Mechanisms of Action of Unoprostone on Trabecular Meshwork Contractility

Abstract

purpose. This study was performed to clarify the possible mechanism behind the ocular hypotensive effect of unoprostone isopropyl (Rescula; Novartis Ophthalmics AG, Basel, Switzerland), a new docosanoid that has been shown to reduce intraocular pressure (IOP) in patients with ocular hypertension or primary open-angle glaucoma. To gain insight into the possible mode of action, the effects of unoprostone on ciliary muscle (CM) and trabecular meshwork (TM) contractility, intracellular calcium levels, and membrane channels were investigated.

methods. The effects of unoprostone (M1 metabolite = free acid, 10⁻⁵ M) and endothelin (ET)-1 (10⁻⁹ M) on bovine TM (BTM) and ciliary muscle (CM) strips were investigated, by using a custom-made force-length transducer system. The effects of unoprostone and ET-1 (5 × 10⁻⁸ M) on intracellular Ca²⁺ mobilization in cultured human TM (HTM) were measured using fura-2AM as a fluorescent probe. Patch–clamp experiments were performed on HTM and BTM cells to investigate the unoprostone-dependent modulation of membrane currents.

results. In isolated TM and CM strips, unoprostone almost completely inhibited ET-induced contractions (TM: 2.9% ± 4.3% vs. 19.6% ± 5.7%, P < 0.05, n = 6; CM: 1.4%± 1.6% vs. 30.1% ± 5.3%, P < 0.01, n= 6; 100% = maximal carbachol-induced (10⁻⁶ M) contraction). However, neither carbachol-induced contraction nor baseline tension was affected by unoprostone. Furthermore, unoprostone had no effect on baseline intracellular calcium levels (baseline: 126 ± 45 nM versus unoprostone: 132 ± 42 nM, n= 8) in HTM cells. The endothelin-induced increase (679 ± 102 nM), however, was almost completely (P < 0.01) blocked by unoprostone (178 ± 40 nM). In patch–clamp recordings, unoprostone could be shown to double the amplitude of outward current (HTM: 200% ± 33%; n = 6; BTM: 179% ± 20%; n= 8). This effect was blocked by the specific inhibitor of maxi-K channels, iberiotoxin.

conclusions. This study presents evidence for direct interaction of unoprostone with the contractility of the TM and CM. This compound may lower IOP by affecting aqueous outflow, most probably conventional outflow pathways (i.e., TM) through inhibition of ET-dependent mechanisms. In addition, unoprostone interacts with the maxi-K channel. Although primarily Ca²⁺-sensitive signal-transduction pathways seem to be involved, effects of unoprostone on Ca²⁺-independent pathways and uveoscleral outflow cannot be excluded.

Unoprostone isopropyl is a novel docosanoid. It resembles the naturally occurring docosanoid metabolites of docosahexaenoic acid (DHA). The latter is an ω-3, polyunsaturated fatty acid abundantly located in the neural tissues of the retina and brain. ^{1 2} Topical application of unoprostone isopropyl ophthalmic solution (Rescula; Novartis Ophthalmics AG, Basel, Switzerland) decreases IOP in rabbits, monkeys, ^{3 4 5} and humans. ^{6 7} Furthermore, the compound has been shown to positively affect optic nerve head circulation in rabbits ⁸ and humans ⁹ and appears to have retinal neuroprotective properties in rats. ¹⁰ There is a debate about whether IOP reduction in humans induced by unoprostone results from increased conventional outflow through the TM or primarily from enhanced uveoscleral flow, ⁵ as it does in primates. 

