Uveitis is an ophthalmologic entity comprising several heterogeneous diseases, all characterized by intraocular inflammation starting initially in the uvea [¹,²]. The different immune pathways involved in this process have not yet been thoroughly described. The participation of T-helper lymphocytes seems to predominate in uveitis as shown by analyses of intraocular fluids of patients with uveitis [³,⁴]. T-helper (Th) lymphocytes are classically composed of two subsets, Th1 and Th2, differentiated by the type of cytokines they secrete [⁵,⁶]. A new lineage of T cells, called Th17, which produces the cytokine, interleukin-17 (IL-17), has recently been described [⁷]. Th17 might be involved in the pathophysiology of experimental autoimmune uveitis (EAU), and Th1 could play a regulation role toward Th17 through its production of IL-27, which inhibits the secretion of IL-17 [⁸,⁹]. Th17 may play a central role in uveitis, but its function and interaction with Th1 remain to be clarified [¹⁰,¹¹]. The Th1 response classically plays a central role in the immunopathological process of experimental autoimmune uveitis [²,¹²]. Analyses of intraocular fluids in human uveitis tend to favor Th1 involvement because they showed Th1 cytokines such as interferon-gamma. However, to date, the analyses have been unable to define a clear predominance of Th1 or Th2 responses [^2-4]. Ocular inflammation primarily involves the uveal tract but can also extend to other ocular structures such as the retina or vitreous. The involvement of the conjunctiva in uveitis has not been studied as much. Little is known about inflammation occurring on the ocular surface of uveitic patients [¹³,¹⁴], but conjunctival injection is frequently observed in anterior uveitis [¹].

Our laboratory developed new, noninvasive, and objective techniques to explore immunoinflammatory markers expressed on the ocular surface by conjunctival epithelial cells, using flow cytometry in impression cytology specimens [¹⁵] in a variety of ocular surface diseases [¹⁶]. Recently, our group has investigated the Th1 and Th2 inflammatory cascades by assessing CC chemokine receptor 4 (CCR4) and CCR5 expression in conjunctival cells in uveitis [¹⁷]. Indeed, CCR4 and CCR5 are known to be chemokine receptors related to the Th2 and Th1 systems, respectively. Their principal ligands are CC chemokine ligand 3(CCL3)/macrophage inflammatory protein 1? (MIP-1?), CCL4/MIP-1?, and CCL5/regulated on activation normal T cell expressed and secreted (RANTES), for CCR4, and CCL17/thymus and activation-regulated chemokine (TARC) and CCL22/macrophage-derived chemokine (MDC), for CCR5.MIP-1? is a chemokine that binds receptors of Th1 cell lines, TARC and MDC are related to Th2 and MIP-1?, and RANTES binds receptors of both Th1 and Th2 cell lines [¹⁸,¹⁹]. CCR4 was overexpressed in the epithelial conjunctival cells of uveitis, and its overexpression suggested involvement of the Th2 system on the ocular surface while participation of Th1 is usually considered to be favored in intraocular pathogenesis, at least in experimental autoimmune uveitis [⁶]. This previous study suggested that there may be striking differences in inflammatory pathways between intraocular tissues and the ocular surface. Therefore, each may react differently to immunogenic inflammation. The aim of the current study was to determine whether the ocular surface mimics or counteracts intraocular inflammatory pathways by directly comparing Th1 and Th2 markers between the conjunctiva (ocular surface component) and the ciliary body (representing intraocular structures) in an experimental model of uveitis, endotoxin-induced uveitis (EIU). This model of acute ocular inflammation has been used for some forms of human anterior uveitis [²⁰] and the inflammatory cell infiltrate of the anterior segment is maximal at 24 h after lipopolysaccharide (LPS) injection [²¹]. In this study, these different inflammatory markers were used to asses Th1 or Th2 activation: chemokine receptor CCR4 and its ligand MDC, for Th2 participation; but also chemokine receptor, CCR5, for Th1 system; and at last its ligand, RANTES, for both the Th1 and Th2 systems. Immunohistochemistry and real-time polymerase chain reaction (PCR) were used to compare the protein and RNA expression of these markers in the conjunctiva and ciliary bodies in EIU 6 h and 24 h after LPS injection.

