Did Allergies Evolve To Save Your Life?

By Christie Wilcox | October 24, 2013 11:00 am
Introduced bee on a native ʻōhiʻa flower

Introduced bee on a native ʻōhiʻa flower.
Photo by Flickr user Zesuri

As most of my friends on the mainland don longer sleeves and more layers, it’s hard not to be a little smug about living in paradise. While, in their neighborhoods, leaves are falling off of trees and icy winds threaten to bring snow, I can throw on a T-shirt and shorts, grab a picnic basket, and hike to a scenic overlook for lunch. But Hawaii’s ever-sunny weather comes with one side-effect that can be deadly serious: year-round, Hawaii has bees.

Don’t get me wrong—bees are wonderful insects, even though most species are not native here in Hawaii. When the first Hawaiians arrived on the shores of these stunning isles, only the yellow-faced bees buzzed around. Honey bees were introduced in the 1850s, and have since become indispensable, wedging their way into the Hawaiian economy and filling the shoes of native pollinators that have become scarce. They’re economically and ecologically vital, not just here in Hawaii, but throughout the US—according to the U.S. Department of Agriculture, these busy bugs pollinate 80% of our flowering crops, and are thus essential for the production of 1/3 of our food. Worldwide, the economic contribution of pollination alone has been valued at over $200 billion.

But for more than 2 million Americans, bees are a dangerous threat. Somewhere between 1% and 7% of human beings are allergic to insect venoms, with their symptoms ranging from mild overreactions to full-blown anaphylactic shock. For those with bee allergies, even the slightest sting can lead to a fight for life. Even more troubling is that, in half of all fatal sting allergy cases, victims had no previous major reactions to venom. Nearly 100 Americans die every year from bee stings, and many more experience severe reactions that require immediate medical treatment.

The allergic response to a bee sting.
From What To Do About Allergies

Allergies are defined as ‘hypersensitive immune responses’—or, in colloquial terms, odd moments when our immune systems flip out. Anaphylaxis is the whole-body manifestation of an allergy, which can range from something as minor as hives to sharp drops in blood pressure and even cardiac arrest. You don’t have an allergic reaction the first time you come in contact with an allergen; instead, like with viruses or other potential invaders, your body takes an immunological picture so it can remember the allergen later. This is what is known as the adaptive immune response, and it’s usually a good thing—when you get the chicken pox, for example, your adaptive immune system remembers what the disease looks like, and can find and kill it should you ever be re-exposed. But when it comes to allergies, the adaptive immune system goes too far. The next time it detects allergens, it sends out hordes of IgE antibodies to destroy them. These IgE antibodies wreak havoc in our bodies—through cascading immunological pathways, IgE antibodies cause the release of histamine and other inflammatory compounds and can lead to anaphylaxis.

In fact, IgE antibodies are so damaging, scientists struggle to understand why—from an evolutionary standpoint—our bodies produce them in the first place. After all, without modern medicine like Epi-Pens, people with severe allergies would most likely die from their condition. Genetic conditions that cause swift death, especially as a child, tend to be weeded out because they prevent individuals from passing along their genes. So how have allergies persisted so long?

For a long time, scientists have viewed allergies as ‘immunological anomalies’—physiological accidents, if you will—genetic disorders in which the body produces IgE antibodies when it shouldn’t. But this view of allergies doesn’t explain why IgE antibodies are found in our genomes in the first place. Given the conservative nature of evolution, it doesn’t make sense for a whole class of destructive antibodies to arise that are never used except by people with a genetic disorder. They have to have some purpose–or, at least, have had one in the past. One group of scientists thinks that IgE antibodies are meant to fight against parasites, making allergies still the unfortunate side-effect of a different body action. Yet even this is unsatisfying—the vast majority of allergens are pollens, foods, drugs, venoms, and metals. Is our parasite defense system really so poorly fashioned that it makes so many mistakes? It was an incongruity that Margie Profet simply couldn’t accept.

