Advertisement

Retinal Neurons Curb Inflammation and Enhance Revascularization in Ischemic Retinopathies via Proteinase-Activated Receptor-2

  • Nicholas Sitaras
    Affiliations
    Department of Pharmacology, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada

    Department of Ophthalmology, Maisonneuve-Rosemont Hospital Research Center, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • José Carlos Rivera
    Correspondence
    Address correspondence to Sylvain Chemtob, M.D., Ph.D., FRCPC, FCAHS, Jean-Sébastien Joyal, M.D., Ph.D., or José Carlos Rivera, Ph.D., CHU Sainte-Justine Research Center, 3175 Côte Sainte-Catherine, Montreal, Quebec, Canada H3T 1C5.
    Affiliations
    Department of Pharmacology, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada

    Department of Ophthalmology, Maisonneuve-Rosemont Hospital Research Center, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • Baraa Noueihed
    Affiliations
    Department of Pharmacology and Therapeutics, McGill University, Montréal, Québec, Canada
    Search for articles by this author
  • Milsa Bien-Aimé
    Affiliations
    Department of Ophthalmology, Maisonneuve-Rosemont Hospital Research Center, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • Karine Zaniolo
    Affiliations
    LOEX-CUO Research Center, Saint-Sacrement Hospital, Québec, Québec, Canada
    Search for articles by this author
  • Samy Omri
    Affiliations
    Department of Pharmacology, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada

    Department of Ophthalmology, Maisonneuve-Rosemont Hospital Research Center, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • David Hamel
    Affiliations
    Department of Pharmacology, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • Tang Zhu
    Affiliations
    Department of Pharmacology, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • Pierre Hardy
    Affiliations
    Department of Pediatrics, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • Przemyslaw Sapieha
    Affiliations
    Department of Ophthalmology, Maisonneuve-Rosemont Hospital Research Center, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • Jean-Sébastien Joyal
    Correspondence
    Address correspondence to Sylvain Chemtob, M.D., Ph.D., FRCPC, FCAHS, Jean-Sébastien Joyal, M.D., Ph.D., or José Carlos Rivera, Ph.D., CHU Sainte-Justine Research Center, 3175 Côte Sainte-Catherine, Montreal, Quebec, Canada H3T 1C5.
    Affiliations
    Department of Pharmacology, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada

    Department of Pharmacology and Therapeutics, McGill University, Montréal, Québec, Canada

    Department of Pediatrics, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
  • Sylvain Chemtob
    Correspondence
    Address correspondence to Sylvain Chemtob, M.D., Ph.D., FRCPC, FCAHS, Jean-Sébastien Joyal, M.D., Ph.D., or José Carlos Rivera, Ph.D., CHU Sainte-Justine Research Center, 3175 Côte Sainte-Catherine, Montreal, Quebec, Canada H3T 1C5.
    Affiliations
    Department of Pharmacology, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada

    Department of Ophthalmology, Maisonneuve-Rosemont Hospital Research Center, University of Montréal, Montréal, Québec, Canada

    Department of Pharmacology and Therapeutics, McGill University, Montréal, Québec, Canada

    Department of Pediatrics, CHU Sainte-Justine Hospital, University of Montréal, Montréal, Québec, Canada
    Search for articles by this author
Open ArchivePublished:December 02, 2014DOI:https://doi.org/10.1016/j.ajpath.2014.10.020
      Ischemic retinopathies are characterized by sequential vaso-obliteration followed by abnormal intravitreal neovascularization predisposing patients to retinal detachment and blindness. Ischemic retinopathies are associated with robust inflammation that leads to generation of IL-1β, which causes vascular degeneration and impairs retinal revascularization in part through the liberation of repulsive guidance cue semaphorin 3A (Sema3A). However, retinal revascularization begins as inflammation culminates in ischemic retinopathies. Because inflammation leads to activation of proteases involved in the formation of vasculature, we hypothesized that proteinase-activated receptor (Par)-2 (official name F2rl1) may modulate deleterious effects of IL-1β. Par2, detected mostly in retinal ganglion cells, was up-regulated in oxygen-induced retinopathy. Surprisingly, oxygen-induced retinopathy–induced vaso-obliteration and neovascularization were unaltered in Par2 knockout mice, suggesting compensatory mechanisms. We therefore conditionally knocked down retinal Par2 with shRNA-Par2–encoded lentivirus. Par2 knockdown interfered with normal revascularization, resulting in pronounced intravitreal neovascularization; conversely, the Par2 agonist peptide (SLIGRL) accelerated normal revascularization. In vitro and in vivo exploration of mechanisms revealed that IL-1β induced Par2 expression, which in turn down-regulated sequentially IL-1 receptor type I and Sema3A expression through Erk/Jnk-dependent processes. Collectively, our findings unveil an important mechanism by which IL-1β regulates its own endothelial cytotoxic actions by augmenting neuronal Par2 expression to repress sequentially IL-1 receptor type I and Sema3A expression. Timely activation of Par2 may be a promising therapeutic avenue in ischemic retinopathies.
      Ischemic retinopathies, such as retinopathy of prematurity and in some circumstances diabetic retinopathy, are the main causes of severe visual impairment in children and working class populations in industrialized countries.
      • Yau J.W.Y.
      • Rogers S.L.
      • Kawasaki R.
      • Lamoureux E.L.
      • Kowalski J.W.
      • Bek T.
      • et al.
      The Meta-Analysis for Eye Disease (Meta-eye) Study Group
      Global prevalence and major risk factors of diabetic retinopathy.
      • Gilbert C.
      • Rahi J.
      • Eckstein M.
      • O'Sullivan J.
      • Foster A.
      Retinopathy of prematurity in middle-income countries.
      Ischemic retinopathies are biphasic diseases characterized by a degeneration of the retinal vascular bed, resulting in tissue hypoxia that triggers a compensatory albeit anarchic and deregulated vessel proliferation into the vitreous, which in severe instances can result in retinal detachment and blindness.
      • Sapieha P.
      • Hamel D.
      • Shao Z.
      • Rivera J.C.
      • Zaniolo K.
      • Joyal J.S.
      • Chemtob S.
      Proliferative retinopathies: angiogenesis that blinds.
      Importantly, retinal ischemia from vascular decay also causes neuronal dysfunction and demise.
      • Fulton A.
      • Hansen R.
      • Moskowitz A.
      • Akula J.
      The neurovascular retina in retinopathy of prematurity.
      Therefore, therapeutic interventions that help alleviate ischemic stress of the neural retina would be desirable to prevent both vascular and neuronal adverse consequences in patients with ischemic retinopathies.
      In ischemic retinopathies, oxidative stress elicits microglia and resident macrophages to release inflammatory cytokines,
      • Demircan N.
      • Safran B.G.
      • Soylu M.
      • Ozcan A.A.
      • Sizmaz S.
      Determination of vitreous interleukin-1 (il-1) and tumour necrosis factor (tnf) levels in proliferative diabetic retinopathy.
      • Mocan M.C.
      • Kadayifcilar S.
      • Eldem B.
      Elevated intravitreal interleukin-6 levels in patients with proliferative diabetic retinopathy.
      • Krady J.K.
      • Basu A.
      • Allen C.M.
      • Xu Y.
      • LaNoue K.F.
      • Gardner T.W.
      • Levison S.W.
      Minocycline reduces proinflammatory cytokine expression, microglial activation, and caspase-3 activation in a rodent model of diabetic retinopathy.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      notably IL-1β, which further recruits and activates inflammatory cells thereby exacerbating retinal inflammation.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • dell'Omo R.
      • Semeraro F.
      • Bamonte G.
      • Cifariello F.
      • Romano M.R.
      • Costagliola C.
      Vitreous mediators in retinal hypoxic diseases.
      • Ishida S.
      • Yamashiro K.
      • Usui T.
      • Kaji Y.
      • Ogura Y.
      • Hia T.
      • Honda Y.
      • Oguchi Y.
      • Adamis A.
      Leukocytes mediate retinal vascular remodeling during dvelopment and vaso-obliteration in disease.
      We found in a murine model of oxygen-induced retinopathy (OIR) that hyperoxia-activated microglia secretes IL-1β through an autocrine feedback loop.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      IL-1β, in turn, induces the release of repulsive cue semaphorin 3A (Sema3A) from neighboring retinal ganglion cells (RGCs), which contributes to vaso-obliteration
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      and hinders normal revascularization.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      This self-sustained and potentially unrelenting destructive inflammatory cascade, if kept unchecked, could result in severe obliteration of the microvascular network of the retina; however, this is not the case because revascularization eventually occurs, albeit slowly (during several days).
      • Dorfman A.
      • Dembinska O.
      • Chemtob S.
      • Lachapelle P.
      Early manifestations of postnatal hyperoxia on the retinal structure and function of the neonatal rat.
      Classic homologous down-regulation of IL receptor type I (Il-1RI) by its own natural ligand has not been observed; in fact, the contrary generally occurs.
      • Teshima S.
      • Nakanishi H.
      • Nishizawa M.
      • Kitagawa K.
      • Kaibori M.
      • Yamada M.
      • Habara K.
      • Kwon A.
      • Kamiyama Y.
      • Ito S.
      • Okumura T.
      Up-regulation of IL-1 receptor through PI3K/Akt is essential for the induction of iNOS gene expression in hepatocytes.
      • Skundric D.S.
      • Bealmear B.
      • Lisak R.P.
      Induced upregulation of IL-1, IL-1RA and IL-1R type I gene expression by Schwann cells.
      Thus, it is reasonable to postulate the existence of an alternate intrinsic regulatory mechanism that counteracts the self-sustained and potentially harmful IL-1β–induced inflammatory cascade.
      Inflammation is associated with the release and activation of various proteases involved in ensued vascularization.
      • Kowluru R.
      • Zhong Q.
      • Santos J.
      Matrix metalloproteinases in diabetic retinopathy: potential role of MMP-9.
      Proteinase-activated receptors (Pars) are G protein–coupled receptors activated by circulating proteases via cleavage of their N-terminus, thus revealing a tethered ligand, which subsequently binds to their second extracellular loop to elicit intracellular signaling.
      • Soh U.J.K.
      • Dores M.R.
      • Chen B.
      • Trejo J.
      Signal transduction by protease-activated receptors.
      Of the four members of the proteinase-activated receptor (Par) family (Par1 through Par4),
      • Soh U.J.K.
      • Dores M.R.
      • Chen B.
      • Trejo J.
      Signal transduction by protease-activated receptors.
      Par2 (official name F2rl1) is recognized for its marked proangiogenic properties in the retina.
      • Uusitalo-Jarvinen H.
      • Kurokawa T.
      • Mueller B.M.
      • Andrade-Gordon P.
      • Friedlander M.
      • Ruf W.
      Role of protease activated receptor 1 and 2 signaling in hypoxia-induced angiogenesis.
      • Belting M.
      • Dorrell M.I.
      • Sandgren S.
      • Aguilar E.
      • Ahamed J.
      • Dorfleutner A.
      • Carmeliet P.
      • Mueller B.M.
      • Friedlander M.
      • Ruf W.
      Regulation of angiogenesis by tissue factor cytoplasmic domain signaling.
      Interestingly, Par2 is robustly activated by IL-1β in different cell types, including neurons and endothelial cells
      • Tsai S.H.
      • Sheu M.T.
      • Liang Y.C.
      • Cheng H.T.
      • Fang S.S.
      • Chen C.
      TGF-β inhibits IL-1β-activated PAR-2 expression through multiple pathways in human primary synovial cells.
      • Ritchie E.
      • Saka M.
      • MacKenzie C.
      • Drummond R.
      • Wheeler-Jones C.
      • Kanke T.
      • Plevin R.
      Cytokine upregulation of proteinase-activated-receptors 2 and 4 expression mediated by p38 MAP kinase and inhibitory kappa B kinase β in human endothelial cells.
      • Boileau C.
      • Amiable N.
      • Martel-Pelletier J.
      • Fahmi H.
      • Duval N.
      • Pelletier J.
      Activation of proteinase-activated receptor 2 in human osteoarthritic cartilage upregulates catabolic and proinflammatory pathways capable of inducing cartilage degradation: a basic science study.
      • Noorbakhsh F.
      • Vergnolle N.
      • McArthur J.C.
      • Silva C.
      • Andrade-Gordon P.
      • Vodjgani M.
      • Hollenberg M.
      • Power C.
      Protease-activated receptor-2 induction by neuroinflammation prevents neuronal death during HIV infection.
      ; IL-1β–mediated inflammation induces the expression of various proteases, which can cleave and activate Par2.
      • Suharti C.
      • van Gorp E.C.
      • Setiati T.E.
      • Dolmans W.M.
      • Djokomoeljanto R.J.
      • Hack C.E.
      • Hugo ten C.
      • van der Meer J.W.
      The role of cytokines in activation of coagulation and fibrinolysis in dengue shock syndrome.
      This may appear rather counterintuitive because amplified IL-1β increases endothelial cell death in OIR,
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      opposing the angiogenic effects of Par2. We therefore hypothesize that in OIR Par2 exerts its angiogenic properties by countering the cytotoxic actions of IL-1β.
      Using a mouse model of OIR,
      • Smith L.E.
      • Wesoloiuski E.
      • McLellan A.
      • Kostyk S.K.
      • D'Amato R.
      • Sullivan R.
      • D'Amore P.A.
      Oxygen-induced retinopathy in the mouse.
      which mimics the cardinal features of ischemic retinopathies, we have uncovered a new mechanism of action for Par2 in retinal revascularization. By conditionally knocking down Par2 [using a lentiviral (Lv)–encoded short hairpin (sh) RNA], we found that IL-1β regulates neuronal Par2 expression, which in turn reduces oxygen-induced vaso-obliteration and enhances desirable revascularization of the retina. We found that Par2 down-regulates IL-1RI specifically in RGCs, which in turn curtails the secretion of Sema3A, thereby facilitating retinal revascularization. Our findings unveil a novel property for Par2 in modulating inflammation, which in the context of ischemic retinopathies limits the vaso-obliterative effects of amplified IL-1β, allowing desirable revascularization and consequently reducing pathologic intravitreal neovascularization.

