Human adipose-derived mesenchymal stem cells (ADSCs) were obtained from lipoaspirates provided by the Biobank, Stem Cell Institute (University of Science, VNU-HCMC, VN). Cells were cultured in ADSCCult I medium (Regenmedlab, VN) and maintained at 37°C with 5% CO₂. Routine passaging involved enzymatic detachment using Deattachment reagent, centrifugation (500 × g, 5 min), and reseeding in T25/T75 flasks. For cryopreservation, cells were suspended in CryoSave I medium (Regenmedlab, VN), cooled sequentially (4°C for 30 min, −20°C for 1–2 hr), and stored at −86°C. Thawing utilized rapid rewarming (37°C water bath), dilution in Thawbest medium (Regenmedlab, VN), and centrifugation (300 × g, 5 min). ADSC identity was validated via flow cytometry (BD FACSCalibur, US) confirming CD44, CD73, CD90, and CD105 positivity (>95%) and CD14, CD34, CD45, and HLA-DR negativity (<2%). Trilineage differentiation potential was assessed using commercial kits: adipogenesis (Oil Red O staining, day 7), osteogenesis (Alizarin Red S staining, day 21), and chondrogenesis (Alcian Blue staining, day 14) All chemicals and media were purchased from Regenmedlab (Ho Chi Minh City, Viet Nam).

Differentiation was induced using a two-phase protocol. Cells were pre-treated for 3 days with StemPro Adipogenesis Differentiation Kit to initiate adipogenesis (Gibco, ThermoFisher scientific, US). Subsequently, 12 experimental groups were exposed to combinatorial treatments of rosiglitazone (0, 100, 500, or 1000 nM; BioGems International, CA) and triiodothyronine (T3; 0, 0.2, or 1 nM; Sigma-Aldrich, Louis St, MO) for 21 days. Morphological changes were tracked via Oil Red O (Sigma-Aldrich, Louis St, MO) staining on days 3, 7, 14, and 21 using a Zeiss inverted microscope.

Table 1

Combinations	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
Rosiglitazone (nM)	-	100	500	1000	-	100	500	1000	-	100	500	1000
T3 (nM)	-	-	-	-	0.2	0.2	0.2	0.2	1.0	1.0	1.0	1.0

Total RNA was extracted using the easy-BLUE Total RNA Kit (iNtRON Biotech, Korea) with chloroform/isopropanol purification. Gene expression of UCP1, PPARγ, and PGC1α (normalized to GAPDH) was quantified via one-step RT-qPCR (Luna Universal Kit; New England Biolabs, MA) on Eppendorf Realplex cycler (Eppendorf, Germany). Primer sequences were as follows: UCP1: F 5′-AGTTCCTCACCGCAGGGAAAGA-3′, R 5′-GTAGCGAGGTTTGATTCCGTGG-3′; PPARγ: F 5′-AGCCTGCGAAAGCCTTTTGGTG-3′, R 5′-GGCTTCACATTCAGCAAACCTGG-3′; PGC1α: F 5′-CCAAAGGATGCGCTCTCGTTCA-3′, R 5′-CGGTGTCTGTAGTGGCTTGACT-3′.

Functional lipolysis was assessed by measuring glycerol release after 18 hr isoproterenol stimulation (100 nM; Lipolysis Assay Kit, Abcam), with absorbance read at 570 nm (Beckman Coulter DTX 880 plate reader).

All procedures were performed in accordance with institutional guidelines and approved by the institutional ethics committee. Male BALB/c mice (6 weeks old; n = 30) were acclimatized for 2 weeks before being divided into two dietary groups: a control diet comprising standard chow (18% protein, 5% lipid) and a high-energy diet consisting of 70% standard chow plus 30% lipid-rich seeds (sunflower/peanut/macadamia).

After following their respective diets for 40 days, overweight mice were randomly assigned to one of three transplant groups (n = 5/group):

Cells were injected subcutaneously in the abdominal region using a 25–30° needle angle. Body weight (Fisher Scientific CLF201 scale), naso-anal length, and BMI [weight (g)/length² (cm²)] were monitored weekly. Thermogenesis was assessed via infrared thermometry (iMediCare iTM-9S) at abdominal and ear sites on days 0, 1, 2, 3, 7, 11, 14, 18, 21, 25, and 28.

