212 JOURNAL OF COSMETIC SCIENCE ISOLATION OF HUMAN ADIPOCYTES Adipose tissue samples from abdominal fat deposits were obtained after plastic surgery. Tissue samples were rinsed in physiological medium to remove adhering blood. Visible fibrous material and blood vessels were carefully dissected. The remaining adipose tissue was cut into small fragments with a scalpel and then incubated in the dissociation medium [MEM medium without phenol red added, with 50 IU/ml of penicillin, 50 pg/ml of streptomycin, and 0.1% (w/v) collagenase], at 37øC under slow shaking for 60 min. The ratio of adipose tissue mass to digestion solution was approximately 4 g for 10 mi. The dispersed tissue was then filtered onto a 250-pm nylon filter to discard non- digested pieces. The suspension was allowed to stand for 10 rain at room temperature to allow adipocytes to float. The medium under the cells was removed and replaced twice by the incubation medium [MEM medium without phenol red added, with 50 IU/ml of penicillin, 50 pg/ml of streptomycin, and 0.5% (w/v) lipid-free bovine serum albu- min]. The integrity of the adipocytes was checked by microscopic observation. cAMP AND NEFA MEASUREMENTS Adipocyte suspensions were incubated 2 hr at 37øC in a water bath in the absence (control) or in the presence of increasing concentrations of SLC. At the end of the incubation period, cell lysates were obtained by sonication and the intra-adipocyte cAMP contents were quantified using a sensitive and specific enzyme immunoassay (EIA) (Amersham, Saclay, France). The NEFA were measured in the incubation medium using a colorimetric enzymatic test (Wako Chemicals, Japan). BINDING EXPERIMENTS Human ot2-adrenoreceptor binding assay. Competition binding experiments were performed as described by Bylund et al. (16), with slight modifications. Briefly, membranes from HT-29 human adrenocarcinoma cells (40 lag protein) were incubated for 20 min at 22øC with 0.7 nM of [3H]RX821002 in the presence of increasing concentrations of SLC in 250 lal of assay buffer containing 1 mM of EDTA, 2 mM of MgC12, and 50 mM of Tris-HC1 (pH 7.4). The reaction was terminated by rapid filtration under vacuum through glass fiber filters (Filtermat A, Wallac), using a 48-sample cell harvester (Much IV, Tomtec). The filters were cooled with 50 mM of ice-cold Tris-HC1 (pH 7.4), dried, and counted for radioactivity in a scintillation counter (1204 Betaplate, Wallac), using a solid scintillant (MeltiLex B/HS, Wallac). Non-specific binding was determined in the presence of 100 laM of (-)epinephrine. The specific binding was defined as the difference between total and non-specific binding. Assays were performed in duplicate and the results expressed as a percent of the control specific binding (mean values). IC50 values and Hill slopes (nil) were calculated by computer-assisted nonlinear regression analysis of the competition curves. ISOLATED ORGAN EXPERIMENTS--o•2-ADRENORECEPTOR ANTAGONISM In vitro ot2-adrenoreceptor bioassay. The bioassay was performed using the rabbit isolated ear vein as described by Duly et al. (17). Male New Zealand white rabbits weighing

LIPOLYTIC EFFECT OF SLIMMING LIPOSOMES 213 about 3 kg (Charles River, St. Aubin les Elbeuf, France) were sacrificed, and the main ear veins were removed quickly, cleaned of connective tissue, and cut into rings 3 mm in length. These preparations were suspended in a 20-ml organ bath containing Krebs solution of the following composition (in mM): NaC1, 118.0 KC1, 4.7 MgSO4, 1.2 CaC12, 2.5 KH2PO 4, 1.2 NaHCO3, 25.0, and glucose 11.0 (pH 7.4). It also contained 1 pM of prazosin, 1 pM of propranolol, and 1 pM of indomethacin to block o•- and [3-adrenergic receptor activation and prostanoid release, respectively. This solution was continuously aerated with a mixture of 95% 0 2 and 5% CO 2 and maintained at 37øC. The tissues were connected to isometric force transducers (UF1, Phymep, Paris, France) for tension recordings, stretched to a resting tension of 1 g, and allowed to equilibrate for 60 min. During this time, they were washed repeatedly and the resting tension was readjusted at 15-min intervals. For the evaluation of o•2-receptor antagonistic activity, the tissues were exposed to a submaximal concentration of clonidine (0.1 pM) to obtain a sustained contraction. Thereafter, they were exposed to cumulative increasing concentrations of SLC. Each concentration was added after the response to the previous one had stabilized, and the resulting change in the clonidine-induced contraction was measured. SLC was thereby investigated on two tissues and the results expressed as a percent of the clonidine- induced contraction (mean values). IN VIVO EXPERIMENTS Eighteen healthy human volunteers were included in this study. All the volunteers applied daily a 3% SLC formulation on one of their thighs their other thigh served as a control. Before and at the end of the study, twenty-eight days later, the volunteers' thigh circumferences were measured. STATISTICAL ANALYSIS Results were presented as means _+ standard error (SE). The level of significance was assessed using a one-way analysis of variance (ANOVA 1), followed by a Dunnett's test. RESULTS AND DISCUSSION The product we have developed consists of a liposome containing esculoside, Centella asiatica triterpenes, caffeine, and L-carnitine. The liposome form allows the enhancement of the penetration of active compounds through the skin (18). In order to assess the magnitude or the SLC lipolytic effect, we first conducted some in vitro experiments on primary normal human adipocytes in suspension. The capability of SLC to induce lipolysis in this model was evaluated by measuring NEFA liberation from adipocytes. As shown in Figure 1, SLC induced a dose-related increase in the NEFA content of the adipocyte incubation media. The SLC effect was detectable for a concen- tration as low as 0.1% (v/v) and reached a maximum at 1% (v/v). To decipher the intracellular signaling pathway underlying this lipolytic effect and because the main intracellular messenger regulating lipolysis is cAMP, we measured