LNX used in the present study was obtained from Daehan Cell Pharm, Inc. (Guri, Republic of Korea). Briefly, the original seeds of 100 g nutmeg (Myristica fragrans) were pulverized and dissolved in 500 ml of 80% ethanol (EtOH). The optimal extraction conditions were determined by monitoring the minimal toxicity of myristicin and maximal bioactivity of nectandrin B. The extract was subjected to a mobile phase of the methanol (MeOH)-H₂O (0-32 min, 63% MeOH; 32-37 min, 63-100% MeOH) mixture. The final extract was then adsorbed through Diaion HP-20 (Merck KGaA), one of the ion exchange resin, and eluted in a serial concentration of 30-90% EtOH to monitor the contents of myristicin and nectandrin B. The nutmeg alcohol extract was detected at 234 nm at a flow rate of 1.5 ml/min using a Waters HPLC Optima Pak C₁₈ reverse-phase column (cat. no. OP C18-51002546; 4.6x250 mm, 5 µm pore size, RS Tech. Corp.). The mobile phase in the analytical column was composed of the mixture (the ratio of 4:6; absolute acetonitrile and 0.1% formic acid). In addition, nectandrin B concentration used in the present study is 4.2 from 91% of dry weight of LNX, which is lower than the concentration used in the previous study (28) with <0.5% of myristicin in a 91% of dry weight of LNX.

Primary antibodies against PPAR-γ (cat. no. sc-7272), C/EBP-α (cat. no. sc-61), phosphorylated (p)-STAT3 (cat. no. sc-8059) and total (T)-STAT3 (cat. no. sc-8019) were purchased from Santa Cruz Biotechnology, Inc. The primary antibody against FAS (cat. no. 9452) was obtained from BD Biosciences. Primary antibodies against p-HSL (S563) (cat. no. 4139) and p-HSL (S565) (cat. no. 4137) were acquired from Cell Signaling Technology, Inc. Primary antibodies against COX-2 (cat. no. 160106) and T-HSL (cat. no. 10006371) were purchased from Cayman Chemical Company. Primary antibody against perilipin A (cat. no. 3948-200) was purchased from BioVision, Inc. The primary antibody of β-actin (A5441), iso-butylmethylxanthine (IBMX), dexamethasone, insulin, Oil Red O solution, Isoproterenol, and Free Glycerol Reagent were obtained from MilliporeSigma. Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), and penicillin/streptomycin were purchased from Welgene, Inc. Fetal calf serum (FCS) and trypan blue dye were purchased from Gibco; Thermo Fisher Scientific, Inc. Enhanced chemiluminescence (ECL) reagents were purchased from Advansta Inc.

3T3-L1 mouse white preadipocytes (American Type Culture Collection) were cultured in DMEM containing 10% heat-inactivated FBS and 1% penicillin/streptomycin at 37^˚C in a humidified atmosphere containing 5% CO₂. 3T3-L1 preadipocytes were seeded with DMEM containing 10% FCS and 1% penicillin/streptomycin mixture and maintained up to the contact inhibition stage for 2 days. The differentiation of 3T3-L1 preadipocytes was initiated by replacing the media with new DMEM containing 10% heat-inactivated FBS plus a cocktail of hormones (MDI): 0.5 mM IBMX (M), 0.5 µM dexamethasone (D) and 5 µg/ml insulin (I) either with or without LNX at the designated concentrations for 2 days (Fig. 1A). After 2 days, the first differentiation medium was removed from the cells and the cells were subsequently cultured in new DMEM supplemented with 10% FBS and 5 µg/ml insulin either with or without LNX at the designated concentrations for an additional 3 days. After 3 days, the second differentiation medium was removed from the cells and the cells were further supplemented with new DMEM containing 10% FBS with or without LNX at the designated concentrations for another 3 days.

