Adipsin is an adipokine that improves β cell function in diabetes

Abstract

A hallmark of type 2 diabetes mellitus (T2DM) is the development of pancreatic β cell failure, which results in insulinopenia and hyperglycemia. We show that the adipokine adipsin has a beneficial role in maintaining β cell function. Animals genetically lacking adipsin have glucose intolerance due to insulinopenia; isolated islets from these mice have reduced glucose-stimulated insulin secretion. Replenishment of adipsin to diabetic mice treated hyperglycemia by boosting insulin secretion. We identify C3a, a peptide generated by adipsin, as a potent insulin secretagogue and show that the C3a receptor is required for these beneficial effects of adipsin. C3a acts on islets by augmenting ATP levels, respiration, and cytosolic free Ca(2+). Finally, we demonstrate that T2DM patients with β cell failure are deficient in adipsin. These findings indicate that the adipsin/C3a pathway connects adipocyte function to β cell physiology, and manipulation of this molecular switch may serve as a therapy in T2DM.

Figure 1. Adipsin Regulates Adipose Tissue Inflammation and Protects Against Diabetes

(A) WT and Adipsin^−/− mice were fed a high fat diet (HFD) for 4 months starting at 1 month of age and visceral adipose tissues were assayed for immune infiltration by qPCR. (B and C) WT and Adipsin^−/− adipose tissues were stained for crown-like structures (CLS). Representative pictures are shown (B) and numbers of CLS (C) were quantitated. Scale bar, 100 µm. (D and E) WT and Adipsin^−/− adipose tissues were stained with toluidine blue for mast cells. Representative pictures are shown (D) and numbers of mast cells (E) were quantitated. Scale bar, 25 µm. (F) Glucose tolerance test (GTT) was performed on WT and Adipsin^−/− mice fed a HFD for 16 weeks with measurement of blood glucose concentrations. (G) Insulin tolerance test (ITT) was performed on WT and Adipsin^−/− mice fed a HFD for 18 weeks with measurement of blood glucose concentrations. For GTT and ITT experiments, n = 12–16 per genotype. *P < 0.05, **P < 0.01. See also Figures S1 and S2.

Figure 2. Adipsin Regulates Insulin Secretion in vivo and in vitro

(A) WT and Adipsin^−/− mice were fed a HFD diet for 4 months starting at 1 month of age and challenged with i.p. glucose injections, and plasma insulin levels were assayed. N = 8–12 mice per genotype. (B and C) Pancreata from WT and Adipsin^−/− mice fed a HFD diet for 4 months were collected and insulin immunohistochemistry staining was performed. Representative pictures are shown (B) and β cell area was quantitated (C). Scale bar, 400 µm. (D) Glucose-stimulated insulin secretion assay on islets from WT and Adipsin^−/− mice fed a HFD diet. *P < 0.05, **P < 0.01. See also Figure S3.

Figure 3. Restoration of Adipsin Improves Insulin Secretion and Glucose Homeostasis

(A) Diabetic db/db mice (3 month old) were treated i.v. with 2 × 10⁹ IFU of control lacZ or adipsin adenovirus vectors and 5 days later serum adipsin was assessed by Western blot. (B and C) Control lacZ- and adipsin-transduced mice were challenged by an i.p. GTT with measurements of blood glucose (B) and plasma insulin (C). N = 6 mice per group. (D) Gluconeogenesis was determined by hepatic Pepck and G6pc gene expression. *P < 0.05, **P< 0.01, ***P< 0.001. See also Figure S4.

Figure 4. C3a Stimulates Pancreatic β Cells to Secrete Insulin

(A and B) Relative expression of Adipsin (A) and C3 (B) in liver, quadriceps muscle (quad), islets, epididymal (epi) and inguinal (ing) fat were quantitated by qPCR. (C to E) Flow cytometry was performed on islet cells with antibodies to receptors C3aR1 (C), C5aR1 (D) and C5L2 (E). (F and G) Islets from WT mice on a chow diet (F) or HFD (G) were subjected to a GSIS assay with recombinant C3a or C5a (100 nM) at the indicated concentrations of glucose. (H and I) db/db mice (3 month old) transduced with lacZ or adipsin adenovirus were treated with vehicle or C3aR antagonist (SB 290157) and subjected to GTT with measurements of blood glucose (H) and plasma insulin (I). Data were pooled from 2 experiments. Statistics for GTT assays between groups: lacZ/vehicle vs. adipsin/vehicle, p = 0.05; adipsin/vehicle vs. adipsin/SB 290157, p < 0.01; lacZ/SB 290157 vs. adipsin/vehicle, p < 0.01. N = 12–15 mice per group. *P < 0.05. See also Figure S5.

Figure 5. C3a Stimulates Cytoplasmic Free Ca²⁺ ([Ca²⁺]_i) and Oxygen Consumption in Islets

(A and B) Islets were treated with C3a (100 nM) in conjunction with 30 mM KCl (A) or 0.25 mM tolbutamide (B) and assayed for insulin secretion. (C–E) [Ca²⁺]_i was measured in islets treated with control or C3a and stimulated with 20 mM glucose, washed with 3 mM glucose, stimulated with KCl and then washed with 3 mM glucose. The peak [Ca²⁺]_i (D) and [Ca²⁺]_i area under the curve (E) are quantified. (F) Intracellular ATP levels were determined after treatment of islets with C3a at 3 or 20 mM glucose. Data were pooled from 3 experiments. (G and H) Oxygen consumption rates (OCR) were measured in islets treated with C3a and the following treatments: 20 mM glucose, oligomycin, rotenone and antimycin A. Statistical analysis was performed on traces prior to oligomycin. Data were pooled from 3 experiments. (H) ATP-coupled respiration of islets is quantified. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 6. Adipsin is Decreased in Patients with Type II Diabetes Mellitus (T2DM) with β Cell Failure

(A to F) T2DM patients were classified according to those on oral metformin therapy (T2DM) or T2DM patients with β cell failure (T2DM-βCF) for those on insulin therapy. (A to C) Visceral and subcutaneous (SubQ) adipose tissue samples from T2DM and T2DM-βCF patients were analyzed for Adipsin (A), Adiponectin (B) and Leptin (C) mRNA levels. (D to F) Circulating adipsin (D), adiponectin (E) and leptin (F) levels were measured from blood samples of T2DM and T2DM-βCF patients. Adiponectin and leptin are plotted according to male and female. *P < 0.05. See also Figure S6.