A Novel Selective Soluble Guanylate Cyclase Activator, MGV354, Lowers Intraocular Pressure in Preclinical Models, Following Topical Ocular Dosing

Abstract

Purpose: The nitric oxide/soluble guanylate cyclase/protein kinase G (NO/sGC/PKG) is known to be involved in the regulation of intraocular pressure (IOP) and may be dysregulated in glaucoma. The purpose is to demonstrate that the sGC activator MGV354 lowers IOP in a monkey model of glaucoma and could be considered as a possible new clinical drug candidate.

Methods: Changes to cGMP were assessed in primary human trabecular meshwork (hNTM) cells and binding studies were conducted using human sGC full-length protein. Ocular safety tolerability, exposure, and efficacy studies were conducted in rabbit and monkey models following topical ocular dosing of MGV354.

Results: sGC was highly expressed in the human and cynomolgus monkey outflow pathways. MGV354 had a 7-fold greater Bmax to oxidized sGC compared to that of reduced sGC and generated an 8- to 10-fold greater cGMP compared to that of a reduced condition in hTM cells. A single topical ocular dose with MGV354 caused a significant dose-dependent reduction of 20% to 40% (versus vehicle), lasting up to 6 hours in pigmented rabbits and 24 hours postdose in a cynomolgus monkey model of glaucoma. The MGV354-induced IOP lowering was sustained up to 7 days following once-daily dosing in a monkey model of glaucoma and was greater in magnitude compared to Travatan (travoprost)-induced IOP reduction. Mild to moderate ocular hyperemia was the main adverse effect noted.

Conclusions: MGV354 represents a novel class of sGC activators that can lower IOP in preclinical models of glaucoma. The potential for sGC activators to be used as effective IOP-lowering drugs in glaucoma patients could be further determined in clinical studies.

The estimated number of glaucoma cases globally is over 60 million, and primary open angle glaucoma accounts for >70% of the total cases. In the United States alone, over 120,000 patients are blind from this disease, accounting for 9% to 12% of all cases of blindness.¹ Lowering of intraocular pressure (IOP) is the current gold standard for reducing visual field (VF) loss in glaucoma. However, up to 45% of patients who achieved an IOP-lowering target of 25% to 30% (6 to 8 mm Hg) from baseline IOP by medical treatment continued to show progressive VF loss. Every 1 mm Hg higher mean IOP in the follow-up period was associated with 13% higher risk of progressive VF loss for the study population, whereas each mm Hg reduction was associated with a 10% reduction in disease progression.^2,3 Inadequate control of IOP can occur due to a dysfunctional outflow pathway and/or improper compliance with topical ocular dosing of antiglaucoma medications. The former, therefore, necessitates the identification of new target mechanisms to complement the existing armamentarium of topical treatments for glaucoma. 

In the United States, recent approval of IOP-lowering glaucoma therapies that can directly increase conventional outflow facility (or trabecular outflow) have spurred an interest in considering additional novel mechanisms targeting this outflow pathway.⁴ These novel therapies include Rhopressa (netarsudil, a rho kinase inhibitor) and Vyzulta (latanoprostene bunod, a dual-action drug that produces nitric oxide [NO] and latanoprost). Recent reports also indicate the completion of phase III clinical trials for Roclatan, a fixed combination of Rhopressa and latanoprost, which lowered IOP by 1 to 3 mm Hg greater than the monotherapy comparators over 3 months of dosing in patients with glaucoma.⁵ IOP lowering via the NO/cGMP signaling pathway is also well known. This is a key pathway in several physiologic processes, including platelet inhibition, smooth muscle relaxation, and neurotransmission. NO is the endogenous ligand of soluble guanylate cyclase (sGC). Binding of NO to the heme domain of sGC results in sGC activation and production of cGMP, a second messenger that initiates downstream signaling by activation of cGMP-dependent protein kinases.^6,7 sGC is a heterodimer composed of either an α1 or α2 subunit combined with the β1 subunit, which has a heme prosthetic group. NO binds to the prosthetic heme of sGC, under physiologic (reduced) conditions, which activates the enzyme to catalyze the conversion of guanosine-5′-triphosphate (GTP) to cGMP.^6,7 There are several reports that NO/sGC/cGMP modulators have a therapeutic effect on IOP reduction in preclinical glaucoma models and in glaucoma patients.^8–15 In addition, NO donors, cGMP analogs, and compounds that increase cGMP in ocular tissues have been well documented to lower IOP via increasing conventional outflow facility.^8,16,17 Vyzulta lowered IOP by an additional 1+ mm Hg better than Xalatan alone in glaucoma patients, and the effect was maintained up to 24 hours on day 28 in patients with glaucoma or ocular hypertension.¹⁸ The preclinical and clinical effects of these and other potential new IOP-lowering drug classes have been recently discussed in a review paper⁴ and by others.^19,20 

The NO/sGC/cGMP pathway also appears to be dysregulated in glaucoma as well as other nonocular diseases/dysfunctions, including pulmonary hypertension, atherosclerosis, thrombosis, inflammation, erectile dysfunction, renal fibrosis/failure, liver cirrhosis, and arterial hypertension.^6,7,13,21 Under the aforementioned pathologic conditions, prolonged oxidative stress can cause the oxidation of the heme group of sGC (from ferrous to ferric state), which is incapable of being activated by NO and can contribute to exacerbation of disease processes. Two classes of compounds that increase sGC activity have been described. sGC stimulators are dependent on heme and their effect is synergistic with NO, but they are not active once sGC becomes oxidized and becomes a heme-free enzyme. sGC activators can activate the enzyme to generate cGMP even in the absence of NO and under oxidative conditions.^6,7 Therefore, pharmacologic activation of sGC could normalize cGMP production and makes the treatment and/or prevention of such disorders possible. It is suggested that the glaucomatous outflow pathway experiences extensive oxidative stress, which can result in its functional impairment leading up to elevated IOP.^22–27 Selective activation of the oxidized form of sGC should target mainly the diseased state of sGC in these putative target tissues, thus offering a highly innovative therapy for glaucoma that should work adjunctively with current therapies. 

In this manuscript, we describe some of the key preclinical attributes of MGV354, a novel sGC activator, that generates cGMP production in normal human trabecular meshwork (NTM) cells under oxidizing conditions and causes sustained IOP lowering when dosed once-daily topically in laser trabeculoplasty-induced ocular hypertensive cynomolgus monkeys. 

Materials and Methods

All animal-related procedures were conducted according to protocols approved by Novartis Institutional Animal Care and Use Committee in compliance with Animal Welfare Act regulations and the Guide for the Care and Use of Laboratory Animals and were in adherence to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. MGV354, an (S)-1-(6-(3-((4-(1-(cyclopropanecarbonyl)piperidin-4-yl)-2-methylphenyl)amino)-2,3-dihydro-1H-inden-4-yl)pyridin-2-yl)-5-methyl-1H-pyrazole-4-carboxylic acid, was prepared as previously described.²⁸ 

Protein Expression/Purification of sGC enzyme

Constructs of full-length sGC were designed with C-terminally Strep-tagged α1 subunit amino acids 1 to 690 and C-terminally FLAG-tagged β1 subunit amino acids 1 to 619. pFastbac1 was used as the vector for virus generation. A full-length sGC protein heterodimer was expressed in insect cells by using the baculovirus system. Sf9 insect cells were cultivated to a concentration of 1 × 10⁶ cells/mL in Sf-900 II SFM media (Gibco). Cells were then infected with α1 and β1 P3 baculovirus and supplemented with 200 μM of the heme precursor 5-aminolevulinic acid (ALA). Cells were harvested at 60% to 70% viability by centrifugation. Pellets were snap frozen in liquid nitrogen and stored at −80°C until purification. 