On day 8 post the differentiation induction, the control or LNX-treated 3T3-L1 cells underwent phosphate-buffered saline (PBS) washing and 10% formaldehyde fixation for 2 h at room temperature (RT). Following a 60% isopropanol washing, the cells were thoroughly dried. The fixed cells were treated with Oil Red O working solution for 1 h at RT, followed by washing with distilled water. Light microscopy was used to identify the lipids or fats that accumulated in the conditioned cells (Nikon Corporation).

On day 8 post differentiation induction, control and LNX-treated 3T3-L1 cells were stained with 0.4% trypan blue dye at RT. Only cells with intact membranes could effectively block the dye. Once dead cells with damaged membranes were exposed to the dye, they were stained and counted under a light microscope.

At the designated time, the control or LNX-treated 3T3-L1 cells were washed with PBS and lysed in a modified RIPA buffer [50 mM Tris-Cl (pH 7.4), 150 mM NaCl, 0.1% SDS, 0.25% sodium deoxycholate, 1% Triton X-100, 1% Nonidet P-40, 1 mM EDTA, 1 mM EGTA, proteinase inhibitor cocktail (1X)]. The whole-cell lysates were collected and centrifuged at 14,000 x g for 15 min at 4°C. The supernatant was collected and protein concentrations were determined using bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific, Inc.).

An aliquot of protein (40 µg per lane) was separated using 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to nitrocellulose membranes (MilliporeSigma). The membranes were washed with Tris-buffered saline (10 mM Tris, 150 mM NaCl) supplemented with 0.05% (v/v) Tween 20 (TBST) and blocked with blocking buffer (TBST containing 5% (w/v) non-fat dried milk) at 4˚C overnight. The membranes were incubated overnight with corresponding primary antibodies for C/EBP-α (1:2,000), PPAR-γ (1:2,000), p-STAT3 (1:2,000), STAT3 (1:2,000), phosphorylated (p)-STAT-5 (1:2,000), STAT-5 (1:2,000), FAS (1:1,000), perilipin A (1:2,000), p-HSL (S563) (1:2,000), p-HSL (S565) (1:2,000), T-HSL (1:2,000), COX-2 (1:2,000) or β-actin (1:10,000) at 4˚C. The membranes were then washed with TBST and incubated with a horseradish peroxidase-conjugated secondary antibody with either anti-mouse IgG (115-035-062; 1:5,000) or anti-rabbit IgG (111-035-045; 1:5,000) (Jackson ImmunoResearch Laboratories, Inc.) for 2 h at RT. The membranes were washed with TBST. ECL reagents were used to develop the images. Equal protein loading per lane was quantified based on the relative intensity of β-actin, an internal control protein. ImageJ software (v.1.6.0.24; National Institutes of Health) was used to intensify the bands for proteins that were standardized to the internal control.

Total RNA was isolated from the control and LNX-treated 3T3-L1 cells after 4 h using RNAiso Plus (Takara Bio, Inc.). A total of three micrograms of total RNA were used to prepare complementary DNA using random hexadeoxynucleotide primers, reverse transcriptase (M-MLV RT), reverse transcriptase buffer (M-MLV RT 5X Buffer) and dNTPs (Promega Corporation). The cDNA was amplified by PCR with the following primers (Bioneer Corporation): COX-2 forward, 5'-TTGAAGACCAGGAGTACAGC-3' and reverse, 5'-GGTACAGTTCCATGACATCG-3'; and β-actin forward, 5'-TCATGAAGTGTGACGTTGACATCCGT-3' and reverse, 5'-CCTAGAAGCACTTGCGGTGCACGATG-3'. The PCR conditions applied for COX-2 were as follows; 30 cycles of denaturation at 95˚C for 30 sec, annealing at 62˚C for 30 sec and extension at 72˚C for 30 sec and for β-actin; 30 cycles of denaturation at 95˚C for 30 sec, annealing at 63˚C for 30 sec and extension at 72˚C for 1 min. Levels of β-actin mRNA expression were used as an internal control.

