Unoprostone Isopropyl and Metabolite M1 Activate BK Channels and Prevent ET-1–Induced [Ca²⁺]_i Increases in Human Trabecular Meshwork and Smooth Muscle

Abstract

Purpose.: Effects of cis-unoprostone isopropyl, its primary metabolite M1, trans-unoprostone isopropyl, latanoprost free acid, andfluprostenol were studied on Ca²⁺-activated K⁺ (BK) channels, plasma membrane potential, [cAMP]_i, [cGMP]_i, and steady state [Ca²⁺]_i, and protection against endothelin-1 (ET-1)–induced steady state [Ca²⁺]_i increases in human cortical neuronal (HCN-1A), trabecular meshwork (HTMC), and pulmonary artery smooth muscle (PASMC) cells. Effects on recombinant human prostaglandin (PG) receptors were determined.

Methods.: BK channel currents were measured using whole-cell patch clamp; [cAMP]_i, [cGMP]_i with ELISAs; [Ca²⁺]_i with indo-1; plasma membrane potential using diBAC₄(3); and PG receptor effects with PG receptor-expressing cells and FLIPR fluo-4 Ca²⁺ assays.

Results.: Unoprostone isopropyl and M1 activated sustained iberiotoxin (IbTX)-sensitive, AL-8810 (FP receptor antagonist)-insensitive BK channel currents with EC₅₀s of 0.51 ± 0.03 nM (n = 5) and 0.52 ± 0.03 nM (n = 6) in HTMCs; 0.61 ± 0.06 nM (n = 8) and 0.46 ± 0.04 nM (n = 5) for M1 in HCN-1A cells and PASMC, respectively. They caused AL-8810–insensitive, IbTX-sensitive membrane hyperpolarization at 10 nM; up to 100 nM had no effect on or decreased [cAMP]_i, [cGMP]_i, and [Ca²⁺]_i; and prevented ET-1–induced [Ca²⁺]_i increases. In contrast, 10 nM latanoprost free acid and fluprostenol caused membrane depolarization; increased [cAMP]_i, [cGMP]_i, and [Ca²⁺]_i; and did not prevent ET-1–induced [Ca²⁺]_i increases. Trans-unoprostone isopropyl had no effects. Unoprostone isopropyl (1.25 μM) had no effect on PG receptors, and neither did M1, except for activating the FP receptor with EC₅₀ = 557.9 ± 55.2 nM (n = 4).

Conclusions.: Prostones, unoprostone isopropyl and M1, are potent AL-8810–insensitive, stereospecific BK channel activators, without [cAMP]_i, [cGMP]_i, or [Ca²⁺]_i involvement, and prevent ET-1–induced steady state Ca²⁺ increases in HTMCs.

Introduction

The aim of this study was to determine the molecular and cellular effects of the IOP-lowering antiglaucoma agent, cis-unoprostone isopropyl (UF-021 ^1,2 ) and its primary metabolite M1, the de-esterified form lacking the isopropyl group ³ (i.e., unoprostone free acid). As was previously demonstrated, unoprostone isopropyl is a potent activator of iberiotoxin (IbTX)-sensitive Ca²⁺ and voltage-activated K⁺ channels (maxi K⁺ or BK channels), causes plasma membrane hyperpolarization, does not increase steady state [Ca²⁺]_i, and counteracts the effects of glutamate-dependent [Ca²⁺]_i deregulation in human cortical neuronal cells (HCN-1A). ⁴ From this study and others, ^4,5 unoprostone isopropyl is thought to have neuroprotective properties. Neuroprotection has been directly shown in preclinical studies on photoreceptor/ganglion cell survival. ⁶ Unoprostone isopropyl's mechanism of action was distinct from that of PGF_2α (which does not activate BK channels) ⁷ and latanoprost, a PGF_2α analog. BK channels are present in trabecular meshwork cells and their activation increases outflow facility and decreases cell volume. ^8,9  

Many studies point to a significant involvement of the potent vasoconstrictor endothelin-1 (ET-1) in the pathophysiology of glaucoma and other ocular diseases. ^10,11 ET-1 has been shown to contract the trabecular meshwork, ⁸ reduce choroidal blood flow via contraction of vascular smooth muscle cells, ¹² and to have neurotoxic effects. ^13,14 Consequently, compounds functionally antagonizing ET-1 effects are viewed as potential IOP-lowering and neuroprotective agents. ¹⁰ Indeed, a human study ¹⁵ and a recent study using a nonhuman primate glaucoma model ¹⁶ showed effective IOP lowering following ET-1 receptor blockade. It has been previously demonstrated that M1 activated BK channel currents in trabecular meshwork cells (TMCs), inhibited ET-1–induced contractions, and inhibited ET-1–induced [Ca²⁺]_i increases in TMC and ciliary muscle strips. ⁸  

