Ilomastat

Extracellular matrix remodeling and matrix metalloproteinase inhibition in visceral adipose during weight cycling in mice

Abstract
Extracellular matrix (ECM) remodeling is necessary for a health adipose tissue (AT) expansion and also has a role during weight loss. We investigate the ECM alteration during weight cycling (WC) in mice and the role of matrix metalloproteinases (MMPs) was assessed using GM6001, an MMP inhibitor, during weight loss (WL). Obesity was induced in mice by a high-fat diet. Obese mice were subject to caloric restriction for WL followed by reintroduction to high-fat diet for weight regain (WR), resulting in a WC protocol. In addition, mice were treated with GM6001 during WL period and the effects were observed after WR. Activity and expression of MMPs was intense during WL. MMP inhibition during WL results in inflammation and collagen content reduction. MMP inhibition during WL period interferes with the period of subsequent expansion of AT resulting in improvements in local inflammation and systemic metabolic alterations induced by obesity. Our results suggest that MMPs inhibition could be an interesting target to improve adipose tissue inflammation during WL and to support weight cyclers.

1.INTRODUCTION
A temporary storage of fat is one of the main roles of adipose tissue. During a positive energy balance, preadipocytes become mature adipocytes and their cell size increase, as well as in a negative energy balance, a reduction of the cells size is registered. Not only adipocyte cell shape but also the initial step of adipocyte differentiation and other process is necessary to adipose tissue development, such as angiogenesis, which depends on extracellular matrix (ECM) remodeling [1, 2]. Matrix metalloproteinases (MMPs), a family of zinc-dependent endopeptidase, play an essential role in regulating ECM remodeling, its deregulation was described in obese humans and in experimental models. An elevated expression of MMP-2 was initially described in adipose tissue of obese mice [3], as well as, higher MMP-2 and -9 expression in subcutaneous adipose tissue of overweight patients [4] or elevated serum MMP-2 and MMP-9 [5] in obese patients. From the initial reports, several works investigated the role of different MMPs in adipose tissue development (for review see Lin et al., [6]). Not much is known about the dynamics of the ECM during a negative energy balance. Weight loss (WL), after bariatric surgery, provided improvements in glucose homeostasis and lipid metabolism, but serum levels of MMP-2, MMP-3 and specific inhibitors of metalloproteinases (TIMPs) were not affected, instead MMP-7 serum levels were increased [7]. In another study that followed patients after bariatric surgery, MMP-2 and -9 activity and picrosirius-red-stained collagen accumulation was observed in parallel of collagen I and III degradation in subcutaneous adipose tissue [8]. In visceral adipose tissue of mice after WL, protein level of MMP-2 was maintained, and a higher activity was observed in zymography assay [9]. If little is known about ECM remodeling in a negative energy balance period, there are no reports about this issue during weight cycling (WC), the repeated loss and regain of body weight [10]. In the modern society, “yo-yo dieting” is a controversial topic of debate, whether WC could lead to adverse health consequences providing more weight gain and increased metabolic risk [10, 11].The present study aimed to investigate the effects of WC on MMPs expression and activity. Adipose tissue inflammation and metabolic parameters were also evaluated. In addition, mice were treated with an MMP inhibitor, GM6001 (ilomastat) during WL period and the effects of ECM remodeling intervention were observed after WR.

2.METHODS
Specific pathogen-free, 5-week-old male Swiss mice were obtained from Centro Multidisciplinar de Investigação Biológica (CEMIB, State University of Campinas, Campinas, São Paulo, Brazil). All experiments were performed in accordance to the principles outlined by the National Council for Animal Experimentation Control (CONCEA, Brazil) and it received approval from Ethics Committee of São Francisco University, Bragança Paulista, SP, Brazil (Protocol CEA/USF 001.05.2014). The mice were initially divided into two groups that received AIN-93 diet (lean group) or high-fat diet (obese group) ad libitum for 8 weeks. Pelletized AIN-93G was purchase from Rhoster (Cat RH 19521), Brazil. HFD was prepared in our lab as described before [12]. The obese group was later subdivided into 3 groups, one group was maintained with HFD ad libitum and two groups were subjected to moderate caloric restriction by providing daily controlled amounts of AIN-93 (the equivalent amount consumed by lean group in the day before) for the subsequent 8 weeks (weight loss groups, WL). One of WL groups was treated with an MMP inhibitor – GM6001 (a broad spectrum MMP inhibitor, 100 mg/kg/day, ip.) throughout the period and it was denominated WLGM. At the end of 16 weeks, the animals were sacrificed resulting in the following experimental groups Lean16, Obese16, WL and WLGM (n=5 each group). The protocol was repeated and WLs groups were reintroduced to HFD ad libitum for additional 8 weeks (weight regain groups, WR). At the end of 24 weeks, Lean24, Obese24, WR and WRGM groups were obtained (n=5 each group). The GM6001 (Tocris bioscience, Bristol, UK) dosage employed in this study was based in previous literature reports [13- 15] and in both protocols (16 and 24 weeks) GM6001 was administered to mice only during WL period. The food intake was recorded in lean16, obese16, lean24, obese24, WR and WRGM groups only in the last two weeks of the protocol (fed ad libitum).

