Scientists led by Wu and colleagues determined eight structural snapshots of UDP-glucose ceramide glucosyltransferase (UGCG), the enzyme that initiates production of nearly all glycosphingolipids in cells. The structures, published in Nature, expose a previously unknown catalytic mechanism and show how existing drugs bind and inhibit the enzyme. They also uncovered a regulatory 'brake' element unique to primate versions of UGCG.
Researchers designed a chemical strategy that links a targeting molecule's binding action directly to the release of an anticancer payload, rather than relying solely on enzymatic cleavage. The team synthesized and validated compounds targeting fibroblast activation protein (FAP) and other markers, testing reaction rates and specificity using UPLC-mass spectrometry and antibody-based assays on proteins including PD-L1, DPPIV, and cathepsin B.
Nearly a decade after Geoffrey Hinton predicted AI would replace radiologists, the profession has instead grown, with practitioner numbers projected to rise more than 26% over three decades. About three-quarters of the 1,400 AI-enabled medical devices cleared by the FDA target radiology, ranging from tools that draft reports and flag urgent cases to systems that can spot abnormalities humans miss, such as AI-assisted colonoscopies detecting more polyps than standard procedures.
Counterfeit versions of retatrutide, an experimental Eli Lilly weight-loss drug not yet submitted for FDA approval, are being sold at spas and on social media despite lacking any regulatory clearance. The drug targets three hormone receptors to suppress appetite and boost calorie burn, with early trial data suggesting significant weight loss, but no legitimate US provider is authorized to prescribe it. ABC Australia reported a case where a counterfeit vial of the drug caused severe liver damage.
Researchers using an automated laboratory in Zhangjiang, Shanghai encountered an unexpected experimental result that led to the identification of a new non-opioid pain treatment candidate. The discovery reportedly emerged from AI-assisted experimentation rather than traditional hypothesis-driven research, suggesting machine learning systems can surface biological insights humans might not have anticipated.