Thrombospondin‐derived peptide attenuates Sjögren's syndrome‐associated ocular surface inflammation in mice

Summary

Sjögren's syndrome is the second most common rheumatic disease in which autoimmune response targets exocrine glands (salivary and lacrimal glands) result in clinical symptoms of dry mouth and dry eye. Inflammation of the lacrimal gland induces tear abnormalities that contribute to the inflammation of the ocular surface, which includes ocular mucosa. Thrombospondin‐1 (TSP‐1) plays a critical regulatory role in the ocular mucosa and as such TSP‐1^–/– mice develop spontaneously chronic ocular surface inflammation associated with Sjögren's syndrome. The autoimmune pathology is also accompanied by a peripheral imbalance in regulatory (T_reg) and inflammatory Th17 effectors. In this study, we demonstrate an in‐vitro effect of a CD47‐binding TSP‐derived peptide in the induction of transforming growth factor (TGF)‐β1‐secreting forkhead box protein 2 (Foxp3⁺) T_regs from activated CD4⁺CD25^– T cells and the inhibition of pathogenic T helper type 17 (Th17)‐promoting interleukin (IL)‐23 derived from antigen‐presenting cells. The in‐vivo administration of this peptide promotes Foxp3⁺ T_reg induction and inhibition of Th17 development. Consistent with these results, topical administration of CD47‐binding TSP peptide, both before and after the onset of the disease, attenuates clinical symptoms of SS‐associated dry eye in TSP‐1^–/– mice. Augmented expression of Foxp3 detected in the draining lymph nodes of TSP peptide ‐treated mice compared to those treated with control peptide suggests the ability of TSP peptide to restore peripheral immune imbalance. Thus, our results suggest that TSP‐derived peptide attenuates Sjögren's syndrome‐associated dry eye and autoimmune inflammation by preventing Th17 development while promoting the induction of T_regs. Collectively, our data identify TSP‐derived peptide as a novel therapeutic option to treat autoimmune diseases.

Introduction

Sjögren's syndrome is a rheumatic disease that afflicts both oral and ocular mucosa, as the salivary and lacrimal glands are targeted by the autoimmune response. The inflammatory damage to the lacrimal gland results in compromised tear quality, which perpetuates the inflammation at the ocular surface. Chronic ocular inflammation associated with autoimmune pathology is a major challenge to treat. Most currently applied treatment options to suppress autoimmune inflammation on a long‐term basis, including corticosteroids, are associated with intolerable side effects. Thus, in addition to a greater risk of visual impairment, chronic ocular inflammation carries a large cost and quality of life burden. Therefore, the availability of better treatment alternatives for chronic ocular inflammation is a pressing need in ophthalmology clinics. Given that an imbalance in regulatory and inflammatory immune response is a major factor underlying most autoimmune diseases, many new therapeutic approaches are designed to restore this balance either by inhibiting or blocking inflammatory effectors and their cytokines or by inducing regulatory effectors.

Thrombospondin‐1 is also an important contributor to regulatory mucosal immunity at the ocular mucosa. Similar to other mucosal surfaces, such as intestinal or airway mucosa, transforming growth factor (TGF)‐β2 is the predominant isoform detectable in conjunctiva 1, 2. We have shown that the conjunctival goblet cells have the ability to secrete TGF‐β2 and activate it in a thrombospondin‐1 (TSP‐1)‐dependent manner, allowing modulation of neighbouring dendritic cell phenotypes towards an immature or tolerogenic state 2. Recently we reported that antigen‐presenting cells (APCs) exposed to TGF‐β2 induce forkhead box P3 (Foxp3)⁺ regulatory T cells (T_regs) in a TSP‐1‐dependent manner 3. Other studies have reported TSP‐1 as a negative regulator of dendritic cell maturation as well as the ability of TSP‐1‐mediated T cell anergy and T_reg homeostasis 4, 5, 6. Moreover, results of in‐vitro experiments have demonstrated that both exogenously provided TSP‐1 and CD47‐binding TSP‐1 peptide inhibit APC‐derived T helper type 1 (Th1)‐promoting interleukin (IL)‐12 secretion 7, 8. Collectively, these findings support the spontaneous development of chronic ocular surface inflammation associated with Sjögren's syndrome in TSP‐1‐deficient mice 9, 10, 11. Our recent identification of a polymorphism in the gene encoding TSP‐1 that results in reduced TSP‐1 expression and its correlation with the susceptibility to chronic ocular surface inflammation inňhumans further confirm findings from murine studies 12.

