communityfix.org

Institute for Protein Design, University of Washington, Seattle, USA (with DTU, Denmark and LSTM, UK)

#00139

OngoingGlobal

Case study of

#00139 Computationally design de novo toxin-neutralizing proteins with deep learning

Implementer

Institute for Protein Design / Baker Lab (University of Washington, HHMI); Technical University of Denmark (Laustsen, Jenkins); Liverpool School of Tropical Medicine (Casewell); University of Northern Colorado (Mackessy)

Location

Institute for Protein Design, University of Washington, Seattle, USA (with DTU, Denmark and LSTM, UK)47.6553, -122.3035

Description

Using the deep-learning method RFdiffusion (with ProteinMPNN sequence design and AlphaFold2 filtering), researchers de novo designed small (~100-amino-acid) proteins to bind three-finger toxins from elapid venom: short-chain α-neurotoxins (design "SHRT"), long-chain α-neurotoxins/α-cobratoxin ("LNG"), and cytotoxins ("CYTX"). Crystal structures closely matched the computational models. The neurotoxin binders were sub-nanomolar and highly thermostable and fully neutralized their targets in patch-clamp assays; in mice they gave complete protection against lethal neurotoxin challenge, including as post-envenoming rescue. The cytotoxin binder neutralized cytotoxicity in vitro but did not reduce dermonecrosis in vivo. The authors position the binders as low-cost, animal-free antivenom components or "fortifying agents." Published in Nature (Jan 2025).

Metrics

4
Short-chain neurotoxin binder (SHRT): mouse survival vs 3× LD50 challenge100% (incl. 15-min post-toxin rescue)
Long-chain neurotoxin binder (LNG): mouse survival vs α-cobratoxin100% (preincubation & 15-min rescue); 60% at 30-min rescue
Cytotoxin binder (CYTX): in vitro protection vs 7 Naja venoms70-90% in vitro, but no reduction of dermonecrosis in vivo
Binder size / thermal stabilityconventional antibodies: larger, animal-derived~100 aa; melting temp up to >95 °C

Lessons learned

  • Deep-learning protein design can produce potent, ultra-stable toxin binders from structure alone — without animal immunization or large-library screening — pointing to cheap microbial manufacturing suited to low-resource settings.
  • In vitro potency did not guarantee in vivo success: the cytotoxin binder neutralized cells but failed to prevent dermonecrosis in mice, so tissue-level endpoints must be tested directly.
  • Because each binder targets one toxin, broad protection needs a designed cocktail; the near-term framing is 'fortifying' existing antivenoms, especially for poorly-immunogenic elapid neurotoxins, rather than standalone replacement.
  • Coverage so far is limited to elapid three-finger toxins; viper metalloproteinases and serine proteases remain to be designed against.

Documented Jul 8, 2026

Author AvatarArnaud Gissinger

communityfix.org