{
  "id": 1427611,
  "title": "Discovery of Selective Small-Molecule Ligands of SV2C by AI-Enhanced Virtual Screening and Experimental Validation",
  "url": "https://urgent.news/2026/08/16/discovery-of-selective-small-molecule-ligands-of-sv2c-by-ai-enhanced",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-16T00:00:00.000Z",
  "source": {
    "name": "bioRxiv",
    "slug": "biorxiv",
    "url": "https://www.biorxiv.org/content/10.64898/2026.08.11.744237v1?rss=1"
  },
  "original_language": "en",
  "account": "Synaptic vesicle glycoprotein 2C (SV2C) is a protein enriched in dopamine-producing neurons of the basal ganglia, playing a role in dopamine storage and release. Disruptions in SV2C are linked to Parkinson's disease (PD), but there are no selective small-molecule probes for SV2C. Researchers employed an AI-enhanced virtual screening (VS) approach coupled with experimental validation to uncover new compounds that selectively interact with SV2C, avoiding interference with SV2A and SV2B.\n\nThe research team used a large, general-purpose commercial library for the VS process. Due to the lack of a full-length, high-resolution SV2C structure, they constructed a homology model from SV2A structures obtained through cryo-electron microscopy (cryo-EM). This model helped analyze SV2C's conformational landscape through molecular dynamics (MD) and Gaussian accelerated MD (GaMD) simulations, both in its apo form and in complex with known SV2 ligands.\n\nA convolutional neural network-based scoring function (CNN_VS), validated using a manually curated benchmark of 39 SV2A ligands (r = 0.72 against experimental pIC50 values), was then applied to filter 5.96 million compounds from the library. These were sequentially narrowed down to 3.19 million, with a focus on those relevant to the CNS. Docking and rescoring narrowed the candidates to 94, from which 71 compounds were experimentally tested using a combination of thermal shift assays (TSA) and a [3H]-padsevonil scintillation proximity assay (SPA). The assays measured Ki values and assessed isoform selectivity against SV2A and SV2B.\n\nThe experiments identified 22 active molecules (31% hit rate), which were grouped into two categories based on their interaction profiles. Some compounds showed primary site competition with SV2C, while others did not, indicating varying levels of selectivity. Notably, compounds 36 and 56 demonstrated strong selectivity for SV2C with Ki values of 24.6 uM and 3.25 uM, respectively, and displayed a greater than 10-fold selectivity over SV2A and approximately 12-fold selectivity relative to SV2B.\n\nAdditionally, a subset of the SV2C-selective compounds acted as padsevonil-site competitors, offering a lead set for further functional characterization and potential drug development targeting dopaminergic signaling disorders. The docking analysis suggested a common binding mode involving conserved tryptophan residues in the SV2 pocket, confirmed by a comparison with an unpublished SV2A-plosaracetam cryo-EM structure. Small structural differences in the luminal domain and transmembrane region were identified as key determinants of isoform selectivity.\n\nThis study demonstrates that an AI-driven VS pipeline, integrated with medium-throughput biophysical assays, can successfully identify selective SV2C binders from a general chemical library, even when dealing with a structurally under-characterized membrane target. The promising hits provide valuable starting points for further lead optimization and research into SV2C's role in PD.",
  "summary": "Synaptic vesicle glycoprotein 2C (SV2C) is a vesicular protein enriched in dopaminergic neurons of the basal ganglia that modulates dopamine storage and release, and its disruption is implicated in Parkinson's disease (PD). Despite strong genetic and pathological links to PD, there are no selective small-molecule probes for SV2C. Here, we describe an AI-enhanced virtual screening (VS) and…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}