In the present studies, effects of unoprostone isopropyl and M1 (both cis forms of the drug) were compared and contrasted with effects of trans-unoprostone isopropyl, which does not lower IOP (Ueno R, personal written communication, 2011), to determine if the effects of unoprostone isopropyl and M1 were stereospecific. HCN1-A cells express BK channel currents and channel protein, ^4,17,18 and were used by numerous investigators to study BK channels and to study unoprostone isopropyl effects. ⁴ Therefore, only effects of unoprostone isopropyl's primary metabolite M1 were studied in HCN-1A cells. Effects of unoprostone isopropyl and M1 on BK channels were measured in human trabecular meshwork cells (HTMCs) and pulmonary artery smooth muscle cells (PASMCs). Because unoprostone isopropyl (Rescula) is used for reducing IOP in ocular hypertension and glaucoma ^1,5,19,20 by increasing blood flow ^12,21,22 and aqueous humor (ocular) outflow facility, ²³ HTMCs were used in the present study as a particularly relevant ocular cell model. BK channels are the predominant K⁺ channels present in vascular smooth muscle, ²⁴ and when they are activated, result in plasma membrane hyperpolarization ultimately causing vasodilation. ^25,26 PASMCs were used for studying BK channel activity in vascular smooth muscle. ^27,28  

There is some controversy as to whether unoprostone isopropyl and M1 are prostaglandin (PG) receptor agonists. ^4,29,30 Therefore, direct measurement of unoprostone isopropyl and M1 agonist effects on recombinant human PG (EP₁–EP₄ and FP) receptors, was carried out to determine whether the prostones, unoprostone isopropyl and M1, are agonists of these receptors. The affinities were then compared with the dose response for BK channel activation. 

PGF_2α and PGF_2α-based antiglaucoma agents increased steady state [Ca²⁺]_i in HCN-1A cells, whereas unoprostone isopropyl did not. ⁴ In bovine TMC, ⁸ M1 did not increase steady state [Ca²⁺]_i, whereas in other studies in HTMCs, ^8,31 human ciliary muscle cells, ³² and rat A7r5³³ cells, M1 (10 nM-10 μM) transiently increased [Ca²⁺]_i over a 3-minute period. Thus, to further investigate cellular effects of unoprostone isopropyl, M1, and trans-unoprostone isopropyl, effects on steady state [Ca²⁺]_i, [cAMP]_i, and [cGMP]_i, as well as effects on plasma membrane potential were determined. The effects of AL-8810, the FP receptor antagonist, ^4,32 on M1 activation of BK channels were measured, as well as on unoprostone isopropyl and M1 effects on plasma membrane potential to investigate involvement of FP receptors in BK channel activation. 

[Ca²⁺]_i deregulation can lead to deleterious effects on cells. ⁴ Because M1 was previously shown to prevent ET-1–induced increases in [Ca²⁺]_i in TMC and ciliary muscle cells, ⁸ effects of unoprostone isopropyl, M1, and trans-unoprostone isopropyl on protection against ET-1–induced [Ca²⁺]_i deregulation ⁴ measured by [Ca²⁺]_i increases were also examined. These effects were compared with effects of the PGF_2α analogs, latanoprost free acid and fluprostenol (travoprost free acid), which are the active forms of latanoprost and travoprost. ³⁴ Unoprostone isopropyl, latanoprost, and travoprost are all isopropyl esters that are hydrolyzed to the active forms, the free acids. ³⁵  

Materials and Methods

Materials

Hank's balanced salt solution (HBSS) ± phenol red, ± Ca²⁺ was from Invitrogen (Carlsbad, CA). Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), and HCN-1A cells were from ATCC (Manassas, VA). HTMCs and Fibroblast medium were from ScienCell Research Laboratories (Carlsbad, CA). Human PASMCs and SMGM-2 BulletKit (smooth muscle basal medium plus growth factors) were from Lonza (Walkersville, MD). Valinomycin, bis-(1,3-diethylthiobarbituric acid) trimethine oxonol (DiBAC₄[3]) and the acetoxymethyl ester (AM) of indo-1 (indo-1/AM), from Molecular Probes (Eugene, OR), were dissolved in ethanol. Latanoprost free acid and fluprostenol (travoprost free acid) were from Cayman Chemical Co (Ann Arbor, MI). Unoprostone isopropyl ([isopropyl[+]-[z]-7-([1R,2R,3R,5S]-3,5-dihydroxy-2-[3-oxodecyl] cyclopentyl) −5-heptenoate]), M1 (unoprostone free acid), and trans-unoprostone isopropyl (trans isomer of unoprostone isopropyl) were from R-Tech Ueno, Ltd. (Sanda, Japan), provided as frozen aliquots of 10-mM solutions in 100% dimethyl sulfoxide (DMSO). In HCN-1A experiments, DMSO was the diluent, 0.1% final DMSO concentration. In all other experiments, 100% ethanol was the diluent to reduce DMSO final concentration to 0.00001%. Vehicle controls were always carried out. IbTX from Tocris Cookson (Ellisville, MO) was dissolved in water. AL-8810 (11 beta-fluoro-15-epiindanyl PGF_2α) and ionomycin (Sigma-Aldrich, St. Louis, MO) were dissolved in DMSO and ethanol respectively. ET-1 (Calbiochem, La Jolla, CA) was dissolved in 5% acetic acid. The Parameter cAMP kit was from R & D Systems (Minneapolis, MN) and the Direct cGMP EIA kit was from Enzo Life Sciences (Farmingdale, NY). 

