Animals
Male heterozygous Ins2Akita (C57BL/6J) mice (originally obtained from the Jackson Laboratory, Bar Harbor, USA, stock number 003548) and age-matched non-diabetic siblings (WT) were used. The Ins2Akita mice develop severe hyperglycaemia by 4 weeks of age (≥ 550 mg/dL or 30.5mM), due pancreatic β-cell loss recapitulating the pathophysiology of type-1 diabetes47. mitoQC Ins2Akita mice were generated as previously reported10 by mating mitoQC+/+ females with Ins2Akita males. The diabetic phenotype in male offspring was corroborated by blood glucose levels (above 550 mg/dL), while detection of Ins2Akita and mitoQC-knockin alleles (mCherry-GFP-mtFIS1101–153) by PCR32. All mice were kept in standard pathogen-free animal housing rooms with 12/12 h light/dark cycle and access to food and water ad libitum.
Animal treatments
Ins2Akita mice were treated via supplementation of KR (60mg/L) or vehicle control (0.1% DMSO) in the drinking water three times per week (given every other day). Age-matched mice not receiving treatment (Ins2Akita and WT) served as full controls. Effective KR dosage was inferred via supplementation of KR to MitoQC Ins2Akita mice from 7.5-months to 8-months of diabetes, by assessing its bioactivity to activate mitochondrial turnover in the diabetic retina (Fig. 6). For long-term treatments, Ins2Akita mice were treated with KR or vehicle control from 4-months to 8-months of diabetes (Fig. 6). No differences were observed in the amount of water consumed by the diabetic groups (30–35 ml per mouse/day; WT group was 4–5 ml per mouse/day). Mice were randomly assigned to each treatment group.
Spectral-Domain Optical Coherence Tomography (SD-OCT)
SD-OCT was performed in 8-months diabetic Ins2Akita mouse groups and age-matched WT mice as previously described5,48. Briefly, mice were anesthetized (ketamine 90 mg/kg and xylazine 10 mg/kg) and pupils dilated using 1% tropicamide and 2.5% phenylephrine (Chauvin, Essex, United Kingdom). The Spectralis-Heidelberg OCT system (Heidelberg Engineering, Heidelberg, Germany) was used to evaluate in vivo the total retinal thickness (from nerve fibre layer [NFL] to the photoreceptor inner segments/outer segments [IS/OS]). Retinal thickness was measured at 600 µm eccentricities from the optic disc in dorso-ventral and nasal-temporal regions.
Electroretinography (ERG)
ERG responses were recorded by 8-months of diabetes in Ins2Akita mouse groups and age-matched WT mice as previously described5,20. In brief, scotopic ERG responses were recorded in anesthetized mice (see above) using a Espion Visual Electrophysiology System (Diagnosys Technologies, Cambridge UK). Scotopic a-wave and b-wave amplitudes were obtained using mouse corneal ERG electrodes (Diagnosys Technologies) in response to single white light-flash intensities (ranging from 0.025–25 cd·s/m2). ERG signals were averaged from 5 responses at each intensity level, with an interstimulus interval of 10 seconds (0.008, 0.025 cd·s/m2), 15 seconds (0.08, 0.25, 0.8 cd·s/m2) or 30 seconds (2.5, 8, 25 cd·s/m2).
Immunohistochemistry (IHC) of human retinas
Age-matched human post-mortem retinas from male and female diabetic donors (n = 6 without clinical retinopathy; 2 with NPDR) and non-diabetic donors (n = 3) were obtained from the National Disease Research Interchange (Philadelphia, Pennsylvania, USA) as previously described10. After deparaffinization, retinal sections were washed in PBS and antigen retrieval conducted by immersion (1 hour) in antigen retrieval buffer (EDTA, pH 8.0; Thermo Fisher Scientific) at 60°C. Sections were then rinsed in PBS and incubated overnight (4°C) with primary antibodies (Cox4, M-opsin, synaptophysin or Calbindin [Supplemental Table 1]) diluted in 0.5% TritonX-100 with 10% normal donkey serum (NDS) in PBS. Following incubation, retinal sections were rinsed in PBS and incubated for 1h with appropriate fluorophore-conjugated secondary antibodies. Retinal sections were cover-slipped with Vectashield/DAPI (Vector Labs, Peterborough, UK) and examined by confocal microscopy (C1 Nikon Eclipse TE200-U, Nikon UK Ltd, Surrey, UK).
IHC of mouse retinas
Eyes were dissected and fixed in 4% paraformaldehyde (Sigma-Aldrich, Dorset, UK) for 2h and processed for IHC as previously described49,50. Briefly, 14 µm-thick retinal cryosections were blocked with 10% NDS in PBS and then incubated overnight (4°C) with primary antibodies (Supplemental Table 1) diluted in 0.5% TritonX-100 with 10% NDS in PBS. Following incubation, samples were processed and examined by confocal microscopy as described for human retinas.
Western blotting
Mouse retinas were lysed in RIPA Lysis Buffer with protease and phosphatase inhibitors cocktails (Sigma-Aldrich). Protein samples (20 µg) were run on 10% (w/v) SDS-PAGE gel and then transferred to PVDF membranes (Millipore, Watford, UK). Samples were then probed for primary antibodies (including β-actin loading controls; Supplemental Table 1) followed by incubation with appropriate HRP-conjugated secondary antibodies, and imaged with G:BOX chemiluminescence system (Syngene; Cambridge, UK) as previously described10.
Cell culture
The human Müller cell line MIO-M1 was obtained from the UCL Institute of Ophthalmology (London, UK). PMCs from mitoQC+/+ mice were isolated and cultured as previously described51 and used for experiments from P2–P4 to avoid variability due to cellular senescence. MIO-M1 and PMCs cultures were maintained in DMEM (containing 10% Fetal Bovine Serum [FBS] and 1% Penicillin/Streptomycin [Thermo-Fisher]), and supplemented with 5.5mM D-glucose (NG) or 30.5mM D-glucose (Sigma-Aldrich; HG) to mimic the hyperglycaemic context in diabetes10. Cells were routinely passaged at 1:4 every 3–4 days endpoint experiments performed in 70–80% confluent cultures. No mycoplasma was detected in the cell cultures.
Cell treatments
MIO-M1 or Mito-QC PMCs were seeded onto 96 or 24-well plates and grown overnight before treatment with compounds of interest. To antagonize mitochondrial fission, cells were pre-incubated with P110 Drp1-inhibitor peptide (1µM; Bio-Techne, Abingdon, UK)22 in NG for 2h. Cultures were then washed in PBS and supplemented with P110 (1µM) in either NG or HG for 72h. Additionally, other compounds of interest, including kinetin (0.3–5 µM, Sigma-Aldrich), KR (0.3–5 µM; Sigma-Aldrich), niclosamide (1 µM; Bio-Techne), urolithin-A (50 µM; Bio-Techne) or vehicle-control (DMSO 0.01%; Sigma-Aldrich) were treated alongside in conditions of NG or HG ± P110. Treatments were changed daily and lasted between 24–72 h depending on the final experimental readout (see below).
Extracellular metabolic flux analysis
Oxygen Consumption rate (OCR) and extracellular acidification rates (ECAR) were obtained using the Seahorse XFe-96 Flux Analyzer (Cell Mito Stress Test; Agilent Technologies, Stockport, UK). MIO-M1 cells and PMCs were seeded at 1200 or 4000 cells/well respectively onto a 96-well microplate, and treated for 72 hours prior to the assay. Cells were then washed and medium replaced with Seahorse XF DMEM assay medium (without phenol red) supplemented with glucose (7 mM), sodium pyruvate (1mM), HEPES (5 mM) and L-glutamine (2 mM). The run consisted of 3 minutes mixing, 3 minutes measuring and subsequent 4 injections as follows: Oligomycin (2µM), FCCP (0.75µM), Rotenone (0.5µM) and Antimycin-A (0.5µM). Following last Antimycin-A run, Hoechst (Thermo Fisher) was injected (final concentration 1ug/mL) and cell nuclei signal imaged (4X magnification) by epifluorescence microscopy (Olympus IX7 coupled to Retiga R6 CCD digital camera). Flux rate data was normalized to total nuclear area in each well.
Luminex® Multiplex Assay
Following treatments for 72h, the production of 5 cytokines (CCL2, VEGF-A, IL-6, IL-8 and GM-CSF) in MIO-M1 cell supernatants were evaluated using the Luminex bead-based assay (Bio-Techne) according to manufacturer’s instruction. In brief, 50µl of the undiluted sample was added to a mixture of colour-coded beads pre-coated with 5 capture antibodies. Following incubation (room temperature for 2h), 50ul of biotinylated-detection antibodies specific to the analytes of interest were added to the mixture and incubated (room temperature for 1h). Phycoerythrin (PE)-conjugated streptavidin was then added to the mixture, incubated at room temperature for 30 min and beads read using the Luminex MAGPIX CCD Imager System (Luminex, Austin, TX). The levels of each analyte in each sample were normalised against DAPI+ cell counts.
JC-1 dye staining
The ψm was assessed in human MIO-M1 cells by ratiometric analysis of JC-1 dye (Thermo Fisher). Following treatments for 72h, cultures were supplemented with 0.5 µg/mL JC-1 (30 minutes at 37°C) and returned to DMEM for live cell epifluorescence microscopy at 37°C, (Olympus IX7 coupled to Retiga R6 CCD digital camera [QImaging, Teledyne]). Positive controls were supplemented for 30 mins with carbonyl cyanide m-chlorophenylhydrazone (CCCP, 100 µM; Abcam, Cambridge, UK) to uncouple mitochondria. The ψm was assessed by calculating the ratio of mean fluorescent intensity (MFI) between red J-aggregates to green monomers.
Immunocytochemistry
Cells were fixed in 4% paraformaldehyde (PFA), rinsed in PBS, permeabilised with Triton X-100 (0.5%) and blocked with 10% FBS in PBS. Cells were then incubated overnight (4°C) with primary antibodies (Supplemental Table 1) diluted in 10% FBS and 0.05% Tween-20 (Sigma-Aldrich) PBS. After incubation, cells were rinsed in PBS, probed with appropriate fluorophore-conjugated secondary antibodies (room temperature for 1h), washed in PBS and stained with DAPI nuclear dye (1:1000; Sigma-Aldrich) for further microscopy imaging.
MIO-M1 neuronal differentiation
MIO-M1 Müller cells were induced for neuronal differentiation using an adapted protocol from Lawrence et al (2011)27. In brief, cells were plated for 24h and then supplemented with PromoCell Neurogenic Differentiation Medium (C-28015, PromoCell GmbH, Heidelberg Germany). Neuronal differentiation occurred rapidly (48h), as exhibited by the expression of mature neuronal markers (heavy-chain neurofilament, β-III tubulin) and by the development of complex neurite networks (Fig. 4G). Following differentiation, cells were treated for further 24h in PromoCell Neurogenic Differentiation Medium supplemented with either NG or HG ± P110. Following treatment, cells were fixed in 4% PFA and stained with Phalloidin-FITC (1:1000; Abcam) for further microscopy and morphometric analysis (see below).
Imaging morphometry
Confocal images were acquired under constant photomultiplier settings (C1-Nikon_Eclipse TE200-U) and analyzed using FIJI software (National Institutes of Health, Bethesda, MD). To avoid experimental bias during imaging, retinal regions were selected based on DAPI nuclear signal from middle-center eccentricities. For cell cultures, epifluorescence images (Olympus IX7 coupled to Retiga R6 CCD digital camera) were selected from equivalent cardinal points using bright-field or DAPI nuclear signal. All images were background subtracted prior to analysis.
Mitochondrial fusion. The AngioTool plugin52 was adapted to evaluate the morphology of the mitochondrial network from Fis1, Cox4 or TOMM20 fluorescence signals. Images from retinal sections (ONL) or individual cells were threshold binarized (constant values were applied for all groups). Images were then assessed for i) average mitochondrial length (equivalent to average vessel length); ii) interconnectivity index (equivalent to junctions density); iii) mitochondrial mass (equivalent to vessels percentage area). Aspect ratio, an established readout of mitochondrial fusion53, was also obtained via particle analysis from same images.
Vimentin and TFAM in human and mouse retinas. For each particular marker different retinal sectors were threshold and binarized (under constant values), including Synaptophysin (OPL), Vimentin (IS-OPL and inner plexiform layer-ganglion cell layer [IPL-GCL]) and TFAM (ONL). Images were then evaluated by particle analysis and the area of positive signal normalized against the total area analysed.
Retinal neurodegeneration in human and mouse retinas. Were quantified in accordance with our previously published methods5,20,21 including i) density of DAPI+ nuclei (ONL), ii) Area of synaptophysin-immunoreactive puncta (OPL), iii) Cone segment length (outer segment plus inner segment), iv) horizontal cell dendritic spine density (OPL) and v) GABAergic amacrine cell density (inner nuclear layer [INL]). Values were normalized to 100 µm retinal length or to the total area analysed.
Quantification of mitolysosomes (mCherry-only foci) at the outer retina and mitoQC PMCs. The total mitolysosome area was quantified in accordance with our previously published methods10, and values normalized to mitochondrial mass as obtained from Fis1-GFP signal.
Quantification of Vimentin and TFAM expression in Müller cell cultures. Were obtained by the mean fluorescence intensity (MFI) per cell after background subtraction.
Neurite retraction in differentiated MIO-M1 cultures. Was obtained by measuring the length of the longest neurite (from soma) in each cell.
Statistics
For each age group, the difference between 2 means was analysed using 2-sided unpaired Student’s t test. For comparisons with more than 2 groups, One-way ANOVA (followed by Bonferroni’s or Dunnet’s post hoc analysis) was used. ERG responses (scotopic a-wave, b-wave) were analyzed using 2-way ANOVA (followed by post hoc Bonferroni pairwise comparisons). Animal sample size was adjusted to n = 4–9 eyes in each mouse group based on 80% power and a 5% significant level. For retinal analyses, 6–8 images (obtained from at least 2 retinal sections/eye) were evaluated and data averaged per eye for statistical analysis. For Müller cell cultures, at least 50 cells (mitochondrial morphology, mitolysosomes, and neurite retraction) or 200 cells (ψm, TFAM and Vimentin expression) per technical replicate were measured and averaged (at least n = 6 technical replicates obtained from n = 2 biological replicates per group). N numbers for and statistical analysis for each readout were specified in Figure legends. Data were expressed as mean ± SEM or as in box-and-whisker plots, showing the median, interquartile ranges (IQRs), and minimum and maximum values. P < 0.05 was considered statistically significant. All statistical analyses were performed using GraphPad Software (La Jolla, CA)
Study approval
The study was approved by the Ethics Committee of the University of Birmingham (Project Licence Number PP6860623), Queen's University Belfast (Project Licence Number PPL2814) and Oklahoma Health Sciences Centre (OUHSC). Human studies were conducted according to the Declaration of Helsinki principles, and written informed consent was received from participants prior to inclusion in the study. All animal procedures were approved by Ethical Review Body (AWERB) and authorized under the UK Animals (Scientific Procedures) Act 1986. Animal use conformed to the standards in the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research and with European Directive 210/63/EU.