Profet isn’t exactly your classic evolutionary biologist. She received degrees in physics, math and philosophy, but not a single one in biology. Yet her grasp of the evolutionary mechanics of allergies is profound. “The specialized mechanisms that collectively constitute the allergic response appear to manifest adaptive design in the precision, economy, efficiency, and complexity with which they achieve the goal of producing allergy,” she wrote. In 1991, Profet proposed a radical explanation for allergies: IgE antibodies, and the allergic reactions they cause, were meant to save our lives from toxins.

Profet laid out four main arguments, which she coined the “toxin hypothesis” for allergies. First, that toxins are ubiquitous and cause acute damage. Thus it would make sense for our bodies to have developed defenses against them. Second, the types of physiological activities that toxins perform—covalent binding to serum proteins, for example—are known to trigger allergic reactions. Third, most allergens are either themselves toxic substances or carrier proteins that bind to smaller toxic substances. Lastly, allergic symptoms could be construed as helpful in the case of envenomation or poisoning, as allergies cause behaviors like vomiting, sneezing and coughing, which could expel toxins, and drops in blood pressure, which could slow the speed at which a toxin moves through the body.

While her ideas won Profet a McArthur Genius Grant in 1993, given her background and the controversial nature of her suggestion, the scientific community was not instantly convinced. Most of her detractors focused on one simple detail: a lack of any experimental evidence. If allergic responses are adaptive, they said, show us. Decades passed with little evidence to support Profet’s radical explanation.

Then, in 2012, Cornell University researchers Paul and Janet Shellman-Sherman connected Profet’s toxin hypothesis to a strange medical phenomenon: people with allergies have lower risks of certain cancers. It’s possible, the team told Psychology Today, that allergies serve to combat potential carcinogens. It was a suggestion made by Profet herself—she noted that the most cancer-causing heavy metals also were the most allergenic, and aflatoxins from fungi that grow on the bane of many allergy sufferers—hay—are some of the most carinogenic substances known to man. If allergies protect the body against such substances, people with them would be less likely to expose themselves to these mutagenic toxins that can cause cancer. Others think the correlation is simpler: the highly overactive immune system of allergic people attacks all invaders, foreign and domestic, and is thus hyper-vigilant against cancer. With so much uncertainty, the connection between allergies, toxins, and cancer remains philosophical.

Today, a paper in the journal Immunity provides even stronger evidence to support Profet’s toxin hypothesis. In it, Stanford University School of Medicine scientists show that small doses of venom and the subsequent allergic pathways triggered serve to protect mice against fatal doses of venom later on.

The team was interested in how adaptive immune responses help in the case of venomous stings and bites. Statements made by postdoc Philipp Starkl, co-first author or co-leader of the study, echoed Profet’s original reasoning: “It was kind of a dogma that most IgE-related responses are detrimental,” he said. “We and others speculated that there should be some very positive evolutionary pressure to keep these cells and these antibodies, because if they were just bad and deleterious, they would have been eliminated.”

Bee stings, in particular, are known for their allergic potential. So, the team designed an experiment using a mouse model to see if allergic responses were beneficial in the case of bee stings. They first injected one group of mice with small amounts of bee venom, equivalent to one or two stings, and another with saline. Then, three weeks later, they injected both groups with a potentially lethal dose of bee venom, and watched to see how the two groups responded.

The mice previously injected with venom fared significantly better. They were three times more likely to survive, had less severe sublethal reactions, and—most importantly—did not develop the anaphylactic reactions characteristic of allergies.

But to really see if an allergic pathway explained this difference, the team ran the same test on three different kinds of mice: mice without IgE, mice without IgE receptors on a special type of immune cell called mast cells (which amplify allergic responses), and mice without those mast cells altogether.

Unlike the normal mice, these three mutant mice types—without the key components of allergic reactions—did not benefit from pre-exposure to venom.

“That was pretty exciting for us,” said Thomas Marichal, lead author. “It was the first time we could see a beneficial function for these IgE antibodies.”

