AMPK Activation by Metformin Suppresses Abnormal Adipose Tissue Extracellular Matrix Remodeling and Ameliorates Insulin Resistance in Obesity

Transcription

1 Page 1 of 51 Diabetes AMPK Activation by Metformin Suppresses Abnormal Adipose Tissue Extracellular Matrix Remodeling and Ameliorates Insulin Resistance in Obesity Ting Luo 1,2, Allison Nocon 1, Jessica Fry 1, Alex Sherban 1, Xianliang Rui 1, Bingbing Jiang 1, X. Jian Xu 1, Jingyan Han 1, Yun Yan 3, Qin Yang 4, Qifu Li 2, Mengwei Zang 1,5,6,7 * 1 Department of Medicine, Boston University School of Medicine, Boston, MA, 02118; 2 Department of Endocrinology, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China. 3 Division of Endocrinology, Department of Pediatrics, Children's Mercy Hospital and University of Missouri-Kansas City, Kansas City, MO 64108; 4 Department of Medicine, Physiology and Biophysics, Center for Diabetes Research and Treatment and Center for Epigenetics and Metabolism, University of California, Irvine, CA 92697; 5 Barshop Institute for Longevity and Aging Studies, Center for Healthy Aging; 6 Department of Molecular Medicine, University of Texas Health Science Center, San Antonio. TX78207; 7 Geriatric Research, Education and Clinical Center, Audie L. Murphy VA Hospital, South Texas Veterans Health Care System, San Antonio. TX Running Title: The anti-fibrotic role of AMPK in adipose tissue *Correspondence to: Mengwei Zang M.D., Ph.D. Associate Professor of Molecular Medicine Barshop Institute for Longevity and Aging Studies, Center for Healthy Aging Department of Molecular Medicine University of Texas Health Science Center at San Antonio Geriatric Research, Education and Clinical Center for Healthy Aging, South Texas Veterans Health Care System 8403 Floyd Curl Dr., Office 292.2, MC8257 STRF-South Texas Research Facility San Antonio, Texas Office: Fax: Zang@uthscsa.edu The word count: 3998 Figures: 8 1 Diabetes Publish Ahead of Print, published online June 9, 2016

2 Diabetes Page 2 of 51 Abstract Fibrosis is emerging as a hallmark of metabolically dysregulated white adipose tissue (WAT) in obesity. Although adipose tissue fibrosis impairs adipocyte plasticity, little is known about how aberrant extracellular matrix (ECM) remodeling of WAT is initiated during the development of obesity. Here we show that treatment with the anti-diabetic drug metformin inhibits excessive ECM deposition in WAT of ob/ob mice and diet-induced obese mice, as evidenced by decreased collagen deposition surrounding adipocytes and expression of fibrotic genes including the collagen cross-linking regulator LOX. Inhibition of interstitial fibrosis by metformin is likely attributed to activation of AMPK and suppression of transforming growth factor-β1 (TGFβ1)/Smad3 signaling, leading to enhanced systemic insulin sensitivity. The ability of metformin to repress TGF-β1-induced fibrogenesis is abolished by the dominant negative AMPK in primary cells from the stromal vascular fraction. TGF-β1-induced insulin resistance is suppressed by AMPK agonists and the constitutively active AMPK in 3T3L1 adipocytes. In omental fat depots of obese humans, interstitial fibrosis is also associated with AMPK inactivation, TGF-β1/Smad3 induction, aberrant ECM production, myofibroblast activation, and adipocyte apoptosis. Collectively, integrated AMPK activation and TGF-β1/Smad3 inhibition may provide a potential therapeutic approach to maintain ECM flexibility and combat chronically uncontrolled adipose tissue expansion in obesity. KEYWORDS: Metformin, AMPK, TGF-β, adipose tissue fibrosis, extracellular matrix remodeling, collagen cross-linking enzyme, adipocytes, myofibroblasts, human obesity, and insulin resistance 2

3 Page 3 of 51 Diabetes Introduction In obesity, visceral adipose tissue dysfunction contributes to the development of insulin resistance and type 2 diabetes mellitus. Physiologically, extracellular matrix (ECM) components such as collagens and fibronectin provide mechanical support for the tissue. During the progression of obesity-induced adipose tissue expansion, chronic, continuous production and deposition of ECM lead to the destruction of normal adipose tissue architecture, which is referred to as adipose tissue ECM remodeling (1). Genetic deletion of collagen VI, which is highly expressed in adipose tissue, improves adipose fibrosis and systemic metabolic dysfunction in high fat diet-fed mice, highlighting the critical role of adipose tissue ECM in regulating metabolic homeostasis (2). Adipose tissue fibrosis is increasingly appreciated as a major contributor of metabolic dysregulation in obese and type 2 diabetic humans (3). However, how persistent aberrant ECM remodeling is initiated by obesity remains poorly understood. The major goal of the present study is to identify the signaling molecules that modulate fat tissue ECM remodeling in obesity. Metformin is the first-line drug that treats obesity-related type 2 diabetes (4; 5). Our initial studies and others indicate that the lipid-lowering effect of metformin is largely attributed to the activation of the energy sensor AMP-activated protein kinase (AMPK) in hepatocytes (6-9). Furthermore, metformin can repress de novo lipogenesis in hepatocytes by activating AMPK (10). Although the adipose tissue acts as a target tissue of AMPK and its activator metformin (11-14), it remains undefined whether AMPK regulates adipose tissue ECM homeostasis. 3

4 Diabetes Page 4 of 51 We report here that the interplay of AMPK and transforming growth factor-β (TGFβ1) signaling is implicated in pericellular fibrosis in omental fat of obese humans. Activation of AMPK by metformin suppresses TGF-β1/Smad3 signaling, resolves interstitial accumulation of ECM in white adipose tissue, and improves local and systemic insulin resistance in genetic and dietinduced obese animals. Pharmacological and molecular activation of AMPK inhibits collagen induction, LOX-related collagen cross-linking, myofibroblast-like cell differentiation, and adipocyte apoptosis and insulin resistance upon TGF-β1 treatment in vitro. Targeting adipose tissue ECM may benefit adipose and whole-body insulin resistance in obesity. 4

5 Page 5 of 51 Diabetes Research Design and Methods Animal studies. Male 8-10 week old ob/ob C57BL/6 mice as well as their lean littermates and wild-type mice were purchased from Jackson Labs (Bar Harbor, Maine). All animal experiments were approved by the University Committee on Use and Care of Animals at Boston University School of Medicine. Primary cell culture from stromal vascular fraction (SVF) of adipose tissue SVF cells in epididymal fat depots from male C57BL/6 mice were isolated and cultured as described previously (15; 16). Human studies Omental adipose tissue samples were collected from obese patients (BMI: 28.7 ± 1.4) and agematched normal weight individuals (BMI: 19.8 ± 1.5). The Ethics Committee of the First Affiliated Hospital of Chongqing Medical University in China approved the study (IRB#15-39). Subjects gave a written informed consent after individual explanation. Statistical analysis Data are presented as means ± standard error (S.E.M). Using GraphPad Prism 5.0 software, results were analyzed by one-way ANOVA between multiple groups, when appropriate, and by a two-tailed Student's t-test between two groups, as described previously (6; 8; 10; 17-20). P < 0.05 was considered statistically significant. 5

6 Diabetes Page 6 of 51 Results Metformin reduces adiposity, alleviates excessive collagen deposition in white adipose tissue (WAT), and ameliorates whole-body insulin resistance in genetically obese ob/ob mice As shown in Fig. 1, body weight gain was moderately but significantly reduced in metformin-treated ob/ob mice, as was seen in early studies (21). While lean mass was unaffected by metformin, the weight loss was mainly attributed to a 14% decrease in fat mass of ob/ob mice, consistent with a moderate weight loss and fat mass reduction in patients with type 2 diabetes after metformin treatment (14; 22; 23). Histologically, H&E staining showed that metformin lessened features of hypertrophic adipocytes in epididymal white adipose tissue (ewat) of ob/ob mice with a switch toward smaller adipocytes. Masson s Trichrome staining showed a ~7-fold increase in collagen deposition and distinct distribution patterns of interstitial collagen in ewat of ob/ob mice, in which fibrous regions appeared denser and thicker, and gave rise to a larger lobe or a lobule-like structure. Sirius Red staining further confirmed that collagen fibers were mainly organized in fibrotic bundles surrounding adipocytes in ob/ob mice. Strikingly, abnormal collagen deposition was reduced approximately 50% in metformin-treated ob/ob mice. Quantification of hydroxyproline concentrations is considered as an accurate indicator of tissue collagen content (24-26). An increase in hydroxyproline content in the ewat of ob/ob mice was also reduced by metformin, reaching a level comparable with that of lean mice. Thus, metformin protects against adiposity and pronounced extracellular matrix (ECM) remodeling caused by obesity. 6

7 Page 7 of 51 Diabetes To determine the effect of metformin on systemic glucose homeostasis, glucose tolerance tests (GTT) showed that glucose clearance was significantly enhanced by metformin, compared to ob/ob mice. Insulin tolerance tests (ITT) also showed that systemic insulin sensitivity was increased by metformin. Consistently, integrated glucose concentrations, as calculated by the area under the curve (AUC) values, were significantly decreased. Likewise, fasting blood glucose levels were declined ~30% by metformin. A positive correlation between fat collagen content and fasting blood glucose was observed in ob/ob mice treated or untreated with metformin (r 2 = , P < 0.01) (Fig. 1M-R). Metformin inhibits aberrant synthesis and degradation of ECM in WAT of ob/ob mice To elucidate the mechanisms by which metformin attenuates pericellular fibrosis, we measured mrna expression of collagen type I (Col1a1), type III (Col3a1), and type VI (Col6a3), all of which were shown to be highly expressed in obese adipose tissue (2). As shown in Fig. 2A, these collagens were increased 5- to 7-fold in ob/ob mice, despite the basal level of Col6a3 being much higher than that of Col1a1 and Col3a1 in lean mice as described previously (2). In contrast, overexpression of these collagens was attenuated ~50% by metformin. Among major fibrogenic genes, expression of fibronectin, a functional integrin ligand, was also stimulated by obesity and was downregulated by metformin. Lysyl oxidase (LOX), a copper-dependent amine oxidase that catalyzes the covalent crosslinking of fibers through post-translational modification of collagens, is essential for the cross-linking of collagen fibrils and fibers and modification of biochemical and mechanical properties of the ECM (27). The anti-fibrotic effect of metformin was also evidenced by downregulation of LOX in ob/ob mice. 7

8 Diabetes Page 8 of 51 The development of tissue fibrosis occurs as a consequence of the dysregulation of matrix metalloproteinases (MMPs) and tissue inhibitors of MMPs (TIMP) (28-31). To test the hypothesis that suppression of fat collagen deposition by metformin might be due to altered matrix degradation, mrna levels of MMP-2, MMP-9, MMP-13, MT1-MMP (MMP14), which are predominantly upregulated in obese adipose tissue (32), were assessed by qrt-pcr. As shown in Fig. 2B-I, expression of different MMPs was induced 5- to 10-fold in ewat of ob/ob mice, suggesting that the perturbation of normal fat ECM structure and/or its replacement with newly synthesized ECM may act as a force driver to promote metabolically unfavorable fat microenvironment. Interestingly, the induction of MMP-9 and TIMP-1 by obesity was reduced ~60% by metformin. The protein expression and distribution of MMP-9 and TIMP-1, as assessed by immunohistochemical and immunoblotting analyses, were also reduced. No significant changes in TIMP-2 and TIMP-3 was evident among the three groups. Collectively, metformin may inhibit abnormal ECM synthesis and degradation in fat tissue, reduce LOX-related collagen cross-linking, and consequently provoke a favorable ECM flexibility. Metformin suppresses aberrant fat ECM remodeling in WAT of diet-induced obese mice To better understand the potential effect of metformin on fat ECM remodeling process in obesity, we expanded our studies to a diet-induced obesity animal model. As shown in Fig. 3, hypertrophic adipocytes in epididymal and inguinal fat of HFHS-induced obese mice were significantly reduced by metformin. Masson s trichrome staining revealed that the collagen fibers surrounding adipocytes were increased 7- and 3-fold in ewat and inguinal white adipose tissue (iwat) of HFHS-fed mice, respectively. To our surprise, the abundance of collagen deposition had a distinct feature between fat depots of obese mice, with the higher amounts 8

