Advertisement

Stress response protein REDD1 promotes diabetes-induced retinal inflammation by sustaining canonical NF-κB signaling

Open AccessPublished:October 26, 2022DOI:https://doi.org/10.1016/j.jbc.2022.102638
      Inflammation contributes to the progression of retinal pathology caused by diabetes. Here, we investigated a role for the stress response protein regulated in development and DNA damage response 1 (REDD1) in the development of retinal inflammation. Increased REDD1 expression was observed in the retina of mice after 16-weeks of streptozotocin (STZ)-induced diabetes, and REDD1 was essential for diabetes-induced pro-inflammatory cytokine expression. In human retinal MIO-M1 Müller cell cultures, REDD1 deletion prevented increased pro-inflammatory cytokine expression in response to hyperglycemic conditions. REDD1 deletion promoted nuclear factor erythroid-2-related factor 2 (Nrf2) hyperactivation; however, Nrf2 was not required for reduced inflammatory cytokine expression in REDD1-deficient cells. Rather, REDD1 enhanced inflammatory cytokine expression by promoting activation of nuclear transcription factor κB (NF-κB). In WT cells exposed to tumor necrosis factor α (TNFα), inflammatory cytokine expression was increased in coordination with activating transcription factor 4 (ATF4)-dependent REDD1 expression and sustained activation of NF-κB. In both Müller cell cultures exposed to TNFα and in the retina of STZ-diabetic mice, REDD1 deletion promoted inhibitor of κB (IκB) expression and reduced NF-κB DNA-binding activity. We found that REDD1 acted upstream of IκB by enhancing both K63-ubiquitination and auto-phosphorylation of IκB kinase complex. In contrast with STZ-diabetic REDD1+/+ mice, IκB kinase complex autophosphorylation and macrophage infiltration were not observed in the retina of STZ-diabetic REDD1-/- mice. The findings provide new insight into how diabetes promotes retinal inflammation and support a model wherein REDD1 sustains activation of canonical NF-κB signaling.

      Keywords

      Abbreviations:

      BCVA (best-corrected visual acuity), DR (diabetic retinopathy), HA (hemagglutinin), ICAM-1 (intercellular adhesion molecule 1), IKK (IκB kinase), IP (immunoprecipitate), IκB (inhibitor of IκB), LPS (lipopolysaccharide), NF-κB (nuclear transcription factor κB), pUb (polyubiquitin), STZ (streptozotocin)
      Diabetic retinopathy (DR) is a significant ocular complication caused by diabetes that can progress to blindness. Retinal complications often develop upon loss of glucose homeostasis, with an estimated 103.12 million individuals suffering from DR and 18.83 million patients having vision threatening retinopathy (
      • Teo Z.L.
      • Tham Y.-C.
      • Yu M.
      • Chee M.L.
      • Rim T.H.
      • Cheung N.
      • et al.
      Global prevalence of diabetic retinopathy and projection of burden through 2045: systematic review and meta-analysis.
      ). The pathogenesis of DR is complex and multi-factorial; however, it is well accepted that inflammation is a crucial factor in the progression of the retinal complications that are caused by diabetes (
      • Rübsam A.
      • Parikh S.
      • Fort P.E.
      Role of inflammation in diabetic retinopathy.
      ). Inflammation is a protective immune response designed to facilitate tissue repair. Nevertheless, the chronic pro-inflammatory state of diabetes contributes to development and progression of DR. In diabetic patients without clinically visible signs of DR, expression of pro-inflammatory adhesion molecules and chemokines, such as intercellular adhesion molecule 1 (ICAM-1), C-C motif chemokine ligand (CCL2, also known as MCP-1), and CCL5 (also known as RANTES) are increased and contribute to leukostasis (
      • Khalfaoui T.
      • Lizard G.
      • Ouertani-Meddeb A.
      Adhesion molecules (ICAM-1 and VCAM-1) and diabetic retinopathy in type 2 diabetes.
      ,
      • Vujosevic S.
      • Micera A.
      • Bini S.
      • Berton M.
      • Esposito G.
      • Midena E.
      Proteome analysis of retinal glia cells-related inflammatory cytokines in the aqueous humour of diabetic patients.
      ,
      • Vujosevic S.
      • Micera A.
      • Bini S.
      • Berton M.
      • Esposito G.
      • Midena E.
      Aqueous humor biomarkers of müller cell activation in diabetic eyes.
      ). This supports that the development of inflammation occurs early in disease progression. Indeed, clinical studies support the benefits of inhibiting specific pro-inflammatory molecules to address the development of neovascularization and macular edema in DR patients (
      • Stahel M.
      • Becker M.
      • Graf N.
      • Michels S.
      Systemic interleukin 1β inhibition in proliferative diabetic retinopathy: a prospective open-label study using canakinumab.
      ).
      An abundance of literature has implicated the transcription factor nuclear factor κ-light-chain enhancer of activated B cells (NF-κB) as a key regulator of immune function and inflammatory responses (
      • Liu T.
      • Zhang L.
      • Joo D.
      • Sun S.-C.
      NF-κB signaling in inflammation.
      ). The NF-κB family of transcription factors [p65 (RelA), c-Rel, RelB, p50, and p52] form homodimers/heterodimers that determine the expression of an array of pro-inflammatory molecules. Canonical NF-κB activation involves phosphorylation of inhibitor of κB (IκB) by IκB kinase (IKK), which promotes IκB proteasomal degradation to allow nuclear translocation of the NF-κB RelA/p50 dimer. Noncanonical activation of RelB/p52 is mediated by IKK-dependent processing of p100 (
      • Sun S.-C.
      The non-canonical NF-κB pathway in immunity and inflammation.
      ). NF-κB activation occurs in preclinical models of type 1 diabetes, and blocking its activity is beneficial in preventing DR pathology (
      • Kanwar M.
      • Chan P.-S.
      • Kern T.S.
      • Kowluru R.A.
      Oxidative damage in the retinal mitochondria of diabetic mice: possible protection by superoxide dismutase.
      ,
      • Noda K.
      • Nakao S.
      • Ishida S.
      • Ishibashi T.
      Leukocyte adhesion molecules in diabetic retinopathy.
      ,
      • Miller W.P.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      REDD1 activates a ROS-generating feedback loop in the retina of diabetic mice.
      ). Limited evidence also supports NF-κB activation in the retina of preclinical models of type 2 diabetes (
      • Ding Y.
      • Yuan S.
      • Liu X.
      • Mao P.
      • Zhao C.
      • Huang Q.
      • et al.
      Protective effects of astragaloside IV on db/db mice with diabetic retinopathy.
      ,
      • Tuzcu M.
      • Orhan C.
      • Muz O.E.
      • Sahin N.
      • Juturu V.
      • Sahin K.
      Lutein and zeaxanthin isomers modulates lipid metabolism and the inflammatory state of retina in obesity-induced high-fat diet rodent model.
      ). Although it was demonstrated nearly 2 decades ago that diabetes promotes NF-κB activation in the retina (
      • Romeo G.
      • Liu W.-H.
      • Asnaghi V.
      • Kern T.S.
      • Lorenzi M.
      Activation of nuclear factor-κb induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes.
      ), the specific signaling events that lead to canonical or noncanonical NF-κB activation in DR have never been thoroughly resolved.
      Diabetes promotes retinal expression of the stress response protein regulated in development and DNA damage 1 (REDD1) (
      • Miller W.P.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      REDD1 activates a ROS-generating feedback loop in the retina of diabetic mice.
      ,
      • Miller W.P.
      • Yang C.
      • Mihailescu M.L.
      • Moore J.A.
      • Dai W.
      • Barber A.J.
      • et al.
      Deletion of the Akt/mTORC1 repressor REDD1 prevents visual dysfunction in a rodent model of type 1 diabetes.
      ,
      • Miller W.P.
      • Sunilkumar S.
      • Giordano J.F.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      The stress response protein REDD1 promotes diabetes-induced oxidative stress in the retina by Keap1-independent Nrf2 degradation.
      ,
      • Miller W.P.
      • Toro A.L.
      • Sunilkumar S.
      • Stevens S.A.
      • VanCleave A.M.
      • Williamson D.L.
      • et al.
      Müller glial expression of REDD1 is required for retinal neurodegeneration and visual dysfunction in diabetic mice.
      ). REDD1 expression is dominant in retinal Müller glia, where the protein contributes to a failed adaptive response of the retina to diabetes that includes gliosis, neurodegeneration, and the development of functional deficits in vision (
      • Miller W.P.
      • Yang C.
      • Mihailescu M.L.
      • Moore J.A.
      • Dai W.
      • Barber A.J.
      • et al.
      Deletion of the Akt/mTORC1 repressor REDD1 prevents visual dysfunction in a rodent model of type 1 diabetes.
      ,
      • Miller W.P.
      • Toro A.L.
      • Sunilkumar S.
      • Stevens S.A.
      • VanCleave A.M.
      • Williamson D.L.
      • et al.
      Müller glial expression of REDD1 is required for retinal neurodegeneration and visual dysfunction in diabetic mice.
      ). Indeed, therapeutic administration of an siRNA for REDD1 knockdown has demonstrated promise for improving best-corrected visual acuity (BCVA) in diabetic patients (
      • Nguyen Q.D.
      • Schachar R.A.
      • Nduaka C.I.
      • Sperling M.
      • Basile A.S.
      • Klamerus K.J.
      • et al.
      Dose-ranging evaluation of intravitreal siRNA PF-04523655 for diabetic macular edema (the DEGAS study).
      ). REDD1 acts in the retina, at least in part, by preventing proper activation of the antioxidant transcription factor nuclear factor erythroid-2-related factor 2 (Nrf2) in response to diabetes (
      • Miller W.P.
      • Sunilkumar S.
      • Giordano J.F.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      The stress response protein REDD1 promotes diabetes-induced oxidative stress in the retina by Keap1-independent Nrf2 degradation.
      ). Beyond the retina, a number of recent studies have also demonstrated a role for REDD1 in the development of inflammation and NF-κB activation (
      • Yoshida T.
      • Mett I.
      • Bhunia A.K.
      • Bowman J.
      • Perez M.
      • Zhang L.
      • et al.
      Rtp801, a suppressor of mTOR signaling, is an essential mediator of cigarette smoke-induced pulmonary injury and emphysema.
      ,
      • Nadon A.M.
      • Perez M.J.
      • Hernandez-Saavedra D.
      • Smith L.P.
      • Yang Y.
      • Sanders L.A.
      • et al.
      Rtp801 suppression of epithelial mTORC1 augments endotoxin-induced lung inflammation.
      ,
      • Pastor F.
      • Dumas K.
      • Barthelemy M.A.
      • Regazzetti C.
      • Druelle N.
      • Peraldi P.
      • et al.
      Implication of REDD1 in the activation of inflammatory pathways.
      ,
      • Lee D.K.
      • Kim J.H.
      • Kim J.
      • Choi S.
      • Park M.
      • Park W.
      • et al.
      REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
      ,
      • Hou X.
      • Yang S.
      • Yin J.
      Blocking the REDD1/TXNIP axis ameliorates LPS-induced vascular endothelial cell injury through repressing oxidative stress and apoptosis.
      ). More specifically, a recent study suggested that REDD1 mediated IKK-independent atypical NF-κB activation by sequestering IκB (
      • Lee D.K.
      • Kim J.H.
      • Kim J.
      • Choi S.
      • Park M.
      • Park W.
      • et al.
      REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
      ). Herein, we investigated a role for REDD1 in the development of diabetes-induced retinal inflammation and macrophage infiltration.

