Hydraulic Pressure Stimulates Adenosine 3′,5′-Cyclic Monophosphate Accumulation in Endothelial Cells from Schlemm’s Canal

Abstract

purpose. Fluid flow across various endothelia results in a variety of intracellular and extracellular adaptations. In the living eye, aqueous humor flows across the surface of endothelial cells on trabecular meshwork (TM) beams and in the juxtacanalicular tissue and through or between a continuous monolayer of endothelial cells that line Schlemm’s canal (SC). This study was undertaken to test the hypothesis that fluid flow induces biochemical changes in the endothelial cells of the outflow pathway that may modify outflow resistance.

methods. Trabecular meshwork and SC cells isolated from the outflow pathway of human cadaveric eyes were seeded onto porous filters, placed in Ussing-type chambers, and subjected to fluid flow driven by a pressure head of 15 mm Hg on their apical surface. Cell lysates were prepared and analyzed for adenosine 3′,5′-cyclic monophosphate (cAMP) accumulation. Barrier function of cell monolayers was examined using transendothelial electrical resistance measurements.

results. Three different SC cell strains in 14 independent experiments responded with at least a threefold increase in cAMP that was both time and pressure dependent. Conversely, flow-treated TM cells failed to respond in six independent experiments in which five different TM cell strains were used. Electrical resistance across cell monolayers positively correlated with cAMP accumulation and was calcium sensitive.

conclusions. cAMP signaling is affected by pressure differentials across SC cell monolayers and provides evidence for the participation of SC cells in the regulation of aqueous outflow.

Mechanotransduction describes the conversion of mechanical force applied outside of a cell into chemical signals inside a cell. ^{1 2 3} This mode of communication enables cells to monitor and impact their external environments. For example, endothelial cells that line blood vessels monitor fluid flow and/or pressure and initiate long-term adaptations in vessel diameter. ⁴ Such adaptations appear to result from biochemical changes that are induced by mechanical deformation of the endothelium. These effects include basic cellular functions such as ion transport, secretion, protein synthesis, expression of specific genes, and second-messenger activation. ^{5 6 7 8 9}  

Similar to endothelial cells that line blood vessels of the vascular system, the endothelial cells that line the outflow pathway of the eye are subject to fluid flow and/or pressure gradients. Unlike a homogenous and continuous monolayer of vascular cells, there are at least two types of outflow pathway cells, trabecular meshwork (TM) and Schlemm’s canal (SC) cells. Only SC cells form a continuous barrier to fluid flow that may produce a pressure gradient across its monolayer. Conversely, TM cells on trabecular beams and in the juxtacanalicular region are located mostly proximal to this pressure gradient across SC and thus probably experience flow but no pressure gradient. We were interested to determine how these two types of outflow cells would respond to fluid flow and/or pressure in vitro and whether they use it as a mechanism to sense and regulate outflow resistance, much as vascular endothelial cells regulate vascular tone. 

Although much is known about the signaling pathways that regulate blood flow through vessels of the vascular system, little is known about the regulation of aqueous outflow in the eye. Studies that implicate the vascular endothelium as a modulator of vascular tone have been performed with a variety of model systems that apply controlled forces to monolayers of vascular endothelial cells in vitro. Studies that implicate a cellular component in the management of aqueous outflow have been limited because indications of pressure effects were described as morphologic changes in the outflow pathway ex vivo. ^{10 11 12 13 14} In the present study, a controlled-force fluid flow was delivered to confluent monolayers of primary cultures of SC and TM cells using hydraulic pressure in an Ussing-type chamber. A signaling molecule, adenosine 3′,5′-cyclic monophosphate, cAMP, was monitored in these cells as a marker for mechanotransduction. 

Methods

Isolation and Culture of Human SC and TM Cells

Human cadaveric eye tissue was obtained from the National Disease Research Interchange (NDRI, Philadelphia, PA) within 48 hours of death for whole eyes stored in moist chambers, and 96 hours for nontransplantable corneal anterior segments stored in solution (Optisol, Chiron Vision, Clairmont, CA). 

