Lipocalin 2 Plays an Important Role in Regulating Inflammation in Retinal Degeneration Free

Several blinding diseases of the retina are characterized by the death of retinal pigmented epithelium (RPE) and photoreceptor cells. In retinitis pigmentosa (RP), an inherited retinal disease characterized by a progressive loss of photoreceptor cells, an activation of prosurvival signaling cascades, involving upregulation of several growth factors, cytokines, and antioxidants, has been observed (1). In age-related macular degeneration (AMD), a major cause of visual impairment in elderly people, several immune responses are activated in the form of inflammatory cytokines, chemokines, Abs, and T cells in both animal models and patients (2). Accumulating evidence suggests that activation of immune responses plays an important role in progression of these blinding diseases. We previously reported an increase in acute phase protein lipocalin 2 (LCN2) coinciding with retinal degeneration in ATP-binding cassette subfamily A member 4 (Abca4)^−/− retinol dehydrogenase 8 (Rdh8)^−/− mice (3). LCN2 is also known as 24p3 or neutrophil gelatinase-associated lipocalin and is a member of the lipocalin superfamily known for its role in cellular transport of lipophilic molecules as fatty acids, iron, retinoids, and steroids. LCN2 is a multifunctional innate immunity protein and can augment cellular tolerance to oxidative stress (4) and, indeed, roles of LCN2 have been suggested under stress conditions and degenerative diseases. In the CNS, LCN2 deficiency has been associated with tissue inflammation (5). Lcn2-deficient mice were found to be highly sensitive to bacterial sepsis (6). Several studies have demonstrated that LCN2 protects against cellular stress, inflammation, and cell death (7–10).

Although a few studies have implicated the involvement of LCN2 in eye disease, the possible roles of LCN2 in retinal degeneration have not been fully elucidated. LCN2 was found upregulated among other acute phase response and inflammatory proteins in the retina of rodent models for diabetes and retinal ischemia/reperfusion injury (11). Increased levels of LCN2 have been reported in a mouse model of Bardet–Biedl syndrome (12). A pivotal role of LCN2 in the development of demyelinating optic neuritis in a mouse model of autoimmune optic neuritis has been demonstrated (13). Valapala et al. (14) identified LCN2 as a contributory factor in inducing chronic inflammatory response in Cryba1 conditional knockout (KO) mice, a mouse model with AMD-like pathology. Lastly, Sinha and colleagues (15) generated genetically engineered mice in which lysosome-mediated clearance in RPE cells is compromised, causing the development of features of early AMD. They further proposed the involvement of an AKT2–NF-κB–LCN2 signaling axis in activating the inflammatory responses in these mice, suggesting this pathway as a potential target for AMD treatment. In humans, Ghosh et al. (15) observed an increased infiltration of LCN2⁺ neutrophils in the choroid and retina of early AMD patients as compared with age-matched controls. These studies, including our observation of increased expression of LCN2 in mouse retinal degeneration (3), reinforce the possible significance of LCN2 in retinal inflammation and retinal degenerative diseases.

In the present study, Lcn2^−/−Abca4^−/−Rdh8^−/− triple-KO mice were generated and RPE cells differentiated from human-induced pluripotent stem cells (hiPS-RPE) were employed to investigate the role of LCN2 in retinal inflammation and degeneration. Our results provide evidence that LCN2 could regulate prosurvival responses and retinal inflammation in mice and humans.

Lcn2^−/− mice were purchased from The Jackson Laboratory (Bar Harbor, ME). All mice were housed in the animal facility at the School of Medicine, Case Western Reserve University, where they were maintained either under complete darkness or in a 12-h light (∼10 lx)/12-h dark cycle environment in a pathogen-free environment. Both male and female mice at 1 mo of age were used in the study. Lcn2^−/− mice were crossed with Abca4^−/−Rdh8^−/− mice to generate Lcn2^−/−Abca4^−/−Rdh8^−/− mice. 129SV or littermates of mutant mice were used as controls. Only the mice with RPE65 leucine variant and free of rd8 mutation were employed in the study. Genotyping for Lcn2^−/− or wild-type (WT) mice was performed using the primers with LCN2-KO and common for KO allele and LCN2-WT and common for WT allele: LCN2-KO, 5′-CCTTCTAT GCCTTCTTGACG-3′; LCN2-common, 5′-TAGGGGATGCCACATCTCA-3′; LCN2-WT, 5′-TGGAGGTGACATTGTAGCTATTG-3′. Genotyping for Abca4^−/−Rdh8^−/− mice was performed as described previously (16). All animal procedures and experiments were approved by the Case Western Reserve University Animal Care Committees and conformed to recommendations of both the American Veterinary Medical Association Panel on Euthanasia and the Association of Research for Vision and Ophthalmology.

