Nitrogen Mustard-Induced Corneal Injury Involves DNA Damage and Pathways Related to Inflammation, Epithelial-Stromal Separation, and Neovascularization

Abstract

Purpose: To evaluate the toxic effects and associated mechanisms in corneal tissue exposed to the vesicating agent, nitrogen mustard (NM), a bifunctional alkylating analog of the chemical warfare agent sulfur mustard.

Methods: Toxic effects and associated mechanisms were examined in maximally affected corneal tissue using corneal cultures and human corneal epithelial (HCE) cells exposed to NM.

Results: Analysis of ex vivo rabbit corneas showed that NM exposure increased apoptotic cell death, epithelial thickness, epithelial-stromal separation, and levels of vascular endothelial growth factor, cyclooxygenase 2, and matrix metalloproteinase-9. In HCE cells, NM exposure resulted in a dose-dependent decrease in cell viability and proliferation, which was associated with DNA damage in terms of an increase in p53 ser15, total p53, and H2A.X ser139 levels. NM exposure also induced caspase-3 and poly ADP ribose polymerase cleavage, suggesting their involvement in NM-induced apoptotic death in the rabbit cornea and HCE cells. Similar to rabbit cornea, NM exposure caused an increase in cyclooxygenase 2, matrix metalloproteinase-9, and vascular endothelial growth factor levels in HCE cells, indicating a role of these molecules and related pathways in NM-induced corneal inflammation, epithelial-stromal separation, and neovascularization. NM exposure also induced activation of activator protein 1 transcription factor proteins and upstream signaling pathways including mitogen-activated protein kinases and Akt protein kinase, suggesting that these could be key factors involved in NM-induced corneal injury.

Conclusions: Results from this study provide insight into the molecular targets and pathways that could be involved in NM-induced corneal injuries laying the background for further investigation of these pathways in vesicant-induced ocular injuries, which could be helpful in the development of targeted therapies.

Figure 1

Effect of NM exposure on apoptotic cell death in rabbit cornea, and cell viability and proliferation, and molecular responses related to DNA damage and apoptotic cell death in HCE cells. The excised rabbit corneas were exposed to 50–200 nmol NM for 2 h, washed and cultured for 12 h or exposed to 100 nmol NM for 2 h, washed and cultured for 12, 24 or 48 h. Thereafter, the corneas were collected, processed, sectioned and subjected to TUNEL staining as detailed under Materials and methods. Representative TUNEL stained corneal sections (A) were further quantified (B & C) as brown colored TUNEL positive cells in 10 randomly selected fields at 400× magnification, and apoptotic cell index was calculated as number of apoptotic cells ×100 divided by total number of cells. HCE cells were either not exposed or exposed to 1–200 μM concentrations of NM and then examined under the light microscope for morphological analysis (D). After similar treatments, HCE cells were subjected to MTT assay (E) or BrdU assay (F) as detailed under the Materials and method section. HCE cells were either not exposed or exposed to 1–200 μM concentrations of NM and cell lysates were prepared. About 60 μg of protein sample was loaded and analyzed by SDS-PAGE followed by western immunoblotting for H2A.X phosphorylation, and P53 phosphorylation and accumulation (G), and cleaved caspase-3 and PARP as detailed under Materials and Methods. Protein loading was checked by stripping and re-probing the membranes with β-actin antibody and the results obtained were quantified by densitometric analysis of the immunoblots. Data presented are mean±SEM (n=3–6); *, p<0.05 as compared to control group; NM, nitrogen mustard; e, epithelial layer; s, stromal layer; red arrows, TUNEL +ve cells.

