Role of Anions in Nitric Oxide-Induced Short-Circuit Current Increase in Isolated Porcine Ciliary Processes

Abstract

purpose. To investigate how nitric oxide (NO) modulates short-circuit current (Isc) in isolated porcine ciliary processes.

methods. Isc changes (Ussing-type chamber) induced either by the NO donors SNP or SIN-1, or by the cGMP analogue 8-pCPT-cGMP were assessed. The effect of inhibitors of guanylate cyclase (10 μM ODQ, 100 μM LY83583), protein kinase G (30 μM Rp-8-pCPT-cGMP, 3 μM KT 5823), protein kinase A (1 μM KT 5720), or protein kinase C (1 μM Go6983) on SNP- or 8-pCPT-cGMP–induced Isc changes were investigated. The effect of inhibitors of anion channel (100 μM niflumic acid, 1 mM DIDS, and 1 mM 9-AC), K⁺-channel (10 mM TEA, 10 mM BaCl₂), Na⁺-channel blockers (1 mM amiloride), Na⁺-K⁺-2Cl⁻ cotransporter inhibitor (0.5 mM bumetanide), or carbonic anhydrase inhibitor (1 mM acetazolamide) was studied. In Cl⁻- or HCO₃ ⁻-free Krebs-Ringer solution, the effect of SNP- or 8-pCPT-cGMP–induced Isc changes was accessed.

results. SNP, SIN-1, or 8-pCPT-cGMP increased Isc with a change in the potential difference that became more negative toward the nonpigmented epithelium (aqueous) side. The Isc increase induced by SNP or SIN-1, but not by 8-pCPT-cGMP, was prevented by ODQ and LY83583. SNP- and 8-pCPT-cGMP–induced Isc increases were prevented by Rp-8-pCPT-cGMP or KT5823 (but not by KT5720 or Go6983), or by niflumic acid, DIDS, 9-AC, or acetazolamide (but not by TEA, BaCl₂, amiloride, or bumetanide). The effect of SNP and 8-pCPT-cGMP was abolished in Cl⁻- and reduced in HCO₃ ⁻-free solutions.

conclusions. NO activates a guanylate cyclase-cGMP-protein kinase G pathway that appears to stimulate stroma-to-aqueous anionic transport, possibly Cl⁻, in porcine ciliary epithelium.

In the eye, ionic transport across the ciliary process epithelium, followed by a secondary passive transport of water, is considered to be instrumental in the process of aqueous humor formation. ¹ In this regard, the anions chloride (Cl⁻) and bicarbonate (HCO₃ ⁻) appear to play a major role. For example, a net flux of Cl⁻ across the ciliary processes (from the stroma to the aqueous side) has been observed in cat, ² rabbit, ³ and bovine ⁴ eyes. Furthermore, at the level of the ciliary pigmented epithelium (PE), at least in rabbit ⁵ and bovine, ⁶ a transmembrane exchange of Cl⁻ with HCO₃ ⁻ appears to be a possible mechanism by which Cl⁻ enters from the stroma into the ciliary epithelium. 

Nitric oxide (NO) modulates transepithelial anionic and fluid transport. For example, through the activation of guanylate cyclase (GC) leading to an increase of guanosine 3′,5′-cyclic monophosphate (cGMP) production, NO stimulates Cl⁻ secretion in salivary glands, ⁷ hepatocytes, ⁸ colon, ⁹ and respiratory tract. ^{10 11} In porcine ciliary epithelium, it has been shown that activation of the NO-cGMP pathway induces transmembrane potential depolarization, an effect prevented by anion channel blockers. ¹² However, to investigate further the role played by NO in the process of aqueous humor formation, the question of whether NO can induce a net transepithelial flux of anions from the PE to the non–pigmented epithelium (NPE) side (stroma-to-aqueous humor) remains open. 

Short-circuit current (Isc), which can be measured in an Ussing-type chamber, can reflect transepithelial ionic flux. ¹³ In the present study we investigated how NO modulates Isc in isolated porcine ciliary processes and/or epithelium bilayers and determined the potential role played by Cl⁻ or HCO₃ ⁻ in this process. 

Materials and Methods

Tissue Preparation

In compliance with the ARVO Statement for the Use of Animals in Ophthalmology and Vision Research, porcine eyes were obtained from a local abattoir immediately after the animal’s death. Eyes were cut at the equator, and the anterior half of each eye was placed in a Petri dish with the cornea facing down. Under a dissecting microscope, the lens and the remnant retina, choroid, and vitreous were removed. Ciliary processes with the underlying stroma (approximately one fifth of the circumferential ciliary body ring) were then dissected from the sclera and the iris. 

