Snakebites claim the lives of more than 50,000 people in India each year, and the antivenoms used today do not work equally well against all snake venoms. Now, researchers have combined just five laboratory-produced antibody fragments into an experimental antivenom that protected mice against venom from several cobras and king cobras from India – even when treatment was delayed until the venom had already begun to take effect.
A serious cobra bite can quickly paralyse the muscles needed for breathing. Worldwide, an estimated 80,000–140,000 people die each year from snakebite envenoming, and around 58,000 of these deaths occur in India.
Yet, treatment is still based on a principle dating back to the 1890s: antibodies extracted from the blood of horses that have been immunised with snake venom. This treatment has saved countless lives, but venom can vary so much between species and geographical regions that an antivenom that works well in one place may work less well in another.
“It is actually quite remarkable that we still treat snakebites using more or less the same basic principle that was developed in the 1890s. It works, and it has saved an enormous number of lives, but today, we can make the treatment much better. The challenge is to create an antivenom that is not limited to one particular snake species from one particular place,” says Andreas H. Laustsen-Kiel, Professor at the Department of Biotechnology and Biomedicine of the Technical University of Denmark in Kongens Lyngby.
Researchers from Denmark and India have now tested a different approach. They combined five carefully selected nanobodies – very small antibody fragments – that target four key groups of toxins found in the venoms of cobras and king cobras. The five nanobodies were originally developed against toxins from African venomous snakes, but many of the same toxin families are also present in cobras and king cobras from India.
“What is quite interesting is that the cocktail protected mice against venom from spectacled cobras from different parts of India, against monocled cobras and against both king cobra species we tested. With spectacled cobra venom from Maharashtra, we could even wait 20 minutes before giving the treatment, and all the mice still survived. Twenty minutes is a very long time in a mouse, so the result is more encouraging than the number might sound at first,” says Andreas H. Laustsen-Kiel.
Snake antivenom still relies on a principle from 1894
The new design aims to solve a problem that traditional antivenom has never fully overcome: the wide variation in snake venom.
“It is quite striking that snakebites still claim around 100,000 lives a year when we have actually had antivenom for more than 100 years. The problem is not that we cannot treat snakebites at all, but that the current treatment is still expensive, difficult to produce and does not always work equally well against the enormous variation in snake venom,” emphasises Andreas H. Laustsen-Kiel.
Traditionally, a horse is given repeated, non-lethal doses of snake venom so that its immune system produces a large mixture of antibodies, which can then be purified from the blood and used as treatment. The method has saved countless lives, but unlike modern recombinant medicines, the end product is not precisely defined: its composition varies from batch to batch, and the treatment can trigger severe immune reactions.
“It is rather similar to what happened with insulin. In the past, insulin came from animals; today, it is produced using biotechnology. That is the direction we would like to take with snake antivenom. The difference is that snake venom contains many different toxins, so here we need to put together the right combination of antibodies,” emphasises Andreas H. Laustsen-Kiel.
In recent years, researchers have therefore begun developing recombinant antivenoms. Instead of the thousands of different antibodies found in animal plasma, these can comprise carefully selected antibodies or smaller antibody fragments.
Nanobodies developed for African snakes could also work against some Indian snakes
In the new study, the researchers used nanobodies (VHHs). They are derived from a special type of antibody found in camelids and are significantly smaller than conventional antibodies. They are highly stable and can be produced in microorganisms.
“Nanobodies are essentially very small and stable antibodies. They can be extremely good at recognising a particular toxin and are robust and relatively simple to produce. That is a rather attractive combination if you want to develop a modern antivenom that can also be produced on a large scale,” emphasises Andreas H. Laustsen-Kiel.
The research team had already demonstrated the principle using venoms from African snake species, where a previous nanobody cocktail protected against 17 of the 18 medically most important elapid snakes in sub-Saharan Africa in animal experiments. Cobras and king cobras from India are evolutionarily and geographically distinct from the African snakes, but their venom contains several of the same major toxin families – including α-neurotoxins, cytotoxins and phospholipase A2.
“The interesting thing is that we did not have to start from scratch. We already had nanobodies from our work on the African snakes, and many of the same types of toxins are also present in cobras and king cobras from India. So, the question was really: how much of what we had already developed could we reuse? It turned out to be quite a lot,” explains Andreas H. Laustsen-Kiel.
The aim is precisely to identify antibodies that do not work against just one specific toxin but recognise features shared by many related toxins. This means researchers can build on antibodies they already have rather than starting from scratch each time they move from one snake species or geographical region to another.
