Only time will tell: Modelling the kinetics of covalent inhibitors.

The attrition rate in drug development is being reduced by continuous advances in science and technology introduced by various academic institutions and pharmaceutical companies. This has been certainly noticeable in reducing the frequency with which unfavorable absorption, distribution, metabolism, and elimination (ADME) characteristics of any candidate drug causes failure in clinical development. Nonetheless, it is important that the objectives in reducing attrition during later stages of development are matched by information generated in the earliest stage of discovery. In this review, we summarize the methodologies employed during the early stages of drug discovery and discuss new findings in the areas of (1) drug metabolism enzymes, (2) the contribution of cytochrome P450 enzymes (P450, CYP) to hepatic metabolism, (3) prediction of hepatic intrinsic clearance, (4) reaction phenotyping, and (5) the metabolic differences between highly homologous enzymes such as CYP3A4 and CYP3A5. The total contribution of P450 and UDP-glucuronosyltransferases to drug metabolism is reported to be more than 80%; therefore, glucuronidation is increasingly recognized as an important clearance pathway in addition to that of P450 enzymes. When estimating the contribution of P450, interpreting the results of inhibition studies using a single P450 inhibitor can lead to false conclusions. For instance, 1-aminobenzotriazole and SKF-525A have a varying range of IC(50) values for inhibition of drug exidation-reaction by different CYP450 enzymes. There are disparities between methodologies at early stage drug discovery and late stage development. For example, although the drug depletion approach for the prediction of hepatic intrinsic clearance may not be desirable at late stages of development, it is suitable at the early drug discovery stage since kinetic characterization and measurement of specific drug metabolites are not required. Data from protein binding assays in plasma and/or liver microsomes is an integral part to predicting hepatic clearance; therefore, the prediction methods for protein binding have been addressed in terms of automation and in silico prediction. The approach to reaction phenotyping using recombinant P450 microsome data are reviewed as this approach enables combining the drug depletion method with appropriate scaling factors to predict clearance values. CYP3A enzymes have broad substrate specificities and are responsible for the oxidative metabolism of more than 50% of clinically used drugs. Although CYP3A4 is the most abundant CYP3A isoform in adult human liver, CYP3A5 may contribute more to CYP3A-mediated drug oxidation by human liver microsomes than CYP3A4 does, especially in Japanese subjects, who typically have a relatively high frequency of genetic CYP3A5 expression. Lack of efficacy and presence of serious side effects in some sub-group of patients remain the biggest sources of drug failure at late stage of drug development. Advances in appreciation of inter-individual variabilities in ADME, by creation of virtual individuals and use of appropriate information from early discovery may lead to a better anticipation of variable clinical and toxicological outcome following administration of any new drug candidate. Thus may also help with dosing strategies which minimize the potential side effects and maximize the clinical benefits. Accordingly, front-loading of efforts for characterizing the candidate drugs at early stages of discovery is recommended.

Artificial intelligence(AI) is an effective tool toacceleratedrug discovery and cut costs in discovery processes. Many successfulAI applications are reported in the early stages of small moleculedrug discovery. However, most of those applications require a deepunderstanding of software and hardware, and focus on a single fieldthat implies data normalization and transfer between those applicationsis still a challenge for normal users. It usually limits the applicationof AI in drug discovery. Here, based on a series of robust models,we formed a one-stop, general purpose, and AI-based drug discoveryplatform, MolProphet, to provide complete functionalities in the earlystages of small molecule drug discovery, including AI-based targetpocket prediction, hit discovery and lead optimization, and compoundtargeting, as well as abundant analyzing tools to check the results.MolProphet is an accessible and user-friendly web-based platform thatis fully designed according to the practices in the drug discoveryindustry. The molecule screened, generated, or optimized by the MolProphetis purchasable and synthesizable at low cost but with good drug-likeness.More than 400 users from industry and academia have used MolProphetin their work. We hope this platform can provide a powerful solutionto assist each normal researcher in drug design and related researchareas. It is available for everyone at https://www.molprophet.com/.

