Background
Uveitis is a common cause of human visual disability and blindness. Although the etiology remains unclear, it is generally believed that a T cell-mediated immune response underlies the pathogenesis, and this is supported by the observation that injection of autoreactive T cells into susceptible, syngeneic rodents induces experimental autoimmune uveitis (tEAU) [
1‐
4]. Moreover, studies in the rodent models of EAU induced by immunization with a well-characterized uveitogenic autoantigen, interphotoreceptor retinoid-binding protein (IRBP), have shown that activation of autoreactive T cells is a key pathogenic event in disease induction, progression, and recurrence [
2,
5‐
7]. While a great deal of information is available about the development and activation of autoimmune T cells in the periphery in EAU, it remains unclear how only a few infiltrating uveitogenic T cells can exert a pathogenic effect inside the eye, an “immune privileged” site, leading to tissue destruction [
8,
9].
Danger signals are initiated by pathogen-associated molecular patterns (PAMPs) or damage-associated molecular patterns (DAMPs) [
10,
11], the latter being actively secreted from inflammatory or stressed cells or passively released from damaged apoptotic or necrotic cells. Both PAMPs and DAMPs can be recognized by pattern-recognition receptors, including TLRs, NOD-like receptors, retinoic acid-inducible gene-like receptors, and receptors for advanced glycation end products [
10,
11]. MyD88 is an intracellular adaptor protein required for signaling by most TLRs [
12]. The importance of DAMPs in T cell-mediated uveitis was identified as a result of our observation that in the chronic uveitis mouse model induced by transfer of IRBP-specific T cells (tEAU), a model induced without the use of microbial products and resembling human chronic uveitis, mice lacking MyD88 (MyD88
−/−) are completely resistant to induction of tEAU [
13].
In a search for DAMPs that interact with TLRs/MyD88 in tEAU, we previously showed that HMGB1 is an early and critical mediator of IRBP-specific T cell-induced intraocular inflammation, since HMGB1 antagonists reduce ocular inflammation by suppression of uveitogenic T cell functions, such as IRBP-specific T cell proliferation and cytokine production [
13]. HMGB1, one of the most important DAMPs released by cells, binds to TLR2 and 4 then MyD88 is recruited by the TLR adaptor protein Mal [
14]. We also found that HMGB1 is actively secreted within the eye within 24 h after IRBP-specific T cell transfer as a consequence of direct cell-cell contact between infiltrating IRBP-specific T cells and viable retinal cells [
13]. These results led us to search for surface molecules expressed on IRBP-specific T cells and retinal cells that mediate the secretion of HMGB1.
Fas (CD95) is well studied as a death receptor that induces apoptosis through the traditional caspase pathway, presumably one of the mechanisms of immune privilege that protect the eye from severe intraocular inflammation [
15‐
17]. However, recent studies have shown that activation of the Fas/FasL system can also induce release of pro-inflammatory cytokines by macrophages that are independent of conventional caspase-mediated apoptotic signaling [
18‐
21]. Moreover, it was found that Fas activation induces rapid TLR4/IRAK4-dependent release of HMGB1, which contributes to Fas-mediated pro-inflammatory cytokine production by viable macrophages [
22]. In the present study, we explored whether Fas also mediated HMGB1 secretion by viable non-lymphoid retinal cells after interaction with activated IRBP-specific T cells. We examined HMGB1 release in the eye and ocular inflammation after injection of IRBP-specific T cells from immunized wild-type (Wt) B6 mice into Fas-deficient (Fas
lpr) mice and also determined whether early interruption of Fas signaling could suppress tEAU in Wt mice. Our results revealed the importance of Fas/FasL activation in the active secretion of HMGB1 from living retinal cells and its involvement in the early event of intraocular inflammation triggered by uveitogenic T cells.
