Do not poison thyself

Abstract

The mechanisms by which poisonous plants and animals protect themselves against their own toxins provide interesting leads for synthetic biology with potential applications in medicine and agriculture.

The use of toxins—either for self‐defence, or to kill or paralyse prey—is widespread among all kingdoms of life, including plants, animals, fungi and bacteria. Yet, as organisms evolved agents of chemical warfare and the metabolic networks to synthesize these, they also had to evolve the means to avoid damage to themselves. Many toxins aim at universal targets, such as the nervous system in animals or crucial metabolic pathways, which means that mechanisms of self‐avoidance are fine‐tuned and sophisticated and could therefore inspire basic and applied research, in particular in synthetic biology.

… as organisms evolved agents of chemical warfare and the metabolic networks to synthesize these, they also had to evolve the means to avoid damage to themselves.

Humans have known of and used many of these toxins for medicinal, cosmetic and recreational use: opium, digitalis, penicillin, nicotine or taxol are just a few examples. During the 19^th and 20^th century, many compounds were isolated, analysed and eventually synthesized. However, it took much longer to unravel the metabolic pathways by which organisms synthesize toxins, let alone the factors that protect them from self‐toxicity. Recent discoveries to elucidate these pathways have shed light on the evolutionary arms race between predators and prey, along with identifying compounds or targets of interest for medicine and agriculture. Overall, this research has led to the co‐inheritance theory, whereby all genes associated with toxin production are clustered together with the protective genes and passed on to offspring as a single package. The hypothesis is that inheriting genes for producing toxic compounds without the corresponding ones for protecting against these would be counterproductive.

The evolution of toxins

The evolution of toxicity for defence probably began with bacteria, which developed an array of chemicals that play various roles ranging from subverting other microorganisms to quorum sensing and biofilm formation 1. In eukaryotes, toxin production probably began in plants and fungi, again serving a variety of functions including self‐defence against consumption by early herbivores or fungivores. In the case of plants, self‐immunity against those toxins arose along with the evolution of chemical defences, according to Fred Rook from the Plant Biochemistry Laboratory at the University of Copenhagen in Denmark. “The most effective chemical defence is one that affects a broad variety of pathogens and pests, and/or that acts on an essential cellular process”, he said. “The flip side of this is that such an effective defence chemical is frequently toxic to the plant itself as well. There are various ways of dealing with this duality. For example, cyanogenic glucosides are a two‐component defence system. The glycosylated molecule is stable and much less toxic. It is activated or cleaved by an enzyme, a beta‐glucosidase, upon tissue damage, releasing the toxic effect”. This protects the plant against herbivore attack while avoiding self‐toxicity because the relevant components are harmless at other times. If cleaved by an enzyme, cyanogenetic glucosides release hydrogen cyanide, or prussic acid, that halts respiration by inhibiting a key mitochondrial enzyme called cytochrome c oxidase.

One of the common cyanogenic glucosides is called dhurrin, which has the unusual property of being unstable at the prevailing pH in plant cell cytosol, which is normally slightly alkaline at around 7.2. Rook pointed out that dhurrin is only stable under slightly acid conditions and is therefore held in the vacuole, a storage unit of plants cells with a pH around 5.0. “Storage in the vacuole, an acidic compartment of the plant cell, then becomes an important feature of this defence system”, Rook explained. “This is the evolutionary reason the transporter is also found in the biosynthetic gene cluster, promoting its co‐inheritance”.

In eukaryotes, toxin production probably begun in plants and fungi, again serving a variety of functions including self‐defence against consumption by early herbivores or fungivores.

Moreover, many compounds produced by plants are not toxic to themselves, because they lack the relevant receptors or enzymes, but are detrimental to the animals that consume them. This is the case for many alkaloids, a broad group of nitrogen‐containing compounds produced by plants alongside glycosides. Examples include nicotine, cocaine, caffeine and codeine that often have other functions, such as deterring rival plants. Plants do not have nicotinic acetylcholine or opioid receptors and are therefore immune to many alkaloids; they similarly lack oestrogen receptors so can safely make phytoestrogens that act against those receptors in animals.

Evolution of toxicity in animals

Co‐inheritance of genes associated with toxicity is common to most if not all organisms that use chemical defences, even though the mechanisms and selective factors differ widely. While various vertebrates produce their own toxins, most animals take them up from organisms they consume. Herbivores would get toxins from the plants they eat, while carnivores higher up the food chain would obtain toxic compounds from other animals that had already accumulated them from plants.

While various vertebrates produce their own toxins, most animals take them up from organisms they consume.

Such animals were therefore under evolutionary pressures to avoid self‐toxicity against the poison from their diet. “Frogs first developed a tolerance to alkaloids in food items”, said Richard Fitch at the Department of Chemistry and Physics, Indiana State University, USA, who studies poisonous frogs. “This is not uncommon in animals and gives them an adaptive advantage, utilizing food sources others cannot consume. Then over some generations, the tolerance grows to the point that the frog can consume enough material that it becomes toxic to predators itself and does not get eaten as often”.