Human SC cells were isolated from cadaveric eye tissue using gelatin-coated cannulas, as described previously. ¹⁵ Briefly, The anterior chamber of human cadaveric eyes was cut into eight equal and radially symmetrical wedge-shaped pieces. Using a dissecting microscope (Carl Zeiss, Thornwood, NY), a gelatin-coated suture (6-0 sterile nylon monofilament, Wilson Ophthalmic, Mustang, OK) was gently inserted into the lumen of SC and advanced into the canal. The cannulated pieces of tissue were placed in culture (Dulbecco’s modified Eagle’s medium [DMEM]), containing 10% fetal bovine serum, 100 U/ml penicillin G sodium, and 100 μg/ml streptomycin sulfate (Life Technologies, Grand Island, NY) and maintained at 37°C in humidified air containing 7% CO₂ for at least 3 weeks. Sutures were removed from SC and cells seeded onto 3-cm culture plates. The cell strains used in this study were isolated from nonglaucomatous donor eye tissue from three different donors (SC3, SC6, and SC7) of ages 55, 45, and 50 years, and all have been characterized previously. ¹⁵  

Human TM cells were isolated using a blunt dissection technique in conjunction with extracellular matrix digestion and cultured as previously described. ¹⁶ The cell strains used in this study were isolated from nonglaucomatous donor eye tissue from five different donors (HTM19, HTM22, HTM23, HTM27, and HTM29) of ages 83, 55, 72, 19, and 0 years, respectively. 

Flow Model

Human SC or TM cell isolates were plated at a density of 2.5 × 10⁴ cells/well onto inserts (Snapwell; Costar, Acton, MA) that consist of plastic supports surrounding polycarbonate or polyester filters (1-cm² surface area with 0.4-μm pore diameter). Confluence was visualized by phase-contrast microscopy using an inverted microscope (IM-35; Zeiss). Cells were maintained for 6 to 7 days in DMEM containing 10% serum at 37°C in humidified air with 7% CO₂. Cells were washed twice with serum-free medium and remained in 20 mM HEPES (Sigma, St. Louis, MO)-buffered DMEM (serum-free; pH 7.4) for 16 hours in an air incubator at 37°C. Medium was aspirated from the inserts, and cells were incubated with 0.5 mM 3-isobutyl-1-methylxanthine (IBMX; Sigma) in HEPES-DMEM for 1 minute at 37°C. Cells on filter supports (Snapwell; Costar) were either placed in an Ussing-type chamber (Ussing system CHM5; World Precision Instruments, Sarasota, FL) and mounted in a clamp housing (Fig. 1) to receive “flow”, or they remained in IBMX medium as“ no-flow” control samples. Some no-flow cells were treated with 10μ M forskolin (Calbiochem, La Jolla, CA) for 10 minutes at 37°C and served as positive controls for AC activation. 

The Ussing-type chamber was connected to a fluid reservoir that was filled with HEPES-DMEM medium (serum-free plus 0.5 mM IBMX) to a height 0, 6.7, 13.4, or 20.1 cm (0, 5, 10, or 15 mm Hg) above the center of the cell monolayers. The Ussing-type chamber had two compartments on either side of the cell monolayers, and flow through cell monolayers was a function of the pressure decrease between compartments equivalent to the height of the column of fluid. The apical surface of cell monolayers faced the front compartment, which was medium filled and continuous with the fluid reservoir; the back compartment was open to atmospheric pressure. Cells were subjected to flow for 5, 10, or 20 minutes, filter supports were removed from the chamber, and filters were excised from the plastic supports and analyzed for intracellular cAMP. Cell monolayers were excluded from the present study if greater than 250 μl of effluent appeared in the back compartment of the flow apparatus (i.e., resulting in flow rates of more than 12.5μ l/min·cm² and indicating a compromised cell monolayer). Typical effluent volumes ranged between 50 and 100 μl. 

Measurement of cAMP

Intracellular cAMP was measured using a protein kinase A (PKA) binding assay, as described previously, with few modifications. ¹⁷ Filters with cells were placed in microfuge tubes containing 120 μl ice-cold Tris/EDTA buffer (50 mM Tris/4 mM EDTA, pH 7.5), boiled for 10 minutes, and centrifuged in a microcentrifuge at maximum speed for 2 minutes (approximately 12,000g). The supernatants (50 μl) were added to 50 μl[ ³H]cAMP (0.9 picomoles/50 μl; NEN, Boston, MA) and 100 μl cold PKA solution (0.06 mg PKA/ml Tris/EDTA; Sigma). After a 2-hour incubation at 4°C, 100 μl of activated charcoal solution (20 mg/ml activated charcoal in Tris/EDTA containing 2% bovine serum albumin) was added and the mixture vortexed and centrifuged at maximum speed in a microcentrifuge for 1 minute at 22°C. Samples were placed on ice, and 200 μl of each supernatant was transferred to scintillation vials for counting. A standard curve was generated by adding 50 μl of cAMP standards (0.25–32.0 picomoles; Sigma), instead of cytosol, to PKA solution with[ ³H]cAMP. 