Mice were dark adapted for 48 h before being exposed to light. Light-induced degeneration was induced by exposing mice to 10,000 lx of diffuse white fluorescent light (150 W spiral lamp; Commercial Electric, Cleveland, OH) for 30 min. Before such light exposure, pupils of mice were dilated with mixture of 0.5% tropicamide and 0.5% phenylephrine hydrochloride (Midorin-P; Santen Pharmaceutical, Osaka, Japan). After exposure, animals were kept in the dark until further evaluation.

All procedures to make sections for retinal histology and immunohistochemistry followed established methods (17, 18). For H&E staining eye cups were fixed in 10% formalin for 72 h followed by paraffin embedding. Five-micrometer-thick sections were cut and stained with H&E for analysis by light microscopy. For immunohistochemistry, eye cups were embedded in 4% paraformaldehyde/1% glutaraldehyde overnight followed by cryosectioning. The following Abs were used for immunohistochemistry; rabbit anti–Iba-1 Ab (1:400; Wako, Osaka, Japan), rabbit anti–glial fibrillary acidic protein (GFAP) Ab (1:400; Dako, Carpinteria, CA), rabbit NF-κB p65 Ab (1:1000; eBioscience, San Diego, CA). Secondary Abs used were anti-mouse, anti-rabbit Alexa Fluor 555 (Invitrogen, Carlsbad, CA). Images were captured by a confocal microscope (LSM; Carl Zeiss, Thornwood, NY).

HRAII (Heidelberg Engineering, Heidelberg, Germany) for scanning laser ophthalmoscopy (SLO) and ultra-high resolution spectral domain–optic coherence tomography (SD-OCT; Bioptigen SD-OCT Envisu C2200; Bioptigen, Research Triangle Park, NC) were employed for in vivo imaging of mouse retinas. Mice were anesthetized by i.p. injection of a mixture (20 μl/g body weight) containing ketamine (6 mg/ml) and xylazine (0.44 mg/ml) in 10 mM sodium phosphate (pH 7.2) with 100 mM NaCl. Pupils were dilated with a mixture of 0.5% tropicamide and 0.5% phenylephrine hydrochloride (Midorin-P; Santen Pharmaceutical). Numbers of autofluorescent particles in SLO images were counted per image.

Total RNA was extracted with the RNeasy mini kit (Qiagen, Germantown, MD). Primers used in the study are listed in Supplemental Table I. All procedures for quantitative RT-PCR (qRT-PCR) were carried out as described previously (19).

Enucleated eyes were harvested from the animals and lysed in ice-cold lysis buffer (150 mM NaCl, 1 mM EDTA, 0.2% Nonidet P-40, and 20 mM Tris-HCl [pH 7.5]) containing protease inhibitor mixture (Sigma-Aldrich, St. Louis, MO). Tissue lysate was spun at 10,000 rpm for 10 min at 4°C. Proteins from each sample were transferred onto Immobilon-P membranes (Millipore, Bedford, MA) after SDS-PAGE gel electrophoresis. Membranes were incubated in 1% BSA solution containing a 1:1000 dilution of either anti-LCN2 rabbit polyclonal Ab (sc-50350; Santa Cruz Biotechnology, Santa Cruz, CA) or anti–β-actin Ab (Santa Cruz Biotechnology). Signals were visualized with alkaline phosphatase (Promega, Madison, WI) at a dilution of 1:10,000. The intensities of the bands were normalized to the actin band using ImageJ software (National Institutes of Health, Bethesda, MD).

Primary mouse RPE cells and retinal microglial cells were prepared from 2-wk-old mice based on previously published methods (19, 20). Enucleated eyes were incubated with 2% dispase (Invitrogen) in DMEM (Invitrogen) for 1 h at 37°C, and neural retinas and eyecups were separated under a surgical microscope (ILLUMIN-i; Endure Medical, Cumming, GA). The RPE layer was peeled from eye cups and cultured in DMEM containing minimal essential medium nonessential amino acids (Invitrogen), penicillin/streptomycin (Invitrogen), 20 mM HEPES (pH 7), and 10% FBS. To enrich microglial cells, neural retinas were homogenized and cultured in DMEM containing minimal essential medium nonessential amino acids (Invitrogen), penicillin/streptomycin (Invitrogen), 20 mM HEPES (pH 7.0), and 10% FBS for 7 d at 37°C. Adherent cells to the plastic surface were treated with 0.05% trypsin (Invitrogen), and less adhesive cells were collected as microglial cells. Human primary RPE cells were purchased from Lonza (Walkersville, MD).

