Advertisement

Hyperglycemia Promotes Mitophagy and Thereby Mitigates Hyperglycemia-Induced Damage

Published:September 02, 2022DOI:https://doi.org/10.1016/j.ajpath.2022.08.004
      The observation that diabetic retinopathy (DR) typically takes decades to develop suggests the existence of an endogenous system that protects from diabetes-induced damage. To investigate the existance of such a system, primary human retinal endothelial cells were cultured in either normal glucose (5 mmol/L) or high glucose (30 mmol/L; HG). Prolonged exposure to HG was beneficial instead of detrimental. Although tumor necrosis factor-α–induced expression of vascular cell adhesion molecule 1 and intercellular adhesion molecule 1 was unaffected after 1 day of HG, it waned as the exposure to HG was extended. Similarly, oxidative stress–induced death decreased with prolonged exposure to HG. Furthermore, mitochondrial functionality, which was compromised by 1 day of HG, was improved by 10 days of HG, and this change required increased clearance of damaged mitochondria (mitophagy). Finally, antagonizing mitochondrial dynamics compromised the cells' ability to endure HG: susceptibility to cell death increased, and basal barrier function and responsiveness to vascular endothelial growth factor deteriorated. These observations indicate the existence of an endogenous system that protects human retinal endothelial cells from the deleterious effects of HG. Hyperglycemia-induced mitochondrial adaptation is a plausible contributor to the mechanism responsible for the delayed onset of DR; loss of hyperglycemia-induced mitochondrial adaptation may set the stage for the development of DR.
      Almost 10% of the population has diabetes mellitus (DM). The incidence of DM is increasing, with no clear path to curb this trend. People with DM develop complications that reduce their quality of life. For instance, diabetic retinopathy (DR) is the leading cause of blindness in working-age individuals.
      • Lee R.
      • Wong T.Y.
      • Sabanayagam C.
      Epidemiology of diabetic retinopathy, diabetic macular edema and related vision loss.
      Not only approaches to prevent DM, but also to protect people with DM from progressing to DR, have been ineffective. Consequently, most patients with DM still eventually develop DR.
      • Aiello L.P.
      • Gardner T.W.
      • King G.L.
      • Blankenship G.
      • Cavallerano J.D.
      • Ferris 3rd, F.L.
      • Klein R.
      Diabetic retinopathy.
      • Klein R.
      • Klein B.E.
      • Moss S.E.
      • Davis M.D.
      • DeMets D.L.
      The Wisconsin epidemiologic study of diabetic retinopathy, II: prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years.
      Diabetes Control and Complications Trial Research Group
      Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus: Diabetes Control and Complications Trial.
      The Diabetes Control and Complications Trial Research Group
      Clustering of long-term complications in families with diabetes in the diabetes control and complications trial.
      • Cruickshanks K.J.
      • Moss S.E.
      • Klein R.
      • Klein B.E.
      Physical activity and proliferative retinopathy in people diagnosed with diabetes before age 30 yr.
      • Chew E.Y.
      • Ambrosius W.T.
      • Davis M.D.
      • Danis R.P.
      • Gangaputra S.
      • Greven C.M.
      • Hubbard L.
      • Esser B.A.
      • Lovato J.F.
      • Perdue L.H.
      • Goff Jr., D.C.
      • Cushman W.C.
      • Ginsberg H.N.
      • Elam M.B.
      • Genuth S.
      • Gerstein H.C.
      • Schubart U.
      • Fine L.J.
      Effects of medical therapies on retinopathy progression in type 2 diabetes.
      • Cheung N.
      • Mitchell P.
      • Wong T.Y.
      Diabetic retinopathy.
      • Yau J.W.
      • Rogers S.L.
      • Kawasaki R.
      • Lamoureux E.L.
      • Kowalski J.W.
      • Bek T.
      • et al.
      Meta-Analysis for Eye Disease Study Group: Global prevalence and major risk factors of diabetic retinopathy.
      The only DR prophylactic is controlling blood sugar, which is fraught with noncompliance and ineffectiveness—even some diligently compliant patients develop DR.
      • Klein R.
      • Lee K.E.
      • Gangnon R.E.
      • Klein B.E.
      The 25-year incidence of visual impairment in type 1 diabetes mellitus the Wisconsin epidemiologic study of diabetic retinopathy.
      While the current understanding of DR (which includes self-amplifying oxidative stress–induced damage) predicts that it will develop soon after the onset of DM,
      • Brownlee M.
      Biochemistry and molecular cell biology of diabetic complications.
      • Stitt A.W.
      • Curtis T.M.
      • Chen M.
      • Medina R.J.
      • McKay G.J.
      • Jenkins A.
      • Gardiner T.A.
      • Lyons T.J.
      • Hammes H.P.
      • Simo R.
      • Lois N.
      The progress in understanding and treatment of diabetic retinopathy.
      • Antonetti D.A.
      • Klein R.
      • Gardner T.W.
      Diabetic retinopathy.
      • Kowluru R.A.
      • Kowluru A.
      • Mishra M.
      • Kumar B.
      Oxidative stress and epigenetic modifications in the pathogenesis of diabetic retinopathy.
      often, DR develops after many years of DM. Furthermore, some patients with DM do not develop complications (including DR) for ≥50 years, and such resistance is unrelated to glycemic control.
      • Bain S.C.
      • Gill G.V.
      • Dyer P.H.
      • Jones A.F.
      • Murphy M.
      • Jones K.E.
      • Smyth C.
      • Barnett A.H.
      Characteristics of type 1 diabetes of over 50 years duration (the Golden Years Cohort).
      These clinical observations suggest the presence of an endogenous system that protects from the deleterious effects of DM. The existence of such a system remains speculative, and the reason for the delay in the onset of DR in patients with DM is unknown.
      Mitochondrial dynamics ensure that the quality and capability of the mitochondria are in sync with the environment of the cells.
      • Pickles S.
      • Vigie P.
      • Youle R.J.
      Mitophagy and quality control mechanisms in mitochondrial maintenance.
      • Skeie J.M.
      • Nishimura D.Y.
      • Wang C.L.
      • Schmidt G.A.
      • Aldrich B.T.
      • Greiner M.A.
      Mitophagy: an emerging target in ocular pathology.
      • Kluge M.A.
      • Fetterman J.L.
      • Vita J.A.
      Mitochondria and endothelial function.
      • Westermann B.
      Mitochondrial fusion and fission in cell life and death.
      This process involves fusion and fission that mixes and redistributes the mitochondrial contents to sequester dysfunctional mitochondria, which are eliminated by mitophagy. Mutation of key governors of mitochondrial dynamics either causes or is associated with diseases that afflict a large percentage of the world's population (eg, neurodegenerative disease, diabetes, retinopathies, cancer, and cardiomyopathies).
      • Skeie J.M.
      • Nishimura D.Y.
      • Wang C.L.
      • Schmidt G.A.
      • Aldrich B.T.
      • Greiner M.A.
      Mitophagy: an emerging target in ocular pathology.
      ,
      • Chan D.C.
      Mitochondrial dynamics and its involvement in disease.
      • Chen Y.
      • Dorn 2nd, G.W.
      PINK1-phosphorylated mitofusin 2 is a Parkin receptor for culling damaged mitochondria.
      • Chen H.
      • McCaffery J.M.
      • Chan D.C.
      Mitochondrial fusion protects against neurodegeneration in the cerebellum.
      This article describes the discovery of a high glucose (HG)–induced system that protects retinal vascular cells from the damaging effects of HG.

      Materials and Methods

      Tissue Culture

      Primary human retinal endothelial cells (HRECs) were purchased from Cell Systems (ACBRI 181; Kirkland, WA). They were isolated from donor A, a 26-year–old White man. Cells were authenticated for cytoplasmic von Willebrand factor/factor VIII, cytoplasmic uptake of acetylated, fluorescently-labeled low-density lipoprotein (Di-I-AC-LDL), and cytoplasmic CD31, glial fibrillary acidic protein, NG2, and platelet-derived growth factor receptor-β by immunofluorescence. Mycoplasma, fungal, and bacterial sterility was confirmed using a culture method. Cells were cultured in endothelial cell basal medium-2 (EBM-2; CC3156; Lonza, Basel, Switzerland) supplemented with microvascular endothelial SingleQuots kit (EGM-2MV; CC4147; Lonza). The medium was refreshed daily. The glucose concentration in normal glucose (NG) and HG medium was 5 and 30 mmol/L d-glucose, respectively. In cells cultured in NG, the concentration after 0, 8, 16, and 24 hours was 5.2, 4.7, 4.3, and 3.4 mmol/L, respectively.
      Primary human glomerular microvascular endothelial cells were purchased from Cell Systems (ACBRI 128; Kirkland, WA). These cells were cultured the same as HRECs.
      Primary human retinal pericyte cells (HRPCs) were purchased from Cell Systems (ACBRI 183). These cells were cultured in Dulbecco’s modified Eagle’s medium with l-glutamine and sodium bicarbonate, supplemented with 10% fetal bovine serum (MT35010CV; Thermo Fisher Scientific, Waltham, MA) and penicillin/streptomycin. The medium was refreshed daily in NG or HG conditions, same as for HRECs.
      The 293T cells were purchased from ATCC (Manassas, VA). These cells were cultured in Dulbecco’s modified Eagle’s medium with l-glutamine and sodium bicarbonate, supplemented with 10% fetal bovine serum and penicillin/streptomycin in a 5% CO2 tissue culture incubator.
      For lentivirus production, 70% confluent 293T cells were transfected with Lipofectamine 2000 (catalog number 11668019; Invitrogen, Waltham, MA) complexed with the packaging plasmid (psPAX2), envelope plasmid (pVSVg), and lenti plasmid of interest [mito-roGFP2-Orp1 or pHAGE-mtKeima (131626; Addgene, Watertown, MA)]. The supernatant containing the virus was collected for 3 consecutive days, aliquoted, and stored at –80°C. Primary cells were infected with lentivirus harboring mito-roGFP2-ORP1 or pHAGE-mtKeima with 8 μg/mL polybrene reagent added to the medium. On the following day, the medium was replaced with complete growth medium. The infection efficiency was routinely >80% across all experiments.

      Glucose Consumption Rate

      NG-HRECs and HG-HRECs were plated into 24-well plates and cultured at full confluency in complete endothelial cell media (Lonza) containing 5 mmol/L (NG) or 30 mmol/L (HG) glucose. Medium was refreshed every 24 hours. The glucose level in the supernatant was measured every 8 hours (0, 8, 16, and 24 hours) using a glucometer with glucose test strips (Contour Next; CVS, Chicago, IL). The glucose values were extrapolated from the standard curve's trendline equation. For the standard curve, Dulbecco’s modified Eagle’s medium (catalog number 103680-100; Agilent, Santa Clara, CA) was used without glucose, supplemented with pyruvate, l-glutamine, and the bullet kit from Lonza to mimic the medium used for culturing HRECs. Multiple dilutions of glucose (0, 10, 20, 30, and 40 mmol/L) were used to plot the standard curve.

      Assessing Cell Death

      LDH

      Lactate dehydrogenase (LDH), a measure of membrane integrity for cell-mediated cytotoxicity, was quantified using colorimetric CytoTox96 nonradioactive cytotoxicity assay (Promega, Madison, WI; G1780). LDH is a stable cytosolic enzyme that is released on cell lysis. The LDH activity that was released in the culture supernatant was measured with a coupled enzymatic assay following the manufacturer's instructions. The OD was determined using a Synergy H1 spectrophotometer (Agilent). The amount of color formed is proportional to the number of lysed cells. For each experimental group, the amount of released LDH was normalized to the total LDH level, which was obtained by lysing cells using the lysis buffer supplied with the kit. To measure basal cell death, LDH was allowed to accumulate in the conditioned medium, which was harvested, and LDH activity was assessed as described above. To determine death in response to oxidative stress, the medium was replaced and then vehicle or tert-butyl hydroperoxide (TBH) (final concentration of 5 mmol/L) was added for 4 hours, whereupon the LDH activity in the medium was quantified.