Measurement of Electrical Resistance

Human SC or TM cell isolates were plated onto inserts (Transwell; Costar) as described above (Transwells are identical with Snapwells except for the design of the plastic support). The cells were maintained in DMEM containing 10% serum at 37°C in humidified air with 7% CO₂. Sixteen hours before measurements were taken, cells were washed and incubated in serum-free HEPES-DMEM in a room-air incubator at 37°C. Filters were transferred to a transendothelial resistance measurement chamber (ENDOHM-12; World Precision Instruments) that was filled with serum-free HEPES-DMEM. Alternating current (20 μA at 12.5 Hz) was delivered between two planar and positionally fixed electrodes on either side of cell monolayers, and electrical resistance was measured using an ohmmeter (Millicell-ERS; Millipore, Bedford, MA). The background electrical resistance generated by medium and filter insert was measured and subtracted from measurements of filters with cells. 

Results

The objective of this study was to test the hypothesis that cells of the outflow pathway of the human eye change biochemically in response to fluid flow and/or a hydraulic pressure gradient. These mechanical forces were modeled using an Ussing-type chamber that contained monolayers of human SC or TM cells on porous filters. Phase-contrast micrographs of SC and TM cell monolayers on porous filters are shown in Figure 2 . Fluid flow through cell monolayers was driven by a pressure gradient of 15 mm Hg for 20 minutes. The rates of fluid flow (2–5μ l/min·cm²) were calculated from the volume of effluent collected in the rear flow chamber. An intracellular signaling molecule, cAMP, was measured as an indicator of the conversion of extracellular mechanical forces to intracellular chemical signals. 

Figure 3 shows the effect of fluid flow and/or pressure on cells of the outflow pathway 2 days (B) or 7 days (A, C) after cell monolayers had reached confluence. Fluid flow/pressure stimulated cAMP accumulation in primary cultures of SC cells (A) but not TM cells (C). After 10 minutes of fluid flow/pressure, three SC cell strains responded with at least a twofold stimulation of cAMP. This increased to threefold after 20 minutes compared with nonflow control samples. Schlemm’s canal cells were also tested 2 days after confluence (B) and failed to respond to fluid flow/pressure with a significant increase in cAMP after 20 minutes. Additionally, monolayers of TM cells were tested 2 weeks after confluence and failed to respond with significant cAMP accumulations (data not shown). In all experiments nonflow cells were treated with 10μ M forskolin for 10 minutes at 37°C to stimulate AC, indicate viability of cells, and serve as positive control for cAMP accumulation. 

Figure 4 shows the effect of increasing pressure on the accumulation of cAMP in SC cells. Monolayers of SC cells were exposed to hydraulic pressures of 5, 10, and 15 mm Hg for 20 minutes. Levels of cAMP accumulation in SC cell monolayers increased in a pressure-dependent manner. Positive, but not significant responses were noted with a pressure of 5 mm Hg compared wtih nonflow controls. Statistically significant responses were measured at pressures of 10 and 15 mm Hg. 

Barrier function of monolayers of human SC and TM cells on porous filters was tested using measurements of transendothelial electrical resistance (TEER). Figure 5 shows that SC cells provide a net resistance of 12 to 15Ω /cm² when tested at least 1 week after confluence (Figs. 5B 5C) . This barrier was significantly different from TM cells whose resistance (approximately 2Ω /cm²) did not differ significantly from the resistance generated by medium and porous filters without cells (Fig. 5A) . The electrical barrier of SC cell monolayers was sensitive to agents that affect cell–cell junctions. Figure 5B shows the effect of 200 μM ethacrynic acid (ECA) on TEER of SC cell monolayers. Similarly, the TEER of SC monolayers is sensitive to the presence of a calcium chelator, EDTA, at a 4-mM concentration (Fig. 5C) . 