Discussion
The Fas/FasL system has been mainly studied for its role in caspase-dependent, apoptotic programmed cell death [
30] to maintain tissue/cell homeostasis. Previous studies on the eye have shown that both FasL and Fas are expressed on ocular tissues and contribute to the immunologically privileged status of the eye by causing apoptosis of invading leukocytes and protects the eye from immune-mediated damage [
15‐
17,
26] or play a role in tissue damage by promoting apoptotic death of ocular cells.[
17,
31,
32]. However, the Fas/FasL system is being increasingly recognized for its ability to trigger inflammation [
29]. Numerous studies have demonstrated that activation of Fas signaling in a variety of non-lymphoid cells, including colonic and lung epithelial cells [
33,
34], hepatocytes [
35], synoviocytes [
36], macrophages [
20], and fibroblasts [
37], can lead to the expression and release of inflammatory factors in vitro and in vivo, in particular, at the early stage of inflammation development. Such factors may, in turn, recruit inflammatory cells, exacerbating the inflammatory process. Accordingly, Fas/FasL-mediated inflammation has been shown to play an important role in the pathogenesis of several diseases, including acute respiratory distress syndrome [
34], cystic fibrosis [
38], arthritis [
36,
39], and cancer [
40], all of which have an underlying inflammatory component. Moreover, many chronic inflammatory diseases are attenuated in mice lacking Fas or FasL (
gld) [
39,
41,
42]. For example, compared to Wt B6 mice,
gld and
lpr mice are highly resistant to the development of experimental autoimmune encephalomyelitis (EAE) [
43‐
45] and EAU [
46], which share essential cellular mechanisms, indicating involvement of Fas/FasL in the T cell-mediated tissue inflammation.
Our results clearly demonstrate that Fas is required for active and rapid release of HMGB1 from tissue retinal cells via cell-cell interaction with activated uveitogenic T cells. Released HMGB1 either alone or in combination with other pro-inflammatory mediators triggers inflammatory cascades in the eye, probably by enhancing and sustaining the pathogenicity of IRBP
1–20-specific T cells. The in vitro results that Fas on retinal cells mediates HMGB1 release was further supported by our in vivo studies using Fas knockout mice, in which transfer of IRBP
1–20-specific T cells failed to trigger HMGB1 release or induce tEAU, while local HMGB1 injection restored susceptibility to induction of ocular inflammation. These data are the first to implicate HMGB1 in Fas-mediated HMGB1 release from tissue cells in response to infiltrating autoreactive T cells and provide a possible explanation for the observation that Fas
lpr mice do not develop EAU after immunization with IRBP antigen or adoptive transfer of IRBP-specific T cells [
46] (Fig.
3). The wild expression of Fas on photoreceptor cells, retinal pigment epithelium [
26], microglia [
47], and astroglia [
48] of the retina correlated with the expression pattern of HMGB1 [
13]. Our study suggested that one of the mechanisms that a few T cells initiate autoimmune uveitis is that IRBP-specific T cells interact with parenchymal cells such as residential DCs [
49], microglia [
50], astrocytes [
51], and Müller cells [
52], resulting in the subsequent production of HMGB1 by those cells, mediated by Fas/FasL, an early event in the pathogenesis of intraocular inflammation. Most adoptively transferred disease-inducing T cells require “licensing for pathogenicity” in the lung and other organs in order to induce disease in target organs [
53,
54]—“a hub-and-spoke pattern” [
55]. Both processes of “licensing for pathogenicity” and of HMGB1 production could take place during the induction of effector phase of EAU. The release of intraocular HMGB1 on day 1 may attract the migration of “licensed” pathogenic T cells to the eye, and the later increase of HMGB1 could be related to the involvement of these T cells and the damage they invoke in the retina.
The form of HMGB1 should be assessed as this has a major effect on promotion of either pro-inflammatory cytokine production or chemo-attraction of inflammatory cells, although neutralization of HMGB1 inhibited tEAU. Our results complement our current understanding of the Fas/FasL system in the eye, i.e., that it is not involved only in induction of apoptosis of infiltrating leukocytes or tissue cells, but also in initiation of inflammation, in particular, during the early stage of disease development in the eye. Different initial molecular events might lead to the activation of different molecular mechanisms resulting in the transmission of either apoptotic or non-apoptotic signals. The molecular process by which Fas is switched from an inflammatory role to an apoptotic function is not known, and the paradoxical roles of the Fas/FasL interaction in stimulating apoptosis of invading leukocytes in the eye and in promoting HMGB1 release and inflammation in the eye need to be explored.
