Accelerated Discovery of Novel Ponatinib Analogs with Improved Properties for the Treatment of Parkinson’s Disease

Abstract

Parkinson’s disease (PD) is a debilitating and common neurodegenerative disease. New insights implicating c-Abl activation as a driving force in PD have opened a new drug development avenue for PD treatment beyond the symptomatic relief by L-DOPA. BCR-Abl inhibitors, which include nilotinib and ponatinib, have been found to inhibit this process, and nilotinib has shown improvement in outcomes in a 12-patient, nonrandomized trial. However, nilotinib is a potent inhibitor of hERG, a cardiac K+ channel whose inhibition increases risk of sudden death. We used our machine learning approach to predict novel molecules that would inhibit c-Abl while also having minimal liability against hERG. Of our six novel compounds tested, we identified two that had c-Abl potencies comparable to nilotinib, but with significantly improved profiles regarding the hERG channel. Our best compound exhibited a hERG IC50 of 12.1 μM (compared to nilotinib with an IC50 of 0.45 μM and ponatinib with IC50 of 0.767 μM). This work is a step forward for a machine learning enabled, multiparameter optimization of a chemical space and represents a significant advance in the development of novel Parkinson’s therapies.

Prospective evaluation and success of a machine learning hit-to-lead drug development program against phosphatidylinositol 3-kinase α

“…If synthetic difficulties were encountered with a given compound, we would move to the next attractive target on the list. Our work on carbonic anhydrase II (CAII) has been published previously; 13 we now wish to report our findings on PI3Kα.…”

“…We have evaluated the possibility of using FRESH as a lead development tool for reducing the intensity of synthetic preparation required in the optimization process. In both CAII 13 and PI3Kα, FRESH was capable of generating tens of thousands of synthesis candidates and reducing that pool to tens of compounds predicted to be highly active and have good pharmacological parameters. In the case of PI3Kα, all of our synthesized compounds were found to have low nanomolar activity against the enzyme and low-to-submicromolar activity against the E545K mutant MCF-7 cell line.…”

“…The prioritized list obtained was further filtered for structures already extant in PubMed, thus removing the Kim et al compounds. Similar to the CAII prospective analysis, 13 we again pursued only those compounds that were quick to synthesize; those that had a high synthetic burden were discarded. General considerations for synthetic candidates could be summarized as follows: 1) the compound must be synthetically accessible from inexpensive building blocks; 2) the compound must have few to no stereocenters; 3) the compound must be outside the established IP; 4) the compound should be drug-like in appearance; 5) the compound should be made within two months or it is abandoned, and the next lower compound should be attempted.…”

“…[6][7][8] There are now abundant virtual screening methods working from purely 2D descriptors like extended connectivity fingerprints and molecular graphs, hybrid 2D/3D methods of virtual screening and purely 3D methods of screening both from a machine learning perspective and from a structural biology perspective. [9][10][11][12][13][14][15][16] As part of our initial investigation of new virtual screening methods, we describe here a second prospective evaluation of our original 2D/3D hybrid cheminformatics protocol termed FRESH. 11 We have now demonstrated across two distinct biological targets and chemical spaces that FRESH can dramatically increase the ease of identifying commercially unavailable, quality lead compound candidates and should, as a consequence, reduce the attrition rate of compounds entering the clinic and increase the cost-effectiveness of drug development.…”

Artificial intelligence to deep learning: machine intelligence approach for drug discovery

“…Here, the QSAR model was built using an optimum set of molecular descriptors, which were sorted out using an amalgamation of ML algorithms, hybridization techniques, backward elimination strategy, and visual analysis [ 469 ]. Similarly, [ 470 ] used a cascade of Naïve Bayes networks to find potent and safe abelson tyrosine-protein kinase 1 (c-Abl) inhibitors, which promote neuroprotection in PD. Likewise, Shao et al 2018 used integration of SVM algorithm and Tanimoto similarity-based clustering, followed by in vitro experiments, to find novel antagonists of both A 2A adenosine receptor as well as Dopamine D 2 receptor, as it has been observed that blocking these two receptors leads to neuroprotection in PD [ 471 ].…”