Data are presented as mean ± SD. Differences between groups were analyzed using two-tailed Student’s t-tests or ANOVA in GraphPad Prism 8, with significance at p < 0.05.

Isolated adipose-derived mesenchymal stem cells (ADSCs) exhibited typical fibroblast-like morphology, adhering to plastic surfaces with elongated spindle shapes (Figure 1A). Flow cytometry confirmed expression of mesenchymal surface markers CD44 (98.65 ± 1.68%), CD73 (98.67 ± 1.04%), CD90 (99.68 ± 0.33%), and CD105 (99.11 ± 0.55%), while lacking hematopoietic markers CD14 (0.06 ± 0.05%), CD34 (0.04 ± 0.05%), CD45 (0%), and HLA-DR (0.003 ± 0.005%) (Figure 1B). Trilineage differentiation assays revealed robust multipotency: adipogenic induction produced lipid-laden cells staining positive with Oil Red O (Figure 1C), osteogenic differentiation induced mineralized matrix deposition (Alizarin Red S; Figure 1D), and chondrogenesis generated proteoglycan-rich aggregates (Alcian Blue; Figure 1E), confirming compliance with ISCT criteria.

Figure 1

Combinatorial treatment with rosiglitazone and T3 induced distinct morphological shifts. By day 7, cells exposed to rosiglitazone (≥ 500 nM) and T3 (≥ 0.2 nM) developed multilocular lipid droplets, contrasting with unilocular droplets in controls (rosiglitazone/T3-free). This phenotype intensified by day 21, where the 1000 nM rosiglitazone + 0.2 nM T3 group showed the highest proportion of cells with small, multilocular droplets (Figure 2). Time-lapse imaging revealed progressive lipid accumulation: nascent droplets appeared by day 3, expanded through day 14, and achieved mature morphology by day 21. Oil Red O staining confirmed lipid retention in all groups, but multilocularity was exclusive to rosiglitazone/T3-treated cells (Figure 3).

Figure 3

qPCR analysis demonstrated dose-dependent upregulation of brown adipocyte markers:

UCP1: The 1000 nM rosiglitazone + 0.2 nM T3 protocol maximally induced UCP1 (19-fold vs. control; p < 0.001) in donor 1 cells (Figure 4 A). While donor variability existed (e.g., 4.9-fold increase in donor 2 with 500 nM rosiglitazone + 0.2 nM T3), the 1000 nM rosiglitazone + 0.2 nM T3 combination consistently outperformed single-agent treatments across donors (p < 0.05).

PPARγ: Expression peaked at 7.48-fold (p < 0.001) under 1000 nM rosiglitazone + 0.2 nM T3 (donor 1), with significant synergy versus rosiglitazone alone (Figure 4 B).

PGC1α: The same protocol elevated PGC1α 3.84-fold (p < 0.05) in donor 1 cells, corroborated in donors 2 (3.62-fold) and 3 (3.32-fold) (Figure 4 C).

Figure 4

Differentiated brown adipocytes exhibited suppressed lipolysis, reflecting their metabolic reprogramming for lipid retention. The 1000 nM rosiglitazone + 0.2 nM T3 group released only 34.19 ± 0.17 nM glycerol versus 39.85 ± 0.52 nM in controls (p < 0.0001; Figure 5). This reduction was consistent across donors and correlated with multilocular morphology, confirming functional maturation.

Figure 5

Mice fed a high-energy diet for 40 days showed significant weight gain (31.58 ± 1.0 g (in model mice) vs. 26.12 ± 1.4 g in controls (normal mice); p < 0.0001) and elevated BMI (0.35 ± 0.015 (in model mice) vs. 0.29 ± 0.004 (normal mice); p < 0.0001), validating successful obesity induction (Figure 6 A,B).

