ChemMedChem

Constructive (generative) machine learning enables the automated generation of novel chemical structures without the need for explicit molecular design rules. This study presents the experimental application of such a deep machine learning model to design membranolytic anticancer peptides (ACPs) de novo. A recurrent neural network with long short‐term memory cells was trained on α‐helical cationic amphipathic peptide sequences and then fine‐tuned with 26 known ACPs by transfer learning. This optimized model was used to generate unique and novel amino acid sequences. Twelve of the peptides were synthesized and tested for their activity on MCF7 human breast adenocarcinoma cells and selectivity against human erythrocytes. Ten of these peptides were active against cancer cells. Six of the active peptides killed MCF7 cancer cells without affecting human erythrocytes with at least threefold selectivity. These results advocate constructive machine learning for the automated design of peptides with desired biological activities.

A synthetic transcriptional activator encompassing both sequence‐specific pyrrole–imidazole polyamides (PIPs) and an epigenetic activator (suberoylanilide hydroxamic acid) was recently shown to induce the endogenous expression of core pluripotency genes in mouse embryonic fibroblasts (MEFs). Microarray data analysis suggested Oct‐3/4 as the probable target pathway of the activator. However, the expression levels in MEFs treated with the activator were relatively lower than those in mouse embryonic stem cells. Herein, we report studies carried out to improve the efficacy of the activator and show that the biological activity was significantly (p<0.05) improved against the core pluripotency genes after the incorporation of an isophthalic acid (IPA) at the C terminus. The resultant IPA conjugate dramatically induced Oct‐3/4 and demonstrated a new chemical strategy for developing PIP conjugates as next‐generation genetic switches.

Acute lung injury (ALI) has a high lethality rate, and interleukin‐6 (IL‐6) and tumor necrosis factor‐α (TNF‐α) contribute most to tissue deterioration in cases of ALI. In this study, we designed and synthesized a new series of thiazolo[3,2‐a]pyrimidine derivatives based on a previously identified lead compound, and we evaluated their anti‐inflammatory activities. Structure–activity relationship studies led to the discovery of two highly potent inhibitors. The two promising compounds were found to inhibit lipopolysaccharide (LPS)‐induced IL‐6 and TNF‐α release in a dose‐dependent manner in mouse primary peritoneal macrophages (MPMs). Furthermore, administration of these compounds resulted in lung histopathological improvements and attenuated LPS‐induced ALI in vivo. Taken together, these data indicate that these novel thiazolo[3,2‐a]pyrimidine derivatives could be developed as candidate drugs for the treatment of ALI.

We synthesized potential inhibitors of farnesyl diphosphate synthase (FPPS), undecaprenyl diphosphate synthase (UPPS), or undecaprenyl diphosphate phosphatase (UPPP), and tested them in bacterial cell growth and enzyme inhibition assays. The most active compounds were found to be bisphosphonates with electron‐withdrawing aryl‐alkyl side chains which inhibited the growth of Gram‐negative bacteria (Acinetobacter baumannii, Klebsiella pneumoniae, Escherichia coli, and Pseudomonas aeruginosa) at ∼1–4 μg mL⁻¹ levels. They were found to be potent inhibitors of FPPS; cell growth was partially “rescued” by the addition of farnesol or overexpression of FPPS, and there was synergistic activity with known isoprenoid biosynthesis pathway inhibitors. Lipophilic hydroxyalkyl phosphonic acids inhibited UPPS and UPPP at micromolar levels; they were active (∼2–6 μg mL⁻¹) against Gram‐positive but not Gram‐negative organisms, and again exhibited synergistic activity with cell wall biosynthesis inhibitors, but only indifferent effects with other inhibitors. The results are of interest because they describe novel inhibitors of FPPS, UPPS, and UPPP with cell growth inhibitory activities as low as ∼1–2 μg mL⁻¹.

Cancer is one of the leading causes of human mortality globally; therefore, intensive efforts have been made to seek new active drugs with improved anticancer efficacy. Indazole‐containing derivatives are endowed with a broad range of biological properties, including anti‐inflammatory, antimicrobial, anti‐HIV, antihypertensive, and anticancer activities. In recent years, the development of anticancer drugs has given rise to a wide range of indazole derivatives, some of which exhibit outstanding activity against various tumor types. The aim of this review is to outline recent developments concerning the anticancer activity of indazole derivatives, as well as to summarize the design strategies and structure–activity relationships of these compounds.

