IL-18 Not Required for IRBP Peptide–Induced EAU: Studies in Gene-Deficient Mice

Abstract

purpose. Interleukin (IL)-18 has been described as a proinflammatory cytokine in rheumatoid arthritis and bacterial infectious diseases. The present study was designed to determine the role of IL-18 in a model of ocular experimental autoimmune uveitis (EAU). The initial studies were conducted to detect the expression of IL-18 in normal mouse eye tissue, and the later studies investigated induction of EAU in mice with an IL-18^−/− phenotype.

methods. IL-18 detection was performed by using 5-bromo-4-chloro-3-indoyl-β--d-galactopyranoside (X-Gal) staining on frozen sections of eyes from mice (129/CD1, DBA1, and Balb/c), either of normal phenotype (+/+) or of deficiency (±, −/−) in the IL-18 gene which had been replaced by introduced genes including LacZ under the control of an IL-18 promotor. Severity of EAU was assessed in DBA1 and 129/CD1 wild-type (WT) or IL-18 knockout (KO) mice after immunization with the uveitogenic antigen: interphotoreceptor retinal binding protein (IRBP) peptide 161-180. Lymphocyte proliferation and cytokine production were also measured in WT and IL-18 KO DBA1 mice 15 days after immunization.

results. IL-18 is constitutively expressed in the epithelial cells in iris, ciliary body, and retina. EAU-resistant mice (129/CD1) with an IL-18^−/− phenotype remained resistant after immunization with IRBP peptide (P161-180). However, EAU-susceptible mice (DBA1) exhibited disease with similar histologic characteristics, despite a generalized reduction of interferon (IFN)-γ and tumor necrosis factor (TNF)-α on an IL-18^−/− phenotype. DBA1 IL-18^−/− also demonstrated reduced IL-10 production.

conclusions. The IL-18 gene is not necessary for the initiation or pathogenesis of EAU induced by IRBP peptide 161-180. IL-18 is expressed in the epithelial cells in iris, ciliary body, and retina in the eyes, but its role in the eye remains undetermined.

Experimental autoimmune uveitis (EAU) is an organ-specific, T helper (Th)1 cell– mediated disease that targets the photoreceptor-associated antigens of the eye. ¹ It can be induced in many species with uveitogenic retinal antigens. ² The pathologic appearance of murine EAU closely resembles the lesions of several human noninfectious uveitic diseases ^{3 4} and serves as an ideal animal model for the study of the mechanisms and therapeutic approaches of human posterior uveitis. 

Interleukin (IL)-18 is a pleiotropic cytokine involved in the activation of Th1 cytokine responses, Fas ligand (Fas-L) expression, and both CC and CXC chemokine induction. ^{5 6} First designated as interferon (IFN)-γ–inducing factor, ⁷ IL-18 is identified as a Th1-type cytokine. ⁸ IL-18 induces proliferation of, upregulates IL-2Rα expression by, and promotes IFN-γ, tumor necrosis factor (TNF)-α, and granulocyte-macrophage colony-stimulating factor (GM-CSF) production by Th1 clones. ^{9 10 11} Although IL-18 shares some biologic activities with IL-12, both cytokines have different receptor-binding activities and signal transduction pathways, ⁸ IL-18 synergizes with IL-12 in the modulation of the development of Th1 and natural killer (NK) cells. ¹² Synthesis of both IL-18 and IL-12 have been described in activated macrophages, ^{13 14} which are important in the effector phase of EAU. ^{15 16} In vivo studies show that IL-18 plays a critical role in the regulation of Th1 and Th2 balance after Leishmania major and Staphylococcus aureus infection. ¹⁷ Furthermore, studies in NOD mice have suggested that IL-18 plays a potential role in predisposition to autoimmunity. ¹⁸ We therefore sought to identify IL-18 expression in the eye tissues, and to investigate whether Th1-cell–mediated EAU develops in the absence of IL-18. 

