VEGF-A–Expressing Adipose Tissue Shows Rapid Beiging and Enhanced Survival After Transplantation and Confers IL-4–Independent Metabolic Improvements

Adipose tissue in the adult organism undergoes periods of dynamic expansion and reduction under different metabolic conditions, depending on the energy needs of the host (1). To support the remodeling of adipose tissue, the plasticity of the embedded vasculature in adipose tissues is crucial to maintain appropriate access of oxygen and nutrients to the tissue (2). Although white adipose tissue (WAT) is vascularized, brown adipose tissue (BAT) is particularly highly vascularized, and the interaction between adipocytes and vascular capillaries is essential for adipocyte homeostasis under physiological and pathological conditions (1,2).

Vascular endothelial growth factor-A (VEGF-A) is classically known to be involved in vascular development during embryogenesis (vasculogenesis), as well as blood vessel formation (angiogenesis) and tissue remodeling in the adult organism (3). Recent findings highlight the roles of adipose tissue VEGF-A in the control of adipose tissue function and systemic energy metabolism through the modulation of the adipose vasculature (4,5). VEGF-A overexpression in WAT facilitates angiogenesis and thereby causes a “beiging effect” in subcutaneous WAT (sWAT), altogether resulting in a healthier expansion of WAT as well as protection against genetically and diet-induced obesity and metabolic dysfunction (4). Furthermore, VEGF-A overexpression in BAT augments vascularization and thermogenesis during chronic cold exposure and protects against systemic metabolic dysfunction induced by a high-fat diet (HFD) challenge (5). VEGF-A therefore exerts a crucial role in adipose tissue homeostasis and adaptation to altered nutrient and environmental conditions through a number of different mechanisms (2). The kinetics and extent to which the VEGF-A–induced beiging of subcutaneous fat resembles mechanistically the process induced by cold are not clear.

In the area of tissue regeneration and plastic surgery, adipose tissue is commonly used for autologous transplantation (6). However, poor survival and a high absorption rate of transplanted adipose tissue are likely caused by ischemia and insufficient adipogenic differentiation. The lack of sufficient proangiogenic activity immediately posttransplantation is the main drawback of adipose cell transplants. To overcome these issues, several attempts have been made through the modulation of the methods used to harvest the tissue, the approach used for injection, and the choice of the injection site as well as combination with adipose-derived regenerative cells, extracellular scaffolds, and embedding of angiogenic cytokines such as VEGF and epidermal growth factor (7–13). These approaches were only partially successful, and proper vascularization of implanted adipose tissue remains a big challenge. We have previously described a unique mouse model that allows us to inducibly express VEGF-A specifically in the adipocyte with very high spatial resolution (4). We examined whether VEGF-A–induced vasculature and metabolic changes in adipocytes are cell autonomous and further investigated if a metabolically healthy VEGF-A–overexpressing adipose tissue can be used for cell/tissue therapy to improve metabolic homeostasis in transplant recipients in vivo. For these studies, we used an adipose tissue–specific, doxycycline (dox)–inducible, VEGF-A–overexpressing transgenic (VEGF-Tg) mouse model (4) and used the mice as donors for fat pad implantation in normal and metabolically challenged recipients.

To generate a mouse model with dox-inducible VEGF-A overexpression in adipocytes, tetracycline-responsive element (TRE)–VEGF-A Tg mice were bred with adiponectin promoter–driven rtTA-Tg mice (Apn-rtTA), as previously described (4). Animals used in this study were all in a pure C57BL/6 background (The Jackson Laboratory). All experiments were conducted using littermate-controlled male mice and started when these mice were 7 weeks old. Mice were housed in cages with a 12-h dark/light cycle with free access to water and regular chow diet (5080; LabDiet, Frenchtown, NJ). For the HFD challenge experiments, mice were fed with a diet containing 60% calories from fat (D12492; Research Diets, Inc., New Brunswick, NJ). For low dosage of dox treatment in combination with HFD experiments, all mice (including controls) were fed with the HFD paste (Research Diets, Inc.) mixed with dox powder (Sigma-Aldrich) to a final concentration of 60 mg/kg. For the normal dox diet, a 600 mg/kg diet (S4107; Bio-Serv, Flemington, NJ) was used. All animal experiments were approved by the Institutional Animal Care and Research Advisory Committee at the University of Texas Southwestern Medical Center.

