New insight into the mechanisms of ectopic fat deposition improvement after bariatric surgery

Non-alcoholic fatty-liver disease (NAFLD) is frequent in obese patients and represents a major risk factor for the development of diabetes and its complications. Bariatric surgery reverses the hepatic features of NAFLD. However, its mechanism of action remains elusive. We performed a comprehensive analysis of the mechanism leading to the improvement of NAFLD and insulin resistance in both obese rodents and humans following sleeve-gastrectomy (SG). SG improved insulin sensitivity and reduced hepatic and monocyte fat accumulation. Importantly, fat accumulation in monocytes was well comparable to that in hepatocytes, suggesting that Plin2 levels in monocytes might be a non-invasive marker for the diagnosis of NAFLD. Both in vitro and in vivo studies demonstrated an effective metabolic regeneration of liver function and insulin sensitivity. Specifically, SG improved NAFLD significantly by enhancing AMP-activated protein kinase (AMPK) phosphorylation and chaperone-mediated autophagy (CMA) that translate into the removal of Plin2 coating lipid droplets. This led to an increase in lipolysis and specific amelioration of hepatic insulin resistance. Elucidating the mechanism of impaired liver metabolism in obese subjects will help to design new strategies for the prevention and treatment of NAFLD.

ATP synthesis 12,13 . Ezquerro et al. 14 found that in rats under a high-fat diet sleeve gastrectomy improved NAFLD by reducing hepatosteatosis and circulating transaminase levels through the downregulation of lipogenesis and the increase of both autophagy and mitochondrial β-oxidation.
NAFLD is characterized by the hepatocyte accumulation of neutral lipids forming lipid droplets (LDs), known as hepatic steatosis, surrounded by proteins of which Perilipin 2 (Plin2) is a major component 15 . Plin2 deletion or suppression of Plin2 expression protect from NAFLD development in genetic models of obesity or in rodents fed a high-fat diet (HFD) 16,17 .
Chaperone-mediated autophagy (CMA) promotes Plin2 catabolism and, therefore, contributes to LD degradation by facilitating lipolysis 18 . The heat shock cognate protein of 70 kDa (Hsc70) recognizes the pentapeptide motif of Plin2 19 and transports this protein to the lysosome where Plin2 binds to the CMA receptor, LAMP2A (lysosome-associated membrane protein 2A), and is translocated into the lysosomal lumen where is degraded 20 . An important step that allows the binding of Plin2 to LAMP2A is its previous phosphorylation that is dependent on AMPK.
We hypothesize that the energy imbalance generated by SG upregulates liver AMPK and promotes Plin2 degradation leading to the improvement of NAFLD/NASH. Therefore, we explored these mechanisms performing SG in diet-induced obesity, NAFLD in rats and in subjects with NAFLD.
We observed that ectopic fat deposition in rodents with diet-induced obesity (DIO) included not only hepatic steatosis but also lipid deposition in circulating monocytes in concert with increased Plin2 expression. We found that SG increased AMPK Thr172 phosphorylation, LAMP2A expression and reduced Plin2 expression in both hepatocytes and monocytes. SG reduced the number of lipid droplets and increased the glycogen storage both in hepatocytes and monocytes; Akt Ser473, Glycogen Synthase Kinase (GSK3αβ) Ser21/9 and Forkhead Box O1 (FoxO1) Thr24 phosphorylation were all increased. Plin2 overexpression in primary cultures of hepatocytes and monocytes reduced Akt phosphorylation and lead to insulin resistance.
Together, these data suggest that liver fat accumulation and hepatic insulin resistance share a common pathway coordinated by Plin2.