Materials and Methods

Mice

Mice with a targeted disruption of IL-18 gene were generated in 129/CD1 as described, ¹⁷ and backcrossed onto DBA1 and Balb/c strains for five generations. Briefly, a targeting vector was designed to replace approximately 5 kb of the IL-18 gene, which contains exons 4, 5, and 6 and part of exons 3 and 7. The start codon for IL-18 was also deleted by the replacement with the LacZ gene encoding a bacterialβ -galactosidase and neomycin-resistant genes, which is confirmed by Southern blot analysis with both 5′ and 3′ external probes. Inbred mice at the age of 10 to 12 weeks were used for the experiments. The procedures adopted conformed to the regulations of the Animal License Act (UK) and to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Analysis for IL-18 in DBA1 Mice

Three mice were chosen randomly from DBA1 wild-type (WT) or IL-18 knockout (KO) groups. Approximately 1 cm of tail tip from each mouse was digested at 55°C overnight in lysis buffer (10 mM Tris[ pH 8.0], 50 mM EDTA, 100 mm NaCl, 0.5% sodium dodecyl sulfate [SDS] and 500 μg/ml proteinase K). The PCR primers were designed according to a targeted and WT gene sequences: Primer (P)1, 5′-ACTCTATAAATCATCCAGCCTCGGGTATTC; P2, 5′-CTCTTAACCCCGAGCCTTTCATCGCTCCTG and P3, 5′-ACCGCTATCAGGACATAGCGTTGCTACCCCGT. A 550-bp polymerase chain reaction (PCR) product was amplified by P1 and P2, which indicates the IL-18 KO mouse, whereas a 400-bp PCR product indicates the WT mouse and was amplified by P1 and P3. Furthermore, spleen cells from naive IL-18 WT and KO DBA1 mice were cultured at 2 × 10⁶ cells/well with 10 ng/ml lipopolysaccharide (LPS) in 24-well plate for 48 hours. IL-18 in the supernatant was estimated by enzyme-linked immunosorbent assay (ELISA; cytokine measurement), with standard and antibodies from R&D Systems (Abingdon, UK). 

Antigen

Interphotoreceptor retinal binding protein (IRBP) peptide 161-180 (SGIPYIISYLHPGNTILHVD; purity >95%) was synthesized by Sigma Genosys (Cambridge, UK). 

Frozen Section Preparation and X-Gal Staining

Eyes from mouse strains of 129/CD1, DBA1 and Balb/c (−/−, ±, or +/+) were frozen in optimal temperature cutting compound (OCT; Miles, Elkhart, IN) immediately after removal. For 5-bromo-4-chloro-3-indoyl-β-d-galactopyranoside (X-Gal) staining, 5- to 8-μm cryostat sections were first fixed for 5 minutes at room temperature with 2% (vol/vol) formaldehyde-0.2% (vol/vol) glutaraldehyde in phosphate-buffered saline (PBS). After rinsing with PBS three times, slides were incubated at 37°C overnight in the stain solution containing 1 mg/ml X-Gal, 5 mM potassium ferrocyanide, and 2 mM MgCl₂ in PBS. Sections were washed and mounted for microscopic evaluation and photography. 

EAU Induction and Evaluation

Mice of the strain of 129/CD1 (+/+,−/−) and DBA1 (+/+,−/−) were immunized subcutaneously with 100 μg IRBP peptide 161-180 emulsified with an equal volume of complete Freund’s adjuvant (CFA, H37Ra; Difco, Detroit, MI) in a total volume of 100 μl. An additional intraperitoneal injection of 0.5 μg of purified Bordetella pertussis toxin (PTX, Strain Wellcome 28; Speywood, Clwyd, UK) in 250 μl was also administered to each animal. 

Animals were killed on day 15 after immunization. Ten eyes from five WT and five KO DBA1 mice were frozen in OCT immediately to obtain frozen sections. The remaining eyes were removed and fixed in 2.5% buffered glutaraldehyde and embedded in resin for standard hematoxylin and eosin (H&E) staining. The intensity of uveoretinitis was evaluated histologically and graded using a modified version of the customized histologic grading system ^{16 19} by independent observers, as described. This grading system allows the observer to differentiate between the inflammatory cell infiltrate and the structural damage to the retina/choroid, because inflammatory cell infiltration does not always lead to retinal damage. 

