Pathological Involvement of Astrocyte-Derived Lipocalin-2 in the Demyelinating Optic Neuritis

Abstract

Purpose.: The current study was done to determine the role of lipocalin-2 (LCN2) in the pathogenesis of demyelinating optic neuritis using an experimental autoimmune optic neuritis (EAON) model.

Methods.: The EAON was induced by subcutaneous immunization with an emulsified mixture of myelin oligodendrocyte glycoprotein (MOG_35–55) peptide in mice. The LCN2 expression was examined in the optic nerve after MOG peptide injection. Degree of demyelination, inflammatory infiltration, glial activation, and expression profile of inflammatory mediators in the optic nerve were compared between LCN2 knockout (KO) animals and wild-type littermates by histological analysis and real-time PCR following EAON induction. Plasma levels of LCN2 in patients with optic neuritis were measured by ELISA.

Results.: The expression of LCN2 was notably increased in the optic nerve after EAON induction. Expression of LCN2 was colocalized with reactive astrocytes. A significant reduction of demyelination, inflammatory infiltration, and gliosis was demonstrated in the optic nerve of LCN2 KO mice. The LCN2 KO mice also showed markedly reduced gene expression associated with the M1-polarized glia phenotype and toll-like receptor signaling in the optic nerve. The LCN2 levels in plasma were significantly higher in optic neuritis patients (71.6 ± 10.6 ng/mL) compared to healthy controls (37.4 ± 9.1 ng/mL, P = 0.0284).

Conclusions.: In this study, we demonstrated a significant induction of LCN2 expression in astrocytes of the optic nerve following EAON induction. Our results imply that astrocyte-derived LCN2 may have a pivotal role in the development of demyelinating optic neuritis, and LCN2 can be a therapeutic target to alleviate immune and inflammatory damage in the optic nerve.

Optic neuritis is the first clinical presentation in 38% of patients with multiple sclerosis (MS) before development of systemic neurological symptoms.^1,2 Optic neuritis manifests as an acute and self-limiting optic nerve inflammation with decreased visual acuity; recovery occurs over several weeks in the majority of patients.^1,3 The pathogenesis of optic neuritis has been reported to be associated with chronic inflammation and demyelination of the optic nerve axons with a relapsing-remitting course, eventually causing neurodegeneration as seen in MS.^4,5 The pathophysiology of MS and associated optic neuritis has been investigated using murine animal models. In particular, the central nervous system (CNS)–specific antigen, such as the myelin oligodendrocyte glycoprotein (MOG) has been demonstrated to induce experimental autoimmune optic neuritis (EAON), as well as experimental autoimmune encephalomyelitis (EAE).^5,6 The MOG antigen is present abundantly within the optic nerve and reacts with immune and inflammatory cells to cause tissue damage. Studies using the EAON/EAE animal models suggest that the preceding microglial activation may have a crucial role in disease development by causing the initial damage to axons, and the following infiltration of MOG-specific T cells induce demyelination and axon pathology.^5,7–11 

Recently, we and others reported that lipocalin-2 (LCN2) expression was highly upregulated during the active phase of the EAE.^12,13 The LCN2 was first described as an acute phase protein involved in the innate immunity given its ability to bind bacterial siderophores.¹⁴ Expression and secretion of LCN2 is increased by toll-like receptor (TLR) signaling in immune cells.¹⁴ The LCN2 has been suggested to participate in the initiation of neuroinflammation and the promotion of leukocyte migration into injured spinal cord and brain through chemokine induction.^15,16 Studies so far have demonstrated the expression and role of LCN2 in the parenchyma of brain and spinal cord during EAE. To our knowledge, however, there is no report on the LCN2 expression and its pathological implication in the optic neuritis or EAON mouse model. The purpose of this study was to examine the expression of LCN2 in optic nerve and to evaluate the pathological role of LCN2 in optic neuritis using the EAON model based on LCN2 knockout (KO) mice. 

