Inhibition of Adipogenesis in 3T3-L1 Cells and Suppression of Abdominal

Home , L1 (protein)

ORIGINAL ARTICLE Inhibition of adipogenesis in 3T3-L1 cells and suppression of abdominal fat accumulation in high-fat diet-feeding C57BL/6J mice after downregulation of hyaluronic acid

EJi1,5, MY Jung1,5, JH Park2, S Kim1, CR Seo1, KW Park1, EK Lee3, CH Yeom4 and S Lee1

OBJECTIVE: Adipogenesis can be spatially and temporally regulated by extracellular matrix (ECM). We hypothesized that the regulation of hyaluronic acid (HA), a component of the ECM, can affect adipogenesis in fat cells. The effects of HA on adipogenesis were investigated in vitro in 3T3-L1 cells and in vivo in high-fat diet-feeding C57BL/6J mice. METHODS: We investigated the effects of HA by degradation of pre-existing or synthesized HA and artiﬁcial inhibition of HA synthesis in adipogenesis. RESULTS: In vitro adipogenesis in 3T3-L1 cells was inhibited by treating them with exogenous hyaluronidase (HYAL) and with 4-methylumbelliferone, which inhibited the synthesis of HA in a concentration-dependent manner. In vivo, abdominal fat accumulation in high-fat diet-feeding C57BL/6J mice was suppressed by exogenous HYAL 104 IU injections, which was associated with reduction of lipid accumulation in liver and increase of insulin sensitivity. CONCLUSION: Changes in the ECM such as accumulation of high molecular weight of HA by HAS and degradation of HA by endogenous HYAL were essential for adipogenesis both in vitro and in vivo.

International Journal of Obesity (2014) 38, 1035–1043; doi:10.1038/ijo.2013.202 Keywords: extracellular matrix; hyaluronic acid; adipogenesis; 3T3-L1 cells; high-fat diet-induced obesity

INTRODUCTION cytoskeletal elements is necessary for the transformation of 11–13 Obesity is defined as a condition in which excessive fat accumulates pre-adipocytes into mature adipocytes. Several reports have in the body, which can result in many health-related problems. examined the relationship between the ECM and the Obesity is one of the main factors responsible for the increased differentiation of 3T3-L1 cells, including changes in the protein incidence of cardiovascular disease, Type 2 diabetes and several types composition and reorganization of the ECM during 3T3-L1 14,15 16 of cancer. The resultant increases in morbidity and reductions in life differentiation. Demeulemeester et al. reported that weight expectancy are both personally and economically damaging.1 As a gain in C57BL/6J mice was reduced by treatment with a matrix result, obesity is a serious public health problem, and mechanisms to metalloproteinase (MMP) inhibitor. Therefore, appropriate control obesity have been actively studied worldwide. Though regulation of the ECM has a role in adipose tissue development several drugs have been developed to treat obesity, an ideal in vivo as well as in vitro. Hyaluronic acid (also known as pharmacological therapy has not yet been implemented. Therefore, hyaluronan or hyaluronate (HA)) is a critical component of the a novel approach for the treatment of obesity is needed. ECM, and is found both at the cell surface and within the cells. Adipocyte differentiation has previously been studied in vitro, HA is a negatively charged, non-sulfated, glycosaminoglycan (GAG) using specific cell lines to establish ideal conditions for composed of repeating disaccharides of D-glucuronic acid and differentiation.2–5 Likewise, adipocytes studied using in vitro N-acetyl-glucosamine. HA is synthesized as a large unbranched systems shared common characteristics and mechanisms of polymer (molecular weight ranging from 20 kDa to 42 MDa) on differentiation with adipocytes in vivo.6,7 For example, it has the cytoplasmic side of the plasma membrane by HA synthase been demonstrated that peroxisome proliferator-activated (HAS), which is an integral membrane glycosyltransferase, and is receptor gamma (PPARg), which is a transcription factor of the subsequently extruded to the extracellular surface.17,18 Despite its ligand-dependent nuclear receptor type,8 and CCAAT-enhancer- simple structure, HA has numerous functions, as it is progressively binding proteins (C/EBPs), which are a family of transcription degraded by enzymes of the hyaluronidase (HYAL) family at the factors of the leucine zipper type,9 play important roles in cellular level.19 Recently, it has been suggested that the specific adipogenesis. In addition, the extracellular matrix (ECM) was found functions of HA fragments are polymer-size-dependent,19 though to be involved in adipogenesis through changes in protein the location and the concentration of the polymers as well as of composition and dynamics.10 various HA-size-specific binding proteins are also variables.20 During adipogenesis, regulation of the expression of various Native high molecular weight (Mw) HA acts as a scaffold on which genes that are specifically involved in the formation of ECM and other macromolecules are assembled and has a remarkable ability

