Autoimmunity in Dry Eye Is Due to Resistance of Th17 to Treg Suppression¹

Dry eye disease (DED)⁴ is a complex, multifactorial autoimmune disorder with a largely unknown pathogenesis. The disease is characterized by an immune and inflammatory process that affects the ocular surface and, in severe cases, may lead to blindness due to corneal ulceration or scarring (1, 2, 3). It has been estimated that >9 million Americans suffer from DED, and tens of millions more have less severe symptoms that are notable during exposure to adverse predisposing factors such as dry air (4, 5, 6).

The ocular surface inflammation in DED is sustained by ongoing activation and infiltration of pathogenic immune cells, primarily of the CD4⁺ T cell compartment (1, 2, 3). The contribution of T cells to the immunopathogenesis of DED is supported by many observations, including: 1) increased T cell infiltration of the conjunctiva in both clinical and experimental dry eye (7, 8); 2) presence of IFN-γ⁺ T cells in the conjunctiva of dry eye mice; and 3) induction of disease in nude mice by transfer of CD4⁺ T cells from dry eye mice (9). In addition, topical application of T cell immunosuppressants, such as cyclosporine (10) or agents inhibiting the pathogenic T cell recruitment to the ocular surface such as α₄β₁ integrin antagonist (11), lead to disease remission.

Despite this strong evidence for T cell involvement in the pathogenesis of dry eye, the majority of studies is focused primarily on the effector T cell responses. The concept that inefficient functioning of regulatory T cells (Tregs) may cause unrestrained generation of pathogenic T cells in DED has not yet been systematically investigated. A functional balance between Tregs and effector T cells is important for maintaining efficient immune responses while avoiding autoimmunity. Foxp3 expressing CD4⁺CD25⁺ Tregs are a naturally occurring suppressor T cell population that inhibits the activity of self-reactive Th1 (IFN-γ⁺), Th2 (IL-4⁺), and Th17 (IL-17⁺, also known as IL-17A) cells that potentially induce inflammation and tissue destruction (12, 13, 14, 15). In this study, we explored a well-characterized mouse model of DED (16) and tested the hypothesis that defective Treg homeostasis leads to a dysregulated activation of self-reactive T cells. Our data demonstrate the presence of dysfunctional Tregs and the resistance of pathogenic T cells, particularly Th17 cells, to Treg suppression in DED. In addition, we investigated whether in vivo blockade of IL-17 would ameliorate the disease and prevent both Th17 cell differentiation and loss of Treg function.

Six- to 8-wk-old female C57BL/6 mice (Taconic Farms and Charles River Laboratories) were used in these experiments. The protocol was approved by the Institutional Animal Care and Use Committee, and all animals were treated according to the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research.

DED was induced by exposure of mice to the controlled-environment chamber (CEC) for 9 days during which mice also received administration of scopolamine to maximize ocular dryness as described previously (16, 17). In brief, 0.5 mg/0.2 ml scopolamine hydrobromide (Sigma-Aldrich) was injected s.c. in the dorsal skin of mice four times daily. The mice placed in the CEC were continuously exposed to a relative humidity <30%, airflow of 15 L/min, and a constant temperature (21–23°C). Age- and sex-matched mice raised in the standard environment of the vivarium were used as normal controls. Induction of dry eye was confirmed by measuring changes in the corneal integrity using corneal fluorescein staining. One microliter of 5% fluorescein was applied to the lateral conjunctival sac of the mice, and eyes were examined for fluorescein staining with a slit lamp biomicroscope under cobalt blue light. Punctate staining was recorded in a masked fashion using the standard National Eye Institute grading system, giving a score from 0 to 3 for each of the five areas of the cornea (18).

Draining submandibular and cervical lymph nodes (LN) were harvested from normal and DED mice, and single-cell suspensions were prepared. Total CD4⁺ T cells and CD4⁺CD25⁺ Tregs were isolated by magnetic separation using CD4 T cell and Treg isolation kits (Miltenyi Biotec). Purity of sorted cells was >95%, and 97% of CD4⁺CD25⁺ Tregs were Foxp3⁺ in all the groups as confirmed by FACS analysis.