Materials and Methods

Animals

Wild-type (WT) C57BL/6 mice were obtained from Samtaco (Osan, Korea), and Lcn2 knockout (LCN2 KO) mice were kindly provided by Kiyoshi Mori (Kyoto University, Kyoto, Japan) and Shizuo Akira (Osaka University, Osaka, Japan). The Lcn2 KO mice were back-crossed for 8 to 10 generations in a C57BL/6 background to generate homozygous and heterozygous animals, as described previously.¹⁴ The absence of Lcn2 in Lcn2-deficient mice was confirmed by PCR of genomic DNA. Animals used in the present study were acquired and cared for in accordance with the procedures approved by the Institutional Animal Care Committee of Kyungpook National University and the animal care guidelines of the National Institute of Health (NIH; Bethesda, MD, USA), and in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

EAON Induction

Mice (7–8 weeks) were immunized subcutaneously with 200 μg of MOG (MOG_35–55; M-E-V-G-W-Y-R-S-P-F-S-R-V-V-H-L-Y-R-N-G-K; GLBiochem, Shanghai, China) in 100 μL of a solution containing 50% of complete Freund's adjuvant (CFA) with 10 mg/mL of heat-killed H37Ra strain Mycobacterium tuberculosis (Difco, Detroit, MI, USA) into areas draining into axillary and inguinal lymph nodes. Pertussis toxin (List Biological Laboratories, Campbell, CA, USA) in PBS (at 200 ng per mouse) was administered intraperitoneally on days of immunization and again 48 hours later. Animals were weighed and examined for disease symptoms daily. Evaluation of disease severity and other experiments were done in a blinded fashion. Disease severity was scored using a 0 to 5 scale, as follows: 0, no symptom; 1, limp tail; 2, weakness and incomplete paralysis of one or two hind limbs; 3, complete hind limb paralysis; 4, forelimb weakness or paralysis; and 5, moribund state or death. 

Optic Nerve Histology

Mice were anesthetized with diethyl ether, transcardially perfused with cold saline, and then perfused with 4% paraformaldehyde (PFA) diluted in 0.1 M PBS. Optic nerves were isolated and fixed using 4% PFA for 3 days, and then cryoprotected with 30% sucrose solution for 3 days. Six animals were used per experimental group. Tissues were embedded in OCT compound (Tissue-Tek; Sakura Finetek, Tokyo, Japan) for frozen section and then cut into 10-μm thick longitudinal sections. To detect LCN2 expression and glial activation, sections were incubated with goat anti-LCN2 antibody (1:500 dilution; R&D Systems, Minneapolis, MN, USA), rabbit anti-glial fibrillary acidic protein (GFAP) antibody (1:500 dilution; Dako, Glostrup, Denmark), and rabbit anti-ionized calcium-binding adapter molecule-1 (Iba-1) antibody (1:500 dilution; WAKO, Osaka, Japan). Sections were visualized with Cy3-conjugated anti-goat or rabbit IgG, or FITC-conjugated anti-goat or rabbit IgG antibody (Jackson Laboratory, Bar Harbor, ME, USA). To assess demyelination and inflammatory cell infiltration, sections were stained with FluoroMyelin (1:300 dilution; Invitrogen, Carlsbad, CA, USA) and hematoxylin/eosin (H&E), respectively. The pathologic findings were graded from 0 to 3 as described: grade 0, no lesion; 1, moderate cellular infiltration into the optic nerve; 2, strong cellular infiltration into the optic nerve; and 3, massive cellular infiltration into the optic nerve.¹⁷ Each section was captured using a CCD color video camera (Olympus D70; Olympus, Tokyo, Japan) attached to a microscope (Olympus BX51) equipped with a ×100 objective lens. Image analysis was conducted by Image J (available in the public domain at http://rsb.info.nih.gov/ij/) software as reported previously.¹² 

Traditional or Real-Time RT-PCR

Total RNA was extracted from optic nerve tissues using TRIzol reagent (Invitrogen), according to the manufacturer's instructions. Total RNA (0.5 μg) was reverse-transcribed into cDNA using Superscript II (Invitrogen) and oligo (dT) primers. The PCR amplification was conducted using a DNA Engine Tetrad Peltier Thermal Cycler (MJ Research, Waltham, MA, USA) at an annealing temperature of 55°C to 60°C for 20 to 30 cycles using specific primer sets. To analyze PCR products, 10 μL of each PCR product was electrophoresed on 1% agarose gel and detected under UV light. We used Gapdh as an internal control. Real-Time PCR was performed using the One Step SYBR PrimeScript RT-PCR Kit (Perfect Real-Time; Takara Bio, Inc., Tokyo, Japan), and detection was performed using the ABI Prism 7500 Sequence Detection System (Applied Biosystems, Foster City, CA, USA). Nucleotide sequences of the primers were based on published cDNA sequences (Table 1). 

