The Cytokine Interleukin-6 and the Chemokines CCL20 and CXCL13 Are Novel Biomarkers of Specific Endogenous Uveitic Entities

Home , CCL20, CXCL13

Immunology and Microbiology The Cytokine Interleukin-6 and the Chemokines CCL20 and CXCL13 Are Novel Biomarkers of Speciﬁc Endogenous Uveitic Entities

Ahmed M. Abu El-Asrar,1,2 Nele Berghmans,3 Saleh A. Al-Obeidan,1 Ahmed Mousa,1 Ghislain Opdenakker,3 Jo Van Damme,3 and Soﬁe Struyf3

1Department of Ophthalmology, College of Medicine, King Saud University, Riyadh, Saudi Arabia 2Dr. Nasser Al-Rashid Research Chair in Ophthalmology, King Saud University, Riyadh, Saudi Arabia 3Department of Microbiology and Immunology, Rega Institute for Medical Research, University of Leuven, Leuven, Belgium

Correspondence: Ahmed M. Abu El- PURPOSE. The purpose of this study was to determine levels of the cytokines IL-1b, IL-6, IL-21, Asrar, Department of Ophthalmolo- IL-22, and IL-23 and the chemokines CXCL13, CCL19, CCL20, and CCL21 in aqueous humor gy, King Abdulaziz University Hospi- (AH) samples from patients with specific uveitic entities. tal, Old Airport Road, P.O. Box 245, Riyadh 11411, Saudi Arabia; METHODS. Paired serum samples (n ¼ 13) and AH samples (n ¼ 111) from patients with active [email protected], or idiopathic granulomatous uveitis (IGU) or with uveitis associated with HLA-B27-related [email protected]. inflammation, Beh¸cet’s disease (BD), Vogt-Koyanagi-Harada (VKH) disease, or sarcoidosis and Submitted: April 18, 2016 control patients were analyzed in two different multiplex assays. Accepted: July 25, 2016 RESULTS. Cytokines IL-1b, IL-21, IL-22, and IL-23 were not detected in any AH sample. Citation: Abu El-Asrar AM, Berghmans Chemokine CCL21 concentrations in serum were significantly higher than those in AH. N, Al-Obeidan SA, et al. The cytokine CCL19 levels in AH and serum were not significantly different. Levels of CCL20 and CXCL13 interleukin-6 and the chemokines in AH were significantly higher than those in serum. IL-6 was not detected in serum samples. CCL20 and CXCL13 are novel bio- IL-6 AH levels were significantly higher in patients with HLA-B27-associated uveitis and in BD markers of specific endogenous patients than in patients with VKH disease, sarcoidosis, and IGU (P < 0.0001). CCL20 AH uveitic entities. Invest Ophthalmol Vis levels were significantly higher in HLA-B27-associated uveitis than in BD, VKH, sarcoidosis, Sci. 2016;57:4606–4613. DOI:10.1167/iovs.16-19758 and IGU (P ¼ 0.001), whereas CXCL13 AH levels were significantly higher in VKH disease and IGU than in HLA-B27-associated uveitis, BD, and sarcoidosis (P ¼ 0.007). CONCLUSIONS. IL-6-driven immune responses are more potent in HLA-B27-associated uveitis and BD than in VKH disease, sarcoidosis, and IGU. CCL20 appears to be a specific biomarker of HLA-B27-associated uveitis, whereas CXCL13 appears to be a biomarker of VKH disease and IGU. Our findings suggest that IL-6, CCL20, and CXCL13 could serve as drug targets for treatment of specific clinical entities of endogenous uveitis. Keywords: Beh¸cet’s disease, chemokines, cytokines, HLA-B27, sarcoidosis, uveitis, Vogt- Koyanagi-Harada disease

ndogenous uveitis is a clinically heterogenous group of for clinical intervention. For example, biological therapies that E potentially blinding intraocular inflammatory diseases. They inhibit the proinflammatory cytokine tumor necrosis factor-a often occur in conjunction with systemic inflammatory (TNF-a) are now widely used in routine clinical practice for the diseases, such as HLA-B27-associated uveitis, Beh¸cet’s disease treatment of several autoimmune inflammatory diseases, (BD), Vogt-Koyanagi-Harada (VKH) disease, and sarcoidosis. including endogenous refractory uveitis associated with BD.7,8 Because endogenous uveitis includes a wide range of diverse TNF-a levels in the aqueous humor (AH) were significantly inflammatory manifestations in the eye, with different clinical higher in patients with Beh¸cet’s disease than in patients with phenotypes, it is possible that different immunopathogenic other causes of endogenous uveitis,3 These therapies display mechanisms are involved in each clinical subtype. Moreover, specific modes of action and have been developed on the basis causes and outcomes of the disease continue to be largely of a detailed understanding of cytokine involvement in unpredictable and unexplained. inflammation. Nevertheless, management of uveitis has steadily improved Dysregulated expression of the proinflammatory cytokine in recent decades, mostly due to enhanced understanding of interleukin-6 (IL-6) has been shown to be involved in the the pathophysiology of the disease, which has been translated development of chronic inflammatory autoimmune diseases. IL- into effective therapies. However, despite the remarkable 6 exerts a variety of biological activities on responsive cell successes achieved, the clinical complexity and the variable populations through its binding to transmembrane IL-6 responses to therapy leave multiple challenges in uveitis receptor (IL-6R) as well as soluble IL-6R (sIL-6R).9–11 Trans- management. Cytokines and chemokines control inflammation, forming growth factor-b (TGF-b) and the inflammatory cyto- and they have important roles in the pathology of various kines, such as IL-1b, IL-6, IL-21, and IL-23 contribute to the 1–6 endogenous uveitis entities, thus representing major targets generation of T helper (Th) 17 cells that play a crucial role in iovs.arvojournals.org j ISSN: 1552-5783 4606