Spontaneously developed ocular mucosal inflammation in TSP‐1‐deficient mice resembles closely that seen in patients with Sjögren's syndrome. Despite appearing normal at birth, these mice develop a chronic ocular surface inflammation with age, with earliest signs detectable in the conjunctiva by 8 weeks and a fully established ocular mucosa inflammation by 12 weeks of age. The disease is characterized by secretory dysfunction of lacrimal gland and conjunctival goblet cells that lead to a gradual decline in tear quality and disruption of the corneal epithelial barrier. These changes are accompanied with the development of a chronic conjunctival inflammation characterized by the presence of inflammatory infiltrates, elevated levels of inflammatory cytokines, loss of mucin secreting goblet cells and reduced tear mucin (MUC5AC) levels 10.

In this study we demonstrate that a TSP‐1‐derived peptide (4N1K), in vitro, induces TGFβ‐secreting Foxp3⁺ T_regs and inhibits Th17‐promoting IL‐23 secretion by inflammatory APCs. This effect is also detected in vivo, as TSP‐1 peptide inhibits Th17 successfully and promotes Foxp3⁺ T_reg development. Furthermore, topical administration of 4N1K in TSP‐1‐deficient mice attenuates clinical symptoms of Sjögren's syndrome‐associated ocular surface inflammation while augmenting the expression of T_reg marker, Foxp3, in draining lymph nodes. Together our results support therapeutic efficacy of a TSP‐1‐derived peptide in the treatment of chronic autoimmune inflammation.

Materials and methods

Mice

C57BL/6 (H‐2b) male mice, aged 4–12 weeks, were purchased from Charles River Laboratories (Wilmington, MA, USA). A breeding pair of TSP‐1^null mice (C57BL/6 background) was purchased from Jackson Laboratories (Bar Harbor, MI, USA). These mice were bred subsequently in‐house in a pathogen‐free facility at Schepens Eye Research Institute (SERI) initially and later at Boston University School of Medicine (BUSM), Boston, MA.

The Institutional Animal Care and Use Committee (IACUC) at Boston University School of Medicine, Boston, approved animal studies described in this manuscript in accordance with the National Institutes of Health (NIH) guide for the care and use of laboratory animals. All animal experiments were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

Antibodies and reagents

Antibodies were anti‐CD3; clone 2C11(BD Biosciences, San Jose, CA, USA), anti‐CD47 (Biolegend, San Diego, CA, USA), fluorescence conjugated anti‐CD4, anti‐CD25 and anti‐Foxp3 (BD Biosciences).

Recombinant proteins were mouse IL‐2 and mouse interferon (IFN)‐γ (R&D Systems, Minneapolis, MN, USA). Peptides, TSP peptide 4N1K (KRFYVVMWKK) and control peptide 4NGG (KRFYGGMWKK), were synthesized by Bio Basic (Markham, Ontario, Canada) and reconstituted in sterile phosphate‐buffered saline (PBS). Enzyme‐linked immunosorbent assay (ELISA) kits were MUC5AC (TSZ ELISA, Waltham, MA, USA), IL‐23 and TGF‐β1 (eBioscience, San Diego, CA, USA). The T_reg isolation kit was purchased from Miltenyi (Bergish Gladbach, Germany). Phorbol myristate acetate (PMA) and ionomycin were purchased from Sigma (St Louis, MO, USA) (PMA) and EMD Millipore (Billerica, MA, USA) (ionomycin). Culture medium RPMI‐1640, 10 mM HEPES, 0·1 mM non‐essential amino acid (NEAA), 1 mM sodium pyruvate, 100 U/ml penicillin, 100 mg/ml streptomycin, 200 mM L‐glutamine (Lonza, Basel, Switzerland), insulin‐transferrin‐selenium (ITS) + 1 culture supplement [1 μg/ml iron‐free transferrin, 10 ng/ml linoleic acid, 0·3 ng/ml Na2Se and 0·2 μg/ml Fe(NO3)3] were purchased from Sigma.

