Molecular Vision 2002; 8:32-44 <http://www.molvis.org/molvis/v8/a5/> Received 24 October 2001 | Accepted 18 February 2002 | Published 6 March 2002

Download Reprint

Gene transfer of dominant-negative RhoA increases outflow facility in perfused human anterior segment cultures

Jason L. Vittitow,¹ Rahul Garg,¹ Laura-Leigh S. Rowlette,¹ David L. Epstein,¹ Edward T. O'Brien² Teresa Borrás¹
 
 

¹Dept. of Ophthalmology, Duke University Medical Center, Durham, NC; ²Dept. of Biology, University of North Carolina, Chapel Hill, NC

Correspondence to: Teresa Borrás, Ph.D., Duke University Medical Center, Wadsworth Building, Erwin Road, Box 3802, Durham, North Carolina 27710; Phone: (919) 684-6639; FAX: (919) 684-8983; e-mail: borra001@mc.duke.edu

Abstract

Purpose: To investigate the regulation of expression and the role of the RhoA gene in the human trabecular meshwork (TM). To attempt to modulate outflow facility by gene transfer of the RhoA gene's dominant-negative mutant protein.

Methods: Total RNA extracted from cultured human trabecular meshwork (HTM) cells treated with outflow facility drugs were analyzed by northern blot hybridization using an amplified human RhoA cDNA from plasmid pZip-RhoA wild type (wt) [1]. A dominant-negative form of RhoA (single amino acid substitution of Thr19 to Asn) was placed under the control of the CMV promoter and inserted into a replication-deficient adenoviral vector by overlapping recombination (AdhRhoA2). AdhRhoA2 was infected into perfused anterior segment cultures from post-mortem human donors and HTM and Schlemm's canal cells in culture. Changes in outflow facility (flow/pressure) were calculated as percent changes from baseline values (C₀), pooled into treated and control groups and expressed as the mean ± standard error. HTM and Schlemm's Canal (SC) cells were fluorescently double-labeled for the RhoA protein and actin, paxillin, or ZO-1.

Results: Transcription of RhoA in HTM cells was not considerably affected by treatment of the cells with cytoskeletal/outflow facility drugs. At 66 h post-injection, anterior segments treated with AdhRhoA2 (n=9) exhibited an increase in outflow facility of 32.5±7.7% while that of the vehicle-injected controls (n=6) was 5.1±4.0% (p=0.02). HTM cells treated with AdhRhoA2 showed a marked change in morphology with a reduction in actin stress fibers and of the focal adhesion-containing protein, paxillin. Confluent monolayers of SC cells infected with AdhRhoA2 were devoid of peripheral ZO-1 staining indicating a loss of intercellular junctions.

Conclusions: In the HTM cells, cytoskeletal/outflow facility drugs do not seem to affect the levels of RhoA mRNA, possibly suggesting the importance of mRNA availability to allow rapid turnover of its function. Gene transfer of inactive RhoA to the intact human TM results in an increase in outflow facility. This increase appears to be correlated with a loosening of the cell-substrate and cell-cell attachments in the cells of the outflow pathway. Adenoviral vectors carrying the dominant negative form of RhoA could potentially be utilized as a gene therapy to modulate outflow facility.

Introduction

Glaucoma, an optic neuropathy most often caused by impairment in drainage of aqueous humor through the trabecular meshwork (TM), causes irreversible blindness in more than 2 million people in the United States and more than 60 million people worldwide. Elevated intraocular pressure (IOP) is the most important risk factor for the development of glaucoma, and reduction of IOP is the only available treatment that lowers the rate of visual field loss [2].

The trabecular meshwork, an avascular, multilaminar tissue located circumferentially in the anterior chamber angle extending from the scleral spur to Schwalbe's line, generates resistance to aqueous humor outflow thus creating an IOP [3]. The tissue consists of different endothelial cell regions, each arranged in a characteristic architecture. The inner uveal and corneoscleral regions, closer to the anterior chamber, are comprised of cells that line closely collagen-elastin trabecular beams, forming a net-like structure and leaving different sized openings in between. The deeper juxtacanalicular region (JCT) is composed of cells that lay relatively free in the area just proximal to Schlemm's canal (SC). Finally, the cells of the inner wall of SC are tightly connected to each other forming what looks to be a barrier to the aqueous humor entering the canal. These SC cells are known to be of vascular origin [4]. The whole TM tissue is embedded in a continuously changing extracellular material and it is exposed to numerous aqueous humor factors capable of triggering many signal transduction pathways. Experimental evidence has indicated that the main site of outflow resistance, both in normal and primary open angle glaucoma, appears to be at, or just proximal (JCT) to the inner wall of SC [5]. However, many studies have indicated that TM cells might contribute to the maintenance of the resistance through a number of biological mechanisms such as changes in cytoskeletal organization [6], phagocytosis of debris [7], secretion of extracellular matrix (ECM) components [8], and/or ability to respond to several types of stress.

Currently, most prescribed medications act to lower elevated IOP by either suppressing aqueous humor production at the ciliary processes or by increasing uveoscleral outflow through the ciliary body. In the laboratory, pharmacological agents that directly affect the cytoskeletal network, have been shown to affect outflow facility [6]. Drugs such as cytochalasin B [9], ethacrynic acid (ECA) [10], vinblastine [11], H7 [12], latrunculin B [6] and 2,3-butanedione 2-monoxime (BDM) [12] produce marked effects on cell shape and outflow facility.

The Rho family of proteins belongs to the larger Ras superfamily of GTP-binding proteins and has been shown to regulate a number of cytoskeleton-dependent cell functions. Rho itself induces formation of stress fibers and re-organization of focal adhesions [13] in epithelial [14] and endothelial cells [15]. Focal adhesions, which consist of a number of proteins, including paxillin, serve as anchoring points between actomyosin filaments and the ECM. The actions of RhoA are caused in part by increasing the phosphorylation of the myosin light chain (MLC), which in non-muscle cells leads to actomyosin contraction [16]. RhoA cycles between an active, GTP-bound form, and an inactive, GDP-bound conformation. Cycling between the two conformations is regulated by guanine nucleotide exchange factors (GEFs), which bind to inactive RhoA to promote the release of GDP and allow GTP to bind, and by GTPase activating proteins (GAPs), that stimulate hydrolysis of GTP to yield GDP [17]. The rapid transformation of RhoA from its inactive to its active form allows for the transient activation of Rho kinase (ROCK) [18], which acts by both directly phosphorylating MLC as well as by inhibiting myosin phosphatase [18,19].

Importantly, RhoA is also known to affect endothelial barrier function, which is dependent on the integrity of intercellular junctions and actomyosin contractility [20,21]. However, Rho regulation of adherens junctions appears to differ between cell types. Whereas inhibition of Rho in epithelial cells leads to a reduction of E-cadherin-containing junctions [22], the same inhibition does not perturb the localization of VE-cadherin in endothelial cells [23]. In epithelial cells, expression of either dominant-negative or constitutively active RhoA perturbs tight junction structure [24].

Given the observed importance of the actin cytoskeleton in influencing outflow facility, and the well-established role of RhoA on cytoskeletal organization, we hypothesized that RhoA might play an important role in the modulation of aqueous humor outflow. To investigate this role, we first looked at the expression of the RhoA gene in human TM (HTM) and SC cells exposed to drugs known to affect outflow facility. Second, we examined the effect of inhibiting RhoA function in outflow facility. Other investigators have studied the ROCK inhibitor Y27632 in rabbit [25] and porcine models [26], where it was found to increase outflow facility, and in bovine preparations, where it decreases TM contraction leading to increased facility [27].

Because of potential problems associated with multiple targets of conventional pharmacological agents, here we sought to directly inhibit RhoA protein function by gene transfer of its dominant-negative form to the TM. In addition, because the aqueous humor outflow pathways of lower mammals are quite different from those of humans [28], we used intact perfused human TM from post-mortem donors, and HTM and SC cell cultures. We searched for a correlation between inactivation of RhoA and cytoskeletal markers for cell-substrate adhesion and cell-cell interactions. Finally, both because of the inability of transfecting plasmids to intact human TM, and the potential application of RhoA gene therapy for the treatment of glaucoma, we performed this study using a replication-deficient adenoviral vector to transfer the dominant-negative RhoA protein (AdhRhoA2).

We found that, in the HTM cells, RhoA mRNA levels do not appear to be influenced by drugs that induce large changes in the actin cytoskeleton. A single injection of AdhRhoA2 efficiently delivered the dominant-negative protein to the intact human TM and caused an increase in outflow facility. HTM and SC cells infected with AdhRhoA2 exhibited a lower number of focal contacts and the disappearance of ZO-1 protein from cellular junctions. These results suggest that the functional role of RhoA in the human TM involves cellular relaxation mechanisms accompanied by loosening of both cell-ECM and cell-cell interactions. Further understanding of this mechanism might lead to the development of gene transfer techniques that could modulate outflow facility.

Methods

Culture of outflow facility cells

Eyes from non-glaucomatous human donors were obtained within 48 h of death from national eye banks. For isolation of HTM cells, the trabecular meshwork from a single individual was isolated from surrounding tissue by making incisions both anterior and posterior to the meshwork and removing it using forceps. The tissue was then cut into small pieces, treated with 1 mg/ml collagenase in phosphate buffered saline (PBS) and incubated at 37 °C in a shaker water-bath for 1 h. Incubation was followed by low speed centrifugation for 5 min. Pellets were resuspended in 5 ml of Improved Minimal Essential Medium (IMEM; Biofluids, Rockville, MD) supplemented with 20% fetal bovine serum (FBS) and 50 mg/ml gentamicin (GIBCO BRL, Rockville, MD). Resuspended tissue was plated on a single, 2% gelatin coated 35 mm well and maintained in a 37 °C, 7% CO₂ incubator. Once confluent (2-3 weeks), cells were passed to a T-25 flask and labeled as passage 1. Subsequently, cells were passed 1:4 at confluency and maintained in the same medium with 10% serum. These non-transformed cells subsist for nine to ten passages. Cells used in these experiments originated from five different donors (ages 18 to 72 years old) and they were used at passages 4 to 7.

For isolation of the SC cells, we used the explant method as previously described [29]. Briefly, after cleaning, sterilizing, and opening each globe, the lens and iris were removed and the anterior segment placed face down in a shallow petri dish. PBS was used to keep the tissue hydrated during dissection. Fine forceps and a scalpel blade were used to remove the uveal and corneoscleral TM. Radial cuts, about 0.8-1.2 cm apart were made across the SC and the JCT-TM section (with a minimum of inner wall endothelium) was carefully removed. Shallow circumferential cuts on both sides of SC were usually necessary to remove all JCT layers. Finally, to remove the canal a deeper incision into the sclera was made circumferentially between the canal and the iris root followed by a clean cut "underneath" (as viewed from inside the eye). The canal and a minimal amount of surrounding tissue were then transferred to a 3 cm plastic culture dish, placed opened face down, and covered with a sterile, uncoated cover slip. Culture medium with FBS and antibiotics were then added, and the tissue placed in an incubator until cells began growing from the explant (mean time to first cells was less than 2 weeks). Only cells that emerged from the canal surface onto the plastic dish were cultured. Any cells that grew onto the cover slip surface were discarded. Once cells had grown to about 1 cm in diameter, the tissue was removed. After filling the 3 cm dish, the cells were passaged using trypsin-EDTA into T-25 culture flasks until passage 3, when they were frozen back until needed. Cells were judged to be SC by morphologically comparing each culture to a video record of all cultures derived by this method. These characteristics most commonly included an elongated, fusiform shape, and relatively small size. The cells formed extensive cell-cell contacts at confluence. The SC cells also grew more rapidly than HTM cells and, due to their fusiform shape and tendency to make extensive intercellular junctions, they would arrange themselves in regular patterns. As shown in O'Brien et al. [30], SC cells exhibited a primary cilium. In this study, to allow formation of cell-cell junctions, SC cells were grown to confluency and maintained confluent for four days before performing experiments.

Drug treatments

For outflow-drug treatment experiments, cultured HTM cells were grown to confluency in complete IMEM (see above) and exposed to the drugs as follows. Treatments with Ethacrynic Acid (ECA; Sodium Edecrin, Merck, Westpoint, PA) and Dexamethasone (DEX; Sigma, St. Louis, MO) were conducted at the final concentrations of 0.25 mM for 2.5 h (serum-free medium) and 0.1 mM for 12 days respectively (serum-containing medium). ECA was reconstituted at 10 mM with PBS just before use then diluted 40 fold into serum-free IMEM. DEX was prepared in absolute ethanol at 0.1 mM and diluted 1000 fold into fresh complete IMEM every other day for the duration of the experiment. Treatments with vinblastine (Sigma, St. Louis, MO), latrunculin B (Calbiochem, San Diego, CA) and H7 (Sigma, St. Louis, MO) were perfomed under three different conditions: serum-starved, serum-free, and in the presence of serum. For the no serum conditions, cells grown in complete IMEM (10% FBS, see above) were washed 3X with serum-free IMEM medium and either maintained without serum for 24 h before treatment (serum-starved) or immediately treated with the drugs in the absence of serum (serum-free). For the serum conditions, complete IMEM was replaced with the same medium containing the drugs. In most cases, drugs were prepared by reconstituting the reagents directly in the commercial bottle. The vinblastine solution was prepared in water at 1 mM and diluted 20 fold (final 50 mM) in the corresponding medium. Latrunculin B was prepared in absolute ethanol at 1 mM and diluted 1000 fold (final 1 mM) into the corresponding medium. H7 was prepared in PBS at 9 mM and diluted 30 fold into the corresponding medium (final 300 mM). All concentrated stocks were prepared fresh each time and discarded after each experiment. The drugs were exposed to the cells for the period of time described in the results.

RNA extraction and northern blot hybridization

Total RNA from cultured HTM cells was extracted using an RNeasy kit (QIAGEN, Valencia, CA) following manufacturer's recommendations. For the organ culture, anterior segments were frozen in liquid nitrogen within two minutes of turning off the perfusion pumps and stored at -80 °C. TMs were later obtained under a dissecting microscope from the frozen anterior segments before the complete thawing of the specimen. The isolated tissue was placed into a 1.5 ml microcentrifuge tube containing 350 ml of guanidine thiocyanate buffer, homogenized with a disposable sterile pestle and loaded onto a QIAshredder column (QIAGEN, Valencia, CA). Extraction continued using the RNeasy kit and RNA molecules selectively bound to the silica gel base were eluted with 30 ml of RNase-free water.

Samples containing 10 mg of total RNA from HTM cells or the total RNA recovered from individual trabecular meshworks were lyophilized to dryness, denatured in 50% formamide, and separated by 2.2 M formaldehyde, 1.25% agarose, 0.05 M 3-(N-morpholino)propanesulphonic acid (MOPS), 1 mM EDTA gel electrophoresis. After running, gels were washed with dH₂O for 30 min and transferred to a nitrocellulose membrane (Schleicher & Schuell, Keene, NH) overnight by capillary action with 10X sodium saline citrate (SSC). After the transfer, UV cross-linked blots were pre-hybridized at 42 °C for 6-24 h in a buffer containing 50% formamide, 5X SSC, 0.5X Denhardt's, 50 mM sodium phosphate pH 7.4, 0.1% sodium dodecyl sulfate (SDS) with 50 mg/ml sheared, denatured salmon sperm DNA (Research Genetics, Huntsville, AL). The filters were then placed in fresh buffer, hybridized overnight to 1-5x10⁶ counts per minute (cpm)/ml of random primed labeled RhoA cDNA (Boehringer Mannheim, Indianapolis, IN) at the same temperature. Human RhoA cDNA was obtained by polymerase chain reaction (PCR) amplification of the plasmid pZip-RhoA (wt) [1] insert using forward (Number 121; 5'-GCGCAAGCTTATGGCTGCCATCCGGAAGAA-3') and reverse (Number 122; 5'-ATATGCGGCCGCTCACAAGACAAGGCAACC-3') primers, followed by gel purification of the fragment using a QIAquick PCR purification kit (QIAGEN, Valencia, CA). After hybridization, filters were washed five times (15-20 min each): four times in 2X SSC-1% SDS (two at room temperature (RT) and two at 52 °C) and once in 2X SSC at 52 °C. Exposure was conducted using BioMax MR X-Ray film (Scientific Imaging Systems, Eastman Kodak, New Haven, CT) at -80 °C with intensifying screens. To monitor RNA degradation and loading, filters were subsequently re-hybridized to 2x10⁶ cpm/ml of a g-³²P ATP end-labeled (T4 Polynucleotide Kinase kit, GIBCO BRL, Rockville, MD) human ribosomal 28S oligonucleotide probe [31] at 42 °C for 2 h, washed twice in 2X SSC-1% SDS (one at RT, one at 37 °C) and exposed at RT. The probe for the gene encoding TIGR/MYOC was obtained by amplification of plasmid pJH5 [32] using forward (Number 98, 5'GCGAAAGCTTTCCAGAGGAAGCCT3', (42-62 nt TIGR/MYOC cDNA)) and reverse (Number 99, 5'CCAGGATCCCTGAGCATCTCCTT3' (1645-1623 nt TIGR/MYOC cDNA)) primers. Images of films were captured using an Arcus II scanner (AGFA division, Bayer Corporation, Wilmington, MA) and hybridization intensities determined by scanning densitometry using the Scion Image (Scion Corporation, Frederick, MD).

Anterior segment perfusion system

Organ cultures were prepared within 30 to 40 h of death as described previously [33,34]. Post-mortem human eyes were obtained from the North Carolina and Old Dominion Eye Banks with the signed consent of the patient's family and following the Tenets of the Declaration of Helsinki. The ages of the donors were between 65-85 years of age (average age 74 years) and none had been diagnosed with glaucoma. Eyes were dissected at the equator and their lens, iris and vitreous were removed. The anterior segments were clamped to a two cannula, modified petri-dish and perfused at constant flow with Dulbecco's modified Eagle's medium containing 4.5 mg/ml glucose (GIBCO BRL), 100 U/ml penicillin, 100 mg/ml streptomycin, 170 mg/ml gentamicin and 250 mg/ml amphotericin B. Organ cultures were maintained at 37 °C, 5% CO₂. Flow rates were between 3 to 4 ml/min and were maintained constant in each case with a Harvard microinfusion pump (Harvard Bioscience, S. Natick, MA). Intraocular pressures were continuously monitored with a pressure transducer connected to the dish's second cannula and recorded with an automated computerized system. Viability of the TM after perfusion was assessed in representative pairs by either morphological examination (light and/or electron microscopy) or biochemical parameters (quality of their RNA or protein).

Recombinant adenoviruses

AdhRhoA2, the replication-deficient recombinant adenovirus encoding the full-coding mutated RhoA cDNA driven by the cytomegalovirus (CMV) promoter was obtained by overlap recombination. The RhoA cDNA contains a point mutation converting the wild-type amino acid 19 from Threonine to Asparagine. This mutant cDNA renders the GTP binding site of the encoded protein inactive thus resulting in a dominant-negative RhoA. The mutated RhoA cDNA originated from plasmid pZIP-RhoA(19N) [1] by amplification of its insert with forward and reverse primers Number 121 and Number 122 (see above). These primers were designed to contain Hind III and Not I sites on their 5' end respectively. Gel purified RhoA insert was digested with Hind III and Not I and cloned into the shuttle vector pGEM-CMV [35]. This vector contains human adenovirus serotype 5 (Ad5) sequences 1-194 (inverted terminal repeat, ITR) that would provide the adenovirus left terminus and Ad5 sequences 3328-5781 for overlap recombination. The resulting recombinant shuttle plasmid, pRG1, was linearized with Age I and co-transfected with an Ad5 arm into 293 cells by calcium phosphate/DNA co-precipitation. The viral arm was obtained from Ad.Pacbgal [35] by digestion with Xba I/ Cla I and isolation of the 27 kb fragment by b-Agarase I (New England Biolabs, Beverly, MA). DNA precipitates were exposed to the cells for 12 h, washed exhaustively and allowed to recombine for two weeks. After recombination, harvested cells were lysed and their supernatant assayed for plaque purification by agar overlay [36]. Recombinant viruses were obtained from well-separated plaques, amplified, and tested for the presence of RhoA cDNA by PCR of the purified viral DNA. The RhoA insert of one positive clone was fully sequenced in both orientations. High-titer viral stocks were obtained by propagation in 293 cells and twice purified by CsCl density centrifugation as described [34,36]. Purified viruses were titered by plaque assay in 293 cells and stored in aliquots at -80 °C. Viral titers were typically between 1-5x10¹⁰ particle forming units (pfu). A map of the shuttle vector and a diagram of the construction are shown in Figure 1.

AdenoGFP is a first-generation replication-deficient adenoviral vector derived from Ad5 with deletions in the E1a and E3 regions. The virus carries a variant of the jellyfish Aequorea vitoria green fluorescent protein (GFP) cDNA. The cDNA of this transgene is under the control of the CMV promoter and has a number of mutations that result in modifications of the spectral properties of the encoded fluorescent protein. The vector was obtained from Qbiogen (Montreal, Canada) and grown and purified in our laboratories.

Delivery of recombinant adenoviruses to cells and organ cultures

For the tissue culture, cells seeded on 6-well dishes and maintained in IMEM supplemented with 10% FBS, 50 mg gentamicin were allowed to reach 80% confluency (HTM cells) or to be confluent for a period of four days (SC cells). For immunofluorescence studies, cells were grown on glass cover slips pre-coated with poly-D-lysine (Sigma, St. Louis, MO). Before infection, cells were extensively washed with PBS and incubated for 1 h with 100-500 pfu of AdhRhoA2 virus in 1 ml of serum-free IMEM. Viral absorption continued for 18 h in the presence of either 1% serum (SC cells) or serum-free media (HTM cells). After viral exposure, media was changed and cultures evaluated 48 h post-infection. Control dishes were treated with AdenoGFP at the same pfu/ml or with the same volume of viral vehicle under identical conditions.

For organ cultures, paired eyes were perfused for an initial period of 24 h with standard medium, which allowed stabilization of outflow facility. Then, perfusion pumps were stopped, pressure allowed to decrease to <5 mm Hg and eyes injected through the cornea with 20 ml of the viral solution (10⁷ pfu) using a 50 ml Hamilton syringe and a 1 cm beveled 30G disposable needle. The dish was then rocked gently two-three times to obtain mixing of the sample inside the chamber, pumps restarted and perfusion continued for 48 to 72 h post-injection. Contralateral eyes were injected with 20 ml of virus vehicle. In one pair, the delivery of Ad.Pacbgal and vehicle was done by anterior chamber exchange. In this case, 20 ml of the given viral dilution or vehicle were diluted into 1 ml of perfusion medium and delivered into the incoming perfusion tubing through a 3-way stopcock.

Light Microscopy

Two pairs of eyes were used for used for histologic examination. Wedge shape specimens from each of two quadrants containing the angle region with the TM were fixed in 4% paraformaldehyde, 2% glutaraldehyde and embedded in glycol methacrylate using the JB-4 embedding kit from Polysciences (Warrenton, PA). Blocks were cut in sections 2 mm thick followed by counterstaining with hematoxylin and eosin (H & E). Representative sections were examined by light microscopy particularly for the presence of cells and preservation of TM architecture.

Outflow facility measurements of the human organ cultures

Pressure values were recorded just prior to injection (baseline) and every 6 h post-injection for a period of 48 to 72 h. Outflow facility (C), also defined as the inverse of the resistance (R), C=1/R, was calculated as the rate of the flow (F) divided by the intraocular pressure, P (Goldmann's equation, C=F/P). Facility was measured in ml/min/mm Hg. Baseline facility (C₀) was calculated from the average of three values obtained from pressure readings recorded at 30 min intervals just before treatment. For C, outflow facility was calculated with pressure values obtained every 6 h post-treatment. Data from each experiment was calculated as percent change of facility from the baseline, combined into a treated and control groups and expressed as the mean ± standard error. Experimental effects were statistically obtained by comparing the percent change in outflow facility of the experimental eyes to the percent change in facility of the controls by the student t-test.

Protein extraction and Western blotting

After viral treatment, HTM cells were washed twice with PBS and harvested in 40 ml of loading buffer (100 mM Tris-HCl, 200 mM DTT, 4% SDS, 20% glycerol and 0.02% bromophenol) with a cell scraper. Proteins were extracted by boiling the samples for 10 min followed by centrifugation at 16,000 g. Supernatants were separated by 12.5% SDS-PAGE and electrophoretically transferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad, Hercules, CA). After blocking with 5% non-fat dry milk in PBS-0.2% Tween 20 (Sigma, St. Louis, MO) for 2-4 h, membranes were washed twice with PBS-0.2% Tween 20 and incubated overnight at 4 °C with a rabbit anti-human RhoA antibody (Santa Cruz Biotechnology, Santa Cruz, CA), diluted 1:1000, followed by treatment with anti-rabbit IgG conjugated to horseradish peroxidase (Pierce, Rockford, IL), diluted 1:6000, for 1 h and washed as above. Immunoreactive bands were visualized by chemiluminescence using ECL Plus (Amersham, Piscataway, NJ) and exposure to Kodak BioMax MR film (Eastman Kodak, Rochester, NY). For organ cultures, the dissected, perfused human TM was washed with PBS, briefly centrifuged at 180 g and resuspended in 20 ml of loading buffer. Tissues were homogenized using a Microson Ultrasonic XL2000 cell disruptor equipped with a 2.4 mm microprobe (Misonix Inc., Farmingdale, NY). The samples were then centrifuged at 16,000 g for 10 min and their supernatants loaded on a 12.5% SDS-PAGE gel, transferred, washed and stained as described above.

Immunofluorescence

Cells were cultured on glass cover slips pre-coated with poly-D-Lysine, fixed and fluorescently double labeled for the RhoA protein and either actin, paxillin or ZO-1. Cells were washed twice with warmed serum-free IMEM, fixed with warmed 3.7% paraformaldehyde for 10 min, washed twice with DPBS, permeabilized with 0.1% Triton-X/DPBS for 10 min and blocked for 20 min with 10% goat serum/DPBS. Cells were again washed with DPBS and incubated at RT with rabbit anti-human RhoA antibody (1:200) for 2 h, followed by an additional 1 h with a FITC-conjugated goat anti-rabbit secondary antibody (1:200; Santa Cruz Biotechnology, Santa Cruz, CA). For double labeling with actin, phalloidin-TRITC (Sigma, St. Louis, MO) was added to the secondary antibody solution at a concentration of 0.5 mg/ml. For double-labeling with paxillin, mouse anti-human paxillin (Transduction Laboratories, Lexington, KY) was added to the RhoA primary antibody solution (1:200) and horse anti-mouse Texas Red-conjugated antibody (Vector, Burlingame, CA) was added to the secondary antibody solution. For double-labeling with ZO-1, cells were incubated simultaneously with a mouse anti-human Rho antibody (1:200; Santa Cruz Biotechnology, Santa Cruz, CA) and a rabbit anti-human ZO-1 diluted 1:200 (Zymed, San Francisco, CA) followed by incubation with Cy2 conjugated donkey anti-mouse diluted 1:100 (Jackson ImmunoResearch Laboratories, West Grove, PA) and Cy3 conjugated donkey anti-rabbit secondaries diluted 1:400 (Jackson ImmunoResearch Laboratories). All antibody solutions were made in 1.5% serum/DPBS and washes performed between all incubation steps (3 times, 5 min each) with DPBS. Cover slips were mounted with 20 ml Fluoromount G (Southern Biotechnology Associates, Birmingham, AL) except for Cy2 and Cy3 labeled cells, which were mounted with 50% glycerol/PBS and sealed with clear enamel. All secondary antibodies were tested for cross-reactivity in each cell type by incubating in the absence of the primary antibodies. Fluorescence was visualized using a ZEISS Axioplan epifluorescence microscope with Plan-Neofluar objectives. Each slide was thoroughly examined and photographs were taken from representative fields. Images were either photographed using FujiChrome Provia 400F slide film (Fuji Film, Tokyo, Japan) or captured with a Zeiss AxioCam CCD camera and analyzed using AxioVision digital imaging sowtware and Adobe Photoshop (Adobe Software, Mountain View, CA).

Results

Effect of outflow facility drugs on the expression of RhoA in HTM cells

To investigate the contribution of RhoA in TM physiology, we first examined the expression of the RhoA gene in cells of the outflow pathway. Total RNA extracted from cultured HTM and SC cells was analyzed by northern blot hybridization using, as a probe, the amplified human RhoA cDNA from plasmid pZip-RhoA using primers Number 121 and Number 122 (see methods). Figure 2 shows that RhoA transcripts are abundantly present in HTM and SC cells. The number and size of these transcripts (1.8 and 1.5 kb) are similar to those present in other eukaryotic cells [37]. Interestingly, drugs known to affect cytoskeletal organization and outflow facility had a minimal effect on the levels of RhoA mRNA. Cells exposed to 50 mM vinblastine, 1 mM latrunculin B, or 300 mM H7 for 1 h under serum-starved conditions showed little change in RhoA gene expression (Figure 2). Upon scanning, the intensity of the hybridization bands revealed a RhoA/28S ratio of the treated versus a RhoA/28S ratio of the control to be 0.99 for vinblastine, 0.83 for latrunculin B, and 0.54 for H7. Experiments were also performed exposing the cells to the same drugs for longer periods of time (6 and 12 h) and under non-starved conditions, with and without the presence of serum. While the widely described changes in cell shape were obvious, the expression of the RhoA gene was not affected under any of these conditions (not-shown). Likewise, cells exposed to ECA (0.25 mM, 2.5 h) in serum-free conditions and to DEX (0.1 mM, 12 days) under serum-containing conditions did not result in evident changes of RhoA expression (Figure 2). The RhoA/28S ratio of the treated versus the RhoA/28S ratio of the control was 0.98 for the HTM cells and 0.91 for the SC cells. In the DEX treated HTM cells, the same ratios of the treated versus the control were 0.73 for RhoA and 4.31 for Tigr/Myoc.

Delivery of AdhRhoA2 to cultured cells and intact trabecular meshwork tissue

A replication-deficient adenovirus (AdhRhoA2) carrying the full coding cDNA region (582 nt) of the mutated RhoA protein (Thr19 to Asn) was constructed as indicated in Methods. The single amino acid substitution in the dominant-negative RhoA protein promotes a tight binding of the mutant protein to GEFs, sequestering these proteins and preventing them from activating either the mutant or the wild type endogenous RhoA. To first test the positive delivery of the mutant protein to the cells of the outflow pathway, we analyzed the presence of the recombinant RhoA mRNA and protein in infected cultured cells and perfused human anterior segment cultures. HTM cells at 70-80% confluency were infected at increasing multiplicity of infection of 10 and 100 pfu/cell and examined for the presence of RhoA mRNA and protein 48 h post-infection. Figure 3 show the northern and western blots where a dose-dependent increase in RhoA message and protein was observed in the infected HTM cells. The virus was also able to penetrate, with great efficiency, the intact human trabecular meshwork after a single-dose injection though the cornea in the perfused model system. One anterior segment from each pair of eyes was injected with 10⁷ pfu of AdhRhoA2 while the contralateral eye was injected with virus vehicle. A northern blot run with the total RNA from the dissected TM revealed a considerable increase of recombinant RhoA mRNA in the injected eye (Figure 3A). RhoA protein was also clearly detected in the infected TM tissue while it was barely seen in the uninfected one (Figure 3B).

Morphology of the perfused trabecular meshwork injected with either AdhRhoA2 or vehicle

Light microscopy photographs from one eye pair injected with either AdhRhoA2 or viral vehicle are shown in Figure 4. The anterior segments were obtained from an 86 year-old female and perfused for 96 h post-injection. The architecture of the TM tissue was well preserved in both the viral and the vehicle injected specimens. A good number of cells are present in all three regions with no apparent difference between treated and untreated eye. In the virus-injected tissue, a slight difference in the packing of the trabecular beams as well as a somewhat narrower SC could be observed.

Dominant-negative RhoA induces increased outflow facility in perfused post-mortem human anterior segments

Baseline outflow facilities for the eyes in this study (n=9 for the infected eyes and n=6 for the controls) ranged from 0.10 to 0.36 for the viral and 0.09 to 0.28 for the vehicle, with an average of 0.20±0.02 (n=15). The pressures of these eyes at baseline were between 11-32 mm Hg for the viral and 12-32 mm Hg for the vehicle injected eyes. Eyes that did not have a stable baseline or facilities between 0.06 and 0.4 were excluded. Six out of the nine infected eyes and four out of the six vehicle treated eyes were perfused for 72 h after infection, while the remaining eyes were perfused for either 42 or 66 h. Figure 5 shows the percent change in outflow facility from baseline of those eyes perfused for 72 h (n=6 for the infected eyes and n=4 for the controls). Values were computed every 6 h, from 6 h to 72 h post-infection, on eyes injected with a single dose of AdRhoA2 and vehicle injected controls. At 6 h, the percent change of C (C/C₀) of the dominant negative RhoA infected eye was 1.4±6.3% versus 3.4±4.0% in the control eyes. The outflow facility of the treated eyes increased moderately with time while that of vehicle injected eyes remained stable (Figure 5). The mean difference between treated and untreated eyes became significant at 66 h post-infection with values of 32.5±7.7% and 5.1±4.0%, respectively (p=0.02). At 72 h post-infection, the percent change from the baseline was of 34.1±9.1% versus 4.2±3.01% in the controls (p=0.03).

A second calculation was performed including all eyes that had been perfused for shorter time periods (different n at different time points). Results were similar to those obtained above. Higher outflow facility values in the treated eyes were first observed at 42 h post-infection, and the difference of the means between those expressing dominant negative Rho and the controls was significant at 72 h (p=0.03; graph not shown).

Perfusion experiments with a recombinant adenovirus carrying a reporter gene did not produce the increase in outflow facility obtained with AdhRhoA2. Eyes injected with a single dose of Ad.Pacbgal carrying the E. coli LacZ gene [35] showed a very small change in outflow facility with respect to their baseline or to their contralateral controls injected with vehicle (6.4±4.1% vs. 0.0±4.3%; n=9 for the infected eyes and n=10 for the controls) at 48 h. At 72 h, the outflow facility values of the reporter virus were 7.4±3.3% versus -10.7±12.8% of those of the vehicle control (n=4 for the infected eyes and n=3 for the controls).

Effect of dominant negative RhoA overexpression on the actin cytoskeleton and focal adhesions of HTM cells

Figure 6 shows the actin network of cells exposed to AdhRhoA2, and to the two controls, AdenoGFP and viral vehicle. Cells infected with the recombinant adenovirus expressing dominant-negative RhoA (AdhRhoA2) show significant changes in cell shape, exhibiting a more flattened shape and cell separation. A reduction in the amount of stress fibers, which appeared to correlate with the intensity of RhoA staining, was also observed (top panels). Cells showing a higher level of expression of the dominant-negative RhoA (more intense staining) appeared to lose cortical actin and develop small bundles of short actin filaments, which form a zigzag pattern throughout the cell. Subconfluent untreated trabecular meshwork cells kept in serum-free medium for 24 h exhibit a well-formed actin cytoskeleton. Numerous actin stress fibers, visualized by TRITC-phalloidin, crossed the body of the cells forming a dense filamentous network (Figure 6, vehicle treated). Infection with the reporter adenovirus encoding GFP had no effect on cell shape or on the formation stress fibers (Figure 6, AdenoGFP).

Figure 7 shows the effect of AdhRhoA2 and two controls (vehicle and AdenoGFP) on the focal adhesions of HTM cells (visualized by immunofluorescence labeling with an anti-paxillin antibody). Paxillin was diminished centrally in cells infected with AdhRho2 with some remaining at the cell's periphery (top panels). This would suggest that reduction of active Rho molecules in HTM cells has an effect on the amount and distribution of focal contacts. Uninfected cells contained abundant focal adhesions that were evenly distributed toward the lower surface of the cell (middle panels). Infection with AdenoGFP had a slight effect on the abundance of focal contacts (lower panels) but clearly different when compared to the AdhRhoA2 infected cells (top panels). In all cases, no fluorescence was observed when HTM cells were stained without the use of the primary antibodies.

The observed data are consistent with an involvement of RhoA inactivation in causing relaxation of cells of the corneoscleral and JCT regions of the intact human TM tissue and potentially influencing outflow facility.

Dominant-negative RhoA alters intercellular attachments in Schlemm's canal cells

Because the endothelial cells of the inner wall of the SC form a tight monolayer and because the expression of dominant negative RhoA increased outflow facility in perfused human anterior segments, we analyzed its effects on intercellular attachments of cultured SC cells. After untreated SC cells were maintained confluent for four days, ZO-1 antibody was clearly located at the cell membrane junctions between untreated cells (Figure 8, middle panels). This staining did not appear as a continuous line but rather as a segmented border. Cells infected with RhoA were devoid of this ZO-1 staining pattern at the intercellular borders and appear somewhat separated from each other (Figure 8, top panels). In contrast, cells infected with AdenoGFP (Figure 8, lower panels) retained the ZO-1 staining at the junctions and appear very similar to those treated with vehicle. No fluorescence was observed when SC cells were stained in the absence of primary antibodies.

Discussion

One of the ways by which the trabecular meshwork appears to influence the resistance to aqueous humor outflow is by changing the cytoskeletal organization of its cells, the quality of its cell-substrate adhesions, and the apparent tightness of its intercellular junctions. In this study, we report that the expression of the actin regulator gene RhoA in the human TM is not influenced by treatments with the cytoskeletal modifying drugs vinblastine and ECA and reduced a little by treatments with latrunculin B and H7. We also have learned that using adenoviral gene transfer of the dominant-negative conformation of the protein RhoA, we are able to induce an increase in outflow facility in human perfused anterior segment cultures.

Little is known regarding the expression of RhoA. Although numerous studies have identified active RhoA as a signal transduction molecule involved in cytoskeletal organization [13,17,38], to our knowledge, none of these studies has addressed the question of whether the expression of the RhoA gene increases or decreases during these changes. Here we show that RhoA mRNA is very abundant in the human TM suggesting that the gene plays an important role in TM physiology. In addition, we found little or no changes in mRNA levels under conditions of profound morphological changes. There was no evidence for a differential utilization of its two polyadenylation sites.

The continuous presence of high levels of mRNA in the HTM cells could be the result of either a very stable transcript or a constitutive expression of the gene. It is also possible that the expression of the gene could be the result of a feedback mechanism in response to a decrease in the amount of active protein inside the cell. However, the fact that cytoskeletal drugs did not considerably affect the expression of RhoA might be an indication that the HTM cell regulates RhoA function by activating and inactivating the protein, rather than by changing the expression of the gene. One could also speculate that the action of these drugs on cytoskelestal organization is either downstream or independent of the RhoA pathway. In contrast, RhoA mRNA levels are important for RhoA's cellular transformation ability and the RhoA gene has been shown to be over-expressed in breast, lung, and colon carcinomas [39].

Because cultured endothelial cells do not transfect well, and because of the importance of maintaining minimal interference in the outflow pathway by the use of transfecting reagents, we have used recombinant adenoviruses. As we have previously shown with other adenoviral vectors [34,40], in the current study we observed that AdhRhoA2 delivered the dominant-negative RhoA protein very efficiently to the intact human TM. Both northern and western blots indicate that RhoA mRNA and protein are abundantly made in the cells of the human TM after a single injection of the recombinant. We also have observed that inactivation of RhoA produced an increase in outflow facility. The difference between treated and untreated eyes became significant at 66 h post-infection (Figure 5) at a time when presumed high levels of dominant negative mRNA were present. At the same viral concentration, an adenoviral vector carrying the LacZ reporter gene showed a slight increase in outflow facility when compared to vehicle-injected eyes. But the facility change values for the anterior segments that received AdhRhoA2 remained about four times higher than for those receiving the Ad.Pacbgal. It has been reported that viral vectors carrying reporter genes have, by themselves, a moderate protective effect [41]. The mechanism by which this happens is still unknown and although it was originally attributed to an adenovirus-triggered immune response [42], recent hypotheses seem to point to other causes, such the anti-apoptotic effect of some viral genes. Its occurrence in an organ culture system where no immune response is present would coincide with this current trend.

Although an exhaustive morphometric study was not perfomed, the results of our light microscopy studies showed that the architecture of the TM of the RhoA infected eyes looked somewhat less organized and with wider spaces between the trabecular beams, reflecting perhaps the cause of the observed increase in outflow facility.

Increases in outflow facility have been correlated with alterations of the TM cytoskeletal network, and changes in outflow pathway morphology. Changes in both ECM composition and presumed permeability of the inner wall of the Schlemm's canal have been hypothesized [43].

In order to explore the way by which inactivation of RhoA could produce an increase in outflow facility we examined its effect on cultured HTM and SC cells. HTM cells infected with dominant-negative RhoA exhibited a definite change in morphology. These cells showed a flattened, spread-out appearance with a clear cell separation. We hypothesized that this might reflect an induced relaxation of the HTM cells that could perhaps contribute to a more flexible architecture and, in turn, facilitate the passage of the aqueous humor through the tissue. It has been proposed that the HTM cells possess smooth-muscle contractile properties [44] and that modulation of their contractility state might influence outflow facility [6,27,45]. In addition, substances that induce TM relaxation independent of the ciliary muscle, such as H7, have been shown to increase outflow facility in monkeys [46] and in an excised eye porcine system [12]. The mode of action by which this presumed cellular or tissue relaxation would contribute to the observed increase in outflow facility is not known, but in the case of H7, it has been hypothesized to be due to a possible change in the dimensions of the flow pathway through the JCT [46] or inner wall of SC [12]. In our experiments, staining of paxillin, a marker for the presence of focal adhesions, was moderately reduced in infected cells but not totally eliminated, especially in the periphery of the cell. This perhaps is an indication that cell-substrate interactions might not be the primary site for RhoA-induced effects on outflow facility. The fact that the remaining focal contacts were located at the cell's leading edge may indicate the formation of Rac-dependent focal complexes, which are not dependent on the function of RhoA [47,48]. Thus, after removing the set of central supports, the edge supports appeared to be increased.

More obvious was the effect of inactivation of RhoA on the intercellular attachments of the SC cells. Because tight junctions rather than adherens junctions are responsible for paracellular permeability, we examined the localization of the tight junction associated protein ZO-1 in the SC cells. Although it has been shown that ZO-1 might be also associated with adherens junctions' cadherin in some cell types [49], ZO-1 proteins appear to be primary regulators of the tight junctions of the blood brain barrier [50]. Confluent SC cells infected with dominant-negative RhoA exhibit a complete loss of peripheral ZO-1 staining. SC cells appear to separate from each other and form intercellular gaps (Figure 8, top panel). ZO-1 has been localized to cellular tight junctions and other cell membrane sites, and increasing evidence has implicated this protein in the regulation of fluid across cellular barriers [51]. In the eye, antisense oligonucleotides to ZO-1 blocked the DEX-induced increase of transendothelial fluid flow resistance across TM and SC monolayers [52]. Our experiments support such a finding. We think that the observed loss of ZO-1 occurring in SC cells upon infection with AdRhoA2 might reflect an opening of the inner wall barrier that either directly, or through a change in "funneling effects" at the level of the JCT [53] could contribute to an increase in outflow facility. In untreated SC cells, the staining pattern of ZO-1 was segmented (Figure 8, middle panel). The inner wall of SC potentially represents an ultimate barrier for aqueous humor outflow. Aqueous humor enters the lumen of the SC through both paracellular and transcellular routes. The existence of paracellular routes has been extensively documented [54], and formation of openings (presumably both transcellular and paracellular) in the inner wall has been shown to play a role in outflow facility in monkeys [55]. It is possible that, similar to other tissues [51] ZO-1 may be involved in fluid flow across the inner wall of SC.

Although the inner wall of SC appears to be similar to vascular endothelium, and it is generally accepted that the SC cells are of vascular origin [4], it is interesting to note that inactivation of RhoA in cells of human vascular origin has been recently shown to enhance, rather than to reduce, endothelial barrier function [20,21]. In those studies, the use of either conventional drugs (Y-27632 and C3) or a dominant-negative adenovirus prevented thrombin-induced disassembly of tight junctions and reduced the thrombin-induced increase in permeability. However, it has also been observed that while inhibition of RhoA preserves tight junction function in endothelial cells it does the opposite and reduces tight junction function in epithelial cells [24]. Thus the regulation of tight junction integrity must differ between epithelial and endothelial cells. Here we observed that inactivation of RhoA causes a loss of ZO-1 from the junctions and appears to loosen intercellular attachments of SC cells. This suggests that although SC cells have, most likely, conserved some vascular cell characteristics [56], they may also have acquired different traits for the function of an outflow system. At the present time we don't know which of the consequences of RhoA inactivation in the TM (relaxation, reduction of stress fibers, loosening of the cell-substrate or cell-cell junctions) plays the most important role in producing the observed increase in outflow. However, the fact that similar outflow effects occur in different species (rabbit, porcine, bovine, and now human) [25-27] suggests that the observed RhoA effect might be specific for the cells of the outflow pathway.

The potential of manipulating the expression of a TM gene to consequently influence outflow facility has important implications for future gene therapy treatment of glaucoma. Using a similar adenoviral vector, we recently reported the successful repeated delivery (3 times) of a GFP protein to the TM of the living monkey [57]. Each delivery lasted for a period of about four weeks. Although further improvements are needed, together these results suggest the feasibility of delivering a dominant-negative RhoA gene to the TM as a treatment for elevated IOP.

In summary, our data indicates that inactivation of RhoA by adenoviral gene transfer to the intact human outflow pathway results in an increase in outflow facility. Although the mechanism of action for this increase is not presently known, we hypothesize that a reduction of intercellular attachments in SC cells together with a relaxation of transduced TM cells might act to decrease aqueous humor outflow resistance. These two actions could influence outflow either dependently or independently of each other, thus giving RhoA activation/inactivation a potentially important role for modulating outflow. Our findings suggest that dominant-negative RhoA could have an important potential as a gene therapy drug for the treatment of glaucoma.

Acknowledgements

The authors thank the North Carolina (Winston-Salem, NC) and the Old Dominion (Richmond, VA) Eye Banks for providing human eyes; Drs. C. Der (University of North Carolina, Chapel Hill) and Dr. P. Gonzalez for providing material for the adenoviral construction, and Dr. M. Caballero for biochemistry advice. Supported by the Research to Prevent Blindness (TB is a Jules and Doris Stein Research to Prevent Blindness Professor Awardee), NIH grants EY11906 (TB), EY13126 (TB), EY12172 (ETO), EY01894 (DLE) and P30EY05722 (Duke University Eye Center Core Grant) and American Health Assistance Foundation (TB).

References

1. Khosravi-Far R, Solski PA, Clark GJ, Kinch MS, Der CJ. Activation of Rac1, RhoA, and mitogen-activated protein kinases is required for Ras transformation. Mol Cell Biol 1995; 15:6443-53.

2. Weinreb RN. A rationale for lowering intraocular pressure in glaucoma. Surv Ophthalmol 2001; 45 Suppl 4:S335-6.

3. Francis BA, Alvarado J. The cellular basis of aqueous outflow regulation. Curr Opin Ophthalmol 1997; 8:19-27.

4. Hamanaka T, Bill A, Ichinohasama R, Ishida T. Aspects of the development of Schlemm's canal. Exp Eye Res 1992; 55:479-88.

5. Grant WM. Experimental aqueous perfusion in enucleated human eyes. Arch Ophthalmol 1963; 69:783-801.

6. Tian B, Geiger B, Epstein DL, Kaufman PL. Cytoskeletal involvement in the regulation of aqueous humor outflow. Invest Ophthalmol Vis Sci 2000; 41:619-23.

7. Polansky JR, Wood IS, Maglio MT, Alvarado JA. Trabecular meshwork cell culture in glaucoma research: evaluation of biological activity and structural properties of human trabecular cells in vitro. Ophthalmology 1984; 91:580-95.

8. Yue BY. The extracellular matrix and its modulation in the trabecular meshwork. Surv Ophthalmol 1996; 40:379-90.

9. Kaufman PL, Barany EH. Cytochalasin B reversibly increases outflow facility in the eye of the cynomolgus monkey. Invest Ophthalmol Vis Sci 1977; 16:47-53.

10. Epstein DL, Freddo TF, Bassett-Chu S, Chung M, Karageuzian L. Influence of ethacrynic acid on outflow facility in the monkey and calf eye. Invest Ophthalmol Vis Sci 1987; 28:2067-75.

11. Epstein DL, Roberts BC, Skinner LL. Nonsulfhydryl-reactive phenoxyacetic acids increase aqueous humor outflow facility. Invest Ophthalmol Vis Sci 1997; 38:1526-34.

12. Epstein DL, Rowlette LL, Roberts BC. Acto-myosin drug effects and aqueous outflow function. Invest Ophthalmol Vis Sci 1999; 40:74-81.

13. Ridley AJ, Hall A. The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 1992; 70:389-99.

14. Ridley AJ, Comoglio PM, Hall A. Regulation of scatter factor/hepatocyte growth factor responses by Ras, Rac, and Rho in MDCK cells. Mol Cell Biol 1995; 15:1110-22.

15. Wojciak-Stothard B, Entwistle A, Garg R, Ridley AJ. Regulation of TNF-alpha-induced reorganization of the actin cytoskeleton and cell-cell junctions by Rho, Rac, and Cdc42 in human endothelial cells. J Cell Physiol 1998; 176:150-65.

16. Ridley AJ. Stress fibres take shape. Nat Cell Biol 1999; 1:E64-6.

17. Hall A. Rho GTPases and the actin cytoskeleton. Science 1998; 279:509-14.

18. Amano M, Chihara K, Kimura K, Fukata Y, Nakamura N, Matsuura Y, Kaibuchi K. Formation of actin stress fibers and focal adhesions enhanced by Rho-kinase. Science 1997; 275:1308-11.

19. Kimura K, Ito M, Amano M, Chihara K, Fukata Y, Nakafuku M, Yamamori B, Feng J, Nakano T, Okawa K, Iwamatsu A, Kaibuchi K. Regulation of myosin phosphatase by Rho and Rho-associated kinase (Rho-kinase). Science 1996; 273:245-8.

20. Carbajal JM, Schaeffer RC Jr. RhoA inactivation enhances endothelial barrier function. Am J Physiol 1999; 277:C955-64.

21. Wojciak-Stothard B, Potempa S, Eichholtz T, Ridley AJ. Rho and Rac but not Cdc42 regulate endothelial cell permeability. J Cell Sci 2001; 114:1343-55.

22. Braga VM, Machesky LM, Hall A, Hotchin NA. The small GTPases Rho and Rac are required for the establishment of cadherin-dependent cell-cell contacts. J Cell Biol 1997; 137:1421-31.

23. Braga VM, Del Maschio A, Machesky L, Dejana E. Regulation of cadherin function by Rho and Rac: modulation by junction maturation and cellular context. Mol Biol Cell 1999; 10:9-22.

24. Jou TS, Schneeberger EE, Nelson WJ. Structural and functional regulation of tight junctions by RhoA and Rac1 small GTPases. J Cell Biol 1998; 142:101-15.

25. Honjo M, Tanihara H, Inatani M, Kido N, Sawamura T, Yue BY, Narumiya S, Honda Y. Effects of rho-associated protein kinase inhibitor Y-27632 on intraocular pressure and outflow facility. Invest Ophthalmol Vis Sci 2001; 42:137-44.

26. Rao PV, Deng PF, Kumar J, Epstein DL. Modulation of aqueous humor outflow facility by the Rho kinase-specific inhibitor Y-27632. Invest Ophthalmol Vis Sci 2001; 42:1029-37.

27. Thieme H, Nuskovski M, Nass JU, Pleyer U, Strauss O, Wiederholt M. Mediation of calcium-independent contraction in trabecular meshwork through protein kinase C and rho-A. Invest Ophthalmol Vis Sci 2000; 41:4240-6.

28. Rohen JW. Morphology of the uveal tract. Int Ophthalmol Clin 1965; 5:581-667.

29. O'Brien TE, Metheney CD, Polansky JR. Immunofluorescence method for quantifying the trabecular meshwork glucocorticoid response (TIGR) protein in trabecular meshwork and Schlemm's canal cells. Curr Eye Res 1999; 19:517-24.

30. O'Brien ET, Ren X, Wang Y. Localization of myocilin to the golgi apparatus in Schlemm's canal cells. Invest Ophthalmol Vis Sci 2000; 41:3842-9.

31. Barbu V, Dautry F. Northern blot normalization with a 28S rRNA oligonucleotide probe. Nucleic Acids Res 1989; 17:7115.

32. Caballero M, Rowlette LL, Borras T. Altered secretion of a TIGR/MYOC mutant lacking the olfactomedin domain. Biochim Biophys Acta 2000; 1502:447-60.

33. Johnson DH. Human trabecular meshwork cell survival is dependent on perfusion rate. Invest Ophthalmol Vis Sci 1996; 37:1204-8.

34. Borras T, Rowlette LL, Erzurum SC, Epstein DL. Adenoviral reporter gene transfer to the human trabecular meshwork does not alter aqueous humor outflow. Relevance for potential gene therapy of glaucoma. Gene Ther 1999; 6:515-24.

35. Channon KM, Blazing MA, Shetty GA, Potts KE, George SE. Adenoviral gene transfer of nitric oxide synthase: high level expression in human vascular cells. Cardiovasc Res 1996; 32:962-72.

36. Borras T, Tamm ER, Zigler JS Jr. Ocular adenovirus gene transfer varies in efficiency and inflammatory response. Invest Ophthalmol Vis Sci 1996; 37:1282-93.

37. Moscow JA, He R, Gudas JM, Cowan KH. Utilization of multiple polyadenylation signals in the human RHOA protooncogene. Gene 1994; 144:229-36.

38. Burridge K, Chrzanowska-Wodnicka M. Focal adhesions, contractility, and signaling. Annu Rev Cell Dev Biol 1996; 12:463-518.

39. Fritz G, Just I, Kaina B. Rho GTPases are over-expressed in human tumors. Int J Cancer 1999; 81:682-7.

40. Borras T, Matsumoto Y, Epstein DL, Johnson DH. Gene transfer to the human trabecular meshwork by anterior segment perfusion. Invest Ophthalmol Vis Sci 1998; 39:1503-7.

41. Di Polo A, Aigner LJ, Dunn RJ, Bray GM, Aguayo AJ. Prolonged delivery of brain-derived neurotrophic factor by adenovirus-infected Muller cells temporarily rescues injured retinal ganglion cells. Proc Natl Acad Sci U S A 1998; 95:3978-83.

42. Richardson PM, Lu X. Inflammation and axonal regeneration. J Neurol 1994; 242:S57-60

43. Damji KF, Epstein DL. History of Outflow Resistance. In: Van Buskirk EM, Shields MB, editors. 100 years of progress in glaucoma. Philadelphia: Lippincott-Raven; 1997. p. 20-58.

44. de Kater AW, Shahsafaei A, Epstein DL. Localization of smooth muscle and nonmuscle actin isoforms in the human aqueous outflow pathway. Invest Ophthalmol Vis Sci 1992; 33:424-9.

45. Gills JP, Roberts BC, Epstein DL. Microtubule disruption leads to cellular contraction in human trabecular meshwork cells. Invest Ophthalmol Vis Sci 1998; 39:653-8.

46. Tian B, Gabelt BT, Peterson JA, Kiland JA, Kaufman PL. H-7 increases trabecular facility and facility after ciliary muscle disinsertion in monkeys. Invest Ophthalmol Vis Sci 1999; 40:239-42.

47. Horwitz AR, Parsons JT. Cell migration--movin' on. Science 1999; 286:1102-3.

48. Rottner K, Hall A, Small JV. Interplay between Rac and Rho in the control of substrate contact dynamics. Curr Biol 1999; 9:640-8.

49. Itoh M, Nagafuchi A, Yonemura S, Kitani-Yasuda T, Tsukita S, Tsukita S. The 220-kD protein colocalizing with cadherins in non-epithelial cells is identical to ZO-1, a tight junction-associated protein in epithelial cells: cDNA cloning and immunoelectron microscopy. J Cell Biol 1993; 121:491-502.

50. Huber JD, Egleton RD, Davis TP. Molecular physiology and pathophysiology of tight junctions in the blood-brain barrier. Trends Neurosci 2001; 24:719-25.

51. Dejana E, Corada M, Lampugnani MG. Endothelial cell-to-cell junctions. FASEB J 1995; 9:910-8.

52. Underwood JL, Murphy CG, Chen J, Franse-Carman L, Wood I, Epstein DL, Alvarado JA. Glucocorticoids regulate transendothelial fluid flow resistance and formation of intercellular junctions. Am J Physiol 1999; 277:C330-42.

53. Johnson M, Shapiro A, Ethier CR, Kamm RD. Modulation of outflow resistance by the pores of the inner wall endothelium. Invest Ophthalmol Vis Sci 1992; 33:1670-5.

54. Epstein DL, Rohen JW. Morphology of the trabecular meshwork and inner-wall endothelium after cationized ferritin perfusion in the monkey eye. Invest Ophthalmol Vis Sci 1991; 32:160-71.

55. Grierson I, Lee WR. Pressure-induced changes in the ultrastructure of the endothelium lining Schlemm's canal. Am J Ophthalmol 1975; 80:863-84.

56. Gonzalez P, Epstein DL, Borras T. Genes upregulated in the human trabecular meshwork in response to elevated intraocular pressure. Invest Ophthalmol Vis Sci 2000; 41:352-61.

57. Borras T, Gabelt BT, Klintworth GK, Peterson JC, Kaufman PL. Non-invasive observation of repeated adenoviral GFP gene delivery to the anterior segment of the monkey eye in vivo. J Gene Med 2001; 3:437-49.