Antiglaucoma EP 2 Agonists: A Long Road That Led Somewhere

Introduction

On September 21, 2018, omidenepag, the first prostanoid EP₂ receptor agonist, was approved in Japan for the treatment of glaucoma and ocular hypertension. Omidenepag is a potent EP₂ agonist^1–4 and is present as the active pharmaceutical ingredient in the product Eybelis. The treatment regimen is once daily dosing. Already approved for manufacture and marketing in Japan, global development of the product is intended. This is a significant accomplishment, especially considering the time it has taken to introduce an EP₂ agonist into the ophthalmologist's armamentarium of antiglaucoma drugs.

The first ocular hypotensive EP₂ agonist was described as long ago as 1993, ⁵ this is contemporaneous with the research, development, and product phases on prostaglandin F_2α (PGF_2α) prodrugs. Over these 25 years, EP₂ agonists have been the subject of antiglaucoma research and development by a number of companies and individuals and much has been learned. In this article, each aspect of the EP₂ research is reviewed and an extensive list of source references is provided. Has it all been worth it? Absolutely.

Pharmacology

Originally and correctly defined by traditional pharmacological experimentation as 1 of 4 receptor subtypes that are preferentially recognized by prostaglandin E₂ (PGE₂), the EP₂ receptor was the final prostanoid receptor to be structurally defined as a single polypeptide with 7 transmembrane spanning domains, 3 extracellular loops with an amino terminus, and 3 intracellular loops with a carboxylate terminus.^6–9

The 25 years of ocular EP₂ research spanned the transition from pharmacological methods using isolated tissue and cell assays to high-throughput methods using human recombinant receptor transfectants.^6,7,10 Elevated intraocular pressure (IOP) was among the earliest medical indications for EP₂ agonists.^6,11,12 Additional therapeutically useful properties included bronchodilation,^13–15 uterine smooth muscle relaxation, ¹⁵ and anti-inflammation at multiple levels.^8,17–20 Neuroprotective indications also were identified. EP₂ receptor activation was shown to protect neurons against cerebral ischemia, oxidative stress, and N-methyl-D-aspartate (NMDA) receptor-mediated cytotoxicity.^21–25 This was followed by studies that indicated neuroprotection associated with glaucoma and other neurodegenerative ocular diseases.^26–29

Ocular Therapeutics

Potential ocular anti-inflammatory,^6,7,10 antineovascularization, ²⁰ and neuroprotective^21–26 effects of EP₂ receptor agonists have received very little attention. The focus for EP₂ agonists has been glaucoma, understandably by virtue of the profound ocular hypotensive effects that are persistent to the point where once-a-day dosing is feasible.

An increase in uveoscleral outflow has been found consistently to significantly contribute to lowering IOP. A seminal study showed that decreased IOP produced by the selective EP₂ receptor agonist, butaprost, exclusively involved an increase in uveoscleral outflow. ³⁰ Butaprost reduced the IOP of laser photocoagulation-induced hypertensive monkey eyes to the baseline levels of the nonlasered contralateral eyes but, nevertheless, no effect on pressure-dependent outflow facility or aqueous humor inflow was apparent. ³⁰ After 1 year treatment with butaprost, an organized widening of the intermuscular spaces in the anterior portion of the ciliary muscle was apparent. ³⁰ The formation of these outflow channels between ciliary muscle bundles was also apparent after a 1 year treatment regimen with another selective EP₂ agonist, AH 13205. ³¹

Numerous studies provide a structural correlate for increased uveoscleral outflow in response to EP₂ receptor agonist treatment. A direct effect on ciliary smooth muscle cells is a decrease in cell stiffness in response to butaprost. ³² Insight into one possible biochemical pathway in EP₂ agonist-induced ciliary muscle remodeling was the finding that butaprost upregulated Cyr61 (cysteine-rich angiogenic protein 61) in human ciliary smooth muscle cells. ³³ Butaprost-sensitive EP₂ receptors were also found functionally present on human trabecular meshwork cells,^32,33 but clearly do not operate to increase aqueous humor outflow facility in ocular normotensive monkeys.^30,31 In contrast, the EP₂ agonist, omidenepag, caused an increase in outflow facility in addition to an increase in uveoscleral outflow in ocular hypertensive monkeys. ³

The most recent antiglaucoma agent JV-GL1 (PGN 9856-isopropyl ester) causes a decrease in aqueous humor secretion and increases in outflow facility and uveoscleral outflow. It was exceptionally long acting with significant reductions in IOP that could persist for as long as 2 weeks after a single dose administered as an eye drop. ³⁴ Other unexpected differences between the ocular activities of selective EP₂ agonists are reported. As yet, these are mostly unexplained. Although alternative mRNA splicing variation and heterodimerization of prostanoid receptors have provided elucidation at the molecular level for naturally occurring eicosanoids such as prostamide Fα^35–38 and 8-epi PGE₂,^39,40 no such molecular phenomena have been reported for EP₂ receptors.