Photoreceptor outer segment (POS) membranes were prepared from 1-mo-old Abca4^−/−Rdh8^−/− and WT mice using a published method (19). LPS was purchased from InvivoGen (San Diego, CA).

Establishment of hiPS cell lines and differentiated to RPE monolayers has been described in detail (21–25). All procedures were approved by the Institutional Review Boards at the Case Western Reserve University (Cleveland, OH) and adhered to the Declaration of Helsinki. All cell culture procedures were approved by Case Western Reserve University Institutional Biosafety Committee. All samples were obtained after donors had given informed consent.

hiPS-RPE cells (30,000–50,000) were seeded in 96-well plates. Cells were treated with 100 μM H₂O₂ and recombinant LCN2 protein (R&D Systems, Minneapolis, MN) in doses of 1, 10, and 100 ng/ml for 24 h. On the next day, 1 drop/ml of CellEvent caspase-3/7 green detection reagent was added to each well for 60 min at 37°C. Cells were then washed and fixed in 4% paraformaldehyde for 15 min, followed by DAPI staining. Cells were visualized and imaged under inverted fluorescence microscope.

Assays were performed on 96-well plates, with 1 × 10⁴ cells seeded in each well. Cells were incubated with and without recombinant LCN2 (R&D Systems) for 24 h. After 2 h incubation, 10  μl of the WST-1 solution (Roche, Mannheim, Germany) was added to the culture medium and incubated for 2 h at 37 °C. Absorbance was measured using a microplate ELISA reader (Multiskan FC microplate reader; Fisher Scientific, Pittsburgh, PA). All experiments were conducted in triplicate and replicated at least three times. Viable cells number was calculated by comparing the absorbance values of the samples after background subtraction.

Amounts of CCL2, TNF-α, and LCN2 in serum, plasma, and tissue homogenates were quantified by ELISA kits (R&D Systems) according to the manufacturer’s instructions. Eyes were homogenized in 500 μl of Nonidet P-40 lysis buffer containing 20 mM Tris (pH 8), 137 mM NaCl, and 1% Nonidet P-40. Protein concentrations were measured using a NanoDrop (Thermo Fisher Scientific, Waltham, MA).

Detection and quantification of gene expression was performed using a mouse inflammatory cytokine and receptors array (PAMM-011ZF; SABiosciences/Qiagen, Frederick, MD) according to the manufacturer’s instructions. This PCR-based array was selected, as it includes 84 diverse genes important in the immune response, including genes encoding CC chemokines, CXC chemokines, IL cytokines, other cytokines, chemokine receptors, and cytokine receptors, as well as other genes involved in the inflammatory response. MicroRNA (miRNA) array was also performed using an inflammatory response miRNA PCR array (MIMM-105Z; SABiosciences/Qiagen) according to the manufacturer’s instructions. Real-time PCR amplification was performed using RT² SYBR Green PCR master mix (Qiagen). Primers were designed using Web tool Primer3 and synthesized by Eurofins MWG Operon (Huntsville, AL). Data analysis was performed using the ΔΔC_T method according to the manufacturer’s protocol (SABiosciences).

hiPS-RPE cells (30,000–50,000) were seeded in 96-well plates. Cells were coincubated with 1 μg/ml LPS and 1 ng/ml LCN2 for 24 h. On the next day, the culture medium was removed and cells were fixed by adding 100 μl of 4% formalin to the wells for 15 min, followed by washing in PBS twice for 5 min. Fixed cells were permeabilized with 100 μl of 0.2% Triton X-100 in PBS for 1 min at room temperature. The cells were then incubated with rabbit anti-mouse p65 (1:100) (eBioscience) in PBS containing 10% goat serum overnight at 4°C. Cells were then washed twice with PBS and incubated with Alexa Fluor 488–labeled goat anti-rabbit IgG Ab (1:200) (Molecular Probes) in PBS at room temperature for 2 h. The cells were washed twice for 5 min with PBS. DAPI (1:500; Vector Laboratories, Burlingame, CA) was added to the cells to visualize the nuclei. Cells were then washed twice with PBS and imaged under fluorescent microscope. Five high-power fields (×100 magnification) were randomly selected in each sample for analysis. Positive nuclear staining in LPS-treated cells was used as positive control for NF-κB staining. The percentage of nuclear staining for NF-κB p65 was scored by counting the positive-stained cells and the total number of cells quantified in random microscopic fields using the Metamorph image analysis software (Molecular Devices, San Jose, CA).

