Xiuhong Lu

Vinyl sulfoximines have emerged as promising electrophilic warheads for targeted covalent inhibitors owing to their tunable S(VI) framework, yet their broader application has been hindered by the lack of practical synthetic methods and systematic evaluation of their intrinsic electrophilicity. Herein, we report a mild one-pot oxidative synthesis of vinyl sulfoximines from readily accessible β-chlorosulfides, providing efficient access to structurally diverse vinyl sulfoximines. The protocol exhibits broad substrate scope and functional-group tolerance, affording the desired products in moderate to excellent yields. Systematic kinetic studies with amino acid derivatives revealed a pronounced preference for thiol nucleophiles and established quantitative structure–reactivity relationships for tuning the electrophilic reactivity of vinyl sulfoximines. By combining practical synthetic access with systematic warhead characterization, this work establishes vinyl sulfoximines as tunable electrophilic motifs for covalent inhibitor design.

Natural products (NPs) are a critical source for drug discovery, and artificial intelligence (AI) is utilized to improve the efficiency of NP-based drug discovery. However, the existing AI-driven models typically generate a library of pseudo-natural products that only covers a small portion of the chemical space and the compounds were also restricted by poor drug-likeness profiles. Herein, the GPT1 is developed to generate diverse pseudo-natural products with excellent validity, uniqueness, and novelty while retaining molecular features similar to the training set. Subsequently, the Augmented Hill-Climb (AHC) strategy is employed to generate synthetically accessible compounds with enhanced drug-likeness. Using the integrated NPDL-GEN model (GPT1 + AHC), compounds G1-G5 were obtained, exhibiting significantly improved drug-likeness profiles. Furthermore, the pseudo-natural products H1-H3 generated via transfer learning also possess potent anti-inflammatory activities. Thus, our developed machine learning models can accelerate NP-based drug discovery.

Tau PET tracers are being developed for imaging Alzheimer's disease (AD), primary tauopathies, and potentially screening of cognitively unimpaired elders. A second‐generation tau tracer PM‐PBB3, currently in Phase 3 clinical trials with FDA Fast Track Designation, shows promise as a broad‐spectrum tau imaging agent, but is limited by photoisomerization and binding to amyloid fibrils. Herein, the study reports the development of a better tau probe, BMP‐7, created by strategically introducing a methyl group at position 2 of the butadiene scaffold to enhance its chemical and biological properties. BMP‐7 exhibits remarkable photostability, showing no significant change in HPLC assays after 6 h of light exposure. Critically, BMP‐7 demonstrates increased sensitivity and 4.9‐fold greater selectivity than PM‐PBB3 for detecting tau pathology in brain sections from transgenic mouse models of AD and 4R tauopathies. Specifically, BMP‐7 binds to MC1‐reactive pathological tau conformations requiring both C‐ and N‐terminal phosphorylation, which is abolished by in vitro dephosphorylation. Furthermore, BMP‐7 readily penetrates the blood‐brain barrier and binds to tau pathology in vivo. These studies demonstrate that BMP‐7 bearing 2‐methyl‐butadiene scaffold improves photostability and significantly enhances the sensitivity and selectivity of tau pathology detection, offering substantial advantages for future applications. Therefore, BMP‐7 shows potential for further clinical development.

Cyclopropane fragments, which widely exist in marketed drugs and natural products, can confer special pharmacological properties to small‐molecule drugs. Therefore, developing methods to construct cyclopropanes is of great significance. Nevertheless, the introduction of cyclopropane primarily relies on already‐formed cyclopropyl groups, which significantly restricts the diversity of cyclopropane skeletons. Late‐stage direct cyclopropanation is still a challenging task. Herein, a photo‐induced intermolecular deoxycyclopropanation reaction that employs alcohols as substrates, and 1 mol.% of 2,3,5,6‐tetrakis(carbazol‐9‐yl)‐1,4‐dicyanobenzene (4CzTPN) as organophotocatalyst is reported. This method proceeds with high transformation efficiency (up to 98% yield) and exhibits broad functional group tolerance, such as primary, secondary, and tertiary alcohols as well as various activated β‐halogenated alkenes. This process is mild, easy to operate, and has low equipment requirements. The power of this technology is demonstrated by the late‐stage functionalization of five marketed drugs and five natural products.