Table 1

 

DNA Sequences of the Primers Used for Traditional or Real-Time RT-PCR

 

LCN2 Measurement in Human Plasma by ELISA

The LCN2 levels were measured in plasma samples of 8 patients with optic neuritis and 8 healthy controls (age- and sex-matched, Table 2). The study was approved by the institutional review board and followed the tenets of the Declaration of Helsinki. Informed consent was obtained from all subjects. Blood samples of the eight patients were collected at the day of diagnosis with optic neuritis. Intravenous high dose steroid therapy was started after the blood sample collection. All eight patients did not have previous neurological diseases or newly occurred neurologic symptoms associated with their visual disturbance. In brain MRI of the eight patients, there was no white matter lesion suggesting multiple sclerosis. Plasma levels of LCN2 were measured using a commercially available Sandwich ELISA Duo-set (R&D Systems). The assays were run in 96-well plates (Corning International, Corning, NY, USA) using 100 μL of plasma (1:400 dilution) per the manufacturer's instructions. For standards, human recombinant LCN2 protein was used at concentrations ranging from 39.06 to 2500 pg/ml. Results were normalized to the total protein content of the plasma samples. All measurements were obtained from duplicated assays. 

Table 2

 

Patient Demographics and Clinical Characteristics

 

Statistical Analysis

All values are expressed as mean ± SEM. Student's t-test was used to determine the statistical significance of gene expression and fluorescence intensity. Clinical scores and categorical variables were analyzed with the Mann-Whitney nonparametric test. All other data sets were analyzed by 1-way or 2-way ANOVA with Bonferroni's post hoc tests using SPSS version 14.0K (SPSS, Inc., Chicago, IL, USA). Statistical significance was accepted for P values <0.05. 

Results

LCN2 Expression in the Optic Nerve of the EAON-Induced Mice

To investigate a correlation between EAON and LCN2 expression in optic nerve, immunofluorescence analysis of optic nerve tissue sections was performed on day 17 after MOG immunization (Fig. 1A), at which time the mice showed typical EAE symptoms (Supplementary Fig. S1). The LCN2 protein staining was significantly increased in the EAON-induced mice compared to those of naïve animals (Figs. 1A, 1B). The LCN2 expression was colocalized with GFAP (astrocyte-specific marker)–positive cells (Fig. 1A), indicating that LCN2 expression was induced in reactive astrocytes in the optic nerve after MOG immunization. Increased LCN2 mRNA expression also was observed in the optic nerve of EAON-induced mice (Fig. 1C). 

Figure 1

 

Expression of LCN2 in the optic nerves of EAON-induced mice. (A) Double immunofluorescence analysis revealed a strong induction of LCN2 expression (red) in GFAP (green)–positive astrocytes in the optic nerve at day 17 after immunization compared to that in naïve mice. Astrocytes were identified by GFAP staining and nuclei were stained with DAPI (blue in the merged images). Boxed regions were magnified and shown in the bottom. Scale bar: 100 μm. (B) Fluorescence intensity of LCN2 staining was measured with the color histogram tool of Image J. Data are mean ± SEM (n = 6 per each group). *P < 0.01. (C) The mRNA levels of LCN2 were determined by traditional RT-PCR in naïve and EAON-induced mice. We used Gapdh as an internal control. Data are mean ± SEM (n = 6 per each group). *P < 0.01.