      Materials and Methods

       Animals

      Adult C57BL6/J mice [wild type (WT)] and Par2 knockout mice (B6.Cg-F2rl1tm1Mslb or Par2−/−) were purchased from Jackson Laboratories (Bar Harbor, ME). Par2−/− mice were genotyped as described by Jackson Laboratories. Adoptive CD-1 lactating females were purchased from Charles Rivers Laboratories (Sainte-Hyacinthe, Canada) to tend to hyperoxia-exposed mice pups. All experiments were conducted in accordance with the Association for Research in Vision and Ophthalmology statement regarding use of animals in ophthalmic and vision research and were approved by Maisonneuve-Rosemont and Sainte-Justine Research Center animal care committees.

       OIR

      Mice pups were exposed from postnatal day 7 to postnatal day 12 to 75% oxygen using a BioSpherix oxycycler (BioSpherix, Lacona, NY). Vaso-obliteration and neovascularization were assessed in hyperoxia-exposed mice pups at postnatal days 12 and 17, respectively, as described previously.
      • Smith L.E.
      • Wesoloiuski E.
      • McLellan A.
      • Kostyk S.K.
      • D'Amato R.
      • Sullivan R.
      • D'Amore P.A.
      Oxygen-induced retinopathy in the mouse.
      • Stahl A.
      • Connor K.M.
      • Sapieha P.
      • Chen J.
      • Dennison R.J.
      • Krah N.M.
      • Seaward M.R.
      • Willett K.L.
      • Aderman C.M.
      • Guerin K.I.
      • Hua J.
      • Lofqvist C.
      • Hellstrom A.
      • Smith L.E.
      The mouse retina as an angiogenesis model.
      Concisely, mice pups were fully anesthetized in 3% isoflurane in oxygen and decapitated using a guillotine. Eyes were enucleated and fixed in 4% paraformaldehyde solution for 1 hour at room temperature. Retinas were dissected and stained overnight at 4°C with fluorescein-labeled Griffonia (Bandeiraea) simplicifolia lectin 1, isolectin B4 (Vector Labs, Burlingame, CA; 1:100). Lectin-stained retinas were whole mounted onto Superfrost/Plus microscope slides (Thermo Fisher Scientific, Waltham, MA) with the photoreceptor side down and imbedded in Fluoro-gel (Electron Microscopy Sciences, Hatfield, PA) and imaged at 10× using a Zeiss AxioObserver.Z1 (Zeiss, San Diego, CA). Images were merged into a single file using the MosiaX option in the AxioVision software version 4.6.5 (Zeiss). Quantification of vaso-obliteration and neovascularization was assessed using the SWIFT_NV methods as previously described.
      • Stahl A.
      • Connor K.M.
      • Sapieha P.
      • Willett K.L.
      • Krah N.M.
      • Dennison R.J.
      • Chen J.
      • Guerin K.I.
      • Smith L.E.
      Computer-aided quantification of retinal neovascularization.
      The commercial IL-1 receptor antagonist (IL-1Ra) anakinra (Kineret; Swedish Orphan Biovitrum AB, Stockholm, Sweden) was administered intraperitoneally from postoperative day 7 (P7) to P8 at 20 mg/kg/d.

       Preparation of Lv Particles

      Third-generation Lv (HIV-1) was prepared as previously described.
      • Dull T.
      • Zufferey R.
      • Kelly M.
      • Mandel R.J.
      • Nguyen M.
      • Trono D.
      • Naldini L.
      A third-generation lentivirus vector with a conditional packaging system.
      A p24 enzyme-linked immunosorbent assay kit (Clontech, Mountain View, CA) was used to quantify Lv titers for Lv shRNA against green fluorescent protein (GFP) (8.5 ng/μL), Lv shPar2 (9.6 ng/μL), and Lv GFP (15.0 ng/μL).

       Intravitreal Injections

      Animals were anesthetized in 3% isolfurane in oxygen and injected intravitreally either at P3 with 1.0 μL of Lv particles (described above) or at P7 with 10 μmol of anakinra or P7 and P9 or P12 and P14 with 100 μmol of NH2-SLIGRL (Elim Biopharm, Hayward, CA) using a Hamilton syringe equipped with 50-gauge glass capillary.

       Immunohistochemistry

      Eyes were enucleated from mice pups and fixed in 4% paraformaldehyde at room temperature for 2 hours and saturated overnight at 4°C in a 30% sucrose solution before embedding in optimal cutting temperature compound (Tissue-Tek, Sakura, Torrance, CA). Coronal sections of 10 μm were sectioned using a Cryostat CM3050S (Leica Microsystems, Concord, Canada). Sections were subsequently washed with phosphate-buffered saline, blocked, and permeabilized for 1 hour at room temperature and subsequently incubated with fluorescein-labeled Griffonia (Bandeiraea) simplicifolia lectin 1, isolectin B4 (Vector Labs; 1:100) for retinal vasculature. Antibodies to rabbit βIII-tubulin (ECM Biosciences, Versailles, KY; 1:1000), mouse βIII-tubulin (Sigma-Aldrich, St Louis, MO; 1:1500), mouse Par2 (SAM11, Invitrogen, Carlsbad, CA; 1:500), rat CD31 (BD Biosciences, Franklin Lakes, NJ; 1:100), rabbit IL-1RI (Santa Cruz Biotechnology, Santa Cruz, CA; 1:400), rabbit Sema3A (Abcam, Cambridge, MA, 1:500), or rabbit vascular endothelial growth factor (VEGF) (Santa Cruz Biotechnology; 1:200), whereas fluoresceinated secondary antibodies (goat anti-mouse IgG Alexa Fluor 488, 594, and/or 647 and goat anti-rabbit IgG Alexa Fluor 488, 594, and/or 647; Invitrogen) were used for localization studies according to the manufacturer's recommendations. Samples were visualized using 30× or 60× objectives with an IX81 confocal microscope (Olympus, Richmond Hill, Canada), and images were obtained with Fluoview software version 3.1 (Olympus). The specificity of Par2 SAM11 monoclonal antibody was confirmed on sagittal sections from Par2−/− mice, revealing absence of immunoreactivity.

       Laser-Capture Microdissection

      Eyes were enucleated and immediately embedded in optimal cutting temperature compound and snap frozen in liquid nitrogen and subsequently cut into 16-μm coronal sections onto MembraneSlide 1.0 PEN nuclease free slides (Zeiss). To visualize vessels, sections were prepared as previously described.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      Retinal sections were laser microdissected with the Zeiss (Observer.Z1) Palm Microbeam laser microscope system. Isolated retinal RNA was transcribed into cDNA for quantitative real-time PCR analysis (see RT-PCR and Quantitative Real-Time PCR).

       Western Blot

      Eyes were enucleated and retinas dissected and placed into commercial radioimmunoprecipitation assay buffer (Cell Signaling Technology, Danvers, MA) and homogenized using Precellys 24 homogenizer (Bertin Technologies, Montigny-le-Bretonneux, France). Samples were centrifuged, and 30 μg of pooled retinal lysate from two different animals was loaded on an SDS-PAGE gel and subsequently electroblotted onto either polyvinylidene difluoride or nitrocellulose membrane (BioRad, Hercules, CA). After blocking, the membranes were blotted with mouse antibody to Par2 (1:400, SAM11, Invitrogen), mouse antibody to β-actin (1:1000, Santa Cruz Biotechnology), rabbit antibody to VEGF (1:200, Santa Cruz Biotechnology), rabbit antibody to IL-1RI (1:400, Santa Cruz Biotechnology), goat antibody to IL-1β (1:400, R&D Systems, Minneapolis, MN), rabbit antibody to Sema3A (1:1000, Abcam), rabbit antibody to total (1:500, Cell Signaling) or phosphorylated Irak1 (1:500, Sigma-Aldrich), rabbit or mouse antibody to total and phosphorylated Erk1/2 (1:1000, Cell Signaling), rabbit antibody to total and phosphorylated p38 (1:1000, Cell Signaling), or rabbit or mouse antibody to total and phosphorylated Jnk (1:1000, Cell Signaling). After washing, membranes were incubated with 1:5000 horseradish peroxidase–conjugated anti-mouse or 1:2000 horseradish peroxidase anti-goat or anti-rabbit secondary antibodies (Millipore, Billerica, MA). Membranes were imaged with LAS-3000 imager, and bands were assessed using densitometry plugins in Multi Gauge software version 4.0 (FujiFilm, Tokyo, Japan). Specificity of SAM11 antibody against Par2 was tested in lysates from Par2−/− mice retina using WT mice retina as control; immunoblots using Par2 SAM11 antibody on cell lysates from various cell lines expressing Par2 or not have also been provided.

       RT-PCR and Quantitative Real-Time PCR

      Freshly dissected whole retina or laser-capture microdissected samples were processed using RiboZol RNA Extraction Reagent (AMRESCO, Solon, Ohio) as indicated in the manufacturer's instructions. Genomic DNA was removed using DNase I (Invitrogen). Approximately 1 μg of total RNA was reverse transcribed into cDNA using iScript RT Supermix (BioRad) as indicated in the manufacturer's instructions. cDNA was analyzed by quantitative real-time PCR using iQ SYBR Green Supermix (BioRad) with primers targeting mouse Par2 (forward 5′-TGACCACGGTCTTTCTTCCG-3′ and reverse 5′-TCAGGGGGAACCAGATGACA-3′), rat Par2 (forward 5′-TGGGAGGTATCACCCTTCTG-3′ and reverse 5′-GGGGAACCAGATGACAGAGA-3′), mouse Sema3A (forward 5′-GCTCCTGCTCCGTAGCCTGC-3′ and reverse 5′-TCGGCGTTGCTTTCGGTCCC-3′), mouse VEGF-A (forward 5′-GCCCTGAGTCAAGAGGACAG-3′ and reverse 5′-CTCCTAGGCCCCTCAGAAGT-3′), mouse and rat IL-1RI (forward 5′-TGAATGTGGCTGAAGAGCAC-3′ and reverse 5′-CGTGACGTTGCAGACAGTT-3′), and mouse IL-1β (forward 5′-CTGGTACATCAGCACCTCACA-3′ and reverse 5′-GAGCTCCTTAACATGCCCTG-3′) [designed using Primer3 (National Center for Biotechnology Information)]. Quantitative gene expression analysis was assessed using the ABI 7500 Real-Time PCR system (Applied Biosystems, Foster City, CA) and compared with control genes cyclophilin A (forward 5′-CAGACGCCACTGTCGCTTT-3′ and reverse 5′-TGTCTTTGGAACTTTGTCTGCAA-3′) or 18S primers (Ambion, Austin, TX) using the ΔΔCT quantification.

       Preparation of Stable Cell Lines

      The RGC-5 cell line was kindly provided by Neeraj Agarwal (National Eye Institute, Bethesda, MD), which was prepared as previously described.
      • Sapieha P.
      • Sirinyan M.
      • Hamel D.
      • Zaniolo K.
      • Joyal J.S.
      • Cho J.H.
      • Honoré J.C.
      • Kermorvant-Duchemin E.
      • Varma D.R.
      • Tremblay S.
      • Leduc M.
      • Rihakova L.
      • Hardy P.
      • Klein W.H.
      • Mu X.
      • Mamer O.
      • Lachapelle P.
      • Di Polo A.
      • Beauséjour C.
      • Andelfinger G.
      • Mitchell G.
      • Sennlaub F.
      • Chemtob S.
      The succinate receptor GPR91 in neurons has a major role in retinal angiogenesis.
      Undifferentiated RGC-5 samples were incubated overnight with Lv particles that contained shRNA. The next day, media was changed and incubated for 48 hours before selection with 5 μg/mL of puromycin (Sigma-Aldrich) for 7 days. Cells were differentiated thereafter with 1 μmol/L staurosporine for 1 hour (Sigma-Aldrich).

       Stimulation of RGC-5 Cell Lines and RBMVECs

      Rat brain microvascular endothelial cells (RBMVECs) were obtained from (Cell Applications Inc, San Diego, CA; catalog number R840-05a) and used between passages 2 and 7 (brain and retina microvascular endothelial cells share numerous common properties
      • Chang-Ling T.
      • Neill A.L.
      • Hunt N.H.
      Early microvascular changes in murine cerebral malaria detected in retinal wholemounts.
      ). RGC-5 samples were cultured in Dulbecco’s modified Eagle’s medium (Invitrogen) supplemented with 10% fetal bovine serum (Cell Applications) and 1% penicillin/streptomycin (Cell Applications) at 37°C and 5% CO2 whereas RBMVECs were cultured as indicated in the manufacturer's instructions. Cells were starved 4 hours before treatment with 1.0 or 10.0 ng/mL of recombinant murine IL-1β (PeproTech, Rocky Hill, NJ) or 0, 40, or 100 μmol/L SLIGRL-NH2. Commercial inhibitors for Erk1/2 (U0126), Jnk (SP600125), and p38 (SB203580) were purchased from Sigma-Aldrich and used at 10 μmol/L approximately 60 minutes before SLIGRL-NH2 treatment.