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Intraocular Cytokines and Chemokines in Uveitis IOVS j September 2016 j Vol. 57 j No. 11 j 4607 the pathogenesis of several autoimmune and inflammatory TABLE 1. Multiplex Assay-Dependent, Specific Detection Range of diseases through production of effector molecules, such as IL- Chemokines and Cytokines 17 and IL-22.12 Chemokines are multifunctional mediators that regulate Assay Detection Range, ng/mL leukocyte recruitment to the inflamed tissue, promote Bio-Plex Pro human chemokine kit* inflammation, and enhance immune responses. Chemokines CCL21 0.039–160.7 are divided into four subgroups (CXC, CC, C, and CX C) 3 CXCL13 0.001–6.7 depending on the arrangement of the conserved cysteine CCL20 0.001–6.7 residues. The specific effects of chemokines are mediated by CCL19 0.003–49.2 binding to distinct members of a family of G protein-coupled receptors.13,14 CCL20 is the only chemokine known to interact Milliplex MAP kit† with CC chemokine receptor 6 (CCR6). The CCL20-CCR6 axis CCL20 0.01–10.0 is responsible for the chemoattraction of immature dendritic IL-22 0.04–161.2 cells, B lymphocytes, and activated and memory CD4þ and IL-1b 0.02–18.9 CD8þ Tlymphocytes.15 CCL20 is implicated in several IL-21 0.02–20.5 autoimmune diseases, such as rheumatoid arthritis16–20 and IL-23 0.37–1522.5 experimental autoimmune encephalomyelitis.21,22 IL-6 0.01–10.2 There is increasing evidence that B lymphocytes play an important role in the pathogenesis of autoimmune diseases, * Bio-Rad Laboratories, Temse, Belgium. † Merck Millipore, Molsheim, France. such as rheumatoid arthritis,23–25 Sjogren’s syndrome,26–28 multiple sclerosis,29–31 myasthenia gravis,32,33 systemic lupus erythematosus34,35 and autoimmune thyroiditis.36 Akey The procedure was performed using a surgical microscope. phenomenon in these diseases is the formation of ectopic The samples were snap frozen and maintained at 708C until lymphoid aggregates with germinal center-like structures in the used. Aqueous humor samples from patients with uveitis were inflamed tissues, which contain proliferating B lymphocytes, obtained before they underwent therapy. All procedures plasma cells, follicular helper CD4þ T lymphocytes, and a followed the tenets of the Declaration of Helsinki, and network of follicular dendritic cells.26–28 Accumulated evi- informed consent was obtained from all patients and the dence has demonstrated that the formation and maintenance control subjects. The study was approved by the Research of ectopic lymphoid structures in these chronic inflammatory Center, College of Medicine, King Saud University. conditions is critically dependent on the ectopic expression of Two sets of samples were assayed for cytokines and the lymphoid chemokines CXCL13, CCL19, CCL21, and chemokines. In the first set of samples (n ¼ 56), CXCL13, CXCL12 and their specific receptors CXCR5 (for CXCL13), CCL19, CCL20, and CCL21 were quantified using a Bio-Plex CCR7 (for CCL19 and CCL21), and CXCR4 (for CXCL12).26–28 Pro human chemokine kit (Bio-Rad Laboratories, Temse, The aim of this study was to identify novel molecules Belgium), whereas in the second set of samples (n ¼ 76), involved in the pathogenic mechanisms of endogenous uveitis CCL20, IL-1b, IL-6, IL-21, IL-22, and IL-23 were quantified using that could serve as potential targets for selective therapy or a Milliplex MAP kit (Merck Millipore, Molsheim, France) that may act as novel biomarkers for different entities of according to the protocols provided by the manufacturers. endogenous uveitis. We, therefore, measured the levels of the The detection ranges of the assays as provided by the inflammatory cytokines IL-1b, IL-6, IL-21, IL-22, and IL-23 and manufacturers are listed in Table 1. the chemokines CXCL13, CCL19, CCL20, and CCL21 in the AH The first set of AH samples was obtained from 13 patients from patients with active uveitis associated with human with HLA-B27-associated uveitis, 11 with BD, 12 with VKH leukocyte antigen (HLA)-B27-related intraocular inflammation, disease, seven with sarcoidosis, and five with IGU. Eight BD, VKH disease, and sarcoidosis and from patients with patients who had undergone cataract extraction served as the idiopathic granulomatous uveitis (IGU). control group. The second set of AH samples was obtained from 17 patients with HLA-B27-associated uveitis, 15 with BD, PATIENTS AND METHODS 15 with VKH disease, six with sarcoidosis, and nine with IGU. Fourteen patients who had undergone cataract extraction Patients with active uveitis seen at the outpatient clinic of King served as a control group. Abdulaziz University Hospital were included in the study. To address whether cytokines and chemokines detected in Patients who had undergone cataract extraction with no the AH could originate from blood circulation, we measured history of uveitis served as the control group. Conditions were the levels of cytokines and chemokines in 13 serum samples. diagnosed using established clinical criteria, with supporting The serum samples were obtained from three patients with laboratory evidence as needed.37,38 Uveitis was further HLA-B27-associated uveitis, four with BD, three with VKH classified as nongranulomatous if fine endothelial keratic disease, two with sarcoidosis, and one with IGU. precipitates were seen in the absence of iris nodules and/or choroidal granulomas, and as granulomatous if larger keratic Statistical Methods precipitates, including large (‘‘mutton-fat’’) keratic precipitates, or Koeppe and/or Busacca iris nodules, and/or optic disc Data management was preliminarily done using Excel 2013 or choroidal granulomas were seen. In each patient, the uveitis (Microsoft, Redmond, WA, USA), and then all statistics were activity was graded according to the criteria of the Standard- analyzed using SPSS version 20.0 software (IBM, Armonk, NY, ization of the Uveitis Nomenclature working Group grading USA). Undetected values were treated as ‘‘zero’’ in terms of scheme.39 None of the patients was taking topical or systemic analysis due to their trivial decimal values, which were quite therapy on presentation. close to zero. Descriptive statistics were calculated so that Aqueous humor (100–200 lL) was aspirated from each categorical variables were presented as frequency (percent) patient by means of limbic paracentesis, using a 27-gauge and continuous variables as means (6standard deviation). The needle attached to a tuberculin syringe, after the application of chi-square test was used to compare the detection rates. The the topical local anesthetic oxybuprocaine hydrochloride 0.4% Kruskal-Wallis test was used to compare means among (Benoxinate; Chauvin Pharmaceuticals Ltd., Kingston, UK). different disease categories. Mann-Whitney U-test was then Intraocular Cytokines and Chemokines in Uveitis IOVS j September 2016 j Vol. 57 j No. 11 j 4608