      Fluorescence-Activated Cell Sorting

      NG-HRECs and HG-HRECs were cultured in a 6-well tissue culture plate. For oxidative stress–induced death, confluent monolayers were treated with vehicle or 5 mmol/L TBH for 4 hours. Cells were collected by trypsinization and stained for annexin V and propidium iodine (Annexin VFITC Assay Kit; item number 600300; Cayman Chemical, Ann Arbor, MI), indicators of apoptosis and necrosis, respectively. Fluorescently stained cells were quantified on a Gallios cell sorting instrument (Beckman Coulter, Indianapolis, IN) with appropriate controls (singly stained and unstained cells). Cell death was defined as the sum of single- and double-positive cells.

      Inflammation Index

      RNA Sequencing

      The goal of this series of experiments was to assess the effect of HG on expression of genes in HRECs. To this end, triplicate dishes of confluent HRECs that were treated with either glucose condition for 1 or 10 days were harvested, and mRNA was isolated (RNeasy isolation kit; catalog number 74106; Qiagen, Hilden, Germany) and submitted to the Bioinformatics Core at University of Illinois at Chicago. Two-way analysis of variance was applied to compare the changes in gene expression between 1-day NG/HG and 10-day NG/HG HRECs. For the 10-day data set, the list of genes with a statistically significant change in expression level was subjected to Gene Ontology analysis to identify inflammatory response genes. Inflammation-related genes were further filtered on the basis of the absolute number of counts, which was set as >1000.

      PCR Analysis

      HRECs cultured in NG or HG for 10 days were harvested, cells were lysed, and mRNA was isolated and used to synthesize cDNA (High-Capacity cDNA Reverse Transcription Kit; catalog number 4368813; Applied Biosystems, Waltham, MA). Quantitative PCR was performed using Fast SYBR Green master mix (catalog number 4385612; Applied Biosystems) and run on the QuantStudio 7 Flex Real-Time PCR system (catalog number 4485701; Applied Biosystems). To calculate the relative amount of mRNA, the threshold cycle (CT) of each transcript was normalized to the average CT of β-actin. Fold change was calculated using the 2-CT method.

      Western Blot Analysis

      Confluent cultures of NG-HRECs and HG-HRECs were treated with tumor necrosis factor (TNF)-α (2.5 ng/mL) for the indicated time periods. After treatment, cells were rinsed with ice-cold phosphate-buffered saline (137 mmol/L NaCl, 2.7 mmol/L KCl, 8 mmol/L Na2HPO4, and 2 mmol/L KH2PO4) and lysed in 2× electrophoresis sample buffer (10 mmol/L EDTA; 2% SDS; 0.2 mol/L 2-mercaptoethanol; 20% glycerol; 200 mmol/L Tris-HCl, pH 6.8; and 0.2% bromophenol blue). Proteins were resolved on SDS-polyacrylamide gel and subjected to Western blot analysis.
      The membrane was blotted for vascular cell adhesion molecule 1 [VCAM1; ab134047; rabbit monoclonal antibody (mAb); 1:1000 dilution; Abcam, Cambridge, UK] and intercellular adhesion molecule 1 (ICAM1; ab109361; rabbit mAb; 1:1000 dilution; Abcam) using primary antibodies. The protein level was normalized to β-actin (8H10D10; mouse mAb; 1:1000 dilution; Cell Signaling, Danvers, MA). Secondary antibodies were used against mouse or rabbit (IRDye 800CW goat anti-rabbit or anti-mouse IgG; 926-32211 or 926-32210, respectively, LI-COR, Lincoln, NE), depending on the primary antibody source, to visualize the band. The blot images were acquired on a LiCor system (LI-COR), and the relative amount of protein was quantified by ImageJ software version 2.0.0-rc-69/1.52v (build 269a0ad53f; NIH, Bethesda, MD; https://imagej.net/software/fiji, last accessed December 4, 2018). To measure the total level of mitochondria, antibodies translocase of outer mitochondrial membrane 20 (TOM20; D8T4N rabbit mAb; number 42406; 1:1000 dilution; Cell Signaling) and cytochrome C oxidase subunit 4I1 (COXIV; 3E11; rabbit mAb; number 4850; 1:1000 dilution; Cell Signaling) were used. Antibodies were validated using Western blot analysis of extracts from various cell lines, immunoprecipitation of the protein of interest, confocal immunofluorescence analysis of mouse tissues, and flow cytometric analysis of cell lines.

      Seahorse Analysis

      HRECs were plated at full confluency onto gelatin-coated XF96 well cell culture microplates (catalog number 102416-100; Agilent) in complete endothelial cell media from Lonza. The optimal cell confluency was determined using an ATP rate assay kit (catalog number 103592-100; Agilent). Before the start of oxygen consumption rate/extracellular acidification rate measurements on the Seahorse XFe96 machine (Agilent), the medium was replaced with Seahorse XF Dulbecco’s modified Eagle’s medium assay medium (catalog number 103680-100; Agilent) supplemented with 1.462 g/L glutamine, 110 mg/L pyruvate, and 5 or 30 mmol/L (for NG or HG, respectively, condition) d-glucose, which is equivalent to the culture conditions using complete Lonza media. Oxygen consumption rate was measured using MitoStress test (catalog number 103015-100; Agilent) following the manufacturer's protocol.

      TBH Challenge Assay

      The mito-roGFP2-Orp1sensor used in the TBH challenge assay was made from pLPCX mito-roGFP2-Orp1, which was a gift from Tobias Dick (Institute of Immunology, Heidelberg University, Heidelberg, Germany) (64992; Addgene).
      • Gutscher M.
      • Sobotta M.C.
      • Wabnitz G.H.
      • Ballikaya S.
      • Meyer A.J.
      • Samstag Y.
      • Dick T.P.
      Proximity-based protein thiol oxidation by H2O2-scavenging peroxidases.
      The mito-roGFP2-ORP1 in the pLPCX plasmid was cut with ClaI, blunted with T4DNA polymerase, and then cut with Bgl2. The resulting 1.5-kb DNA fragment was ligated into the Hpa1/BamH1-cut pLV-EF1a vector. Insert-containing constructs were detected on the basis of diagnostic BamH1/EcoR1 restriction fragments. Because the ClaI site in the pLPCX mito-roGFP2-Orp1 plasmid is blocked by methylation, this construct was propagated in dam-/dcm- bacteria. The mito-roGFP2-Orp1pLV-EF1a plasmid was used to make lentivirus (as described previously
      • Li Y.
      • Baccouche B.
      • Olayinka O.
      • Serikbaeva A.
      • Kazlauskas A.
      The role of the Wnt pathway in VEGF/anti-VEGF-dependent control of the endothelial cell barrier.
      ), which was used to stably express mito-roGFP2-Orp1 in HRECs. The mito-roGFP2-Orp1 colocalized with mitochondrial proteins (data not shown), which is consistent with reports from other groups using this roGFP sensor in other cell types.
      • Dey S.
      • Sidor A.
      • O'Rourke B.
      Compartment-specific control of reactive oxygen species scavenging by antioxidant pathway enzymes.
      ,
      • Waypa G.B.
      • Marks J.D.
      • Guzy R.
      • Mungai P.T.
      • Schriewer J.
      • Dokic D.
      • Schumacker P.T.
      Hypoxia triggers subcellular compartmental redox signaling in vascular smooth muscle cells.
      Cells stably expressing mito-roGFP2-ORP1 were plated onto gelatin-coated black 96-well tissue culture plates. Each experimental condition was plated in triplicate, and three to four areas per well were selected for image acquisition. For time-lapse imaging, the medium was replaced with phenol red–free endothelial cell growth basal medium (CC-3129; Lonza) supplemented with microvascular endothelial SingleQuots kit (EGM-2MV; CC4147; Lonza) and 25 mmol/L HEPES. The 96-well tissue culture plate with a confluent cell monolayer was placed inside the temperature-controlled microscope chamber at 37°C for at least 2 hours before the start of time-lapse imaging. The 20× objective, 405-nm, and 488-nm lasers were used for image acquisition on Zeiss microscope equipped with spinning disk (Zeiss, Jena, Germany). After establishing the basal level of oxidative stress, TBH, vehicle, dithiothreitol, or diamide was added into each of the designated wells. The images were processed on ImageJ software and normalized following a published protocol.
      • Morgan B.
      • Sobotta M.C.
      • Dick T.P.
      Measuring E(GSH) and H2O2 with roGFP2-based redox probes.
      Briefly, images were thresholded manually per each channel, and 405 nm was divided into 488 nm. The data from each experimental condition were normalized [Rnormalized = (0.2/Rred) × value] to the fully reduced state (Rred) induced by dithiothreitol. The effect of TBH on oxidative stress was defined as a fold change over vehicle.

      Measuring Mitophagy

      The mitochondrially localized fluorescent protein Keima (MtKeima) was used to measure mitophagy in HRECs, human glomerular microvascular endothelial cells, and HRPCs. MtKeima has an excitation spectrum that changes according to pH.
      • Sun N.
      • Malide D.
      • Liu J.
      • Rovira I.I.
      • Combs C.A.
      • Finkel T.
      A fluorescence-based imaging method to measure in vitro and in vivo mitophagy using mt-Keima.
      HRECs were transduced with lentivirus carrying pHAGE-mtKeima [gift from Richard Youle (National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD)] (131626; Addgene)
      • Vargas J.N.S.
      • Wang C.
      • Bunker E.
      • Hao L.
      • Maric D.
      • Schiavo G.
      • Randow F.
      • Youle R.J.
      Spatiotemporal control of ULK1 activation by NDP52 and TBK1 during selective autophagy.
      using the procedure described above. HRECs stably expressing MtKeima were plated at full confluency onto gelatin-coated 35-mm glass-bottom dishes (P35G-1.5-14-C; MatTek, Ashland, MA). The next day, cells were transfected with siRNA targeting mitophagy regulators and then subjected to imaging. For live cell imaging on Zeiss confocal microscope, the medium was replaced to phenol red–free endothelial cell growth basal medium (CC-3129; Lonza) supplemented with microvascular endothelial SingleQuots kit (EGM-2MV; CC4147; Lonza) and 25 mmol/L HEPES. The 20× objective and Z-stack with six slices were used for image acquisition; 488 nm was used for green mitochondria and 565 nm for red lysosomes. Because the mitophagic process includes the engulfment of mitochondria by the lysosome, the colocalization between mitochondria and lysosome was determined as mitophagy. At least five randomly selected sections of the dish were imaged for each experimental condition. Images were then converted into maximal intensity projection for further analysis on ImageJ software. Mitophagy was quantified by dividing red pixels to green pixels using Pearson correlation coefficient [Colocalization Threshold Analysis (Rcol)] on ImageJ software.
      HRECs stably expressing mitochondrially localized mito-roGFP2-Grx1
      • Gutscher M.
      • Pauleau A.L.
      • Marty L.
      • Brach T.
      • Wabnitz G.H.
      • Samstag Y.
      • Meyer A.J.
      • Dick T.P.
      Real-time imaging of the intracellular glutathione redox potential.
      were cultured in either NG or HG for at least 10 days and stained for 30 minutes with lysosomal marker LysoTracker-red (L7528; Invitrogen) following manufacturer's protocol. Medium was then replaced to phenol red–free endothelial cell medium supplemented with 25 mmol/L HEPES. The immunofluorescence images were acquired and processed as described above. Colocalization between red lysosomes and green mitochondria was quantified using Colocalization Threshold Analysis (Rcol) on ImageJ software, same as for MtKeima.
      Urolithin A was purchased from Sigma (St. Louis, MO; SML1791).

