Neuroscience

A paradoxical relationship between mitochondrial calcium regulation and retinal ganglion cell degeneration after axon damage

Sean McCracken
Minglei Zhao
Kyler Squirrell
Christopher Zhao
Saman Behboodi Tanourlouee
Michelle Aum
Philip R Williams author has email address

John F. Hardesty, MD Department of Ophthalmology and Visual Sciences, Washington University School of Medicine, St Louis, United States
Department of Neuroscience, Washington University School of Medicine, St Louis, United States
Hope Center for Neurological Disorders, Washington University School of Medicine, St Louis, United States

https://doi.org/10.7554/eLife.111321.1

Open access
Copyright information

Figures and data

In vivo RGC Mitochondrial Ca²⁺ levels.
(A) Example in vivo 2-photon max intensity projections (mip) of cytoplasmic (Cyto, left) and mitochondrial (Mito, right) Twitch-2b expressed in retinal ganglion cells (RGCs) of VGlut2-Cre mice. (B) Swarm plots of homeostatic cyto- and mito-Ca²⁺ levels, as measured by cpVenus (YFP) to mCerulean (CFP) FRET ratios, MWU test (cyto-T2b n = 868 RGCs from 14 retinas and 8 mice, mito-T2b n = 1066 RGCs from 19 retinas and 15 mice). (C) Average intensity projections (aip) of mito-T2b in vivo of the same RGCs at baseline and 4 days later. Magenta arrows indicate RGCs with higher baseline mito-Ca²⁺ and green arrows indicate RGCs with lower mito-Ca²⁺ levels. (D) Scatterplot of mito-Ca²⁺ levels pre and 4 day follow-up timepoints. R = 0.45, (n = 205 RGCs from 5 retinas and 3 mice). (E) Swarm plots of change in mito-Ca²⁺ levels from baseline to 4 day follow-up replotted from panel (D). (F) Example aips of mito-T2b at baseline and 10 min following intravitreal Ru265 injection. White arrows indicate RGCs with reduced mito-Ca²⁺. (G) Line graphs of mito-Ca²⁺ levels before and 10 min after Ru265 injection. Individual RGCs are in blue, and mean is in black (n = 175 RGCs from 4 retinas and 4 mice). (H) Example aips of cyto-T2b at baseline and 10 min after intravitreal Ru265. White arrows indicate RGCs with increased cyto-Ca²⁺. (I) Line graphs of cyto-Ca²⁺ levels before and 10 min after Ru265 injection. Individual RGCs are in orange, and mean is in black (n = 123 RGCs from 4 retinas and 4 mice). (J) Swarm plot comparing changes in Ca²⁺ levels after Ru265 between cyto- and mito-T2b. Black bars are mean +/- SEM replotted from panels (G and I). (K) Example aips of KCNG4-Cre mito-T2b acquired with in vivo transscleral imaging. GCL = ganglion cell layer, IPL = inner plexiform layer, NFL = nerve fiber layer. Box indicates magnified region. (L) Representative confocal mips of fixed retinal wholemounts showing endogenous mito-T2b (yellow) and TOMM20 (cyan) immunostaining. Scale bars = 100 μm.

Mitochondrial Ca²⁺ levels are higher in some well-surviving RGC subtypes.
(A) Example in vivo aips of mito-T2b from VGlut2-Cre transgenic mouse and the same imaged region in fixed retinal wholemount immunostained for SPP1 (red, αRGCs) and TBR2 (cyan, ip-RGCs). Mito-T2b is greyscale. Arrows represent cells identified as positive for the different stains: cyan = TBR2 only, red = SPP1 only, and white = SPP1 / TBR2 positive. (B) Swarm plots of mito-Ca²⁺ levels in individual RGCs positive for indicated RGC markers. Means are labeled with a black bar +/- SEM (n = 479 RGCs from 6 retinas and 6 mice). (C) Representative aips of mito-T2b expression in VGlut2-Cre (all RGCs) and KCNG4-Cre transgenic mice (αRGCs). (D) Swarm plots of mito-Ca²⁺ levels from VGlut2-Cre and KCNG4-Cre transgenic mice (n = 79 RGCs from 6 retinas and 3 mice). Scale bars = 100 μm.