Animals were fasted for 6 h, and glucose homeostasis was evaluated by glucose blood level and insulin tolerance test (ITT) as previously described [16]. The rate constant for glucose disappearance during an insulin tolerance test (kITT) was calculated using the formula 0.693/t1/2. The glucose t1/2 was calculated from the slope of the least square analysis of the plasma glucose concentrations during the linear decay phase [17].Drops of blood were employed for total cholesterol and triglycerides measurement using Accutrend Plus (Roche Diagnostics, Mannheim, Germany).Indirect calorimetry was carried out using the Oxylet/Physiocage system (Panlab, Barcelona, Spain). Animals were individually placed in respiratory chambers (temperature: 22-23°C, humidity: 45-55% light cycle dark 12/12 hours) with an air flow of 0.5 L/min. During the 24 hours of analysis, O2 (%) and CO2 (%) were measured every 9 minutes. The Software Metabolism (Panlab) calculated the O2 consumption, CO2 and energy expenditure (kCal/h/kg0.75).Mice were fasted for 6 h and anesthetized by xylasine/ketamine overdose (0.1 mL/30g body weight of 1:1 v/v of 2% xylasine and 10% ketamine); blood samples were collected by cardiac puncture in tubes without anticoagulant and centrifuged to isolate the serum. Adipose tissue depots (epididymal, subcutaneous, perirenal and mesenteric) and liver were carefully dissected, weighted and expressed as a percentage of body weight (b.w.). Epididymal adipose tissue samples were collected and stored at -80°C for further analyses or employed immediately for stromal vascular fraction isolation.

Epididymal adipose tissue samples were digested by collagenase method [18]. The stromal vascular fraction (SVF) was used for cell analysis. Cells (106 cells) were incubated with anti-CD45PerCP (BD Biosciences, CA, USA) for total leukocyte count or with anti-CD14FITC/anti-F4/80PE/anti-CD11bPerCP for macrophage identification (BD Biosciences). For each sample, 10,000 events were collected on a Guava Easy-Cyte HT (Millipore, Hayward, CA, USA) cytometer, defining FSC, SSC on a linear scale and FL1, FL2, and FL3 on a logarithmic scale. Light scatter profiles were obtained for each candidate population using InCyte software (Millipore).Epididymal adipose tissue biopsies were homogenized in solubilization buffer as described [16]. Total protein extracts were obtained and used in Multiplex Assay kit for MMPs quantification (Milliplex Mouse MMP3MAG-79K, Merck Millipore, MA, USA). MMP activity was determined by gelatin zymography as described [9]. Gels were stained with 0.25% Coomassie brilliant blue R-250 and then destained with 10% acetic acid in 40% methanol. Gels were photographed using Gel Documentation system (Gel DocTM XR+ v. 5.0, Bio-Rad, CA, USA) and bands with MMP activity identified by the Image LabTM Software (Bio-Rad).
Insulin was measured in serum using Multiplex kit (Mouse Adipokine, Millipore). HDL and LDL cholesterol serum levels were determined using a commercial kit (LABORLAB, Sao Paulo, Brazil). Adiponectin, leptin, IL-6 and MCP-1 levels in adipose tissue were measured using Milliplex kit (Mouse Adipocyte, Millipore). IL-10 and TNF- was measured by EIA kit (Quantikine Elisa mouse IL-10, R&D Systems, MN, USA; TNF alpha Mouse Elisa Kit, Abcam, Cambridge, UK).