How the Fas signaling pathway triggers HMGB1 release is not known. Met 12, a 12-amino acid peptide containing the YLGA motif in the N-terminal region of the extracellular domain of the α chain of Met, a tyrosine kinase receptor for hepatocyte growth receptor, is a small molecular weight inhibitor of Fas [
23,
24] and acts as an inhibitor of the apoptosis-activating Fas/FasL pathway in a transformed photoreceptor cell line (661 W cells) and in an animal model of retinal detachment [
23]. We tested the inhibitory effect of Met 12 on Fas-mediated HMGB1 release, and the subsequent intraocular inflammation and the results, shown in Figs.
4 and
5, support our hypothesis that Fas on retinal cells mediates HMGB1 release, which can be blocked by Met 12. Our results showed that adoptive transfer of IRBP-specific T cells reduced EAU severity in the Fas
lpr mice compared with WT mice (Fig.
3), whereas blocking of Fas systemically using Met 12 did not have a similar effect on EAU disease compared with controls (Fig.
4c). It has been repeatedly observed that experimental results using KO mice are frequently different when compared to blocking agents (or antibodies). In Fas KO mice, there is both systemic and intraocular deletion of Fas so that infiltrating T cells cannot interact with Fas on parenchymal ocular cells, resulting in inhibition of EAU. In contrast, the dose of Met used systemically may be too low to block Fas expressed within the eye, and/or Met administered systemically may be degraded before entering the eye. Support for these explanations is provided by the observation that the intraocular injection of Met does block Fas/FasL interaction between retinal cells and infiltrating T cells.
Another new finding was that RIP2, a receptor-interacting serine/threonine kinase with a C-terminal caspase activation and recruitment domain (CARD) that plays a critical role in Fas-mediated apoptosis [
27], is also involved in Fas-mediated HMGB1 release (Fig.
6). Our finding that downstream molecules of the Fas signaling pathways, such as Met and Rip2, regulate both apoptosis and inflammation complicates the use of pharmaceutical agonists or antagonists for Fas-mediated pathological events. Further studies on the role of Fas/FasL in inflammation and apoptosis in immune responses should result in improved immunotherapies based in Fas/FasL and their signaling molecules.
Although the Fas activator Jo2 triggered release of both HMGB1 and IL-1 from retinal cells (Figs.
1 and
7), we have previously shown that HMGB1 antagonists reduce intraocular inflammation induced by injection of IRBP
1–20-specific T cells [
13], whereas, transfer of IRBP
1–20-specific T cells into IL-1RKO mice induced disease in all six mice (Fig.
7). Together, these results show that HMGB1, but not IL-1, is required for intraocular inflammation triggered by infiltrating effector autoreactive T cells. In contrast, IL-1RKO mice are resistant to active induction of EAU by immunization with IRBP
1–20 and complete Freund’s adjuvant (data not shown and [
56,
57]. These results suggest that the IL-1R is required for the generation of pathogenic T cells, in particular Th17 cells [
58,
59] but is not needed for the subsequent pathogenic events occurred in the eye, i.e., in the effector phase of EAU. Similarly, in an EAE model mice used to study multiple sclerosis in humans, the IL-1R was found to be required for induction of EAE by active immunization with the antigen MOG but not for induction of EAE induced by transfer of MOG-specific T cells [
60]. Together, these results suggest that, once disease induced by pathogenic Th17 cells has been established, reducing IL-1 levels may not be an effective means of treatment, whereas blockade of HMGB1 and its related signaling molecules might achieve the therapeutic goal.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
GJ performed experiments and generated and analyzed the data. YW, JY, AH, and YZ assisted with the experiments and the data analysis. DS and HK helped with the designing of the experiments and edited the manuscript. And HS directed the study, planned experiments, interpreted the data, and wrote the manuscript. All authors read and approved the final manuscript.