For the preparation of ciliary epithelium bilayers, eyes were first perfused through the two posterior ciliary arteries for 2 minutes with oxygenated Dulbecco’s modified Eagle’s medium (DMEM) and then for 15 minutes with DMEM containing 0.1% collagenase and 0.05% hyaluronidase. After this procedure, eyes were prepared as just described, except that after dissecting the remnant vitreous, the epithelium bilayer was obtained by peeling it from the underlying stroma. Some epithelium bilayer preparations were subject to a histologic examination at the end of the experimental protocol described below to confirm that the bilayers were free of the underlying stroma and that the gross morphologic integrity of the tissue had not been altered during the experiments (Fig. 1) . 

Ussing-Type Chamber Experiments

The technique for the Isc measurement has been described in the literature. ¹³ In brief, experiments were conducted with a self-modified, four-port Ussing-type chamber that had an exposure area of 0.10 cm². The chamber was made of two hemichambers, each with two ports, one for a voltage electrode and the other one for a current electrode. Electrodes (Ag/AgCl) filled with 3% agar-3 M KCl gel were connected to a DVC-1000 voltage/current clamp apparatus (World Precision Instruments, Sarasota, FL). The two hemichambers were bathed with 5 mL modified Krebs-Ringer physiologic solution (mM: NaCl, 118.05; KCl, 4.69; KH₂PO₄, 1.20; MgSO₄, 1.20; NaHCO₃, 25.01; EDTA, 0.026; CaCl₂, 2.51; glucose, 11.10; and HEPES, 10.00; in 37°C and 95% O₂-5% CO₂; pH 7.4). The system was calibrated (∼20 minutes) by monitoring any voltage difference between the two electrodes and adjusting any difference to zero. Tissues were then mounted between the two hemichambers filled with modified Krebs-Ringer physiologic solution. The short-circuit current (Isc) was measured with the current electrodes and continuously monitored on a chart recorder. From time to time, the transepithelial potential difference (PD) was also measured through the two voltage electrodes (tips placed within 2 to 3 mm of the tissue surface). Electrical resistance (ER) was determined from the current deflection induced by a 500-μV voltage offset through the voltage electrodes. 

In Cl⁻-free Krebs-Ringer solution, Cl⁻ was replaced by equimolar gluconate. In HCO₃ ⁻-free Krebs-Ringer solution, HCO₃ ⁻ was replaced by equimolar Cl⁻ and bubbled with 100% O₂. 

Experimental Protocols

Once mounted in the Ussing-type chamber, preparations were left quiescent for approximately 45 minutes until Isc reached a steady state (baseline). Only preparations with a negative baseline PD on the NPE side when bathed with physiologic Krebs-Ringer solution were considered for further experimental protocols. In the first set of experiments, either on the NPE, the PE, or both sides, ciliary processes and epithelium bilayer preparations were exposed to the NO-donors sodium nitroprusside (SNP), or 3-morpholinosydnonine (SIN-1), or the cGMP analogue 8-pCPT-cGMP. First, the response over time after a single dose of SNP (100 μM), SIN-1 (1 mM), or 8-pCPT-cGMP (100 μM) was assessed. Then, dose–response curves were constructed in a cumulative and increasing concentration manner (SNP: 100 nM to 100 μM, SIN-1: 100 nM to 1 mM, 8-pCPT-cGMP: 10 nM to 100 μM). In the second set of experiments, on the NPE side, ciliary processes were incubated (10 minutes) with the GC inhibitors 1-H-(1,2,4)oxadiazole(4,3-α)quinoxalin-1-1 (ODQ; 10 μM) or LY83583 (100 μM) and then exposed to a single dose of SNP (100 μM), SIN-1 (1 mM), or 8-pCPT-cGMP (0.1 mM). In the third set of experiments, on the NPE side, ciliary processes were incubated (20 minutes) with either the protein kinase G (PKG) inhibitor Rp-8-pCPT-cGMP (30 μM), KT 5823 (3 μM), the protein kinase A (PKA) inhibitor KT 5720 (1 μM), or the protein kinase C (PKC) inhibitor Go6983 (1 μM). Then, the preparations were exposed to SNP (100 nM to 100 μM) or 8-pCPT-cGMP (10 nM to 100 μM). In the fourth set of experiments, ciliary processes were incubated (20–30 minutes) either with the anion channel blockers niflumic acid (100 μM), 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS; 1 mM), anthracene-9-carboxylic acid (9-AC; 1 mM), the K⁺ channel blockers tetraethylammonium chloride (TEA; 10 mM), BaCl₂ (10 mM), or the Na⁺ channel blocker amiloride (1 mM). All anion channel blockers were applied to the NPE side, and all cation channel blockers were applied to both sides of the ciliary processes. Preparations were then exposed to increasing concentrations of either SNP (100 nM to 100 μM) or 8-pCPT-cGMP (10 nM to 100 μM). In the fifth set of experiments, ciliary processes were bathed in either Cl⁻- or HCO₃ ⁻-free Krebs-Ringer solution until a steady baseline of Isc was achieved. Then, the dose–response curve of SNP (100 nM to 100 μM) or 8-pCPT-cGMP (10 nM to 100 μM) on the NPE side was constructed. In the sixth and last set of experiments, either on the NPE or the PE side, ciliary processes were incubated (30 minutes) with either the Na⁺-K⁺-2Cl⁻ cotransporter inhibitor bumetanide (0.5 mM) or the carbonic anhydrase inhibitor acetazolamide (1 mM). Preparations were then exposed to increasing concentrations of either SNP (100 nM to 100 μM) or 8-pCPT-cGMP (10 nM to 100 μM). 