The thawed cell pellets were resuspended in 10× vol/g pellet of lysis buffer (25 mM Tris with pH 7.5, 100 mM NaCl, 10% glycerol, 1 mM Tris[2-carboxyethyl]phosphine [TCEP], 1 mM EDTA, 0.1 mM ferric citrate, and complete protease inhibitor tablets) and stirred for 15 minutes at room temperature (RT). Cells were mechanically lysed by a polytron blender (Fisher Scientific) at power setting 3 for 3 minutes. The lysate was centrifuged for 20 minutes in a JLA-16.250 rotor at 38,400g (Beckman Coulter, Brea, CA, USA). The resulting supernatant was collected and mixed with Strep-tactin Superflow Plus resin (Qiagen, Hilden, Germany) followed by inversion for 2 hours at 4°C. The supernatant/resin mix was added to a gravity column and washed with 40 column volumes of wash buffer (25 mM Tris with pH 7.5, 100 mM NaCl, 10% glycerol, and 1 mM TCEP) followed by elution with 4 × 5 column volumes of elution buffer (25 mM Tris with pH 7.5, 100 mM NaCl, 10% glycerol, 1 mM TCEP, and 2.5 mM desthiobiotin). Elutions were run on SDS-PAGE 4% to 12%, and fractions containing sGC were pooled and concentrated through a Vivaspin 20, 30-kDa molecular weight cutoff (MWCO) spin concentrator until the volume was <1 mL. Concentrated sGC was injected onto a Superdex 200 10/300 GL column equilibrated in sizing buffer (25 mM Tris with pH 7.5, 100 mM NaCl, and 1 mM TCEP). Peak fractions were collected, pooled, and concentrated through the spin concentrator to reach approximately 2 mg/mL. Final purified protein was analyzed and verified by liquid chromatography-mass spectrometry (LC-MS), Nanodrop UV-Vis (Thermo Scientific), LC-UV, and SDS-PAGE. 

For the affinity selection-mass spectrometry (AS-MS) compound binding assay, purified sGC protein was desalted into assay buffer (25 mM Tris with pH 7.5 and 150 mM NaCl) using a PD-10 column to remove the TCEP reducing agent. Protein was diluted to a 0.5-μM concentration and treated in one of the following two ways: (1) readdition of reducing agent TCEP at 0.5 mM or (2) addition of oxidizing agent 1H-[1,2,4]-oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) at 200 μM. Protein was aliquoted for each condition in a 96-well plate and mixed with compound MGV354 in an 8-point 1:2 dilution series from 20 μM to 0.156 μM. The protein-compound mix was incubated at RT for 1 hour prior to transfer to the AS-MS plate by using a CyBi-Well liquid handler. AS-MS was run using an AutoScreen-96A Sephadex G-50 plate (GE Healthcare), which was preequilibrated in assay buffer. The protein-compound mix was spun on a plate for 2 minutes at 910g, and the flow-through was collected for MS detection. Unbound compound was retained on the plate, and bound compound was detected in the flow-through. Heme bound to protein was also detected in the same denaturing MS run. 

Samples were run on an Acquity ultra performance liquid chromatography (Waters, Milford, MA, USA) and MS acquisition was on a G2 Xevo quadrupole time of flight MS in positive mode (mobile phase A, water 0.05% formic acid; mobile phase B, acetonitrile 0.04% formic acid; Waters). MassLynx (Waters) browser software was used to process the raw data and in-house RPT (Novartis) software used to detect bound compound and heme. Dose response curves and binding constants for bound compound/heme were plotted and calculated using Prism software (GraphPad Software, San Diego, CA, USA). 

Cell Culture of NTM Cells and cGMP Assay

Primary hNTM cells (derived in Alcon) were originally obtained from donor eyes as previously described²⁹ and cultured in Dulbecco's modified Eagle's medium (DMEM) without sodium pyruvate (Gibco; Catalog no. 11965-092) that contained 10% fetal bovine serum (Gibco; Catalog no. 10082-147), 1% PenStrep (Gibco; Catalog no. 15140-122), and 1% GlutaMAX Supplement (Gibco, Catalog no. 35050-061). Cells used were between passage 6 and 10. 

Dose Response for MGV354-Induced cGMP Production

NTM cells were plated in 96-well plates at a density of 20,000 cells/well in 100 μl and incubated overnight at 37°C. Cells were then washed with Dulbecco's PBS (DPBS) (Gibco; Catalog no. 14190-136) and incubated for 15 minutes at RT with 20 μl of assay buffer (serum-free DMEM [Gibco; Catalog no. 11965-092] that contained 1 mM isobutylmethylxanthine [IBMX] [Sigma-Aldrich Corp., St. Louis, MO, USA; Catalog no. I5839-1G], 0.1% BSA [Sigma-Aldrich Corp.; Catalog no. A9576-50ML], and 20 μM ODQ [Sigma-Aldrich Corp.; Catalog no. O3636]). MGV354 was serially diluted in dimethyl sulfoxide (DMSO) starting at a concentration of 10 μM, and then diluted 1:500 in assay buffer. Twenty microliters of compound diluted in assay buffer was then added to the cells. The final volume was 40 μl/well, and the final concentration of DMSO was 0.1%. Each sample was run in duplicate. Cells were incubated with MGV354 for 60 minutes at 37°C, and then processed according to instructions of the CatchPoint cGMP fluorescent assay kit (Molecular Devices, Menlo Park, CA, USA; Catalog no. R8074). Briefly, 40 μl of lysis buffer was added to the cells, and plates were incubated on a shaker for 15 minutes at RT. Forty microliters of lysates were transferred to a coated plate provided in the kit, followed by the addition of 40 μl of rabbit anti-cGMP antibody and 40 μl of horseradish peroxidase (HRP)-conjugated cGMP. After 2 hours of incubation at RT, plates were washed and incubated with stoplight substrate for 1 hour at RT. Plates were read on a Synergy H1 reader (Biotek, Winooski, VT, USA) at 530 excitation/590 emission. cGMP concentrations were generated as described below. 

Time Course Evaluation of cGMP in Response to MGV354 Treatment of NTM Cells

NTM cells were plated in 96-well plates at a density of 20,000 cells/well in 100 μl and incubated overnight at 37°C. At time 0, cells were washed with DPBS (Gibco; Catalog no. 14190-136) and incubated with 40 μl of assay buffer (serum-free DMEM [Gibco; Catalog no. 11965-092) that contained 1 mM IBMX [Sigma-Aldrich Corp., Catalog no. I5839-1G], 0.1% BSA [Sigma-Aldrich Corp., Catalog no. A9576-50ML], and with or without 20 μM ODQ [Sigma-Aldrich Corp., Catalog no. O3636]). At the indicated timepoints, the assay buffer was removed and replaced with 40 μl of fresh assay buffer containing 1 μM MGV354. The total incubation time for all samples was 180 minutes. All reactions (ran in duplicate) were ended at the same time by adding 40 μl of lysis buffer, and samples were processed as described in the previous paragraph. cGMP concentrations were generated as described below. 