On day 8 of the post-differentiation induction, 3T3-L1 cells were serum-starved for 2 h and further treated without or with LNX (6 µg/ml) or isoproterenol (ISO; 20 µM), a known lipolysis inducer for another 3 h. The culture medium from the control, LNX- or ISO-treated cells was stored and the glycerol content in the respective culture medium was measured using a Free Glycerol Reagent according to the manufacturer's instructions. The absorbance was measured at 540 nm using a microplate reader.

Cell counting, western blotting and RT-PCR analysis were conducted in triplicates and repeated two times. The mean and SD values were used to express the data. One-way ANOVA was employed to compare the significance of the differences, followed by a Bonferroni test for the post hoc analysis. P<0.05 was considered to indicate a statistically significant difference. The statistical software used in the present study was IBM SPSS Statistics 25 software (IBM Corp.).

Initially, in the preliminary studies, LNX was tested at higher concentrations such as 50, 100 and 200 µg/ml for 3T3-L1 preadipocyte differentiation (data not shown). However, LNX treatment at 100 and 200 µg/ml had been demonstrating cytotoxicity in differentiating 3T3-L1 cells. Hence, LNX was tested in lower concentrations as revealed in the present study (6.25, 12.5 and 25 µg/ml) to further check the deposition rate of LDs during 3T3-L1 differentiation. Of interest, it was found that the accumulation of LDs appeared to be increased in the higher concentration of LNX in a dose-dependent manner (data not shown). Therefore, since the LNX effects on the inhibition of lipid accumulation during 3T3-L1 preadipocyte differentiation using Oil Red O staining had to be assessed, various concentrations of LNX (1.56, 3.12 and 6.25 µg/ml) have been selected for the present study. The timetable for 3T3-L1 pre-adipocyte differentiation used in the present study is depicted in Fig. 1A. As demonstrated in Fig. 1B, Oil Red O staining revealed markedly higher lipid accumulation in differentiated 3T3-L1 cells (D8) than in undifferentiated cells (D0). Notably, compared with mock-treated cells, treatment with LNX led to a dose-dependent inhibition of lipid accumulation in differentiated 3T3-L1 cells. The ability of LNX to downregulate lipid accumulation in differentiated 3T3-L1 cells was confirmed using phase-contrast microscopy (Fig. 1C). Additionally, the effects of 1.56, 3.12 or 6.25 µg/ml LNX on cell growth during the differentiation of 3T3-L1 preadipocytes into adipocytes, were examined by using cell count analysis. As illustrated in Fig. 1D, treatment with LNX at the tested doses did not significantly affect cell growth during 3T3-L1 preadipocyte differentiation. Quantification of the Oil Red O staining images to determine the LDs percentage using ImageJ software also revealed that LNX reduced lipid accumulation in a dose-dependent manner (Fig. 1E). Hence, based on the maximal lipid-lowering effect on differentiating 3T3-L1 preadipocytes with no cytotoxicity, the concentration of 6 µg/ml of LNX was selected for further experiments.

Subsequently, to explore the fundamental molecular and cellular mechanisms underlying this LNX-induced anti-adipogenic effect, LNX (6 µg/ml) was investigated on the expression and phosphorylation levels of major adipogenic transcription factors, including C/EBP-α, PPAR-γ and STAT3/5, using immunoblotting analysis. As revealed in Fig. 2A, LNX treatment did not modulate significantly the expression levels of C/EBP-α and PPAR-γ proteins in 3T3-L1 preadipocyte differentiation at the assessed timepoints; however, it led to the slight elevation of the expression levels of C/EBP-α and PPAR-γ on D5 and D8, respectively. In addition, C/EBP-β was also assessed during 3T3-L1 differentiation. However, a band for this protein could not be obtained, which was possibly due to an antibody error.

