What is the problem?
Snakebite remains a neglected tropical disease, claiming over 100,000 lives a year and causing severe or lasting disabilities for many more, according to the authors. Among the culprits, the three-finger toxins of some snakes (the elapids) can destroy tissue or block nerve receptors, with a risk of paralysis.
Today the only treatment is a serum made of antibodies taken from the plasma of immunised animals. It is costly and of limited effect against these toxins.
Cytotoxins cause severe tissue damage, while neurotoxins block nicotinic acetylcholine receptors, the docking points nerves use to signal muscles, which can be life-threatening.
What did the researchers do?
They used deep learning methods to design from scratch (de novo) proteins able to bind three kinds of three-finger toxins: short-chain neurotoxins, long-chain neurotoxins and cytotoxins.
With limited lab screening, they obtained proteins that are very stable to heat, bind strongly to their target, and whose shape matches the computer model very closely.
Because the proteins were designed by computer, only a limited number had to be tested in the lab to find good ones.
What were the results?
The designed proteins neutralised all three families of three-finger toxins in lab tests. They also protected mice from a lethal dose of neurotoxin.
The authors stress that these proteins are potent, stable and easy to manufacture, which could pave the way for a cheaper, more widely available antivenom. Good stability to heat matters for shipping and storing a medicine. Also, the manufactured proteins had almost exactly the shape the computer had drawn, which shows the design approach really works.
"In vitro" means in a dish or test tube. In the mouse test, the animals faced a lethal neurotoxin challenge, a dose that would otherwise be deadly.
Key figures
- over 100,000deaths a year from snakebite, according to the authors
- 3 familiesof three-finger toxins neutralised in the lab
What does this change for patients?
For now, nothing for a person who has been bitten: this is laboratory research. But the idea is to one day replace animal-derived serums with manufactured proteins that are more stable and cheaper to produce.
The authors also hope computer-based design will help develop treatments for neglected tropical diseases in places with limited resources.
Beyond snakebite, the authors argue that computational design could help democratise therapeutic discovery by substantially reducing costs and resource requirements, particularly where resources are limited.
What are the limits?
The study stops at lab tests and mice given an isolated toxin, not a whole venom with its many components. It targets one large family of toxins, whereas a venom contains many others.
Safety in humans, the right dose and real-world effect after a bite must still be shown through clinical trials yet to come. The authors present this work as a possible basis for next-generation antivenoms, not as a finished product.
There is no human data yet.
After a bite or a sting: a word from Mima
These proteins exist only in the lab, so if you are bitten, nothing new is available. Emergency services and hospital remain the only answer. The study opens a route for future treatments.
For your care in general, Mima can keep your prescriptions, appointments and questions, and explain treatments in plain words. A question for your doctor before a trip: "What should I do after a bite in the region I am going to?"
Frequently asked questions
Can AI create an antivenom?
In this study, an AI designed proteins that neutralise snake venom toxins in the lab and protect mice from a lethal neurotoxin. It is not yet a treatment for humans.
Why are current antivenoms limited?
They are made of antibodies taken from the plasma of immunised animals. According to the authors they are costly and of limited effect against three-finger toxins.
What is a de novo designed protein?
It is a protein designed from scratch by computer, rather than copied or modified from a protein found in nature.
When will this treatment be available?
The study gives no date. Safety tests and clinical trials in humans are still needed.
The original study
Vázquez Torres S, Benard Valle M, Laustsen AH, Casewell NR, Baker D, et al. “De novo designed proteins neutralize lethal snake venom toxins.” Nature, 15 January 2025. DOI : 10.1038/s41586-024-08393-x
Read the full study in NatureThis article summarises a scientific study and is not medical advice.