Dexamethasone disrupts intercellular junction formation and cytoskeleton organization in human trabecular meshwork cells

Abstract

Purpose: Patients reproduce symptoms of primary open-angle glaucoma (POAG) when treated with glucocorticoids (GCs) topically on the eyes. Here we investigated the effects of GCs on junctional protein expression and cytoskeleton organization in primary human trabecular meshwork (TM) cultures to understand the molecular pathologies of POAG.

Methods: Human TM cells from POAG (GTM) and age-matched nondiseased (NTM) individuals were obtained by standard surgical trabeculectomy. Some of the cultures were treated with dexamethasone (DEX), a synthetic GC, at 1-5 x 10(-7) mol/l for 1-7 days. The expression levels of zonula occluden-1 (ZO-1) and connexin43 (Cx43) in TM cells with or without DEX treatment were measured using reverse transcription (RT)-PCR, immunocytochemistry, and western blot analysis.

Results: mRNA and proteins of ZO-1 and Cx43 were found in both NTM and GTM cells. ZO-1 and Cx43 were located on the plasma membrane, especially along the border of adjacent cells. ZO-1 had no marked changes in localization in NTM and GTM cells after treatment with 10(-7) mol/l DEX for 48 h, whereas Cx43 appeared to increase in the cytoplasm. mRNA of two ZO-1 isoforms, alpha+ and alpha-, were present in TM cells, and the former was expressed less than the latter. Only ZO-1 alpha- isoform protein was expressed in NTM cells, whereas proteins of both isoforms were found in GTM cells. DEX increased the protein levels of ZO-1 and Cx43 in both NTM and GTM cells. DEX also altered the F-actin architecture and promoted cross-linked actin network formation, the effects of which were more pronounced in GTM cells.

Conclusions: Our findings not only provide molecular insights to the pathogenesis of GC-induced glaucoma but also suggest that junctional proteins ZO-1 and Cx43 as well as F-actin are targets for developing new modalities in glaucoma therapy.

Figure 1

The morphology of trabecular meshwork cells before and after dexamethasone treatment. Normal human trabecular meshwork (NTM) and primary open angle glaucoma trabecular meshwork (GTM) cells were obtained and cultured under identical culture protocol. GTM cells were slightly larger compared to NTM cells. Cell morphology shows no significant changes after treatment with 10⁻⁷ mol/l DEX for 1 week (1W) compared to the untreated control. Scale bar=50 μm.

Figure 2

Trabecular meshwork cells express zonula occludens 1 and Cx43. NTM cells have lower zonula occludens 1 (ZO-1) α– isoform levels but higher connexin 43 (Cx43) levels compared to GTM cells, yet NTM and GTM cells have similar ZO-1 α+ levels, as illustrated by RT–PCR. GAPDH was used as the internal loading control. M stands for molecular size ladder. n=3.

Figure 3

Effects of dexamethasone on zonula occludens 1 and connexin 43 in normal trabecular meshwork cells. Fixed normal trabecular meshwork (NTM) cells were immunolabeled with zonula occludens 1 (ZO-1) and connexin 43 (Cx43) (green fluorescence) and observed under a confocal microscope, using identical parameters. The ZO-1 antibody recognizes both α+ and α– isoforms of ZO-1. Both ZO-1 and Cx43 were plasma membrane bound. Treatment with 10⁻⁷ mol/l dexamethasone (DEX) for 2 days resulted in no significant changes in expression and distribution of ZO-1. However, DEX increased the cytoplasmic pool of Cx43. Nuclei are shown by DAPI staining (blue fluorescence). The scale bar=20 μm.

Figure 4

Dexamethasone increases zonula occludens 1 and connexin 43 expression in trabecular meshwork cells. Western blot analysis showed the expression of zonula occludens 1 (ZO-1) and connexin 43 (Cx43) protein in trabecular meshwork (TM) cells. NTM cells express the ZO-1 α– isoform and Cx43 protein, whereas GTM cells express both ZO-1 isoforms and Cx43 protein. Treatment with 10⁻⁷ mol/l dexamethasone (Dex) caused a time-dependent increase in both ZO-1 and Cx43, with peak expression at 4 and 5 days, followed by a gentle decline at 6 days; however, the levels of these proteins were not at the basal level after 7 days. GAPDH was used as the internal loading control and n=3.

Figure 5

The cytoskeleton architecture of glaucomatous trabecular meshwork cells is tangled. The organization of the filamentous actin (F-actin; red fluorescence), and the distribution of vinculin (green fluorescence), the actin focal adhesion point on plasma membrane, were observed using confocal microscopy. The glaucomatous trabecular meshwork cells have a more irregular actin architecture and vinculin distribution compared to normal trabucular meshwork cells.

Figure 6

Treatment with 10⁻⁷ mol/l dexamethasone (DEX) for 7 days increased cross-linked actin networks formation in both normal trabecular meshwork and trabecular meshwork cells derived from POAG individuals cells. cross-linked actin networks (CLANs) were also observed in the untreated trabecular meshwork cells derived from POAG individuals (GTM) cells. Green squares indicate areas of enlargement. Nuclei were shown by DAPI staining (blue fluorescence).

Figure 7

Histograms represent the fold changes in cross-linked actin networks to cell number ratios. The ratio for untreated NTM cells was arbitrarily set as 1. DEX caused a significant increase in cross-linked actin network (CLAN) numbers in both NTM and GTM cells compared to their respective untreated control. The asterisk indicates data significantly different from the respective untreated control at p≤0.05; the scale bar is 20 μm.