Surviving RGCs have higher homeostatic mitochondrial Ca²⁺ levels.
(A) Timeline depicting in vivo imaging and ONC experiments. (B) Example in vivo 2-photon aips of mito-Ca²⁺ expressing RGCs before ONC and at the indicated days post ONC (dpc). (C) Swarm plots of pre-injury mito-Ca²⁺ levels in RGCs that survived (green) or died (red) following ONC. Means +/- SEM are labeled with black bars (survive n = 100 RGCs, die n = 669 RGCs from 5 retinas and 3 mice,). (D) Survival curves of RGCs with high (magenta) or low (green) pre-injury mito-Ca²⁺ levels. Error bars are SEM (n = 295 RGCs from 3 retinas and 3 mice). (E) Dynamic analysis of mito-Ca²⁺ levels in RGCs that survived (green) or died (red) during in vivo imaging time course. Statistics comparing surviving RGCs (green) or dying RGCs (red) to their pre-injury mito-Ca²⁺ levels, or surviving vs. dying RGCs (black) (from n = 3 retinas and 3 mice including 45 +/- 4.7 surviving RGCs and 119 +/- 38 dying RGCs per sample). (F) Dynamics of mito-Ca²⁺ levels between 2-day time points either prior to death (pre-death) (n = 126 RGCs from 3 retinas and 3 mice) or from all other timepoints (non-death) (n = 225 RGCs from 3 retinas and 3 mice). Scale bar = 100 μm.

Longitudinal Ru265 injections protect RGCs from ONC.
(A) Timeline depicting ONC and Ru265 injection experiments. (B) Representative mips of retinal wholemounts immunostained with RBPMS at 14 dpc with indicated treatments. (C) Quantification of RBPMS densities from indicated experimental conditions. Points are individual RGCs, bars are mean +/-SEM, MWU test (vehicle n = 12 retinas from 12 mice, 20 μM Ru265 n = 7 retinas from 7 mice, 200 μM Ru265 n = 5 retinas from 5 mice, 2 mM Ru265 n = 8 retinas from 8 mice). Scale bar = 100 μm.

Mitochondrial Calcium Uniporter expression regulates RGC survival after ONC.
(A) Schematic illustrating AAV constructs used to knockdown (shMCU) or overexpress (MCU-OX) mitochondrial calcium uniporter protein (MCU). (B) Time course of AAV injection and ONC experiments. (C) Representative mips of fixed retinal wholemounts immunostained for RBPMS at 14 dpc. (D-E) Quantification of RBPMS+ cell density in mice with intravitreal injections of indicated AAV MCU gene expression manipulations compared to littermate AAV-mCherry treated controls collected 14 dpc. Points indicated individual eyes bars indicate mean +/- SEM, MWU test (D: shMCU n = 8 retinas from 8 mice and mCherry n = 7 retinas from 7 mice, 6 eye matched; E: MCU-OX n = 8 retinas from 8 mice and mCherry n = 8 retinas from 8 mice, all eye matched). (F-G) Paired-eye comparison of RGC density following indicated AAV treatment or mCherry control. Samples are also included in panels D-E. Dashed line is y = x. Scale bar = 50 μm.

Ru265 decreases Mito-Ca²⁺ levels in a dose dependent manner.
(A) Example in vivo 2-photon aips of mito-T2b before and 10 min after injection of vehicle control (50% DMSO and 50% PBS). (B) Line graphs of mito-Ca²⁺ levels before and 10 min after vehicle injection. Individual RGCs are in grey, and mean is in black, MWU test (n = 95 RGCs from 3 retinas and 3 mice). (C) Swarm plot of changes in mito-Ca²⁺ levels after vehicle injection. (D) Example in vivo 2-photon aips of mito-T2b before and 10 min after injection of 20 μM Ru265. Arrow indicates RGC with reduction in mito-Ca²⁺. (E) Line graphs of mito-Ca²⁺ levels before and 10 min after 20 μM Ru265 injection. Individual RGCs are in grey, and mean is in black, MWU test (n = 101 RGCs from 4 retinas and 4 mice). (F) Swarm plot of changes in mito-Ca²⁺ levels after 20 μM Ru265injection. (G) Example in vivo 2-photon aips of mito-T2b before and 10 min after injection of 200 μM Ru265. Arrows indicate RGCs with reduction in mito-Ca²⁺. (H) Line graphs of mito-Ca²⁺ levels before and 10 min after 200 μM Ru265 injection. Individual RGCs are in grey, and mean is in black, MWU test (n = 159 RGCs from 4 retinas and 4 mice). (I) Swarm plot of changes in mito-Ca²⁺ levels after 200 μM Ru265 injection. Scale bars = 50 μm.