9 Page 9 of 51 Diabetes found in ewat and much lower amounts seen in iwat. Since the basal expression levels of Col1a1, Col6a3, and fibronectin in epididymal fat, a visceral fat associated with metabolic disease, were ~5-fold higher than those in subcutaneous fat, it may reflect depot-specific effect of fat fibrosis. The excessive collagen deposition of ewat, as assessed histologically and biochemically by hydroxyproline content, was significantly decreased by metformin. The beneficial effect of metformin on interstitial fibrosis was also evidenced by reduced fibrogenic gene expression and concurrently inhibited ECM regulators such as MMP-2, MMP-9, MMP-14, and TIMP-1. The modulation of ECM components also occurred in the iwat of metformintreated mice. Decreased hydroxyproline concentrations in ewat was positively correlated with lowered fasting glucose levels in metformin-treated HFHS-fed mice with metformin (r 2 = , P < 0.01). Thus, metformin benefits both moderate obesity in HFHS-fed mice and severe obesity in leptin-deficient ob/ob mice. Metformin stimulates AMPK activity and inhibits transforming growth factor β1 (TGF-β1) signaling in WAT of obese mice As shown in Fig. 4, phosphorylation of AMPK at Thr-172, required for AMPK activation (33), declined ~50% in ewat of both ob/ob mice and HFHS-fed mice and the impairment was restored by metformin, returning to nearly normal levels. Stimulation of fat AMPK activity was further confirmed by concomitantly increased phosphorylation of ACC at Ser79, a well-known substrate of AMPK, consistent with metformininduced AMPK in insulin-resistant cells (6; 7; 34) and in the livers of diabetic mice (8; 35). The fibrogenic regulator TGF-β binds to TGF-β receptor I and II and subsequently phosphorylates and activates downstream effectors, Smad2 and Smad3, which translocate into 9

10 Diabetes Page 10 of 51 the nucleus and transcriptionally activate fibrogenic target genes (36-38). Because TGF-β1 is implicated in the progression of liver and heart fibrosis (36-38), we determined whether adipose tissue AMPK could be functionally linked to the modulation of TGF-β1. In agreement with elevated plasma TGF-β1 levels in obese humans (39), mrna amounts of TGF-β1 were increased 5.0- and 3.4-fold in ewat of ob/ob mice and HFHS-fed mice, respectively. Consistently, positive staining for TGF-β1 was highly elevated and located in hypertrophic adipocytes and other cells in the stromal space of ob/ob mice. Strong positive staining for Ser- 423/425 phosphorylation of Smad3, an indicator of TGF-β1 activity, was primarily located in nuclei of adipocytes and other cells surrounding adipocytes, which was overlapped on areas of TGF-β1-positive cells. The induction of TGF-β1 expression and Smad3 phosphorylation were evident in ewat and iwat of HFHS-fed mice. Conversely, TGF-β1 activity, as determined by phosphorylation and nuclear translocation of Smad3, was inhibited by metformin. Taken together, metformin activates AMPK, inactivates TGF-β1/Smad3 signaling, and ameliorates fat ECM accumulation in obesity. MMP-mediated degradation of the ECM not only facilitates tissue remodeling but also generates bioactive cleaved peptides and releases chemokines that promote inflammation and apoptosis (31). The crown-like structure (CLS) positively stained by MAC-2, a lectin that is expressed by activated macrophages (40), was dramatically increased in ob/ob mice (Fig. S3A-B). Accordingly, an apoptotic phenotype, as determined by cleaved caspase-3, a key effector that executes the apoptotic program (41), was predominantly located in adipocytes surrounded by CLS (Fig. 4O-P). Consistent with decreased fat inflammatory response in ob/ob mice lacking collagen VI (2), the inflammatory response and adipocyte apoptotic phenotypes in ob/ob mice 10

11 Page 11 of 51 Diabetes were attenuated by metformin. Interestingly, mrna levels of pro-inflammatory cytokines, including tumor necrosis factor-α, (TNF-α), interleukin-1 (IL-1) β, and monocyte chemotactic protein (MCP)-1, were downregulated ~60% by metformin (Fig. S3C). Taken together, metformin decreases TGF-β/Smad3 signaling, and relieves continuous and progressive ECM remodeling in visceral adipose tissue of obesity, which are associated with repressed macrophage recruitment and adipocyte apoptosis. Aberrant ECM deposition and pericellular fibrosis in WAT of obese humans are associated with AMPK inhibition and TGF-β1 induction To gain direct evidence for the relationship among AMPK, TGF-β1, and ECM components of fibrotic fat tissue in obese humans, histological and pathological changes in omental fat depots were characterized and compared between obese subjects (n = 8) and age-matched normal weight subjects (n = 10). As shown in Fig. 5, H&E staining revealed that the pathologically expanded adipose tissue with much greater adipocyte areas of obese subjects than those of control subjects. Masson s Trichrome and Sirius Red staining showed elevated collagen fibers around adipocytes of obese adipose tissue with a higher percentage fibrotic area, revealing fat fibrosis as a pathological process that likely reduces tissue plasticity and function in human obesity. Notably, pericellular fibrotic areas of obese subjects were evident in both macrophage-rich and non-macrophage-rich areas. As previously described (42), immunofluorescent analysis showed increased collagen I staining around adipocytes in visceral adipose tissue from obese patients, but the minimal staining was seen in lean subjects. In agreement with elevated fat LOX in ob/ob mice (Fig. 2A), the upregulation of LOX was seen in omental fat of obese humans. The distribution patterns of positive staining for LOX was similar to those of collagen I staining, indicating that activation of LOX-mediated 11

12 Diabetes Page 12 of 51 collagen cross-linking may create an insoluble fibrotic matrix of collagen fibers in omental fat depots. Furthermore, immunohistochemistry of adjacent fat sections for the expression and distribution of TGF-β1 and α-smooth muscle actin (α-sma), a well-recognized marker of myofibroblasts, was conducted. TGF-β1-positive cells were highly increased and primarily located in adipocytes and other cell types of fibrotic adipose tissue of obese subjects. Some inappropriately activated α-sma-positive myofibroblast-like cells were observed. Strikingly, the distribution of TGF-β1-positive cells was overlapped in that of myofibroblast-like cells, suggesting a possible role of TGF-β1 in the differentiation of myofibroblast-like cells that were presumably derived from resident fibroblasts and other cells. Moreover, the expression and distribution of MMP-9- and TIMP-1-positive cells were increased in obese subjects, suggesting that excessive ECM degradation and remodeling may contribute to the pathological destruction of normal matrix structure of fat depots. Similarly, the accumulated ECM deposition was associated with elevated apoptotic adipocytes, as reflected by increased staining for cleaved caspase-3. Similar to inactivation of AMPK in obesity-induced insulin resistant mice (Fig. 4), phosphorylation of AMPK was diminished in WAT of obese individuals without affecting AMPKα. Fasting glucose levels were elevated in obese subjects as compared to controls. Collectively, suppression of AMPK and overproduction of TGF-β1 contribute to excessive fat ECM remodeling in obese patients, characterized by collagen deposition, myofibroblast-like cell activation, and adipocyte apoptosis. Pharmacological activation of AMPK inhibits TGF-β1-induced ECM expression in primary murine SVF cells Adipose tissue, in addition to adipocytes, contains stromal vascular fraction (SVF) cells including preadipocytes, fibroblasts, vascular cells, and immune 12

13 Page 13 of 51 Diabetes cells. Given hyperactivation of adipose tissue TGF-β1 in obese mice and humans (Fig. 4-5), primary SVF cells isolated from epididymal fat pads of lean C57BL/6 mice were treated with TGF-β1, in an effort to develop an in vitro model that mimics in vivo conditions of obesityinduced fat fibrosis. As shown in Fig. 6A-I, mrna amounts of collagen I, III, and VI were markedly increased 6- to 8-fold over the basal level upon TGF-β1 treatment. This rise was paralleled by the elevation of fibronectin and α-sma. Like MMP-2, expression of MMP-9 was stimulated over 10-fold by TGF-β1, and the induction closely correlated with an elevation of TIMP-1, but not of TIMP-2. We further determined whether AMPK regulates the ECM homeostasis in a cell autonomous manner. Except the complete inhibition of Col6a3, the stimulation of fibrogenesis by TGF-β1 was dose-dependently reduced by metformin, to an extent to ~ 50% levels of TGF-β1-treated cells. In accord with repression of collagen synthesis, metformin had an inhibitory effect on ECM degradation regulators (MMP-2, MMP-9 and TIMP- 1), consistent with the in vivo effect of metformin (Fig. 2). AICAR, an analog of adenosine monophosphate (AMP) that activates AMPK, also reduced fibrogenic response, to an extent similar to that of metformin. A specific activator of AMPK, A , repressed induction of collagen and TIMP-1 by TGF-β, even though to a lesser extent. Collectively, AMPK agonists attenuate TGF-β1-induced fibrogenic response. Metformin inhibits TGF-β1 activation of myofibroblast-like cells in vitro To identify a possible source of fibrogenic cells within the obese adipose tissue, the transformation of myofibroblast-like cells derived from cell origins of the SVF population was assessed upon TGF-β1 induction. As shown in Fig. 6J-L, immunocytochemical analysis showed that in the absence of TGF-β1, α-sma + cells were negligible; exposure to TGF-β1 led to an elevation in α- 13

14 Diabetes Page 14 of 51 SMA + cells. After a 96-h incubation of TGF-β1, uniformly spindle-shaped α-sma + cells were highly increased, presuming that these cells are likely differentiated from resident fibroblasts and other cell populations in primary SVF cells. Strikingly, the amounts of α-sma + myofibroblastlike cells were reduced by metformin. The reduction of α-sma + cells and collagen production also coincided with the suppression of LOX. Therefore, metformin attenuates the development of myofibroblast-like cells that constitute a pool of collagen-producing cells in fibrotic adipose tissue. A constitutively active form of AMPK (CA-AMPK) is sufficient to reduce TGF-β1 activation of fibrogenesis in primary SVF cells To further characterize the role of AMPK in modulating ECM homeostasis, primary adipose tissue SVF cells were transduced with adenoviruses expressing control GFP or constitutively active AMPK (CA-AMPK), a truncated mutation of AMPKα1 [α1 (1 312) T172D] that retains significant kinase activity without requiring its interaction with β- and γ-subunits (7). Overexpression of the CA-AMPK enhanced AMPK activity as described previously (6; 7; 10). Like the metformin action (Fig. 6), overexpression of CA-AMPK effectively suppresses phosphorylation of Smad3 and subsequently mitigates the transcription of fibrogenic genes in response to TGF-β1 (Fig. 7A-H). Metformin and AICAR suppress TGF-β1 induction of fibrogenic response in an AMPKdependent manner in primary SVF cells The results described above suggest that activation of AMPK, either through pharmacological or genetic means, exerts an anti-fibrotic effect. To rigorously illustrate the requirement of AMPK for metformin actions, an adenovirus expressing the dominant-negative AMPK (DN-AMPK) (6; 7; 10; 34) was transduced into primary SVF 14