Subcutaneous adipose tissues were taken from the inguinal fat pads of 7-week-old male mice. Following washing of the pads in PBS, 200-mg tissue pieces were surgically implanted into interscapular area of isogenic C57/BL6J male wild-type mice. The fat grafts were retrieved at the indicated time points for further analysis.

Adipose tissues were harvested and homogenized in TRIzol (Invitrogen, Carlsbad, CA) using a MagNA Lyser (Roche, Basel, Switzerland) and ceramic beads. For both cell culture (stromal cells) and adipose tissue, total RNAs were lyzed in TRIzol (Invitrogen) and isolated using the RNeasy RNA extraction kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The cDNAs were prepared by reverse transcription with 1 μg total RNA and Maloney murine leukemia virus (Invitrogen). Quantitative real-time PCRs were performed with TaqMan or SYBR gene-specific primers on an ABI Prism 7900 HT sequence detection system (Applied Biosystems, Foster City, CA). Primer sequences are listed in Supplementary Table 1. The relative amounts of all mRNAs were calculated, and GAPDH and β₂-microglobulin mRNA levels were used for the internal controls.

Tissues were fixed with 10% formalin and embedded in paraffin for histological analysis. Deparaffinized tissue slides were stained with primary antibodies for uncoupling protein-1 (UCP1) (1:200; ab23841; Abcam, Cambridge, MA), CBP/p300-interacting transactivator 1 (CITED1) (1:300; ab15096; Abcam), MAC-2 (1:500; CL8942AP; Cedarlane Laboratories), and endomucin (1:200; sc-65495; Santa Cruz Biotechnology). Biotin-labeled secondary antibodies were used. The reaction was visualized by the DAB Chromogen A system (DakoCytomation, Carpinteria, CA) and counterstained with hematoxylin. Images were acquired using a Coolscope (Nikon, Tokyo, Japan). Hematoxylin and eosin (H&E) staining and Masson’s Trichrome C staining were performed by Dr. John Shelton at the University of Texas Southwestern Medical Center Histology Core.

Formalin-fixed, paraffin-embedded slides were used. Deparaffinized slides were stained with primary antibodies for MAC-2 (1:500; CL8942AP; Cedarlane Laboratories) and perilipin (20R-PP004; Fitzgerald Industries International). Fluorescence-labeled secondary antibodies were used and counterstained with DAPI. Images were acquired using the Leica confocal microscope (Leica Microsystems) and analyzed with ImageJ software (National Institutes of Health).

Fasting glucose levels were determined after 3 h of fasting, and serum samples were collected from tail-vein blood samples. For the oral glucose tolerance tests (OGTTs), mice were fasted for 5 h prior to administration of glucose (2.5 g/kg body weight by gastric gavage). At each of the indicated time points, serum samples were collected from tail veins. Glucose concentrations were measured using an oxidase-peroxidase assay (Sigma-Aldrich). Serum triglycerides (TGs), cholesterol (Infinity TG or cholesterol kit; Thermo Fisher Scientific, Waltham, MA), and free fatty acid (FFA) (NEFA-HR [2]; Wako Diagnostics, Tokyo, Japan) levels were measured by the kits following the manufacturer’s instructions. The examination of serum adiponectin and insulin levels was done by ELISA (ALPCO, Salem, NH).