Serum free medium

Serum‐free medium (SFM) was used for in‐vitro assays. The medium contained RPMI‐1640, 10 mM HEPES, 0·1 mM NEAA, 1 mM sodium pyruvate, 100 U/ml penicillin, 100 mg/ml streptomycin, 200 mM L‐glutamine, 0·1% BSA) and ITS + 1 culture supplement.

In‐vitro stimulation of T cells

Lymph nodes and spleens were harvested from C57BL/6 mice and a single‐cell suspension was prepared by homogenizing them through a 70‐μm nylon membrane. The cells were then washed and the red blood cells (RBC) were lysed using RBC lysis buffer [155 mM NH₄Cl, 10 mM KHCO₃ and 0·1 mM ethylenediamine teraacetic acid (EDTA)]. Using a T_reg cell isolation kit, CD4⁺CD25^– T cells were isolated according to the manufacturer's protocol. Purity of isolated cells as determined by flow cytometry was > 97%. Isolated CD4⁺CD25^– T cells (2 × 10⁵ per well) were cultured in a 96‐well plate and stimulated with plate‐bound anti‐CD3 (1 μg/ml) antibody together with soluble anti‐CD28 (1 μg/ml) in the presence of either control or TSP peptide (10 μM) with or without anti‐CD47 (10 μg/ml) antibody. After 48 h of culture, supernatants were collected from peptide‐stimulated T cells and tested for levels of total TGF‐β1 by ELISA. Activated T cells were isolated, washed and stained for cell surface CD4, CD25 and intracellular Foxp3 using a cell permeabilizing kit (eBioscience). In some experiments activated T cells were stimulated further with anti‐CD3 (1 μg/ml) and recombinant mouse (rm)IL‐2 (50 U/ml) every 2–3 days for a period of 9 days prior to harvesting. Stained cells were analysed by flow cytometry (BD LSR II; BD Biosciences). Flow cytometric analysis was performed using FlowJo software (Treestar, Inc, Ashland, OR, USA).

In‐vitro stimulation of APCs

Macrophages (> 95% F4/80+) were harvested from peritoneal fluid of C57BL/6 mice injected with 2 ml of 3% thioglycollate solution (Sigma) 3 days earlier. These APCs (2 × 10⁵ per well) were stimulated with IFN‐γ (500 ng/ml) and lipopolysaccharide (LPS) (10 μg/ml) in SFM for 24 h prior to adding control or TSP peptide (50 μg/ml). Culture supernatants were collected at 18 h to detect IL‐23 levels in an ELISA assay and RNA was isolated from stimulated cells to detect expression of p40 in a real‐time polymerase chain reaction (PCR).

Real time PCR

Total RNA was isolated from cervical lymph nodes harvested from TSP peptide and control peptide‐treated mice, and from APCs using TRIzol Reagent (Life Technologies, Carlsbad, CA, USA), according to the manufacturer's instructions. cDNA was synthesized using the SuperScript VILO cDNA kit (Life Technologies). Real‐time PCR was performed on a 7200 Real Time System (Applied Biosystems, Carlsbad, CA, USA) using SYBR Green PCR Master Mix (Life Technologies) to determine relative quantitative expression levels of Foxp3 and IL‐12 (p40). Foxp3 primers (forward, 5′‐GGAGAGGCAGAGGACACTCAAT‐3′ and reverse, 5′‐GTGGTTTCTGAAGTAGGCGAACAT‐3′), IL‐12 (p40) primers (forward, 5′‐GTTCGAATCCAGCGCAAGAA‐3′ and reverse, 5′‐TTTGCATTGGACTTCGGTAGATGT‐3′) and glyceraldehyde‐3‐phosphate dehydrogenase primers (forward, 5′‐CGAGAATGGGAAGCTTGTCA‐3′ and reverse, 5′‐AGACACCAGTAGACTCCACGACAT‐3′) were used. Amplification reactions were set up in quadruplicate with the following thermal profile: 95ºC for 3 min, 40 cycles at 95ºC for 20 s, 53ºC for 30 s and 72ºC for 40 s. To verify the specificity of the amplification reaction, a melting curve analysis was performed. The fluorescence signal generated at each cycle was analysed using system software. The threshold cycle values were used to determine relative quantification of gene expression with glyceraldehyde‐3‐phosphate dehydrogenase as a reference gene.

