Advertisement

Peroxisome Proliferator–Activated Receptor α Protects Capillary Pericytes in the Retina

Open AccessPublished:August 06, 2014DOI:https://doi.org/10.1016/j.ajpath.2014.06.021
      Pericyte degeneration is an early event in diabetic retinopathy and plays an important role in progression of diabetic retinopathy. Clinical studies have shown that fenofibrate, a peroxisome proliferator–activated receptor α (PPARα) agonist, has robust therapeutic effects on diabetic retinopathy in type 2 diabetic patients. We evaluated the protective effect of PPARα against pericyte loss in diabetic retinopathy. In streptozotocin-induced diabetic mice, fenofibrate treatment significantly ameliorated retinal acellular capillary formation and pericyte loss. In contrast, PPARα−/− mice with diabetes developed more severe retinal acellular capillary formation and pericyte dropout, compared with diabetic wild-type mice. Furthermore, PPARα knockout abolished the protective effect of fenofibrate against diabetes-induced retinal pericyte loss. In cultured primary human retinal capillary pericytes, activation and expression of PPARα both significantly reduced oxidative stress–induced apoptosis, decreased reactive oxygen species production, and down-regulated NAD(P)H oxidase 4 expression through blockade of NF-κB activation. Furthermore, activation and expression of PPARα both attenuated the oxidant-induced suppression of mitochondrial O2 consumption in human retinal capillary pericytes. Primary retinal pericytes from PPARα−/− mice displayed more apoptosis, compared with those from wild-type mice under the same oxidative stress. These findings identified a protective effect of PPARα on retinal pericytes, a novel function of endogenous PPARα in the retina.
      Diabetic retinopathy (DR) is a major sight-threatening microvascular complication of both type 1 and type 2 diabetes.
      • Fong D.S.
      • Sharza M.
      • Chen W.
      • Paschal J.F.
      • Ariyasu R.G.
      • Lee P.P.
      Vision loss among diabetics in a group model Health Maintenance Organization (HMO).
      DR is a chronic, progressive, and multifactorial disorder, with retinal microvascular dysfunction being the major component.
      • Gardner T.W.
      • Antonetti D.A.
      • Barber A.J.
      • LaNoue K.F.
      • Levison S.W.
      Diabetic retinopathy: more than meets the eye.
      Retinal capillary pericytes play an essential role in the maintenance of microvascular stability and regulation of endothelial proliferation.
      • Hammes H.P.
      • Lin J.
      • Renner O.
      • Shani M.
      • Lundqvist A.
      • Betsholtz C.
      • Brownlee M.
      • Deutsch U.
      Pericytes and the pathogenesis of diabetic retinopathy.
      It has been shown that pericyte dropout in diabetes correlates with the development of DR.
      • Mizutani M.
      • Kern T.S.
      • Lorenzi M.
      Accelerated death of retinal microvascular cells in human and experimental diabetic retinopathy.
      Pericyte loss is a hallmark of early DR.
      • Engerman R.L.
      Pathogenesis of diabetic retinopathy.
      Relative to healthy subjects, plasma levels of free fatty acids are usually elevated in both type 1 and type 2 diabetic patients.
      • Boden G.
      • Shulman G.I.
      Free fatty acids in obesity and type 2 diabetes: defining their role in the development of insulin resistance and beta-cell dysfunction.
      • Herrero P.
      • Peterson L.R.
      • McGill J.B.
      • Matthew S.
      • Lesniak D.
      • Dence C.
      • Gropler R.J.
      Increased myocardial fatty acid metabolism in patients with type 1 diabetes mellitus.
      Many studies have demonstrated that DR is associated with insulin resistance, which is associated with high plasma free fatty acid levels.
      • Parvanova A.
      • Iliev I.
      • Filipponi M.
      • Dimitrov B.D.
      • Vedovato M.
      • Tiengo A.
      • Trevisan R.
      • Remuzzi G.
      • Ruggenenti P.
      Insulin resistance and proliferative retinopathy: a cross-sectional, case-control study in 115 patients with type 2 diabetes.
      • Roden M.
      • Price T.B.
      • Perseghin G.
      • Petersen K.F.
      • Rothman D.L.
      • Cline G.W.
      • Shulman G.I.
      Mechanism of free fatty acid-induced insulin resistance in humans.
      Palmitate, a saturated fatty acid, has been implicated in dysfunctions and apoptosis in many cell types, including retinal pericytes via activation of NAD(P)H oxidase and NF-κB.
      • Cacicedo J.M.
      • Benjachareowong S.
      • Chou E.
      • Ruderman N.B.
      • Ido Y.
      Palmitate-induced apoptosis in cultured bovine retinal pericytes: roles of NAD(P)H oxidase, oxidant stress, and ceramide.
      • Staiger K.
      • Staiger H.
      • Weigert C.
      • Haas C.
      • Haring H.U.
      • Kellerer M.
      Saturated, but not unsaturated, fatty acids induce apoptosis of human coronary artery endothelial cells via nuclear factor-kappaB activation.
      Peroxisome proliferator–activated receptor α (PPARα), a hormone-activated nuclear receptor, is known as an important modulator of lipid metabolism.
      • Pyper S.R.
      • Viswakarma N.
      • Yu S.
      • Reddy J.K.
      PPARalpha: energy combustion, hypolipidemia, inflammation and cancer.
      PPARα has also been shown to have anti-inflammatory and antioxidant activities.
      • Simo R.
      • Hernandez C.
      Advances in the medical treatment of diabetic retinopathy.
      • Li S.
      • Gokden N.
      • Okusa M.D.
      • Bhatt R.
      • Portilla D.
      Anti-inflammatory effect of fibrate protects from cisplatin-induced ARF.
      Fenofibrate, a potent PPARα agonist, has been used clinically to treat dyslipidemia and cardiovascular disease for >30 years.
      • Drouin P.
      • Lambert D.
      • Mejean L.
      • Pointel J.P.
      • Debry G.
      [Study of lipid metabolic coefficient K2 in patients with hyperlipoproteinemia type IV before and after reduction of triglyceride level by adapted diet therapy].
      Previous studies support that fenofibrate has anti-inflammatory and antioxidant effects.
      • Belfort R.
      • Berria R.
      • Cornell J.
      • Cusi K.
      Feno reduces systemic inflammation markers independent of its effects on lipid and glucose metabolism in patients with the metabolic syndrome.
      • Olukman M.
      • Sezer E.D.
      • Ulker S.
      • Sozmen E.Y.
      • Cinar G.M.
      Feno treatment enhances antioxidant status and attenuates endothelial dysfunction in streptozotocin-induced diabetic rats.
      Two recent large, randomized, placebo-controlled clinical trials, The Fenofibrate Intervention in Event Lowering in Diabetes (FIELD) study and The Action to Control Cardiovascular Risk in Diabetes (ACCORD) study, demonstrated robust therapeutic effects of fenofibrate on microvascular complications of diabetes, including DR in type 2 diabetic patients.
      • Keech A.C.
      • Mitchell P.
      • Summanen P.A.
      • O'Day J.
      • Davis T.M.
      • Moffitt M.S.
      • Taskinen M.R.
      • Simes R.J.
      • Tse D.
      • Williamson E.
      • Merrifield A.
      • Laatikainen L.T.
      • d'Emden M.C.
      • Crimet D.C.
      • O'Connell R.L.
      • Colman P.G.
      FIELD study investigators
      Effect of Feno on the need for laser treatment for diabetic retinopathy (FIELD study): a randomised controlled trial.
      • 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.
      Interestingly, the beneficial effect of fenofibrate on DR was not associated with changes in circulating lipid levels. Our previous study showed that the protective effect of fenofibrate on DR can be achieved through ocular administration, suggesting a local drug target in ocular tissues.
      • Chen Y.
      • Hu Y.
      • Lin M.
      • Jenkins A.J.
      • Keech A.C.
      • Mott R.
      • Lyons T.J.
      • Ma J.X.
      Therapeutic effects of PPARalpha agonists on diabetic retinopathy in type 1 diabetes models.
      The mechanism for the protective effect of fenofibrate on DR is not fully elucidated.
      In our recent studies, we demonstrated that diabetes-induced down-regulation of PPARα plays an important role in retinal inflammation and microvascular dysfunction in DR,
      • Hu Y.
      • Chen Y.
      • Ding L.
      • He X.
      • Takahashi Y.
      • Gao Y.
      • Shen W.
      • Cheng R.
      • Chen Q.
      • Qi X.
      • Boulton M.E.
      • Ma J.X.
      Pathogenic role of diabetes-induced PPAR-alpha down-regulation in microvascular dysfunction.
      and fenofibrate has therapeutic effects on DR in type 1 diabetes models.
      • Chen Y.
      • Hu Y.
      • Lin M.
      • Jenkins A.J.
      • Keech A.C.
      • Mott R.
      • Lyons T.J.
      • Ma J.X.
      Therapeutic effects of PPARalpha agonists on diabetic retinopathy in type 1 diabetes models.
      The function of PPARα in the retina has not been well understood. In this study, we evaluated the protective effect of fenofibrate against pericyte degeneration in DR using primary pericytes, streptozotocin (STZ)-induced diabetic animals, and PPARα knockout (PPARα−/−) mice, and investigated the underlying molecular mechanism.

