GPR142 Controls Tryptophan-Induced Insulin and Incretin Hormone Secretion to Improve Glucose Metabolism

Abstract

GPR142, a putative amino acid receptor, is expressed in pancreatic islets and the gastrointestinal tract, but the ligand affinity and physiological role of this receptor remain obscure. In this study, we show that in addition to L-Tryptophan, GPR142 signaling is also activated by L-Phenylalanine but not by other naturally occurring amino acids. Furthermore, we show that Tryptophan and a synthetic GPR142 agonist increase insulin and incretin hormones and improve glucose disposal in mice in a GPR142-dependent manner. In contrast, Phenylalanine improves in vivo glucose disposal independently of GPR142. Noteworthy, refeeding-induced elevations in insulin and glucose-dependent insulinotropic polypeptide are blunted in Gpr142 null mice. In conclusion, these findings demonstrate GPR142 is a Tryptophan receptor critically required for insulin and incretin hormone regulation and suggest GPR142 agonists may be effective therapies that leverage amino acid sensing pathways for the treatment of type 2 diabetes.

Fig 1. Selective activation of GPR142 signaling by L-Tryptophan and L-Phenylalanine and GPR142 mRNA expression in mouse and human tissues.

(A) IP-1 levels in HEK293 cells expressing human GPR142 treated with amino acids and metabolites at varying concentrations. (B) IP-1 levels in HEK293-hGPR142 cells and untransfected control HEK293 cells treated with L-Tryptophan or L-Phenylalanine. Data are mean ± SD of 2 replicate wells. Data representative of 3 independent experiments are shown. (C-E) Mouse Gpr142 mRNA expression determined by quantitative RT-PCR in (C) tissues of normal C57BL/6 male mice, (D) pancreatic islets isolated from male Gpr142 KO mice and WT littermate controls, and (E) pancreatic islets isolated from male 8-week-old db/db mice and age-matched C57 control mice. (F, G) Human GPR142 mRNA expression determined by quantitative RT-PCR in (F) a panel of human tissues and (G) pancreatic islets isolated from healthy non-diabetic or type 2 diabetic donors. Data are normalized against Rplp0 mRNA in each tissue sample, then normalized to the mean value of islet expression (C, F) or control group (D, E, G) set as 1, and expressed as fold expression. N = 2–4 (C-D), N = 5 per group (E), N = 1 pooled biological samples from ≥ 3 donors per tissue type (F), or N = 12–15 donors per group (G). Error bars represent SEM. ND: not detectable. **: p<0.01 control vs. db/db.

Fig 2. Effects of amino acids on insulin secretion in isolated pancreatic islets.

(A-C) Islets isolated from Gpr142KO mice and WT littermate controls were incubated in the presence of L-Tryptophan (A), L-Phenylalanine (B), Exendin-4, or L-Arginine (C) for 1 hour, and insulin concentrations in the culture media were measured and expressed as fold change compared to control treatment (11.1 mM glucose). For each experiment, islets were isolated from 8–10 mice, pooled, then plated for treatment. Data are mean ± SEM. N = 5 replicates per treatment group. *,**: p<0.05, 0.01 KO vs. WT; ###: p<0.001 WT islets treatment group vs. 11.1mM glucose control group; ^^,^^^: KO islets p<0.01, 0.001 treatment group vs. 11.1mM glucose control group. (D) Islets isolated from a non-diabetic human donor were incubated at 11.1mM glucose in the presence of varying concentrations of L-Tryptophan, and insulin concentrations in the culture media were measured. Data are mean ± SEM. N = 5 replicates per group. *,****: p<0.05, 0.0001 vs. control (no L-Trp).

Fig 3. Glucose tolerance tests after L-Tryptophan and L-Phenylalanine oral dosing in Gpr142KO mice and WT littermate controls.

(A-D) Vehicle (15% HP-β-CD) or L-Trp (500 mg/kg), or (E-F) vehicle or L-Phe (500 mg/kg) was dosed orally at 20 ml/kg to 5hr daytime fasted mice 15 minutes prior to the glucose challenge. Glucose (2 g/kg) was injected i.p. (A-B, E-F) or given by oral gavage (C-D) at indicated time points. Tail blood glucose was monitored for 120 minutes after glucose injection. Animals used were male, 5–8 months of age, and maintained on standard chow diet. Data are mean ± SEM. N = 7 per group. *,**,***: p<0.05, 0.01, 0.001 treatment group vs. vehicle.

Fig 4. In vivo effects of L-Tryptophan and L-Phenylalanine on plasma hormones in Gpr142KO mice and WT littermate controls.

Vehicle, L-Trp (500 mg/kg), or L-Phe (500 mg/kg) was dosed orally to overnight fasted male mice and cardiac blood was collected 10 minutes after dosing. Animals were 6–9 months of age, and maintained on standard chow diet. Plasma levels of GIP (A), total GLP-1 (B), and insulin (C) were measured. Data are mean ± SEM. N = 6–7 per group. *,***,****: p<0.05, 0.001, 0.0001 between indicated groups; NS: not significant.

Fig 5. Fasted and refed levels of plasma hormones and blood glucose in Gpr142KO mice and WT littermate controls.

Plasma levels of insulin (A), GIP (B), total GLP-1 (C), and tail blood glucose (D) were measured after overnight fasting (T = 0), 30-minute refeeding, or 90-minute refeeding. 7-month-old male mice maintained on standard chow diet were used for the study. Data are mean ± SEM. N = 8 per genotype per time point. *,***: p<0.05, 0.001 KO vs. WT.

Fig 6. Effects of compound A on insulin secretion, plasma hormones and glycemia during oral glucose tolerance test.

(A-B) Pancreatic islets isolated from normal C57 mice (A) or a non-diabetic human donor (B) were incubated in the presence of 11 mM glucose and varying concentrations of compound A for 1 hour. Insulin concentrations in the culture media were measured. Data representative of 2–3 independent experiments are shown. Data are mean ± SEM. N = 5 replicates per group. *,**,***: p<0.05, 0.01, 0.001 vs. control. (C-D) Vehicle (1% w/v HEC, 0.25% v/v Tween80, 0.05% v/v Antifoam in DI water) or compound A (30 mg/kg) was dosed orally to overnight fasted normal male C57 mice. 30 minutes later, glucose (2 g/kg) was orally dosed, and cardiac blood was collected at T = 0 (no glucose dosing), 3, or 15min after glucose challenge. Plasma levels of GIP (C) and total GLP-1 (D) were measured. (E-F) Vehicle or compound A (10 mg/kg) was dosed orally to 5hr daytime fasted diet-induced obese male C57 mice. 30 minutes later, glucose (2 g/kg) was orally dosed. Blood glucose was monitored for the next 120 minutes (E) and plasma insulin at 15 minutes after glucose challenge was measured (F). Data are mean ± SEM. N = 6 per group. *,**,***: p<0.05, 0.01, 0.001 compound A vs. vehicle.