Epididymal and subcutaneous adipose tissue samples were fixed with paraformaldehyde. Subsequently, the specimens were processed, and embedded in paraffin. Histological sections of 5.0 µm were stained with hematoxylin–eosin to determine the size of adipocytes. The area of each intact cell on each image (500 cells per group) was measured by drawing the circumference and the adipocyte area was then extrapolated using ImageJ software (http://rsbweb.nih.gov/ij/). Additional sections were stained with Picro-Sirius Red for collagen analysis (mainly collagen I and III) and it was also analyzed using Image J software.Hydroxyproline was measured using a hydroxyproline colorimetric assay as described [19]. Epididymal adipose samples were weighted, dehydrated in acetone, degreased in petroleum ether and dried. Samples were hydrolyzed in 6 N HCl for 18 hours at 110ºC and neutralized with 6 N NaOH. Twenty microliters of supernatant was treated with of chloramine T solution and Erlich’s solution (4-dimethylamino-benzaldehyde in perchloric acid) and incubated for 20 minutes at 65ºC. The absorbance was measured at 540 nm and the concentration was determined using the standard curve created with hydroxyproline (0.2 to 6 µg/mL).The relative expression levels of different genes in the epididymal adipose tissue samples were quantified by real-time polymerase chain reaction (PCR). Total RNA extraction, cDNA synthesis and quantitative PCR were performed as previously described [20], using a 7500 real-time PCR system (Applied Biosystems). Expression levels were determined using 2−ΔΔCt relative quantification and were normalized to control gene (18S) and represented as fold change with respect to control (lean16 and lean24) group. Primers sequences were as follows: Col1a1 – FW 5’- GTGCTCCTGGTATTGCTGGT-3’; Col1a1 – RV 5’-GGCTCCTCGTTTTCCTTCTT-3’; Col3a1 – FW 5’-GGGTTTCCCTGGTCCTAAAG-3’; Col3a1 – RV 5’- CCTGGTT TCCCATTTTCTCC-3’; Col6a1 – FW 5’-GATGAGGGTGAAGTGGGAGA-3’; Col6a1 – RV 5’-CAGCACGAAGAGGATGTCAA-3’; Eln – FW 5’- TGGTATTGGTGGCATCGG-3’; Eln – RV 5’-CCTTGGCTTTGACTCCTGTG-3’; Lox – 5’-CCACAGCATGGACGAATTCA-3’; Lox – RV 5’-AGCTTGCTTTGTGGCCTTCA-3’;18S – FW 5’- AAACGGCTACCACATCCAAG-3’ and 18S – RV 5’- CAATTACAGGGCCTCGAAAG-3’.Data were expressed as mean±SEM. Statistically significant differences (p values
<0.05) were determined using analysis of variance (ANOVA) followed by Dunnett's test for multiple comparisons or Kruskal-Wallis test using GraphPad Instat (GraphPad Software, Inc., La Jolla, CA, USA). 3.RESULTS Mice fed on HFD for 8 weeks presented increase body weight (Figure 1) and insulin resistance resulting in high fasted blood glucose levels and reduced decay of glucose blood levels during insulin tolerance test (data not shown). After a moderate caloric restriction with AIN-93, mice lost weight and reached the body weight of age-matched lean controls (Figure 1). Reintroduction of HFD for additional 8 weeks reestablished the obese status (Figure 1). The mice’s body composition was not different when compared to mice that lost weight (WL) with the same age lean mice (Lean16). Mice’s body composition after weight regain (WR) was not different from age-matched obese mice (Obese24; Table 1). The WL resulted in reductions of fasted glucose and insulin blood levels and, insulin resistance if compared with age-matched obese but these parameters were also different from age-matched lean mice. WR reestablished the insulin resistance with a more pronounced hyperinsulinemia (Table 2). Total cholesterol, LDL cholesterol, triglycerides blood levels were also reduced in mice that lost weight, but again the levels did not reach the levels observed in lean mice. In mice that regained weight, we did not observe differences in cholesterol and triglycerides level when compared with age-matched obese (Table 2).MMP-2, -3, -8, pro-9 and -12 expressions were increased in epididymal adipose tissue of obese animals and after WL, MMP-2 and -12 expressions were not restored to levels found in lean mice with the same age (Figure 2A). The gelatinases activity was higher in mice’s adipose tissue that lost weight (Figure 2B). The adipocyte size was reduced after weight loss as expected (Figure 3G). Pericellular collagen (mainly I and III) analysis by Sirius red stain and hydroxyproline content (total collagen) were increased by obesity and they were not different in WL if compared with obese (Figure 3E and F, respectively). Genes encoding proteins involved in ECM structural components, such as collagen I (Col 1a1), collagen III (Col 3a1), collagen VI (Col 6a1) and elastin (Eln) were increased in obese group and remained unchanged after WL, with the exception of elastin which was reduced after WL (Figure 3H). In addition, we observed an increase in the gene expression of the cross-linking enzyme, lysyl oxidase (Lox) in the obese group, but after WL this gene was down regulated (Figure 3H). An brief analysis of subcutaneous adipose tissue revealed the increased of pericellular collagen only in obese