      Results

      Diabetes-induced REDD1 expression promotes markers of retinal inflammation

      After 16 weeks of diabetes, WT (REDD1+/+) and REDD1 knockout (REDD1−/−) mice exhibited similar increases in fasted blood glucose concentrations (Fig. 1A). REDD1 mRNA abundance was increased in retinal lysates from REDD1+/+ mice (Fig. 1B). Increased REDD1 mRNA expression in the retina of diabetic mice was visualized throughout retinal layers by in situ hybridization (Fig. 1C). The abundance of mRNA transcripts encoding ICAM-1 (Fig. 1D), CCL5 (Fig. 1E), and CCL2 (Fig. 1F) were also increased in the retina of diabetic mice in a manner that was REDD1-dependent. CCL2 protein expression was increased in the retina of diabetic REDD1+/+ mice, but not diabetic REDD1−/− (Fig. 1G). Moreover, ICAM-1 protein expression in the inner retina and glial activation, as characterized by GFAP expression, were upregulated in diabetic REDD1+/+ mice (Figs. 1H and S1). Diabetes-induced expression of ICAM-1 and GFAP in the retina were attenuated in REDD1−/− mice, as compared to REDD1+/+ mice.
      Figure thumbnail gr1
      Figure 1REDD1 was required for increased inflammatory cytokine expression in the retina of diabetic mice. Diabetes was induced in REDD1+/+ and REDD1−/− mice by administration of streptozotocin (STZ). All analyses were performed 16 weeks after mice were administered STZ or a vehicle (Veh). A, fasting blood glucose concentrations were determined. B, REDD1 mRNA abundance in retinal tissue homogenate was determined by RT-PCR. C, REDD1 mRNA localization in the retina was examined by RNA-ISH (400× magnification; scale bar represents 50 μm). DF, abundance of mRNAs encoding ICAM-1 (D), CCL5 (E), and CCL2 (F) was determined in retinal tissue homogenate by RT-PCR. G, CCL2 protein expression in retinal tissue homogenate was determined by Western blotting. Representative blots are shown. Molecular mass in kDa is indicated at right of each blot. H, ICAM-1 (green) and GFAP (red) were examined in retinal sections by immunofluorescence. Hoechst 33342 (blue) was used to visualize nuclei. Representative micrographs are shown (400× magnification; scale bar represents 75 μm). Values are presented as means ± SD. Data were analyzed in (A, D, and EG) by two-way ANOVA, and pairwise comparisons were made using the Tukey's test for multiple comparisons. Data in (B) were analyzed by unpaired Student’s t test. ∗p < 0.05 versus Veh; #, p < 0.05 versus REDD1+/+; nd, not detected; CCL2, C-C motif chemokine ligand; GCL, ganglion cell layer; INL, inner nuclear layer; ICAM-1, intercellular adhesion molecule 1; ISH, in situ hybridization; ONL, outer nuclear layer.

      REDD1 deletion reduces inflammatory cytokine expression independently of Nrf2

      We recently demonstrated a role for REDD1 expression specifically in Müller glia in the development of retinal dysfunction in diabetic mice (
      • Miller W.P.
      • Toro A.L.
      • Sunilkumar S.
      • Stevens S.A.
      • VanCleave A.M.
      • Williamson D.L.
      • et al.
      Müller glial expression of REDD1 is required for retinal neurodegeneration and visual dysfunction in diabetic mice.
      ). To investigate a role for REDD1 in inflammatory cytokine production by human cells, MIO-M1 Müller glia were exposed to hyperglycemic culture conditions. Hyperglycemic conditions promoted REDD1 protein expression (Fig. 2A) and increased the abundance of mRNAs encoding ICAM-1 (Fig. 2B), CCL5 (Fig. 2C), and CCL2 (Fig. 2D). Increased inflammatory cytokine expression in cells exposed to hyperglycemic conditions required REDD1. Inflammation has long been associated with oxidative stress and we recently demonstrated protective effects of REDD1 deletion that were mediated by enhanced Nrf2 signaling (
      • Miller W.P.
      • Sunilkumar S.
      • Giordano J.F.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      The stress response protein REDD1 promotes diabetes-induced oxidative stress in the retina by Keap1-independent Nrf2 degradation.
      ). In REDD1-deficient cells exposed to hyperglycemic conditions, Nrf2 knockdown reduced expression of the Nrf2 target gene Heme Oxygenase-1 (HO-1) (Fig. 2E). However, Nrf2 knockdown did not impact the abundance of mRNAs encoding ICAM-1 (Fig. 2F), CCL5 (Fig. 2G), or CCL2 (Fig. 2H) in REDD1-deficient cells exposed to hyperglycemic conditions.
      Figure thumbnail gr2
      Figure 2Inflammatory cytokine expression was reduced in REDD1-deficient Müller glial cultures independently of Nrf2. AD, WT and REDD1 KO human MIO-M1 cells were cultured in medium containing 5 mM glucose and exposed to medium containing either 30 mM glucose (HG) or 5 mM glucose and 25 mM mannitol as an osmotic control (LG) for 4 h. REDD1 protein expression in cell lysates was determined by Western blotting (A). Representative blots are shown. Molecular mass in kDa is indicated at left of each blot. ns, not specific. E–H, Nrf2 was knocked down in REDD1 KO MIO-M1 cells by stable expression of an shRNA (shNrf2). Control REDD1 KO cells expressed a scramble shRNA (shScr). All cells were exposed to HG for 4 h. Abundance of mRNAs encoding HO-1 (E), ICAM-1 (B and F), CCL5 (C and G), and CCL2 (D and H) was determined in cell lysates by RT-PCR. Values are presented as means ± SD. Data were analyzed by two-way ANOVA, and pairwise comparisons were made using the Tukey's test for multiple comparisons. ∗p < 0.05 versus LG; #, p < 0.05 versus WT; $, p < 0.05 versus shScr. CCL2, C-C motif chemokine ligand; ICAM-1, intercellular adhesion molecule 1; Nrf2, nuclear factor erythroid-2-related factor 2.

      REDD1 sustains TNFα-induced NF-κB signaling

      To explore the impact of REDD1 on NF-κB signaling, WT and REDD1 KO MIO-M1 cells were exposed to TNFα. In WT cells, REDD1 expression was increased after 1 to 4 h of exposure to TNFα (Fig. 3, A and B). Acute TNFα exposure (i.e., 0.25 h) promoted rapid phosphorylation of NF-κB at S536 and reduced expression of IκBα in both WT and REDD1 KO cells (Fig. 3A). However, after 4 h of TNFα exposure, phosphorylation of NF-κB (Fig. 3, A and C) and reduced Iκ-Bα expression (Fig. 3, A and D) were sustained in WT, but not in REDD1 KO cells. In both WT and REDD1 KO cells, TNFα caused rapid localization of NF-κB to the nucleus (Fig. 3E). With more prolonged exposure to TNFα, NF-κB remained localized to the nucleus in WT cells, whereas the transcription factor returned to the cytoplasm in REDD1 KO cells. The data support that REDD1 is required for sustained activation of NF-κB.
      Figure thumbnail gr3
      Figure 3REDD1 sustained NF-κB signaling in Müller glial cultures exposed to TNF⍺. WT and REDD1 KO human MIO-M1 cells were exposed to medium supplemented with TNF⍺ for up to 4 h. A, REDD1 expression, NF-κB phosphorylation at S536, and IκB⍺ expression were determined in cell lysates by Western blotting. Representative blots are shown. Molecular mass in kDa is indicated at right of each blot. REDD1 expression (B), NF-κB phosphorylation (C), and IκB⍺ expression (D) were quantified after 4 h of exposure to TNF⍺ or vehicle (Veh). Values are presented as means ± SD. Data were analyzed by two-way ANOVA, and pairwise comparisons were made using the Tukey's test for multiple comparisons. ∗p < 0.05 versus Veh; #, p < 0.05 versus WT. E, nuclear localization of NF-κB p65 was determined by immunofluorescence. Nuclei were visualized with DAPI (600× magnification; scale bar represents 25 μm). IκB, inhibitor of IκB; ns, not specific; NF-κB, nuclear transcription factor κB.