Drugs

Collagenase, hyaluronidase, SNP, SIN-1, niflumic acid, 9-AC, TEA, amiloride, BaCl₂, sodium gluconate, potassium gluconate, calcium gluconate, bumetanide, acetazolamide, and dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich (Buchs, Switzerland). ODQ, LY83583, KT5823, KT5720, Go6983, and DIDS were obtained from Calbiochem-Novabiochem (Läufigen, Switzerland). 8-pCPT-cGMP and Rp-8-pCPT-cGMP were bought at Biolog-life Science Institute (Bremen, Germany). All drugs were freshly prepared on the day of the experiments. The drugs were dissolved in modified Krebs-Ringer physiologic solution except for 9-AC, TEA, amiloride, ODQ, LY83583, Go6983, and DIDS, which were dissolved in DMSO. The final concentration of DMSO in the solution never exceeded 1%. 

Data and Statistical Analysis

Results are expressed as the mean ± SEM with the abbreviation n corresponding to the number of eyes assessed (one preparation per eye). The EC₅₀, or the concentration causing 50% of the maximum response, was calculated with the sigmoid-fit function of commercial software (S-Plus, ver. 2000; Statsoft, Tulsa, OK). Statistical comparisons were conducted with one-way ANOVA multiple-comparison followed by the Bonferroni correction with a two-tailed P < 0.05 considered to be statistically significant. 

Results

SNP-, SIN-1–, and 8-pCPT-cGMP–Induced Isc Changes

In ciliary process preparations, when applied to the NPE side, a single dose of SNP (100 μM), or SIN-1 (1 mM), or the cGMP analogue 8-pCPT-cGMP (100 μM) induced a rapid peak increase in Isc (∼ 2 minutes) followed by a plateau that was sustained for at least 30 minutes (Fig. 2) . When these drugs were applied to the PE side of the preparations, only a slight increase in Isc was observed (Fig. 2) . 

It has to be stressed that in the present study attention has only been focused on the peak Isc increase evoked by SNP, SIN-1, or 8-pCPT-cGMP. When different concentrations of these three drugs were applied to the NPE side, the concentration–response curve of the peak Isc increase could be constructed (Fig. 3) for SNP (EC₅₀: 25.3 ± 1.0 μM), SIN-1 (EC₅₀: 162.9 ± 5.6 μM), and 8-pCPT-cGMP (EC₅₀: 4.5 ± 0.3 μM). In contrast, when applied to the PE side, the concentration–response curve of the peak Isc increase for SNP, SIN-1, and 8-pCPT-cGMP were significantly (P < 0.001) lower than when applied to the NPE side (Fig. 3) . When these drugs were applied simultaneously on the both sides of the ciliary processes, the peak Isc increase was not significantly different from the one observed when the drugs were applied only on the NPE side (Fig. 3) . 

In addition, it has to be noted that the increase of Isc was associated with a change in PD from baseline that always became more negative toward the NPE side (100 μM SNP: −622.2 ± 58.7 μV; 1 mM SIN-1: −618.2 ± 69.4 μV; 100 μM 8-pCPT-cGMP: −755.8 ± 38.8 μV). The latter observation indicates that the Isc increase induced by these drugs could be associated with a net excess transport of negative ionic charges toward the NPE side, a net excess transport of positive ionic charges toward the PE side, or both. Finally, no significant (P > 0.32) change in electrical resistance was observed when SNP, SIN-1, or 8-pCPT-cGMP were applied to the NPE, the PE, or both sides, probably because baseline resistances were so low that precise changes in transepithelial resistance might have been underestimated. 

As was the case with the ciliary processes, in ciliary epithelium bilayer preparations, SNP (EC₅₀: 28.8 ± 1.3 μM), SIN-1 (EC₅₀: 164.5 ± 7.2 μM), or 8-pCPT-cGMP (EC₅₀: 4.0 ± 0.3 μM) induced a concentration-dependent increase of Isc when applied to the NPE side (Fig. 4) . Also, the effect of SNP, SIN-1, and 8-pCPT-cGMP on the Isc increase was significantly (P < 0.001) lower when the drugs were applied to the PE than when applied to the NPE side (Fig. 4) . These results indicate that apparently the underlying stroma is not responsible for the Isc response difference observed after these drugs have been applied either on the NPE or the PE side. 