Cell Culture

HCN-1A cells (passage 4) were grown in DMEM medium with 10% FBS to 85% to 90% confluence on 9 × 22-mm glass cover slips and 35-mm Petri dishes as described previously. ⁴ HTMC were grown in poly-L-lysine-coated flasks in the recommended defined Fibroblast medium containing 2% FBS and then seeded onto poly-L-lysine–coated 35-mm Petri dishes and glass coverslips for experiments. Cell passages 2 to 10 (of 15 guaranteed by the company) were used. PASMCs (passages 1 and 2) were grown in defined smooth muscle cell culture medium on 9 × 22-mm glass coverslips or in 35-mm Petri dishes. 

Patch Clamp Measurement of Whole-Cell BK Currents

Whole-cell patch clamp measurement of K⁺ currents was carried out as previously described. ⁴ Borosilicate glass pipettes were pulled by a 2-stage Narashige puller (Narishige International, USA, Inc., East Meadow [Long Island], NY) to produce 1.0- to 1.5-M Ω resistance. Data were acquired with an Axopatch CV-4 headstage (Molecular Devices, Sunnyvale, CA), a Digidata 1200 digitizer (Molecular Devices), and an Axopatch 1D amplifier (Molecular Devices). Sampling frequency was 1 KHz, filter setting was 1 KHz, and seal resistances were 10 G Ω. Data were analyzed using pClamp 6.04 (Axon Instruments, Union City, CA), Excel, and Origin 5 (OriginLab, Northampton, MA). 

Measurement of Plasma Membrane Potential

Plasma membrane potential was determined as described. ⁴ Cells grown on coverslips were incubated with 100 nM DiBAC₄(3) in HBSS for 5 minutes in the dark. Experimental procedures including calibration were as described. ⁴  

Measurement of [cAMP]_i, [cGMP]_i, and Steady State [Ca²⁺]_i

The [cAMP]_i and [cGMP]_i were measured with ELISA assays using the Parameter cAMP kit and a Direct cGMP EIA kit. For [Ca²⁺]_i measurements, cells grown on coverslips were incubated at 37°C with 2.5 μM indo-1/AM in HBSS for 30 minutes and then mounted in the ISS K2 fluorometer (ISS Inc., Champaign, IL) in indo-1/AM-free HBSS at 22°C. Fluorescence was measured with excitation at 338 nm and emission at 401 nm (390–410 nm). Ca²⁺ scans occurred over 190 seconds. Procedures for [Ca²⁺]_i determination were as described. ⁴  

Measurement of Prostaglandin Receptor Activity

Unoprostone isopropyl and M1 agonist effects on human cloned recombinant EP_1–4 and FP receptors were assayed by Millipore Corporation, Bioscience Division (St. Charles, MO), using ChemiScreen Ca²⁺-Optimized GPCR Stable Cell Lines expressing EP₁, EP₂, EP₃, EP₄, and FP receptors and a fluorometric imaging plate reader (FLIPR) fluo-4 fluorescent assay of [Ca²⁺]_i changes. This method ^36,37 has been used for a variety of receptors. ³⁸ These GPCR cell lines containing high levels of the promiscuous G protein, Gα15, to enhance coupling of the receptor to the calcium signaling pathway, were transfected with cDNAs of either full-length human EP₁, EP₂, splice variant 6 of EP₃, EP₄, and FP receptors. Using fluo-4 fluorescent assays of [Ca²⁺]_i changes, agonist effects of unoprostone isopropyl and M1 were measured using the various cell lines with the respective recombinant prostaglandin receptors. Results were expressed in relative fluorescence units and plotted as % maximum response. PGE₂ and PGF_2α were used as positive controls for EP and FP receptors respectively. 

Statistical Analysis

Statistical significance was determined using Student's t-test. Origin 5 (OriginLab, Northampton, MA) was used to fit curves. The number of independent experiments (n) and P values for statistical significance are indicated on graphs, in legends, and in the text. For BK current measurements, n is the number of cells patched. For plasma membrane potential and [Ca²⁺]_i measurements, n is the number of coverslips used. For all other experiments, n is the number of assays. 

Results

HCN-1A Cells: Effects of M1 and Trans-Unoprostone Isopropyl on BK Channel Currents and Plasma Membrane Potential

BK channel currents in HCN-1A cells were activated by 10 nM M1 and inhibited by 100 nM of the specific high-affinity BK channel inhibitor, IbTX ³⁹ (Fig. 1A). M1 activated these BK channel currents with an EC₅₀ = 0.61 ± 0.06 nM (n = eight cells) (Fig. 1B), similar to the EC₅₀ = 0.6 ± 0.2 nM (n = six cells) for unoprostone isopropyl previously reported. ⁴ Trans-unoprostone isopropyl (10 nM) had no effect (Fig. 1C). As previously shown for 100 nM unoprostone isopropyl, ⁴ 10 nM unoprostone isopropyl and 10 nM M1 also caused IbTX-inhibitable, AL-8810–insensitive plasma membrane hyperpolarization, whereas 10 nM trans-unoprostone isopropyl was not significantly different from the vehicle control (Fig. 1D). In contrast, 10 nM latanoprost free acid and 10 nM fluprostenol both caused large plasma membrane depolarization that was inhibited by 0.5 μM AL-8810 (Fig. 1D). 

