IκB Kinase-β Inhibitor IMD-0354 Beneficially Suppresses Retinal Vascular Permeability in Streptozotocin-Induced Diabetic Mice

Abstract

Purpose.: The purpose of the present study is to evaluate the effect of selective IKK-β inhibition by IMD-0354 on inflammation, apoptosis, and angiogenesis in diabetic retinopathy (DR).

Methods.: Six weeks after administration of a streptozotocin (STZ) injection, before diabetic retinopathy (DR) was evident, one group of STZ-induced diabetic mice was systemically administered with IMD-0354 (30 mg/kg) daily for another 6 weeks. Ten weeks after the STZ injection, with DR already present, another group of STZ-induced diabetic mice was administered IMD-0354 for 2 weeks. As controls, nondiabetic mice of the same age were treated with IMD-0354 for 6 weeks, and diabetic mice were treated with 10 μL of dimethyl sulfoxide (DMSO) for 6 weeks. Using these groups of mice, the following effects of IMD-0354 were analyzed: (1) inhibition of nuclear factor-κB (NF-κB) activation, (2) retinal morphology, (3) apoptotic signaling by cleaved caspase-3, (4) retinal vascular permeability, (5) angiogenesis of the retina, and (6) retinal production of VEGF.

Results.: Systemic administration of IMD-0354 for 6 weeks to week-6 diabetic mice caused significant reduction in the loss of retinal ganglion cells and apoptotic signaling, with preservation of retinal vascular integrity and suppression of retinal VEGF expression. When inhibition of NF-κB activation treatment started after the onset of STZ-induced DR (week 10), IMD-0354 was still effective in preventing further DR progression while the vascular integrity was preserved.

Conclusions.: The present data indicate that NF-κB activation is the key step in the development of DR. Its suppression by IMD-0354 may present a promising therapeutic strategy for DR, especially in the early stages of the disease.

Introduction

Diabetic retinopathy (DR) is one of the leading causes of blindness in working-age individuals and is a serious concern to individuals with diabetes mellitus (DM) types 1 and 2. ¹ Microvascular damage results in higher metabolic and oxygen demands, which lead to retinal ischemia, while hypoxia and oxidative stress can damage the retina further as they are common stimuli for both chronic inflammation and angiogenesis. Vascular endothelial growth factor is known to play a central role in the progression of pathological DR. ² Following activation of the VEGF signaling pathway, vessel leakage and angiogenesis occur in the retina, leading to a local inflammatory response that results in vascular sprouting. ³ These new vessels are easily disrupted, causing vitreous hemorrhage and retinal detachment, which often lead to vision loss. ⁴ Upstream from the VEGF, signaling pathway and activation of nuclear factor-κB (NF-κB) induce the transcription of a variety of proinflammatory proteins, such as inducible nitric oxide synthase (iNOS), cyclooxygenase-2 (COX-2), and intercellular adhesion molecule 1 (ICAM-1). ⁵ Recent studies have further shown that VEGF expression in human macrophages is downregulated by inhibition of NF-κB activation, ⁶ and this limits the diabetes-induced increase in the expression of ICAM-1 and VEGF in vivo. ⁷  

In the diabetic retina, NF-κB localized in the subretinal membranes and microvessels ^8,9 is considered to be responsible for accelerated loss of the pericytes, ¹⁰ as well as the degradation of the tight junction proteins of the retinal vessels, such as ZO-1 and claudin-5. ¹¹ Under normal conditions, NF-κB exists in the cytoplasm in an inactive form, associated with regulatory proteins called inhibitors of κB (IκB). The phosphorylation of IκB, an important step in NF-κB activation, is mediated by IκB kinase (IKK) under inflammatory stimuli or stress. The IKK complex consists of at least three subunits, including the kinases IKK-α and IKK-β (also called IKK-1 and IKK-2) ¹² and the regulatory subunit IKK-γ. ¹³ IKK activation initiates IκB phosphorylation; the IκB is then ubiquitinated, which makes it a target for degradation by the 26S proteasome, ¹⁴ thereby releasing NF-κB dimers from the cytoplasmic NF-κB–IκB complex and allowing them to translocate to the nucleus. Nuclear factor-κB then binds to the κB enhancer elements of the target genes, inducing the transcription of proinflammatory genes. Proinflammatory cytokines, such as IL-1β and TNF-α, are regulated by NF-κB activation and are known to be stimuli that cause the activation of IKK. Since NF-κB is the main factor in the positive feedback loop of inflammation, inhibition of its activation may prove to be an effective therapy for intraocular inflammation such as DR. However, recent research has indicated that NF-κB also plays an important role in cell cycle and survival, ^15–17 neuron function, and even higher functions such as memory. ¹⁸  