      Materials and Methods

      Animals

      PPARα−/− mice and age- and genetic background–matched C57/BL6J mice [wild-type (WT) mice] were purchased from Jackson Laboratories (Bar Harbor, ME). All of the animal experiments were in strict agreement with The Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research and approved by the Institutional Animal Care and Use Committee of The University of Oklahoma (Oklahoma City, OK).

      STZ-Induced Diabetic Animal Model

      To induce diabetes in mice, PPARα−/− and WT mice received five consecutive daily i.p. injections of 50 mg/kg STZ. Blood glucose levels were measured 72 hours after the last STZ injection and monitored monthly thereafter. Only mice with consistently elevated glucose levels (>400 mg/dL) were used as diabetic mice.

      Retinal Digestion

      Retinal digestion was performed according to a procedure described previously, with some modifications.
      • Kim J.
      • Kim C.S.
      • Sohn E.
      • Lee Y.M.
      • Jo K.
      • Kim J.S.
      KIOM-79 protects AGE-induced retinal pericyte apoptosis via inhibition of NF-kappaB activation in vitro and in vivo.
      Briefly, the eyeballs were fixed in 10% neutral-buffered formalin for at least 3 days. The retina was dissected and digested with 5% trypsin for 60 minutes. The retinal vasculature was gently isolated and stained with Periodic Acid–Schiff (PAS) Staining System (Sigma-Aldrich, St. Louis, MO). Acellular capillaries and pericyte ghosts in flat-mounted retinal vasculature were quantified from eight random fields per retina, according to a documented protocol.
      • Barile G.R.
      • Pachydaki S.I.
      • Tari S.R.
      • Lee S.E.
      • Donmoyer C.M.
      • Ma W.
      • Rong L.L.
      • Buciarelli L.G.
      • Wendt T.
      • Horig H.
      • Hudson B.I.
      • Qu W.
      • Weinberg A.D.
      • Yan S.F.
      • Schmidt A.M.
      The RAGE axis in early diabetic retinopathy.

      Immunofluorescence Staining

      The digested retinas were stained with the antibodies against NG2 (chondroitin sulfate proteoglycan, a marker for microvasculature pericytes)
      • Ozerdem U.
      • Grako K.A.
      • Dahlin-Huppe K.
      • Monosov E.
      • Stallcup W.B.
      NG2 proteoglycan is expressed exclusively by mural cells during vascular morphogenesis.
      and CD31 (a marker for vascular endothelial cells), as previously described.
      • Kim O.S.
      • Kim J.
      • Kim C.S.
      • Kim N.H.
      • Kim J.S.
      KIOM-79 prevents methyglyoxal-induced retinal pericyte apoptosis in vitro and in vivo.
      Rabbit anti-NG2 antibody was a generous gift from Dr. William B. Stallcup (Sanford-Burham Medical Research Institute, La Jolla, CA), and goat anti-CD31 antibody was from R&D Systems Inc. (Minneapolis, MN). Pericytes were counted in eight random fields per retina.

      Human Retinal Capillary Pericytes

      Primary human retinal capillary pericytes (HRCPs) were purchased from Cambrex Bio Science Walkersville Corp. (Walkersville, MD) and cultured in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum and 1% antibiotic-antimycotic. Cells of passages 2 to 4 were used for experiments.

      Mouse Retinal Pericyte Culture

      Mouse retinal pericytes were cultured from 3-week-old mice, as previously described.
      • Scheef E.A.
      • Sorenson C.M.
      • Sheibani N.
      Attenuation of proliferation and migration of retinal pericytes in the absence of thrombospondin-1.
      Briefly, the retinas were digested with collagenase type II (1 mg/mL; Worthington, Lakewood, NJ) at 37°C for 1 hour. Retinal pericytes from PPARα−/− mice (PPARα−/− pericyte) and those from WT mice (WT pericyte) were resuspended in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum, 1% antibiotic-antimycotic, and 44 U/mL murine recombinant interferon-γ (R&D Systems) and seeded in collagen-coated dishes, respectively. Cells of passages 2 to 4 were used for experiments.

      Identification of Pericytes

      The purity of HRCP and mouse retinal pericytes was confirmed by immunofluorescence with antibodies against specific pericyte markers, α-smooth muscle actin (α-SMA; Sigma-Aldrich) and NG2.

      Recombinant Adenovirus Expressing PPARα and Infection

      A human PPARα cDNA clone (http://www.ncbi.nlm.nih.gov/genbank; GenBank accession number NM_001001928) was purchased from GeneCopeia (Rockville, MD). The full-length coding region of the PPARα cDNA was amplified by PCR using primers (forward primer, 5′-GCGGCCGCCACCATGGTGGACACGGAAAGCCCAC-3′; and reverse primer, 5′-AGCGCTGTACATGTCCCTGTAGATCTCC-3′). The PCR product was cloned into the pGEM-T easy vector (Promega, Madison, WI), and the sequence was confirmed. The clone without any mutation was digested with NotI and AfeI, and then subcloned into pBluscript with a 1D4 epitope sequence (ETSQVAPA, encoded by 5′-ATCAGCGCTGAGACCAGCCAAGTGGCGCCTGCCTAAGTCGACC-3′) fused to the carboxyl terminus of the PPARα cDNA. The 1D4 epitope is a peptide from bovine rhodopsin, which is used as a detection tag in this study. The full-length PPARα-1D4 was digested by KpnI, and its sticky ends were blunt ended using Klenow fragment (New England BioLabs, Ipswich, MA). After heat inactivation of Klenow fragment, the insert was further digested with NotI and subcloned between NotI and EcoRV sites of pShuttle-Cmv IRES-hrGFP (Agilent Technologies, Santa Clara, CA) for construction of adenovirus. Preparation, amplification, and titration of the recombinant adenovirus expressing PPARα-1D4 (Ad-PPARα) were performed as described previously.
      • Moiseyev G.
      • Chen Y.
      • Takahashi Y.
      • Wu B.X.
      • Ma J.X.
      RPE65 is the isomerohydrolase in the retinoid visual cycle.
      HRCPs were infected with either a control adenovirus (Ad-β-gal) or Ad-PPARα [multiplicity of infection (MOI), 50 each], mixed with polyethylenimine (Sigma-Aldrich), following a previously described procedure.
      • Cacicedo J.M.
      • Benjachareowong S.
      • Chou E.
      • Ruderman N.B.
      • Ido Y.
      Palmitate-induced apoptosis in cultured bovine retinal pericytes: roles of NAD(P)H oxidase, oxidant stress, and ceramide.

      TUNEL Assay

      Cells were seeded in an 8-well chamber (BD Biosciences, San Jose, CA) at 1 × 104 cells per well for 24 hours. The cells were fixed with 4% paraformaldehyde for 1 hour, and used for terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL; Roche, Indianapolis, IN), following the manufacturer's instructions.

      Measurement of ROS

      HRCPs were seeded in 24-well plates at 2 × 104 cells per well. Intracellular reactive oxygen species (ROS) were measured using a chloromethyl derivative of H2DCFDA (CM-H2DCFDA; Life Technologies Corp., Carlsbad, CA), according to the manufacturer's recommendation.

      Mitochondrial Respiration Measurement

      HRCPs (1 × 106) were harvested, washed with phosphate-buffered saline, and diluted in 0.75 mL Hanks' balanced salt solution (Life Technologies Corp.). Respiration was measured polarographically at 37°C with a Clark-style oxygen electrode (Instech, Plymouth Meeting, PA). Carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (1 μmol; Sigma-Aldrich) was added to stimulate maximal respiration after 10 minutes. The addition of 2.0 μmol/L rotenone inhibited >95%, thus demonstrating that oxygen consumption was mitochondrial dependent. The starting amount of molecular oxygen in the 0.6-mL electrode chamber was based on the assumption that 213 nmol/mL of molecular oxygen is dissolved at atmospheric pressure and 37°C.

      Western Blot Analysis

      Equal amounts of total protein from each sample were loaded for Western blot analysis with the following primary antibodies: anti–caspase-3 (Cell Signaling, Danvers, MA), anti-Bax (Cell Signaling), anti–Bcl-2 (Cell Signaling), anti-NAD(P)H oxidase 4 (NOX4; Santa Cruz, Dallas, TX), anti–phospho–NF-κB (Cell Signaling), anti-IκBα (Cell Signaling), anti-PPARα (Abcam, Cambridge, MA), anti-succinate dehydrogenase complex, subunit A, flavoprotein (SDHA; Santa Cruz), and anti-NDUFS3 (Life Technologies Corp.). Membranes were stripped and reblotted with anti–β-actin (Sigma-Aldrich) for loading control.