      siRNA

      Confluent HRECs, plated onto a 6-well tissue culture plate, were transfected with siRNA in an antibiotic-free complete endothelial cell medium (Lonza). To this end, 10 nmol/L of ON-TARGETplus Human siRNA SMARTpool (Horizon Discovery, Waterbeach, UK) targeting mitofusin 2 (MFN2; L-012961-00), dynamin 1 like (DNM1L; L-012092-00), optineurin (OPTN; L-016269-00), OPA1 mitochondrial dynamin like GTPase (OPA1; L-005273-00), or a nontargeting pool of siRNA (Scr; D-001810-10-05) was complexed with DharmaFECT 1 transfection reagent (catalog identifier T-2001; Horizon Discovery) at 1:2 ratio in reduced serum Opti-MEM medium (31985070; Gibco, Grand Island, NY) and added to the medium. The next day, transfected cells were trypsinized and plated into a 96-well tissue culture plate (for TBH-challenge assay or LDH assay) or 8-well chamber slide (for transendothelial electrical resistance assay) in complete endothelial cell media. At 48-hour time point after transfection, transfected cells were subjected to the designated experimental assay. The extent of silencing was determined on both the protein and mRNA level using Western blot analysis and quantitative RT-PCR, respectively, methods described above. The sequences of the primers used for quantitative RT-PCR are listed in Table 1. Antibodies used for Western blot analysis were as follows: MFN2 (D2D10; rabbit mAb; number 9482; Cell Signaling), DNM1L (DRP1; D6C7; rabbit mAb; number 8570; Cell Signaling), OPA1 (D6U6N; rabbit mAb; number 80471; Cell Signaling), and OPTN (E4P8C; rabbit mAb; number 70928; Cell Signaling).
      Table 1Sequence of Primers Used in Quantitative RT-PCR
      Gene namePrimer sequence
      OPA1 FWD5′-GAGGACAGCTTGAGGGTTATTC-3′
      OPA1 REV5′-CTGCAGAGCCTCTTCCATAAA-3′
      DNM1L FWD5′-CGCAGAACCCTAGCTGTAATC-3′
      DNM1L REV5′-CTGGAATAACCCTTCCCATCAA-3′
      OPTN FWD5′-CCTAAGGGAAGGGAATCAGAAG-3′
      OPTN REV5′-CCTCTGTCTGGGTTTCAATCT-3′
      MFN2 FWD5′-GCCCTCCAGCACTACTTATTT-3′
      MFN2 REV5′-CCATAGCTGTCGCTGAAAGT-3′
      ACTB FWD5′-GGATCAGCAAGCAGGAGTATG-3′
      ACTB REV5′-AGAAAGGGTGTAACGCAACTAA-3′
      TFAM FWD5′-GGGAAGGAGGGTTGTGTATTT-3′
      TFAM REV5′-AGGAGTTAGCCAAACGCAATA-3′
      PINK1 FWD5′-GGCTTGGCAAATGGAAGAAC-3′
      PINK1 REV5′-CTCAGTCCAGCCTCATCTACTA-3′
      TOMM20 FWD5′-GGTGAATATGAGAAGGGCGTAG-3′
      TOMM20 REV5′-ACTGGTGGTGGAAGAGTTTG-3′
      FWD, forward; REV, reverse.

      Barrier Function

      Cell permeability was assessed by measuring changes in transendothelial electrical resistance using an electrical cell-substrate impedance sensing ZThera instrument (Applied Biophysics, Troy, NY), which was housed in a standard tissue culture incubator that was maintained at 37°C and 5% CO2. Briefly, transfected cells were plated into an 8-well chamber at full confluency. The following day, the medium was refreshed before the transendothelial electrical resistance measurements. Basal resistance was established within 2 hours, whereupon cells were stimulated with 2 nmol/L vascular endothelial growth factor or vehicle (phosphate-buffered saline) and continued to be monitored for up to 48 hours, as previously described.
      • Li Y.
      • Baccouche B.
      • Olayinka O.
      • Serikbaeva A.
      • Kazlauskas A.
      The role of the Wnt pathway in VEGF/anti-VEGF-dependent control of the endothelial cell barrier.
      ,
      • Li Y.
      • Yan Z.
      • Chaudhry K.
      • Kazlauskas A.
      The renin-angiotensin-aldosterone system (RAAS) is one of the effectors by which vascular endothelial growth factor (VEGF)/anti-VEGF controls the endothelial cell barrier.

      Statistics

      Each experiment included three sample repeats. Statistical significance was assessed by two-tailed t-test with unpaired two-sample equal variance.

      Results

      The Duration of Exposure to Hyperglycemia Influences Susceptibility to Damage Resulting from DM-Related Insults

      The effect of varying the duration of HG on several parameters associated with DR pathogenesis (inflammation and cell death) was considered. These experiments were performed in HRECs cultured in either 5 mmol/L (90 mg/dL) or 30 mmol/L (540 mg/dL) d-glucose. Exposure of endothelial cells to inflammatory cytokines increases expression of ICAM1 and VCAM1, thereby promoting extravasation of immune cells, a key step in inflammation.
      • Liao J.K.
      Linking endothelial dysfunction with endothelial cell activation.
      Stimulation of HRECs with TNF-α increased the expression of ICAM1 and VCAM1, and the extent of expression was comparable in cells treated with HG for either 6 hours or 1 day (Figure 1A and Supplemental Figure S1). Prolonging the exposure to HG reduced the magnitude of TNF-α–induced expression of ICAM1 (10 days) and VCAM1 (3 and 10 days) (Figure 1A). These results indicate that extended exposure of cells to HG attenuated TNF-α–induced activation.
      Figure thumbnail gr1
      Figure 1The duration of exposure to high glucose (HG) influences susceptibility to damage caused by diabetes mellitus–related insults. A: Human retinal endothelial cells (HRECs) incubated in either normal glucose (NG) or HG for the indicated duration were subsequently exposed to vehicle (veh) or tumor necrosis factor (TNF)-α (2.5 ng/mL) for 6 hours. The vascular cell adhesion molecule 1 (VCAM1) and intercellular adhesion molecule 1 (ICAM1) signal was normalized to the β-actin signal. The resulting data are expressed as a percentage of the response in NG cells. B: Top panel: The bar graph indicating basal cell death is from a single representative experiment in which each experimental condition was assayed in triplicate. Bottom panel: HRECs that had been incubated in either NG or HG for the indicated duration were either left resting or treated with vehicle or 5 mmol/L TBH for 4 hours. Oxidative stress–induced cell death is shown via LDH activity. C: Same as B, except death was assessed by fluorescence-activated cell sorting (FACS) analysis of cells stained with annexin V and propidium iodine (PI). Cell death was quantified as the sum (three quadrants in and ) of single-positive (annexin V or PI) and double-positive (annexin V and PI) cells. Data are expressed as the average ± SD (A and C). n = at least 3 independent experiments (A and C). Statistical significance was assessed by two-tailed t-test with unpaired two-sample equal variance. ∗P < 0.05.
      The effect of HG on steady-state expression of inflammation-related genes was investigated. RNA-sequencing analysis indicated that expression of 55 genes underwent a statistically significant change after 10 days of exposure to HG (Supplemental Figure S2). In contrast, the expression of only four of these genes changed after 1 day of HG (Supplemental Figure S2). This indicated that the duration of exposure to HG affected both the steady-state expression of inflammation-related genes and responsiveness to cytokine-induced activation. The observation that prolonging the exposure to HG reduced TNF-α–driven activation was counter to the dogmatic expectation that HG causes progressive damage.
      The following experiments were conducted to test whether the duration of exposure to HG influenced basal- or insult-induced death. As shown in Figure 1B, after initially increasing (after 3 days of HG), basal cell death was no longer elevated after 10 days of HG. Similarly, oxidative stress–induced cell death was higher in cells that had been cultured in HG for 1 day, and then became lower after 10 days of HG (Figure 1B). In these experiments, cell death was assessed by measuring release of LDH. Evaluating cell death by expression of annexin V and permeability to propidium iodine confirmed the key observation that extended exposure to HG made cells resistant to cell death (Figure 1C and Supplemental Figures S3 and S4). These observations resonate with a previous report that HG induced a slight increase in HREC death at 4 and 6 days of treatment, and prolonging the exposure for 4 weeks no longer activated cell death.
      • Busik J.V.
      • Mohr S.
      • Grant M.B.
      Hyperglycemia-induced reactive oxygen species toxicity to endothelial cells is dependent on paracrine mediators.
      Although Figure 1B shows that cell death increased after 3 days of HG, a decline in the total protein (Figure 1A) was not detected. More importantly, LDH data were normalized for cell number by expressing the results as a fraction of the total LDH. Furthermore, in the course of daily maintenance of the cells, the cell monolayer was invariably confluent, and the small number of floating cells did not differ across experimental conditions (data not shown). A plausible explanation for the inability to detect a decline in the overall cell number in case of cell death could be that cells were continuously cultured in 5% serum-containing medium, which supports proliferation. Consequently, cell proliferation likely replaced cells that died, and hence maintained a comparable number of cells across experimental conditions.
      Taken together, these findings indicate that prolonged exposure of HRECs to HG made them resilient to the deleterious effects of HG instead of making them more vulnerable. Such adaptation was most apparent when cells were challenged, and involved resistance to cytokine-induced activation and oxidative stress–driven cell death.