       Preparation of RGC-5 Conditioned Media

      Terminally differentiated RGC-5 conditioned media were seeded (106 cells) and starved 4 hours before treatment with 0 or 100 μmol/L NH2-SLIGRL and exposed to 5.0 ng/mL of recombinant murine IL-1β (PeproTech). Supernatants were collected 24 hours later, centrifuged briefly, and filtered through 0.22-μm filters (Millipore) and distributed for proliferation assays (see below).

       Cell Survival Assay

      Approximately 104 RBMVECs per well were seeded in 24-well plates and starved 4 hours before exposure to RGC-5 conditioned media. Neutralizing antibody to Sema3A was used at 2 μg/mL (Abcam). After 24 hours, 50 μL of a 5 μg/mL solution of thiazolyl blue tetrazolium bromide (Sigma-Aldrich) was added and cells incubated for 2 to 3 hours. Supernatants were aspirated and cells were lyzed and resuspended in acidified isopropanol. Duplicate absorbance readings were taken at 565 nm using an Infinite M1000 Pro plate reader (Tecan, San Jose, CA).

       Aortic Explant Microvascular Growth Assay

      Aortae were isolated from adult Par2−/− mice, sectioned into 1-mm rings, and placed into growth factor–reduced Matrigel (BD Biosciences) in 24-well plates. Rings were cultured in RBMVEC supplemented endothelial basal medium (Cell Applications Inc, San Diego, CA) 4 to 5 days before a 48 hours exposure to RGC-5 conditioned media. Treated rings were photomicrographed using AxioObserver (Zeiss) and microvascular growth assessed using Image Pro version 4.5 (Media Cybernetics, Rockville, MD). Neutralizing antibody to mouse VEGF164 was used at concentrations of 20 μg/mL (R&D Systems).

       Statistical Analysis

      Results are presented as means ± SEM. for all studies. One-way or two-way analysis of variance with significance α = 0.05 was used for processing data. Bonferroni post hoc analysis was used for calculating significance between groups. Two-tailed Student's t-tests were used to test for significance between two means.

      Results

       Neuronal Par2 Expression Augments during OIR

      Retinas from WT mice subjected to 75% oxygen for 5 days (P7 to P12)
      • Smith L.E.
      • Wesoloiuski E.
      • McLellan A.
      • Kostyk S.K.
      • D'Amato R.
      • Sullivan R.
      • D'Amore P.A.
      Oxygen-induced retinopathy in the mouse.
      were analyzed at different time points during OIR. During the early phases of vaso-obliteration (at P8), Par2 protein levels surged approximately threefold compared with normoxic controls (normalized versus P5) (Figure 1, A and B). Par2 expression in retina remained high during the hyperoxic phase at P12 (Figure 1, A and B); by P17, Par2 levels were comparable to normoxic controls. Par2−/− mice had no immunoreactivity to SAM11 Par2 monoclonal antibody (Supplemental Figure S1).
      Figure thumbnail gr1
      Figure 1Neuronal proteinase-activated receptor (Par)-2 increases in mice retina during oxygen-induced retinopathy (OIR). A: Western blot analysis of wild-type mice whole retina at different time points of OIR showing increases of Par2 at postnatal day 8 (P8) and P12 compared with room-air raised mice retinas (two retinas per time point). B: Densitometry quantification of A. Values are normalized to P5 retinas. C: Immunohistochemical analysis on retinal coronal sections from P8, P12, and P17 wild-type mice pups exposed to either normoxia or OIR. Par2 (green) immunohistochemistry increases substantially in βIII-tubulin–expressing retinal ganglion cells (red) during OIR at P8 and P12. Nuclei are counterstained with DAPI (blue). D: Laser-capture microdissected retinas from P8 pups exposed to normoxia or OIR. Quantitative real-time PCR on microdissected retinal layers reveal robust expression of Par2 mRNA in GCL, compared with microdissected vessels, which increases specifically in this layer after 24-hour exposure to high oxygen. P < 0.05, ∗∗∗P < 0.001. n = 3 independent experiments (A); n = 4 (D). Scale bar = 50 μm (C). Original magnification: ×300 (C). GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; ND, not detected; NFL, nerve fiber layer; ONL, outer nuclear layer; OPL, outer plexiform layer.
      Immunofluorescence analysis on sagittal sections from P8 and P12 normoxic retinas had preferential distribution of Par2 in βIII-tubulin positive RGCs (Figure 1C and Supplemental Figure S2A), which robustly increased in these cells during OIR (Figure 1C and Supplemental Figure S2A). Par2 also slightly co-localized with retinal endothelial cells at P8 (CD31+) (Supplemental Figure S2B). Laser-capture microdissection of retinal tissue followed by real-time quantitative PCR analysis revealed significantly higher Par2 mRNA in the ganglion cell layer (GCL) relative to microdissected vessels (Figure 1D and Supplemental Figure S2, C and D), particularly in OIR-subjected animals (approximately 3.5-fold increase, P < 0.001). Low levels of Par2 mRNA were also detected in the inner nuclear layer likely owing to the invading blood vessels that protrude into this layer in retinas from P8 mice pups (Figure 1D and Supplemental Figure S2, C and D); in OIR, the vaso-obliteration led to the disappearance of Par2 expression in this layer. At P17, Par2 expression was localized mostly to the retinal endothelium in both room air and hyperoxia-raised mice. Altogether, our data indicate that Par2 increases sharply during the early phases of OIR, and its expression is localized preferentially on RGCs.

       Par2 Protects the Retina from Oxygen-Induced Vaso-Obliteration and Induces Retinal Revascularization

      To explore the role of Par2 in ischemic retinopathy, we exposed Par2 knockout mice (Par2−/−) to the OIR model. As previously reported,
      • Uusitalo-Jarvinen H.
      • Kurokawa T.
      • Mueller B.M.
      • Andrade-Gordon P.
      • Friedlander M.
      • Ruf W.
      Role of protease activated receptor 1 and 2 signaling in hypoxia-induced angiogenesis.
      mice lacking the Par2 gene surprisingly had no change in vaso-obliteration or preretinal neovascularization compared with age-matched WT mice (Supplemental Figure S3A). Because, compensatory mechanisms are regularly reported with germ cell line gene modulation,
      • Barbaric I.
      • Miller G.
      • Dear T.
      Appearances can be deceiving: phenotypes of knockout mice.
      and signaling redundancies exist among different Pars
      • Soh U.J.K.
      • Dores M.R.
      • Chen B.
      • Trejo J.
      Signal transduction by protease-activated receptors.
      (as reported in Par2 transgenic mice
      • Schmidlin F.
      • Amadesi S.
      • Dabbagh K.
      • Lewis D.E.
      • Knott P.
      • Bunnet N.W.
      • Gater P.R.
      • Geppetti P.
      • Bertrand C.
      • Stevens M.E.
      Protease-activated receptor 2 mediates eosinophil infiltration and hyperreactivity in allergic inflammation of the airway.
      ), we attempted to circumvent this drawback by conditional knockdown of Par2 using Lv constructs bearing shRNAs targeting Par2 injected intravitreally in WT mice pups at P3. Lv vectors had high tropism for RGCs [see co-localization of green fluorescent protein (encoded in Lv) with βIII-tubulin] (Supplemental Figure S3B).
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      • Sapieha P.
      • Sirinyan M.
      • Hamel D.
      • Zaniolo K.
      • Joyal J.S.
      • Cho J.H.
      • Honoré J.C.
      • Kermorvant-Duchemin E.
      • Varma D.R.
      • Tremblay S.
      • Leduc M.
      • Rihakova L.
      • Hardy P.
      • Klein W.H.
      • Mu X.
      • Mamer O.
      • Lachapelle P.
      • Di Polo A.
      • Beauséjour C.
      • Andelfinger G.
      • Mitchell G.
      • Sennlaub F.
      • Chemtob S.
      The succinate receptor GPR91 in neurons has a major role in retinal angiogenesis.
      • Binet F.
      • Mawambo G.
      • Sitaras N.
      • Tetreault N.
      • Lapalme E.
      • Favret S.
      • Cerani A.
      • Leboeuf D.
      • Tremblay S.
      • Rezende F.
      • Juan A.
      • Stahl A.
      • Joyal J.
      • Milot É.
      • Kaufman R.
      • Guimond M.
      • Kennedy T.
      • Sapieha P.
      Neuronal ER stress impedes myeloid-cell-induced vascular regeneration through ire1-alpha degradation of netrin-1.
      • Cerani A.
      • Tetreault N.
      • Menard C.
      • Lapalme E.
      • Patel C.
      • Sitaras N.
      • Beaudoin F.
      • Leboeuf D.
      • De Guire V.
      • Binet F.
      • Dejda A.
      • Rezende F.
      • Miloudi K.
      • Sapieha P.
      Neuron-derived Semaphorin 3A is an early inducer of vascular permeability in diabetic retinopathy via neuropilin-1.
      In WT mice Lv shPar2 successfully down-regulated retinal Par2 expression by P7 compared with contralateral eyes injected with control Lv shGFP (Supplemental Figure S3C). Par2 knockdown caused an increase in retinal vaso-obliteration at P12 and P17 (Figure 2, A and B), resulting in a significant reduction in the rate of revascularization between these corresponding ages (by 20.9%, P < 0.001) (Figure 2C). Consequently, Lv shPar2-treated animals exhibited increased aberrant preretinal neovascularization at P17 compared with control Lv shGFP injected pups (Figure 2B); Lv shPar2 was ineffective in Par2−/− pups (Supplemental Figure S4), consistent with the specificity of the shPar2.
      Figure thumbnail gr2
      Figure 2Modulation of proteinase-activated receptor (Par)-2 activity during oxygen-induced retinopathy (OIR) affects vaso-obliteration (VO) and neovascularization (NV). A: Representative photomicrographs from isolectin B4–stained retinal flat mounts from postnatal day 12 (P12) wild-type (WT) mice pups exposed to OIR. VO areas are outlined with solid black lines. Pups injected intravitreally with LV containing shRNA against Par2 (approximately 9.6 ng) had significant increases in VO compared with vehicle or control Lv shRNA against green fluorescent protein (shGFP) (approximately 8.5 ng) injected animals. Mice pups injected intravitreally with 100 μmol Par2 agonist peptide SLIGRL-NH2, however, have a significant decrease in VO. VO areas are represented in the bar graph on the right. B: Photomicrographs of P17 WT mice retina exposed to OIR and injected intravitreally with Lv shPar2 have increased VO and NV compared with control vehicle or Lv shGFP–treated animals, whereas SLIGRL-treated mice pups have significant decreases in the avascular area and NV. Avascular areas and NV areas are outlined with solid black lines and white hashed lines, respectively. The bar graphs show the quantification of avascular areas and NV areas by SWIFT_NV. C: Retinal revascularization rate from P12 to P17 after OIR in mice previously injected with Lv shPar2 or SLIGRL. Change in VO was calculated for each group using VO at P12 as starting value (100%). Samples at P17 were compared relative to singly injected vehicle-treated animals. Lv shPar2 delays whereas SLIGRL accelerates retinal revascularization. P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. n = 6 to 12 retinas (A); n = 6 to 12 retinas (B); n = 18 to 23 retinas (C). Original magnification: ×100 (A and B).
      We next proceeded to evaluate the effects of exogenous Par2 stimulation with the Par2 activating peptide SLIGRL (corresponding to tethered ligand in mice)
      • Kaufmann R.
      • Hollenberg M.
      Proteinase-activated receptors (PARs) and calcium signaling in cancer.
      injected intravitreally. Because expression of Par2 increased robustly at P8 and P12, we administered SLIGRL via intravitreal injection during early (P7) and late (P12) hyperoxic exposure. Administration of SLIGRL at P7 caused a significant reduction in vaso-obliteration at P12 compared with vehicle-injected animals (Figure 2A); a successive injection at P9 further diminished vaso-obliteration (Figure 2A). Likewise, intravitreal injections of SLIGRL after hyperoxic exposure (P7-P12) at P12 and P14 reduced dose dependently the size of the avascular area measured at P17 (Figure 2B), which translated to an accelerated rate of revascularization between P12 and P17 (20.6% and 34.3% for single and double injections, P < 0.001) (Figure 2C). As expected, SLIGRL treatment diminished pathologic preretinal neovascularization at P17 (Figure 2B). Par2−/− mice were unresponsive to SLIGRL (Supplemental Figure S4). Collectively, these data indicate that Par2 enhances the integrity of the inflammation-challenged
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      retinal vascular bed during hyperoxic stress by promoting normal revascularization, which in turn reduces undesirable aberrant preretinal neovascularization.