Ru265 induces Ca²⁺ elevations at 2 minutes post injection, likely due to DMSO vehicle.
(A) Example in vivo 2-photon aips of mito-T2b before and 2 min after injection of 2 mM Ru265 (left). Line graphs of mito-Ca²⁺ levels before and 2 min after 2 mM Ru265 injection. Individual RGCs are in blue, and mean is in black (n = 154 RGCs from 4 retinas and 4 mice). (B) Example in vivo 2-photon aips of cyto-T2b before and 2 min after injection of 2 mM Ru265 (left). Line graphs of cyto-Ca²⁺ levels before and 2 min after 2 mM Ru265 injection. Individual RGCs are in orange, and mean is in black (n = 96 RGCs from 4 retinas and 4 mice). (C) Swarm plots comparing changes in Ca²⁺ levels 2 min after 2 mM Ru265 between cyto- and mito-T2b. Black bars are mean +/- SEM, MWU test. (D) Example in vivo 2-photon aips of mito-T2b before and 2 min after injection of DMSO/PBS vehicle (left). Line graphs of mito-Ca²⁺ levels before and 2 min after vehicle injection. Individual RGCs are in blue, and mean is in black (n = 73 RGCs from 3 retinas and 3 mice). (E) Example in vivo 2-photon aips of cyto-T2b before and 2 min after injection of behicle (left). Line graphs of cyto-Ca²⁺ levels before and 2 min after vehicle injection. Individual RGCs are in orange, and mean is in black (n = 87 RGCs from 3 retinas and 3 mice). (F) Swarm plots comparing changes in Ca²⁺ levels 2 min after vehicle injection between cyto- and mito-T2b, MWU test. Scale bars = 50 μm.

Mitochondrial targeted Twitch2b mislocalization occurs after longer expression durations.
(A) Example confocal mips of fixed retinal wholemount showing endogenous mito-T2b expression (green) and TOMM20 immunostaining (magenta) at the indicated time points after AAV injection. Yellow arrow indicates RGC with partial mislocaliztion and orange arrows indicate RGCs with complete mislocaliztion of mito-T2b. (B) Example confocal mips of ‘healthy’ RGCs expressing mito-T2b co-localized with TOMM20, co-stained with RBPMS (blue). (C) Example confocal mips of RGCs expressing mito-T2b that is ‘partially mislocalizated’ and immunostained for TOMM20 and RBPMS. (D) Example confocal mips of RGCs expressing mito-T2b that is ‘completely mislocalizated’ and immunostained for TOMM20 and RBPMS. (E-F) Pie charts showing the proportion of RGCs in each of the qualitative mito-T2b localization groups at indicated time points after AAV injection (< 30 dpi n = 917 RGCs from 5 retinas and 3 mice; > 60 dpi n = 446 RGCs from 4 retinas from 4 mice mice). (G) Percentage of RGCs in indicated mislocalization groups at indicated time points. Graph is sample means +/-SEM (MWU test). (H-I) Z-score of mito-T2b expression intensity measure from fixed wholemounts representing indicated qualitative localization groups and time points after AAV injection. Graph is sample means +/- SEM. Scale bars = 10 μm.

Endoplasmic Reticulum Ca²⁺ levels are not correlated with survival to ONC.
(A) Example in vivo 2-photon mips of cytoplasmic cyto-T2b (left) and ER-2b (right) expressed in RGCs of VGlut2-Cre mice. (B) Swarm plots of homeostatic cyto-, mito-, and ER-Ca²⁺ levels, as measured by cpVenus (YFP) to mCerulean (CFP) FRET ratios (cyto-T2b and mito-T2b replotted from Figure 1; ER-T2b = 717 RGCs from 9 retinas and 5 mice. MWU test). (C) Example in vivo 2-photon aips of ER-Ca²⁺ expressing RGCs before ONC and at the indicated days post ONC (dpc). (D) Swarm plots of pre-injury ER-Ca²⁺ levels in RGCs that survived (green) or died (red) following ONC. Means +/- SEM are labeled with black bars (survive n = 28 RGCs from 4 retinas and 3 mice, die n = 133 RGCs. MWU test). Scale bars = 100 μm.