      Adaptation Is Associated with Improved Mitochondrial Functionality

      The study focused on the mitochondria to further characterize adaptation. Existing publications demonstrating that mitochondria adapt to environmental changes
      • Pickles S.
      • Vigie P.
      • Youle R.J.
      Mitophagy and quality control mechanisms in mitochondrial maintenance.
      ,
      • Kluge M.A.
      • Fetterman J.L.
      • Vita J.A.
      Mitochondria and endothelial function.
      were the basis to check whether the duration of exposure to HG influenced mitochondrial functionality. Seahorse analysis revealed that HG was initially detrimental; it reduced basal oxygen consumption (Figure 2A), the spare respiratory capacity (Figure 2B), and ATP production (Figure 2C). All of these parameters improved (instead of further declining) with a prolonged exposure to HG (Figure 2).
      Figure thumbnail gr2
      Figure 2Exposure to high glucose (HG) initially compromises the oxygen consumption rate (OCR), and then improves it. Human retinal endothelial cells incubated in either normal glucose (NG) or HG for the indicated duration were subsequently assayed for oxygen consumption rate, as described in . A: Basal respiration was calculated after subtracting nonmitochondrial respiration. B: Spare respiratory capacity, which indicates the capability of the cell to respond to an energetic demand, was calculated based on the difference between the basal respiration and maximal respiration. C: ATP production in cells cultured in NG and HG. The asterisks indicate statistical significance, which was assessed by two-tailed t-test with unpaired two-sample equal variance. The data shown in this figure are from a single representative experiment in which 10 to 12 wells were used for each experimental condition. Similar results were obtained from three independent experiments. ∗P < 0.05.
      The effect of HG on mitochondrial functionality was investigated using a newly developed approach that evaluated the mitochondria's ability to resolve acute oxidative stress. This technique uses an roGFP sensor, which has the following four features. i) The green fluorescent protein (GFP) portion of the sensor is mutated so that its fluorescence reflects oxidative status. ii) It is tagged with a mitochondrial localization signal. iii) It is fused with the yeast peroxidase ORP1 to impart preference for hydrogen peroxide. iv) It is reversible and, therefore, can be used to observe both an increase, and subsequent decrease, in oxidative stress in live cells and in real time. Figure 3A shows images of HRECs stably expressing the mito-roGFP2-Orp1 sensor, and the oxidation-dependent change in signal intensity. Such roGFP sensors have been characterized extensively and used to evaluate changes in oxidative stress in a variety of experimental conditions.
      • Meyer A.J.
      • Dick T.P.
      Fluorescent protein-based redox probes.
      • Roma L.P.
      • Deponte M.
      • Riemer J.
      • Morgan B.
      Mechanisms and applications of redox-sensitive green fluorescent protein-based hydrogen peroxide probes.
      • Albrecht S.C.
      • Barata A.G.
      • Grosshans J.
      • Teleman A.A.
      • Dick T.P.
      In vivo mapping of hydrogen peroxide and oxidized glutathione reveals chemical and regional specificity of redox homeostasis.
      Figure thumbnail gr3
      Figure 3High glucose (HG)–induced adaptation improves mitochondrial functionality. A: Mito-roGFP2-Orp1 human retinal endothelial cells (HRECs) stably expressing mito-roGFP2-Orp1, and emitting a fluorescent signal when exposed to light at the 488- and 405-nm wavelengths. An increase in oxidative stress (addition of H2O2) altered the fluorescent intensity in both channels; arrows to the right indicate the direction of the change. The level of mitochondrial oxidative stress in mitochondria was calculated as the ratio of 405 nm/488 nm. B: Normal glucose (NG) HRECs stably expressing mito-roGFP2-Orp1 challenged with an escalating dose of TBH. The 405 nm/488 nm ratio was normalized to Rred and expressed as a fold change over vehicle. C: Mito-roGFP2-Orp1–expressing HRECs cultured in NG or HG with vehicle or TBH (270 μmol/L) added (arrow). The resulting data were normalized to Rred and expressed as a fold change over vehicle. The asterisks indicate time points at which differences between NG and HG were statistically significant, which was calculated by two-tailed t-test with unpaired two-sample equal variance. Data show the average of triplicates ± SD of a single representative experiment (B) or the average ± SD for four independent experiments (C). ∗P < 0.05. Scale bar = 20 μm (A).
      The mito-roGFP2-Orp1 sensor was stably expressed in HRECs, and the resulting cells were exposed to TBH. The change (increase and resolution) in mitochondrial oxidative stress was monitored for at least 5 hours. TBH caused an acute and transient increase in mitochondrial hydrogen peroxide; the resolution took longer at higher doses of TBH (Figure 3B). These experimental conditions (dose and duration of exposure to TBH) had no effect on cell viability (data not shown). This TBH challenge assay was used to assess the capacity of the mitochondria to resolve oxidative stress. Unless indicated otherwise, the assay was performed using 270 μmol/L TBH.
      The TBH challenge assay was used to determine the effect of different durations of exposure to HG on mitochondrial functionality. The performance in the TBH challenge assay was unaffected by exposure to HG for 0.5 days; the response curves of cells cultured in NG and HG overlapped (Figure 3C). After 1 day of HG, the TBH-induced increase in oxidative stress was higher and persisted longer in HG versus NG cells (Figure 3C). Further increasing the duration of the exposure to HG reversed this effect. After 10 days, the response of HG cells improved relative to NG cells; TBH induced a smaller increase and faster resolution of oxidative stress (Figure 3C). The TBH challenge assay results confirm the conclusions from the Seahorse analysis that exposure to HG initially compromised mitochondrial functionality, which improved (instead of declining further) as the duration of exposure was prolonged. These mechanistic insights further support the novel concept that HRECs adapt to HG in ways that improve their ability to withstand its deleterious effect.
      Mitochondrial oxidative stress did not increase even after 10 days of exposure to HG; the basal level of H2O2 (before addition of TBH) was comparable in NG and HG cells (Figure 3C). Other reactive oxygen species, measured with additional roGFPs or fluorescent oxidation-reduction–sensitive dyes, were also unchanged by HG (data not shown). The mitochondria's resistance to HG-induced oxidative stress appears to be an additional feature of adaptation.

      The Mechanism of Hyperglycemia-Induced Mitochondrial Adaptation Involved Mitochondrial Dynamics

      Prolonged Exposure to HG Elevated Mitophagy

      Mitophagy, a key component of mitochondrial dynamics, is one of the ways that mitochondria are able to adapt to environmental changes.
      • Pickles S.
      • Vigie P.
      • Youle R.J.
      Mitophagy and quality control mechanisms in mitochondrial maintenance.
      Mitophagy was investigated to study the role of prolonged exposure to HG in mitochondrial functionality. The fluorescent mitophagy sensor (MtKeima) was stably expressed in NG or HG HRECs. In this experimental system, mitochondria at neutral pH are green, whereas mitochondria undergoing mitophagy are red because they are exposed to acidic pH when they fuse with the lysosome.
      • Sun N.
      • Malide D.
      • Liu J.
      • Rovira I.I.
      • Combs C.A.
      • Finkel T.
      A fluorescence-based imaging method to measure in vitro and in vivo mitophagy using mt-Keima.
      Mitophagy increased following exposure of NG cells to carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP) (Figure 4, A and B ), which compromises the mitochondrial membrane potential and thereby increases mitophagy.
      • Heytler P.G.
      • Prichard W.W.
      A new class of uncoupling agents--carbonyl cyanide phenylhydrazones.
      Furthermore, mitophagy was greater in HG cells compared with their NG counterparts (Figure 4, A and B). A lysotracker-based approach confirmed that mitophagy was higher in HG versus NG cells (Supplemental Figure S5). Time-course experiments indicated that mitophagy increased after ≥7 days of exposure to HG (Supplemental Figure S6).
      Figure thumbnail gr4
      Figure 4Mitophagy is elevated in adapted cells. A: Representative confocal images showing the overlay of red and green fluorescence in MtKeima-expressing human retinal endothelial cells (HRECs) cultured in normal glucose (NG) or high glucose (HG). NG cells treated with FCCP (10 μmol/L), a mitochondrial membrane protonophore, were used as a positive control. The red signals (white arrows) point to mitochondria within an acidic compartment, such as the lysosome. B: The extent of mitophagy in MtKeima-expressing HRECs subjected to the indicated experimental conditions was quantified. The data presented are from at least three independent experiments, which were normalized to the level of mitophagy in NG-HRECs. C: PTEN induced kinase 1 (PINK1), transcription factor A, mitochondrial (TFAM), and translocase of outer mitochondrial membrane 20 (TOMM20) mRNA levels in HRECs cultured in NG or HG. D: Cytochrome C oxidase subunit 4I1 (COX4) and TOMM20 protein levels in HRECs treated with NG or HG. The signal was quantified using ImageJ software (version 2.0.0-rc-69/1.52v; build 269a0ad53f); COX4 and TOMM20 were normalized to β-actin. Each experiment included triplicate samples. Data indicate the average ± SD of triplicates of a single representative experiment (C) or the average ± SD of three independent experiments (D). Statistical significance was calculated by two-tailed t-test with unpaired two-sample equal variance (B). n = 3 (C). ∗P < 0.05. Scale bar = 20 μm (A).
      The effect of HG on biogenesis and total mitochondrial content was similarly investigated. HG increased expression of genes that are involved with both mitophagy [PTEN induced kinase 1 (PINK1)] and biogenesis [transcription factor A, mitochondrial (TFAM) and PPARG coactivator 1 alpha (PGC1α)], without changing the total mitochondrial content (Figure 4, C and D; data not shown). Together, these data indicate that prolonged exposure to HG resulted in increased mitophagy and an accompanying increase in mitochondrial biogenesis.
      The morphology of the mitochondrion shown in Figure 4 (elongated in NG versus punctate in HG) also supports the conclusion that HG increased mitophagy. However, although this was apparent for the cells shown in Figure 4A, the high degree of heterogeneity of the mitochondrial morphology among cells in either glucose condition precluded use of mitochondrial morphology as a parameter to assess mitophagy (Supplemental Figure S7). In contrast, the quantitative approaches measuring the degree of mitophagy of the entire population of cells showed that HG elevated mitophagy (Figure 4 and Supplemental Figures S5 and S6).

      Mitochondrial Dynamics Are Involved in the Maintenance of HIMA

      To assess the role of mitochondrial dynamics in the maintenance of hyperglycemia-induced mitochondrial adaptation (HIMA), expression of four genes involved in various aspects of this quality control system: fission (DNM1L), fusion (OPA1 and MFN2), and mitophagy (OPTN and MFN2) was suppressed. HIMA was compromised (reduced performance in the TBH challenge assay) by suppression of any one of these four genes (Table 2, Figure 5, and Supplemental Figure S8, A–D). Furthermore, reducing expression of fission genes suppressed mitophagy (Supplemental Figure S8), which demonstrates the interrelatedness of the various component of mitochondrial dynamics. The effect on HIMA was greatest in MFN2 knockdown cells (Figure 5 and Table 2), perhaps because MFN2 is required for multiple components of mitochondrial dynamics.
      • Chan D.C.
      Mitochondrial dynamics and its involvement in disease.
      These data indicate that maintenance of HIMA requires enhanced mitochondrial dynamics.
      Table 2Importance of Mitochondrial Dynamics for Maintenance of HIMA
      VariableDNM1LOPA1OPTNMFN2
      Role in mitochondrial dynamicsFissionFusionMitophagyMitophagy; fusion
      Extent of siRNA-mediated knockdown (mRNA/protein), %>80>80>80>80
      Compromised performance in the TBH challenge assayY/N/Y/YY/N/Y/YY/Y/YY/Y/Y
      Suppression of mitophagyY/Y/YY/N/Y/YY/N/Y/NY/Y/Y
      Each Y or N designation indicates the response of an independent experiment.
      HIMA, hyperglycemia-induced mitochondrial adaptation; N, no; TBH, tert-butyl hydroperoxide; Y, yes.
      Figure thumbnail gr5
      Figure 5Mitophagy is required for maintenance of hyperglycemia-induced mitochondrial adaptation. A: MFN2 mRNA from human retinal endothelial cells (HRECs) cultured in high glucose (HG) mock transfected [control (Cntr)] or transfected with either scrambled siRNA (siScr) or MFN2-targeting siRNA (siMFN2) was normalized to β-actin, and expressed as a fold change relative to control. The P value indicates a statistically significant difference between the two groups. B: Western blot analysis of the cells described in A using the indicated antibodies. C: Mitophagy in MtKeima-expressing HRECs cultured in HG for 10 days. The P value was determined using t-test. Similar results were obtained on at least three independent occasions. D: Cells that had been treated as described in A were subjected to the lactate dehydrogenase activity assay. The P value was determined using t-test. Similar results were obtained on at least three independent occasions. E: Mito-roGFP2-Orp1–expressing HRECs were cultured in HG, transfected, and subjected to the TBH challenge assay. The experiment was repeated at least three times, and the statistical significance (P < 0.05) for a single time point is indicated by asterisks. Data indicate the average ± SD of triplicates of a single representative experiment (C and D). n = 3 (C and D). ∗P < 0.05.