      IKK activity is required for REDD1-dependent NF-κB activation and inflammatory cytokine expression

      To further examine NF-κB activity in Müller glia, a NF-κB luciferase reporter system was employed. TNFα promoted NF-κB reporter activity in WT cells, whereas REDD1 deletion prevented the effect (Fig. 4A). TNFα also enhanced the abundance of mRNAs encoding ICAM-1 (Fig. 4B), CCL5 (Fig. 4C), and CCL2 (Fig. 4D) in WT cells. Remarkably, REDD1 deletion prevented an increase in inflammatory cytokine expression in cells exposed to TNFα. Similarly, pharmacological IKKβ inhibition suppressed TNFα-induced NF-κB reporter activity in WT cells (Fig. 4E) and reduced expression of ICAM-1 (Fig. 4F), CCL5 (Fig. 4G), and CCL2 (Fig. 4H).
      Figure thumbnail gr4
      Figure 4REDD1-dependent NF-κB activation and inflammatory cytokine expression required IKK. AH, WT and REDD1 KO human MIO-M1 cells were exposed to medium supplemented with TNF⍺ or vehicle (Veh) for 4 h. EH, WT MIO-M1 cells were exposed to the IKKβ inhibitor IMD0354 (IMD) for 1 h prior to the addition of TNF⍺. NF-κB activity was measured in lysates from cells expressing NF-κB firefly luciferase/Renilla luciferase reporter plasmids by dual luciferase assay (A and E.). Abundance of mRNAs encoding ICAM-1 (B and F), CCL5 (C and G), and CCL2 (D and H) was determined in cell lysates by RT-PCR. Values are presented as means ± SD. Data were analyzed by either two-way (A–D) or one way (E–H) ANOVA, and pairwise comparisons were made using the Tukey's test for multiple comparisons. ∗p < 0.05 versus Veh; #, p < 0.05 versus WT or TNF⍺ alone. CCL2, C-C motif chemokine ligand; IKK, IκB kinase; ICAM-1, intercellular adhesion molecule 1; NF-κB, nuclear transcription factor κB.

      REDD1 promotes canonical NF-κB signaling

      Canonical NF-κB signaling involves activation of the IKK complex that consists of two kinases (IKKα and IKKβ) and a regulatory subunit (NEMO/IKKγ). TNFα exposure enhanced phosphorylation of both NEMO at S376 (Fig. 5A) and IKKα/β at S176/S180 (Fig. 5B) in WT MIO-M1 cells, but not in REDD1 KO cells. K63-linked nondegradable polyubiquitin (pUb) chains regulate canonical NF-κB signaling (
      • Deng L.
      • Wang C.
      • Spencer E.
      • Yang L.
      • Braun A.
      • You J.
      • et al.
      Activation of the IκB kinase complex by TRAF6 requires a dimeric ubiquitin-conjugating enzyme complex and a unique polyubiquitin chain.
      ,
      • Chen Z.J.
      Ubiquitin signalling in the NF-κB pathway.
      ), with direct ubiquitination of NEMO positively modulating IKK complex activation (
      • Israël A.
      NF-κB activation: nondegradative ubiquitination implicates NEMO.
      ,
      • Zhu G.
      • Wu C.-J.
      • Zhao Y.
      • Ashwell J.D.
      Optineurin negatively regulates TNFα- induced NF-κB activation by competing with NEMO for ubiquitinated RIP.
      ,
      • Laplantine E.
      • Fontan E.
      • Chiaravalli J.
      • Lopez T.
      • Lakisic G.
      • Véron M.
      • et al.
      NEMO specifically recognizes K63-linked poly-ubiquitin chains through a new bipartite ubiquitin-binding domain.
      ). To determine the impact of REDD1 on K63-specific ubiquitination of NEMO, hemagglutinin (HA)-tagged NEMO was expressed in HEK239 cells. K63-pUb was detected in the HA-NEMO immunoprecipitate (IP) from cell lysates (Fig. 5C). As compared to WT cells, K63-pUb was reduced in the HA-NEMO IP from REDD1 KO cells. Surprisingly, IKKα and IKKβ coimmunoprecipitation with HA-NEMO was unaffected by REDD1 deletion or TNFα exposure (Fig. 5D). However, enhanced phosphorylation of IKKα/β was observed in the HA-NEMO IP from WT cells after TNFα exposure, whereas a similar effect was not observed in REDD1 KO cells. Further, phosphorylation of IKKα/β (S176/S180) and IκB at S32 were enhanced upon TNFα exposure in a manner that was dependent on REDD1 expression.
      Figure thumbnail gr5
      Figure 5REDD1 promoted canonical NF-κB signaling. A and B, WT and REDD1 KO MIO-M1 cells were exposed to medium supplemented with TNFα or vehicle (Veh) for 4 h. Phosphorylation of NEMO at S376 (A) and IKKα/β at S176/180 (B) was determined by Western blotting. Values are presented as means ± SD. Data were analyzed by two-way ANOVA, and pairwise comparisons were made using the Tukey's test for multiple comparisons. ∗p < 0.05 versus Veh; #, p < 0.05 versus WT. C and D, HA-tagged NEMO (HA-NEMO) or an empty vector control (EV) was expressed in HEK293 cells by transient transfection. Cells were exposed to medium supplemented with TNFα or Veh for 1 h. HA-NEMO was immunoprecipitated from cell lysates by affinity purification. Cell lysate (5% of input) and HA-tag immunoprecipitate (IP: HA) was examined by Western blotting. Representative blots are shown. Molecular mass in kDa is indicated at left of each blot. HA, hemagglutinin; IKK, IκB kinase; NF-κB, nuclear transcription factor κB; TNFα, tumor necrosis factor α.

      REDD1 deletion suppressed canonical NF-κB signaling in the retina of diabetic mice

      To investigate a role for REDD1 in diabetes-induced NF-κB activation, nuclear isolates from whole retina were examined using an ELISA that measures the binding of NF-κB p65 to an oligonucleotide encoding the κB consensus motif. Diabetes promoted NF-κB activity (Fig. 6A) and increased nuclear NF-κB p65 expression (Fig. 6B) in the retina of REDD1+/+ mice, but not in REDD1−/− mice. IκB expression was also reduced in the retina of diabetic REDD1+/+ mice, as compared to nondiabetic REDD1+/+ mice (Fig. 6C). As compared to diabetic REDD1+/+ mice, IκB expression was increased in the retina of diabetic REDD1−/− mice. Diabetes also promoted phosphorylation of NF-κB at S536 (Fig. 6D) and NEMO at S376 (Fig. 6E), which were also dependent on REDD1. To correlate changes in NF-κB with retinal inflammation, leukostasis was examined in retinal sections. In the retina of REDD1+/+ mice, diabetes resulted in macrophage infiltration into the inner retina (Fig. 6, F and G). However, as compared to nondiabetic REDD1−/− mice, macrophage infiltration was not altered in the retina of diabetic REDD1−/− mice.
      Figure thumbnail gr6
      Figure 6REDD1 deletion prevented NF-κB activation and macrophage infiltration in the retina of diabetic mice. Diabetes was induced in REDD1+/+ and REDD1−/− mice by administration of streptozotocin (STZ) or a vehicle (Veh) control. All analyses were performed after 16 weeks of diabetes. A, NF-κB activity was determined in nuclear isolates obtained from retinal lysates by DNA-binding ELISA. B, NF-κB p65 protein expression in nuclear isolates was determined by Western blotting. Representative blots are shown. Molecular mass in kDa is indicated at right of each blot. C–E, IκBα expression (C), NF-κB phosphorylation at S536 (D), and NEMO phosphorylation at S376 (E) were evaluated in retinal tissue homogenate by Western blotting. F, F4/80 (green) and IBA1 (red) were examined in retinal sections by immunofluorescence. Hoechst 33342 (blue) was used to visualize nuclei. Representative micrographs are shown (400× magnification; scale bar represents 75 μm). G, F4/80-positive cells in (F) were counted. Values are presented as means ± SD. Data were analyzed by two-way ANOVA, and pairwise comparisons were made using the Tukey's test for multiple comparisons. ∗p < 0.05 versus Veh; #, p < 0.05 versus REDD1+/+; GCL, ganglion cell layer; IκB, inhibitor of IκB; INL, inner nuclear layer; NF-κB, nuclear transcription factor κB; ONL, outer nuclear layer; OS, photoreceptor outer segments.

      Activating transcription factor 4 is necessary for TNFα-induced REDD1 expression

      To investigate potential upstream mediators that promote REDD1 mRNA expression in the retina, a role for the transcription factor activating transcription factor 4 (ATF4) was examined. ATF4 is a well-documented regulator of stress-induced REDD1 mRNA transcription (
      • Whitney M.L.
      • Jefferson L.S.
      • Kimball S.R.
      ATF4 is necessary and sufficient for ER stress-induced upregulation of REDD1 expression.
      ,
      • Kimball S.R.
      • Jefferson L.S.
      Induction of REDD1 gene expression in the liver in response to endoplasmic reticulum stress is mediated through a PERK, eIF2α phosphorylation, ATF4-dependent cascade.
      ,
      • Xu D.
      • Dai W.
      • Kutzler L.
      • Lacko H.A.
      • Jefferson L.S.
      • Dennis M.D.
      • et al.
      ATF4-Mediated upregulation of REDD1 and Sestrin2 suppresses mTORC1 activity during prolonged leucine deprivation.
      ). ATF4 protein expression was increased in retinal lysates from streptozotocin (STZ)-diabetic mice as compared to their littermate controls (Fig. 7A). When MIO-M1 cells were exposed to TNFα, ATF4 and REDD1 protein expression coordinately increased (Fig. 7B). In cells expressing an shRNA targeting the ATF4 mRNA, ATF4 expression was suppressed (Fig. 7C), and the stimulatory effect of TNFα on REDD1 expression was absent (Fig. 7D). In addition to reducing REDD1 mRNA abundance (Fig. 7E), ATF4 knockdown also suppressed the abundance of mRNAs encoding ICAM-1 (Fig. 7E), CCL5 (Fig. 7F), and CCL2 (Fig. 7G). Together, the data support ATF4-dependent REDD1 expression in the regulation of inflammatory cytokine production.
      Figure thumbnail gr7
      Figure 7ATF4 was necessary for TNFα-induced REDD1 expression. A, diabetes was induced in mice by administration of streptozotocin (STZ) or a vehicle (Veh) control. After 16 weeks of diabetes, ATF4 protein expression was examined in retinal lysates by Western blotting. Representative blots are shown. Molecular mass in kDa is indicated at right of each blot. B, MIO-M1 cells were exposed to medium supplemented with TNFα or vehicle (Veh) for up to 4 h. ATF4 and REDD1 protein expression were examined by Western blotting. C, ATF4 mRNA was examined in MIO-M1 cells expressing an shRNA targeting the ATF4 mRNA (shATF4) versus WT control cells by RT-PCR. D, WT and shATF4 cells were exposed to TNFα for 2 h and REDD1 protein expression was determined by Western blotting. E–H, REDD1 (E), ICAM-1 (F), CCL5 (G), and CCL2 (H) mRNA abundances were determined in WT and shATF4 cells by RT-PCR. Values are presented as means ± SD. Data in (A, C, and EH) were analyzed by unpaired Student’s t test. Data in (D) were analyzed by two-way ANOVA, and pairwise comparisons were made using the Tukey's test for multiple comparisons. ∗p < 0.05 versus Veh; #, p < 0.05 versus WT. ATF4, activating transcription factor 4; CCL2, C-C motif chemokine ligand; ICAM-1, intercellular adhesion molecule 1; TNFα, tumor necrosis factor α.