An alternative biomimetic tool – Dual hydrophilic/reversed-phase interaction mode

Avoiding drug-drug interactions (DDIs) mediated through inhibition of cytochrome P450 (CYP) activity is highly desirable. Direct inhibition (DI) of CYP through new chemical entities (NCEs) or time-dependent inhibition (TDI) through reactive metabolites should be elucidated at an early stage of drug discovery research. In particular, TDI of CYP occurring through reactive metabolites may be irreversible and even sustained, causing far more serious DDIs for TDIs than for DIs. Furthermore, it is important to ascertain whether an NCE inhibits multiple CYP isoforms. Hence, using a cocktail-substrate approach that we previously established (in which the activity of 8 CYP isoforms is simultaneously evaluated in a single run), we evaluated the IC50 values of direct inhibitors and TDI parameters (kobs, shifted IC50, KI and kinact) of time-dependent inhibitors that affect multiple CYP isoforms. The IC50 values for 8 CYP isoforms obtained using the cocktail-substrate approach were nearly identical to values previously reported. The TDI parameters for CYP1A2, 2C9, 2C19, 2D6, and CYP3A4/5 obtained using the cocktail-substrate approach were also nearly identical to those obtained using a single-substrate approach. Thus, the cocktail-substrate approach is useful for evaluating DI and TDI in the early stages of drug discovery and development processes.

Multidimensional compound optimization is a new paradigm in the drug discovery process, yielding efficiencies during early stages and reducing attrition in the later stages of drug development. The success of this strategy relies heavily on understanding this multidimensional data and extracting useful information from it. This paper demonstrates how principled visualization algorithms can be used to understand and explore a large data set created in the early stages of drug discovery. The experiments presented are performed on a real-world data set comprising biological activity data and some whole-molecular physicochemical properties. Data visualization is a popular way of presenting complex data in a simpler form. We have applied powerful principled visualization methods, such as generative topographic mapping (GTM) and hierarchical GTM (HGTM), to help the domain experts (screening scientists, chemists, biologists, etc.) understand and draw meaningful decisions. We also benchmark these principled methods against relatively better known visualization approaches, principal component analysis (PCA), Sammon's mapping, and self-organizing maps (SOMs), to demonstrate their enhanced power to help the user visualize the large multidimensional data sets one has to deal with during the early stages of the drug discovery process. The results reported clearly show that the GTM and HGTM algorithms allow the user to cluster active compounds for different targets and understand them better than the benchmarks. An interactive software tool supporting these visualization algorithms was provided to the domain experts. The tool facilitates the domain experts by exploration of the projection obtained from the visualization algorithms providing facilities such as parallel coordinate plots, magnification factors, directional curvatures, and integration with industry standard software.

Although designed covalent inhibitors as drug candidates offer several unique advantages over conventional reversible inhibitors, including high potency and the potential for less frequent dosing, there is a general tendency to avoid the covalent mode of action in drug discovery programs due to concerns regarding immune-mediated toxicity that can arise from indiscriminate reactivity with off-target proteins. Therefore, the ability to assess off-target reactivity relative to target specificity is desirable for optimizing covalent drug candidates in the early discovery stage. One concern with current surrogate nucleophile trapping approaches is that they employ a simplistic model nucleophile such as glutathione, which may not reliably reflect the covalent interactions with cellular or extracellular proteins. One way to get a more relevant reactivity assessment is to directly measure the ability of an inhibitor to covalently modify nucelophilic amino acids on biologically relevant proteins, both on- and off-target. In this article, we describe a label-free bottom-up proteomic workflow for simultaneous evaluation of target binding and off-target reactivity of covalent drug candidates to selected proteins at the peptide level. Ibrutinib, a covalent drug targeting the active site of BTK protein, was used as a model compound to demonstrate the feasibility of the workflow. The compound was incubated with a mixture of target protein, Bruton's tyrosine kinase (BTK), and two abundant proteins in blood, hemoglobin (Hb) and human serum albumin (HSA), and then the ibrutinib modification sites were determined utilizing a bottom-up proteomic approach. A non-BTK specific model compound (1) known to modify cysteine residues was also included. By comparing the extent of off-target modifications to the targeted BTK C481 binding in a wide compound concentration range, we were able to determine the concentration where maximum target binding was achieved with minimal off-target reactivity. The generic label-free bottom-up proteomics workflow described in this article should be useful in the rank order assessment of off-target reactivity vs on-target reactivity of covalent drug candidates in the early drug discovery stage.