hiPS-RPE cells (30,000–50,000) were seeded in 96-well plates. Cells were coincubated with 1 μg/ml LPS and 1 ng/ml LCN2 for 24 h. On the next day, media were removed and cells were rinsed once with ice-cold PBS. Next, PBS was removed and 100 μl of Nonidet P-40 lysis buffer was added and incubated on the plate on ice for 5 min. Cells were scraped, collected, and centrifuged for 10 min (14,000 rpm) at 4°C. Supernatant was collected and total protein was measured using a Nanodrop (Thermo Fisher Scientific). Total endogenous levels of total NF-κB p65 protein was then measured using the PathScan total NF-κB p65 sandwich ELISA kit (no. 7174; Cell Signaling Technology, Danvers, MA) according to the manufacturer’s instructions. Endogenous levels of phospho–NF-κB p65 protein were measured using the PathScan phospho–NF-κB p65 sandwich ELISA kit (no. 7173; Cell Signaling Technology) according to the manufacturer’s instructions. ELISA results were normalized to the total protein content per well.

All participants in the study underwent a complete ophthalmic examination and visual function tests by ophthalmologists. Informed consent was obtained from each person for the present study. Procedures followed the Declaration of Helsinki guidelines and were approved by the Institutional Review Boards of Case Western Reserve University, RIKEN, Institute of Biomedical Research and Innovation Hospital, and Cleveland Clinic. Nonfasting blood samples were collected at the Institute of Biomedical Research and Innovation Hospital or at the Cole Eye Institute from patients diagnosed with Stargardt disease (n = 11), RP (n = 117), and the wet form of AMD (n = 57). Blood was also collected from healthy controls (n = 77) who had no retinal degeneration as determined by ophthalmologic examination. Healthy controls include family members of patients with inherited retinal disorder and individuals who underwent cataract surgery. Plasma was prepared as previously described and stored under argon at −80°C until analysis (26).

pcDNA3.1-LCN2 (OHu27037D) was purchased from GenScript (Piscataway, NJ). ARPE19 cells (American Type Culture Collection, Manassas, VA) were transiently transfected with 1 μg of pcDNA3.1-LCN2 or pcDNA3.1 plasmids using Lipofectamine 3000 (Thermo Fisher Scientific) according to the manufacturer’s protocol. Expression of LCN2 mRNA was then examined by RT-PCR 72 h after transfection using primers shown in Supplemental Table I.

Statistical analyses were performed using the t test for comparing two groups, and one-way ANOVA was used to detect differences among three or more groups. Mann–Whitney U statistics were employed to analyze human plasma samples. Results were presented as mean ± SD. The results were considered statistically significant at p < 0.05.

To examine expressional changes of acute stress protein LCN2 in various tissues under stress conditions that can induce retinal degeneration, 1-mo-old Abca4^−/−Rdh8^−/− mice were exposed to light at 10,000 lx for 30 min and qRT-PCR was performed with isolated tissues, including the eye, liver, heart, brain, spleen, lung, and kidney. A 2-fold or higher increase of Lcn2 was observed in the eye (15.38 ± 3.86) and brain (4.97 ± 0.55) when compared with dark-adapted controls (Fig. 1A). Lcn2 expression was not obvious in these tissues prior to light exposure except spleen, where low levels of Lcn2 expression were detected. To determine the kinetics of LCN2 levels in the eye after light exposure, LCN2 expression was examined by immunoblot 1, 3, and 7 d after illumination at 10,000 lx for 30 min. LCN2 protein in the eye was found most upregulated 1 d after light exposure (Fig. 1B, 1C). Greatest LCN2 increase in serum was observed 3 d after light exposure in mice (Fig. 1D). Immunohistochemistry indicated that secreted LCN2 protein was detected in RPE and microglial cells in the inner nuclear layer of the retina (Fig. 1E). Because incubation with POS can induce production of inflammatory cytokines and chemokines from RPE and microglial cells (19), RPE and microglia cells were isolated from 2-wk-old Abca4^−/−Rdh8^−/− mice. When these cells were incubated with POS for 24 h, a 14-fold or 5-fold increase in Lcn2 was observed in RPE or microglial cells, respectively (Fig. 2).

