A Free Energy Perturbation Approach to Estimate the Intrinsic Solubilities of Drug-like Small Molecules

Abstract

<p>Optimizing the solubility of small molecules is important in a wide variety of contexts, including in drug discovery where the optimization of aqueous solubility is often crucial to achieve oral bioavailability. In such a context, solubility optimization cannot be successfully pursued by indiscriminate increases in polarity, which would likely reduce permeability and potency. Moreover, increasing polarity may not even improve solubility itself in many cases, if it stabilizes the solid-state form. Here we present a novel physics-based approach to predict the solubility of small molecules, that takes into account three-dimensional solid-state characteristics in addition to polarity. The calculated solubilities are in good agreement with experimental solubilities taken both from the literature as well as from several active pharmaceutical discovery projects. This computational approach enables strategies to optimize solubility by disrupting the three-dimensional solid-state packing of novel chemical matter, illustrated here for an active medicinal chemistry campaign.</p>

Discovery of a Potent and Selective Tyrosine Kinase 2 Inhibitor: TAK-279

“…To address this challenge, we implemented a predictive RRCK model for passive permeability 41 in conjunction with a predictive FEP+ solubility model. 42 We recognized that optimization of kinetic solubility with these neutral molecules might be difficult, so it was important to ensure that they maintained good permeability in addition to adequate solubility to ensure high oral absorption. To improve solubility, we retained the fluoropyridyl-pyridone while exploring amide replacements for the cyclopropane ring (Table 3).…”

“…The majority of aryl and alkyl groups we evaluated maintained excellent binding potency to the TYK2 JH2 domain (K D values < 7 pM), with the notable exception of 2-methyl pyridone 33. As an initial triage strategy, kinetic solubility 42 and human hepatocyte stability were used to narrow our focus to compounds with an opportunity for good absorption and the potential for a lower predicted human dose relative to 11. The majority of compounds in Table 5 possess reduced human hepatocyte stability relative to 11 and compounds with promising stability (i.e., CL pred < 1.8 mL/min/kg) include 30, 31, 34, 38, 39, 45, and 47.…”

“…42 [M + H] + , LCMS purity: 99.70%, HPLC purity: 99.53%, CHIRAL HPLC purity: 100%,1 H NMR (DMSO-d 6 , 400 MHz): 9.05 (s, 1H), 8.27 (s, 1H), 8.37−8.35 (m, 2H), 8.23 (s, 1H), 8.12−8.05 (m, 1H), 7.97 (s, 1H), 7.88−7.87 (d, J = 8.0 Hz, 1H), 7.65−7.63 (d, J = 6.0 Hz, 1H), 7.58−7.56 (m, 1H), 6.48−6.45 (t, J = 6.8 Hz, 1H), 6.25 (s, 1H), 4.37−4.33 (m, 1H), 3.76− 3.74 (d, J = 7.6 Hz, 1H), 3.23 (s, 3H), 2.93 (s, 3H), 2.82−2.79 (m, 2H), 2.16−2.08 (m, 2H). FR-b was concentrated under reduced pressure at 30 °C to afford pure FR-b as compound ent-30 (0.026 g).…”