(A) Primary inguinal white preadipocytes were differentiated in the presence of full or partial agonists for PPARγ, such as rosiglitazone (rosi), MRL24, nTZDpa, Mbx-102, and BVT.13. mRNA levels for brown fat-selective genes (ucp1 and cidea) were analyzed by qRT-PCR. * P<0.05, ** P<0.01 relative to control. (B) Wild type C57BL/6J mice were injected IP with saline or rosiglitazone at 10mg/kg for 10days. mRNA levels for ucp1and cidea were measured by qRT-PCR in interscapular BAT, inguinal WAT (Ing), anterior subcutaneous WAT (AntSC), epidydimal WAT (Epi), retroperitoneal WAT (retroP) and mesenteric WAT (Mesent). *P<0.05, **P<0.01 relative to saline control. (C) mRNA levels for ucp1, cidea, pgc1a and prdm16 were measured by qRT-PCR in primary adipocytes differentiated from SVF derived from inguinal WAT and from epidydimal WAT in the presence or absence of rosiglitazone. *P<0.05, **P<0.01 relative to control. § P<0.01 relative to epidydimal cells treated with rosiglitazone. Data are expressed as means ± SEM.

(A) mRNA levels for Prdm16 were measured by qRT-PCR in primary inguinal adipocytes infected with a control scramble shRNA (sh-scr) or shRNA targeting PRDM16 (sh-PRDM16). **P<0.01 relative to sh-scr. (B) Oil-red O staining of primary inguinal cells at day 7 of differentiation in the presence or absence of rosiglitazone. (C) Microarray analysis of the differentiated inguinal adipocytes expressing sh-scr or sh-PRDM16 in the presence or absence of rosiglitazone. (D) mRNA levels for the indicated genes were analyzed by qRT-PCR. * P<0.05, **P<0.01 relative to control cells. § P<0.01 relative to the cells expressing sh-PRDM16 treated with rosiglitazone. (E) UCP1 protein levels were analyzed by Western blotting. The cells were treated with or without forskolin (cAMP), at 10 μM for 6 hrs. (F) Total and uncoupled cellular respiration were measured in primary inguinal adipocytes expressing sh-scr or sh-PRDM16 in the presence or absence of rosiglitazone and/or. dibutyryl cAMP. ** P<0.01. Data are expressed as means ± SEM.

(A) PRDM16 protein levels were analyzed by Western blotting in epidydimal WAT, inguinal WAT and interscapular BAT from wild type and PRDM16 transgenic mice. Pol-II was blotted as a loading control. (B) mRNA levels for ucp1, cidea, pgc1a, and cox8b were measured by qRT-PCR in primary epidydimal adipocytes from wild type and PRDM16 transgenic mice. **P<0.01 relative to control. § P<0.01 relative to wild type cells treated with rosiglitazone. (C) mRNA levels for elovl3 and fabp4 were measured by qRT-PCR. (D) Wild type mice and PRDM16 transgenic mice were injected IP with saline or rosiglitazone at 10mg/kg for 10days. The inguinal WAT depots were fixed and stained by H&E. Scale bar, 100 μm. (E) Immunohistochemistry for UCP1 expression from samples in (D). Scale bar, 100 μm. (F) mRNA levels for ucp1 and prdm16 were quantified by qRT-PCR from samples in (D). *P<0.05, **P<0.01. Data are expressed as means ± SEM.

(A) Primary inguinal cells from SVF were differentiated in the presence or absence of rosiglitazone at 1μM for 4 days (Day4R), for 8 days (Day8R), and for 4 days with rosiglitazone and additional 4 days without rosiglitazone (Day4R-D4off). PRDM16 protein levels were analyzed by Western blotting. Pol-II was blotted as a loading control. (B) Primary inguinal WAT-derived adipocytes were treated with full or partial agonists for PPARγ. PRDM16 protein levels were analyzed by Western blotting. β-actin was blotted as a loading control. (C) PRDM16 protein levels were analyzed by Western blotting in primary adipocytes treated with rosiglitazone at different doses. (D) PRDM16 protein levels were analyzed by Western blotting in BAT and inguinal WAT isolated from wild-type mice. Mice were injected IP with saline or rosiglitazone at 10mg/kg for 10 days. (E) PRDM16 protein levels in a cycloheximide chase experiment were analyzed by Western blotting. Top panel: control cells, Bottom panel: rosiglitazone treated cells (F) Regression analysis of PRDM16 protein stability in (E). (G) F442A cells expressing flag-tagged PRDM16 were differentiated in the presence or absence of rosiglitazone. Cells were treated with a proteasome inhibitor MG132 (10 μM) for 16 hours prior to harvesting the cells. PRDM16 was immunoprecipitated using flag antibody, and ubiquitin was detected by Western blotting. Total PRDM16 protein is shown in the bottom panel. (H) Correlation analysis between PRDM16 protein levels and ubiquitinated PRDM16 levels. (I) PRDM16 protein levels were analyzed by Western blotting in primary inguinal adipocytes infected with control scramble shRNA or shRNA targeting PRDM16 at low- or high-tires adenovirus. (J) mRNA levels for prdm16 and ucp1 were measured by qRT-PCR in (I). **P<0.01 relative to sh-scr control cells treated with rosiglitazone. § P<0.01 relative to cells without rosiglitazone treatment. (K) Correlation analyses between ucp1 mRNA levels and PRDM16 protein levels (left panel), and between ucp1 mRNA levels and prdm16 mRNA levels (right panel). Data are expressed as means ± SEM.