Figure 1

 

HCN1A cells: (A) Effect of M1 on BK channel currents; (B) Dose response of effect of M1 on BK channel currents; (C) Effect of trans-unoprostone isopropyl, followed by M1 on BK channel current at +130 mV; (D) Effect of unoprostone isopropyl, M1, trans-unoprostone isopropyl, latanoprost free acid, and fluprostenol on plasma membrane potential. (A) Representative BK channel currents: control, 10 nM M1 followed by 100 nM IbTX and (B) dose response of effect of M1 on BK channel currents (n = 5 cells). EC₅₀ = 0.61 ± 0.06 nM (n = 5 cells). (C) Control BK channel currents at +130 mV, followed by 10 nM trans-unoprostone isopropyl (trans-uno), then 10 nM M1 followed by 100 nM IbTX (n = 5 cells). *P < 0.0005 wrt control and trans-uno; #P < 0.001 wrt IbTX; ns, not significant wrt control. In (D) plasma membrane potential change was measured with DiBAC₄(3) after addition of 10 nM unoprostone isopropyl (uno), M1, trans-unoprostone isopropyl (trans-uno), latanoprost free acid (latFA), and fluprostenol (flup). AL-8810 (0.5 μM) was added followed by 1 nM IbTX as indicated. For all compounds, n = 3 coverslips, except for uno where n = 8 coverslips and uno+AL-8810 where n = 4 coverslips. *P < 0.0005, **P < 0.005, #P < 0.02 wrt drug alone. ns, not significant wrt vehicle; NS, not significant wrt drug alone. In (B–D), data are plotted as mean ± SEM.

HTMC: Effects of Unoprostone Isopropyl, M1, and Trans-Unoprostone Isopropyl on BK Channel Currents, Plasma Membrane Potential, [cAMP]_i, [cGMP]_i, and Steady State [Ca²⁺]_i without and with ET-1

In HTMCs, BK channel currents were activated by 0.5 nM M1 and inhibited by 100 nM IbTX (Figs. 2A, 2B). BK channel currents were activated by M1 with an EC₅₀ = 0.51 ± 0.04 nM (n = six cells) (Fig. 2C), similar to that measured for HCN-1A cells (Fig. 1B). Trans-unoprostone isopropyl (10 nM) had no effect (Fig. 2D). HTMC BK channel currents were also activated by unoprostone isopropyl (Fig. 3A) with an EC₅₀ = 0.51 ± 0.03 (n = five cells), not significantly different from M1 effects or unoprostone isopropyl ⁴ and M1 (Fig. 1) effects in HCN-1A cells. Trans-unoprostone isopropyl, 10 nM, had no effect (Fig. 3B). Unoprostone isopropyl (10 nM) and 10 nM M1 both caused IbTX-inhibitable, AL-8810–insensitive plasma membrane hyperpolarization (Fig. 3C), as previously shown for HCN-1A cells, ⁴ whereas 10 nM trans-unoprostone isopropyl had no effect. In contrast, 10 nM latanoprost free acid and 10 nM fluprostenol both caused plasma membrane depolarization that was inhibited by 0.5 μM AL-8810. 

Figure 2. 

 

HTMC: (A) Effect of M1 on BK channel currents; (B) I-V curves; (C) dose response of effect of M1 on BK channel currents; (D) Effect of trans-unoprostone isopropyl, followed by M1, on BK channel current at +130 mV. (A) Representative BK channel currents: control, 10 nM M1 followed by 100 nM IbTX. (B) I-V curves, n = 4 cells and inset shows control BK channel currents at +130 mV, followed by 10 nM M1, and then 100 nM IbTX (n = 4 cells). *P < 0.005 wrt control, #P < 0.05 wrt IbTX. ns, not significant wrt control. (C) Dose response of effect of M1 on BK channel currents (n = 6 cells). EC₅₀ = 0.52 ± 0.03 nM (n = 6 cells). (D) Control BK channel currents at +130 mV, followed by 10 nM trans-unoprostone isopropyl (trans-uno) and then 10 nM M1 followed by 100 nM IbTX (n = 5 cells). *P < 0.005 wrt control and trans-uno, #P < 0.001 wrt IbTX. In (B–D), data are plotted as mean ± SEM.

Figure 3. 

 

HTMC: (A) Dose response of effect of unoprostone isopropyl on BK channel currents; (B) Effect of trans-unoprostone isopropyl, followed by unoprostone isopropyl, on BK channel current at +130 mV. (C) Effect of unoprostone isopropyl, M1, trans-unoprostone isopropyl, latanoprost free acid and fluprostenol on plasma membrane potential. (A) Dose response of effect of M1 on BK channel currents (n = 5 cells). EC₅₀ = 0.51 ± 0.03 nM (n = 5 cells). (B) Control BK channel currents at +130 mV, followed by 10 nM trans-unoprostone isopropyl (trans-uno) and then 10 nM unoprostone isopropyl (uno) followed by 100 nM IbTX (n = 5 cells). *P < 0.001 wrt control, trans-uno, and IbTX. ns, not significant wrt control. In (C), plasma membrane potential change was measured with DiBAC₄(3) after addition of 10 nM unoprostone isopropyl (uno), M1, trans-unoprostone isopropyl (trans-uno), latanoprost free acid (latFA), and fluprostenol (flup). AL-8810 (0.5 μM) was then added, followed by 1 nM IbTX as indicated. For all data plotted, n = 3 coverslips except for unoprostone isopropyl where n = 6 coverslips. *P < 0.005, **P < 0.0005, #P < 0.01 wrt drug alone. ns, not significant wrt vehicle; NS, not significant wrt drug alone. In (A–C), data are plotted as mean ± SEM.