A Russell's viper, one of the most dangerous snakes in the world. From Wikipedia user Shyamal

A Russell’s viper, one of the most dangerous snakes in the world. From Wikipedia user Shyamal

But for the hypothesis to hold up, IgE responses would have to be beneficial for much more than just bee stings. So, the team tried the same experiment with venom from Russell’s vipers. Russell’s vipers cause more snakebite incidents and deaths than any venomous snake in the world, and are one of the “big four” snakes of high concern in India. Again, pre-exposure to the venom protected the mice.

“Our findings support the hypothesis that this kind of venom-specific, IgE-associated, adaptive immune response developed, at least in evolutionary terms, to protect the host against potentially toxic amounts of venom, such as would happen if the animal encountered a whole nest of bees, or in the event of a snakebite,” affirmed co-author Stephen Galli.

“This is the first evidence, that we know of, indicating that IgE-associated ‘allergic-type’ immune responses can actually reduce the toxicity of naturally occurring venoms.”

The connection between toxins and allergies may have gone unnoticed because, in modern times, the substances that cause allergies have become less threatening in our daily lives. “We experience allergies in a much cleaner world, where we don’t have the same threats of venomous creatures and potentially toxic food that existed for much of our evolutionary history,” said Galli. “So we’re left with this residual type of reactivity that seems completely mysterious and pointless and harmful.”

If allergies are supposed to be helpful, though, why are some people so prone to anaphylaxis? “Anaphylaxis probably represents the extreme end of a spectrum of IgE-associated reactivity,” said Galli, “which in some unfortunate individuals is either poorly regulated or excessively robust, so the reaction itself can become dangerous.”

Further research is needed to see if the pattern holds with the diversity of toxins that cause allergies. But if it does, it says something very interesting about our evolutionary past. For such a dangerous defense system to have evolved, the threat of deadly toxins—like those from venomous animals—must have created a strong selective pressure. Lynne Isbell once proposed that fear of venomous snakes drove the evolution of primate eyes and brains, tying the very elements that make us human to these infamous animals. If IgE antibodies, too, are a defense against venoms, it would further suggest a very intimate relationship between our ancestors and dangerous animals. Perhaps the biblical tale connecting early humans to snakes is much closer to reality than we thought.

Citation: Marichal T., Starkl P., Reber L., Kalesnikoff J., Oettgen H., Tsai M., Metz M. & Galli S. (2013). A Beneficial Role for Immunoglobulin E in Host Defense against Honeybee Venom, Immunity, DOI:

  • meer atik
  • http://blogs.discovermagazine.com Brent Bestwick

    Great stuff! Always gratifying to see a reason for something that appeared randomly harmful…

    Also, found a typo: the sentence, “The bees previously injected with venom fared significantly better.” should read, “The mice…”

    Thanks for article.

  • Andrew padilla

    I think it’s cool that the ieg antibodies protect are body’s. In some ways it hurts are body’s but in other ways it doesn’t .allergies are a serious thing and no one should ever take them for granted im allergic to bee’s and I know If I get stung it is mot a pretty sight. So don’t take allergies for granted

  • gil

    if a self replicate car cant evolve into an airplan, how can a bacteria can evolve into human ?
    the evolution say that small steps for milions years become a big steps. but according to this a lots of small steps in self replicat car (with dna) will evolve into a airplan.
    but there is no step wise from car to airplan

    • amphiox

      What makes you think a self-replicating car CAN’T evolve into an airplane?

      Have you ever observed a self-replicating car?

      • gil

        because there is no step wise from a car inrto airplan.
        if we will see this kind of car. is this car is evidence for design?

        • amphiox

          How do you know there is NO stepwise sequence from a car into an airplane? Have you tried to figure one out? Have you tested to see if any of the steps of the sequence are in fact possible, or impossible?

          As a matter of fact it is not that difficult to imagine potential stepwise sequences by which a self-replicating car COULD in fact evolve into an airplane. Transitional forms even exist in real life!