Sjögren's syndrome model – treatment regimen and disease monitoring

A mouse model of Sjögren's syndrome characterized in TSP‐1^null mice was used 11. Mice (total n = 10 per group) at the ages of 6 (prior to the onset of ocular surface inflammation) or 12 weeks (after disease onset) were treated topically with control or TSP peptide (10 μg/mouse), bilaterally (5 µl/eye), once a day for 2 weeks. Disease progression was monitored by corneal fluorescein staining and determining tear content of MUC5AC before starting the treatment (baseline) and after 1 and 2 weeks of treatment in the case of 12‐week‐old mice. In the case of 6‐week‐old mice disease was monitored at baseline and at 12 weeks of age (with peptide treatment from weeks 6–8). Corneal fluorescein staining was performed as described previously 11. Briefly, 1% sodium fluorescein (Sigma‐Aldrich, St Louis, MO, USA) was applied to the cornea of mice under anaesthesia. Three minutes later, eyes were flushed with PBS to remove excess fluorescein, and corneal staining was evaluated with a slit‐lamp microscope using a cobalt blue light. Punctate staining was recorded using a standardized National Eye Institute grading system of 0–3 for each of the five areas of the cornea 13. Pilocarpine‐induced tears were collected as described previously 11 and pooled samples were analysed to determine MUC5AC content in an ELISA. The assay was performed according to the manufacturer's instructions. Detected mucin content was normalized to tear volume, and the result was presented as MUC5AC ng/ml of tear volume.

Statistical analysis

The unpaired Student's t‐test was used to determine significant differences between mean values of experimental and control groups. Error bars enclosing mean values in figures represent ± standard error of the mean (s.e.m.). P < 0·05 was considered statistically significant.

Results

TSP peptide promotes induction of regulatory immune response

We examined the effect of CD47‐binding TSP peptide 4N1K (or control peptide 4NGG) on purified CD4⁺CD25^– T cells activated with anti‐CD3/CD28 antibodies for 48 h. To assess the induction of T_regs we analysed Foxp3 expression in activated T cells by flow cytometry and measured levels of TGF‐β secreted in culture supernatants. As shown in Fig. 1a, significantly increased numbers of T cells, activated in the presence of TSP peptide, expressed Foxp3 compared with those activated in the presence of the control peptide. Consistently, T cell activation in the presence of TSP peptide, and not control peptide, resulted in significantly increased levels of TGF‐β detectable in culture supernatant, as shown in Fig. 1b. Further culturing of peptide‐treated activated T cells with IL‐2 resulted in expansion and stable Foxp3 expression in TSP peptide‐treated T cells compared to those treated with control peptide (Fig. 1c). Considering that TSP peptide, 4N1K, binds TSP receptor CD47, we assessed if T_regs were induced via ligation of CD47 on T cells. We activated CD4⁺CD25^– T cells in the presence of CD47 blocking antibody and TSP‐ or control peptides. As shown in Fig. 1a, CD47 blockade abrogated TSP peptide‐induced Foxp3 expression. Similarly, in CD47‐deficient T cells TSP peptide failed to induce Foxp3 expression (data not shown). Together, these results indicate that TSP peptide can induce Foxp3 T_regs and therefore have the potential to alter peripheral immune response in autoimmune pathology.

Figure 1.

Thrombospondin (TSP) peptide promotes induction of forkhead box protein 3 (Foxp3)⁺ regulatory T cells. Expression of Foxp3 was evaluated (a) by flow cytometry in CD4⁺CD25^– T cells from C57BL/6 mice stimulated with CD3/CD28 in the presence of control or TSP peptide for 48 h. The bar graph represents mean % of CD25⁺Foxp3⁺ cells from three individual experiments ± standard error of the mean (s.e.m.). Representative flow cytometry plots of CD25 and Foxp3 staining of CD4 gated cells are shown. (b) Mean transforming growth factor (TGF)‐β1 levels ± s.e.m. in culture supernatants of stimulated CD4⁺CD25^– T cells in the presence of control or TSP peptide. (c) Flow cytometric analysis of CD4 gated cells was performed to determine % of CD25⁺Foxp3⁺ cells in CD4⁺CD25^– T cells after stimulation for 48 h in the presence of control or TSP peptide followed by expansion with recombinant mouse interleukin rmIL‐2 (50 U/ml); (d) after stimulation for 48 h in the presence of TSP peptide and isotype or anti‐CD47 antibody (10 μg/ml) (*P < 0·05).

