Explaining Accurate Predictions of Multitarget Compounds with Machine Learning Models Derived for Individual Targets

In drug discovery, compounds with well-defined activity against multiple targets (multitarget compounds, MT-CPDs) provide the basis for polypharmacology and are thus of high interest. Typically, MT-CPDs for polypharmacology have been discovered serendipitously. Therefore, over the past decade, computational approaches have also been adapted for the design of MT-CPDs or their identification via computational screening. Such approaches continue to be under development and are far from being routine. Recently, different machine learning (ML) models have been derived to distinguish between MT-CPDs and corresponding compounds with activity against the individual targets (single-target compounds, ST-CPDs). When evaluating alternative models for predicting MT-CPDs, we discovered that MT-CPDs could also be accurately predicted with models derived for corresponding ST-CPDs; this was an unexpected finding that we further investigated using explainable ML. The analysis revealed that accurate predictions of ST-CPDs were determined by subsets of structural features of MT-CPDs required for their prediction. These findings provided a chemically intuitive rationale for the successful prediction of MT-CPDs using different ML models and uncovered general-feature subset relationships between MT- and ST-CPDs with activities against different targets.

Scaffold Hybridization Strategy Leads to the Discovery of Dopamine D3 Receptor-Selective or Multitarget Bitopic Ligands Potentially Useful for Central Nervous System Disorders

In the search for novel bitopic compounds targeting the dopamine D₃ receptor (D₃R), the N-(2,3-dichlorophenyl)piperazine nucleus (primary pharmacophore) has been linked to the 6,6- or 5,5-diphenyl-1,4-dioxane-2-carboxamide or the 1,4-benzodioxane-2-carboxamide scaffold (secondary pharmacophore) by an unsubstituted or 3-F-/3-OH-substituted butyl chain. This scaffold hybridization strategy led to the discovery of potent D₃R-selective or multitarget ligands potentially useful for central nervous system disorders. In particular, the 6,6-diphenyl-1,4-dioxane derivative 3 showed a D₃R-preferential profile, while an interesting multitarget behavior has been highlighted for the 5,5-diphenyl-1,4-dioxane and 1,4-benzodioxane derivatives 6 and 9, respectively, which displayed potent D₂R antagonism, 5-HT_1AR and D₄R agonism, as well as potent D₃R partial agonism. They also behaved as low-potency 5-HT_2AR antagonists and 5-HT_2CR partial agonists. Such a profile might be a promising starting point for the discovery of novel antipsychotic agents.

Clock/Sleep-Dependent Learning and Memory in Male 3xTg-AD Mice at Advanced Disease Stages and Extrinsic Effects of Huprine X and the Novel Multitarget Agent AVCRI104P3

A new hypothesis highlights sleep-dependent learning/memory consolidation and regards the sleep-wake cycle as a modulator of β-amyloid and tau Alzheimer's disease (AD) pathologies. Sundowning behavior is a common neuropsychiatric symptom (NPS) associated with dementia. Sleep fragmentation resulting from disturbances in sleep and circadian rhythms in AD may have important consequences on memory processes and exacerbate the other AD-NPS. The present work studied the effect of training time schedules on 12-month-old male 3xTg-AD mice modeling advanced disease stages. Their performance in two paradigms of the Morris water maze for spatial-reference and visual-perceptual learning and memory were found impaired at midday, after 4 h of non-active phase. In contrast, early-morning trained littermates, slowing down from their active phase, exhibited better performance and used goal-directed strategies and non-search navigation described for normal aging. The novel multitarget anticholinesterasic compound AVCRI104P3 (0.6 µmol·kg^-1, 21 days i.p.) exerted stronger cognitive benefits than its in vitro equipotent dose of AChEI huprine X (0.12 μmol·kg^-1, 21 days i.p.). Both compounds showed streamlined drug effectiveness, independently of the schedule. Their effects on anxiety-like behaviors were moderate. The results open a question of how time schedules modulate the capacity to respond to task demands and to assess/elucidate new drug effectiveness.

Antinociceptive and Cytotoxic Activity of Opioid Peptides with Hydrazone and Hydrazide Moieties at the C-Terminus

In the present contribution, we analyze the influence that C-terminal extension of short opioid peptide sequences by organic fragments has on receptor affinity, in vivo analgesic activity, and antimelanoma properties. The considered fragments were based on either N-acylhydrazone (NAH) or N'-acylhydrazide motifs combined with the 3,5-bis(trifluoromethyl)phenyl moiety. Eleven novel compounds were synthesized and subject to biological evaluation. The analyzed compounds exhibit a diversified range of affinities for the µ opioid receptor (MOR), rather low δ opioid receptor (DOR) affinities, and no appreciable neurokinin-1 receptor binding. In three out of four pairs, N-acylhydrazone-based derivatives bind MOR better than their N'-acylhydrazide counterparts. The best of the novel derivatives have similar low nanomolar MOR binding affinity as the reference opioids, such as morphine and biphalin. The obtained order of MOR affinities was compared to the results of molecular docking. In vivo, four tested compounds turned out to be relatively strong analgesics. Finally, the NAH-based analogues reduce the number of melanoma cells in cell culture, while their N'-acylhydrazide counterparts do not. The antimelanoma properties are roughly correlated to the lipophilicity of the compounds.

Dual Inhibition of PDK1 and Aurora Kinase A: An Effective Strategy to Induce Differentiation and Apoptosis of Human Glioblastoma Multiforme Stem Cells

The poor prognosis of glioblastoma multiforme (GBM) is mainly attributed to drug resistance mechanisms and to the existence of a subpopulation of glioma stem cells (GSCs). Multitarget compounds able to both affect different deregulated pathways and the GSC subpopulation could escape tumor resistance and, most importantly, eradicate the stem cell reservoir. In this respect, the simultaneous inhibition of phosphoinositide-dependent kinase-1 (PDK1) and aurora kinase A (AurA), each one playing a pivotal role in cellular survival/migration/differentiation, could represent an innovative strategy to overcome GBM resistance and recurrence. Herein, the cross-talk between these pathways was investigated, using the single-target reference compounds MP7 (PDK1 inhibitor) and Alisertib (AurA inhibitor). Furthermore, a new ligand, SA16, was identified for its ability to inhibit the PDK1 and the AurA pathways at once, thus proving to be a useful tool for the simultaneous inhibition of the two kinases. SA16 blocked GBM cell proliferation, reduced tumor invasiveness, and triggered cellular apoptosis. Most importantly, the AurA/PDK1 blocker showed an increased efficacy against GSCs, inducing their differentiation and apoptosis. To the best of our knowledge, this is the first report on combined targeting of PDK1 and AurA. This drug represents an attractive multitarget lead scaffold for the development of new potential treatments for GBM and GSCs.