Improving therapeutic synergy score predictions with adverse effects using multi-task heterogeneous network learning.

Drug combinations could trigger pharmacological therapeutic effects (TEs) and adverse effects (AEs). Many computational methods have been developed to predict TEs, e.g. the therapeutic synergy scores of anti-cancer drug combinations, or AEs from drug-drug interactions. However, most of the methods treated the AEs and TEs predictions as two separate tasks, ignoring the potential mechanistic commonalities shared between them. Based on previous clinical observations, we hypothesized that by learning the shared mechanistic commonalities between AEs and TEs, we could learn the underlying MoAs (mechanisms of actions) and ultimately improve the accuracy of TE predictions. To test our hypothesis, we formulated the TE prediction problem as a multi-task heterogeneous network learning problem that performed TE and AE learning tasks simultaneously. To solve this problem, we proposed Muthene (multi-task heterogeneous network embedding) and evaluated it on our collected drug-drug interaction dataset with both TEs and AEs indications. Our experimental results showed that, by including the AE prediction as an auxiliary task, Muthene generated more accurate TE predictions than standard single-task learning methods, which supports our hypothesis. Using a drug pair Vincristine-Dasatinib as a case study, we demonstrated that our method not only provides a novel way of TE predictions but also helps us gain a deeper understanding of the MoAs of drug combinations.

Design, synthesis, and biological evaluation of dihydroartemisinin-fluoroquinolone conjugates as a novel type of potential antitubercular agents.

Tuberculosis remains a global public health problem in recent years. To develop novel type of potential antitubercular agents, twelve novel dihydroartemisinin-fluoroquinolone (DHA-FQ) conjugates (three types of molecules) were gradually designed and conveniently synthesized. All the newly synthesized conjugates were well characterized and evaluated against different Mycobacterium tuberculosis strains in vitro. The screening results showed that five DHA-FQ conjugates were active toward M. tuberculosis H37Rv, and compound 3a exhibited the strongest inhibitory activity (MIC=0.0625 µg/mL), which was comparable to the positive control Moxifloxacin and even stronger than Ofloxacin. Conjugates 2a and 3a also displayed comparable activities against various clinically isolated sensitive and resistant M. tuberculosis strains (MIC=0.125-16 µg/mL) to Moxifloxacin. All target compounds possessed selective anti-M. tuberculosis ability. Preliminary structure-activity relationship demonstrated that short linker between DHA and FQ was favorable for strong antitubercular activity. This study provides a new clue for the development of novel antitubercular lead molecules.

Cobdock: an accurate and practical machine learning-based consensus blind docking method.

Probing the surface of proteins to predict the binding site and binding affinity for a given small molecule is a critical but challenging task in drug discovery. Blind docking addresses this issue by performing docking on binding regions randomly sampled from the entire protein surface. However, compared with local docking, blind docking is less accurate and reliable because the docking space is too largetly sampled. Cavity detection-guided blind docking methods improved the accuracy by using cavity detection (also known as binding site detection) tools to guide the docking procedure. However, it is worth noting that the performance of these methods heavily relies on the quality of the cavity detection tool. This constraint, namely the dependence on a single cavity detection tool, significantly impacts the overall performance of cavity detection-guided methods. To overcome this limitation, we proposed Consensus Blind Dock (CoBDock), a novel blind, parallel docking method that uses machine learning algorithms to integrate docking and cavity detection results to improve not only binding site identification but also pose prediction accuracy. Our experiments on several datasets, including PDBBind 2020, ADS, MTi, DUD-E, and CASF-2016, showed that CoBDock has better binding site and binding mode performance than other state-of-the-art cavity detector tools and blind docking methods.

FLONE: fully Lorentz network embedding for inferring novel drug targets.

Motivation: To predict drug targets, graph-based machine-learning methods have been widely used to capture the relationships between drug, target and disease entities in drug-disease-target (DDT) networks. However, many methods cannot explicitly consider disease types at inference time and so will predict the same target for a given drug under any disease condition. Meanwhile, DDT networks are usually organized hierarchically carrying interactive relationships between involved entities, but these methods, especially those based on Euclidean embedding cannot fully utilize such topological information, which might lead to sub-optimal results. We hypothesized that, by importing hyperbolic embedding specifically for modeling hierarchical DDT networks, graph-based algorithms could better capture relationships between aforementioned entities, which ultimately improves target prediction performance. Results: We formulated the target prediction problem as a knowledge graph completion task explicitly considering disease types. We proposed FLONE, a hyperbolic embedding-based method based on capturing hierarchical topological information in DDT networks. The experimental results on two DDT networks showed that by introducing hyperbolic space, FLONE generates more accurate target predictions than its Euclidean counterparts, which supports our hypothesis. We also devised hyperbolic encoders to fuse external domain knowledge, to make FLONE enable handling samples corresponding to previously unseen drugs and targets for more practical scenarios. Availability and implementation: Source code and dataset information are at: https://github.com/arantir123/DDT_triple_prediction. Supplementary information: Supplementary data are available at Bioinformatics Advances online.

