Characterization of neutrophil and monocyte specific chemotactic factors derived from the cornea in response to hydrogen peroxide injury

Abstract

To unravel the contributions of corneal tissue in endocular inflammation, we examined the capability of corneal endothelial cells (CECs) to release leukocyte chemotactic factors (LCFs) following injury induced by the leukocyte product hydrogen peroxide. Hydrogen peroxide (H2O2) was chemically generated by the interactions of glucose and glucose oxidase prepared in serum-free minimal essential medium (MEM). For these studies, endothelial surfaces of isolated bovine corneas were incubated with glucose (1 mg/ml) and glucose oxidase (20 U/ml) for 4, 6, and 10 hours at 37 C/in a 5% CO2 atmosphere. Supernatants were then removed and assayed for bovine neutrophil or mononuclear cell chemotactic activity. C5 fragment was our positive control for 100% chemotactic response. Corneas were also fixed in buffered formalin for histopathologic evaluation. Results of these studies indicated that 1) 6-hour interactions of the glucose (G)/glucose oxidase (GO) mixture with endothelial surfaces resulted in both endothelial cell injury (cytoplasmic vacuolization and convoluted nuclei) and production of chemotactic factors (via checkerboard analysis) specific for both neutrophils (58% maximum chemotactic response [MCR]) and mononuclear cells (75% MCR); 2) control corneas treated with either G or GO for 4 and 6 hours produced low levels of LCFs (5-15% MCR); 3) preliminary molecular weight characterization of cornea-derived LCFs obtained from corneas incubated with G/GO for 6 hours revealed the detection of chemotactic activity specific for mononuclear cells in two major fractions, one near the void volume (greater than 130,000 daltons) and one near the elution volume (less than or equal to 10,000-15,000 daltons). Chemotactic activity specific for neutrophils was detected only in one major fraction near the elution volume (less than or equal to 10,000-15,000 daltons); and 4) the production of these LCFs by isolated corneas was significantly inhibited, in a dose-response fashion, when the enzyme catalase (1200-6000 U/ml) was added to corneas incubated with G/GO for 6 hours. To investigate whether isolated CECs were capable of producing LCFs in response to G/GO injury, the authors incubated cultured bovine CECs with G/GO for 3, 6, and 20 hours at 37 C in a 5% CO2 atmosphere. Similar to isolated corneas, cultured CECs incubated with G/GO for 6 hours produced significant levels of LCFs specific for neutrophils (67% MCR) and mononuclear cells (75% MCR). Furthermore, the addition of catalase (3000 U/ml) to corneas incubated with G/GO for 6 hours markedly reduced the production of LCFs. These in vitro studies suggest that corneal cells and/or corneal matrix may play important roles in the initiation and amplification of endocular inflammation in vivo by elaborating factors that control leukocyte influx to the anterior chamber region.