LAUSR

Obesity is caused by an imbalance between energy intake and expenditure, and has become a global epidemic with over 650 million adults affected. Adipose tissues in mammals are composed of white adipose tissue (WAT) and classical brown adipose tissue (BAT), and their balance is highly related to the occurrence of obesity. The browning of white adipocytes results in “beige” or “brite” adipocytes, which appear functionally similar to classical brown adipocytes, and can be detected in WAT deposits of animals that have been exposed to cold or other inducers (Fu et al., 2015). Accumulating evidence has emphasized the role of brown and beige adipocytes in counteracting obesity by dissipating chemical energy as heat (Kajimura et al., 2015). This process is reported to be mediated by uncoupling protein 1 (UCP1), which is abundant within the inner mitochondrial membrane of brown and beige adipocytes (Nowack et al., 2017). UCP1 functions as an uncoupler of oxidative phosphorylation by increasing the permeability of the inner mitochondrial membrane that leads to a dissipation of the proton gradient. Several transcriptional regulators act as powerful activators for recruitment of brown adipocytes in WAT, such as PRDM16, and PGC1α (Ohno et al., 2012; Fu et al., 2015). Compounds such as celastrol (cela), berberine, and artemisinin have been identified as inducing brown-like adipocytes in WAT (Zhang et al., 2014; Ma et al., 2015; Lu et al., 2016). Despite the beneficial effects in thermogenesis, the direct protein targets and specific molecular mechanisms of these compounds are still illusive. Limited efficacy, diminished specificity, and multiple side effects remain major challenges in clinic uses of these compounds. Identifying novel compounds and clarifying direct specific targets for activating beige adipocytes are of great significance. In previous work, we reported that artemether can induce browning of adipocytes (Lu et al., 2016). In the same study, we also found that ethacrynic acid (Edecrine, EA) (Fig. 1A) activates the expression of Ucp1 during adipogenesis of 3T3-L1 and mesenchymal cell line C3H10T1/2 (Fig. S1A and S1B). We found that EA treatment led to the smaller lipid droplets (Fig. 1B), higher mRNA and protein levels of UCP1 (Figs. 1C and S1C). Subsequently, we examined the metabolic profile elicited by EA in vivo. Mice maintained on high fat diet (HFD; 60% calories from fat) were treated with EA (5 mg/kg) by intraperitoneal injection for 1 week, and then sacrificed with cold exposure. Surprisingly, mice with EA administration showed significantly reduced body weight (Fig. S1D) and improved cold resistance (Fig. 1D). After cold exposure, iWAT were separated and consumption of oxygen was measured. EA significantly increased the oxygen consumption rate (OCAR) in iWAT (Fig. 1E), implying augmented energy expenditure with EA administration. Meanwhile, iWAT of EA administered mice displayed distinctly smaller lipid droplets (Fig. 1F) and increased UCP1 level (Fig. 1F–H), as well as other browning related genes, including Pgc1α, Prdm16, Cd137, Tmem26, and particularly Tbx1 (Fig. 1G), which is considered as a strong marker distinguishing beige adipocytes from either white adipocytes or brown adipocytes (Roh et al., 2018). In addition, mRNA levels of Pgc1α and Prdm16 were detected to increase in BAT with EA treatment, but the mRNA level of Ucp1 remained unchanged (Fig. S1E). Taken together, these results strongly indicate a distinct effect of EA on enhancing the browning of iWAT induced by cold exposure. EA is one of the most potent diuretic agents. By binding to the Na-K-2Cl-co-transporter (NKCC2) in the ascending loop of Henle, EA inhibits sodium reabsorption, by a mechanism similar to furosemide (FS) (Milne et al., 2007) Besides, EA is generally known as an inhibitor of glutathione-Stransferases (GSTs) (Mary Schultz et al., 1997). However, the potential role of NKCC2 or GSTs in iWAT browning remains unknown. We began by investigating the function of NKCC2 by treating C3H10T1/2 cells with FS, which is an inhibitor of NKCC2. As shown in Fig. S2A–C, FS did not lead to neither morphological changes of lipid droplets nor increased expression of browning related genes, suggesting the browning effect of EA is most likely not related to NKCC2. We next asked whether the browning effect of EA is related to GSTs. Cytosolic GST proteins are grouped into of seven classes, including GST alpha (GSTA), mu (GSTM), pi (GSTP), omega (GSTO), kappa (GSTK), zeta (GSTZ) and