1Department of Biotechnology and Bioengineering, Sungkyunkwan University, Suwon, Korea; 2Institute of Cancer Research, College of Medicine, The Catholic University of Korea, Seoul, Korea; 3Department of Biochemistry, College of Medicine, The Catholic University of Korea, Seoul, Korea and 4Yeom’s Family Medicine Clinic, Seoul, Korea. Correspondence: Dr S Lee, Department of Biotechnology and Bioengineering, Sungkyunkwan University, 300, Cheoncheon-dong, Jangan-gu, Suwon, 440-746, Korea. E-mail: [email protected] 5These authors contributed equally to this work. Received 3 May 2013; revised 24 September 2013; accepted 4 October 2013; accepted article preview online 31 October 2013; advance online publication, 26 November 2013 Reduction of hyaluronic acid blocks adipogenesis EJiet al 1036 to retain water, thus contributing to the organization and light/dark cycle. When mice were 6.5 weeks of age, they were divided into maintenance of the integrity and hydrodynamic properties of five groups (n ¼ 5 in all groups except group B, in which n ¼ 4). Mice in one the ECM.21 In 1993, it was reported first by Calvo et al.22 that HA of two groups fed NFD were injected with phosphate-buffered saline (PBS) as a negative control for fat accumulation, while the other group was existed in the extracellular media of differentiated 3T3-L1 4 pre-adipocyte cultures and after that a few papers regarding injected with 10 IU of HYAL (BMI Korea, Jeju, Korea) as a negative control for high-dose HYAL injection. Three groups were given a high-fat diet the relationship between HA and adipogenesis have 23,24 containing 60% fat (HFD; Research Diets Inc., New Brunswick, NJ, USA). Two been published. Interestingly, it was shown by Allingham groups of mice fed HFD were injected with 102 IU and 104 IU of HYAL, et al. that the expression of genes for the biosynthesis and respectively. The third group was injected with PBS every 3 days as a degradation of HA was positively involved in the differentiation positive control for HFD-induced fat accumulation. Food intake was of adipocytes.23 Also, HA-modified matrix has been studied measured every 3 days, and body weight was monitored every 6 days for in vitro and in vivo for adipose tissue engineering.25–27 2 months. Therefore, it has been expected that HA is important in adipogenesis. Intraperitoneal glucose tolerance test (IPGTT) In this study, we examined if the regulation of HA inhibited IPGTTs were performed in all mice at 13.5 weeks of age. After overnight adipogenesis and thus lipid accumulation based on this associa- fasting, animals received an intraperitoneal injection of D-glucose tion between HA and adipogenesis. We observed the inhibitory (2 g kg À 1) in 0.9% NaCl.30 A drop of blood was taken from the tail vein effects of HA on adipogenesis in 3T3-L1 cells in vitro, and the before the glucose injection (0), as well as 15, 30, 60, 90, and 120 min after prevention of insulin resistance and non-alcoholic fatty liver the injection in order to determine blood glucose levels using a disease (NAFLD) caused by excess accumulation of abdominal fat glucometer (Accu-Chek Active; Roche Diagnostics GmbH). in HFD-feeding C57BL/6J mice in vivo by artificially reducing HA levels. Micro-CT At 14 weeks of age, the mice were imaged using Micro-CT (Skyscan model 1076; Skyscan, Kontich, Belgium) under anesthesia. The resolution of the MATERIALS AND METHODS micro-CT was 35 mm. Culture and differentiation of 3T3-L1 pre-adipocytes 3T3-L1 pre-adipocytes were purchased from American Type Culture Sample and serum analysis Collection (ATCC; Manassas, VA, USA). The pre-adipocytes were maintained At 14.5 weeks of age, the mice were put under deep anesthesia after in Dulbecco’s modified Eagle’s medium (DMEM; Hyclone, South Logan, UT, overnight fasting. Blood samples were taken from the inferior vena cava, USA), and supplemented with 10% newborn calf serum (Hyclone) and and serum samples were collected by centrifuging at 14 000 r.p.m. for antibiotics (Hyclone) at 37 1Cin5%CO2. To induce adipocyte differentia- 5 30 min after being kept at room temperature for 1 h. Aspartate tion, the pre-adipocytes were plated at a density of 2 Â 10 cells per well transaminase (AST), alanine transaminase (ALT), alkaline phosphatase on six-well plates. After confluence, cells were incubated in DMEM (ALP), total cholesterol (T.cho), triglycerides (TG), high-density lipoprotein supplemented with 10% fetal bovine serum (Hyclone) and antibiotics, (HDL-C), low-density lipoprotein (LDL-C) and insulin levels were analysed and treated with DMI (1 mM dexamethasone (Sigma-Aldrich, St Louis, MO, À 1 by the Korea Animal Medical Science Institute (KAMESI, Seoul, Korea). USA), 0.5 mM 3-iso-butyl-1-methylxantine (Sigma-Aldrich) and 1 mgml The weights of fat tissues and organs were also measured. insulin (Roche Diagnostics GmbH, Mannheim, Germany)) for 3 days followed by treatment with insulin alone. To maintain artificially low levels of HA during adipogenesis, we used testicular type hyaluronidase Quantitative real-time PCR (PH-20; HYAL; Sigma-Aldrich) and 4-methylumbelliferone (4-MU; Sigma- Total RNA was prepared directly from cells or tissues using Tri reagent 28 Aldrich), which inhibited the HA synthesis. 3T3-L1 cells were treated (Molecular Research Center, Inc., Cincinnati, OH, USA) according to the with various concentrations of HYAL and 4-MU after DMI induction, manufacturer’s instructions. Moloney murine leukemia virus (M-MLV) respectively. Quantification of HA in culture media was measured reverse transcriptase (Superbio Co., Daejeon, Korea) was used for cDNA using Hyaluronan Enzyme-Linked Immunosorbent Assay kit (HA-ELISA; synthesis, and qReal-time PCR was performed using SYBR green premix Echelon, Salt Lake City, UT, USA), according to the manufacturer’s (Takara bio Inc., Shiga, Japan). Mouse b-2 microglobulin (or b2M) mRNA or instructions. For subsequent experiments, the cells and cultured media 36B4 was used as the internal standard for error correction between were harvested at the following times: during the contact-inhibited samples. Oligonucleotide sequences for each primer are listed in phase (growth arrest) prior to induction of differentiation (days 2 and 3), Supplementary Table 1. qReal-time PCR amplification of each cDNA was during the clonal expansion phase (day 6), post-differentiation (day 8), performed independently in triplicate.31 and during terminal differentiation (day 10).23 On day 10, cells were stained with Oil Red O (Sigma-Aldrich). Western blot After terminal differentiation, 3T3-L1 cells were lysed using Pro-prep Transfection of small interfering RNA (Intron Biotechnology, Seongnam, Korea). Western blot was followed as Small interfering RNAs (siRNAs) were used to silence mouse hyaluronic acid described by Waki et al.32 Antibodies against PPARg, Fabp4 (or aP2), and synthase 2 (HAS2) expression. Scramble control and HAS2-specific siRNAs HRP-conjugated anti-goat IgG were obtained from Santa Cruz were synthesized by Genolution Pharmaceuticals, Inc. (Seoul, Korea). Biotechnology (Santa Cruz, CA, USA), and antibodies against b-actin and The sense sequence of HAS2-specific siRNA was 50-CUGAAACUCCCAUA HRP-conjugated anti-mouse IgG were purchased from Abcam (Cambridge, GAAUAUU-30. The sense sequence of control non-specific scramble RNA UK) and Cell Signaling Technology, Inc. (Beverly, MA, USA), respectively. was 50-CCUCGUGCCGUUCCAUCAGGUAGUU-30. Cells plated at a density of 2 Â 105 cells per well in a six-well plate were transfected with 40 pmol of scramble RNA or HAS2-specific siRNAs using RNAiMAX (Invitrogen, H&E staining Carlsbad, CA, USA) as previously described.29 Cells were treated with Epididymal fat and liver tissue were fixed with 4% paraformaldehyde, siRNA for 6 h, and then the medium was exchanged. After 48 h, cells were dehydrated, and embedded in paraffin. Sections 10 mm thick were processed using differentiation protocols. Transfection was carried out in subjected to hematoxylin and eosin (H&E) staining.33 duplicated wells and repeated at least three times. Statistical analysis Fat accumulation induced by a high-fat diet feeding and HYAL Data were analysed using GraphPAD Prism statistical software (version injections 5.01; GraphPAD Software Inc., La Jolla, CA, USA), and were expressed as Male C57BL/6J Jms Slc (5.5 weeks old) mice were purchased from Central means±s.e.m. Significant differences between groups were determined Lab Animal Inc. (Seoul, Korea). Upon arrival, mice were fed a Rodfeed using the Student’s t-test and one-way ANOVAs followed by post hoc (normal-fat diet, NFD; DBL, Eumsung, Korea) for 1 week. The animals were Dunnett’s multiple comparisons tests, or two-way ANOVAs with Bonferroni housed in individual cages in a temperature-controlled room with a 12-h correction.

