C., Cheng, J.T., 1997. Subtyping of alpha1adrenoceptors responsible for the contractile response in the rat corpus carvenosum. Neurosci. Lett. 228, 159 – 162. Trendelenburg, U., 1972. Classification of sympathomimetic amines. In: Blaschko, H., Muscholl, E. (Eds.), Catecholamines. Handbook of Experimental Pharmacology, 33, pp. 336 – 362. Umekawa, T., Yoshida, T., Sakane, N., Kondo, M., 1996. Effect of CL316,243, a highly specific beta(3)adrenoceptor agonist, on lipolysis of human and rat adipocytes. Horm. Metab. Res. 28, 394 – 396. Yen, S.T., Li, M.H., Hsu, C.T., Lee, T.L., Cheng, J.T., 1998. Stimulatory effect of octopamine on beta3adrenoceptors to lower the uptake of [14C]-deoxy-Dglucose into rat adipocytes in vitro. J. Auton. Pharmacol. 18, 13 – 19. Yu, P.H., 1986. Monoamine oxidase. In: Boulton, A.A., Baker, G.B., Yu, P.H., (Ed.) Neuromethods, Neurotransmitter Enzymes 5, 235 – 272.