      Statistical Analysis

      All quantitative data were expressed as means ± SD. Statistical analysis was performed using a Student's t-test for comparison of two groups and using one-way analysis of variance for studies of more than two groups. Statistical significance was set at P < 0.05.

      Results

      Fenofibrate Reduces Retinal Acellular Capillary Formation and Pericyte Dropout in Diabetic Mouse Retinas

      Diabetic mice at 3 months after the diabetes onset were fed regular chow (DM) or special chow containing 0.06% fenofibrate (DM + Feno; LabDiet/TestDiet, Fort Worth, TX) for another 3 months. There were no significant differences in blood glucose levels (mg/dL), body weight (g), and food consumption (g) between the DM and DM + Feno groups (Table 1).
      Table 1Summary of Blood Glucose, Body Weight, and Food Intake of Diabetic Mice
      VariableWT-NDMWT-DMWT-DM + FenoPPARα−/−-NDMPPARα−/−-DMPPARα−/−-DM + Feno
      Blood glucose (mg/dL)
       Duration of Diabetes (months)
      0125 ± 13.8113 ± 10.4120 ± 15.2128 ± 18.6107 ± 15.0114 ± 10.6
      1130 ± 10.0497 ± 38.7
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      494 ± 45.2
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      121 ± 17.6492 ± 34.7
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      482 ± 16.4
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      3137 ± 15.1521 ± 30.8
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      515 ± 34.1
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      141 ± 14.6512 ± 41.6
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      528 ± 23.2
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      6126 ± 6.1514 ± 42.2
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      552 ± 50.6
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      129 ± 8.5544 ± 62.7
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      524 ± 56.2
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      Body weight (g)
       Duration of Diabetes (months)
      021 ± 2.820 ± 3.521 ± 2.420 ± 2.221 ± 1.820 ± 1.1
      123 ± 1.520 ± 1.3
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      22 ± 1.9
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      23 ± 2.221 ± 2.0
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      21 ± 0.8
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      330 ± 1.920 ± 2.3
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      20 ± 1.3
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      30 ± 3.520 ± 1.0
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      19 ± 1.4
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      638 ± 1.919 ± 1.9
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      19 ± 1.6
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      39 ± 1.918 ± 1.9
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      18 ± 1.1
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      Food intake (g/day)
       Duration of Diabetes (months)
      02.0 ± 0.221.9 ± 0.142.0 ± 0.302.1 ± 0.122.0 ± 0.241.9 ± 0.13
      12.3 ± 0.192.5 ± 0.152.4 ± 0.252.2 ± 0.202.4 ± 0.362.4 ± 0.29
      33.0 ± 0.323.2 ± 0.273.2 ± 0.203.1 ± 0.473.1 ± 0.233.0 ± 0.16
      64.1 ± 0.353.1 ± 0.28
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      3.1 ± 0.17
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      4.04 ± 0.333.2 ± 0.31
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      3.1 ± 0.15
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      Values are means ± SD.
      Feno, fenofibrate.
      P < 0.05, ∗∗P < 0.01 versus WT-NDM group.
      P < 0.01 versus PPARα−/−-NDM group (n = 5).
      Acellular capillaries and pericyte ghosts were quantified in the trypsin-digested retina of these diabetic mice and age-matched nondiabetic mice (NDM). Retinal acellular capillaries and pericyte ghosts were greatly increased in the DM group, compared with the NDM group (Figure 1, A–D). The DM + Feno group showed significantly reduced retinal acellular capillaries and pericyte ghosts, compared with the DM group with regular chow (Figure 1, A–D).
      Figure thumbnail gr1
      Figure 1Effects of Feno on retinal acellular capillary formation and pericyte loss in diabetic mouse retina. STZ-induced diabetic mice at 3 months after the diabetes onset were fed regular chow or chow containing Feno for another 3 months. Trypsin digestion of the retina was performed in NDM, DM, and DM + Feno (n = 5), followed by PAS staining to view retinal vasculature and evaluate acellular capillaries (A and B) and pericytes ghosts (C and D). AD: Acellular capillaries and pericyte ghosts counted in eight random fields per retina in a double-blind manner and expressed as numbers per mm2 of retina. E and F: Immunostaining of the trypsin-digested retina with antibodies for NG2 (red) and CD31 (green). NG2-positive pericytes were counted in eight random fields and expressed as numbers per mm2 of retina. All values are means ± SD. Acellular capillaries (black arrows, A), pericyte ghost (red arrow, C), and pericytes (white arrows, E). ∗∗P < 0.01 versus NDM group; ††P < 0.01 versus DM group. Original magnification, ×400 (A, C, and E). Feno, fenofibrate.
      Retinal vascular pericyte density was quantified in retinal vasculature with double immunostaining of NG2, a pericyte marker, and CD31, an endothelial cell marker. The results showed that pericyte numbers were decreased in the retina of DM mice compared with that of NDM mice, and fenofibrate significantly increased pericyte density in the retina of DM mice (Figure 1, E and F).

      Fenofibrate Protects Human Retinal Capillary Pericytes from Oxidative Stress–Induced Cell Death

      Palmitate, a diabetic stressor, significantly decreased HRCP viability (Figure 2A). Fenofibrate significantly blunted palmitate-induced cell loss of HRCP in a concentration-dependent manner (Figure 2B). Under normal conditions (without palmitate), however, fenofibrate did not significantly affect HRCP viability, suggesting the increased cell viability is not through enhanced cell growth (Figure 2C).
      Figure thumbnail gr2
      Figure 2Feno improves viability of HRCPs under diabetic stress. A: The HRCPs were treated with different concentrations of Pal for 24 hours, and the cell viability was measured (n = 4). Cell viability of HRCP after incubation with indicated concentrations of Feno for 4 hours, followed by 400 μmol/L Pal for another 24 hours (B) and after treatment with Feno at indicated concentrations for 48 hours (C). Cell viability was measured using MTT assay and expressed as percentage of control. Data are given as means ± SD (n = 4). ∗∗P < 0.01 versus control (A) or versus Pal alone group (B). P < 0.05 versus Pal alone group. Feno, fenofibrate; Pal, palmitate.
      Fenofibrate significantly attenuated palmitate-induced HRCP apoptosis, as observed by TUNEL (Figure 3, A and B). Furthermore, fenofibrate decreased levels of cleaved caspase-3 in palmitate-treated HRCP (Figure 3C) and attenuated the increases of the Bax/Bcl-2 ratio in palmitate-treated pericytes (Figure 3D), supporting the effect of fenofibrate on pericyte apoptosis. Moreover, treatment with fenofibrate after pre-exposure to palmitate for 8 hours also decreased pericyte apoptosis (Supplemental Figure S1).
      Figure thumbnail gr3
      Figure 3Feno protects pericytes from oxidative stress–induced apoptosis. The HRCPs were treated with 40 μmol/L Feno for 4 hours, and then co-incubated with 400 μmol/L Pal for another 48 hours. Detection of apoptotic cells by TUNEL (red) (A) and expression as percentage of total cells (blue, DAPI nucleus staining; n = 4) (B). The HRCPs were treated with 40 μmol/L Feno for 4 hours. Pal (400 μmol/L) was added to and incubated with the cells for another 24 hours. Western blot and densitometric analyses of cleaved caspase-3 (C) and Bax and Bcl-2 (D) in the cells (n = 3). All values are means ± SD. ∗∗P < 0.01 versus control; ††P < 0.01 versus with Pal alone group. Original magnification, ×200 (A). Feno, fenofibrate; Pal, palmitate.