      HG-Induced Adaptation Is Beneficial

      The study further investigated whether loss of HIMA would compromise the cells' ability to endure HG. To this end, MFN2 expression (Figure 5) was suppressed and the impact on susceptibility to insult-induced death and functionality of the cells was assessed. Loss of HIMA increased vulnerability to both basal- (Figure 5D) and oxidative stress–induced cell death (Figure 6A). Furthermore, barrier function was compromised (ie, increased the permeability of confluent monolayers) (Figure 6B). Although the increase in basal cell death may contribute to enhanced permeability of MFN2 knockdown cells, reducing the level of DNM1L reduced barrier function without influencing cell death (Supplemental Figure S8). Finally, MFN2 knockdown cells lost their ability to relax the barrier further in response to vascular endothelial growth factor (Figure 6B). Taken together, these results indicate that HRECs adapted to HG and thereby acquired protection from HG-induced harm.
      Figure thumbnail gr6
      Figure 6Loss of hyperglycemia-induced mitochondrial adaptation compromises the viability and functionality of high glucose (HG) human retinal endothelial cells (HRECs). A: HRECs cultured in HG were mock transfected [control (Cntr)], or transfected with either scrambled siRNA (siScr) or MFN2-targeting siRNA (siMFN2). Cells were treated with vehicle (Veh) or TBH and subjected to the lactate dehydrogenase (LDH) activity assay. The P value indicates a statistically significant difference between the two groups. B: Representative Western blot of HRECs cultured in HG, transfected with either siScr or siMFN2, and treated with tumor necrosis factor (TNF)-α . The experiment was repeated four times on independent occasions. C: The transendothelial electrical resistance (TEER) of cells treated as described in A was assessed (as described in ) under either unstimulated (left panel) or vascular endothelial growth factor (VEGF)–stimulated conditions (right panel). The experiment was repeated at least three times; representative TEER tracings are shown. ∗P < 0.05. ICAM1, intercellular adhesion molecule 1; VCAM1, vascular cell adhesion molecule 1.
      The recent discovery that MFN2 contributes to barrier function
      • Kim Y.M.
      • Krantz S.
      • Jambusaria A.
      • Toth P.T.
      • Moon H.G.
      • Gunarathna I.
      • Park G.Y.
      • Rehman J.
      Mitofusin-2 stabilizes adherens junctions and suppresses endothelial inflammation via modulation of beta-catenin signaling.
      raises the possibility that the enhanced permeability of MFN2 knockdown cells is caused by factors other than dysfunctional mitochondrial dynamics. However, barrier function was also compromised when DNM1L (required for fission) expression was suppressed (Supplemental Figure S8E).
      Although suppressing mitochondrial dynamics eliminated certain features of HIMA, others persisted. Cells remained resistant to TNF-α–induced activation, which was monitored by ICAM1 and VCAM1 expression, as in Figure 1 (Figure 6B). These findings suggest that although elevated mitochondrial dynamics are an essential component of HIMA, they are not the only one.
      In an attempt to eliminate HIMA in HRECs by an alternative approach, the exposure to HG was prolonged for up to 15 days, which approached the passage limit of these primary cells. Prolonged exposure to HG did not eliminate HIMA (data not shown), suggesting that this is a relatively stable state.

      Urolithin A Promotes Partial Adaptation of NG Cells

      The study tested whether increasing mitophagy in NG cells would be sufficient to induce adaptation. Exposure of NG cells to urolithin A, an agent that enhances mitophagy and improves mitochondrial health,
      • Andreux P.A.
      • Blanco-Bose W.
      • Ryu D.
      • Burdet F.
      • Ibberson M.
      • Aebischer P.
      • Auwerx J.
      • Singh A.
      • Rinsch C.
      The mitophagy activator urolithin A is safe and induces a molecular signature of improved mitochondrial and cellular health in humans.
      increased mitophagy and performance in the TBH challenge assay (Figure 7, A and B ). However, it did not attenuate TNF-α–induced activation or reduce basal- or insult-induced death (Figure 7, C and D, and Supplemental Figure S9). Incomplete adaptation was also observed with other mitophagy-promoting agents (Torin2, mammalian target of rapamycin inhibitor; and SR3677, Rho kinase inhibitor; data not shown). These data indicate that although enhanced mitophagy is essential for cells to maintain key features of HIMA, it is not sufficient to acquire it.
      Figure thumbnail gr7
      Figure 7Urolithin A (UA) induces partial hyperglycemia-induced mitochondrial adaptation. A: Mitophagy in MtKeima-expressing normal glucose (NG) human retinal endothelial cells (HRECs) treated with vehicle or UA. The experiment was repeated at least three times. The statistical significance (P < 0.05) for a single time point is indicated by asterisks. B: Representative results from TBH assay in mito-roGFP2-Orp1–expressing NG HRECs treated with vehicle or UA. Similar results were observed in at least three independent experiments. The statistical significance for a single time point is indicated with an asterisk. C: Western blot of NG HRECs treated with vehicle or UA and stimulated with tumor necrosis factor (TNF)-α. The bar graph is the compilation of three independent experiments. D: Basal- and TBH-induced cell death in NG HRECs treated with vehicle or UA. Similar results were obtained on at least three independent occasions. Data are expressed as the average ± SD for three independent experiments, normalized to control (Cntr; A), or the average ± SD of triplicates of a single representative experiment (D). ∗P < 0.05. ICAM1, intercellular adhesion molecule 1; VCAM1, vascular cell adhesion molecule 1.

      The Effect of Osmolality

      The effect of osmolality was tested by comparing the behavior of cells cultured for 10 days in 5 mmol/L d-glucose, 30 mmol/L d-glucose, or 5 mmol/L d-glucose (present in the culture medium) + 25 mmol/L l-glucose (LG) (Supplemental Figure S10). The results show that 10 days of treatment with LG did not induce basal cell death, but promoted resistance to oxidative stress–induced cell death as in HG. The resistance to TBH-induced death was not due to increased mitophagy; moreover, LG had no effect on mitochondrial respiratory capacity.

      The Response of Other Cell Types to Hyperglycemia

      Muller and human retinal pigment epithelium cells respond to HG by up-regulating the level of oxidative stress and cytokine production.
      • Busik J.V.
      • Mohr S.
      • Grant M.B.
      Hyperglycemia-induced reactive oxygen species toxicity to endothelial cells is dependent on paracrine mediators.
      To investigate whether pericytes, the other cell type in retinal capillaries, also underwent adaptation, they were subjected to a similar series of experiments as with HRECs. Prolonged exposure of primary HRPCs to HG resulted in resistance to HG-induced death (Figure 8A) but not to oxidative stress–induced death (Figure 8B). Furthermore, the underlying mechanism by which HRPCs adapted was not the same as for HRECs; there was no improvement in the TBH challenge assay, and mitophagy was not increased (Figure 8, C and D). This indicates that both vascular cell types of retinal capillaries adapt to HG, but the degree of protection and process by which they achieve this state are not the same.
      Figure thumbnail gr8
      Figure 8Hyperglycemia-induced mitochondrial adaptation does not occur in all cell types. A and E: Primary human retinal pericyte cells (HRPCs) and human glomerular microvascular endothelial cells (HGECs) were cultured in either normal glucose (NG) or high glucose (HG) for the indicated duration and then cell death was quantified as described in . B and F: Same as described in the legend of . C and G: mito-roGFP2-Orp–expressing HRPCs and HGECs were cultured in either NG or HG for 10 days and then subjected to the TBH challenge assay, as described in . No statistically significant differences were observed between NG and HG cells in three independent experiments. D and H: MtKeima-expressing HRPCs and HGECs were cultured in either NG or HG for 10 days and then the level of mitophagy was quantified, as described in . No statistically significant differences were observed between NG and HG cells in three independent experiments. Data are expressed as the average ± SD (A and E). n = 3 (A and E). ∗P < 0.05. FACS, fluorescence-activated cell sorting.
      In contrast to HRECs, primary human glomerular microvascular endothelial cells did not adapt to HG (Figure 8, E–H). HG promoted death of unchallenged cells and did not protect them from oxidative stress–induced death. Furthermore, there was no improvement of performance in the TBH challenge assay or change in the level of mitophagy. Thus, HIMA is not a universal feature of endothelial cells, even among those that reside in organs that undergo DM-induced dysfunction after a considerable delay.