Tissue collagen content was measured by assessing 4-hydroxyproline levels with a kit from BioVision (Milpitas, CA). Briefly, 50 mg fat tissue was homogenized in distilled water and then mixed with 6 N HCl at 120°C for 6 h. Supernatants were dried and further incubated with chloramine-T at 25°C for 10 min, and then DMAB was added to each well and incubated at 90°C for 60 min. The absorbance was measured at 560 nm using a microplate reader as suggested by the manufacturer.

Serum VEGF-A levels were measured by ELISA. Mouse whole blood was collected and serum prepared. Sera were assessed by a mouse-specific VEGF-A ELISA kit (Abcam) and analyzed according to the manufacturer's instructions.

All results are provided as means ± SEM. All statistical analyses were performed using Prism software (GraphPad Software, La Jolla, CA). Differences between the two groups over time were determined by a two-way ANOVA for repeated measures. For comparisons between two independent groups, a Student t test was used. Significance was accepted as P < 0.05.

Previous reports have suggested a potential role of VEGF-A in adipose tissue function and systemic energy metabolism, in part through VEGF-A’s effects on stimulation of angiogenesis, macrophage M2-subtype inflammation, and BAT differentiation and its function (4). To address more directly what the kinetics of VEGF-A action on adipose tissue are, we took advantage of the high temporal resolution that our genetic approach offers. We used the adipose tissue-specific dox-inducible VEGF-A Tg mouse model. This model is a combination of two Tg mouse lines, the TRE-driven VEGF-A Tg mice (TRE-VEGF) and the adipose tissue–specific tetracycline on (Tet-on) Apn-rtTA mice. In this model, the TRE can be activated by the rtTA transcription factor in the presence of dox (4). Notably, expression of the transgene in this inducible system is strictly limited to the mature adipocyte and is not present in other cell types, such as adipogenic precursor cells and macrophages. In our previous studies, chronic overexpression of VEGF-A in adipose tissue under an HFD challenge for 8 weeks triggered beige adipose tissue-like properties in sWAT and decreased the expression of inflammatory factors such as interleukin-6 (IL-6), F4/80, and tumor necrosis factor-α in epididymal WAT (4) compared with wild-type mice. A reduction of these inflammatory markers is frequently seen in metabolically healthy adipose tissue (14,15). The new class of adipocytes observed in sWAT under these conditions is referred to as “beige” adipocytes, which emerge in sWAT and are characterized by displaying elevated levels of proteins characteristic of BAT, such as expression of UCP1 and an elevated respiratory rate (16).

To further connect elevated local VEGF-A expression to the beiging effects observed in WAT, we analyzed sWAT histologically during a time-course experiment after exposing mice to dox (Fig. 1). Histological examination of sWAT in VEGF-Tg mice revealed that within 2 days of VEGF-A exposure, widespread beiging is apparent. Smaller adipocyte size and multilocular appearance can be seen, which became even more evident at 6 days post–VEGF-A induction by H&E staining (Fig. 1A). We further confirmed this phenotypic change in sWAT by immunohistochemically assessing UCP1 expression and additional beige adipocyte cell markers, such as CITED1 (Fig. 1B, C, and E) and the endothelial cell maker endomucin. Within 2 days post–dox-diet exposure, beige adipocyte markers and endomucin are apparent in VEGF-A–expressing mice. There was very limited macrophage accumulation present in areas with pronounced beiging as determined by immunostaining with the macrophage marker MAC-2 (Fig. 1D). These results indicate that the rate of appearance of beige adipocytes induced by VEGF-A in sWAT occurs with rapid kinetics, within as short as 2 days after VEGF-A exposure, leading to a widespread beiging phenotype in sWAT within 6 days of induction. Importantly, although there was a transient increase in macrophage infiltration on day 2, this signal was lost as time went on. To achieve a better temporal resolution that would allow us to address the relationship of vascularization per se versus beiging, we performed an experiment within an even shorter time course. The result clearly suggests that the increased vascularization precedes the bulk of the UCP1 induction, as judged by the apparent endomucin stain that only overlaps at the 36-h time point with the emerging UCP1 stain (Fig. 1F–I).