Data Analyses to Determine cGMP Concentrations in NTM Cells

Background signal, determined from the average of 2 wells incubated without rabbit anti-cGMP antibody, was subtracted from all wells. A standard curve for cGMP was generated by plotting raw data (minus background) from known cGMP concentrations (a cGMP standard curve was run on each plate) by using a 4-parameter sigmoidal dose-response; unknown cGMP concentrations from experimental samples (each run in duplicate) were interpolated from the standard curve. MGV354 EC₅₀ values were generated by plotting cGMP concentrations by using a 4-parameter sigmoidal dose-response. All statistical analyses were performed using GraphPad Prism. 

Broad Ligand Profiling of MGV354 and cAMP Assay

MGV354 was also submitted to an assay panel comprising 85 targets to assess selectivity performed per procedures developed by an external contract research organization (Eurofins-Panlabs, Pharmacology Laboratories, Taipei, Taiwan). Compounds submitted to this panel were tested at a single concentration in duplicate, typically 10 μM. The percent inhibition of native ligand binding in the presence of MGV354 was determined. Binding with MGV354 at ≥50% was considered noteworthy. Chinese hamster ovary (CHO) cells or human embryonic kidney (HEK293) cells overexpressing the various ligands were used for assessing the binding properties of MGV354. 

Functional assays for cAMP mobilization or suppression were performed in-house, as described below. The cAMP assay was performed in a white, 384-well flat bottom assay plate (Nunc Catalog no. 262360; Thermo Scientific) by using the homogenous time resolved fluorescence technology (HTRF), following standard procedure as previously described.³⁰ HTRF cAMP immunoassay kits were purchased from Cisbio International (Bagnols-sur-Cèze, France; Catalog no. 62AM4PEC). Aliquots of cryopreserved cells overexpressing the target of interest were used. 

Agonist Mode

On the day of the experiment, the cells were thawed rapidly, resuspended in Hanks' balanced salt solution (HBSS)/20 mM HEPES with 1 mM IBMX buffer at a concentration of 2.5 × 10⁵/mL, and plated at 20 μL/well into the assay plate containing 100 nL of test compounds serially diluted in 100% DMSO. Low controls (unstimulated cells) and high controls (reference agonist at EC₁₀₀) were also included on each plate. The final DMSO concentration in the assay plate was 0.5%. 

Antagonist Mode

On the day of the experiment, frozen cells were resuspended in HBSS/20 mM HEPES buffer at a concentration of 5 × 10⁵/mL and plated at 10 μL/well into the assay plate containing 100 nL of test compounds serially diluted in 100% DMSO. Plates were incubated at RT for 5 minutes, and 10 μL of the reference compound at EC₈₀ was diluted in assay buffer, with 2 mM IBMX added to the appropriate wells. Low controls were determined with a known antagonist at high concentration (IC₁₀₀), and high controls (reference agonist EC₈₀) were included on each plate. The final DMSO concentration in the assay plate was 0.5%. HTRF reagents were prepared by diluting stock solutions of anti-cAMP cryptate and cAMP D2 1:20 in lysis buffer supplied with the kit. After 30 minutes of incubation at RT, 10 μL of cAMP-D2 and 10 μL of anti-cAMP cryptate were added to the assay plates. After 1 hour of incubation time at RT, the plates were read on an Envision reader (PerkinElmer, Turku, Finland; excitation wavelength, 330 nm; emission wavelengths, 620 and 665 nm). Results were calculated from the 665 nm/620 nm ratio, following standard protocols.³⁰ 

Ocular Safety Following Topical Dosing of MGV354 in Rabbits and Cynomolgus Monkeys

A proprietary formulation of MGV354 (0.01%, 0.1%, and 1.0%) was dosed topically (35-μl ocular drops) in both eyes of Dutch-belted rabbits and cynomolgus monkeys for up to 4 weeks, with a 4-week recovery period. These concentrations resulted in doses of 7, 70, and 700 μg, respectively, for each animal; 3 males/3 females per dose were included for the 0.01% and 0.1% groups and 5 males/5 females in the control and 1% dose groups. Ocular assessments included visual and slit lamp examinations. In addition, MGV354 (0.1% and 1.0%) was dosed topically for 14 days in cynomolgus monkeys, and changes to anterior chamber depth (ACD), axial length (AXL), central corneal thickness (CCT), anterior chamber inflammation (cells/flare), and corneal cloudiness/damage were assessed. Following predosing baseline assessments, additional data were collected on day 1 at 1 hour and 6 hours postdose, on day 7 prior to dosing and 1 hour postdose, and on day 14 prior to dosing and 1 hour postdose. The topical formulation consisted of hydroxypropyl methylcellulose (0.5%), dibasic sodium phosphate anhydrous (0.2%), sodium chloride (0.65%), polysorbate 80 (0.05%), sodium hydroxide and/or hydrochloric acid (to adjust pH 7.4), and purified water (q.s. 100%). 

Ocular and Systemic Exposure Assessments in Rabbits and Systemic Exposure in Cynomolgus Monkeys, Following Topical Ocular Dosing of MGV354

Following a topical ocular instillation of 1% MGV354 (300 μg), rabbits were sacrificed at 1, 3, 6, and 24 hours postdose to harvest both the left and right eyes as well as blood samples. The blood was drawn from the central ear artery and was centrifuged at 10,000 rpm for 2 minutes. Approximately 40 μL of plasma was transferred to tubes or a PCR-96-AB-C plate, capped with tube caps or a PCR strip cap and stored at −20°C for compound analysis. The eye samples were kept separate according to animal number and marked for left or right eye. Upon enucleating, eyes were frozen in liquid nitrogen for 30 seconds and put on ice until stored at −80°C. Frozen eye tissues were subsequently dissected to remove the cornea, aqueous humor, iris-ciliary body, vitreous humor, choroid, and retina. Aqueous humor and cornea were transferred directly to microcentrifuge tubes. Vitreous humor was weighed and transferred to microcentrifuge tubes containing 3-mm tungsten beads. The retina, choroid, and iris-ciliary body were weighed and transferred to centrifuge tubes containing 6 volumes of acetonitrile in PBS and 3-mm tungsten beads. 

In cynomolgus monkeys, approximately 1 mL of whole blood was collected from the femoral vein at specified timepoints following a single bilateral dose of 0.01%, 0.1%, or 1% topical ocular instillation. For a repeated dose study, approximately 1 mL of whole blood was collected from the femoral vein at 1 hour following the last dose (14th) of 0.1% or 1% topical ocular instillation. An LC-MS/MS method was used to quantitate MGV354 drug levels in plasma and ocular tissues. Aliquots (20 μL) of experimental samples (5 times diluted with plasma), the prepared standards, and quality control (QC) samples were deproteinated by adding acetonitrile (120 μL) containing 50 ng/mL of the internal standard (glyburide). They were then vortexed for 5 minutes and centrifuged for 5 minutes at 4°C at 4000 rpm. A total of 100 μL of the supernatant was transferred to a new plate to which 100 μL of water was added. Then, 5 μL of the aliquot was subjected to HPLC/MS analysis. 

