- Email: [email protected]
Social immunity in insects
occur in other species (although female spotted hyenas might beg to differ because of their high rates of maternal morbidity and infant mortality in first pregnancies). Explanations of human exceptionalism in this regard commonly invoke consequences of bipedalism and large neonatal skulls (we do take inordinate pride in our large brains), but cephalopelvic disproportion is a less important cause of human maternal mortality than postpartum hemorrhage, puerperal sepsis, and preeclampsia. One unusual feature of human birth was the frequent presence of helpers who assisted mothers and newborns. The survival of most mammalian infants is absolutely dependent upon their mother being able to care for them immediately after birth and this selective premium on maternal health will have acted as a constraint on how much fetuses could demand from their mothers during pregnancy. An evolutionary history of birth attendance may have contributed to the difficulties human mothers now experience during pregnancy because the survival of a baby was no longer tightly dependent on the rapid recovery of its mother after birth. Help was available to support both mother and infant. As a consequence, the indirect costs to babies of increased demands on mothers during pregnancy were relaxed in the human lineage and fetuses responded evolutionarily by increasing their demands. FURTHER READING Abrams, E.T., and Rutherford, J.N. (2011). Framing postpartum hemorrhage as a consequence of human placental biology: an evolutionary and comparative perspective. Am. Anthropol. 113, 417–430. Boddy, A., Fortunato, A., Sayres, M.W., and Aktipis, A. (2015). Fetal microchimerism and maternal health: a review and evolutionary analysis of cooperation and conflict beyond the womb. Bioessays 37, 1106–1118. Elliot, M.G., and Crespi, B.J. (2009). Phylogenetic evidence for early hemochorial placentation. Placenta 30, 949–967. Haig, D. (2007). Putting up resistance: maternal-fetal conflict over the control of uteroplacental blood flow. In Endothelial Biomedicine, W.C. Aird, ed. (Cambridge: Cambridge University Press). Haig, D. (2010). Fertile soil or no man’s land: cooperation and conflict in the placental bed. In Placental Bed Disorders, R. Pijnenborg, I. Brosens, and R. Romero, eds. (Cambridge: Cambridge University Press). Villie, P., Dommergues, M., Brocheriou, I., Piccoli, G.B., Tourret, J., and Hertig, A. (2018). Why kidneys fail post-partum: a tubulocentric viewpoint. J. Nephrol. 31, 645–651.
Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA 02138, USA. E-mail: [email protected]
R458
Primer
Social immunity in insects Sylvia Cremer When animals become sick, infected cells and an armada of activated immune cells attempt to eliminate the pathogen from the body. Once infectious particles have breached the body’s physical barriers of the skin or gut lining, an initially local response quickly escalates into a systemic response, attracting mobile immune cells to the site of infection. These cells complement the initial, unspecific defense with a more specialized, targeted response. This can also provide long-term immune memory and protection against future infection. The cell-autonomous defenses of the infected cells are thus aided by the actions of recruited immune cells. These specialized cells are the most mobile cells in the body, constantly patrolling through the otherwise static tissue to detect incoming pathogens. Such constant immune surveillance means infections are noticed immediately and can be rapidly cleared from the body. Some immune cells also remove infected cells that have succumbed to infection. All this prevents pathogen replication and spread to healthy tissues. Although this may involve the sacrifice of some somatic tissue, this is typically replaced quickly. Particular care is, however, given to the reproductive organs, which should always remain disease free (immune privilege). Similarly, when an ant colony is infected, the contaminated or infected ants, along with their healthy nestmates, fight the infection collectively. A colony of social insects is hence protected by both the individual defenses of its members and a systemic, colony-wide response, providing the colony with a protection known as ‘social immunity’. Individual defenses comprise behaviors that prevent pathogen contamination (e.g., pathogen avoidance and selfgrooming) and those that fight infections. The latter utilizes the innate immune system, which is capable of raising specific and long-lasting responses, or immune memory (‘immune priming’). In addition, entry of pathogens into the colony and infection of individuals is quickly detected by nestmates. These
Current Biology 29, R425–R473, June 3, 2019 © 2019 Elsevier Ltd.