International Journal of Obesity (2014) 1035 – 1043 & 2014 Macmillan Publishers Limited Reduction of hyaluronic acid blocks adipogenesis EJiet al 1037 RESULTS The effects of HYAL on adipogenesis prompted us to Downregulation of HA concentration by exogenous HYAL and investigate the roles of HAS2, which is a core enzyme on HA 4-MU during induction of adipogenesis in 3T3-L1 cells accumulation in adipocyte differentiation.23 To show the effects Prior to investigating the downregulation of HA during adipoge- of HAS2 in adipogenesis, siRNA-mediated HAS2 silencing was nesis, we confirmed that cell viability was not affected by performed. siRNAs were transiently transfected into 3T3-L1 cells exogenous HYAL or 4-MU up to concentrations of 200 mM and differentiated into adipocytes for 5 days. Consistent with the (Supplementary Figures 1a and 1b). When exogenous HYAL and effects of HYAL, lipid accumulation was significantly impaired by 4-MU were provided continuously in cultured media after DMI silencing HAS2 expression compared with the control non-specific induction, HA concentrations were reduced compared with the scramble siRNA transfected cells (Figures 3a and c). The degree of untreated condition during adipogenesis (Figures 1a and b). In the inhibitory effects on HAS2 expression by introducing two different positive control group (PBS), HA concentration increased 11-fold doses of siRNAs mirrors the degree of anti-lipogenic effects during (between days 2 and 5). However, HA concentration decreased to adipocyte differentiation (Figure 3b). Furthermore, the expression 54% and to 8% after treatment with exogenous HYAL 100 and 200 of adipocyte markers such as PPARg and Fabp4 was also dose IU, respectively, compared to PBS treatment (100%). In addition, dependently inhibited by silencing HAS2 expression (Figures 3d HA concentrations after treatment with 100 and 200 mM 4-MU and e). Taken together, these data demonstrate that inhi- were reduced to 38% and to 6%, respectively. Thus, exogenous bition of HA accumulation by pharmacological, enzymatic and HYAL and 4-MU were suitable for downregulation of HA levels genetical suppression affects lipid accumulation and adipocyte during adipogenesis in 3T3-L1 cells. differentiation.

Suppression of increases in body weight and body fat by HYAL Reduction in HA levels and inhibition of adipogenesis in 3T3-L1 treatment in HFD-feeding C57BL/6J mice We treated 3T3-L1 cells with exogenous HYAL or 4-MU, To confirm the effects of downregulation of HA in vivo, we used an which inhibited the formation and accumulation of lipid droplets animal model of HFD-feeding C57BL/6J mice. We intended to in a concentration-dependent manner (Figures 2a and b). decrease HA levels in the mice using IP injections of HYAL every 3 Adipogenesis was inhibited by up to 96% (Po0.0001) and up days while providing a HFD. We prepared 5 groups of mice (n ¼ 5 to 80% (Po0.0001) with exogenous HYAL 200 IU and 200 mM in all groups except group B, where, n ¼ 4 ). Group A, fed NFD and 4-MU treatment, respectively (Supplementary Figures 2a and b). injected with PBS; group B, fed NFD and injected with HYAL 104 IU; However, as the timing of treatment with exogenous HYAL and group C, fed HFD and injected with PBS; group D, fed HFD and 4-MU were delayed, their abilities to suppress adipogenesis injected with HYAL 102 IU; and group E, fed HFD and injected with were poor (Supplementary Figures 2c and d). Adipogenesis HYAL 104 IU. Body weight and food intake were monitored every 6 was inhibited most effectively in the case in which exogenous days and every 3 days, respectively. These data showed that HYAL and 4-MU were administered after DMI induction. weight gain decreased with injections of HYAL (Figure 4a) without We wondered whether the reduction of HA during DMI induction significant differences in food intake (Supplementary Figure 4a). was involved in clonal expansion but it was revealed that clonal At last measurement, the body weight and the net weight gained expansion was independent of HA reduction (Supplementary in group C and group E were as follows: group C (32.60±0.81 g, Figures 3a and b). We also investigated whether these inhibitory 12.20±0.49 g); and group E (29.80±0.58 g, 10.40±0.51 g). These effects corresponded to the expression profiles of adipogenic data indicate that body weight and net weight gained in group E marker genes in 3T3-L1. At the mRNA level, the expression were down by B9% and 15%, respectively, compared with group C. of adipogenic master marker genes such as PPARg and The distribution and volume of abdominal fat in group E were C/EBPa8,34,35 decreased with exogenous HYAL and 4-MU in a determined using micro-CT (resolution ¼ 35 mm) for comparison concentration-dependent manner (Figures 2c and d). Fabp4 and with that of control groups (Figure 4b). Abdominal fat in group E, AdipoQ induced by PPARg and C/EBPa36 showed a similar pattern especially epididymal fat (Epi fat, black arrows), was lesser than of expression to that of PPARg and C/EBPa (Figures 2c and d). It that of the control groups. In comparing the abdominal views was also determined that downregulation of HA by exogenous and the abdominal fat, Epi fat (white arrows) and retroperitoneal HYAL and 4-MU resulted in inhibition of differentiation of pre- fat (RP fat) appeared to be inhibited by injections of HYAL adipocytes into mature adipocytes at the protein level as well 104 IU (Figure 4c). The weights of Epi fat in groups C and E were (Figure 2e). 1.05±0.19 and 0.60±0.04 g, respectively (Po0.01). Thus, expansion