      Discussion

      Prolonged exposure of primary human retinal vascular cells to HG decreased (instead of increased) their vulnerability to the deleterious effects of HG. The underlying mechanism in endothelial cells involves enhanced mitochondrial dynamics and mitophagy, although elevating mitophagy in normal glucose cells was insufficient to induce this state. Although HRPCs also undergo adaptation in response to HG, the degree of protection from death was attenuated and the underlying mechanism appears to be distinct from HIMA.
      Preconditioning, a brief exposure to a sub-threshold level of injury, protects organs, such as the brain and heart, from subsequent insult. The underlying mechanism of this phenomenon involves changes in the mitochondria, which include elimination of dysfunctional mitochondrial (by mitophagy).
      • Correia S.C.
      • Santos R.X.
      • Perry G.
      • Zhu X.
      • Moreira P.I.
      • Smith M.A.
      Mitochondria: the missing link between preconditioning and neuroprotection.
      ,
      • Gottlieb R.A.
      • Gustafsson A.B.
      Mitochondrial turnover in the heart.
      Recent studies have elucidated the molecular mediators and relevance to additional organs. For instance, the exercise capacity of skeletal muscle is dependent on autophagy to clear dysfunctional mitochondria. These events are driven by the protein encoded by the DDIT4 gene/thioredoxin interacting protein (REDD1/TXNIP) complex, which increases oxidative stress and thereby promotes autophagy of mitochondria.
      • Qiao S.
      • Dennis M.
      • Song X.
      • Vadysirisack D.D.
      • Salunke D.
      • Nash Z.
      • Yang Z.
      • Liesa M.
      • Yoshioka J.
      • Matsuzawa S.
      • Shirihai O.S.
      • Lee R.T.
      • Reed J.C.
      • Ellisen L.W.
      A REDD1/TXNIP pro-oxidant complex regulates ATG4B activity to control stress-induced autophagy and sustain exercise capacity.
      Similarly, ischemic post-conditioning of the retina improves the retina's neural function and is dependent on autophagy, which also strongly suggests a mitochondrial involvement.
      • Mathew B.
      • Chennakesavalu M.
      • Sharma M.
      • Torres L.A.
      • Stelman C.R.
      • Tran S.
      • Patel R.
      • Burg N.
      • Salkovski M.
      • Kadzielawa K.
      • Seiler F.
      • Aldrich L.N.
      • Roth S.
      Autophagy and post-ischemic conditioning in retinal ischemia.
      Together, these findings indicate that mitochondria are able to adapt to environmental conditions in ways that protect from pathology.
      In contrast to the extensive literature focused on preconditioning in cardiovascular disease, there are only a handful of publications investigating DM/HG and the concept of preconditioning. Patients with type I DM are protected from ischemia-induced injury of skeletal muscle.
      • Engbersen R.
      • Riksen N.P.
      • Mol M.J.
      • Bravenboer B.
      • Boerman O.C.
      • Meijer P.
      • Oyen W.J.
      • Tack C.
      • Rongen G.A.
      • Smits P.
      Improved resistance to ischemia and reperfusion, but impaired protection by ischemic preconditioning in patients with type 1 diabetes mellitus: a pilot study.
      Similarly, 1 week of DM protected mice from ischemia/reperfusion-induced heart injury compared with non-DM mice.
      • Ravingerova T.
      • Adameova A.
      • Matejikova J.
      • Kelly T.
      • Nemcekova M.
      • Kucharska J.
      • Pechanova O.
      • Lazou A.
      Subcellular mechanisms of adaptation in the diabetic myocardium: relevance to ischemic preconditioning in the nondiseased heart.
      The underlying mechanism of this phenomenon has not been addressed. Compared with ischemia- or hypoxia-induced preconditioning, HG-induced preconditioning is a largely unexplored area of research.
      There is compelling evidence that mitochondrial dysfunction is associated with pathogenesis of DR. Single-nucleotide polymorphisms in genes that regulate mitochondrial functionality (UPC1 and UPC2) are associated with diabetic retinopathy in patients.
      • Brondani L.A.
      • de Souza B.M.
      • Duarte G.C.
      • Kliemann L.M.
      • Esteves J.F.
      • Marcon A.S.
      • Gross J.L.
      • Canani L.H.
      • Crispim D.
      The UCP1 -3826A/G polymorphism is associated with diabetic retinopathy and increased UCP1 and MnSOD2 gene expression in human retina.
      ,
      • Crispim D.
      • Fagundes N.J.
      • dos Santos K.G.
      • Rheinheimer J.
      • Boucas A.P.
      • de Souza B.M.
      • Macedo G.S.
      • Leiria L.B.
      • Gross J.L.
      • Canani L.H.
      Polymorphisms of the UCP2 gene are associated with proliferative diabetic retinopathy in patients with diabetes mellitus.
      In early DR, the mitochondrial content declines in the retina of mice and human donor eyes. As the disease progresses, mitophagy is impaired and dysfunctional mitochondria accumulate.
      • Zhao L.
      • Zhang C.L.
      • He L.
      • Chen Q.
      • Liu L.
      • Kang L.
      • Liu J.
      • Luo J.Y.
      • Gou L.
      • Qu D.
      • Song W.
      • Lau C.W.
      • Ko H.
      • Mok V.C.T.
      • Tian X.Y.
      • Wang L.
      • Huang Y.
      Restoration of autophagic flux improves endothelial function in diabetes through lowering mitochondrial ROS-mediated eNOS monomerization.
      ,
      • Hombrebueno J.R.
      • Cairns L.
      • Dutton L.R.
      • Lyons T.J.
      • Brazil D.P.
      • Moynagh P.
      • Curtis T.M.
      • Xu H.
      Uncoupled turnover disrupts mitochondrial quality control in diabetic retinopathy.
      Similarly, damaged mitochondria accumulate as the homeostasis between biogenesis and turnover of mitochondria is compromised by aging.
      • Palikaras K.
      • Lionaki E.
      • Tavernarakis N.
      Coordination of mitophagy and mitochondrial biogenesis during ageing in C. elegans.
      Furthermore, mitochondrial disintegration/fragmentation activates NLR family pyrin domain containing 3 (NLRP3)–mediated production of inflammatory cytokines and necroptosis.
      • Rocha M.
      • Apostolova N.
      • Diaz-Rua R.
      • Muntane J.
      • Victor V.M.
      Mitochondria and T2D: role of autophagy, ER stress, and inflammasome.
      Finally, mitochondrial dysfunction-driven death of vascular cells is likely to compromise the vasculature's barrier function, triggering a self-perpetuating cycle of leakage and inflammation. Attempts to prevent diabetic retinopathy with mitochondria-targeted therapy are beneficial in some, but not all, cases.
      • Hombrebueno J.R.
      • Cairns L.
      • Dutton L.R.
      • Lyons T.J.
      • Brazil D.P.
      • Moynagh P.
      • Curtis T.M.
      • Xu H.
      Uncoupled turnover disrupts mitochondrial quality control in diabetic retinopathy.
      ,
      • Zhou P.
      • Xie W.
      • Meng X.
      • Zhai Y.
      • Dong X.
      • Zhang X.
      • Sun G.
      • Sun X.
      Notoginsenoside R1 ameliorates diabetic retinopathy through PINK1-dependent activation of mitophagy.
      • Hinder L.M.
      • Sas K.M.
      • O'Brien P.D.
      • Backus C.
      • Kayampilly P.
      • Hayes J.M.
      • Lin C.M.
      • Zhang H.
      • Shanmugam S.
      • Rumora A.E.
      • Abcouwer S.F.
      • Brosius 3rd, F.C.
      • Pennathur S.
      • Feldman E.L.
      Mitochondrial uncoupling has no effect on microvascular complications in type 2 diabetes.
      • Osorio-Paz I.
      • Uribe-Carvajal S.
      • Salceda R.
      In the early stages of diabetes, rat retinal mitochondria undergo mild uncoupling due to UCP2 activity.
      Thus, DM-induced damage of the mitochondria within the retina, which triggers a self-amplifying loop of dysfunction, appears to be a driver of DR pathogenesis.
      The compelling role of mitochondrial dysfunction in the development of DR supports this report's central concepts that cells harbor a system to protect from DM-induced damage, and that this system's purpose is to preserve the functionality of the mitochondria (Figure 9). Exposure to HG initially compromised the health of the mitochondria, which is detectable by decreased performance in the TBH challenge assay (Figure 3C). A subsequent increase in mitochondrial dynamics is at least one of the drivers of improved mitochondrial health observed after 10 days of HG. Finally, we posit that loss of HIMA sets the stage for advancing to DR (Figure 9).
      Figure thumbnail gr9
      Figure 9Overview of the discoveries and their potential clinical relevance. Culturing human retinal endothelial cells in high glucose (HG) initially compromised their mitochondria. The cells respond by adapting, which includes clearance of the dysfunctional mitochondria via mitophagy. Such adaptation is a plausible contributor to the underlying mechanism responsible for the long delay between the onset of diabetes and manifestation of diabetic retinopathy. Furthermore, loss of adaptation may be a prerequisite for development of retinopathy in patients with diabetes.
      Although the discovery of HIMA is an important first step, the work described herein does not provide a comprehensive description of this process. Open questions include what is the trigger(s) that initiates and commits cells to HIMA? It may be a multistep process instead of a single event that occurs at a precise time point; we observe distinct changes at different times (attenuation of TNF-α–induced expression of VCAM1 at 3 days and elevation of mitophagy at 7 days). Elucidating the full spectrum of HIMA-associated changes and their functional relationships will enable progress in this arena.
      In addition to HRECs, it appears that the prolonged exposure of primary human retinal pericytes to HG also increases their ability to resist HG-induced death, although by a different mechanism. In contrast, primary human glomerular endothelial cells did not adapt under our experimental conditions. We chose this additional endothelial cell type because like DR, diabetic nephropathy does not develop quickly following the onset of DM.
      • Andersen A.R.
      • Christiansen J.S.
      • Andersen J.K.
      • Kreiner S.
      • Deckert T.
      Diabetic nephropathy in type 1 (insulin-dependent) diabetes: an epidemiological study.
      ,
      • Krolewski A.S.
      • Warram J.H.
      • Christlieb A.R.
      • Busick E.J.
      • Kahn C.R.
      The changing natural history of nephropathy in type I diabetes.
      If an endogenous system to protect the kidney from DM-driven dysfunction exists, then it may not reside within the glomerular endothelium.
      An inherent component of the HIMA concept is that improving the functionality of a subset of retinal cells will be beneficial for the entire retina. Previous publications report that even a small reduction in the degree/type of insult that the retina experiences is sufficient to protect from developing retinopathy in animals that have DM. Suppressing the inflammatory process within myeloid cells,
      • Portillo J.A.
      • Greene J.A.
      • Okenka G.
      • Miao Y.
      • Sheibani N.
      • Kern T.S.
      • Subauste C.S.
      CD40 promotes the development of early diabetic retinopathy in mice.
      • Portillo J.A.
      • Schwartz I.
      • Zarini S.
      • Bapputty R.
      • Kern T.S.
      • Gubitosi-Klug R.A.
      • Murphy R.C.
      • Subauste M.C.
      • Subauste C.S.
      Proinflammatory responses induced by CD40 in retinal endothelial and Muller cells are inhibited by blocking CD40-Traf2,3 or CD40-Traf6 signaling.
      • Portillo J.C.
      • Lopez Corcino Y.
      • Miao Y.
      • Tang J.
      • Sheibani N.
      • Kern T.S.
      • Dubyak G.R.
      • Subauste C.S.
      CD40 in retinal Muller cells induces P2X7-dependent cytokine expression in macrophages/microglia in diabetic mice and development of early experimental diabetic retinopathy.
      altering phototransduction or visual cycle within photoreceptor cells/retinal pigment epithelial cells,
      • Liu H.
      • Tang J.
      • Du Y.
      • Lee C.A.
      • Golczak M.
      • Muthusamy A.
      • Antonetti D.A.
      • Veenstra A.A.
      • Amengual J.
      • von Lintig J.
      • Palczewski K.
      • Kern T.S.
      Retinylamine benefits early diabetic retinopathy in mice.
      ,
      • Liu H.
      • Tang J.
      • Du Y.
      • Saadane A.
      • Samuels I.
      • Veenstra A.
      • Kiser J.Z.
      • Palczewski K.
      • Kern T.S.
      Transducin1, phototransduction and the development of early diabetic retinopathy.
      or overexpressing metallothionein in endothelial cells
      • Wang K.
      • Dai X.
      • He J.
      • Yan X.
      • Yang C.
      • Fan X.
      • Sun S.
      • Chen J.
      • Xu J.
      • Deng Z.
      • Fan J.
      • Yuan X.
      • Liu H.
      • Carlson E.C.
      • Shen F.
      • Wintergerst K.A.
      • Conklin D.J.
      • Epstein P.N.
      • Lu C.
      • Tan Y.
      Endothelial overexpression of metallothionein prevents diabetes-induced impairment in ischemia angiogenesis through preservation of HIF-1alpha/SDF-1/VEGF signaling in endothelial progenitor cells.
      attenuates or completely prevents animals with DM from succumbing to DR. Together, these discoveries suggest that manifestation of DR involves a relatively small shift in the balance between endogenous systems that prevent DM-driven damage and drivers of pathogenesis.
      Does HIMA exist in vivo, does it protect patients from DR, and is its demise a prerequisite for progression to DR? Our ongoing research is focused on addressing these open questions.

      Acknowledgments

      We thank Tobias Dick for providing pLPCX mito Grx1-roGFP2 (64977; Addgene, Watertown, MA) and pLPCX mito roGFP2-Orp1 (64992; Addgene); Richard Youle for providing pHAGE-mt-mKeima (131626; Addgene); Sara Pastor Puente for running flow cytometry; Lauren Kalinoski for generating the model diagram; and Vrathasha Vrathasha, Monica Lee, and Jalees Rehman for critical evaluation of the manuscript.

      Supplemental Data

      • Supplemental Figure S1

        Response to tumor necrosis factor (TNF)-α after 6 hours exposure to either normal glucose (NG) or high glucose (HG). Same as Figure 1A, except cells were cultured in NG or HG for 6 hours. ICAM1, intercellular adhesion molecule 1; norm, normalized; VCAM1, vascular cell adhesion molecule 1.

      • Supplemental Figure S2

        The effect of high glucose (HG) on expression of inflammation-related genes. Triplicate dishes of human retinal endothelial cells were cultured in normal glucose (NG) or HG for either 1 or 10 days and then lysed; RNA was extracted and subjected to RNA-sequencing analysis, as described in Materials and Methods. LogFC indicates the fold change in expression level between the NG and HG groups. Statistical significance was calculated with one-way analysis of variance (Q value < 0.05). Gene names were selected on the basis of Gene Ontology analysis.

      • Supplemental Figure S3

        Raw fluorescence-activated cell sorting (FACS) data for the experiments shown in Figure 1; 1-day exposure to normal glucose (NG) or high glucose (HG). Human retinal endothelial cells were treated with HG for 1 day, stained with annexin V–fluorescein isothiocyanate (FITC)/propidium iodine (PI), and subjected to flow cytometry analysis, as described in Materials and Methods. The basal group was cultured in media containing the indicated concentration of glucose for 24 hours before the start of the experiment. In the vehicle (veh) and TBH groups, the medium was refreshed, vehicle or TBH was added for 4 hours, and then the cells were stained with annexin V–FITC/PI and subjected to flow cytometry analysis. The scatterplots are the raw data; the bar graph shows the sum of three quadrants: D-+, D+-, and D++.

      • Supplemental Figure S4

        Raw fluorescence-activated cell sorting (FACS) data for the experiments shown in Figure 1; 10-day exposure to normal glucose (NG) or high glucose (HG). Same as Supplemental Figures S3, except human retinal endothelial cells were treated with HG for 10 days instead of 1 day. FITC, fluorescein isothiocyanate; PI, propidium iodine.

