IL-36α/IL-36RA/IL-38 signaling mediates inflammation and barrier disruption in human corneal epithelial cells under hyperosmotic stress

Abstract

Purpose: To explore the distinct expression and diverse roles of IL-36 cytokines in dry eye disease using an in vitro hyperosmolarity model of human corneal epithelial cells (HCECs).

Methods: Primary HCECs were cultured from fresh donor limbal explants. Hyperosmolarity model was established by switching HCECs from isosmotic (312 mOsM) to hyperosmotic medium (350-500 mOsM) alone or with addition of rhIL-36RA or rhIL-38 for 2-48 h. Some cultures were treated with IL-36α (1-10 ng/ml) with or without rhIL-36RA or rhIL-38. Gene expression was detected by RT-qPCR; and protein production and barrier disruption were evaluated by ELISA and/or immunofluorescent staining.

Results: IL-36 cytokines were differential expressed in primary HCECs. Among 3 pro-inflammatory agonists, IL-36α, but not IL-36β and IL-36γ, was distinctly induced at osmolarity-dependent manner while two antagonist IL-36RA and IL-38 were significantly suppressed in HCECs exposed to hyperosmotic stress. IL-36α increased to 4.4-fold in mRNA and 6.9-fold at protein levels (116.0 ± 36.33 pg/ml vs 16.79 ± 6.51 pg/ml in controls) by 450 mOsM, but dramatically inhibited by addition of rhIL-36RA or rhIL-38. Exogenous rhIL-36α stimulated expression of TNF-α and IL-1β at mRNA and protein levels and disrupted tight junction proteins ZO-1 and occludin. However, rhIL-36RA or rhIL-38 suppressed TNF-α and IL-1β production and protected HCECs from barrier disruption in response to IL-36α or hyperosmolarity.

Conclusions: Our findings demonstrate that the stimulated pro-inflammatory IL-36α with the suppressed antagonists IL-36RA and IL-38 is a novel mechanism by which hyperosmolarity induces inflammation in dry eye. IL-36RA and IL-38 may have a therapeutic potential in dry eye.

Keywords: Barrier; Corneal epithelium; Dry eye; Hyperosmolarity; IL-36; IL-36RA; IL-38; Inflammation.

Fig. 1.

Gene expression of IL-36 cytokines in human corneal epithelial cells (HCECs). Five of IL-36 cytokines, three proinflammatory agonists IL-36α, IL-36β, IL-36γ and two IL-36R antagonists IL-36RA and IL-38 were detected at different mRNA levels in HCECs based on the threshold cycle (Ct) value in real-time PCR with house-keeping gene GAPDH as an internal control. Data were summarized as mean ± SD from 10 separated experiments.

Fig. 2.. Differential responses of five IL-36 cytokines to hyperosmolarity in HCECs by RT-qPCR.

A. The time course of IL-36 family response to hyperosmolarity was performed in HCECs exposed to 450 mOsM medium for 2–14 h. IL-36α mRNA reached peak level at 4 h, while the expression of IL-36β and IL-36γ did not change in the period of 2–24 h. The mRNA levels of IL-36RA and IL-38 were suppressed with lowest levels at 4 h. B. The dose-response experiment showed that IL-36α mRNA increased osmolarity-dependently with peak level by 450 and 500 mOsM while the expression of IL-36RA and IL-38 decreased osmolarity-dependently with lowest levels by 450 mOsM medium. Data were summarized as mean ± SD from 5 separated experiments. *P < 0.05, **P < 0.01, as compared with 312 mOsM; ^ P < 0.05, ^^ P < 0.01, as compared with 312 mOsM.

Fig. 3.