To examine effects of Lcn2 deficiency in retinal degeneration, 1-mo-old Lcn2^−/−Abca4^−/−Rdh8^−/− and Abca4^−/−Rdh8^−/− mice were exposed to light at 10,000 lx for 30 min. WT littermates were subjected to the same light exposure. Histology sections of the retina demonstrated a decrease in outer nuclear layer thickness in Lcn2^−/−Abca4^−/−Rdh8^−/− mice as compared with Abca4^−/−Rdh8^−/− mice 7 d after light exposure (Fig. 3A, upper panels, 3B). In vivo imaging of the retina by SLO was performed 7 d after light exposure (Fig. 3A, lower panels). Increased number of autofluorescent spots, which are activated microglial cells and macrophages (19), were counted in Lcn2^−/−Abca4^−/−Rdh8^−/− mice as compared with Abca4^−/−Rdh8^−/− mice (Fig. 3C). Lcn2^−/− mice did not develop light-induced retinal degeneration (see “Deficiency of Lcn2 is associated with inflammatory changes” below).

To investigate effects of Lcn2 deficiency to retinal glial cells after light exposure, we examined the expression of Iba-1 for microglial cells and GFAP for Müller cells. Activation of glial cells is associated with retinal inflammation (27). Lcn2^−/−Abca4^−/−Rdh8^−/− and Abca4^−/−Rdh8^−/− mice at 1 mo of age were exposed to light at 10,000 lx for 30 min, and eyes were collected 24 h thereafter. Immunohistochemistry with anti–Iba-1 Ab showed increased protein expression and more Iba-1⁺ microglial cell numbers in Lcn2^−/−Abca4^−/−Rdh8^−/− mice as compared with Abca4^−/−Rdh8^−/− mice (Fig. 4A, upper panels). Immunohistochemical staining with anti-GFAP Ab revealed stronger gliosis with an increased expression of GFAP in Müller cells of Lcn2^−/−Abca4^−/−Rdh8^−/− mice (Fig. 4A, lower panels). qRT-PCR of Gfap displayed a similar pattern of increase as observed in immunohistochemistry (Fig. 4B). These results indicate that Lcn2 deficiency contributes to stronger response of retinal glial cells that are associated with retinal inflammation.

To further dissect LCN2-mediated effects on inflammatory genes and immune pathways, we employed the RT² real-time PCR array kit for 84 key immune genes involved in mediating inflammation. Lcn2^−/−Abca4^−/−Rdh8^−/− and Abca4^−/−Rdh8^−/− mice (1 mo old) were exposed to light at 10,000 lx for 30 min. Dark-adapted mice were used for controls. The PCR array was performed 1 d (24 h) after light exposure. Total RNA was isolated from all groups of mice. Equal amounts were converted to cDNA and then subjected to pathway-focused gene expression profiling using real-time PCR. Scatter plots comparing the differential expression of genes in light-exposed Lcn2^−/−Abca4^−/−Rdh8^−/− and Abca4^−/−Rdh8^−/− mice were generated (Fig. 5A). Each symbol represents an individual gene. The boundary lines indicate a 2-fold difference. Genes outside the boundary lines have ≥2-fold altered expression in Lcn2^−/−Abca4^−/−Rdh8^−/− mice compared with Abca4^−/−Rdh8^−/− mice. Table I lists the fold changes of the genes compared in the two groups of mice. Lcn2^−/−Abca4^−/−Rdh8^−/− mice displayed upregulation of 22 inflammatory mediators, including chemokines, chemokine receptors, and ILs, as compared with Abca4^−/−Rdh8^−/− mice. Conversely, Cx3cl1, Vegfa, and Cxcl15 were downregulated in Lcn2^−/−Abca4^−/−Rdh8^−/− mice. Validation by qRT-PCR for Ccl8, Ccl2, Cxcl10, Ccr5, Tnf, and Il2a was carried out using a separate set of mice (Fig. 5B). Production of CCL2 and TNF was quantified with eyes of 1-mo-old Lcn2^−/−Abca4^−/−Rdh8^−/− and Abca4^−/−Rdh8^−/− mice 24 h after light exposure at 10,000 lx for 30 min. Increased levels of CCL2 and TNF in Lcn2^−/−Abca4^−/−Rdh8^−/− mice (CCL2, 34.75 ± 1.5 pg/ml; TNF, 6.02 ± 0.04 pg/ml) compared with Abca4^−/−Rdh8^−/− mice (CCL2, 5.75 ± 0.7 pg/ml; TNF, 1.26 ± 0.02 pg/ml) were measured (Fig. 5C, 5D).