Chemical Structure, Ocular Activity Relationships

The earliest selective EP₂ agonists were butaprost ⁴¹ and AH 13205, ⁴² which retained the typical C20 eicosanoid backbone of PGE₂. 19(R)-OH PGE₂ is a naturally occurring EP₂-selective agonist. ⁴³ All lower IOP but butaprost appears to be the most efficacious, likely by virtue of the carboxylic acid group being derivatized to a methyl ester with the intention to enhance ocular penetration. A number of compounds were later designed and chemically synthesized that have a clear structural resemblance to PGE₂. An important early discovery was that the 11-OH group on the cyclopentenone ring of PGE₁ and PGE₂ could be removed without a decrease in EP₂ receptor potency.^44,45 The selectivity for EP₂ receptors relative to EP₄ receptors was modest, increasing the potential for a greater degree of unwanted clinical side effects. ⁴⁵

An important consideration for ophthalmic drug delivery was that 11-deoxy compounds were thereby made stable in aqueous formulations by preventing elimination of H₂O from the cyclopentenone ring to form chemically reactive compounds of the prostaglandin analog series. Interestingly, 11-deoxy PGE₁ was less efficacious and shorter acting than butaprost in lowering monkey IOP.^30,44 On reflection, the differential effects of butaprost, AH 13205, and 11-deoxy PGE₁ on IOP were an early portend of unexpected findings between compounds in the eye.

The ability of α and ω chain substituents to confer receptor-specific selectivity and potency is well known and the cyclopropyl moiety in the lower chain of butaprost ⁴¹ was conserved in many EP₂ agonist designs.^46–49 In the design of EP₂ agonists for ophthalmic use, replacement of the cyclopentenone ring was of critical importance. Thus, numerous compounds were designed that involved permutations of the positions and/or presence of the 9-keto and 11-OH moieties, with Cl being a widely used substituent at both the 9 and 11 positions.^50–68

One chemical synthetic program of EP2 agonists involved a large series of lactams (cyclic amides). No biological data are reported for β-lactams,^69,70 but the pentenone γ-lactams were a particularly successful series of potent and selective EP₂ agonists.^71–82 Curiously, the first γ-lactams were EP₄ selective. ⁷¹ The increase from a 5-membered ring to the 6-membered ring of the δ-lactams resulted in a complete loss of EP₂ agonist activity but interestingly resulted in selective EP₄ agonists.^83–87 In short, the γ-lactams provided an invaluable scaffold, which tolerated a range of substitutions and ring insertions to create novel heterocycles and led to an even wider range of EP₂-selective compounds.^88–91

The cyclic amide ring was opened up in one series of compounds to replace the cyclic structure with an amide, the N atom positioned at the start of the lower chain. ⁹² Information used from elaborating PGE-like structures enabled a nonprostanoid series based on indenone and tetralone; this series included some potent and selective EP₂ agonists. ⁹³ A unique nonprostanoid was based on a naphthylamide at the core of the scaffold. ⁹⁴ This series also comprised some highly active and selective EP₂ agonists. ⁹⁴ Many other potent and selective EP₂ agonists have been designed.^95–97 Since no information was uncovered indicating that glaucoma was the therapeutic target disease, they are beyond the scope of this article.

Clinical success has been achieved with a nonprostanoid-like structure. ⁹⁸ Omidenepag retains most of the chemical structural features of the first nonprostanoid EP₂ agonist to be reported, CP 533536. ⁹⁹ The scaffold of omidenepag closely resembles that of CP 533536^98,99 with the following substitutions: pyrazole for tertiary butyl and pyridine-2-ylamino acetic acid for phenyl-2-oxo-acetic acid.^98,99 For omidenepag, an ∼10-fold increase in activity was achieved at the EP₂ receptor compared with CP 533536.^98,99 This does not appear to translate into a meaningful difference in ocular hypotensive efficacy, although no head-to-head comparison is available. It appears to be a side effect profile that “won the day” for omidenepag.