15 Page 15 of 51 Diabetes cells. As shown in Fig. 7I-N, Metformin and AICAR prevented TGF-β1-induced overproduction of collagens; their effects were largely abolished by the DN-AMPK. The inhibitory effect of metformin and AICAR on MMP-9 and TIMP-1 was almost completely diminished by the DN- AMPK, highlighting AMPK as a negative regulator of fibrogenesis. The data demonstrate that metformin limits excess ECM accumulation in response to TGF-β1 at least partially through AMPK. AMPK reduces TGF-β1/Smad3 signaling and improves insulin sensitivity in adipocytes Because adipose tissue TGF-β1 is implicated in the pathogenesis of obesity-induced metabolic dysfunction in humans (39), we hypothesized that adipocytes might be involved in this pathological process. To this end, differentiated 3T3L1 adipocytes were treated with TGF-β1 in the absence and presence of AMPK activators. As shown in Fig. 8, metformin largely decreased TGF-β1-stimulated phosphorylation of Smad3. Metformin-treated adipocytes were also resistant to TGF-β1-induced apoptosis, as evidenced by reduced cleaved caspase 3. Like SVF cells (Fig. 7), CA-AMPK also blocked induction of fibrogenic genes caused by TGF-β1 in the 3T3L1 adipocytes, suggesting the effect of AMPK on fibrogenesis occurs in different cells of adipose tissue. Furthermore, TGF-β1 suppressed insulin-induced phosphorylation of Akt in adipocytes, suggesting a causal relationship between the reduction of WAT fibrosis and insulin resistance in response to metformin. Conversely, metformin or AICAR had similar effects to attenuate TGFβ1-induced insulin resistance. As an independent assessment of insulin sensitivity, insulinmediated suppression of basal- or isoproterenol-stimulated lipolysis was measured by assessing glycerol release from 3T3-L1 adipocytes. TGF-β1 largely abolished insulin responsiveness, representing a state of insulin resistance in 3T3-L1 adipocytes. Co-treatment with AMPK 15

16 Diabetes Page 16 of 51 activators restored the ability of insulin to inhibit lipolysis in adipocytes exposed to TGF-β1. Additionally, the ability of metformin to enhance insulin action was recapitulated by the CA- AMPK. Consistently (Fig. 4), impaired phosphorylation of Akt in ewat of ob/ob mice was substantially counteracted by metformin. Thus, AMPK appears to inhibit TGF-β1 activity and enhance insulin sensitivity in adipocytes. 16

17 Page 17 of 51 Diabetes Discussion The present study provides the first biochemical evidence that metformin suppresses the TGFβ1-induced fibrogenic response at least partially through AMPK in vitro, which may explain the beneficial effects of metformin on fat fibrosis and systemic insulin resistance in obesity in vivo (Fig. 8K). In visceral white adipose tissue of obese humans, suppression of AMPK and activation of TGF-β1/Smad3 contribute to persistent and aberrant ECM remodeling, characterized by excess collagen synthesis and deposition, LOX-related collagen crosslink, SMA + myofibroblast-like cell activation, adipocyte death, and MAC-2 + macrophage recruitment. In genetic and diet-induced obese mice, metformin stimulates AMPK, reduces TGF-β1 induction and periadipocyte fibrosis, and improves local and systemic insulin resistance. Pharmacological and molecular activation of AMPK effectively inhibit TGF-β1-dependent fibrogenesis and enhance insulin sensitivity in vitro. The concomitant dysregulation of AMPK and TGF-β1/Smad3 signaling contributes to excessive ECM remodeling in WAT of obese humans and animals The most important implication of the present study is the characterization of a previously unrecognized role of AMPK on omental adipose fibrogenesis in obese humans, Pathological changes include insoluble fibrotic matrix of collagen fibers with activation of LOX-mediated collagen crosslinking as well as an elevation of SMA + cells and apoptotic adipocytes close to the fibrosis bundle. Studies by the Ruderman group also indicate that defective AMPK in visceral adipose tissue correlates with insulin resistance in morbidly obese patients (43). Increased fat mass and collagen deposition, as assessed by trichrome staining, hydroxyproline content, and fibrogenic 17

18 Diabetes Page 18 of 51 gene expression, are likely attributed to AMPK inhibition in ob/ob mice, as AMPKα1- or AMPKα2- deficient mice exhibit adiposity and metabolic dysfunction in their adipose tissues (44; 45). Expounding on previous observations that elevated plasma TGF-β1 correlates with adiposity in obese patients (39), the present study reveals TGF-β1 hyperactivation may contribute to the initiation of ECM remodeling in WAT of obese mice and humans. It is worth noting that ob/ob mice develop less fibrosis in response to chronic toxic liver injury than wildtype mice (46). Further research is required to understand the tissue-specific role of leptin in regulating ECM homeostasis. Since TGF-β1 positive staining was evident in both macrophage infiltrated areas and non-macrophage-rich areas in WAT of ob/ob mice, we cannot exclude the secondary effect of fat inflammation on fat fibrosis, However, our in vitro studies delineate that TGF-β1 acts to induce adipocyte insulin resistance and stimulate myofibroblast-like cells, a major source of excessive ECM production, which is widely shared by the pathogenesis of liver fibrosis (29; 30). These studies support the proposed modeling: the dysregulation of AMPK and TGF-β1 plays a causal role in promoting pathologically uncontrolled adipose tissue expansion and insulin resistance. Future studies are required to elucidate the precise role of each cell type from SVF in fibrotic fat tissue. Metformin treatment: a potential applicant for the improvement of adipose fibrosis and dysfunction in obesity While metformin has been the mainstay of anti-diabetic therapy for over 50 years (4; 5), its underlying mechanisms are not well defined. One major finding of the present study is the identification of a novel function of metformin a switch of visceral fat to subcutaneous fat properties, i.e., decreased fibrogenic and inflammatory responses, which mirrors anti-fibrotic phenotypes in ob/ob mice lacking collagen VI (2). Fibrogenic cell phenotypes, e.g., TGF-β1/Smad3 stimulation, LOX induction, α-sma-expression, and collagen 18

19 Page 19 of 51 Diabetes induction in obese fat, are lessened by metformin. Since DN-AMPK abolishes the anti-fibrotic effect of metformin in SVF cells, metformin represses TGF-β1-induced fibrogenesis in vitro via an AMPK-dependent mechanism. Decreased fat fibrosis and lowered fasting blood glucose levels are highly correlated in metformin-treated mice. Metabolic improvements seen in metformin-treated mice may be a secondary consequence of reduced fat mass and inflammation. Fibrosis also negatively affects human adipocyte function via mechanosensitive molecules involving connective tissue growth factor (CTGF) and the mechanosensitive transcriptional complex YAP/TEAD (42). Further studies will investigate whether CTGF and YAP may be involved in the role of metformin in limiting fat fibrogenesis in human obesity. Therapeutically, activating AMPK by metformin may ameliorate pericellular fibrosis of white adipose tissue, systemic metabolic dysregulation, and insulin resistance. Future studies will explore whether AMPK-targeting agents, like metformin, can be used to design new clinical trials to improve the health of patients with obesity or other fibrotic diseases. In conclusion, the convergence of AMPK inhibition and TGF-β1/Smad3 activation may contribute to the development of WAT fibrosis and insulin resistance, leading to the metabolic unfavorable microenvironment in obese humans. Targeting AMPK represents a potential therapeutic approach to combat adipose tissue fibrosis and metabolic dysfunction associated with obesity. 19

20 Diabetes Page 20 of 51 Acknowledgments Author Contribution: Dr. Mengwei Zang is the guarantor of this work and has 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. T. L., A.N., and M.Z. contributed to experimental design and research data; A.N., J.F., A.S., X.R., X.J.X., B.J. contributed to research data and provided technique assistance; B.J. and Q.L. provided reagents and materials; T.L., J.F., B.J., X.J.X, J.H. Y.Y., and Q.Y. contributed to the discussion and reviewed the manuscript; M.Z. obtained the funding and wrote the manuscript. This work has been supported by the National Institutes of Health Grants RO1 DK076942, R01 DK100603, and R21AA (to M.Z.), and the American Diabetes Association Award 1-15-BS-216 (to M.Z.). M.Z. is also the recipient of the Boston University Clinical and Translational Science Institute fund (1UL1TR001430). We are grateful to Dr. Susan K Fried at Boston Nutrition Obesity Research Center (BNORC P30DK046200) for important advice in adipocyte work. We also thank Dr. Richard A. Cohen at Boston University School of Medicine for insightful discussion. We thank Jinyan Zang at Harvard University for editorial assistance. 20

21 Page 21 of 51 Diabetes References 1. Sun K, Kusminski CM, Scherer PE: Adipose tissue remodeling and obesity. The Journal of clinical investigation 2011;121: Khan T, Muise ES, Iyengar P, Wang ZV, Chandalia M, Abate N, Zhang BB, Bonaldo P, Chua S, Scherer PE: Metabolic dysregulation and adipose tissue fibrosis: role of collagen VI. Molecular and cellular biology 2009;29: Divoux A, Tordjman J, Lacasa D, Veyrie N, Hugol D, Aissat A, Basdevant A, Guerre-Millo M, Poitou C, Zucker JD, Bedossa P, Clement K: Fibrosis in human adipose tissue: composition, distribution, and link with lipid metabolism and fat mass loss. Diabetes 2010;59: He L, Wondisford Fredric E: Metformin Action: Concentrations Matter. Cell Metabolism 21: Ferrannini E: The target of metformin in type 2 diabetes. The New England journal of medicine 2014;371: Zang MW, Xu SQ, Maitland-Toolan KA, Zuccollo A, Hou XY, Jiang BB, Wierzbicki M, Verbeuren TJ, Cohen RA: Polyphenols stimulate AMP-activated protein kinase, lower lipids, and inhibit accelerated atherosclerosis in diabetic LDL receptor-deficient mice. Diabetes 2006;55: Zang MW, Zuccollo A, Hou XY, Nagata D, Walsh K, Herscovitz H, Brecher P, Ruderman NB, Cohen RA: AMP-activated protein kinase is required for the lipid-lowering effect of metformin in insulin-resistant human HepG2 cells. Journal of Biological Chemistry 2004;279: Fullerton MD, Galic S, Marcinko K, Sikkema S, Pulinilkunnil T, Chen ZP, O'Neill HM, Ford RJ, Palanivel R, O'Brien M, Hardie DG, Macaulay SL, Schertzer JD, Dyck JR, van Denderen BJ, Kemp BE, Steinberg GR: Single phosphorylation sites in Acc1 and Acc2 regulate lipid homeostasis and the insulin-sensitizing effects of metformin. Nat Med 2013;19: Shaw RJ, Lamia KA, Vasquez D, Koo SH, Bardeesy N, Depinho RA, Montminy M, Cantley LC: The kinase LKB1 mediates glucose homeostasis in liver and therapeutic effects of metformin. Science 2005;310: Li Y, Xu SQ, Mihaylova MM, Zheng B, Hou XY, Jiang BB, Park O, Luo ZJ, Lefai E, Shyy JYJ, Gao B, Wierzbicki M, Verbeuren TJ, Shaw RJ, Cohen RA, Zang MW: AMPK Phosphorylates and Inhibits SREBP Activity to Attenuate Hepatic Steatosis and Atherosclerosis in Diet-Induced Insulin-Resistant Mice. Cell Metabolism 2011;13: Daval M, Foufelle F, Ferre P: Functions of AMP-activated protein kinase in adipose tissue. J Physiol-London 2006;574:

22 Diabetes Page 22 of Gauthier MS, Miyoshi H, Souza SC, Cacicedo JM, Saha AK, Greenberg AS, Ruderman NB: AMP-activated protein kinase is activated as a consequence of lipolysis in the adipocyte - Potential mechanism and physiological relevance. Journal of Biological Chemistry 2008;283: Xu XJ, Gauthier MS, Hess DT, Apovian CM, Cacicedo JM, Gokce N, Farb M, Valentine RJ, Ruderman NB: Insulin sensitive and resistant obesity in humans: AMPK activity, oxidative stress, and depot-specific changes in gene expression in adipose tissue. Journal of Lipid Research 2012;53: Virtanen KA, Hallsten K, Parkkola R, Janatuinen T, Lonnqvist F, Viljanen T, Ronnemaa T, Knuuti J, Huupponen R, Lonnroth P, Nuutila P: Differential effects of rosiglitazone and metformin on adipose tissue distribution and glucose uptake in type 2 diabetic subjects. Diabetes 2003;52: Aune UL, Ruiz L, Kajimura S: Isolation and differentiation of stromal vascular cells to beige/brite cells. Journal of visualized experiments : JoVE 2013;73:doi: / Wang QA, Scherer PE, Gupta RK: Improved methodologies for the study of adipose biology: insights gained and opportunities ahead. J Lipid Res 2014;55: Madiraju AK, Erion DM, Rahimi Y, Zhang X-M, Braddock DT, Albright RA, Prigaro BJ, Wood JL, Bhanot S, MacDonald MJ, Jurczak MJ, Camporez J-P, Lee H-Y, Cline GW, Samuel VT, Kibbey RG, Shulman GI: Metformin suppresses gluconeogenesis by inhibiting mitochondrial glycerophosphate dehydrogenase. Nature 2014;510: Miller RA, Chu Q, Xie J, Foretz M, Viollet B, Birnbaum MJ: Biguanides suppress hepatic glucagon signalling by decreasing production of cyclic AMP. Nature 2013;494: Li Y, Xu SQ, Giles A, Nakamura K, Lee JW, Hou XY, Donmez G, Li J, Luo ZJ, Walsh K, Guarente L, Zang MW: Hepatic overexpression of SIRT1 in mice attenuates endoplasmic reticulum stress and insulin resistance in the liver. Faseb Journal 2011;25: Sun K, Park J, Gupta OT, Holland WL, Auerbach P, Zhang N, Goncalves Marangoni R, Nicoloro SM, Czech MP, Varga J, Ploug T, An Z, Scherer PE: Endotrophin triggers adipose tissue fibrosis and metabolic dysfunction. Nature communications 2014;5:doi: /ncomms Matsui Y, Hirasawa Y, Sugiura T, Toyoshi T, Kyuki K, Ito M: Metformin reduces body weight gain and improves glucose intolerance in high-fat diet-fed C57BL/6J mice. Biol Pharm Bull 2010;33: Desilets AR, Dhakal-Karki S, Dunican KC: Role of metformin for weight management in patients without type 2 diabetes. The Annals of pharmacotherapy 2008;42:

23 Page 23 of 51 Diabetes 23. Tiikkainen M, Hakkinen AM, Korsheninnikova E, Nyman T, Makimattila S, Yki-Jarvinen H: Effects of rosiglitazone and metformin on liver fat content, hepatic insulin resistance, insulin clearance, and gene expression in adipose tissue in patients with type 2 diabetes. Diabetes 2004;53: Jeong WI, Park O, Gao B: Abrogation of the antifibrotic effects of natural killer cells/interferon-gamma contributes to alcohol acceleration of liver fibrosis. Gastroenterology 2008;134: Yang L, Chan C-C, Kwon O-S, Liu S, McGhee J, Stimpson SA, Chen LZ, Harrington WW, Symonds WT, Rockey DC: Regulation of peroxisome proliferator-activated receptor-γ in liver fibrosis. American Journal of Physiology - Gastrointestinal and Liver Physiology 2006;291:G902-G Halberg N, Khan T, Trujillo ME, Wernstedt-Asterholm I, Attie AD, Sherwani S, Wang ZV, Landskroner-Eiger S, Dineen S, Magalang UJ, Brekken RA, Scherer PE: Hypoxia-inducible factor 1alpha induces fibrosis and insulin resistance in white adipose tissue. Molecular and cellular biology 2009;29: Cox TR, Bird D, Baker AM, Barker HE, Ho MW, Lang G, Erler JT: LOX-mediated collagen crosslinking is responsible for fibrosis-enhanced metastasis. Cancer Res 2013;73: Cox TR, Erler JT: Remodeling and homeostasis of the extracellular matrix: implications for fibrotic diseases and cancer. Disease models & mechanisms 2011;4: Wynn TA: Common and unique mechanisms regulate fibrosis in various fibroproliferative diseases. The Journal of clinical investigation 2007;117: Schuppan D, Kim YO: Evolving therapies for liver fibrosis. The Journal of clinical investigation 2013;123: Hu J, Van den Steen PE, Sang QX, Opdenakker G: Matrix metalloproteinase inhibitors as therapy for inflammatory and vascular diseases. Nature reviews Drug discovery 2007;6: Chavey C, Mari B, Monthouel MN, Bonnafous S, Anglard P, Van Obberghen E, Tartare- Deckert S: Matrix metalloproteinases are differentially expressed in adipose tissue during obesity and modulate adipocyte differentiation. The Journal of biological chemistry 2003;278: Kahn BB, Alquier T, Carling D, Hardie DG: AMP-activated protein kinase: Ancient energy gauge provides clues to modern understanding of metabolism. Cell Metabolism 2005;1: Hou X, Xu S, Maitland-Toolan KA, Sato K, Jiang B, Ido Y, Lan F, Walsh K, Wierzbicki M, Verbeuren TJ, Cohen RA, Zang M: SIRT1 regulates hepatocyte lipid metabolism through activating AMP-activated protein kinase. JBiolChem 2008;283:

24 Diabetes Page 24 of Zhou GC, Myers R, Li Y, Chen YL, Shen XL, Fenyk-Melody J, Wu M, Ventre J, Doebber T, Fujii N, Musi N, Hirshman MF, Goodyear LJ, Moller DE: Role of AMP-activated protein kinase in mechanism of metformin action. Journal of Clinical Investigation 2001;108: Heldin CH, Miyazono K, ten Dijke P: TGF-beta signalling from cell membrane to nucleus through SMAD proteins. Nature 1997;390: Massague J, Blain SW, Lo RS: TGFbeta signaling in growth control, cancer, and heritable disorders. Cell 2000;103: Wakefield LM, Roberts AB: TGF-beta signaling: positive and negative effects on tumorigenesis. Current opinion in genetics & development 2002;12: Yadav H, Quijano C, Kamaraju AK, Gavrilova O, Malek R, Chen W, Zerfas P, Zhigang D, Wright EC, Stuelten C, Sun P, Lonning S, Skarulis M, Sumner AE, Finkel T, Rane SG: Protection from obesity and diabetes by blockade of TGF-beta/Smad3 signaling. Cell Metab 2011;14: Ho MK, Springer TA: Mac-2, a novel 32,000 Mr mouse macrophage subpopulation-specific antigen defined by monoclonal antibodies. Journal of immunology (Baltimore, Md : 1950) 1982;128: Porter AG, Janicke RU: Emerging roles of caspase-3 in apoptosis. Cell death and differentiation 1999;6: Pellegrinelli V, Heuvingh J, du Roure O, Rouault C, Devulder A, Klein C, Lacasa M, Clement E, Lacasa D, Clement K: Human adipocyte function is impacted by mechanical cues. The Journal of pathology 2014;233: Xu XJ, Gauthier MS, Hess DT, Apovian CM, Cacicedo JM, Gokce N, Farb M, Valentine RJ, Ruderman NB: Insulin sensitive and resistant obesity in humans: AMPK activity, oxidative stress, and depot-specific changes in gene expression in adipose tissue. J Lipid Res 2012;53: Zhang W, Zhang X, Wang H, Guo X, Li H, Wang Y, Xu X, Tan L, Mashek MT, Zhang C, Chen Y, Mashek DG, Foretz M, Zhu C, Zhou H, Liu X, Viollet B, Wu C, Huo Y: AMP-activated protein kinase alpha1 protects against diet-induced insulin resistance and obesity. Diabetes 2012;61: Villena JA, Viollet B, Andreelli F, Kahn A, Vaulont S, Sul HS: Induced adiposity and adipocyte hypertrophy in mice lacking the AMP-activated protein kinase-alpha2 subunit. Diabetes 2004;53: Leclercq IA, Farrell GC, Schriemer R, Robertson GR: Leptin is essential for the hepatic fibrogenic response to chronic liver injury. J Hepatol 2002;37:

25 Page 25 of 51 Diabetes Figure Legends Fig. 1. Metformin treatment reduces body weight gain, decreases adiposity, and relives excessive collagen deposition in white adipose tissue (WAT) of genetically obese ob/ob mice. A-B. Administration of metformin (250 mg/kg/day) for 4 weeks causes a moderate but significant body weight loss in ob/ob mice. Mice were divided into three groups: lean littermates (n = 15), ob/ob mice (n = 11), and ob/ob mice treated with metformin (ob/ob +Metformin, n = 10). C-D. Body composition in three groups of the mice. E. Representative Hematoxylin-Eosin (H&E) staining of epididymal adipose tissue (ewat) of the mice. F. Average adipocyte areas of ewat in H&E-stained sections calculated using ImageJ software were shown. G. The bar graphs represent adipocyte size distribution in ewat of three groups of mice. Adipocyte size distribution was determined from 5-6 mice per group (>500 cells counted per group). H-I. Representative Masson s Trichrome staining (collagen as blue, nuclei as black, and cytoplasm as red) and Sirius Red staining (collagen as red) show the fibrotic structural characteristics in epididymal fat depots of obese mice. Collagen fibers were organized in bundles of various thicknesses. Thinner collagen fibrils were also seen in the interstitial space between adipocytes, demonstrating pericellular fibrosis in fat depots. Thicker collagen fibers surrounding hypertrophic adipocytes were located in macrophage-abundant areas (blue arrow) in the stromal space of ob/ob mice. Notably, interstitial deposition of collagen fibrils around adipocytes was also evident in non-macrophage-rich regions (green arrow) of ob/ob mice. Original magnification 10 and 20. J-K. The bar graphs represent relatively positive staining areas to total tissue areas calculated by Image J software from NIH. The data are presented as the mean ± SEM, n = 7, *P < 0.05, vs. control mice; # P < 0.05, vs. ob/ob mice. L. Hydroxyproline content of epididymal fat, an indicator of total collagen content, was measured. M. Metformin lowers fasting blood glucose levels in ob/ob mice. N. Areas of collagen-positive staining in WAT are correlated with fasting glucose levels in ob/ob mice treated or untreated with metformin (R 2 = , P = ). O. Glucose tolerance tests (GTT, i.p., glucose 2 g/kg) in mice following a 16-h fast. P. Areas under the curve (AUC) for GTT were estimated by summing the numerical integration values of successive linear segments of the glucose disposal curve above baseline glucose, n = 7-8. Q. Insulin tolerance tests (ITT, i.p., insulin 0.75 U/kg) in mice following a 6-h fast. R. Summed ITT results from Q, n = 5-6. Fig. 2. Metformin administration ameliorates abnormal synthesis and degradation of ECM components in WAT of ob/ob mice. A. The transcription of fibrogenic genes encoding collagen I (Col1a1), III (Col3a1), VI (Col6a3), fibronectin, and the collagen cross-linking enzyme LOX in epididymal fat are induced in ob/ob mice and suppressed by metformin. B-C. The mrna levels of MMP-2, MMP-9, MMP-13, and MMP-14 as well as TIMP-1, TIMP-2, and TIMP-3 in ewat of three groups of mice are measured by qrt-pcr. D-I. Identification of possible ECM regulators involved in fat fibrosis. Representative immunohistochemical staining and immunoblots for MMP-9 (D-F) and TIMP-1 (G-I) are shown. Positive immunostaining appeared a brown color, similar to a collagen distribution pattern in fat depots. Notably, positive staining for MMP-9 and TIMP-1 was visualized mainly in some of adipocytes (red arrow) and other cells (blue arrow) in the stromal space of ob/ob mice. The induction of MMP-9 and TIMP- 1 was decreased in metformin-treated mice. The minimal staining was noted in lean littermates. There was no detectable staining with a nonspecific IgG isotype control in the adjacent adipose tissue sections (data not shown), indicating the specificity of MMP-9 and TIMP-1 staining. 25