For drugs that directly act on targets in the central nervous system (CNS), sufficient drug delivery into the brain is a prerequisite for drug action. Systemically administered drugs can reach CNS by passage across the endothelium of capillary vasculatures, the so-called blood-brain barrier (BBB). Literature data suggest that most marketed CNS drugs have good membrane permeability and relatively high plasma unbound fraction, but are not good P-glycoprotein (P-gp) substrates. Therefore, it is important to use the in vitro parameters of P-gp function activity, membrane permeability and plasma unbound fraction as key criteria for lead optimization during the early stage of drug discovery. Evidence from preclinical and clinical studies suggests that drug concentration in cerebrospinal fluid (CSF) appears to be reasonably accurate in predicting unbound drug concentration in the brain. Therefore, CSF can be used as a useful surrogate for in vivo assessment of CNS exposure and provides an important basis for the selection of drug candidates for entry into development. However, it is important to point out that CSF drug concentration is not always an accurate surrogate for predicting unbound drug concentration in the brain. Depending on the physicochemical properties of drugs and the site/timing of CSF sampling, the unbound drug concentration at the biophase within the brain could differ significantly from the corresponding CSF drug concentration.

Small molecules, including natural compounds, in aqueous buffer that self-associate into colloidal aggregates is the main cause of false results in the early stage of drug discovery. Here we reported resonant waveguide grating (RWG) based assays to identify natural compound aggregation and characterize its influence on membrane receptors in living cells. We first applied a cell-free aggregation assay to determine compound critical aggregation concentration (CAC) values. Then we characterized the aggregators' influence on membrane receptors using three types of dynamic mass redistribution (DMR) assays. Results showed that colloidal aggregates may cause false activity in DMR desensitization assays; some of the false activities can be implied by the large response in DMR agonism assays and can further be identified by DMR antagonism assays. Furthermore, the aggregation mechanism was confirmed by addition of 0.025% tween-80, with cell signals attenuated and potency decreased. Finally, these observations were used for aggregate examination and promiscuity investigation of a traditional herbal medicine, Rhodiola rosea, which ultimately led to the revealing of the true target and reduced the risk of a bioactivity tracking process at the very first stage. This study highlights that the RWG based assays can be used as practical tools to distinguish between real and false hits to provide reliable results in the early stage of drug discovery.

It has been proposed that the increase in the instances of idiosyncratic adverse drug reactions (IADRs) and black box warnings may be attributed to the occurrence of reactive metabolites. Consequently, a high-throughput screen for reactive metabolites formed from liver microsome extracts with added glutathione (GSH) was developed for use in the early stages of drug discovery. To enhance sensitivity and specificity, as well as accelerate data processing, a mixture of a stable-isotope probe consisting of natural GSH (light GSH) and stable-isotope-labeled [(15) N,(13) C(2)] GSH (heavy GSH) at a ratio of 1:1 was used. Any metabolite that reacted with the GSH results in the formation of light and heavy GSH conjugates with a 3 Da difference. Employing a precursor-ion scan using negative ion electrospray ionization (ESI) corresponding to the expected fragments, signals with the appropriate ratio in the precursor ion scan are then further examined. The new method greatly simplifies data collection by assuming molecules containing GSH will fragment to form specific ions. As such, this approach accelerates data processing (and collection) at the risk of missing compounds that do not fragment as expected. The assay was validated with 33 diverse drugs known to form GSH conjugates, 5 drugs known to not form GSH adducts and over 100 samples containing components of the normal in vitro matrix. In all cases data collected matched the expected result. The observed sensitivity, specificity, and fast data processing make this assay an excellent fit for high-throughput screening of reactive metabolites in the early stages of drug discovery. This method is not intended to eliminate compounds or terminate their development. Instead, it is to bring forward molecules with one less liability and thus a greater probability of ultimate success.