ants will try to remove contaminations from the nest, or from contaminated nestmates, to prevent infection. In cases where infections cannot be prevented, sick individuals may either remove themselves from the colony or are isolated by other ants. Similar to immune cells targeting infected cells, social immunity may also involve the sacrifice of infected colony members. This prevents pathogen replication and spread of pathogens to healthy colony members. Additionally, there is extra protection directed at the queen, the sole reproducing member of the colony. Metazoan bodies and the colonies of social insects (the social bees and wasps, the ants and termites) are hence protected from disease by strikingly similar principles, despite the fact that they represent very different types of biological organization. A central feature common to both, however, is the presence of a distinct germline and soma. Both are interdependent and require one another for reproduction. In an insect colony, workers are typically sterile themselves, but rear the eggs laid by the queen. Successful colony reproduction relies on both tasks. Even if cells can survive in cell culture when isolated from the body, and individual ants can survive when removed from their colony, reproduction – that is, the formation of new bodies or new colonies – requires cooperation between the subunits of the whole reproductive entity. Due to these organizational similarities, insect colonies are often referred to as ‘superorganisms’ (Figure 1). Just like organisms, superorganisms fight disease in a stepwise manner, meaning they employ distinct mechanisms to fight pathogens at the different steps of disease progression. Infectious particles enter the body or colony, establish and begin to spread around the host. This occurs by infecting a cell or colony member, replicating inside them, and then producing transmissible stages that spread to others. Host defenses have evolved to try to break this process at each step. If host defense is successful, the cycle is interrupted and the infection will die out; if not, disease progresses and has to be fought at the next step (Figure 2). Collective nest hygiene Even in the absence of pathogens, social insects keep their nests
meticulously clean of any potential sources of infection (Figure 2, Step 1). For example, when garden ants move into a new nest, they treat the nest material, including the new brood chambers, with a self-produced disinfectant, a formic acid-rich poison. Other species, like wood ants and honeybees, use antimicrobials collected from the environment to enrich their nest, with antimicrobial tree resin being the most commonly used. Termites incorporate their own feces into their nests, which, at first sight may seem unexpected. Yet, as they contain a lot of microbiota that produce antimicrobials, termite feces are not only a good building material, but also ensure cleanliness. Waste is removed before it can accumulate and form a niche for fungi, bacteria, or parasites such as mites to grow. Similarly, dying nestmates are removed from the nest. All corpses and debris are brought to specific graveyards or middens, located either in peripheral nest chambers or outside the nest. Waste removal is particularly sophisticated in the massive colonies of leaf-cutter ants, which process enormous masses of leaf material on a daily basis, to rear their symbiotic fungus as food. These fungal crops are continually weeded and treated with fungicides to remove errant, parasitic fungi, which would otherwise wipe out the ants’ monoculture. Honeybees, faced with nest contamination, employ a tactic used by our own bodies: they increase the temperature. This social fever is the result of the bees shivering their flight muscles, and their collective effort increases the overall hive temperature. This behavior has been shown to kill heat-sensitive pathogens. Nest hygiene is often a general precaution that is employed prophylactically and non-specifically. It relies on the mechanical removal of potentially infectious material, often combined with the application of antimicrobials with a broad spectrum of activity. Nest hygiene hence plays a similar role to the general grooming and prophylactic use of disinfectants (e.g., medical anointing in monkeys), observed in many species across the animal kingdom. Sanitary care If an individual is contaminated with a pathogen, either due to inefficient
Organism
Superorganism
Soma
Germline Current Biology
Figure 1. Insect colonies as superorganisms. Insect colonies form a reproductive entity consisting of one or a few reproductive individuals (the queen(s), pink), and their sterile workers (blue), which rear the brood and maintain the colony. Colony reproduction is dependent on the interaction of both these functional subunits, similar to the interdependency between the germline (reproductive organs; pink) and the somatic tissue (blue) in a multicellular, metazoan organism. Due to this analogous functional division of insect colony members into germline (queen) and soma (workers), insect colonies are often termed superorganisms. Graphics by ppm-vi.
nest hygiene or from foraging outside, sanitary care functions as the second step in the colony’s social immune response (Figure 2, Step 2). Similar to delousing monkeys, honeybees remove ectoparasites, such as the Varroa mite, by grooming their nestmates. This mite – the scourge of beekeepers – is not only damaging due its feeding on the blood (hemolymph) of the bees, but because it is also a vector of many diseases, including several viral diseases. Ants and termites also groom fungal spores from the body of contaminated nestmates, usually before the spores can penetrate the cuticle and cause internal infection. This allogrooming is more effective at preventing infection than selfgrooming, as additional body parts can be reached, such as the thorax. Garden ants combine the mechanical removal of infectious particles with the application of antimicrobials, increasing their chances of effectively preventing infection. Grooming is one of the most common sanitary care behaviors in social insects, and is very effective. Yet,
does allogrooming increase the risk of infection for groomer? Grooming ants collect the infectious material they remove into pouches inside the mouth, termed the infrabuccal pockets. Here, it is compacted and sterilized using antimicrobial gland compounds. The compacted pellets are eventually ejected and have an extremely reduced ability to germinate. Despite these measures, pathogen transfer can still occur from the contaminated individual to its nestmates. However, although many nestmates contract infections, only a small proportion of these cause disease. This is because only a small fraction of pathogen is transferred to each ant, leading to non-lethal, lowlevel infections. Low-level infections have been shown to trigger a protective immunization, in both ants and termites. This is similar to the use of low-level infections in early human medicine, before modern vaccination was developed. Known as ‘variolation’, it was used to immunize people against diseases such as small pox.