Discussion

The assumption that the cells of the outflow pathway actively participate in aqueous outflow regulation suggests that they sense and respond to changes in either pressure or fluid flow. ^{11 18 19 20 21} The link between the cellular-sensing mechanism(s) and mode(s) of response most likely involves an intracellular signaling pathway. In the present study, the intracellular signaling molecule, cAMP, was monitored in monolayers of TM and SC cells after treatments. cAMP was chosen as a marker because it is a convergence point for many second-messenger pathways and because of known associations between aqueous outflow function and cAMP. ^{22 23 24 25 26 27} In the present study, we observed cAMP accumulation in SC but not TM endothelial cells. Fluid flow driven by a pressure gradient across cell monolayers stimulated cAMP by at least twofold over the level in control samples in all three SC cell strains tested. Conversely, cells of five different TM strains failed to respond. 

The cAMP accumulations corresponded to significant TEER measurements across SC monolayers. The magnitude of TEER across monolayers of SC cells at least 1 week after confluence was similar to that in other endothelium, such as bovine aortic or human umbilical vein but less than certain tight epithelium. ^{28 29} At earlier time points (e.g., 2 days after confluence), SC cells failed to demonstrate either significant cAMP accumulation or TEER. This suggested the involvement of cell–cell associations that typically require several days to several weeks to mature. ³⁰ Furthermore, we observed that both ECA and EDTA significantly diminished TEER in SC monolayers. The latter effect indicated the presence of calcium-sensitive junctions in SC cells. Taken together, these observations suggest that pressure gradients across confluent SC cell monolayers require at least a week to form. Only after this time did SC cells respond with significant cAMP accumulation, indicating sensitivity to a pressure differential rather than fluid flow in this system. 

Several studies have suggested a regulatory role of SC cells in outflow resistance that involves either flow paracellularly through cell junctions or intracellularly through pores by means of a“ funneling” mechanism. ^{23 31 32} Evidence for the involvement of SC cell–cell junctions in outflow mechanisms comes from models in which drugs such as epinephrine, ECA, EDTA, and H7 have been shown to affect SC cell–cell associations and to influence SC monolayer permeability in vitro or aqueous outflow function in vivo. ^{23 33 34 35 36} These effects in vivo clearly involve SC but may also be in series with resistance changes in the juxtacanalicular region. Regardless, such functional data are consistent with in vivo morphologic data that demonstrate the presence of both gap and tight junctions in SC monolayers, whereas TM monolayers contain only gap junctions. ^{14 37 38}  

Differences in cell–cell attachments of TM in vitro compared with SC cell monolayers can explain differences in TEER and cAMP responses. For SC cells it appears that cell–cell attachments were important to generate, sense, and respond (in the form of cAMP accumulation) to pressure gradients across cell monolayers. For TM cell monolayers, a pressure gradient was probably never established as indicated by insignificant TEER and cAMP accumulations even 2 weeks after confluence (data not shown). However, because TM cells are positioned in the outflow pathway where they will more likely see fluid flow rather than pressure gradients, it is possible that TM cells respond only to fluid flow, and that signaling mechanisms other than cAMP are involved. 

In vivo, aqueous humor generally flows in an apical to basal direction across TM endothelium and a basal to apical direction across SC endothelium. Our system, which delivered flow in the apical to basal direction, modeled fluid flow more appropriately across TM rather than SC endothelium. This orientation of fluid flow was used for both cell types, because early experiments showed that fluid flow across SC monolayers in the basal to apical direction dislodged sections of SC cells off of their filter substratum using pressures as low as 5 mm Hg (data not shown). The necessity of another system to explore this specific variable became clear. In this context, our system may better represent backflow on the venous side of SC. Therefore, the cAMP stimulation observed may be related to a “one-way valve” response of SC cells in vivo. ³⁹ Alternatively, results in the present study may simply indicate a general mechanism by which SC cells interact with their neighboring cells and substratum and respond to changes in pressure. 

This hypothesis would be consistent with a proposed mechanism for the participation of AC in the process of mechanotransduction. ³ Forces such as fluid flow or pressure induce cellular deformations that are transmitted and focused onto AC through cytoskeletal filaments. The result is a perturbation of the tertiary structure of AC that alters the energetic equilibrium of its catalytic domain and results in the conversion of more adenosine triphosphate to cAMP. Consequently, deformation of the plasma membrane is dependent on cell–cell and cell–matrix anchorage, an arrangement that is supported by our flow-response data, TEER measurements, and sensitivity of TEER to EDTA. 