          When we examine a car, we infer that it is designed from 3 features. 1) upon disassembly we find that it has no mechanism for self-replication with heritable variation, 2) it’s various parts exhibit lineage chimerism – the steering wheel is an elaboration from a part from ships, the tires are elaborations from bicycles, and so forth, which is something that designers do very often but evolution does only rarely, and 3) we ALREADY KNOW beforehand that humans exist and that humans are capable of and do, design and build cars. Absent the above three features, an alien looking at a car would not be able to conclude that it was an designed artifact and not an evolved organism.

          • gil

            .”How do you know there is NO stepwise sequence from a car into an airplane? Have you tried to figure one out?”-
            yes. you can try to do this with any complex systems that you whant.
            even with a digital hand watch and analog watch.
            try to convert one into another. its impossible in small functional steps. take a car for example. lets say that we have a self replicat material that we need to evolve this into a car. what will be the first step?
            “As a matter of fact it is not that difficult to imagine potential stepwise sequences by which a self-replicating car COULD in fact evolve into an airplane. Transitional forms even exist in real life!”-
            not. this is still need a lots of parts to one into another. even with flying car.
            “When we examine a car, we infer that it is designed from 3 features. 1) upon disassembly we find that it has no mechanism for self-replication with heritable variation, 2) it’s various parts exhibit lineage chimerism – the steering wheel is an elaboration from a part from ships, the tires are elaborations from bicycles, and so forth, which is something that designers do very often but evolution does only rarely, and 3) we ALREADY KNOW beforehand that humans exist and that humans are capable of and do, design and build cars. Absent the above three features, an alien looking at a car would not be able to conclude that it was an designed artifact and not an evolved organism.”-
            so according to all of this. if we will find a new kind of car in another star. a car that can be self replicat with dna. you will think that this kind of car dont need a designer?

          • amphiox

            Show your work. Demonstrate that it is IMPOSSIBLE for a car to become an airplane through a series of small functional steps.

            Just because you with your limited imagination cannot envision such a transition does not mean it is impossible.

            Just declaring that it “can’t” happen does not make that true, no matter how loudly you try to shout it.

            And if we find a “car” on another star that is self-replicating (doesn’t even need dna, by the way), and this self-replication allows for inheritance of variation, then yes, absolutely, said “car” does NOT need any designer.

          • amphiox

            I for example, with no engineering background at all, can easily envision a POSSIBLE stepwise sequence that turns a car into a glider.

            Start with small decorative winglets on the car’s body. The car functions perfectly as a car, so this is functional. The winglets are decorative, but some people like it, making it easier to sell. So it is even advantageous.

            Increase the size of the winglets. At each step, there is decorative appeal. The car remains functional as a car throughout.

            Once the winglet size reaches a certain threshold, the car can function as a weak glider. This means it can jump gaps slightly greater than before. This is a functional advantage. It also means that if it falls off a cliff, it will not fall quite as quickly and sustain less devastating damage on impacting the ground. It will cost less to repair. This too is an advantage.

            Now the gliding function can be improved by gradually decreasing the weight of the car, stripping out unnecessarily heavy padding and components, and replacing some of the heavy steel parts, one bit at a time, with lighter aluminum parts.

            We can further improve the gliding function by slowly, step by step, warping the shape of the car body, making it more streamlined and rounder.

            Through all these steps the car still functions perfectly as a car, and becomes an ever better glider. Meanwhile, the increased streamlining and lighter weight increase its fuel efficiency. Yet more functional stepwise improvement.

            Finally let the wings increase stepwise until it becomes a full fledged glider.

            Next you want to turn the glider into a plane. All you need for that is to give it its own motive power. So you must change the car’s ICE into a prop-propeller plane’s ICE. Not conceptually difficult since the two types of engines are in fact similar enough that they can and have been interchangeable.

            Give a slight rotation to one of the wheels so that as it spins, it will generate a weak force backwards, pushing the glider forward. Not enough to be fully powered, but enough to extend its glide ratio just a little bit. Another functional improvement.

            Complete the rotation stepwise until the one wheel is now facing forward/backward entirely. Stepwise change the shape of the tire rim until it is a propeller. Each step increases the motive force, and even when it is not yet a fully functional propeller, it will still serve to let the glider glide further. More fully functional advantageous stepwise changes.

