Deep learning driven de novo drug design based on gastric proton pump structures

Large language models open new way of AI-assisted molecule design for chemists

Recent advancements in artificial intelligence (AI)-based molecular design methodologies have offered synthetic chemists new ways to design functional molecules with their desired properties. While various AI-based molecule generators have significantly advanced toward practical applications, their effective use still requires specialized knowledge and skills concerning AI techniques. Here, we develop a large language model (LLM)-powered chatbot, ChatChemTS, that assists users in designing new molecules using an AI-based molecule generator through only chat interactions, including automated construction of reward functions for the specified properties. Our study showcases the utility of ChatChemTS through de novo design cases involving chromophores and anticancer drugs (epidermal growth factor receptor inhibitors), exemplifying single- and multiobjective molecule optimization scenarios, respectively. ChatChemTS is provided as an open-source package on GitHub at https://github.com/molecule-generator-collection/ChatChemTS . Scientific contribution ChatChemTS is an open-source application that assists users in utilizing an AI-based molecule generator, ChemTSv2, solely through chat interactions. This study demonstrates that LLMs possess the potential to utilize advanced software, such as AI-based molecular generators, which require specialized knowledge and technical skills.

Molecular Structure of the Na+,K+-ATPase α4β1 Isoform in Its Ouabain-Bound Conformation

Na+,K+-ATPase is the active ion transport system that maintains the electrochemical gradients for Na+ and K+ across the plasma membrane of most animal cells. Na+,K+-ATPase is constituted by the association of two major subunits, a catalytic α and a glycosylated β subunit, both of which exist as different isoforms (in mammals known as α1, α2, α3, α4, β1, β2 and β3). Na+,K+-ATPase α and β isoforms assemble in different combinations to produce various isozymes with tissue specific expression and distinct biochemical properties. Na+,K+-ATPase α4β1 is only found in male germ cells of the testis and is mainly expressed in the sperm flagellum, where it plays a critical role in sperm motility and male fertility. Here, we report the molecular structure of Na+,K+-ATPase α4β1 at 2.37 Å resolution in the ouabain-bound state and in the presence of beryllium fluoride. Overall, Na+,K+-ATPase α4 structure exhibits the basic major domains of a P-Type ATPase, resembling Na+,K+-ATPase α1, but has differences specific to its distinct sequence. Dissimilarities include the site where the inhibitor ouabain binds. Molecular simulations indicate that glycosphingolipids can bind to a putative glycosphingolipid binding site, which could potentially modulate Na+,K+-ATPase α4 activity. This is the first experimental evidence for the structure of Na+,K+-ATPase α4β1. These data provide a template that will aid in better understanding the function Na+,K+-ATPase α4β1 and will be important for the design and development of compounds that can modulate Na+,K+-ATPase α4 activity for the purpose of improving male fertility or to achieve male contraception.

How ligands modulate the gastric H,K-ATPase activity and its inhibition by tegoprazan

The introduction of potassium-competitive acid blockers (P-CABs) has been a major innovation in gastric H,K-ATPase inhibition and many laboratories are actively engaged in the development of novel molecules within this class. This work investigates the interaction between H,K-ATPase and tegoprazan, a representative of the P-CABs group, in terms of K+ and H+ binding, through functional and structural analyses. First, by studying the H,K-ATPase activity, we found a model to describe the non-Michaelis-Menten kinetics through a "ping-pong" mechanism that explains a stoichiometry of 1 H+, 1 K+, and 1 ATP molecule, but also considering the influence of H+ on the ionization states of the protein. A kinetic evaluation of the inhibition of tegoprazan denotes the binding to two different intermediates states with apparent Kd (μM) 0.56 ± 0.04 and 2.70 ± 0.24 at pH 7.2. Molecular dynamics simulations revealed important changes in the interactions of tegoprazan with the transmembrane residues depending on whether the site contains K+ or not. This explains the decrease in affinity as a function of K+ concentration observed in the kinetic experiments. On the other hand, the structures predict that the protonation of tegoprazan is responsible for the change in its dihedral angle. The rotation of the benzimidazole ring allows the inhibitor to be positioned further into the luminal cavity, a situation compatible with the higher inhibition affinity of H,K-ATPase measured at low pH. Results presented herein will provide a basis for the rational design of novel P-CABs ligands.