Recently, our group synthesized the low-molecular-weight compound, IMD-0354 (N-[3,5-Bis-trifluoromethyl-phenyl]-5-chloro-2-hydroxy-benzamide). IMD-0354 is a non–adenosine triphosphate–binding (ATP-binding) competitive selective IKK-β inhibitor, particularly when it is induced by proinflammatory cytokines such as TNF-α and IL-1β. ^19–21 Unlike other nonselective NF-κB inhibitors, such as dehydroxymethylepoxyquinomicin (DHMEQ), which block the DNA-binding activity of NF-κB, ²² IMD-0354 prevents ATP attachment to IKK-β. It thus decreases its ability to phosphorylate NF-κB–IκB dimers without a complete systemic disruption of the NF-κB pathway, as IKK-α is still able to activate NF-κB. In this way, administration of IMD-0354 could be a suitable and safe strategy to control inflammation and neovascularization in DR. 

Recent reports have shown that IMD-0354 is effective in acute and subacute inflammatory diseases such as myocardial ischemia/reperfusion injury, ²¹ insulin resistance, ²³ allergic inflammation in an acute mouse model of asthma, ²⁰ and bleomycin-induced lung fibrosis in mice through the suppression of plasminogen activator inhibitor-1 (PAI-1) production via the suppression of NF-κB activation. ¹⁹ In addition, we also previously demonstrated that IMD-0354 is effective in ameliorating endotoxin-induced uveitis (EIU) in rats. ²⁴ No adverse effects of IMD-0354 were detected in any of these reports, either in vitro or in vivo. ^19–21 Preclinical toxicology studies also confirmed that IMD-0354 has a wide safety margin and good tolerability, in compliance with Good Laboratory Practice regulations. Studying the effect of selective IKK-β inhibition by IMD-0354 on inflammation, apoptosis, and angiogenesis in diabetic retinas is therefore of great medical interest. To investigate this in the present study, we evaluate the effect of IMD-0354 on streptozotocin (STZ)-induced diabetic retinopathy in mice. 

Materials and Methods

Animals and Induction of Diabetes

Nine-week-old C57BL/6 male mice (Hokudo Co., Sapporo, Japan) were maintained under specific pathogen-free conditions. All procedures involving the animals were performed in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research resolution on the use of animals in research. 

The timeline of the experiment is presented in Figure 1. One week after the animals' arrival, in order to reduce stress associated with transfer, mice were randomly assigned to diabetic and nondiabetic groups. Diabetes was induced by daily intraperitoneal (i.p.) injections of STZ (Sigma-Aldrich Corp., St. Louis, MO) in 100 μL Na-citrate buffer, pH 4.5 (40 mg/kg) to 10-week-old mice for 4 days. The development of diabetes was defined by the presence of glucose in the urine at concentrations > 500 mg/dL by 4 weeks after the first STZ injection (week 5) as evaluated using Aution Sticks (10EA; Arkray, Kyoto, Japan) (data not shown). Blood sugar levels in venous blood samples obtained from the tail vein were measured with a Medisafe GR-102 glucometer (Terumo, Tokyo, Japan) before the injection of the STZ and at the end point of the experiment. Body weight was measured at the end point of the experiment. 

Figure 1

 

Timeline of the experiment.

In Vivo IMD-0354 Treatment

IMD-0354 was dissolved in 10 μL of a dimethyl sulfoxide (DMSO) vehicle (Sigma-Aldrich, Tokyo, Japan) immediately before use. One group of diabetic mice was systemically administered with IMD-0354 (30 mg/kg) daily from week 6 of STZ injection (week 7) for 6 consecutive weeks until the end point of the experiment (week 13); another group of diabetic animals received IMD-0354 (30 mg/kg) daily for the last 2 weeks of the experiment (weeks 11–13). To evaluate the safety and side effects of the IMD-0354, nondiabetic mice were also administered with IMD-0354 (30 mg/kg) daily for 6 weeks from week 7 of the experiment. Since there have been reports that DMSO at high doses may have slight inhibiting effects on NF-κB activation, ²⁵ an additional group of diabetic mice was treated daily with 10 μL DMSO for 6 weeks from week 7 of the experiment. Naïve mice were used as controls. Twelve weeks after the initial STZ injection (week 13), mice were euthanized by i.p. injection of sodium pentobarbital at 100 mg/kg (Sigma-Aldrich) and the eyes were harvested. 