            Then you keep the remaining three wheels as landing gear, and you have a prop plane.

            A trained engineer could probably provide an even more likely scenario to turn a internal combustion car engine into a prop propeller engine, or even turn the ICE/grill/exhaust system stepwise into a jet engine.

            Or maybe even simpler, turn it into a gasoline-powered rocket engine.

            Of course the scenario is dependent on certain environmental factors. It will need a population of car buyers who find little winglets on cars attractive for the early steps. And it will need a population of car buyers who value fuel efficiency and aerodynamic shapes over plush and heavy luxury acroutements for the middle steps.

            But that is true with real world evolution too.

          • gil

            its not so simple. you must stuck in some point. because even the engine of the airplan is diferent from the engine of the car. so you need to change the engine in small steps. its impossible.

          • amphiox

            It is rather presumptious of you to declare something “impossible” without even bothering to try to find out if it actually is.

            One more time I ask you: SHOW YOUR WORK.

          • gil

            but according to the evolution a very diferent kind of systems evolve from each ohter. so we will need to show how a car engine change into airplan engine. lets say even something much simple. can you add a gps to a car in small fucntional steps?

          • amphiox

            A gps can very easily evolve in small functional steps from a car radio.

            The onus is on YOU to show that it is IMPOSSIBLE, not on me to demonstrate every imaginable small step. All I need to do is show that it is possible, and I have already done so.

            SHOW YOUR WORK, please.

            (I notice that you still don’t realize that the ORIGINAL REAL LIFE rotary propeller AND jet engines did in fact evolve by small steps from the original internal combustion engine.)

          • gil

            hi again. you said:
            “A gps can very easily evolve in small functional steps from a car transistor radio.”-

            how? there is no step wise from radio to gps. and even i so- what if the original car dont has radio? and what about the radio? its is also evolve step wise from what?

            “(An evolved GPS system likely won’t have a display, though. A display does not benefit the car/GPS system, only the driver.”-
            yep. but the human is also part of nature. so the question is what the use of part gps.
            do you think that a robot with dna itsnt evidence for design?

          • amphiox

            Once more you claim an impossibility without even trying. The onus is ON YOU to SHOW YOUR WORK and demonstrate that it is impossible.

            Again and again you refuse to do so.

            I suspect it is because, deep down, you know in your heart that you are wrong.

          • amphiox

            If the robot can self-replicate, that is not evidence for design at all.

            In fact, lifeforms on earth can be said to BE self-replicated nanotech robots with DNA.

            Enough of this, you are moving the goalposts. First a plane, then a gps, now a robot with dna? When you still haven’t addressed the requirement that you present actual evidence for the hard claim of impossibility you originally made.

            This is called moving the goalposts, and it is a disgustingly intellectually dishonest way to behave.

            We are done with this conversation until you SHOW YOUR WORK. There shall be no response for me, and nothing to talk about until you DEMONSTRATE hard EVIDENCE for your claim of impossibility.

            Reality does not have to abide by your personal incredulity and inability to envision step-wise processes.

          • amphiox

            1. There is no law of engineering or physics that says a plane and a car cannot have the same kind of engine. A car engine in a plane wouldn’t be the most efficient or effective engine, but anything that can spin a propeller would still work, and a car engine can certainly do that. The plane would be an imperfect plane, but IT IS STILL A PLANE.

            2. The M1 Abrams tank famously uses a jet engine instead of a regular motor engine. So I could turn that multi-tonne tank into an airplane even easier than I can turn a car into one.

            3. There is no law of engineering or physics that says you cannot turn any component into any other in a stepwise process. Worst case scenario I will move the atoms one at a time, one planck length at a time. Hard? Sure. Impossible? Not without further EVIDENCE (not assertion) from you.

            4. There is no law of engineering or physics that says I HAVE to turn the ENGINE of the car into the ENGINE of the plane. I could just as easily turn a different component of the car into the engine of the plane, step by small step, while simultaneously REMOVING the engine of the car, step by small step. And while the airplane engine is being built, step by small step, it doesn’t even have to function. It can be a decorative ornament on the roof that spins a little propeller to attract customers who like spinning gizmos, while the whole thing still works as a car using its other engine.

