Hyperglycemia-induced Sirt3 downregulation increases microglial aerobic glycolysis and inflammation in diabetic neuropathic pain pathogenesis

Abstract

Background: Hyperglycemia-induced neuroinflammation significantly contributes to diabetic neuropathic pain (DNP), but the underlying mechanisms remain unclear.

Objective: To investigate the role of Sirt3, a mitochondrial deacetylase, in hyperglycemia-induced neuroinflammation and DNP and to explore potential therapeutic interventions.

Method and results: Here, we found that Sirt3 was downregulated in spinal dorsal horn (SDH) of diabetic mice by RNA-sequencing, which was further confirmed at the mRNA and protein level. Sirt3 deficiency exacerbated hyperglycemia-induced neuroinflammation and DNP by enhancing microglial aerobic glycolysis in vivo and in vitro. Overexpression of Sirt3 in microglia alleviated inflammation by reducing aerobic glycolysis. Mechanistically, high-glucose stimulation activated Akt, which phosphorylates and inactivates FoxO1. The inactivation of FoxO1 diminished the transcription of Sirt3. Besides that, we also found that hyperglycemia induced Sirt3 degradation via the mitophagy-lysosomal pathway. Blocking Akt activation by GSK69093 or metformin rescued the degradation of Sirt3 protein and transcription inhibition of Sirt3 mRNA, which substantially diminished hyperglycemia-induced inflammation. Metformin in vivo treatment alleviated neuroinflammation and diabetic neuropathic pain by rescuing hyperglycemia-induced Sirt3 downregulation.

Conclusion: Hyperglycemia induces metabolic reprogramming and inflammatory activation in microglia through the regulation of Sirt3 transcription and degradation. This novel mechanism identifies Sirt3 as a potential drug target for treating DNP.

Keywords: Akt/FoxO1; Sirt3; diabetic neuropathic pain; glycolysis; metformin; neuroinflammation.

FIGURE 1

Identification of the role of Sirt3 in DNP Pathogenesis. (A) Volcano plot shows DEGs between SDH of diabetic mice and control; n = 3. (B) Chord plot shows DEGs related to Glycolysis/Gluconeogenesis. (C) GO terms and KEGG enrichment of the DEGs. (D–F) the transcription (D) and protein production (E, F) of Sirt3 in spinal dorsal horn after STZ injection (n = 4, **p < 0.01). (G, H) Paw withdrawal percentage in response to von Frey filament (0.4 g, G; 0.07 g, H) in Sirt3^−/− group and Sirt3^+/+ group after STZ infection; n = 10, *p < 0.05 compared Sirt3^−/− DNP group versus Sirt3^+/+ DNP group. (I) Paw withdrawal latency in response to heat stimulus in Sirt3^−/− group and Sirt3^+/+ group after STZ infection; n = 10, *p < 0.05 compared Sirt3^−/− DNP group versus Sirt3^+/+ DNP group. (J) Blood glucose in Sirt3^−/− group and Sirt3^+/+ group after STZ infection; n = 10, ns p > 0.05 compared Sirt3^−/− DNP group versus Sirt3^+/+ DNP group.

FIGURE 2

Sirt3 deficiency exacerbated neuroinflammation in the SDH in vivo. (A) Co‐expression of Sirt3 and IBA‐1, GFAP, and NeuN in SDH after STZ infection. (B, C) Protein level of IBA‐1 (B) and GFAP (C) in spinal dorsal horn (SDH) of Sirt3 deficient DNP mice. (D) Immunofluorescence shows the upregulation of IBA‐1 in SDH of Sirt3 deficient DNP mice. (E) Transcriptions of IL‐1β, TNF‐α and IL‐6 in SDH of Sirt3 deficient DNP mice (n = 4, *p < 0.05, **p < 0.01). (F, G) Activation of MAPK (F) and NF‐κb (G) pathway in SDH of Sirt3 deficient DNP mice. (H) Transmission electron microscopy shows myelin sheath (MS) and mitochondria (Mi) in the SDH of Sirt3 deficient DNP mice.

FIGURE 3

Sirt3 deficiency enhanced glycolysis in the SDH in vivo. (A) Protein expression levels of key enzymes (HK2, PKM2, LDHA) in glycolysis in SDH of Sirt3 deficient DNP mice. (B, C) glucose metabolism intermediates were changed in the SDH. Lactate levels (B) and pyruvate levels (C) increased (n = 4, **p < 0.01). (D, E) Paw withdrawal percentage in response to von Frey filament (0.4 g, D; 0.07 g, E) was measured in DNP mice treated intraperitoneally with 2‐DG; n = 10 *p < 0.05 compared 2‐DG DNP group versus vehicle DNP group. (F) Paw withdrawal latency in response to heat stimulus was measured in DNP mice treated intraperitoneally with 2‐DG; n = 10 *p < 0.05 compared 2‐DG DNP group versus vehicle DNP group. (G) Transmission electron microscopy shows myelin sheath (MS) and mitochondria (Mi) in the SDH of DNP mice treated intraperitoneally with 2‐DG.