26 Diabetes Page 26 of 51 Original magnification 20 and 40. The bar graphs represent relatively positive staining areas to total tissue areas. The data are presented as the mean ± SEM, n = 7. *P < 0.05, vs. lean mice; # P < 0.05, vs. ob/ob mice. Fig. 3. Metformin treatment inhibits abnormal interstitial deposition of collagens in WAT of high fat, high sucrose (HFHS) diet-induced obese mouse model. Mice were fed normal chow diet (ND, n = 7), HFHS diet (n = 7), or HFHS diet supplemented with metformin (HFHS + Metformin, n = 7). Metformin-treated mice were initially fed on a HFHS diet for 8 weeks, followed by HFHS feeding plus metformin treatment (250 mg/kg/day) for additional 4 weeks. A- B. Representative H&E staining of epididymal and inguinal white adipose tissue (ewat and iwat). C-D. Representative Masson s Trichrome staining of ewat (C) and iwat (D) is shown. Notably, interstitial deposition of collagen fibrils around adipocytes was evident in both macrophage-rich (blue arrow) and non-macrophage-rich regions (green arrow). E. Representative immunofluorescent staining for LOX (red) and nuclear staining with DAPI (blue). Notably, immunostaining for LOX displayed a pattern similar to the collagen distribution of obese mice (green arrows). Original magnification 10 or 20. F. Administration of metformin reduces adiposity in C57BL/6 mice fed with the HFHS diet. The average adipocyte areas of ewat were measured. G. The bar graphs represent relative positive staining areas to total tissue areas, the mean ± SEM, n = 7. *P < 0.05, vs. normal chow diet mice; # P < 0.05, vs. HFHS-fed mice. H. Hydroxyproline content analysis of epididymal fat depots. I. Metformin lowers fasting blood glucose levels in HFHS-fed mice. J. Hydroxyproline levels of ewat are correlated with fasting blood glucose levels in ob/ob mice treated with or without metformin (R 2 = , P = ). K. Metformin inhibits expression of fibrogenic genes including collagen I, III, and VI, fibronectin, or LOX in ewat and iwat of HFHS-fed mice. L-M. Metformin suppresses the induction of fibrogenic regulators in ewat and iwat of obese mice, the mean ± SEM, n = 5-6. Fig. 4. Metformin stimulates AMPK activity and inhibits TGF-β1/Smad3 signaling in WAT of obese mice. A-D. The phosphorylation of AMPK and its downstream target, ACC, is decreased in epididymal fat depots of ob/ob mice and this impairment is restored by metformin. The data are presented as the mean ± SEM, n = 5-7. *P < 0.05, vs. control mice; # P < 0.05, vs. obese mice. E-F. The phosphorylation of AMPK is stimulated in the ewat of metformintreated obese mice. G-H. The induction of TGF-β1 is suppressed in ewat or iwat of metformin-treated mice either on ob/ob background or with a HFHS diet. I-J. Representative immunohistochemical staining of epididymal fat sections for TGF-β1. Strong staining for TGFβ1 was visualized mainly in adipocytes (red arrows) and other cells (blue arrows) in the WAT of ob/ob mice and this elevation was inhibited by metformin. The data are presented as the mean ± SEM, n = 7. *P < 0.05, vs. control mice; # P < 0.05, vs. obese mice. K-N. Metformin reduces the elevation in phosphorylated Smad3 in ewat and iwat of HFHS-fed mice. Representative immunohistochemical staining of epididymal fat sections and immunoblots for phosphorylation of Smad3 and its nuclear translocation in ob/ob mice. O-P. Metformin reduces dead adipocytes in ob/ob mice. Adipocytes of lean mice were not immunoreactive for cleavage caspase-3, an apoptotic marker, whereas more dead adipocytes (red arrows) surrounded by crown-like structure (blue arrows) in obese mice are immunoreactive for cleavage caspase-3. There was no detectable staining with a nonspecific IgG isotype control in adipose tissue sections (data not shown), indicating the specific staining for TGF-β1 expression, Smad3 phosphorylation, and 26

27 Page 27 of 51 Diabetes cleaved caspase-3. Original magnification 20 and 40. Q-R. Metformin reduces cleaved casepase-3 levels in ewat of HFHS-fed mice. Fig. 5. Pericellular fibrosis in the visceral adipose tissue of obese humans with abnormal ECM accumulation and dead adipocytes is associated with AMPK suppression and TGFβ1 induction. A. Representative H&E staining of omental fat depots in age-matched normal weight subjects (n = 10) and obese subjects (n = 8). B. Masson s Trichrome staining shows minimal staining of collagen fibers in control subjects but numerous collagen fibers (yellow arrows) extended into hypertrophic adipocytes in obese subjects. C. Sirius Red staining shows distinct characteristics of fibril amounts and distribution between control and obese subjects. Large amounts of collagens in the extracellular matrix surrounding hypertrophic adipocytes (blue arrows) were observed in obese subjects. Sirius Red staining could also be visualized by polarized light microscopy, demonstrating type I and type III collagen fibers present in bundles of fibrosis (blue arrows). D-E. Representative immunofluorescent staining for type I collagen (red) and LOX (red) and nuclear staining with DAPI (blue). Notably, immunostaining for type I collagen displayed a pattern similar to LOX distribution in obese subjects (blue arrows). F-I. Representative immunostaining for fibrogenic regulators and components. Omental fat tissue sections shown were serial but not always consecutive. Positive immunostaining appeared a brown color. The expression and distribution patterns of positive staining for TGF-β1 (F) and α- SMA (G) were similar to those for MMP-9 (H) and TIMP-1 (I). Notably, immunostaining for TGF-β1 was located in both adipocytes (red arrows) and macrophage infiltration areas (blue arrows) in obese subjects. Immunostaining for α-sma displayed the intense staining at sites of adipose tissue damage/repair, indicating the proliferation and migration of activated myofibroblast-like cells into impaired adipocytes for the secretion of excess ECM components. Immunostaining for MMP-9 and TIMP-1 also showed intense staining in the regions of macrophage-rich areas and non-macrophage-rich areas, indicating that the different cell types may be involved in collagen deposition in fibrotic adipose tissue. The minimal staining for fibrogenic regulators was present in fat depots of control subjects. J. Representative immunostaining for cleaved caspase-3. Dead adipocytes were highly immunoreactive for cleaved caspase-3 in visceral WAT of obese subjects. There was no detectable staining with a nonspecific IgG isotype control in human adipose tissue sections (data not shown). K. Positive staining for phosphorylated AMPK and total AMPK (green) is localized primarily in most of adipocytes of control subjects and the intensity of phosphorylated AMPK is less evident in obese subjects. Original magnification 20 or 40. L-M. The bar graphs represent adipocyte size and relatively positive staining areas to total tissue areas, the mean ± SEM, n = 7. *P < 0.05, vs. control subjects. N. Fasting glucose levels in control and obese subjects, the mean ± SEM, n = Fig. 6. Pharmacological activation of AMPK suppresses the transcription of TGF-β1- dependent fibrogenic genes in vitro. A-B. A representative immunoblot and densitometry quantification of Smad3 phosphorylation. The cells of the stromal vascular fraction (SVF) were isolated from epididymal fat tissue of lean C57BL/6 mice. The cells were incubated for 24 h with TGF-β1 (2 ng/ml) in the absence or presence of different AMPK activators, such as metformin (1 mm) and AICAR (200 µm). The phosphorylation levels of Smad3 were normalized to endogenous Smad3 levels and expressed as a percentage of those under untreated conditions. Data are presented as the mean ± SEM, n = 3 4. *P < 0.05, vs. untreated cells. # P < 0.05, vs. 27

28 Diabetes Page 28 of 51 TGF-β1-treated cells. C-F. Metformin suppresses the expression of fibrogenesis-related genes in response to TGF-β1. The mrnas encoding fibrogenic components, including different types of collagen (Col1a1, Col3a1, and Col6a3), fibronectin, and α-sma, as well as expression of major fibrogenic regulators, such as MMP-2, MMP-9, TIMP-1 or TIMP-2, were analyzed by real-time PCR. G-H. AICAR represses the transcription of TGF-β1-dependent fibrogenic genes in a dosedependent manner. I. A inhibits the effect of TGF-β1 on fibrogenic genes. J. TGF-β1 induces the transformation of myofibroblast-like cells in a time-dependent manner. Representative immunofluorescence images show α-sma-staining (green) and nuclear staining with DAPI (blue) in cultured SVF cells. K. Treatment with metformin (1 mm) for 24 h inhibits the transformation of myofibroblast-like cells in the SVF cells exposed to TGF-β1. L. Treatment with metformin (1 mm) and AICAR (200 µm) for 24 h inhibits the induction of LOX in the SVF cells exposed to TGF-β1. Representative immunofluorescence images show LOX-staining (red) and nuclear staining with DAPI (blue). Fig. 7. AMPK is necessary for metformin or AICAR to inhibit TGF-β1-dependent fibrogenesis in vitro. A-D. Adenovirus-mediated overexpression of the constitutively active form of AMPK (CA-AMPK) is sufficient to decrease phosphorylation of Smad2 and Smad3, which are well-known downstream effectors of TGF-β1. Primary SVF cells were infected with an adenovirus expressing myc-tagged CA-AMPK (Ad-CA-AMPK) or a control adenovirus (Ad- GFP). Cells were treated with TGF-β1 (2 ng/ml) for another 24 h in serum-free medium. The dose-dependent expression of the CA-AMPK (~31 kda) was detected by immunoblotting with anti-myc antibody. E-H. TGF-β1-induced fibrogenesis is prevented by overexpression of the CA-AMPK. I-J. The stimulatory effect of AICAR on AMPK is diminished by overexpression of a dominant negative form of AMPK (DN-AMPK). Primary SVF cells were infected with an adenovirus expressing the DN-AMPK (Ad-DN-AMPK) or Ad-GFP. Cells were then incubated for another 24 h with TGF-β1 (2 ng/ml) in the absence or presence of metformin (1 mm) or AICAR (200 µm). Overexpression of the DN-AMPK was confirmed by immunoblots with anti- AMPKα antibody. K-L. The effect of AICAR on TGF-β1-induced phosphorylation of Smad3 and TIMP-1 induction is abrogated by the DN-AMPK. M-N. The effect of metformin and AICAR on TGF-β1-dependent expression of fibrogenic genes is diminished by the DN-AMPK. Fig. 8. AMPK produces anti-fibrotic effects and protects against TGF-β1-induced insulin resistance in differentiated 3T3L1 adipocytes. A-C. Metformin represses TGF-β1-induced signaling and cell apoptosis. 3T3-L1 adipocytes on day 10 of differentiation were incubated for 24 h with 2 ng/ml TGF-β1 in the absence or presence of metformin (1 mm). Representative immunoblots and quantification of Smad3 phosphorylation and cleaved caspase-3 are shown. The data are expressed as the fold change over the basal level (mean ± SEM, n = 3-4). *P < 0.05, vs. untreated cells; # P < 0.05, vs. TGF-β1-treated cells. D. Overexpression of the CA-AMPK is sufficient to suppress TGF-β1-induced expression of fibrogenic genes and the collagen crosslinking enzyme LOX. Differentiated 3T3-L1 adipocytes were infected with Ad-CA-AMPK or Ad-GFP for 24 h and subsequently incubated with or without 2 ng/ml TGF-β1 for additional 24 h. The mrna levels of Col1a1, Col6a3, TIMP-1, and LOX were determined by quantitative RT- PCR. E-F. Pharmacological activation of AMPK rescues TGF-β1-induced insulin resistance. Differentiated 3T3-L1 adipocytes were pretreated for 24 h with 2 ng/ml TGF-β1 in the absence or presence of metformin (1 mm) or AICAR (200 µm). Cells were quiesced in serum-free DMEM medium for 5 h and subsequently stimulated with 25 nm insulin for 10 min. The levels 28