Figure 2

Effect of NM exposure on epithelial thickness and epithelial-stromal separation in rabbit cornea. The excised rabbit corneas were exposed to 50–200 nmol NM for 2 h, washed and cultured for 12 h or exposed to 100 nmol NM for 2 h, washed and cultured for 12, 24 or 48 h. Thereafter, the corneas were collected, processed, sectioned and subjected to H&E staining as detailed under Materials and methods. H&E stained sections were evaluated for epithelial thickness (A–C) and epithelial-stromal separation (D-F). Representative H&E stained corneal sections epithelial thickness (A) were quantified (B and C) as detailed under Materials and methods. Representative H&E stained corneal sections showing the incidence of epithelial-stromal separations (D) were measured at 400× magnification and classified into small (less than 100 μm²), medium (100–1000 μm²) or large (more than 1000 μm²) epithelial-stromal separations (E & F). Data presented are mean±SEM (n=3–6); *, p<0.05 as compared to control group; NM, nitrogen mustard; e, epithelial layer; s, stromal layer; red arrows, epithelial-stromal separation.

Figure 3

Effect of NM exposure on inflammatory, proteolytic and angiogenic mediators in rabbit cornea and HCE cells. The excised rabbit corneas were exposed to 50–200 nmol NM for 2 h, washed and cultured for 12 h or exposed to 100 nmol NM for 2 h, washed and cultured for 12, 24 or 48 h. Thereafter, the corneas were collected, processed, sectioned and subjected to VEGF staining as detailed under Materials and methods. Representative VEGF stained cells in corneal sections (A) were scored for the brown color cytoplasmic staining (B and C). Following desired NM exposures, lysates were prepared from corneal tissue and equal amount of protein was subjected to western immunoblotting for COX-2 and MMP-9 expression (D). HCE cells were either not exposed or exposed to 50 or 100 μM concentrations of NM and cell lysates were prepared. About 60 μg of protein sample was loaded and analyzed by SDS-PAGE followed by western immunoblotting for VEGF COX-2, MMP-9 and iNOS levels in HCE cells (E) as detailed under Materials and methods. Protein loading was checked by stripping and re-probing the membranes with β-actin antibody and the results obtained were quantified by densitometric analysis of the immunoblots as detailed in Materials and methods. Data presented are mean±SEM (n=3–6); *, p<0.05 as compared to control group; NM, nitrogen mustard; e, epithelial layer; s, stromal layer.

Figure 4

Effect of NM exposure on phosphorylation of MAPKs, AKT and AP-1 transcription factor in HCE cells. HCE cells were either not exposed or exposed to 50 or 100 μM concentrations of NM and cell lysates were prepared. About 60 μg of protein sample was loaded and analyzed by SDS-PAGE followed by western immunoblotting for phosphorylated and total MAPKs: ERK, p38, SAPK/JNK (A), AKT (B), AP-1 subunits c-jun and c-fos (C) as detailed under Materials and Methods. Protein loading was checked by stripping and re-probing the membranes with β-actin, tubulin or TBP antibody and the results obtained were quantified by densitometric analysis of the immunoblots as detailed in Materials and methods. NM, Nitrogen mustard.

Figure 5

Effect of NM exposure on DNA damage and cleaved PARP levels, and on inflammatory and proteolytic mediators in HCE cells. HCE cells were exposed to NM (100 and 200 μM) for 2 h and the cells were washed cultured for 12, 24 and 48 h time period. Thereafter, cell lysates were prepared and about 60 μg of protein sample was loaded and analyzed by SDS-PAGE followed by western blotting for H2A.X and p53 phosphorylation and accumulation (A), cleaved PARP (A), and COX-2 and MMP-9 levels (B) as detailed under Materials and methods. Protein loading was checked by stripping and re-probing the membranes with β-actin antibody and the results obtained were quantified by densitometric analysis of the immunoblots as detailed under Materials and methods. NM, Nitrogen mustard.

Figure 6

Effect of NM exposure on lipid peroxidation and DMPO nitrone protein adduct formation in HCE cells. HCE cells were either not exposed or exposed to 50 or 100 μM concentrations of NM and cell lysates were prepared. About 60 μg of protein sample was loaded and analyzed by SDS-PAGE followed by western immunoblotting with 4-HNE antibody (A) or anti-DMPO antibody (B). Protein loading was checked by stripping and re-probing the membranes with β-actin antibody as detailed in Materials and methods. NM, nitrogen mustard; red arrows, protein adducts.