sWAT and BAT were collected after 2 days of dox exposure and assayed by Western blot analysis for UCP1 expression (Fig. 2A). To assess whether VEGF-A–induced phenotypic changes in sWAT are tissue autonomous, we performed adipose tissue transplantations. sWAT taken from either VEGF-Tg mice or wild-type mice given a dox diet for 7 days prior to tissue harvest were subcutaneously implanted into isogenic C57/BL6J wild-type mice and maintained on a dox diet (Fig. 2B). Gross examination of fat grafts 7 days after implantation revealed that VEGF⁺ sWAT grafts appeared darker in color, and more blood vessels were generated compared with control sWAT transplants (Fig. 2B). H&E-stained tissues also showed increased vascularization and a multilocular appearance of VEGF⁺ sWAT compared with control tissues (Fig. 2C). Immunostaining for UCP1 confirmed that BAT-like phenotypic changes in VEGF⁺ sWAT grafts were still intact in the setting of 7 days after implantation (Fig. 2D). Furthermore, the Tg transplants were significantly better vascularized, as judged by the much denser endomucin stain seen in the transplanted fat pads expressing VEGF-A (Fig. 2E). These results indicate that VEGF-A–induced phenotypic changes in sWAT reflect tissue-autonomous effects, and these are sustained for at least 1 week postimplantation.

However, if stromal vascular cells are harvested from the donor and in vitro–differentiated into adipocytes, there is no difference with respect to the induction of any of the beige marker genes in the continued presence of VEGF-A induction during differentiation in vitro (Fig. 2F), indicating that the effects are not adipocyte autonomous, consistent with our previous report demonstrating that in vitro treatment of mature adipocytes with exogenous VEGF-A did not result in an induction of a beiging program (17).

In our previous study, we found that VEGF-A overexpression in sWAT ameliorates hypoxia, fibrosis, and proinflammatory responses induced by an HFD challenge (4). To test whether exogenously implanted VEGF-A⁺ fat tissues can contribute to systemic metabolism of HFD-challenged host mice, either control or VEGF-A⁺ sWAT pads were taken from wild-type or VEGF-Tg mice 1 week after dox-diet exposure, respectively, and subsequently implanted into isogenic wild-type mice. Following sWAT implantation, these mice were fed with an HFD combined with dox for 3 weeks (Fig. 3A). Fasting glucose levels in circulation were not significantly different in VEGF-A⁺ fat–implanted mice compared with control fat–implanted mice at the end of this period (Fig. 3B). Body weight changes were also comparable in both groups (data not shown). To provide an additional carbohydrate challenge, we performed OGTTs on mice carrying for VEGF-A⁺ or wild-type fat implants. Mice carrying VEGF-A⁺ fat transplants showed enhanced glucose tolerance compared with mice receiving control transplants (Fig. 3B). VEGF-A–expressing transplanted fat pads can therefore exert a significant positive impact on metabolic homeostasis of recipient mice. Extracts of the transplanted fat pads were analyzed at the end of the study period for a number of parameters. There were significantly higher levels of VEGF-A mRNA seen in the Tg versus wild-type fat transplants (Fig. 3C). As expected, higher Ucp1 mRNA levels were observed in the VEGF⁺ transplants (Fig. 3D). VEGF-A action also translated into the expected higher vascular density as judged by the higher levels of Cd31 mRNA (Fig. 3E). There were significant lower levels of inflammatory markers found in the VEGF-A⁺ fat pads. Reduced F4/80 levels along with reduced serum amyloid A3 and lower tumor necrosis factor-α mRNA were observed, whereas Il6 levels seemed unchanged (Fig. 3F). These reduced inflammatory markers reflect an overall healthier fat pad, and consistent with these improvements, there is significantly less fibrosis seen in the VEGF-A⁺ fat pad, as judged by gene expression of extracellular matrix proteins and total collagen levels in the transplanted tissue (Fig. 3G and H). To address systemic effects of transplantation with VEGF-A⁺ fat tissue, we measured serum VEGF-A levels of the host animals. The effects of VEGF-A production in the transplanted fat pads seem to be restricted to the local microenvironment, as there is no significant difference in circulating VEGF-A levels compared with mice receiving wild-type transplants (Fig. 3I). In addition, there was no evidence suggesting that the transplantation of VEGF-A⁺ fat pads led to beiging of WAT of the host (Supplementary Fig. 1). The VEGF-A⁺ fat transplants led to significant improvements in systemic parameters as well, beyond the improvements in the OGTT. Adiponectin levels were increased in circulation (Fig. 3J). Although circulating cholesterol levels were unaffected, significant improvements in circulating TGs, nonesterified FFAs, and insulin were observed (Fig. 3K–N).