FIGURE 1. Comparisons of mean chemokine levels in serum with levels in aqueous humor samples from patients with uveitis, using the Bio-Plex multiplex assay kit. used to compare means from two independent groups. CCL20 and CXCL13, respectively). Pairwise comparisons Spearman correlation coefficients were computed to investi- (Mann-Whitney U-test) indicated that CCL20 levels in patients gate correlations between variables. A P value less than 0.05 with HLA-B27-associated uveitis were significantly higher than indicated statistical significance. in patients with BD, VKH disease, sarcoidosis, and IGU (P ¼ 0.014; P < 0.0001; P ¼ 0.002; and P ¼ 0.012, respectively). The CXCL13 levels in VKH disease were significantly higher than RESULTS the levels in HLA-B27-associated uveitis and BD (P ¼ 0.004; and Analysis of Aqueous Humor From Uveitis Patients P ¼ 0.007, respectively). In addition, mean CXCL13 levels in IGU were significantly higher than the mean levels in HLA-B27- by Chemokine Bio-Plex Multiplex Assay associated uveitis (P ¼ 0.022). Among the chemokines analyzed in the AH from patients, only When patients were divided into those with nongranulo- CCL20 and CXCL13 levels were significantly higher than those matous uveitis (HLA-B27-associated uveitis and BD) (n ¼ 24) in serum samples (Fig. 1). These findings suggest that local and those with granulomatous uveitis (VKH disease, sarcoid- chemokine production is the authentic source of CCL20 and osis, and IGU) (n ¼ 24), CCL20 levels in the AH from CXCL13 within the ocular microenvironment and that a nongranulomatous uveitis (4.9 6 7.5 ng/mL) were significantly systemic inflow mechanism is rather unlikely. CCL20 and higher than those in granulomatous uveitis cases (0.1 6 0.2 CXCL13 were not detected in any of the AH samples from the ng/mL; P < 0.0001). On the other hand, CXCL13 levels in control group. granulomatous uveitis (2.4 6 4.1 ng/mL) were significantly The incidence rates for CCL20 and CXCL13 detection are higher than those in nongranulomatous uveitis (0.7 6 2.4 ng/ shown in Table 2. The incidence rate for detection of CCL20 in mL; P < 0.0001). AH samples from HLA-B27-associated uveitis patients (100%) was significantly higher than in patients with VKH disease (58.3%) and sarcoidosis (42.9%) (P ¼ 0.036; P ¼ 0.014, Correlations Between CCL20 Levels in AH and respectively; chi-square test). Disease Activity The chemokine levels in the five disease groups were compared (Kruskal-Wallis test), and the results are shown in The levels of CCL20 in AH samples correlated significantly with Table 2. CCL20 and CXCL13 levels in AH samples differed the disease activity in all patients (r ¼ 0.6; P < 0.0001) (Fig. 2), significantly among HLA-B27-associated uveitis, BD, VKH in patients with BD (r ¼ 0.88; P < 0.0001) and in patients with disease, sarcoidosis, and IGU (P ¼ 0.001 and P ¼ 0.007 for nongranulomatous uveitis (r ¼ 0.63; P ¼ 0.001). A significant

TABLE 2. Summary Data for CCL20 and CXCL13 Levels by Bio-Plex Multiplex Assay Kit

CCL20 CXCL13

No. of Samples With Mean 6 SD Levels No. of Samples With Mean 6 SD Levels Disease Detectable Levels (%) Detected, ng/mL Detectable Levels (%) Detected, ng/mL

HLA-B27-associated uveitis (n ¼ 13) 13 (100) 6.4 6 8.3 8 (61.5) 0.2 6 0.3 Behcet’s disease (n ¼ 11) 8 (72.7) 1.6 6 5.4 8 (72.7) 1.0 6 3.3 VKH disease (n ¼ 12) 7 (58.3) 0.1 6 0.1 11 (91.7) 2.3 6 4.2 Sarcoidosis (n ¼ 7) 3 (42.9) 0.09 6 0.2 5 (71.4) 0.9 6 1.2 Idiopathic granulomatous uveitis (n ¼ 5) 4 (80) 0.3 6 0.3 5 (100) 3.6 6 5.1 P value (Kuskal-Wallis test) 0.001* 0.007* VKH, Vogt-Koyanagi-Harada disease. * Statistically signiﬁcant at 5% level of signiﬁcance. Intraocular Cytokines and Chemokines in Uveitis IOVS j September 2016 j Vol. 57 j No. 11 j 4609

FIGURE 2. Correlations between aqueous humor (AH) cytokine and chemokine levels and disease activity. In these analyses, all patients were included. The levels of CCL20 in AH (measured by Bio-Plex multiplex assay kit) (A) and by Milliplex multiplex assay kit (B) correlated significantly with disease activity. In addition, CCL20 levels were significantly higher in patients with anterior chamber reaction >2þ than those in patients with anterior chamber reaction 2þ. Also, the levels of IL-6 in AH (measured by Milliplex multiplex assay kit) correlated significantly with disease activity (C). However, although IL-6 levels were higher in patients with anterior chamber reaction >2þ than those in patients with anterior chamber reaction, 2þ statistical significance was not reached. Intraocular Cytokines and Chemokines in Uveitis IOVS j September 2016 j Vol. 57 j No. 11 j 4610