29 Page 29 of 51 Diabetes of Akt phosphorylation were normalized to those of endogenous Akt and expressed as the fold change over the basal level, mean ± SEM, n = 3. *P < 0.05, vs. untreated group; # P < 0.05, vs. insulin stimulation group; $ P < 0.05, TGF-β1 treatment and insulin stimulation group. G. AMPK prevents TGF-β1-induced impairment of insulin action in adipocytes. Differentiated 3T3-L1 adipocytes were pretreated for 24 h with 2 ng/ml TGF-β1 in the absence or presence of metformin (1 mm) or AICAR (200 µm). Cells were serum-starved in DMEM containing 0.2% BSA for 2 h and subsequently incubated in nutrient-free KRP containing 4% fatty acid free-bsa for an additional 2 h in the presence or absence of 1 µm isoproterenol and/or 25 nm insulin as indicated. Glycerol levels in medium aliquots were measured and normalized to protein concentrations in whole-cell lysates, the mean ± SEM, n = 4. *P < 0.05, vs. untreated or isoproterenol stimulation group; # P < 0.05, vs. insulin-treated group; $ P < 0.05, TGF-β1- pretreated and insulin-treated group. H. Overexpression of the CA-AMPK is sufficient to restore insulin sensitivity in TGF-β1-treated adipocytes. Differentiated 3T3-L1 adipocytes were infected with Ad-CA-AMPK or Ad-GFP for 24 h. Cells were pretreated with 2 ng/ml TGF-β1 for 24 h and subjected to a glycerol release assay with 1 µm isoproterenol in the absence or presence of 25 nm insulin as indicated. I-J. Metformin enhances insulin signaling in ewat of ob/ob mice, as reflected by increased phosphorylation of Akt. K. The proposed mechanism for the beneficial effect of AMPK and its activator metformin on adipose tissue fibrosis, a hallmark of metabolically challenged adipose tissue in obesity. Adipose tissue fibrosis is characterized by pathologically uncontrolled accumulation of ECM, induction of collagen cross-linking, activation of myofibroblasts, and adipocyte insulin resistance and death in human obesity. Adipose tissue AMPK inhibition is likely linked to TGF-β1-dependent fibrogenesis and ECM remodeling in obesity. Therapeutically, AMPK activation by metformin suppresses TGF-β1- activated fibrogenesis in adipocytes and other-related cells, which in turn resolves adipose tissue fibrosis and dysfunction as well as improves local and systemic insulin resistance caused by obesity. 29

30 Figure 1 Luo T, et al. A E 'l c; 60 # fso ~ 40r ~~Ln ' r::: ~ ~ ~10~X;::: ~ 'l Lean Lean * I ' n I I I ob/ob ob/ob +Metformin ====;:::: ~ B_s ~4 "' 'iii ~ 3.c "' ~ 2 "' '8 1 Ill ob/ob Lean,., ',..., '" 'i- \1.~ /.... * c # g: 30 ~ "' 20 LL "' * # ob/ob ob/ob Lean ob/ob ob/ob Lean ob/ob ob/ob +Metformin +Metformin +Metformin ob/ob + Metformin G ~) ' --., '..,... -~.. D ::0 "'...J ""' ewat D Lean Lean ob/ob ob/ob +Metformin ob/ob H r::: s!/) Cll r E 1.0X e.s:: (,J ;: 1- / \. ) D ob/ob+met L Lean ob/ob ob/ob +Metformin Lean ob/ob ob/ob +Metformin M Q 350 ::i' 300 :!;!.[ 250 "' 0 "::1 Ci "C 0 0 1ii Lean ob/ob ob/ob +Metformin + Lean ob/ob * ob/ob+melformin ~: N Time(min) R ~250.. "' 200 "' 0 g g> 100 ~ ob/ob ob/ob+metformin ~. ~ so , G' ~... t: "C " E E ::1 "' Trichrome Area (% ofwat Area) 20 Lean ob/ob ob/ob +Metformin

31

32

33

34

35

36

37

38 Diabetes Page 38 of 51 Supplemental Experimental Procedures Luo, et al. AMPK Activation by Metformin Suppresses Abnormal Adipose Tissue Extracellular Matrix Remodeling and Ameliorates Insulin Resistance in Obesity Ting Luo 1,2, Allison Nocon 1, Jessica Fry 1, Alex Sherban 1, Xianliang Rui 1, Bingbing Jiang 1, X. Jian Xu 1, Jingyan Han 1, Yun Yan 3, Qin Yang 4, Qifu Li 2, Mengwei Zang 15,6,7 * Animal studies In the first cohort experiment with a genetically obese ob/ob mouse model (1), mice were divided into three groups: lean mice in group 1 (n = 15) consumed a nutritionally replete liquid mouse diet ad libitum (BioServ, Frenchtown, New Jersey); obese ob/ob mice in group 2 (n = 11) were pair-fed the same volume of liquid diet as those in the metformin-treated ob/ob mice in group 3; obese ob/ob mice in group 3 (n = 10) were treated with metformin in the liquid mouse diet at a dose of 250 mg/kg/day as described previously (2-4). In the second cohort experiment with a diet-induced obese mouse model, age-matched, male C57BL/6 mice were divided into three groups: mice in group 1 (n = 7) were fed a chow diet ad libitum for 12 weeks; mice in group 2 (n = 7) were rendered obese by ad libitum feeding of a high fat, high sucrose (HFHS) diet consisting of 36.0% fat, 36.2% carbohydrate, and 20.5% protein (BioServ #F1850, Frenchtown, NJ) for 12 weeks as described previously (1; 5); mice in group 3 (n = 7) were fed on the HFHS diet for 8 weeks, followed by 4 weeks of HFHS feeding plus metformin treatment (250 mg/kg/day) administered via drinking water, which was calculated based on the average daily water intake. All mice were weighed at the beginning of the feeding period and weekly thereafter until killed. At the time of sacrifice, epididymal white adipose tissue (ewat) and inguinal white adipose tissue (iwat) were snap frozen in liquid nitrogen and stored at -80C. For histology and immunohistochemistry, part of the ewat and iwat were fixed in 10% phosphate-buffered formalin acetate, processed, and then embedded in paraffin. All mice were housed in a temperature-controlled environment with a 12-h light/dark cycle. All animal experiments were approved by the University Committee on Use and Care of Animals at Boston University School of Medicine. Body composition analysis Body composition was determined using the body composition analyzer EchoMRI700, a non-invasive quantitative magnetic resonance system (Echo Medical Systems, Houston, TX) in the Metabolic Phenotyping Core at Boston University. Fat mass, lean mass, free water, and total body water were measured in conscious mice as described previously (6; 7). Glucose tolerance tests and insulin tolerance tests Glucose Tolerance Tests (GTTs) and Insulin Tolerance Tests (ITTs) were performed to assess whole-body glucose homeostasis and insulin sensitivity, respectively (1). Blood glucose

39 Page 39 of 51 Diabetes concentrations were measured using the AlphaTrak Blood Glucose Monitoring System. After a 16-h fast, GTTs were performed with an intraperitoneal injection of 10% glucose solution (2 g/kg body weight) into the mice. Blood glucose levels were measured at 0, 15, 30, 60, 90, and 120 min after glucose administration. After a 6-h fast, ITTs were performed with an intraperitoneal injection of human insulin (0.75 U/kg body weight) into the mice. Blood glucose levels were measured at 0, 15, 30, 60, 90, and 120 min after insulin administration. Primary cell culture isolation from stromal vascular fraction (SVF) of adipose tissue Male 6-week-old lean C57BL/6 mice were sacrificed and epididymal fat depots were isolated as described previously (8; 9). Briefly, fat tissues were minced with surgical scissors for 2 min (at 3 openings and closings of blades per second), resulting in small pieces (1 mm 3 ) of adipose tissue. The adipose tissue samples were digested by collagenase (type I, 1 mg/ml, Worthington Biochemical, #LS004194) at a 1:1 ratio of milliliters of digestion solution to grams of adipose tissue under agitation for min at a 37 C water bath. The tissue samples were centrifuged at 700 g for 10 min to separate the stromal cell fraction (SVF) in a brownish pellet on the bottom from the top layer of mature adipocytes. The SVF pellets were collected and suspended in DMEM/F12 supplemented with 10% FBS (Hyclone, Logan, Utah), 100 U/ml penicillin, and 100 µg/ml streptomycin and filtered through a 40 µm diameter cell strainer (BD Falcon) After the centrifugation at 700 g for 10 min, the final SVF pellets were resuspended in DMEM/F12 supplemented with 10% FBS. Two hours after plating cells, cells were washed with PBS twice to remove red blood cells, immune cells, and other contaminants, and fresh media were added. Primary SVF cells between 3-5 passages were used for experiments. Cell treatment and adenoviral Infection Primary SVF cells were isolated and treated as described in Fig. S4. Briefly, cells were treated for 24 h with or without TGF-β1 (2 ng/ml) in the absence or presence of different AMPK activators, such as metformin (1 and 2 mm), AICAR (200 and 500 µm), or A (100 µm) in serum-free DMEM/F12 medium as described previously (5; 10). In some cases, SVF cells were infected with adenoviral vectors encoding GFP (Ad-GFP), a myc-tagged constitutively active form of AMPK (Ad-CA-AMPK), or a myc-tagged dominant negative form of AMPK (Ad-DN-AMPK) (5; 10-12). After adenovirus infection, these cells were quiesced in serum-free media overnight and incubated for 24 h with TGF-β1 (2 ng/ml) in the absence or presence of AMPK activators. Purification and infection of adenoviruses have been previously described (10; 12). 3T3-L1 cells culture and differentiation 3T3-L1 cells (American Type Culture Collection) were cultured in Dulbecco s modified Eagle s medium (DMEM) with 10% Bovine Calf Serum (BCS; Hyclone) in 5% CO 2 as described previously (13-15). Two days post confluence, cells were exposed to DMEM with 10% Fetal Bovine Serum (FBS; Invitrogen) containing 0.5 mm isobutylmethylxanthine (IBMX, Sigma#I5879), 1 µm dexamethasone (Sigma#D4902), and 1 µg/ml insulin (Sigma#I9278). After 2 days, cells were maintained in medium containing FBS only. For adenovirus infection, 3T3-L1 adipocytes were

40 Diabetes Page 40 of 51 transduced with the adenoviruses encoding GFP or the CA-AMPK for 24 h, as described previously (10; 11). For insulin regulation experiments, differentiated 3T3-L1 adipocytes in 6-well plates were serum-starved in DMEM containing 0.2% bovine serum albumin (BSA) for 5 h at 37 C. Cells were incubated in serum-free DMEM containing 2% fatty acid-free BSA with or without 25 nm insulin as indicated in figure legends. Glycerol release assays Differentiated 3T3-L1 adipocytes in 12-well plates were serum-starved in DMEM containing 0.2% bovine serum albumin (BSA) for 2 h at 37 C. Adipocytes were washed in Kreb's Ringer phosphate (KRP) buffer (136 mm NaCl, 4.7 mm KCl, 1 mm NaPO 4, ph 7.4, 0.9 mm MgSO 4, and 0.9 mm CaCl 2 ) and then incubated for additional 2 h at 37 C in 0.5 ml of KRP-4% fatty acid-free BSA in the presence or absence of 1 µm isoproterenol and/or 25 nm insulin as described previously (13; 14). Each treatment condition was performed in duplicate. The glycerol content in the medium was measured colorimetrically by using glycerol reagent (Sigma#F6428) according to the manufacturer's protocol. The optical density at 540 nm was measured using a plate reader (SPETRAmax340 Microplate Spectrophotometer, Molecular Devices Corporation). The cells were washed with cold phosphate-buffered saline (PBS) and lysed in lysate buffer and protein concentrations were determined. Glycerol release levels were normalized to cellular protein content (ug glyceriol/mg protein). Histological Analysis For histology studies, 6-µm thick fat tissue sections from mice and humans were stained with Hematoxylin and Eosin (H&E) as described previously (1; 5). For the quantification of adipocyte size, tissue sections from epididymal and inguinal adipose tissues of mice were stained with H&E. Adipocyte areas were measured from3-4 sections of each mouse and 5-6 mice per group (>500 cells/group) as described previously (15; 16). National Institutes of Health (NIH) Image J software ( was used to measure adipocyte areas, which are represented as the average adipocyte area in µm 2 (16). Each adipocyte size was quantified with the NIH Image J software. Once an image of H&E stained fat tissue sections was opened by Image J, the Set Scale function was selected under the Analyze menu. Then, the scale parameters were set as µm 2. Using the freehand sketch tool in the toolbar, the area of each lipid droplet was encircled. Subsequently, Measure tool was selected under the Analyze menu. The adipocyte areas were calculated and presented as µm 2 of each adipocyte. To assess morphological and fibrotic changes, Masson s Trichrome staining was performed in 6-µm thick cross sections of fat tissues. Following deparaffinization and rehydration, tissue sections were stained by Weigert's iron hematoxylin (Sigma#HT25A) for 10 min at room temperature and rinsed with distilled water for 3 min. Sections were also incubated in Biebrich Scarlet-Acid Fuchsin Solution (Sigma#HT151) for 5 min. After that, sections were incubated in freshly prepared phosphotungstic acid-phosphomolybdic acid solution (Sigma#152 and #153) for 5 min and rinsed three times with distilled water. Sections were transferred into Aniline Blue Solution (Sigma#SLBD6394V) and stained for 10 min. After rinsing three times with distilled water, sections were differentiated in 1% Acetic Acid Solution (Electron Microscopy Sciences # ) for 3 min