This in turn encourages the frog to become brightly coloured and conspicuous, unlike non‐toxic species, which tend to evolve camouflage. “Conspicuousness such as bright coloration is ubiquitous across life for attracting mates and warning competitors”, Fitch explained. “If the frog is already toxic, then the conspicuousness is no longer a liability and additionally serves the warning function we associate with aposematism”. This is the process of communicating danger such as toxicity with a visual warning to reduce the risk of accidental consumption. The success of such warning strategies has been exploited by other animals in mimicking the appearance of similar‐looking species without evolving the toxicity themselves, such as insects that imitate wasps and bees without being able to sting.

Resistance in poison frogs

Progress has been made untangling how poison frogs protect themselves against the toxin while minimizing side effects of that adaptation. The frogs first evolved toxicity by isolating chemicals from their arthropod prey and storing them in their skin. One of the most potent ones is epibatidine, an alkaloid neurotoxin that exerts stronger pain relief than morphine and eventually leads to numbness and paralysis but which is harmless to the frogs themselves. This avoidance of self‐toxicity evolved through amino acid substitutions in the toxin's target, the nicotinic and muscarinic acetylcholine (nAChR) receptors, which changed the protein structure sufficiently to avoid binding with epibatidine. Initially, this came at the cost of reduced ability of the receptor to bind the neurotransmitter acetylcholine, which was a serious deficiency since motor neurons use it to signal activation to muscles. Some frogs underwent further amino changes in their nAChR receptor to restore the ability to bind acetylcholine, while still being insensitive to epibatidine, as researchers at the University of Texas in Austin, USA, showed 2. “The paper was impressive in that it picked apart the costs and benefits of specific amino acid substitutions that contribute to resistance”, commented Butch Brodie, Director of the Mountain Lake Biological Station at the University of Virginia, USA, who was not involved in the study. “They were able to get the precise effects of individual switches within a protein to show how some compromised basic function, and others compensated for that lost function. We've always suspected that dynamics, but this is the first demonstration I'm aware of”.

The same paper also found that only some frogs had restored the ability of the nicotinic acetylcholine receptor to bind acetylcholine, while others, such as the Oophaga genus (Fig 1) that comprise nine known species of poison dart frog, had not. The question of why they had not restored the function of the receptor and how they coped with that deficiency is still open, according to Rebecca Tarvin, the lead author on that paper. “Clearly these animals survive without nearby additional replacements, suggesting one of the following scenarios”, she explained. “Firstly, Oophaga may tolerate the cost of the original amino acid substitution (called S>C) because of a trade‐off with the benefit from resistance S>C provides or the higher levels of defence S>C could allow. Secondly, Oophaga may have other compensatory mechanisms such as increased acetylcholine synthesis or changes in the expression of nicotinic acetylcholine receptors. Thirdly, Oophaga may have other replacements we did not identify that ameliorate the cost of the S>C replacement, or fourthly Oophaga do experience a cost in function of the receptor and thus represent some intermediate stage in the evolution of resistance to epibatidine”.

Figure 1. Oophaga pumilio .

Wikipedia/Marshal Hedin, San Diego, USA.

Some frogs underwent further amino changes in their nAChR receptor to restore the ability to bind acetylcholine, while still being insensitive to epibatidine…

This apparent lack of the receptor function in Oophaga frogs may also reflect the weakness of focusing on single amino acid substitutions, according to Brodie. “Those without the compensatory changes we would expect to either evolve them later or lose the toxicity and therefore resistance against it altogether”, he said. “It's also possible they have some other compensatory mechanism we don't know about. The risk of trying to evaluate function of a protein alone is that we don't know whether other aspects of the organism have evolved along with it in an interactive way”.

Tarvin acknowledged the need to consider the larger genetic background. “By genetic background here we mean the entire sequence of the nAChR receptor would directly impact the potential effect of a specific amino acid replacement, given that different regions of the protein interact during protein folding and activity”, she explained. “Indeed, we do quantify drastic differences in the cost of the S>C replacement when we point‐mutate human versus frog receptors”. In the human case, the point mutations decrease the sensitivity of the receptor 100 times, but only by a factor of two in frogs.

Poison frogs capture toxins with their diet, in this case, ants, which, in some cases, can manufacture toxins themselves, unlike most animals 3. This too has evolutionary significance because the frogs must actively seek out poisonous arthropods. “These frogs need to be very active predators and eat a lot of these already toxic prey that might be rare in the first place, hard to digest or poor in nutritious value”, commented Juan Santos from St. John's University in New York and co‐author on the paper. “In contrast, close relatives that are non‐toxic are not very active predators, eat whatever is available and abundant, and do not need to worry about sequestering toxins. However, these frogs have to be very good at hiding themselves through camouflage. Both strategies, toxic or non‐toxic/camouflage, are effective anti‐predator strategies, but in the long run toxicity, such as in poison frogs, is more specialized. This complexity most likely will make it not widespread among anurans”. Tarvin added that at least poison frogs do not have to share their diet with many others: “Once the frogs evolved resistance, they may have unique access to this enormous resource – ants make up a large portion of the biomass in the rainforest”.

