How do ants stop diseases spreading?
We have all been there. You have just sat down, perhaps on a train or in a lecture, when the person next to you sneezes: they have a cold. You grimace and hope for the best. But, sure enough, by the end of the day you begin to notice the tell-tale signs that you have caught their cold and are falling ill yourself. Though a typical scenario, I am sure we have all unwillingly played our part in many times. This story highlights an important issue faced by human societies and those of other animals that live in groups: infectious diseases...
Infectious diseases, usually bacterial, viral or fungal in origin, reproduce by infiltrating the defences of a host – be that a human or some other creature – replicating within their body, and then transmitting themselves to other hosts nearby. Termed a parasitic life cycle, the process of jumping from an infectious to a healthy host is more likely to be successful if the potential victims are in close, regular contact with one another. In theory, as the number of hosts and contacts between them increases, the easier it is for pathogens to spread. And if the spread is great enough that a significant portion of a host population falls ill, the disease is said to have caused an epidemic. A good example is the bubonic plague, which killed tens of millions of people in Europe during the 14th Century.
Although social animals might be at greater risk of catching diseases from others, this is not to say that they are constantly suffering epidemics. In response to the selection pressure applied by infectious diseases, social animals have evolved countermeasures to reduce their chances of infection. For example, many animals will avoid sick individuals and humans are able to recognise subtle differences in facial features to tell when someone is under the weather. In addition, some social animals will participate in mutual grooming to remove pathogens and there is even evidence of animals administering medicine to one another.
Perhaps the most sophisticated social disease defences, aside from our own modern medicine, are found in the social insects: the wasps, bees, ants and termites. These creatures quietly dominate most ecosystems on earth, forming colonies that range from just a few individuals to super-colonies containing tens of millions of individuals and spanning several kilometres. Like human societies, there can be thousands of contacts between individuals on a daily basis, from brief touches of the antennae to the intimate regurgitation and sharing of food. This creates a vast network of interactions that results in a complex, coordinated social behaviour. Yet, this network can be hijacked by pathogens to spread between social insect hosts.
Living in crowded, busy conditions, social insects at first glance appear to be particularly susceptible to epidemics, but such events are rare. In fact, attempts to eradicate social insect pests using biological controls – for example, exposing a termite infestation to fungal pathogens – have failed time and again. This is because social insects have evolved social immunity. This is a layer of protection that functions alongside the individual defences of each insect and shields the colony against disease. It emerges through the cooperation of all colony members and can be thought of as the colony’s “immune system”. Most often, social immunity behaviours aim to prevent infection. For example, colonies will incorporate antimicrobial tree resin into their nests, workers clean themselves before entering sensitive areas of the nest and the dead are carefully and meticulously disposed off, all to reduce the likelihood of a pathogen invading. However, less is known about what happens when these first lines of defence fail. If a nestmate falls sick, how do social insects prevent it spreading and causing an epidemic?
To answer this question, we worked with the invasive garden ant and exposed their cocooned pupae to the spores of a fungal pathogen. When ants encounter contaminated brood, they perform sanitary care: meticulous grooming and the application of their antiseptic poison to prevent infection. This poison, comprised mostly of formic acid, is emitted from a small opening at the tip of their abdomens, termed the acidopore. The same poison is sprayed at would-be predators coming too close to the nest, but during sanitary care, the ants suck this poison up into their mouths and spit it back out in small, carefully applied dosages whilst grooming.
Sanitary care prevents infections by removing and sterilising the fungal spores before they can germinate and grow in the body of the brood, like a seed sending roots into soil. However, we noticed that some contaminated pupae were removed from their cocoons within a few days of pathogen exposure, behaviour that we called “unpacking”. Initially, we assumed that, by taking off the cocoons, the ants were trying to remove the spores we had applied. Yet, when we filmed the ants to study unpacking in more detail, we were astonished to find that instead of caring for their brood, the ants were doing quite the opposite.
Working methodologically and using their sharp mandibles, the ants began by cutting a pupa out of its cocoon. They then started chewing off the pupa’s limbs and making incisions in its body. At the same time, the ants sprayed their poison, directly from the acidopore, into the wounds they created and onto the severed limbs. As if this behaviour was not startling enough, the majority of pupae were still alive when the ants began dismembering them. To reflect its fatal and noxious nature, we called this sequence of events “destructive disinfection”.
By taking pupae away from the ants just as this process began, we found that the majority were infected: the fungus successfully grew out of their bodies and sprouted new infectious spores within a few days. We also isolated pupae that the ants had not unpacked and found that they were typically disease free ¬– the ants only destroyed pupae that already have infections. However, at the time of unpacking, infected and non-infected pupae look the same. How then are the ants able to tell who is sick and who is not?
We speculated that the ants might smell when a pupa has fallen ill. This might sound far-fetched, but ants rely heavily on chemical odours to communicate with one another, and many animals, including lobsters, frogs and humans, also use scents to tell the difference between healthy and sick individuals. To put this hypothesis to the test, we dipped infected pupae into a chemical bath that, just like when we bathe, removed any odours sticking to its body. When these pupae were placed back with the ants, they were no longer unpacked, indicating that the aromas triggering destructive disinfection had been removed.
Using an approach known as gas chromatography-mass spectrometry, or GC–MS for short, we then analysed the leftover “bath water” from the washes to see what was in it. The GC–MS analysis showed that infected pupae have higher amounts of four chemicals compared to healthy pupae and those that had been exposed to the pathogen, but were not yet sick. Surprisingly, these chemicals were not a fungal odour, but were part of the normal smell of the pupae, only amplified. What is more, they seem to be triggered by an immune reaction to the fungus: pupae that we injected with the building blocks of a fungal cell wall – mimicking an infection but not actually causing one – also had higher amounts of two of these smells. After reading more about them, we realised that both of these chemicals are also elevated on sick honeybees, suggesting that social insects may use a common set of odours to identify sick nest mates.
Having worked out how the ants tell who is sick, we wanted to know why they then perform destructive disinfection. To answer this question, we carried out a simple experiment, where contaminated pupae were kept alone or with a group of ants. Pupae alone did not receive any destructive disinfection and so, unsurprisingly, almost all became infectious within a few days. However, when pupae were kept with ants and did receive destructive disinfection, the number that became infectious was drastically reduced by 95%. It is clear then that by performing this behaviour, the ants are not only killing the pupae but are eradicating the infection along with it. To understand destructive disinfection in more detail, we decided to become ants ourselves: using forceps to peel away the cocoon and micro-scissors to surgically remove the pupae’s limbs and injure its cuticle, we mimicked the unpacking and wounding the ants perform during destructive disinfection. In addition, we created a synthetic ant poison that we applied to pupae in place of poison spraying. By mixing these “behaviours” in different combinations, we were able to investigate how each contributes to the overall extermination of the fungus.
Regardless of what other behaviours were performed, the fungus always grew out of a pupa’s corpse if poison spraying was missing. In other words, the poison is the active antimicrobial ingredient that prevents the fungus growing. However, the other behaviours still play an important role: the poison only worked if the pupa was first unpacked and its body damaged. This is because, firstly, the cocoon acts like a raincoat, keeping the poison out and from touching the pupae inside. Secondly, because the fungus is growing within the pupae at the time of destructive disinfection, removing the pupa’s limbs and wounding its cuticle allows the poison to penetrate the pupa’s body more easily, in the same way that marinating a piece of meat works better if it is first pierced several times. By ensuring that the poison passes through both the cocoon and cuticle of the pupae, the ants are delivering their poison directly to the site of infection, inhibiting fungal growth from the inside out.
In a final experiment, we asked how important destructive disinfection is and whether, really, a single infectious pupa is that harmful to the overall health of the colony. We kept groups of ants with either a destructively disinfected pupa or a pupa that had already become infectious and so was covered in newly contagious spores. This latter situation simulates what would happen if the ants failed to detect and eradicate the infection. Whilst ants housed with a disinfected pupa survived perfectly fine, those kept with an infectious individual did not fare well. Over 40% of the ants contracted an infection and died within a few days. We removed these dead ants before they became contagious to confirm infection, but work by other researchers has shown diseased ants left in the colony quickly infect others, and those individuals then spread their infection to even more ants, resulting in a snowball effect and the breakout of an epidemic that leads to colony collapse.
Clearly, detecting infections early and treating them before they can spread is as important in preventing epidemics within ant societies as it is in humans. But here the ants are acting less like humans and more like immune cells in a body. We can think of social insect colonies are being a single organism, made up of lots of individual organisms behaving and functioning as one, in the same way that our bodies consist of lots of cooperating cells. These “superorganisms”, just like our bodies, need to prevent pathogens from invading and making them ill. When a pathogen attacks one of our cells, immune cells detect the threat and respond by piercing holes in the infected cell to administer chemicals that kill it, along with the pathogen. Destructive disinfection in ants is remarkably similar and has likely evolved for the same reason: the need to protect the whole over its parts. This means that individuals, be they insects or cells, are sacrificed when necessary to ensure that the colony or body continues to survive and reproduce. Our work, therefore, raises interesting questions about what common evolutionary processes were at play during the emergence of multicellular organisms and superorganismal insect societies...