Figure 1. HA concentration after exogenous HYAL and 4-MU treatment. (a and b) HA concentration in culture media as measured by HA-ELISA kits was decreased by exogenous HYAL and 4-MU in a concentration-dependent manner, respectively. Data are expressed as mean±s.e.m., and were analysed using one-way ANOVAs with Bonferroni correction (###Po0.001 between PBS and concentration HYAL 100 IU or 4-MU 100 mM, ***Po0.001 between PBS and concentration HYAL 200 IU or 4-MU 200 mM).

& 2014 Macmillan Publishers Limited International Journal of Obesity (2014) 1035 – 1043 Reduction of hyaluronic acid blocks adipogenesis EJiet al 1038

Figure 2. Effects of exogenous HYAL and 4-MU on adipogenesis. (a and b) Adipogenesis in 3T3-L1 cells depended on the treatment concentration of exogenous HYAL or 4-MU (upper box: scanning of six-well plates, lower box: view under inverted microscope). (c and d) The mRNA expression levels of PPARg, C/EBPa, Fabp4, and AdipoQ measured by qReal-time PCR decreased with HYAL or 4-MU treatment in a concentration-dependent manner. The experiments were repeated in triplicate. (e) The protein expression level of PPARg and Fabp4 as conﬁrmed by western blot were also reduced by HYAL or 4-MU treatment. The data are expressed as mean±s.e.m. and were analysed using one-way ANOVAs with Dunnett’s multiple comparisons tests (*Po0.05, **Po0.001, ***Po0.0001).

of Epi fat was suppressed by up to 40% by injections of HYAL H&E staining was smaller than that of group C, indicating that the 104 IU. There were no differences, however, between the weights degree of lipid accumulation in fat cells was reduced by HYAL 104 of organs in each group (Supplementary Figure 4b). The tendency IU injection (Figure 4e). for increased body fat to be suppressed by exogenous HYAL treatment was somewhat reﬂected in the total cholesterol (T.cho) and low-density lipoprotein (LDL-C) levels (Figure 4d) in serum. Reductions in lipid accumulation in the liver after HYAL treatment The T.cho levels in groups E and C were 151.50±3.12 and Liver damage is a serious problem brought about by NAFLD. 168.75±8.07 mg dl À 1, respectively (Po0.05). In addition, it We analysed the liver tissue from mice in each group to conﬁrm appeared that the size of Epi fat cells in group E as seen on whether the symptoms of NAFLD were reduced by HYAL

International Journal of Obesity (2014) 1035 – 1043 & 2014 Macmillan Publishers Limited Reduction of hyaluronic acid blocks adipogenesis EJiet al 1039