Effect of GC Inhibitors on SNP, SIN-1, or 8-pCPT-cGMP–Induced Isc Increase

In the ciliary process preparations, in the presence of the guanylate cyclase inhibitors ODQ or LY83583, the Isc increase induced by SNP (100 μM) or SIN-1 (1 mM), but not by 8-pCPT-cGMP (100 μM), was significantly (P < 0.001) prevented (Fig. 5) . These results indicate that the increase of Isc in porcine ciliary processes induced by SNP or SIN-1 involve the activation of a guanylate cyclase. 

Effect of Protein Kinases Inhibitors on the SNP- or 8-pCPT-cGMP–Induced Isc Increase

The increase of Isc induced by SNP or 8-pCPT-cGMP was significantly (P < 0.05–0.001) prevented by the PKG inhibitors Rp-8-pCPT-cGMP and KT5823, but was almost unaffected by the PKA and PKC inhibitors KT5720 and Go6983, respectively (Fig. 6) . In particular, the concentration–response curve of 8-pCPT-cGMP (EC₅₀: 5.0 ± 0.3 μM) was significantly shifted to the right by Rp-8-pCPT-cGMP (EC₅₀: 81.9 ± 20.8 μM, P < 0.001) or KT5823 (EC₅₀: 14.0 ± 2.5 μM, P < 0.05; Fig. 6B ). These results indicate that the Isc increase induced by SNP or 8-pCPT-cGMP is likely to involve the activation of a PKG. 

Effect of Ion Channel- or Transporter Blockers on the SNP- or 8-pCPT-cGMP–Induced Isc Increase

The Isc increase induced by SNP or 8-pCPT-cGMP was virtually abolished (P < 0.001) in the presence of DIDS (an assumed Cl⁻ channel inhibitor, ¹⁴ Cl⁻-HCO₃ ⁻ exchanger inhibitor, ¹⁴ and/or Na⁺-HCO₃ ⁻ co-transporter inhibitor ¹⁵ ) or niflumic acid (a potent Cl⁻ channel blocker ^{14 16} ), and were strongly (P < 0.001) reduced by 9-AC (a potent Cl⁻ channel blocker ¹⁷ ) when applied to the NPE side (Fig. 7) . In contrast, neither the presence of the Na⁺ channel blocker amiloride nor the K⁺ channel blocker TEA, BaCl₂ applied to the NPE side (data not shown) even on both sides of the preparations had any significant effect on the Isc increase induced by SNP or 8-pCPT-cGMP (Fig. 7) . These results indicate that in porcine ciliary processes, the increase of Isc induced by SNP or 8-pCPT-cGMP is likely to involve anions rather than cations. If one refers to the observation made herein that the increase in Isc induced by these drugs was associated with an increase in the negativity of the PD on the NPE side, it can be deduced that, apparently, the NO-donor and the cGMP analogue induce a net transfer of anions from the PE (stroma) to the NPE (aqueous humor) side. 

Effect of Anion Substitution on an SNP- or 8-pCPT-cGMP–Induced Isc Increase

In experiments in which Cl⁻ was replaced by gluconate in Krebs-Ringer solution (Cl⁻-free Krebs-Ringer solution), an increase in Isc baseline was observed. Once this new steady state baseline Isc has been achieved, the Isc increase induced by SNP or 8-pCPT-cGMP was abolished only when the solution was applied to the PE (or on both sides) but not on the NPE side of ciliary processes (Fig. 8) . In experiments in which HCO₃ ⁻ was replaced by Cl⁻ in Krebs-Ringer solution (HCO₃ ⁻-free Krebs-Ringer solution) the Isc increase induced by SNP or 8-pCPT-cGMP was significantly reduced when this solution was applied to the PE, the NPE, or both sides (Fig. 8) . 

Regarding the experiments with HCO₃ ⁻-free Krebs-Ringer solution, it has to be noted that the reduction in Isc change was less pronounced when it was applied to the PE side than when applied to the NPE side. It addition, the effect was even more pronounced when the HCO₃ ⁻-free Krebs-Ringer solution was applied to both sides at the same time. 

In other words, these results further support the involvement of anions, in particular Cl⁻ and HCO₃ ⁻, in the Isc increase induced by the NO donor SNP and the cGMP analogue 8-pCPT-cGMP. These results also indicate that these drugs may evoke a transfer of Cl⁻ from the stromal (PE) to the aqueous humor (NPE) side of porcine ciliary processes. 

Effect of Na⁺-K⁺-2Cl⁻ Cotransporter Inhibitor or Carbonic Anhydrase Inhibitor on the SNP- or 8-pCPT-cGMP–Induced Isc Increase

It has been recently reported in the literature that the Na⁺-K⁺-2Cl⁻ cotransporter inhibitor, in particular Cl⁻, plays an important role in ion transport in the ciliary processes either on the PE ¹⁸ or the NPE ^{14 16} side. In the present study, the increase of Isc induced by SNP or 8-pCPT-cGMP was not significantly affected by bumetanide when applied either to the NPE or the PE side (Fig. 9) . The 0.5 mM concentration of bumetanide tested was higher than concentrations reported to have an inhibitory effect in bovine ¹⁸ or rabbit ¹⁴ ciliary processes and also in porcine airways. ¹⁹ These results indicate that an Na⁺-K⁺-2Cl⁻ cotransporter is not likely to be involved in transepithelial Cl⁻ transport induced by SNP and 8-pCPT-cGMP. 