Ocular side effects can be an issue associated with any drug and these may not always be exposed until dedicated safety evaluation is conducted. Safety and side effect information has been provided for taprenepag (CP 533536-isopropyl ester). The taprenepag ocular side effects were published, such information is not always available in the public domain. Iritis and related photophobia were both reported for trapenepag. ¹⁰⁰ Iritis seems unlikely to be the result of either EP₂ receptor stimulation or nonspecific irritation: it could be the result of off-target activity at the prostanoid DP₂ (CRTh2) receptor, since this receptor is proinflammatory.^101,102 Prostanoid EP₂ agonists would not be expected to initiate inflammatory responses, since they exhibit broad anti-inflammatory properties.^{17–20,103–106} Taprenepag also has been reported to cause increased corneal thickness, which appears to be reversible and independent of evidence for corneal epithelial or endothelial toxicity.^107,108 These side effects are not reported with omidenepag.

For existent EP₂ agonists, polar ester modification of EP₂ receptor agonists has been proposed as a method of attenuating increased corneal thickness.^109,110 Application of EP₂ agonists to the periorbital skin ³⁴ would preferentially deliver drugs to the limbal/scleral areas of the globe, thereby limiting corneal exposure and corneal thickening. Ocular discomfort and reflex eyelid closure also would be reduced by periorbital drug delivery. ³⁴

Ocular Biodisposition and Drug Delivery

It is regrettable that there is a paucity of drug metabolism/pharmacokinetic data on EP₂ receptor agonists provided by research/drug development organizations. It follows that important information, as detailed hereunder, may not be taken into adequate account when furthering drug development and delivery. The sclera may be a more important route of accession to the globe than the cornea for topically applied glaucoma medications.^111,112 Most of the drug applied to the ocular surface distributes into the eyelids and periorbital skin.^113,114 Most surprising results were obtained when studying the ocular biodisposition of JV-GL1: (a) after application of a single eye drop, the ocular hypotensive effects persisted for several days after the drug was undetectable in ocular tissues and (b) a simple bioisosteric replacement in the JV-GL1 molecule resulted in an inactive compound, which was explained as a virtual absence of ocular bioavailability.

Owing to poor corneal/scleral penetration, patient adherence issues, and adverse events, a litany of alternative drug delivery methods have been introduced over many years, only 2 of the newest and most promising are mentioned herein. Intraocular implants, injected into the anterior chamber, seem very promising.^115,116 A 6-month IOP-lowering effect was achieved with a bimatoprost sustained release implant. ¹¹⁶ Implants may be used to deliver EP₂ agonists^43–99 in the future, although corneal thickening and inflammation would need to be closely monitored to ensure adequate safety. Drug delivery to the periorbital skin is intended for the long acting compound, JV-GL1, ³⁴ and should represent a drug delivery method that is noninvasive and more convenient than eye drops. In the future, periorbital drug delivery might be the norm for all ocular EP₂ agonists.^43–99

EP₂ agonists have been widely reported to exert neuroprotective properties,^21–29 but sufficient retinal tissue exposure is difficult to achieve by using eye drops or anterior chamber implants. Although EP₂ agonist molecules may possess dual ocular hypotensive and neuroprotective activities, current drug delivery obstacles prevent providing to patients this advantageous therapeutic property.

The Past and the Future

In the case of PGF analogs, several different compounds were approved by drug regulatory authorities and launched into clinical practice within a relatively short space of time. In contrast, omidenepag is the only EP₂ agonist available to ophthalmologists for the foreseeable future. Most organizations that invented EP₂ agonists either never entered or dropped out of the antiglaucoma “race.” The inventors of omidenepag should be credited with an appreciation of the wide range of activity profiles related to structurally diverse compounds interacting with a singular target, namely the EP₂ receptor in this instance. This credit is particularly due since the initial clinical findings related to inflammation and increased corneal thickness were a discouragement,^100,107,108 especially as EP₂ receptor density is relatively high in the corneal epithelium. ¹¹⁷ The recent renaming of PGN 9856-isopropyl ester ³⁴ to JV-GL1 indicates that this long-acting and potent EP₂ receptor agonist is now under development as an antiglaucoma drug.