      Discussion

      Herein, we investigated a role for REDD1 in the development of diabetes-induced retinal inflammation. Increased REDD1 expression was observed in the retina of mice after 16-weeks of STZ-diabetes, as well as in human Müller glia cultures exposed to hyperglycemic culture conditions. Diabetes promoted NF-κB signaling, enhanced pro-inflammatory cytokine expression, and lead to the development of retinal leukostasis. However, REDD1 deletion reduced IKK activation, suppressed nuclear localization of NF-κB, and attenuated both inflammatory cytokine expression and macrophage infiltration of the inner retina in diabetic mice. Overall, the studies provide evidence that REDD1 acts to promote retinal inflammation by sustaining activation of canonical NF-κB signaling (Fig. 8).
      Figure thumbnail gr8
      Figure 8REDD1 sustains NF-κB signaling to promote retinal inflammation. Working model for the role of REDD1 in diabetes-induced activation of NF-κB signaling and enhanced inflammatory cytokine production by Müller glia. Graphic created with BioRender.com. NF-κB, nuclear transcription factor κB.
      DR is diagnosed by clinically visible microvascular abnormalities in the retina and thus has principally been investigated as a microvascular complication. However, evidence supports that gliosis, neuroinflammation, neurodegeneration, and loss of neurovascular coupling are critical elements of disease pathogenesis (
      • Antonetti D.A.
      • Barber A.J.
      • Bronson S.K.
      • Freeman W.M.
      • Gardner T.W.
      • Jefferson L.S.
      • et al.
      JDRF Diabetic Retinopathy Center Group Diabetic retinopathy: seeing beyond glucose-induced microvascular disease.
      ). In fact, neuroglial deficits can precede and even predict the visible signs of microvascular disease in diabetic patients (
      • Han Y.
      • Bearse Jr., M.A.
      • Schneck M.E.
      • Barez S.
      • Jacobsen C.H.
      • Adams A.J.
      Multifocal electroretinogram delays predict sites of subsequent diabetic retinopathy.
      ,
      • Adams A.J.
      • Bearse Jr., M.A.
      Retinal neuropathy precedes vasculopathy in diabetes: a function-based opportunity for early treatment intervention?.
      ). Specificity of REDD1 expression in human retina is consistent with markers of Muller glia, and conditional deletion of REDD1 in Muller glia of mice prevents the increase in retinal REDD1 expression with STZ-diabetes (
      • Miller W.P.
      • Toro A.L.
      • Sunilkumar S.
      • Stevens S.A.
      • VanCleave A.M.
      • Williamson D.L.
      • et al.
      Müller glial expression of REDD1 is required for retinal neurodegeneration and visual dysfunction in diabetic mice.
      ). Enhanced REDD1 mRNA was observed throughout the retinal layers after 16 weeks of STZ-diabetes. Muller glia extend radially across the entire retina, thus the localization of REDD1 mRNA in retinal sections was not unexpected. Retinal Müller glia play a central role in the development of inflammation in DR (
      • Rübsam A.
      • Parikh S.
      • Fort P.E.
      Role of inflammation in diabetic retinopathy.
      ). Müller glia provide critical homeostatic and trophic support to maintain both activity of retinal neurons and integrity of the blood-retinal barrier. In response to diabetes, Müller cells become activated and secrete a range of pro-inflammatory factors that are NF-κB target genes (
      • Coughlin B.A.
      • Feenstra D.J.
      • Mohr S.
      Müller cells and diabetic retinopathy.
      ). Herein, expression of ICAM-1, CCL5, and CCL2 were increased in both the retina of STZ-diabetic mice and in retinal Müller glia cultures exposed to TNF⍺ in coordination with activation of NF-κB.
      NF-κB serves as a key mediator of the retinal inflammatory response to diabetes (
      • Nagai N.
      • Izumi-Nagai K.
      • Oike Y.
      • Koto T.
      • Satofuka S.
      • Ozawa Y.
      • et al.
      Suppression of diabetes-induced retinal inflammation by blocking the angiotensin II type 1 receptor or its downstream nuclear factor-κb pathway.
      ,
      • Kowluru R.A.
      • Kanwar M.
      Effects of curcumin on retinal oxidative stress and inflammation in diabetes.
      ,
      • Lennikov A.
      • Hiraoka M.
      • Abe A.
      • Ohno S.
      • Fujikawa T.
      • Itai A.
      • et al.
      IκB kinase-β inhibitor IMD-0354 beneficially suppresses retinal vascular permeability in streptozotocin-induced diabetic mice.
      ). It has long been established that diabetes promotes retinal NF-κB activation in both diabetic patients (
      • Romeo G.
      • Liu W.-H.
      • Asnaghi V.
      • Kern T.S.
      • Lorenzi M.
      Activation of nuclear factor-κb induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes.
      ) and in preclinical models (
      • Kowluru R.A.
      • Koppolu P.
      • Chakrabarti S.
      • Chen S.
      Diabetes-induced activation of nuclear transcriptional factor in the retina, and its inhibition by antioxidants.
      ,
      • Kowluru R.A.
      • Chakrabarti S.
      • Chen S.
      Re-institution of good metabolic control in diabetic rats and activation of caspase-3 and nuclear transcriptional factor (NF-kB) in the retina.
      ). However, the underlying mechanisms responsible for enhanced nuclear localization of NF-κB in the context of DR are not well delineated. Only recently was enhanced NF-κB phosphorylation reported in the retina of Ins2Akita mice (
      • Homme R.P.
      • Sandhu H.S.
      • George A.K.
      • Tyagi S.C.
      • Singh M.
      Sustained inhibition of NF-κB activity mitigates retinal vasculopathy in diabetes.
      ) and STZ-diabetic rats (
      • Liang W.-J.
      • Yang H.-W.
      • Liu H.-N.
      • Qian W.
      • Chen X.-L.
      HMGB1 upregulates NF-kB by inhibiting IKB-α and associates with diabetic retinopathy.
      ) and mice (
      • Zhang T.
      • Mei X.
      • Ouyang H.
      • Lu B.
      • Yu Z.
      • Wang Z.
      • et al.
      Natural flavonoid galangin alleviates microglia-trigged blood–retinal barrier dysfunction during the development of diabetic retinopathy.
      ). Limited reports of enhanced IκB phosphorylation and reduced IκB expression in the retina of STZ-diabetic rodents support activation of canonical NF-κB signaling (
      • Zhang T.
      • Mei X.
      • Ouyang H.
      • Lu B.
      • Yu Z.
      • Wang Z.
      • et al.
      Natural flavonoid galangin alleviates microglia-trigged blood–retinal barrier dysfunction during the development of diabetic retinopathy.
      ,
      • Zhai J.
      • Li Z.
      • Zhang H.
      • Ma L.
      • Ma Z.
      • Zhang Y.
      • et al.
      Berberine protects against diabetic retinopathy by inhibiting cell apoptosis via deactivation of the NF-κB signaling pathway.
      ). Canonical NF-κB signaling depends on activation of the IKK complex, which includes two kinase subunits (IKKα/β) and the regulatory subunit NEMO. In particular, IKKβ and NEMO mediate canonical NF-κB signaling, whereas noncanonical signaling is dependent on IKKα. Increased IKKβ phosphorylation was recently reported in retinal lysates from STZ-diabetic rats (
      • Li W.
      • Shen X.
      • Wang Y.
      • Zhang J.
      The effect of Shengpuhuang-tang on retinal inflammation in streptozotocin-induced diabetic rats by NF-κB pathway.
      ). Herein, increased NF-κB DNA-binding activity was observed in the retina of STZ-diabetic mice in coordination with nuclear localization and phosphorylation of NF-κB p65. Moreover, diabetes attenuated retinal IκB expression and enhanced autophosphorylation of NEMO within the IKK complex.
      A growing body of literature supports that the stress response protein REDD1 plays a critical role in the development of diabetes-induced retinal defects. Indeed, REDD1 deletion prevents retinal pathology and functional deficits in vision in STZ-diabetic mice (
      • Miller W.P.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      REDD1 activates a ROS-generating feedback loop in the retina of diabetic mice.
      ,
      • Miller W.P.
      • Yang C.
      • Mihailescu M.L.
      • Moore J.A.
      • Dai W.
      • Barber A.J.
      • et al.
      Deletion of the Akt/mTORC1 repressor REDD1 prevents visual dysfunction in a rodent model of type 1 diabetes.
      ,
      • Miller W.P.
      • Sunilkumar S.
      • Giordano J.F.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      The stress response protein REDD1 promotes diabetes-induced oxidative stress in the retina by Keap1-independent Nrf2 degradation.
      ). It is well established that REDD1 mediates the cellular stress response to a number of adverse conditions, including hyperglycemia (
      • Lipina C.
      • Hundal H.S.
      Is REDD1 a metabolic éminence grise?.
      ,
      • Miller W.P.
      • Sunilkumar S.
      • Dennis M.D.
      The stress response protein REDD1 as a causal factor for oxidative stress in diabetic retinopathy.
      ). In both the retina of diabetic mice and in human retinal cell cultures exposed to hyperglycemic conditions, REDD1 expression was upregulated in association with pro-inflammatory signaling. Deletion of REDD1 prevented a pro-inflammatory response to diabetes and hyperglycemic conditions, supporting a role for REDD1 in development of inflammation. This is consistent with prior studies on endotoxemia pathophysiology that demonstrate an essential role for REDD1 in lipopolysaccharide (LPS)-induced inflammation (
      • Pastor F.
      • Dumas K.
      • Barthelemy M.A.
      • Regazzetti C.
      • Druelle N.
      • Peraldi P.
      • et al.
      Implication of REDD1 in the activation of inflammatory pathways.
      • Lee D.K.
      • Kim J.H.
      • Kim J.
      • Choi S.
      • Park M.
      • Park W.
      • et al.
      REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
      ).
      Oxidative stress resulting from an imbalance between the production of reactive oxygen species and their elimination by antioxidants contributes to the development of retinal inflammation with diabetes (
      • Kowluru R.A.
      • Koppolu P.
      • Chakrabarti S.
      • Chen S.
      Diabetes-induced activation of nuclear transcriptional factor in the retina, and its inhibition by antioxidants.
      ). Upregulation of REDD1 has been linked to multiple disease models that involve the development of oxidative stress, including DR (
      • Miller W.P.
      • Sunilkumar S.
      • Dennis M.D.
      The stress response protein REDD1 as a causal factor for oxidative stress in diabetic retinopathy.
      ). Retinal cells combat the development of oxidative stress through activation of the Nrf2 antioxidant response. Nrf2 deletion promotes the development of inflammation with ischemia-reperfusion injury (
      • Wei Y.
      • Gong J.
      • Yoshida T.
      • Eberhart C.G.
      • Xu Z.
      • Kombairaju P.
      • et al.
      Nrf2 has a protective role against neuronal and capillary degeneration in retinal ischemia–reperfusion injury.