TABLE 3. Summary Data for CCL20 and IL-6 Levels by Milliplex Multiplex Assay Kit

CCL20 IL-6

No. of Samples With Mean 6 SD Levels No. of Samples With Mean 6 SD Levels Disease Detectable Levels (%) Detected, ng/ml Detectable Levels (%) Detected, ng/ml

HLA-B27-associated uveitis (n ¼ 17) 13 (88.2) 0.74 6 0.81 17 (100) 29.2 6 12.5 Behcet’s disease (n ¼ 15) 3 (20) 0.2 6 0.6 15 (100) 16.3 6 16.7 VKH disease (n ¼ 15) 4 (26.7) 0.02 6 0.04 12 (80) 3.9 6 6.9 Sarcoidosis (n ¼ 6) ND ND 6 (100) 0.7 6 0.6 Idiopathic granulomatous uveitis (n ¼ 9) 2 (22.2) 0.04 6 0.08 9 (100) 11.9 6 13.1 P value (Kruskal-Wallis test) <0.0001* <0.0001* ND, not detected; VKH, Vogt-Koyanagi-Harada disease. * Statistically significant at 5% level of significance. negative correlation was found between AH levels of CCL20 than those in granulomatous uveitis (5.7 6 9.5 ng/mL; P < and CXCL13 in all patients (r ¼ 0.38; P ¼ 0.008). 0.0001). Similarly, CCL20 levels in AH from nongranulomatous uveitis (0.49 6 0.77 ng/mL) were significantly higher than that Determination of Cytokine and Chemokine in granulomatous uveitis (0.02 6 0.05 ng/mL; P < 0.0001). Concentrations in Aqueous Humor Samples From Uveitis Patients by Milliplex Multiplex Assay Correlations Between IL-6 and CCL20 Aqueous Humor Levels and Disease Activity Among the cytokines and chemokines analyzed in uveitis, only IL-6 and CCL20 were detected in AH samples but not in serum The levels of IL-6 (r ¼ 0.55; P < 0.0001) and CCL20 (r ¼ 0.59; P samples. IL-1b, IL-21, IL-22, and IL-23 were not detected in any < 0.0001) in AH samples correlated significantly with the of the AH samples from uveitis patients and the control group. disease activity in all patients (Fig. 2). In addition, the levels of IL-6 was detected in only five of 14 control AH samples, IL-6 correlated significantly with disease activity in patients whereas CCL20 was not detected in any of the control AH with HLA-B27-associated uveitis (r ¼ 0.58; P ¼ 0.014). Finally, a samples. IL-6 was detected in all AH samples from patients significant positive correlation was found between aqueous with uveitis, except in three samples from patients with VKH humor levels of IL-6 and CCL20 in all patients (r ¼ 0.69; P < disease(Table3).Whenthewholestudygroupwas 0.0001). considered, IL-6 levels were significantly higher in the AH of patients (n ¼ 62) (14.7 6 15.7 ng/mL) than in that of controls (n ¼ 14; 0.3 6 0.9 ng/mL; P < 0.0001). In a comparison DISCUSSION between individual patient groups and the control group, In the present study, IL-6 levels in the AH of uveitis patients significantly increased IL-6 levels were observed for HLA-B27- were significantly enhanced compared to those in controls and associated uveitis, BD, VKH disease, sarcoidosis, and IGU (P < were higher in patients with HLA-B27-associated uveitis and 0.0001; P < 0.0001; P ¼ 0.002; P ¼ 0.003; and P < 0.0001, BD than in patients with VKH disease, sarcoidosis, and IGU. respectively). Furthermore, a significantly positive correlation between IL-6 Similar to the results obtained with the first set of AH levels in the AH from patients with uveitis and the clinical samples (Table 2), the incidence rate of CCL20 detection in disease activity was detected in all patients and in patients with HLA-B27-associated uveitis in the second set of AH samples HLA-B27-associated uveitis. IL-6 was originally identified as B- (Table 3) was significantly higher than in BD, VKH disease, cell stimulatory factor 2, inducing differentiation of activated B sarcoidosis, and IGU (P ¼ 0.0004; P ¼ 0.029; P ¼ 0.0007; and P cells into immunoglobulin-producing plasma cells. However, it ¼ 0.0034, respectively; chi-square test). To evaluate distribution is a multipotent cytokine, regulating both B and T cell of the IL-6 and CCL20 levels among the five disease groups, we activation/differentiation.10,11,40–42 IL-6 in combination with þ used the Kruskal-Wallis test, and the results are shown in Table TGF-b, is essential for Th17 differentiation from na¨ıve CD4 T 3. IL-6 and CCL20 levels in AH samples differed significantly cells, whereas IL-6 inhibits TGF-b-induced regulatory T cells among patients with HLA-B27-associated uveitis, BD, VKH (Treg) development.11,12,43 The resultant predominance of disease, sarcoidosis, and IGU (P < 0.001 for both compari- Th17 cells over Treg caused by IL-6 may be responsible for the sons). Pairwise comparisons indicated that IL-6 levels in HLA- disruption of immunological tolerance and the development of B27-associated uveitis were significantly higher than in VKH inflammatory autoimmune diseases. Numerous studies have disease, sarcoidosis, and IGU (P < 0.0001; P ¼ 0.006; and P ¼ confirmed the pathological roles of IL-6 in inflammatory and 0.016, respectively) and that IL-6 levels in BD were significantly autoimmune diseases.9–11 Our findings suggest a role for IL-6 in higher than in VKH disease and sarcoidosis (P ¼ 0.007; and P ¼ the development of endogenous uveitis, particularly in HLA- 0.002, respectively). As shown in the first set of AH samples B27-associated uveitis and BD and imply that the IL-6–IL-6R analyzed with the Bio-Plex assay (Table 2), the Milliplex assay axis might be a target pathway for the treatment of these also confirmed that, in the second set of patients (Table 3), diseases. Clinical trials have demonstrated the efficacy of CCL20 levels were