Animal study. SG reduces hepatic fat accumulation in diet-induced obesity and significantly reduced Plin2
protein expression in both liver and monocytes. Body weight was significantly lower in the SG than in the sham group (534.10 ± 25.71 vs. 336.40 ± 10.31 g, P = 0.0003). Compared with sham operation, SG rats drastically reduced fat accumulation in both liver and monocytes of DIO rats (Fig. 1, Panels A-D).
Human study. The anthropometric characteristics of the obese patients, before and after SG, as well as those of controls are summarized in Table 1. The weight reduction was ca. 32% but the weight of the patients remained significantly higher than that of controls. Sleeve gastrectomy improved significantly the lipid profile and reduced significantly transaminases.
Plin2 protein expression in monocytes is significantly reduced after sleeve gastrectomy and strongly correlates with the NAS score. Compared with controls, subjects before SG show high fat accumulation (Fig. 3, Panels A-D). Plin2 protein expression of both liver and monocytes significantly and positively correlated with the NAS score (Liver: R = 0.52, P = 0.0025; Monocytes: R = 0.58, P = 0.00099). Moreover, the expression of Plin2 in monocytes strongly and positively correlated with Plin2 expression in the liver (Before SG: R = 0.70, P < 0.0001; Controls: R = 0.90, P = 0.0007). Figure 3, Panel E shows that Plin2 protein expression in the liver was significantly lower in controls than in obese subjects before SG (0.76 ± 0.13 vs. 0.28 ± 0.15, P = 0.03). Moreover, Plin2 protein expression in monocytes decreased from 0.49 ± 0.16 vs. 0.21 ± 0.04, P = 0.01) after SG (Fig. 3, Panel F).
AKT Ser473 GSK3αβ and FoxO1 phosphorylation significantly decrease by overexpressing Plin2 in primary cultures of hepatocytes and monocytes. In order to prove that a high Plin2 expression is responsible for the Akt Ser473, GSK3αβ Ser21/9 and FoxO1 Thr24 reduced phosphorylation, we overexpressed Plin2 by transfecting human Plin2 plasmid in primary cultures of hepatocytes and monocytes from controls.
Moreover, Nile Red staining shows a reduction in lipid droplet accumulation after both pioglitazone (Fig. 7, Panel C) and metformin (Fig. 7, Panel D) exposure in human hepatocytes primary cultures.

Discussion
Non-alcoholic fatty-liver disease (NAFLD) is currently the most frequent liver disorder in the Western world 1 . Diabetes is a major risk factor for the development of NAFLD. Currently, there are no FDA-approved drugs specifically tailored for NAFLD or NASH. Bariatric/metabolic surgery is effective in reversing the hepatic features of NAFLD and NASH, as shown by retrospective or prospective but non randomized studies 11 . However, the mechanism of action of bariatric/metabolic surgery remains elusive.
In the present study, we provide evidences both in rodents and in humans that after SG there is a net increase of phosphorylated AMPK and LAMP2A expression, causing a decrease in Plin2 expression and a reduction in LDs accumulation in both liver and circulating monocytes. In addition, we observed an increased phosphorylation of Akt Ser473, GSK3αβ Ser21/9 and FoxO1 Thr24 and an increased glycogen deposition in liver and monocytes of both rodents and humans after SG.
We also report that fat accumulation in monocytes was well comparable to that in hepatocytes, suggesting that Plin2 levels in monocytes might be a non-invasive marker for the diagnosis of NAFLD.
The changes in phosphorylation of proteins of interest were independent of changes in body weight both in humans and in rats suggesting a weight-independent mechanism of action of SG. Most beneficial effects of sleeve gastrectomy are, in fact, achieved in a weight-independent manner as shown by Frühbeck 21  www.nature.com/scientificreports www.nature.com/scientificreports/ the mechanism of action is not completely elucidated. One of the pathways affected by caloric restriction is the AMPK one. AMPK controls glucose disposal by inhibiting liver gluconeogenesis and increasing muscle glucose uptake [22][23][24][25] . Targeting hepatic AMPK with an AMPK-specific small-molecule activator reduces liver steatosis and improves glucose disposal in obese mice 26,27 . Recently, Garcia et al. 28 showed that liver-specific AMPK activation in DIO mice protects against liver steatosis.
Our data show a net increase of AMPK phosphorylation after SG in both hepatocytes and monocytes either in rodents or in humans. AMPK phosphorylation is a crucial step for the binding of Plin2 to LAMP2A that leads to an increased Plin2 catabolism and lipid droplets exposure to lypolysis 19 . Indeed, after SG we found a significant reduction of Plin2 expression and lipid droplets accumulation.
Bariatric/metabolic surgery procedures that reduce caloric intake, including SG, have shown high rates of remission of type 2 diabetes mostly due to the improvement of insulin sensitivity 7 . In fact, here we report a significant increase of Akt Ser473, GSK3αβ Ser21/9 and FoxO1 Thr24 phosphorylation in both liver and monocytes, which points towards an improved hepatic insulin sensitivity with reduced gluconeogenesis and increased glycogen synthesis. Indeed, the inhibition of GSKαβ activity through its phosphorylation leads to increased glycogen synthesis, while the inhibition of FoxO1 activity, again through its phosphorylation, reduces gluconeogenesis 29 . www.nature.com/scientificreports www.nature.com/scientificreports/ These results are supported by in vitro data showing that overexpression of Plin2, using a human-Plin2 plasmid in primary hepatocyte and monocyte cultures, significantly affects insulin signaling decreasing Akt Ser473, GSK3αβ Ser21/9 and FoxO1 Thr24 phosphorylation.
Several studies have shown that AMPK is activated by two major antidiabetic drugs, metformin and pioglitazone 19,30 . We report that treatment of primary hepatocytes with pioglitazone or metformin reduces Plin2 expression, LDs accumulation and improves insulin signaling.
When insulin binds rodent or human hepatocytes it is internalized and preferentially associated with lysosomal structures 20,31 . The upregulation of LAMP2A we found might, thus, reflect the improvement of hepatic insulin resistance after sleeve gastrectomy that was shown by the significant reduction of HOMA-IR values.
The lack of post-surgery liver biopsies and the lack of a pair-fed group of rats represent two limitations of our study, although our results in humans are supported by the similar findings associated with SG in rodents. We showed, in fact, a high correlation between LDs and Plin2 content in the liver and in monocytes that were available also after SG. Furthermore, the surrogate marker HOMA-IR, that has been shown to be highly correlated with hepatic insulin resistance as assessed by the gold standard euglycemic hyperinsulinemic clamp 32 , was used to assess hepatic insulin resistance.
In conclusion, we propose that excessive caloric intake, which is associated with NAFLD and obesity, reduces AMPK activation, causing a decrease of Plin2 catabolism mediated by chaperone-mediated autophagy that leads to an increase of ectopic fat deposition and simultaneously to hepatic insulin resistance. After SG, the www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ energy imbalance due to reduced caloric intake causes an increase in AMPK phosphorylation, a crucial step in Plin2-LAMP2A binding, leading to an enhanced Plin2 autophagy that exposes LD triglycerides to intracellular lipases. In addition, AMPK increased phosphorylation improves insulin signaling and promotes glycogen synthesis (Fig. 8).