The ciliary muscle (CM) and the trabecular meshwork (TM) have been shown to be contractile elements in the conventional outflow pathway that contribute to the regulation of outflow facility (for review see Ref. ¹¹ ). Because DHA, as a parent unoprostone molecule, has been shown to exert vasorelaxing effects, ^{12 13 14 15 16} we attempted to identify whether unoprostone has similar effects and what the molecular mechanisms are. DHA’s smooth muscle–relaxing properties have been suggested to be caused by suppressing receptor-mediated Ca²⁺ influx in vascular smooth muscle cells. ¹⁷ Improved endothelium-dependent relaxation and inhibition of endothelium-dependent contraction was demonstrated in a coronary artery model using normal animals fed a diet rich in ω-3 fatty acids. ^{18 19} A similar enhancement of endothelium-mediated vasodilation in atherosclerotic human coronary arteries was later demonstrated after only 3 weeks of treatment with fish oil rich inω -3 fatty acids (among which DHA is the most abundant). ¹⁴  

To clarify the ocular hypotensive effects of unoprostone, contractility measurements of bovine TM (BTM) and CM strips were performed. We attempted to clarify the signal-transduction pathways affected by unoprostone within these smooth muscle–like tissues, by using functional contractility measurements, measurements of intracellular calcium [Ca²⁺]_i, and patch–clamp recordings. 

Materials and Methods

Contractility Measurements

Enucleated bovine eyes were obtained from an abattoir and placed on ice. Small TM and CM strips were carefully dissected according to methods described previously. ²⁰ Isolated BTM and CM strips were approximately 2 to 4 mm long and 0.5 mm wide. The effects of agents on BTM and CM tension were obtained after the tissues had been allowed to rest under control conditions for at least 1 hour. Experiments were performed under constant temperature of 37°C and stable pH conditions. Isometric contractions were expressed relative to the response obtained with a maximally effective carbachol concentration of 10⁻⁶ M, which was tested in each tissue strip as a control (100% contractility). To determine the activity of unoprostone and/or endothelin (ET) or vehicle alone, the compounds were administered either to the precontracted tissues or to tissues at basal tension. The ionic concentrations (in millimolar) of Ringer’s solution were: 151 Na⁺, 5 K⁺, 1.7 Ca²⁺, 0.9 Mg²⁺, 131 Cl⁻, 0.9 SO₄ ²⁻, 1 H₂PO₄ ⁻, 28 HCO₃ ⁻, and 5 glucose. All solutions were gassed with 5% CO₂ in air, which resulted in a pH of 7.4. 

Cell Cultures

Human TM (HTM) cells were isolated by methods based on those of Grierson et al. ²¹ and Siegner et al. ²² Human eyes were obtained from multiorgan donors of our hospital who were identified according to age, gender, and time of death. Donor ages ranged from 20 to 68 years, and a history of glaucoma was ruled out by screening patients’ histories. The tenets of the Declaration of Helsinki were followed, informed consent obtained, and institutional human experimentation committee approval granted for the studies. BTM cells were obtained in a similar fashion with methods previously described. ²³  

HTM and BTM cells were cultured in Dulbecco’s modified Eagle’s minimum essential medium (DMEM) supplemented with 20% fetal calf serum (FCS), 100 U/ml penicillin, and 100 μg/ml streptomycin (all cell culture material was obtained from Biochrom, Berlin, Germany). The cells were kept in an atmosphere containing 95% air-5% CO₂ at 37C° temperature and were passaged using the trypsin-EGTA method. Only well characterized, normal HTM cells from early passages (2–4) were used for patch–clamp recordings and fura-2AM dye calcium measurements. ¹¹  

Measurement of [Ca²⁺]_i

Measurements of[ Ca²⁺]_i were performed using the Ca²⁺-sensitive dye fura-2AM (Sigma, Deisenhofen, Germany) based on methods described by Grynkiewicz et al. ²⁴ Cells were trypsinized as described earlier and remained on coverslips in culture medium for at least 1 week. Before each experiment, the semiconfluent cells were incubated in control solution and 10 mM fura-2AM for 20 minutes at room temperature. The dye was loaded by diffusion and intracellular cleavage of fura-2AM to fura-2. The coverslip was then placed in a perfusion chamber on the stage of the inverted microscope, which was also used for the patch–clamp recordings. Cells were perfused with control solution for at least 30 minutes to wash out extracellular dye. The excitation light was generated by a xenon lamp (XPO 75 W/2; Osram, Carl Zeiss, Oberkochen, Germany) filtered by two rotating filters (six per second) at 340 nm and 380 nm. Relative fluorescence of the dye after excitation was registered at 510 nm by means of a photomultiplier (928 SF; Hamamatsu, Hamamatsu City, Japan) with consequent signal detection with a patch–clamp amplifier (EPC-9; Heka Electronics, Lambrecht, Germany). Data storage and data processing were performed on computer (TIDA for Windows; Heka Electronics). Changes in the 340:380-nm fluorescence ratio represent relative changes in[ Ca²⁺]_i. Absolute[ Ca²⁺]_i was calculated using the equation and dissociation constant of Grynkiewicz et al. ²⁴  