      • Supplemental Figure S5

        High glucose (HG) promotes mitophagy; lysotracker-based approach. Mytophagy in human retinal endothelial cells (HRECs) stably expressing a fluorescent mitochondrial marker (mito-GRX1-roGFP2) cultured in either normal glucose (NG) or HG. The bar graph on the left shows the level of mitophagy in NG or HG (10 days) HRECs quantified using ImageJ software (version 2.0.0-rc-69/1.52v; build 269a0ad53f). The images on the right indicate colocalization between mitochondria (green) and lysosome (red). The green and red signals within a given field were overlayed, and Pearson correlation coefficient [Colocalization Threshold Analysis (Rcol)] was quantified. Five random areas per well were imaged. The data that are presented were normalized to NG-HRECs. Statistical significance was calculated using the t-test. Data are expressed as the average ± SD of Rcol for three independent experiments. n = 3 NG or HG (10 days) HRECs. ∗P < 0.05. Scale bar = 20 μm.

      • Supplemental Figure S6

        The duration of exposure to high glucose (HG) that elevates mitophagy. MtKeima-expressing human retinal endothelial cells (HRECs) were cultured in either normal glucose (NG) or HG for the indicated duration and then subjected to mitophagy analysis, as described in Figure 4. The data are expressed as a fold change in mitophagy relative to NG HRECs for each time point. The average ± SD of at least three independent experiments is shown in the bar graph. ∗P < 0.05 (t-test).

      • Supplemental Figure S7

        Morphology of mitochondria in 10-days normal glucose (NG) and high glucose (HG) cells. mitoORP1-roGFP2–expressing human retinal endothelial cells were cultured in either NG or HG for 10 days and then imaged. The images show that the morphology of the mitochondria was heterogeneous regardless of the glucose concentration, which impaired identification of morphologic features that were unique to a given glucose concentration. Scale bar = 20 μm.

      • Supplemental Figure S8

        DNM1L (fission) is required for hyperglycemia-induced mitochondrial adaptation. At 48 hours after transfection with scrambled siRNA (siScr) or DNM1L-targeting siRNA (siDNM1L) or non-transfected [control (cntr)], high glucose (HG) human retinal endothelial cells (HRECs) were analyzed as follows. A and B: Western blot analysis and RT-qPCR analysis, as in Figure 5. C: Transfected HG-HRECs that were stably expressing MtKeima were subjected to mitophagy analysis, as described in Figure 4. The P value indicates a statistically significant difference between groups. D: Transfected HG-HRECs that stably expressed mito-roGFP2-Orp1 were subjected to the TBH challenge assay, as described in Figure 3. The P value indicates a statistically significant difference between the siScr and siDNM1L groups. E: Same as the left panel of Figure 6C, except with DNM1L siRNA instead of MFN2 siRNA. The transendothelial electrical resistance tracing of a single experiment is presented; similar results were observed in at least three independent experiments. ∗P < 0.05.

      • Supplemental Figure S9

        Raw fluorescence-activated cell sorting (FACS) data for the experiments shown in Figure 7D. Same as Supplemental Figures S3, except that normal glucose (NG) human retinal endothelial cells were treated with vehicle (blue) or urolithin A (UA; black) for 3 days before analyzing cell death. FITC, fluorescein isothiocyanate; PI, propidium iodine.

      • Supplemental Figure S10

        The effect of osmolarity on endothelial cells. Human retinal endothelial cells were cultured in complete endothelial cell media containing normal glucose (NG; 5 mmol/L d-glucose), high glucose (HG; 30 mmol/L d-glucose), or 25 mmol/L l-glucose + 5 mmol/L d-glucose (LG) for at least 10 days. Cells were plated at full confluency for all experiments described in Materials and Methods. A: Basal- and TBH-induced cell death. B: Seahorse analysis of mitochondrial functionality. C: TBH challenge to measure mitochondrial functionality. D: Western blot analysis for mitochondrial proteins that are diagnostic for the total amount of mitochondria. E: Mitophagy. Each set of experiments was repeated on at least three independent occasions. Statistical significance was calculated using t-test. The data are expressed as average ± SD (AE). ∗P < 0.05. COX4, cytochrome C oxidase subunit 4I1; LDH, lactate dehydrogenase; OCR, oxygen consumption rate; Ox., oxidized; TOMM20, translocase of outer mitochondrial membrane 20; veh, vehicle.