Methods
Animal model. Thirty Wistar rats of both sexes aged 10 weeks were housed individually in a controlled room at 22 °C with a 12-h day/night cycle (lights on 0700-1900 h). The animals received a purified tripalmitin-based High fat liquid diet (HFD) ad libitum (Rieper AG, Bolzano, Italy) supplying 71% of energy from saturated fat, although corn oil (1.9/100 g diet) was present in order to prevent essential fatty acid deficiency, 20% from carbohydrate comprising cornstarch and sucrose (2:1 weight for weight) and 10% from protein. The HFD was continued for 10 weeks before and 15 weeks after the operation. The animals randomly underwent SG or sham operation. Survival rates were 90% after sham operation, and 75% after SG. All experimental procedures were approved by the Catholic University of Rome Institutional Animal Care Committee and all methods were performed in accordance with the relevant guidelines and regulations 33 . interventions. The rats were anesthetized using ketamine (75 mg/kg intramuscularly) and xylazine (10 mg/ kg intramuscularly). Ten milliliters of sterile 0.9% NaCl were administered subcutaneously before surgery. Access to the peritoneal cavity was obtained by 3-cm laparotomy 33 .
Vertical sleeve gastrectomy. A laparotomy incision was made in the abdominal wall and the stomach was isolated outside the abdominal cavity and placed on saline-moistened gauze pads. The gastric to spleen connections were released along the greater curvature. The lateral 80% of the stomach was excised leaving a tubular gastric remnant in continuity with the esophagus and with the pylorus and duodenum. After the operation, the abdominal wall was closed in layers 33 .
In sham-operated rats, a midline laparotomy was performed, and the stomach was exposed and gently manipulated. The abdominal cavity was kept open for the same amount of time required to perform the other operations. A 1-cm gastrotomy was performed and then closed as in the SG group. postoperative care. At the end of the surgical procedures, all rats received sterile 0.9% NaCl 10 mL i.p. and 10 mL s.c. to maintain hydration during healing. The animals received ketoprofen 5 mg/kg as an analgesic. They were placed on a heated mat until they recovered and then were returned to their home cages. The rats were allowed to drink purified water for 12 h after surgery, and a liquid diet containing 5% glucose and 0.2% KCl was provided for the next 48 h. Thereafter, they received the HFD until 15 weeks after surgery 34 . oral glucose tolerance test. The oral glucose tolerance test (OGTT) was performed at kill. Animals were fasted overnight and then received a 50% D-glucose solution (1 g/kg body weight) by oral gavage. Blood was collected from the tail vein for measurement of glucose and insulin concentrations at 0, 15, 30, 60, 90,120 and 180 min at the end of the study. After centrifugation, plasma was divided into appropriate subsamples and stored at −20 °C until analyses.
Analytical methods. Blood glucose levels were analyzed by glucometer (Accu-Chek, Roche Diagnostics Division, Grenzacherstrasse, Switzerland). Serum insulin was measured by a rat insulin ultrasensitive ELISA (Biovendor GmbH, Kassel, Germany), with a sensitivity of 0.025 ng/ml and an intra-and inter-assay precision of 10%.
Hepatocytes isolation. Hepatocytes were isolated by collagenase perfusion 35 of the liver. Cells were cultured for 24 hours in DMEM with 10% FBS.