Thrombospondin‐derived peptide inhibits APC‐derived IL‐23 and the induction of Th17 effectors

It is known that exogenously provided TSP‐1 and CD47‐binding TSP‐1 peptide both inhibit APC‐derived Th1‐promoting IL‐12 secretion 7, 8. In this study we assessed the effect of TSP peptide, 4N1K, on Th17‐promoting IL‐23 secretion by APCs. Thioglycollate‐elicited peritoneal macrophages stimulated with IFN‐γ and LPS were treated with control or TSP peptide. The level of IL‐23 secreted in culture supernatants was determined by ELISA. As shown in Fig. 2a, TSP peptide reduced IL‐23 secretion by APCs significantly compared with the control peptide. This observation was confirmed by inhibition of the p40 subunit (common to IL‐12 and IL 23 cytokines) in TSP peptide‐treated APCs as determined by real‐time PCR (Fig. 2b). These results suggest that TSP peptide inhibits the proinflammatory phenotype of APCs by preventing their ability to support inflammatory effectors.

Figure 2.

Thrombospondin (TSP)‐derived peptide inhibits antigen‐presenting cell (APC)‐derived interleukin IL‐23. Thioglycollate‐elicited macrophages were stimulated in vitro with interferon (IFN)‐γ (50 ng/ml)/lipopolysaccharide (LPS) (1 μg/ml) in the absence or presence of control or TSP peptide for a period of 48 h. (a) Levels of IL‐23 in culture supernatants were determined by enzyme‐linked immunosorbent assay (ELISA); and (b) p40 message levels were determined by real‐time polymerase chain reaction (PCR) using RNA isolated from control or TSP peptide‐treated macrophages. Samples were analysed in triplicate. Shown are mean ± standard error of the mean (s.e.m.). *P < 0·05 compared to control peptide.

In addition to IL‐12 inhibition in APCs, the ligation of CD47 on T cells was reported to inhibit selectively the development of Th1 effectors 14. Therefore, based on IL‐23 inhibition noted in our experiments, we next tested if TSP peptide inhibits induction of Th17 effectors in vivo. To test this possibility we immunized a group of C57BL/6 mice to induce Th17 effectors. We treated these mice (n = 3 per group) with intraperitoneal injection of either control or TSP peptide on days 1, 3 and 5 post‐immunization. On day 7 post‐immunization lymph node cells harvested from the peptide‐injected mice were stimulated with PMA/ionomycin and analysed by flow cytometry to detect intracellular IL‐17 and Foxp3 in CD4⁺ cells. As shown in Fig. 3, reduced numbers of IL‐17‐expressing CD4⁺ T cells were detectable in TSP‐peptide‐treated mice compared with the control group. This change was also accompanied by an increase in Foxp3‐expressing CD4⁺ T cells. These results are consistent with our in‐vitro results and confirm the ability of TSP peptide to restore the peripheral balance of the regulatory immune response, supporting its potential anti‐inflammatory therapeutic effect in autoimmune pathology.

Figure 3.

Administering thrombospondin (TSP)‐derived peptide to immunized mice alters their peripheral balance between regulatory T cell (T_reg) and T helper type 17 (Th17) effectors. Ovalbumin‐immunized mice (n = 3 per group) were treated intraperitoneally (i.p.) with control or TSP peptide (10 μg) on days 1, 3 and 5 post‐immunization. On day 6 pooled lymph node cells (1 × 10⁶) were stimulated with phorbol myristate acetate (PMA) plus ionomycin in the presence of brefeldin A for 20 h. Cell surface CD4 and intracellular interleukin IL‐17 and forkhead box protein 3 (Foxp3) was stained with fluorescence‐conjugated antibodies and cells were analysed by flow cytometry. Percentage of positive cells are indicated in each plot.