      Fenofibrate Inhibits the NF-κB/NOX4/ROS Pathway

      Fenofibrate significantly reduced palmitate-induced ROS production (Figure 4A). To identify the source of ROS production, we examined the expression of NOX4 in HRCP. Palmitate markedly up-regulated NOX4 expression in HRCP, which was significantly attenuated by fenofibrate (Figure 4B).
      Figure thumbnail gr4
      Figure 4Effects of Feno on ROS production, IκBα down-regulation, phosphorylation of NF-κB, and NOX4 overexpression induced by Pal. The HRCPs were treated with 40 μmol/L Feno for 4 hours and then cotreated with Feno and 400 μmol/L Pal for another 24 hours. A: Quantification of intracellular ROS using CM-H2DCFDA and expression as percentage of control (n = 4). Western blot and densitometric analyses of NOX4 (B), phosphorylated NF-κB (p-NF-κB; C), and IκBα (D). All values are means ± SD (n = 3). ∗∗P < 0.01 versus control; ††P < 0.01 versus Pal alone group. Feno, fenofibrate; Pal, palmitate.
      NF-κB is an important transcription factor and a regulator of NOX4 expression,
      • Lu X.
      • Murphy T.C.
      • Nanes M.S.
      • Hart C.M.
      PPAR{gamma} regulates hypoxia-induced Nox4 expression in human pulmonary artery smooth muscle cells through NF-{kappa}B.
      and NF-κB activation contributes to palmitate-induced pericyte apoptosis.
      • Cacicedo J.M.
      • Benjachareowong S.
      • Chou E.
      • Ruderman N.B.
      • Ido Y.
      Palmitate-induced apoptosis in cultured bovine retinal pericytes: roles of NAD(P)H oxidase, oxidant stress, and ceramide.
      Fenofibrate significantly prevented palmitate-induced NF-κB phosphorylation and rescued IκBα levels, which is the NF-κB inhibitory protein (Figure 4, C and D), suggesting that the antioxidant effect of fenofibrate may be mediated by blockade of the NF-κB/NOX4 pathway.

      PPARα Protects HRCP under Diabetic Stress

      To explore the role of PPARα in pericyte apoptosis induced by oxidant, we evaluated the effect of palmitate on expression of endogenous PPARα in HRCP. PPARα levels were significantly decreased in palmitate-treated HRCP, which was rescued by fenofibrate (Figure 5A).
      Figure thumbnail gr5
      Figure 5Effect of PPARα on Pal-induced apoptosis in HRCPs. The HRCPs were treated with 40 μmol/L Feno for 4 hours and then cotreated with Feno and 400 μmol/L Pal for another 24 hours. A: Western blot and densitometric analyses of PPARα levels. BE: The HRCPs were infected with Ad-PPARα for 24 hours to overexpress PPARα, with Ad-β-gal as control. The HRCPs were treated with 400 μmol/L Pal after incubation with Ad-PPARα for 48 hours, and apoptotic cells were detected with TUNEL staining (red) (B), and quantified and expressed as percentage of total cells (blue, DAPI staining; n = 4) (C). The HRCPs were treated with 400 μmol/L Pal after incubation with Ad-PPARα for 24 hours. Western blot and densitometric analyses of cleaved caspase-3 (D) and Bax and Bcl-2 (E). All values are means ± SD (n = 3). ∗∗P < 0.01 versus control; ††P < 0.01 versus Pal alone group (A) or Ad-β-gal + Pal group (CE). Original magnification, ×200 (B). Feno, fenofibrate; Pal, palmitate.
      To evaluate the role of PPARα in palmitate-induced pericyte apoptosis, PPARα was overexpressed in HRCP using an adenovirus vector (Supplemental Figure S2A). Overexpression of PPARα significantly attenuated palmitate-induced pericyte apoptosis (Figure 5, B and C; Supplemental Figure S2B). Ad-PPARα markedly decreased cleaved caspase-3 and the Bax/Bcl-2 ratio in HRCP exposed to palmitate (Figure 5, D and E).
      Similar to fenofibrate, PPARα overexpression also attenuated palmitate-induced ROS production in HRCP (Figure 6A), and inhibited palmitate-induced NOX4 expression and activation of NF-κB in HRCP (Figure 6, B and C). Overexpression of PPARα prevented the palmitate-induced down-regulation of IκBα levels in HRCP (Figure 6D).
      Figure thumbnail gr6
      Figure 6Effects of PPARα on ROS production, down-regulation of IκBα, NF-κB activation, and NOX4 overexpression induced by Pal. The HRCPs were infected with Ad-PPARα at an MOI of 50 for 24 hours to overexpress PPARα, with the same titer of Ad-β-gal as control. The cells were then exposed to 400 μmol/L Pal for 24 hours. A: Quantification of intracellular ROS using CM-H2DCFDA and expression as percentage of control (n = 4). Western blot and densitometric analyses of NOX-4 (B), phosphorylated NF-κB (p-NF-κB; C), and IκBα (D) (n = 3). Values are means ± SD. ∗∗P < 0.01 versus control; ††P < 0.01 versus Ad-β-gal + Pal group. Pal, palmitate.

      Fenofibrate Prevents Palmitate-Induced Suppression of Mitochondrial Function in Pericytes

      Mitochondria are a major source of ROS production in physiological conditions or under metabolic stress in numerous pathological conditions.
      • Stowe D.F.
      • Camara A.K.
      Mitochondrial reactive oxygen species production in excitable cells: modulators of mitochondrial and cell function.
      Damaged and/or inactive mitochondria produce more ROS than healthy mitochondria.
      • Murphy M.P.
      How mitochondria produce reactive oxygen species.
      Thus, the increase of ROS in HRCP exposed to palmitate may also be ascribed to decreased mitochondrial respiration. Measurement of the rate of mitochondrial O2 consumption revealed that palmitate suppressed both of the basal and maximal (uncoupled) rates of mitochondrial O2 consumption by approximately 50% (Figure 7, A and B). Fenofibrate treatment significantly reversed the decline of mitochondrial O2 consumption rate induced by palmitate. Maximal rates of mitochondrial respiration were not significantly increased by the addition of carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone, a commonly used uncoupler, in any of the treatment conditions, suggesting that the basal level of mitochondria oxygen consumption was near maximal.
      Figure thumbnail gr7
      Figure 7Effects of Feno and PPARα on the Pal-induced suppression of the mitochondrial O2 consumption. AD: The HRCPs were treated with 40 μmol/L Feno for 4 hours and then cotreated with Feno and 400 μmol/L Pal for another 24 hours. EH: The HRCPs were infected with Ad-PPARα for 24 hours to overexpress PPARα, with Ad-β-gal as control, followed by exposure to 400 μmol/L Pal for 24 hours. The basal (A and E) and maximal (B and F) rates of mitochondrial O2 consumption were measured by a Clark-style oxygen electrode and expressed as follows: nmol O2 min−1 million cells−1. Amount of mitochondria was normalized by measurement of NDFUS3 and SDHA (C, D, G, and H) using Western blot and densitometry analyses (n = 3). All values are means ± SD. P < 0.05, ∗∗P < 0.01 versus control; P < 0.05, ††P < 0.01 versus Pal alone group (A and B) or Ad-β-gal + Pal group (E and F). Feno, fenofibrate; Pal, palmitate.
      To determine the impact of PPARα expression on mitochondria function, we infected HRCP with Ad-PPARα, and then exposed the cells to palmitate for 24 hours. Ad-PPARα, but not control adenovirus, significantly attenuated the palmitate-induced decline of mitochondrial O2 consumption, similar to fenofibrate (Figure 7, E and F).
      We also examined whether the fenofibrate-induced changes in mitochondrial respiration are due to changes in total mitochondrial content. NDUFS3 and SDHA, subunits of complexes I and II of the electron transport chain, respectively, were measured with Western blot analysis. No significant changes in NDUFS3 or SDHA levels were detected after 24 hours of palmitate treatment, with or without fenofibrate or with or without infection of Ad-PPARα (Figure 7, C, D, G, and H).

      PPARα Knockout Exacerbate Diabetes-Induced Capillary Degeneration and Pericyte Loss