α-Tocopherol Pharmacokinetics in Adults with Cystic Fibrosis: Benefits of Supplemental Vitamin C Administration.

BACKGROUND: Numerous abnormalities in cystic fibrosis (CF) could influence tocopherol absorption, transportation, storage, metabolism and excretion. We hypothesized that the oxidative distress due to inflammation in CF increases vitamin E utilization, which could be positively influenced by supplemental vitamin C administration. METHODS: Immediately before and after receiving vitamin C (500 mg) twice daily for 3.5 weeks, adult CF patients (n = 6) with moderately advanced respiratory tract (RT) disease consumed a standardized breakfast with 30% fat and a capsule containing 50 mg each hexadeuterium (d6)-α- and dideuterium (d2)-γ-tocopheryl acetates. Blood samples were taken frequently up to 72 h; plasma tocopherol pharmacokinetics were determined. During both trials, d6-α- and d2-γ-tocopherols were similarly absorbed and reached similar maximal plasma concentrations ~18-20 h. As predicted, during vitamin C supplementation, the rates of plasma d6-α-tocopherol decline were significantly slower. CONCLUSIONS: The vitamin C-induced decrease in the plasma disappearance rate of α-tocopherol suggests that vitamin C recycled α-tocopherol, thereby augmenting its concentrations. We conclude that some attention should be paid to plasma ascorbic acid concentrations in CF patients, particularly to those individuals with more advanced RT inflammatory disease and including those with severe exacerbations.

Optimization of a series of 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidine inhibitors of IGF-1R: elimination of an acid-mediated decomposition pathway.

Initial evaluation of a series 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidines revealed a C(1') carboxamide was preferred for sub-micromolar in vitro potency against IGF-1R. Subsequent solution stability studies with 1 revealed a susceptibility toward acid-induced intramolecular cyclization with the C(1') carboxamide. Herein, we describe several successful approaches toward generating both potent and acid-stable inhibitors of IGF-1R within the 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidine template.

Discovery of 3,5-disubstituted-1H-pyrrolo[2,3-b]pyridines as potent inhibitors of the insulin-like growth factor-1 receptor (IGF-1R) tyrosine kinase.

Exploration of the SAR around a series of 3,5-disubstituted-1H-pyrrolo[2,3-b]pyridines led to the discovery of novel pyrrolopyridine inhibitors of the IGF-1R tyrosine kinase. Several compounds demonstrated nanomolar potency in enzyme and cellular mechanistic assays.

Optimization of 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidine IGF-1R tyrosine kinase inhibitors towards JNK selectivity.

The SAR of C5' functional groups with terminal basic amines at the C6 aniline of 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidines is reported. Examples demonstrate potent inhibition of IGF-1R with 1000-fold selectivity over JNK1 and 3 in enzymatic assays.

Discovery of 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidines: potent inhibitors of the IGF-1R receptor tyrosine kinase.

The evaluation of a series of 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidines as inhibitors of the IGF-1R (IGF-IR) receptor tyrosine kinase is reported. Examples demonstrate nanomolar potencies in in vitro enzyme and mechanistic cellular assays as well as promising in vivo pharmacokinetics in rat.

Reversible carboxamide-mediated internal activation at C(6) of 2-chloro-4-anilino-1H-pyrrolo[2,3-d]pyrimidines.

A synthetic route to bisanilino-1H-pyrrolo[2,3-d]pyrimidines has been discovered, wherein the C(6)-chloride reactivity is necessarily enhanced via reversible acid-catalyzed internal activation of the pyrimidine ring by a C(1')-carboxamide moiety. Subsequent selective nucleophilic displacements at C(6) and C(1') constitute a one-pot tandem protocol for the rapid assembly of bisanilino-1H-pyrrolo[2,3-d]pyrimidines.

Identification of NVP-TNKS656: the use of structure-efficiency relationships to generate a highly potent, selective, and orally active tankyrase inhibitor.

