Peripancreatic adipose tissue protects against high-fat-diet-induced hepatic steatosis and insulin resistance in mice

Background/objectives Visceral adiposity is associated with increased diabetes risk, while expansion of subcutaneous adipose tissue may be protective. However, the visceral compartment contains different fat depots. Peripancreatic adipose tissue (PAT) is an understudied visceral fat depot. Here, we aimed to define PAT functionality in lean and high-fat-diet (HFD)-induced obese mice. Subjects/methods Four adipose tissue depots (inguinal, mesenteric, gonadal, and peripancreatic adipose tissue) from chow- and HFD-fed male mice were compared with respect to adipocyte size (n = 4–5/group), cellular composition (FACS analysis, n = 5–6/group), lipogenesis and lipolysis (n = 3/group), and gene expression (n = 6–10/group). Radioactive tracers were used to compare lipid and glucose metabolism between these four fat depots in vivo (n = 5–11/group). To determine the role of PAT in obesity-associated metabolic disturbances, PAT was surgically removed prior to challenging the mice with HFD. PAT-ectomized mice were compared to sham controls with respect to glucose tolerance, basal and glucose-stimulated insulin levels, hepatic and pancreatic steatosis, and gene expression (n = 8–10/group). Results We found that PAT is a tiny fat depot (~0.2% of the total fat mass) containing relatively small adipocytes and many “non-adipocytes” such as leukocytes and fibroblasts. PAT was distinguished from the other fat depots by increased glucose uptake and increased fatty acid oxidation in both lean and obese mice. Moreover, PAT was the only fat depot where the tissue weight correlated positively with liver weight in obese mice (R = 0.65; p = 0.009). Surgical removal of PAT followed by 16-week HFD feeding was associated with aggravated hepatic steatosis (p = 0.008) and higher basal (p < 0.05) and glucose-stimulated insulin levels (p < 0.01). PAT removal also led to enlarged pancreatic islets and increased pancreatic expression of markers of glucose-stimulated insulin secretion and islet development (p < 0.05). Conclusions PAT is a small metabolically highly active fat depot that plays a previously unrecognized role in the pathogenesis of hepatic steatosis and insulin resistance in advanced obesity.


List of contents
Supplementary method descriptions Figure S1. Additional PAT and PAT adipocyte data.    Table S1. Adipose depot weight in the HFD time course study.  capacity RNA-to-DNA kit (Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions. Real-time PCR products were detected using SYBR Green (Life Technologies) and quantified using the relative Ct-method. Suitable reference genes (Actb and/or Tbp) were identified by the NormFinder algorithm using GenEx (MultiD) as described (27). Primer sequences are available in Supplementary Table   3.
Cells were washed, resuspended in DMEM 10% FBS and counted on automatic cell counter (Countess 2, Thermo Fisher, Waltham, MA, USA) after which they were stained for flow cytometry analysis. PAT samples were pooled from 20 mice to a total n=5 due to their small size.

Perilipin-1 staining of paraffin-embedded sections
Paraffin-embedded sections were rehydrated and antigen retrieval was performed by heating samples for 20 min with 10 mM citric acid 0.05% Tween (pH=6). Non-specific staining was blocked with 3% donkey serum in PBS for 30 min on RT. Sections were

Oral glucose tolerance test (OGTT), glucose, insulin and glucagon measurements
OGTT was performed in PAT-ectomized and control mice as indicated in

Adipocyte, hepatic lipid droplet, pancreatic islet and intrapancreatic adipocyte size measurements
GWAT, IWAT, MWAT, PAT, liver and pancreas samples were fixed for 72h in 4% phosphate-buffered formalin solution (VWR Chemicals, Stockholm, Sweden), stored in 50% ethanol and embedded in paraffin. Paraffin sections (four 7µm sections/mouse and 4-5 mice/group for adipose tissues and liver & two 7µm sections/mouse, 17-38 islets/section, and 3-4 mice/groups for pancreas) were stained with hematoxylin&eosin (H&E) solution. Five 20x images per animal were obtained using a light microscope (Olympus BX60 & PlanApo, 20x/0.7, Olympus, Tokyo, Japan). Adipocyte area, hepatic lipid droplet area and number, islet area and number, and intrapancreatic adipocyte area and number were measured using ImageJ v1.47 software (National Institutes of Health, Bethesda, MD, USA) as previously described (25). Average adipocyte and islet size were presented as µm 2 , and liver lipid droplet size distribution was presented as % of total. The investigator was blinded to the group allocation.
Samples were slowly heated to 85°C for 2-5min and then cooled down to room temperature. They were centrifuged at maximum speed to remove insoluble material before proceeding with the triglyceride content assay following instructions of manufacturer (Randox Laboratories, Dublin, United Kingdom).

Islet and PAT co-culture
Pancreatic islets (12 per well) were put on top of the 24-mm Transwell membrane inserts (Costar, Washington DC, USA) and co-cultured with finely minced (~1mm 3 pieces) PAT explants (~10 mg/sample, bottom well) in DMEM at 11 mM glucose overnight at 37°C, 5% CO2. After incubation, medium was collected for insulin measurements and pancreatic islets were collected for glucose-stimulated insulin secretion assay.

Glucose-stimulated insulin secretion (GSIS)
Pancreatic islets were washed in Krebs buffer 0.1% BSA with no glucose, before placing them in individual tubes (N=12 islets/tube). In order to measure GSIS, islets were incubated with Krebs buffer 2.8 mM or 16.8 mM glucose in an orbital-shaker incubator at 80-100 rpm during 1h at 37°C. The supernatant was collected and stored at -80°C and islets were homogenized in HCl and frozen pending later measurements of insulin.