LCN2 KO Mice Are Partially Resistant to EAON Induction

The LCN2 KO mice were used next to determine the role of LCN2 in EAON development. The WT and LCN2 KO mice were immunized with MOG, and then optic nerve tissue sections were examined for demyelination. Compact myelin of the optic nerve was visualized by FluoroMyelin in naïve WT and LCN2 KO mice (Fig. 2A). Demyelination of the optic nerve was observed in WT mice at day 17 after MOG immunization. In the EAON-induced WT mice, demyelination of the optic nerve was demonstrated as partial irregularity of the linear neurofilament structure and irregularly stained axon expanded throughout the optic nerve. However, demyelination of the optic nerve was significantly reduced in the EAON-induced LCN2 KO mice (Fig. 2A). Similarly, EAON-induced reduction in the intensity of myelin staining of the optic nerve was attenuated by LCN2 deficiency (Fig. 2A, lower panel, P < 0.05). Next, we examined the immune/inflammatory infiltration in the optic nerves of WT mice and LCN2 KO mice at day 17 after MOG immunization with H&E staining. Histological data revealed that WT mice displayed characteristic EAON histological alteration, including massive parenchymal infiltration (Fig. 2B). However, this infiltration of inflammatory cells was minimally detected in the optic nerves of LCN2 KO mice. These observations were presented by histological scores (Fig. 2C). Real-time PCR analysis was performed to compare the expression levels of cytokines and chemokines in the optic nerves isolated from EAON-bearing WT and LCN2 KO mice (Fig. 2D). Among the cytokines and chemokines examined, relative expression of IL-1β, IFN-γ, and CCL2 was significantly decreased by LCN2 deficiency in the optic nerves of the EAON-induced animals. 

Figure 2

 

Deficiency of LCN2 ameliorates EAON. (A) Upper, myelin of optic nerve was visualized by FluoroMyelin staining in naïve or EAON-induced WT and LCN2 KO mice (day 17 after immunization). Scale bar: 100 μm. Lower, quantified data of demyelination in naïve or EAON-induced WT and LCN2 KO mice. Data are mean ± SEM (n = 6 per each group). *P < 0.05, EAON-induced WT versus LCN2 KO mice. (B) Representative photographs of the optic nerve stained with H&E of naïve or EAON-induced WT and LCN2 KO mice (day 17 after immunization). Scale bar: 100 μm. (C) Histopathological score of optic neuritis based on H&E staining. The scoring was conducted using a scale from 0 to 3: grade 0, no lesion; 1, moderate cellular infiltration into the optic nerve; 2, strong cellular infiltration into the optic nerve; and 3, massive cellular infiltration into the optic nerve. Data are mean ± SEM (n = 6 per each group). *P < 0.05, EAON-induced WT versus LCN2 KO mice. (D) Expression levels of cytokines and chemokines were determined by real-time RT-PCR analysis of optic nerve tissues isolated from EAON-bearing WT and LCN2 KO mice. Data are mean ± SEM (n = 6). *P < 0.05 and **P < 0.01, EAON-bearing WT versus KO mice.

LCN2 Deficiency Attenuates Activation of Astrocytes and Microglia in the Optic Nerve of the EAON-Induced Mice

Immunofluorescence analysis of glia-specific markers revealed that LCN2 deficiency reduced activation of astrocytes and microglia in the optic nerve following MOG immunization (Fig. 3). The LCN2 KO mice showed reduced astrocytosis in the optic nerve compared to that of WT mice, as determined by GFAP staining of EAON-induced optic nerve tissue sections (Fig. 3A, upper panel). The fluorescence intensity of GFAP staining in the optic nerve also was reduced in the LCN2 KO mice compared to WT animals (P < 0.05, Fig. 3A, lower panel). Microglial activation in the EAON-induced optic nerve was similarly reduced in LCN2 KO mice, as determined by Iba-1 (a microglial cell marker) staining (Fig. 3B, upper panel). In the normal optic nerve of naïve animals, microglia had long branched (ramified) processes, whereas they appeared amoeboid with few short and thick processes in the optic nerve of the EAON-induced mice at day 17 after immunization. Microglial activation involved not only expansion of microglia in number, but also a change in cell morphology into an activated state (amoeboid microglia). This EAON-induced microglial activation was reduced in the LCN2 KO mice (P < 0.05, Fig. 3B, lower panel). 