       IL-1β Induces Par2 Expression in Neuronal and Endothelial Cells

      Augmented levels of inflammatory cytokines, such as IL-1β, IL-6, and tumor necrosis factor-α, are found in the vitreous of patients with ischemic retinopathies and corresponding animal models.
      • Demircan N.
      • Safran B.G.
      • Soylu M.
      • Ozcan A.A.
      • Sizmaz S.
      Determination of vitreous interleukin-1 (il-1) and tumour necrosis factor (tnf) levels in proliferative diabetic retinopathy.
      • Mocan M.C.
      • Kadayifcilar S.
      • Eldem B.
      Elevated intravitreal interleukin-6 levels in patients with proliferative diabetic retinopathy.
      • Krady J.K.
      • Basu A.
      • Allen C.M.
      • Xu Y.
      • LaNoue K.F.
      • Gardner T.W.
      • Levison S.W.
      Minocycline reduces proinflammatory cytokine expression, microglial activation, and caspase-3 activation in a rodent model of diabetic retinopathy.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      Of relevance, IL-1β increases Par2 expression in cultured endothelial cells,
      • Ritchie E.
      • Saka M.
      • MacKenzie C.
      • Drummond R.
      • Wheeler-Jones C.
      • Kanke T.
      • Plevin R.
      Cytokine upregulation of proteinase-activated-receptors 2 and 4 expression mediated by p38 MAP kinase and inhibitory kappa B kinase β in human endothelial cells.
      chondrocytes,
      • Boileau C.
      • Amiable N.
      • Martel-Pelletier J.
      • Fahmi H.
      • Duval N.
      • Pelletier J.
      Activation of proteinase-activated receptor 2 in human osteoarthritic cartilage upregulates catabolic and proinflammatory pathways capable of inducing cartilage degradation: a basic science study.
      synovial cells,
      • Tsai S.H.
      • Sheu M.T.
      • Liang Y.C.
      • Cheng H.T.
      • Fang S.S.
      • Chen C.
      TGF-β inhibits IL-1β-activated PAR-2 expression through multiple pathways in human primary synovial cells.
      and neuronal cells.
      • Noorbakhsh F.
      • Vergnolle N.
      • McArthur J.C.
      • Silva C.
      • Andrade-Gordon P.
      • Vodjgani M.
      • Hollenberg M.
      • Power C.
      Protease-activated receptor-2 induction by neuroinflammation prevents neuronal death during HIV infection.
      We explored the association between IL-1β and Par2 in OIR. Consistent with the intimate link between oxidative stress and inflammation,
      • Kim Y.
      • West X.Z.
      • Byzova T.V.
      Inflammation and oxidative stress in angiogenesis and vascular disease.
      IL-1β levels increased during hyperoxia, particularly at P8 (Figure 3A), when Par2 expression peaked in OIR (Figure 1). Pretreatment of hyperoxia-exposed mice with the IL-1 receptor antagonist (IL-1Ra) anakinra reduced IL-1β mRNA expression (Supplemental Figure S5A) as reported
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      and significantly lowered Par2 mRNA levels at P8 (Figure 3B), suggesting that IL-1β may regulate Par2 expression in vivo.
      Figure thumbnail gr3
      Figure 3Increased expression of IL-1β in the retina during oxygen-induced retinopathy (OIR) stimulates proteinase-activated receptor (Par)-2 expression. A: Western blot analysis of IL-1β in whole retina from mice pups at various time points of OIR indicating an increase at postnatal day 8 (P8) compared with room-air raised wild-type (WT) mice pups (two retinas per time point). The bar graph shows the densitometry quantification. Values are normalized to P5 retinas. B: Quantitative real-time PCR analysis of whole retina from P8 WT mice pups demonstrating an increase in Par2 mRNA upon exposure to hyperoxia for 24 hours. Animals treated i.p. with IL-1 receptor antagonist (IL-1Ra; Kineret) have significantly less Par2 mRNA. IL-1β stimulated cultured retinal neurons [retinal ganglion cell (RGC)-5] (C) and rat brain microvascular endothelial cells (RBMVECs) (D) exhibit time- and dose-dependent increases in Par2 mRNA as analyzed by quantitative real-time PCR. Administration of IL-1Ra abrogates IL-1β–dependent increases in Par2 mRNA in both RGC-5 (E) and RBMVECs (F) as determined by quantitative real-time PCR analysis. G: RGC-5 samples stimulated with IL-1β have a time-dependent increase in Par2 protein levels. Densitometry quantifications are illustrated in the bar graphs. H: RBMVECs stimulated with IL-1β have an increase in Par2 protein levels at 12 hours only. The bar graph represent densitometry quantifications. n = 3 independent experiments (A); n = 3 (B); n = 4 to 8 (C and D); n = 2 independent experiments (G); n = 2 independent experiments (H). P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.
      In vivo findings were ascertained in vitro. Par2-expressing cultured murine retinal neuronal cells (RGC-5) and rat brain microvascular endothelial cells (RBMVECs) (Supplemental Figure S5, B and C) had dose-dependent increases in Par2 mRNA expression on treatment with recombinant murine IL-1β (Figure 3, C and D); these effects were abolished by co-treatment with IL-1Ra (Figure 3, E and F). Par2 mRNA increased within 2 to 4 hours after stimulation with IL-1β in both RGC-5 and RBMVECs but was not sustained thereafter, whereas increased Par2 protein translation was detected by 6 to 12 hours in both cells but continued to increase at 24 hours only in RGC-5 (Figure 3, G and H), consistent with changes in Par2 expression in microdissected retinas of animals subjected or not to OIR (Figure 1D). Hence, IL-1β up-regulates Par2 expression in retinal ganglion neurons in a sustained manner but yields a limited and transient response in vascular endothelial cells.

       Par2 Is Involved in Negative Feedback Regulation of IL-1 Response

      Chronic inflammation can be detrimental to the nascent vessels, either directly via cytokine or chemokine signaling
      • Gardiner T.A.
      • Gibson D.S.
      • de Gooyer T.E.
      • de la Cruz V.F.
      • McDonald D.M.
      • Stitt A.
      Inhibition of tumor necrosis factor-α improves physiological angiogenesis and reduces pathological neovascularization in ischemic retinopathy.
      • Meyer N.
      • Christoph J.
      • Makrinioti H.
      • Indermitte P.
      • Rhyner C.
      • Soyka M.
      • Eiwegger T.
      • Chalubinski T.
      • Wanke K.
      • Fujita H.
      • Wawrzyniak P.
      • Burgler S.
      • Zhang S.
      • Akdis M.
      • Menz G.
      • Akdis C.
      Inhibition of angiogenesis by IL-32: possible role in asthma.
      • Massague J.
      • Blain S.
      • Lo R.
      TGF-beta signaling in growth control, cancer, and heritable disorders.
      • Kermorvant-Duchemin E.
      • Sennlaub F.
      • Sirinyan M.
      • Brault S.
      • Andelfinger G.
      • Kooli A.
      • Germain S.
      • Ong H.
      • d'Orleans-Juste P.
      • Gobeil F.
      • Zhu T.
      • Boisvert C.
      • Hardy P.
      • Jain K.
      • Falck J.R.
      • Balazy M.
      • Chemtob S.
      Trans-arachidonic acids generated during nitrative stress induce a thrombospondin-1–dependent microvascular degeneration.
      • Minuzzo S.
      • Moserle L.
      • Indraccolo S.
      • Amadori A.
      Angiogenesis meets immunology: cytokine gene therapy of cancer.
      or indirectly through inflammatory cell activation.
      • dell'Omo R.
      • Semeraro F.
      • Bamonte G.
      • Cifariello F.
      • Romano M.R.
      • Costagliola C.
      Vitreous mediators in retinal hypoxic diseases.
      • Ishida S.
      • Yamashiro K.
      • Usui T.
      • Kaji Y.
      • Ogura Y.
      • Hia T.
      • Honda Y.
      • Oguchi Y.
      • Adamis A.
      Leukocytes mediate retinal vascular remodeling during dvelopment and vaso-obliteration in disease.
      • Shen H.
      • Yao P.
      • Lee E.
      • Greenhalgh D.
      • Soulika A.
      Interferon-gamma inhibits healing post scald burn injury.
      We previously uncovered that IL-1β, predominantly generated by microglia in OIR, interfered with normal retinal revascularization by eliciting RGCs to secrete Sema3A, a proapoptotic endothelial repulsive cue.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      Paradoxically, we found that IL-1β also up-regulates Par2 (Figure 3), which enhances normal revascularization of the retina (Figure 2C). We therefore set out to determine how Par2 promotes retinal revascularization in OIR by examining its modulation of IL-1 signaling. We determined IL-1β and IL-1RI levels in retinas from P9 WT mouse pups exposed to OIR and injected intravitreally with SLIGRL or Lv shPar2. Both IL-1β and IL-1RI expression increased in vehicle-injected retinas of animals exposed to hyperoxia (Figure 4, A and B). Par2 stimulation with SLIGRL attenuated the OIR-provoked increase in IL-1RI expression predominantly localized on RGC (Figure 4, A and B); normoxic animals treated with SLIGRL also exhibited moderate reductions in IL-1RI (Supplemental Figure S6, A and B). Conversely Lv shPar2 further augmented IL-1RI expression in OIR (Figure 4A). SLIGRL and Lv shPar2 had no effect on IL-1β expression (Figure 4A).
      Figure thumbnail gr4
      Figure 4Proteinase-activated receptor (Par)-2 activation results in down-regulation of IL-1 receptor type I (IL-1RI) in vivo. A: Western Blot analysis of IL-1β and IL-1RI in postnatal day 9 (P9) wild-type (WT) mice whole retina exposed to oxygen-induced retinopathy (OIR) and injected with vehicle, 100 μmol SLIGRL, Lv shRNA against green fluorescent protein (shGFP), or Lv shPAR-2 (two retinas per group). Increases in IL-1β and IL-1RI are observed in OIR exposed retinas; however, treatment with SLIGRL significantly reduces IL-1RI expression but not IL-1β. Conversely, administration of Lv shPar2 trends toward an increase in IL-1RI levels; however, the results are not significant. IL-1β levels remained the same. The bar graphs indicate densitometry quantification. B: Immunohistochemical analysis of P9 WT mice retinas exposed to OIR (middle panels) have significant immunoreactivity to IL-1RI (green) localized primarily in the ganglion cell layer (βIII-tubulin; red) compared with normoxic-exposed retinas (upper panels). SLIGRL administration reverses this process (lower panels). Nuclei are counterstained with DAPI (blue) and vessels with isolection B4 (magenta). C: Representative images from isolectin B4–stained retinal flat mounts from P12 WT mice pups exposed to OIR and injected intravitreally with either 100 μmol SLIGRL or 10 μmol IL-1Ra or both. Vaso-obliteration (VO) areas are outlined with solid blue lines. Individually, administration of Par2 agonist and Il-1RI antagonist successfully inhibits hyperoxia-induced VO. However, there is no additive effect from co-administration of both SLIGRL and IL-1Ra. n = 3 independent experiments (A); n = 6 to 8 retinas (C). P < 0.05, ∗∗∗P < 0.001. Scale bar = 50 μm (B). Original magnification: ×300 (B); ×100 (C). GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; NFL, nerve fiber layer; ONL, outer nuclear layer; OPL, outer plexiform layer.
      To determine whether inhibition of IL-1β and stimulation of Par2 acted synergistically in enhancing revascularization, WT mice subjected to OIR were treated with IL-1Ra, SLIGRL, or both. The effects on vaso-obliteration were comparable (Figure 4C), suggesting that Par2 stimulation did not induce any additional mechanism to that through IL-1RI suppression to enhance retinal revascularization in OIR. As anticipated, Par2−/− mice were only responsive to IL-1Ra (Supplemental Figure S6C). Consistent with comparable vaso-obliteration in WT and Par2−/− mice (Supplemental Figure S3A), Par2−/− mice had similar IL-1RI levels compared with both normoxic and OIR WT mice, supporting the presence of compensatory mechanism in Par2−/− mice with OIR (Supplemental Figure S6D).

       Par2 Stimulation Reduces IL-1RI Expression in Retinal Neurons via Erk1/2 and Downstream Jnk Signaling