Rabbit vitreous humor, choroid, retina, and iris-ciliary body were homogenized by using a Qiagen tissue homogenizer with a volume of 10% acetonitrile in water 6 times the weight of the tissue. A total of 20 μL of spiking standard solutions was added to 20 μL of blank matrix to prepare matrix standard solutions and QCs. Aliquots (20 μL) of unknowns, with 20 μL of 50:50 acetonitrile:water added, the prepared standards, and the QC samples were deproteinated by adding acetonitrile (150 μL) containing 100 ng/mL of the internal standard (glyburide). They were vortexed and centrifuged for 10 minutes at 4°C at 4000 rpm. A total of 125 μL of the supernatant was transferred to a new plate to which 100 μL of water was added to each well. Then, 10 μL of the aliquot was subjected to HPLC/MS analysis. An Agilent 1200 binary HPLC pump (Agilent Technologies, Waldbronn, Germany) with a Leap CTC HTS Pal autosampler (Sparta, NJ, USA) was used for all liquid chromatography separations. The chromatographic separation of analytes was achieved on an ACE C18 column (3 μm, 2.1 × 30 mm) from MAC-MOD Analytical, Inc. (Chadds Ford, PA, USA), in conjunction with fast gradient conditions and mobile phases A (water containing 0.1% formic acid) and B (acetonitrile containing 0.1% formic acid). A Sciex API-4000 triple quadrupole (MS/MS) mass spectrometer equipped with an electrospray ionization was used for detection. The instrument was operated in the positive ion multiple reaction monitoring (MRM) mode, employing argon, a collision gas. The following MRM transitions were monitored: m/z 576.5 to 318.2 and m/z 494.2 to 169.2 for MGV354 and the internal standard (glyburide), respectively. 

Data were acquired and processed by Sciex Analyst 1.4.2 software. Standard regression and back-calculation of unknown concentrations were performed with Thermo Watson 7.4.2 software purchased from Thermo Fisher Scientific, Inc. (Philadelphia, PA, USA). Quantification of the parent compound in rat plasma was based on an at least 5-point calibration curve. The dynamic range was set from 2.0 nM lower limit of quantitation (LLOQ) to 20 μM. The bias of all calibration standards and quality control samples was within the acceptance criteria of ±30%. All pharmacokinetic (PK) parameters were derived from concentration-time data by noncompartmental analyses. All PK parameters were calculated with WinNonlin Phoenix version 6.4 (Certara, St. Louis, MO, USA). Although the ocular tissues and plasma samples were collected by sacrificing animals at each timepoint (i.e., composite sampling), each eye tissue sample was treated as a single subject for ocular PK and each animal was treated as a single subject for systemic PK, respectively. Noncompartmental PK analyses were performed on each subject and PK parameters were reported as mean ± SD of 3 (for plasma PK parameters) or 6 subjects (for ocular PK parameters). The peak concentrations (Cmax) and the times they occurred (Tmax) were recorded. The area under the concentration-time curve (AUClast) was calculated using the linear trapezoidal rule. 

Effect of MGV354 on IOP in Dutch-Belted Rabbits, Following a Single Topical Ocular Dose

MGV354 (0.1%, 30 μg) or vehicle was instilled as a 30-μl drop in one eye of Dutch-belted rabbits (n = 8 rabbits/group), while the fellow eye was untreated. IOP was measured noninvasively as described previously³¹ by using an Alcon computerized pneumatonometer after the application of a drop of 0.1% proparacaine to anesthetize the cornea. IOP readings were taken before (baseline, 0 hours) and after dosing with vehicle or drug from 0.5 to 6 hours and again at 24 and 48 hours postdose. Changes from baseline IOP were determined and converted to percent change from baseline (which was set at 100%). Comparisons in IOP differences were made between vehicle and drug-dosed eyes in the two groups of rabbits. Data are represented as average percent IOP change (±SEM) versus time postdose. 

Measurement of IOP in Laser Trabeculoplasty-Induced Unilateral Ocular Hypertensive Cynomolgus Monkey Model of Glaucoma, Following Topical Ocular Dosing of MGV354

MGV354, Travatan, or vehicle were dosed once daily as topical ocular formulations in both eyes of a unilateral monkey model of glaucoma, and IOP was measured (noninvasively) before dosing (considered baseline) and at 1, 3, 6, and 24 hours postdose by using an applanation pneumatonometer on conscious monkeys, as described previously.³¹ Dosing was either done during the daytime (8 AM) or nighttime (8 PM), where indicated. IOP was stably elevated in the monkeys, following a unilateral laser-induced trabeculoplasty of the TM tissue up to a 325° axis, and scarring of 90% of the tissue left 10% of TM tissue intact. IOP was elevated from 25 to 27 mm Hg (prelaser baseline IOP) and became gradually stable at 33 to 40 mm Hg over a few months. Efficacy studies were performed after stable IOP elevation was reached. The fellow eye was left intact and exhibited IOP that was comparable to naïve monkey eyes (normotensive). All IOP studies were conducted in a masked manner. Between 8 and 10 cynomolgus monkeys were used per group, and typically an IOP study will include 2 groups (n = 16 to 20 monkeys/study). For all IOP studies, percent IOP change was determined by subtracting IOP readings at indicated timepoints postdosing from the predose baseline IOP value, which was set at 100%. Statistical analyses to detect differences in IOP between starting baselines and vehicle-dosed group was performed with repeated measures (RM) ANOVA, followed by the Bonferroni t-test, and significance was set at P < 0.05. 

Immunohistochemic Localization of sGC in Cynomolgus Monkey and Human Anterior Chamber Tissues

Five-μm-thick sections of human anterior segments were purchased from Excalibur Pathology (Norman, OK, USA). The preservation interval was 5 to 6 hours. Samples were fixed in 10% neutral buffered formalin (NBF) for 24 to 48 hours. Monkey eyes were collected and fixed in 10% NBF for 4 days and were paraffin embedded, sectioned at a 5-μm thickness, and mounted onto slides (Superfrost Plus, Fisher Scientific, Waltham, MA, USA). Immunohistochemistry was performed using BenchMark ULTRA (Ventana IHC/ISH slide staining system; Ventana Medical Systems, Inc., Tucson, AZ, USA). Slides were treated by Cell Conditioner #1 (Catalog no. 950–124; Ventana Medical Systems, Inc.) for 56 minutes and incubated with primary antibodies (rabbit anti-guanylyl cyclase α1 [ab50358], 7.5 μg/ml; rabbit anti-guanylyl cyclase β1, [ab154841], 5 μg/ml; AbCam, Cambridge, UK) for 60 minutes, and then were incubated with OmniMap anti-rabbit IgG (Catalog no. 760–149; Ventana Medical Systems, Inc.) and Discovery ChromoMap DAB kit (Catalog no. 76–159; Ventana Medical Systems, Inc.) for 12 minutes. Control sections were only incubated with anti-IgG and stained. Finally, stained slides were covered and scanned with the Aperio AT2 slide scanner (Leica, Wetzlar, Germany). 

Results

Expression of sGC Subunits in Cynomolgus Monkey and Human Anterior Chamber Tissues

The cynomolgus monkey glaucoma model was the key preclinical glaucoma model used to screen and identify sGC activators, including MGV354, and it was therefore important to determine the protein expression of the target enzyme. The intracellular sGC enzyme exists as a functional heterodimer consisting of α and β subunits. As shown in Figure 1, both human and cynomolgus monkey anterior chamber tissues express high levels of sGC (α/β), including TM and Schlemm's canal. sGC was also expressed in ciliary smooth muscle tissues in both species (data not shown). 

Figure 1

 

Detected by immunohistochemistry, sGC α and β subunits are highly expressed in TM and Schlemm's canal of human and monkey ocular tissues. Robust staining was also observed in the ciliary muscle tissues of both species (data not shown).