Patch–Clamp Recordings

The patch–clamp experiments were performed at 37°C, as previously described. ^{25 26} A coverslip with trypsinized cells was introduced into a perfusion chamber on the stage of an inverted microscope (Axiovert 35; Carl Zeiss) and superfused continuously with variable solutions. Borosilicate patch pipettes (Clark GC 150T-15; Harvard Apparatus, Kent, UK) were pulled and polished using a universal puller (DMZ; Zeitz, Augsburg, Germany). The input resistance of the pipette filled with the standard solution was 4 to 5 MΩ. Nystatin (150 μg/ml) in the patch pipette was used to obtain perforated patches. Membrane currents were recorded using a patch–clamp amplifier (EPC-9; Heka Electronics) patch–clamp amplifier. Electrical stimulation, data storage, and processing were performed on computer (TIDA for Windows; Heka Electronics). The same software was used to automatically calculate membrane capacitance and access resistance. Potentials were referenced to the bath, by using an Ag-AgCl electrode connected to the bath solution by an agar bridge electrode, so that a negative potential corresponded to a negative pipette potential. Liquid junction potentials were determined and corrected according to the method described by Neher. ²⁷  

Unless indicated otherwise, positive ions flowing into the pipette correspond to a negative current and are depicted in figures as going downward, whereas positive ions flowing out of the pipette are designated by a positive current in the upward direction. Currents were monitored using a protocol that generated steps of 200-msec duration to various voltages between −80 and 100 mV. After each step, the voltage returned to a holding potential of −40 mV for 200 msec. Currents were continuously sampled at 100 Hz throughout the entire duration of this protocol. Control perfusion solution (Ringer’s) contained (in millimolar): 151 NaCl, 4 KCl, 1.7 CaCl₂, 1 KH₂PO₄, 0.9 MgSO₄, 10 HEPES, and 5 glucose adjusted to pH 7.4 (NaOH). Standard pipette solution contained (in millimolar): 119 potassium glutamate, 10 NaCl, 1 KH₂PO₄, 0.9 MgSO₄, 3.3 EGTA, 6.6 CaEGTA, and 10 HEPES, adjusted to pH 7.2 with NaOH. 

Chemicals and Solution

The following reagents were used for the experiments: ET-1, (Alexis Deutschland GmbH, Grünberg, Germany) and unoprostone isopropyl (unoprostone free acid, stock solution of the M1 metabolite dissolved in dimethyl sulfoxide; Novartis). All other chemicals were purchased from Merck (Darmstadt, Germany), Sigma, and Serva (Heidelberg, Germany). 

Calculations and Statistical Analysis

Data are presented as mean ± SEM and were analyzed for significance using the paired Student’s t-test. Significance was assumed at the following probabilities, as shown in the figures and table: *P < 0.05, **P < 0.01, ***P < 0.001, on computer (Sigma Plot Scientific Graph System, ver. 1.02, SPSS Science, San Rafael, CA). 

Results

Effects of Unoprostone on Carbachol- and ET-Induced Contractility

As has been shown before, ²⁰ the muscarinic agonist carbachol 10⁻⁶ M led to a contraction in TM and CM strips (Figs. 1 2 3 and 4) . Unoprostone 10⁻⁵ M had no influence on baseline tension in both tissues (TM: −0.5% ± 4.2% vs. 100%; CM:− 1.1% ± 1.2% vs. 100% precontracted tissue with carbachol 10⁻⁶ M, n = 5; P < 0.8 vs. zero contraction, data not shown), nor did it influence carbachol-induced contraction (TM: 89.7% ± 3.6% vs. 100%; n= 3; CM: 100.8% ± 4.9% vs. 100%; n = 5; data not shown). ET (10⁻⁹ M) caused contractions from baseline level in both tissues (TM: 19.6% ± 5.7%; CM: 30.1% ± 5.3%, n = 6; Figs. 1 3 ), which were completely blocked by unoprostone 10⁻⁵ M (TM: 2.9% ± 4.4%; CM: 1.4% ± 1.6%, n = 6, P < 0.1 vs. zero contraction, Figs. 2 4 ). A summary of results in all contractility experiments performed with unoprostone and ET in TM and CM strips is shown in Figure 5 . 

