Alpha-internexin as a Corneal Autoantigen in Spontaneous Autoimmune Keratitis Mouse Model

Abstract

purpose. The purpose of this study was to identify target antigens of autoimmune keratitis with a disease-prone mouse model.

methods. BALB/c nude mice grafted with embryonic rat thymi (TG nude mice) develop various organ-localized autoimmune lesions, including keratitis. A hybridoma producing a monoclonal antibody (OT-20), specific for corneal epithelium was established by using spleen cells from this model mouse of keratitis, and the target of OT-20 was identified by immunoblot analysis. Then, using the antigen, T-cell proliferation and cytokine production by TG nude mice with keratitis were examined.

results. Immunoblot analysis revealed α-internexin to be the target antigen of OT-20 that specifically recognizes corneal epithelium. Sera from TG nude mice with keratitis reacted with α-internexin on Western blot analysis, and the T cells of these mice on stimulation with α-internexin exhibited proliferation responses and produced IL-2, IFN-γ, and TNF-α, but not IL-4 or IL-5.

conclusions. These results suggest that α-internexin is one of the corneal antigens associated with keratitis, developing spontaneously in TG-nude mice, with a probable pathogenic role.

Corneal melting disease, such as Mooren’s ulceration in humans, is a chronic sight-threatening ailment that occurs either in isolation or in association with systemic changes. The pathogenesis remains poorly understood, but there is evidence for an autoimmune etiology. ^{1 2 3 4 5} Histologic studies have revealed an accumulation of numerous neutrophils, plasma cells, and lymphocytes in the ulcerative cornea and the conjunctiva adjacent to lesions. ^{4 6} Immunoglobulin and complement deposits in the conjunctival epithelium, as well as circulating antibodies to both corneal and conjunctival epithelium, have been detected in affected patients. ^{1 2 3 7 8 9 10 11 12} In addition, clinical studies have established an epidemiologic relationship between chronic human hepatitis C virus (HCV) infection and Mooren’s ulcers. ^{13 14}  

For the analysis of human disease, the use of animal models is extremely useful, and nude mice have been engineered in which deficient T-cell functions are partially reconstituted by grafting of rat thymic rudiments (TG nude mice). Interestingly, multiple-organ localized autoimmune diseases with individual organ-specific autoantibodies develop in such animals. ^{15 16 17 18 19} Autoimmune keratitis is one of the conditions found at a high incidence in TG nude mice, and histologic and immuno-histologic studies have shown similarities with human ocular lesions. ²⁰  

Recently, we have established a hybridoma (OT-20) producing monoclonal antibodies specifically targeting the corneal epithelium using splenic B cells obtained from a TG nude mouse with keratitis. In the present study, we aimed to identify the corneal target antigen recognized by OT-20 and examined humoral and cellular immune responses of TG nude mice against the corneal antigen. As a result, α-internexin was identified as the specific corneal antigen reacting with OT-20, also being recognized by circulating antibodies in TG nude mice developing keratitis. Moreover, T cells from these mice exhibited proliferation responses and produced Th1-type cytokines on stimulation with α-internexin. 

Materials and Methods

Generation of TG Nude Mice

Four-week-old female BALB/c nude (nu/nu) mice (Charles River Japan, Inc. Atsugi, Japan) were used as recipients. Thymi were dissected from 15-day-old F344 rat embryos (Charles River Japan), and two thymic lobes were grafted under each renal capsule. Mice were housed in a specific pathogen-free room and were killed at 1, 2, 3, 4, 5, and 7 months after thymus grafting. Blood samples were collected via the axially artery with the mice under ether anesthesia, and sera from individual mice were stored at −80°C. The animal care and experimentation were performed in accordance with local regulations for use of animals in research, including the ARVO statements for the use of Animals in Ophthalmic and Vision Research. 

Histology

Organs were fixed in Bouin’s fixative, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E) for histologic examination. 

Corneal Protein Purification

Corneal tissues were separated from 32 BALB/c mice eyeballs and were homogenized in 1 mL of 50 mM Tris-HCL (pH 7.8), containing 15% glycerol, 150 mM NaCl, 0.1% Tween 20, and proteinase inhibitor, using a homogenizer (Eppendorf AG, Hamburg, Germany). The homogenized corneal extracts were centrifuged at 10,000g for 10 minutes, and the supernatants were collected and stored at −20°C. 

