Adipocyte PHLPP2 inhibition prevents obesity-induced fatty liver

Abstract

Increased adiposity confers risk for systemic insulin resistance and type 2 diabetes (T2D), but mechanisms underlying this pathogenic inter-organ crosstalk are incompletely understood. We find PHLPP2 (PH domain and leucine rich repeat protein phosphatase 2), recently identified as the Akt Ser473 phosphatase, to be increased in adipocytes from obese mice. To identify the functional consequence of increased adipocyte PHLPP2 in obese mice, we generated adipocyte-specific PHLPP2 knockout (A-PHLPP2) mice. A-PHLPP2 mice show normal adiposity and glucose metabolism when fed a normal chow diet, but reduced adiposity and improved whole-body glucose tolerance as compared to Cre- controls with high-fat diet (HFD) feeding. Notably, HFD-fed A-PHLPP2 mice show increased HSL phosphorylation, leading to increased lipolysis in vitro and in vivo. Mobilized adipocyte fatty acids are oxidized, leading to increased peroxisome proliferator-activated receptor alpha (PPARα)-dependent adiponectin secretion, which in turn increases hepatic fatty acid oxidation to ameliorate obesity-induced fatty liver. Consistently, adipose PHLPP2 expression is negatively correlated with serum adiponectin levels in obese humans. Overall, these data implicate an adipocyte PHLPP2-HSL-PPARα signaling axis to regulate systemic glucose and lipid homeostasis, and suggest that excess adipocyte PHLPP2 explains decreased adiponectin secretion and downstream metabolic consequence in obesity.

Fig. 1. Blocking adipocyte PHLPP2 attenuates HFD-induced adipocyte hypertrophy.

a PHLPP1 and 2 protein levels in eWAT and iWAT from normal chow diet (NCD) or high-fat diet (HFD)-fed C57BL/6 J, and lean or db/db mice, with quantification of PHLPP2 levels (right, n = 4–6 independent mice per group). b Gene expression in adipocytes and SVF isolated from eWAT from HFD-fed Cre- and A-PHLPP2 mice sacrificed after an 18 h fast (n = 3 independent mice per group). c Body weight curves, d body composition, and e tissue weights in NCD-fed or HFD-fed Cre-PHLPP2 or A-PHLPP2 mice (n = 6–7 independent mice per group). f eWAT and iWAT triglyceride (TG) content (n = 6 independent mice per group), and g H&E staining (left) and adipocyte cell size-frequency distribution (right) (n = 3 independent mice per group) in HFD-fed Cre- and A-PHLPP2 mice. Scale bars, 100 μm. All data are shown as the means ± SEM. Statistical significance was determined by unpaired two-tailed Student’s t-test.

Fig. 2. Adipocyte-specific PHLPP2 ablation improves glucose homeostasis.

a Glucose tolerance test (GTT) and (b) insulin tolerance test (ITT) with increasing lengths of NCD or HFD feeding in A-PHLPP2 and control mice (n = 6–7 independent mice per group) after tamoxifen injection. c Plasma insulin levels in NCD-fed or HFD-fed A-PHLPP2 and control mice after a 5 h fast (n = 6–7 independent mice per group. All data are shown as the means ± SEM. Statistical significance was determined by unpaired two-tailed Student’s t-test.

Fig. 3. Adipocyte PHLPP2 regulates energy expenditure and fat lipolysis.

a Body weight curves in HFD-fed Cre- and A-PHLPP2 mice after initiation of pair-feeding (n = 7–8 independent mice per group). b, c VO₂ consumed and VCO₂ generated, as measured by indirect calorimetry in HFD-fed Cre- and A-PHLPP2 mice in response to the β3-adrenergic agonist CL316,243 (0.5 mg kg⁻¹) (n = 6–7 independent mice per group). d Serum glycerol and non-esterified fatty acid (NEFA) levels in mice following injection of CL316,243 at 0.5 mg kg⁻¹ (n = 6 independent mice per group). e Ex vivo lipolysis assay in isolated eWAT from Cre-PHLPP2 and A-PHLPP2 mice following treatment with 1 μM CL316,243 (n = 3 independent mice per group). f In vitro lipolysis in differentiated control or PHLPP2-repressed (using two different sgRNA sequences) 3T3-L1 adipocytes, with or without 10 μM forskolin (n = 3 independent experiments). g In vitro lipolysis in differentiated 3T3-L1 adipocytes expressing either empty vector (EV) or PHLPP2 with or without 1 μM CL316,243. (n = 3 independent experiments. All data are shown as the means ± SEM. Statistical significance was determined by unpaired two-tailed Student’s t-test.