Morphological Analyses

The eyes of the mice were fixed with Superfix solution (Kurabo, Osaka, Japan) and processed for paraffin sectioning. Sagittally cut sections (5-μm thick) through the optic nerve were stained with hematoxylin and eosin (H&E). To isolate the optic nerve disc, several sections of roughly the same size were made; these were then stained with H&E. The neuronal cell number in the ganglion cell layer (GCL) was quantified by counting the nuclei. The retinal thickness was evaluated in two areas, 200 μm apart nasally or temporally. Data for number of cells in the GCL and retinal thickness in each sample were evaluated from three consecutive sections and averaged. Unless otherwise stated, imaging and evaluation of sections were performed under an Olympus BX51 fluorescence microscope (Olympus, Tokyo, Japan) by a masked researcher using the proprietary software. 

Immunohistochemical Studies for Cleaved Caspase-3 and NF-κB Activity

In one section per eye, a cut was made through the optic nerve disc for immunohistochemical study. Unstained sections were then rehydrated and subjected to heat-induced antigen retrieval in a microwave oven for 5 minutes in a 10-mM Na citrate buffer, pH 6.0. The sections were incubated in a blocking solution composed of 10% normal bovine serum in PBS (8 g NaCl, 2.9 g [Na₂HPO₄]12H₂0, 0.2 g KCl, 0.2 g KH₂PO₄, in 1 L H₂O; pH 7.4) and incubated with a rabbit anticleaved caspase-3 antibody (1:100; Cell Signaling Technology, Danvers, MA, USA) for 12 hours. After washing the sections with PBS three times, specific antibody staining was visualized by a secondary antibody dye conjugate (1:1000; Invitrogen, Carlsbad, CA, USA) with red florescence. 

For NF-κB immunostaining, unstained sections were treated with an NF-κB p65 antibody (1:100; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and then a secondary antibody dye conjugate (1:1000; Invitrogen) with red fluorescence as above. Since NF-κB is widely present in intercellular tissue, to detect its specific expression within cell nuclei the sections were further stained with a 1:1000 dilution of 4′,6-diamidino-2-phenylindole (DAPI; Invitrogen), which binds strongly to A-T–rich regions in the DNA. 

Visualization of Retinal Vessel Breakdown

Retinal vascular leakage was visualized using 1 mg/mouse FITC-conjugated concanavalin A (Vector Labs, Burlingame, CA, USA). Under deep anesthesia by i.p. injection of 50 mg/kg sodium pentobarbital (Sigma-Aldrich), the thoracic cavities were opened and FITC-conjugated concanavalin A diluted in 300 μL PBS was perfused via the left ventricle of the heart on a preparation table. After 15 minutes, the eyes were enucleated and fixed with 4% paraformaldehyde (Sigma-Aldrich) for 12 hours. The retinas were then carefully isolated using an Olympus SZ-STB1 microscope (Olympus) and examined under an Eclipse Ni fluorescence microscope (Nikon, Tokyo, Japan). 

Vascular Endothelial Growth Factor Expression in Retinal Tissues

Freshly isolated mice retinas were placed in 700 μL PBS followed by sonication on ice for 50 seconds. The lysates were then centrifuged at 5000g for 10 minutes at 4°C. The total protein concentration was measured with a bicinchoninic acid protein assay kit (Pierce, Rockford, IL, USA). All samples were adjusted to contain a total of 30 μg protein for subsequent assay. The concentration of VEGF in the supernatant was determined using a commercially available mouse VEGF ELISA kit (R&D Systems, Minneapolis, MN, USA). 

Statistical Analyses

All values were expressed as the mean ± SD for the respective groups. Statistical analyses were carried out using the nonparametric Mann-Whitney U test. P values less than 0.05 were considered significant. 