Unoprostone isopropyl and M1 (up to 100 nM) did not increase [cAMP]_i significantly above vehicle alone, but latanoprost free acid and fluprostenol caused significant and large increases (Fig. 4A). Unoprostone isopropyl and M1 significantly decreased [cGMP]_i, whereas latanoprost free acid and fluprostenol significantly increased [cGMP]_i (Fig. 4B). Figure 4C shows steady state [Ca²⁺]_i effects in the absence and presence of ET-1. Unoprostone isopropyl (100 nM) and M1 (100 nM) significantly reduced steady state [Ca²⁺]_i (P < 0.01 and P < 0.025, respectively) and 100 nM trans-unoprostone isopropyl had no effect on steady state [Ca²⁺]_i. ET-1 (1 nM) caused a large steady state [Ca²⁺]_i increase, which was prevented by 100 nM unoprostone isopropyl and 100 nM M1, but not by 100 nM trans-unoprostone isopropyl. In contrast, 100 nM latanoprost free acid and 100 nM fluprostenol both increased steady state [Ca²⁺]_i (P < 0.005 for both), consistent with the findings of others in ciliary muscle ³² and did not prevent ET-1–induced steady state [Ca²⁺]_i increases. 

Figure 4. 

 

HTMC: Effect of unoprostone isopropyl, M1, trans-unoprostone isopropyl, latanoprost free acid, and fluprostenol on (A) [cAMP]_i, (B) [cGMP]_i, and (C) steady state [Ca²⁺]_i without and with ET-1. For (A) and (B), HTMCs were treated for 30 minutes with the indicated concentrations of drug or vehicle, then (A) [cAMP]_i and (B) [cGMP]_i were measured (n = 4 assays, except for vehicle n = 5 assays). Forskolin/IBMX (FSK/IBMX, 5 μM/20 μM) and NaNP (1 μM) were used as positive controls for [cAMP]_i and [cGMP]_i respectively. In (A), *P < 0.0005, #P < 0.0025, **P < 0.001. ns, not significant, all wrt vehicle. It must be noted that with unoprostone isopropyl and M1, [cAMP]_i does not change; whereas with latanoprost free acid and fluprostenol, [cAMP]_i increases significantly. In (B), *P < 0.0005. ns, not significant, all wrt vehicle. It must be noted that with unoprostone isopropyl and M1, [cGMP]_i decreases significantly; whereas with latanoprost free acid and fluprostenol, [cGMP]_i increases significantly. (C) Steady state [Ca²⁺]_i was measured using indo1/AM before and after treatment with 100 nM drug ± 1 nM ET-1 for 30 minutes (n = 3 coverslips). Drugs: unoprostone isopropyl (uno), M1, trans-unoprostone isopropyl (trans-uno), latanoprost free acid (latFA), and fluprostenol (flup). *P < 0.0005 wrt uno+ET-1 & M1+ET-1, #P < 0.001 wrt uno+ET-1, P < 0.005 wrt M1+ET-1. In (A–C), data are plotted as mean ± SEM.

HTMC: Effect of the FP Receptor Antagonist, AL-8810, on M1-Activated BK Channel Currents: A Timed Study

The effect of the FP receptor antagonist AL-8810 on BK channel currents activated by M1 was examined in HTMCs. Because transient [Ca²⁺]_i increases occur with M1, ^8,31–33 such that within 2 to 3 minutes [Ca²⁺]_i was below control values, the effect of 100 nM M1 at 1, 5, and 10 minutes after addition of M1, in the presence and absence of 30 μM AL-8810 on BK channel currents was measured. In Figure 5A, BK channel currents are shown before and after addition of 30 μM AL-8810, followed by 100 nM M1. Current recordings were made at 1, 5, and 10 minutes after M1 addition. Then, 100 nM IbTX was added. Figure 5B shows the summarized data ± AL-8810. AL-8810 had no effect on the BK channel current and 100 nM M1 caused a large and significant (P < 0.0005) increase in the BK channel current at 1 minute that was maintained over 10 minutes. Subsequent IbTX addition inhibited the current. Parallel experiments were performed without AL-8810, wherein vehicle alone was added. There were no significant differences between HTMC incubated with or without 30 μM AL-8810. 

Figure 5

 

HTMC: Effect of the FP receptor antagonist, AL-8810 on M1 activated BK channel currents: timed study. (A) Representative current recordings of BK channels in HTMC. After the control recording (a), 30 μM AL-8810 was added and the current recorded (b), followed by 100 nM M1. Currents were then recorded at 1 minute (c), 5 minutes (d), and 10 minutes (e) after M1 addition and then after 100-nM IbTX addition (f). (B) The summarized current data plotted as mean ± SEM with AL-8810 (n = 6 cells) and without AL-8810 (n = 5 cells). Data are plotted as current at +130 mV normalized to cell capacitance. ns, all values with AL-8810 not significant versus all values with vehicle.