Antibodies and Immunohistochemical Staining

Purified rat anti-mouse antibodies against CD4 (L3T4, H129) and biotinylated-CD11c (HL3) were from PharMingen (San Diego, CA). Other antibodies for F4/80 (C1:A3-1), sialoadhesin (3d6.112), and major histocompatibility complex (MHC)-II (P7/7) were from Serotec (Oxford, UK). 

Immunostaining staining was performed as described previously. ¹⁶ Briefly, 8-μm cryostat sections were first fixed with 100% acetone and then rehydrated in Tris-buffered solution (TBS), followed by incubation with the primary purified monoclonal antibody for at least 1 hour. Samples were then incubated with biotinylated rabbit anti-rat antibody and streptavidin-alkaline phosphatase (AP) for 30 minutes each. All three were purchased from Dako (High Wycombe, UK). Finally, the signal was visualized with substrate fast red and naphthol AS-BI phosphate in Tris buffer (pH 9.7), which produced a bright red color. Levamisole was added as usual to the AP substrate to block endogenous alkaline phosphatase activity. All these chemicals were from Sigma (Poole, UK). All procedures were conducted at room temperature. 

For dual staining of X-Gal and CD11c, the slide was treated for X-Gal staining as described, followed by the standard immunohistochemical staining steps. 

Lymphocyte Proliferation

DBA1 mice of IL-18^+/+ and ^−/− phenotypes were immunized with IRBP peptide and killed on day 15. The lymph nodes draining the site of immunization (inguinal) or eyes (cervical) were collected and pooled within each group separately. Triplicate cultures of 2 × 10⁵ cells/well were stimulated with different concentrations of peptide in 96-well plates in RPMI supplemented with l-glutamine, antibiotics, sodium pyruvate, nonessential amino acids, 2-mercaptoethanol, and 10% fetal calf serum (FCS). The cultures were incubated for 60 hours and were pulsed with[ ³H]-thymidine for the last 16 hours. 

Cytokine Measurement

Cell culture supernatants were obtained from the inguinal lymph node lymphocytes cultured at 2 × 10⁶ cells/well in 24-well plate after 72 hours’ stimulation with 50μ g/ml peptide. TNF-α, IFN-γ, IL-4, IL-10, and IL-12 were measured by ELISA using antibodies and standards from PharMingen. Briefly, 96-well plates were coated with the appropriate anti-cytokine antibodies overnight. After blocking the plates with bovine serum albumin and a further 2-hour incubation with supernatants or standard, the plates were developed using biotin-conjugated anti-cytokine antibodies. Horseradish peroxidase–conjugated streptavidin was added before development with substrate. 

Statistical Analysis

Statistical analysis was performed by computer (SPSS software; SPSS, Chicago, IL). Analysis of disease incidence was performed byχ ², and EAU grades (nonparametric) were analyzed by Mann–Whitney test. Analysis of lymphocyte proliferation responses and cytokine production was performed by independent Student’s t-test. P < 0.05 was considered statistically significant. 

Results

IL-18 Expression in the Epithelial Cells in Mice Iris, Ciliary Body, and Retina

IL-18 KO mice were generated by homologous recombination and backcrossed onto other backgrounds. Homozygous IL-18 KO and WT mice were bred from littermates. Naive IL-18 KO or WT mice have similar percentages of CD4⁺, CD8⁺ T cells, B cells, and NK cells by flow cytometric analysis (data not shown). Splenic T cells also produced similar levels of proliferative response and cytokines, such as IFN-γ and IL-4, when activated with concanavalin A (ConA) or anti-CD3 antibody (data not shown). Our PCR result confirmed the genotype of the mice. Moreover, the culture supernatant from DBA1 mice spleen cells proved that a high level of IL-18 was produced by WT mice, but no IL-18 was produced by the KO mice (Fig. 1) . 