In Vitro Assessments of MGV354

Binding Affinity to Oxidized sGC

An AS-MS experiment was performed with 0.5-μM full-length human sGC (n = 2) and MGV354 (Fig. 2A) and tested at varying concentrations to evaluate its binding to the enzyme under reducing and oxidizing conditions The specificity of MGV354 to preferentially bind to the oxidized form of sGC was demonstrated by pretreatment with a known oxidizing agent, ODQ, and compared with that by using a reducing agent, TCEP. ODQ binds to the prosthetic heme moiety of the sGCβ subunit, changes it from ferrous (Fe²⁺) to ferric (Fe³⁺), and causes heme dissociation, which renders sGC unresponsive to NO.⁸ Under reducing conditions with TCEP, the heme moiety remains intact and is responsive to NO. The compound MGV354 was tested for binding to sGC under reducing and oxidizing conditions, and it demonstrated specific and saturable binding in both cases (Fig. 2B). With treatment of the oxidizing agent ODQ, MGV354 had a dissociation constant (K_d) of 0.49 μM (±0.11 μM SEM) and a Bmax of 4340 (±210 SEM). By contrast, in reducing conditions with TCEP, MGV354 had a similar specific binding with a K_d of 0.15 μM (±0.04 SEM), but a lower Bmax of 630 (±26 SEM). Thus, when sGC is oxidized, the capacity of the protein to bind the compound is increased by ∼7 fold, presumably by displacement of the oxidized heme group (Fig. 2B). Detection of heme in the same assay shows loss of heme as a function of increased compound concentration. 

Figure 2

 

(A) Structure of MGV354. (B) MGV354 binds to oxidized, heme-free human sGC enzyme. After treatment of the sGC enzyme with an oxidizing agent, ODQ, MGV354 has a K_d of 0.49 μM (±0.11 SEM) and a Bmax of 4340 (±210 SEM). Under reducing conditions with TCEP, MGV354 has similar specific binding, with a K_d of 0.15 μM (±0.04 SEM), but a lower Bmax of 630 (±26 SEM). Using AS-MS, we detected a 7-fold increase in maximal binding of MGV354 to the full-length human sGC (815/819) under oxidizing conditions with ODQ treatment compared to that of reducing conditions with TCEP treatment.

Broad Ligand Profiling

When profiled for determining off-target binding and/or function in a broad ligand panel, MGV354 (0.1, 1, and 10 μM) did not exhibit any significant binding in 79 receptors (Eurofins Panlabs, Taipei City, Taiwan). However, >50% binding was noted with the prostanoid receptors DP1, EP3, and IP at 10 μM of MGV354. MGV354 caused functional agonism of opioid-κ (65%; EC₅₀, 5.2 μM), alpha 2A adrenergic receptor (8.9 μM), and muscarinic receptor-M2 (EC₅₀, 3.9 μM). 

cGMP Production and Potency Determination in NTM Cells

MGV354 treatment for 1 hour caused a dose-dependent increase in cGMP levels in NTM cells (average EC₅₀ ± SD, 0.0025 ± 0.0016 μM; n = 6, independent experiments) following treatment with ODQ (20 μM). A representative figure is shown in Figure 3A. The average cGMP produced by MGV354 in human NTM cells was 46 ± 28 nM. MGV354 (1 μM) caused a linear increase in intracellular cGMP levels in a time-dependent manner (assay conditions up to 3 hours) in NTM cells (Fig. 3B). MGV354-induced cGMP production in the presence of ODQ was 8- to 10-fold greater than that observed in the absence of ODQ (Fig. 3B). These data indicated that the compound was preferentially more active under conditions wherein heme on sGC was oxidized (Fe3+) than when it was reduced (Fe²⁺-bound native state; absence of ODQ) (Fig. 3B). IBMX (pan-phosphodiesterase inhibitor, 1 mM) was included in these in vitro assessments. 

Figure 3

 

Production of intracellular cGMP by MGV354 in primary NTM cells. (A) Dose response curve for MGV354-induced cGMP production in NTM cells treated for 1 hour with different doses of MGV354; intracellular cGMP levels were measured thereafter by using an ELISA (Molecular Devices). IBMX (1 mM), a nonselective phosphodiesterase inhibitor, was included in all the treatment conditions to prevent cGMP hydrolysis. The EC₅₀ value of MGV354 in this representative experiment was 0.005 μM. The average EC₅₀ value was 0.0025 ± 0.0016 μM (n = 6 independent experiments). (B) Production of cGMP by MGV354 (1 μM) was determined in the absence or presence of ODQ (20 μM). Maximal cGMP production was observed in NTM cells in the presence of both ODQ and IBMX. MGV354 produced time- and ODQ-dependent increases in cGMP in primary NTM cells. NTM cells were incubated with MGV354 up to 3 hours in the presence and absence of ODQ.

Ocular and Systemic Exposure of MGV354 After Topical Ocular Dosing in Preclinical Species

The ocular and/or systemic exposure of MGV354 following a topical ocular administration (eye drop) was also assessed both in rabbits (ocular and systemic) and in monkeys (systemic) by using a Carbopol-Tyloxapol suspension formulation. 

In rabbits, 30 μl of 1% (300 μg) MGV354 was given to both eyes (bilateral, 600 μg total), which resulted in aqueous humor exposure (Cmax, 479 ng/mL; AUC, 1889 ng*h/mL) well above the in vitro EC₅₀ value (1.1 ng/mL), even at 24 hours postdose. MGV354 achieved the highest concentrations in the cornea (Fig. 4A). Of note, the back of eye exposure in the retina and choroid was substantially lower than what was observed in the anterior segment. The systemic exposure was also low in rabbits (Cmax of 5 ng/mL), as shown in Figure 4A. The TM or outflow tissue levels were not determined in this study due to technical challenges. 

Figure 4

 

Ocular and systemic exposure of MGV354 in (A) Dutch-belted rabbits and (B) systemic exposure in cynomolgus monkeys following topical ocular dosing. (A) Time-concentration profiles are displayed for MGV354 in rabbit ocular tissues and systemic circulation, following a topical ocular administration of 1% (300 μg) MGV354 (in Carbopol-Tylaxypol suspension eye drop, 30 μl, bilateral) (n = 3 rabbits/timepoint). Time-concentration profiles of MGV354 in systemic circulation in cynomolgus monkeys following a topical ocular administration of 0.01% (3 μg), 0.1% (30 μg), or 1% (300 μg) of MGV354 (in Carbopol-Tylaxypol suspension eye drop, 30 μl) to both eyes (bilateral) (n = 3 per dose).

The systemic exposure of MGV354 following a topical ocular administration (eye drop) in monkeys was assessed with 0.01%, 0.1%, and 1% dosing. The increase in systemic exposure in monkeys following 0.01% to 1% dose was roughly dose proportional. The dose-normalized AUCinf was 22, 17, and 25.67 h*ng/mL for 0.01%, 0.1%, and 1% doses, respectively. Cmax was also increased in a dose-dependent manner. The dose-normalized Cmax was 2.0, 1.88, and 0.806 ng/mL, respectively. The absorption via eye appears to be rapid. Tmax values for 0.01% and 0.1% doses were 0.5 and 0.8 hours. However, at 1%, Tmax was 2.3 hours, suggesting a slightly delayed absorption compared to that of lower doses. 