Figure 3

 

Deficiency of LCN2 attenuates gliosis in EAON mice. (A) Upper, immunofluorescence staining of GFAP as an astrocytic marker, to observe the activation of astrocytes in naïve or EAON-induced WT and LCN2 KO mice (day 17 after immunization). Lower, comparison of the fluorescence intensity of GFAP staining between WT and KO mice. Data are mean ± SEM (n = 6 per each group). *P < 0.05, EAON WT versus EAON KO mice. (B) Upper, immunofluorescence staining of Iba-1 as a microglia marker. Magnified images of Iba-1–positive microglia are shown in the inset. Scale bars: white, 50 μm; yellow, 100 μm. Lower, the fluorescence intensity of Iba-1 staining in the optic nerve was compared between the two genotypes. Data are mean ± SEM (n = 6 per each group). *P < 0.05, EAON WT versus EAON KO mice.

LCN2 Deficiency Attenuates the Expression of M1 Markers

Real-time RT-PCR analysis was performed to assess the expression of glial M1 and M2 markers in the optic nerves isolated from EAON-bearing WT and LCN2 KO mice (Fig. 4). The EAON-induced expression of M1 markers (CD68 and CD86), but not M2 marker mannose receptor (Mrc1), was significantly decreased in the optic nerves of the LCN2 KO mice compared to those of WT mice (P < 0.01). Interestingly, EAON-induced LCN2 KO mice also showed decreased expression of TLR2, TLR4, and MyD88 (P < 0.01), suggesting that LCN2 may have a critical role in TLR signaling in the optic nerve of the EAON-induced mice. 

Figure 4

 

Deficiency of LCN2 reduces the expression of M markers and TLR2/4-dependent innate immune signaling components. Real-time RT-PCR analysis was performed for the M1 and M2 markers (CD68, CD86, and MRC1) or genes related to TLR2 and TLR4 signaling pathways using optic nerves isolated from EAON-bearing WT or LCN2 KO mice. Data are mean ± SEM (n = 6). **P < 0.01, EAON WT versus EAON KO mice.

Increased LCN2 Levels in Plasma Samples of Patients With Optic Neuritis

On the day of diagnosis with optic neuritis, plasma samples were obtained from all patients. Age- and sex-matched controls did not have any infection, inflammatory disease, diabetes, or cancer that can affect plasma levels of LCN2, and they all demonstrated their best-corrected visual acuities of 20/20. Mean plasma LCN2 levels were significantly higher in the patients with optic neuritis (71.6 ± 10.6 ng/mL) when compared to controls (37.4 ± 9.1 ng/mL, P = 0.0284, Fig. 5). 

Figure 5

 

Quantification of plasma LCN2 levels by Sandwich ELISA in controls and optic neuritis subjects. The horizontal bar in each column indicates the mean values of LCN2 levels, with statistically significant differences between the two groups: control subjects (n = 8); optic neuritis patients (n = 8). The LCN2 values (mean and range in ng/mL) were normalized to the total protein content. Statistical comparison between the groups is shown with P values.

Discussion

In this study, we investigated LCN2 expression and its role in optic neuritis using the EAON mouse model. The expression of LCN2 was notably increased in the optic nerve astrocytes after EAON induction. Comparative analysis of the WT and LCN2 KO mice after EAON induction revealed that LCN2 deficiency ameliorated demyelination and decreased reactive astrocytosis in the optic nerve of the EAON-induced animals. In addition, LCN2 deficiency attenuated microglial activation in the optic nerve. Compared to WT animals, the LCN2 KO mice showed a markedly reduced M1- and TLR signaling-related gene expression, which may be responsible for the less severe optic neuritis. Finally, higher plasma levels of LCN2 in patients with optic neuritis compared to control subjects supported the important role of LCN2 in the pathogenesis. 