Human F-ATP synthase as a drug target

Practical and conceptual barriers have kept human F-ATP synthase out of reach as a target for the treatment of human diseases. Although this situation has persisted for decades, it may change in the near future. In this review the principal functionalities of human F-ATP synthase--proton motive force / ATP interconversion, membrane bending and mitochondrial permeability transition--are surveyed in the context of their respective potential for pharmaceutical intervention. Further, the technical requirements necessary to allow drug designs that are effective at the multiple levels of functionality and modality of human F-ATP synthase are discussed. The structure-based development of gastric proton pump inhibitors is used to exemplify what might be feasible for human F-ATP synthase. And finally, four structural regions of the human F-ATP synthase are examined as potential sites for the development of structure based drug development.

Clinical Implications of Proton Pump Inhibitors and Vonoprazan Micro/Nano Drug Delivery Systems for Gastric Acid-Related Disorders and Imaging

Excessive stomach acid or bacterial infection are the root causes of gastric acid-related disorders, such as peptic ulcer disease and gastroesophageal reflux disease. Proton pump inhibitors including lansoprazole, omeprazole, esomeprazole, rabeprazole, etc. are medications used to treat gastric acid-related diseases. One of the most effective drugs for treating gastroesophageal reflux disease is vonoprazan, owing to its ability to strongly inhibit gastric acid. Proton pump inhibitors and vonoprazan work in distinct ways to prevent the production of stomach acid. Vonoprazan inhibits acid secretion by blocking the potassium-competitive acid blocker receptor, whereas proton pump inhibitors function by irreversibly blocking the proton pump in the parietal cells of the stomach. Delayed release tablets, delayed release capsules, minitablets, pellets, bilayer, floating, mucoadhesive tablets and nanoparticles, are some of the methods used in the development of micro/nano formulations with proton pump inhibitors and vonoprazan. Diagnosis and therapy of gastric acid-related illnesses, particularly those treated with drugs such as vonoprazan and proton pump inhibitors, rely heavily on imaging modalities such as CT scans, X-rays, endoscopy, fluorescence and HRM imaging. This review provides a comprehensive update on various micro/nanoformulations of proton pump inhibitors and vonoprazan. Moreover, we provide an outlook on clinical imaging of proton pump inhibitors and vonoprazan formulation for gastric acid related diseases. We have limited our discussion to case studies and clinical trials on proton pump inhibitors and vonoprazan for gastric acid related disease.

Understanding the Landscape: The Emergence of Artificial Intelligence (AI), ChatGPT, and Google Bard in Gastroenterology

Introduction Artificial intelligence (AI) integration in healthcare, specifically in gastroenterology, has opened new avenues for enhanced patient care and medical decision-making. This study aims to assess the reliability and accuracy of two prominent AI tools, ChatGPT 4.0 and Google Bard, in answering gastroenterology-related queries, thereby evaluating their potential utility in medical settings. Methods The study employed a structured approach where typical gastroenterology questions were input into ChatGPT 4.0 and Google Bard. Independent reviewers evaluated responses using a Likert scale and cross-referenced them with guidelines from authoritative gastroenterology bodies. Statistical analysis, including the Mann-Whitney U test, was conducted to assess the significance of differences in ratings. Results ChatGPT 4.0 demonstrated higher reliability and accuracy in its responses than Google Bard, as indicated by higher mean ratings and statistically significant p-values in hypothesis testing. However, limitations in the data structure, such as the inability to conduct detailed correlation analysis, were noted. Conclusion The study concludes that ChatGPT 4.0 outperforms Google Bard in providing reliable and accurate responses to gastroenterology-related queries. This finding underscores the potential of AI tools like ChatGPT in enhancing healthcare delivery. However, the study also highlights the need for a broader and more diverse assessment of AI capabilities in healthcare to leverage their potential in clinical practice fully.