The deleted IL-18 gene was replaced with a bacterialβ -galactosidase (LacZ) gene which can be visualized as blue-stained foci at the single-cell level. X-Gal staining was performed on normal WT, heterozygous, and homozygous mice with different backgrounds (129/CD1, DBA1, Balb/c). No blue X-Gal staining was seen in the WT Balb/c iris, ciliary body, and retina (Fig. 2A ) nor in 129/CD1 and DBA1 mice (data not shown). X-Gal–positive foci (representing IL-18) were clearly expressed in the retina, ciliary body, and iris of IL-18 KO mice (Fig. 2) . The even distribution and location of the blue foci indicates that positive staining was associated with epithelial cells in the iris, ciliary body, and retina. In the ciliary body, some staining appeared to localize in the stroma. However, further staining of the eye sections with F4/80 or CD11c showed no positive colocalization with dendritic cells or macrophages ²⁰ at these sites, further indicating that most of the X-Gal–positive staining was associated with the local epithelial cells. Control dual staining (Fig. 2F) on the lymph nodes from IL-18 KO mice showed similar absence of colocalization of CD11c-positive cells and X-Gal–positive cells, indicating that in this system dendritic cells do not show constitutive expression of IL-18. There was no difference of LacZ expression in the eye tissues among the three strains. 

EAU in IL-18 KO and WT DBA1 Mice

DBA1 mice are moderately susceptible to IRBP peptide 161-180–induced EAU. ²¹ Therefore we investigated how EAU develops in the DBA1 IL-18 KO mice. Immunization with IRBP peptide 161-180 induced EAU in the IL-18 homozygous (−/−) mice with a severity similar to that in the control animals (+/+) (Fig. 3 , P > 0.05). The characteristics of the disease and infiltrating cell phenotype stained with macrophage makers (F4/80), T cell markers (CD4), and dendritic cell markers (MHC-II) showed no significant difference between the two groups (Fig. 4) . Moreover, macrophages stained with sialoadhesin and dendritic cells stained with CD11c were also easily detected in the diseased retina from the IL-18 KO mice (data not shown). 

EAU in IL-18 KO and WT 129/CD1 Mice

IL-18, as a Th1-inducing cytokine, is considered to act synergistically with IL-12 during antigen presentation. However, as shown earlier, IL-18 gene deletion did not diminish the severity of EAU in EAU-susceptible mice. In contrast, both IL-18 KO and WT 129/CD1 mice remained resistant after IRBP peptide immunization, indicating that IL-18 gene deletion does not confer susceptibility on EAU-resistant mice by IRBP peptide 161-180. These results suggest that IL-18 does not play a major role in protection against the disease. 

Proliferation and Cytokine Response in IL-18 KO DBA1 Mice

To test whether IL-18 KO mice have a different cytokine profile, draining lymph nodes were collected and lymphocyte isolated and cultured on day 15 after immunization for assay of cell proliferation and cytokine levels in the supernatant. The results (Fig. 5) show that IL-18–deficient lymphocytes from the inguinal lymph nodes proliferated well to the immunizing antigen with no statistically significant difference from WT mice, whereas both cervical lymphocytes failed to proliferate. However, cytokine production by T cells from IL-18 deficient mice showed a generalized reduction of both Th1 (IFN-γ, TNF-α) and Th2 (IL-10) cytokines, but no difference in IL-12 level was seen between the two groups (Fig. 6) . IL-4 levels were too low to be detected in either group. 

