Mitochondrial pyruvate carrier inhibitors improve metabolic parameters in diet-induced obese mice

Abstract

The mitochondrial pyruvate carrier (MPC) is an inner mitochondrial membrane complex that plays a critical role in intermediary metabolism. Inhibition of the MPC, especially in liver, may have efficacy for treating type 2 diabetes mellitus. Herein, we examined the antidiabetic effects of zaprinast and 7ACC2, small molecules which have been reported to act as MPC inhibitors. Both compounds activated a bioluminescence resonance energy transfer-based MPC reporter assay (reporter sensitive to pyruvate) and potently inhibited pyruvate-mediated respiration in isolated mitochondria. Furthermore, zaprinast and 7ACC2 acutely improved glucose tolerance in diet-induced obese mice in vivo. Although some findings were suggestive of improved insulin sensitivity, hyperinsulinemic-euglycemic clamp studies did not detect enhanced insulin action in response to 7ACC2 treatment. Rather, our data suggest acute glucose-lowering effects of MPC inhibition may be due to suppressed hepatic gluconeogenesis. Finally, we used reporter sensitive to pyruvate to screen a chemical library of drugs and identified 35 potentially novel MPC modulators. Using available evidence, we generated a pharmacophore model to prioritize which hits to pursue. Our analysis revealed carsalam and six quinolone antibiotics, as well as 7ACC1, share a common pharmacophore with 7ACC2. We validated that these compounds are novel inhibitors of the MPC and suppress hepatocyte glucose production and demonstrated that one quinolone (nalidixic acid) improved glucose tolerance in obese mice. In conclusion, these data demonstrate the feasibility of therapeutic targeting of the MPC for treating diabetes and provide scaffolds that can be used to develop potent and novel classes of MPC inhibitors.

Keywords: diabetes; gluconeogenesis; metabolic disease; mitochondrial metabolism; pyruvate.

Figure 1

Zaprinast and 7ACC2 are MPC inhibitors that do not activate PPARγ.A, the schematic depicts the BRET-based MPC biosensor (RESPYR) system with MPC1–Venus and MPC2-RLuc fusion proteins in the absence (left) or presence (right) of MPC inhibitors. Created with BioRender.com. B, dose–response effects of zaprinast or 7ACC2 in a RESPYR assay. Values are presented as mean ± standard error of the mean. n = 5 per group. C, pyruvate-stimulated mitochondrial respiration with increasing doses of zaprinast or 7ACC2. D, the effects of zaprinast on mitochondrial pyruvate metabolism require MPC. The effects of vehicle or zaprinast on pyruvate-stimulated respiration by cardiac mitochondria from WT or cardiac-specific MPC2-knockout mice are shown. E, zaprinast and 7ACC2 do not activate PPARγ. The effects of zaprinast, 7ACC2, rosiglitazone, and pioglitazone on the activity of a Gal4-PPARγ–driven luciferase reporter are shown. ∗p < 0.01 versus WT vehicle and MPC2−/−, ∗∗p < 0.01 versus all other groups. BRET, bioluminescence resonance energy transfer; MPC, mitochondrial pyruvate carrier; MPP, mitochondrial pyruvate carrier; OCR, oxygen consumption rate; PPARγ, peroxisome proliferator–activated receptor γ; RESPYR, reporter sensitive to pyruvate.

Figure 2

Zaprinast and 7ACC2 improve glucose tolerance in DIO mice.A and B, WT and LS-Mpc2−/− mice were fed a high-fat (HF) diet for 12 weeks and then administered a single dose of zaprinast (A) or 7ACC2 (B) or vehicle control. Glucose tolerance was then assessed 16 h later after an overnight fast. ∗p < 0.05 compared to WT vehicle. ∗∗p < 0.05 compared to WT vehicle and all zaprinast-treated mice. C, insulin tolerance test in WT mice fed low-fat (LF) or HF diet after 3 days treatment with vehicle, zaprinast, or 7ACC2 treatment. Values are presented as mean ± standard error of the mean. n = 7 to 10 per group. D, plasma insulin concentrations in WT mice fed LF or HF diet after 3 days treatment with vehicle, zaprinast, or 7ACC2 treatment. Values are presented as mean ± standard error of the mean. n = 8 to 18 per group. ∗p < 0.05 compared to LF vehicle. ∗∗p < 0.05 compared to LF and DIO vehicle. E, WT mice were fed a HF diet or LF control diet and then received 3 days of zaprinast or 7ACC2 treatment. An insulin bolus was injected 10 min prior to sacrifice. Liver insulin signaling was assessed using S473 phosphorylated–specific AKT antibodies by Western blot. The ratio of pAKT-s473/total AKT was quantified using densitometric analysis of band intensity, and the values are presented as mean ± standard error of the mean below the Western blot images. DIO, diet-induced obese.