For Treg frequencies, LN cells were stained with the anti-CD4 FITC, anti-CD25 PE, and anti-FoxP3 PECy5. For intracellular cytokine detection, cells were first stimulated with PMA plus ionomycin (Sigma-Aldrich) for 6 h in the presence of GolgiStop (BD Biosciences) and then stained with anti-IL-17A PE or anti-IFN-γ PE and anti-CD4 FITC. All the Abs with their matched isotype controls and fix-perm buffer were purchased from eBioscience. Stained cells were analyzed on an EPICS XL flow cytometer (Beckman Coulter).

For all the suppression assays including intracellular cytokine analysis of proliferating cells, Tregs and responder CD4⁺ T (Tresp) cells were isolated from the draining LN. Purified total CD4⁺ Tresp cells (1 × 10⁵) were cocultured with CD4⁺CD25⁺ Tregs (5 × 10⁴), T cell-depleted syngeneic splenocytes (1 × 10⁵), and 1 μg/ml anti-CD3 Ab for 3 days. Proliferation of CD3-stimulated Tresp cells without adding Tregs was considered control proliferation with 0% suppression. Proliferation was measured using the BrdU proliferation kit (Millipore), and percentage of suppression was calculated using the following formula: %suppression = [(Tresp proliferation without Tregs − Tresp proliferation with Tregs)/Tresp proliferation without Tregs] × 100.

Draining LN, cornea, and conjunctiva of normal and dry eye mice were harvested. RNA was isolated with RNeasy Micro kit (Qiagen) and reverse transcribed using Superscript III kit (Invitrogen). Real-time PCR was performed using TaqMan Universal PCR Mastermix and preformulated primers for FoxP3 (assay ID: Mm00475156_m1), IL-17A (Mm00439619_m1), IL-6 (Mm00446190_m1), IL-17 receptor A (IL-17RA; Mm00434214_m1), and GAPDH (Mm99999915_g1; Applied Biosystems). The results were analyzed by the comparative threshold cycle method and normalized by GAPDH as an internal control.

Acetone-fixed cryosections of whole eyeballs were blocked with 2% BSA and anti-FcR Ab (eBioscience). The sections were immunostained with primary rabbit anti-mouse IL-17RA or isotype-matched control Abs (Santa Cruz Biotechnology overnight at 4°C and then with rhodamine-conjugated secondary Ab (The Jackson Laboratory) for 1 h at room temperature. The stained cross-sections were mounted using Vector Shield 4′,6-diamidino-2-phenylindole mounting medium (Vector Laboratories) and examined under the epifluorescence microscope (Nikon).

Lyophilized rat anti-mouse IL-17 and isotype mAbs (R&D Systems) were resuspended in PBS to yield a concentration of 100 μg/100 μl. Anti-IL-17 or isotype Ab was administered to mice by i.p. injections at 1 day before (200 μg/mouse) placing them into the CEC and then at days 1, 3, and 5 (100 μg/mouse). Corneal fluorescein staining was performed daily to score the disease. At day 6, mice were euthanized, and the draining LN and conjunctiva were harvested for further analyses.

Student’s t test was used for comparison of mean between the groups, and p < 0.05 was considered significant.

Regional draining LN were collected from DED and normal mice. Single-cell suspensions prepared from the LN were subjected to triple-color FACS analysis for CD4⁺CD25⁺Foxp3⁺ Tregs and to immunomagnetic sorting for isolation of Tregs. We did not observe significant differences in the frequencies of Tregs either as a proportion of total CD4⁺ cells (range, 9–11%) or of the total LN cells (range, 3–4%) between normal and DED mice (Fig. 1, a and b). Using an in vitro coculture suppression assay system, we next compared the regulatory potential of purified Tregs to suppress the proliferation of CD3 Ab-stimulated naive T cells (isolated from the LN of normal mice) and primed T cells (isolated from the LN of DED mice; Fig. 1 c). Tregs isolated from DED mice showed significantly reduced potential in suppressing proliferation of both naive T cells (p = 0.015) and primed T cells (p = 0.008) compared with those from normal mice. In addition, Tregs of DED mice were significantly less effective in suppressing the proliferation of primed T cells than those of naive T cells (p = 0.041). However, Tregs of normal mice were found to be equally effective in suppressing the proliferation of both the naive and primed T cells. These results clearly suggest a defect in Treg function, not in their numbers, in mice with dry eye disease.

