Our prior data demonstrated an important protective role for IFNγ in uveitis in a murine model of progressive spondyloarthritis. Here, we tested the extent to which IL-17 mediates uveitis occurring in the absence of IFNγ. Antigen-specific T cell production of IL-17A was exacerbated in vitro as a consequence of IFNγ deficiency. Assessment of leukocyte trafficking within the iris microvasculature and tissue by intravital videomicroscopy in mice treated with anti-IL-17 neutralizating antibody revealed impaired intravascular responses, such as rolling and adherence that coincided with fewer infiltrating cells in the iris tissue. The histopathology demonstrated that IL-17 blockade significantly reduced the number of infiltrating leukocytes within the aqueous humor and vitreous body. The peripheral arthritis that accompanies PG-induced uveitis was also diminished in mice that received the anti-IL-17 blocking antibody. However, despite protective effects of IL-17 inhibition on PG-induced uveitis, unexpected retinal toxicity was observed. Collectively, our data support an important role for IL-17 in the pathogenesis of PG-induced uveitis, which is negatively regulated by IFNγ.
Uveitis encompasses a heterogeneous group of diseases, and even though it can arise from microbial or viral infection, it is frequently diagnosed in inflammatory diseases involving the joints, such as ankylosing spondylitis, juvenile idiopathic arthritis, Behçet's disease and many others. The underlying mechanisms that result in the co-existence of uveitis and arthritis/spondylitis are poorly understood. An important association exists between HLA-B27 and anterior uveitis in patients with ankylosing spondylitis. Approximately 90% of affected individuals have the inherited B27 allele [27
], but HLA-B27 does not account for all of the genetic risk for disease, nor the extent to which it influences the onset and severity of uveitis versus spondyloarthritis. For example, HLA-B27 transgenic rats spontaneously develop spondyloarthropathy but do not develop uveitis [28
]. This underscores the complexity of multisystem, immune-mediated polygenic diseases, in which environmental factors and/or endogenous microflora can contribute to the pathogenesis of disease within the different target organs.
Although it is not known whether the pathogenic mechanisms in uveitis are the same as arthritis and/or spondylitis, we initially hypothesized that the underlying basic mechanisms would be shared between these two organs in mice immunized with PG. Somewhat paradoxically, however, we found that IFNγ deficiency reduced disease in the joint and spine but worsened disease in the eye [4
]. The discordant effects of IFNγ underscore the complexity of how immune-mediated diseases target multiple organs differentially. This phenomenon has presented considerable challenges in the treatment of uveitis in patients with multisystem diseases, such as Behçet's disease or AS. For example, the TNFα inhibitor, etanercept, diminishes arthritic symptoms in a subgroup of patients with AS [30
], but it does not always have comparable beneficial effects for uveitis in the same patients [32
]; or it can even lead to increased severity of uveitis [33
]. Understanding the unique aspects of individually affected tissues is clearly pertinent to the design of optimal therapies.
The data presented here suggest that IFNγ is protective in the eye by virtue of being counter-regulatory to the IL-17 response. Studies with various KO mice or blocking antibodies have dissected the separate roles of Th1 and Th17 effector responses in the peripheral arthritis aspect of the PG-induced disease model [11
], wherein IFNγ serves an an important regulatory role in arthritis [26
]. These studies indicate a dominant pathogenic role for IFNγ and the Th1 effector response over that of the Th17 response in arthritis. We recently demonstrated a similar pathogenic role for IFNγ in peripheral arthritis and extended its contribution to the axial disease that develops in TCR-Tg mice [4
]. Our findings would indicated that although the Th17 response is not considered critical in the development of PG-induced joint disease, it does appear to be a prevailing response contributing to uveitis.
Th17 cells have been reported to play a role in induction of the animal model of posterior uveitis, experimental autoimmune uveitis (EAU) [34
] and have been implicated in uveitis clinically [35
]. IL-17 is thought to contribute to disease through its ability to promote recruitment and activation of neutrophils, which is consistent with our previously reported increase in granulocyte numbers in PG-immunized mice lacking IFNγ [4
]. The production of IL-17 from T cells is well-described, yet additional cellular sources including γδ T cells, NKT cells, macrophages, mast cells and neutrophils may be just as important as the T cells [37
]. Indeed, a recent report supports the role of innate immune cells rather than T cells in expression of IL-17 in the joints of patients with spondyloarthritis [41
]. Thus, the pathology of uveitis observed in response to PG-immunization could involve coordinated events between the T cells of the Th1 and Th17 lineages and innate immune cells, which then propagate tissue damage within uveitic eyes. Future studies that compare the cytokine sources along with their targets within uveitic eyes versus the arthritic joints will be informative.
The IL-23/IL-17 axis is a major focus of spondyloarthritis research, and many pharmaceutical companies are actively working on strategies to block IL-17 in uveitis, AS and related spondyloarthropathies [14
]. Therefore, it is critical to understand how this cytokine can contribute to immune-mediated uveitis models in mice, and possibly in humans. Although reduced clinical signs of arthritis and ocular disease following anti-IL-17 treatment suggest great therapeutic potential for the use of this antibody, our observations of retinal toxicity indicate the need to proceed cautiously. We were unable to measure visual acuity, however the observed depletion of photoreceptors and ONL would be expected to result in near to complete vision loss. This observation brings to question why inhibition of IL-17 may result in toxicity within the eye. We are not aware of any reports that demonstrate a role or involvement of IL-17A in the survival of retinal cells. Nonetheless, a number of studies implicate IL-17A involvement in apoptotic or cell survival responses in different cell types and diseases [43
], suggesting its potential role in cellular viability within the eye. Retinal toxicity derived from anti-IL-17 antibody treatments has not been previously reported in other murine uveitis models [34
], indicating that the toxicity could possibly involve a combination of events within this particular disease model, such as glial cell activation within the retina, the presence of PG-antigen within the retina, and/or ongoing systemic disease [4
]. At the time examined here (three weeks post PG-immunization), the photoreceptor toxicity did not coincide with increased expression of GFAP (glial fibrillary acidic protein, a marker of glial activation, data not shown). However, we cannot rule out the contribution of glial activation at times preceding photoreceptor toxicity. Studies are underway to elucidate whether inhibiting the Th17 response through targeting IL-17R or the upstream cytokine IL-23 reduces uveitis without photoreceptor damage. These studies would inform us to the potential toxicity effects of this particular antibody versus the contribution of the Th17 pathway in photoreceptor viability. The observations presented here yield valuable information regarding contrasting mechanisms of disease within different portions of the body and could have an impact on how we approach therapy in complex, inflammatory diseases.