Figure 6

The abdominal skin temperature measurements are shown in Figure 7A. Prior to transplantation, mice in all three groups had similar abdominal skin temperatures (36.62–36.68°C). Over the first three days post-transplantation, the body temperature of mice receiving either white adipose cells (WAC) or brown adipose cells (BAC) increased significantly. In the WAC group, temperature peaked at 37.83°C on Day 1 and reached a lower peak (37.59°C) on Day 3. On Day 7, the WAC group’s temperature dropped sharply (36.57°C), whereas the BAC group maintained a significantly higher value (37.51°C; p ≤ 0.001). From Day 11 onward, WAC-group temperatures stabilized between 36.63 and 36.72°C. BAC-group temperatures gradually decreased until Day 18 (36.77°C, p > 0.05), then remained stable. In the PBS-injected control group, temperatures rose during the first three days—though less markedly than in the cell-transplanted groups—and reached a maximum of 37.21°C on Day 1. From Day 7 onward, this group maintained abdominal skin temperatures between 36.62 and 36.75°C.

Figure 7

Figure 7B shows ear (pinna) temperature measurements. The PBS control group had a modest temperature rise on Day 1 (from 37.19 to 37.59°C), returning to baseline by Day 3 (37.19°C). In contrast, both cell-transplanted groups showed sustained elevated ear temperatures over the first three days (37.64–37.87°C). Thereafter, the WAC group’s temperature trended downward, eventually aligning statistically with the PBS group. On Day 7 (37.51°C; p ≤ 0.001) and Day 11 (37.53°C; p ≤ 0.01), the BAC group maintained significantly higher ear temperatures than the other two groups, but from Day 14 onward, BAC temperatures were no longer statistically different. In summary, both cell-transplanted groups initially exhibited elevated body temperature, most likely due to an immune response. After peaking around Day 3, body temperature generally declined by Day 7. However, the BAC-transplanted group continued to show a higher temperature, gradually tapering off around Day 14. A single dose of one million BAC cells can therefore elevate mouse body temperature for approximately two weeks, consistent with previous studies on brown adipose tissue transplantation in obese mice. Notably, ear temperature was consistently higher than abdominal temperature (e.g., on Day 0: 37.15 ± 0.01°C vs. 36.62 ± 0.03°C), indicating that BAC transplantation increases temperature both locally (in the abdominal area) and systemically (as reflected at the ear site).

Figure 8

Body weight changes are illustrated in Figure 8A. All groups experienced weight loss within the initial three days post-transplantation. While body weights in the PBS and WAC groups began recovering on Day 7, mice in the BAC group continued to lose weight. Statistically significant differences in body weight emerged beginning on Day 11 (BAC group: 30.4 ± 0.68 g versus PBS group: 31.5 ± 0.59 g, p ≤ 0.05; WAC group: 32.26 ± 0.5 g, p ≤ 0.01). The largest weight reduction in the BAC group occurred on Day 18 (29.76 ± 0.55 g, p ≤ 0.001). From Day 21 onward, BAC weights began to rise again but remained significantly lower than those in the control groups.

The initial weight loss in all three groups during the first three days likely reflects common post-injection reactions, including fever and local inflammation at the injection site. After Day 7, the PBS and WAC groups began gaining weight, whereas the BAC group continued losing weight, demonstrating that a dose of one million BAC significantly reduces body weight in the transplanted mice.

Changes in BMI are presented in Figure 8B. All groups showed a downward trend in BMI during the first three days post-transplantation, with no significant differences at that time. Starting on Day 7, BMI trajectories diverged. The PBS and WAC groups exhibited a modest increase in BMI by Day 7 (fluctuating between 0.35 and 0.36 until Day 28), with no statistical difference between them. This indicates that WAC transplantation did not influence BMI changes.

In contrast, the BAC group’s BMI continued to decrease, reaching its lowest value on Day 18 (0.31 ± 0.01, p ≤ 0.01 compared to WAC). Although the BAC group’s BMI began to rise again from Day 21 onward (0.33 ± 0.02 on Day 28), it remained significantly lower than that of the WAC group (p ≤ 0.05). The initial BMI decline observed in all groups likely stemmed from common post-injection effects. Subsequently, only the BAC group maintained a downward BMI trend, achieving maximal reduction around Day 18. From Day 21 onward, the BMI-lowering effect started to wane, consistent with the trends in body weight. These findings align with reported inverse correlations between brown adipose tissue activity and BMI. Therefore, transplanting one million BAC cells reduces BMI in diet-induced weight-gain mice, sustaining the effect for approximately three weeks.