Because Lcn2^−/−Abca4^−/−Rdh8^−/− mice displayed more severe light-induced retinal degeneration and inflammation as compared with Abca4^−/−Rdh8^−/− mice (Figs. 3, 4, 5), roles of LCN2 in retinal inflammation after light exposure were examined using Lcn2^−/− mice. When 6-wk-old Lcn2^−/− mice were exposed to light at 10,000 lx for 30 min, increased levels of inflammatory changes were observed 24 h after illumination as compared with WT mice, including increased expression of Iba-1/Aif-1 and Gfap (Fig. 6A). Although increased immune reactions were observed in Lcn2^−/− mice, this light condition did not cause obvious retinal degenerative changes in Lcn2^−/− and WT mice; in contrast, Abca4^−/−Rdh8^−/− mice showed retinal structural changes 7 d after illumination (Fig. 6B). No retinal degeneration was observed in 6-wk-old and 4-mo-old Lcn2^−/− mice when they were kept under regular lighting conditions. Because inflammatory changes in light-exposed Lcn2^−/− mice suggest regulatory roles of LCN2 in inflammation, expression of miRNA, which can modulate transcriptional expression of inflammatory molecules, was examined by miRNA array. Changes in miRNA expression obtained from Lcn2^−/− and WT mice are presented in Table II. These results indicate that loss of Lcn2 is prone to accelerating inflammation.

The LCN2 gene promoter region contains binding sites for several transcription factors, including NF-κB (28), which prompted us to examine the effects of LCN2 on NF-κB signaling in RPE cells. Notably, the NF-κB pathway is a prototypical proinflammatory signaling pathway involved in the expression of multiple proinflammatory genes such as cytokines, chemokines, and adhesion molecules (29–32). hiPS-RPE cells were incubated with 1 μg/ml LPS for 24 h to investigate whether LCN2 can inhibit the nuclear translocation of NF-κB p65. Immunocytochemistry with anti-p65 Ab revealed that LPS-stimulated hiPS-RPE cells showed p65 staining in the nuclei, whereas unstimulated cells showed stronger staining in the cytoplasm. Supplementation of LCN2 in the culture medium decreased the numbers of cells with LPS-induced nuclear translocation of p65 (Fig. 7A, 7B). Furthermore, phosphorylation of NF-κB p65, which indicates activation of NF-κB, was examined using ELISA. In contrast to the finding that levels of total NF-κB p65 remained unchanged, a decrease in phosphorylated NF-κB p65 was observed when LCN2 was supplemented to hiPS-RPE cells (Fig. 7C). These observations suggest that LCN2 contributes to inhibition of NF-κB activation in human RPE cells. hiPS-RPE cells can express LCN2 as well as murine RPE cells when these cells were incubated with POS (Supplemental Fig. 1).

Several in vitro studies have demonstrated that LCN2 protects against cellular stress and exposure to H₂O₂ (33–37). Other studies have reported that oxidative stress plays a major role in degenerative retinal diseases such as RP and AMD (38–41). To investigate the effects of LCN2 on H₂O₂-induced oxidative stress in RPE cells, 1 ng/ml LCN2 was incubated with hiPS-RPE cells under oxidative stress conditions. Incubation with 100 μM H₂O₂ for 24 h in the absence of LCN2 resulted in reduced cell viability to 49.51 ± 11.59%, indicating that about half of the cell population died (Fig. 8A). When 1 ng/ml LCN2 was added to these incubation conditions, protective effects of LCN2 against H₂O₂-induced cell death were observed. LCN2 supplementation maintained cell viability nearly to the control level at 87.62 ± 12.74%. With increasing doses of recombinant LCN2, increased expression of antioxidant enzymes heme oxygenase 1 (HMOX1) and superoxide dismutase 2 (SOD2) in hiPS-RPE cells was observed (Fig. 8B, 8C), suggesting that an upregulation of HMOX1and SOD2 is an adaptive mechanism to protect cells from oxidative damage. HMOX1 and SOD2 are known to act as the first-line antioxidant enzyme defense system against reactive oxygen species (ROS) and particularly superoxide anion radicals (40).

Retinal inflammation has detrimental effects on cell viability (27, 42–44). To examine whether LCN2 could protect RPE cells from inflammation-associated cell death, hiPS-RPE cells were incubated with 1 μg/ml LPS for 24 h in the absence and presence of LCN2. Caspase-3/7 expression was measured to assess cell apoptosis. Increased numbers of apoptotic cells were detected in LPS-treated cells in the absence of LCN2. When hiPS-RPE cells were pretreated with 10 or 100 ng/ml LCN2 for 24 h before incubation with 1 μg/ml LPS, a reduction in the number of apoptotic cells was observed as compared with the LPS only–treated group (Fig. 9). Anti-immune reactions of LCN2 were observed in human RPE-derived cells when these cells were cultured with LCN2 (Supplemental Fig. 2). These results suggest that LCN2 provides protection from inflammation-associated cell death.