Iminoheterocycle as a druggable motif: BACE1 inhibitors and beyond

pH-gradient PAMPA-based in vitro model assay for drug-induced phospholipidosis in early stage of drug discovery

Stability of compounds in the human plasma is crucial for maintaining sufficient systemic drug exposure and considered an essential factor in the early stages of drug discovery and development. The rapid degradation of compounds in the plasma can result in poor in vivo efficacy. Currently, there are no open-source software programs for predicting human plasma stability. In this study, we developed an attention-based graph neural network, PredPS to predict the plasma stability of compounds in human plasma using in-house and open-source datasets. The PredPS outperformed the two machine learning and two deep learning algorithms that were used for comparison indicating its stability-predicting efficiency. PredPS achieved an area under the receiver operating characteristic curve of 90.1%, accuracy of 83.5%, sensitivity of 82.3%, and specificity of 84.6% when evaluated using 5-fold cross-validation. In the early stages of drug discovery, PredPS could be a helpful method for predicting the human plasma stability of compounds. Saving time and money can be accomplished by adopting an in silico-based plasma stability prediction model at the high-throughput screening stage. The source code for PredPS is available at https://bitbucket.org/krict-ai/predps and the PredPS web server is available at https://predps.netlify.app.

High-throughput screening (HTS) is one of the most powerful approaches available for identifying new lead compounds for the growing catalogue of validated drug targets. However, just as virtual and experimental HTS have accelerated lead identification and changed drug discovery, they have also introduced a large number of peculiar molecules. Some of these have turned out to be interesting for further optimization, others to be dead ends when attempts are made to optimize their activity, typically after a great deal of time and resources have been devoted. Such false positive hits are still one of the key problems in the field of HTS and in the early stages of drug discovery in general. Many studies have been devoted to understanding the origins of false-positives, and the findings have been incorporated in filters and methods that can predict and eliminate problematic molecules from further consideration. This paper will focus on the structural classes and known mechanisms of nonleadlike false positives, together with experimental and computational methods for identifying such compounds.

High‐throughput screening for stability and inhibitory activity of compounds toward cytochrome P450‐mediated metabolism

In the early stages of drug discovery, adequate evaluation of the potential drug–drug interactions (DDIs) of drug candidates is important. Several CYP3A activators are known to lead to underestimation of DDIs. These compounds affect midazolam 1'-hydroxylation but not midazolam 4-hydroxylation. We used both metabolic reactions of midazolam to evaluate the activation and inhibition of CYP3A activators simultaneously. For our CYP inhibition assay using cocktail probe substrates, simultaneous liquid chromatography-tandem mass spectrometry monitoring of 1'-hydroxymidazolam and 4-hydroxymidazolam for CYP3A was established in addition to monitoring of 4-hydroxydiclofenac and 1'-hydroxybufuralol for CYP2C9 and CYP2D6. The results of our cocktail inhibition assay were well correlated with those of a single inhibition assay, as were the estimated inhibition parameters for typical CYP3A inhibitors. In our assay, a proprietary compound that activated midazolam 1'-hydroxylation and tended to inhibit 4-hydroxylation was evaluated along with known CYP3A activators. All compounds were well characterised by comparison of the results of midazolam 1'- and 4-hydroxylation. In conclusion, our CYP cocktail inhibition assay can detect CYP3A activation and assess the direct and time-dependent inhibition potentials for CYP3A, CYP2C9, and CYP2D6. This method is expected to be very efficient in the early stages of drug discovery.