      To elucidate signaling pathways by which Par2 regulates IL-1RI expression, we corroborated our in vivo observations on cultured cells. In RGC-5 samples, Par2 stimulation with SLIGRL suppressed time and dose dependently IL-1RI mRNA and protein expression (Figure 5, A and B); concordantly, SLIGRL also attenuated IL-1β–induced Irak-1 phosphorylation (Supplemental Figure S7A). Cells infected with Lv shPar2 were unresponsive to SLIGRL (Figure 5, A and B, and Supplemental Figure S7B). Interestingly, in contrast to RGCs, in RBMVECs IL-1RI mRNA and protein expression increased at 1 hour and 2 hours after stimulation with SLIGRL, respectively (Supplemental Figure S7, C and D), suggesting distinct signaling pathways in different cell types.
      Figure thumbnail gr5
      Figure 5Erk1/2 activates downstream Jnk, which is required for proteinase-activated receptor (Par)-2–dependent inhibition IL-1 receptor type I (IL-1RI) in retinal neurons. A: Retinal ganglion cell (RGC)-5 cell lines exhibit dose-dependent decreases in IL-1RI mRNA after 1-hour stimulation with SLIGRL. RGC-5 infected with Lv shPar2 abrogates effect. Lv shGFP–infected cells served as control RGC-5. B: RGC-5 cell lines have a time-dependent decrease in IL-1RI protein levels after 100 μmol/L SLIGRL treatment; Lv shPar2-infected RGC-5 do not respond to SLIGRL treatment. Densitometry quantification is shown below. C: Administration of SLIGRL activates mitogen-activated protein kinase (MAPK) signaling pathways: Erk1/2, Jnk, and p38. D: RGC-5 cell lines pretreated with either MEK1/2 inhibitor (U0125) or Jnk inhibitor (SP600125) but not p38 inhibitor (SB203580) abrogate the SLIGRL-dependent decreases in IL-1RI mRNA in RGC-5. E: Activation of MAPK by 100 μmol/L SLIGRL in RGC-5 is abolished using specific inhibitors to MEK1/2 (U0125), p38 (SB203580), and Jnk (SP600125); inhibition of Erk1/2 also inhibits Jnk phosphorylation. Conversely, Erk1/2 phosphorylation is not affected by Jnk inhibitor, indicating that Erk1/2 activates Jnk, which sequentially down-regulates IL-1RI mRNA. n = 6 to 10 (A); n = 2 independent experiments (B). P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.
      Par2 stimulation results in activation of several downstream effectors, including Ca2+ transients and mitogen-activated protein kinase (MAPK) signaling.
      • Soh U.J.K.
      • Dores M.R.
      • Chen B.
      • Trejo J.
      Signal transduction by protease-activated receptors.
      • Rothmeier A.
      • Ruf W.
      Protease-activated receptor 2 signaling in inflammation.
      • Ossovskaya V.S.
      • Nunnet N.
      Protease-activated receptors: contribution to physiology and disease.
      RGC-5 samples treated with SLIGRL exhibited a time-dependent increase in MAPK signaling, including Erk1/2, Jnk, and p38 phosphorylation (Figure 5C). Increased MAPK activation was also detected in retina of WT mice subjected to OIR (Supplemental Figure S7E); similar observations were made in Par2−/− mice, again consistent with compensatory mechanisms in these knockout mice. Inhibition of p38 (using SB203580) in RGC-5 samples did not affect SLIGRL-dependent suppression of IL-1RI mRNA, whereas pretreatment with MEK1/2 inhibitor (U0125) or Jnk inhibitor (SP600125) prevented SLIGRL-induced IL-1RI mRNA suppression (Figure 5D). Western blot analysis revealed that the selective MEK1/2 inhibitor (U0125) inhibited Erk1/2 and Jnk phosphorylation, whereas the specific Jnk inhibitor (SP600125) only inhibited Jnk phosphorylation. Together these observations suggest that Par2 stimulation suppresses IL-1RI in RGCs by signaling sequentially via Erk1/2 and Jnk (Figure 5E), as previously reported.
      • Paumelle R.
      • Tulasne D.
      • Leroy C.
      • Coll J.
      • Vandenbunder B.
      • Fafeur V.
      Sequential activation of ERK and repression of JNK by scatter factor/hepatocyte growth factor in Madin-Darby canine kidney epithelial cells.
      In contrast, in RBMVEC SLIGRL-induced IL-1RI up-regulation was unaffected by the MAPK kinase inhibitor (PD98059) substantiating distinct signaling pathways of Par-2 in different cells types (Supplemental Figure S7, F and G).

       Par2 Activation Inhibits IL-1β–Mediated Sema3A Release and Ensuing Endothelial Cell Death and Promotes Angiogenesis by Concomitant VEGF Release

      OIR-triggered increase in IL-1β leads to Sema3A release and ensuing endothelial cell death, resulting in increased vasoobliteration
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      and hindered normal revascularization.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      Because Par2 negatively regulates IL-1RI signaling (Figures 4 and 5), we proceeded to investigate whether this was associated with suppressed Sema3A expression. OIR-triggered Sema3A expression was indeed suppressed by Par2 stimulation with SLIGRL (Figure 6, A and B); as expected, Lv shPar2 further augmented Sema3A levels compared with Lv shGFP (Figure 6A). Concomitantly, VEGF-A levels increased sharply in RGCs on SLIGRL stimulation (Supplemental Figure S8, A and B), facilitating angiogenesis (Supplemental Figure S8, C and D).
      Figure thumbnail gr6
      Figure 6Activation of proteinase-activated receptor (Par)-2 abolishes IL-1β–mediated semaphorin 3A (Sema3A) release, reducing endothelial cell death and promoting vascular sprouting. A: Western blot analysis of Sema3A in postnatal day 9 (P9) wild-type (WT) mice whole retinas exposed to oxygen-induced retinopathy (OIR) and injected intravitreally. OIR exposure augments Sema3A expression in P9 vehicle-injected animals; however, OIR retinas treated with 100 μmol SLIGRL have significantly reduced Sema3A levels compared with vehicle-injected eyes (2 retinas per group). The bar graph shows densitometry quantification. Conversely, Lv shPar2 treatment augments Sema3A levels compared with Lv shRNA against green fluorescent protein (shGFP)–injected retinas. B: Immunohistochemical analysis on P9 WT mice retina exposed to hyperoxia reveals significant immunoreactivity to Sema3A, particularly in the ganglion cell layer (middle panels) compared to normoxic-exposed retinas (upper panels). Treatment with SLIGRL, however, significantly reduces Sema3A levels in this layer (lower panels). C: Retinal ganglion cell (RGC)-5 cell lines stimulated with IL-1β exhibit time-dependent increases in Sema3a mRNA. D: Control Lv shGFP-infected RGC-5 stimulated with 100 μmol/L SLIGRL abrogates IL-1β–dependent increases in Sema3a mRNA after 24 hours; however, this effect is abolished in Lv shPar2–infected RGC-5. E: MTT cell survival assay on rat brain microvascular endothelial cells (RBMVECs) exposed to conditioned media (CM) from RGC-5 cell lines stimulated with IL-1β reveals significant reduction in endothelial cell survival. Treatment with an antibody to Sema3A abrogates this effect. F: Similar experiment as in E but using CM from Lv shGFP– or Lv shPar2–infected RGC-5 stimulated with IL-1β, SLIGRL, or both. IL-1β–stimulated CM significantly reduces endothelial cell survival, whereas CM from Lv shGFP RGC-5 co-stimulated with SLIGRL and IL-1β reverses this effect. Lv shPar2 RGC-5 co-stimulated with SLIGRL and IL-1β does not salvage the IL-1β–induced endothelial cell death. G: Aortic explants exposed to the same CM as in F. Explants treated with CM from SLIGRL-stimulated RGC-5 have increased sprouting, whereas explants exposed to CM from IL-1β–stimulated RGC-5 exhibit diminished sprouting growth. CM from Lv shGFP RGC-5 co-stimulated with SLIGRL and IL-1β translate to normal aortic sprouting angiogenesis, whereas CM from Lv shPar2 RGC-5 does not. The bar graph shows quantification of aortic vascular sprouting. n = 3 independent experiments (A); n = 6 (C); n = 7 to 8 (D); n = 5 to 6 (E); n = 6 (F); n = 6 to 9 (G). P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Scale bar = 50 μm (B). Original magnification: ×300 (B); ×40 (G). GCL, ganglion cell layer; INL; inner nuclear layer; IPL, inner plexiform layer; NFL, nerve fiber layer; ONL, outer nuclear layer; OPL, outer plexiform layer.
      To ascertain in vivo findings, we stimulated RGC-5 with IL-1β and measured Sema3A in the presence or absence of SLIGRL. Sema3A mRNA was markedly induced by IL-1β (Figure 6C), and this effect was abrogated by SLIGRL, which was ineffective in Lv shPar2-infected cells (Figure 6D); importantly, SLIGRL in the absence of IL-1β did not affect Sema3A expression.
      Next, we determined the functional role of Par2 in angiogenesis by assessing RBMVEC survival, using conditioned media from RGC-5 stimulated with IL-1β in the presence or absence of SLIGRL. Endothelial cell death triggered by IL-1β stimulation of RGC-5 was Sema3A dependent because this was abrogated using a neutralizing antibody to Sema3A (Figure 6E). IL-1β–induced endothelial cell death was also abolished by Par2 stimulation (Figure 6F). Similarly, conditioned media from Par2 stimulated RGC-5 elicited aortic explant vascular sprouting (Figure 6G) due to concomitant VEGF-A release from these cells (Supplemental Figure S8, C and D), whereas conditioned media from IL-1β stimulated RGC-5 diminished aortic explant vascular sprouting, and this effect was abolished by co-treatment of RGC-5 with SLIGRL (Figure 6G). Altogether these data indicate that the actions of Par2 stimulation in the retina negatively modulate IL-1β actions and thus prevent IL-1β–induced Sema3A release while maintaining elevated levels of VEGF-A; these effects dampen endothelial cell death during oxygen-induced vaso-obliteration and augment desirable revascularization through retinovascular proliferation.