LCN2 is known to modulate cell homeostasis by its interaction with specific cell-surface receptors, namely murine 24p3r (45–47) and the solute carrier family 22 member 17 (SLC22A17), which belongs to the major cation transporter family in humans. To explore the possible involvement of a receptor-mediated effect, expression of LCN2 receptor in RPE cells was examined. Expression of SLC22A17 was detected in human RPE cells, including hiPS-RPE, ARPE19 (a human RPE derived cell line), and primary human RPE cells (Fig. 10A). When Abca4^−/−Rdh8^−/− mice were exposed to light exposure at 10,000 lx for 30 min, qRT-PCR revealed that 24p3r expression increased in the RPE cells after light exposure (Fig. 10B), suggesting the possible involvement of RPE–LCN2 interaction during retinal degeneration pathogenesis.

To investigate roles of LCN2 in human retinal degenerative diseases, LCN2 plasma levels in patients with Stargardt disease, RP, and AMD were measured and compared with age-matched control individuals without any retinal diseases (Supplemental Fig. 3). Determined LCN2 plasma concentrations were as follows: 45.13 ± 5.37 ng/ml in Stargardt disease (n = 11), 44.78 ± 31.73 ng/ml in RP (n = 117), 48.41 ± 28.52 ng/ml in AMD (n = 57), and 17.04 ± 11.9 ng/ml in controls (n = 77). Patients with these retinal degenerative diseases exhibited higher plasma levels of LCN2 as compared with control individuals, suggesting roles of LCN2 in human diseases.

Lastly, protective effects of LCN2 from RPE cell death were examined. ARPE19 cells were transfected with plasmids for LCN2 expression, and then these cells were cultured with 1 μg/ml LPS for 24 h to induce inflammation-associated cell death. LCN2-transfected cells demonstrated better cell viability as compared with control vector–transfected cells (Fig. 11A). Transfection of LCN2 successfully increased LCN2 levels in ARPE19 cells (Fig. 11B).

Millions of people around the world suffer from retinal degenerative diseases such as inherited retinal dystrophies and AMD, with major debilitating impact on daily life. Apoptosis and inflammation play important roles in the pathogenesis of these diseases. Previous studies have demonstrated that retinal inflammation contributes to retinal degeneration (19) and that increased expression of an acute stress protein LCN2 is observed in Abca4^−/−Rdh8^−/− mice after light exposure (3). In the present study, Lcn2 expression was increased in the brain as well as in the eye 1 d after light exposure in Abca4^−/−Rdh8^−/− mice with LCN2 production observed in RPE and retinal microglial cells. We also compared Lcn2^−/−Abca4^−/−Rdh8^−/− mice with Abca4^−/−Rdh8^−/− mice after exposure to light and found that deficiency of Lcn2 resulted in more severe retinal damage and increased expression of inflammatory cytokines, including Ccl8, Ccl2, and Cxcl10. Lcn2 loss also resulted in increased expression of one of the receptors of CCL3, namely Ccr5. Our previous work has shown that CCL2 and CCL3 have distinct roles in the pathogenesis of retinal degeneration in Abca4^−/−Rdh8^−/− mice (27). In other published studies, mice lacking Lcn2 exhibited impaired migration of astrocytes to injury sites with decreased Cxcl10 expression (48). This observation suggests that LCN2 protein, secreted under inflammatory conditions, could amplify neuroinflammation by inducing neural immune cells to secrete chemokines such as CXCL10 for recruiting additional inflammatory cells. Unexpectedly, we observed in the present study more activated microglial cells in light-exposed Lcn2^−/−Abca4^−/−Rdh8^−/− mice as compared with Abca4^−/−Rdh8^−/− mice, with Lcn2^−/−Abca4^−/−Rdh8^−/− mice exhibiting increased Cxcl10 expression. Activation of microglial cells in response to chemokines, including CCL2 produced from the RPE, is a hallmark of inflammation in retinal degeneration (27). Activated microglial cells in the subretinal space display increased production of proinflammatory and chemotactic cytokines (27).