significantly higher in HLA-B27-associated tocilizumab, a humanized anti-IL-6R antibody, for patients with uveitis than in BD, VKH disease, sarcoidosis, and IGU (P ¼ rheumatoid arthritis, Castleman’s disease, and systemic juve- 0.001; P < 0.0001; P ¼ 0.001; and P ¼ 0.001, respectively). nile idiopathic arthritis, leading to approval of this innovative When patients were divided into those with nongranulo- drug for the treatment of these diseases.9–11 Small retrospec- matous uveitis (HLA-B27-associated uveitis and BD; n ¼ 32) and tive studies demonstrated the efficacy of tocilizumab in the those with granulomatous uveitis (VKH disease, sarcoidosis, treatment of refractory uveitic macular edema.44–46 As an and IGU) (n ¼ 30), IL-6 levels in the AH from nongranulo- alternative for IL-6, Th17 polarization might be driven through matous uveitis (23.1 6 15.8 ng/mL) were significantly higher an IL-1b/IL-1R signaling pathway downstream of Mincle/ Intraocular Cytokines and Chemokines in Uveitis IOVS j September 2016 j Vol. 57 j No. 11 j 4611 caspase activation and recruitment domain 9 (CARD 9)- of inflammation. CXCL13 expressed by follicular dendritic cells is dependent signaling and inflammasome activation. The latter central in this process as it attracts CXCR5-expressing B pathway is activated by, for example, mycobacterial peptido- lymphocytes and follicular helper CD4þ T lymphocytes to the glycan47 and also plays a central role in autoimmune diseases of ectopic follicles.25–28 In rheumatoid arthritis synovia containing the eye.48 lymphoid aggregates, elevated levels of CXCL13 are observed,25 Several studies identified the CCL20-CCR6 axis as a and several studies identified CXCL13 as a new potential marker potential mediator of inflammatory cell recruitment in the for severity of inflammation in rheumatoid arthritis,23,24 systemic rheumatoid arthritis joint. Rheumatoid arthritis synovial fluid lupus erythematosus,34,35 multiple sclerosis,30,31 and myasthenia contained significantly more CCL20 than that in osteoarthri- gravis.32 Supportive for a role for CXCL13 as an important tis,16–18 and CCL20 and CCR6 expression levels were up- mediator of human autoimmune disorders, CXCL13 neutralizing regulated in rheumatoid arthritis synovial tissue.16,17 The monoclonal antibody significantly reduced disease severity in upregulated production of CCL20 induced the recruitment of animal models of rheumatoid arthritis and experimental autoim- 52 CCR6-expressing mononuclear cells including Th17 cells to mune encephalomyelitis. inflamed joints in rheumatoid arthritis and its animal mod- It must be noted that, although research on other 18,20 el. In addition to expressing CCR6, Th17 cells also extraocular autoimmune disorders can be very informative, synthesize CCL20, thereby establishing a positive feedback the eye is an immune-privileged site that actively regulates and mechanism, which may perpetuate the chronic inflammatory directs immune responses that take place in its ‘‘territory.’’53 process in rheumatoid arthritis.18 Several studies reported that The immune privilege of the eye limits tissue damage after an the proinflammatory cytokines IL-1b,TNF-a,andIL-17 episode of immune activation and allows allogenic, unmatched stimulate the production of CCL20 by rheumatoid fibroblast- corneal transplantation. It implies strong intraocular tolerance, like synovial cells17–19 and that IL-6 increased IL-1b-, TNF-a-, which is achieved through production of the anti-inflammatory and IL-17-induced CCL20 production.19 On the other hand, cytokine TGF-b and through the presence of different subsets of CCL20 enhanced the production of IL-6 in fibroblast-like Tregs in the ocular surface tissues.54 However, when immune synovial cells.20,49 In agreement with these studies, our privilege is faced with a serious challenge, it fails dismally to analysis identified a significant positive correlation between prevent severe destruction and permanent structural damage in the AH levels of CCL20 and those of IL-6. Administration of uncontrolled uveitis.55 In those occasions, immune therapy blocking anti-CCR6 monoclonal antibody inhibited inflamma- might be beneficial. However, the cytokine patterns in uveitis tion in a mouse model of arthritis, supporting the implication seem to be nonredundant, and tailored immune therapy might of CCL20-CCR6 axis in the pathogenesis of rheumatoid be required. Indeed in the present study, levels of the B-cell arthritis.18 In experimental autoimmune encephalomyelitis, a chemoattractant CXCL13 were significantly higher in patients model of T lymphocyte-dependent inflammation, both CCL20 with VKH disease and IGU than in patients with BD, HLA-B27- and its receptor CCR6 were upregulated in the spinal cord and associated uveitis, and sarcoidosis. These findings suggest that B draining lymph nodes. Disease severity and accumulation of lymphocytes may be pathogenetically important in uveitis mononuclear cells were significantly reduced by administra- associated with VKH disease and IGU and that CXCL13 levels in tion of specific neutralizing anti-CCL20 and anti-CCR6 antibod- AH might be a useful biomarker for this pathology. Interestingly, ies and in gene-targeted CCR6-deficient mice.21,22 several reports described predominance of B lymphocytes in In this study, the concentrations of CCL20 in the AH were the