      ). Moreover, knockdown of Nrf2 and its downstream targets HO-1 and NQO1 increases NF-κB–mediated inflammation in monocytes exposed LPS (
      • Rushworth S.A.
      • MacEwan D.J.
      • O’Connell M.A.
      Lipopolysaccharide-induced expression of NAD(P)H:quinone oxidoreductase 1 and Heme oxygenase-1 protects against excessive inflammatory responses in human monocytes.
      ). REDD1 acts to negatively regulate Nrf2 activity in the retina of diabetic mice by promoting nuclear exclusion and degradation of the transcription factor (
      • Miller W.P.
      • Sunilkumar S.
      • Giordano J.F.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      The stress response protein REDD1 promotes diabetes-induced oxidative stress in the retina by Keap1-independent Nrf2 degradation.
      ). As compared to WT Müller cells exposed to hyperglycemic conditions, REDD1 deletion promoted Nrf2 hyperactivation in coordination with blunted expression of ICAM-1, CCL5, and CCL2. Nrf2 knockdown in REDD1-deficient cells restored HO-1 expression to a level that was similar to that observed in WT cells. However, Nrf2 knockdown did not enhance expression of ICAM-1, CCL5, and CCL2 in REDD1-deficient cells exposed to hyperglycemic conditions. This supports that the reduced inflammatory cytokine expression with REDD1 ablation was independent of Nrf2 hyperactivation.
      A role for REDD1 in NF-κB–mediated inflammation has been noted in mice exposed to LPS or cigarette smoke (
      • Yoshida T.
      • Mett I.
      • Bhunia A.K.
      • Bowman J.
      • Perez M.
      • Zhang L.
      • et al.
      Rtp801, a suppressor of mTOR signaling, is an essential mediator of cigarette smoke-induced pulmonary injury and emphysema.
      ,
      • Pastor F.
      • Dumas K.
      • Barthelemy M.A.
      • Regazzetti C.
      • Druelle N.
      • Peraldi P.
      • et al.
      Implication of REDD1 in the activation of inflammatory pathways.
      ,
      • Lee D.K.
      • Kim J.H.
      • Kim J.
      • Choi S.
      • Park M.
      • Park W.
      • et al.
      REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
      ). Lee et al. (2018) demonstrated that REDD1 promotes atypical NF-κB activation by IκBα sequestration (
      • Lee D.K.
      • Kim J.H.
      • Kim J.
      • Choi S.
      • Park M.
      • Park W.
      • et al.
      REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
      ). In macrophages exposed to LPS for 24 h, REDD1 interferes with IκBα binding to NF-κB to promote nuclear translocation and activity of the transcription factor (
      • Lee D.K.
      • Kim J.H.
      • Kim J.
      • Choi S.
      • Park M.
      • Park W.
      • et al.
      REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
      ). Prior work suggests that LPS activates NF-κB via both an early IKK-dependent mechanism, as well as through delayed IKK-independent mechanism (
      • Li X.
      • Su J.
      • Cui X.
      • Li Y.
      • Barochia A.
      • Eichacker P.Q.
      Can we predict the effects of NF-κB inhibition in sepsis? Studies with parthenolide and ethyl pyruvate.
      ). Notably, Lee et al. observed a suppressive effect of REDD1 on NF-κB p65 phosphorylation in macrophages exposed to LPS for only 30 min, despite absence of IκBα expression at that timepoint (
      • Lee D.K.
      • Kim J.H.
      • Kim J.
      • Choi S.
      • Park M.
      • Park W.
      • et al.
      REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
      ). The observation suggests the possibility that REDD1 may also act independently of IκBα sequestration. The studies herein extend on the prior work by demonstrating that REDD1 sustained canonical NF-κB signaling by enhancing IKK activation to suppress IκBα expression. REDD1 deletion reduced both auto-phosphorylation and K63-ubiquitination of the IKK complex. Together with the prior report (
      • Lee D.K.
      • Kim J.H.
      • Kim J.
      • Choi S.
      • Park M.
      • Park W.
      • et al.
      REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
      ), the findings support a model wherein REDD1 acts by both enhancing IKK-dependent degradation of IκBα to promote NF-κB nuclear translocation and through IKK-independent sequestration of newly synthesized IκBα. In addition to promoting the transcription of pro-inflammatory cytokines, NF-κB also mediates transcription of IκBα, which facilitates feedback inhibition of the signaling pathway (
      • Sun S.-C.
      • Ganchi P.A.
      • Ballard D.W.
      • Greene W.C.
      NF-κB controls expression of inhibitor IκBα: evidence for an inducible autoregulatory pathway.
      ). In REDD1-deficient Müller glia cultures, TNFα-induced NF-κB phosphorylation and the reduction in IκBα expression were resolved more quickly than in WT cells. This suggests that REDD1 uncouples autoregulation of NF-κB activation and provides an additional molecular checkpoint for the termination of pro-inflammatory signaling.
      We recently provided evidence that increased retinal REDD1 protein expression results from impaired lysosomal proteolysis of REDD1 protein as a consequence of redox stabilization (
      • Miller W.P.
      • Sha C.M.
      • Sunilkumar S.
      • Toro A.L.
      • VanCleave A.M.
      • Kimball S.R.
      • et al.
      Activation of disulfide redox switch in REDD1 promotes oxidative stress under hyperglycemic conditions.
      ). After 6 weeks of STZ-diabetes, enhanced REDD1 protein expression in retinal lysates is observed in the absence of a change in REDD1 mRNA expression (
      • Miller W.P.
      • Sha C.M.
      • Sunilkumar S.
      • Toro A.L.
      • VanCleave A.M.
      • Kimball S.R.
      • et al.
      Activation of disulfide redox switch in REDD1 promotes oxidative stress under hyperglycemic conditions.
      ). By contrast, 16 weeks of STZ-diabetes resulted in upregulated REDD1 mRNA abundance. The observation suggests that with prolonged diabetes duration, redox stabilization of REDD1 is potentially exacerbated by a transcriptional effect. Prior reports demonstrate that retinal REDD1 expression is transcriptionally upregulated in response to hypoxic stress in a murine model of retinopathy of prematurity (
      • Brafman A.
      • Mett I.
      • Shafir M.
      • Gottlieb H.
      • Damari G.
      • Gozlan-Kelner S.
      • et al.
      Inhibition of oxygen-induced retinopathy in rtp801-deficient mice.
      ). Moreover, evidence supports that REDD1 is a transcriptional target of ATF4 (
      • Whitney M.L.
      • Jefferson L.S.
      • Kimball S.R.
      ATF4 is necessary and sufficient for ER stress-induced upregulation of REDD1 expression.
      ), HIF-1α (
      • Shoshani T.
      • Faerman A.
      • Mett I.
      • Zelin E.
      • Tenne T.
      • Gorodin S.
      • et al.
      Identification of a novel hypoxia-inducible factor 1-responsive gene, RTP801, involved in apoptosis.
      ), p53 (
      • Ellisen L.W.
      • Ramsayer K.D.
      • Johannessen C.M.
      • Yang A.
      • Beppu H.
      • Minda K.
      • et al.
      REDD1, a developmentally regulated transcriptional target of p63 and p53, links p63 to regulation of reactive oxygen species.
      ), and even RelB (
      • Ramachandiran S.
      • Adon A.
      • Guo X.
      • Wang Y.
      • Wang H.
      • Chen Z.
      • et al.
      Chromosome instability in diffuse large B cell lymphomas is suppressed by activation of the noncanonical NF-κB pathway.
      ). Herein, increased retinal REDD1 mRNA abundance was observed coincident with enhanced ATF4 protein expression. ATF4 knockdown prevented an increase in REDD1 expression in Müller cells exposed to TNFα and attenuated inflammatory cytokine expression. The data support a prior study that identified an essential role for ATF4 in diabetes-induced inflammatory cytokine production by Müller glial (
      • Zhong Y.
      • Li J.
      • Chen Y.
      • Wang J.J.
      • Ratan R.
      • Zhang S.X.
      Activation of endoplasmic reticulum stress by hyperglycemia is essential for müller cell–derived inflammatory cytokine production in diabetes.
      ). More specifically, ATF4 expression localizes to Müller cells in the retina of STZ-diabetic mice, and expression of a dominant negative ATF4 variant reduces inflammatory cytokine expression in both the retina of diabetic mice and in Müller cells exposed to hyperglycemic conditions (
      • Zhong Y.
      • Li J.
      • Chen Y.
      • Wang J.J.
      • Ratan R.
      • Zhang S.X.
      Activation of endoplasmic reticulum stress by hyperglycemia is essential for müller cell–derived inflammatory cytokine production in diabetes.
      ). The studies here extend on the prior work by demonstrating a key role for REDD1 and NF-κB acting downstream of ATF4 to promote Müller glial expression of pro-inflammatory cytokines.
      Overall, the findings here provide new insight into the molecular mechanisms whereby diabetes contributes to the development of retinal inflammation. In the retina of diabetic mice, REDD1 expression was increased in coordination with NF-κB activation, pro-inflammatory cytokine expression, and macrophage infiltration. REDD1 deletion reduced NF-κB activation in the retina of diabetic mice and prevented the development of leukostasis. The data support that therapeutic intervention to suppress REDD1 expression may improve retinal pathology in diabetic patients. Therapeutically, DR is principally addressed by blockade of vascular endothelial growth factor (
      • Brown D.M.
      • Nguyen Q.D.
      • Marcus D.M.
      • Boyer D.S.
      • Patel S.
      • Feiner L.
      • et al.
      Long-term outcomes of ranibizumab therapy for diabetic macular edema: The 36-month results from two phase III trials: RISE and RIDE.
      ). A major limitation of these interventions is that vascular endothelial growth factor suppression largely addresses the microvascular dysfunction and neovascularization that characterize later stages of disease progression. The studies here support that REDD1 suppression may represent a therapeutic strategy to prevent the development of retinal inflammation. Indeed, administration of an intravitreally administered siRNA targeting the REDD1 mRNA (PF-04523655) in patients with diabetic macular edema showed a trend toward improvement in BCVA when compared to focal/grid laser (+5.8 letters with 3 mg PF-04523655 versus +2.4 letters with laser, p = 0.08) (
      • Nguyen Q.D.
      • Schachar R.A.
      • Nduaka C.I.
      • Sperling M.
      • Basile A.S.
      • Klamerus K.J.
      • et al.
      Dose-ranging evaluation of intravitreal siRNA PF-04523655 for diabetic macular edema (the DEGAS study).
      ). In a secondary analysis of patients that completed a 12-months follow up, mean improvement in BCVA with 3 mg PF-04523655 was superior to laser intervention. The proof-of-concept studies here are consistent with a mechanism of action wherein REDD1 suppression improves visual acuity in diabetic patient by preventing sustained NF-κB activation and reducing the retinal inflammatory response to diabetes.

