Retraction Note: Crystal structures of agonist-bound human cannabinoid receptor CB1

Tian Hua, Kiran Vemuri, Spyros P. Nikas, Robert B. Laprairie, Yiran Wu, Lu Qu, Mengchen Pu, Anisha Korde, Shan Jiang, Jo‐Hao Ho, Gye Won Han, Kang Ding, Xuanxuan Li, Haiguang Liu, Michael A. Hanson, Suwen Zhao, Laura Bohn, Alexandros Makriyannis, Raymond C. Stevens, Zhi‐Jie Liu, Zhijie LiuNature 2025 · DOI · OpenAlex

Introduction

In 2017, a landmark study published in Nature provided high-resolution crystal structures of the human cannabinoid receptor $CB_1$ bound to agonists. This work was pivotal for understanding the molecular basis of $CB_1$ activation, a receptor central to the effects of cannabis and a major target for treating pain, obesity, and neurological disorders. Recently, the authors of this study issued a formal retraction of the original article (nature23272) due to errors identified in the functional signaling data. This retraction note serves to correct the scientific record while introducing a revised, peer-reviewed manuscript containing the corrected data.

The Discrepancy in Functional Data

The primary reason for the retraction involves "cell-based cAMP accumulation data." In structural biology, while X-ray crystallography provides the atomic coordinates of a protein, functional assays are required to demonstrate that the structural model accurately reflects the protein's behavior in a biological context.

The $CB_1$ receptor is a G protein-coupled receptor (GPCR) that typically inhibits the production of cyclic adenosine monophosphate (cAMP) when activated. The authors identified issues in the following data subsets from the 2017 publication:

  • Functional signaling plots (Figures 1d and 3c)
  • Supplemental signaling data (Extended Data Figure 4)
  • Quantitative metrics (Extended Data Table 2), including $EC_{50}$ values which measure the potency of receptor activation.

Re-execution and Validation

To address these discrepancies, the authors independently repeated the cAMP accumulation assays. The goal was to provide a rigorous "litmus test" to determine if the mutant versions of the receptor used in the structural studies functioned as originally reported.

Crucially, the re-execution of these experiments revealed that the core structural conclusions of the original paper—the 3D maps of the $CB_1$ receptor and the identification of the agonist binding pocket—remain valid. The errors were confined to the functional validation metrics rather than the crystallographic data itself. This represents a "clean" retraction where the authors identified a specific technical flaw, corrected it, and found that the new data still supported their primary structural interpretations.

Impact on the Scientific Record

The retraction and subsequent publication of a revised manuscript in 2025 (DOI: 10.1038/s41586-025-09454-5) highlight a commitment to scientific transparency. By withdrawing the flawed data, the authors ensure that future drug-discovery pipelines and molecular docking simulations are based on accurate signaling values.

While the relative potencies and the roles of specific amino acids in the binding pocket remain consistent with the original model, the updated $EC_{50}$ values are now the definitive reference for the community. This long-tail process of verification—spanning eight years from the original publication—underscores the self-correcting nature of high-impact scientific research.

Conclusion and Future Directions

The retraction of the 2017 $CB_1$ study is a significant event for the GPCR research community. Moving forward, researchers must:

  1. Update Citations: Ensure all future work references the 2025 revised manuscript rather than the retracted 2017 version.
  2. Refine Models: Use the corrected functional data to refine computational models and drug-screening algorithms.
  3. Verify Derivative Studies: Any research that relied specifically on the quantitative metrics from the original Extended Data Table 2 should be re-evaluated against the new dataset to confirm that derivative conclusions remain sound.

The structural integrity of the $CB_1$ agonist-bound model remains a cornerstone of cannabinoid research, now bolstered by a corrected and verified functional foundation.

Summary updated

BibTeX
@inproceedings{hua2025retraction,
  author = {Tian Hua and Kiran Vemuri and Spyros P. Nikas and Robert B. Laprairie and Yiran Wu and Lu Qu and Mengchen Pu and Anisha Korde and Shan Jiang and Jo‐Hao Ho and Gye Won Han and Kang Ding and Xuanxuan Li and Haiguang Liu and Michael A. Hanson and Suwen Zhao and Laura Bohn and Alexandros Makriyannis and Raymond C. Stevens and Zhi‐Jie Liu and Zhijie Liu},
  title = {{Retraction Note: Crystal structures of agonist-bound human cannabinoid receptor CB1}},
  booktitle = {Nature},
  year = {2025},
  doi = {10.1038/s41586-025-09538-2},
}
Figure