The present study demonstrated protective roles of LCN2 in retinal degeneration as other reported studies; however, there are conflicting observations (49–56), and both pro- and anti-inflammatory properties of this glycoprotein have been reported. LCN2 expression can be induced by several cytokines and growth factors, including IL-6, IL-1β, IL-10, IL-17, TNF, and TGF (57–62). Such LCN2 induction is dependent on the activation of NF-κB transcriptional activity, which is suggested as a positive regulator of LCN2 expression itself (28). Another report suggested that LCN2 plays a role as an anti-inflammatory regulator of macrophage polarization and NF-κB/STAT3 pathway activation (52). This study demonstrated that LPS stimulation elicited an increase in the activation of NF-κB, c-Jun, and STAT3 signaling pathways in Lcn2^−/− bone marrow–derived macrophages. Pretreatment with recombinant LCN2 attenuated LPS-stimulated degradation of Ik-Ba and STAT3 phosphorylation as well as LPS-induced gene expression of IL-6 and inducible NO synthase in Lcn2^−/− bone marrow–derived macrophages. In this study, we investigated the regulation of LCN2 in RPE cells. Pretreatment of hiPS-RPE cells with LCN2 before exposing to LPS resulted in a significant decrease in nuclear translocation of NF-κB p65, suggesting a negative feedback loop involving LCN2. We also found that LCN2 murine receptor 24p3r expression in RPE cells increased in a similar fashion as LCN2 after light exposure in Abca4^−/−Rdh8^−/− mice, implicating LCN2 receptor and agonist interactions in RPE cells. Although LCN2 signaling pathways are not fully elucidated, it is noteworthy that the LCN2 promoter region contains the binding sites of several transcription factors such as STAT1, STAT3, CREB, and C/EBPβ and NF-κB (28).

The antioxidant function of LCN2 has been well characterized (33, 35, 63) and the protein has been shown to be cytoprotective against oxidative stress (37, 64, 65). LCN2 protects against cellular stress from exposure to H₂O₂, and that overexpression of LCN2 allows cells to better tolerate oxidative stress conditions (6). Our present results show that expression of LCN2 in hiPS-RPE cells suppresses H₂O₂-induced cell death and prolongs cell survival. We found that the transcript levels of the key oxidative stress–catalyzing enzymes HMOX1 and SOD2 increased with increasing LCN2 concentrations, supporting an antioxidant role for LCN2 in RPE cells. HMOX1 not only regulates the cellular content of the pro-oxidant heme, but it also produces catabolites with regulatory and protective functions (66). SOD mRNA levels increase following a wide range of mechanical, chemical, and biological stimuli that increase ROS, such as UVB and x-irradiation, ozone, and LPS (67). Increased gene expression of SODs and heme oxygenases are adaptive cellular defense mechanisms against oxidative stress. In our animal study, loss of Lcn2 in mice resulted in more severe retinal degeneration after light exposure. In the retina, LCN2 may serve to protect from a broad array of ROS produced by light exposure (68–70), including ROS generated in the visual cycle from the conversion of 11-cis-retinal to all-trans-retinal (19, 68).

LCN2 concentration in biological fluids under healthy conditions is low, but the protein can be upregulated by inflammation and becomes detectable at various stages in several diseases (71–74). We found LCN2 plasma levels elevated >2-fold in patients with Stargardt disease, RP, and AMD compared with healthy controls. In this study, LCN2 expression was observed in the RPE and in the inner retina. An earlier report demonstrated that RPE is a main source of secreted LCN2 in the eye (14). Because breakdown of the blood retinal barrier can be observed during the process of retinal degeneration (19), such comprised blood retinal barrier might contribute to leakage of LCN2 produced in the eye and to increased levels of LCN2 in plasma of patients. Activated immune cells, which produce LCN2, could migrate into the blood vessels and produce LCN2 into the blood. This result not only implies potentially important roles for circulating LCN2 in the pathogenesis of retinal degenerative diseases, but also the possibility that LCN2 could serve as a biomarker of early disease onset and progression (75).

In conclusion, this study provides evidence that LCN2 may serve to protect the retina from inflammation-induced degeneration through regulation of cytokine and chemokine production, and by preserving cell viability and attenuating apoptosis through regulation of anti-oxidant enzymes. LCN2, secreted mainly from RPE cells in the retina, could be a critical mediator in retinal inflammation and degeneration processes.

We thank Catherine Dollar and Scott Howell (Visual Science Research Center, Case Western Reserve University), Tatiana Reidel (Department of Ophthalmology and Visual Sciences, University Hospital of Cleveland), Dr. Yuki Arai, Dr. Akiko Yoshida, and Kanako Kawai (RIKEN) for technical assistance and comments.

The authors have no financial conflicts of interest.