Topically administered TSP‐derived peptide limits the disruption of the ocular surface barrier successfully in a mouse model of Sjögren's syndrome

To evaluate the efficacy of TSP‐derived peptide, 4N1K, as a therapeutic intervention for autoimmune ocular inflammation, we used a well‐established and validated mouse model of Sjögren's syndrome – TSP‐1^–/– mice. Although normal at birth, these mice develop chronic ocular surface disease spontaneously by the age of 12 weeks, which increases progressively in severity with age. The onset of the disease is marked by the disrupted corneal barrier integrity detectable by increased fluorescein staining in all 12‐week‐old TSP‐1^–/– mice 11. Concurrently, the inflammation in the conjunctiva leads to a significant loss of mucin‐secreting goblet cells and reduced soluble mucin (MUC5AC) levels in tears 10. Considering that mice deficient in TSP‐1 remain healthy until the age of 6 weeks, we first evaluated the ability of TSP peptide treatment to prevent or attenuate the progression of the spontaneous ocular surface disease. We treated 6‐week‐old TSP‐1‐deficient mice, before the onset of ocular surface disease, with topical TSP peptide for 2 weeks, and then evaluated the disease progression at the age of 12 weeks by determining corneal barrier integrity and tear quality.

As shown in Fig. 4a,b, the corneal fluorescein score of untreated and control peptide‐treated 12‐week‐old TSP‐1‐deficient mice is increased significantly compared with that seen in age‐matched wild‐type (WT) mice. However, significantly reduced corneal fluorescein staining is detected in 12‐week‐old TSP‐1‐deficient mice that received TSP peptide treatment earlier compared with controls. Furthermore, TSP peptide treatment also resulted in preserving tear MUC5AC levels and therefore tear quality. While control peptide‐treated mice showed a progressive decrease in tear MUC5AC levels with age, no such decline was detectable in mice treated with TSP peptide (Fig. 4c). In fact, at the age of 12 weeks, tear MUC5AC level in TSP peptide‐treated mice was significantly higher than that detected in mice treated previously with control peptide.

Figure 4.

Topically administered thrombospondin (TSP)‐derived peptide successfully limits the disruption of ocular surface barrier in a mouse model of Sjögren's syndrome. Beginning at 6 weeks of age (prior to the disease onset), TSP‐1‐deficient mice (n = 10 per group) were treated topically with either control or TSP peptide (10 µg/mouse) for a period of 2 weeks. (a) Integrity of corneal barrier was monitored by fluorescein staining at 12 weeks of age in untreated wild‐type (WT), TSP‐1‐deficient and peptide‐treated TSP‐1‐deficient mice (*P < 0·05 compared to WT; #P < 0·05 compared to untreated TSP‐1‐deficient mice); (b) representative images of fluorescein staining noted, clockwise, in mice (WT) with intact corneal barrier, with disrupted barrier (untreated and control peptide‐treated TSP‐1^null mice) and with improved barrier integrity (TSP peptide‐treated TSP‐1^null mice); (c) tear MUC5AC levels [mean ± standard error of the mean (s.e.m.)] were evaluated by enzyme‐linked immunosorbent assay (ELISA) before the initiation (baseline) of treatment and at 12 weeks of age in peptide‐treated TSP‐1‐deficient mice (P < 0·05 *compared to control peptide and **compared to baseline); (d) cervical lymph node cells harvested at the end of the study period were analysed for the expression of Foxp3 message levels by real‐time polymerase chain reaction (PCR). Data are presented as fold change relative to control peptide‐treated group. [Colour figure can be viewed at wileyonlinelibrary.com]

An imbalance in peripheral immune regulation is believed to contribute to autoimmune pathology and, as such, a significant decline in the peripheral Foxp3⁺ T_reg population was reported in TSP‐1‐deficient mice 10, 11. To determine if the therapeutic efficacy of TSP peptide, 4N1K, correlates with restoration of peripheral regulatory response, we harvested draining cervical lymph nodes from TSP‐1^–/– mice that were treated with control or TSP peptide. These lymph nodes were harvested at the termination of disease monitoring period. As shown in Fig. 4d, we noted significantly increased levels of message for Foxp3 in mice treated with TSP peptide compared with controls. These results suggest that early TSP peptide treatment prevents successfully the disruption of the corneal barrier associated with Sjögren's syndrome otherwise detectable by the age of 12 weeks in TSP‐1‐deficient mice.

