A beta cell ATGL-lipolysis/adipose tissue axis controls energy homeostasis and body weight via insulin secretion in mice

Aims/hypothesis To directly assess the role of beta cell lipolysis in insulin secretion and whole-body energy homeostasis, inducible beta cell-specific adipose triglyceride lipase (ATGL)-deficient (B-Atgl-KO) mice were studied under normal diet (ND) and high-fat diet (HFD) conditions. Methods Atgl flox/flox mice were cross-bred with Mip-Cre-ERT mice to generate Mip-Cre-ERT/+;Atgl flox/flox mice. At 8 weeks of age, these mice were injected with tamoxifen to induce deletion of beta cell-specific Atgl (also known as Pnpla2), and the mice were fed an ND or HFD. Results ND-fed male B-Atgl-KO mice showed decreased insulinaemia and glucose-induced insulin secretion (GSIS) in vivo. Changes in GSIS correlated with the islet content of long-chain saturated monoacylglycerol (MAG) species that have been proposed to be metabolic coupling factors for insulin secretion. Exogenous MAGs restored GSIS in B-Atgl-KO islets. B-Atgl-KO male mice fed an HFD showed reduced insulinaemia, glycaemia in the fasted and fed states and after glucose challenge, as well as enhanced insulin sensitivity. Moreover, decreased insulinaemia in B-Atgl-KO mice was associated with increased energy expenditure, and lipid metabolism in brown (BAT) and white (WAT) adipose tissues, leading to reduced fat mass and body weight. Conclusions/interpretation ATGL in beta cells regulates insulin secretion via the production of signalling MAGs. Decreased insulinaemia due to lowered GSIS protects B-Atgl-KO mice from diet-induced obesity, improves insulin sensitivity, increases lipid mobilisation from WAT and causes BAT activation. The results support the concept that fuel excess can drive obesity and diabetes via hyperinsulinaemia, and that an islet beta cell ATGL-lipolysis/adipose tissue axis controls energy homeostasis and body weight via insulin secretion. Electronic supplementary material The online version of this article (doi:10.1007/s00125-016-4105-2) contains peer-reviewed but unedited supplementary material, which is available to authorised users.

Proteins were extracted as described previously [4]. For adipose tissue samples, homogenates were centrifuged at 10, 000 g for 10 min at 4°C and infranatant was collected to quantify proteins. Antibodies and dilutions are listed below.

Insulin secretion ex-vivo
Islets were distributed in 12-well plates (10 islets/well) in RPMI medium containing 3 mmol/l glucose for 2 h and preincubated for 45 min at 37°C in KRB medium with 10 mmol/l Hepes (KRBH) containing 0.5% defatted-BSA and 3 mmol/l glucose. Islets were then incubated for 20 min at 3 or 16 mmol/l glucose and at 3 mmol/l glucose plus 35 mmol/l KCl. For experiments on isolated islets from 23-week-old HFD-fed mice, islets were incubated for 1 h at 3 or 16 mmol/l glucose, in the presence or absence of palmitate/oleate (0.15 mmol/l each). For rescue experiment with 1-PG, islets were incubated in the presence or absence of 100 µmol/l 1-PG during the pre-incubation at 3 mmol/l glucose and the incubation at 3 or 16 mmol/l glucose. Stock solution of 1-PG was prepared in DMSO. Each condition was run in 3 replicates. Insulin release was normalized for the total islet insulin content. 3

Islet metabolism
Lipolysis was measured on 100 islets isolated from 10-week-old male mice and incubated for 1h in KRBH 0.5% defatted-BSA at 3 or 16 mmol/l glucose. At the end of the incubation, media were kept to measure glycerol [5] and NEFA [6] release.
Glucose oxidation and utilization were assessed as described previously on islets isolated from 10-week-old male mice [4].

Intracellular Ca 2+
After isolation, islets were dispersed into single cells by trypsin digestion [8] and were seeded at a density of 60,000 cells per well in 96-well black plates with clear bottom coated with extracellular matrix derived from 804G cells (804G-ECM; gift from Philippe Halban, Geneva, Switzerland) [9]. After 3h incubation in RPMI medium containing 6 mmol/l glucose and 10% FBS to allow cell attachment, cells were loaded with Fura-2 AM (6 µmol/l) dissolved in pluronic F-127 for 75 min and calcium was measured as previously described [10]. Each condition was run in 5 replicates. Cytosolic calcium was calculated according to Grynkiewicz et al. [11].  Wako diagnostic NEFA-HR, respectively).
All the variables were determined in 10-week-old male mice (2 weeks after tamoxifen treatment) fed a ND. Triacylglycerol (TG), non-esterified fatty acids (NEFA), and glycerol were measured in plasma from anesthetized overnight fasted and fed male mice. *p<0.05 vs fl/fl (one-way ANOVA and Bonferroni post hoc test).