Figure 3. Inhibition of adipogenesis by knockdown of HAS2.(a) Knockdown of HAS2 expression in 3T3-L1 cells with transfection of siRNA inhibited lipid accumulation as assessed by Oil Red O staining. 3T3-L1 cells were induced to differentiate into adipocytes for 5 days. (b) Transient transfection of 3T3-L1 cells with siRNA targeted to HAS2 dose dependently reduces the expression of HAS2.(c) Quantification of lipid accumulation of differentiated cells by measuring the extracted dye at 520 nm. (d, e) Knockdown of HAS2 inhibits the expression of adipocyte markers. Expression of PPARg (d) and its target gene Fabp4 (e) was measured by real-time PCR. NC: growth media, scr: scrambled siRNA-transfected cells in differentiating condition, siHAS2: HAS2-specific siRNA-transfected cells in differentiating condition. Data shown represent the mean±s.e.m. Statistical significance was determined relative to scrambled RNA-transfected cells by the Student’s t-test (*Po0.05; **Po0.005; ***Po0.0005). treatment. On comparing the color of the liver tissue, the livers of There was a significant difference in blood glucose levels at 60 min mice in group E appeared healthier than those of group C (Po0.05) between groups C and E. In addition, the area under the (Figure 5a). This was clearly shown on H&E staining, as the lipid curve for group E was 23% less than that of group C, based on the spots in livers from group E were fewer and smaller than that of IPGTT graph (Figure 6a). The insulin level for group E was group C (Figure 5b). The degree of liver injury was estimated 4.18±0.46 mIU ml À 1, and there were no significant differences numerically through serum levels of liver enzymes, including between groups A and E (Figure 6b). Collectively, these results aspartate aminotransferase (AST), alanine transaminase (ALT) and indicate that insulin resistance was alleviated after treatment with alkaline phosphatase (ALP)37 (Figure 5c). Enzyme levels in group E of HYAL 104 IU. were somewhat lower than those of group C, and there was no evidence of damage to normal livers from injection of HYAL 104 IU. Thus, our results showed that the accumulation of lipid in the liver DISCUSSION was decreased by treatment with HYAL 104 IU, and that NAFLD Adipogenesis is regulated both spatially and temporally by did not develop in mice that received these injections due to the sophisticated processes that maintain and alter the ECM via absence of progression of obesity. cell–cell and cell–environment interactions. The correlation between collagen and adipogenesis has been extensively studied Effects of HYAL treatment on insulin resistance caused by HFD- in terms of the regulation of collagen synthesis and degradation. induced abdominal obesity in C57BL/6J mice In these reports, adipogenesis was inhibited by the synthesis of Obesity is significantly associated with metabolic syndromes such certain types of collagen, and their degradation was artificially as Type 2 diabetes and insulin resistance is a typical symptom of manipulated using ethyl-3,4-dihydroxybenzoate as a specific Type 2 diabetes. To investigate whether insulin resistance was inhibitor of collagen synthesis and general inhibitor of MMP improved with exogenous HYAL treatment, we performed IPGTTs activity.38,39 The ECM, however, includes a variety of components and measured insulin levels in the serum. In group C, blood in addition to HA, and thus more sophisticated research glucose peaked at 30 min (524.00±26.95 mg dl À 1) and was approaches are necessary to understand the role of the ECM in maintained at a higher level at 90 min (338.60±25.33 mg dl À 1) adipogenesis. In this report we showed that the downregulation than in group A (Figure 6a). In addition, the insulin level of group C of HA in the ECM inhibited adipogenesis in vitro and in vivo and was 10.80±1.73 mIU ml À 1, which was much higher than that of this finding would be a new angle to the adipogenesis research. group A (9.00±0.71 mIU ml À 1) (Figure 6b). Thus, insulin resistance The relationship between adipogenesis and HA was shown via appeared to be greater in the HFD-feeding mice group C. On the gene expression, as well as synthesis and degradation of HA other hand, blood glucose in group E peaked at 15 min during adipogenesis in 3T3-L1 cells.23 We focused on the effects of (477.20±17.96 mg dl À 1) and subsequently decreased over time. HA during adipogenesis both in vitro and in vivo, and the

& 2014 Macmillan Publishers Limited International Journal of Obesity (2014) 1035 – 1043 Reduction of hyaluronic acid blocks adipogenesis EJiet al 1040

Figure 4. The effects of HYAL treatment on C57BL/6 J mice. (a) Weight gain was decreased by injections of HYAL 104 IU compared with the positive controls. Body weight was monitored every 6 days for 8 weeks. Groups A (empty circle) and B (gray circle) were negative controls for fat accumulation and high dose HYAL, respectively. Group C (black square) was a positive control for fat accumulation. Groups D (gray triangle) and E (black inverted triangle) were induced obesity groups that were simultaneously injected with HYAL 102 IU and104 IU, respectively (n ¼ 5 in all groups except group B, n ¼ 4 in group B). (b)InjectionsofHYAL104 IU inhibited adipogenesis in body fat, especially epididymal fat (Epi fat, black arrows). Abdominal fat tissue was measured by micro-CT (resolution ¼ 35 mm). (c) Increases in body fat were suppressed by injections of HYAL 104 IU. The volume and weight of Epi fat (white arrows) and retroperitoneal fat (RP fat) were diminished in HYAL-injected mice. (d) Total cholesterol (T.cho) in serum showed a mild decrease after HYAL injection. Total cholesterol (T.cho), triglyceride (TG), high density lipoprotein (HDL-C), and low density lipoprotein (LDL-C) levels were evaluated by serum analysis. (e) Fat cells were smaller in size than those of the positive controls on H&E staining after paraformaldehyde ﬁxation (scale bar ¼ 50 mm). The data are expressed as mean±s.e.m. and were analysed using two-way ANOVAs with Bonferroni correction for body weight, and one-way ANOVAs with Dunnett’s multiple comparisons tests for the weight of fat, serum analysis, and qReal-time PCR (*Po0.05, **Po0.001). The full colour version of this ﬁgure is available at International Journal of Obesity online.

International Journal of Obesity (2014) 1035 – 1043 & 2014 Macmillan Publishers Limited Reduction of hyaluronic acid blocks adipogenesis EJiet al 1041

Figure 5. The effects of HYAL injections on NAFLD on C57BL/6 J mice. (a and b) The accumulation of lipid in the liver was reduced by treatment with HYAL 104 IU. The livers were histologically analysed using H&E staining after paraformaldehyde ﬁxation (scale bar ¼ 50 mm). (c) Hepatotoxicity caused by progression of obesity was alleviated by HYAL 104 IU injections. Liver enzymes including aspartate aminotransferase (AST), alanine transaminase (ALT) and alkaline phosphatase (ALP) were measured by serum analysis (n ¼ 4 in each group). The data are expressed as mean±s.e.m. and were analysed using one-way ANOVAs with Dunnett’s multiple comparisons tests (*Po0.05).