      Discussion

      Sustained inflammation is an established cause of vessel network injury.
      • Demircan N.
      • Safran B.G.
      • Soylu M.
      • Ozcan A.A.
      • Sizmaz S.
      Determination of vitreous interleukin-1 (il-1) and tumour necrosis factor (tnf) levels in proliferative diabetic retinopathy.
      • Mocan M.C.
      • Kadayifcilar S.
      • Eldem B.
      Elevated intravitreal interleukin-6 levels in patients with proliferative diabetic retinopathy.
      • Krady J.K.
      • Basu A.
      • Allen C.M.
      • Xu Y.
      • LaNoue K.F.
      • Gardner T.W.
      • Levison S.W.
      Minocycline reduces proinflammatory cytokine expression, microglial activation, and caspase-3 activation in a rodent model of diabetic retinopathy.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • dell'Omo R.
      • Semeraro F.
      • Bamonte G.
      • Cifariello F.
      • Romano M.R.
      • Costagliola C.
      Vitreous mediators in retinal hypoxic diseases.
      • Ishida S.
      • Yamashiro K.
      • Usui T.
      • Kaji Y.
      • Ogura Y.
      • Hia T.
      • Honda Y.
      • Oguchi Y.
      • Adamis A.
      Leukocytes mediate retinal vascular remodeling during dvelopment and vaso-obliteration in disease.
      • Gardiner T.A.
      • Gibson D.S.
      • de Gooyer T.E.
      • de la Cruz V.F.
      • McDonald D.M.
      • Stitt A.
      Inhibition of tumor necrosis factor-α improves physiological angiogenesis and reduces pathological neovascularization in ischemic retinopathy.
      • Tremblay S.
      • Miloudi K.
      • Chaychi S.
      • Favret S.
      • Binet F.
      • Polosa A.
      • Lachapelle P.
      • Chemtob S.
      • Sapieha P.
      Systemic inflammation perturbs developmental retinal angiogenesis and neuroretinal function.
      Resident microglial cells, a dominant source of proinflammatory IL-1β, are autoactivated by this cytokine, which contributes to exacerbating vaso-obliteration in ischemic retinopathies.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      Despite this autoamplified inflammation, vaso-obliteration eventually ceases to progress, giving rise instead to normal retinal revascularization,
      • Dorfman A.
      • Dembinska O.
      • Chemtob S.
      • Lachapelle P.
      Early manifestations of postnatal hyperoxia on the retinal structure and function of the neonatal rat.
      • Lange C.
      • Ehlken C.
      • Stahl A.
      • Martin G.
      • Hansen L.
      • Agostini H.
      Kinetics of retinal vaso-obliteration and neovascularisation in the oxygen-induced retinopathy (OIR) mouse model.
      implying the presence of an intrinsic regulatory mechanism. We found that neuronal Par2 is an important negative regulator of IL-1β–induced vascular degeneration, thereby fostering the revascularization of vaso-obliterated retinal tissue. In this context, rapid IL-1β release induces Par2 expression, which in turn negatively regulates IL-1RI signaling by abrogating its expression; these effects essentially counteract the relentless inflammatory autoamplification cascade
      • Teshima S.
      • Nakanishi H.
      • Nishizawa M.
      • Kitagawa K.
      • Kaibori M.
      • Yamada M.
      • Habara K.
      • Kwon A.
      • Kamiyama Y.
      • Ito S.
      • Okumura T.
      Up-regulation of IL-1 receptor through PI3K/Akt is essential for the induction of iNOS gene expression in hepatocytes.
      • Skundric D.S.
      • Bealmear B.
      • Lisak R.P.
      Induced upregulation of IL-1, IL-1RA and IL-1R type I gene expression by Schwann cells.
      that would otherwise cause irreversible retinal damage.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      IL-1RI down-regulation occurs in Par2-expressing RGCs where it curtails proapoptotic Sema3A production and concomitantly enhances VEGF secretion, together promoting revascularization of vaso-obliterated retinal regions (Figure 7).
      Figure thumbnail gr7
      Figure 7Proteinase-activated receptor (Par)-2 on retinal ganglion cells (RGCs) enhances revascularization by suppressing IL-1 receptor type I (IL-1RI). Augmented levels of secreted IL-1β stimulate rapid production of Par2 to counteract the hyperoxic stress. However, exhaustive inflammation causes RGCs to suspend Par2 production and instead induces semaphorin 3A (Sema3A) production, which leads to vaso-obliteration and contributes to pathological preretinal neovascularization. Conversely, timely activation of Par2, using peptide agonist SLIGRL, abrogates this process by suppressing IL-1RI via Erk1/2 and Jnk-dependent processes in RGCs to blunt Sema3A production and concomitantly stimulate vascular endothelial growth factor (VEGF) production to augment the revascularization process of the ischemic retina.
      There is increasing evidence that neurons play a crucial role in the formation of retinal vessel network in development and disease. For instance, mice pups that lack retinal ganglion cells (brn3bZ-dta/+; six3-cre mice) are devoid of a retinal capillary network.
      • Sapieha P.
      • Sirinyan M.
      • Hamel D.
      • Zaniolo K.
      • Joyal J.S.
      • Cho J.H.
      • Honoré J.C.
      • Kermorvant-Duchemin E.
      • Varma D.R.
      • Tremblay S.
      • Leduc M.
      • Rihakova L.
      • Hardy P.
      • Klein W.H.
      • Mu X.
      • Mamer O.
      • Lachapelle P.
      • Di Polo A.
      • Beauséjour C.
      • Andelfinger G.
      • Mitchell G.
      • Sennlaub F.
      • Chemtob S.
      The succinate receptor GPR91 in neurons has a major role in retinal angiogenesis.
      Early in OIR, neurons appear to contribute to neurovascular survival by increasing proangiogenic signals, such as GPR91, VEGF, and Netrin1, in an attempt to offset damages from ischemic stress.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      • Sapieha P.
      • Sirinyan M.
      • Hamel D.
      • Zaniolo K.
      • Joyal J.S.
      • Cho J.H.
      • Honoré J.C.
      • Kermorvant-Duchemin E.
      • Varma D.R.
      • Tremblay S.
      • Leduc M.
      • Rihakova L.
      • Hardy P.
      • Klein W.H.
      • Mu X.
      • Mamer O.
      • Lachapelle P.
      • Di Polo A.
      • Beauséjour C.
      • Andelfinger G.
      • Mitchell G.
      • Sennlaub F.
      • Chemtob S.
      The succinate receptor GPR91 in neurons has a major role in retinal angiogenesis.
      • Binet F.
      • Mawambo G.
      • Sitaras N.
      • Tetreault N.
      • Lapalme E.
      • Favret S.
      • Cerani A.
      • Leboeuf D.
      • Tremblay S.
      • Rezende F.
      • Juan A.
      • Stahl A.
      • Joyal J.
      • Milot É.
      • Kaufman R.
      • Guimond M.
      • Kennedy T.
      • Sapieha P.
      Neuronal ER stress impedes myeloid-cell-induced vascular regeneration through ire1-alpha degradation of netrin-1.
      However if unchecked, sustained hypoxia and inflammation can result in suppression of VEGF and secretion of vasorepulsive cue Sema3A, in addition to activation of endoplasmic reticulum stress pathways that translates into an inositol-requiring enzyme-1α–dependent cleavage of Netrin1 mRNA.
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      • Binet F.
      • Mawambo G.
      • Sitaras N.
      • Tetreault N.
      • Lapalme E.
      • Favret S.
      • Cerani A.
      • Leboeuf D.
      • Tremblay S.
      • Rezende F.
      • Juan A.
      • Stahl A.
      • Joyal J.
      • Milot É.
      • Kaufman R.
      • Guimond M.
      • Kennedy T.
      • Sapieha P.
      Neuronal ER stress impedes myeloid-cell-induced vascular regeneration through ire1-alpha degradation of netrin-1.
      Together, these molecular events divert metabolic stores away from perishing tissue to preserve more viable retinal regions.
      • Joyal J.S.
      • Sitaras N.
      • Binet F.
      • Rivera J.C.
      • Stahl A.
      • Zaniolo K.
      • Shao Z.
      • Polosa A.
      • Zhu T.
      • Hamel D.
      • Djavari M.
      • Kunik D.
      • Honoré J.C.
      • Picard E.
      • Zabeida A.
      • Varma D.R.
      • Hickson G.
      • Mancini J.
      • Klagsbrun M.
      • Costantino S.
      • Beauséjour C.
      • Lachapelle P.
      • Smith L.E.
      • Chemtob S.
      • Sapieha P.
      Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
      Once perishing tissues have been sequestered, mechanisms to prevent the relentless inflammatory destruction of the retina must exist to initiate neurovascular retinal repair. We found that amplified IL-1β signals for increased Par2 expression in RGCs, likely as an attempt to overcome excessive vascular decay and initiate revascularization. In this regard, neuronal Par2 functions as an intrinsic checkpoint to avoid excess proinflammatory cytokine signaling. Specifically, Par2 exerts its beneficial effects by suppressing IL-1RI expression in RGCs via an Erk/Jnk-dependent pathway, which in turn limits IL-1β–induced Sema3A secretion, despite high IL-1β retinal levels during OIR. Altogether this study highlights a central role for Par2 and uncovers an important regulatory mechanism in neurons that avoids excessive vascular decay and reestablishes normal retinal revascularization in these conditions.
      Par2 is ubiquitously expressed in various organs and can have distinct cellular functions depending on the cell type where it is expressed.
      • Bucci M.
      • Roviezzo F.
      • Cirino G.
      Protease-activated receptor-2 (PAR2) in cardiovascular system.
      • Ramachandran R.
      • Hollenberg M.D.
      Proteinases and signalling: pathophysiological and therapeutic implications via PARs and more.
      We found that Par2 in the retina is most prominently localized on RGCs and to a lesser extent on endothelium, where it exhibits functionally distinct cellular processes. Tissue factor, which is expressed on endothelium (but not on neurons), has been found, when constitutively activated, to robustly stimulate angiogenesis of the vascular plexus during development via Par2
      • Belting M.
      • Dorrell M.I.
      • Sandgren S.
      • Aguilar E.
      • Ahamed J.
      • Dorfleutner A.
      • Carmeliet P.
      • Mueller B.M.
      • Friedlander M.
      • Ruf W.
      Regulation of angiogenesis by tissue factor cytoplasmic domain signaling.
      ; however, in OIR, endothelial Par2 activation by tissue factor fails to curb oxygen-induced vaso-obliteration and subsequent pathologic neovascularization.
      • Uusitalo-Jarvinen H.
      • Kurokawa T.
      • Mueller B.M.
      • Andrade-Gordon P.
      • Friedlander M.
      • Ruf W.
      Role of protease activated receptor 1 and 2 signaling in hypoxia-induced angiogenesis.
      Correspondingly, we found that activation of Par2 on endothelial cells increases IL-1RI (and IL-1β
      • Ritchie E.
      • Saka M.
      • MacKenzie C.
      • Drummond R.
      • Wheeler-Jones C.
      • Kanke T.
      • Plevin R.
      Cytokine upregulation of proteinase-activated-receptors 2 and 4 expression mediated by p38 MAP kinase and inhibitory kappa B kinase β in human endothelial cells.
      ) expression, which may potentiate vascular injury,
      • Rivera J.C.
      • Sitaras N.
      • Noueihed B.
      • Hamel D.
      • Madaan A.
      • Zhou T.
      • Honore J.
      • Quiniou C.
      • Joyal J.
      • Hardy P.
      • Sennlaub F.
      • Lubell W.
      • Chemtob S.
      Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
      and does so seemingly via distinct intracellular signaling pathways. Neuronal Par2, on the other hand, dampens oxygen-induced endothelial cell death by curbing IL-1β–mediated Sema3A release while maintaining elevated levels of VEGF in the retina, thus allowing normal regrowth of intraretinal neovessels (and consequently avoiding extraretinal neovessel formation). The exact reason why this dichotomy exists in the retina, however, remains elusive. We hypothesize that in pathologic conditions (as seen in ischemic retinopathy), endothelial Par2 does not account for the exaggerated immunologic response and, rather than counteracting this response, exacerbates vascular decay, whereas neuronal Par2 functions as a regulator of inflammation to avoid excessive neurovascular degeneration. In the brain and other organs, a similar dichotomy is apparent; Par2 activation can have opposing effects, depending on the tissue or cell where it is expressed.
      • Luo W.
      • Wang Y.
      • Reiser G.
      Protease-activated receptors in the brain: receptor expression, activation, and functions in neurodegeneration and neuroprotection.
      • Ramachandran R.
      • Noorbakhsh F.
      • DeFea K.
      • Hollenberg M.D.
      Targeting proteinase-activated receptors: therapeutic potential and challenges.
      For instance, experimental models of neurodegenerative disorders reveal that Par2 on glial cells evoke distinct cellular mechanisms from neuronal Par2 to initiate neurodegeneration and neuroprotection, respectively.
      • Noorbakhsh F.
      • Vergnolle N.
      • McArthur J.C.
      • Silva C.
      • Andrade-Gordon P.
      • Vodjgani M.
      • Hollenberg M.
      • Power C.
      Protease-activated receptor-2 induction by neuroinflammation prevents neuronal death during HIV infection.
      • Afkhami-Goli A.
      • Noorbakhsh F.
      • Keller A.J.
      • Vergnolle N.
      • Westaway D.
      • Jhamandas J.H.
      • Andrade-Gordon P.
      • Hollenberg M.D.
      • Arab H.
      • Dyck R.H.
      • Power C.
      Proteinase-activated receptor-2 exerts protective and pathogenic cell type-specific effects in Alzheimer's disease.
      • Noorbakhsh F.
      • Tsutsui S.
      • Vergnolle N.
      • Boven L.A.
      • Shariat N.
      • Vodjgani M.
      • Warren K.G.
      • Andrade-Gordon P.
      • Hollenberg M.D.
      • Power C.
      Proteinase-activated receptor 2 modulates neuroinflammation in experimental autoimmune encephalomyelitis and multiple sclerosis.
      • Jin G.
      • Hayashi T.
      • Kawagoe J.
      • Takizawa T.
      • Nagata T.
      • Nagano I.
      • SyojI M.
      • Abe K.
      Deficiency of PAR-2 gene increases acute focal ischemic brain injury.
      • Wang H.
      • Ubl J.
      • Reiser G.
      Four subtypes of protease-activated receptors, co-expressed in rat astrocytes, evoke different physiological signaling.
      Importantly, this study also highlights the drawback associated with embryonic gene deletion. As previously reported,
      • Uusitalo-Jarvinen H.
      • Kurokawa T.
      • Mueller B.M.
      • Andrade-Gordon P.
      • Friedlander M.
      • Ruf W.
      Role of protease activated receptor 1 and 2 signaling in hypoxia-induced angiogenesis.
      OIR-induced vaso-obliteration and neovascularization were unaltered in mice having embryonic deletion of the Par2 gene. We found that conditionally knocking down Par2 in wild-type mice retina invokes a distinct vascular phenotype compared with Par2−/− mice, whereupon they are exposed to OIR. Furthermore, mice conditionally knocked down for Par2 display divergent molecular signatures with regard to Par2−/− mice, indicating that embryonic gene deletion may, in some circumstances, unobtrusively induce compensatory mechanisms that may guise the native function of genes under investigation.
      Overall, our findings uncover an important unprecedented functional mechanism induced by inflammation to limit vascular injury. We hereby identify Par2 as an important physiologic mediator that regulates the inflammatory hypoxic-ischemic retinal environment in ischemic retinopathies, thus allowing for normal vascular regeneration of the retina. To date considerable attention has been focused on inhibiting pathologic neovascularization,
      • Hellström A.
      • Smith L.E.
      • Dammann O.
      Retinopathy of prematurity.
      • Miller J.
      • Couter J.L.
      • Strauss E.
      • Ferrara N.
      Vascular endothelial growth factor A in intraocular vascular disease.
      • Wang S.
      • Park J.
      • Duh E.
      Novel targets against retinal angiogenesis in diabetic retinopathy.
      • Hernandez C.
      • Simo R.
      Strategies for blocking angiogenesis in diabetic retinopathy: from basic science to clinical practice.
      whereby anti-VEGF administration is currently accepted as a possible treatment regimen for patients with ischemic retinopathies.
      • Stewart M.W.
      The expanding role of vascular endothelial growth factor inhibitors in ophthalmology.
      However, there are no clinical treatments that prompt desirable revascularization in an effort to reduce ischemic stress that is crucial to neuroretinal dysfunction and preretinal neovascularization. The current study offers an approach for alleviating ischemic stress using Par2 agonists.

      Acknowledgments

      We acknowledge Hensy Fernandez and Isabelle Lahaie for valuable technical assistance. We thank Dr. Christian Beauséjour for providing us with Lv constructs.
      N.S., J.C.R., J.-S.J., and S.C. conceived and designed the experiments; N.S., J.C.R., B.N., M.B.-A., K.Z., S.O., D.H., and T.Z. performed the experiments; N.S., J.C.R., and S.C. analyzed the data; J.C.R., P.H., P.S., and J.-S.J. provided expert advice; N.S. and J.C.R prepared the figures; N.S., J.C.R., and S.C. wrote the manuscript.

      Supplemental Data

      • Supplemental Figure S1

        Specificity of SAM11 proteinase-activated receptor (Par)-2 mouse antibody. RT-PCR (A) and Western blot (B) data from postoperative day 12 (P12) Par2 intact (wild-type) and transgenic (Par2−/−) mice retina showing the presence or absence of the Par2 gene (F2rl1) and protein (using SAM11 antibody), respectively. C: Comparison of wild-type and Par2 null P8 mice retina stained with SAM11 Par2 antibody. Par2−/− mice are devoid of immunofluorescent signal using SAM11 antibody. Scale bar = 30 μm. Original magnification, ×100 (C).

      • Supplemental Figure S2

        Proteinase-activated receptor (Par)-2 expression in neural retina. A: High-magnification immunohistochemical images of sagittal sections from postoperative day 8 (P8) wild-type mice showing colocalization of Par2 (green) in βIII-tubulin–positive retinal ganglion cells (red). B: Par2 (green) also colocalizes with endothelial cells (CD31; red) albeit to a lesser extent. Nuclei are counterstained with DAPI (blue). C: Cross sectional image of a postoperative day 8 mouse retina depicting the delineated areas captured using laser-capture microdissection. Vessels are stained with isolectin B4 (red), whereas nuclei are counterstained with DAPI (blue). D: Expression profile of laser dissected retinal layers probed for various retinal cell markers CD31 (vessels), Thy1.1 (GCL), βIII-tubulin (GCL and INL), and rhodopsin (ONL). Scale bars: 30 μm (A and B); 50 μm (C). Original magnification: ×600 (A and B); ×300 (C). GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; ND, not detected; NFL, nerve fiber layer.