Chronic Ru265 treatment does not kill RGCs.
(A) Representative mips of retinal wholemounts immunostained with RBPMS at 14 days after sham ONC and indicated treatments. (B) Quantification of RBPMS densities from indicated experimental conditions. Points are individual retinas bars are mean +/- SEM (Sham ONC + Vehicle n = 5 retinas from 4 mice; Sham ONC + Sham injection n = 4 retinas from 4 mice; Sham ONC + Ru265 n = 5 retinas from 4 mice). (C) Representative mips of retinal wholemounts immunostained with RBPMS at 14 days after ONC and indicated sham treatments. (D) Quantification of RBPMS densities from indicated experimental conditions. Points are individual retinas bars are mean +/- SEM (ONC + Sham n = 6 retinas from 6 mice; ONC + PBS n = 5 retinas from 5 mice; ONC + Vehicle n = 6 retinas from 6 mice). Scale bars = 100 μm.

Validation of AAV mediated manipulation of MCU expression.
(A) Example confocal mips of retinal wholemounts immunostained for MCU (green) and RBPMS (blue) with endogenous mCherry reporter signal from AAV-MCU-OX treatment (red). (B) Scatterplot of MCU immunofluorescence intensity and mCherry reporter signal measured from confocal wholemount images. Dashed line represents trendline for all RGCs with a 95% confidence interval. Grey background shading indicates threshold for flagging RGCs as mCherry- or mCherry+. (C) Box and whisker plaots of MCU immunostaining intensity compared between MCU-OX mCherry reporter positive and negative RGCs. Points are individual RGCs, boxes depict the lower quartiles, the median, and the upper quartiles while the whiskers represent the minimum and maximum datapoint at 1.5 x the interquartile range. MWU test (n = 504 RGCs from 3 retinas and 3 mice). (D) Example confocal mips of retinal wholemounts immunostained for MCU (green) and RBPMS (blue) with endogenous mCherry reporter signal from AAV-shMCU treatment (red). (E) Scatterplot of MCU immunofluorescence intensity and mCherry reporter signal measured from confocal wholemount images. Dashed line represents trendline for all RGCs with a 95% confidence interval. Grey shading indicates threshold for flagging cells as mCherry- or mCherry+. (F) Box and whisker plots of MCU immunostaining intensity compared between shMCU mCherry reporter positive and negative RGCs. Points are individual cells, boxes depict the lower quartiles, the median, and the upper quartiles while whiskers represent the minimum and maximum datapoint at 1.5 x the interquartile range, MWU test (n = 489 RGCs from 3 retinas and 3 mice). Scale bars = 100 μm.

MCU-OX leads to slow RGC degeneration.
(A) Representative mips of fixed retinal wholemounts immunostained for RBPMS at 14 days after sham surgery. (B) Quantification of RBPMS+ cell density in mice with intravitreal injections of indicated AAV MCU gene expression manipulations compared to littermate AAV-mCherry treated controls. Points indicate individual eyes, bars indicate mean +/- SEM. (mCherry n = 6 retinas from 6 mice, shMCU n = 3 retinas from 3 mice, MCU-OX n = 7 retinas from 7 mice). (C) Representative mips of fixed retinal wholemounts immunostained for RBPMS at 8 weeks after sham surgery. (D) Quantification of RBPMS+ cell density in mice with intravitreal injections of indicated AAV MCU gene expression manipulations compared to littermate AAV-mCherry treated controls. Points indicate individual eyes, bars indicate mean +/- SEM, MWU test. (mCherry n = 4 retinas from 4 mice, shMCU n = 3 retinas from 3 mice, MCU-OX n = 3 retinas from 3 mice). Scale bars = 100 μm.

Sign up for email alerts