uveal inflammatory infiltrate in granulomatous uveitic quantified with the use of two multiplex assays. Both of these entities, such as sympathetic ophthalmia,56–59 multifocal commercial assays showed that CCL20 AH levels significantly choroiditis60 and progressive subretinal fibrosis.61 Additionally, correlated with disease activity. Use of the Milliplex assay B lymphocytes were identified in choroidal inflammatory showed that the incidence of detection of CCL20 in AH from infiltrate in two cases of VKH disease with ‘‘sunset glow patients with HLA-B27-associated uveitis was significantly fundus.’’62 B lymphocyte aggregates were described in the uvea higher than in patients with BD, VKH disease, sarcoidosis, of one patient with end-stage VKH disease.63 These findings and IGU. In addition, the Bio-Plex assay showed that the suggest that treatment with rituximab, an antibody against incidence of CCL20 detection in HLA-B27-associated uveitis CD20 leading to depletion of B lymphocytes, might be effective was significantly higher than in patients with VKH disease and in patients with VKH disease and IGU. Several clinical trials sarcoidosis. The detection limit in the Bio-Plex assay was lower have demonstrated the effectiveness of rituximab in treating than that in the Milliplex assay. Moreover, the Bio-Plex assay signs and symptoms of multiple autoimmune diseases.26,64 yielded higher concentrations than the Milliplex assay. It Recently, rituximab was reported to be effective in the should be noted that other studies have reported differences in treatment of 2 cases with diffuse subretinal fibrosis65 and one the absolute concentrations of analytes when comparing case of resistant VKH disease.66 different multiplex cytokine analysis kits.50,51 By using both In summary, our findings suggest that intraocular levels of the Bio-Plex and Milliplex assays, we found CCL20 levels were cytokines and chemokines differ depending on the cause of significantly higher in patients with HLA-B27-associated uveitis uveitis. In the ocular inflammatory microenvironment of patients than in the other uveitic entities. Our findings suggest that with endogenous uveitis, CCL20 might serve as a biomarker of CCL20 might serve as a biomarker for HLA-B27-associated inflammation in HLA-B27-associated uveitis, and the B cell uveitis and that the CCL20–CCR6 axis may serve as an chemoattractant CXCL13 might serve as a biomarker of uveitis excellent drug target for HLA-B27-associated uveitis. associated with VKH disease. IL-6-driven immune responses are Among the lymphoid chemokines analyzed, only CXCL13 more potent in HLA-B27-associated uveitis and BD than in VKH levels in the AH from patients with uveitis were significantly disease and sarcoidosis. Our findings also suggest that CCL20, higher than in sera. These findings suggest that CXCL13 is CXCL13, and IL-6 could serve as excellent drug targets for the locally produced within the ocular microenvironment and that treatment of specific clinical entities of endogenous uveitis. a systemic inflow mechanism is rather unlikely. In addition, CXCL13 was not detected in any of the control AH samples. Acknowledgments A sizeable subset of patients with autoimmune diseases, such as rheumatoid arthritis, Sjogren’s syndrome, multiple sclerosis, The authors thank Connie B. Unisa-Marfil for secretarial work. myasthenia gravis, systemic lupus erythematosus, and autoim- Supported by King Saud University through Vice Deanship of mune thyroiditis develop ectopic lymphoid structures at the sites Research Chair, Dr. Nasser Al-Rashid Research Chair in Ophthal- Intraocular Cytokines and Chemokines in Uveitis IOVS j September 2016 j Vol. 57 j No. 11 j 4612 mology (AMAE-A), the Fund for Scientific Research of Flanders, the via CCL20 in rheumatoid arthritis and its animal model. J Exp Interuniversity Attraction Poles Program, Belgian Science Policy Med. 2007;204:2803–2812. Office (project P7/40), and the Concerted Research Actions (2013/ 19. Kawashiri SY, Kawakami A, Iwamoto N, et al. Proinflammatory 015) of the Regional Government of Flanders (SS and GO). cytokines synergistically enhance the production of chemo- Disclosure: A.M. Abu El-Asrar, None; N. Berghmans, None; S.A. kine ligand 20 (CCL20) from rheumatoid fibroblast-like Al-Obeidan, None; A. Mousa, None; G. Opdenakker, None; J. synovial cells in vitro and serum CCL20 is reduced in vivo Van Damme, None, S. Struyf, None by biologic disease-modifying antirheumatic drugs. J Rheu- matol. 2009;36:2397–2402. References 20. Tanida S, Yoshitomi H, Nishitani K, et al. CCL20 produced in the cytokine network of rheumatoid arthritis recruits CCR6þ 1. Abu El-Asrar AM, Struyf S, Descamps FJ, et al. Chemokines and mononuclear cells and enhances the production of IL-6. gelatinases in aqueous humor of patients with active uveitis. Cytokine. 2009;47:112–118. Am J Ophthalmol. 2004;138:401–411. 21. Kohler RE, Caon AC, Willenborg DO, Clark-Lewis I, McColl SR. 2. Abu El-Asrar AM, Al-Obeidan SA, Kangave D, et al. CXC A role for macrophage inflammatory protein-3a/CC chemo- chemokine expression profiles in aqueous humor of patients kine ligand 20 in immune priming during T cell-mediated with different clinical entities of endogenous uveitis. Immu- inflammation of the central nervous system. J Immunol. 2003; nobiology. 2011;216:1004–1009. 170:6298–6306. 