      • Supplemental Figure S3

        Proteinase-activated receptor (Par)-2 intact and transgenic mice share similar vascular phenotypes in oxygen-induced retinopathy; shRNA-bearing lentiviral (Lv) constructs target Par2 in retinal ganglion cells. A: Representative flat mounts of isolectin B4–stained retinas from Par2 intact [wild-type (WT)] and transgenic (Par2−/−) mice at postoperative day 12 (P12), P15, and P17 previously exposed to hyperoxia from P7 to P12. Vaso-obliteration and neovascularization areas are outlined with solid blue lines and yellow-hashed lines, respectively. Avascular areas and neovascularization are shown in the right panels; no significant differences were observed between mice populations. B: Immunohistochemical staining showing colocalization of green fluorescent protein (GFP) (green) with retinal ganglion cells (RGCs) (βIII-tubulin; red) in P8 WT mice retina after intravitreal injection (postoperative day 3) with GFP reporter Lv (approximately 15.0 ng). Nuclei are counterstained with DAPI (blue). C: Western blot analysis of whole retina from P7 mice demonstrating effective knockdown of Par2 after intravitreal injection with lentiviral shPar2; contralateral eyes injected with lentiviral shGFP served as control (two retinas per group). n = 5 to 12 retinas (A). Scale bar = 30 μm (B). Original magnification 100× (A); ×600 (B). GCL, ganglion cell layer; IPL, inner plexiform layer.

      • Supplemental Figure S4

        Proteinase-activated receptor (Par)-2 transgenic mice do not respond to SLIGRL or Lv shRNA Par2 treatment. A: Representative flat mounts from isolectin B4–stained postoperative day 12 Par2−/− mice exposed to oxygen-induced retinopathy (OIR) and injected intravitreally at postoperative day 7 (P7) with vehicle (phosphate-buffered saline), 100 μmol SLIGRL, approximately 8.5 ng of Lv shGFP, or approximately 9.6 ng of Lv shPar2 show no change in vaso-obliteration (VO). B: Similar effects were observed in P17 Par2−/− OIR mice injected intravitreally at P12 showing no change in VO or neovascularization (NV). VO and NV areas are outlined with solid blue lines and yellow-hashed lines, respectively. n = 7 to 9 retinas (A); n = 7 to 8 retinas (B). Original magnification, ×100 (A and B).

      • Supplemental Figure S5

        Anakinra treatment in vivo successfully dampens interleukin-1β (IL-1β) in whole retina and proteinase-activated receptor (Par)-2 expression in various cell lines. A: Anakinra i.p. administration translated to decreased IL-1β mRNA in postoperative day 8 wild-type mice whole retina exposed to oxygen-induced retinopathy compared with saline-treated animals. B: Cell lysates from rat brain microvascular endothelial cells (RBMVECs), retinal ganglion cells RGC-5 cell line, and rat cerebral astrocytes (RCA) immunoblotted with Par2 SAM11 antibody. RCA devoid of Par2 mRNA do not reveal a band at 55 kDa. C: Par2 expression in RBMVECs, RGC-5, and RCA; Par2 mRNA is detected in RBMVECs and RGC-5 only. ∗∗∗P < 0.001 (A). n = 3 (A).

      • Supplemental Figure S6

        Proteinase-activated receptor (Par)-2–activating peptide SLIGRL treatment in normoxia raised mice pups; Par2 transgenic mice respond to interleukin-1 receptor antagonist (IL-1Ra) but not to SLIGRL treatment. Western blot (A) and immunohistochemical (B) analysis of retinas from normoxia raised postoperative day 9 (P9) wild-type (WT) mice injected with vehicle or 100 μmol SLIGRL (contralateral eye) demonstrate slight decreases in IL-1 receptor type I (IL-1RI) levels. Il-1RI stained in green, retinal ganglion cells in red (βIII-tubulin), and nuclei are counterstained in blue; vessels are visualized with isolection B4 staining (magenta). C: Representative flat mounts from isolectin B4–stained P12 Par2−/− exposed to oxygen-induced retinopathy and injected intravitreally at P7 with vehicle, 100 μmol SLIGRL, 10 μmol IL-1Ra, or both SLIGRL and IL-1Ra. Vaso-obliteration (VO) areas are outlined with solid blue lines. D: Western blot analysis of P9 mice retina from WT and Par2−/− mice showing an effective increase in IL-1RI on exposure to oxygen-induced retinopathy; however, there is no variation in IL-1RI levels in Par2−/− compared with WT mice retina. Densitometry quantification at right. n = 6 to 8 retina (C); n = 3 independent experiments (D). P < 0.05 (D), ∗∗P < 0.01 (C and D), and ∗∗∗P < 0.001 (C and D). Scale bar = 50 μm (B). Original magnification: ×300 (B and C). GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; NFL, nerve fiber layer; ONL, outer nuclear layer; OPL, outer plexiform layer.

      • Supplemental Figure S7

        Proteinase-activated receptor (Par)-2–induced reduction of Irak1 phosphorylation in retinal ganglion cell (RGC)-5 cell line; Par2-mediated regulation of interleukin-1 receptor type I (IL-1RI) in rat brain microvascular endothelial cells (RBMVECs). A: Control RGC-5 pretreated with 100 μmol/L SLIGRL successfully abolish IL-1β–dependent Irak1 phosphorylation, whereas Par2-deficient RGC-5 does not. The bar graph shows densitometry quantification. B: RGC-5 infected with Lv short hairpin Par2 (shPar2) effectively decreased Par2 protein levels; Lv shGFP–infected RGC-5 served as controls. C: Par2-expressing RBMVECs display dose-dependent increases in IL-1RI mRNA after 1-hour stimulation with SLIGRL as analyzed by quantitative PCR. D: RBMVECs had increased IL-1RI protein levels after 100 μmol/L SLIGRL stimulation at 2 hours only. E: Western blot analysis of postoperative day 9 mice retina from wild-type (WT) and Par2−/− mice showing an effective increase in Erk1/2 phosphorylation on exposure to oxygen-induced retinopathy; however, there is no variation in phospho-Erk1/2 levels in Par2−/− compared with WT mice retina. The bar graph shows densitometry quantification. F: RBMVECs pretreated with the MET inhibitor PD98059 have no increase in SLIGRL-dependent Erk1/2 phosphorylation. G: Erk1/2 inhibition using PD98059 does not affect SLIGRL-dependent increases in IL-1RI mRNA in RBMVECs. n = 8 (C); n = 3 independent experiments (D); n = 6 (G). P < 0.05 (E), ∗∗P < 0.01 (C), and ∗∗∗P < 0.001 (C and G).

      • Supplemental Figure S8

        Proteinase-activated receptor (Par)-2 stimulates production of vascular endothelial growth factor (VEGF)-A in retinal ganglion cells and promotes VEGF-A–dependent vascular sprouting. A: Western blot analysis showing effective down-regulation of VEGF-A in whole retina from postoperative day 9 (P9) wild-type (WT) mice exposed to oxygen-induced retinopathy (OIR). Intravitreal injection of 100 μmol SLIGRL, however, increases VEGF-A levels compared with vehicle-treated animals. Densitometry quantification is on the right. B: Immunohistochemical analysis of coronal sections from P9 WT mice retina exposed to OIR and injected intravitreally with vehicle or SLIGRL. Whereas hyperoxia treatment decreases VEGF-A (green) immunoreactivity in the retina proper, intravitreal administration of SLIGRL augments VEGF-A specifically in the βIII-tubulin–expressing retinal ganglion cells (RGCs) (red). Nuclei are counterstained in blue; vessels are stained with isolection B4 (magenta). C: Control lentiviral (Lv) shGFP–infected RGC-5 samples treated with SLIGRL exhibit increased VEGF-A mRNA compared with vehicle controls after 24 hours; Lv shPar2-infected RGC-5, however, reveals no increases in VEGF-A mRNA after treatment with SLIGRL. D: Aortic explants stimulated with conditioned media from SLIGRL-stimulated RGC-5 reveal increased vascular sprouting; however, treatment with neutralizing antibody to VEGF-A abrogates SLIGRL-induced aortic sprouting. n = 3 independent experiments (A); n = 8 (C); n = 4 to 5 (D). P < 0.05 (A, C, and D). Scale bar = 50 μm (B). Original magnification: ×300 (B); ×40 (D). GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; NFL, nerve fiber layer; ONL, outer nuclear layer; OPL, outer plexiform layer.