Results

The naïve control mice (30.7 ± 2.1 g) exhibited a significantly higher body weight by the end of the study period compared to nontreated diabetic mice (19.4 ± 1.6 g) (P < 0.01). The administration of IMD-0354 at 30 mg/kg did not significantly affect the body weight; both groups of diabetic mice treated with IMD-0354, for either 6 (20.2 ± 1.0 g) or 2 (20.7 ± 3.4 g) weeks, had a similar body weight to nontreated STZ mice (P > 0.05). IMD-0354 treatment also had no apparent effect on body weight in normoglycemic mice (28.5 ± 2.2 g) when compared with naïve controls (P > 0.05). Dimethyl sulfoxide treatment for 6 weeks was shown to have a slight but significant (P < 0.05) impact on body weight (15.5 ± 4.3 g) compared to nontreated diabetic mice. This may be attributed to the injection stress to which mice in the DMSO group were subjected. 

All mice were normoglycemic at the beginning of the study (data not shown); however, the blood glucose levels were significantly increased (P < 0.01) in diabetic untreated mice (718.5 ± 61.2 mg/dL) compared to the naïve control mice (141.8 ± 57.1 mg/dL). The IMD-0354 treatment had no apparent effect on the blood glucose concentrations in diabetic mice treated for 6 (736.6 ± 39.3 mg/dL) and 2 (734.4 ± 55.2 mg/dL) weeks, compared with untreated diabetic mice (P > 0.05). Blood glucose remained normoglycemic in nondiabetic mice treated with IMD-0354 for 6 weeks to the extent that there was no significant difference compared to naïve controls (P > 0.05). 

At first we used the immunohistochemistry of NF-κB p65 to test whether systemically administered IMD-0354 really penetrated into the retina beyond the blood–retina barrier and suppressed NF-κB activity (Fig. 2). Immunoreactivity toward NF-κB p65 was markedly observed in the nuclei of cells in the diabetic mice retinas (Fig. 2A) compared to naïve controls (Fig. 2D). Nuclear factor-κB p65 nuclear colocalization was reduced in diabetic retinas treated with IMD-0354 for 6 weeks (Fig. 2B). In DR mice treated with IMD-0354 for 2 weeks, at week 10, after STZ injection, NF-κB activity increased in spite of the treatment (Fig. 2C). Nuclear factor-κB–positive cells were scarce in naïve controls (Fig. 2D). The number of the NF-κB–immunopositive cells was quantified and presented in Figure 3. Nuclear factor-κB–positive cells were significantly more numerous in the nontreated diabetic animals compared to naïve controls (P < 0.05). Six weeks of IMD-0354 treatment significantly reduced the number of NF-κB–positive cells (P < 0.05) in the diabetic mice when compared with nontreated STZ animals, bringing the cell number back to the normal levels, with no significant difference between diabetic animals treated with IMD-0354 for 6 weeks and naïve controls (P > 0.05). IMD-0354 treatment for 2 weeks, initiated at week 10 after STZ injection, failed to downregulate NF-κB activation in the retina, with the number of NF-κB–immunopositive cells significantly higher than in naïve controls (P < 0.05), and there was no significant reduction of NF-κB colocalization compared to nontreated diabetic animals (P > 0.05). Therefore, these data indicate that systemically administered IMD-0354 could effectively inhibit NF-κB activity within the diabetic retina if treated before the onset of DR but would not be effective in patients that have it already developed. 

Figure 2

 

Effect of IMD-0354 on NF-κB p65 activation in the retinas of STZ-induced diabetic mice. Multiple NF-κB–positive cells, which colocalized with the nucleus (marked by arrows), are seen in diabetic mouse retinas (A). Nuclear factor-κB nuclear colocalization is downregulated by 6 weeks of treatment with IMD-0354 (pre-DR onset) (B). Nuclear factor-κB–positive cells are widely present in the retinas of diabetic mice treated with IMD-0354 for 2 weeks from week 10 after STZ injection (after DR onset) (C). Virtually no NF-κB colocalization is present in the naïve control group (D).

Figure 3

 

Quantitative analysis of NF-κB p65–positive cells within the retinas of STZ-induced mice treated with IMD-0354. Cells with nuclear colocalization of NF-κB (marked by arrows in Fig. 2) were counted in one section per eye from the whole retinal layer and averaged. Data are shown as the mean ± SD (n = 5); not significant (n.s.) = P > 0.05; *P < 0.05 = significantly different.