Indirect Immunofluorescence

Six-μm-thick cryostat tissue sections were prepared from the eyeball of a normal adult BALB/c mouse, fixed in acetone, and incubated with test serum (1:40 to 1:2560 dilutions) for 40 minutes, washed in PBS, and then incubated with FITC-labeled anti-mouse IgG (Cappel, Durham, NC) for 40 minutes. 

Establishment of a Hybridoma

Spleen cells from a TG nude mouse with keratitis were fused with the SP2/0 mouse myeloma cells, ²¹ and stably hybridized cells were selected with HAT media supplement (Sigma Chemical Co., St Louis, MO) as described. ²¹ Hybridomas producing antibodies against murine corneal epithelia were selected by immunohistochemistry using cryostat tissue sections, and, finally, one clone (OT-20: IgM) was established after cloning limiting dilution. 

Western Blot Analysis

Ten micrograms of protein from corneal tissue lysates was loaded on SDS-PAGE using 12% polyacrylamide gels (Daiichi Pure Chemicals, Tokyo, Japan) and was electrophoretically transferred to nitrocellulose membranes (Hybond ECL; Amersham Pharmacia Biotech, Arlington Heights, IL) (60 V for 4 hours at 4°C). These membranes were blocked with 3% skimmed milk for 1 hours at 37°C and were incubated with various dilutions of sera from TG nude mice or BALB/c mice, or with the monoclonal antibody OT-20 overnight at 4°C. After washing three times with PBS, the membranes were then incubated with an alkaline phosphatase-labeled anti-mouse IgG secondary antibody (Promega, Tokyo, Japan) for 1 hour at 37°C. After washing three times with PBS, bands were visualized with substrate (BCIP/NBT Color Development Substrate; Promega) according to the manufacturer’s specifications. Molecular weight markers were included to determine the sizes of bands. 

Recombinant α-internexin

α-internexin cDNA fragments were synthesized for production of a glutathione-S-transferase (GST) fusion protein by using a vector (pGEX4T-4; Amersham Pharmacia Biotech). For this purpose, α-internexin cDNA was amplified by PCR from the pRSVi-α-internexin vector using primers that contained the desired restriction enzyme recognition sites (primer 1: GGCACCGAATTCATGAGCTTCGGATCAGAG, primer 2: TAGCAACAGTCGACTTACATTTTTTGG). The primers were designed from the c-terminal sequence of the rat α-internexin protein. 

Primer I contained a synthetic EcoR I site, and primer II contained a synthetic Sal I site, and PCR products were digested with these restriction enzymes and subcloned into the pGEX4T-4 vector. Fusion proteins produced in XL1-blue Escherichia coli (Novagen, Darmstadt, Germany) according to the manufacturer’s recommendations were purified with a column (GSTrap; Amersham Pharmacia Biotech) according to the manufacturer’s protocol. 

RT-PCR

Corneas and livers were removed from adult BALB/c mice, and embryonic brains were removed from 15-day-old mouse embryos. Total RNAs were isolated from these tissues with reagent (Isogen; Nippon Gene, Tokyo, Japan) and reverse transcribed to cDNAs with a cDNA synthesis kit (Super Script First-Strand Synthesis System for RT-PCR; Invitrogen, Tokyo, Japan). 

PCR was performed in a 50 μL reaction mixture containing 10×PCR buffer (Applied Biosystems), 800 mM dNTP Mix (TAKARA, Tokyo, Japan), 1.25 U Taq polymerase (Applied Biosystems), and 25 mM of each primer (α-internexin: forward primer 5′-GTATGAGTCCCTGGCCGCTAA-3′, reverse primer 5′-GAGAACGTAAGGGGTCGGTACAA-3′, size: 841 bp. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH): forward primer 5′-ATGGTGAAGGTCGGTGTGAAC-3′ reverse primer 5′-GCCTTGACTGTGCCGTTGAAT-3′, size: 158 bp.) using a thermal schedule of 95°C for 30 seconds followed by 59°C for 30 seconds and 72°C for 30 seconds for α-internexin or 95°C for 30 seconds followed by 60°C for 30 seconds and 72°C for 30 seconds for GAPDH in a thermal cycler (PCR2700; Applied Biosystems) for 25 cycles. PCR products were separated on 1×TAE (40 mM Tris-acetate, 1 mM EDTA) 2% agarose gels that contained 0.5 mg/mL ethidium bromide. 