Discussion

To detect IL-18 expression in the eye tissues, the LacZ gene was knocked into the murine IL-18 locus, thereby linkingβ -galactosidase expression to transcription through the target promoter. The expression pattern of the IL-18–targeted LacZ gene mirrored the sites of IL-18 gene expression. We used X-Gal staining to identify the single-cell expression of IL-18 in the eye tissues using this LacZ knock-in strategy. Of note, there was a constitutive expression of IL-18 in the epithelial cells in iris, ciliary body, and retina in BALB/c, 129/CD1, and DBA1 mice. This was also confirmed by our reverse transcription–polymerase chain reaction (RT-PCR) detection of IL-18 mRNA in human retinal pigment epithelial (RPE) cell lines (data not shown). RPE cells are an important component of the blood–retinal barrier and have an immunoregulatory role in retinal autoimmunity by phagocytosing and recycling autoantigen-rich rod outer segment. ²² The expression of IL-18 in the RPE cells may indicate a role for IL-18 in autoimmune retinal diseases similar to its role in other inflammatory diseases where it is expressed by epithelial cells in intestinal epithelium. However, our studies of the EAU model do not support a significant role for ocular IL-18 in this circumstance. 

IL-18 has been shown to play a proinflammatory role in human rheumatoid arthritis (RA) and in the RA mouse model, ²³ but it is not essential for IFN-γ production by microbial agents. ²⁴ Also, Xu et al. ²⁵ reported that IL-18 receptor (R) is selectively expressed by Th1 cells, but not Th2 cells and suggested IL-18R as a specific marker for Th1 cells. Overall, it has been hypothesized that IL-18 is an important cytokine in the Th1-mediated immune response in collaboration with IL-12. In the present study, by using IL-18–deficient mice backcrossed into DBA1 mice, which are susceptible to IRBP peptide 161-180, we showed that EAU was inducible in DBA1 IL-18 KO mice, and remained noninducible in the resistant 129/CD1 IL-18–deficient mice. These findings differ from those in the L. major infection mouse model, in which the immune response shifted from Th1 type to Th2 type with the disruption of the IL-18 gene. ¹⁷ Previous reports have suggested that Th2 mechanisms may come into play when Th1 mechanisms are inhibited in EAU. ²⁶ However, we have not found in the IL-18^−/− mice any evidence for Th2 upregulation; the IL-4 cytokine was not detectable, and IL-10 levels were reduced. Moreover, the cellular response of the IL-18 KO mice, measured by antigen-specific lymphocyte proliferation, showed no difference between the two groups. However, a reduced level of IFN-γ and TNF-α observed in our EAU model is in agreement with the evidence that IL-18 is a costimulator for the Th1 cytokines (e.g., IFN-γ, TNF-α production). ^{8 17}  

There was a notable corresponding reduction in IL-10 cytokine in IL-18 KO mice, thus confirming the pleiotropic nature of IL-18 cytokine. This supports a recent study that revealed an induction of the Th2 cytokine IL-13 by IL-18. ²⁷ Taken together, our results suggest that IL-18 may act as a coinducer of both Th1 and Th2 cytokines in our EAU model. Because a balance of cytokine production is likely to determine the outcome of the immune responses, ¹³ these effects in EAU have no resultant influence on the disease’s severity. 

No difference in IL-12 secretion in the lymphocyte culture supernatant was observed (Fig. 6) between IL-18^+/+ and ^−/− mice after EAU immunization, indicating that IL-12 production is independent of IL-18, as suggested by Kohno and Kurimoto. ¹¹ IL-12 is required for induction and expression of EAU. ²⁸ From theses results, we may conclude that IL-12 rather than IL-18 is the key cytokine that is essential for EAU induction. In view of the data by Gieni et al. ²⁹ that antigen-driven IL-12 is a mechanism by which the genetic background can affect the pattern of cytokine synthesis by T cells during the development of adaptive immune responses, the present observations that EAU is inducible in IL-18 KO DBA1 mice and not in the 129/CD1 mice can be explained. 

A further conclusion from the present data is that macrophages and dendritic cells with no IL-18 expression can function normally. Functional IL-18 can be produced by different subtypes of murine and human dendritic cells, ³⁰ and dendritic cell-derived IL-18 can enhance IL-12–dependent Th1 development. ³¹ Moreover IL-18 expressed on osteoblasts inhibits osteoclast-like multinucleated cell formation, not through production of IFN-γ, but through production of GM-CSF, ³² an important cytokine for dendritic cell differentiation and maturation. Our data show that autoimmunity to retinal antigens can be induced in the absence of IL-18 and in the context of the described cytokine profile in IL-18 KO mice when compared with WT mice. In addition, no difference was observed between WT mice and KO mice from the immunohistochemical staining results of infiltrating MHC-II- and F4/80-positive cells in the retina. This suggests that IL-18 is neither necessary for dendritic cells to present antigen, nor for macrophages to damage the target tissue. 