PASMC: Effects of Unoprostone Isopropyl, M1, and Trans-Unoprostone Isopropyl on BK Currents, Plasma Membrane Potential, and Steady State [Ca²⁺]_i without and with ET-1; and Comparison with Latanoprost Free Acid and Fluprostenol

M1 activated IbTX-inhibitable BK channel currents in PASMCs with EC₅₀ = 0.46 ± 0.04 nM (n = six cells) (Figs. 6A, 6B), not significantly different from those measured in HTMCs and in HCN-1A cells. Trans-unoprostone isopropyl (10 nM) had no effect (Fig. 6B). In PASMCs, 10 nM unoprostone isopropyl and 10 nM M1 both caused IbTX-inhibitable, AL-8810–insensitive hyperpolarization (Fig. 6C). Unoprostone isopropyl decreased steady state [Ca²⁺]_i (P < 0.025), M1 had no effect on steady state [Ca²⁺]_i, and both unoprostone isopropyl and M1 abolished ET-1–induced steady state [Ca²⁺]_i increases (Fig. 6D). Trans-unoprostone isopropyl (10 nM) had no effect on the plasma membrane potential, [Ca²⁺]_i, or ET-1–induced steady state [Ca²⁺]_i increases. In contrast, 10 nM latanoprost free acid and 10 nM fluprostenol both caused large plasma membrane depolarization inhibitable by 0.5 μM AL-8810, caused large steady state [Ca²⁺]_i increases (P < 0.0005 and P < 0.025, respectively) and did not prevent ET-1–induced steady state [Ca²⁺]_i increases. 

Figure 6. 

 

PASMC: (A) Dose response of effect of M1 on BK channel currents; (B) Effect of trans-unoprostone isopropyl, followed by M1 on BK channel currents at +130mV. Effect of unoprostone isopropyl, M1, trans-unoprostone isopropyl, latanoprost free acid, and fluprostenol on (C) plasma membrane potential and (D) steady state [Ca²⁺]_i without and with ET-1. (A) Dose response of effect of M1 on BK channel currents (n = 5 cells). EC₅₀ = 0.46 ± 0.04 (n = 5 cells). (B) Control BK channel currents at +130 mV, followed by 10 nM trans-unoprostone isopropyl (trans-uno) and then 10 nM M1 followed by 100 nM IbTX (n = 6 cells). *P < 0.001 wrt trans-uno and IbTX, #P < 0.005 wrt control. ns, not significant wrt control. (C) Plasma membrane potential change was measured with DiBAC₄(3) after addition of 10 nM unoprostone isopropyl (uno), M1, trans-unoprostone isopropyl (trans-uno), latanoprost free acid (latFA), and fluprostenol (flup). AL-8810 (0.5 μM) was added followed by 1 nM IbTX as indicated. For all compounds, n = 3 coverslips. *P < 0.005, #P < 0.01 wrt drug alone. ns, not significant wrt vehicle; NS, not significant wrt drug alone. (D) Steady state [Ca²⁺]_i was measured using indo1/AM before and after treatment with drug ± 1 nM ET-1 for 30 minutes (n = 3 coverslips). Drugs: 100 nM unoprostone isopropyl (uno), M1, trans-unoprostone isopropyl (trans-uno), latanoprost free acid (latFA), and fluprostenol (flup). *P < 0.01 wrt uno+ET-1 and P < 0.025 wrt M1+ET-1, #P < 0.001 wrt uno+ET-1 and P < 0.0005 wrt M1+ET-1; **P < 0.0005 wrt uno+ET-1 and M1+ET-1; ##P < 0.001 wrt uno+ET-1 and P < 0.0005 wrt M1+ET-1. In (A–D), data are plotted as mean ± SEM.

Effects of Unoprostone Isopropyl and M1 on Human Cloned PG Receptors

Figure 7 shows agonist effects of unoprostone isopropyl and M1 on EP₁–EP₄ and FP receptors with appropriate positive controls. Both unoprostone isopropyl and M1 were not EP₁–EP₄ receptor agonists (EC₅₀ > 1.25 μM). M1, but not unoprostone isopropyl, was a weak FP receptor agonist with an EC₅₀ = 557.9 ± 55.2 nM (n = four assays), approximately 1000 times higher than its EC₅₀ of 0.6 nM for activation of BK channels. 

Figure 7. 

 

Agonist effects of unoprostone isopropyl and M1 on human cloned PG receptors: (A) EP₁, (B) EP₂, (C) EP₃, (D) EP₄, and (E) FP. Dose response of effect of unoprostone isopropyl and M1 on change in [Ca²⁺]_i in EP₁, EP₂, EP₃, EP₄, and FP receptor–expressing cells, calculated as Δ relative fluorescence units (% maximum activation). PGE₂ was the positive control agonist for EP₁–EP₄ receptors and PGF_2α was the positive control agonist for the FP receptor. Data are plotted as mean ± SEM, n = 5 assays for unoprostone isopropyl (uno), n = 4 assays for M1, and n = 6 assays for PGE₂ and PGF_2α. *P < 0.001 wrt uno. EC₅₀ for M1 on FP receptor = 557.9 ± 55.2 nM (n = 4 assays).