In the presence of the carbonic anhydrase inhibitors acetazolamide, SNP- or 8-pCPT-cGMP–induced Isc increase was significantly reduced (Fig. 9) . These results suggest that carbonic anhydrase, potentially by modulating HCO₃ ⁻ formation and eventually possibly Cl⁻-HCO₃ ⁻ exchange, is involved in the mechanism of transepithelial anion transport induced by SNP or 8-pCPT-cGMP. 

Baseline Electrophysiological Parameters

To be complete and for information, as limited information is available in the literature regarding measurements conducted in porcine ciliary epithelium preparations in an Ussing-chamber, the electrophysiological parameters (Isc, PD, and ER) in baseline conditions are shown in Table 1 . It should be noted that when Cl⁻-free or HCO₃ ⁻-free Krebs-Ringer solution was applied to the PE side or on both sides of the ciliary processes, the polarities of PD reversed and ER increased. Such a change in PD has also been described in rabbit ²⁰ and bovine ⁴ ciliary processes after depletion of HCO₃ ^{− 20 21} or reduction of Cl^{− 4} in the bathing solution. Although the components as well as the mechanisms of baseline Isc are still not fully understood, these results suggest that both Cl⁻ and HCO₃ ⁻ are essential in transepithelial Isc and/or PD built up in porcine ciliary processes. One of the explanations for the PD polarity reversal in bilateral Cl⁻-free Krebs-Ringer solution could be, as reported in the literature on the rabbit ciliary epithelium, ^{22 23} that the electrogenic Na⁺-K⁺ pump current is greater on the NPE than the PE side. If this is the case in the porcine ciliary epithelium, such an asymmetry in pump–current activity could reverse the polarity in preparations that are not mediating a net Cl⁻ transport. Further investigation is needed to test this hypothesis. Finally, it has also to be noted that in ciliary epithelium bilayer preparations, as expected, the baseline ERs were significantly (P < 0.001) lower than those of the ciliary process preparations, indirectly reflecting the ER of the stroma. 

The influence of the ion channel/transporter blockers on baseline Isc and/or electrical resistance are summarized in Table 2 . All ion channel/transporter blockers had induced a change in baseline Isc, while significant increase in ER were observed only when 10 mM BaCl₂ was applied simultaneously to both sides of the ciliary processes. As is the case in other animal species, ¹ these results stress the involvement of ion channel and transporters in ion transport in porcine ciliary epithelium. For DMSO (vehicle-control), it has to be noted that only when 1% DMSO was applied to the NPE side of the tissue preparations was a notable decrease in the baseline Isc observed, without any significant change in ER (Table 2) . Nevertheless, even at the maximum concentration used (<1%), DMSO never had any significant effect on the Isc or potential difference evoked by SNP, SIN-1, or 8-pCPT-cGMP. 

Discussion

The present study shows that the NO donors SNP and SIN-1 andthe cGMP analogue 8-pCPT-cGMP increased Isc in isolated porcine ciliary processes as well as in epithelium bilayers. Furthermore, with these drugs the PD became more negative toward the NPE side. The SNP- and SIN-1–induced (but not the 8-pCPT-cGMP–induced) increase of Isc was prevented by the GC inhibitors ODQ and LY83583. The SNP- and 8-pCPT-cGMP–induced Isc increases were significantly prevented by the PKG inhibitors Rp-8-pCPT-cGMP or KT5823, but not by the PKA inhibitor KT5720 or the PKC inhibitor Go6983. The SNP- and 8-pCPT-cGMP–induced Isc increase was significantly prevented by the anionic channel blockers DIDS, 9-AC, or niflumic acid but not by the cationic channel blockers amiloride, TEA or BaCl₂. Finally, the SNP- and 8-pCPT-cGMP–induced increase in Isc was almost abolished when Cl⁻-free Krebs-Ringer solution was applied to the PE side of ciliary processes and significantly prevented when HCO₃ ⁻-free Krebs-Ringer solution was applied either on the PE, NPE, or both sides of the preparation. Taken together, these observation suggests that NO activates a GC-cGMP-PKG pathway that induces a transepithelial flux of anions, that appears to involve at least the anion Cl⁻, from the stromal (PE) to the aqueous humor (NPE) side of isolated porcine ciliary processes. 