Current Biology 29, R425–R473, June 3, 2019
R459
Care–kill dichotomy Together, nest hygiene and sanitary care are effective measures to prevent pathogens from establishing within the colony. Most studies of social immunity describe examples of these early steps to stop disease, but they are not always fully effective. What happens when these initial defenses fail to prevent infection and disease takes hold? In these situations, social immunity switches from preventative measure aimed at protecting the whole colony and all its members, to preventing the replication and transmission of disease. These latter measures focus on the exclusion or elimination of infected individuals, akin to immune cells targeting infected cells in a body. Hence, once an infection has developed to the point of no return, social immunity switches from a ‘care’ to a ‘kill’ strategy. For the colony, the outcome is similar — for the infected individual, they face being either rescued or sacrificed for the colony. How do social insects resolve this apparent conflict of interest between the individual and the colony? In superorganisms, fitness is not measured at the level of the individual, as no single ant can reproduce independently. Instead, the fitness of each individual depends entirely on the overall fitness of the colony. If removing infected individuals to prevent disease
Collective hygiene Step 1: Contamination
Sanitary care
Step 2: Infection
Infection elimination
Care–kill breakpoint Step 3: Replication
Network modulation
Step 4: Transmission
Reproductive entity
Pathogen
Subunit
Cooperative defense Current Biology
R460
Current Biology 29, R425–R473, June 3, 2019
Figure 2. Disease stage-dependent course of social immunity. Social immunity acts to interfere at each possible step of disease progression (steps 1–4). To protect the reproductive entity from disease, distinct cooperative defense mechanisms are employed at each step (green upward-bent arrows). If they fail, the disease progresses to its next step (black downward-bent arrows), and the pathogen (blue diamonds) may be able to infect individuals (green circles), replicate (multiple diamonds), and transmit to new colony members (small green arrows). Following disease progression, the mechanisms of social immunity involve collective nest hygiene (e.g., ants spraying antimicrobial poison into their nest), sanitary care of contaminated individuals (e.g., removal of infectious particles by allogrooming), elimination of the infection (e.g., application of disinfectant after piercing the cuticle), and modulation of the social interaction network of the colony (by behavioral changes of colony members). Note that social immunity actions progress from initial steps that aim to protect the colony by protecting its individual members, to protecting the colony by preventing disease transmission. The latter involves the sacrifice of infected individuals, indicated by the ‘care–kill breakpoint’. Graphics by ppm-vi.
spread can maximize fitness, it will be to the benefit of not only the healthy portion of the colony, but also that of the sacrificed individual. Being sterile, social insect workers cannot reproduce directly, and hence they gain their fitness through caring for their relatives, with whom they share a significant portion of the same genes. Any action that increases the chance of their own genes being passed on into the next generation – be that sanitary care or being sacrificed due to infection – will indirectly increase their own inclusive fitness. We should therefore expect social insect workers to signal honestly about their health status, regardless of disease progression. That is, they should signal for help when contaminated and signal for elimination when infected. As in the metazoan body, where cells signal about their status to the immune system, we should expect superorganisms to have evolved similar mechanisms that protect the reproductive entity over its individual parts. What are these mechanisms (Figure 2, Step 3)? Infection elimination It has been known since the 1960s that honeybees can detect infected brood through the sealed brood comb, uncap it, and then drop the brood to the ground away from the hive. The bees performing this behavior have highly sensitive olfaction and have even been artificially selected for this hygienic behavior, in the hopes of combatting some honeybee diseases. More recently, this removal behavior has also been described for a number of ant species. Unlike bees, ants pool their brood in open piles, where they can single out infected individuals and remove them from the brood chamber. However, unlike the winged honeybee, simply dropping infectious brood outside the nest would run the risk of re-infection, since ants forage in the same territories they place their dead, if not even placing them in chambers inside the nest itself. To overcome this problem, garden ants have evolved a complex multicomponent behavior to rid colonies of brood infections. First, they slice open an infected pupa’s cocoon, and then bite and pierce its soft cuticle. Having made a wound, the ants then bend their abdomen over the pupae and emit a poisonous spray, rich in formic acid. Together, this behavior ensures the