      References

        • Yau J.W.Y.
        • Rogers S.L.
        • Kawasaki R.
        • Lamoureux E.L.
        • Kowalski J.W.
        • Bek T.
        • et al.
        • The Meta-Analysis for Eye Disease (Meta-eye) Study Group
        Global prevalence and major risk factors of diabetic retinopathy.
        Diabetes Care. 2012; 35: 556-564
        • Gilbert C.
        • Rahi J.
        • Eckstein M.
        • O'Sullivan J.
        • Foster A.
        Retinopathy of prematurity in middle-income countries.
        Lancet. 1997; 350: 12-14
        • Sapieha P.
        • Hamel D.
        • Shao Z.
        • Rivera J.C.
        • Zaniolo K.
        • Joyal J.S.
        • Chemtob S.
        Proliferative retinopathies: angiogenesis that blinds.
        Int J Biochem Cell Biol. 2010; 42: 5-12
        • Fulton A.
        • Hansen R.
        • Moskowitz A.
        • Akula J.
        The neurovascular retina in retinopathy of prematurity.
        Prog Retin Eye Res. 2009; 28: 452-482
        • Demircan N.
        • Safran B.G.
        • Soylu M.
        • Ozcan A.A.
        • Sizmaz S.
        Determination of vitreous interleukin-1 (il-1) and tumour necrosis factor (tnf) levels in proliferative diabetic retinopathy.
        Eye (Lond). 2006; 20: 1366-1369
        • Mocan M.C.
        • Kadayifcilar S.
        • Eldem B.
        Elevated intravitreal interleukin-6 levels in patients with proliferative diabetic retinopathy.
        Can J Ophthalmol. 2011; 41: 747-752
        • Krady J.K.
        • Basu A.
        • Allen C.M.
        • Xu Y.
        • LaNoue K.F.
        • Gardner T.W.
        • Levison S.W.
        Minocycline reduces proinflammatory cytokine expression, microglial activation, and caspase-3 activation in a rodent model of diabetic retinopathy.
        Diabetes. 2005; 54: 1559-1565
        • Rivera J.C.
        • Sitaras N.
        • Noueihed B.
        • Hamel D.
        • Madaan A.
        • Zhou T.
        • Honore J.
        • Quiniou C.
        • Joyal J.
        • Hardy P.
        • Sennlaub F.
        • Lubell W.
        • Chemtob S.
        Microglia and IL-1β in ischemic retinopathy elicit microvascular degeneration through neuronal Semaphorin3A.
        Arterioscler Thromb Vasc Biol. 2013; 133: 1881-1891
        • dell'Omo R.
        • Semeraro F.
        • Bamonte G.
        • Cifariello F.
        • Romano M.R.
        • Costagliola C.
        Vitreous mediators in retinal hypoxic diseases.
        Mediators Inflamm. 2013; : 1-16
        • Ishida S.
        • Yamashiro K.
        • Usui T.
        • Kaji Y.
        • Ogura Y.
        • Hia T.
        • Honda Y.
        • Oguchi Y.
        • Adamis A.
        Leukocytes mediate retinal vascular remodeling during dvelopment and vaso-obliteration in disease.
        Nat Med. 2003; 9: 781-789
        • Joyal J.S.
        • Sitaras N.
        • Binet F.
        • Rivera J.C.
        • Stahl A.
        • Zaniolo K.
        • Shao Z.
        • Polosa A.
        • Zhu T.
        • Hamel D.
        • Djavari M.
        • Kunik D.
        • Honoré J.C.
        • Picard E.
        • Zabeida A.
        • Varma D.R.
        • Hickson G.
        • Mancini J.
        • Klagsbrun M.
        • Costantino S.
        • Beauséjour C.
        • Lachapelle P.
        • Smith L.E.
        • Chemtob S.
        • Sapieha P.
        Ischemic neurons prevent vascular regeneration of neural tissue by secreting semaphorin 3A.
        Blood. 2011; 117: 6024-6035
        • Dorfman A.
        • Dembinska O.
        • Chemtob S.
        • Lachapelle P.
        Early manifestations of postnatal hyperoxia on the retinal structure and function of the neonatal rat.
        Invest Ophthalmol Vis Sci. 2008; 49: 458-466
        • Teshima S.
        • Nakanishi H.
        • Nishizawa M.
        • Kitagawa K.
        • Kaibori M.
        • Yamada M.
        • Habara K.
        • Kwon A.
        • Kamiyama Y.
        • Ito S.
        • Okumura T.
        Up-regulation of IL-1 receptor through PI3K/Akt is essential for the induction of iNOS gene expression in hepatocytes.
        J Hepatol. 2004; 40: 616-623
        • Skundric D.S.
        • Bealmear B.
        • Lisak R.P.
        Induced upregulation of IL-1, IL-1RA and IL-1R type I gene expression by Schwann cells.
        J Neuroimmunol. 1997; 74: 9-18
        • Kowluru R.
        • Zhong Q.
        • Santos J.
        Matrix metalloproteinases in diabetic retinopathy: potential role of MMP-9.
        Expert Opin Investig Drugs. 2012; 21: 797-805
        • Soh U.J.K.
        • Dores M.R.
        • Chen B.
        • Trejo J.
        Signal transduction by protease-activated receptors.
        Br J Pharmacol. 2010; 160: 191-203
        • Uusitalo-Jarvinen H.
        • Kurokawa T.
        • Mueller B.M.
        • Andrade-Gordon P.
        • Friedlander M.
        • Ruf W.
        Role of protease activated receptor 1 and 2 signaling in hypoxia-induced angiogenesis.
        Arterioscler Thromb Vasc Biol. 2007; 27: 1456-1462
        • Belting M.
        • Dorrell M.I.
        • Sandgren S.
        • Aguilar E.
        • Ahamed J.
        • Dorfleutner A.
        • Carmeliet P.
        • Mueller B.M.
        • Friedlander M.
        • Ruf W.
        Regulation of angiogenesis by tissue factor cytoplasmic domain signaling.
        Nat Med. 2004; 10: 502-509
        • Tsai S.H.
        • Sheu M.T.
        • Liang Y.C.
        • Cheng H.T.
        • Fang S.S.
        • Chen C.
        TGF-β inhibits IL-1β-activated PAR-2 expression through multiple pathways in human primary synovial cells.
        J Biomed Sci. 2009; 16: 97-111
        • Ritchie E.
        • Saka M.
        • MacKenzie C.
        • Drummond R.
        • Wheeler-Jones C.
        • Kanke T.
        • Plevin R.
        Cytokine upregulation of proteinase-activated-receptors 2 and 4 expression mediated by p38 MAP kinase and inhibitory kappa B kinase β in human endothelial cells.
        Br J Pharmacol. 2007; 150: 1044-1054
        • Boileau C.
        • Amiable N.
        • Martel-Pelletier J.
        • Fahmi H.
        • Duval N.
        • Pelletier J.
        Activation of proteinase-activated receptor 2 in human osteoarthritic cartilage upregulates catabolic and proinflammatory pathways capable of inducing cartilage degradation: a basic science study.
        Arthritis Res Ther. 2007; 9: R121
        • Noorbakhsh F.
        • Vergnolle N.
        • McArthur J.C.
        • Silva C.
        • Andrade-Gordon P.
        • Vodjgani M.
        • Hollenberg M.
        • Power C.
        Protease-activated receptor-2 induction by neuroinflammation prevents neuronal death during HIV infection.
        J Immunol. 2005; 174: 7320-7329
        • Suharti C.
        • van Gorp E.C.
        • Setiati T.E.
        • Dolmans W.M.
        • Djokomoeljanto R.J.
        • Hack C.E.
        • Hugo ten C.
        • van der Meer J.W.
        The role of cytokines in activation of coagulation and fibrinolysis in dengue shock syndrome.
        Thromb Haemost. 2002; 87: 42-46
        • Smith L.E.
        • Wesoloiuski E.
        • McLellan A.
        • Kostyk S.K.
        • D'Amato R.
        • Sullivan R.
        • D'Amore P.A.
        Oxygen-induced retinopathy in the mouse.
        Invest Ophthalmol Vis Sci. 1994; 35: 101-111
        • Stahl A.
        • Connor K.M.
        • Sapieha P.
        • Chen J.
        • Dennison R.J.
        • Krah N.M.
        • Seaward M.R.
        • Willett K.L.
        • Aderman C.M.
        • Guerin K.I.
        • Hua J.
        • Lofqvist C.
        • Hellstrom A.
        • Smith L.E.
        The mouse retina as an angiogenesis model.
        Invest Ophthalmol Vis Sci. 2010; 51: 2813-2826
        • Stahl A.
        • Connor K.M.
        • Sapieha P.
        • Willett K.L.
        • Krah N.M.
        • Dennison R.J.
        • Chen J.
        • Guerin K.I.
        • Smith L.E.
        Computer-aided quantification of retinal neovascularization.
        Angiogenesis. 2009; 12: 297-301
        • Dull T.
        • Zufferey R.
        • Kelly M.
        • Mandel R.J.
        • Nguyen M.
        • Trono D.
        • Naldini L.
        A third-generation lentivirus vector with a conditional packaging system.
        J Virol. 1998; 72: 8463
        • Sapieha P.
        • Sirinyan M.
        • Hamel D.
        • Zaniolo K.
        • Joyal J.S.
        • Cho J.H.
        • Honoré J.C.
        • Kermorvant-Duchemin E.
        • Varma D.R.
        • Tremblay S.
        • Leduc M.
        • Rihakova L.
        • Hardy P.
        • Klein W.H.
        • Mu X.
        • Mamer O.
        • Lachapelle P.
        • Di Polo A.
        • Beauséjour C.
        • Andelfinger G.
        • Mitchell G.
        • Sennlaub F.
        • Chemtob S.
        The succinate receptor GPR91 in neurons has a major role in retinal angiogenesis.
        Nat Med. 2008; 14: 1067-1076
        • Chang-Ling T.
        • Neill A.L.
        • Hunt N.H.
        Early microvascular changes in murine cerebral malaria detected in retinal wholemounts.
        Am J Pathol. 1992; 140: 1121-1130
        • Barbaric I.
        • Miller G.
        • Dear T.
        Appearances can be deceiving: phenotypes of knockout mice.
        Brief Funct Genomic Proteomic. 2007; 6: 91-103
        • Schmidlin F.
        • Amadesi S.
        • Dabbagh K.
        • Lewis D.E.
        • Knott P.
        • Bunnet N.W.
        • Gater P.R.
        • Geppetti P.
        • Bertrand C.
        • Stevens M.E.
        Protease-activated receptor 2 mediates eosinophil infiltration and hyperreactivity in allergic inflammation of the airway.
        J Immunol. 2002; 169: 5315-5321
        • Binet F.
        • Mawambo G.
        • Sitaras N.
        • Tetreault N.
        • Lapalme E.
        • Favret S.
        • Cerani A.
        • Leboeuf D.
        • Tremblay S.
        • Rezende F.
        • Juan A.
        • Stahl A.
        • Joyal J.
        • Milot É.
        • Kaufman R.
        • Guimond M.
        • Kennedy T.
        • Sapieha P.
        Neuronal ER stress impedes myeloid-cell-induced vascular regeneration through ire1-alpha degradation of netrin-1.
        Cell Metab. 2013; 17: 353-371
        • Cerani A.
        • Tetreault N.
        • Menard C.
        • Lapalme E.
        • Patel C.
        • Sitaras N.
        • Beaudoin F.
        • Leboeuf D.
        • De Guire V.
        • Binet F.
        • Dejda A.
        • Rezende F.
        • Miloudi K.
        • Sapieha P.
        Neuron-derived Semaphorin 3A is an early inducer of vascular permeability in diabetic retinopathy via neuropilin-1.
        Cell Metab. 2013; 18: 505-518
        • Kaufmann R.
        • Hollenberg M.
        Proteinase-activated receptors (PARs) and calcium signaling in cancer.
        Adv Exp Med Biol. 2012; 740: 979-1000
        • Kim Y.
        • West X.Z.
        • Byzova T.V.
        Inflammation and oxidative stress in angiogenesis and vascular disease.
        J Mol Med. 2013; 91: 323-328
        • Gardiner T.A.
        • Gibson D.S.
        • de Gooyer T.E.
        • de la Cruz V.F.
        • McDonald D.M.
        • Stitt A.
        Inhibition of tumor necrosis factor-α improves physiological angiogenesis and reduces pathological neovascularization in ischemic retinopathy.
        Am J Pathol. 2005; 166: 637-644
        • Meyer N.
        • Christoph J.
        • Makrinioti H.
        • Indermitte P.
        • Rhyner C.
        • Soyka M.
        • Eiwegger T.
        • Chalubinski T.
        • Wanke K.
        • Fujita H.
        • Wawrzyniak P.
        • Burgler S.
        • Zhang S.
        • Akdis M.
        • Menz G.
        • Akdis C.
        Inhibition of angiogenesis by IL-32: possible role in asthma.
        J Allergy Clin Immunol. 2012; 129: 964-973.e7
        • Massague J.
        • Blain S.
        • Lo R.
        TGF-beta signaling in growth control, cancer, and heritable disorders.
        Cell. 2000; 103: 295-309
        • Kermorvant-Duchemin E.
        • Sennlaub F.
        • Sirinyan M.
        • Brault S.
        • Andelfinger G.
        • Kooli A.
        • Germain S.
        • Ong H.
        • d'Orleans-Juste P.
        • Gobeil F.
        • Zhu T.
        • Boisvert C.
        • Hardy P.
        • Jain K.
        • Falck J.R.
        • Balazy M.
        • Chemtob S.
        Trans-arachidonic acids generated during nitrative stress induce a thrombospondin-1–dependent microvascular degeneration.
        Nat Med. 2005; 11: 1339-1345
        • Minuzzo S.
        • Moserle L.
        • Indraccolo S.
        • Amadori A.
        Angiogenesis meets immunology: cytokine gene therapy of cancer.
        Mol Aspects Med. 2007; 28
        • Shen H.
        • Yao P.
        • Lee E.
        • Greenhalgh D.
        • Soulika A.
        Interferon-gamma inhibits healing post scald burn injury.
        Wound Repair Regen. 2012; 20: 580-591
        • Rothmeier A.
        • Ruf W.
        Protease-activated receptor 2 signaling in inflammation.
        Semin Immunopathol. 2012; 34: 133-149
        • Ossovskaya V.S.
        • Nunnet N.
        Protease-activated receptors: contribution to physiology and disease.
        Physiol Rev. 2004; 84: 579-621
        • Paumelle R.
        • Tulasne D.
        • Leroy C.
        • Coll J.
        • Vandenbunder B.
        • Fafeur V.
        Sequential activation of ERK and repression of JNK by scatter factor/hepatocyte growth factor in Madin-Darby canine kidney epithelial cells.
        Mol Biol Cell. 2000; 11: 3751-3763
        • Tremblay S.
        • Miloudi K.
        • Chaychi S.
        • Favret S.
        • Binet F.
        • Polosa A.
        • Lachapelle P.
        • Chemtob S.
        • Sapieha P.
        Systemic inflammation perturbs developmental retinal angiogenesis and neuroretinal function.
        Invest Ophthalmol Vis Sci. 2013; 54: 8125-8139
        • Lange C.
        • Ehlken C.
        • Stahl A.
        • Martin G.
        • Hansen L.
        • Agostini H.
        Kinetics of retinal vaso-obliteration and neovascularisation in the oxygen-induced retinopathy (OIR) mouse model.
        Graefes Arch Clin Exp Ophthalmol. 2009; 247: 1205-1211
        • Bucci M.
        • Roviezzo F.
        • Cirino G.
        Protease-activated receptor-2 (PAR2) in cardiovascular system.
        Vascul Pharmacol. 2005; 43: 247-253
        • Ramachandran R.
        • Hollenberg M.D.
        Proteinases and signalling: pathophysiological and therapeutic implications via PARs and more.
        Br J Pharmacol. 2008; 153: S263-S282
        • Luo W.
        • Wang Y.
        • Reiser G.
        Protease-activated receptors in the brain: receptor expression, activation, and functions in neurodegeneration and neuroprotection.
        Brain Res Rev. 2007; 56: 331-345
        • Ramachandran R.
        • Noorbakhsh F.
        • DeFea K.
        • Hollenberg M.D.
        Targeting proteinase-activated receptors: therapeutic potential and challenges.
        Nat Rev Drug Discov. 2012; 11: 69-86
        • Afkhami-Goli A.
        • Noorbakhsh F.
        • Keller A.J.
        • Vergnolle N.
        • Westaway D.
        • Jhamandas J.H.
        • Andrade-Gordon P.
        • Hollenberg M.D.
        • Arab H.
        • Dyck R.H.
        • Power C.
        Proteinase-activated receptor-2 exerts protective and pathogenic cell type-specific effects in Alzheimer's disease.
        J Immunol. 2007; 179: 5493-5503
        • Noorbakhsh F.
        • Tsutsui S.
        • Vergnolle N.
        • Boven L.A.
        • Shariat N.
        • Vodjgani M.
        • Warren K.G.
        • Andrade-Gordon P.
        • Hollenberg M.D.
        • Power C.
        Proteinase-activated receptor 2 modulates neuroinflammation in experimental autoimmune encephalomyelitis and multiple sclerosis.
        J Exp Med. 2006; 203: 425-435
        • Jin G.
        • Hayashi T.
        • Kawagoe J.
        • Takizawa T.
        • Nagata T.
        • Nagano I.
        • SyojI M.
        • Abe K.
        Deficiency of PAR-2 gene increases acute focal ischemic brain injury.
        J Cereb Blood Flow Metab. 2005; 25: 302-313
        • Wang H.
        • Ubl J.
        • Reiser G.
        Four subtypes of protease-activated receptors, co-expressed in rat astrocytes, evoke different physiological signaling.
        Glia. 2002; 37: 53-63
        • Hellström A.
        • Smith L.E.
        • Dammann O.
        Retinopathy of prematurity.
        Lancet. 2013; 382: 1445-1457
        • Miller J.
        • Couter J.L.
        • Strauss E.
        • Ferrara N.
        Vascular endothelial growth factor A in intraocular vascular disease.
        Ophthalmology. 2012; 120: 106-114
        • Wang S.
        • Park J.
        • Duh E.
        Novel targets against retinal angiogenesis in diabetic retinopathy.
        Curr Diab Rep. 2012; 12: 355-363
        • Hernandez C.
        • Simo R.
        Strategies for blocking angiogenesis in diabetic retinopathy: from basic science to clinical practice.
        Expert Opin Investig Drugs. 2007; 16: 1209-1226
        • Stewart M.W.
        The expanding role of vascular endothelial growth factor inhibitors in ophthalmology.
        Mayo Clin Proc. 2012; 87: 77-88