Effects of Unoprostone and ET on [Ca²⁺]_i

Baseline intracellular calcium in HTM cells was 127 ± 45 nM (n = 5) and is within the range published by our and other groups. ^{28 29} ET 5 × 10⁻⁸ M resulted in an increase of intracellular calcium to 679 ± 102 nM from baseline (n = 5; P < 0.05; Fig. 6A ). Unoprostone (10⁻⁵ M) itself led to a slight but not statistically significant increase in intracellular calcium from baseline (112 ± 42 nM to 132 ± 41 nM, n = 8) but was able to almost completely block the ET-induced increase in intracellular calcium (178 ± 40 nM, n = 8; P < 0.01, Fig. 6B ). All[ Ca²⁺]_i measurement data are summarized in Figure 7 . 

Unoprostone Effects on Membrane Currents

Membrane currents of both human and BTM cells were investigated using the patch–clamp technique. To simulate the conditions used in the experiments measuring contractility, the muscarinic agonist acetylcholine or ET were added to the perfusing Ringer’s solution in concentrations of 5 × 10⁻⁵ M or 5 × 10⁻⁸ M, respectively, during the entire experiment. After formation of the seal and establishment of the perforated patch configuration, cells were allowed to stabilize for a few minutes. Then, unoprostone was added in a concentration of 10⁻⁵ M. 

In all cases studied, addition of unoprostone resulted in a highly significant increase (P < 0.001) in outward current to almost twice the original level (Fig. 8) . Conversely, addition of unoprostone in the presence of iberiotoxin (10⁻⁷ M) resulted in only an insignificant increase (Fig. 9) . No species-dependent differences could be found in the response to unoprostone (see Table 1 for details). 

Discussion

Today, treatment of glaucoma still focuses on lowering intraocular pressure (IOP) either pharmacologically or surgically. New glaucoma medications, such as the selective prostaglandin F2α receptor (FP) agonists (latanoprost) have recently become commercially available as antiglaucoma drugs, but despite good IOP control, they appear to be prone to ophthalmic side effects, such as increased iris pigmentation, excessive eyelash growth or macular edema. ^{30 31} Unoprostone isopropyl (Rescula; Novartis) is the first synthetic docosanoid introduced for the treatment of glaucoma. The substance has been in clinical use as an antiglaucoma drug in Japan for several years and appears to be well tolerated. ⁶ There is still some debate concerning how this compound lowers IOP. Although some investigators suggest a modulation of conventional outflow facility, others see the primary target site in the uveoscleral outflow pathway. ^{3 4 5 6 7 8 9 10}  

The TM has smooth muscle–like properties and is actively involved in aqueous humor dynamics through contractile mechanisms. We have described a functional antagonism between TM and CM (for review see Ref. ¹¹ ) with CM contraction leading to a distension of the TM with subsequent reduction in IOP, and with TM contraction leading to the opposite effect. Furthermore contractility appears to be differently regulated in TM and CM in respect to various signal transduction pathways involved in the regulation of smooth muscle contractility. ^{11 32 33} To investigate the mode of action of a compound that lowers IOP, the effects on the modulation of contractility within the major outflow pathway must be determined. ¹¹ This study presents evidence for the modulation of TM and CM contractility through unoprostone. This is in line with investigations on the effects of unoprostone on the conventional outflow facility in rabbits, ³⁴ primates, and humans. ^{6 7} The fact that unoprostone was not able to influence baseline contractility or pathways that involve G-protein–linked acetylcholine receptors (G_q) suggests involvement of different G-proteins (perhaps G_i) or additional G-protein–independent pathways in mediating the relaxing effects of unoprostone. This compound does not bind to known prostaglandin receptors ³⁵ where it would have induced relaxation in TM and CM, as has been demonstrated with various prostaglandin agonists. ³⁶  