group, considering that cell size is increased and therefore the amount of fluorescence by Sirius red stain is reduced in each high-power field (Figure 5A-F).MMP-2, -8, pro-9 and -12 expression was higher in WL/WR animals if compared with age-matched obese, but the gelatinases activity showed no difference between the groups (Figure 2C and D). The adipocyte size at the end or regain period was not different from age-matched obese (Figure 4G). Pericellular collagen I and III content was reduced but hydroxyproline concentrations (total collagen) was similar if compared with obese mice (Figure 4E and F; respectively). The collagen I (Col 1a1), collagen III (Col 3a1), collagen VI (Col 6a1) and elastin (Eln) genes as well as lysyl oxidase (Lox) presented a large variation after WR, becoming indifferent to the lean group, but also to the obese group (Figure 4H). Again, an analysis of subcutaneous adipose tissue indicate a deposition of pericellular collagen I and III if considered that adipocyte size is increased in obese and WR group (Figure 5G - L). If we consider the local leptin and adiponectin production, inflammation was observed in visceral adipose tissue in obese mice, and it was reduced in mice that lost weight (Table 3). Moreover, a higher number of leukocytes and macrophages remained in stromal vascular fraction of this mice group as well as a higher level of TNF- production (Table 3). WR reestablished the local adipokine production at the same levels of mice that remained obese during 24 weeks. A higher number of leukocytes and macrophages infiltrated was observed if compared to obese mice (Table 3).Mice treated with GM6001 during WL period presented no changes in body weight (Figure 1), adiposity (Table 1) and biochemical parameters (Table 2) if we compare to mice that lost weight without treatment. However, we observed a reduction in the basal insulin level and energy expenditure (EE) (Table 2) in these mice. MMPs protein expression was not altered by GM treatment (Figure 2A) and MMP-2 and -9 activities were inhibited in adipose tissue after GM treatment (Figure 2B). Epididymal adipose tissue from mice treated with GM during WL showed no differences in the final adipocyte size (Figure 3G). We observed a reduction in the pericellular collagen I and III (Figure 3D) and total collagen estimated by hydroxyproline concentrations in visceral adipose tissue (Figure 3F) when compared to matched controls. However, collagen I (Col 1a1), collagen III (Col 3a1), collagen VI (Col 6a1), elastin (Eln) and lysyl oxidase (Lox) gene expression were not different when compared with non-treated WL group (Figure 3H). In subcutaneous adipose tissue no differences were observe regarding pericellular collagen deposition and final adipocyte size between GM-treated or WL groups (Figure 5). GM treatment reduced MCP-1 and TNF- in adipose tissue (Table 3) although the number of leukocytes and macrophages remained in stromal vascular fraction (Table 3). Mice that received GM during WL and were submitted to WR showed no difference in body weight or adiposity, (Figure 1; Table 1) only the liver weight was significantly reduced compared to the matched controls (Table 1). No alterations in food intake were registered during GM treatement (Table 2).WR in mice treated with GM during WL resulted in reduction of triglycerides and LDL cholesterol blood levels and improvements in glucose tolerance (Table 2). MMP-12 expression was decreased in epididymal adipose tissue with GM treatment, but the gelatinases activity were not different between the groups (Figure 2C and D). GM treatment did not altered the adipocyte size (Figure 4G), pericellular or total collagen or gene expression of collagen I (Col 1a1), collagen III (Col 3a1), collagen VI (Col 6a1), elastin (Eln) and lysyl oxidase (Lox) in visceral adipose tissue if compared with non-treated group (Figure 4D-F and H). However, the adipocyte size was reduced in subcutaneous adipose tissue by GM treatment (Figure 5 J-L). Besides that, after WR in the GM treated mice, an increased adiponectin and IL-10 expression was observed in visceral adipose tissue, as well as, an increase in leptin, but the TNF- level was reduced (Table 3). An increased infiltration of leukocytes and macrophages was still observed in this local (Table 3). 