Toward a first drug therapy for the treatment of retinitis pigmentosa.

Abstract

The goal of this project was to contribute to the pharmaceutical development of CN03, the lead candidate for drug treatment of retinitis pigmentosa (RP) and other inherited retinal degenerations (IRDs). RP is an inherited neurodegeneration of the retina with with currently no treatments available on the market. In recent years, a molecular mechanism for the degeneration pathway of IRDs such as RP was found to be triggered by overactivation of protein kinase G (PKG) by cGMP, as a result of an unnatural build-up of intracellular cGMP. CN03, a phosphorothioate analogue of cGMP (cGMPS) with an RP configuration at phosphorus, was subsequently found to inhibit PKG and lead to preservation of photoreceptors.

The first task undertaken was to develop a chemical process for manufacture of this active pharmaceutical ingredient (API) at larger scales. A synthetic strategy was designed which makes use of H-phosphonate chemistry to construct the required 3′,5′-cyclic RP-phosphorothioate. The internal cyclization and sulfurization of the nucleoside 5′-H-phosphonate monoester to afford the 3′,5′-cyclic phosphorothioate could be steered to obtain the desired RP-diastereomer with 90% selectivity over the thermodynamically favored SP-phosphorothioate. To enable this chemistry, a strategy for protection of the 2′-OH nucleoside hydroxyl was needed. The silylation procedure was developed using triisopropylsilyl chloride, and achieved an impressive 80% selectivity for the 2′-OH over the 3′-OH. CN03 and intermediates were isolated by crystallization. Intermediate 5 was the only compound which could not be crystallized, but the crude could be telescoped to the following step. This allowed for a departure from chromatographic operations and enabled further scale-up. After optimization work, the process was scaled up to 14 liter scale (640 grams of the starting 8-bromoguanosine), which afforded ~125 g of the target CN03 in >99.9% purity. Access to larger amouts of the drug compound and a scalable process for future manufacture enabled further pharmaceutical development.

Next, solid forms of CN03 with low aqueous solubility were investigated. This was sought in order to extend the half-life of the drug compound in the eye, which is critical to aid in reducing the frequency of intravitreal injections – the most likely route of administration. Eight salt forms of CN03 as well as its free acid form (CN03-H) were synthesized. A total of 21 crystal modifications were identified with XRPD, some of which were characterized further with NMR. Thermal analyses of all salt forms showed no melting points despite their crystalline character, as they instead underwent thermal decomposition. Their hygroscopicity was studied with dynamic vapor sorption, where each API form exhibited unique sorption properties. The salts of calcium, benethamine, and benzathine (CN03-Ca, -Bnet, and -BZ respectively) achieved sub-milligram solubilities in water, with respective values of 300, 50, and 10 µg/mL. CN03-BZ was identified as the most promising CN03 solid form based on the lowest solubility as well as lowest hygroscopicity. A single crystal X-ray diffraction (SXRD) measurement was also made which led to the first solved crystal structure for a cGMPS analogue in the literature. Determination of the absolute configuration confirmed the RP-chirality of CN03, which had previously been indirectly inferred. SXRD also served to reveal the stabilizing interactions in the crystal lattice.

Finally, three closely related analogues of CN03 were prepared and their effect on in vitro models of RP was studied. The phosphate analogue and the SP-phosphorothioate diastereomer (oxo-CN03 and SP-CN03) were expected to promote photoreceptor cell degeneration, while the activity of the corresponding phosphorodithioate (dithio-CN03) was unknown. All compounds could be readily prepared using the same synthetic process as for CN03 with only minor adaptations to one step, which marked the first synthesis of 3′,5′-cyclic phosphorodithioate nucleosides using H-phosphonothioates reported in the literature. In vitro studies found that SP-CN03 was not toxic to photoreceptors, but instead preserved photoreceptors to a similar level as CN03 itself, alleviating concerns of its presence as an impurity in the drug product. Oxo-CN03 was not significantly more toxic than CN03 or SP-CN03, and as expected provided no neuroprotection. The most promising compound was the novel dithio-CN0, which was identified as the least toxic and provided significantly more protection to photoreceptors than current drug candidate CN03. The aqueous solubility of dithio-CN03 was also measured to explore its viability as a drug candidate for intravitreal applications. The solubility of dithio-CN03 as a triethylammonium salt was measured at 1.8 mM, compared to 13.7 mM for the same salt of CN03, which further establishes dithio-CN03 as a potentially more effective drug compound as part of sustained-release formulations for treatment of RP via intravitreal injections.