A qualitative histological examination of fat grafts after the HFD challenge revealed that VEGF-A⁺ fat grafts are histologically quite distinct from control transplants (Fig. 4A). Enhanced vascularization was further confirmed with endothelial markers, such as an antiendomucin stain (Fig. 4B). UCP1 staining for VEGF⁺ fat grafts further reflects extensive browning of sWAT compared with control (Fig. 4C). These results revealed that VEGF-A–induced angiogenesis and beiging effects in sWAT remain intact in fat grafts even 3 weeks after implantation.

Adipose vasculature and beiging effects are associated with metabolic benefits to adipose tissues. To see if these properties confer better survival posttransplantation, VEGF-A⁺ fat grafts were stained with the adipocyte marker perilipin (Fig. 4D, red). Only live adipocytes retain a perilipin-positive stain on their lipid droplets. The macrophage marker MAC-2 was used in parallel to monitor the degree of local inflammation (Fig. 4D, green). The survival rate of adipocytes was dramatically increased in VEGF-A⁺ fat grafts with reduced macrophage infiltration relative to control fat grafts (Fig. 4D and E). These results suggest that VEGF-A overexpression in adipocyte increases the survival of adipocytes in fat grafts, thereby explaining the reduced levels of inflammatory markers as well as the reduced fibrotic response seen in Fig. 3.

Previous work suggested that both adiponectin and IL-4 are important components in the context of beiging of sWAT. Adiponectin overexpression leads to enhanced levels of beiging (18,19), whereas cold-induced beiging and browning were postulated to involve IL-4. Genetic loss of IL-4 signaling impairs cold-induced biogenesis of beige fat (20). We wanted to test whether the VEGF-A–induced beiging process has similar requirements as the cold-induced beiging, or whether we can highlight mechanistic differences. We were therefore breeding the adipocyte-specific inducible VEGF-A mice into the IL-4– and adiponectin-null backgrounds. Histologically, it did not make a difference whether the VEGF-A induction was done in a wild type (not shown), an IL-4–heterozygous state, or in an IL-4–null background (Fig. 5A). Comparable areas of UCP1 induction were seen independent of IL-4 genotype (Fig. 5B). We confirmed that IL-4 is indeed absent in the IL-4–null background (Fig. 5C), whereas the degrees of Ucp1 mRNA induction were comparable even in the IL-4–null background (Fig. 5D). Similarly, the beiging process was completely independent of the presence or absence of adiponectin (Fig. 5E–H). When we tested the requirement of IL-4 for cold-induced beiging, we saw indeed that both BAT and sWAT were visibly less “dark,” potentially reflecting reduced induction of the mitochondrial program (Fig. 6A). Post–cold exposure, the entire beiging and browning program in BAT was reduced (Fig. 6B and C), and this was further confirmed histologically by H&E stains and anti-UCP1 immunostains (Fig. 6D–F), confirming in our hands the previously established findings that the process of beiging and browning critically relies on IL-4.