4.DISCUSSION Regulated extracellular matrix deposition and degradation is necessary for a health adipose tissue expansion, without adverse metabolic consequences [21]. However, interstitial fibrosis is observed in adipose tissue during obesity development and it has been associated with adipocyte dysfunction and local inflammation [22]. During WL induced by bariatric surgery in obese patients, fibrosis and collagen accumulation in adipose tissue persist in some individuals and it was associated with reduced fat mass loss combined with additional predictive factor such as age, diabetes and IL-6 levels [23]. However, in an abrupt reduction of adipocyte cell size observed in cancer patients with cachexia, it is described a significant increase of fibrosis in adipose tissue [24, 25]. Metabolic disturbances improve after WL by bariatric surgery, but WR is common and potentially compromises the health benefits of the surgery [26]. Weight cycling or “yo- yo diet” is also common between obese patients that intentionally lose weight, and this lost is followed by WR. Weight cyclers had higher fasting insulin [27], impaired glucose tolerance [28, 29] and greater risk for metabolic syndrome [30] regardless of BMI if compared to non-cyclers, but there is no information about fibrosis/inflammation on adipose tissue in this context. Therefore, our primary subject in this work was to study extracellular matrix alterations in visceral adipose tissue in a model of WC in mice. Caloric restriction was able to provide a significant WL in mice accompanied by improvements in glucose and lipid homeostasis, although the final metabolic status was still different from lean mice. Our results are in agreement with previous published data showing that after WL mice retain alterations in glucose homeostasis [31].Inflammatory markers in adipose tissue was not totally absent, because we could still observe an intense macrophage infiltration beside the MCP-1 and TNF-α higher levels after WL, suggesting that adipose tissue macrophage phenotype is predominantly M1 after WL. The protein expression of MMP-2 and MMP-12, gelatinase activity (MMP-2 and MMP-9), pericellular and total collagen content, as well as a mRNA expression of Col 1a1, Col 3a1 and Col 6a1 were higher in visceral adipose tissue, suggesting that an intense adipose tissue remodeling was maintained during WL. Our results are in agreement with a recent published data, where authors registered inflammation and fibrosis even after WL in mice [32]. An increased MMP-2 activity and mRNA expression were also described previously in subcutaneous and gonadal fat mass of mice after 6 week of caloric restriction [9]. Visceral adipose tissue fibrosis could represent a maladaptive mechanism that contributing to deleterious metabolic complications not only during obesity development but also during weight loss period. Adipocyte progenitors (platelet-derived growth factor receptor- + CD9high cells) seem to be the responsible for ECM deposition and fibrosis in adipose tissue during obesity [33]. These adipocyte progenitors are also identified in omental adipose tissue and correlated with local fibrosis level and with insulin resistance in humans [33].After a restriction caloric period, mice were reintroduced to a high-fat diet for 8 weeks and they reached the final body weight observed for mice fed with high-fat diet during 24 weeks. Only basal insulin blood levels were worsened in WC mice when compared to age-matched obese control. While the observed adipose tissue inflammation was similar in WC and obese mice, the protein expression of several MMPs was up- regulated in WC mice. However, gelatinase activity was not increased. Total collagen content was increased, but pericellular collagen was reduced in adipose tissue of WC mice, suggesting that fibrosis was not strictly related to inflammation in the adipose tissue but it was also modified by adipose tissue expansion and regression cycles.Alterations of pericellular collagen deposition were more observable in visceral than in subcutaneous adipose tissue in Swiss mice. Different strains of mice have different fibrotic profiles in adipose tissue and it is related to fibrotic progenitors presence [33]. These fibrotic progenitors cells (CD9highPDGFR+) also have a pro-inflammatory phenotype, but in PdgfraKI mutants, Marcelin and coworkers observed that fibrosis induced by PDGFR activation was not associated with inflammation markers [33], corroborating our hypothesis that fibrosis was not strictly related to inflammation in the adipose tissue during weight cycle. We could not observe a correlation between cell size and pericellular collagen deposition in our experiments. In a recent work, the reduction of cell size in subcutaneous adipose tissue of bariatric patients was followed by an intense pericellular collagen deposition, but no alterations in adipose tissue stiffness were recorded by the authors [34]. A down-regulation of cross-linking enzymes as lysyl oxidase (LOX) was observed indicating that matrix fibers cross- linking was reduced as well as degraded collagen I and III presence [34]. Our study has limitations by not identify the fibrotic cell source or measured the stiffness and presence of degraded collagen in adipose tissue, but also showed a reduction in Lox and Eln gene expression, suggesting that ECM could be less rigid during WL and that pericelullar collagen deposition could be due an intense ECM remodeling and not due a deleterious fibrosis. Additionally, we treated mice during WL period with non-specific synthetic inhibitors of MMPs, GM6001, because we observed an intense gelatinase activity after this period. Our results showed that GM6001 treatment during caloric restriction