Therapeutic efficacy of topically administered TSP‐derived peptide in treating established ocular surface disease in Sjögren's syndrome

While TSP peptide, 4N1K, prevented progression of autoimmune ocular diseases successfully, we next evaluated whether this peptide is also effective in resolving an established disease. To assess this, we took advantage of extensively characterized disease stages in the mouse model of Sjögren's syndrome. Treatment with topical eye drops containing either control or TSP peptide was initiated in 12‐week‐old TSP‐1‐deficient mice with an established ocular surface disease. Mice were treated daily for a period of 2 weeks.

We assessed corneal barrier by fluorescein staining and measured tear MUC5AC levels by ELISA before initiating the treatment (baseline) and after 1 and 2 weeks of treatment. Treatment with TSP peptide led to a significant progressive decline in corneal fluorescein scores compared with those detected in mice treated with control peptide (Fig. 5a). Such significantly improved corneal barrier integrity from the baseline was detectable within 1 week of TSP peptide treatment, which persisted after completion of the treatment at 2 weeks. Similarly, tear MUC5AC levels also improved significantly in mice treated with TSP peptide for 2 weeks, but not 1 week, compared with those treated with the control peptide (Fig. 5b). Also, as shown in Fig. 5c, significantly increased expression of Foxp3 was detected in the cervical lymph nodes harvested from TSP peptide‐treated mice compared to those treated with control peptide. The recovery of tear MUC5AC levels correlates with restoration of systemic regulation, suggesting inadequate suppression of inflammation achieved by TSP peptide within the first week of treatment. These results demonstrate clearly that the treatment with TSP peptide improves significantly corneal barrier integrity and tear quality, thereby attenuating the ocular surface disease in TSP‐1‐deficient mice. Overall, these data indicate therapeutic efficacy of TSP peptide in preventing as well as resolving clinical manifestations of autoimmune ocular inflammation.

Figure 5.

Therapeutic efficacy of topically administered thrombospondin (TSP)‐derived peptide in treating established ocular surface disease in Sjögren's syndrome. Twelve‐week‐old TSP‐1‐deficient mice (n = 10 per group) with an established ocular surface inflammation were treated topically with control or TSP peptide (10 µg/mouse) for 2 weeks. (a) Corneal barrier integrity was assessed by fluorescein staining before initiating the treatment (baseline) and after 1 and 2 weeks of treatment, and (b) tear MUC5AC levels [mean ± standard error of the mean (s.e.m.)] were evaluated by enzyme‐linked immunosorbent assay (ELISA) (*P < 0·05); (c) cervical lymph node cells harvested at the end of the study period were analysed for the expression of forkhead box protein 3 (Foxp3) message levels by real‐time polymerase chain reaction (PCR). Data are presented as fold change relative to control peptide‐treated group.

Discussion

We have reported previously that spontaneous development of Sjögren's syndrome‐associated chronic ocular surface inflammation in TSP‐1‐deficient mice is accompanied with a peripheral imbalance in T_reg and pathogenic Th17 effectors 11. Additionally, the induction of regulatory immune response by APCs exposed to the TGF‐β2 in the microenvironment is dependent upon their TSP‐1 expression 3. In this study, using a validated murine model of Sjögren's‐associated ocular surface inflammation, we now demonstrate that a peptide derived from the C‐terminal domain of TSP‐1 (4N1K) is remarkably effective in treating autoimmune ocular surface inflammation. Our results suggest that TSP peptide can target both APCs and T cells to help restore peripheral immune regulation by preventing the development of pathogenic Th17 effectors while promoting the development of Foxp3⁺ T_regs.

