Soluble Fas ligand blocks destructive corneal inflammation in mouse models of corneal epithelial debridement and LPS induced keratitis.

Neutrophil-mediated inflammation plays a critical role in corneal damage following injury or infection. Previous studies demonstrated that membrane-bound FasL (mFasL) induces neutrophil chemokine production. However, the extracellular domain of mFasL is normally cleaved by matrix metalloproteinases to release a soluble form of FasL (sFasL) and sFasL antagonizes mFasL-mediated chemokine production. Therefore, we hypothesized that sFasL could be used to prevent neutrophil-mediated corneal inflammation associated with injury and bacterial keratitis. To test this hypothesis, GFP-only, sFasL-GFP, or mFasL-GFP were expressed in the corneal stroma of C57BL/6 mice, using intra-stromal injections of plasmid DNA or adenoviral vectors (AV) and the role of mFasL and sFasL in corneal inflammation was examined in models of corneal injury and LPS-induced keratitis. Our work addresses an important area of disagreement in the field of FasL, with regard to the mechanism by which sFasL regulates ocular inflammation. Herein, we demonstrate that an intrastromal injection of GFP-only, sFasL-GFP, or mFasL-GFP plasmid DNA resulted in GFP expression throughout the corneal stroma for up to two weeks with little to no evidence of inflammation in the GFP-only and sFasL-GFP groups and mild corneal inflammation in the mFasL-GFP group. Similarly, following epithelial debridement, corneas expressing GFP-only or sFasL-GFP showed no significant signs of corneal inflammation, with clear corneas at 15 days post debridement. By contrast, epithelial debridement of corneas expressing mFasL-GFP triggered persistent corneal inflammation and the development of central corneal opacities that was blocked by sFasL. Similar to the mFasL-GFP plasmid DNA, intrastromal injection of mFasL-GFP AV triggered mild corneal inflammation, but it was transient and resolved by day 10 with corneas remaining clear out to 30 days post injection. Nevertheless, intrastromal expression of mFasL-GFP AV exacerbated LPS-induced keratitis, corneal opacity, and neovascularization, while sFasL-GFP AV expression prevented LPS-induced keratitis, resulting in a clear cornea. Histological analysis of corneas with LPS-induced keratitis revealed a robust infiltration of macrophages and neutrophils and sFasL expression specifically blocked the neutrophil influx. Overall, our data demonstrate that stromal expression of mFasL is inflammatory, while sFasL is non-inflammatory, and opposes the effects of mFasL in mouse models of epithelial debridement and LPS-induced keratitis. These data demonstrate that a delicate balance between sFasL and mFasL regulates ocular inflammation. This study further identifies sFasL as a potent inhibitor of neutrophil-mediated corneal damage, and supports the potential use of sFasL in the treatment of neutrophil-mediated keratitis. These results strongly support the hypothesis that, in the immune privileged environment of the eye, the isoform of FasL regulates immune privilege and determines the extent of inflammation: mFasL promotes inflammation and sFasL blocks inflammation.