As mentioned before, unoprostone is not a prostanoid but represents a new docosanoid. Concerning prostanoids we recently tested the effects of various agonists as well as antagonists on TM contraction and found virtually no effect of PGF_2α on TM contractility. However, EP₂ agonists relaxed the TM, whereas a thromboxane agonist contracted it. ³⁶  

ETs represent a group of 21-amino-acid–containing peptides and were first described by Yanagisawa et al. ³⁷ ET-1 mediates contraction in various smooth muscle systems, such as ocular vasculature, as well as the TM and CM. ^{38 39 40} Using RT-PCR-techniques the ET-A but not ET-B receptor was found to be expressed in human CM and TM cells. ²⁸ Here, an involvement in the mobilization of internal calcium ^{28 40} and an activation of phospholipase C could be demonstrated. ^{28 41} Levels of ET-1 were found to be increased in the aqueous humor of patients with primary open-angle glaucoma (pooled samples). ⁴² The role of ET in the pathogenesis of glaucoma remains unclear; however, it can be speculated that this vasoactive compound may influence aqueous humor outflow by mediating contractility in the conventional outflow pathway. It is interesting to note that unoprostone was able to inhibit ET-induced contraction but not the contraction induced by another G-protein–linked receptor agonist (carbachol), suggesting a distinct modulation of contractility by these two receptor agonists. Similar results of a unoprostone-mediated ET antagonist (ET-1) were detected in perfused pig retinal arteries ⁴³ as well as in the choroid of humans in a dose-dependent manner. ⁴⁴  

Changes in internal calcium mediate contraction and relaxation in TM and CM through L-type calcium channels ²⁹ and maxi-K channels. ^{25 26} In TM, contractility is partly dependent and partly independent of calcium and uses PKC and rho-A/ROCK-mediated pathways based on pharmacomechanical coupling events. ^{29 39 45} In our studies, unoprostone almost completely inhibited the ET-induced increase in[ Ca²⁺]_i. There was no statistically significant effect on[ Ca²⁺]_i in HTM cells through unoprostone alone, which underlines the absence of effect of the compound on baseline contractility. 

In patch–clamp experiments, we were able to show that unoprostone stimulates maxi-K channels–that is, high-conductance, calcium-activated potassium channels in TM cells of both bovine and human origin. In previous publications, ^{25 26} we have shown that the stimulation of potassium efflux through maxi-K channels leads to hyperpolarization with subsequent closure of L-type calcium channels in TM. ²⁹ This should lead to lowered values of cytosolic calcium, explaining the relaxation of TM in response to unoprostone in strips of BTM precontracted by ET. According to this hypothesis, an effect of unoprostone on contractility can only be expected in situations in which L-type calcium channels are open before the administration of the substance. In unstimulated tissue with physiological membrane voltage, maxi-K and L-type channels show only minimal activity. ^{25 26 29} Correspondingly, we were unable to observe any effect of unoprostone on unstimulated TM cells. Even if unoprostone opens maxi-K channels in this situation, hyperpolarizing the tissue, no effect can be expected on L-type channels, because these are already closed. 

In contrast to experiments of contractility in which an effect of unoprostone could only be observed when the tissue was precontracted by ET, in patch–clamp experiments, the maxi-K channel was also stimulated when the cells were pretreated with acetylcholine or ET. As mentioned, there appear to be marked differences in the signaling pathways (perhaps through different G-protein receptors) leading to the contraction of TM after stimulation of either muscarinic or ET receptors. Thus, we were able to show that, although 42% of carbachol-mediated contraction is dependent on the presence of external calcium, only 23% of ET-mediated contraction depends on the influx of calcium into the tissue. ³⁹ Carbachol-mediated contraction does not appear to involve calcium influx to the degree observable in ET-mediated responses. 