Figure 6. The effects of HYAL injections on insulin resistance in C57BL/6J mice. (a) Blood glucose levels in the HFD-fed, HYAL 104 IU-injected mice were more smoothly regulated than the positive controls during intraperitoneal glucose tolerance tests (IPGTT) (n ¼ 5 in each group except group B, n ¼ 4 in group B). Groups A (empty circle) and B (gray circle) were negative controls for obesity and high dose HYAL, respectively. Group C (black square) was a positive control for obesity, and Group E (black inverted triangle) had induced obesity and was injected with HYAL 104 IU. (b)HYAL104 IU injections decreased insulin levels in the group fed HFD, but the insulin level of the negative control was not affected by HYAL. The values were expressed as mean±s.e.m. and analysed using two-way ANOVAs with Bonferroni correction (*Po0.05). experiments were designed to downregulate HA during induction Given that adipogenesis was inhibited by exogenous HYAL and of adipogenesis. Downregulation of HA implies a reduction in high 4-MU treatment in a concentration-dependent manner in vitro, Mw HA (42 MDa, the intact HA formation synthesized by HAS), we are able to predict the role of HA in adipogenesis. The function the non-natural degradation of pre-existing or newly synthesized of HA is size (Mw)-dependent,19 and is thus determined by a HA, and the artificial inhibition of HA synthesis. We first confirmed regulated balance between synthesis by HAS and degradation by whether the HA concentration in culture media was reduced after endogenous HYAL.40 It was confirmed that HAS2 and endogenous treatment with exogenous HYAL and 4-MU in a concentration- HYAL2 were primarily involved in adipogenesis in 3T3-L1 cells.25 dependent manner (Figure 1). Our results showed that treatment HA concentration increased rapidly during the period from growth with HYAL and 4-MU suppressed the formation of lipid droplets arrest to clonal expansion (Figure 1). However, HA accumulation and the accumulation of triglycerides, and reduced the expression did not affect clonal expansion (Supplementary Figures 3a and b). of adipogenic marker genes in differentiated 3T3-L1 cells During this time frame, it was reported that HA synthesized by (Figure 2). We also confirmed that HA reduction by HAS2 HAS2 had an average Mw of over 2 Â 103 kDa, which is the longest knockdown inhibited adipogenesis (Figure 3). In addition, it was chain synthesized by endogenous HAS and which corresponds to demonstrated that the abdominal fat accumulation and the serum the HA Mw previously reported as being present during 3T3-L1 lipid contents were reduced by exogenous HYAL 104 IU in differentiation.23 It was previously thought that the accumulation HFD-feeding C57BL/6J mice (Figure 4). of high Mw HA was required for the change in cellular

& 2014 Macmillan Publishers Limited International Journal of Obesity (2014) 1035 – 1043 Reduction of hyaluronic acid blocks adipogenesis EJiet al 1042 morphology from a fibroblast-like shape to a round shape for the as hypertrophy is the main contributing factor to adipose tissue accumulation of lipid within the cytoplasm, similar to the early stages expansion in obesity. However, recent remarkable studies by of differentiation of other cell types during morphogenesis.41,42 tracking adipocytes show that HFD-induced obesity can expand This occurs in a hydrated environment, in which cells are prevented epididymal adipose tissue by both hypertrophy and hyperplasia. by structural barriers from undergoing morphogenetic changes The studies show that de novo adipogenesis in epididymal and by receiving signals from HA and its associated factors.43 adipose tissue is preferentially initiated in the prolonged At the same time, expression of endogenous HYAL2 mRNA also HFD-induced obesity.52 In line with this, inhibition of hyperplasia increased (data not shown), which may be explained by the role of can affect HFD-induced obesity similar to the effects by HA HYAL2 in the tethering of high Mw HA to CD44 on cell surfaces for inhibition of the current study. It is also possible that the inhibition intracellular signal transduction.44,45 It has been reported that the of obesity by suppression of HA can be attributed todecreased HA-CD44-HYAL2 complex induces invagination of the plasma food intake. We failed to discriminate the food intake between membrane, and that HYAL2 subsequently cleaves high Mw control and HYAL-treated obese mice, suggesting that the HA to a smaller 20 kDa size (50 disaccharide units), which is energy consumption may not be a potential contributing then internalized and delivered to the endosomes for signal factor. The increased energy expenditure by HA suppression is transduction.40 The 20-kDa fragments of HA are highly another possibility for the anti-obese effects of HA suppression. angiogenic46 and induce transcription of MMPs.47 The details of Collectively, it will be intriguing to elucidate the target tissues for HA-induced signal transduction for adipogenesis are unknown, the anti-obese effects of HA inhibition. The underlying though the results of our study suggest that HA fragments are mechanisms of a reduction in insulin resistance, fatty liver and involved in PPARg-mediated adipogenesis, as well as in the inflammation would also be very interesting lines of study to be angiogenesis and ECM alteration required for adipogenesis. pursued. A previous study suggested that multivalent binding of the HA ligand to CD44 at the cell surface may be an important first step in intracellular signaling by HA, and that it depended on the size of CONFLICT OF INTEREST the HA ligand, as well as the quantity and density of cell surface The authors declare no conflict of interest. CD44.48 The exogenous HYAL used in our experiments was testicular type HYAL; thus, it produced even-numbered oligosaccharides containing mostly tetrasaccharides as the ACKNOWLEDGEMENTS smallest fragments, with N-acetylglucosamine at the reducing This study was supported by a grant from KAVITA Co. Mr. Hyenho Myeong from end, by random cleavage of b-N-acetyl-hexosamine (1-4) Korea BMI supplied the HYAL. glycosidic bonds.41 It has been reported that HA oligomers having 6–18 sugars can bind monovalently to CD44, and that HA oligomers composed of 20–38 sugars can form divalent bonds REFERENCES with CD44.48 Both bonds have been shown to attenuate 1 Haslam DW, James WP. Obesity. Lancet 2005; 366: 1197–1209. intracellular signaling of HA.49 As a result, it was supposed that 2 Green H, Kehinde O. Spontaneous heritable changes leading to increased adipose HA tetrasaccharides produced by exogenous HYAL treatment conversion in 3T3 cells. Cell 1976; 7: 105–113. cannot sufficiently stimulate intracellular signaling through 3 Green H, Meuth M. An established pre-adipose cell line and its differentiation in culture. Cell 1974; 3: 127–133. monovalent binding with CD44. 4-MU is an indirect inhibitor of 4 Chapman AB, Knight DM, Dieckmann BS, Ringold GM. Analysis of gene expression HA synthesis, which works by depleting UDP-glucuronic acid in during differentiation of adipogenic cells in culture and hormonal control of the cells. Therefore, the production of HA is downregulated by developmental program. J Biol Chem 1984; 259: 15548–15555. 28 blocking elongation of HA at the extracellular surface. The 5 Negrel R, Grimaldi P, Ailhaud G. Establishment of pre-adipocyte clonal line from common effects of exogenous HYAL and 4-MU treatment during epididymal fat pad of ob/ob mouse that responds to insulin and to lipolytic adipogenesis included inhibition of timely accumulation of high hormones. Proc Natl Acad Sci USA 1978; 75: 6054–6058. Mw HA, resulting in a low-affinity interaction of HA ligand with 6 Green H, Kehinde O. Formation of normally differentiated subcutaneous fat pads CD44. Thus, downregulation of HA affected both morphogenesis by an established preadipose cell line. J CellPhysiol 1979; 101: 169–171. and intracellular signal transduction of HA-induced adipogenesis. 7 Vannier C, Gaillard D, Grimaldi P, Amri EZ, Djian P, Cermolacce C et al. Adipose conversion of ob17 cells and hormone-related events. Int J Obes 1985; 9(Suppl 1): Our studies showed that the inhibition of adipogenesis by 41–53. downregulation of HA in vitro was also reflected on abdominal fat 8 Rosen ED, Sarraf P, Troy AE, Bradwin G, Moore K, Milstone DS et al. PPAR gamma is accumulation in HFD-feeding C57BL/6J mice. Our results showed required for the differentiation of adipose tissue in vivo and in vitro. Mol Cell 1999; that the epididymal fat tissue included in abdominal fat failed to 4: 611–617. develop after treatment with exogenous HYAL due to inhibition 9 Cao Z, Umek RM, McKnight SL. Regulated expression of three C/EBP isoforms of the accumulation of triglycerides in fat cells (Figure 4e). during adipose conversion of 3T3-L1 cells. Genes Dev 1991; 5: 1538–1552. In the obese population, fat cell size is an important factor due to 10 Mariman EC, Wang P. Adipocyte extracellular matrix composition, dynamics and the ability of large fat cells to secrete more adipokines. In addition, role in obesity. Cell Mol Life Sci 2010; 67: 1277–1292. the distribution of fat tissue is more significant than the total fat 11 Soukas A, Socci ND, Saatkamp BD, Novelli S, Friedman JM. Distinct transcriptional 50,51 profiles of adipogenesis in vivo and in vitro. J Biol Chem 2001; 276: 34167–34174. mass, as fat tissue has endocrine functions. It has been shown 12 Gregoire FM, Smas CM, Sul HS. Understanding adipocyte differentiation. Physiol that abdominal fat distribution is strongly correlated with the Rev 1998; 78: 783–809. 50 metabolic complications of obesity. Our results suggest that 13 McBeath R, Pirone DM, Nelson CM, Bhadriraju K, Chen CS. Cell shape, cytoskeletal exogenous HYAL treatment decreased fat cell size and reduced tension, and RhoA regulate stem cell lineage commitment. Dev Cell 2004; 6: abdominal fat tissue, eventually alleviating the insulin resistance 483–495. and fatty liver caused by abdominal obesity. 14 Aratani Y, Kitagawa Y. Enhanced synthesis and secretion of type IV collagen and To date, the effects of blocking HA have not been studied in entactin during adipose conversion of 3T3-L1 cells and production of unorthodox adipogenesis. Our in vitro and in vivo studies highlight the laminin complex. J Biol Chem 1988; 263: 16163–16169. potential therapeutic use of HA manipulation in obesity and its 15 Kawaguchi N, Sundberg C, Kveiborg M, Moghadaszadeh B, Asmar M, Dietrich N et al. related metabolic diseases. However, the anti-obese effects in ADAM12 induces actin cytoskeleton and extracellular matrix reorganization during early adipocyte differentiation by regulating beta1 integrin function. J Cell Sci HFD-diet-induced obese mice and anti-adipogenic effects in 2003; 116(Pt 19): 3893–3904. 3T3-L1 cells are difficult to reconcile as inhibition of adipogenesis 16 Demeulemeester D, Collen D, Lijnen HR. Effect of matrix metalloproteinase does not inhibit obesity. It has been believed that suppression of inhibition on adipose tissue development. Biochem Biophys Res Commun 2005; adipogenesis does not always correlate with inhibition of obesity 329: 105–110.