      To further confirm the role of PPARα in fenofibrate's protection against pericyte loss in DR, diabetic PPARα−/− mice were fed normal chow or fenofibrate chow, as described previously for 3 months, starting at 3 months after diabetes onset. There were no significant differences in blood glucose levels (mg/dL), body weight (g), and food intake (g) between diabetic PPARα−/− mice and age-matched diabetic WT mice, with and without fenofibrate treatment (Table 1).
      In mice with 6 months of diabetes, retinal pericyte density was significantly lower in diabetic PPARα−/− mice than that in diabetic WT mice, with similar levels of hyperglycemia. Fenofibrate had no protective effect on the retinal pericyte loss in diabetic PPARα−/− mice (Figure 8, A and B). Consistently, diabetic PPARα−/− mice developed significantly more acellular capillaries and pericyte ghosts than diabetic WT mice (Figure 8, C and D). Fenofibrate markedly reduced the numbers of acellular capillaries and pericyte ghosts in diabetic WT mice, but not in diabetic PPARα−/− mice (Figure 8, C and D).
      Figure thumbnail gr8
      Figure 8Lack of beneficial effects of Feno on retinal acellular capillary formation and pericyte loss in diabetic PPARα−/− mice and on Pal-induced apoptosis in PPARα−/− pericytes. STZ-induced diabetic WT mice and diabetic PPARα−/− mice at 3 months after the diabetes onset were fed chow containing Feno for another 3 months. AD: The retinas were digested with trypsin. A and B: Immunostaining of the trypsin-digested retina with antibodies for NG2 (red) and CD31 (green). NG2-positive pericytes were counted in eight random fields and expressed as numbers per mm2 of retina. Quantification of retinal acellular capillaries (C) and pericyte ghosts (D) in trypsin-digested retina with PAS staining and expressed as numbers per mm2 of retina (n = 5). Arrows in A indicate Feno Pericytes. E and F: Primary retinal pericytes were cultured from PPARα−/− mice and age-matched WT mice. The cells were treated with 40 μmol/L Feno for 4 hours and then cotreated with Feno and 400 μmol/L Pal for 48 hours. Apoptotic cells were identified by TUNEL (red), and the nuclei were counterstained with DAPI (blue). Apoptotic cells were quantified and expressed as percentage of total cells. G: Mouse primary pericytes were treated with 40 μmol/L Feno for 4 hours and then cotreated with Feno and 400 μmol/L Pal for 24 hours, and cell viability was quantified by MTT assay and expressed as percentage of control (n = 4). All values were expressed as means ± SD. ∗∗P < 0.01 versus NDM group (BD) or control group (F and G); ††P < 0.01 versus DM group (BD) or Pal alone group (F and G); ‡‡P < 0.01 versus WT-DM group (BD) or WT + Pal group (F and G). Original magnification: ×400 (A); ×200 (E). PPARα−/−-DM, diabetic PPARα−/− mice fed regular chow; PPARα−/−-DM + Feno, diabetic PPARα−/− mice fed Feno chow; PPARα−/− + Pal, PPARα−/− pericytes treated with Pal; PPARα−/− + Pal + Feno, PPARα−/− pericytes cotreated with Pal and Feno; WT-DM, diabetic WT mice fed regular chow; WT + Pal, WT pericytes treated with Pal; WT + Pal + Feno, WT pericytes cotreated with Pal and Feno. Feno, fenofibrate; Pal, palmitate.

      PPARα−/− Pericytes Are More Susceptible to Palmitate-Induced Apoptosis

      To confirm that fenofibrate protects pericytes against palmitate-induced apoptosis in a PPARα-dependent manner, we cultured primary retinal pericytes from PPARα−/− mice and age-matched WT mice. More than 90% of primary cells were NG2 or α-SMA–positive cells (Supplemental Figure S3). HRCPs were used as positive control, in which >95% of cells were positive in the pericyte marker staining (Supplemental Figure S3). Palmitate induced a significantly higher apoptotic rate in PPARα−/− pericytes, compared with that in WT pericytes (Figure 8, E and F; Supplemental Figure S4). Fenofibrate dramatically decreased palmitate-induced apoptosis in WT pericytes, but not in PPARα−/− pericytes (Figure 8, E and F). The cell viability of PPARα−/− pericytes was significantly lower than that of WT pericytes after palmitate treatment (Figure 8G). Fenofibrate rescued WT pericytes, but not PPARα−/− pericytes, from palmitate exposure (Figure 8G). These data further confirmed that the pericyte-protective effect of fenofibrate is through a PPARα-dependent mechanism.