          • Voice of reason

            you are arguing you can’t think of a design path to get from one to the other, I agree that intelligent design is difficult to explain.

        • jh

          What about the possibility that airplanes and cars will undergo convergent evolution – that is, evolve to be one and the same thing?

          Look what happened with computers and phones. Compare a computer and a phone circa 1960. They didn’t look the least bit alike, nor did they serve a similar purpose, have similar functions, or even use the same types of electronics.

          Who’d have imagined in 1960 a “stepwise” change of a phone into a computer or vice-versa?

          • gil

            hi/ look my coment above. you made my point.
            have a nice day.

          • amphiox

            No he didn’t. He utterly destroyed your point. But you just haven’t realized it.

          • http://blogs.discovermagazine.com Longmire

            I know that was a long article to read but it was actually about allergies and the bodies response to them. So your for “Intelligent design” ok so why was a bee (which was obviously created to pollenate our food crop) created to be deadly to us, what can that teach us, and how does that lead to truth.

    • Thianar

      Not knowing all details on how life started doesn’t mean that you should jump to a wild conclusion (Intelligent design).
      It just means that we need more research and discussions . One size fits all final declarations offered by religious zealous of all origins should stay away from science centers.

      • gil

        so what if you will see a self replicat robot with dna?

  • Kevin Bonham

    So, I have a lot of problems with this explanation.

    We know several circumstances where IgE antibodies are useful, e.g. parasitic worm infections. This is just one other example where they’re useful. This does not suggest allergic reactions are a good thing, only that IgE reactions are a good thing – this is not surprising, else why would they have evolved?

    Allergic reactions are to type 2 immunity as autoimmune disorders are to type 1 and type 17 immunity. We know IgG is used for all kinds of things, like defenses against viral infections. We also know that rheumatoid arthritis is largely mediated by IgG immune complexes. This does not imply arthritis is somehow a good thing, just because it’s mediated by something that does other good things.

    The observation that humans not living under the ultra-sanitized conditions of modern life do not get allergies is further evidence that they are an abnormal, not an adaptive (in the evolutionary, not immunological sense) response. Our immune systems have been evolutionarily calibrated for exposure to microbes and parasites, and when we don’t encounter those things, some of the regulation can be thrown off.

    I’m not saying this finding is wrong, the data in the paper look pretty convincing. But the interpretation that the purpose of “allergic responses… is to protect the host against venoms and other environmental toxins,” seems overblown. If we had uncovered the mechanisms behind rheumatoid arthritis before antiviral immunity, would be say that “arthritic responses evolved to protect against viruses?”

    Allergies are a pathogenic over-reaction of type 2 immunity. Type 2 immunity may have evolved to combat toxins, but there’s no need to dress it up further than that.

  • Holly Harris

    I both agree and disagree with this article. The body does not recognize allergens as foreign like they do pathogens so it does have a different response to allergens. IgE’s are also released when the allergy, such as dust or pollen, enters the body. However, I do not think this saves your life. Having high IgE levels can harm the body, for example you can get Job’s Syndrome. I do think that if you have minor allergies it is a good thing because it keeps your immune system activated and turned on. People are so focused on having clean air, water, and an environment that our body is not getting the built up immune system that we need to fight off illnesses later in life.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Science Sushi

Real Science. Served Raw.

About Christie Wilcox

Christie Wilcox is a science writer and PhD Student at the University of Hawaii, where she studies the protein toxins in venomous fish. She is renowned in the science blogosphere for her delicate balance of contemporary science and scientific perspective seasoned with just the right amount of wit. Her award-winning posts have been featured in The Open Laboratory: The Best Science Writing On Blogs four years running and landed on the pages of major media outlets including The New York Times and Scientific American. To learn more about her life and work, check out her webpage or follow her on Twitter, Google+, or Facebook.


See More


@NerdyChristie on Twitter

Collapse bottom bar

Login to your Account

E-mail address:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »