Free
Physiology and Pharmacology  |   October 2011
Cytoskeletal Dependence of Adenosine Triphosphate Release by Human Trabecular Meshwork Cells
Author Affiliations & Notes
  • Ang Li
    From the Departments of Physiology and
  • Chi Ting Leung
    From the Departments of Physiology and
  • Kim Peterson-Yantorno
    From the Departments of Physiology and
  • W. Daniel Stamer
    the Department of Ophthalmology and Vision Science, University of Arizona, Tucson, Arizona.
  • Mortimer M. Civan
    From the Departments of Physiology and
    Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania; and
  • Corresponding author: Mortimer M. Civan, Department of Physiology, University of Pennsylvania, Richards Building, Philadelphia, PA 19104-6085; civan@mail.med.upenn.edu
Investigative Ophthalmology & Visual Science October 2011, Vol.52, 7996-8005. doi:10.1167/iovs.11-8170
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Ang Li, Chi Ting Leung, Kim Peterson-Yantorno, W. Daniel Stamer, Mortimer M. Civan; Cytoskeletal Dependence of Adenosine Triphosphate Release by Human Trabecular Meshwork Cells. Invest. Ophthalmol. Vis. Sci. 2011;52(11):7996-8005. doi: 10.1167/iovs.11-8170.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: To test whether adenosine triphosphate (ATP) release links cytoskeletal remodeling with release of matrix metalloproteinases (MMPs), regulators of outflow facility and intraocular pressure.

Methods.: ATP release was measured by luciferin-luciferase. Ecto-ATPases from transformed human trabecular meshwork (TM) cells (TM5) and explant-derived TM cells were identified by RT-PCR. Actin was visualized by phalloidin staining. Cell viability was assayed by lactate dehydrogenase and thiazolyl blue tetrazolium bromide methods and propidium iodide exclusion, gene expression by real-time PCR, and MMP release by zymography. Cell volume was monitored by electronic cell sorting.

Results.: Hypotonicity (50%) and mechanical stretch increased ATP release with similar pharmacologic profiles. TM cells expressed ecto-ATPases E-NPP1–3, E-NTPD2, E-NTPD8, and CD73. Prolonged dexamethasone (DEX) exposure (≥2 weeks), but not brief exposure (3 days), increased cross-linked actin networks and reduced swelling-triggered ATP release. Cytochalasin D (CCD) exerted opposite effects. Neither DEX nor CCD altered the cell viability, gene expression, or pharmacologic profile of ATP-release pathways. DEX accelerated, and CCD slowed, the regulatory volume decrease after hypotonic exposure. Activating A1 adenosine receptors (A1ARs) increased total MMP-2 and MMP-9 release. DEX reduced total A1AR-triggered MMP release, and CCD increased the active form of MMP-2 release. The A1AR agonist CHA and the A1AR antagonist DPCPX partially reversed the effects of DEX and CCD, respectively.

Conclusions.: Cytoskeletal restructuring modulated swelling-activated ATP release, in part by changing the duration of cell swelling after hypotonic challenge. Modifying ATP release is expected to modulate MMP secretion by altering ecto-enzymatic delivery of adenosine to A1ARs, linking cytoskeletal remodeling and MMP-mediated modulation of outflow facility.

Glaucoma is frequently associated with elevated intraocular pressure (IOP) that leads to the loss of retinal ganglion cells and atrophy of the optic nerve. The only intervention documented to slow the onset and progression of irreversible blindness is to lower IOP, even in patients with normotensive glaucoma. 1 4 IOP can be reduced by decreasing inflow rate, increasing outflow rate through the uveoscleral exit pathway, or enhancing outflow facility (reducing resistance to outflow) of aqueous humor through the pressure-sensitive trabecular outflow route. Of particular promise in reducing outflow resistance are cytoskeleton-disrupting drugs that can lower IOP in humans 5 and monkeys. 6  
Depolymerization of the actin cytoskeleton, directly with agents such as cytochalasin D (CCD) or indirectly with agents such as 1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H-7), reduces IOP 5,6 and outflow resistance. 6 The converse is also true. Enhancing cytoskeletal polymerization with dexamethasone (DEX) 7 increases IOP in 30% to 40% of healthy humans. 8 The responsive eyes display structural changes in the trabecular meshwork similar to those observed in primary open angle glaucoma. 9  
How cytoskeletal disruption reduces outflow resistance is uncertain. However, altered activity of matrix metalloproteinases (MMPs) not only accompanies cytoskeletal remodeling but likely mediates the effects of such remodeling on outflow resistance. 10 Disrupting the cytoskeleton with CCD or latrunculin A activates MMP-2 in human trabecular meshwork (TM) cells, 10 whereas enhancing polymerization with dexamethasone reduces the expression 11 and secretion 12 of MMPs in TM cells. The link between cytoskeletal remodeling and modulation of MMP activity and secretion is unknown. 
The role of MMPs in modulating outflow resistance is even more strongly established in mediating the purinergic regulation of outflow. In the absence of cytoskeleton-remodeling agents, TM cells release MMP-2 after agonist activation of A1 adenosine receptors (A1ARs). 13 The activation of A1ARs also reduces outflow resistance in nonhuman primates 14 and perfused bovine anterior segments 15 and lowers IOP in several species. 14,16 18 The outflow effect can be blocked by inhibiting MMP activity. 15 The cascade of events linking A1AR activation to MMP-2 release by human TM cells has been recognized. 19 The event triggering this sequence is the delivery of adenosine to the A1 receptors, through ecto-enzymatic metabolism of adenosine triphosphate (ATP) released by the TM cells. The mechanisms underlying swelling-activated ATP release by human TM cells have recently been identified. 20  
We wondered whether cell ATP release, triggering purinergic regulation of MMP-mediated outflow resistance, might also be modulated by cytoskeletal modeling, thereby potentially influencing outflow. The present study tested whether assembly and disassembly of the actin cytoskeleton alters ATP release by human TM cells. 
Materials and Methods
Cellular Models
Transformed normal human trabecular meshwork (HTM) cells (TM5; Alcon Research Inc., Fort Worth, TX) and primary HTM cells were cultured as previously reported 20 22 and studied from passages 20 to 40 and 4 to 7, respectively. Cytoskeleton-remodeled TM5 cells were obtained by the addition of DEX (1 μM) or CCD (25 μM) to the culture media for the periods specified. 
Solutions and Pharmacologic Reagents
As previously described, 20,23 the isotonic solution (295–305 mOsm/kg) with 0.1 mM external free Ca2+ was composed of 110 mM NaCl, 4.7 mM KCl, 5.1 mM CaCl2, 1.2 mM MgCl2, 30 mM NaHCO3, 1.2 mM KH2PO4, 15 mM HEPES, 5 mM EGTA, and 10 mM glucose. Selectively omitting NaCl reduced osmolality to approximately 100 mOsm/kg (67% hypotonicity). Intermediate osmolalities were generated by appropriate mixing of the isotonic and hypotonic solutions. In some experiments, isotonicity was restored by combining 1 part of mannitol stock solution (1.5 M) with 10 parts of the hypotonic solution (50%). The final osmolalities were verified, and pH values were adjusted to 7.4 before each experiment. Biochemical reagents were purchased from Sigma-Aldrich (St. Louis, MO), except for probenecid (Alfa Aesar, War Hill, MA). Chemicals and media for cell culture were obtained from Gibco (Invitrogen, Carlsbad, CA). Ethanol (<0.1%) was used to solubilize hydrophobic drugs, exposing controls to the same concentration of vehicle. Unless otherwise noted, all experiments were conducted at room temperature. 
Confocal Microscopy of F-Actins by Phalloidin Staining
Control and DEX-treated (1 μM for either 3 days or ≥2 weeks) TM5 cells were trypsinized and seeded on noncoated coverslips for 12 hours to allow firm attachment to the underlying surface. CCD-treated samples were obtained by applying CCD (25 μM) to coverslips with adherent cells for 1 hour before fixation. The cells were then fixed with 4% paraformaldehyde for 10 minutes and permeabilized with 100 μM digitonin for 10 minutes, followed by blocking with 1% bovine serum albumin-PBS for 1 hour. Samples were probed with Alexa Fluor-488 phalloidin (1:2000; Invitrogen) for 2 hours and rinsed with PBS three times. DAPI (0.15 μg/mL) was added to counterstain the nuclei fluorescently. Coverslips mounted with reagent (Prolong Gold Antifade; Invitrogen) were observed under a confocal laser scanning microscope (FluoView 1000; Olympus America, Center Valley, PA). Single layers of 0.5-μm thickness were photographed. 
ATP Measurement
ATP release was measured by bioluminescent luciferin-luciferase reaction with light emission recorded using a microplate luminometer (Synergy 2; BioTek, Winooski, VT), as previously documented. 20,23 Cells of the DEX group were treated continuously with 1 μM DEX for either 3 days or over 14 days, as indicated, whereas those of the CCD group were treated with 25 μM CCD for 1 hour. In some experiments, measurements were paused at designated time points for less than 1 minute before recordings resumed to restore isosmolality by the addition of hypertonic mannitol. Separate standard curves were used in experiments involving changes in ionic strength. No test substance interfered with the ATP assay at the specified concentration used in this study. Inhibition of the hypotonicity-induced enhancement of ATP release was calculated from the following equation, as previously described 20,23 :   C max was the maximal ATP concentration after hypotonic treatment without inhibitor, C con was the control ATP concentration in the isotonic bath at the same time point, and C exp was the maximal ATP concentration after hypotonic treatment in the presence of inhibitor. 
Stretch Chamber Experiment
Mechanical perturbation was performed by stretching TM5 cells in the chambers in accordance with the method of Tschumperlin et al., 24 with some modification. Briefly, silicone membranes (Specialty Manufacturing Inc., Saginaw, MI) were autoclaved and coated with fibronectin (50 μg/mL; Millipore, Billerica, MA). A sterile hollow glass cylinder (1-cm diameter; Fisher Scientific, Pittsburgh, PA) was placed on the center of the membrane surface to keep cells growing within this designated area. TM5 cells were trypsinized and plated into the cylinder at a density of 1.27 million/cm2 membrane (i.e., 0.1 million per cylinder-delimited area) and were cultured for 2 days until cells securely adhered to the membrane. Before study, the cylinder was removed, and the culture medium was replaced with 200 μL isotonic solution with or without drugs for a 1-hour preincubation. Thereafter, 50 μL solution was collected as the control (Stretch). By screwing down the chamber for 7 full turns, the membrane was extended to 137%, and another 50 μL solution was collected after stretch was maintained for 10 minutes (Stretch+). Sample ATP concentrations were determined using the method described in ATP Measurement. Cytoskeleton-remodeled cells were generated by treatment with DEX (1 μM, ≥14 days) or CCD (25 μM, 1 hour) by the time of stretching. 
Cell Viability Assays
To quantify cell viability after treatment, the lactose dehydrogenase (LDH) assay and the thiazolyl blue tetrazolium bromide (MTT) assay were adapted for microplate-based experiments. 
Activity of released LDH was measured with a cytotoxicity detection kit (Roche Diagnostic, Indianapolis, IN) in accordance with the manufacturer's instructions, 23 and the cell metabolic states were estimated by the MTT assay as previously described. 20 For stretch experiments, cells grown on the membrane after stretching were stained directly with propidium iodide (2 μg/mL) for 10 minutes, and photographed under a fluorescence microscope (Eclipse Ti; Nikon Instrument Inc., Melville, NY). 
Real-Time Quantitative PCR
Total RNA isolation, reverse transcription, and the following TaqMan gene expression assays were performed, as previously described. 20,23 FAM-labeled MGB TaqMan probes used in the assays are listed in Supplementary Table S1A). The expression levels of indicated genes were all normalized to that of PX1 in specified control cells after 2−ΔΔCt calculation, with human glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as the endogenous control. For samples with cytoskeletal remodeling, TM5 cells were initially treated with DEX (1 μM, ≥14 days) or CCD (25 μM, 1 hour) before RNA extraction. 
Cell Volume Measurement
Cell volume was monitored by electronic cell sorting, using a Coulter counter (ZBI-Channelyzer II; Beckman Coulter Inc., Brea, CA) with a 100-μm aperture as previously reported. 25 TM5 cells were trypsinized and resuspended in isotonic solutions with or without drugs for 1 hour. Thereafter, 50% hypotonicity was applied, and cell volume changes were recorded at indicated time points. Samples with remodeled cytoskeletons were obtained by pretreating cells with either DEX (1 μM, ≥14 days or ≤3 days) or CCD (25 μM, 1 hour). 
Reverse Transcription–Polymerase Chain Reaction
Cell cDNA templates were obtained as described. PCR was performed with a kit (AccuPrime Taq DNA polymerase High Fidelity; Invitrogen) according to the manufacturer's recommendations. Primers used for gene-specific amplification are shown in Supplementary Table S1B. PCR products were separated on 1% agarose gels containing 0.05% ethidium bromide. Bands were visualized under ultraviolet light, sized, and photographed (Molecular Imager Gel Doc XR+ System; Bio-Rad, Hercules, CA). The successfully amplified products were recovered by gel extraction and further verified by sequencing in the DNA Sequencing Facility of the University of Pennsylvania. 
Gelatin Zymography of Matrix Metalloproteinases
The secretion of MMP-2 and MMP-9 into external media was analyzed by gelatin zymography, as previously described, 10,26 with minor modifications. In brief, cells were plated onto 48-well plates at a density of 0.2 million per well and were allowed to grow to confluence, followed by starvation for 24 hours in serum-free media (catalog no. 21063; Gibco). Thereafter, 140 μL fresh DMEM with or without drugs was added to each well to condition cells for 20 hours at 37°C. The conditioned media were completely collected and cleared by centrifugation (10,000g) for 20 minutes. The supernatants were then mixed with protein sample buffer (Zymogram Sample Buffer; Bio-Rad) as suggested by the instructions, and 30 μL per sample was loaded onto each lane of the precast polyacrylamide gel (10% Precast Zymogram Gels; Bio-Rad) for SDS-PAGE separation. After electrophoresis, gels were washed sequentially (Zymogram Renaturing Buffer for 3 hours, Zymography Developing Buffer for 24 hours at 37°C, and Coomassie Brilliant Blue R-250 Staining Solution for 8 hours; Bio-Rad) according to the manufacturer's recommendations. Gels were destained in the destaining solution (Coomassie Brilliant Blue R-250; Bio-Rad) until clear bands were visible against the blue background. The gels were subsequently scanned (Scanjet 3570c, Hewlett-Packard, Palo Alto, CA), and band density was analyzed by ImageJ software (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available at http://rsb.info.nih.gov/ij/index.html). 
Statistical Analysis
Student's t-test, paired or unpaired as appropriate, was applied in comparing two sets of data, and one- or two-way ANOVA was applied to compare three or more sets of data. Statistical analyses were performed with a statistical analysis program (SigmaStat; Aspire Software International, Ashburn, VA). Unless otherwise stated, the results are presented as mean ± SE, with n and N indicating the number of wells studied and the number of independent experiments performed, respectively. P < 0.05 was considered statistically significant. 
Results
Effects of Cytoskeletal Remodeling on Swelling-Activated ATP Release
Continuous exposure to 1 μM DEX for 14 days or more produced substantial cytoskeletal changes in the phalloidin-stained actin cytoskeleton of TM5 cells. Polymerization of F-actin filaments was enhanced, and cross-linked actin networks (CLANs) were extensively formed (Figs. 1B, 1E). CLANs were rarely seen in control TM5 cells (Figs. 1A, 1D). Consistent with earlier reports, 7,8,27 short-term addition (3 days) of DEX (1 μM) did not lead to such structural changes (data not shown). In contrast, CCD (25 μM) rapidly disrupted the actin cytoskeleton within 1 hour, so that F-actin filaments remained visible in only isolated, sparse areas (Figs. 1C, 1F), as previously reported by Sanka et al. 10  
Figure 1.
 
Cytoskeletal remodeling produced by prolonged DEX and brief CCD exposure. TM5 cells were treated with cytoskeleton-remodeling agents, including DEX (1 μM, ≥14 days) and CCD (25 μM, 1 hour). F-actin filaments were visualized by phalloidin staining (green), and nuclei were counterstained with DAPI (blue). Compared with nontreated controls (NT; A, D), DEX treatment strengthened actin filaments (B) and generated CLANs (E, red arrow). In contrast, CCD rapidly disrupted cytoskeleton, making the actin filaments visible only in isolated, sparse areas (C, F). Scale bar, 20 μm.
Figure 1.
 
Cytoskeletal remodeling produced by prolonged DEX and brief CCD exposure. TM5 cells were treated with cytoskeleton-remodeling agents, including DEX (1 μM, ≥14 days) and CCD (25 μM, 1 hour). F-actin filaments were visualized by phalloidin staining (green), and nuclei were counterstained with DAPI (blue). Compared with nontreated controls (NT; A, D), DEX treatment strengthened actin filaments (B) and generated CLANs (E, red arrow). In contrast, CCD rapidly disrupted cytoskeleton, making the actin filaments visible only in isolated, sparse areas (C, F). Scale bar, 20 μm.
TM5 cells release ATP after hypotonic swelling. 20,28 We tested whether cytoskeletal remodeling modifies that release (Fig. 2). The ATP concentration was 9.7 ± 0.3 nM (n = 404 wells) after incubation of TM5 cells in control isotonic solution. This low baseline level was stable during the 2 hours of measurement. Hypotonicity (50%) triggered ATP release, raising the bath concentration by 3.2-fold to 41 ± 1 nM (n = 569; P < 0.001). Neither DEX (1 μM for ≥14 days, n = 164) nor CCD (25 μM for 1 hour, n = 234) treatment significantly affected baseline (P = 0.882 by one-way ANOVA). However, stimulating actin polymerization with 1 μM DEX for ≥14 days substantially decreased swelling-elicited ATP release by 36% ± 3% (n = 246; P < 0.001 vs. normal hypotonic control); depolymerizing actin filaments with CCD increased release by 45.9% ± 4.8% (n = 303; P < 0.001). Incubating cells with 1 μM DEX for 3 days did not alter isotonic (n = 32; P = 0.194) or hypotonic (n = 48; P = 0.162) bath ATP levels compared with nontreated controls. The observed effects of DEX and CCD suggested that the actin cytoskeleton is important in some way in regulating swelling-activated ATP release. 
Figure 2.
 
Effects of DEX and CCD on swelling-induced ATP release. TM5 cells were exposed to cytoskeleton-remodeling agent DEX (1 μM, ≥14 days or 3 days) or CCD (25 μM, 1 hour). Prolonged DEX treatment enhanced actin polymerization and decreased swelling-triggered ATP release, whereas CCD produced opposite results. Short-term DEX treatment was ineffective. Isotonic baselines were unaffected by the drugs, compared with nontreated (NT) controls. *P < 0.05, one-way ANOVA. Numbers of wells analyzed are entered under the x-axis.
Figure 2.
 
Effects of DEX and CCD on swelling-induced ATP release. TM5 cells were exposed to cytoskeleton-remodeling agent DEX (1 μM, ≥14 days or 3 days) or CCD (25 μM, 1 hour). Prolonged DEX treatment enhanced actin polymerization and decreased swelling-triggered ATP release, whereas CCD produced opposite results. Short-term DEX treatment was ineffective. Isotonic baselines were unaffected by the drugs, compared with nontreated (NT) controls. *P < 0.05, one-way ANOVA. Numbers of wells analyzed are entered under the x-axis.
Cytoskeletal Remodeling Did Not Change the ATP-Releasing Pathways
We wondered whether DEX- and CCD-induced changes of ATP release after cell swelling might be attributed to alterations in these conduits. Hypotonicity-evoked ATP efflux is released through three major conduits in TM5 cells: pannexin-1 (PX1) and connexin (Cx) hemichannels and the purinergic P2X7 receptor (P2RX7). 20 Consistent with our previous data, 20 relatively selective blockers (Discussion) of PX1 (probenecid [PRO] 0.1 mM) and Cx hemichannels (heptanol [HEP] 1 mM), and P2RX7 (KN62, 1 μM) inhibited swelling-induced ATP release by 35% ± 1% (n = 75), 49% ± 3% (n = 72), and 37% ± 2% (n = 71), respectively. The nonselective chloride-channel blocker 5-nitro-2-(3-phenylpropylamino) benzoic acid (NPPB) decreased the release by 32% ± 2% (n = 64) at the low concentration of 30 μM. The inhibitor of vesicular ATP release, bafilomycin A1 (BAF, 2 μM), had no effect (1% ± 5%, n = 48; P > 0.05). Application of KN62 together with either PRO or HEP enhanced the inhibition to 53% ± 2% (n = 64) and to 70% ± 1% (n = 80), respectively. These less than additive enhancements are considered in the Discussion. Nearly complete inhibition (91% ± 1%, n = 48) of the hypotonically stimulated ATP release from TM5 cells was achieved by simultaneous blocking of PX1, Cx hemichannels, and P2RX7 using a combination of 30 μM CBX, 1 mM HEP, and 1 μM KN-62. 
The ATP-releasing pharmacologic profile of the cytoskeleton-remodeled TM5 cells was similar to that of control cells (Fig. 3). We also tested whether gene expression of the three major conduits had been altered by DEX or CCD. The results obtained with real-time PCR indicated that the mRNA levels of PX1, Cx, and P2RX7 remained unchanged after cytoskeletal remodeling (Supplementary Fig. S1). These data indicated that cytoskeletal remodeling did not alter the number or properties of the ATP-releasing conduits in TM cells. 
Figure 3.
 
Unchanged efficacy of blockers of swelling-evoked ATP release in TM5 cells after cytoskeletal restructuring. Inhibitors of previously identified conduits for hypotonicity-evoked ATP release, including probenecid (PRO, 0.1 mM, PX1 blocker), heptanol (HEP, 1 mM, Cx blocker), NPPB (30 μM, nonselective chloride channel blocker), and KN62 (1 μM, P2RX7 blocker) displayed the same efficacy in nontreated (NT), DEX-treated (1 μM, ≥14 days), and CCD-treated (1 μM, 1 hour) cells. The combination of carbenoxolone (CBX, 30 μM), HEP (1 mM), and KN62 (1 μM) nearly abolished this release, whereas bafilomycin A1 (BAF, 2 μM, vesicular ATP release blocker) was ineffective. P > 0.05 vs. NT controls by one-way ANOVA. The numbers of wells analyzed in the study are entered under the x-axis.
Figure 3.
 
Unchanged efficacy of blockers of swelling-evoked ATP release in TM5 cells after cytoskeletal restructuring. Inhibitors of previously identified conduits for hypotonicity-evoked ATP release, including probenecid (PRO, 0.1 mM, PX1 blocker), heptanol (HEP, 1 mM, Cx blocker), NPPB (30 μM, nonselective chloride channel blocker), and KN62 (1 μM, P2RX7 blocker) displayed the same efficacy in nontreated (NT), DEX-treated (1 μM, ≥14 days), and CCD-treated (1 μM, 1 hour) cells. The combination of carbenoxolone (CBX, 30 μM), HEP (1 mM), and KN62 (1 μM) nearly abolished this release, whereas bafilomycin A1 (BAF, 2 μM, vesicular ATP release blocker) was ineffective. P > 0.05 vs. NT controls by one-way ANOVA. The numbers of wells analyzed in the study are entered under the x-axis.
Because the different amounts of ATP released by cytoskeleton-remodeled cells might have reflected cell death caused by drug toxicity, LDH and MTT assays were performed. There was no indication from the LDH assay that remodeling agents decreased cell viability or from the MTT assay that the drugs altered the metabolic state in the concentrations specified (Supplementary Fig. S2A). The nearly complete abolition of hypotonicity-triggered ATP release from all cells by combined blockade of PX1, Cx hemichannels, and P2RX7, together with the findings from the LDH and MTT assays, provided no indication of measurable loss of cell viability under our experimental conditions. 
Role of Regulatory Volume Decrease in Swelling-Activated ATP Release
Cytoskeletal restructuring did not alter the exit pathways of hypotonicity-triggered ATP release; therefore, we considered the possibility that DEX and CCD might have altered either the magnitude or the duration of swelling produced by the hypotonic challenge. Figure 4 presents the time courses of cell volumes after application of 50% hypotonicity. The control and two experimental groups swelled rapidly, peaked at approximately 4 minutes, and thereafter exhibited a regulatory volume decrease (RVD). The time to and the magnitude of maximal swelling were identical among the three groups. Nevertheless, the RVD recovery phases of the control, DEX-treated, and CCD-treated cells were significantly different (P < 0.05, two-way ANOVA with Dunnett's multiple comparisons test). The times (T1/2) for 50% return to isotonic volume were 21.2 ± 1.9 minutes (control), 16.3 ± 0.2 minutes (DEX), and 25.1 ± 1.7 minutes (CCD). Treatment with DEX for <3 days did not alter the T1/2 (n = 8; P > 0.08 vs. control group). 
Figure 4.
 
Effects of prolonged DEX treatment and transient CCD exposure on the RVD of TM5 cells. TM5 cells were treated with DEX (1 μM, ≥14 days) and CCD (1 μM, 1 hour) before challenge by hypotonicity (50%). The trajectories are three-parameter exponential fits. Results were significantly different among the DEX-treated (n = 4), CCD-treated (n = 4), and nontreated (NT) control (n = 3) cells (P < 0.05 by two-way ANOVA). The DEX group had a much shorter period for 50% recovery after hypotonic exposure (T1/2) compared with the NT control group, whereas the CCD group had the longest.
Figure 4.
 
Effects of prolonged DEX treatment and transient CCD exposure on the RVD of TM5 cells. TM5 cells were treated with DEX (1 μM, ≥14 days) and CCD (1 μM, 1 hour) before challenge by hypotonicity (50%). The trajectories are three-parameter exponential fits. Results were significantly different among the DEX-treated (n = 4), CCD-treated (n = 4), and nontreated (NT) control (n = 3) cells (P < 0.05 by two-way ANOVA). The DEX group had a much shorter period for 50% recovery after hypotonic exposure (T1/2) compared with the NT control group, whereas the CCD group had the longest.
RVD strongly influences the rate of hypotonicity-triggered ATP release in certain cells. 29 We accordingly tested whether shortening the period of hypotonic swelling could reduce ATP release from TM5 cells. We have previously reported that ATP release from TM5 cells triggered by 50% hypotonicity peaks at 26.6 ± 0.3 minutes, with a time to half-peak response of 8.5 ± 0.1 minutes. 20 We have now found that restoring isotonicity by adding hypertonic mannitol at 0 minute, 5 minutes, and 20 minutes after hypotonic shock decreased ATP release by 89% ± 2% (n = 32; P < 0.001 vs. the hypotonic control), 66% ± 3% (n = 32; P < 0.001), and −2% ± 4% (n = 32; P > 0.05), respectively. These data strongly suggested that cytoskeletal remodeling modulated ATP release from TM5 cells by modifying the swelling produced by hypotonic challenge rather than targeting downstream conduits. 
TRPV4 channels, members of the transient receptor potential channels, are stimulated by hypotonicity, leading to Ca2+ influx and activation of downstream events, 30 including the release of large molecules such as diadenosine tetraphosphate (Pintor JJ, et al. IOVS. 2011;52:ARVO E-Abstract 2052). We tested the potential mediation of TRPV4 in hypotonicity-triggered ATP release by applying the nonselective inhibitor of TRPV channels ruthenium red (RUR). RUR (20 μM) had no effect on ATP from TM5 cells, either under isotonic conditions (n = 16; P > 0.9) or after hypotonic challenge (n = 32; P = 0.997). Therefore, TRPV4 does not play a role in hypotonicity-initiated ATP release from TM5 cells. 
Conduits for Stretch-Activated ATP Release
The foregoing data were obtained by stimulating ATP release with cell swelling. In addition, pannexins 31 and connexins 32 are known to be mechanosensitive. Given that the ATP-releasing pathways recruited by cells are likely stimulus dependent, 20 we identified the release mechanisms elicited by mechanical stretch. 
Mechanically stretching the TM5 cells by 30% increased ATP release by nearly twofold (194% ± 22%, n = 82; P < 0.001). The responses to separate application of inhibitors of PX1 (PRO), Cx (HEP), P2RX7 (KN62), vesicular release (BAF), and the combined inhibition of PX1, Cx, and P2RX7 (CBX + HEP + KN62) were identical with those to hypotonic swelling (Fig. 5). No detectable cell damage was found after stretching (Supplementary Fig. S2B). The pharmacologic results indicated that swelling- and stretch-elicited ATP efflux from TM5 cells were mediated mainly through PX1 and Cx hemichannels, together with P2X7 ionoreceptors. The effects of cytoskeletal remodeling on stretch-activated ATP release were also similar to those on swelling-activated release. Increasing actin polymerization by prolonged DEX application (1 μM, ≥14 days) reduced stretch-triggered release by 81% ± 1% (n = 33; P < 0.001 vs. normal controls). Depolymerizing actin with CCD (25 μM, 1 hour) increased ATP efflux by 32% ± 3% (n = 43; P < 0.001). The results indicate that mechanical stretch and hypotonic swelling activate the same ATP-releasing pathways. 
Figure 5.
 
Correlation of drug inhibitions on hypotonicity- and stretch-elicited ATP release. TM5 cells were mechanically stretched in the chambers, and the following drugs were used to block the release: PRO (0.1 mM, n = 21), HEP (1 mM, n = 9), KN62 (1 μM, n = 18), BAF (2 μM, n = 11), and a combination of 30 μM CBX, 1 mM HEP, and 1 μM KN62 (n = 21). The trajectory is a linear least-squares fit, with a slope of 1.04 ± 0.02 (P < 0.001) and a correlation coefficient of 0.999, indicating that the same conduits are recruited in both swelling- and stretch-elicited ATP release.
Figure 5.
 
Correlation of drug inhibitions on hypotonicity- and stretch-elicited ATP release. TM5 cells were mechanically stretched in the chambers, and the following drugs were used to block the release: PRO (0.1 mM, n = 21), HEP (1 mM, n = 9), KN62 (1 μM, n = 18), BAF (2 μM, n = 11), and a combination of 30 μM CBX, 1 mM HEP, and 1 μM KN62 (n = 21). The trajectory is a linear least-squares fit, with a slope of 1.04 ± 0.02 (P < 0.001) and a correlation coefficient of 0.999, indicating that the same conduits are recruited in both swelling- and stretch-elicited ATP release.
Effects of Cytoskeletal Remodeling on A1 Adenosine Receptor-Mediated Secretion of MMP-2 and MMP-9
Agonist activation of A1 adenosine receptors (A1ARs) increases secretion of MMP-2 in cultured TM cells, 13 resulting in the reduction of outflow resistance. 14,15 In principle, ecto-enzymatic conversion of released ATP can lead to A1AR activation. However, the expression profile of ecto-ATPases in TM cells is unknown. By using RT-PCR followed by DNA sequencing, we found the expression of multiple ecto-ATPases in explant-derived TM cells and transformed TM5 cells (Fig. 6). The identification of ecto-enzymes CD73, E-NPP1–3, E-NTPD2, and E-NTPD8 provides the first molecular evidence that conversion of ATP to adenosine is physiologically feasible in TM cells. 
Figure 6.
 
Multiple membrane-type nucleotide ectoenzymes expressed by trabecular meshwork cells. RT-PCR with subsequent DNA sequencing verified that both TM5 and HTM cells express CD73, E-NPP1, E-NPP2, E-NPP3, E-NTPD2, and E-NTPD8, and several transcript variants were identified as well. No product was detectable on omitting reverse-transcriptase [RTase(-)], confirming that the cDNA derived from total RNA was free of genomic DNA contamination.
Figure 6.
 
Multiple membrane-type nucleotide ectoenzymes expressed by trabecular meshwork cells. RT-PCR with subsequent DNA sequencing verified that both TM5 and HTM cells express CD73, E-NPP1, E-NPP2, E-NPP3, E-NTPD2, and E-NTPD8, and several transcript variants were identified as well. No product was detectable on omitting reverse-transcriptase [RTase(-)], confirming that the cDNA derived from total RNA was free of genomic DNA contamination.
We next confirmed that A1AR activation stimulates MMP-2 secretion. TM5 cells were incubated either with the selective A1AR agonist CHA (100 nM) or with the nonselective AR agonist NECA (1 μM) for 20 hours, with or without the highly selective A1AR antagonist DPCPX (20 nM). The results of gelatin zymography (Fig. 7A) indicated that adding CHA and NECA increased the secretion of MMP-2 by 38% ± 5% (n = 23; P < 0.001 vs. the nontreated control) and 70% ± 5% (n = 12; P < 0.001), respectively. The CHA- and NECA-stimulated enhancements were significantly reduced to 16% ± 5% (n = 23) and 25% ± 5% (n = 12), respectively, by the A1AR antagonist DPCPX. DPCPX had no effect on baseline MMP secretion (n = 18; P > 0.05 vs. the nontreated control). For the first time, we also found that MMP-9, which was also abundant in external media, was upregulated by stimulating A1ARs (Fig. 7A). The antagonist DPCPX strongly decreased both CHA- and NECA-stimulated MMP-2 and MMP-9 secretion, indicating that A1ARs are the dominant AR mediators of this purinergic stimulation. 
Figure 7.
 
MMP-2 and MMP-9 secretion modulated by adenosine A1 receptors and cytoskeletal remodeling. (A) Stimulating A1ARs with either specific agonist CHA (100 nM) or wide-spectrum AR agonist NECA (1 μM) increased MMP-2 and MMP-9 secretion. Both stimulations were blocked by concomitant application of the selective A1AR antagonist DPCPX (20 nM). Numbers of independent experiments are indicated along the abscissa. (B) Increasing actin polymerization with DEX (1 μM, ≥14 days) reduced MMP-2 and MMP-9 secretion considerably, and the reductions were significantly reversed by activating A1ARs with CHA (100 nM). Applying CCD (25 μM, 1 hour) raised MMP-9 secretion without affecting total MMP-2. DPCPX (20 nM) lowered both MMP-9 and total MMP-2 secretion of CCD-treated cells. Numbers of independent experiments are indicated along the abscissa. (C). The zymographic gel image shows the changes in MMP secretion produced by CCD (1 μM), an increase in the proform of MMP-9, and the appearance of the active form of MMP-2.
Figure 7.
 
MMP-2 and MMP-9 secretion modulated by adenosine A1 receptors and cytoskeletal remodeling. (A) Stimulating A1ARs with either specific agonist CHA (100 nM) or wide-spectrum AR agonist NECA (1 μM) increased MMP-2 and MMP-9 secretion. Both stimulations were blocked by concomitant application of the selective A1AR antagonist DPCPX (20 nM). Numbers of independent experiments are indicated along the abscissa. (B) Increasing actin polymerization with DEX (1 μM, ≥14 days) reduced MMP-2 and MMP-9 secretion considerably, and the reductions were significantly reversed by activating A1ARs with CHA (100 nM). Applying CCD (25 μM, 1 hour) raised MMP-9 secretion without affecting total MMP-2. DPCPX (20 nM) lowered both MMP-9 and total MMP-2 secretion of CCD-treated cells. Numbers of independent experiments are indicated along the abscissa. (C). The zymographic gel image shows the changes in MMP secretion produced by CCD (1 μM), an increase in the proform of MMP-9, and the appearance of the active form of MMP-2.
Cytoskeleton-remodeling agents strongly affected both baseline and A1AR-stimulated secretion of MMP-2 and MMP-9 (Figs. 7B, 7C). Prolonged DEX treatment drastically inhibited MMP-2 and MMP-9 secretion by 70% ± 3% (n = 32) and 92% ± 1% (n = 32), respectively, in comparison with controls. In contrast, MMP-9 secretion was significantly increased by 48% ± 3% (n = 26; P < 0.05) by CCD treatment. CCD application did not alter total MMP-2 secretion, in agreement with Sanka et al., 10 but did produce the appearance of the active form of MMP-2 (Fig. 7C). It is possible that the appearance of the active form of MMP-2 might have raised the total digesting ability of MMP-2, enzymatically reducing the proactive form secreted. 
We also examined the interaction of cytoskeleton restructuring and A1AR-active agents. Applying CHA (100 nM) to DEX-treated cells partially reversed DEX inhibition, stimulating the secretion of MMP-2 and MMP-9 from 8% ± 1% and 30% ± 3% in controls to 21% ± 2% (n = 22) and 75% ± 3% (n = 22), respectively (Fig. 7B). Similarly, adding DPCPX (20 nM) partially reversed the stimulation by CCD, reducing MMP-9 secretion from 148% ± 3% to 119% ± 2% (n = 32). DPCPX (20 nM) reduced the total and active forms of MMP-2 secretion after CCD treatment by 20% ± 4% (n = 36) and 23% ± 5% (n = 36; P < 0.05 vs. CCD group), respectively. The persistence of the active form of MMP-2 after the inhibition of A1ARs suggested that CCD treatment stimulated MMP-2 secretion or activation through additional A1AR-independent signaling pathways. 
Discussion
ATP release is the enabling step in purinergic regulation of outflow resistance and IOP. Consistent with our previous studies of TM5 and explant-derived human TM cells, 20 separate inhibition of PX1 hemichannels, Cx hemichannels, and P2X7 receptors reduced swelling-activated ATP release from TM5 cells to comparable extents (Fig. 8). Simultaneous inhibition of all three pathways reduced that release by >90%. Stretch triggered ATP release through the same pharmacologically defined pathways. Stretch may be an important physiological stimulus for TM cell ATP release. TM cells are likely to be displaced and stretched during oscillations of 2.4 ± 0.7 mm Hg produced by the cardiac cycle in the IOP of healthy humans. 33 Elevated ATP levels in the aqueous humor have been associated with primary angle closure glaucoma, 34 as well as chronic angle closure glaucoma, 35 possibly because of increased stretch of the anterior segment resulting from the very high intraocular pressures recorded. 
Figure 8.
 
Cytoskeleton-dependent purinergic regulation of aqueous humor outflow. Swelling-activated ATP release is recognized to proceed through PX1 and Cx hemichannels and P2X7 purinergic receptors. 20 The present study tests showed mechanical stretch triggers release through the same conduits, whereas remodeling actin cytoskeleton does not modulate this process. The present study also demonstrated ecto-enzymes CD73, E-NPP1–3, E-NTPD2, and E-NTPD8, which convert ATP to adenosine, are expressed in TM cells, consequently activating A1AR to stimulate the secretion of MMP-9 in addition to MMP-2.
Figure 8.
 
Cytoskeleton-dependent purinergic regulation of aqueous humor outflow. Swelling-activated ATP release is recognized to proceed through PX1 and Cx hemichannels and P2X7 purinergic receptors. 20 The present study tests showed mechanical stretch triggers release through the same conduits, whereas remodeling actin cytoskeleton does not modulate this process. The present study also demonstrated ecto-enzymes CD73, E-NPP1–3, E-NTPD2, and E-NTPD8, which convert ATP to adenosine, are expressed in TM cells, consequently activating A1AR to stimulate the secretion of MMP-9 in addition to MMP-2.
As previously reported, 20 comparable inhibition of ATP release was observed by selectively blocking PX1 and Cx hemichannels of TM5 and explant-derived TM cells. We also reported that concurrent inhibition of PX1 and Cx hemichannels produced an additive reduction in ATP release from TM5 cells. 20 Estimating the contribution of P2X7 receptors is more uncertain. In the present study, a separate block of the P2X7 ionorecepter with KN62 reduced swelling-activated ATP release by 37% ± 2%. However, KN62 produced a less than additive inhibition when administered simultaneously with an inhibitor of either PX1 or Cx hemichannels. In the latter conditions, KN62 appeared to be inhibiting ATP release by 16% to 19%. Evidently, P2RX7 contributes more greatly to ATP release when the parallel ATP release pathways are operative, consistent with the expectation that the P2X7 ATP receptor is activated in the presence of higher baseline ATP levels. In contrast to its importance in ATP release from TM cells, P2RX7 appears to play no role in the release of ATP by the ciliary epithelial cells responsible for the formation of aqueous humor. 23  
Pharmacologic estimations of the relative contributions of the parallel pathways to total ATP release are necessarily approximate, given possible cross-target inhibitions. However, our previous results obtained with 21 inhibitors and PX1 knockdown suggest that these estimations are likely correct for human TM cells. 20 An alternative strategy based on mimetic peptides has proved unsuccessful in generating selective inhibitors. 36 In contrast, probenecid is relatively selective in blocking PX1 and not Cx hemichannels, 37 and CBX also preferentially blocks PX1 hemichannels in low concentrations 38 with a reported IC50 value 5- to 20-fold lower than that required to block Cx hemichannels. We found that 0.1 mM PRO and 3 μM CBX produced identical inhibition (∼36%) of ATP release by TM5 cells, consistent with an inhibition of 31% ± 2% produced by partial knockdown of PX1. 20 Heptanol is a relatively selective blocker of Cx hemichannels at 1 mM, 39 but it also blocks PX1 hemichannels at 3 mM. 40 External Ca2+ selectively blocks Cx hemichannels. 41 Once again, the two relatively selective blockers of Cx hemichannels produced similar inhibitions of ATP release, 1 mM heptanol inhibiting by 44% ± 3% and 2.5 mM Ca2+ inhibiting by 36% ± 7%. 20 Blocking PX1 and Cx hemichannels concurrently produced an additive inhibition. 20 We have assayed the contribution of P2X7 receptors with the blocker KN-62, 42 which also inhibits calcium/calmodulin-dependent protein kinase II (Ca/CaMKII). However, other P2RX7 inhibitors, brilliant blue G (BBG) 42 and A438079, 43 produced the same inhibition of ATP release by TM5 cells, and the selective inhibitor of Ca/CaMKII lavendustin C 44 was ineffective. 20 Simultaneously blocking PX1, Cx, and P2RX7 pathways inhibited ATP release from TM5 cells by >90%. 
We have used BAF to assess the vesicular release of ATP. This inhibitor, like monensin, interferes with vesicular uptake of ATP within the cell and is frequently used to reduce vesicular ATP transfer to the extracellular fluid. 45 48 Neither BAF nor monensin significantly affected ATP release from TM5 cells. 20 BAF was also tested on explant-derived TM cells and was found ineffective. Interestingly, Luna et al. 49 reported that monensin nearly completely inhibited ATP release from porcine TM cells stimulated with cyclic mechanical stretch. The basis for the difference is unclear, possibly reflecting species origin (porcine vs. human cells), stimulus used for ATP release, or other unidentified methodological factors. 
Cytoskeletal remodeling modified ATP release. Disrupting the actin cytoskeleton with CCD increased swelling-activated ATP release by 46% ± 5%. The cytochalasins act by capping the barbed end of actin polymers, preventing elongation. 6,50 Stimulating actin polymerization with glucocorticoid had the opposite effect. DEX inhibited swelling-activated ATP release by 36% ± 3% after incubation for ≥2 weeks but had no effect on that release after incubation for 3 days. In agreement with previous reports, 7,8 prolonged DEX administration increased actin polymerization and the number of CLANS, which were rarely seen in untreated cells. CLANS usually appear after 3 to 4 days of exposure to DEX, 7 possibly by increasing the expression and activation of β3 integrins. 51 The temporal association of CLAN formation with the modulation of ATP release raises the possibility that CLANs or a co-temporal structural change may play a role in modifying ATP release. 
We wondered whether cytoskeletal remodeling might modulate swelling-activated ATP release by altering the gene expression or membrane trafficking of the ATP release pathways. However, neither gene expression nor the pharmacologic profile of the release mechanisms was changed. Alternatively, cytoskeletal restructuring could act by altering the signaling cascade linking swelling to ATP release. One possible mechanism would be to modify the cell volume response to hypotonic challenge. In most cells, swelling is not sustained during continued hypotonic exposure; rather, it displays a regulatory volume decrease (RVD) mediated by the release of solute and secondarily by water. 52,53 Thus, the duration of hypotonic swelling is limited by the RVD so that slowing the RVD might be expected to increase swelling-activated ATP release. In support of this expectation, most ATP was found to be released during maximal swelling, before the initiation of RVD by A549 and 16HBE14o cells. 29 Similarly, we have now found that early restoration of isotonicity by the addition of mannitol within 5 minutes after hypotonic challenge reduced ATP release. We observed that CCD prolonged the period of hypotonic swelling by slowing the RVD, whereas prolonged exposure to DEX had the opposite effect. Brief incubation with DEX did not alter the RVD; the time required for 50% recovery from peak swelling (T1/2) was unchanged. In contrast, prolonged DEX administration significantly reduced T1/2 by approximately 5 minutes 
The hypothesis under study has been that changes in outflow resistance triggered by cytoskeletal remodeling might be mediated, at least partly, by modifying ATP release and thereby altering the delivery of adenosine to A1 adenosine receptors. The tacit assumption is that TM cells express ecto-enzymes capable of converting ATP to adenosine, for which no evidence has yet been available. Our RT-PCR analysis, followed by DNA sequencing, has now demonstrated gene expression of multiple ecto-enzymes in both explant-derived human TM cells and the TM5 cell line, which can metabolize ATP to adenosine. 54 E-NPP1–3 are members of the ecto-nucleotide pyrophosphatase/phosphodiesterase family of enzymes that convert ATP to adenosine monophosphate (AMP). E-NTPD2 and E-NTPD8 are members of the ecto-nucleoside triphosphate diphosphohydrolases that convert both ATP to ADP and ADP to AMP. Ecto-5′-nucleotidase (CD73) can convert AMP to adenosine. The observed expression of these six ecto-enzymes supports the plausibility that modifications of ATP release can alter the delivery of adenosine to the TM cells. 
Activation of A1ARs with the selective agonist CHA increased MMP-2 secretion, confirming the earlier observation by Shearer and Crosson. 13 Furthermore, the selective A1AR antagonist DPCPX abolished the increase in MMP-2 secretion initiated by the nonselective agonist A1AR agonist NECA. In addition, we have observed for the first time that CHA and NECA enhance MMP-9 secretion, an enhancement that is almost entirely blocked by DPCPX. A1 adenosine receptors appear to play a role in the altered MMP secretion produced by cytoskeletal remodeling. The A1AR antagonist DPCPX sharply reduced the CCD-initiated stimulation of MMP-9 secretion but had no effect on baseline secretion of either MMP-9 or MMP-2. DPCPX also reduced secretion of the proactive form and the active form of MMP-2, but to a lesser extent, suggesting a role for additional, unidentified regulators of MMP-2 release. 
We conclude that stretch and swelling of TM5 cells trigger ATP release through pharmacologically identical pathways, PX1 and Cx hemichannels, and P2X7 receptors (Fig. 8). The role of P2RX7 is enhanced by concurrent activity of the hemichannels. Both explant-derived TM cells and TM5 cells express ecto-ATPases capable of converting released ATP to adenosine. Adenosine activation of A1 adenosine receptors stimulates TM-cell secretion of MMP-9, as well as MMP-2, gelatinase. Thus, the modulation of ATP release triggered by cytoskeletal restructuring can play a role in altering MMP release by the TM cells, leading to modified outflow resistance and intraocular pressure. The quantitative significance of this role is under investigation. 
Supplementary Materials
Figure sf01, PDF - Figure sf01, PDF 
Figure sf02, PDF - Figure sf02, PDF 
Table st1, PDF - Table st1, PDF 
Footnotes
 Supported by National Institutes of Health Grants EY13624 (MMC) and EY17007 (WDS) and Core Grant EY01583 (University of Pennsylvania).
Footnotes
 Disclosure: A. Li, None; C.T. Leung, None; K. Peterson-Yantorno, None; W.D. Stamer, None; M.M. Civan, None
The authors thank Iok-hou Pang and Alcon Laboratories Inc. for generously providing the TM5 cell line, Susan S. Margulies (Department of Bioengineering, University of Pennsylvania) for the generous loan of stretch chambers and guidance in their use, Yao Yao for gel documentation, and Yuting Zhao for zymography assistance. 
References
Collaborative Normal-Tension Glaucoma Study Group. The effectiveness of intraocular pressure reduction in the treatment of normal-tension glaucoma. Am J Ophthalmol. 1998;126(4):498–505. [CrossRef] [PubMed]
Collaborative Normal-Tension Glaucoma Study Group. Comparison of glaucomatous progression between untreated patients with normal-tension glaucoma and patients with therapeutically reduced intraocular pressures. Am J Ophthalmol. 1998;126(4):487–497. [CrossRef] [PubMed]
Kass MA Heuer DK Higginbotham EJ . The Ocular Hypertension Treatment Study: a randomized trial determines that topical ocular hypotensive medication delays or prevents the onset of primary open-angle glaucoma. Arch Ophthalmol. 2002;120(6):701–713, discussion 829–730. [CrossRef] [PubMed]
The AGIS Investigators. The Advanced Glaucoma Intervention Study (AGIS), 7: the relationship between control of intraocular pressure and visual field deterioration. Am J Ophthalmol. 2000;130(4):429–440. [CrossRef] [PubMed]
Tanihara H Inatani M Honjo M . Intraocular pressure-lowering effects and safety of topical administration of a selective ROCK inhibitor, SNJ-1656, in healthy volunteers. Arch Ophthalmol. 2008;126(3):309–315. [CrossRef] [PubMed]
Tian B Geiger B Epstein DL Kaufman PL . Cytoskeletal involvement in the regulation of aqueous humor outflow. Invest Ophthalmol Vis Sci. 2000;41(3):619–623. [PubMed]
Clark AF Wilson K McCartney MD . Glucocorticoid-induced formation of cross-linked actin networks in cultured human trabecular meshwork cells. Invest Ophthalmol Vis Sci. 1994;35(1):281–294. [PubMed]
Clark AF Brotchie D Read AT . Dexamethasone alters F-actin architecture and promotes cross-linked actin network formation in human trabecular meshwork tissue. Cell Motil Cytoskeleton. 2005;60(2):83–95. [CrossRef] [PubMed]
Clark AF Wilson K de Kater AW Allingham RR McCartney MD . Dexamethasone-induced ocular hypertension in perfusion-cultured human eyes. Invest Ophthalmol Vis Sci. 1995;36(2):478–489. [PubMed]
Sanka K Maddala R Epstein DL Rao PV . Influence of actin cytoskeletal integrity on matrix metalloproteinase-2 activation in cultured human trabecular meshwork cells. Invest Ophthalmol Vis Sci. 2007;48(5):2105–2114. [CrossRef] [PubMed]
Spiga MG Borras T . Development of a gene therapy virus with a glucocorticoid-inducible MMP1 for the treatment of steroid glaucoma. Invest Ophthalmol Vis Sci. 2010;51(6):3029–3041. [CrossRef] [PubMed]
el-Shabrawi Y Eckhardt M Berghold A . Synthesis pattern of matrix metalloproteinases (MMPs) and inhibitors (TIMPs) in human explant organ cultures after treatment with latanoprost and dexamethasone. Eye. 2000;14(pt 3A):375–383. [CrossRef] [PubMed]
Shearer TW Crosson CE . Adenosine A1 receptor modulation of MMP-2 secretion by trabecular meshwork cells. Invest Ophthalmol Vis Sci. 2002;43(9):3016–3020. [PubMed]
Tian B Gabelt BT Crosson CE Kaufman PL . Effects of adenosine agonists on intraocular pressure and aqueous humor dynamics in cynomolgus monkeys. Exp Eye Res. 1997;64(6):979–989. [CrossRef] [PubMed]
Crosson CE Sloan CF Yates PW . Modulation of conventional outflow facility by the adenosine A1 agonist N6-cyclohexyladenosine. Invest Ophthalmol Vis Sci. 2005;46(10):3795–3799. [CrossRef] [PubMed]
Avila MY Stone RA Civan MM . A(1)-, A(2A)- and A(3)-subtype adenosine receptors modulate intraocular pressure in the mouse. Br J Pharmacol. 2001;134(2):241–245. [CrossRef] [PubMed]
Crosson CE . Adenosine receptor activation modulates intraocular pressure in rabbits. J Pharmacol Exp Ther. 1995;273(1):320–326. [PubMed]
Crosson CE Gray T . Characterization of ocular hypertension induced by adenosine agonists. Invest Ophthalmol Vis Sci. 1996;37(9):1833–1839. [PubMed]
Husain S Shearer TW Crosson CE . Mechanisms linking adenosine A1 receptors and extracellular signal-regulated kinase 1/2 activation in human trabecular meshwork cells. J Pharmacol Exp Ther. 2007;320(1):258–265. [CrossRef] [PubMed]
Li A Leung CT Peterson-Yantorno K . Mechanisms of ATP release by human trabecular meshwork cells, the enabling step in purinergic regulation of aqueous humor outflow. J Cell Physiol. 2011 Mar 4. [Epub ahead of print]
Pang IH Shade DL Clark AF Steely HT DeSantis L . Preliminary characterization of a transformed cell strain derived from human trabecular meshwork. Curr Eye Res. 1994;13(1):51–63. [CrossRef] [PubMed]
Stamer WD Seftor RE Williams SK Samaha HA Snyder RW . Isolation and culture of human trabecular meshwork cells by extracellular matrix digestion. Curr Eye Res. 1995;14(7):611–617. [CrossRef] [PubMed]
Li A Leung CT Peterson-Yantorno K Mitchell CH Civan MM . Pathways for ATP release by bovine ciliary epithelial cells, the initial step in purinergic regulation of aqueous humor inflow. Am J Physiol Cell Physiol. 2010;299(6):C1308–C1317. [CrossRef] [PubMed]
Tschumperlin DJ Margulies SS . Equibiaxial deformation-induced injury of alveolar epithelial cells in vitro. Am J Physiol. 1998;275(pt 1):L1173–L1183. [PubMed]
Yantorno RE Coca-Prados M Krupin T Civan MM . Volume regulation of cultured, transformed, non-pigmented epithelial cells from human ciliary body. Exp Eye Res. 1989;49(3):423–437. [CrossRef] [PubMed]
Hawkes SP Li H Taniguchi GT . Zymography and reverse zymography for detecting MMPs and TIMPs. Methods Mol Biol. 2010;622:257–269. [PubMed]
Zhuo YH He Y Leung KW . Dexamethasone disrupts intercellular junction formation and cytoskeleton organization in human trabecular meshwork cells. Mol Vis. 2010;16:61–71. [PubMed]
Fleischhauer JC Mitchell CH Stamer WD . Common actions of adenosine receptor agonists in modulating human trabecular meshwork cell transport. J Membr Biol. 2003;193(2):121–136. [CrossRef] [PubMed]
Boudreault F Grygorczyk R . Cell swelling-induced ATP release is tightly dependent on intracellular calcium elevations. J Physiol. 2004;561(pt 2):499–513. [CrossRef] [PubMed]
Harteneck C Reiter B . TRP channels activated by extracellular hypo-osmoticity in epithelia. Biochem Soc Trans. 2007;35(pt 1):91–95. [PubMed]
Bao L Locovei S Dahl G . Pannexin membrane channels are mechanosensitive conduits for ATP. FEBS Lett. 2004;572(1–3):65–68. [CrossRef] [PubMed]
Bao L Sachs F Dahl G . Connexins are mechanosensitive. Am J Physiol Cell Physiol. 2004;287(5):C1389–C1395. [CrossRef] [PubMed]
Evans DW Hosking SL Embleton SJ Morgan AJ Bartlett JD . Spectral content of the intraocular pressure pulse wave: glaucoma patients versus normal subjects. Graefes Arch Clin Exp Ophthalmol. 2002;240(6):475–480. [CrossRef] [PubMed]
Zhang X Li A Ge J . Acute increase of intraocular pressure releases ATP into the anterior chamber. Exp Eye Res. 2007;85(5):637–643. [CrossRef] [PubMed]
Li A Zhang X Zheng D Ge J Laties AM Mitchell CH . Sustained elevation of extracellular ATP in aqueous humor from humans with primary chronic angle-closure glaucoma. Exp Eye Res. 2011 Jul 2. [Epub ahead of print]
Wang J Ma M Locovei S Keane RW Dahl G . Modulation of membrane channel currents by gap junction protein mimetic peptides: size matters. Am J Physiol Cell Physiol. 2007;293(3):C1112–C119. [CrossRef] [PubMed]
Silverman W Locovei S Dahl G . Probenecid, a gout remedy, inhibits pannexin 1 channels. Am J Physiol Cell Physiol. 2008;295(3):C761–C767. [CrossRef] [PubMed]
Bruzzone R Barbe MT Jakob NJ Monyer H . Pharmacological properties of homomeric and heteromeric pannexin hemichannels expressed in Xenopus oocytes. J Neurochem. 2005;92(5):1033–1043. [CrossRef] [PubMed]
Ma W Hui H Pelegrin P Surprenant A . Pharmacological characterization of pannexin-1 currents expressed in mammalian cells. J Pharmacol Exp Ther. 2009;328(2):409–418. [CrossRef] [PubMed]
Nishida M Sato Y Uemura A . P2Y6 receptor-Galpha12/13 signalling in cardiomyocytes triggers pressure overload-induced cardiac fibrosis. EMBO J. 2008;27(23):3104–3115. [CrossRef] [PubMed]
Ebihara L Liu X Pal JD . Effect of external magnesium and calcium on human connexin46 hemichannels. Biophys J. 2003;84(1):277–286. [CrossRef] [PubMed]
North RA . Molecular physiology of P2X receptors. Physiol Rev. 2002;82(4):1013–1067. [CrossRef] [PubMed]
Donnelly Roberts DL Jarvis MF . Discovery of P2X7 receptor-selective antagonists offers new insights into P2X7 receptor function and indicates a role in chronic pain states. Br J Pharmacol. 2007;151(5):571–579. [CrossRef] [PubMed]
Watanabe H Takahashi R Tran QK . Increased cytosolic Ca(2+) concentration in endothelial cells by calmodulin antagonists. Biochem Biophys Res Commun. 1999;265(3):697–702. [CrossRef] [PubMed]
Coco S Calegari F Pravettoni E . Storage and release of ATP from astrocytes in culture. J Biol Chem. 2003;278:1354–1362. [CrossRef] [PubMed]
Feranchak AP Lewis MA Kresge C . Initiation of purinergic signaling by exocytosis of ATP-containing vesicles in liver epithelium. J Biol Chem. 2010;285:8138–8147. [CrossRef] [PubMed]
Thyssen A Hirnet D Wolburg H Schmalzing G Deitmer JW Lohr C . Ectopic vesicular neurotransmitter release along sensory axons mediates neurovascular coupling via glial calcium signaling. Proc Natl Acad Sci U S A. 2010;107:15258–15263. [CrossRef] [PubMed]
Tokunaga A Tsukimoto M Harada H Moriyama Y Kojima S . Involvement of SLC17A9-dependent vesicular exocytosis in the mechanism of ATP release during T cell activation. J Biol Chem. 2010;285:17406–17416. [CrossRef] [PubMed]
Luna C Li G Qiu J Challa P Epstein DL Gonzalez P . Extracellular release of ATP mediated by cyclic mechanical stress leads to mobilization of AA in trabecular meshwork cells. Invest Ophthalmol Vis Sci. 2009;50(12):5805–5810. [CrossRef] [PubMed]
Cooper JA . Effects of cytochalasin and phalloidin on actin. J Cell Biol. 1987;105(4):1473–1478. [CrossRef] [PubMed]
Filla MS Schwinn MK Nosie AK Clark RW Peters DM . Dexamethasone-associated cross-linked actin network formation in human trabecular meshwork cells involves beta3 integrin signaling. Invest Ophthalmol Vis Sci. 2011;52(6):2952–2959. [CrossRef] [PubMed]
Kregenow FM . The response of duck erythrocytes to nonhemolytic hypotonic media: evidence for a volume-controlling mechanism. J Gen Physiol. 1971;58(4):372–395. [CrossRef] [PubMed]
Hoffmann EK Lambert IH Pedersen SF . Physiology of cell volume regulation in vertebrates. Physiol Rev. 2009;89(1):193–277. [CrossRef] [PubMed]
Yegutkin GG . Nucleotide- and nucleoside-converting ectoenzymes: important modulators of purinergic signalling cascade. Biochim Biophys Acta. 2008;1783(5):673–694. [CrossRef] [PubMed]
Figure 1.
 