Expression of the inflammatory markers, RANTES, MDC, CCR5, and CCR4, was analyzed in the conjunctiva and ciliary bodies using two different techniques, immunostaining and molecular biology. A positive staining in immunohistochemistry was observed with all the markers studied in the conjunctiva and ciliary body of EIU rats, but this technique provided more qualitative than quantitative analyses of chemokines and chemokine receptors. To compare the level of expression of these markers, we combined the molecular biology technique with semi-quantitative real-time PCR since it is a more precise tool to quantify gene expression. In our study, semi-quantitative PCR confirmed the immunostaining results. Indeed, the protein and RNA expressions of RANTES, MDC, and CCR5 were higher in the conjunctiva and ciliary body of EIU rats (at 6 h and 24 h) than of the control rats on. In immunohistochemistry, RANTES, MDC, and CCR5 expressions peaked 24 h after the LPS injection while in PCR, their RNA expression peaked at 6 h. The time difference between the peak of protein and RNA expression could be explained by the time needed to synthesize protein from RNA. Both methods found that there were no differences in the CCR4 level between EIU rats and controls, which is quite different from our previous report in humans [¹⁷] where CCR4 was overexpressed by conjunctival epithelial cells. This difference is probably explained by the fact that EIU is a quickly systemic induced, short-term uveitis model that does not fully reproduce the complexity of human uveitis.. In this model, it is possible that in addition to the local ocular inflammation, there is an upregulation of epithelial cells by LPS systemic stimulation, which could therefore corroborate to our results published in patients with uveitis. Nevertheless, our goal was to compare surface and uvea immunopathological changes, and this model showed that expression of RANTES, MDC, CCR4, and CCR5 in EIU was quite similar in the conjunctiva and ciliary body, indicating that conjunctival inflammation might reproduce the intraocular inflammation generated in this uveitis model probably by local extension and diffusion. This study showed an overexpression of RANTES, MDC, and CCR5 in EIU animals compared to controls and confirmed previous publications in which RANTES was found by RT?PCR in the ciliary body in EIU animals [²²,²³]. In the literature, high levels of interleukin-1 (IL-1), IL-6, IL-10, tumor necrosis factor (TNF)-?, RANTES, interferon (IFN)-?, monocyte chemotactic protein 1 (MCP-1), and macrophage inflammatory protein 2 (MIP-2) were found in EIU animals [^21-26]. Therefore, Th1 and Th2 cytokine profiles in EIU have not been well defined, probably because inflammatory pathways in EIU mostly involve neutrophils and macrophages [^27-29]. Nevertheless, lymphocytes also infiltrate the iris and ciliary body and seem to have a fundamental role in the pathogenesis of EIU [²⁷,³⁰,³¹], so discrimination of Th1/Th2 systems should be helpful in understanding lymphocyte functions in EIU. In our study, Th1 activation seemed to predominate in this model with high levels of CCR5 overexpression and no increased expression of CCR4, but Th2 participation with MDC was also noted. The Th1 response is usually thought to play a central role in other animal model studies of uveitis. Indeed, in experimental autoimmune uveitis (EAU), which is a Th1 inflammation induced by a retinal antigen, Th1 participation is recognized [²] with the overexpression of chemokines related to the Th1 system such as MIP-1 ? and RANTES as well as the chemokine receptor, CCR5 [³²]. In the current study, the analysis of chemokines and chemokine receptors in EIU partially confirmed the Th1 activation in the immunopathological process of EAU. There was no difference for the expression of the Th2-specific chemokine receptor, CCR4, between the EIU and control animals in our study, which is similar to findings for EAU where low levels of CCR4 were found [³³].

In conclusion, chemokine and chemokine receptor analysis showed that inflammation on the ocular surface reproduced intraocular phenomena by extension in this experimental model of uveitis, EIU. Further studies on human uveitis and other experimental models of uveitis could compare the chemokine profile in intraocular fluid and ocular surface samples. If the ocular surface mimics intraocular inflammatory pathways, the conjunctiva could provide a new and easier access for uveitis studies with less invasive exams than anterior chamber paracentesis or vitrectomy samples. Chorioretinal biopsies, like conjunctival impressions, could be developed to investigate inflammatory activity and pathophysiological mechanisms in uveitis.

This study was supported by unrestricted grants from Quinze-Vingts National Ophthalmology Hospital (Paris, France) and INSERM UMR S 872, Cordeliers Biomedical Institute, Pierre et Marie Curie University ? Paris 6, Paris Descartes University (Paris, France).