      References

        • Lee R.
        • Wong T.Y.
        • Sabanayagam C.
        Epidemiology of diabetic retinopathy, diabetic macular edema and related vision loss.
        Eye Vis (Lond). 2015; 2: 17
        • Aiello L.P.
        • Gardner T.W.
        • King G.L.
        • Blankenship G.
        • Cavallerano J.D.
        • Ferris 3rd, F.L.
        • Klein R.
        Diabetic retinopathy.
        Diabetes Care. 1998; 21: 143-156
        • Klein R.
        • Klein B.E.
        • Moss S.E.
        • Davis M.D.
        • DeMets D.L.
        The Wisconsin epidemiologic study of diabetic retinopathy, II: prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years.
        Arch Ophthalmol. 1984; 102: 520-526
        • Diabetes Control and Complications Trial Research Group
        Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus: Diabetes Control and Complications Trial.
        J Pediatr. 1994; 125: 177-188
        • The Diabetes Control and Complications Trial Research Group
        Clustering of long-term complications in families with diabetes in the diabetes control and complications trial.
        Diabetes. 1997; 46: 1829-1839
        • Cruickshanks K.J.
        • Moss S.E.
        • Klein R.
        • Klein B.E.
        Physical activity and proliferative retinopathy in people diagnosed with diabetes before age 30 yr.
        Diabetes Care. 1992; 15: 1267-1272
        • Chew E.Y.
        • Ambrosius W.T.
        • Davis M.D.
        • Danis R.P.
        • Gangaputra S.
        • Greven C.M.
        • Hubbard L.
        • Esser B.A.
        • Lovato J.F.
        • Perdue L.H.
        • Goff Jr., D.C.
        • Cushman W.C.
        • Ginsberg H.N.
        • Elam M.B.
        • Genuth S.
        • Gerstein H.C.
        • Schubart U.
        • Fine L.J.
        Effects of medical therapies on retinopathy progression in type 2 diabetes.
        N Engl J Med. 2010; 363: 233-244
        • Cheung N.
        • Mitchell P.
        • Wong T.Y.
        Diabetic retinopathy.
        Lancet. 2010; 376: 124-136
        • Yau J.W.
        • Rogers S.L.
        • Kawasaki R.
        • Lamoureux E.L.
        • Kowalski J.W.
        • Bek T.
        • et al.
        Meta-Analysis for Eye Disease Study Group: Global prevalence and major risk factors of diabetic retinopathy.
        Diabetes Care. 2012; 35: 556-564
        • Klein R.
        • Lee K.E.
        • Gangnon R.E.
        • Klein B.E.
        The 25-year incidence of visual impairment in type 1 diabetes mellitus the Wisconsin epidemiologic study of diabetic retinopathy.
        Ophthalmology. 2010; 117: 63-70
        • Brownlee M.
        Biochemistry and molecular cell biology of diabetic complications.
        Nature. 2001; 414: 813-820
        • Stitt A.W.
        • Curtis T.M.
        • Chen M.
        • Medina R.J.
        • McKay G.J.
        • Jenkins A.
        • Gardiner T.A.
        • Lyons T.J.
        • Hammes H.P.
        • Simo R.
        • Lois N.
        The progress in understanding and treatment of diabetic retinopathy.
        Prog Retin Eye Res. 2016; 51: 156-186
        • Antonetti D.A.
        • Klein R.
        • Gardner T.W.
        Diabetic retinopathy.
        N Engl J Med. 2012; 366: 1227-1239
        • Kowluru R.A.
        • Kowluru A.
        • Mishra M.
        • Kumar B.
        Oxidative stress and epigenetic modifications in the pathogenesis of diabetic retinopathy.
        Prog Retin Eye Res. 2015; 48: 40-61
        • Bain S.C.
        • Gill G.V.
        • Dyer P.H.
        • Jones A.F.
        • Murphy M.
        • Jones K.E.
        • Smyth C.
        • Barnett A.H.
        Characteristics of type 1 diabetes of over 50 years duration (the Golden Years Cohort).
        Diabet Med. 2003; 20: 808-811
        • Pickles S.
        • Vigie P.
        • Youle R.J.
        Mitophagy and quality control mechanisms in mitochondrial maintenance.
        Curr Biol. 2018; 28: R170-R185
        • Skeie J.M.
        • Nishimura D.Y.
        • Wang C.L.
        • Schmidt G.A.
        • Aldrich B.T.
        • Greiner M.A.
        Mitophagy: an emerging target in ocular pathology.
        Invest Ophthalmol Vis Sci. 2021; 62: 22
        • Kluge M.A.
        • Fetterman J.L.
        • Vita J.A.
        Mitochondria and endothelial function.
        Circ Res. 2013; 112: 1171-1188
        • Westermann B.
        Mitochondrial fusion and fission in cell life and death.
        Nat Rev Mol Cell Biol. 2010; 11: 872-884
        • Chan D.C.
        Mitochondrial dynamics and its involvement in disease.
        Annu Rev Pathol. 2020; 15: 235-259
        • Chen Y.
        • Dorn 2nd, G.W.
        PINK1-phosphorylated mitofusin 2 is a Parkin receptor for culling damaged mitochondria.
        Science. 2013; 340: 471-475
        • Chen H.
        • McCaffery J.M.
        • Chan D.C.
        Mitochondrial fusion protects against neurodegeneration in the cerebellum.
        Cell. 2007; 130: 548-562
        • Gutscher M.
        • Sobotta M.C.
        • Wabnitz G.H.
        • Ballikaya S.
        • Meyer A.J.
        • Samstag Y.
        • Dick T.P.
        Proximity-based protein thiol oxidation by H2O2-scavenging peroxidases.
        J Biol Chem. 2009; 284: 31532-31540
        • Li Y.
        • Baccouche B.
        • Olayinka O.
        • Serikbaeva A.
        • Kazlauskas A.
        The role of the Wnt pathway in VEGF/anti-VEGF-dependent control of the endothelial cell barrier.
        Invest Ophthalmol Vis Sci. 2021; 62: 17
        • Dey S.
        • Sidor A.
        • O'Rourke B.
        Compartment-specific control of reactive oxygen species scavenging by antioxidant pathway enzymes.
        J Biol Chem. 2016; 291: 11185-11197
        • Waypa G.B.
        • Marks J.D.
        • Guzy R.
        • Mungai P.T.
        • Schriewer J.
        • Dokic D.
        • Schumacker P.T.
        Hypoxia triggers subcellular compartmental redox signaling in vascular smooth muscle cells.
        Circ Res. 2010; 106: 526-535
        • Morgan B.
        • Sobotta M.C.
        • Dick T.P.
        Measuring E(GSH) and H2O2 with roGFP2-based redox probes.
        Free Radic Biol Med. 2011; 51: 1943-1951
        • Sun N.
        • Malide D.
        • Liu J.
        • Rovira I.I.
        • Combs C.A.
        • Finkel T.
        A fluorescence-based imaging method to measure in vitro and in vivo mitophagy using mt-Keima.
        Nat Protoc. 2017; 12: 1576-1587
        • Vargas J.N.S.
        • Wang C.
        • Bunker E.
        • Hao L.
        • Maric D.
        • Schiavo G.
        • Randow F.
        • Youle R.J.
        Spatiotemporal control of ULK1 activation by NDP52 and TBK1 during selective autophagy.
        Mol Cell. 2019; 74: 347-362.e6
        • Gutscher M.
        • Pauleau A.L.
        • Marty L.
        • Brach T.
        • Wabnitz G.H.
        • Samstag Y.
        • Meyer A.J.
        • Dick T.P.
        Real-time imaging of the intracellular glutathione redox potential.
        Nat Methods. 2008; 5: 553-559
        • Li Y.
        • Yan Z.
        • Chaudhry K.
        • Kazlauskas A.
        The renin-angiotensin-aldosterone system (RAAS) is one of the effectors by which vascular endothelial growth factor (VEGF)/anti-VEGF controls the endothelial cell barrier.
        Am J Pathol. 2020; 190: 1971-1981
        • Liao J.K.
        Linking endothelial dysfunction with endothelial cell activation.
        J Clin Invest. 2013; 123: 540-541
        • Busik J.V.
        • Mohr S.
        • Grant M.B.
        Hyperglycemia-induced reactive oxygen species toxicity to endothelial cells is dependent on paracrine mediators.
        Diabetes. 2008; 57: 1952-1965
        • Meyer A.J.
        • Dick T.P.
        Fluorescent protein-based redox probes.
        Antioxid Redox Signal. 2010; 13: 621-650
        • Roma L.P.
        • Deponte M.
        • Riemer J.
        • Morgan B.
        Mechanisms and applications of redox-sensitive green fluorescent protein-based hydrogen peroxide probes.
        Antioxid Redox Signal. 2018; 29: 552-568
        • Albrecht S.C.
        • Barata A.G.
        • Grosshans J.
        • Teleman A.A.
        • Dick T.P.
        In vivo mapping of hydrogen peroxide and oxidized glutathione reveals chemical and regional specificity of redox homeostasis.
        Cell Metab. 2011; 14: 819-829
        • Heytler P.G.
        • Prichard W.W.
        A new class of uncoupling agents--carbonyl cyanide phenylhydrazones.
        Biochem Biophys Res Commun. 1962; 7: 272-275
        • Kim Y.M.
        • Krantz S.
        • Jambusaria A.
        • Toth P.T.
        • Moon H.G.
        • Gunarathna I.
        • Park G.Y.
        • Rehman J.
        Mitofusin-2 stabilizes adherens junctions and suppresses endothelial inflammation via modulation of beta-catenin signaling.
        Nat Commun. 2021; 12: 2736
        • Andreux P.A.
        • Blanco-Bose W.
        • Ryu D.
        • Burdet F.
        • Ibberson M.
        • Aebischer P.
        • Auwerx J.
        • Singh A.
        • Rinsch C.
        The mitophagy activator urolithin A is safe and induces a molecular signature of improved mitochondrial and cellular health in humans.
        Nat Metab. 2019; 1: 595-603
        • Correia S.C.
        • Santos R.X.
        • Perry G.
        • Zhu X.
        • Moreira P.I.
        • Smith M.A.
        Mitochondria: the missing link between preconditioning and neuroprotection.
        J Alzheimers Dis. 2010; 20: S475-S485
        • Gottlieb R.A.
        • Gustafsson A.B.
        Mitochondrial turnover in the heart.
        Biochim Biophys Acta. 2011; 1813: 1295-1301
        • Qiao S.
        • Dennis M.
        • Song X.
        • Vadysirisack D.D.
        • Salunke D.
        • Nash Z.
        • Yang Z.
        • Liesa M.
        • Yoshioka J.
        • Matsuzawa S.
        • Shirihai O.S.
        • Lee R.T.
        • Reed J.C.
        • Ellisen L.W.
        A REDD1/TXNIP pro-oxidant complex regulates ATG4B activity to control stress-induced autophagy and sustain exercise capacity.
        Nat Commun. 2015; 6: 7014
        • Mathew B.
        • Chennakesavalu M.
        • Sharma M.
        • Torres L.A.
        • Stelman C.R.
        • Tran S.
        • Patel R.
        • Burg N.
        • Salkovski M.
        • Kadzielawa K.
        • Seiler F.
        • Aldrich L.N.
        • Roth S.
        Autophagy and post-ischemic conditioning in retinal ischemia.
        Autophagy. 2021; 17: 1479-1499
        • Engbersen R.
        • Riksen N.P.
        • Mol M.J.
        • Bravenboer B.
        • Boerman O.C.
        • Meijer P.
        • Oyen W.J.
        • Tack C.
        • Rongen G.A.
        • Smits P.
        Improved resistance to ischemia and reperfusion, but impaired protection by ischemic preconditioning in patients with type 1 diabetes mellitus: a pilot study.
        Cardiovasc Diabetol. 2012; 11: 124
        • Ravingerova T.
        • Adameova A.
        • Matejikova J.
        • Kelly T.
        • Nemcekova M.
        • Kucharska J.
        • Pechanova O.
        • Lazou A.
        Subcellular mechanisms of adaptation in the diabetic myocardium: relevance to ischemic preconditioning in the nondiseased heart.
        Exp Clin Cardiol. 2010; 15: 68-76
        • Brondani L.A.
        • de Souza B.M.
        • Duarte G.C.
        • Kliemann L.M.
        • Esteves J.F.
        • Marcon A.S.
        • Gross J.L.
        • Canani L.H.
        • Crispim D.
        The UCP1 -3826A/G polymorphism is associated with diabetic retinopathy and increased UCP1 and MnSOD2 gene expression in human retina.
        Invest Ophthalmol Vis Sci. 2012; 53: 7449-7457
        • Crispim D.
        • Fagundes N.J.
        • dos Santos K.G.
        • Rheinheimer J.
        • Boucas A.P.
        • de Souza B.M.
        • Macedo G.S.
        • Leiria L.B.
        • Gross J.L.
        • Canani L.H.
        Polymorphisms of the UCP2 gene are associated with proliferative diabetic retinopathy in patients with diabetes mellitus.
        Clin Endocrinol (Oxf). 2010; 72: 612-619
        • Zhao L.
        • Zhang C.L.
        • He L.
        • Chen Q.
        • Liu L.
        • Kang L.
        • Liu J.
        • Luo J.Y.
        • Gou L.
        • Qu D.
        • Song W.
        • Lau C.W.
        • Ko H.
        • Mok V.C.T.
        • Tian X.Y.
        • Wang L.
        • Huang Y.
        Restoration of autophagic flux improves endothelial function in diabetes through lowering mitochondrial ROS-mediated eNOS monomerization.
        Diabetes. 2022; 71: 1099-1114
        • Hombrebueno J.R.
        • Cairns L.
        • Dutton L.R.
        • Lyons T.J.
        • Brazil D.P.
        • Moynagh P.
        • Curtis T.M.
        • Xu H.
        Uncoupled turnover disrupts mitochondrial quality control in diabetic retinopathy.
        JCI Insight. 2019; 4: e129760
        • Palikaras K.
        • Lionaki E.
        • Tavernarakis N.
        Coordination of mitophagy and mitochondrial biogenesis during ageing in C. elegans.
        Nature. 2015; 521: 525-528
        • Rocha M.
        • Apostolova N.
        • Diaz-Rua R.
        • Muntane J.
        • Victor V.M.
        Mitochondria and T2D: role of autophagy, ER stress, and inflammasome.
        Trends Endocrinol Metab. 2020; 31: 725-741
        • Zhou P.
        • Xie W.
        • Meng X.
        • Zhai Y.
        • Dong X.
        • Zhang X.
        • Sun G.
        • Sun X.
        Notoginsenoside R1 ameliorates diabetic retinopathy through PINK1-dependent activation of mitophagy.
        Cells. 2019; 8: 213
        • Hinder L.M.
        • Sas K.M.
        • O'Brien P.D.
        • Backus C.
        • Kayampilly P.
        • Hayes J.M.
        • Lin C.M.
        • Zhang H.
        • Shanmugam S.
        • Rumora A.E.
        • Abcouwer S.F.
        • Brosius 3rd, F.C.
        • Pennathur S.
        • Feldman E.L.
        Mitochondrial uncoupling has no effect on microvascular complications in type 2 diabetes.
        Sci Rep. 2019; 9: 881
        • Osorio-Paz I.
        • Uribe-Carvajal S.
        • Salceda R.
        In the early stages of diabetes, rat retinal mitochondria undergo mild uncoupling due to UCP2 activity.
        PLoS One. 2015; 10: e0122727
        • Andersen A.R.
        • Christiansen J.S.
        • Andersen J.K.
        • Kreiner S.
        • Deckert T.
        Diabetic nephropathy in type 1 (insulin-dependent) diabetes: an epidemiological study.
        Diabetologia. 1983; 25: 496-501
        • Krolewski A.S.
        • Warram J.H.
        • Christlieb A.R.
        • Busick E.J.
        • Kahn C.R.
        The changing natural history of nephropathy in type I diabetes.
        Am J Med. 1985; 78: 785-794
        • Portillo J.A.
        • Greene J.A.
        • Okenka G.
        • Miao Y.
        • Sheibani N.
        • Kern T.S.
        • Subauste C.S.
        CD40 promotes the development of early diabetic retinopathy in mice.
        Diabetologia. 2014; 57: 2222-2231
        • Portillo J.A.
        • Schwartz I.
        • Zarini S.
        • Bapputty R.
        • Kern T.S.
        • Gubitosi-Klug R.A.
        • Murphy R.C.
        • Subauste M.C.
        • Subauste C.S.
        Proinflammatory responses induced by CD40 in retinal endothelial and Muller cells are inhibited by blocking CD40-Traf2,3 or CD40-Traf6 signaling.
        Invest Ophthalmol Vis Sci. 2014; 55: 8590-8597
        • Portillo J.C.
        • Lopez Corcino Y.
        • Miao Y.
        • Tang J.
        • Sheibani N.
        • Kern T.S.
        • Dubyak G.R.
        • Subauste C.S.
        CD40 in retinal Muller cells induces P2X7-dependent cytokine expression in macrophages/microglia in diabetic mice and development of early experimental diabetic retinopathy.
        Diabetes. 2017; 66: 483-493
        • Liu H.
        • Tang J.
        • Du Y.
        • Lee C.A.
        • Golczak M.
        • Muthusamy A.
        • Antonetti D.A.
        • Veenstra A.A.
        • Amengual J.
        • von Lintig J.
        • Palczewski K.
        • Kern T.S.
        Retinylamine benefits early diabetic retinopathy in mice.
        J Biol Chem. 2015; 290: 21568-21579
        • Liu H.
        • Tang J.
        • Du Y.
        • Saadane A.
        • Samuels I.
        • Veenstra A.
        • Kiser J.Z.
        • Palczewski K.
        • Kern T.S.
        Transducin1, phototransduction and the development of early diabetic retinopathy.
        Invest Ophthalmol Vis Sci. 2019; 60: 1538-1546
        • Wang K.
        • Dai X.
        • He J.
        • Yan X.
        • Yang C.
        • Fan X.
        • Sun S.
        • Chen J.
        • Xu J.
        • Deng Z.
        • Fan J.
        • Yuan X.
        • Liu H.
        • Carlson E.C.
        • Shen F.
        • Wintergerst K.A.
        • Conklin D.J.
        • Epstein P.N.
        • Lu C.
        • Tan Y.
        Endothelial overexpression of metallothionein prevents diabetes-induced impairment in ischemia angiogenesis through preservation of HIF-1alpha/SDF-1/VEGF signaling in endothelial progenitor cells.
        Diabetes. 2020; 69: 1779-1792

      Linked Article

      • This Month in AJP
        The American Journal of PathologyVol. 192Issue 12
        • Preview
          Plasma biomarkers of hepatocyte function may help with diagnosis and prognosis of alcohol-associated hepatitis (AH). Using state-of-the-art in vitro analysis techniques, Argemi et al (Am J Pathol 2022, 1658–1669) identified plasma protein signatures of hepatocyte function for mild and severe AH patients. These proteins were reduced in AH patients. Tested protein combinations differentiated between severe and nonsevere AH with high sensitivity and specificity. These plasma biomarkers may serve as a novel noninvasive molecular tool for research and clinical use in studying and managing AH.
        • Full-Text
        • PDF