Discussion

This is the first comprehensive study of the concentration dependence of unoprostone isopropyl and M1's activation of BK channels and their effects on plasma membrane potential, [cAMP]_i, [cGMP]_i, and steady state [Ca²⁺]_i in HTMCs. Similar EC₅₀s and plasma membrane hyperpolarization effects were seen with HCN-1A cells and PASMCs. In HTMCs, concentrations as high as 100 nM M1 (200 times EC₅₀ for BK channel activation) did not increase [cAMP]_i, [cGMP]_i, or steady state [Ca²⁺]_i. This mechanism of action of BK channel activation without the need for second messengers is different from that of the PGF_2α analogs, latanoprost free acid, and fluprostenol. Indeed, PGF_2α itself does not activate IbTX-sensitive BK channels, ⁷ and prostaglandins are known to depolarize neuronal cells ^4,32 and variously increase [cAMP]_i [cGMP]_i, ^29,30,40 and [Ca²⁺]_i. ^32,33  

Steady state [Ca²⁺]_i was reduced by M1 in HTMCs (Fig. 4C). This is consistent with the findings of others that M1 did not increase steady state [Ca²⁺]_i in HTMCs ⁸ and that unoprostone isopropyl did not increase steady state [Ca²⁺]_i in HCN-1A cells. ⁴ However, several studies showed Ca²⁺ mobilization (reported as nonratiometric, relative Ca²⁺ signals) by M1 in HTMCs, ³¹ human ciliary muscle cells, ³² and rat A7r5 cells. ³³ The time course of Ca²⁺ mobilization by M1 was fundamentally different from that caused by other agents. Thus, in A7r5 cells, ³³ travoprost free acid, PGF_2α, and bimatoprost acid caused a peak increase in Ca²⁺ signal at 15 to 30 seconds and a sustained increase in Ca²⁺, which persisted over the 3-minute time course of the experiments, whereas M1 gave only transient increases in Ca²⁺, which decayed to resting levels or below by 2 to 3 minutes (with no increase in steady state Ca²⁺). These responses were sensitive to AL-8810. At 10 nM M1, where the BK channel is fully activated, there was only a very small increase in the Ca²⁺ signal compared with the highest M1 concentration (3 μM). The largest Ca²⁺ signal at 3 μM M1 also decreased to baseline or below within minutes. ³³ Similarly, transient AL-8810–sensitive Ca²⁺ increases with M1 occurred in HTMCs, ³¹ although Ca²⁺ remained elevated at 10 μM M1 in the single tracing shown. The M1 response in A7r5 cells is similar to that in bradykinin-treated HTMCs in the absence of extracellular Ca²⁺, where Ca²⁺ release from intracellular stores gave a transient [Ca²⁺]_i increase from 100 nM to 400 nM, but then fell within 2 to 3 minutes to baseline. ⁴¹ This suggests that M1 might cause Ca²⁺ release from intracellular Ca²⁺ stores, but does not subsequently allow Ca²⁺ entry from the extracellular medium through plasma membrane Ca²⁺ channels. Shimura et al. ⁴² reported inhibition of Ca²⁺ release-activated Ca²⁺ currents at high concentrations of M1 in monkey TMC. Thieme et al. ⁴³ reported inhibition of L-type Ca²⁺ channels in HTMCs, over the concentration range of M1 from 1 μM to 100 μM and attributed this in part to plasma membrane hyperpolarization by BK channels. M1 also induced plasma membrane hyperpolarization at the low concentrations used in the present studies (Figs. 1D, 3C, and 6C). Whereas travoprost acid (fluprostenol), latanoprost free acid, and PGF_2α promoted mitogen-activated protein (MAP) kinase phosphorylation, a Ca²⁺ dependent process, ⁴⁴ in a concentration-dependent manner, M1 effects were biphasic and minimal compared with, for example, PGF_2α. ³² This difference could result from the lack of sustained [Ca²⁺]_i increases with M1 and is unlike the cat iris sphincter smooth muscle cells' response to thapsigargin, an initial transient [Ca²⁺]_i increase followed by a sustained [Ca²⁺]_i ⁴⁵ resulting in MAP kinase activation to similar levels as elicited by PGF_2α. ⁴⁴  

Figure 5 shows that BK channel activation by M1 persists long after transient [Ca²⁺]_i increases have decayed and is similar in the presence and absence of AL-8810, thus apparently being independent of FP receptor occupation. BK channel activation by other long-chain cis fatty acids occurs without a change in [Ca²⁺]_i ⁴⁶ and arachidonic acid activates BK channels without a change in Ca²⁺ sensitivity. ⁴⁷ Ca²⁺-independent BK channel activation also occurs with a variety of fatty acids and negatively charged lipids. ⁴⁸ Cymo4, a dehydroabietic acid, activates BK channels in “the virtual absence of Ca^2+”49 and other agents also activate BK channels without [Ca²⁺]_i increases. ⁵⁰ In addition, an endogenous ancillary protein in nonexcitable cells activates BK channels at resting membrane potential and without Ca²⁺. ⁵¹ Genistein, a tyrosine kinase inhibitor, stimulated BK channel currents in bovine TMCs, and this was unaffected by depleting the cells of Ca²⁺. ⁵² Genistein also inhibited L-type Ca²⁺ channels in both bovine and human TMCs. ⁵³  