Cytoskeletal remodeling produced by prolonged DEX and brief CCD exposure. TM5 cells were treated with cytoskeleton-remodeling agents, including DEX (1 μM, ≥14 days) and CCD (25 μM, 1 hour). F-actin filaments were visualized by phalloidin staining (green), and nuclei were counterstained with DAPI (blue). Compared with nontreated controls (NT; A, D), DEX treatment strengthened actin filaments (B) and generated CLANs (E, red arrow). In contrast, CCD rapidly disrupted cytoskeleton, making the actin filaments visible only in isolated, sparse areas (C, F). Scale bar, 20 μm.
Figure 1.
 
Cytoskeletal remodeling produced by prolonged DEX and brief CCD exposure. TM5 cells were treated with cytoskeleton-remodeling agents, including DEX (1 μM, ≥14 days) and CCD (25 μM, 1 hour). F-actin filaments were visualized by phalloidin staining (green), and nuclei were counterstained with DAPI (blue). Compared with nontreated controls (NT; A, D), DEX treatment strengthened actin filaments (B) and generated CLANs (E, red arrow). In contrast, CCD rapidly disrupted cytoskeleton, making the actin filaments visible only in isolated, sparse areas (C, F). Scale bar, 20 μm.
Figure 2.
 
Effects of DEX and CCD on swelling-induced ATP release. TM5 cells were exposed to cytoskeleton-remodeling agent DEX (1 μM, ≥14 days or 3 days) or CCD (25 μM, 1 hour). Prolonged DEX treatment enhanced actin polymerization and decreased swelling-triggered ATP release, whereas CCD produced opposite results. Short-term DEX treatment was ineffective. Isotonic baselines were unaffected by the drugs, compared with nontreated (NT) controls. *P < 0.05, one-way ANOVA. Numbers of wells analyzed are entered under the x-axis.
Figure 2.
 