To further study this effect, the impact of IMD-0354 on retinal morphology in diabetic mice was evaluated. As shown in Figure 4, a decrease was observed in the overall cell number, and patches of complete cell loss were visible in untreated diabetic mice retinas; in contrast, densities and distributions of cells in the GCL in IMD-0354–treated mice retinas were well preserved in the mice treated with IMD-0354 for 6 weeks. Two weeks of treatment after 10 weeks of STZ injection with IMD-0354 and 6-week treatment with DMSO showed little to no effect on retinal morphology in diabetic animals. IMD-0354 itself did not cause morphological alterations in nondiabetic retinas. 

Figure 4

 

The effect of IMD-0354 on retinal morphology in the STZ-induced diabetic mice. Representative H&E staining of the retinas of STZ-induced mice treated with IMD-0354 or a vehicle solution is shown. Retinal ganglion cell (RGC) loss (marked by an asterisk) is visible in the STZ-induced diabetic mice (A). Retinal ganglion cell loss is well preserved in the group treated with IMD-0354 for 6 weeks from week 6 (B). Some reduction in RGC is visible in the group treated with IMD-0354 for 2 weeks from week 10 after STZ injection (C). Patches of RGC loss (marked by an asterisk) are seen in the group treated with DMSO for 6 weeks from week 6 after STZ injection (D). No pathologic changes are observed in nondiabetic mice treated with IMD-0354 (E) or naïve controls (F).

Quantitative data in Table 1 demonstrates that the retinas of untreated diabetic mice showed significant loss of cells in the GCL compared with naïve controls (P < 0.05). Reduction in the number of GLC cells was prevented in diabetic retinas treated with IMD-0354 for 6 weeks (P < 0.05); however, 2 weeks of IMD-0354 treatment started at week 10 after STZ injection did not have a significant ameliorative effect on retinal GLC cell loss compared with nontreated diabetic animals (P > 0.05). Dimethyl sulfoxide treatment for 6 weeks in diabetic animals also had no significant effect (P > 0.05) compared to nontreated STZ mice. Retinal thickness was not significantly affected in any of the experimental groups (Table 2). 

Table 1

 

Number of Nuclei in the Ganglion Cell Layer of Diabetic and Control Mice That Were Either Not Treated or Treated With IMD-0354

	Eyes, n	Cells per Section	95% Confidence Interval	P Value
DM	5	263 ± 21.47	245.14–282.77
DM + IMD-0354, 6 weeks	5	306 ± 20.83	288.4–324.92	0.02
DM + IMD-0354, 2 weeks	5	285 ± 24.49	264.19–307.13	0.14
DM + DMSO, 6 weeks	5	274 ± 7.63	267.14–282.1	0.08
IMD-0354, 6 weeks	5	315 ± 32.61	287.0–344.19	0.02
Naïve	9	305 ± 9.17	297.61–313.69	0.02

Table 2

 

Retinal Thickness 200 μm From the Optic Nerve Disc in Diabetic and Control Mice That Were Either Not Treated or Treated With IMD-0354

	Eyes, n	Micrometers	95% Confidence Interval	P Value
DM	5	135 ± 16.64	118.85–151.48
DM + IMD-0354, 6 weeks	5	142 ± 23.93	118.73–165.65	0.39
DM + IMD-0354, 2 weeks	5	128 ± 10.3	117.12–140.44	0.25
Naïve	9	143 ± 7.56	134.55–151.67	0.25

As shown in Figures 5 and 6, the active form of cleaved caspase-3 was upregulated in the GCL cells and weakly upregulated in the inner nuclear layer (INL) cells of diabetic retinas compared to those of nondiabetic controls (P < 0.05). Such upregulation of cleaved caspase-3 was significantly reduced in the group treated with IMD-0354 for 6 weeks (P < 0.05). However, 2 weeks of treatment since week 10 of STZ injection had no significant effect on reduction of cleaved caspase-3–positive cell number when compared to nontreated diabetic retinas (P > 0.05). Dimethyl sulfoxide administration had no effect on the number of cleaved caspase-3 cells compared with nontreated diabetic animals (P > 0.05). These results suggest that IMD-0354 may reduce retinal apoptosis when treatment begins before the onset of DR, but fails to prevent apoptosis in retinal tissue of diabetic mice when treatment starts in already developed DR. 