In the present study, the day courses such as D2, D5 and D8 in 3T3-L1 preadipocyte differentiation were observed to determine the p/T-STAT3 and FAS protein expression (Fig. 2A and B). It was revealed that STAT3 was highly phosphorylated during 3T3-L1 preadipocyte differentiation. Of note, p-STAT3 expression level was decreased on D5. In addition, on D8, STAT3 protein levels were strongly phosphorylated again during differentiation period (as compared with the control group of D5). Notably, the mid-phase of STAT3 decline affected the adipocyte differentiation. Meanwhile, FAS protein was greatly attenuated on D2 compared with D5 or D8. Hence, based on these day course experiment results, different day courses were investigated on each target (D8 for STAT3 and D2 for FAS protein expression levels) (Fig. 2C-F).

Interestingly, while treatment with LNX had no effect on the phosphorylation and total expression levels of STAT3 on D2 of 3T3-L1 cell differentiation, it markedly reduced the phosphorylation of STAT3 without altering its total expression levels on D5 of the cell differentiation. However, LNX treatment further reduced the phosphorylation and total expression levels of STAT3 on D8 of cell differentiation. Moreover, LNX treatment slightly inhibited the expression of FAS on D2, D5 and D8, during 3T3-L1 preadipocyte differentiation. LNX treatment did not influence perilipin A expression during cell differentiation. Expression levels of the control actin protein remained constant under these experimental conditions. The densitometric data from Fig. 2A are demonstrated in Fig. 2B. The results of the triplicate experiments, as demonstrated in Fig. 2C and D, revealed the capability of LNX (6 µg/ml) to vastly inhibit STAT3 phosphorylation and FAS expression on D8 and D2 of 3T3-L1 preadipocyte differentiation, respectively. The densitometric data from Fig. 2C and D are shown in Fig. 2E and F, respectively.

Subsequently, it was explored whether treatment with LNX (6 µg/ml) modulated lipolysis in differentiated 3T3-L1 adipocytes by measuring the levels of glycerol content and HSL phosphorylation, two hallmarks of lipolysis. The experimental scheme and timescale for measuring glycerol content and HSL phosphorylation are exhibited in Fig. 3A. To this end, differentiated 3T3-L1 adipocytes on D8 were serum-starved (0% FBS) for 2 h and incubated in a fresh culture media containing 10% FBS in the absence or presence of LNX (6 µg/ml) or ISO (20 µM), a known inducer of lipolysis (31), for an additional 3 h. Culture media and whole-cell lysates from each condition were used to measure the two lipolysis hallmarks aforementioned. As illustrated in Fig. 3B, ISO treatment for 3 h significantly enhanced glycerol release from differentiated 3T3-L1 adipocytes, whereas LNX treatment did not. In addition, while ISO treatment for 3 h significantly elevated p-HSL levels at residues S563 and S565 in differentiated 3T3-L1 adipocytes, LNX treatment had no stimulatory effect. The total HSL protein expression levels remained unchanged under these experimental conditions. The densitometric data from Fig. 3C are demonstrated in Fig. 3D.

Next, it was further explored whether TNF-α at 10 ng/ml induces COX-2 expression in 3T3-L1 preadipocytes and whether treatment of LNX at different concentrations (1.5, 3 or 6 µg/ml) interferes with it by using RT-PCR analysis. As expected, the exposure of TNF-α for 4 h greatly induced COX-2 mRNA expression in 3T3-L1 preadipocytes. However, LNX treatment at the examined concentrations did not significantly influence cytokine-induced COX-2 mRNA expression in these cells (Fig. S1). Using immunoblotting analysis (Fig. 4A), it was investigated whether LNX treatment at higher doses (10, 25 and 50 µg/ml) could modulate cytokine-induced COX-2 expression in 3T3-L1 preadipocytes. As revealed in Fig. 4, while treatment with TNF-α for 4 h markedly enhanced COX-2 protein expression in 3T3-L1 preadipocytes, LNX treatment at such high doses tested did not interfere with the TNF-α-induced COX-2 protein expression in 3T3-L1 preadipocytes. Expression levels of the control actin protein remained constant under these experimental conditions. The densitometric data from Fig. 4A are illustrated in Fig. 4B.