Foldit Players Help Model a Spider Toxin for Pain Therapy

Started by beta_helix

beta_helix Staff Lv 1

Voltage-gated sodium channels (NaVs) are essential proteins for bioelectrical signaling playing a crucial role in many physiological processes and one subtype, NaV1.7, has emerged as a particularly attractive target to develop novel non-opioid pain therapies. However, its close structural similarity to other NaV subtypes has made selective targeting NaV1.7 without affecting other subtypes remarkably difficult. Peptide toxins from animal venoms, like the tarantula-derived Protoxin-2 (PTx2), offer a powerful scaffold for designing selective inhibitors, especially when paired with modern structure-based design methods.

Our lab at UC Davis previously engineered a variant of this toxin, PTx2-3127, with improved selectivity for NaV1.7. To guide further design, we collaborated with cryo-EM experts to visualize the toxin bound to the full channel. Unfortunately, the resulting experimental EM map lacked sufficient resolution near the toxin’s binding site in the channel (VSD2), making it impossible to build a confident atomic model using standard approaches.

To overcome this, we turned to the Foldit community. In a pair of puzzles, players were asked to fit the toxin variant into the experimental EM density while preserving its native disulfide bonds, a challenge requiring creativity and precision. The first round produced promising models but struggled with maintaining disulfide integrity; a second round, which included constraints for the disulfide bonds, led to high quality models that maintained the structural integrity of the peptide toxin variant. The final top model was compelling: all disulfide bonds were intact, and the interface between the peptide and the VSD2 was well-resolved and biophysically plausible. This model provides critical atomic-level insight into how PTx2-3127 modulates the channel and offers a solid foundation for our future design efforts.

With continued collaboration between scientists and citizen designers, we move closer to the long-sought goal of designing potent and non-addictive pain therapeutics!

beta_helix Staff Lv 1

The design used by the UC Davis researchers was by Galaxie and gmn :-)

Based on our observations of Foldit solutions for cryo-EM puzzles, the "best" structures from Foldit players were also the best according to the Foldit score function, which has not been the case with all other Foldit puzzle types. We reported this observation in the Foldit cryo-EM paper: https://doi.org/10.1371/journal.pbio.3000472

Because of this, we started by giving our collaborators at UC Davis the top-scoring Foldit solution at first… and planned to send them more, but they were already satisfied with its accuracy!

You will hopefully be able to ask the UC Davis researchers questions about this puzzle series at our next Foldit Office Hour (we're aiming for the end of this month/early June).