This observation is in line with a series of other reports indicating that NO may be involved in the production of aqueous humor in porcine ciliary processes. Indeed, by immunostaining, it has been demonstrated that neuronal NO synthase (NOS) and inducible NOS are present in the porcine ciliary epithelium. ²⁴ Furthermore, reflecting the biological activity of NOS, the presence of a basal NO production has been demonstrated in fresh and thawed, isolated porcine ciliary processes by a porphyrinic NO microsensor. ²⁵ It has also been shown that in thawed isolated porcine ciliary processes, but not in the iris, norepinephrine (α-, β-adrenoreceptor agonist) increases nitrite production (a metabolite of NO), an effect prevented by N ^G-nitro-l-arginine methyl ester (l-NAME; an inhibitor of NO formation) or the ocular hypotensive drug propranolol (β-adrenoreceptor antagonist). ²⁶ Using the same experimental setup, it was also shown that drugs increasing cAMP stimulate nitrite production through a mechanism involving both PKA and NOS. ²⁷ It was further demonstrated that the α₂-adrenoreceptor agonist brimonidine (another ocular hypotensive drug) inhibits nitrite production through a mechanism involving a pertussis toxin (PTX)-sensitive G₁ protein and an α₂-adrenoreceptor. ²⁸ Finally, by electrophysiology, it has been shown that NO, through the activation of GC and an increase in cGMP production, depolarizes porcine ciliary epithelium membrane potential. ¹²  

In addition to these reports, in the present study, NO increased at least the transepithelial flux of Cl⁻ from the stromal to the aqueous humor side of isolated porcine ciliary processes. Indeed, the NO donor SNP and the cGMP analogue 8-pCPT-cGMP induced a change of PD that became more negative toward the NPE side of the preparation. Such a change in PD either reflects the presence of a net transport of anions from the PE to the NPE side or, on the contrary, a net transport of cations from the NPE to the PE side, or both. As the increase of Isc induced by SNP and 8-pCPT-cGMP was prevented by anionic channel blockers or by Cl⁻ depletion on the PE side, it appears that anionic currents, and to a certain extent Cl⁻, are essential in the effect elicited by these two drugs. In other words, these observations provide indirect evidence that NO leads to an increase in aqueous humor production, at least in isolated porcine ciliary processes. 

In line with the findings in the present study, it has to be stressed that not only in ciliary processes but also in nonocular tissues it has been reported that NO can modulate anionic transepithelial transport and therefore fluid secretion and/or reabsorption. For example, in human airway epithelial cells, it has been demonstrated that through a guanylyl cyclase-dependent pathway, the NO donors S-nitrosoglutathione and S-nitroso-N-acetyl-d,l-penicillamine increase whole-cell Cl⁻ currents. ¹¹ In rat colon it has also been shown that SNP increases anion secretion and inhibits sodium chloride reabsorption. ⁹ Finally, in rat salivary glands ⁷ and hepatocytes, ⁸ stimulation of anionic secretion induced by activation of the NO-cGMP pathway has also been demonstrated. 

Anions such as Cl⁻ and HCO₃ ⁻ are considered to play an important role in the process of aqueous humor formation. In particular, active transport of Cl⁻ across ciliary epithelium is regarded to be the rate-limiting process in aqueous humor formation. ²¹ Indeed, a net transport of Cl⁻ across ciliary epithelium has been observed in feline, ² rabbit, ³ and bovine eyes. ^{4 29} In contrast to Cl⁻, until now it has been difficult to demonstrate the existence of a net transepithelial transport of HCO₃ ⁻. However, HCO₃ ⁻ is believed to play a potential role in the modulation of Cl⁻ secretion, such as through the involvement of Cl⁻-HCO₃ ⁻ exchanger or carbonic anhydrase located either on the PE or the NPE side of the ciliary epithelium. ^{5 6} The results of the present study indirectly suggest that the two latter mechanisms are involved in the modulation of Isc increase induced by SNP or 8-pCPT-cGMP. Indeed, the increase of Isc induced by these two drugs was significantly inhibited in the presence of HCO₃ ⁻-free Krebs-Ringer solution or the carbonic anhydrase inhibitor acetazolamide. Further investigations should be conducted to clarify the role played by HCO₃ ⁻ in the Isc modulation induced by NO. 

It is interesting to note that in the present study the Isc increase induced by the NO donors or the cGMP analogue was different when applied to the NPE or the PE side of the ciliary processes (epithelial bilayer). A peak increase of Isc followed by a plateau of the increase was observed when these drugs were applied to the NPE side, whereas this biphasic response was not observed when drugs were applied to the PE side. Furthermore, the peak response induced either by the NO donors or the cGMP analogue was more pronounced (P < 0.001) when drugs were applied to the NPE than when applied to the PE side. One may assume that the characteristics of the increase in the Isc peak and the plateau periods, as well as their response to the pathway inhibitors, may be very different. The reasons for those differences are still unclear, especially as it is usually assumed that NO, as well as the cell-permeable cGMP analogue 8-pCPT-cGMP, should in theory be able to pass rather freely through cell membranes. As this observation could be made both in ciliary processes (with the underlying stroma on the PE side) and in ciliary epithelium bilayers, it can be excluded that the stroma plays a role in such a difference. It has to be noted that an asymmetry in the response to the cGMP analogue when the drug was applied to the NPE or the PE side has also been observed, but in isolated rabbit ciliary processes. ³⁰  