The systemic accumulation of MGV354 following multiple repeated dosing was assessed in monkeys after either 1% (unilateral, 300 μg total) or 0.1% (unilateral, 30 μg total) topical ocular dosing for 14 days. The systemic exposure at 1 hour after the last dose was ∼11 to 19 ng/mL or ∼50 to 123 ng/mL for 0.1% and 1%, respectively, which was comparable to that of the 1 hour exposure following a single dose. The results suggest that no accumulation of MGV354 occurred in monkeys following daily dosing for 14 days. 

Ocular Tolerability of MGV354 in Dutch-Belted Rabbits and Cynomolgus Monkeys Following Topical Ocular Dosing

Following 4 weeks of once-daily topical ocular dosing, slight to moderate ocular irritation was noted that generally resolved within 24 hours. Mild ocular hyperemia was noted in about one-third of drug-dosed animals at 3 to 16 hours postdose in pharmacology studies. No other local or systemic adverse effects were noted following topical ocular dosing. Other signs of discomfort such as eye rubbing were not observed following MGV354 dosing. No changes in pupil diameter were noted with MGV354 in either species. No other serious adverse effects including systemic and cardiovascular changes were noted with topical dosing of MGV354. 

In addition, other ocular assessments following MGV354 (0.1% and 1%) treatment in cynomolgus monkeys were made following once-daily dosing for 14 days. There were no signs of corneal damage, corneal cloudiness, or ocular inflammation (i.e., no flare or cells in the anterior chamber) with MGV354 dosing. No significant differences in CCT, ACD, or AXL between treated and untreated eyes were noted at any timepoint during the study (Supplementary Figs. S1–S3). 

Effect of MGV354 on IOP in Dutch-Belted Rabbits Following a Single Topical Ocular Dose

MGV354 (0.1%, 30 μg) caused an apparent maximal percent IOP lowering of 22% (±6%; P < 0.01, versus vehicle and baseline IOP) or a 6.4-mm Hg reduction from baseline at 6 hours postdose and remained lower at 11.8% (±6%; NS) at 24 hours postdose (Fig. 5). During the same period, vehicle did not cause any change in percent IOP. Baseline IOP was 29 ± 0.4 mm Hg in drug-dosed eyes, whereas baseline IOP in vehicle-dosed eyes was 30 ± 0.4 mm Hg. Mild hyperemia was the main ocular adverse effect noted in MGV354-dosed eyes that lasted up to 3 hours postdose in 6 out of 8 rabbits. 

Figure 5

 

Effect of MGV354 (0.1%, 30 μg) on IOP in Dutch-belted rabbits following a single topical ocular dose. Two groups of Dutch-belted rabbits were used in this study (n = 8 per group). One eye of each rabbit received either drug or vehicle, whereas the fellow eye was untreated. Baseline IOP was 29 ± 0.4 mm Hg in drug-dosed eyes, whereas that in vehicle-dosed eyes was 30 ± 0.4 mm Hg. *Denotes a significant difference in IOP between the indicated timepoint versus starting baseline, and **denotes a significant difference in IOP between the vehicle versus drug-treated eyes (RM ANOVA and Bonferroni test for comparison between treatment groups, P < 0.05).

Effect of MGV354 on IOP in Unilateral Laser Trabeculoplasty-Induced Ocular Hypertensive Cynomolgus Monkey Model of Glaucoma

When dosed in the morning as a single topical ocular formulation (30 μL drop), MGV354 (0.3 μg, 3 μg, and 30 μg at 0.001%, 0.01%, and 0.1%, respectively) caused a significant IOP reduction of 12% to 15% (5 to 6 mm Hg) at 3 hours and 19% to 26% (7 to 10 mm Hg) at 6 hours postdose in a monkey model of glaucoma (predose baseline IOP was 34 to 39 mm Hg) (RM ANOVA and Bonferroni test for multiple comparisons, P < 0.05). At 24 hours postdose, a dose-dependent IOP reduction was noted; 0.3 μg of MGV354 caused a 9% (3 mm Hg) IOP reduction (P < 0.01, versus vehicle and baseline IOP), 3 μg of MGV354 caused an 18% IOP reduction (7 mm Hg) (P < 0.001, versus vehicle and baseline IOP), whereas 30 μg of MGV354 resulted in a 24% IOP reduction (8 mm Hg) (P < 0.005, versus vehicle and baseline IOP) (Fig. 6A). In the normotensive fellow eye in monkeys, MGV354 (3 and 30 μg doses) caused a transient increase in IOP (by 6%–11%; 1.5 to 3 mm Hg) (P < 0.05, versus vehicle and baseline IOP) that lasted up to 1 to 3 hours postdose; however, at later timepoints (3 to 24 hours), IOP lowering of <5% below the predose baseline was noted (Fig. 6A [only 3 μg dose shown]). 

Figure 6

 

(A) Dose-response effect of MGV354 (0.001%, 0.01%, and 0.1%; 0.3 μg, 3 μg, and 30 μg) and (B) effect of Travatan (0.004%; 1 μg) following the administration of a single topical ocular drop at 8AM in the laser trabeculoplasty induced ocular hypertensive eyes of cynomolgus monkeys. The effect of 0.01% (3 μg) MGV354 on the normotensive cynomolgus monkey fellow eye is also provided to demonstrate the transient IOP increase (dashed line). IOP was measured at indicated timepoints postdose (n = 9 monkeys per dose). Vehicle was administered in a separate group of monkeys and the hypertensive eye IOP response is shown (n = 9). Percent IOP change was determined by subtracting IOP readings at indicated timepoints postdosing from the predose baseline IOP value, which was set at 100%. Data are average percent IOP change ± SEM. Baseline IOP readings are 34 to 39 mm Hg in the hypertensive eyes of monkeys, whereas that for normotensive eyes was 24 to 27 mm Hg. Statistical analyses for (A) are provided in the Results section. In (B), *Denotes a significant difference in IOP between the indicated timepoint and starting baseline, and **denotes a significant difference in IOP between the vehicle and drug-treated eyes (RM ANOVA and Bonferroni test for comparison between treatment groups, P < 0.05).

From a comparative perspective, Travatan (1 μg, 0.004%), a prostaglandin FP agonist, caused sustained IOP lowering of 15% to 19% (6 to 7 mm Hg) from 6 hours through 24 hours after a single AM dose in hypertensive eyes in a monkey model of glaucoma (Fig. 6B) (P < 0.01, versus vehicle and baseline IOP). Travatan-induced IOP reduction ranged between 5% and 10% compared to that of the predose baseline IOP readings (baseline IOP, 23 to 27 mm Hg) in the normotensive fellow eye of these monkeys (data not shown). 

MGV354 (30 μg, 0.1%) was also evaluated in a multiday (once-daily AM dosing for 7 days) IOP efficacy study in a monkey model of glaucoma. MGV354-induced IOP lowering was sustained from day 1 through day 7, with a percent IOP reduction ranging between 15% and 25% at 3 hours, 32% and 38% at 6 hours, and 23% and 38% at 24 hours postdose in the hypertensive eyes of cynomolgus monkeys (baseline IOP, 39 mm Hg) (P < 0.05, versus vehicle and baseline IOP) (Fig. 7). In normal eyes (baseline IOP, 25 mm Hg), a 6% to 10% increase at 1 to 3 hours and a 10% decrease in IOP at 24 hours postdose were noted on days 1 through 7 with MGV354 (Fig. 7). In similar multiday IOP studies in the same monkey model of glaucoma, Travatan (dosed once daily for 8 days) caused an IOP reduction of 3% to 14% at 3 hours, 12% to 18% at 6 hours, and 9% to 15% at 24 hours postdose on days 1 and 8 (P < 0.05, versus vehicle and baseline IOP) (Fig. 8). 

