Mitochondrial fission and mitophagy are independent mechanisms regulating ischemia/reperfusion injury in primary neurons

Abstract

Mitochondrial dynamics and mitophagy are constitutive and complex systems that ensure a healthy mitochondrial network through the segregation and subsequent degradation of damaged mitochondria. Disruption of these systems can lead to mitochondrial dysfunction and has been established as a central mechanism of ischemia/reperfusion (I/R) injury. Emerging evidence suggests that mitochondrial dynamics and mitophagy are integrated systems; however, the role of this relationship in the context of I/R injury remains unclear. To investigate this concept, we utilized primary cortical neurons isolated from the novel dual-reporter mitochondrial quality control knockin mice (C57BL/6-Gt(ROSA)26Sortm1(CAG-mCherry/GFP)Ganl/J) with conditional knockout (KO) of Drp1 to investigate changes in mitochondrial dynamics and mitophagic flux during in vitro I/R injury. Mitochondrial dynamics was quantitatively measured in an unbiased manner using a machine learning mitochondrial morphology classification system, which consisted of four different classifications: network, unbranched, swollen, and punctate. Evaluation of mitochondrial morphology and mitophagic flux in primary neurons exposed to oxygen-glucose deprivation (OGD) and reoxygenation (OGD/R) revealed extensive mitochondrial fragmentation and swelling, together with a significant upregulation in mitophagic flux. Furthermore, the primary morphology of mitochondria undergoing mitophagy was classified as punctate. Colocalization using immunofluorescence as well as western blot analysis revealed that the PINK1/Parkin pathway of mitophagy was activated following OGD/R. Conditional KO of Drp1 prevented mitochondrial fragmentation and swelling following OGD/R but did not alter mitophagic flux. These data provide novel evidence that Drp1 plays a causal role in the progression of I/R injury, but mitophagy does not require Drp1-mediated mitochondrial fission.

Fig. 1. Mitophagic flux is increased following OGD/R.

A Schematic diagram of experimental design. B Utilizing primary cortical neurons from MitoQC reporter mice, cells were analyzed for the presence of mCherry puncta. Cells containing MitoQC were also stained with LAMP1, a lysosomal marker, to determine colocalization of mCherry puncta with lysosomes. C Quantification of mCherry puncta normalized to cell count. D Mander’s correlation coefficient was used to quantify colocalization of mCherry with LAMP1 throughout OGD/R. LAMP1 particle counts were analyzed and normalized to cell count. Differences across time were analyzed using one-way ANOVA with Tukey post-hoc analysis for multiple comparisons. R post-reoxygenation; *p < 0.05, **p < 0.01; ***p < 0.001 vs controls; n = 6 biological replicates. Scale bar = 10 µm.

Fig. 2. PINK1/Parkin-mediated mitophagy.

A Cells labeled with PINK1 and ATPB for colocalization of PINK1 with mitochondria, n = 4. B Cells labeled with Parkin and ATPB for colocalization of Parkin with mitochondria, n = 3. C Quantification of colocalization using Pearson’s correlation coefficient (PCC). D Western Blot of mitochondrial and cytosolic fractions for PINK1, Parkin, and ubiquitin. E Quantitation of protein levels, normalized to VDAC and GAPDH, Ubiquitin: n = 6, PINK1: n = 5, Parkin: n = 8. F Western Blot of LC3 (autophagy marker) and Rab5 (endosome marker) in mitochondrial and cytosolic fractions. G Quantitation of LC3 conversion and Rab5, normalized to VDAC and GAPDH, LC3: n = 4, Rab5: n = 4. Differences between groups were computed using one-way ANOVA with Dunnett’s post-hoc analysis for comparisons versus control. R post-reoxygenation; *p < 0.05; **p < 0.01; ***p < 0.001. Scale bar = 10 µm.

Fig. 3. Mitochondria morphology classification of mCherry puncta.

A Mitochondrial morphology was classified using a machine learning-based classification system. Mitochondria were classified into four different morphologies consisting of: networks (green), unbranched (blue), swollen (yellow), and punctate (red). Scale bar = 5 µm. B mCherry puncta were segmented from images of MitoQC neurons and analyzed using our mitochondrial morphology classification system. C Total mCherry puncta counts throughout reoxygenation (R). D Stacked bar graphs representing percent red puncta throughout R. mCherry puncta: n = 4 biological replicates, 44,536 objects. Scale bar = 10 µm.

Fig. 4. Drp1 KO stabilizes mitochondrial morphology during OGD/R.

A Schematic diagram of experimental design. B Western blot of Drp1 knockout after Lentiviral transduction in Drp1^fl/fl cortical neurons. C Viability of Drp1^fl/fl cortical neurons after 2.5 h OGD and 6 h reoxygenation (R). Comparison was made by Student’s t-test, n = 10. D Representative images of Drp1^fl/fl/QC^Ki/+ primary cortical neurons (GFP channel), infected with either LV-EF1a-empty or LV-EF1a-cre, and subjected to OGD/R. E Individual comparisons of percent mitochondrial objects for each morphology over OGD/R. F Individual comparisons of percent mitochondrial area for each morphology over OGD/R. Two-way ANOVA was used to detect differences across time and between conditions. Multiple comparisons across time were assessed by comparing the means of each time point with the mean of the control and were calculated using Dunnett’s post-hoc analysis. Multiple comparisons between conditions were computed using Sidak’s post-hoc analysis. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 across time, ^#p < 0.05; ^##p < 0.01; ^###p < 0.001; ^####p < 0.0001 between condition, n = 8 biological replicates per group. Scale bar = 10 µm.

Fig. 5. Mitophagic flux is independent of Drp1.

A Representative images of Drp1^fl/fl/QC^Ki/+ primary cortical neurons, infected with either LV-EF1a-empty or LV-EF1a-cre, subjected to OGD/R and analyzed for the presence of mCherry puncta. B Quantification of mCherry puncta per cell and total LAMP1 puncta per cell during OGD/R. Two-way ANOVA was used to detect differences across time and between conditions. Multiple comparisons across time were assessed by comparing the means of each time point with the mean of the control and were calculated using Dunnett’s post-hoc analysis. Multiple comparisons between conditions were computed using Sidak’s post-hoc analysis. *p < 0.05; **p < 0.01 across time, ^#p < 0.05 between condition, n = 8 biological replicates per group. Scale bar = 10 µm.