Non-steroidal anti-inflammatory drug delays corneal wound healing by reducing production of 12-hydroxyheptadecatrienoic acid, a ligand for leukotriene B4 receptor 2

Abstract

Non-steroidal anti-inflammatory drugs (NSAIDs) are widely used to reduce inflammation by suppressing cyclooxygenases (COXs). NSAID eye drops are frequently prescribed after ocular surgery to reduce inflammation and pain, but this treatment has clinically significant side effects, including corneal ulcer and perforation. The molecular mechanisms underlying these side effects remain unknown. Recently, the COX product 12(S)-hydroxyheptadeca-5Z,8E,10E-trienoic acid (12-HHT) was identified as an endogenous ligand for leukotriene B₄ receptor 2 (BLT2), which is important in maintenance of epithelial homeostasis. We hypothesized that NSAID-dependent corneal damage is caused by reduced production of 12-HHT. Diclofenac eye drops decreased the abundance of downstream products of COX and delayed corneal wound healing in BALB/c mice. Expression of BLT2 was observed in murine ocular tissues including cornea, and in human corneal epithelial cell line and human primary corneal epithelial cells. In BLT2-knockout mice, corneal wound healing was delayed, but the diclofenac-dependent delay in corneal wound healing disappeared. 12-HHT accelerated wound closure both in BLT2-transfected corneal cell line and human primary corneal epithelial cells. Thus, our results reveal that NSAIDs delay corneal wound healing by inhibiting 12-HHT production, and suggest that stimulation of the 12-HHT/BLT2 axis represents a novel therapeutic approach to corneal wound healing.

Figure 1

The NSAID diclofenac suppresses 12-HHT production and delays corneal wound healing in BALB/c mice. (A) Representative macroscopic images of fluorescein-stained corneal wounds in BALB/c mice after treatment with or without 0.1% diclofenac in PBS for 2 days (four times/day). (B) Wound area was measured every 8 hr after wounds were introduced in 13week old mice (n = 5–6). Error bars indicate means ± S.E. Data were analyzed by two-way ANOVA: ***p < 0.001. (C,D) Mice were treated with 0.1% diclofenac in PBS (Diclofenac) or vehicle control (Control) for 2 days (four times/day). Levels of eicosanoids and 12-HHT in the eye were quantified by LC-MS/MS (n = 3). Error bars indicate means ± S.E. (E) Expression of COX-1 mRNA in mouse cornea, conjunctiva, lens, and retina (n = 3).

Figure 2

Expression of BLT2 in murine eye. (A) RT-PCR was performed to examine BLT2 and EP receptors in cornea. (B,C) Mouse Blt2 mRNA expression was determined by qPCR analysis in mouse eye, skin, intestine (B), cornea, conjunctiva, lens, and retina (C) (n = 3). β-actin was used as an internal control. (D) Immunohistochemical staining of BLT2 in cornea of BLT2 WT or BLT2 KO mice. (Magnification: 400X) (E,F) Expression levels of human BLT2 mRNA were measured by qPCR in the human corneal epithelial cell line HCET (E) and primary human corneal epithelial cells (F). GAPDH was used as an internal control. Error bars indicate means ± S.E. (n = 3). RT: reverse transcription. Str: stromal layer; Epi: epithelial layer of cornea.

Figure 3

BLT2 KO cornea has no overt phenotype at steady state. (A) Representative HE-stained sections of cornea from BLT2 WT (left panel) and BLT2 KO mice (right panel). (B) Electron micrographs of cornea from BLT2 WT (left panel) and BLT2 KO mice (right panel). (C,E) Representative electron micrographs of superficial layer (C) and basal layer (E) in BLT2 WT (left panel) and BLT2 KO cornea (right panel). Arrows indicate desmosomes. (D,F) Quantification of the number of desmosomes in the superficial layer (E) and basal layer (F). (G–I) mRNA expression of mouse E-cadherin (Cdh1), N-cadherin (Cdh2), and vimentin (Vim) was determined by qPCR analysis of mouse cornea. Error bars indicate means ± S.E. (n = 3). n.s.: not significant. Scale bars: 50 µm (A), 10 µm (B), 1 µm (C,E).

Figure 4

Corneal wound healing is delayed in BLT2 KO mice. (A) Representative macroscopic images of fluorescein-stained corneal wounds in BLT2 WT and BLT2 KO mice at the indicated times after wounding. (B) Corneal wound area was measured in 11–12 weeks old BLT2 WT and BLT2 KO mice. Data were analyzed by two-way ANOVA: ***p < 0.001. (n = 8–12) (C) Corneal wounds were measured in BLT2 KO mice after treatment with 0.1% diclofenac for 2 days. Data were analyzed by two-way ANOVA; n.s.: not significant. (D) Corneal wounds were measured in 10 week old WT mice after treatment with 0.1% diclofenac for 2 days and treatment of 100 nM 12-HHT or 10 µM CAY10583 (BLT2 agonist). Data were analyzed by two-way ANOVA (8–32 hour); *p < 0.05 **p < 0.01 (n = 5–6).

Figure 5

12-HHT/BLT2 signaling accelerates wound closure in corneal epithelial cells in vitro. (A) Flow cytometric analysis of HCET cells stably expressing FLAG-tagged human BLT2. NS: no staining. (B) 12-HHT–dependent Ca²⁺ mobilization was assessed after loading of Fluo-8 AM in BLT2-overexpressing HCET cells. Error bars indicate means ± S.E. (n = 4). (C) Cell growth of HCET-Mock and HCET-BLT2 (n = 3). (D) Representative fields of HCET-Mock (upper) and HCET-BLT2 cells (bottom) at 0 and 15 hr after scratching. (E) HCET-Mock and HCET-BLT2 cells were subjected to scratch assay in medium containing 0.5% FBS. Wound area was measured and is shown as wound closure rate (%). (F) HCET cells were subjected to scratch assay in the presence of 1 µM 12-HHT. (G) Primary human corneal epithelial cells were subjected to scratch assay. 12-HHT (10 nM) was added, and the wound closure rate was measured. All data represent means ± S.E. Data were analyzed by two-way ANOVA: *p < 0.05, **p < 0.01, ****p < 0.0001; n.s.: not significant.

Figure 6

Schematic illustration of how the 12-HHT/BLT2 signaling pathway functions in corneal wound healing, and how NSAIDs delay healing. (A) 12-HHT activates BLT2 signaling and subsequently accelerates corneal epithelial cell migration, resulting in healing of corneal wounds. (B) NSAID eye drops inhibit production of 12-HHT in the eye, resulting in suppression of downstream events and ultimately delaying the corneal wound healing process.