To determine the subsets of primed T cells, which are resistant to Treg suppression, we analyzed intracellular cytokine by FACS in proliferating subsets of primed T cells (isolated from the LN of DED mice) cocultured with Tregs of normal and DED mice (Fig. 2). The main proliferating subset of primed T cells in the presence of Tregs of DED mice was found to be IL-17-producing CD4⁺ T cells, which proliferate ∼2.5 times more avidly compared with their proliferation in the presence of Tregs of normal mice. A modest increase in IFN-γ-producing CD4⁺ T cell proliferation was also observed in the presence of Tregs of DED mice than those of normal mice. However, no significant difference was observed in the percentages of any of the subsets in proliferating naive T cells (isolated from the LN of normal mice) in the presence of Tregs from either normal or DED mice (data not shown).

To further confirm the in vivo presence of Th17 cells in DED mice, we examined both the LN and conjunctiva for IL-17 and its receptor (IL-17RA) along with IL-6 and Foxp3 expression using FACS and real-time PCR. The LN of DED mice showed a significant increase (4- to 5-fold) in the frequency of IL-17⁺CD4⁺ T cells compared with those of normal mice (p = 0.026; Fig. 3,a). Similarly, real-time PCR analysis confirmed an increase in the expression levels of IL-17 (2- to 3-fold) and IL-6 (1.2- to 1.5-fold) mRNA in the LN of DED mice compared with normal mice. However, a minor decrease in the Foxp3 mRNA levels was observed in the LN of DED mice compared with those of normal mice (Fig. 3,b). As observed in the LN, a similar trend was found in the mRNA expression levels of IL-17, IL-6, and Foxp3 in the conjunctiva of DED mice (Fig. 3,c). In addition, we investigated the presence of IL-17Rs on the ocular surface by immunostaining and real-time PCR. Immunostaining performed on cross-sections of eye using anti-IL-17RA Ab showed that corneal and conjunctival epithelia of both normal and DED mice express IL-17RA (Fig. 3,d). Real-time PCR analysis for IL-17RA mRNA expression in cornea and conjunctiva showed no significant difference in the expression levels of IL-17Rs in both the groups (Fig. 3 e).

We next studied the effects of in vivo neutralization of IL-17 by anti-IL-17 Ab on the clinical signs of dry eye disease. A schematic diagram of the experimental design is shown in Fig. 4,a. Corneal fluorescein staining scores, a readout for the principal clinical sign of dry eye inflammation, was significantly lower in anti-IL-17 Ab-treated mice compared with isotype Ab-treated and untreated mice during both the induction (day 2) and the progression phase of the disease (Fig. 4 b).

In addition to the clinical signs of DED, we analyzed the effects of IL-17 blockade on Th17 cell frequency and Treg function. The LN of anti-IL-17 Ab-treated mice showed a >3-fold decrease in the frequency of IL-17⁺CD4⁺ T cells compared with that of isotype Ab-treated mice (Fig. 5,a). Similarly, an ∼2-fold decrease in the expression levels of IL-17 mRNA in the conjunctiva of anti-IL-17 Ab-treated mice was observed compared with those of isotype Ab-treated mice (Fig. 5,b). The in vitro Treg suppression assay using CD3-stimulated, primed T cells (isolated from the LN of dry eye mice) and Tregs (isolated from the LN of mice treated with anti-IL-17 or isotype Abs) showed a significant recovery in the suppressor potential of Tregs only in mice treated with anti-IL-17 Ab compared with those isolated from the isotype Ab-treated groups (p = 0.029; Fig. 5 c). These results clearly suggest an inverse relationship between Th17 cells and Treg function in DED.

The demonstration of impaired Treg function and pathogenic Th17 cells in DED is a novel finding that extends our understanding of immunopathogenesis of the disease. Specifically, we show herein that CD4⁺CD25⁺Foxp3⁺ Tregs are functionally incompetent in down-regulating the activation of pathogenic T cells in DED, and in vivo blockade of IL-17 significantly reduces the severity and progression of disease as well as markedly prevents the loss of Treg function.