3. Abu El-Asrar AM, Struyf S, Kangave D, et al. Cytokine profiles 22. Liston A, Kohler RE, Townley S, et al. Inhibition of CCR6 in aqueous humor of patients with different clinical entities of function reduces the severity of experimental autoimmune endogenous uveitis. Clin Immunol. 2011;139:177–184. encephalomyelitis via effects on the priming phase of the 4. Kogiso M, Tanouchi Y, Miki S, Mimura Y. Characterization of T- immune response. J Immunol. 2009;182:3121–3130. cell subsets, soluble interleukin-2 receptors and interleukin-6 23. Bugatti S, Manzo A, Vitolo B, et al. High expression levels of in Vogt-Koyanagi-Harada disease. Jpn J Ophthalmol. 1992;36: the B cell chemoattractant CXCL13 in rheumatoid synovium 37–43. are a marker of severe disease. Rheumatology (Oxford). 2014; 5. Takase H, Futagami Y, Yoshida T, et al. Cytokine profile in 53:1886–1895. aqueous humor and sera of patients with infectious or 24. Greisen SR, Schelde KK, Rasmussen TK, et al. CXCL13 noninfectious uveitis. Invest Ophthalmol Vis Sci. 2006;47: predicts disease activity in early rheumatoid arthritis and 1557–1561. couldbeanindicatorofthetherapeutic‘‘window of 6. Ooi KG, Galatowicz G, Calder VL, Lightman SL. Cytokines and opportunity.’’ Arthritis Res Ther. 2014;16:434. chemokines in uveitis: is there a correlation with clinical 25. Shi K, Hayashida K, Kaneko M, et al. Lymphoid chemokine B phenotype? Clin Med Res. 2006;4:294–309. cell-attracting chemokine-1 (CXCL13) is expressed in germinal 7. Al Rashidi S, Al Fawaz A, Kangave D, Abu El-Asrar AM. Long- center of ectopic lymphoid follicles within the synovium of term clinical outcomes in patients with refractory uveitis chronic arthritis patients. J Immunol. 2001;166:650–655. associated with Beh¸cet disease treated with infliximab. Ocul 26. Finch DK, Ettinger R, Karnell JL, Herbst R, Sleeman MA. Effects Immunol Inflamm. 2013;21:468–474. of CXCL13 inhibition on lymphoid follicles in models of 8. Maya JR, Sadig MA, Zapata LJ, et al. Emerging therapies for autoimmune disease. Eur J Clin Invest. 2013;43:501–509. noninfectious uveitis: what may be coming to the clinics. J 27. Corsiero E, Bombardieri M, Manzo A, Bugatti S, Uguccioni M, Ophthalmol. 2014;2014:310–329. Pitzalis C. Role of lymphoid chemokines in the development 9. Schaper F, Rose-John S. Interleukin-6: biology, signaling and of functional ectopic lymphoid structures in rheumatic strategies of blockade. Cytokine Growth Factor Rev. 2015;26: autoimmune diseases. Immunol Lett. 2012;145:62–67. 475–487. 28. Pitzalis C, Jones GW, Bombardieri M, Jones SA. Ectopic 10. Tanaka T, Narazaki M. Ogata A, Kishimoto T. A new era for the lymphoid-like structures in infection, cancer and autoimmu- treatment of inflammatory autoimmune diseases by interleu- nity. Nat Rev Immunol. 2014;14:447–462. kin-6 blockade strategy. Semin Immunol. 2014;26:88–96. 29. Haugen M, Frederiksen JL, Degn M. B cell follicle-like 11. Yao X, Huang J, Zhong H, et al. Targeting interleukin-6 in structures in multiple sclerosis—with focus on the role of B inflammatory autoimmune diseases and cancers. Pharmacol cell activating factor. J Neuroimmunol. 2014;273:1–7. Ther. 2014;141:125–139. 30. Sellebjerg F, Börnsen L, Khademi M, et al. Increased 12. Maddur MS, Miossec P, Kaveri SV, Bayry J. Th17 cells. Biology, cerebrospinal fluid concentrations of the chemokine CXCL13 pathogenesis of autoimmune and inflammatory diseases, and in active MS. Neurology. 2009;73:2003–2010. therapeutic strategies. Am J Pathol. 2012;181:8–18. 31. Krumbholz M, Theil D, Cepok S, et al. Chemokines in multiple 13. Murphy PM, Baggiolini M, Charo IF, et al. International Union sclerosis: CXCL12 and CXCL13 up-regulation is differentially of Pharmacology, XXII. Nomenclature for chemokine recep- linked to CNS immune cell recruitment. Brain. 2006;129(Pt tors. Pharmacol Rev. 2000;52:145–176. 1):200–211. 14. Struyf S, Proost P, Van Damme J. Regulation of the immune 32. Shiao YM, Lee CC, Hsu YH, et al. Ectopic and high CXCL13 response by the interaction of chemokines and proteases. Adv chemokine expression in myasthenia gravis with thymic Immunol. 2003;81:1–44. lymphoid hyperplasia. J Neuroimmunol. 2010;221:101–106. 15. Schutyser E, Struyf S, Van Damme J. The CC chemokine CCL20 33. Meraouna A, Cizeron-Clairac G, Panse RL, et al. The chemo- and its receptor CCR6. Cytokine Growth Factor Rev. 2003;14: kine CXCL13 is a key molecule in autoimmune myasthenia 409–426. gravis. Blood. 2006;108:432–440. 16. Ruth JH, Shahrara S, Park CC, et al. Role of macrophage 34. Schiffer L, Worthmann K, Haller H, Schiffer M. CXCL13 as a inflammatory protein-3a and its ligand CCR6 in rheumatoid new biomarker of systemic lupus erythematosus and lupus arthritis. Lab Invest. 2003;83:579–588. nephritis—from bench to bedside? Clin Exp Immunol. 2014; 17. Matsui T, Akahoshi T, Namai R, et al. Selective recruitment of 179:85–89. CCR6-expressing cells by increased production of MIP-3a in 35. Wong CK, Wong PTY, Tam LS, Li EK, Chen DP, Lam CWK. rheumatoid arthritis. Clin Exp Immunol. 2001;125:155–161. Elevated production of B cell chemokine CXCL13 is correlated 18. Hirota K, Yoshitomi H, Hashimoto M, et al. Preferential with systemic lupus erythematosus disease activity. J Clin recruitment of CCR6-expressing Th17 cells to inflamed joints Immunol. 2010;30:45–52. Intraocular Cytokines and Chemokines in Uveitis IOVS j September 2016 j Vol. 57 j No. 11 j 4613

