Search

Abstract Generative deep learning is reshaping drug design. Chemical language models (CLMs) – which generate molecules in the form of molecular strings – bear particular promise for this endeavor. Here, we introduce a recent deep learning architecture, termed Structured State Space Sequence (S4) model, into de novo drug design. In addition to its unprecedented performance in various fields, S4 has shown remarkable capabilities to learn the global properties of sequences. This aspect is intriguing in chemical language modeling, where complex molecular properties like bioactivity can ‘emerge’ from separated portions in the molecular string. This observation gives rise to the following question: Can S4 advance chemical language modeling for de novo design? To provide an answer, we systematically benchmark S4 with state-of-the-art CLMs on an array of drug discovery tasks, such as the identification of bioactive compounds, and the design of drug-like molecules and natural products. S4 shows a superior capacity to learn complex molecular properties, while at the same time exploring diverse scaffolds. Finally, when applied prospectively to kinase inhibition, S4 designs eight of out ten molecules that are predicted as highly active by molecular dynamics simulations. Taken together, these findings advocate for the introduction of S4 into chemical language modeling – uncovering its untapped potential in the molecular sciences.

Abstract Background Advances in treatment and optimization of chemotherapy protocols have greatly improved survival in cancer patients. Unfortunately, treatment can cause a reduction in left ventricular (LV) ejection fraction (EF) leading to cancer therapy‐related cardiac dysfunction (CTRCD). We conducted a scoping review of published literature in order to identify and summarize the reported prevalence of cardiotoxicity evaluated by noninvasive imaging procedures in a wide‐ranging of patients referred to cancer treatment as chemotherapy and/or radiation therapy. Methods Different databases were checked (PubMed, Embase, and Web of Science) to identify studies published from January 2000 to June 2021. Articles were included if they reported data on LVEF evaluation in oncological patients treated with chemotherapeutic agents and/or radiotherapy, measured by echocardiography and/or nuclear or cardiac magnetic resonance imaging test, providing criteria of CTRCD evaluation such as the specific threshold for LVEF decrease. Results From 963 citations identified, 46 articles, comprising 6841 patients, met the criteria for the inclusion in the scoping review. The summary prevalence of CTRCD as assessed by imaging procedures in the studies reviewed was 17% (95% confidence interval, 14–20). Conclusions The results of our scoping review endorse the recommendations regarding imaging modalities to ensure identification of cardiotoxicity in patients undergoing cancer therapies. However, to improve patient management, more homogeneous CTRCD evaluation studies are required, reporting a detailed clinical assessment of the patient before, during and after treatment.

The cannabinoid receptor 1 (CB1R) plays a pivotal role in regulating various physiopathological processes, thus positioning itself as a promising and sought-after therapeutic target. However, the search for specific and effective CB1R ligands has been challenging, prompting the exploration of drug repurposing (DR) strategies. In this study, we present an innovative DR approach that combines computational screening and experimental validation to identify potential Food and Drug Administration (FDA)-approved compounds that can interact with the CB1R. Initially, a large-scale virtual screening was conducted using molecular docking simulations, where a library of FDA-approved drugs was screened against the CB1R’s three-dimensional structures. This in silico analysis allowed us to prioritize compounds based on their binding affinity through two different filters. Subsequently, the shortlisted compounds were subjected to in vitro assays using cellular and biochemical models to validate their interaction with the CB1R and determine their functional impact. Our results reveal FDA-approved compounds that exhibit promising interactions with the CB1R. These findings open up exciting opportunities for DR in various disorders where CB1R signaling is implicated. In conclusion, our integrated computational and experimental approach demonstrates the feasibility of DR for discovering CB1R modulators from existing FDA-approved compounds. By leveraging the wealth of existing pharmacological data, this strategy accelerates the identification of potential therapeutics while reducing development costs and timelines. The findings from this study hold the potential to advance novel treatments for a range of CB1R -associated diseases, presenting a significant step forward in drug discovery research.

Four NDM-1 mutants (L218T, L221T, L269H and L221T/Y229W) were generated in order to investigate the role of leucines positioned in L10 loop. A detailed kinetic analysis stated that these amino acid substitutions modified the hydrolytic profile of NDM-1 against some β-lactams. Significant reduction of kcat values of L218T and L221T for carbapenems, cefazolin, cefoxitin and cefepime was observed. The stability of the NDM-1 and its mutants was explored by thermofluor assay in real-time PCR. The determination of TmB and TmD demonstrated that NDM-1 and L218T were the most stable enzymes. Molecular dynamic studies were performed to justify the differences observed in the kinetic behavior of the mutants. In particular, L218T fluctuated more than NDM-1 in L10, whereas L221T would seem to cause a drift between residues 75 and 125. L221T/Y229W double mutant exhibited a decrease in the flexibility with respect to L221T, explaining enzyme activity improvement towards some β-lactams. Distances between Zn1-Zn2 and Zn1-OH- or Zn2-OH- remained unaffected in all systems analysed. Significant changes were found between Zn1/Zn2 and first sphere coordination residues.

Accumulated evidence suggests that enhancing the endocannabinoid (eCB) tone, in particular of anandamide (N-arachidonoylethanolamine, AEA), has therapeutic potential in many human diseases. Fatty acid amide hydrolase (FAAH) is a membrane-bound enzyme principally responsible for the degradation of AEA, and thus it represents a relevant target to increase signaling thereof. In recent years, different synthetic and natural compounds have been developed and tested on rat FAAH, but little is known of their effect on the human enzyme. Here, we sought to investigate six major cannabis-derived compounds to compare their action on rat and human FAAHs. To this aim, we combined an in silico analysis of their binding mode and affinity, with in vitro assays of their effect on enzyme activity. This integrated approach allowed to disclose differences in efficacy towards rat and human FAAHs, and to highlight the role of key residues involved in the inhibition of both enzymes. This study suggests that the therapeutic efficacy of compounds targeted towards FAAH should be always tested in vitro on both rat and human enzymes.