Figure 7

 

Effect of once-daily dosing of MGV354 for 7 days on IOP in a cynomolgus monkey model of glaucoma. Once-daily bilateral topical ocular instillation of 30 μl of 0.1% (30 μg) MGV354 suspension (n = 9 monkeys) or vehicle solution (n = 9 monkeys) was performed at 8 AM. Data are average percent IOP change ± SEM (n = 9/group). Baseline IOP readings in hypertensive eyes was 39 mm Hg and that for normotensive fellow eyes was 25 mm Hg. *Denotes statistical significance versus baseline IOP, and **denotes statistical significance versus vehicle by using RM ANOVA followed by Bonferroni t-test (P < 0.05).

Figure 8

 

Effect of once-daily dosing of Travatan for 8 days on IOP in a cynomolgus monkey model of glaucoma. The once-daily bilateral topical ocular instillation of 30 μl of 0.004% (1.2 μg) Travatan or vehicle solution was performed at 8 AM. Data are average percent IOP change ± SEM (n = 9/group). Baseline IOP readings in hypertensive eyes was 33 mm Hg and that for normotensive fellow eyes was 24 mm Hg. *Denotes statistical significance versus baseline IOP, and **denotes statistical significance versus vehicle by using RM ANOVA followed by Bonferroni t-test (P < 0.05).

MGV354 (3 and 30 μg doses) was also dosed in the nighttime (8 to 9 PM), and IOP was measured at 12 hours (8 AM), 16 hours (1 PM), 20 hours (5 PM), the following day, and 36 hours (9 AM) postdose. Both doses caused a robust IOP reduction of 30% to 37% (11 to 14 mm Hg) at the 12 to 20 hour timepoints (P < 0.001 for both doses versus vehicle and baseline IOP) and 6% to 13% (2 to 5 mm Hg) at 36 hours (P < 0.001 only for the 3 μg dose versus vehicle and baseline IOP) (Fig. 9). In a separate study, IOP reduction following a single PM administration of Travatan (1 μg, 0.004%) caused an IOP reduction of 28% (10 mm Hg) at 12 to 18 hours postdose (P < 0.001, versus vehicle and baseline IOP) (Fig. 9). The 36-hour timepoint was not measured with Travatan. It appeared that both doses of MGV354 caused a comparable IOP lowering seen with that of Travatan following a single PM-dosing regimen. 

Figure 9

 

A single topical ocular instillation of 0.01% (3 μg) or 0.1% (30 μg) MGV354, Travatan (0.004%, 1 μg), or vehicle solution in cynomolgus monkeys was performed at 8 PM, and IOP readings were measured the following day at 8 AM (12 hours), 12 noon (16 hours), 4 PM (20 hours), and 8 AM on the third day (36 hours) (mean ± SEM, n = 9/group). Data provided above are from three independent studies. Baseline IOP readings in hypertensive eyes of the MGV354 (30 μg) group were 33 mm Hg and 37 mm Hg for both the MGV354 (3 μg) group and the Travatan group. The baseline IOP reading in hypertensive eyes of the vehicle group was 34 mm Hg. *Denotes statistical significance versus baseline IOP, and **denotes statistical significance versus vehicle by using RM ANOVA followed by Bonferroni t-test (P < 0.05).

Discussion

The present study has identified MGV354 as a novel and selective sGC activator capable of lowering IOP for a sustained period in normal rabbits (20% reduction up to 6 hours) and in a monkey model of glaucoma (25%–35% IOP reduction up to 24 hours with a once-daily dose). sGC activators provide a new therapeutic target for affecting the conventional outflow pathway and the uveoscleral pathway in a glaucomatous eye that is likely subjected to oxidative stress, reduced NO/cGMP production, and excessive matrix deposition. 

sGC is a cytosolic enzyme consisting of an α/β heterodimer with a highly conserved full-length protein across human, cynomolgus monkey, and rabbit; the β1 subunit shows greater than 98% identity across species, whereas for the α1 subunit, the identity is 99% between human and cynomolgus monkey and 92% between human and rabbit. Therefore, we were able to use the rabbit and cynomolgus monkey as relevant species for safety and efficacy assessments. In addition, the expression of sGC in the anterior segment tissues of humans and cynomolgus monkeys enabled the use of latter species as the primary means of profiling of sGC activators, leading to the identification of MGV354 as a clinical candidate. 

As mentioned before, the NO/sGC/cGMP pathway is critically involved in the endogenous regulation of IOP based on the following findings in preclinical in vivo models and ex vivo/in vivo assessments in humans. NO donors enhance aqueous outflow in human explants.³² Organic nitrates lower IOP in normal and glaucomatous humans when administered intravenously or orally.³³ Latanoprostene bunod (Vyzulta; NO-donating latanoprost) was shown to lower IOP by 1.3 mm Hg better than latanoprost.¹⁸ In preclinical species, NO donors and a cGMP analog lower IOP in rabbits and nonhuman monkeys.^9,10,14 sGCα1 knockout mice exhibit elevated IOP and had retinal ganglion cell loss.¹³ Overexpression of NO synthase 3 (NOS-3) lowers IOP in mice.¹² Intracameral infusion of 8-bromo cGMP (cell-permeable and nonhydrolyzable cGMP analog) was shown to increase outflow facility by 25% to 30% in normal cynomolgus monkey eyes.³⁴ Recently, a sGC stimulator, IWP-953, was shown to increase conventional outflow facility in enucleated normal mouse eyes.³⁵ Furthermore, Ellis et al.,³⁶ demonstrated in porcine anterior segment perfusion studies that the DETA-NO (NO donor)-mediated increase in outflow facility was blocked by ODQ, suggesting that the oxidation state of sGC may be critical for proper outflow facility. 

It is also well known that perturbations in the anterior chamber, oxidative stress, and differential ECM upregulation can negatively affect TM/Schlemm's canal physiology and reduce outflow facility. What is unclear at this time is whether or not the aforementioned perturbations may also directly reduce levels of NO/cGMP and other intrinsic IOP regulating mechanisms, which in turn could exacerbate conventional outflow dysfunction in glaucoma. Thus, sGC activators likely offer the advantage of lowering IOP without the need for NO or nonoxidative environment, in contrast to a stimulator. A patent application from GlaxoSmithKline in 2015 also revealed the use of the sGC activator to lower IOP in Japanese white rabbits following topical and intraocular injections.³⁷ 

However, a key limitation about ascertaining the effectiveness of sGC activator versus stimulator in IOP lowering is the inability to directly measure the ratio of reduced versus oxidized state of sGC in situ/in vivo in preclinical models or in human glaucomatous eyes. This gap exists in the field of sGC biology, and its dysregulation under pathologic conditions has not been detected directly in vivo.³⁸ However, data from anterior segment perfusion studies in porcine eyes demonstrate that hydrogen peroxide (3 mM) perfusion at elevated IOP (i.e. 30 mm Hg) can reduce outflow facility.³⁹ The glaucomatous TM and outflow tissue appear to have higher levels of mitochondrial DNA damage, an indication of increased sensitivity to oxidative stress response.^40,41 Taken together, these data suggest that oxidative damage and elevated IOP could have a combinatorial effect in exacerbating TM/outflow pathway dysfunction. Therefore, the lesser magnitude and shorter duration of MGV354-induced IOP reduction in normotensive eyes of Dutch-belted rabbits and cynomolgus monkeys compared to the more pronounced and sustained IOP lowering observed in the hypertensive eyes of a cynomolgus monkey model of glaucoma could be due to a more prominent oxidative stress condition in hypertensive eyes, especially in the TM/outflow pathway. However, no assessments of oxidative stress were performed in either species that would allow a firm conclusion. As mentioned above, the inability to assess reduced versus oxidized state of sGC in vivo in the outflow pathway tissues is also a limitation. Aqueous humor levels of NO/cGMP have known to be reduced in glaucomatous patients compared to nonglaucomatous patients.^21,22 Therefore, it is interesting to speculate whether a reduction in endogenous NO results in elevated IOP in addition to increased oxidative stress in the glaucomatous anterior segment, which would also likely abrogate sGC-mediated relaxation of the outflow pathway. More studies are needed to prove these hypotheses. 