Human Studies
Study population. We enrolled 15 obese subjects with NAFLD of both sexes (age 39.7 ± 2.1) before and one year after laparoscopic SG. Eight subjects who underwent laparoscopic elective cholecystectomy, but otherwise in healthy conditions, served as controls. Venous blood samples were collected, after 12 hours fasting, at the time of the patients' enrolment and one year after SG.
Comprehensive history was obtained from all subjects. Physical exam, including height, weight and anthropometric measurements were conducted. Body mass index (BMI) was calculated by dividing the weight (kg) by height (cm). A detailed medication list was obtained. All participants underwent biochemical tests: ALT, AST, fasting glucose, insulin, triglycerides, HDL and LDL-cholesterol.
The study was approved by the Ethical Committee of the University of Rome "La Sapienza" and the patients signed an informed consent prior the study enrolment; in addition, they signed a specific informed consent for the surgical operations and all methods were performed in accordance with the relevant guidelines and regulations Human sleeve gastrectomy. Sleeve Gastrectomy (SG) consists in a vertical resection of the stomach in order to obtain a residual gastric capacity ranged from 60 to 80 ml. The procedures were carried out by laparoscopy using 5 trocars. The resection was performed using linear stapler with sequential cartridges alongside an oro-gastric calibration tube placed against the lesser curvature. Resection was begun 4 to 6 cm proximal to the pylorus and was continued up-ward to the angle of His with complete excision of the fundus, part of the body and antrum.
During the first month after surgery patients had a semiliquid diet with about 850 kcal per day for the first month and thereafter they gradually introduced solid foodstuffs having a free diet. Liver histology. Human liver biopsy material, obtained during the operation, was mounted on archival slides originally prepared from 4% formaldehyde-fixed paraffin-embedded tissue and stained with hematoxylin-eosin to evaluate the percentage of steatosis. Blind analysis of biopsy specimens was performed by a liver pathologist (CG).
Human primary hepatocytes isolation. Hepatocytes from obese, NAFLD subjects and healthy subjects were isolated as described elsewhere 38 . Briefly, tissue was diced and washed in HBSS to remove excess blood. Tissue was transferred to a 50 ml tube containing EGTA buffer (HBSS, 0.5 micromol/L EGTA, 0.5% BSA) and agitated at 100 rpm in a water bath with shaking bed for 10 min at 37 °C. Tissue was then placed in digestion buffer (HBSS, 0.05% collagenase IV, 0.5% fatty acid free BSA, 10 micromol/L CaCl2) and agitated in a water bath with shaking bed for 30 min at 37 °C. Supernatant was collected and filtered through 100 μm cell strainer and the cell suspension centrifuged at 80 g for 5 min at 4 °C and the supernatant discarded. Cells were cultured in DMEM in collagen coated dish at 37 °C and 5%CO 2 .

Figure 8.
Proposed mechanism of NAFLD and insulin resistance reversal after Sleeve Gastrectomy. Excessive caloric intake, associated with NAFLD and obesity, reduces AMPK activation, causing a decrease of Plin2 catabolism mediated by chaperone-mediated autophagy leading to ectopic fat deposition increase and simultaneously to hepatic insulin resistance. After SG, the energy imbalance due to reduced caloric intake causes an increase in AMPK phosphorylation, a crucial step in Plin2-LAMP2A binding, leading to an enhanced Plin2 autophagy that exposes LD triglycerides to intracellular lipases. In addition, AMPK increased phosphorylation improves insulin signaling and promotes glycogen synthesis.