Epithelial cells at the ocular surface as well as APCs are known to express TSP‐1 2, 15, and both these cell types activate their endogenous TGF‐β in a TSP‐1‐dependent manner 2, 9. This activation requires binding of TSP‐1 to its receptor CD36 16. Thrombospondin peptide used in this study is derived from the CD47‐binding C‐terminal domain that is distant from the TGF‐β binding portion of the TSP‐1 molecule 17. Therefore, peptide 4N1K does not contain any TGF‐β‐activating sequence. Indeed, we detected a TGF‐β‐independent ability of 4N1K peptide to induce TGF‐β‐secreting Foxp3⁺ T_regs. We have reported previously that CD36‐deficient APCs, although failing to activate latent TGF‐β, manage to induce Foxp3 expression in activated T cells, presumably via their TSP‐1 expression 3. However, such activated T cells are not functional, as they do not secrete TGF‐β. Thus, TGF‐β‐independent induction of Foxp3 by CD36‐deficient APCs differs from that induced by CD47‐binding 4N1K peptide. This difference may be due to the ability of APC‐derived TSP‐1 as a whole molecule to engage multiple receptors on T cells against CD47‐driven induction of Foxp3 by 4N1K. Moreover, T_reg induction by TSP peptide 4N1K correlates with its efficacy in limiting and resolving autoimmune ocular pathology in the Sjögren's syndrome mouse model used in this study.

Highly proinflammatory Th17 effector T cells have been associated with a number of autoimmune diseases 18. These cells are also known to induce strong B cell proliferation and antibody production 19. The cytokine IL‐23 is pivotal for stabilizing the pathogenicity of such Th17 cells 20. To our knowledge, our experiments demonstrate for the first time the ability of TSP peptide, 4N1K, to inhibit such a critical heterodimeric cytokine (p19/p40) derived from APCs. We show that such inhibition is achieved by suppressing expression of the p40 subunit, which is also a component of the Th1‐promoting heterodimeric cytokine IL‐12 (p35/p40). Our results not only confirm previously reported inhibition of IL‐12 by 4N1K 7, but also extend this inhibitory effect to Th17‐promoting IL‐23. Reduced frequency of IL‐17 expressing CD4⁺ cells in mice injected with 4N1K is in line with our in‐vitro results. These results suggest that an inhibitory effect of 4N1K on Th17 may enhance further its efficacy of T_reg induction. Blockade of IL‐17 or IL‐23 21 and the induction of T_regs 22 are emerging as independently attractive therapeutic strategies to treat autoimmune diseases. In our study the combinatorial effect of 4N1K clearly supports its observed efficacy in treating autoimmune disease in a mouse model.

For ocular diseases, the topical instillation of drugs is the most common and preferred method of drug administration due to ease of access and patient compliance 23. Moreover, greater permeability of the conjunctival epithelium in combination with greater tissue surface area for drug contact 24 favoured our choice of the topical application of TSP peptide. Penetration and distribution of topically administered TSP peptide to target immune cells in the conjunctiva are probably some of the reasons for its efficacy in improving ocular surface manifestations of Sjögren's syndrome. In particular, the local anti‐inflammatory effect of TSP peptide potentially relieves inflammatory cytokine‐mediated inhibition of tear MUC5AC secretion by conjunctival goblet cells, thus restoring tear mucin levels in TSP‐1‐deficient mice. The improved tear quality in TSP‐peptide‐treated mice can help to restore corneal barrier integrity. While, in this study, we demonstrated functional modulation of APCs and T cells by TSP peptide, other immune cells as potential targets remain to be evaluated.

Overall, our results suggest that a TSP‐1‐derived peptide, 4N1K, attenuates autoimmune ocular inflammation by changing the peripheral immune response to favour T_reg induction. Thus, our study takes advantage of the immunomodulatory function of TSP‐1 to develop a least invasive method that achieves inhibition of inflammation while restoring peripheral immune regulation. In conclusion, topically administered TSP peptide, 4N1K, represents an effective approach to treat chronic ocular inflammation associated with autoimmune Sjögren's syndrome. Currently available therapeutic options to treat chronic ocular surface inflammation, as seen in Sjögren's syndrome, involve a broad immunosuppression that is accompanied with several undesirable side effects. However, drugs focused upon selective ligands in order to avoid such side effects are often limited in their ability to treat complex and multi‐factorial chronic inflammation associated with autoimmune pathology. A molecule with multiple functionalities has the potential to serve as a highly effective therapeutic. Therefore, a TSP peptide such as 4N1K with multiple targets, as demonstrated in this study, is a promising candidate for development as a therapeutic entity for chronic ocular inflammation. Our results in preclinical mouse models strongly support this possibility.

Disclosure

The authors have declared that no disclosures exist.