      Discussion

      Two large clinical studies, FIELD and ACCORD, have independently reported the major therapeutic effects of the PPARα agonist, fenofibrate, on DR in people with type 2 diabetes.
      • Keech A.C.
      • Mitchell P.
      • Summanen P.A.
      • O'Day J.
      • Davis T.M.
      • Moffitt M.S.
      • Taskinen M.R.
      • Simes R.J.
      • Tse D.
      • Williamson E.
      • Merrifield A.
      • Laatikainen L.T.
      • d'Emden M.C.
      • Crimet D.C.
      • O'Connell R.L.
      • Colman P.G.
      FIELD study investigators
      Effect of Feno on the need for laser treatment for diabetic retinopathy (FIELD study): a randomised controlled trial.
      • 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.
      Our recent studies suggest that down-regulation of PPARα expression in the retina of diabetic models contributes to retinal inflammation and vascular pathological features in DR.
      • Hu Y.
      • Chen Y.
      • Ding L.
      • He X.
      • Takahashi Y.
      • Gao Y.
      • Shen W.
      • Cheng R.
      • Chen Q.
      • Qi X.
      • Boulton M.E.
      • Ma J.X.
      Pathogenic role of diabetes-induced PPAR-alpha down-regulation in microvascular dysfunction.
      However, the function of PPARα in the retina has not been well understood. In this study, we demonstrated, for the first time to our knowledge, that fenofibrate attenuates retinal pericyte dropout and capillary degeneration in diabetic mice through ameliorating diabetes-induced pericyte apoptosis. Regarding a mechanism underlying these protective actions, our data show that fenofibrate blocks ROS generation and inhibits retinal inflammation by suppressing the NF-κB/NOX4 pathway in pericytes. Furthermore, fenofibrate ameliorates mitochondrial dysfunction induced by a diabetic stressor, a major cause of increased ROS production.
      • Stowe D.F.
      • Camara A.K.
      Mitochondrial reactive oxygen species production in excitable cells: modulators of mitochondrial and cell function.
      In addition, we demonstrated that PPARα knockout exacerbates diabetes stress–induced pericyte loss, indicating that PPARα is an endogenous protective factor in pericytes under a diabetic milieu. Our results from both PPARα overexpression and PPARα knockout mice and primary cells suggest that the beneficial effects of fenofibrate on DR are through a PPARα-dependent mechanism. These findings reveal a novel function of PPARα in the retina.
      DR is a chronic, progressive, and multifactorial disorder with retinal microvascular dysfunction as a major component.
      • Gardner T.W.
      • Antonetti D.A.
      • Barber A.J.
      • LaNoue K.F.
      • Levison S.W.
      Diabetic retinopathy: more than meets the eye.
      Pericytes, smooth muscle–like cells surrounding capillaries, play a pivotal role in maintaining vascular architecture and function.
      • Bergers G.
      • Song S.
      The role of pericytes in blood-vessel formation and maintenance.
      In the adult retina, pericytes do not replicate
      • Engerman R.L.
      Pathogenesis of diabetic retinopathy.
      ; thus, pericyte apoptosis is the main cause of the formation of pericyte ghosts and acellular capillaries in the diabetic retina.
      • Barber A.J.
      • Gardner T.W.
      • Abcouwer S.F.
      The significance of vascular and neural apoptosis to the pathology of diabetic retinopathy.
      Pericyte loss is a hallmark of early DR.
      • Engerman R.L.
      Pathogenesis of diabetic retinopathy.
      Therefore, the purpose of the present study was to determine whether activation and overexpression of PPARα have direct protection of pericytes in DR. Our results demonstrate that fenofibrate markedly attenuated pericyte loss and decreased acellular capillaries in a diabetic model, indicating a protective effect of fenofibrate on diabetes-induced retinal pericyte loss. To determine whether the pericyte-protective effect of fenofibrate in diabetic animals is a direct action on pericytes, we used cultured primary pericytes. Our results showed that fenofibrate enhanced survival of pericytes and blunted pericyte apoptosis under oxidative stress, which is similar to N-acetyl-cysteine (NAC), a commonly used ROS scavenger (Supplemental Figure S5).
      Fenofibrate a synthetic ligand of PPARα, has been used clinically for >30 years for the treatment of dyslipidemia.
      • Drouin P.
      • Lambert D.
      • Mejean L.
      • Pointel J.P.
      • Debry G.
      [Study of lipid metabolic coefficient K2 in patients with hyperlipoproteinemia type IV before and after reduction of triglyceride level by adapted diet therapy].
      However, fenofibrate has been reported to confer some PPARα-independent biological activities.
      • Araki H.
      • Tamada Y.
      • Imoto S.
      • Dunmore B.
      • Sanders D.
      • Humphrey S.
      • Nagasaki M.
      • Doi A.
      • Nakanishi Y.
      • Yasuda K.
      • Tomiyasu Y.
      • Tashiro K.
      • Print C.
      • Charnock-Jones D.S.
      • Kuhara S.
      • Miyano S.
      Analysis of PPARalpha-dependent and PPARalpha-independent transcript regulation following Feno treatment of human endothelial cells.
      • Yamasaki D.
      • Kawabe N.
      • Nakamura H.
      • Tachibana K.
      • Ishimoto K.
      • Tanaka T.
      • Aburatani H.
      • Sakai J.
      • Hamakubo T.
      • Kodama T.
      • Doi T.
      Feno suppresses growth of the human hepatocellular carcinoma cell via PPARalpha-independent mechanisms.
      To test if the protective effect of fenofibrate on pericytes is dependent on PPARα, PPARα expression levels were first examined in palmitate-treated pericytes. Our data showed that PPARα is significantly down-regulated in pericytes under diabetic conditions. All of the beneficial effects of fenofibrate can be achieved by overexpression of PPARα. Knockout of PPARα exacerbates diabetes-induced retinal pericyte loss. PPARα knockout also abolished the pericyte-protective effects of fenofibrate in diabetic PPARα−/− mice and in primary pericytes isolated from PPARα−/− mice. Taken together, these observations support that PPARα functions like an endogenous protective factor in pericytes under diabetic stress, and the pericyte-protective effect of fenofibrate is through activation of endogenous PPARα.
      Growing evidence suggests that oxidative stress and inflammation induced by hyperglycemia are the key initiating events in DR
      • Kern T.S.
      Contributions of inflammatory processes to the development of the early stages of diabetic retinopathy.
      • Rosen P.
      • Nawroth P.P.
      • King G.
      • Moller W.
      • Tritschler H.J.
      • Packer L.
      The role of oxidative stress in the onset and progression of diabetes and its complications: a summary of a Congress Series sponsored by UNESCO-MCBN, the American Diabetes Association and the German Diabetes Society.
      and promote apoptosis of retinal cells, including pericytes.
      • Haanen C.
      • Vermes I.
      Apoptosis and inflammation.
      • Kannan K.
      • Jain S.K.
      Oxidative stress and apoptosis.
      Several lines of evidence support that NOX4 plays a key role in ROS over-production in DR.
      • Li J.
      • Wang J.J.
      • Yu Q.
      • Chen K.
      • Mahadev K.
      • Zhang S.X.
      Inhibition of reactive oxygen species by Lovastatin downregulates vascular endothelial growth factor expression and ameliorates blood-retinal barrier breakdown in db/db mice: role of NADPH oxidase 4.
      Our results demonstrated that knockdown of NOX4 by siRNA decreased palmitate-induced ROS production (Supplemental Figure S6). Furthermore, overexpression of NOX4 increased ROS production and suppressed the oxygen consumption rate of mitochondria in pericytes (Supplemental Figure S7). In our study, activation and overexpression of PPARα both significantly reduced ROS generation and NOX4 expression in pericytes under diabetic stress and in diabetic retina (Supplemental Figure S8). These data suggested that down-regulation of NOX4 may be an important mechanism for the antioxidant activity of PPARα, contributing to its protective effect on pericytes. Although NOX4 is not a direct target gene regulated by PPARα, it has been reported that NF-κB, a transcription factor, stimulates NOX4 promoter activity. Also, knockdown of NF-κB expression results in down-regulation of NOX4 expression, suggesting that NOX4 is a target gene of NF-κB,
      • Lu X.
      • Murphy T.C.
      • Nanes M.S.
      • Hart C.M.
      PPAR{gamma} regulates hypoxia-induced Nox4 expression in human pulmonary artery smooth muscle cells through NF-{kappa}B.
      which is consistent with our observation using NF-κB inhibitor (Supplemental Figure S9). Fenofibrate and PPARα both inhibit phosphorylation of NF-κB, suggesting that PPARα blocks NF-κB activation, which subsequently leads to down-regulation of NOX4 expression.
      Retinal inflammation is also a key pathogenic factor in capillary degeneration.
      • Chen Y.
      • Hu Y.
      • Lin M.
      • Jenkins A.J.
      • Keech A.C.
      • Mott R.
      • Lyons T.J.
      • Ma J.X.
      Therapeutic effects of PPARalpha agonists on diabetic retinopathy in type 1 diabetes models.
      NF-κB signaling plays a key role in up-regulation of inflammatory factors in DR. NF-κB activates transcription of many inflammation-related genes, which accelerates apoptosis of retinal pericytes.
      • Romeo G.
      • Liu W.H.
      • Asnaghi V.
      • Kern T.S.
      • Lorenzi M.
      Activation of nuclear factor-kappaB induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes.
      Together with the documented studies,
      • Delerive P.
      • De Bosscher K.
      • Besnard S.
      • Vanden Berghe W.
      • Peters J.M.
      • Gonzalez F.J.
      • Fruchart J.C.
      • Tedgui A.
      • Haegeman G.
      • Staels B.
      Peroxisome proliferator-activated receptor alpha negatively regulates the vascular inflammatory gene response by negative cross-talk with transcription factors NF-kappaB and AP-1.
      this result suggests that the anti-inflammatory effect of fenofibrate and of PPARα through suppression of NF-κB activation may also contribute to pericyte protection.
      There is no documented evidence suggesting that PPARα directly regulates NF-κB transcription in pericytes. It is known that activation of NF-κB is inhibited by the noncovalent interaction with inhibitory proteins, named as IκBs. In an inflammation milieu, IκBα is degraded, which leads to phosphorylation and nuclear translocation of NF-κB, activating transcription of multiple inflammatory factors.
      • Malek S.
      • Huxford T.
      • Ghosh G.
      Ikappa Balpha functions through direct contacts with the nuclear localization signals and the DNA binding sequences of NF-kappaB.
      Our results show that IκBα, the direct target gene of PPARα, is induced by fenofibrate and PPARα, consistent with previous observations.
      • Delerive P.
      • Gervois P.
      • Fruchart J.C.
      • Staels B.
      Induction of IkappaBalpha expression as a mechanism contributing to the anti-inflammatory activities of peroxisome proliferator-activated receptor-alpha activators.
      These findings suggest that fenofibrate suppresses activation of NF-κB, possibly through up-regulating IκBα. The inhibition of NF-κB by fenofibrate may contribute to its antioxidant and anti-inflammatory effects, leading to pericyte protection in diabetes. The present data suggested that palmitate-induced activation of NF-κB might be upstream of up-regulation of NOX4 expression, consistent with a documented study in pulmonary artery smooth muscle cells.
      • Lu X.
      • Murphy T.C.
      • Nanes M.S.
      • Hart C.M.
      PPAR{gamma} regulates hypoxia-induced Nox4 expression in human pulmonary artery smooth muscle cells through NF-{kappa}B.
      However, some studies have suggested that ROS can also regulate NF-κB activation in some cell types.
      • Nakajima S.
      • Kitamura M.
      Bidirectional regulation of NF-kappaB by reactive oxygen species: a role of unfolded protein response.
      The reciprocal regulations between NF-κB and ROS production in pericytes remain to be studied in the future.
      The mitochondrial electron-transport chain and NAD(P)H oxidase are both major sources of ROS during normal metabolism, and the mitochondrial ROS production is increased in a variety of pathological conditions.
      • Stowe D.F.
      • Camara A.K.
      Mitochondrial reactive oxygen species production in excitable cells: modulators of mitochondrial and cell function.
      Mitochondrial dysfunction also plays an important role in cell apoptosis.
      • Jeong S.Y.
      • Seol D.W.
      The role of mitochondria in apoptosis.
      Herein, we present evidence suggesting that palmitate also induces mitochondrial dysfunction, which may contribute to apoptosis of HRCP. Our results demonstrate that activation and overexpression of PPARα both ameliorate palmitate-induced mitochondrial dysfunction, which is not through increasing mitochondria density in pericytes. It has been reported that NOX4 contributes to mitochondria dysfunction and mitochondrial superoxide generation,
      • Ago T.
      • Kuroda J.
      • Pain J.
      • Fu C.
      • Li H.
      • Sadoshima J.
      Upregulation of Nox4 by hypertrophic stimuli promotes apoptosis and mitochondrial dysfunction in cardiac myocytes.
      which indicated that fenofibrate might improve mitochondrial function by inhibiting NOX4 expression.
      In conclusion, we have identified a novel protective effect of PPARα against retinal pericyte loss in DR. Because of the crucial role of retinal pericytes in the progression of DR, the protective effects of fenofibrate on retinal pericyte may account for its clinical benefits on DR. These findings suggest that fenofibrate may be applied to early stages of DR because it alleviates pericyte loss and, thus, may arrest DR progression. Future studies are warranted to investigate how fenofibrate improves the mitochondrial dysfunction in retinal pericytes under diabetic stress.

      Acknowledgment

      We thank Dr. William B. Stallcup (Sanford-Burham Medical Research Institute, La Jolla, CA) for generously offering rabbit anti-NG2 antibody for the study.

      Supplemental Data

      • Supplemental Figure S1

        Interventive effect of Feno on oxidative stress–induced apoptosis. The HRCPs were incubated with 400 μmol/L Pal for 8 hours, followed by the addition of 40 μmol/L Feno, and treated for another 40 hours. Detection of apoptotic cells by TUNEL (red) (A) and quantification as percentage of total cells (blue, DAPI nucleus staining; n = 4) (B). HRCPs were incubated with 400 μmol/L Pal for 8 hours (C), and 40 μmol/L Feno was added and cells were incubated for another 16 hours (D). C and D: Quantification of intracellular ROS using CM-H2DCFDA and expression as percentage of control (n = 4). All values are means ± SD. ∗∗P < 0.01 versus control; ††P < 0.01 versus with Pal alone group. Original magnification, ×200 (A). Feno, fenofibrate; Pal, palmitate.