In summary, this study suggests that the ocular hypotensive effects of unoprostone appear to be the result of a direct relaxation of TM, a smooth muscle–like tissue in the outflow pathway. However, effects of this compound on uveoscleral outflow cannot be excluded from our studies. Our studies show that the CM is also affected by unoprostone, suggesting a possible influence on uveoscleral outflow. This relaxation is probably mediated by a stimulation of maxi-K channels involving changes in cytosolic calcium in tissues of both human and bovine origin. In addition, the involvement of ET-dependent pathways in Ca²⁺-sensitive mechanisms seems to play a major role. The possibility of involvement of Ca²⁺-independent pathways in the regulation of contractility cannot be excluded. 

 Submitted for publication June 22, 2001; accepted August 13, 2001.

 Commercial relationships policy: Commercial relationship policy: C (AO, CLP, GNL); N (all others).

 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.

 

Figure 1.

 

Original recording of isometric force developed in a bovine TM strip. After a carbachol (10⁻⁶ M)-induced peak, ET (10⁻⁹ M) resulted in a sustained contraction that returned to baseline after the washout period.

Figure 2.

 

Original trace of a BTM strip demonstrating that unoprostone (10⁻⁵ M) inhibited the ET (10⁻⁹ M)-induced contraction shown in Figure 1 .

Figure 3.

 

Original recording of a bovine CM strip showing the effects of carbachol and ET on contractility. Experimental conditions are identical with those shown in Figure 1 .

Figure 4.

 

Unoprostone (10⁻⁵ M) was effective in bovine CM strips in inhibiting the ET (10⁻⁹ M)-induced contraction. Experimental conditions are identical with those in TM experiments shown in Figure 2 .

Figure 5.

 

Summary of data obtained with single TM (A) and CM strips (B) after application of ET (10⁻⁹ M) and unoprostone (10⁻⁵ M). Control bar represents the force generated with 10⁻⁶ M carbachol set to 100%. The number of experiments is in brackets above the bars (*P < 0.05, **P < 0.01).

Figure 6.

 

Measurement of [Ca²⁺]_i in HTM cells. (A) Increase of[ Ca²⁺]_i after application of ET (5 × 10⁻⁸ M). This effect was almost completely inhibited by application of unoprostone 10⁻⁵ M (B).

Figure 7.

 

Summary of [Ca²⁺]_i measurements in HTM cells. The number of individual experiments are in brackets above the bars. Unoprostone (10⁻⁵ M) almost completely blocked the ET (5 × 10⁻⁸ M)-induced increase in[ Ca²⁺]_i (**P < 0.01).

Figure 8.

 

Whole-cell patch–clamp recording of an HTM cell. Unoprostone (10⁻⁵ M) enhanced current through maxi-K channels in cells pretreated with ET (5 × 10⁻⁸ M). This current was almost completely blocked by the maxi-K specific inhibitor iberiotoxin (10⁻⁷ M).

Figure 9.

 

Summary of patch–clamp recordings in HTM cells after application of unoprostone and/or iberiotoxin in the presence of ET (5 × 10⁻⁸ M; *P < 0.05,** P < 0.01, ***P < 0.001).

Table 1.

 

Summary of Patch–Clamp Data in Percentage of Initial Outward Current

Solutions and Substances	HTM Pretreated with Acetylcholine (5 × 10−5 M)			BTM Pretreated with Acetylcholine (5 × 10−5 M)
Ringer’s	100 (n = 4)	***	100 (n = 8)	, **
Unoprostone (10−5 M)	173 ± 8	, ** * >	179 ± 20	, **
Unoprostone+ Iberiotoxin (10−7 M)	67 ± 17	NS	37 ± 7	NS
Iberiotoxin (10−7 M)	60 ± 23	, *	42 ± 12	, *
Ringer’s	118 ± 32		106 ± 29

 Data are percentages.

^***  P < 0.001;

^**  P < 0.01;

*  P < 0.05.

The authors thank Olaf Strauss for invaluable discussion and advice, Lars Choritz, Marianne Boxberger, Heiko Fuchs, and Kirsten Steinhausen for help performing the patch–clamp and fura-2 experiments, and Helga Höffken for expert technical assistance.