IOP is the result of a balance between inflow and outflow of aqueous humor in the eye. There are observations indicating that NO reduces IOP and increases aqueous humor outflow. ^{31 32} These findings are not in contradiction with the observation indicating that NO increases aqueous humor production. Indeed, even if NO increases both outflow and inflow of aqueous humor and if it increases more the former than the latter, the overall effect would then be a decrease in IOP. Such a situation is for example known to occur with the ocular hypotensive epinephrine (α-, β-adrenoreceptor agonist) used to reduce IOP in patients with glaucoma. ³³ Indeed, epinephrine increases both aqueous inflow ³⁴ and outflow ³⁵ ; but, because it increases the outflow more than the inflow, it leads to an overall decrease of IOP. 

In conclusion, through the activation of a GC-cGMP-PKG pathway it appears that NO increases the transepithelial transport of anion, in particular Cl⁻, from the stromal side to the aqueous humor side of porcine ciliary epithelium. This observation supports the hypothesis that NO may play a role in the modulation of aqueous humor production. ^{12 22 23 24 25 26}  

 Supported by the Swiss National Science Foundation 32-61495.00, Bern, Switzerland.

 Submitted for publication November 17, 2003; revised April 23, 2004; accepted May 6, 2004.

 Disclosure: R. Wu, None; K. Yao, None; J. Flammer, None; I.O. Haefliger, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 Corresponding author: Ivan O. Haefliger, Laboratory of Ocular Pharmacology and Physiology, University Eye Clinic Basel, Mittlere Strasse 91, PO Box CH-4012 Basel, Switzerland.

 

Figure 1.

 

Histologic preparation of porcine ciliary epithelium bilayers obtained at the end of an experiment conducted in an Ussing-type chamber. It clearly shows that the pigmented epithelium was free from the underlying stroma and that the gross morphologic integrity of the epithelium bilayers was maintained. Hematoxylin-eosin staining. Magnification: (A) ×50; (B) ×200.

Figure 2.

 

Original tracings of single Isc measurements over time in isolated porcine ciliary processes after exposure to a single dose of SNP, SIN-1, or 8-pCPT-cGMP. When applied to the NPE side, SNP, SIN-1, and 8-pCPT-cGMP induced an increase of Isc (A, C, E). When SNP, SIN-1, or 8-pCPT-cGMP was applied to the PE side, the increase in Isc was much lower (B, D, F).

Figure 3.

 

Effect of SNP, SIN-1, or 8-pCPT-cGMP on Isc measured in isolated porcine ciliary processes when drugs were applied to the NPE, the PE, or both sides of the preparation. In a concentration-dependent manner, SNP (A), SIN-1 (B), or 8-pCPT-cGMP (C) increased Isc, an effect that was more pronounced when the drugs were applied to the NPE side or to both sides than when applied to the PE side. Peak Isc increase: *P < 0.05, **P < 0.01, ***P < 0.001 versus the PE side.

Figure 4.

 

SNP, SIN-1, or 8-pCPT-cGMP on Isc measured in isolated porcine ciliary epithelium bilayers. In a concentration-dependent manner, SNP (A), SIN-1 (B), and 8-pCPT-cGMP (C) increased Isc, an effect more pronounced when the drugs were applied to the NPE side than when applied to the PE side. Peak Isc increase: *P < 0.05, **P < 0.01, ***P < 0.001 versus the PE side.

Figure 5.

 

Effect of the guanylate cyclase inhibitors ODQ and LY83583 on the peak Isc increase induced by application of SNP, SIN-1, and 8-pCPT-cGMP on the NPE side of isolated porcine ciliary processes. The presence of ODQ or LY83583 significantly prevented the increase of Isc induced by SNP or SIN-1, but not that induced by 8-pCPT-cGMP. ***P < 0.001 versus control.

Figure 6.

 

Effect of different PK inhibitors on SNP- or 8-pCPT-cGMP–induced peak Isc increase in isolated porcine ciliary processes when drugs were applied to the NPE side of the preparation. When preincubated with the PKG inhibitor Rp-8-pCPT-cGMP or KT5823, the increase in Isc induced by SNP (A) was inhibited, and the dose–response curves of 8-pCPT-cGMP (B) was shifted to the right. In contrast, neither the PKA inhibitor KT5720 (C) nor the PKC inhibitor Go6983 (D) had a significant effect on the SNP- or 8-pCPT-cGMP–induced Isc increase. *P < 0.05, **P < 0.01, ***P < 0.001 versus control.

Figure 7.

 

Effect of ion channel transporter blockers on SNP- or 8-pCPT-cGMP–induced peak Isc increase in isolated porcine ciliary processes. The Isc increase induced by SNP (A) or 8-pCPT-cGMP (B) was significantly prevented by the anionic channel blockers, 9-AC, niflumic acid, or DIDS, when these blockers were applied to the NPE side. When the Na⁺ channel blocker amiloride or the K⁺ channel blockers TEA or BaCl₂ were applied to both sides of the ciliary processes, the effect of SNP (C) or 8-pCPT-cGMP (D) was not significantly influenced. **P < 0.01, ***P < 0.001 versus control.