41 Page 41 of 51 Diabetes and quickly rinsed with distilled water. After the slides were dehydrated and cleared in xylene, the sections were mounted. Collagen fibers were also stained by Picrosirius Red. Following deparaffinization and rehydration, tissue sections were stained for 1 h in 0.1% (w/v) Picrosirius Red in a saturated aqueous solution of 1.0 % picric acid (Poly Scientific#S2365). After staining, sections were rinsed twice in 1% Acetic Acid Solution (Electron Microscopy Sciences # ) to remove any unbound dye. Following dehydration and clearing with xylene, the sections were mounted. The Picrosirius Red stains were also visualized under polarized light, providing an additional way to visualize fibrillar collagen. Positive staining areas containing fibrillar collagens, as shown either with blue in Masson s Trichrome stain or with bright red in Picrosirius Red stain, as well as whole section areas were quantified by NIH Image J software. Immunohistochemistry Immunohistochemistry of fat tissue sections from mice and humans was performed as described previously (1; 5). After deparaffinization and rehydration, adjacent 6-µm thick fat sections were incubated with 10 mmol/l citric acid (ph 6.0) and heated in a microwave (2 min at 700 W, repeated 3 times) to recover antigenicity. Tissue sections were incubated in 3% H 2 O 2 solution to quench peroxidase for 15 min. Nonspecific binding was blocked with 10% normal goat serum (Vector Laboratories, Burlingame, CA) in phosphate-buffered saline (PBS, ph 7.4) for 30 min at room temperature. The fat tissue sections were incubated with MMP-9 antibody (ab#137867, 1:3000 dilution); TIMP-1 antibody (ab#38978, 1:6000 dilution for mice, 1:3000 dilution for humans); α-sma antibody (Sigma#A2547, 1:5000 dilution); TGF-β1 antibody (CST#3711S, 1:2000 dilution for mice; 1:1000 dilution for humans); phospho-smad3 (Ser423/425) antibody (CST#9514, 1:50 dilution); cleaved caspase-3 (Asp-175) antibody (CST#9661, 1:1000 for mice; 1:500 dilution for humans); and MAC-2 antibody (ab#31707, 1:800 dilution) in PBS with 3% BSA and 0.1% Tween-20 at 4 C overnight. The sections were washed and subsequently incubated at room temperature for 30 min with HRP-conjugated anti-rabbit or anti-mouse IgG secondary antibody (GTVision TM III Detection System/Mo & Rb Kit). Positive immunoreactivity was visualized by a dark brown color of DAB reaction product (GTVision TM III Detection System/Mo & Rb Kit). Sections were counterstained with hematoxylin, cleared with xylene, and mounted. All positive staining was recognized by comparing with the negative controls stained with nonimmune rabbit or mouse isotype control IgG (Vector) under the same conditions. Staining images were captured and digitalized using an Olympus HC5000 digital camera attached to an Olympus microscope (Olympus, Tokyo, Japan). Immunofluorescence Adjacent 6-µm thick fat tissue sections from humans were prepared and immunohistofluorescence of tissue sections was performed as described previously (17). After deparaffinization and rehydration, tissue sections were incubated with 10 mmol/l citric acid (ph 6.0) and heated in a microwave (2 min at 700 W, repeated 3 times) to recover antigenicity. Permeabilization was performed in PBS with 0.2% Triton X-100. Nonspecific binding was blocked with 10% normal goat serum in PBS for 30 min at room temperature. Fat sections were incubated with phospho-ampk α (Thr-172) antibody

42 Diabetes Page 42 of 51 (CST#2531, 1:500 dilution), AMPKα 2 antibody (Bathyl#A300, 1:500 dilution), Collagen I antibody (sc#25974, 1:200 dilution), and LOX antibody (EMD Millipore#ABT112, 1:1000) in PBS with 3% BSA and 0.1% Tween-20 at 4 C overnight. Tissue sections were incubated for 1 h at room temperature with Fluorescein (FITC)-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch# , green color, 1:500 dilution), Taxas Red dye-conjugated AffiniPure Rabbit Anti-Goat IgG (Jackson ImmunoResearch # , red color, 1:500 dilution), or Taxas Red dye-conjugated AffiniPure Mouse Anti-Rabbit IgG (Jackson ImmunoResearch## , red color, 1:500 dilution). The cell nuclei were visualized by the mounting solution with DAPI (VECTORSHIELD). The staining signals for phospho-ampkα and AMPK-α2 were specific inasmuch as incubation with non-immuno IgG showing no detectable fluorescence under similar conditions. Additionally, immunofluorescent staining for α-sma in SVF cells was performed to determine α-sma expression. The α-sma staining was visualized with goat-anti mouse IgG conjugated to Fluorescein (FITC) (Jackson ImmunoResearch# , 1:500 dilution). Staining images were captured and digitalized using the Nikon Eclipse DS-Qi1MC digital camera attached to a Nikon Eclipse 80i microscope. Immunoblotting analysis Immunoblotting analysis was conducted as described previously (10; 12). Briefly, adipose tissues or cultured cells were lysed at 4 C in lysis buffer (20 mm Tris-HCl, ph 8.0; 1% (v/v) Nonidet P-40; 150 mm NaCl; 1 mm EDTA; 1 mm EGTA; 1 mm sodium orthovanadate; 25 mm β -glycerolphosphate; 1 mm dithiothreitol; 1 mm phenylmethylsulfonyl fluoride; 2 µg/ml aprotinin; 2 µg/ml leupeptin; and 1 µg/ml pepstatin). Immunoblotting experiments were performed with ~30 µg proteins of adipose tissues or cell lysates. Phosphorylation levels of AMPK (CST#2531), ACC (EMD Millipore #07-303), Akt (CST#9271), Smad3 (CST#9514), or Smad2 (CST3101) were quantified by using scanning densitometry with NIH Image J software, normalized to those of endogenous AMPKα (Bathyl#A300), ACC (CST#3661), Akt (sc#1619), Smad3 (CST#9523), or Smad2 (CST#3101) and presented as relative to the basal or the control level. The protein levels of TIMP-1 (ab#38978) and cleaved caspase-3 (Asp-175) antibody (CST#9661) were quantified, normalized to those of GAPDH, and presented as a value relative to the control. Quantitative RT-PCR Total RNA was extracted from mouse adipose tissue or cultured cells using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer s instruction. One microgram of total RNA was used to synthesize cdna using High Capacity cdna Reverse Transcription Kit (Applied Biosystems, Carlsbad, CA, USA). The transcripts were quantified with StepOnePlus Real-Time PCR System (Applied Biosystems, Carlsbad, CA, USA) by using the SYBR Green PCR master mix and the CT threshold cycle method. Gene expression levels were normalized to those of GAPDH and presented relative to the control (1; 5; 7). The specificity of the PCR amplification was verified by melting curve analysis of the final products and by running products on an agarose gel. The following primers were used:

43 Page 43 of 51 Diabetes Col1a1, TGCTGGTCCCAAAGGTTC (F) and CAGGGCGACCATCTTGAC (R); Col3a1, ATGCCCACAGCCTTCTACAC (F) and ACCAGTTGGACATGATTCACAG (R); Clo6a3, CTGTCGCCTGCATTCATCC (F) and ACAACCCTCTGC ACAAAGTC (R); Fibronectin, TTGAGGAACATGGCTTTAGGC (F) and CTGGGAACATGACCGATTTG (R); α-sma (or Acta2), TCCCAGACATCAGGGAGTAA (F) and TCGGATACTTCAGCGTCAGGA (R); MMP-2, GGGGAGGCTGACATCATG (F) and AGGGTCCACAGCTCATCATC (R); MMP-9, CCCAAAGACCTGAAAACCTC (F) and ACTGCTTCTCTCCCATCATC (R); MMP-13, CCCTTCCCTATGGTGATGATG (F) and TTCTCGGAGCCTGTCAACTG (R); MMP-14, CCTAAGGTGGGCGAGTATGC (F) and TGTCAGCCTGCTTCTCATGTC (R); TIMP-1, ATCTGGCATCCTCTTGTTG (F) and CGCTGGTATAAGGTGGTCTC (R); TIMP-2, GCAGAAGGAGATGGCAAGATG (F) and CGGGGAGGAGATGTAGCAAG (R); TIMP-3, ACACGGAAGCCTCTGAAAGTC (F) and CCACCTCTCCACAAAGTTGC (R); LOX, TCCGCAAAGAGTGAAGAACC (F) and CATCAAGCAGGTCATAGTGG (R) β-actin, CCACAGCTGAGAGGGAAATC (F) and AAGGAAGGCTGGAAAAGAGC (R) Biochemical measurement of tissue hydroxyproline content Total hydroxyproline content in adipose tissue was quantified colorimetrically as described previously (18-20). Briefly, adipose tissue samples (~50 mg) were homogenized in 100 µl of distilled H 2 O. The homogenates were transferred to pressure-tight vial with PTFE-lined cap, and the samples were hydrolyzed by adding 100 µl of concentrated hydrochloric acid (12 N HCl) (1:1, v/v) and subsequently incubated at 110 C for 3-5 h. The hydrolysate was evaporated under vacuum, and the sediment was dissolved in 33% of isopropanol. After the supernatant (50 µl) was transfer to a 96-well plate, 16 µl of freshly prepared chloramine-t/citrate buffer (ph 6.0), which contains g of chloramine-t (Sigma, #C9887) dissolved in 2 ml of 50% n-propanol and 8 ml of acetate-citrate buffer (0.05M, ph6.0), was added into each reaction. After the incubation at room temperature for 4 min, 184 µl of freshly prepared Ehrlich's Reagent Solution, which contains 2.25 g p-dimethylaminobenzaldehyde (DMAB, Aldrich #156477) dissolved in 10 ml of n-propanol (Aldrich ) and 5 ml of perchloric acid (Aldrich #319228), was added into each reaction, followed by the incubation at 65 C for 15 min. After cooling, the optical density at 560 nm was measured in a microplate reader (SpectrMax M5 Microplate Reader, Molecular Devices, LLC). Hydroxyproline content was calculated from a standard curve prepared with high-purity hydroxyproline (Sigma, #H5534) and expressed as µg hydroxyproline per mg of adipose tissue protein. Human studies Omental adipose tissue specimens were obtained from 18 patients (Male = 6, Female = 12) undergoing laparoscopic or open cholecystectomy. Clinical data are presented in Supplemental Table 1. Omental adipose tissue specimens were collected from a total of eight (three male and five female) obese, Chinese patients (BMI: 28.7 ± 1.4 kg/m 2, age ± 2.0 years, range years). Using the same clinical protocol, omental adipose tissue specimens were collected from ten (three males and five females), age-matched normal weight individuals (BMI: 19.8 ± 1.5 kg/m 2, age ± 3.9 years, range years). BMI parameters for Asian adults are different from those applied in the U.S. (21-23): overweight (23.23 BMI 27.5 kg/m 2 ) and obese (BMI 27.5 kg/m 2 ). All subjects underwent laparoscopic or open cholecystectomy, in which 17 subjects suffered from cholecystolithiasis and 1