Figure 5

 

Cleaved caspase-3 immunohistochemical staining of the retinas of STZ-induced mice treated with IMD-0354. Multiple apoptotic cells (marked by arrows) are visible in the cleaved caspase-3 immunostained retinas of the diabetic mice (A). Apoptotic signaling is downregulated by treatment with IMD-0354 (30 mg/kg) for 6 weeks from week 6 of STZ injection (B). Some reduction in apoptotic signaling is visible in diabetic mice treated with IMD-0354 for 2 weeks from week 10 after STZ injection (C). Dimethyl sulfoxide–treated diabetic mice showed multiple apoptotic cells (D). Apoptotic cells are scarce in IMD-0354–treated nondiabetic mice (E) and naïve controls (F).

Figure 6

 

Quantitative analysis of cleaved caspase-3–positive cells in the STZ-induced mice retinas treated with IMD-0354. Cleaved caspase-3–positive cells (marked by arrows in Fig. 5) were counted in one section per eye from the whole retinal layer. Data are shown as the mean ± SD (n = 5); n.s. = P > 0.05; *P < 0.05 = significantly different.

To evaluate the effects of IMD-0354 on the increased hyperpermeability of retinal vessels, the mice were perfused with FITC-conjugated concanavalin A and the flat-mounted retinal sections were analyzed. In the diabetic mice retinas, prominent leakage of the fluorescence dye from retinal vessels could be observed (Fig. 7A). This leakage was decreased in the DR mice treated with IMD-0354 for 6 weeks (Fig. 7B). Similar but less notable effects of IMD-0354 were observed in DR animals treated with IMD-0354 10 weeks after STZ injection for 2 weeks (Fig. 7C). The DMSO treatment had no ameliorative effects on the retinal vessel leakage in DR mice (Fig. 7D). No vascular integrity loss was observed in nondiabetic animals with (Fig. 7E) or without IMD-0354 (Fig. 7F) treatment. 

Figure 7

 

Flat-mounted retinal sections of the STZ-induced mice treated with IMD-0354 after perfusion with FITC-conjugated concanavalin A. Retinal vascular leakage associated with DR was visualized in flat-mounted retinal sections after perfusion with FITC-conjugated concanavalin A. The blood–retinal barrier is compromised in diabetic mice as multiple sites of vascular leakage (marked by arrows) are observed (A). Microvascular meshwork integrity is well preserved in diabetic mice treated with IMD-0354 for 6 weeks from week 6 (B), as well as in those treated for 2 weeks from week 10 after STZ injection (C). Dimethyl sulfoxide–treated diabetic mice show multiple leakage sites similar to nontreated diabetic animals (D). Neither nondiabetic mice treated with IMD-0354 (E) nor naïve controls (F) show any signs of blood–retinal barrier failure.

The effect of IMD-0354 treatment on the concentrations of VEGF in the retinas of STZ-induced diabetic mice was determined using a VEGF ELISA kit (Fig. 8). The concentrations of VEGF in the diabetic retinas at the end point (12 weeks after the initial injection of STZ) were significantly increased compared with those in nondiabetic retinas (P < 0.05). In diabetic mice treated for 6 and 2 weeks with IMD-0354, VEGF concentrations were significantly reduced compared to diabetic retinas (P < 0.05). 

Figure 8

 

Vascular endothelial growth factor production in the retinas of STZ-induced mice treated with IMD-0354. Vascular endothelial growth factor protein levels within retinas were determined using a specific ELISA kit. All analyses were performed at the end point of each experiment group. Data are shown as the mean ± SD (n = 4); *P < 0.05 = significantly different.

Discussion

In our study, we evaluated the effects of the NF-κB inhibitor IMD-0354 on STZ-induced diabetes. Streptozotocin is a glucosamine-nitrosourea compound that is cytotoxic to pancreatic beta cells after being transported through the glucose transporter-2 (GLUT2), ²⁶ and therefore it is used to generate diabetic model animals. Damage to the pancreatic beta cells leads to increased blood glucose, which causes metabolic changes in various tissues and organs, including the eyes. The increased retinal expressions of iNOS, COX-2, and VEGF are known to be key factors responsible for diabetes-induced retinal inflammation and neovascularization. Hypoxia results in worsening retinal ischemia and is a common stimulus for the accumulation of macrophages and other immune cells. ²⁷ The activation of these cells results in the production of cytokines, such as TNF-α and IL-1β, which act to induce the expression of proinflammatory proteins. The increased expression of many inflammatory proteins is regulated at the level of gene transcription through the activation of proinflammatory transcription factors, including NF-κB, which eventually leads to the synthesis of many cytokines, chemokines, acute phase proteins, and proinflammatory molecules that also act as NF-κB activators, thus closing the inflammatory loop. 