Proliferation Analysis

Aliquots of 1 × 10⁶ splenocytes obtained from normal BALB/c or TG nude mice were plated in a 96-well plates and cultured in triplicate with or without various concentrations of α-internexin recombinant protein in culture medium composed of RPMI 1640 medium, 10 mM HEPES, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 100 U/mL penicillin, 100 μg/mL streptomycin (all from BioWhttaker, Walkersville, MD), and 1 × 10⁻⁵ M 2-ME (Sigma Chemical), and supplemented with 10% fetal bovine serum (Sigma Chemical). Cells were cultured for 72 hours at 37°C in an atmosphere of 5% CO₂, then pulsed with 0.5 μCi [³H]thymidine for 12 hours before termination of culture, and harvested onto glass filters using an automated cell harvester (Tomtec, Orange, CT). Radioactivity was assessed by liquid scintillation spectrometry. The data are presented as stimulation indices, the mean cpm in cultures, with stimulus/mean cpm in control cultures without stimulus. 

Cytokine Assays

Aliquots of 1 × 10⁶ splenocytes obtained from normal BALB/c mice or TG nude were cultured with or without 5 μg/mL of α-internexin, and supernatants were collected after 72 hours for detection of IL-2, -4, -5, TNF-α, and IFN-γ by cytometric-bead-array immunoassay (BD Biosciences, San Jose, CA) as described previously. ²² Briefly, with this method, particles (polystyrene beads, 7.5 m; Bang’s Laboratories, Fishers, IN) are dyed to five different fluorescence intensities, with a proprietary dye having an emission wavelength of ∼650 nm (FL-3). Each particle is coupled via a covalent linkage based on thiol-maleimide chemistry with an antibody (Ab) (PharMingen, San Diego, CA) against IL-2, -4, -5, TNF-α, or IFN-γ as discrete populations, unique in their FL-3 intensity. The Ab particles serve to capture the given cytokines as an immunoassay panel and can be detected simultaneously in mixtures by direct immunoassay by using five different antibodies coupled to phycoerythrin (PE), which emits at ∼585 nm (FL-2). Two-color flow cytometric analysis was performed by using a flow cytometer (FACSCalibur flow cytometer; Becton Dickinson Immunocytometry Systems, San Jose, CA). 

Statistical Analysis

Statistical analyses were performed by the independent t-test. P < 0.05 was considered significant (significance is denoted by an asterisk in the figures). 

Results

Clinical Manifestations of Keratitis Observed in TG Nude Mice

Table 1summarizes data for the incidence, the onset time, and the clinical features of keratitis spontaneously developing in TG nude mice as evaluated by slit lamp and histologic examinations. Keratitis was clinically observed from 2 months after thymus grafting, and the incidence increased and reached almost 100% at 3 months after the grafting. 

Figure 1Dpresents typical clinical features of keratitis developing in TG nude mice at 3 months after thymus grafting. Corneal opacity was caused by inflammation and angiogenesis extended to the circle through all corneal limbus, which covered over the iris and the lens. Inflammatory lesions characterized by infiltrating cells, a high vessel density, and corneal edema were initially observed in peripheral cornea (Fig. 1E)and were gradually extended to the central zone (Fig. 1F) . In addition, keratitis merged with corneal ulceration along the limbus (Fig. 1E) . As indicated in Table 1 , although only 36% of TG nude mice featured inflammation in the center of the cornea at 2 months after thymus grafting, most of the TG nude mice had developed inflammation over the entire surface and limbus ulcers at 7 months. 

