Anna Durner

The ATP-gated P2X7 receptor (P2X7R) activates inflammatory signaling pathways in the central nervous system. In particular, P2X7Rs drive chronic glia-mediated neuroinflammation, which is increasingly recognized as a key contributor to Alzheimer’s disease, a neurodegenerative disorder that lacks effective disease-modifying therapies. Here we identify a potent and selective negative allosteric modulator of P2X7Rs with therapeutic potential. We synthesize a series of small molecules based on a polycyclic scaffold and confirm blood-brain barrier penetration by testing a radiolabeled analogue using positron emission tomography imaging. Through a structure-guided medicinal chemistry campaign centered on our scaffold, we identify four promising P2X7R antagonists. Of these, UB-ALT-P2 exhibits the most favorable safety profile, high oral bioavailability and robust brain penetration. High-resolution cryo-EM structures of UB-ALT-P2 bound to human, mouse, and rat P2X7Rs reveal a conserved antagonist binding mode with steric features that favor prolonged binding to human receptors. In the 5xFAD mouse model of AD, oral UB-ALT-P2 blunts weight loss, improves short- and long-term memory, reduces amyloid-β plaque burden, lowers hyperphosphorylated tau, and diminishes oxidative and inflammatory markers. These results establish UB-ALT-P2 as a potent and safe P2X7R antagonist that can mitigate core AD pathologies, providing a compelling foundation for further development.

PURPOSE: Rapidly expanding medical literature challenges oncologists seeking targeted cancer therapies. General-purpose large language models (LLMs) lack domain-specific knowledge, limiting their clinical utility. This study introduces the LLM system Medical Evidence Retrieval and Data Integration for Tailored Healthcare (MEREDITH), designed to support treatment recommendations in precision oncology. Built on Google's Gemini Pro LLM, MEREDITH uses retrieval-augmented generation and chain of thought . METHODS: We evaluated MEREDITH on 10 publicly available fictional oncology cases with iterative feedback from a molecular tumor board (MTB) at a major German cancer center. Initially limited to PubMed -indexed literature (draft system), MEREDITH was enhanced to incorporate clinical studies on drug response within the specific tumor type, trial databases, drug approval status, and oncologic guidelines. The MTB provided a benchmark with manually curated treatment recommendations and assessed the clinical relevance of LLM-generated options (qualitative assessment). We measured semantic cosine similarity between LLM suggestions and clinician responses (quantitative assessment). RESULTS: MEREDITH identified a broader range of treatment options (median 4) compared with MTB experts (median 2). These options included therapies on the basis of preclinical data and combination treatments, expanding the treatment possibilities for consideration by the MTB. This broader approach was achieved by incorporating a curated medical data set that contextualized molecular targetability. Mirroring the approach MTB experts use to evaluate MTB cases improved the LLM's ability to generate relevant suggestions. This is supported by high concordance between LLM suggestions and expert recommendations (94.7% for the enhanced system) and a significant increase in semantic similarity from the draft to the enhanced system (from 0.71 to 0.76, P = .01). CONCLUSION: Expert feedback and domain-specific data augment LLM performance. Future research should investigate responsible LLM integration into real-world clinical workflows.

The large intracellular C-terminus of the pro-inflammatory P2X7 ion channel receptor (P2X7R) is associated with diverse P2X7R-specific functions. Cryo-EM structures of the closed and ATP-bound open full-length P2X7R recently identified a membrane-associated anchoring domain, an open-state stabilizing “cap” domain, and a globular “ballast domain” containing GTP/GDP and dinuclear Zn²⁺-binding sites with unknown functions. To investigate protein dynamics during channel activation, we improved incorporation of the environment-sensitive fluorescent unnatural amino acid L-3-(6-acetylnaphthalen-2-ylamino)–2-aminopropanoic acid (ANAP) into Xenopus laevis oocyte-expressed P2X7Rs and performed voltage clamp fluorometry. While we confirmed predicted conformational changes within the extracellular and the transmembrane domains, only 3 out of 41 mutants containing ANAP in the C-terminal domain resulted in ATP-induced fluorescence changes. We conclude that the ballast domain functions rather independently from the extracellular ATP binding domain and might require activation by additional ligands and/or protein interactions. Novel tools to study these are presented.