Presently, topical preparations of epinephrine-like compounds are the only commercially available drugs that increase aqueous outflow by directly affecting the conventional outflow pathway. ²⁷ The mechanism of action of epinephrine involves an adrenergic receptor-cAMP–dependent process in these outflow tissues and cells, implicating an active rather than passive regulation of outflow facility. ^{22 23 24 25 26} In the present study, the outflow pathway was modeled by delivering fluid flow driven by hydraulic pressure to TM and SC cell monolayers. The data presented suggest that a pressure decrease across SC cell monolayers stimulated cAMP accumulation in a pressure-dependent fashion. This finding provides evidence for the possibility that the outflow pathway of the human eye self-regulates in response to changes in pressure. 

 Reprint requests: W. Daniel Stamer, University of Arizona, 1801 N. Campbell Avenue Tucson, AZ 85719.

 Supported in part by Grants F32EY0685, RO1EY01894, and P30EY05722 from the National Eye Institute and grants from the American Health Assistance Foundation and Research to Prevent Blindness.

 Submitted for publication September 22, 1998; revised March 1, 1999; accepted March 23, 1999.

 Proprietary interest category: N.

¹  Present address: University of Arizona, Tucson, Arizona.

 

Figure 1.

 

Schematic representation of the perfusion apparatus used to deliver fluid flow and/or pressure to monolayers of endothelial cells. Fluid flows in the direction indicated by arrows from a reservoir to the apical surface of endothelial cell monolayers grown on a porous filter (oval inset) and housed in an Ussing-type chamber. The reservoir is positioned 20.1 cm above the center of cell monolayers to deliver a pressure equivalent to 15 mm Hg to their apical surface.

Figure 2.

 

Phase-contrast microscopy of cultured endothelial cells from the human outflow pathway grown on polyester filters. Monolayers of SC (SC3 passage 4, A) and TM (HTM27 passage 4, B) cells are shown 7 days after confluence. (C) A polyester filter without cells (inset is a ×10 enlargement of filter to demonstrate 0.4-μm pores [arrowhead]). Bar, 200μ m.

Figure 3.

 

The effect of fluid flow on cAMP accumulation in monolayers of endothelial cells from the human outflow pathway. (A) The effect of 10 and 20 minutes (F10′ and F20′, respectively) of fluid flow on cAMP accumulation in SC cells (SC6, passage 2, 7 days [7D ] after confluence) compared with nonflow control samples (BSL). (B) cAMP accumulation in SC cells 2 days (2D) after confluence. (C) The effect of 20 minutes of fluid flow on cAMP accumulation in TM cells 7 days after confluence. Treatment of cells with 10 μM forskolin (FSK) for 10 minutes served as a positive control in all cases for cAMP accumulation. The data represent the mean ± SEM of one experiment that has been repeated 14 times for (A), 4 times for (B), and 6 times for (C). *P < 0.01 and** P < 0.001 versus nonflow, one-atmospheric-pressure controls.

Figure 4.

 

Pressure–response relationship for cAMP accumulation in cultured endothelial cells from Schlemm’s canal. Fluid flow was driven through monolayers of SC cells by pressure heads of 0, 5, 10, or 15 mm Hg. Shown are the combined data ± SEM of four experiments using two SC cell strains (SC3 and SC6). Treatment of cells with 10 μM forskolin (FSK) for 10 minutes served as a positive control for cAMP accumulation. *P < 0.01 and** P < 0.001 versus nonflow controls.

Figure 5.

 

Transendothelial electrical resistance measurements across monolayers of cells of the outflow pathway. The electrical resistance (in ohms per square centimeter) of TM (HTM29, A) and SC (SC3, B, C) are shown 2 days and 9 days after confluence. Nine-day monolayers were tested, treated with either 200μ M ECA (A, B) or 4 mM EDTA (C), washed with medium, and tested again 1 minute later. The data represent the mean of one experiment ± SEM that has been repeated six times for (B) and four times for (A) and (C). *P < 0.001 versus control,** P < 0.001 versus 9-day experimental samples.

The authors thank Jeff Harris for his superb technical assistance.