Tankyrase 1 and 2 have been shown to be redundant, druggable nodes in the Wnt pathway. As such, there has been intense interest in developing agents suitable for modulating the Wnt pathway in vivo by targeting this enzyme pair. By utilizing a combination of structure-based design and LipE-based structure efficiency relationships, the core of XAV939 was optimized into a more stable, more efficient, but less potent dihydropyran motif 7. This core was combined with elements of screening hits 2, 19, and 33 and resulted in highly potent, selective tankyrase inhibitors that are novel three pocket binders. NVP-TNKS656 (43) was identified as an orally active antagonist of Wnt pathway activity in the MMTV-Wnt1 mouse xenograft model. With an enthalpy-driven thermodynamic signature of binding, highly favorable physicochemical properties, and high lipophilic efficiency, NVP-TNKS656 is a novel tankyrase inhibitor that is well suited for further in vivo validation studies.

1-amino-4-benzylphthalazines as orally bioavailable smoothened antagonists with antitumor activity.

Abnormal activation of the Hedgehog (Hh) signaling pathway has been linked to several types of human cancers, and the development of small-molecule inhibitors of this pathway represents a promising route toward novel anticancer therapeutics. A cell-based screen performed in our laboratories identified a new class of Hh pathway inhibitors, 1-amino-4-benzylphthalazines, that act via antagonism of the Smoothened receptor. A variety of analogues were synthesized and their structure-activity relationships determined. This optimization resulted in the discovery of high affinity Smoothened antagonists, one of which was further profiled in vivo. This compound displayed a good pharmacokinetic profile and also afforded tumor regression in a genetic mouse model of medulloblastoma.

Synthesis and biological evaluation of amino-pyridines as androgen receptor antagonists for stimulating hair growth and reducing sebum production.

A series of amino-pyridines were synthesized and evaluated for androgen antagonist activities. Among these compounds, (R)-(+)-6-[methyl-(1-phenyl-ethyl)-amino]-4-trifluoromethyl-nicotinonitrile was the most active example of this class. This compound displayed potent androgen receptor antagonist activity as well as favorable pharmacokinetic characteristics for a potential topical agent. It also demonstrated remarkable potency for stimulating hair growth in a male C3H mouse model as well as reducing sebum production in the male Syrian hamster ear model.

(1R,2S)-4-(2-cyano-cyclohexyl-oxy)-2-trifluoromethyl-benzonitrile, a potent androgen receptor antagonist for stimulating hair growth and reducing sebum production.

Synthesis, pharmacology, and pharmacokinetic profiles of (1R, 2S)-4-(2-cyano-cyclohexyl-oxy)-2-trifluoromethyl-benzonitrile are reported. This compound demonstrated remarkable potency for stimulating hair growth in a male C3H mouse model as well as reducing sebum production in the male Syrian hamster ear model.

The design and synthesis of human branched-chain amino acid aminotransferase inhibitors for treatment of neurodegenerative diseases.

The inhibition of the cytosolic isoenzyme BCAT that is expressed specifically in neuronal tissue is likely to be useful for the treatment of neurodegenerative and other neurological disorders where glutamatergic mechanisms are implicated. Compound 2 exhibited an IC50 of 0.8 microM in the hBCATc assays; it is an active and selective inhibitor. Inhibitor 2 also blocked calcium influx into neuronal cells following inhibition of glutamate uptake, and demonstrated neuroprotective efficacy in vivo. SAR, pharmacology, and the crystal structure of hBCATc with inhibitor 2 are described.

Expression and refolding of recombinant human alpha-tocopherol transfer protein capable of specific alpha-tocopherol binding.

alpha-Tocopherol transfer protein (alpha-TTP) is a cytosolic protein found predominantly in mammalian liver that is proposed to be responsible for the stereoselective uptake of alpha-tocopherol from the diet. Although recombinant alpha-TTP has been reported previously, little detail has been provided about the yields and competency of the recovered protein at binding tocopherols and other ligands. In this work, we report the successful expression and refolding of a recombinant human alpha-TTP. Ligation-independent cloning generated a construct in pET-30 encoding an alpha-TTP fusion protein (pET-30/ttp) containing a six-histidine tag and an S-tag, each cleavable by a separate protease upon expression in Escherichia coli. Overexpression of the protein led to the formation of inclusion bodies that were solubilized in 8 M urea and purified by metal chelate affinity chromatography. Another construct in pET-28b (pET-28b/ttp) provided a soluble protein product after expression that contained a 40-amino-acid N-terminal extension, which can be reduced to 21 amino acids by cleavage with thrombin. The success of different refolding experiments was assessed using a Lipidex gel-based tocopherol binding assay. The best recovery of refolded recombinant alpha-TTP fusion capable of binding alpha-tocopherol was provided by matrix-assisted refolding in the presence of 0.5 M arginine. Cleavage of the fusion protein with Factor Xa successfully generated the full-length wild-type protein with no additional N-terminal amino acids. The resulting purification scheme provides recombinant alpha-TTP in good yield and purity for investigation of both its structure and its binding affinities for different ligands including natural and synthetic tocols.