A transient increase in IOP (5%–10%) was noted in normotensive monkey eyes, which lasted up to 3 hours. There was no dose-dependent increase in the IOP elevation (data not shown). Studies were performed to determine whether or not this was due to changes in the blood aqueous barrier, but no signs of cells, flare, or inflammation were noted in the anterior chamber. No changes to pupil diameter were noted in either rabbits or monkeys with MGV354. No changes to ACD or AXL were observed with the 14-day dosing of MGV354 in monkeys (Supplementary Figs. S2, S3). There could be a few reasons for this transient rise in IOP in normotensive eyes, which include a rapid choroidal expansion, a rise in episcleral venous pressure, and a transient increase in aqueous humor formation in response to elevated levels of cGMP caused by MGV354.^42,43 From a safety perspective, this transient increase in IOP needs to be monitored in humans following MGV354 dosing. Regardless, the impressive 24-hour IOP reduction observed in a cynomolgus monkey model of glaucoma following a single topical ocular dose and its apparent sustained IOP lowering following 7 days of once-daily dosing provide the potential for MGV354 to achieve similar efficacy as that of prostaglandin analogs, including Travatan. Unfortunately, statistical comparisons between Travatan and MGV354 effects could not be made due to the large animal numbers required for these analyses. However, a trend toward improved efficacy of MGV354 over Travatan was noted, particularly following the evening-dosed IOP studies. These putative differences can be better assessed in adequately powered clinical trials. Whether IOP lowering with MGV354 is primarily due to an increase in outflow facility or uveoscleral outflow or suppression of aqueous humor inflow or a combination also needs to be determined in future preclinical studies and may be confirmed in clinical studies involving glaucoma patients. 

There are several potential mechanisms that can be activated downstream of sGC activation, including protein kinase G-mediated inhibition of Rho kinase, which in turn results in myosin light chain inactivation, BK_Ca2+-mediated hyperpolarization, and inhibition of Ca²⁺ influx via L-type channels (see Ref. 15 for review). Whether MGV354-mediated IOP lowering involves these and/or additional mechanisms are the subject of additional investigations. Some of the potential confounding factors and limitations in our current findings include the lack of measurement of aqueous humor or ocular tissue cGMP, a pharmacodynamic marker following target engagement with MGV354, and the potential for negative feedback signaling mechanisms, including activation of phosphodiesterases (PDEs) in outflow pathway tissues, which can dampen the MGV354-mediated IOP reduction. We were unable to measure cGMP levels in the aqueous humor of hypertensive eyes of cynomolgus monkeys, because this entailed an invasive procedure. Although it is possible that negative feedback loops, including PDE activation, may play a role in downregulation of cGMP-mediated signaling by causing its hydrolysis, the fact that we observed sustained IOP lowering with once-daily dosing for 7 days in a cynomolgus monkey model of glaucoma was indicative that such a negative effect was unlikely to occur. Confirmation, however, is necessary in human glaucomatous patients following dosing with MGV354. 

The main ocular adverse effects noted was ocular hyperemia lasting up to 12 hours postdose. This finding is potentially an “on-target” effect. The mechanism of cGMP-mediated vasodilation could be responsible for observed conjunctival hyperemia and suggestive of the existence of oxidized sGC in these vessels. Cinaciguat (BAY-58-2667), a sGC activator, when administered intravenously over 4 hours in healthy individuals, caused reduction in blood pressure and increased heart rate.⁴⁴ This study indicated that sGC exists in both oxidized and reduced state even in general circulatory tissues of healthy individuals, similar to that of conjunctival blood vessels in normal and glaucomatous eyes. Ocular hyperemia following topical dosing of MGV354 needs to be carefully monitored in the clinic. It is important to note that ocular hyperemia is also an ocular adverse effect for prostaglandin analogs and rho kinase inhibitors following topical dosing. 

Due to the sustained cardiovascular depressive effects seen with Cinaciguat following intravenous infusion in healthy volunteers⁴⁴ and in patients with acute heart failure,⁴⁵ it was important to determine the systemic exposure of MGV354. Topical ocular administration of 1% MGV3541 (bilateral, 600 μg total) to cynomolgus monkeys afforded a Cmax of only 9 nM and an AUCinf of 60 nM*h. The systemic exposure was decreased by 6.5× when a 0.1% dose was administered. With a dose of 0.01% (the lowest efficacious dose), the systemic exposure was lower by 45×. Based on the relatively low systemic exposure after topical ocular dosing, it is anticipated that cardiovascular adverse events are unlikely when employing a dose aimed at achieving maximal IOP lowering with MGV354. 

In conclusion, the sGC activator MGV354 lowers IOP in preclinical models of glaucoma with minimal ocular adverse effects and offers a new target of IOP reduction that may be evaluated in patients with ocular hypertension and glaucoma. 

Acknowledgments

The authors thank Rad Daly, Glen Travis Jernigan, Terri Krause, Chris Long, Levi Martin, Sarah Webb, and Shenouda Yacoub for the in vivo studies, including the rabbit ocular safety/tolerability, rabbit IOP, and cynomolgus monkey ocular safety/tolerability and IOP studies. Contributions from Viral Kansara, Tim Drew, Melissa Prentiss, and Ann Brown on conducting and analyzing PK data from in vivo studies are also recognized. Winnie Yang and John Pelletier are also acknowledged for their help with generating the human full-length sGC protein and performing AS-MS to determine and quantitate the affinity of compounds in the presence of oxidizing and reducing agents. The authors thank Jillian Queen for her role as the project manager. Doug Bevan and Nan Ji from Global Discovery Chemistry are also thanked for their efforts. Chenying Guo's help with some of the immunohistochemistry efforts are also appreciated. Contributions from Randall Kolega (technical research and development) on the development of a topical ocular formulation are greatly appreciated. The authors thank Cynthia Grosskreutz (translational medicine) and Thaddeus P. Dryja (ophthalmology research) for their guidance on the MGV354 program. 

Disclosure: G. Prasanna, Novartis (E), P; L. Ferrara, Novartis (E), P; C. Adams, Novartis (E), P; T. Ehara, Novartis (E), P; B. Li, Novartis (E); L. Yang, Novartis (E); C. Xiang, Novartis (E); C.T.H. Ng, Novartis (E); S. Kim, Novartis (E); C. Towler, Novartis (E); T. Topley, Novartis (E); C. McAllister, Novartis (E); M. Ghosh, Novartis (E); R. Newton, Novartis (E); R. Stacy, Novartis (E); D.S. Rice, Novartis (E); M. Mogi, Novartis (E)