      • Supplemental Figure S2

        Effects of overexpression of PPARα on pericyte apoptosis. The HRCPs were infected with Ad-PPARα at an MOI of 50 for 24 hours, with the same titer of Ad-β-gal as control. A: Western blot and densitometric analyses of PPARα levels (normalized by β-actin levels; n = 3). B: Detection of apoptotic cells by TUNEL (red); nuclei counterstained with DAPI (blue; n = 4). All values are means ± SD (n = 3). ∗∗P < 0.01 versus Ad-β-gal group. Original magnification, ×200 (B).

      • Supplemental Figure S3

        Identification of HRCPs and primary mouse pericytes using immunostaining of pericyte markers. Immunostaining of HRCPs and mouse retina microvascular pericytes separately with α-SMA (red) and NG2 (red) antibody. The nuclei were counter-stained with DAPI (blue). Original magnification, ×200.

      • Supplemental Figure S4

        Effects of PPARα deficiency on pericyte apoptosis. Primary pericytes were cultured from the retinas of PPARα−/− mice and age-matched WT mice. Detection of apoptotic cells by TUNEL (red); nuclei were counter-stained with DAPI (blue; n = 4). Original magnification, ×200 (TUNEL staining).

      • Supplemental Figure S5

        Attenuation of ROS production reduces pericyte apoptosis. The HRCPs were treated with 1 mmol/L NAC for 4 hours, followed by 400 μmol/L Pal for another 48 hours. Detection of apoptotic cells by TUNEL (red) (A) and expression as percentage of total cells (B), using blue DAPI nucleus staining (n = 4). HRCPs were treated with 1 mmol/L NAC for 4 hours, followed by 400 μmol/L Pal for another 24 hours. C: Quantification of intracellular ROS using CM-H2DCFDA and expression as percentage of control (n = 4). All values are means ± SD. ∗∗P < 0.01 versus control; ††P < 0.01 versus Pal alone group. Original magnification, ×200 (A). Pal, palmitate.

      • Supplemental Figure S6

        Effects of knockdown of NOX4 on ROS production and NF-κB activation induced by Pal. Expression of NOX4 was knocked down using siRNA in HRCPs. A: Western blot and densitometric analyses of NOX4 expression (n = 3). B: Quantification of intracellular ROS using CM-H2DCFDA and expression as percentage of control (Con; n = 4). C: Western blot and densitometric analyses of p-NF-κB after knockdown of NOX4 and treatment with 400 μmol/L Pal for 24 hours (n = 3). Values are means ± SD. ∗∗P < 0.01 versus with con-siRNA group; ††P < 0.01 versus con-siRNA + Pal group (B) or versus with NOX4-siRNA group (C). Pal, palmitate.

      • Supplemental Figure S7

        Effects of adenovirus expressing NOX4 (Ad-NOX4) on ROS and mitochondrial O2 consumption in HRCPs. The HRCPs were infected with Ad-NOX4 at an MOI of 50 for 24 hours to overexpress NOX4, with Ad-β-gal as control. A: Western blot and densitometric analyses of NOX4 levels (n = 3). B: Quantification of intracellular ROS using CM-H2DCFDA and expression as percentage of control (n = 4). Quantification of the basal (C) and maximal (D) rates of mitochondrial O2 consumption measured by a Clark-style oxygen electrode and expressed as follows: nmol O2 min−1 million cells−1. All values are means ± SD. ∗∗P < 0.01 versus Ad-β-gal group.

      • Supplemental Figure S8

        Effects of Feno on NOX4 overexpression in the diabetic mouse retina. STZ-induced diabetic WT mice at 3 months after the diabetes onset were fed chow containing Feno for another 3 months. The retinas were dissected from age-matched NDM, DM, and Feno DM + Feno (n = 5). Western blot (A) and densitometric (B) analyses of NOX4 expression (n = 3). Values are means ± SD. ∗∗P < 0.01 versus NDM group; ††P < 0.01 versus DM group. Feno, fenofibrate.

      • Supplemental Figure S9

        Effects of NF-κB on NOX4 expression in HRCPs. The HRCPs were treated with 10 μmol/L NF-κB inhibitor Bay 11-7082 [Bay for 24 hours, followed by 400 μmol/L Pal for another 24 hours]. Western blot and densitometric analyses of phosphorylated NF-κB (p-NF-κB) (A) and NOX4 (B). All values are means ± SD (n = 3). ∗∗P < 0.01 versus control; ††P < 0.01 versus Pal alone group. Pal, palmitate.

