The Protective Role of Microglial PPARα in Diabetic Retinal Neurodegeneration and Neurovascular Dysfunction

Abstract

Microglial activation and subsequent pathological neuroinflammation contribute to diabetic retinopathy (DR). However, the underlying mechanisms of microgliosis, and means to effectively suppress pathological microgliosis, remain incompletely understood. Peroxisome proliferator-activated receptor alpha (PPARα) is a transcription factor that regulates lipid metabolism. The present study aimed to determine if PPARα affects pathological microgliosis in DR. In global Pparα mice, retinal microglia exhibited decreased structural complexity and enlarged cell bodies, suggesting microglial activation. Microglia-specific conditional Pparα^-/- (PCKO) mice showed decreased retinal thickness as revealed by optical coherence tomography. Under streptozotocin (STZ)-induced diabetes, diabetic PCKO mice exhibited decreased electroretinography response, while diabetes-induced retinal dysfunction was alleviated in diabetic microglia-specific Pparα-transgenic (PCTG) mice. Additionally, diabetes-induced retinal pericyte loss was exacerbated in diabetic PCKO mice and alleviated in diabetic PCTG mice. In cultured microglial cells with the diabetic stressor 4-HNE, metabolic flux analysis demonstrated that Pparα ablation caused a metabolic shift from oxidative phosphorylation to glycolysis. Pparα deficiency also increased microglial STING and TNF-α expression. Taken together, these findings revealed a critical role for PPARα in pathological microgliosis, neurodegeneration, and vascular damage in DR, providing insight into the underlying molecular mechanisms of microgliosis in this context and suggesting microglial PPARα as a potential therapeutic target.

Keywords: diabetic retinopathy; glycolysis; inflammation; microglial metabolism; mitochondrion; neurodegeneration; pericyte loss.

Figure 1

Microglial activation in Ppara^−/− retinas. (A) Representative flat-mounted retina from 9-month-old Wild-type (WT) and Pparα^−/− mice stained with anti-Iba1 antibody. (B) Microglia counts in WT and Pparα^−/− retinas. (C–F) Microglial morphological analysis with Fiji ImageJ. Images were skeletonized (D) and overlaid with original images (E) to verify accuracy. Branches and intersections were labeled (F). (G,H) Quantification of microglial branch length (G) and endpoints (H) in WT and Ppara^−/− retinas. (I,J) Cell bodies were isolated from images of WT (I) and Pparα^−/− (J) retinas. (K) Quantification of WT and Pparα^−/− microglial soma size. n = 5–6. Data are presented as means ± SD. * p < 0.05, *** p < 0.001, **** p < 0.0001, Student’s t-test. Scale bar: 50 μm.

Figure 2

Decreased retinal thickness in PCKO mice. (A,B) Diagrams of PCKO-Cx3Cr1Cre (A) and PCTG-Cx3Cr1Cre (B) mouse lines constructs. (C) Representative retina fundus and OCT images of non-diabetic WT, PCKO, and PCTG mice (8 months of age). (D) Total retinal thickness quantified by OCT. n = 5–9. * p < 0.05, ** p < 0.01, One-way ANOVA with Tukey’s posthoc comparison. Abbreviations: IPL: inner plexiform layer, INL: inner nuclear layer; OPL: outer plexiform layer, ONL: outer nuclear layer; GCL: ganglion cell layer; RPE: retinal pigment epithelium, WT: wild-type, PCKO: Microglial Pparα conditional knockout, PCTG: Microglial Pparα conditional transgenic mice.

Figure 3

Increased microglial density in diabetic PCKO mice. (A) Representative immunostaining images with Iba1⁺ microglia (arrows) in retina sections. Scale bar: 100 μm. (B) Quantification of retinal microglia density. n = 4–7 mice. (C) Subretinal microglial density on RPE/Choroid flat mounts, n = 4–7 mice. One-way ANOVA with Tukey’s posthoc comparison. Data are presented as mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. Abbreviations: IPL: inner plexiform layer, INL: inner nuclear layer; OPL: outer plexiform layer, ONL: outer nuclear layer; GCL: ganglion cell layer; RPE: retinal pigment epithelium, DM: diabetes mellitus, NDM: non-diabetes, WT: wild-type, PCKO: Microglial Pparα conditional knockout, PCTG: Microglial Pparα conditional transgenic mice.

Figure 4

Impaired retinal function in diabetic PCKO mice. A-wave (A,C) and b-wave (B,D) amplitudes of scotopic ERG (Rod, (A,B)) and photopic ERG (Cone, (C,D)) were recorded in diabetic mice and age-matched non-diabetic controls at 6 months after diabetes onset. n = 10–12. Data are presented as mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, one-way ANOVA with Tukey’s posthoc comparison. Abbreviations, DM: diabetes mellitus, NDM: non-diabetes, WT: wild-type, PCKO: Microglial Pparα conditional knockout, PCTG: Microglial Pparα conditional transgenic mice.

Figure 5

Exacerbated pericyte loss in diabetic PCKO mice. (A) Retinal vasculature was prepared by trypsin digestion followed by periodic-acid-Schiff staining. Scale bar: 20 μm. Pericytes (arrows) were counted and compared. Scale bar: 20 μm. (B) Quantitative analysis of retinal pericyte density. n = 5–6. Data are presented as mean ± SD. * p < 0.05, **** p < 0.0001, one-way ANOVA with Tukey’s posthoc comparison. Abbreviations, DM: diabetes mellitus, NDM: non-diabetes, WT: wild-type, PCKO: Microglial Pparα conditional knockout, PCTG: Microglial Pparα conditional transgenic mice.

Figure 6

Altered metabolic profile with Pparα deletion. (A) Representative real-time trace of oxygen consumption rates (OCR) using a Mito stress test in human microglia exposed to 4-HNE for 24 h following Pparα siRNA knockdown. (B–D) Quantification of basal OCR (B), Maximal respiration (C), and ATP production (D) in Pparα siRNA knockdown HMC3 cells relative to HMC3 cells transfected with scramble siRNA, n = 5–8. Student’s t-test. * p < 0.05, ** p < 0.01. (E) Glycolytic proton efflux rate (glycoPER) measurement obtained from microglia cells exposed to 4-HNE for 24 h following Pparα knockdown and subject to the glycolytic rate assay. n = 5–8. * p < 0.05, Student’s t-test. (F,G) Mito stress test assay in mouse WT and Pparα^−/− mouse primary microglial cells exposed to 4-HNE for 24 h. Quantification of basal OCR (F) and ATP production (G). (H) Glycolytic rate assay in mouse WT and Pparα^−/− mouse primary microglial cells exposed to 4-HNE for 24 h. n = 5–8. Student’s t-test. Data are presented as mean ± SD. *** p < 0.001, **** p < 0.0001.

Figure 7

Increased cytokine release in microglia with PPARα deletion. (A,B) STING and TNFα protein levels in primary WT and Pparα^−/− microglial cells. n = 3. ** p < 0.01, *** p < 0.001, Student’s t-test. (C) PPARα protein levels as quantified by Western blot analysis in HMC3 cells following 24 h 4-HNE exposure. n = 3. ** p < 0.01, Student’s t-test. (D) STING/CD11b co-immunostaining (arrows) in WT-DM, PCKO-DM, and PCTG-DM retinas. Scale bar: 50 µm. (E) Quantification of STING expression per microglial cell. n = 4−6. **** p < 0.0001, One-way ANOVA with Tukey’s posthoc comparison. Data are presented as mean ± SD.