FIGURE 4

Sirt3 deficiency amplified glycolytic activity in primary microglia and intensifies their inflammatory activation in vitro. (A) Protein level of IBA‐1 in Sirt3 deficient primary microglia. (B, C) Activation of NF‐κb (B) and MAPK (C) pathway in Sirt3 deficient primary microglia. (D) Transcriptions of IL‐1β, TNF‐α and IL‐6 in Sirt3 deficient primary microglia (n = 4, *p < 0.05, **p < 0.01). (E) Protein expression levels of key enzymes (HK2, PKM2, LDHA) in glycolysis in Sirt3 deficient primary microglia. (F, G) Lactate levels (F) and pyruvate levels (G) in Sirt3‐deficient primary microglia (n = 4, *p < 0.05, **p < 0.01). (H, I) The experimental program of extracellular acidification rate (ECAR, H) and the oxygen consumption rate (OCR, I) of primary microglia, measured by Seahorse XFe24 Extracellular Flux Analyzer; n = 5, *p < 0.05, **p < 0.01 compared Sirt3^−/− high glucose group versus Sirt3^+/+ high glucose group.

FIGURE 5

Sirt3 overexpression in BV‐2 microglia cells reduced glycolysis and reversed their inflammatory activation in vitro. (A) Protein level of IBA‐1 in Sirt3‐overexpressed BV‐2 cells. (B) cell proliferation was tested by colorimetric determination of CCK8 in Sirt3‐overexpressed BV‐2 cells; n = 4 **p < 0.01. (C, D) Activation of NF‐κb (C) and MAPK (D) pathway in Sirt3‐overexpressed BV‐2 cells. (E) Transcriptions of IL‐1β, TNF‐α and IL‐6 in Sirt3‐overexpressed BV‐2 cells; n = 4 **p < 0.01. (F) Protein expression levels of key enzymes (HK2, PKM2, LDHA) in glycolysis in Sirt3‐overexpressed BV‐2 cells. (G, H) Lactate levels (G) and pyruvate levels (H) in Sirt3‐overexpressed BV‐2 cells; n = 4 **p < 0.01. (I, J) Extracellular acidification rate (ECAR, I) and the experimental program of the oxygen consumption rate (OCR, J) of BV‐2 cells, measured by Seahorse XFe24 Extracellular Flux Analyzer; n = 5 *p < 0.05 compared LV‐Sirt3 high glucose group versus LV‐vector high glucose group.

FIGURE 6

Inactivation of FoxO1 decreased Sirt3 transcription upon high glucose stimulation. (A) Network model describing protein–protein interactions between the FOXO1 and Sirt3. (B) Correlation analysis of FoxO1 and Sirt3 targets in human spinal cord tissue with GEPIA database (http://gepia.cancer‐pku.cn). (C) Luciferase reporting system evaluates the targeting effects of 293 T cells after co‐transfection with Sirt3 promoter plasmids and overexpressing‐FoxO1 plasmids; n = 3 **p < 0.01. (D) Ch‐IP analysis shows that the Sirt3 promoter sequence was pulled down by the anti‐FoxO1 antibody; n = 4 **p < 0.01. (E) Western blotting shows nuclear exclusion of FoxO1 in primary microglia under high glucose stimulation. (F) Western blotting reveals the Sirt3‐dependent impact of FoxO1 on microglial glycolysis under high glucose conditions. (G, H) the transcription (G) and protein production (H) of Sirt3 in primary microglia after AS1842856 treatment; n = 4 **p < 0.01.

FIGURE 7

Akt inactivated FoxO1 upon high glucose stimulation in microglia. (A) Time‐dependent phosphorylation of Akt and FoxO1 in the spinal dorsal horn of DNP mice. (B) Phosphorylation of Akt and FoxO1 in primary microglia under high glucose conditions. (C) Immunofluorescence analysis of FoxO1 translocation in microglia upon treatment with GSK690693. (D) Western blot shows the impact of GSK690693 on FoxO1 phosphorylation and Sirt3 downregulation under high glucose stimulation. (E) quantitative PCR analysis of Sirt3 expression under high glucose stimulation with and without GSK690693 treatment; n = 4 **p < 0.01. (F, G) Western blot shows impact of CQ (F) and MG132 (G) on Sirt3 expression in microglia under high glucose stimulation. (H) Effects of chloroquine on Sirt3 protein levels in microglial cells under high glucose conditions with GSK690693 treatment. (I, J) The transcription (I) and protein level (J) of Sirt3 in microglia after metformin treatment; n = 4 **p < 0.01.

FIGURE 8

Metformin alleviated neuroinflammation and DNP by regulating Sirt3 expression. (A) Blood glucose was measured in DNP mice treated intraperitoneally with metformin; n = 10, ns p > 0.05 compared metformin DNP group versus vehicle DNP group. (B, C) Paw withdrawal percentage in response to von Frey filament (0.07 g, B; 0.4 g, C) was measured in DNP mice treated intraperitoneally with metformin; n = 10 *p < 0.05, **p < 0.01 compared metformin diabetic group versus vehicle diabetic group. (D) Paw withdrawal latency in response to heat stimulus was measured in DNP mice treated intraperitoneally with metformin; n = 10 *p < 0.05 compared metformin diabetic group versus vehicle diabetic group. (E, F) The transcription (E) and protein level (F) of Sirt3 in spinal dorsal horn (SDH) after metformin treatment; n = 4 **p < 0.01. (G) Immunofluorescence shows the upregulation of IBA‐1 in SDH of DNP mice treated intraperitoneally with metformin. (H) Protein level of IBA‐1 in SDH of DNP mice treated intraperitoneally with metformin. (I) Transcriptions of inflammatory cytokines (IL‐1β, TNF‐α, IL‐6) in SDH of DNP mice treated intraperitoneally with metformin; n = 4 **p < 0.01.