Increased levels of TNF-α and other inflammatory cytokines lead to the degradation of the tight junction complex proteins of retinal vessels, such as ZO-1 and claudin-5. Along with tissue ischemia, this leads to increased retinal endothelial cell permeability, loss of vascular integrity, and neovascularization. ¹¹ Although loss of integrity of the blood barrier seems to be mostly dependent on the NF-κB activation pathway, other mechanisms, such as increased production of reactive oxygen species (ROS), are involved in retinal cell death in DR. ²⁸  

In the current study, we have demonstrated that the IKK-β inhibitor IMD-0354 reduced the activation of NF-κB in retinal tissues of diabetic mice. Furthermore, our results indicate that this downregulation of NF-κB activation by treatment with IMD-0354 for 6 weeks in mice that had not yet developed DR (from week 6) significantly reduced the loss of ganglion cells and the cleavage of caspase-3, as well as retinal vascular leakages. Similar beneficial effects were also induced after the onset of STZ-induced DR (in mice that received 2 weeks of treatment from week 10 after STZ injection). Additionally, IMD-0354 treatment reduced the expression of VEGF within the diabetic retinas treated for 6 or 2 weeks. Since it is widely known that NF-κB activation is associated with VEGF expression, ^6,7 it is reasonable to speculate that well-preserved vascular integrity in IMD-0354–treated mice retinas may be observed, along with better preservation of tight junction proteins, and this was well supported by our flat mount data. 

Dimethyl sulfoxide was used as a solvent in our study, and this has been reported to inhibit NF-κB activation at high doses (10 mL/kg). ²⁵ However, the amount of DMSO used in the present study was 30 times smaller than the amount previously reported to have inhibited NF-κB activation. No significant effects were observed in an additional group of week 6 diabetic mice treated with 10 μL of DMSO daily for 6 weeks. Therefore, we assume that the inhibition of NF-κB activation and DR ameliorative effects were indeed due to the IMD-0354 treatment and not the effects of its solvent. 

In terms of any adverse effects of the 30 mg/kg of IMD-0354 given daily to treated mice, no noticeable side effects have been shown to date. In fact, in our study, the IMD-0354 treatment had no effect on blood glucose concentrations, and there was no weight loss/gain in either diabetic or nondiabetic animals, nor were there any systemic side effects or pathologic changes in the eyes of nondiabetic mice treated with IMD-0354 compared to naïve controls. This is consistent with the reports of other authors. ²⁹  

The present data on the role of NF-κB in diabetes-induced retinal inflammation and neovascularization, as well as the benefits of its downregulation, confirm and clarify the observations of several authors, which previously indicated that NF-κB activation is crucially involved the development of DR and neovascularization in general. ^10,11,29 Additionally, it is well known that hyperglycemia stimulates production of PAI-1 in vascular small endothelial cells ^30,31 and induces VEGF production in endothelial cells. ³² Since high glucose levels generate oxidative stress, hyperglycemia also plays a pivotal role in inducing retinal inflammation, vascular damage, and neovascularization through VEGF, inflammatory cytokines, and PAI-1 via NF-κB activation in DR. These results suggested that IMD-0354 significantly suppresses retinal inflammation, vascular damage, and neovascularization. 

IMD-0354 is also available as an oral form prodrug known as IMD-1041, which has similar efficiency and may prove promising in the clinical supportive treatment of DR, especially during the early stages of diabetes. As IMD-1041 has been proven to be safe and well tolerated in phase 2 clinical studies, for our next project we intend to study the efficacy of IMD-1041 on diabetic retinopathy. 

Acknowledgments

We thank Fumihito Hikage (Sapporo Medical University) for technical assistance. 

This work was supported in part by the Northern Advances Center for Science & Technology (NOASTEC) Foundation. 

Disclosure: A. Lennikov, None; M. Hiraoka, None; A. Abe, None; S. Ohno, None; T. Fujikawa, None; A. Itai, None; H. Ohguro, None