The glucocorticoid receptor (GR) is phosphorylated at multiple sites within its N terminus (S203, S211, S226), yet the role of phosphorylation in receptor function is not understood. Using a range of agonists and GR phosphorylation site-specific antibodies, we demonstrated that GR transcriptional activation is greatest when the relative phosphorylation of S211 exceeds that of S226. Consistent with this finding, a replacement of S226 with an alanine enhances GR transcriptional response. Using a battery of compounds that perturb different signaling pathways, we found that BAPTA-AM, a chelator of intracellular divalent cations, and curcumin, a natural product with antiinflammatory properties, reduced hormone-dependent phosphorylation at S211. This change in GR phosphorylation was associated with its decreased nuclear retention and transcriptional activation. Molecular modeling suggests that GR S211 phosphorylation promotes a conformational change, which exposes a novel surface potentially facilitating cofactor interaction. Indeed, S211 phosphorylation enhances GR interaction with MED14 (vitamin D receptor interacting protein 150). Interestingly, in U2OS cells expressing a nonphosphorylated GR mutant S211A, the expression of IGF-binding protein 1 and interferon regulatory factor 8, both MED14-dependent GR target genes, was reduced relative to cells expressing wild-type receptor across a broad range of hormone concentrations. In contrast, the induction of glucocorticoid-induced leucine zipper, a MED14-independent GR target, was similar in S211A- and wild-type GR-expressing cells at high hormone levels, but was reduced in S211A cells at low hormone concentrations, suggesting a link between GR phosphorylation, MED14 involvement, and receptor occupancy. Phosphorylation also affected the magnitude of repression by GR in a gene-selective manner. Thus, GR phosphorylation at S211 and S226 determines GR transcriptional response by modifying cofactor interaction. Furthermore, the effect of GR S211 phosphorylation is gene specific and, in some cases, dependent upon the amount of activated receptor.

Machine learning (ML) models to predict the toxicity of small molecules have garnered great attention and have become widely used in recent years. Computational toxicity prediction is particularly advantageous in the early stages of drug discovery in order to filter out molecules with high probability of failing in clinical trials. This has been helped by the increase in the number of large toxicology databases available. However, being an area of recent application, a greater understanding of the scope and applicability of ML methods is still necessary. There are various kinds of toxic end points that have been predicted in silico. Acute oral toxicity, hepatotoxicity, cardiotoxicity, mutagenicity, and the 12 Tox21 data end points are among the most commonly investigated. Machine learning methods exhibit different performances on different data sets due to dissimilar complexity, class distributions, or chemical space covered, which makes it hard to compare the performance of algorithms over different toxic end points. The general pipeline to predict toxicity using ML has already been analyzed in various reviews. In this contribution, we focus on the recent progress in the area and the outstanding challenges, making a detailed description of the state-of-the-art models implemented for each toxic end point. The type of molecular representation, the algorithm, and the evaluation metric used in each research work are explained and analyzed. A detailed description of end points that are usually predicted, their clinical relevance, the available databases, and the challenges they bring to the field are also highlighted.

Ligand-docking-based methods are starting to play a critical role in lead discovery and optimization, thus resulting in new ‘drug-candidates’. They offer the possibility to go beyond the pool of existing active compounds, and thus find novel chemotypes. A brief tutorial on ligand docking and structure-based virtual screening is presented highlighting current problems and limitations, together with the most recent methodological and algorithmic developments in the field. Recent successful applications of docking-based tools for hit discovery, lead optimization and target-biased library design are also presented. Special consideration is devoted to ongoing efforts to account for protein flexibility in structure-based virtual screening.