Negative regulation of rhIL-36RA and rhIL-38 on IL-36α production induced by hyperosmolarity. IL-36α expression increased by medium at 400-500 mOsM while significantly inhibited by rhIL-36RA and rhIL-38 at10-50 ng/ml, as evaluated by RT-qPCR for mRNA (A) and by ELISA for protein levels (B). Data were summarized as mean ± SD from 5 separated experiments. *P < 0.05, **P < 0.01, as compared with 312 mOsM; ^ P < 0.05, ^^ P < 0.01, as compared with 450 mOsM.

Fig. 4.

IL-36α stimulates the production of TNF-α and IL-1β in HCECs. A. B. TNF-α production increased dose-dependently by IL-36α (1, 5, 10 ng/ml) while suppressed by IL-36RA and IL-38 (10, 50 ng/ml), as evaluated at mRNA by RT-qPCR (A) and protein levels by ELISA (B). C. D. IL-1β production increased dose-dependently by IL-36α (1, 5, 10 ng/ml) while suppressed by IL-36RA and IL-38 (10, 50 ng/ml), as evaluated at mRNA by RT-QPCR (C) and protein levels by ELISA (D). Data were summarized as mean ± SD from 5 separated experiments. *P < 0.05, **P < 0.01, as compared with 312 mOsM; ^ P < 0.05, ^^ P < 0.01, as compared with 450 mOsM.

Fig. 5.

IL-36α/IL-36RA/IL-38 signaling mediated the production of TNF-α and IL-1β in HCECs under hyperosmotic stress. (A) mRNA and protein expression of TNF-α in hyperosmotic medium (400 and 450 mOsM) with 312 mOsM as control. IL-36RA and IL-38 suppressed the production of TNF-α. (B). IL-1β in hyperosmotic medium (400 and 450 mOsM) with 312 mOsM as control. IL-36RA and IL-38 suppressed the production of IL-1β. mRNA expression evaluated by RT-qPCR, protein levels by ELISA. Data were summarized as mean ± SD from 5 separated experiments. *P < 0.05, **P < 0.01, as compared with 312 mOsM; ^ P < 0.05, ^^ P < 0.01, as compared with 450 mOsM.

Fig. 6.. Corneal epithelial barrier was disrupted by IL-36α but protected by IL-36RA and IL-38.

A. C. Representative images showed that IL-36α (10–50 ng/ml) dose-dependently disrupted corneal epithelial tight junction proteins ZO-1 (A) and occluding (C), which were largely restored by IL-36RA and IL-38 restored corneal barrier from disruption. B. D. The integrated or remaining areas of fluorescent stain per field were quantified as diagrams for ZO-1 (C) and occludin (D). Data were summarized as mean ± SD from 5 separated experiments. **P < 0.01, ***P < 0.001, as compared with 312 mOsM.

Fig. 7.. IL-36α/IL-36RA/IL-38 signaling mediated the corneal epithelial barrier disruption by hyperosmolarity.

Representative images of immunofluoreseent staining showed that tight junction proteins ZO-1 (A) and occludin (C) were significantly disrupted by hyperosmotic medium at 450 mOsM, but largely restored by IL-36RA or IL-38 at 10 ng/ml. B. D. The integrated or remaining areas of fluorescent stain per field were quantified as diagrams for ZO-1 (C) and occludin (D). Data were summarized as mean ± SD from 5 separated experiments. **P < 0.01 as compared with 312 mOsM.

Fig. 8.. Negative regulation of rhIL-1RA on mRNA and production of IL-36α induced by hyperosmolarity.

IL-36α expression increased by medium at 450 mOsM while partially inhibited by rhIL-1RA at10-50 ng/ml, as evaluated by RT-qPCR for mRNA (A) and by ELISA for protein levels (B). IL-1β expression increased by 450 mOsM while largely inhibited by rhIL-1RA at10-50 ng/ml, as evaluated by RT-qPCR for mRNA (C) and by ELISA for protein levels (D). Data were summarized as mean ± SD from 5 separated experiments. **P < 0.01, as compared with 312 mOsM; ^ P < 0.05, ^^ P < 0.01, as compared with 450 mOsM.