Five nanobodies target four key groups of toxins
The five nanobodies were not developed de novo. The researchers took antibodies from previous programmes targeting African elapids and adapted the mixture to the toxin families that predominate in cobras and king cobras from India. Three mamba-specific components were removed (as mambas only live in Africa), and one nanobody was replaced with a variant that binds more broadly.
“This left us with five nanobodies, which together target four of the most important toxin groups: long- and short-chain α-neurotoxins, cytotoxins, and phospholipases A2. We had identified several of them using phage display technology and selected them because they could target the toxins we wanted to cover,” says Andreas H. Laustsen-Kiel.
For some toxin families, the researchers used what is known as a consensus toxin: an artificial protein that combines features shared by many related toxins. This increases the chance of finding antibodies that recognise many natural variants within the same toxin family rather than just a single toxin.
“With phage display technology, we have an enormous library of different antibodies we can fish in. The toxin acts a bit like bait: the antibodies that can recognise it and stick to it are pulled out and tested further. But what is really interesting for us is catching those that do not just bind to one toxin but can recognise many variants within the same family,” adds Andreas H. Laustsen-Kiel.
For α-neurotoxins, a nanobody must do more than simply recognise the toxin. These toxins can block signals between nerves and muscles and, in severe cases, paralyse the muscles needed for breathing. The nanobody must therefore bind in such a way that the toxin can no longer block the signal.
“The other toxin families work in different ways, but the principle is the same: it is not enough for the antibody to recognise the toxin – it must also prevent it from causing harm. That is why we then tested whether our nanobodies could both recognise and neutralise toxins in venoms from different snake species and from different parts of India,” says Andreas H. Laustsen-Kiel.
From binding toxins in the laboratory to protecting mice
Andreas H. Laustsen-Kiel and colleagues tested venom from five geographical populations of spectacled cobras, two populations of monocled cobras, and the two king cobra species, Ophiophagus kaalinga and Ophiophagus hannah. All five nanobodies bound to the relevant venoms. The researchers also investigated whether the nanobody targeting phospholipase A2 not only bound to the toxin but also inhibited its enzymatic activity.
“First, we want to see whether they bind broadly across the different venoms. Then we need to show that they can actually neutralise the function we want to block. That is what determines whether the binding is also biologically relevant,” says Andreas H. Laustsen-Kiel.
They also tested whether the nanobodies had some of the properties required if they are eventually to be used as medicines. They showed no measurable non-specific binding to a range of biological molecules, and the mixture was tested at high protein concentrations and temperatures of up to 50 °C. Even after 16 hours at 40 °C, most of it remained soluble.
The crucial test was whether the cocktail could protect mice. In the initial experiments, the researchers mixed three times the median lethal dose (3 × LD50) of each snake venom with the nanobodies before injecting the mixture. This enabled them to measure how effectively the nanobodies neutralised the venom under controlled conditions.
They also performed more realistic rescue experiments, where the venom was administered first. The mice received it subcutaneously and were then given the nanobody cocktail intravenously. The researchers tested both lower doses and delayed treatment to determine whether the cocktail still worked once the envenoming had already begun.
“Pre-incubation tells us whether the nanobodies can neutralise the venom. But in the real world, of course, you get bitten first and treated afterwards. That is why the rescue experiment is far more interesting: can we still stop the envenoming once the venom has already entered the body?” explains Andreas H. Laustsen-Kiel.
The antivenom still works after envenoming has begun
When the researchers mixed the nanobody cocktail with three times the median lethal dose of the various cobra and king cobra venoms, all treated mice survived without any visible signs of envenoming. This was the case for venom from spectacled cobras from five different parts of India, two populations of monocled cobras, and both king cobra species studied. In the control groups, the mice typically died within 15 minutes.
In the rescue experiments, the mice were first given three times the median lethal dose of venom subcutaneously from either a spectacled cobra, a monocled cobra, or Ophiophagus kaalinga. Five minutes later, they received the nanobody cocktail intravenously. Untreated mice rapidly developed clear neurotoxic symptoms – including breathing difficulties, paralysis and loss of reflexes – and died within 25–60 minutes. All the treated mice survived the 24-hour observation period.
“The crucial point here is not just that the animals survive. We administer the venom first and then give the cocktail once the envenoming has begun. This provides a far more relevant picture of how a treatment actually needs to work after a snakebite,” says Andreas H. Laustsen-Kiel.
Using venom from spectacled cobras from Maharashtra, the researchers went a step further. All mice survived even when the cocktail was not administered until 20 minutes after the venom. The researchers were also able to reduce the dose to a 1:1 molar ratio between toxin and nanobody and still achieve full protection.