We exposed wild-type and VEGF-A Tg mice to dox for 3 days to fully induce the beiging program. We then removed the dox from the food and analyzed the sWAT after 7 and 12 days of VEGF-A washout (Fig. 7A). Dox is usually washed out within 12–16 h after its removal (21,22). The sWAT from the Tg mice looks visibly browner after 3 days of induction of VEGF-A with dox. The color of the sWAT appears paler and reaches normal after 12 days of dox washout (Fig. 7A). Histologically, the characteristic multilocular appearance disappears within the same time frame, as judged by H&E stains (Fig. 7B). These observations are further confirmed at the gene expression level (Fig. 7C and D). UCP1 and VEGF-A reach levels of controls after 7 days and in fact fall even below normal levels in the controls by day 12, likely because of a compensatory response to the previous elevation of the levels during dox exposure.

Angiogenesis is an important component of adipose tissue remodeling under both normal and pathophysiological conditions (1). Proper adipose tissue vascularization is essential to maintain normal tissue homeostasis. Improvements in adipose tissue angiogenesis offer a potential therapeutic avenue for metabolic diseases (23). Various cytokines secreted from adipocyte, such as leptin, adiponectin, hepatocyte growth factor, fibroblast growth factor, platelet-derived growth factor, transforming growth factor-β, and VEGF as well as the butyric acid derivative monobutyrin, exert established angiogenic activities in adipose tissue (2). Particularly, a host of recent studies have suggested that adipose tissue–derived VEGF, a bona fide endothelial growth factor, plays a crucial role in adipose tissue plasticity and, secondary to that, in systemic metabolism (4,5,24,25). In this study, we focused on the temporal aspects of VEGF-A action in adipocytes and the advantages this confers upon transplanting the tissue. To our surprise, local overexpression of VEGF-A in adipose tissue very rapidly triggers angiogenesis and browning of sWAT, with initial signs apparent as early as 2 days after exposure to higher levels of VEGF-A. We were particularly surprised how quickly the transformation of sWAT occurred, and this happens with VEGF-A levels well within the physiological range, because massive overexpression of VEGF-A can lead to edema formation and needs to be avoided (4). As such, VEGF-A is probably one of the most potent “beiging” factors described to date or minimally, the only one for which such a refined time course of action is being described. Our previous data (26) have shown that either cold-induced or β3-adrenergic agonist–mediated beiging of sWAT is primarily because of de novo recruitment of precursor cells. Whether a similar mechanism is in place for VEGF-A–mediated beiging is difficult to address, as two parallel inducible systems would be required to do the traditional pulse/chase experiments. However, because mature adipocytes do not express VEGF-A receptors, and the addition of VEGF-A in tissue culture does not cause any degree of browning (4), it seems unlikely that a simple “trans-differentiation” mechanism is in place despite the unexpectedly rapid kinetics observed in this study. Interesting recent data also point toward an intracrine-signaling mechanism of VEGF. VEGF can control differentiation in mesenchymal stem cells that are osteogenic precursors by acting through the activation of the transcription factors RUNX2 and peroxisome proliferator–activated receptor γ2 as well as through interactions with the nuclear envelope proteins lamin A/C. This intracrine signaling mechanism is distinct from the role of secreted VEGF-A and its receptors (27). We cannot exclude that such a mechanism is in place in this study as well. What is, however, quite clear is that the expansion of the vasculature precedes the appearance of UCP1-positive cells, highlighting an interesting temporal relationship between the two processes. Future experiments will have to highlight whether the mere expansion of the vasculature is sufficient for the process of beiging or whether there is an additional VEGF-A–dependent component to the process. Also quite apparent but less surprising is the observation that the beige adipocytes induced by VEGF-A require the continued presence of VEGF-A. Loss of ectopic expression of VEGF-A results in the disappearance of the beige adipocytes within 7 to 12 days, presumably through conversion to white adipocytes.

We chose adipose tissue transplantation as a model system to probe for adipose tissue health and survival benefits induced by VEGF. Fat implantation is a common technique for the repair of tissue defects in plastic and reconstructive surgery. However, the survival rates of adipocytes in fat grafts are varying depending on numerous conditions during surgery (28–30). Therefore, establishment of stable and optimized conditions to sustain adipocyte health and survival is critical for a successful surgical fat implantation. In this study, we show that VEGF-overexpressing adipose tissue confers enhanced adipocyte survival with reduced chronic inflammation, leading to improvements in systemic metabolic parameters in diet-challenged obese hosts. The experiments demonstrate that the original phenotype of the adipose tissue from the donor is maintained in the host over prolonged periods and can convey significant metabolic benefits to the host. Similar results have been reported by Min et al. (31) for human beige fat transplants, commented on by Wang and Scherer (32). These authors demonstrated that they can differentiate human beige precursors in vitro into beige adipocytes. These cells can be transplanted into nude mice, in which they can positively affect systemic glucose homeostasis. The findings we report in this study with Tg VEGF-A are very much in line with the results reported by Wang and Scherer (32) and Min et al. (31), highlighting how small patches of transplanted fat can cause significant improvements in metabolic homeostasis.

Finally, we also studied this model with an eye on deeper mechanistic events leading to the beiging process. In particular, Qiu et al. (20) described an efferent beige fat thermogenic circuit that consists of eosinophils, the type 2 cytokines IL-4/13, and alternatively activated macrophages. In this model, macrophages are recruited to cold-stressed subcutaneous adipose tissue, undergo alternative activation, and induce tyrosine hydroxylase activity along with catecholamine production, factors critical for the beiging process. We tested whether the VEGF-A–mediated beiging process has similar requirements by breeding the inducible adipocyte-specific VEGF-A system into the IL-4–null background. IL-4–null mice indeed showed a reduced propensity toward beiging and browning in the cold. However, the VEGF-A–mediated beiging was unaffected by the absence of IL-4. This is the first example in which the beiging process has been studied outside the scope of cold exposure with respect to the postulated involvement of macrophages and the associated catecholamine production. During the time course, we indeed observe a transient infiltration of macrophages after 24 h, which subsequently rapidly dissipates. Altogether, this hints at important differences between classic cold-mediated effects and at least VEGF-A–induced beiging. Similar results held true for VEGF-A effects in adiponectin null mice.

As additional beiging factors are being described in the literature, it will be interesting to see what commonalities and distinctions can be found regarding the mechanisms driving the beiging process among the different factors and the cold-induced beiging process.

Acknowledgments. The authors thank the University of Texas Southwestern Medical Center Molecular Pathology Core for help with histology and the Metabolic Phenotyping Core Facility for help with the phenotyping.

Funding. This work was supported by the National Institutes of Health (R01-DK55758, R01-DK099110, and P01-DK088761) as well as a grant from the Cancer Prevention and Research Institute of Texas (CPRIT RP140412 to P.E.S.). This work was also supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute funded by the Ministry of Health & Welfare, Republic of Korea (HI14C1277) and a grant from the Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education (NRF-2014R1A1A2054914 to J.P.). M.K. was supported by the Priority Research Centers Program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (2010-0020224). K.S. was supported by a pilot award from the Center for Clinical and Translational Sciences at The University of Texas Health Science Center at Houston (UL1-TR-000371).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. J.P., M.K., K.S., Y.A.A., and X.G. contributed to the data collection and analysis. J.P., M.K., K.S., and P.E.S. contributed to the study conception and design and writing of the manuscript. P.E.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in poster form and as a portion of the keynote lecture at the Keystone Symposia on Diabetes, joint with Obesity and Adipose Tissue Biology, Keystone, CO, 22–26 January 2017.