Keratocan and lumican regulate neutrophil infiltration and corneal clarity in lipopolysaccharide-induced keratitis by direct interaction with CXCL1

Abstract

Keratocan and lumican are keratan-sulfate proteoglycans (KSPG), which have a critical role in maintaining corneal clarity. To determine whether these KSPGs have a role in corneal inflammation, we examined Kera(-/-) and Lum(-/-) mice in a model of lipopolysaccharide (LPS)-induced keratitis in which wild-type mice develop increased corneal thickness and haze due to neutrophil infiltration to the corneal stroma. Corneal thickness increases caused by LPS mice were significantly lower in Kera(-/-) and Lum(-/-) than wild-type mice. Further, LPS-injected Lum(-/-) mice had elevated corneal haze levels compared with that of Kera(-/-) and wild-type. At 24 h post-injection, total enhanced green fluorescent protein-positive bone marrow-derived inflammatory cells in chimeric mice was significantly lower in Kera(-/-) mice and Lum(-/-) mice compared with wild-type mice. Neutrophil infiltration was inhibited in Kera(-/-) and Lum(-/-) mice at 6 and 24 h post-stimulation, with Lum(-/-) corneas having the most profound defect in neutrophil migration. Reconstitution of keratocan and lumican expression in corneas of Kera(-/-) and Lum(-/-) mice using adeno-keratocan and adeno-lumican viral vectors, respectively, resulted in normal neutrophil infiltration in response to LPS. Immunoprecipitation/Western blot analysis showed that lumican and keratocan core proteins bind the CXC chemokine KC during a corneal inflammatory response, indicating that corneal KSPGs mediate neutrophil recruitment to the cornea by regulating chemokine gradient formation. Together, these data support a significant role for lumican and keratocan in a corneal inflammatory response with respect to edema, corneal clarity, and cellular infiltration.

FIGURE 1. In vivo confocal microscopy analysis of stromal thickness (A, B) and haze (C, D) in wild-type, Kera^−/−, and Lum^−/− mice 24 h following a PBS or LPS (500 ng) intrastromal injection

KSPG-null corneas injected with PBS exhibited an overall thinner corneal stroma compared with wild-type controls (A). This difference in stromal thickness increased further following an LPS-induced inflammatory response (B), demonstrating a resistance to increases in stromal thickness. Corneal haze was not significantly different between Kera^−/− and wild-type controls, whereas Lum^−/− mice exhibited significantly greater corneal haze following PBS injection (C). Corneal haze in LPS-injected Kera^−/− mice did not differ from wild-type mice; however, Lum^−/− mice further increased an already apparent difference in opacity (D). Note the difference in y-axis scales. Error bars, standard error of the mean.

FIGURE 2. Kinetics of bone marrow-derived inflammatory cell infiltration into EGFP chimeric Lum^−/− and Kera^−/− mice inflammation

A, in vivo fluorescent stereomicrographs of B6/129 wild-type (B6/129), Kera^−/−, and Lum^−/− × EGFP chimeric mouse corneas 24 h after intrastromal injection of LPS (1 μg). B, quantitation of EGFP-positive cells in chimeric mouse whole flat-mount corneas. Error bars, standard error of the mean.

FIGURE 3. Impaired infiltration of neutrophils into Lumican and Kera^−/− mice corneas 6 and 24 h following LPS-induced inflammation

The total number of neutrophils infiltrating wild-type, Kera^−/−, and Lum^−/− corneas 6 (A) and 24 h (B) following intrastromal injection of 500 ng of LPS was quantitated as the number of NIMP-R14-positive cells/5-μm section. Error bars, standard error of the mean.

FIGURE 4. CXC chemokine expression in cornea of wild-type, Kera^−/−, and Lum^−/− corneas

KC (A) and MIP-2 (B) expression measured by enzyme-linked immunosorbent assay in KSPG-null and wild-type corneas 6 h following intrastromal injection of LPS (500 ng). Error bars, standard error of the mean.

FIGURE 5. Reconstitution of Kera^−/− and Lum^−/− mice corneas in vivo was performed by intrastromal injection of adenoviral constructs expressing human Lumican (Ad-hLum), keratocan (Ad-hKera), or an EGFP control (Ad-EGFP)

EGFP expression was quantitated at 18 h post-injection (A) and corneas imaged using fluorescent stereomicroscopy (B) to confirm transgene expression. Following confirmation of transgene expression, 500 ng of LPS was injected intrastromally and neutrophils were quantitated using NIMP-R14 immunohistochemistry at 6 and 24 h for Kera^−/− (C, E) and Lum^−/− (D, F). Error bars, standard error of the mean.

FIGURE 6

Immunoprecipitation of the CXC chemokine KC from corneal extracts 24 h after LPS (1 μg) intrastromal injection followed by SDS-PAGE on a 12.5% gel and Western blots (A) for keratocan (top) and lumican (bottom). Rat IgG2a control immunoprecipitated samples show the absence of keratocan and lumican signals. Extracts from corneas excised 24 h following LPS injection were digested with keratanase prior to immunoprecipitation, and an equal aliquot was immunoprecipitated without keratanase digestion. Immunoprecipitates were then subjected to SDS-PAGE on a 10% gel and Western blot (B) for keratocan (top) and lumican (bottom). Exposure times for Western blots in panel A were greater in order to demonstrate the absence of signal in the IgG2a control lanes.