      Experimental procedures

      Animals

      Male WT (REDD1+/+) and REDD1 KO (REDD1−/−) B6;129 mice (
      • Brafman A.
      • Mett I.
      • Shafir M.
      • Gottlieb H.
      • Damari G.
      • Gozlan-Kelner S.
      • et al.
      Inhibition of oxygen-induced retinopathy in rtp801-deficient mice.
      ) were maintained on a 12:12-h reverse light dark cycle. Diabetes was induced at 6 weeks of age by administering 50 mg/kg STZ intraperitoneally for five consecutive days. Littermate control mice were injected with equivalent volumes of sodium citrate buffer. Two weeks after injections, diabetic phenotype was confirmed by fasting blood glucose concentrations >250 mg/dL. At 16 weeks of diabetes, mice were euthanized, and retina and whole eyes were extracted. All procedures were approved by the Penn State College of Medicine Institutional Animal Care and Use Committee and were in accordance with the ARVO statement on the ethical use of animals in ophthalmological research.

      Cell culture

      MIO-M1 human Müller cells were obtained from the UCL Institute of Ophthalmology. MIO-M1 cells deficient for REDD1 (REDD1 KO) were generated using CRISPR/Cas9 genome editing as previously described (
      • Dai W.
      • Miller W.P.
      • Toro A.L.
      • Black A.J.
      • Dierschke S.K.
      • Feehan R.P.
      • et al.
      Deletion of the stress-response protein REDD1 promotes ceramide-induced retinal cell death and JNK activation.
      ,
      • Dennis M.D.
      • Coleman C.S.
      • Berg A.
      • Jefferson L.S.
      • Kimball S.R.
      REDD1 enhances protein phosphatase 2A-mediated dephosphorylation of Akt to repress mTORC1 signaling.
      ). MIO-M1 cultures were maintained in Dulbecco’s modified Eagle’s medium (DMEM, Thermo Fisher Scientific) containing 5.6 mM glucose and supplemented with 10% heat inactivated fetal bovine serum and 1% penicillin-streptomycin. MIO-M1 cells stably expressing an shRNA targeting Nrf2 (5′-CCGGG-CTCCTACTGTGATGTGAAATCTCGAGATTTCACATCACAGTAGGAGCTTTTT-3′) or ATF4 (5′-CCGGGCCTAGGTCTCTTAGATGATTCTCGAGAATCATCTAAGAGACCTAGGCTTTTT-3′) were generated as previously described (
      • Miller W.P.
      • Sunilkumar S.
      • Giordano J.F.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      The stress response protein REDD1 promotes diabetes-induced oxidative stress in the retina by Keap1-independent Nrf2 degradation.
      ). Cells expressing pLKO.1-TRC [provided by David Root (Addgene Plasmid #10879)] were used as an shRNA control. To model hyperglycemia, culture medium was supplemented with either 24.4 mM glucose or 24.4 mM mannitol as an osmotic control. Cells were transfected using Lipofectamine 2000 (Life Technologies). Plasmids included pCMV5 vector, pCMV-HA-REDD1, pRL-Renilla luciferase (Promega). NF-κB-TATA luciferase reporter and pCMV4-HA-NEMO (IKKγ) plasmids were kindly provided by Dr Edward Harhaj (Penn State College of Medicine). Where indicated, cell culture medium was supplemented with recombinant human TNFα (20 ng/ml; Sigma) or 1 μM N-(3,5-Bis-trifluoromethylphenyl)-5-chloro-2-hydroxybenzamide (IMD0354; Cayman chemicals).

      Immunofluorescent microscopy

      Whole eyes were excised, and corneas were punctured, followed by incubation in 4% paraformaldehyde (pH 7.5) for 30 min. Eyes were washed with PBS and incubated at 4 °C in 30% sucrose solution containing 0.05% sodium azide. Eyes were embedded in optimal cutting temperature compound, flash frozen, and sectioned. Cryosections (10 μm) were fixed in 2% paraformaldehyde, then permeabilized in PBS with 0.1% Triton-X-100, and blocked in 10% normal donkey serum. Sections were labeled with the appropriate antibodies (Table S1). All sections were counter stained with 1.6 μmol/L Hoechst. Slides were mounted with Fluoromount aqueous mounting media (Sigma-Aldrich) and imaged with a confocal laser microscope (Leica SP8; Leica) using frame-stack sequential scanning. As a control for nonspecific secondary antibody binding, primary antibody was omitted in analysis of some retinal sections (Fig. S2). ImageJ was used to obtain macrophage counts after thresholds were set to isolate positive cells and single stained population counts were then confirmed manually.

      In Situ hybridization

      REDD1 mRNA was visualized using the RNAscope 2.5 HD Assay-RED detection kit and the RNAscope Probe-Mm-Ddit4 probe (Advanced Cellular Diagnostics) targeting nucleotides 133-1480 of NM_029083.2. RNA-in situ hybridization was carried out on 10 μm whole eye cryosections following the manufacturer's protocol. REDD1 mRNA hybridization was carried out on protease and heat-treated sections. Sequential hybridization with preamplifier, amplifier, and alkaline phosphatase–labeled oligos was carried out followed by the application of a chromogenic substrate. Tissue was counter-stained with Mayer’s Hematoxylin and 0.02% ammonia water, and micrographs were captured using an AmScope T720Q brightfield microscope.

      Nuclear fractionation

      Retinas were homogenized in ice-cold hypotonic buffer (20 mM Hepes pH 7.5, 5 mM NaF, 10 μM sodium molybdate, 0.1 mM EDTA, 0.5% Nonidet P-40, and 1% protease-phosphatase inhibitors). Homogenates were centrifuged at 228g for 5 min at 4 °C to pellet nuclei. The nuclear pellet was resuspended in 50 μL complete lysis buffer (Active Motif) and incubated at 4 °C for 30 min on a shaking platform. Lysates were centrifuged at 14,000g for 10 min at 4 °C. Nuclear extract was collected as the supernatant fraction.

      DNA-binding ELISA

      NF-κB activity was quantified using a colorimetric NF-κB p65 DNA-binding ELISA (Trans AM NF-κB p65; Active Motif). Briefly, 20 μg of nuclear protein was incubated for 1 h in the presence of an immobilized oligonucleotide encoding the NF-κB consensus sequence. Binding of the p65 subunit was quantified using an anti-p65 primary antibody and horseradish peroxidase–conjugated secondary. The absorbance (λmax 450 nm) was recorded using Spectra Max M5 plate reader (Molecular Devices).

      Immunoprecipitations

      Immunoprecipitations were performed on 1000g supernatant fractions of cell lysate. HA-tag immunoprecipitation was performed using EZview Red Anti-HA affinity gel (Sigma). Beads were washed with lysis buffer [1 mM EDTA, 5 mM EGTA, 10 mM MgCl2, 50 mM β-glycerophosphate, and 0.1% NP-40] and blocked with lysis buffer containing 1% bovine serum albumin. Cells were harvested in lysis buffer supplemented with 2 mM N-ethylmaleimide, 10 mM sodium pyrophosphate, 1 mM benzamidine, 200 mM sodium vanadate, and protease inhibitor mixture (10 μL/mL) and lysed for 30 min at 4 °C. Cell supernatants were collected by centrifuging lysates for 3 min at 1000g and incubated with the appropriate affinity resin overnight at 4 °C. Affinity resins were washed with cold lysis buffer, resuspended in 1× SDS sample buffer, and boiled for 5 min. The IP was subjected to Western blot analysis.