Using cloned human EP_1–4 and FP receptors, the only activity found was weak agonist activity of M1 on the FP receptor, an EC₅₀ = 557.2 ± 55.2 nM. These results were similar to those found by others using different methods ^29,30 and are approximately 1000 times higher than the EC₅₀ of 0.52 ± 0.03 nM for BK channel activation. The maximal M1 concentration achieved in therapy is approximately 425 ± 99 nM³. The EC₅₀ for transient M1-induced Ca²⁺ mobilization was 2400 nM in HTMCs, ³¹ 306 nM in rat A7r5 cells, ³³ 407 nM in mouse Swiss 3T3 cells, ³³ and 1270 nM in human embryonic kidney cells expressing cloned human FP receptors. ³³ The EC₅₀ for rat uterus contraction by M1 was 310 nM, ⁵⁴ and the EC₅₀ values for M1 binding to bovine corpus luteal homogenate FP receptors, M1-induced phosphoinositol (PI) turnover in a variety of cells, and M1-induced cat and bovine iris contraction were approximately 300 nM to several micromolars in a variety of studies. ^31-33,54 The EC₅₀s were 3280 nM for M1-induced Al-8810–sensitive PI turnover in HTMCs ³¹ ; 3860 nM for M1 binding to the human cloned FP receptor, ³³ and 4192 nM for M1 binding to the cloned human ciliary body FP receptor. ³⁵ Thus the EC₅₀ for BK channel activation by M1 is very low compared with the various EC₅₀ values for M1 reported in the above studies. However, the EC₅₀ for FP receptor activation by M1 in the present study is within the ranges reported by others for a variety of effects. 

In FP receptor knockout mice, the IOP lowering effect of unoprostone isopropyl is lost, suggesting that FP receptors are likely involved in IOP lowering. ⁵⁵ Whatever the role of FP receptor occupation by M1 in IOP lowering, sustained [Ca²⁺]_i increases are not needed for BK channel activation (Fig. 5). Understanding the mechanism of action of unoprostone isopropyl and M1 in IOP reduction will require an understanding of how BK channels are involved in lowering IOP, and how activation of BK channels by M1 without either FP receptor occupation or sustained elevated [Ca²⁺]_i might play a role in IOP reduction. 

BK channels are ubiquitously expressed, ⁵¹ are found in cells and tissues involved in volume regulation, including HTMCs ^8,9 and Schlemm's canal cells, ⁵⁶ and are involved in outflow facility and decreasing trabecular meshwork cell volume. ^8,9,57 In the present study, BK channels in HTMCs were activated by unoprostone isopropyl and M1 with similar EC₅₀s of 0.5 nM. Treatment of trabecular meshwork with unoprostone isopropyl has been previously shown to inhibit ET-1–induced contractility, likely contributing to IOP lowering by increasing aqueous outflow. ⁸ It is of interest that glaucomatous TMCs exhibit mitochondrial defects wherein the cells are abnormally sensitive to Ca²⁺ stress and thus Ca²⁺ regulation dysfunction may contribute to the inability to control IOP. ⁵⁸ BK channel activation has also been shown to cause dilation of ET-1–treated retinal arterioles, resulting in increased blood flow. ⁵⁹ In the present study, PASMCs also exhibited unoprostone isopropyl and M1-stimulated BK channels. BK channels also modulate pre- and postsynaptic signaling at reciprocal synapses in retina ⁶⁰ and are involved in receptor potential regulation of rod receptors. ⁶¹ BK channel activation could be responsible not only for increased blood flow in the eye ²² and ocular outflow facility, ²³ as it results in membrane hyperpolarization and Ca²⁺ entry is prevented, which could lead to smooth muscle relaxation, but also for neuroprotective effects seen in patients with glaucoma ^19,20 and retinitis pigmentosa (Yamamoto S, et al. IOVS 2011;52:ARVO E-Abstract 4992; Sugawara T, et al. IOVS 2011;52:ARVO E-Abstract 4994). M1 has recently been shown to reduce oxidative stress– and light-induced damage in the human retinal pigment epithelial cell line, and this effect is mediated by BK channels, as it is inhibited by IbTX. ⁶² These findings support the view that BK channel activators are neuroprotective agents. ^63,64 The present findings show that at the concentrations used (up to 100 nM), effects of the prostones, unoprostone isopropyl, and M1, were similar in neuronal, ocular trabecular meshwork and smooth muscle cells and provide insight into the molecular and cellular effects and mechanism of action of unoprostone isopropyl and its primary metabolite M1. These findings also provide additional insight that may help explain/understand published clinical findings. ^12,19-22 These findings confirm and extend our previous findings that unoprostone isopropyl activates BK channels in HCN-1A cells. Together, these findings may provide a rationale for understanding unoprostone isopropyl's efficacy in the treatment of glaucoma. 

Footnotes

 Supported by a grant from Sucampo AG, Zug, Switzerland (JC).

Footnotes

 Disclosure: J. Cuppoletti, Sucampo (F, I, C, R); D.H. Malinowska, Sucampo (C); K.P. Tewari, None; J. Chakrabarti, None; R. Ueno, Sucampo (I, S), P