Figure 8.

 

Effect of Cl⁻- or HCO₃ ⁻- free Krebs-Ringer solution on the SNP- and 8-pCPT-cGMP–induced peak Isc increases in isolated porcine ciliary processes. In the presence of chloride-free Krebs-Ringer solution on the PE side or on both sides, but not on the NPE side, of the preparation, the Isc increase induced by either SNP (A) or 8-pCPT-cGMP (B) was abolished. In the presence of bicarbonate-free Krebs-Ringer solution, the Isc increase induced by either SNP (C) or 8-pCPT-cGMP (D) was significantly reduced. This effect was more pronounced when the bicarbonate-free Krebs-Ringer solution was applied to both sides > NPE > PE. *P < 0.05, **P < 0.01, ***P < 0.001 versus control (physiological Krebs-Ringer solution).

Figure 9.

 

Effect of the Na⁺-K⁺-2Cl⁻ cotransporter inhibitor bumetanide or the carbonic anhydrase inhibitor acetazolamide on the SNP- or 8-pCPT-cGMP–induced peak Isc increase in isolated porcine ciliary processes. When bumetanide was applied on either the NPE or the PE side, the Isc increase induced by SNP (A) and 8-pCPT-cGMP (B) was not significantly affected. When acetazolamide was applied to either the NPE or the PE side, the effect of SNP (C) and 8-pCPT-cGMP (D) was significantly reduced. *P < 0.05, **P < 0.01, ***P < 0.001 versus control.

Table 1.

 

Electrophysiological Parameters Measured in Baseline Conditions with an Ussing-Type Chamber in Isolated Porcine Ciliary Processes and Ciliary Epithelium Bilayers

	Ciliary Processes													Ciliary Epithelium Bilayers
		Krebs-Ringer	CI−-free Krebs-Ringer					HCO3 −-free Krebs-Ringer					Krebs-Ringer
				NPE Side	PE Side	Both Sides	NPE Side	PE Side	Both Sides
Eyes (n)	350	13	12	12	12	12	12	36
Isc (μA/cm2)	13.12 ± 0.36	25.42 ± 2.15	−35.70 ± 4.33	−16.91 ± 1.59	27.54 ± 2.90	−33.31 ± 3.93	−14.71 ± 1.36	12.24 ± 0.81
PD (μV)	−830.7 ± 23.5	−1782.8 ± 207.6	2592.0 ± 336.7	1231.7 ± 143.6	−1812.0 ± 187.0	2474.8 ± 330.0	1038.8 ± 100.1	−551.9 ± 40.3
ER (Ohm · cm2)	64.26 ± 0.71	69.25 ± 4.48	73.13* ± 4.13	72.35* ± 3.62	66.90 ± 3.78	73.41* ± 3.17	71.32 ± 2.56	45.24, *** ± 1.38

*   P < 0.05,

***   P < 0.001 vs. Krebs-Ringer solution (ANOVA).

Table 2.

 

Influence of Ion Channel/Transporter Inhibitors on Baseline Isc and ER in Isolated Porcine Ciliary Processes

	Eyes (n)	Change in Baseline Isc (μA/cm2)	Change in ER (Ohm · cm2)
Control with K-R (5 μl, NPE side)	6	0.08 ± 0.24	0.17 ± 1.57
Control with K-R (5 μl, both sides)	8	0.05 ± 0.36	−0.26 ± 0.96
DMSO (0.1%, NPE side)	7	−0.06 ± 0.21	0.26 ± 0.85
DMSO (0.1%, both sides)	5	0.18 ± 0.25	−0.12 ± 1.76
DMSO (1%, NPE side)	6	−2.34 ± 0.55	−0.42 ± 1.43
Amiloride (1 mM, both sides)	6	1.62 ± 0.49	1.25 ± 0.78
BaCl2 (10 mM, both sides)	12	12.55 ± 0.76	4.79 ± 0.31
TEA (10 mM, both sides)	10	−4.12 ± 0.37	0.22 ± 0.78
9-AC (1 mM, both sides)	8	−7.05 ± 1.05	1.10 ± 1.48
DIDS (1 mM, NPE side)	9	−21.42 ± 2.0	0.48 ± 0.58
Niflumic acid (100 μM, NPE side)	9	−4.12 ± 0.37	0.64 ± 1.48
Bumetanide (0.5 mM, NPE side)	9	−2.70 ± 0.35	0.91 ± 0.81
Bumetanide (0.5 mM, PE side)	6	1.22 ± 0.34	1.20 ± 0.68

 K-R, Krebs-Ringer physiologic solution.

The authors thank Andreas Schötzau for help with statistical analyses and Peter Meyer, MD, for help in the histologic assessment of the ciliary epithelium bilayers.