Fig. 4. Adipocyte PHLPP2 regulates HSL phosphorylation and localization.

a Western blots in lysates from eWAT (top) or iWAT (bottom) from Cre- or A-PHLPP2 mice fed on NCD or HFD, with quantification of various HSL phosphorylation sites (right) (n = 3 independent mice per group). b, c Western blots in lysates from differentiated PHLPP2-repressed 3T3-L1 adipocytes (b), or 3T3-L1 adipocytes expressing HA-tagged PHLPP2 (c) with quantification of p-HSL levels (right), with or without CL316,243 (n = 2 independent experiments). d Coomassie brilliant blue staining of purified human HSL (hHSL) or PHLPP2 with or without 10 μM forskolin (FSK) (top), for phosphatase assay of HSL Ser563 and Ser660 with purified PHLPP2 (bottom) (n = 2 independent experiments). e Immunofluorescence of HSL (red), Bodipy (green), and DAPI (blue) in 3T3-L1 adipocytes expressing control or HA-tagged PHLPP2 with or without CL316,243 (top). Quantification of HSL localization on lipid droplets was performed by directly counting Bodipy-positive cells from images of randomly chosen fields (bottom) (n = 3 independent samples). Scale bars, 10 μm. f, g In vitro lipolysis in differentiated control or PHLPP2-repressed adipocytes with or without CL316,243 and/or CAY10499 (f) or HSL deletion (g) (n = 3 independent mice per group). All data are shown as the means ± SEM. Statistical significance was determined by unpaired two-tailed Student’s t-test for a, e, and two-way ANOVA with Tukey’s multiple comparison test for f, g. n.s. not significant (P > 0.05).

Fig. 5. Loss of adipocyte PHLPP2 induces fatty acid oxidative machinery.

a Gene expression in eWAT (left) and iWAT (right) of HFD-fed Cre- and A-PHLPP2 mice sacrificed after an 18 h fast (n = 6–7 independent mice per group). b–d Gene expression in differentiated 3T3-L1 adipocytes with PHLPP2 overexpression (n = 4 independent mice per group). b or PHLPP2 sg with or without CL316,243 and/or CAY10499 (c) or HSL knockdown (d) (n = 3 independent mice per group). e, f Relative molar ratio of NEFA to glycerol release from eWAT or iWAT explants from HFD-fed Cre-PHLPP2 and A-PHLPP2 mice treated with 1 μM CL316,243 for 4 h (n = 3 independent mice per group) (e) or differentiated control or PHLPP2-repressed 3T3-L1 adipocytes with or without CL316,243 (n = 4 per group) (n = 3 independent mice per group) (f). g FAO of ¹⁴C-oleate to ¹⁴CO₂ in eWAT from Cre- and A-PHLPP2 mice (n = 5–6 independent mice per group). All data are shown as the means ± SEM. Statistical significance was determined by unpaired two-tailed Student’s t-test for a, c–e, g, and two-way ANOVA with Tukey’s multiple comparison test for b, f.

Fig. 6. Blocking adipocyte PHLPP2 protects from diet-induced fatty liver.

a–c H&E and oil-red O (ORO) staining (Scale bars, 100 μm) (a), TG (b), and gene expression (c) from livers of HFD-fed Cre- and A-PHLPP2 mice (n = 6–7 independent mice per group). d, e Adipoq and aP2 gene expression in eWAT (left) or iWAT (right) (d), and representative serum adiponectin (e) from HFD-fed Cre- and A-PHLPP2 mice (n = 6–7 independent mice per group). f Western blots in control and PHLPP2-repressed 3T3-L1 adipocytes. g Western blots of conditioned medium from control and PHLPP2-overexpressing 3T3-L1 adipocytes. h Hepatocyte gene expression in response to saline, or serum from Cre- or A-PHLPP2 mice (n = 3 independent samples per group). i Western blots in primary hepatocytes treated with serum from Cre- or A-PHLPP2 mice (n = 3 independent samples). j Hepatocyte gene expression in response to saline, or serum from Cre- or A-PHLPP2 mice with or without anti-Adiponectin neutralizing antibody (n = 4 independent samples per group). k Plasma adiponectin is inversely correlated to adipose PHLPP2:β-actin ratio, as analyzed by linear regression, with the r² and P values indicated (n = 36 per independent samples). l Model representing the effect of adipocyte PHLPP2 on lipolysis to regulate whole-body glucose homeostasis and hepatic lipid accumulation. All data are shown as the means ± SEM. Statistical significance was determined by unpaired two-tailed Student’s t-test.