44 Diabetes Page 44 of 51 subject suffered from cholecystic polypus. All subjects had no metformin administration. As described previously (1; 5), fresh adipose tissue samples were rapidly fixed in 10% phosphate-buffered formalin acetate for 48 h, processed, and embedded in paraffin, from which adjacent 6-µm adipose tissue sections were prepared. Obese and normal omental adipose tissues were stained simultaneously, and a total of 3-4 different fields of images per individuals were captured with identical settings using an Olympus HC5000 digital camera attached to an Olympus microscope (Olympus, Tokyo, Japan) and quantified for positive staining areas. The clinical investigations (IRB number, #15-39) were approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University in China approved. Subjects gave a written informed consent after individual explanation. Statistical analysis Data are presented as means ± standard error (S.E.M). Using GraphPad Prism 5.0 software, results were analyzed by one-way ANOVA between multiple groups, when appropriate, and a two-tailed Student's t-test between two groups, as described previously in recent studies of metformin and obesity (4; 5; 10; 24-26). P < 0.05 was considered statistically significant. References 1. Li Y, Xu SQ, Giles A, Nakamura K, Lee JW, Hou XY, Donmez G, Li J, Luo ZJ, Walsh K, Guarente L, Zang MW: Hepatic overexpression of SIRT1 in mice attenuates endoplasmic reticulum stress and insulin resistance in the liver. Faseb Journal 2011;25: Lin HZ, Yang SQ, Chuckaree C, Kuhajda F, Ronnet G, Diehl AM: Metformin reverses fatty liver disease in obese, leptin-deficient mice. NatMed 2000;6: Duca FA, Cote CD, Rasmussen BA, Zadeh-Tahmasebi M, Rutter GA, Filippi BM, Lam TK: Metformin activates a duodenal Ampk-dependent pathway to lower hepatic glucose production in rats. Nat Med 2015;21: Madiraju AK, Erion DM, Rahimi Y, Zhang X-M, Braddock DT, Albright RA, Prigaro BJ, Wood JL, Bhanot S, MacDonald MJ, Jurczak MJ, Camporez J-P, Lee H-Y, Cline GW, Samuel VT, Kibbey RG, Shulman GI: Metformin suppresses gluconeogenesis by inhibiting mitochondrial glycerophosphate dehydrogenase. Nature 2014;510: Li Y, Xu SQ, Mihaylova MM, Zheng B, Hou XY, Jiang BB, Park O, Luo ZJ, Lefai E, Shyy JYJ, Gao B, Wierzbicki M, Verbeuren TJ, Shaw RJ, Cohen RA, Zang MW: AMPK Phosphorylates and Inhibits SREBP Activity to Attenuate Hepatic Steatosis and Atherosclerosis in Diet-Induced Insulin-Resistant Mice. Cell Metabolism 2011;13: Li Y, Wong K, Giles A, Jiang J, Lee JW, Adams AC, Kharitonenkov A, Yang Q, Gao B, Guarente L, Zang M: Hepatic SIRT1 Attenuates Hepatic Steatosis and Controls Energy Balance in Mice by Inducing Fibroblast Growth Factor 21. Gastroenterology 2014;146: e Li Y, Wong K, Walsh K, Gao B, Zang MW: Retinoic Acid Receptor beta Stimulates Hepatic Induction of Fibroblast Growth Factor 21 to Promote Fatty Acid Oxidation and Control Whole-body

45 Page 45 of 51 Diabetes Energy Homeostasis in Mice. Journal of Biological Chemistry 2013;288: Aune UL, Ruiz L, Kajimura S: Isolation and differentiation of stromal vascular cells to beige/brite cells. Journal of visualized experiments : JoVE 2013; 9. Wang QA, Scherer PE, Gupta RK: Improved methodologies for the study of adipose biology: insights gained and opportunities ahead. J Lipid Res 2014;55: Zang MW, Xu SQ, Maitland-Toolan KA, Zuccollo A, Hou XY, Jiang BB, Wierzbicki M, Verbeuren TJ, Cohen RA: Polyphenols stimulate AMP-activated protein kinase, lower lipids, and inhibit accelerated atherosclerosis in diabetic LDL receptor-deficient mice. Diabetes 2006;55: Hou X, Xu S, Maitland-Toolan KA, Sato K, Jiang B, Ido Y, Lan F, Walsh K, Wierzbicki M, Verbeuren TJ, Cohen RA, Zang M: SIRT1 regulates hepatocyte lipid metabolism through activating AMP-activated protein kinase. JBiolChem 2008;283: Zang MW, Zuccollo A, Hou XY, Nagata D, Walsh K, Herscovitz H, Brecher P, Ruderman NB, Cohen RA: AMP-activated protein kinase is required for the lipid-lowering effect of metformin in insulin-resistant human HepG2 cells. Journal of Biological Chemistry 2004;279: Liu M, Bai J, He S, Villarreal R, Hu D, Zhang C, Yang X, Liang H, Slaga TJ, Yu Y, Zhou Z, Blenis J, Scherer PE, Dong LQ, Liu F: Grb10 promotes lipolysis and thermogenesis by phosphorylation-dependent feedback inhibition of mtorc1. Cell Metab 2014;19: Choi SM, Tucker DF, Gross DN, Easton RM, DiPilato LM, Dean AS, Monks BR, Birnbaum MJ: Insulin regulates adipocyte lipolysis via an Akt-independent signaling pathway. Molecular and cellular biology 2010;30: Kraus D, Yang Q, Kong D, Banks AS, Zhang L, Rodgers JT, Pirinen E, Pulinilkunnil TC, Gong F, Wang Y-c, Cen Y, Sauve AA, Asara JM, Peroni OD, Monia BP, Bhanot S, Alhonen L, Puigserver P, Kahn BB: Nicotinamide N-methyltransferase knockdown protects against diet-induced obesity. Nature 2014;508: Khan T, Muise ES, Iyengar P, Wang ZV, Chandalia M, Abate N, Zhang BB, Bonaldo P, Chua S, Scherer PE: Metabolic dysregulation and adipose tissue fibrosis: role of collagen VI. Molecular and cellular biology 2009;29: Li Y, Xu S, Jiang B, Cohen RA, Zang M: Activation of sterol regulatory element binding protein and NLRP3 inflammasome in atherosclerotic lesion development in diabetic pigs. PLoS One 2013;8:e Jeong WI, Park O, Gao B: Abrogation of the antifibrotic effects of natural killer cells/interferon-gamma contributes to alcohol acceleration of liver fibrosis. Gastroenterology 2008;134: Yang L, Chan C-C, Kwon O-S, Liu S, McGhee J, Stimpson SA, Chen LZ, Harrington WW,

46 Diabetes Page 46 of 51 Symonds WT, Rockey DC: Regulation of peroxisome proliferator-activated receptor-γ in liver fibrosis. American Journal of Physiology - Gastrointestinal and Liver Physiology 2006;291:G902-G Halberg N, Khan T, Trujillo ME, Wernstedt-Asterholm I, Attie AD, Sherwani S, Wang ZV, Landskroner-Eiger S, Dineen S, Magalang UJ, Brekken RA, Scherer PE: Hypoxia-inducible factor 1alpha induces fibrosis and insulin resistance in white adipose tissue. Molecular and cellular biology 2009;29: Hsu WC, Araneta MR, Kanaya AM, Chiang JL, Fujimoto W: BMI cut points to identify at-risk Asian Americans for type 2 diabetes screening. Diabetes Care 2015;38: Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004;363: Lee JW, Brancati FL, Yeh HC: Trends in the prevalence of type 2 diabetes in Asians versus whites: results from the United States National Health Interview Survey, Diabetes Care 2011;34: Miller RA, Chu Q, Xie J, Foretz M, Viollet B, Birnbaum MJ: Biguanides suppress hepatic glucagon signalling by decreasing production of cyclic AMP. Nature 2013;494: Fullerton MD, Galic S, Marcinko K, Sikkema S, Pulinilkunnil T, Chen ZP, O'Neill HM, Ford RJ, Palanivel R, O'Brien M, Hardie DG, Macaulay SL, Schertzer JD, Dyck JR, van Denderen BJ, Kemp BE, Steinberg GR: Single phosphorylation sites in Acc1 and Acc2 regulate lipid homeostasis and the insulin-sensitizing effects of metformin. Nat Med 2013;19: Sun K, Park J, Gupta OT, Holland WL, Auerbach P, Zhang N, Goncalves Marangoni R, Nicoloro SM, Czech MP, Varga J, Ploug T, An Z, Scherer PE: Endotrophin triggers adipose tissue fibrosis and metabolic dysfunction. Nature communications 2014;5:doi: /ncomms4485

47 Page 47 of 51 Diabetes Supplemental Figure 1 A % Total Adipocytes % Total Adipocytes % T o tal Ad ip o cytes 40 ND Body Weight Gain (g) < < < ND * ewat Cell Area (µm 2 ) Cell Area (µm 2 ) Cell Area (µm 2 ) HFHS >18000 HFHS+Metformin >18000 >18000 % Total Adipocytes % Total Adipocytes 40 ND < < < C D E # HFHS HFHS +Metformin Adipocyte Size ( m 2 ) ND * B % Total Adipocytes # HFHS HFHS +Metformin Cell Area (µm 2 ) Cell Area (µm 2 ) Trichrome Positive Area (% of iwat Area) iwat Cell Area (µm 2 ) ND Luo T, et al HFHS * # HFHS HFHS +Metformin > >18000 HFHS+Metformin Supplemental Fig. 1. The effect of metformin on body weight gain as well as adipocyte size and distribution in white adipose tissue of HFHS-induced obese mice treated or untreated with metformin. A. The bar graph shows adipocyte size distribution of epididymal white adipose tissue (ewat) from three groups of mice. C57BL/6 mice were placed on a normal chow diet (ND, n=7) and a high fat, high sucrose diet (HFHS, n =7), and the a HFHS diet supplemented with metformin treatment (250 mg/kg/day, HFHS + Metformin, n =7). The adipocyte areas of ewat were calculated in haematoxylin-eosin (H&E)-stained sections using ImageJ software. Adipocyte areas were determined from 5-6 mice/group (>500 cells counted per group). B. The bar graph shows adipocyte size distribution of inguinal white adipose tissue (iwat) from the animals. C. Body weight gain of the animals, mean ± SEM, n = 7. D. Average adipocyte area of inguinal adipose tissue analyzed in (H&E)-stained sections from the animals. E. Quantification of collagen deposition in inguinal adipose tissue sections stained with Trichrome. Data are presented as the mean ± SEM, n = 7. P< 0.05, vs. normal chow diet mice; # P < 0.05, vs. HFHS-fed mice. >18000

48 Diabetes Page 48 of 51 Supplemental Figure 2 Luo T, et al. 8 ewat 7 iwat Relative mrna Levels Col1a1 Col6a3 Fibronectin Supplemental Fig. 2. Real-time PCR analysis of expression of fibrogenic genes encoding collagen I (Col1a1), VI (Col6a3), and fibronectin in epididymal white adipose tissue (ewat) and inguinal white adipose tissue (iwat) of C57BL/6 mice. Data are presented as the mean ± SEM, n = 5-7.

49 Page 49 of 51 Diabetes Supplemental Figure 3 Luo T, et al. A 40X MAC-2 20X * Lean ob/ob ob/ob + Metformin ob/ob ob/ob + Metformin * * * * B MAC-2 Positive Area (% of WAT Area) Lean * # ob/ob ob/ob +Metformin C Relative mrna Levels 14 Lean 12 ob/ob ob/ob + Metformin * # * # * # 0 TNF- IL-1 MCP-1 Supplemental Fig. 3. Metformin represses inflammatory response in ob/ob mice. A. Representative immunohistochemical staining of epididymal fat sections for MAC-2, a specific marker for mature macrophages. Positive staining for MAC-2 was visualized mainly in macrophages infiltrated into WAT of ob/ob mice. Notably, aggregated macrophages were located in crown-like structures (CLS), which contained up to 15 macrophages surrounding what appeared to be dead adipocytes. In lean animals, MAC-2-positive macrophages were uniformly small, dispersed, and rarely seen in aggregates. MAC-2-positive macrophages were presented in hypertrophic (or degenerating) adipocytes (blue arrows) and aggregated to form numerous CLS (asterisks) in ob/ob mice. There was no detectable staining with a non-specific IgG in adipose tissue sections (data not shown), indicating the specific staining of MAC-2 and cleaved caspase-3. Original magnification 20 and 40. B. The bar graphs represent relatively positive staining areas to total tissue areas. C. Real-time PCR analysis of pro-inflammatory regulators including TNF-α, IL-1β, and MCP-1 in WAT of ob/ob mice treated without or with metformin. Data are presented as the mean ± SEM, n = 6-7. P<0.05,vs. lean mice; # P < 0.05, vs. ob/ob mice.