      Western blotting

      Retinas were flash frozen in liquid nitrogen and homogenized as previously described (
      • Miller W.P.
      • Toro A.L.
      • Barber A.J.
      • Dennis M.D.
      REDD1 activates a ROS-generating feedback loop in the retina of diabetic mice.
      ). Retinal protein was quantified by DC protein assay. Equal protein from cell lysates, retinal homogenates, or nuclear extracts were combined with Laemmli buffer, boiled, and fractionated in Criterion Precast 4 to 20% gels (Bio-Rad Laboratories). Proteins were transferred to a polyvinylidene fluoride membrane, blocked in 5% milk in Tris-buffered saline Tween 20, and evaluated with appropriate antibodies (Table S1).

      Luciferase reporter assay

      Cells were cotransfected with NF-κB-TATA luciferase (500 ng) and pRL-Renilla luciferase (50 ng) plasmids. After 24 h, transfection media was removed, and cells were exposed to culture medium as indicated. Luciferase activity was measured on a FlexStation3 (Molecular Devices) using a Dual-Luciferase Assay Kit (Promega).

      PCR analysis

      Total RNA was extracted with TRIzol (Invitrogen). RNA (1 μg) was reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems) and subjected to quantitative real-time PCR (QuantStudio 12K Flex Real-Time PCR System, Thermo Fisher Scientific; RRID:SCR_021098) using QuantiTect SYBR Green Master Mix (Qiagen). Primer sequences are listed in Table S2. Mean cycle threshold values were determined for control and experimental samples. Changes in mRNA expression were normalized to GAPDH mRNA expression using the 2−ΔΔCT method.

      Statistical analysis

      Data are expressed as mean ± SD. Statistical analysis of data with more than two groups were analyzed with two-way ANOVA, and pairwise comparisons were made using the Tukey's test for multiple comparisons. Difference between two groups was determined by unpaired Student’s t test. Significance was defined as p < 0.05 for all analyses. Specific p-values for experimental groups with different means are in Table S3.

      Data availability

      All data for this publication are included in the article or are available from the corresponding author upon request.

      Supporting information

      This article contains supporting information.

      Conflict of interest

      M. D. D is guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis. The authors declare that they have no conflicts of interest with the contents of this article.

      Acknowledgments

      We thank Elena Feinstein (Quark Pharmaceuticals) for permission to use the REDD1 knockout mice and Edward Harhaj (Penn State College of Medicine) for generously providing plasmids to evaluate NF-κB signaling.

      Author contributions

      S. S., S. R. K, and M. D. D. conceptualization; S. S., A. L. T., and M. D. D. data curation; S. S., A. L. T., C. M. M., and M. D. D. formal analysis; S. S. and M. D. D. funding acquisition; S. S., A. L. T., C. M. M., A. M. V., S. A. S., and W. P. M. investigation; S. S. and M. D. D. visualization; S. S., S. R. K., and M. D. D. methodology; S. S. and M. D. D. writing–original draft; S. S., A. L. T., S. R. K., and M. D. D. writing–reviewing and editing; S. R. K., and M. D. D. resources; S. R. K. and M. D. D. supervision.

      Funding and additional information

      This research was supported by the National Institutes of Health grants R01 EY029702 and R01 EY032879 (to M. D. D.) and a Children’s Miracle Network Trainee Research Grant (to S. S.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

