When a patient presents with uveitis, the physician has no way of knowing when and where the putative exposure to self antigen or a mimic antigen has occurred (with the exception of sympathetic ophthalmia, a uveitic entity that follows a physical trauma to an eye and release of retinal antigens to the draining lymph node).
3 However, we believe that we know the conditions under which exposure to retinal Ag or its mimic will result in disease. Immune stimuli from microorganisms and/or from damaged tissue signal “danger” to the immune system by activating pattern-recognition receptors on leukocytes and elicit an inflammatory environment. Antigen-specific immune responses induced in such an environment develop along pathways designed to destroy invading microbes. They also tend to be harmful to tissue and can result in induction of autoimmunity. In experimental animals, we purposely create such “danger” conditions by immunizing with retinal antigen emulsified in CFA, which contains heat-killed tuberculosis bacteria, so that it drives differentiation of retina-specific T cells to a proinflammatory phenotype.
Activation of retina-specific T cells occurs in the periphery, away from the eye, and they must reach and enter the eye for disease to be induced. This process is frequently referred to as “homing,” implying specific attraction to the target organ. Our data indicate that this term is a misnomer and that entry of the first activated T cells into the eye occurs at random. In experiments in which activated and fluorescently labeled retina-specific or nonspecific T cells were infused into recipient animals, a small but equal number of cells entered the retina within hours of infusion.
27,28 While only the specific cells resulted in EAU development several days later, indicating that induction of disease requires antigen recognition within the eye, this experiment shows that the first antigen-specific T cells entering the eye have no advantage over nonspecific T cells.
Since the healthy eye is protected by a blood–retinal barrier, which prevents free trafficking of cells and even of larger molecules, how are these first T cells able to enter the eye? It is important to keep in mind that activated T cells are sticky and invasive, somewhat resembling metastatic tumor cells. They express adhesion molecules, produce matrix-degrading enzymes, adhere to blood vessels and extravasate into tissues. Therefore, the blood–retinal barrier, which can stop quiescent naïve T cells from entering the eye, is ineffective against activated T cells. It is also important to emphasize that the retina-specific T cells that drive disease are by themselves not sufficient to inflict tissue damage. Recognition of their specific antigen within the eye maintains the activated state of the retina-specific T cells. The products that they secrete, including cytokines and chemokines, cause inflammatory changes in the retina, stimulate nearby ocular resident cells, and activate the retinal microvasculature. This attracts leukocytes from the circulation, which enter the eye in increasing numbers resulting in an inflammatory cascade that builds on itself. In this sense, the antigen-specific T cells are the orchestrators of the inflammatory process but it is the infiltrating blood-borne leukocytes/monocytes, granulocytes, natural killer (NK) cells, and NK T cells (NKT), as well as γδ T cells, that amplify the inflammation and that mediate the final tissue damage.
Indeed, treatments that inhibit recruitment of blood-borne leukocytes into the eye (e.g., blockade of adhesion molecules, depletion of particular leukocyte types), also inhibit EAU. However, while treatments targeting inflammatory processes in general may be effective in ameliorating disease, they do not cure the underlying problem of breakdown in tolerance and presence of autopathogenic T cells, which will cause the disease to rebound as soon as treatment is discontinued. It is therefore of utmost importance to understand the nature of the eliciting retina-specific T cells to develop therapeutic regimens that target them effectively.
Current immunologic knowledge indicates that Ag-specific effector T lymphocytes fall into several major lineages that differ in phenotype and function, known as Th1, Th2, and Th17. They arise from a common precursor after different conditions of stimulation, and each produces a distinct profile of cytokines and chemokines, which is in line with their biological roles in dealing with different types of infections (
Fig. 3). The natural role of Th2 cells (IL-4–, IL-5–, and IL-13–producing) is to deal with worm infections, Th1 (IFN-γ–producing) control intracellular microorganisms such as tuberculosis and Th17 (IL-17–producing) are needed for defense against extracellular microorganisms, such as extracellular bacteria and fungi.
29,30 Importantly, each one of these cell types can also be involved in tissue pathology. Th2 cells are involved in allergy and asthma, whereas Th1 and Th17 cells participate in various inflammatory and autoimmune diseases such as uveitis, multiple sclerosis, rheumatoid arthritis, and psoriasis
29,30 (
Fig. 3).
Autoimmune uveitis in humans is a heterogeneous group of diseases that can differ in their clinical presentation and course, even though patients may respond to the same retinal antigen(s).
3 Associations have been described with Th1 or with Th17 cytokines (reviewed in Refs.
29,
31) and may underlie heterogeneity. In human disease, the causal relationships are hard or impossible to prove, but results obtained in animal models support the notion that diverse cytokine response profiles may lead to ocular disease. This is exemplified by two distinct clinical and immunologic forms of murine EAU: the classic EAU model induced by immunization with IRBP (or its peptide, IRBP
161–180) emulsified in CFA and a recently developed model induced by injection of IRBP
161–180 pulsed, in vitro-matured dendritic cells (DCs).
19,32 Both forms of exposure to IRBP result in uveitis and retinal disease, but the two EAU models differ in fundus appearance, in clinical course and duration, in the nature of the inflammatory infiltrate recruited into the eye, which is monocytic in the classic CFA-EAU and granulocytic in DC-EAU, and most important, in their effector cytokine dependence: Th17 in CFA- and Th1 in DC-EAU. This last conclusion stems from experiments showing that treatment with neutralizing antibodies to IL-17, but not to IFN-γ, prevents and reverses CFA/EAU. Conversely, uveitogenic DC injected into IFN-γ
−/− mice, which are unable to generate a fully functional Th1 effector response, fail to induce disease.
19,32
These experiments demonstrate that uveitis can be either Th17-driven or Th1-driven. The type of response that will predominate appears to be determined by conditions during first exposure to Ag, more specifically, by the quantity and/or quality of Toll-like receptor and other innate receptor signals as well as by the type and diversity of cells involved in antigen presentation (
Fig. 4). Notably, Th17 becomes dominant when antigen is recognized in the context of CFA. We propose that the mycobacteria stimulate multiple microbial pattern recognition receptors on diverse cells residing in the lymph node draining the site of immunization. These are cells that participate in or influence the antigen presentation process and include DCs, monocytes, γδ T cells, and possibly others. This intense and multifaceted stimulation leads to a Th17-dominant immune response. In contrast, Th1 becomes dominant when Ag is presented by in vitro-matured, antigen-pulsed DCs. Experimental evidence shows that in that situation the DCs themselves must migrate into the regional lymph nodes and present the antigen to the T cells (i.e., antigen introduced on the DCs is not transferred to host cells for cross-presentation).
19 The DCs are matured before transfer, only with soluble endotoxin and anti-CD40 antibodies, resulting in a much more focused stimulus than CFA, leading to a Th1-dominant response. In human uveitis, these initial eliciting events are mostly unknown, as they occur long before the patient presents with disease. However, if our findings in animals are relevant to the human situation, these events could be critical to the course of the subsequent disease and would have direct implications for therapy. Thus, the Th1 pathway may be an appropriate target in some types of uveitis, whereas in other uveitic diseases, the focus should perhaps be on Th17.