Detection of Circulating Autoantibodies against Cornea

To examine whether cornea-specific circulating autoantibodies are present in the serum of TG nude mice, indirect immunofluorescence was performed using cryostat tissue sections of normal BALB/c mouse cornea. Corneal epithelium and/or stroma were selectively stained with sera obtained from TG nude mice but not with sera from intact BALB/c mice (Fig. 2) . Sera from TG nude mice, which had not yet developed keratitis, also did not react with corneal epithelium and/or stroma. Although the target tissue (corneal epithelium or stroma) and the intensity of staining were not proportional to the severity and the extent of disease, ∼70% of TG nude mice with keratitis possessed both autoantibodies (autoantibodies to epithelium and to stroma). 

Identification of the Target Antigen

To identify the target antigen for autoimmune lesions developing in TG nude mice, we established hybridoma cell lines producing antibodies against target tissues of autoimmunity such as parietal cells of the stomachs, epithelial cells of lachrymal glands, and corneal epithelia. Some of these monoclonal antibodies reacted with intermediate filaments such as keratin, vimentin, and desmin (data not shown). One monoclonal antibody (OT-20) specifically reacted with corneal epithelium (Fig. 3A) . Representative results of Western blot analysis using OT-20 and corneal extracts are shown in Figure 3B . A 66-kDa protein was preferentially recognized by OT-20 and also reacted with sera from ∼70% of TG nude mice with keratitis but not with serum from either TG nude mice with normal corneas or normal BALB/c mice. To identify the 66kDa protein, we studied several intermediate filament proteins known as autoantigens of several autoimmune diseases for reaction with OT-20 and thereby identified α-internexin as the target antigen. 

α-internexin Expression in Normal Cornea

α-internexin, which is a 66-kDa neurofilament (NF) protein, is widely distributed in the early developing peripheral and central nervous system ²³ and is considered to serve as scaffolding for the coassembly of other neurofilaments like NF-H, -M, or -L. We focused on this protein from the results of immunoblot analysis by using OT-20 and examined mRNA expression of α-internexin in the cornea by RT-PCR. Results of representative experiments are displayed in Figure 4 . Because α-internexin is known to be expressed in the embryonic brain, mRNA extracted from the brains of 15-day-old mouse embryos was used as a positive control. RT-PCR analysis provided evidence that mRNA of α-internexin is expressed in normal corneal tissue at the same level as in the embryonic brain. 

Autoantibodies Specific for α-internexin Protein in TG Nude Mice

Subsequently, we generated a recombinant α-internexin protein in E. coli using the pGEX-4 to 1 vector. The resulting GST fusion protein was expressed as a 92-kDa band on SDS-PAGE, consistent with a GST moiety fused with the 66-kDa putative mature form of α-internexin (data not shown). Immunoblotting confirmed that the recombinant α-internexin fusion protein was recognized by OT-20 (Fig. 5) . Similarly, when sera from TG nude mice with keratitis were examined, autoantibodies specific for the recombinant α-internexin protein were detected in cases that reacted with corneal epithelium but not to stroma (data not shown). These results further indicated α-internexin to be the target antigen in corneal epithelium for circulating autoantibodies in TG nude mice with keratitis. 

T-Cell Responses of TG Nude Mice to α-internexin

If α-internexin is a pathogenic autoantigen for keratitis in TG nude mice, their T cells should recognize this antigen. Therefore, we investigated whether proliferation and cytokine production are observed when T cells obtained from TG nude mice are stimulated in vitro with α-internexin. Representative results of T-cell-proliferation assays are displayed in Figure 6and for cytokine production in Figure 7 . 

As expected, splenic T cells from TG nude mice developing keratitis with circulating antibodies against corneal epithelium manifested proliferative responses to α-internexin in a dose-dependent manner, this not being observed with T cells from normal BALB/c mice (Fig. 6) . Splenic T cells from TG nude mice that did not develop keratitis also exhibited a proliferation response to α-internexin stimulation, but the magnitude was much lower than the case with TG nude mice with keratitis. Moreover, corresponding to the results of T-cell proliferation responses, splenic T cells from TG nude mice with keratitis produced markedly more IL-2, IFN-γ, and TNF-α than those of normal BALB/c mice on stimulation with α-internexin (Fig. 7) . T cells from TG nude mice produced barely detectable levels of IL-4 and IL-5, and there were no significant differences when compared with α-internexin stimulation or with normal BALB/c mice. 