Effects of DEX and CCD on swelling-induced ATP release. TM5 cells were exposed to cytoskeleton-remodeling agent DEX (1 μM, ≥14 days or 3 days) or CCD (25 μM, 1 hour). Prolonged DEX treatment enhanced actin polymerization and decreased swelling-triggered ATP release, whereas CCD produced opposite results. Short-term DEX treatment was ineffective. Isotonic baselines were unaffected by the drugs, compared with nontreated (NT) controls. *P < 0.05, one-way ANOVA. Numbers of wells analyzed are entered under the x-axis.
Figure 3.
 
Unchanged efficacy of blockers of swelling-evoked ATP release in TM5 cells after cytoskeletal restructuring. Inhibitors of previously identified conduits for hypotonicity-evoked ATP release, including probenecid (PRO, 0.1 mM, PX1 blocker), heptanol (HEP, 1 mM, Cx blocker), NPPB (30 μM, nonselective chloride channel blocker), and KN62 (1 μM, P2RX7 blocker) displayed the same efficacy in nontreated (NT), DEX-treated (1 μM, ≥14 days), and CCD-treated (1 μM, 1 hour) cells. The combination of carbenoxolone (CBX, 30 μM), HEP (1 mM), and KN62 (1 μM) nearly abolished this release, whereas bafilomycin A1 (BAF, 2 μM, vesicular ATP release blocker) was ineffective. P > 0.05 vs. NT controls by one-way ANOVA. The numbers of wells analyzed in the study are entered under the x-axis.
Figure 3.
 
Unchanged efficacy of blockers of swelling-evoked ATP release in TM5 cells after cytoskeletal restructuring. Inhibitors of previously identified conduits for hypotonicity-evoked ATP release, including probenecid (PRO, 0.1 mM, PX1 blocker), heptanol (HEP, 1 mM, Cx blocker), NPPB (30 μM, nonselective chloride channel blocker), and KN62 (1 μM, P2RX7 blocker) displayed the same efficacy in nontreated (NT), DEX-treated (1 μM, ≥14 days), and CCD-treated (1 μM, 1 hour) cells. The combination of carbenoxolone (CBX, 30 μM), HEP (1 mM), and KN62 (1 μM) nearly abolished this release, whereas bafilomycin A1 (BAF, 2 μM, vesicular ATP release blocker) was ineffective. P > 0.05 vs. NT controls by one-way ANOVA. The numbers of wells analyzed in the study are entered under the x-axis.
Figure 4.
 
Effects of prolonged DEX treatment and transient CCD exposure on the RVD of TM5 cells. TM5 cells were treated with DEX (1 μM, ≥14 days) and CCD (1 μM, 1 hour) before challenge by hypotonicity (50%). The trajectories are three-parameter exponential fits. Results were significantly different among the DEX-treated (n = 4), CCD-treated (n = 4), and nontreated (NT) control (n = 3) cells (P < 0.05 by two-way ANOVA). The DEX group had a much shorter period for 50% recovery after hypotonic exposure (T1/2) compared with the NT control group, whereas the CCD group had the longest.
Figure 4.
 
Effects of prolonged DEX treatment and transient CCD exposure on the RVD of TM5 cells. TM5 cells were treated with DEX (1 μM, ≥14 days) and CCD (1 μM, 1 hour) before challenge by hypotonicity (50%). The trajectories are three-parameter exponential fits. Results were significantly different among the DEX-treated (n = 4), CCD-treated (n = 4), and nontreated (NT) control (n = 3) cells (P < 0.05 by two-way ANOVA). The DEX group had a much shorter period for 50% recovery after hypotonic exposure (T1/2) compared with the NT control group, whereas the CCD group had the longest.
Figure 5.
 
Correlation of drug inhibitions on hypotonicity- and stretch-elicited ATP release. TM5 cells were mechanically stretched in the chambers, and the following drugs were used to block the release: PRO (0.1 mM, n = 21), HEP (1 mM, n = 9), KN62 (1 μM, n = 18), BAF (2 μM, n = 11), and a combination of 30 μM CBX, 1 mM HEP, and 1 μM KN62 (n = 21). The trajectory is a linear least-squares fit, with a slope of 1.04 ± 0.02 (P < 0.001) and a correlation coefficient of 0.999, indicating that the same conduits are recruited in both swelling- and stretch-elicited ATP release.
Figure 5.
 
Correlation of drug inhibitions on hypotonicity- and stretch-elicited ATP release. TM5 cells were mechanically stretched in the chambers, and the following drugs were used to block the release: PRO (0.1 mM, n = 21), HEP (1 mM, n = 9), KN62 (1 μM, n = 18), BAF (2 μM, n = 11), and a combination of 30 μM CBX, 1 mM HEP, and 1 μM KN62 (n = 21). The trajectory is a linear least-squares fit, with a slope of 1.04 ± 0.02 (P < 0.001) and a correlation coefficient of 0.999, indicating that the same conduits are recruited in both swelling- and stretch-elicited ATP release.
Figure 6.
 
Multiple membrane-type nucleotide ectoenzymes expressed by trabecular meshwork cells. RT-PCR with subsequent DNA sequencing verified that both TM5 and HTM cells express CD73, E-NPP1, E-NPP2, E-NPP3, E-NTPD2, and E-NTPD8, and several transcript variants were identified as well. No product was detectable on omitting reverse-transcriptase [RTase(-)], confirming that the cDNA derived from total RNA was free of genomic DNA contamination.
Figure 6.
 
Multiple membrane-type nucleotide ectoenzymes expressed by trabecular meshwork cells. RT-PCR with subsequent DNA sequencing verified that both TM5 and HTM cells express CD73, E-NPP1, E-NPP2, E-NPP3, E-NTPD2, and E-NTPD8, and several transcript variants were identified as well. No product was detectable on omitting reverse-transcriptase [RTase(-)], confirming that the cDNA derived from total RNA was free of genomic DNA contamination.
Figure 7.
 
MMP-2 and MMP-9 secretion modulated by adenosine A1 receptors and cytoskeletal remodeling. (A) Stimulating A1ARs with either specific agonist CHA (100 nM) or wide-spectrum AR agonist NECA (1 μM) increased MMP-2 and MMP-9 secretion. Both stimulations were blocked by concomitant application of the selective A1AR antagonist DPCPX (20 nM). Numbers of independent experiments are indicated along the abscissa. (B) Increasing actin polymerization with DEX (1 μM, ≥14 days) reduced MMP-2 and MMP-9 secretion considerably, and the reductions were significantly reversed by activating A1ARs with CHA (100 nM). Applying CCD (25 μM, 1 hour) raised MMP-9 secretion without affecting total MMP-2. DPCPX (20 nM) lowered both MMP-9 and total MMP-2 secretion of CCD-treated cells. Numbers of independent experiments are indicated along the abscissa. (C). The zymographic gel image shows the changes in MMP secretion produced by CCD (1 μM), an increase in the proform of MMP-9, and the appearance of the active form of MMP-2.
Figure 7.
 
MMP-2 and MMP-9 secretion modulated by adenosine A1 receptors and cytoskeletal remodeling. (A) Stimulating A1ARs with either specific agonist CHA (100 nM) or wide-spectrum AR agonist NECA (1 μM) increased MMP-2 and MMP-9 secretion. Both stimulations were blocked by concomitant application of the selective A1AR antagonist DPCPX (20 nM). Numbers of independent experiments are indicated along the abscissa. (B) Increasing actin polymerization with DEX (1 μM, ≥14 days) reduced MMP-2 and MMP-9 secretion considerably, and the reductions were significantly reversed by activating A1ARs with CHA (100 nM). Applying CCD (25 μM, 1 hour) raised MMP-9 secretion without affecting total MMP-2. DPCPX (20 nM) lowered both MMP-9 and total MMP-2 secretion of CCD-treated cells. Numbers of independent experiments are indicated along the abscissa. (C). The zymographic gel image shows the changes in MMP secretion produced by CCD (1 μM), an increase in the proform of MMP-9, and the appearance of the active form of MMP-2.
Figure 8.
 
Cytoskeleton-dependent purinergic regulation of aqueous humor outflow. Swelling-activated ATP release is recognized to proceed through PX1 and Cx hemichannels and P2X7 purinergic receptors. 20 The present study tests showed mechanical stretch triggers release through the same conduits, whereas remodeling actin cytoskeleton does not modulate this process. The present study also demonstrated ecto-enzymes CD73, E-NPP1–3, E-NTPD2, and E-NTPD8, which convert ATP to adenosine, are expressed in TM cells, consequently activating A1AR to stimulate the secretion of MMP-9 in addition to MMP-2.
Figure 8.
 
Cytoskeleton-dependent purinergic regulation of aqueous humor outflow. Swelling-activated ATP release is recognized to proceed through PX1 and Cx hemichannels and P2X7 purinergic receptors. 20 The present study tests showed mechanical stretch triggers release through the same conduits, whereas remodeling actin cytoskeleton does not modulate this process. The present study also demonstrated ecto-enzymes CD73, E-NPP1–3, E-NTPD2, and E-NTPD8, which convert ATP to adenosine, are expressed in TM cells, consequently activating A1AR to stimulate the secretion of MMP-9 in addition to MMP-2.
Figure sf01, PDF
Figure sf02, PDF
Table st1, PDF
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×