36. Aust G, Sittig D, Becherer L, et al. The role of CXCR5 and its 50. Lash GE, Scaife PJ, Innes BA, et al. Comparison of three ligand CXCL13 in the compartmentalization of lymphocytes in multiplex cytokine analysis systems: Luminex, SearchLight and thyroids affected by autoimmune thyroid diseases. Eur J FAST Quant. J Immunol Methods. 2006;309:205–208. Endocrinol. 2004;150:225–234. 51. Fu Qin, Zhu J, Van Eyk JE. Comparison of multiplex 37. Al Dhahri H, Al Rubaie K, Hemachandran S, et al. Patterns of immunoassay platforms. Clin Chem 2010;56:314–318. uveitis in a university-based tertiary referral center in Riyadh, 52. Klimatcheva E, Pandina T, Rilly C, et al. CXCL13 antibody for Saudi Arabia. Ocul Immunol Inflamm. 2015;23:311–319. the treatment of autoimmune disorders. BMC Immunol. 2015; 38. Abu El-Asrar AM, Al Tamimi M, Hemachandran S, Al-Mezaine 16:6. HS, Al-Muammar A, Kangave D. Prognostic factors for clinical 53. Stein-Streilein J, Caspi RR. Immune privilege and the philos- outcomes in patients with Vogt-Koyanagi-Harada disease ophy of immunology. Front Immunol. 2014;5:110. treated with high-dose corticosteroids. Acta Ophthalmol. 2013;91:e486–e493. 54. Stern ME, Schaumburg CS, Dana R, Calonge M, Niederkorn JY, Pflugfelder SC. Autoimmunity at the ocular surface: pathogen- 39. Standardization of Uveitis Nomenclature (SUN) Working esis and regulation. . 2010;3:425–442. Group: standardization of uveitis nomenclature for reporting Mucosal Immunol clinical data: results of the first international workshop. Am J 55. Forrester JV, Xu H. Good news-bad news: the yin and yang of Ophthalmol. 2005;140:509–516. immune privilege in the eye. Front Immunol. 2012;3: article 40. Taga T, Kawanishi Y, Hardy RR, Hirano T, Kishimoto T. 338. Receptors for B cell stimulatory factor 2. Quantification, 56. Abu El-Asrar AM, Struyf S, Van den Broeck C, et al. Expression specificity, distribution, and regulation of their expression. J of chemokines and gelatinase B in sympathetic ophthalmia. Exp Med. 1987;166:967–981. Eye. 2007;21:649–657. 41. Van Damme J, Opdenakker G, Simpson RJ, et al. Identification 57. Shah DN, Piacentini MA, Burnier MN, McLean IW, Nussenblatt of the human 26-kD protein, interferon beta 2 (IFN-beta 2), as RB, Chan CC. Inflammatory cellular kinetics in sympathetic a B cell hybridoma/plasmacytoma growth factor induced by ophthalmia a study of 29 traumatized (exciting) eyes. Ocul interleukin 1 and tumor necrosis factor. J Exp Med. 1987;165: Immunol Inflamm. 1993;1:255–262. 914–919. 58.Auw-HaedrichC,LoefflerKU,WitschelH.Sympathetic 42. Uyttenhove C, Coulie PG, Van Snick J. T cell growth and ophthalmia: an immunohistochemistry study of four cases. differentiation induced by interleukin-HP1/IL-6, the murine Ger J Ophthalmol. 1996;5:98–103. hybridoma/plasmacytoma growth factor. J Exp Med. 1988; 59. Aziz HA, Flynn HW Jr, Young RC, Davis JL, Dubovy SR. 167:1417–1427. Sympathetic ophthalmia: clinicopathologic correlation in a 43. Romagnani S. Regulation of the T cell response. Clin Exp consecutive case series. Retina. 2015;35:1696–1703. Allergy. 2006;36:1357–1366. 60. Dunlop AA, Cree IA, Hague S, Luthert PJ, Lightman S. 44. Mesquida M, Molins B, Lloren¸c V, Sainz de la Maza M, Adán A. Multifocal choroiditis: clinicopathologic correlation. Arch Long-term effects of tocilizumab therapy for refractory uveitis- Ophthalmol. 1998;116:801–803. related macular edema. Ophthalmology. 2014;121:2380–2386. 61. Lim WK, Chee SP, Sng I, Nussenblatt RB, Chan CC. 45. Deroux A, Chiquet C, Bouillet L. Tocilizumab in severe and Immunopathology of progressive subretinal fibrosis: a variant refractory Beh¸cet’s disease: four cases and literature review. of sympathetic ophthalmia. Am J Ophthalmol. 2004;138:475– Semin Arthritis Rheum. 2016;45:733–737. 77. 46. Deuter CM, Zierhut M, Igney-Oertel A, et al. Tocilizumab in 62. Inomata H and Sakamoto T. Immunohistochemical studies of uveitic macular edema refractory to previous immunomodu- Vogt-Koyanagi-Harada disease with sunset sky fundus. Curr latory treatment. Ocul Immunol Inflamm. 2016;5:1–6. Eye Res. 1990;(suppl 9):35–40. 47. Shenderov K, Barber DL, Mayer-Barber KD, et al. Cord factor 63. Chan CC, Palestine AG, Kuwabara T, Nussenblatt RB. and peptidoglycan recapitulate the Th17-promoting adjuvant Immunopathologic study of Vogt-Koyanagi-Harada syndrome. activity of mycobacteria through Mincle/CARD9 signaling and Am J Ophthalmol. 1988;105:607–611. the inflammasome. J Immunol. 2013;190:5722–5730. 48. Lee EJ, Brown BR, Vance EE, et al. Mincle activation and the 64. Oligino TJ, Dalrymple SA. Targeting B cells for the treatment of Syk/Card9 signaling axis are central to the development of rheumatoid arthritis. Arthritis Res Ther. 2003;5(suppl 4):S7– autoimmune disease of the eye. J Immunol. 2016;196:3148– S11. 3158. 65. Cornish KS, Kuffova L, Forrester JV. Treatment of diffuse 49. Alaaeddine N, Hilal G, Baddoura R, Antoniou J, Di Battista JA. subretinal fibrosis uveitis with rituximab. Br J Ophthalmol. CCL20 stimulates proinflammatory mediator synthesis in 2015;99:153–154. human fibroblast-like synoviocytes through a MAP kinase- 66. Caso F, Rigante D, Vitale A, et al. Long-lasting uveitis remission dependent process with transcriptional and posttranscription- and hearing loss recovery after rituximab in Vogt-Koyanagi- al control. J Rheumatol. 2011;38:1858–1865. Harada disease. Clin Rheumatol. 2015;34:1817–1820.