did not modify the final body composition, but the insulin level was reduced when compared to non- treated WL group. Lipid homeostasis and M1 macrophage markers (TNF-/MCP-1) were similar to lean mice, suggesting an improvement in metabolism and adipose tissue inflammation. The energy expenditure, measured by indirect calorimetry, indicated that obese animals had a decreased energy expenditure and it was not reversed after WL. Interestingly, the treatment with GM6001 reduced even more this parameter. The reduction in basal energy expenditure is a metabolic adaptation to WL, contributing to survival of the organism during periods of food restriction, but this adjustment during intentional WL increases the propensity to WR [10, 11]. The hyperinsulinemia, insulin resistance and thermogenesis suppressed in individuals who have undergone restrictive diet are also related to propensity to WR [36, 37] and in this case, GM6001 has a beneficial effect in reducing insulinemia. The improvements of collagen deposition and inflammation in adipose tissue induced by GM6001 treatment during WL can explain the improvements in insulinemia. However, our in vivo data did not permit us to confirm if the TNF- reduction is a consequence of reductions in matrix remodeling in adipose tissue, or if GM6001 can directly affect macrophage activity by reducing TNF- release in parallel with MMP inhibition. The MMP role in inflammatory cell migration is well recognized, but the relationship of MMPs and cytokine production is not fully explored. TNF- , leukotriene D4, prostaglandin E2 and toll-like receptors were able to induce MMPs expression and activity in macrophages, as well as, anti- inflammatory cytokines, IL-10 and transforming growth factor-β inhibit it [38, 39]. Conditioned macrophage medium, from cultures, increases TNF- production in 3T3- L1 cultures and it was abolished by GM6001 or NNGH (N-isobutyl-N-(4- methoxyphenylsulfonyl)-glycylhydroxamic acid), an MMP-3 inhibitor, suggesting that, the active MMP-3 could work as a pro-inflammatory factor signaling in macrophage/adipocyte crosstalk [40]. An interesting work, using a 3D-culture model of decellularized ECM and human mature adipocytes from obese and obese diabetes mellitus type 2 patients (DM2), demonstrate that adipocytes derived from subjects without DM2 when differentiated in ECM from DM2 subjects had insulin-stimulated glucose uptake suppressed [41], suggesting that ECM regulates adipocyte metabolism and it could also explain our results with GM6001 during weight loss.GM6001 administration during WL did not alter visceral adipocytes size and mRNA expression of ECM components, but reduced total and pericellular collagen, suggesting that MMP has an important role in ECM remodeling during weight loss in adipose tissue. In a previous published work, mice treated with GM6001 during weight gain presented small adipocytes associated to a reduced weight gain, but a higher amount of collagen [13], contrasting with our results. Curiously, in other experimental models GM6001 treatment was also associated with reductions in fibrosis/collagen deposition [42], as registered in our experimental model. When GM6001-treated mice (only during WL) were reintroduced to a HFD, visceral adipose tissue had expanded normally, but they produced lower levels of IL-6 and TNF, high levels of adiponectin and IL-10 and also high levels of leptin. We also observed an improvement in insulin resistance and lower triglycerides, LDL-cholesterol and total cholesterol levels in this obese group, that were not related to changes in energy expenditure, but probably, they were associated to a reduction on inflammation of visceral adipose tissue. Although, the characteristics (genes expression ECM components, collagen content, MMPs activity and adipocytes size) were not different between animals treated or untreated with GM6001, at the end of the regain period, we must consider that, the beginning of WR process occurred in a modified environment, (reduced ECM remodeling, inhibited MMP’s and less inflammation) and these characteristics could have driven the additional adipose tissue expansion cycle by a beneficial way. A previous description of a short term treatment of diabetes people with an MMP inhibitor, doxycycline, also demonstrated an anti-inflammatory activity and improvements in insulin sensitivity [43], suggesting that MMP inhibitors as GM6001 could have beneficial effects on obesity. 5.CONCLUSION Our results allow us to conclude that after WL an inflammation and an intense MMP activity can remain in adipose tissue. MMP blockade during WL interferes with ECM remodeling and inflammation in visceral adipose tissue, suggesting that ECM is also important during cell size reduction and MMPs could act as important players for inflammation Ilomastat signaling. Metabolic and adipose tissue inflammation improvements are observed when these mice become obese by WR suggesting that the MMPs inhibition could be an interesting target to support weight cyclers.