Bài tổng quan này mô tả một số phương pháp và kỹ thuật hiện đang được sử dụng để đưa ra các mô hình in silico nhằm dự đoán các thuộc tính ADMET. Bài báo cũng thảo luận một số yêu cầu cơ bản đối với việc tạo ra các mối quan hệ ADMET có tính toán học có cơ sở thống kê và dự đoán, cũng như một số cạm bẫy và vấn đề đã gặp phải trong các nghiên cứu này. Ý định của các tác giả là giúp người đọc nhận thức rõ hơn về một số thách thức liên quan đến việc phát triển các mô hình in silico ADMET có ích cho quá trình phát triển thuốc.

We are reporting a short and convenient pathway for the synthesis of novel β‐carboline‐bisindole hybrid compounds from relatively cheap and commercially available chemicals such as tryptamine, dialdehydes and indoles. These newly designed compounds can also be prepared in high yields with the tolerance of many functional groups under mild conditions. Notably, these β‐carboline‐bisindole hybrid compounds exhibited some promising applications as anticancer agents against the three common cancer cell lines MCF‐7 (breast cancer), SK‐LU‐1 (lung cancer), and HepG2 (liver cancer). The two best compounds 5b and 5g inhibited the aforementioned cell lines with the same range of the reference Ellipticine at less than 2 µM. A molecular docking study to gain more information about the interactions between the synthesized molecules and the kinase domain of the EGFR was performed. Therefore, this finding can have significant impacts on the development of future research in medicinal chemistry and drug discovery.

The development of robust and predictive QSAR models is highly dependent on the use of molecular descriptors that contain information relevant to the property being modelled. Selection of these relevant features from a large pool of possibilities is difficult to achieve effectively. Modern Bayesian methods provide substantial advantages over conventional feature selection methods for feature selection and QSAR modelling. We illustrate the importance of descriptor choice and the beneficial properties of Bayesian methods to select context‐dependent relevant descriptors and build robust QSAR models, using data on anaesthetics. Our results show the effectiveness of Bayesian feature selection methods in choosing the best descriptors when these are mixed with less informative descriptors. They also demonstrate the efficacy of the Abraham descriptors and identify deficiencies in ParaSurf descriptors for modelling anaesthetic action.

The glucocorticoid receptor (GR) is a member of the nuclear receptor superfamily that affects immune response, development, and metabolism in target tissues. Glucocorticoids are widely used to treat diverse pathophysiological conditions, but their clinical applicability is limited by side effects. A prediction of the binding affinity toward the GR would be beneficial for identifying glucocorticoid‐mediated adverse effects triggered by drugs or chemicals. By identifying the binding mode to the GR using flexible docking (software Yeti) and quantifying the binding affinity through multidimensional QSAR (software Quasar), we validated a model family based on 110 compounds, representing four different chemical classes. The correlation with the experimental data (cross‐validated r²=0.702; predictive r²=0.719) suggests that our approach is suited for predicting the binding affinity of related compounds toward the GR. After challenging the model by a series of scramble tests, a consensus approach (software Raptor), and a prediction set, it was incorporated into our VirtualToxLab and used to simulate and quantify the interaction of 24 psychotropic drugs with the GR.

Curcumin binds to the amyloid β peptide (Aβ) and inhibits or modulates amyloid precursor protein (APP) metabolism. Therefore, curcumin‐derived isoxazoles and pyrazoles were synthesized to minimize the metal chelation properties of curcumin. The decreased rotational freedom and absence of stereoisomers was predicted to enhance affinity toward Aβ₄₂ aggregates. Accordingly, replacement of the 1,3‐dicarbonyl moiety with isosteric heterocycles turned curcumin analogue isoxazoles and pyrazoles into potent ligands of fibrillar Aβ₄₂ aggregates. Additionally, several compounds are potent inhibitors of tau protein aggregation and depolymerized tau protein aggregates at low micromolar concentrations.