The protein LCN2, a secreted glycoprotein, has been found in plasma, serum, urine, and cerebrospinal fluid (CSF).^18–20 The LCN2 is secreted from epithelial cells,²¹ macrophages,²² neutrophils,²³ and tumor cells²⁴ under various conditions, such as metastatic cancer, acute injury, and obesity/type 2 diabetes.²⁵ In the CNS, LCN2 expression has been observed in reactive astrocytes mainly after lipopolysaccharide administration,²⁶ EAE,¹² stroke,²⁷ and other inflamed conditions.²⁸ The LCN2 now is considered an acute response gene in neuroinflammation and CNS injury. It has been involved in modulating immune activation of microglia,²⁹ astrocytes,³⁰ and endothelial cells,¹⁵ as well as neuronal cell death.³¹ Additionally, LCN2 may act as a chemokine inducer. In particular, LCN2 leads to upregulation of CXCL10 by Janus kinase (JAK)2/signal transducers and activators of transcription (STAT)3 and I κB kinase (IKK)/nuclear factor κ-light-chain-enhancer of activated B cells (NF-κB) pathways in astrocytes.¹⁵ Thus, LCN2 proteins, which are acutely induced, may amplify neuroinflammation by recruiting additional inflammatory cells in the CNS. Our data showed LCN2 expression in reactive astrocytes of demyelinated optic nerve after MOG immunization, and demonstrated a LCN2 deficiency attenuated EAON pathogenesis. In this respect, astrocytes-derived LCN2 may have an important role in recruiting peripheral immune and inflammatory cells into the optic nerve making proinflammatory environment. Additionally, the lack of classical blood–brain barrier properties in the optic nerve may accelerate LCN2-induced recruitment of immune and inflammatory cells. Furthermore, our previous study implicated LCN2 induced in the spinal cord and peripheral lymphoid tissues in the development of EAE pathogenesis.¹² 

The general function of glia is to support neuronal activity under healthy conditions. Reactive glia, however, have been considered as one of the pathological hallmarks in neuroinflammation or neurodegenerative conditions, such as MS.³² Microglia and astrocytes under these conditions show two functionally distinct phenotypes: a classical activation (M1 polarization) that is associated with the production of proinflammatory cytokines and chemokines, and an alternative activation (M2 polarization) that is involved in adaptive immunity, tissue repair, and remodeling.^26,29,33 This study shows the alteration of inflammatory factors and M1 phenotype marker gene expression in the optic nerves of LCN2 KO mice after EAON induction. A significant reduction in the expression of proinflammatory cytokines (Il1b and Ifng), chemokine (Ccl2), and M1 polarization markers (Cd68, Cd86) was observed in the LCN2 KO optic nerve. Our data suggested that LCN2 may promote M1-polarization of glial cells in the inflamed optic nerve. 

The results obtained from the EAON animal studies could be translated into human patients. A critical role of LCN2 in the pathogenesis of optic neuritis was corroborated by the findings that significantly increased LCN2 plasma levels were observed in patients with optic neuritis compared to those of healthy controls. In addition, our results are in line with a previous study demonstrating that increased LCN2 levels in the serum of multiple sclerosis patients relative to healthy controls.³⁴ Brain MRI of optic neuritis patients did not show white matter lesions, which were diagnostic sign of multiple sclerosis. However, the risk of optic neuritis patients for development of multiple sclerosis is still higher than that of controls. Thus, we will follow the patients included in this study for several years to find out whether an initial increase of plasma LCN2 levels in optic neuritis patients can be a prognostic factor for progression into multiple sclerosis. 

In conclusion, the present study demonstrated that LCN2 has a pivotal role in the development of EAON. The main cellular source of LCN2 in the optic nerve of the EAON-induced mice is astrocytes. Astrocyte-derived LCN2 seems to promote the proinflammatory nature of reactive glia through M1 polarization and TLR signaling pathways. Thus, LCN2 can be therapeutically targeted for the treatment of demyelinating optic neuritis. 

Acknowledgments

Supported by a grant of the Korea Healthcare Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A111345, HI14C3331), the 2012 Cheil-Nammyung Foundation Research Fund, and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP; No. 2008-0062282 and 2012R1A2A2A02046812). 

Disclosure: B.Y. Chun, None; J.-H. Kim, None; Y. Nam, None; M.-I. Huh, None; S. Han, None; K. Suk, None