In summary, our studies show that IL-18 was constitutively expressed in the local cells in the eye tissues, which may contribute to the local cytokine profile for the development of EAU. Also the data showed that IL-18 could be a coinducer of both Th1 and Th2 cytokines in the EAU model. It is possible that the balance of cytokines determines whether EAU was inducible in the moderately susceptible DBA1 mice, but remained uninducible in the 129/CD1 mice with the disruption of the IL-18 gene. 

 Submitted for publication June 13, 2000; revised September 13, 2000; accepted September 29, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

(A) In three tail tips from mice in each group, the genotype was confirmed by PCR. The 550-bp PCR product indicates IL-18 knockout genotype, and the 400-bp product indicates the WT genotype. (B) Spleen cells from naive DBA1 WT and KO mice were cultured and stimulated with 10 ng/ml LPS for 48 hours. ELISA results show that IL-18 was not detectable (ND) in all three homozygous IL-18 KO mice, whereas WT littermate spleen cells produced a significantly high level of IL-18 after LPS stimulation.

Figure 2.

 

X-Gal staining (bright blue) of the eye tissues from WT or KO Balb/c, 129/CD1, and DBA1 mice. (A) Absence of X-Gal staining in the WT Balb/c retina; the same result was observed in retina from 129/CD1 and DBA1 mice. (B) Positive X-Gal staining (arrows) in IL-18 KO DBA1 mouse ciliary body. (C, D, and E) positive staining (arrows) in the retina from IL-18 KO Balb/c, DBA1, and 129/CD1 mice, respectively. (F) Dual staining of X-Gal and CD11c in a DBA1 IL-18 KO mouse lymph node, showing no colocalization of the expression of IL-18 (arrows) and CD11c (arrowheads).

Figure 3.

 

EAU grades in IRBP peptide 161-180–immunized IL-18 KO and WT DBA1 mice. Mice were immunized with 100 μg of peptide and 0.5 μg of PTX. Eyes were collected for histopathology 15 days after immunization. EAU was graded using a previously reported grading system. ^{16 19} There was no statistically significant difference between the two groups in the incidence and grades of the disease. Incidence of EAU is labeled on the top of each column (diseased/total mice).

Figure 4.

 

Glutaraldehyde-fixed DBA1 mice eyes with standard H&E staining. (A) Showing normal retina structure in an IL-18 KO DBA1 mouse. ILM, inner limiting membrane; INL, inner nuclear layer; ONL, outer nuclear layer; ROS, photoreceptor rod outer segment; RPE, retinal pigment epithelium; Ch, choroid; Sc, sclera. No significant difference was shown in the severity of IRBP-induced EAU between WT mice (B) and KO mice (C). (D through I) Immunohistochemical staining of the diseased DBA1 retina from WT and KO group mice on day 15 after immunization with a panel of anti-mouse monoclonal antibodies. (D, E, and F) WT retina; (G, H, and I) IL-18 KO retina. No obvious differences were observed with the staining of MHC-II (D and G), CD4 (E and H), and F4/80 (F and I) between the two groups.

Figure 5.

 

Inguinal and cervical lymph node cells from WT and KO DBA1 mice 15 days after immunization cultured with IRBP peptide161-180 at different concentrations. T-cell proliferation was determined. Lymphocytes from the WT and KO cervical lymph nodes failed to proliferate, and the lines overlap. Data are means ± SD, n = 3. cLN, cervical lymph nodes; iLN, inguinal lymph nodes.

Figure 6.

 

Production of cytokines by inguinal draining lymph node cells from WT and KO DBA1 mice cultured with 50 μg/ml of peptide at 15 days after injection. Data are means ± SD, n = 3;* P < 0.05.