Tregs play a central role in maintaining peripheral tolerance and controlling organ-specific autoimmunity by suppressing pathogenic autoreactive T cells (12, 13). A defective Treg homeostasis has previously been associated with the pathogenesis of several autoimmune diseases such as experimental autoimmune encephalomyelitis, multiple sclerosis, and rheumatoid arthritis (19, 20, 21). However, data on the role of Tregs in DED are very limited, and the homeostasis of Tregs in disease condition has not been investigated. There is only one report in which authors showed that adoptive transfer of CD4⁺ T cells from DED mice induces more severe disease in Treg-deficient mice compared with normal mice (9). Thus, until now, it has remained unclear whether the exacerbation of the disease is merely due to the absence of normal Tregs or a functional Treg defect. In the present study, we have analyzed Tregs both quantitatively and qualitatively. Our data demonstrate that Treg frequencies remain unchanged, but their potential to suppress the activation of effector T cells is markedly decreased in DED.

Previously, involvement of Th1 cells has been demonstrated in the pathogenesis of DED (9). In the current study, we show a second major Th subset (Th17), which is associated specifically with the Treg dysfunction and pathogenesis of DED. Many of the classic Th1-mediated autoimmune diseases are now also being associated with Th17 cells, suggesting that autoimmune disorders are not necessarily restricted to just one type of immune response (14, 15). Our data demonstrate that DED Tregs are relatively more inefficient in suppressing the activation of DED T effectors than that of naive T cells. These findings prompted us to analyze the proliferating subsets of T effectors. Interestingly, the main proliferating subset of DED T effectors in the presence of Tregs is found to be IL-17-secreting CD4⁺ T cells, confirming previous reports that suggest that in contrast to Th1, Th17 cells may not be suppressed by Tregs (22, 23, 24).

IL-17 is known to induce epithelial, stromal, and immune cells to secrete proinflammatory cytokines such as IL-6, TNF-α, IL-1, and IL-8 (15). Indeed, these cytokines have also been shown to be increased in the DED (25, 26) and, along with our demonstration of Th17 cells and expression of IL-17R by the ocular surface epithelia, point toward the importance of an IL-17-mediated cytokine cascade in the immunopathogenesis of DED. The constitutive expression of IL-17Rs by corneal and conjunctival epithelia makes the ocular surface more vulnerable to IL-17-mediated inflammation and tissue damage. The presence of proinflammatory cytokines in the milieu may actually favor the generation of autoreactive T cells. For instance, IL-6 in the presence of TGF-β leads to the differentiation of autoreactive Th17 cells, which can restrain Treg function and favor autoreactivity (27). Recently, it has also been reported that Tregs can be self-induced to become Th17 cells in the presence of IL-6 (28). However, our data do not demonstrate any decrease in Treg frequency in DED, suggesting that generation of Th17 cells in DED mice is primarily not due to the self-conversion of Tregs into Th17 cells.

The functional relevance of Th17 cells in the immunopathogenesis of DED is fundamentally confirmed by the finding that in vivo neutralization of IL-17 results in a markedly attenuated induction and severity of disease, which is paralleled by a reduction in the expansion of Th17 cells. Remission of clinical disease by IL-17 blockade has also been reported in other Th17-associated autoimmune disorders (29, 30), but this is the first in vivo demonstration that blockade of IL-17 not only inhibits the generation of Th17 cells but importantly also notably prevents the loss of Treg function. Previous studies have demonstrated that anti-IL-17 treatment can directly inhibit the T cell expansion and IL-6 production by IL-17 target (epithelial, stromal, and immune cells) cells (31). IL-6, which controls a reciprocal relationship in the development of Tregs and Th17 cells, inhibits Treg function and facilitates Th17 cell expansion (27). Taken together, these findings suggest that Th17 cell-secreted IL-17 antagonizes the Treg function, which in turn may further unleash Th17 and Th1 cells to expand, release inflammatory cytokines, and cause pathological damage.

In summary, these findings provide new insights on the role of both regulatory and effector subsets of T cells in the pathogenesis of DED. The induction of Th17 cells and the lack of Treg control over pathogenic T effectors, in combination, lead to the induction of autoimmunity and sustained inflammation in DED. The demonstration that IL-17 blockade leads to diminished disease severity suggests potential novel approaches for the treatment of dry eye disease. Strategies designed to target IL-17 or inhibit Th17 cell function and, thereby favor Treg functionality, may be viable approaches for the treatment of dry eye.

The authors have no financial conflict of interest.