International Journal of Obesity (2014) 1035 – 1043 & 2014 Macmillan Publishers Limited Reduction of hyaluronic acid blocks adipogenesis EJiet al 1043 17 Coutinho PM, Deleury E, Davies GJ, Henrissat B. An evolving hierarchical family 33 Lynch MJ. Medical Laboratory Technology and Clinical Pathology. 2nd edn, classification for glycosyltransferases. J Mol Biol 2003; 328: 307–317. (W. B. Saunders Co.: Philadelphia, PA, USA, 1969. 18 Prehm P. Hyaluronate is synthesized at plasma membranes. Biochem J 1984; 220: 34 Farmer SR. Transcriptional control of adipocyte formation. Cell Metab 2006; 4: 597–600. 263–273. 19 Stern R, Asari AA, Sugahara KN. Hyaluronan fragments: an information-rich 35 Rosen ED, Hsu CH, Wang X, Sakai S, Freeman MW, Gonzalez FJ et al. C/EBPalpha system. Eur J Cell Biol 2006; 85: 699–715. induces adipogenesis through PPARgamma: a unified pathway. Genes Dev 2002; 20 Day AJ, Prestwich GD. Hyaluronan-binding proteins: tying up the giant. J Biol 16: 22–26. Chem 2002; 277: 4585–4588. 36 Lowe CE, O’Rahilly S, Rochford JJ. Adipogenesis at a glance. J Cell Sci 2011; 124: 21 Laurent TC, Fraser JR. Hyaluronan. FASEB J 1992; 6: 2397–2404. 2681–2686. 22 Calvo JC, Gandjbakhche AH, Nossal R, Hascall VC, Yanagishita M. Rheological 37 Rector RS, Thyfault JP, Wei Y, Ibdah JA. Non-alcoholic fatty liver disease and the effects of the presence of hyaluronic acid in the extracellular media of differ- metabolic syndrome: an update. World JGastroenterol 2008; 14: 185–192. entiated 3T3-L1 pre-adipocyte cultures. Arch Biochem Biophys 1993; 302: 468–475. 38 Ibrahimi A, Bonino F, Bardon S, Ailhaud G, Dani C. Essential role of collagens for 23 Allingham PG, Brownlee GR, Harper GS, Pho M, Nilsson SK, Brown TJ. Gene terminal differentiation of pre-adipocytes. Biochem Biophys ResCommun 1992; expression, synthesis and degradation of hyaluronan during differentiation of 187: 1314–1322. 3T3-L1 adipocytes. Arch Biochem Biophys 2006; 452: 83–91. 39 Lijnen HR, Maquoi E, Hansen LB, Van Hoef B, Frederix L, Collen D. Matrix 24 Zizola CF, Julianelli V, Bertolesi G, Yanagishita M, Calvo JC. Role of versican and metalloproteinase inhibition impairs adipose tissue development in mice. hyaluronan in the differentiation of 3T3-L1 cells into pre-adipocytes and mature Arterioscler ThrombVasc Biol 2002; 22: 374–379. adipocytes. Matrix Biol 2007; 26: 419–430. 40 Stern R. Hyaluronan catabolism: a new metabolic pathway. Eur J Cell Biol 2004; 83: 25 Halbleib M, Skurk T, de Luca C, von Heimburg D, Hauner H. Tissue engineering 317–325. of white adipose tissue using hyaluronic acid-based scaffolds. I: in vitro 41 Menzel EJ, Farr C. Hyaluronidase and its substrate hyaluronan: biochemistry, differentiation of human adipocyte precursor cells on scaffolds. Biomaterials 2003; biological activities and therapeutic uses. Cancer Lett 1998; 131: 3–11. 24: 3125–3132. 42 Hay ED. Development of the vertebrate cornea. Int Rev Cytol 1980; 63: 263–322. 26 Hemmrich K, von Heimburg D, Rendchen R, Di Bartolo C, Milella E, Pallua N. 43 Toole BP. Hyaluronan in morphogenesis. Semin Cell Dev Biol 2001; 12: 79–87. Implantation of pre-adipocyte-loaded hyaluronic acid-based scaffolds into nude 44 Bourguignon LY, Singleton PA, Diedrich F, Stern R, Gilad E. CD44 interaction mice to evaluate potential for soft tissue engineering. Biomaterials 2005; 26: with Na þ -H þ exchanger (NHE1) creates acidic microenvironments leading to 7025–7037. hyaluronidase-2 and cathepsin B activation and breast tumor cell invasion. J Biol 27 Flynn L, Prestwich GD, Semple JL, Woodhouse KA. Adipose tissue engineering Chem 2004; 279: 26991–27007. in vivo with adipose-derived stem cells on naturally derived scaffolds. J Biomed 45 Culty M, Nguyen HA, Underhill CB. The hyaluronan receptor (CD44) participates in Mater Res A 2009; 89: 929–941. the uptake and degradation of hyaluronan. J Cell Biol 1992; 116: 1055–1062. 28 Kultti A, Pasonen-Seppanen S, Jauhiainen M, Rilla KJ, Karna R, Pyoria E et al. 46 West DC, Hampson IN, Arnold F, Kumar S. Angiogenesis induced by degradation 4-Methylumbelliferone inhibits hyaluronan synthesis by depletion of cellular products of hyaluronic acid. Science 1985; 228: 1324–1326. UDP-glucuronic acid and downregulation of hyaluronan synthase 2 and 3. Exp Cell 47 Fieber C, Baumann P, Vallon R, Termeer C, Simon JC, Hofmann M et al. Res 2009; 315: 1914–1923. Hyaluronan-oligosaccharide-induced transcription of metalloproteases. J Cell Sci 29 Song NJ, Yoon HJ, Kim KH, Jung SR, Jang WS, Seo CR et al. Butein is a novel 2004; 117: 359–367. anti-adipogenic compound. J Lipid Res 2013; 54: 1385–1396. 48 Lesley J, Hascall VC, Tammi M, Hyman R. Hyaluronan binding by cell surface CD44. 30 Soria B, Roche E, Berna G, Leon-Quinto T, Reig JA, Martin F. Insulin-secreting cells J Biol Chem 2000; 275: 26967–26975. derived from embryonic stem cells normalize glycemia in streptozotocin-induced 49 Toole BP. Hyaluronan: from extracellular glue to pericellular cue. Nat Rev 2004; 4: diabetic mice. Diabetes 2000; 49: 157–162. 528–539. 31 Yeom CH, Lee G, Park JH, Yu J, Park S, Yi SY et al. High dose concentration 50 Jensen MD. Role of body fat distribution and the metabolic complications of administration of ascorbic acid inhibits tumor growth in BALB/C mice implanted obesity. J Clin Endocrinol Metab 2008; 93: S57–S63. with sarcoma 180 cancer cells via the restriction of angiogenesis. Journa Transl 51 Mirza MS. Obesity, visceral fat, and nafld: querying the role of adipokines in the Med 2009; 7:70. progression of nonalcoholic fatty liver disease. ISRNGastroenterol 2011; 2011: 32 Waki H, Yamauchi T, Kamon J, Ito Y, Uchida S, Kita S et al. Impaired multi- 592404. merization of human adiponectin mutants associated with diabetes. 52 Wang QA, Tao C, Gupta RK, Scherer PE. Tracking adipogenesis during white Molecular structure and multimer formation of adiponectin. J Biol Chem 2003; adipose tissue development, expansion and regeneration. Nat Med 2013; 19: 278: 40352–40363. 1338–1344.

Supplementary Information accompanies this paper on International Journal of Obesity website (http://www.nature.com/ijo)

& 2014 Macmillan Publishers Limited International Journal of Obesity (2014) 1035 – 1043