      Supporting information

      References

        • Teo Z.L.
        • Tham Y.-C.
        • Yu M.
        • Chee M.L.
        • Rim T.H.
        • Cheung N.
        • et al.
        Global prevalence of diabetic retinopathy and projection of burden through 2045: systematic review and meta-analysis.
        Ophthalmology. 2021; 128: 1580-1591
        • Rübsam A.
        • Parikh S.
        • Fort P.E.
        Role of inflammation in diabetic retinopathy.
        Int. J. Mol. Sci. 2018; 19: 942
        • Khalfaoui T.
        • Lizard G.
        • Ouertani-Meddeb A.
        Adhesion molecules (ICAM-1 and VCAM-1) and diabetic retinopathy in type 2 diabetes.
        J. Mol. Histol. 2008; 39: 243-249
        • Vujosevic S.
        • Micera A.
        • Bini S.
        • Berton M.
        • Esposito G.
        • Midena E.
        Proteome analysis of retinal glia cells-related inflammatory cytokines in the aqueous humour of diabetic patients.
        Acta Ophthalmol. 2016; 94: 56-64
        • Vujosevic S.
        • Micera A.
        • Bini S.
        • Berton M.
        • Esposito G.
        • Midena E.
        Aqueous humor biomarkers of müller cell activation in diabetic eyes.
        Invest. Ophthalmol. Vis. Sci. 2015; 56: 3913-3918
        • Stahel M.
        • Becker M.
        • Graf N.
        • Michels S.
        Systemic interleukin 1β inhibition in proliferative diabetic retinopathy: a prospective open-label study using canakinumab.
        Retina. 2016; 36: 385-391
        • Liu T.
        • Zhang L.
        • Joo D.
        • Sun S.-C.
        NF-κB signaling in inflammation.
        Signal. Transduct. Target. Ther. 2017; 2: 17023
        • Sun S.-C.
        The non-canonical NF-κB pathway in immunity and inflammation.
        Nat. Rev. Immunol. 2017; 17: 545-558
        • Kanwar M.
        • Chan P.-S.
        • Kern T.S.
        • Kowluru R.A.
        Oxidative damage in the retinal mitochondria of diabetic mice: possible protection by superoxide dismutase.
        Invest. Ophthalmol. Vis. Sci. 2007; 48: 3805-3811
        • Noda K.
        • Nakao S.
        • Ishida S.
        • Ishibashi T.
        Leukocyte adhesion molecules in diabetic retinopathy.
        J. Ophthalmol. 2012; 2012: 279037
        • Miller W.P.
        • Toro A.L.
        • Barber A.J.
        • Dennis M.D.
        REDD1 activates a ROS-generating feedback loop in the retina of diabetic mice.
        Invest. Ophthalmol. Vis. Sci. 2019; 60: 2369-2379
        • Ding Y.
        • Yuan S.
        • Liu X.
        • Mao P.
        • Zhao C.
        • Huang Q.
        • et al.
        Protective effects of astragaloside IV on db/db mice with diabetic retinopathy.
        PLoS One. 2014; 9: e112207
        • Tuzcu M.
        • Orhan C.
        • Muz O.E.
        • Sahin N.
        • Juturu V.
        • Sahin K.
        Lutein and zeaxanthin isomers modulates lipid metabolism and the inflammatory state of retina in obesity-induced high-fat diet rodent model.
        BMC Ophthalmol. 2017; 17: 129
        • Romeo G.
        • Liu W.-H.
        • Asnaghi V.
        • Kern T.S.
        • Lorenzi M.
        Activation of nuclear factor-κb induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes.
        Diabetes. 2002; 51: 2241-2248
        • Miller W.P.
        • Yang C.
        • Mihailescu M.L.
        • Moore J.A.
        • Dai W.
        • Barber A.J.
        • et al.
        Deletion of the Akt/mTORC1 repressor REDD1 prevents visual dysfunction in a rodent model of type 1 diabetes.
        Diabetes. 2018; 67: 110-119
        • Miller W.P.
        • Sunilkumar S.
        • Giordano J.F.
        • Toro A.L.
        • Barber A.J.
        • Dennis M.D.
        The stress response protein REDD1 promotes diabetes-induced oxidative stress in the retina by Keap1-independent Nrf2 degradation.
        J. Biol. Chem. 2020; 295: 7350-7361
        • Miller W.P.
        • Toro A.L.
        • Sunilkumar S.
        • Stevens S.A.
        • VanCleave A.M.
        • Williamson D.L.
        • et al.
        Müller glial expression of REDD1 is required for retinal neurodegeneration and visual dysfunction in diabetic mice.
        Diabetes. 2022; 71: 1051-1062
        • Nguyen Q.D.
        • Schachar R.A.
        • Nduaka C.I.
        • Sperling M.
        • Basile A.S.
        • Klamerus K.J.
        • et al.
        Dose-ranging evaluation of intravitreal siRNA PF-04523655 for diabetic macular edema (the DEGAS study).
        Invest. Ophthalmol. Vis. Sci. 2012; 53: 7666-7674
        • Yoshida T.
        • Mett I.
        • Bhunia A.K.
        • Bowman J.
        • Perez M.
        • Zhang L.
        • et al.
        Rtp801, a suppressor of mTOR signaling, is an essential mediator of cigarette smoke-induced pulmonary injury and emphysema.
        Nat. Med. 2010; 16: 767-773
        • Nadon A.M.
        • Perez M.J.
        • Hernandez-Saavedra D.
        • Smith L.P.
        • Yang Y.
        • Sanders L.A.
        • et al.
        Rtp801 suppression of epithelial mTORC1 augments endotoxin-induced lung inflammation.
        Am. J. Pathol. 2014; 184: 2382-2389
        • Pastor F.
        • Dumas K.
        • Barthelemy M.A.
        • Regazzetti C.
        • Druelle N.
        • Peraldi P.
        • et al.
        Implication of REDD1 in the activation of inflammatory pathways.
        Sci. Rep. 2017; 7: 7023
        • Lee D.K.
        • Kim J.H.
        • Kim J.
        • Choi S.
        • Park M.
        • Park W.
        • et al.
        REDD-1 aggravates endotoxin-induced inflammation via atypical NF-kappaB activation.
        FASEB J. 2018; 32: 4585-4599
        • Hou X.
        • Yang S.
        • Yin J.
        Blocking the REDD1/TXNIP axis ameliorates LPS-induced vascular endothelial cell injury through repressing oxidative stress and apoptosis.
        Am. J. Physiol. Cell Physiol. 2019; 316: C104-C110
        • Deng L.
        • Wang C.
        • Spencer E.
        • Yang L.
        • Braun A.
        • You J.
        • et al.
        Activation of the IκB kinase complex by TRAF6 requires a dimeric ubiquitin-conjugating enzyme complex and a unique polyubiquitin chain.
        Cell. 2000; 103: 351-361
        • Chen Z.J.
        Ubiquitin signalling in the NF-κB pathway.
        Nat. Cell Biol. 2005; 7: 758-765
        • Israël A.
        NF-κB activation: nondegradative ubiquitination implicates NEMO.
        Trends Immunol. 2006; 27: 395-397
        • Zhu G.
        • Wu C.-J.
        • Zhao Y.
        • Ashwell J.D.
        Optineurin negatively regulates TNFα- induced NF-κB activation by competing with NEMO for ubiquitinated RIP.
        Curr. Biol. 2007; 17: 1438-1443
        • Laplantine E.
        • Fontan E.
        • Chiaravalli J.
        • Lopez T.
        • Lakisic G.
        • Véron M.
        • et al.
        NEMO specifically recognizes K63-linked poly-ubiquitin chains through a new bipartite ubiquitin-binding domain.
        EMBO J. 2009; 28: 2885-2895
        • Whitney M.L.
        • Jefferson L.S.
        • Kimball S.R.
        ATF4 is necessary and sufficient for ER stress-induced upregulation of REDD1 expression.
        Biochem. Biophys. Res. Commun. 2009; 379: 451-455
        • Kimball S.R.
        • Jefferson L.S.
        Induction of REDD1 gene expression in the liver in response to endoplasmic reticulum stress is mediated through a PERK, eIF2α phosphorylation, ATF4-dependent cascade.
        Biochem. Biophys. Res. Commun. 2012; 427: 485-489
        • Xu D.
        • Dai W.
        • Kutzler L.
        • Lacko H.A.
        • Jefferson L.S.
        • Dennis M.D.
        • et al.
        ATF4-Mediated upregulation of REDD1 and Sestrin2 suppresses mTORC1 activity during prolonged leucine deprivation.
        J. Nutr. 2020; 150: 1022-1030
        • Antonetti D.A.
        • Barber A.J.
        • Bronson S.K.
        • Freeman W.M.
        • Gardner T.W.
        • Jefferson L.S.
        • et al.
        JDRF Diabetic Retinopathy Center Group Diabetic retinopathy: seeing beyond glucose-induced microvascular disease.
        Diabetes. 2006; 55: 2401-2411
        • Han Y.
        • Bearse Jr., M.A.
        • Schneck M.E.
        • Barez S.
        • Jacobsen C.H.
        • Adams A.J.
        Multifocal electroretinogram delays predict sites of subsequent diabetic retinopathy.
        Invest. Ophthalmol. Vis. Sci. 2004; 45: 948-954
        • Adams A.J.
        • Bearse Jr., M.A.
        Retinal neuropathy precedes vasculopathy in diabetes: a function-based opportunity for early treatment intervention?.
        Clin. Exp. Optom. 2012; 95: 256-265
        • Coughlin B.A.
        • Feenstra D.J.
        • Mohr S.
        Müller cells and diabetic retinopathy.
        Vis. Res. 2017; 139: 93-100
        • Nagai N.
        • Izumi-Nagai K.
        • Oike Y.
        • Koto T.
        • Satofuka S.
        • Ozawa Y.
        • et al.
        Suppression of diabetes-induced retinal inflammation by blocking the angiotensin II type 1 receptor or its downstream nuclear factor-κb pathway.
        Invest. Ophthalmol. Vis. Sci. 2007; 48: 4342-4350
        • Kowluru R.A.
        • Kanwar M.
        Effects of curcumin on retinal oxidative stress and inflammation in diabetes.
        Nutr. Metab. 2007; 4: 8
        • Lennikov A.
        • Hiraoka M.
        • Abe A.
        • Ohno S.
        • Fujikawa T.
        • Itai A.
        • et al.
        IκB kinase-β inhibitor IMD-0354 beneficially suppresses retinal vascular permeability in streptozotocin-induced diabetic mice.
        Invest. Ophthalmol. Vis. Sci. 2014; 55: 6365-6373
        • Kowluru R.A.
        • Koppolu P.
        • Chakrabarti S.
        • Chen S.
        Diabetes-induced activation of nuclear transcriptional factor in the retina, and its inhibition by antioxidants.
        null. 2003; 37: 1169-1180
        • Kowluru R.A.
        • Chakrabarti S.
        • Chen S.
        Re-institution of good metabolic control in diabetic rats and activation of caspase-3 and nuclear transcriptional factor (NF-kB) in the retina.
        Acta Diabetol. 2004; 41: 194-199
        • Homme R.P.
        • Sandhu H.S.
        • George A.K.
        • Tyagi S.C.
        • Singh M.
        Sustained inhibition of NF-κB activity mitigates retinal vasculopathy in diabetes.
        Am. J. Pathol. 2021; 191: 947-964
        • Liang W.-J.
        • Yang H.-W.
        • Liu H.-N.
        • Qian W.
        • Chen X.-L.
        HMGB1 upregulates NF-kB by inhibiting IKB-α and associates with diabetic retinopathy.
        Life Sci. 2020; 241: 117146
        • Zhang T.
        • Mei X.
        • Ouyang H.
        • Lu B.
        • Yu Z.
        • Wang Z.
        • et al.
        Natural flavonoid galangin alleviates microglia-trigged blood–retinal barrier dysfunction during the development of diabetic retinopathy.
        J. Nutr. Biochem. 2019; 65: 1-14
        • Zhai J.
        • Li Z.
        • Zhang H.
        • Ma L.
        • Ma Z.
        • Zhang Y.
        • et al.
        Berberine protects against diabetic retinopathy by inhibiting cell apoptosis via deactivation of the NF-κB signaling pathway.
        Mol. Med. Rep. 2020; 22: 4227-4235
        • Li W.
        • Shen X.
        • Wang Y.
        • Zhang J.
        The effect of Shengpuhuang-tang on retinal inflammation in streptozotocin-induced diabetic rats by NF-κB pathway.
        J. Ethnopharmacol. 2020; 247: 112275
        • Lipina C.
        • Hundal H.S.
        Is REDD1 a metabolic éminence grise?.
        Trends Endocrinol. Metab. 2016; 27: 868-880
        • Miller W.P.
        • Sunilkumar S.
        • Dennis M.D.
        The stress response protein REDD1 as a causal factor for oxidative stress in diabetic retinopathy.
        Free Radic. Biol. Med. 2021; 165: 127-136
        • Wei Y.
        • Gong J.
        • Yoshida T.
        • Eberhart C.G.
        • Xu Z.
        • Kombairaju P.
        • et al.
        Nrf2 has a protective role against neuronal and capillary degeneration in retinal ischemia–reperfusion injury.
        Free Radic. Biol. Med. 2011; 51: 216-224
        • Rushworth S.A.
        • MacEwan D.J.
        • O’Connell M.A.
        Lipopolysaccharide-induced expression of NAD(P)H:quinone oxidoreductase 1 and Heme oxygenase-1 protects against excessive inflammatory responses in human monocytes.
        J. Immunol. 2008; 181: 6730
        • Li X.
        • Su J.
        • Cui X.
        • Li Y.
        • Barochia A.
        • Eichacker P.Q.
        Can we predict the effects of NF-κB inhibition in sepsis? Studies with parthenolide and ethyl pyruvate.
        Expert Opin. Investig. Drugs. 2009; 18: 1047-1060
        • Sun S.-C.
        • Ganchi P.A.
        • Ballard D.W.
        • Greene W.C.
        NF-κB controls expression of inhibitor IκBα: evidence for an inducible autoregulatory pathway.
        Science. 1993; 259: 1912-1915
        • Miller W.P.
        • Sha C.M.
        • Sunilkumar S.
        • Toro A.L.
        • VanCleave A.M.
        • Kimball S.R.
        • et al.
        Activation of disulfide redox switch in REDD1 promotes oxidative stress under hyperglycemic conditions.
        Diabetes. 2022; 28db22-0355
        • Brafman A.
        • Mett I.
        • Shafir M.
        • Gottlieb H.
        • Damari G.
        • Gozlan-Kelner S.
        • et al.
        Inhibition of oxygen-induced retinopathy in rtp801-deficient mice.
        Invest. Ophthalmol. Vis. Sci. 2004; 45: 3796-3805
        • Shoshani T.
        • Faerman A.
        • Mett I.
        • Zelin E.
        • Tenne T.
        • Gorodin S.
        • et al.
        Identification of a novel hypoxia-inducible factor 1-responsive gene, RTP801, involved in apoptosis.
        Mol. Cell. Biol. 2002; 22: 2283-2293
        • Ellisen L.W.
        • Ramsayer K.D.
        • Johannessen C.M.
        • Yang A.
        • Beppu H.
        • Minda K.
        • et al.
        REDD1, a developmentally regulated transcriptional target of p63 and p53, links p63 to regulation of reactive oxygen species.
        Mol. Cell. 2002; 10: 995-1005
        • Ramachandiran S.
        • Adon A.
        • Guo X.
        • Wang Y.
        • Wang H.
        • Chen Z.
        • et al.
        Chromosome instability in diffuse large B cell lymphomas is suppressed by activation of the noncanonical NF-κB pathway.
        Int. J. Cancer. 2015; 136: 2341-2351
        • Zhong Y.
        • Li J.
        • Chen Y.
        • Wang J.J.
        • Ratan R.
        • Zhang S.X.
        Activation of endoplasmic reticulum stress by hyperglycemia is essential for müller cell–derived inflammatory cytokine production in diabetes.
        Diabetes. 2012; 61: 492-504
        • Brown D.M.
        • Nguyen Q.D.
        • Marcus D.M.
        • Boyer D.S.
        • Patel S.
        • Feiner L.
        • et al.
        Long-term outcomes of ranibizumab therapy for diabetic macular edema: The 36-month results from two phase III trials: RISE and RIDE.
        Ophthalmology. 2013; 120: 2013-2022
        • Dai W.
        • Miller W.P.
        • Toro A.L.
        • Black A.J.
        • Dierschke S.K.
        • Feehan R.P.
        • et al.
        Deletion of the stress-response protein REDD1 promotes ceramide-induced retinal cell death and JNK activation.
        FASEB J. 2018; 32: 6883-6897
        • Dennis M.D.
        • Coleman C.S.
        • Berg A.
        • Jefferson L.S.
        • Kimball S.R.
        REDD1 enhances protein phosphatase 2A-mediated dephosphorylation of Akt to repress mTORC1 signaling.
        Sci. Signal. 2014; 7: ra68