      References

        • Fong D.S.
        • Sharza M.
        • Chen W.
        • Paschal J.F.
        • Ariyasu R.G.
        • Lee P.P.
        Vision loss among diabetics in a group model Health Maintenance Organization (HMO).
        Am J Ophthalmol. 2002; 133: 236-241
        • Gardner T.W.
        • Antonetti D.A.
        • Barber A.J.
        • LaNoue K.F.
        • Levison S.W.
        Diabetic retinopathy: more than meets the eye.
        Surv Ophthalmol. 2002; 47: S253-S262
        • Hammes H.P.
        • Lin J.
        • Renner O.
        • Shani M.
        • Lundqvist A.
        • Betsholtz C.
        • Brownlee M.
        • Deutsch U.
        Pericytes and the pathogenesis of diabetic retinopathy.
        Diabetes. 2002; 51: 3107-3112
        • Mizutani M.
        • Kern T.S.
        • Lorenzi M.
        Accelerated death of retinal microvascular cells in human and experimental diabetic retinopathy.
        J Clin Invest. 1996; 97: 2883-2890
        • Engerman R.L.
        Pathogenesis of diabetic retinopathy.
        Diabetes. 1989; 38: 1203-1206
        • Boden G.
        • Shulman G.I.
        Free fatty acids in obesity and type 2 diabetes: defining their role in the development of insulin resistance and beta-cell dysfunction.
        Eur J Clin Invest. 2002; 32: 14-23
        • Herrero P.
        • Peterson L.R.
        • McGill J.B.
        • Matthew S.
        • Lesniak D.
        • Dence C.
        • Gropler R.J.
        Increased myocardial fatty acid metabolism in patients with type 1 diabetes mellitus.
        J Am Coll Cardiol. 2006; 47: 598-604
        • Parvanova A.
        • Iliev I.
        • Filipponi M.
        • Dimitrov B.D.
        • Vedovato M.
        • Tiengo A.
        • Trevisan R.
        • Remuzzi G.
        • Ruggenenti P.
        Insulin resistance and proliferative retinopathy: a cross-sectional, case-control study in 115 patients with type 2 diabetes.
        J Clin Endocrinol Metab. 2004; 89: 4371-4376
        • Roden M.
        • Price T.B.
        • Perseghin G.
        • Petersen K.F.
        • Rothman D.L.
        • Cline G.W.
        • Shulman G.I.
        Mechanism of free fatty acid-induced insulin resistance in humans.
        J Clin Invest. 1996; 97: 2859-2865
        • Cacicedo J.M.
        • Benjachareowong S.
        • Chou E.
        • Ruderman N.B.
        • Ido Y.
        Palmitate-induced apoptosis in cultured bovine retinal pericytes: roles of NAD(P)H oxidase, oxidant stress, and ceramide.
        Diabetes. 2005; 54: 1838-1845
        • Staiger K.
        • Staiger H.
        • Weigert C.
        • Haas C.
        • Haring H.U.
        • Kellerer M.
        Saturated, but not unsaturated, fatty acids induce apoptosis of human coronary artery endothelial cells via nuclear factor-kappaB activation.
        Diabetes. 2006; 55: 3121-3126
        • Pyper S.R.
        • Viswakarma N.
        • Yu S.
        • Reddy J.K.
        PPARalpha: energy combustion, hypolipidemia, inflammation and cancer.
        Nucl Recept Signal. 2010; 8: e002
        • Simo R.
        • Hernandez C.
        Advances in the medical treatment of diabetic retinopathy.
        Diabetes Care. 2009; 32: 1556-1562
        • Li S.
        • Gokden N.
        • Okusa M.D.
        • Bhatt R.
        • Portilla D.
        Anti-inflammatory effect of fibrate protects from cisplatin-induced ARF.
        Am J Physiol Renal Physiol. 2005; 289: F469-F480
        • Drouin P.
        • Lambert D.
        • Mejean L.
        • Pointel J.P.
        • Debry G.
        [Study of lipid metabolic coefficient K2 in patients with hyperlipoproteinemia type IV before and after reduction of triglyceride level by adapted diet therapy].
        Nutr Metab. 1979; 23 (French): 416-428
        • Belfort R.
        • Berria R.
        • Cornell J.
        • Cusi K.
        Feno reduces systemic inflammation markers independent of its effects on lipid and glucose metabolism in patients with the metabolic syndrome.
        J Clin Endocrinol Metab. 2010; 95: 829-836
        • Olukman M.
        • Sezer E.D.
        • Ulker S.
        • Sozmen E.Y.
        • Cinar G.M.
        Feno treatment enhances antioxidant status and attenuates endothelial dysfunction in streptozotocin-induced diabetic rats.
        Exp Diabetes Res. 2010; 2010: 828531
        • Keech A.C.
        • Mitchell P.
        • Summanen P.A.
        • O'Day J.
        • Davis T.M.
        • Moffitt M.S.
        • Taskinen M.R.
        • Simes R.J.
        • Tse D.
        • Williamson E.
        • Merrifield A.
        • Laatikainen L.T.
        • d'Emden M.C.
        • Crimet D.C.
        • O'Connell R.L.
        • Colman P.G.
        • FIELD study investigators
        Effect of Feno on the need for laser treatment for diabetic retinopathy (FIELD study): a randomised controlled trial.
        Lancet. 2007; 370: 1687-1697
        • 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
        • Chen Y.
        • Hu Y.
        • Lin M.
        • Jenkins A.J.
        • Keech A.C.
        • Mott R.
        • Lyons T.J.
        • Ma J.X.
        Therapeutic effects of PPARalpha agonists on diabetic retinopathy in type 1 diabetes models.
        Diabetes. 2013; 62: 261-272
        • Hu Y.
        • Chen Y.
        • Ding L.
        • He X.
        • Takahashi Y.
        • Gao Y.
        • Shen W.
        • Cheng R.
        • Chen Q.
        • Qi X.
        • Boulton M.E.
        • Ma J.X.
        Pathogenic role of diabetes-induced PPAR-alpha down-regulation in microvascular dysfunction.
        Proc Natl Acad Sci U S A. 2013; 110: 15401-15406
        • Kim J.
        • Kim C.S.
        • Sohn E.
        • Lee Y.M.
        • Jo K.
        • Kim J.S.
        KIOM-79 protects AGE-induced retinal pericyte apoptosis via inhibition of NF-kappaB activation in vitro and in vivo.
        PLoS One. 2012; 7: e43591
        • Barile G.R.
        • Pachydaki S.I.
        • Tari S.R.
        • Lee S.E.
        • Donmoyer C.M.
        • Ma W.
        • Rong L.L.
        • Buciarelli L.G.
        • Wendt T.
        • Horig H.
        • Hudson B.I.
        • Qu W.
        • Weinberg A.D.
        • Yan S.F.
        • Schmidt A.M.
        The RAGE axis in early diabetic retinopathy.
        Invest Ophthalmol Vis Sci. 2005; 46: 2916-2924
        • Ozerdem U.
        • Grako K.A.
        • Dahlin-Huppe K.
        • Monosov E.
        • Stallcup W.B.
        NG2 proteoglycan is expressed exclusively by mural cells during vascular morphogenesis.
        Dev Dyn. 2001; 222: 218-227
        • Kim O.S.
        • Kim J.
        • Kim C.S.
        • Kim N.H.
        • Kim J.S.
        KIOM-79 prevents methyglyoxal-induced retinal pericyte apoptosis in vitro and in vivo.
        J Ethnopharmacol. 2010; 129: 285-292
        • Scheef E.A.
        • Sorenson C.M.
        • Sheibani N.
        Attenuation of proliferation and migration of retinal pericytes in the absence of thrombospondin-1.
        Am J Physiol Cell Physiol. 2009; 296: C724-C734
        • Moiseyev G.
        • Chen Y.
        • Takahashi Y.
        • Wu B.X.
        • Ma J.X.
        RPE65 is the isomerohydrolase in the retinoid visual cycle.
        Proc Natl Acad Sci U S A. 2005; 102: 12413-12418
        • Lu X.
        • Murphy T.C.
        • Nanes M.S.
        • Hart C.M.
        PPAR{gamma} regulates hypoxia-induced Nox4 expression in human pulmonary artery smooth muscle cells through NF-{kappa}B.
        Am J Physiol Lung Cell Mol Physiol. 2010; 299: L559-L566
        • Stowe D.F.
        • Camara A.K.
        Mitochondrial reactive oxygen species production in excitable cells: modulators of mitochondrial and cell function.
        Antioxid Redox Signal. 2009; 11: 1373-1414
        • Murphy M.P.
        How mitochondria produce reactive oxygen species.
        Biochem J. 2009; 417: 1-13
        • Bergers G.
        • Song S.
        The role of pericytes in blood-vessel formation and maintenance.
        Neuro Oncol. 2005; 7: 452-464
        • Barber A.J.
        • Gardner T.W.
        • Abcouwer S.F.
        The significance of vascular and neural apoptosis to the pathology of diabetic retinopathy.
        Invest Ophthalmol Vis Sci. 2011; 52: 1156-1163
        • Araki H.
        • Tamada Y.
        • Imoto S.
        • Dunmore B.
        • Sanders D.
        • Humphrey S.
        • Nagasaki M.
        • Doi A.
        • Nakanishi Y.
        • Yasuda K.
        • Tomiyasu Y.
        • Tashiro K.
        • Print C.
        • Charnock-Jones D.S.
        • Kuhara S.
        • Miyano S.
        Analysis of PPARalpha-dependent and PPARalpha-independent transcript regulation following Feno treatment of human endothelial cells.
        Angiogenesis. 2009; 12: 221-229
        • Yamasaki D.
        • Kawabe N.
        • Nakamura H.
        • Tachibana K.
        • Ishimoto K.
        • Tanaka T.
        • Aburatani H.
        • Sakai J.
        • Hamakubo T.
        • Kodama T.
        • Doi T.
        Feno suppresses growth of the human hepatocellular carcinoma cell via PPARalpha-independent mechanisms.
        Eur J Cell Biol. 2011; 90: 657-664
        • Kern T.S.
        Contributions of inflammatory processes to the development of the early stages of diabetic retinopathy.
        Exp Diabetes Res. 2007; 2007: 95103
        • Rosen P.
        • Nawroth P.P.
        • King G.
        • Moller W.
        • Tritschler H.J.
        • Packer L.
        The role of oxidative stress in the onset and progression of diabetes and its complications: a summary of a Congress Series sponsored by UNESCO-MCBN, the American Diabetes Association and the German Diabetes Society.
        Diabetes Metab Res Rev. 2001; 17: 189-212
        • Haanen C.
        • Vermes I.
        Apoptosis and inflammation.
        Mediators Inflamm. 1995; 4: 5-15
        • Kannan K.
        • Jain S.K.
        Oxidative stress and apoptosis.
        Pathophysiology. 2000; 7: 153-163
        • Li J.
        • Wang J.J.
        • Yu Q.
        • Chen K.
        • Mahadev K.
        • Zhang S.X.
        Inhibition of reactive oxygen species by Lovastatin downregulates vascular endothelial growth factor expression and ameliorates blood-retinal barrier breakdown in db/db mice: role of NADPH oxidase 4.
        Diabetes. 2010; 59: 1528-1538
        • Romeo G.
        • Liu W.H.
        • Asnaghi V.
        • Kern T.S.
        • Lorenzi M.
        Activation of nuclear factor-kappaB induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes.
        Diabetes. 2002; 51: 2241-2248
        • Delerive P.
        • De Bosscher K.
        • Besnard S.
        • Vanden Berghe W.
        • Peters J.M.
        • Gonzalez F.J.
        • Fruchart J.C.
        • Tedgui A.
        • Haegeman G.
        • Staels B.
        Peroxisome proliferator-activated receptor alpha negatively regulates the vascular inflammatory gene response by negative cross-talk with transcription factors NF-kappaB and AP-1.
        J Biol Chem. 1999; 274: 32048-32054
        • Malek S.
        • Huxford T.
        • Ghosh G.
        Ikappa Balpha functions through direct contacts with the nuclear localization signals and the DNA binding sequences of NF-kappaB.
        J Biol Chem. 1998; 273: 25427-25435
        • Delerive P.
        • Gervois P.
        • Fruchart J.C.
        • Staels B.
        Induction of IkappaBalpha expression as a mechanism contributing to the anti-inflammatory activities of peroxisome proliferator-activated receptor-alpha activators.
        J Biol Chem. 2000; 275: 36703-36707
        • Nakajima S.
        • Kitamura M.
        Bidirectional regulation of NF-kappaB by reactive oxygen species: a role of unfolded protein response.
        Free Radic Biol Med. 2013; 65: 162-174
        • Jeong S.Y.
        • Seol D.W.
        The role of mitochondria in apoptosis.
        BMB Rep. 2008; 41: 11-22
        • Ago T.
        • Kuroda J.
        • Pain J.
        • Fu C.
        • Li H.
        • Sadoshima J.
        Upregulation of Nox4 by hypertrophic stimuli promotes apoptosis and mitochondrial dysfunction in cardiac myocytes.
        Circ Res. 2010; 106: 1253-1264