Discussion

The present study indicated that α-internexin is one of corneal antigens involved in keratitis developing spontaneously in TG-nude mice, as evidenced by the specific recognition by circulating antibodies and T cells of affected TG nude mice. 

When embryonic rat thymi are grafted into nude mice, the T-cell immune system becomes established in the mice, the thymic epithelial cells being of donor origin, while bone-marrow–derived cells, such as T cells and dendritic cells, are replaced by cells of the recipients. ¹⁵ Therefore, the T-cell maturation process may be disturbed at some stage, because of rat thymic epithelial and mouse immature T-cell xenogeneic interactions. Because MHC-restricted T-cell immune responses are preserved in TG nude mice, positive selection is considered to function. However, several severe organ-specific autoimmune diseases, such as dacryoadenitis, uveoretinitis, thyroiditis, gastritis, and orchitis, as well as keratitis, develop spontaneously, ^{15 16 17 18 19 20 24} suggesting that negative selection, by which autoreactive T cells are eliminated at the maturation process in thymus, ^{25 26} may be inappropriate or that generation of regulatory T cells such as CD25⁺ regulatory T cells may not be substantial. ²⁷  

α-internexin is a type IV intermediate filament protein that is expressed abundantly in neurons during development of the peripheral and central nervous systems. Only a few neurons express α-internexin in adult neural tissues. ^{28 29 30} However, upregulation has been reported in injured motoneurons. ³¹ In the eye, retinal horizontal cells, amacrine cells, and ganglion nerve fibers are known to express α-internexin in the early stage of development, while in the late stage, it is replaced by neurofilament subunits (NF-L, -M, -H) and vimentin. ³² However, information on expression of α-internexin in corneal tissue has remained unclear. Therefore, our novel observation that the cornea and the embryonic brain express α-internexin in equal amounts (Fig. 4) , in contrast to scanty α-internexin expression in the retina, is a significant finding. Because the cornea is a nonlymphatic, nonvascular, and immune privileged tissue, ^{33 34} corneal proteins are protected from systemic immunity under normal circumstances. This characteristic might contribute to the observed abundant expression of α-internexin. In fact, the brain is also an immune privileged site. However, we are not able, at present, to provide an appropriate explanation for the role of α-internexin in the cornea, and further experiments are needed to answer this question. 

The monoclonal antibody OT-20 specific for α-internexin reacts with the corneal epithelium but not with the stroma in indirect immunofluorescence. On the other hand, autoantibodies specific for corneal stroma presented in the sera of TG nude mice with keratitis (Fig. 2) , indicating that other corneal target antigens are also involved in the development of keratitis, a possibility currently under investigation in our laboratory. We are also trying to establish a murine model of keratitis by immunization of α-internexin. 

In TG nude mice, keratitis occurs bilaterally and progresses chronically and irreversibly. Interestingly, histologic and serologic features, and the course of corneal lesion development in TG nude mice are strikingly reminiscent of the lesions described in a variety of human immunogenic keratitis cases. In the early stage in TG nude mice, infiltrating inflammatory cells are observed only at the periphery of the cornea (Fig. 1E) , and, as the disease progresses, infiltration extends gradually to the center of the cornea, accompanied by angiogenesis and stromal edema (Fig. 1F) . This histopathology is, in part, similar to Mooren’s ulcer and rejection reactions after keratoplasty. 

Because 30% of Mooren’s ulcer patients possessed antibodies to HCV, it has been suggested that Mooren’s ulcer might be associated with chronic HCV infection. ¹³ Interestingly, the amino-acid sequence of human α-internexin shares homology with that of HCV (our recent computer analysis). We are undertaking a study to examine whether α-internexin acts as an autoantigen, with production of specific antibodies in patients suffering from various forms of autoimmune keratitis, including Mooren’s ulcer and postkeratoplasty rejection. 

 Submitted for publication June 20, 2005; revised July 21, 2005; accepted November 16, 2005.