Figure 3

7ACC2 does not improve insulin sensitivity in DIO mice. C57BL/6J mice were fed a high-fat diet for 12 weeks and then treated with vehicle or 7ACC2 for 2 days before undergoing a hyperinsulinemic–euglycemic clamp. Graphs display: A, fasting arterial glucose; B, fasting or clamped insulin concentrations; C, glucose infusion rate; D, glucose flux (Rd); E, endogenous rate of appearance (Ra); F, Ra versus fasting and clamped insulin concentrations; and G, uptake of ¹⁴C 2-deoxyglucose into various insulin target tissues. Values are presented as mean ± standard error of the mean. n = (7) per group. ∗p < 0.05 compared to WT vehicle, ∗∗p < 0.01 versus fasting values in same treatment group. DIO, diet-induced obese.

Figure 4

7ACC2 and zaprinast suppress hepatocyte glucose production.A, pyruvate-stimulated hepatocyte glucose production assay in hepatocytes from littermate WT and LS-Mpc2−/− mice after treatment with vehicle, UK-5099, 7ACC2, or zaprinast. n = (3) per group. ∗p < 0.05 compared to pyruvate plus vehicle. B, ¹³C- and ¹²C-pyruvate was administered to WT hepatocytes treated with vehicle, UK-5099, 7ACC2, or zaprinast. Created with BioRender.com. C, total enrichment of ¹³C into the indicated metabolites measured by mass spectrometry after 3 h. Data presented as mean +standard error of the mean. n = (3) per group of a representative experiment (of two). ∗p ≤ 0.05 versus vehicle-treated hepatocytes. D and E, C57BL6/J mice were administered a single dose of vehicle, zaprinast, or 7ACC2, and a lactate/pyruvate tolerance test was conducted beginning 30 min later. Blood glucose (D) or lactate (E) concentrations are shown. ∗p < 0.05 compared to zaprinast and 7ACC2 concentrations at the same time point. MPC, mitochondrial pyruvate carrier.

Figure 5

A high-throughput screen identifies novel modulators of the MPC.A, the experimental workflow of the high-throughput screen of the Pharmakon 1600 library using the RESPYR system is shown. Created with BioRender.com. B, the first technical replicate is graphed on the y-axis, and the second replicate is graphed on the x-axis. Average UK-5099 reads for the screen are indicated with the orange open circle. Positive hits are indicated with blue dots. Compounds that altered the signal ratio in cells expressing only MPC2-RLuc8 in the absence of the acceptor, MPC1–Venus, are indicated in gray and were excluded from further analysis. C and D, the chemical structures of known MPC inhibitors and compounds in the Pharmakon 1600 library with similar chemical structures (C) or 7ACC1 (D) are shown. MPC, mitochondrial pyruvate carrier; RESPYR, reporter sensitive to pyruvate.

Figure 6

The novel MPC inhibitors nalidixic acid, 7ACC1, and carsalam inhibit pyruvate-stimulated mitochondrial respiration, attenuate hepatocyte glucose production, and improve glucose tolerance in DIO mice.A, pyruvate-stimulated mitochondrial respiration with 10 μM of the indicated compounds are shown. Respiration values were normalized to vehicle control treatment. B, mitochondrial respiration dose–response curves of 7ACC2, 7ACC1, nalidixic acid, and carsalam. C, the effects of 7ACC2, 7ACC1, nalidixic acid, and carsalam on pyruvate-stimulated hepatocyte glucose production are shown. D, C57BL/6J mice were fed a high-fat diet for 14 weeks and then administered a daily dose of nalidixic acid or vehicle control for 3 days prior to glucose tolerance testing. Area under the curve was calculated and is displayed inset in the bar graph. ∗p < 0.05 compared to vehicle. DIO, diet-induced obese; OCR, oxygen consumption rate.