Producing versus acquiring toxins

Though snakes are more related to frogs, venomous snakes employ significantly different strategies to avoid self‐toxicity. Snake venom is modified saliva produced by adapted glands in the mouth and injected via fangs for stunning or killing prey or as defence against predators. Snakes that have their fangs at the front of the mouth tend to use venom both for defence and predation, enabling them to capture animals much larger in girth than themselves. This includes the elapids—cobras, mambas, coral snakes and kraits—and the viperids, such as rattlesnakes and adders, which use the same venom and delivery system.

Some snakes are rear‐fanged and they tend to use their venom in defence while feeding on small animals such as newts and toads that do not need to be stunned or killed. One particular rear‐fanged species, Rhabdophis tigrinus (Fig 2), has enabled scientists to identify the source of snake venom, which comes from its favoured diet of poison toads 4. Indeed, many other snakes feed on the same toads, which defend themselves chemically with bufadienolides, cardiotonic steroids that act as heart stimulants through the Na/K‐ATPase ion pump. “Snakes that are exposed to these toxins through their diet have gained mutations in the gene coding for this protein, making them resistant against the toxin”, explained Shab Mohammadi, who has been involved in studies of self‐toxicity in Rhabdophis tigrinus at the University of Nebraska–Lincoln in the United States. The toads themselves have a similar mutation that renders them immune. Unlike in the case of the poison dart frogs, the mutation was present in a wide range of taxa including many species that do not feed on those toads at all and therefore do not need protection against bufadienolides 5. Mohammadi concluded that either there is little performance cost associated with these mutations or they provide an unknown benefit.

Figure 2. Rhabdophis tigrinus .

Wikipedia/Public Domain.

Epibatidine, used by poisonous frogs, attracted much attention as a potential pain‐killer because of its strong action, but it proved to be too toxic.

Most vertebrates using chemical defences tend to acquire their venom from their diet, whereas many invertebrates manufacture their own. This is the case for Hymenoptera, the order of arthropods including bees, wasps, hornets and ants. Their females use a modified ovipositor—originally evolved for laying eggs—to inject the venom. The insects have not evolved immunity to their venoms, and self‐toxicity is conferred by physical separation.

Exploiting the metabolic pathways

Given the long history and widespread use of natural toxins, there is great interest in exploiting both the metabolic pathways to synthesize these compounds as well as the mechanisms that provide immunity against them. Epibatidine, used by poisonous frogs, attracted much attention as a potential pain‐killer because of its strong action, but it proved to be too toxic. “Epibatidine holds enormous potential for improving our understanding of nicotinic acetylcholine receptors”, Tarvin commented. “Any information about these proteins and their functions could improve our medical toolbox. But it is not likely to be used as a pain‐killer”. Nonetheless, related compounds have been tested to help smokers give up tobacco by desensitizing them to nicotine based on their action on the nicotinic acetylcholine receptor 6.

There is greater potential though by exploiting knowledge of self‐toxicity avoidance in agriculture. One objective is to reduce the toxicity of food crops. A recent project funded by the European Union found that it was possible to eliminate glucosinolates from oil seed rape that the plant uses to defend themselves against herbivores 7. This was achieved by mutating two genes encoding glucosinolate transporters (GTRs), but translation of this loss‐of‐function mutation into the plant proved challenging, because Brassica is polyploid. One benefit of such works is to increase the nutritional value of plants, in this case to produce safer animal feed cake from oil seed rape.

Rook also noted that knowledge of the genes involved in self‐toxicity has broader potential for identifying biosynthetic enzymes and in plant breeding. This exploits the co‐inheritance theory that self‐toxicity genes tend to be clustered together so that they are inherited as a unit. “The advantage of this genomic organization is that it provides candidate genes for biosynthetic enzymes and transporters for testing, requiring availability of continuous genomic DNA sequences”, Rook explained. “This speeds up the elucidation of biosynthetic pathways of interest and their agricultural or biotechnological application. Knowing the genes involved in a chemical defence pathway allows their use as genetic markers in traditional breeding programs. If they are organized in a gene cluster, then the pathway will be closely linked genetically and can be transferred as a complete functional unit”. He added that it may also be possible to transfer these units to other plant species to fend off herbivores: “A benefit there would be that the natural pests of the receiving species, such as a crop, will not be adapted to the introduced defence system”.

… knowledge of the genes involved in self‐toxicity has broader potential for identifying biosynthetic enzymes and in plant breeding.

For many millennia, humans have been exploiting nature's chemical warfare agents for their own purpose, but it was only during the past few decades that science was able to unravel their mechanisms of action so as to improve their benefits for human use. A better understanding of how plants, animals and microorganisms can evade the effects of the very poisons that they produce could help to further improve human use of natural toxins not just for agriculture.

EMBO Reports (2018) 19: e46756