 Disclosure: T. Hattori, None; M. Takeuchi, None; K. Ohno, None; M. Ogawara, None; T. Asatani, None; Y. Usui, None; R. Muramatsu, None; M. Inagaki, None; M. Usui, None; O. Taguchi, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Incidence of Autoimmune Keratitis in TG Nude Mice

Months after Thymus Grafting	No. of Mice Used	No. of Mice with Disease (%)
				P	W	P + S	W + S	A	U
1	10	0	0	0	0	0	0
2	14	2 (14)	0	4 (29)	5 (36)	10 (71)	4 (29)
3	47	8 (22)	1 (3)	4 (11)	21 (57)	29 (78)	36 (97)
4	28	7 (25)	1 (4)	2 (7)	18 (64)	23 (82)	26 (93)
5	24	6 (25)	1 (4)	4 (17)	13 (54)	20 (83)	23 (96)
7	28	3 (11)	0	0	25 (89)	26 (93)	28 (100)

 P, peripheral cornea inflammation; W, whole cornea inflammation; S, stromal edema; A, angiogenesis; U, ulcer.

Figure 1.

 

Keratitis developing in TG nude mice. (A, D) Stereomicroscopic observations of normal BALB/c mouse cornea (A) and TG nude mouse cornea (D). (B, C) Micrographs of a normal BALB/c mouse limbus (B) and the center (C) of a cornea. (E, F) Micrographs of a limbus (E) and the center of a cornea (F) of a TG nude mouse with keratitis. (E) Early stage keratitis in a TG nude mouse (2 months after thymus grafting). Infiltration of lymphocytes and neutrophils is only apparent at the periphery. (F) Late stage keratitis in a TG nude mouse (7 months after thymus grafting). Infiltration of lymphocytes and neutrophils are apparent in the center of the cornea. (B, C, E, F) H&E staining; magnification ×200.

Figure 2.

 

Indirect immunofluorescence of normal BALB/c mouse cornea using sera from a normal BALB/c mouse (A), and from TG nude mice (B, C). Note negative staining with BALB/c serum (A), while the corneal epithelium (B) or stroma (C) are selectively positive with TG nude mice sera. Magnification, ×200.

Figure 3.

 

Monoclonal antibody OT-20 produced by a hybridoma reacts specifically with corneal epithelium and a 66 kDa corneal protein. (A) Indirect immunofluorescence using the monoclonal antibody OT-20 established from a TG nude mouse with keratitis. (B) Western blot analysis using corneal extracts and OT-20 (lane 1), serum from a TG nude mouse with keratitis (lane 2), or serum from a normal BALB/c mouse (lane 3).

Figure 4.

 

Expression of α-internexin mRNA in murine tissue extracts. mRNA was isolated from the cornea (lane 1), embryonic brain (lane 2), and liver (lane 3), and expression of α-internexin was analyzed by RT-PCR.

Figure 5.

 

The GST α-internexin fusion protein (94 kDa) detected by serum from TG nude mice with keratitis. Western blot analysis was performed using the GST α-internexin fusion protein (94 kDa) and OT-20 (lane 1), serum from a normal BALB/c mouse (lane 2), or sera from TG nude mice with autoimmune keratitis (lanes 3 and 4).

Figure 6.

 

Proliferation responses to in vitro α-internexin stimulation in spleen cell cultures. Spleens were obtained from normal BALB/c mice (▴), TG nude mice with keratitis (•), and TG nude mice without keratitis (▪). Each data point represents a result obtained with pooled splenic T-cell samples from groups of four mice. Splenic T cells were stimulated with the indicated concentrations of α-internexin for 96 hours. The results are expressed as stimulation indices. The experiment was repeated three times with similar patterns of results.

Figure 7.

 

Cytokine production by splenic T cells stimulated with α-internexin. Spleen cells were obtained from normal BALB/c mice or TG nude mice and cultured with or without 5 μg/mL of α-internexin. Supernatants were collected 72 hours after the initiation of culture, and IL-2, IFN-γ, TNF-α, IL-4, and IL-5 concentrations were assayed with the cytometric bead array system (BD Biosciences). Bar 1: TG nude T cells cultured with α-internexin; Bar 2: TG nude T cells culture without α-internexin; Bar 3: Normal BALB/c T cells cultured with α-internexin; Bar 4: Normal BALB/c mice T cells cultured without α-internexin. The asterisk indicates a P value <0.05 between TG nude T cells and BALB/c mice T cells cultured with α-internexin.