“For me, the 20-minute experiment is one of the most important findings. Here, the venom has already been in the body for some time before we administer the treatment, and yet all the mice survive. Of course, it is still a mouse model, but it is much closer to the situation in which a treatment actually needs to work,” emphasises Andreas H. Laustsen-Kiel.
The experiments also revealed the limits of protection
The effect had a clear limit. When the researchers reduced the molar ratio further to around 0.3:1 nanobody to toxin, all the mice eventually died, but around two hours later than the untreated animals. The lower dose had an effect but was not enough to save the animals.
“It is also important to know the point at which the treatment is no longer strong enough. Here we can see a clear dose–response relationship: first full protection, and when we go low enough, only a delay in the time of death,” says Andreas H. Laustsen-Kiel.
The cocktail also did not work against everything. It offered no protection against common krait venom, which was consistent with the laboratory tests. Krait venom is dominated by beta-bungarotoxins, which the cocktail was not designed to target.
“It was no great surprise that the cocktail did not work against krait venom, because it simply does not contain antibodies against some of the toxins that are most important in krait venom. If we want to include krait, we need to identify the missing components and add them. This actually illustrates very well how we think about antivenom,” adds Andreas H. Laustsen-Kiel.
Getting from mice to people will take more
The new study was carried out in mice, but Andreas H. Laustsen-Kiel hopes that the first clinical trials can begin within the next 3–5 years. The first treatment to reach people, however, does not need to be a fully recombinant antivenom. One possibility is to supplement existing horse-derived antivenoms with a few carefully selected antibodies that fill some of the gaps in their protection.
“Of course, we can work towards completely replacing the old antivenom, but there is also a very pragmatic route. We can take an antivenom that already works and is already being used and then boost it with a few of our antibodies where it has gaps. If, for example, it does not neutralise a particular important toxin well enough, we can add a component that does exactly that. In this way, we could potentially improve an existing treatment quite substantially without having to build the whole thing from scratch – and it might also be a faster route through clinical development,” explains Andreas H. Laustsen-Kiel.
One challenge will be to ensure that the nanobodies both reach the tissues quickly and remain in the body long enough. Their small size helps them reach the toxins quickly but also means that they can disappear from the blood relatively quickly. Other molecular formats may therefore be worth exploring – for example, multimers or Fc fusions that remain in the body for longer, although they may penetrate tissues more slowly.
“There is a trade-off. The advantage of nanobodies is that they are small and can reach and penetrate the tissues quickly. The downside is that they can also disappear from the blood quickly. So, we need to find the right balance, and this may well mean that the final version looks a little different from the molecules we are testing here,” says Andreas H. Laustsen-Kiel.
Could the same building blocks make antivenom cheaper and broader?
Production could be another advantage. Nanobodies can be produced in microorganisms, and the research team has already shown that this is feasible at pilot scale. India also has substantial capacity for producing biological medicines.
Calculations based on data from the team’s pilot study suggest that recombinant antivenoms could be cheaper to produce than some of the high-quality products used today.
“Our calculations suggest that this type of antivenom could potentially be produced for perhaps 25 to 125 euros per treatment. By comparison, a high-quality traditional antivenom has a price tag of around 1,000 euros. These are, of course, still estimates, but if we can make the treatment both better and significantly cheaper, it starts to become really interesting for the areas where snakebites hit hardest,” explains Andreas H. Laustsen-Kiel.
However, the current cocktail is far from covering all of India’s medically important venomous snakes. In addition to kraits, several viper species have other important toxins, so broader mixtures will require additional nanobodies, other antibodies, or small-molecule toxin inhibitors.
“We have previously demonstrated the same principle for African elapids and American coral snakes. The interesting question is whether we can build antivenoms around toxin families that recur across many snakes rather than having to start from the venom profile of each individual snake. This does not necessarily mean that we will end up with a single universal antivenom,” says Andreas H. Laustsen-Kiel.
Some antibodies may be useful across many snake species, whereas others will need to be added depending on which toxins are most important in a particular region.
“That is where I think it starts to get really interesting. If some of the same broadly neutralising antibodies can be reused against snakes in Africa, India, and other parts of the world, we will not have to invent a completely new antivenom from scratch every time. We can build on components we already know and combine them according to the toxins we need to cover. But ultimately, even a better antivenom will only save lives if patients can reach treatment and hospitals have access to it,” concludes Andreas H. Laustsen-Kiel.