poison can enter the infected pupae, where it disinfects the host from the inside out. This effectively prevents pathogen replication before new transmissible stages can be produced. This process of ‘destructive disinfection’ is not only functionally similar to the elimination of infected cells in a body, but also mechanistically equivalent. For example, infected cells can be made porous through perforin proteins, which make holes into which cytotoxins are injected. Infected cells emit a ‘find me, eat me’ signal that attracts immune cells towards them. Similarly, the infected pupae have an altered chemical odor that is detected by the workers. In both cases, this allows for the accurate detection of the sick cells or individuals, respectively, and prevention of collateral damage of healthy tissue or colony members (‘immunopathology’). Whilst the brood is immobile and requires active removal by workers, adult insects can leave the nest of their own accord. It has indeed been found that sick or generally moribund ants leave the colony shortly before death. The basis for this behavior is not yet well understood, but seems to include a loss of the insects’ ability to integrate social cues. Nest departure may also be prompted by the healthy nestmates, with honeybees expressing aggression towards infected individuals. In termites, initial grooming by nestmates can sometimes escalate to aggressive grooming, that then leads to cannibalism, and hence the ‘removal’ of the infected individual, via the antimicrobial environment of the termites’ gut. Modulation of the social interaction network Whilst the cells in a body are highly connected and embedded in an extracellular matrix, insects in a colony can move freely. This means they can flexibly adjust their interactions with the other colony members to potentially reduce disease transmission. Epidemiological modeling forecasts disease spread through a network of hosts, which is typically visualized as dots representing individuals (nodes) and interconnecting lines (edges) showing their contacts. These models predict higher disease spread in simulated colonies that lack a structured network, as if all individuals interacted randomly,
compared with those with structure. In fact, social insects express structured interaction networks, due to a clustering according to the tasks they perform, and where these are carried out in the nest. Most strikingly, young workers, the nurses, take care of the brood and the queen in the center of the nest, whereas older workers leave the nest to forage. This leads to an inherent network structure that may have, in part, been shaped by selection to limit infectious disease transmission, as predicted by the ‘organizational immunity hypothesis’. Is it possible for social insects to enhance their constitutively expressed transmission-inhibiting network properties even further upon pathogen exposure? Experimental evidence for induced organizational immunity by behavioral modulation has indeed been recently observed (Figure 2, Step 4). Pathogencontaminated garden ant foragers were found to change their spatial preferences, staying outside the nest for longer and interacting less with the nurse ants in the center of the network. In response to forager contamination, the nurses brought the brood even closer to the nest center. This led to an even stronger disassociation between foragers and the nurses, resulting in a reduced risk of infection for the nurses and queen. Pathogen spread was thus highest towards other individuals of the same age- and task-group – other foragers – whereas most nurses and queens received only low pathogen amounts, which typically cause immunization rather than disease. Host susceptibility Whether or not hosts develop disease when encountering a pathogen depends on the number of infectious particles and the hosts’ disease susceptibility. Hosts can differ greatly in susceptibility, which is an inherent property and can also be altered by social context and infection history. Further, pathogen strains are often more or less virulent against different host genetic backgrounds. In social insects, colony members are usually highly related, resulting in a correlated susceptibility to pathogens. Genetic similarity across colony members is highest in colonies headed by a sole singly mated queen. This means that a pathogen strain able to infect one
Current Biology 29, R425–R473, June 3, 2019
R461
Totipotent society
Superorganismal society
All individuals
Soma
Germline only
Germline
Reproduction Current Biology
Figure 3. Totipotent versus superorganismal societies. In societies of totipotent individuals (nodes: circles; edges: black lines), each society member integrates germline (pink) and soma (blue), so that all individuals can reproduce to form a next generation. In superorganismal societies, members belong either to the germline (the queen(s)) or the soma (sterile workers). At each reproduction event, the network collapses to the single queen state, which then builds up a new society by its sterile worker offspring. Graphics by ppm-vi.
colony member is likely able to infect all others in a colony. Multiple mating in queens or multiple queens reproducing in a colony has been suggested to evolve so as to reduce susceptibility to disease, by increasing a colony’s genetic diversity. Still, the overall high relatedness in social insects increases the risk of colony collapse. Disease susceptibility can change dynamically over an individual’s lifetime. It can depend on an animal’s overall condition (e.g., nutrition, stress), but also its infection history. As insects show both individual and transgenerational immune priming, previous infections earlier in life or in the parents can induce protective immunization. Interestingly, in social insects, changes in disease susceptibility induced by previous pathogen contact seem to affect how they care for nestmates in the future. For example, garden ants adjust their grooming and disinfection behaviors in a risk-averse manner, reducing their risk of contracting diseases they have an increased susceptibility to. R462
Overall, social and individual defenses are tightly interconnected and the overall outcome of disease dynamics in the colony will be shaped by how effectively sanitary care and infection elimination reduce pathogen numbers in the colony, how easily the remaining pathogens can spread across the social interaction network, and the disease susceptibility of the colony. Evolution of social immunity Since the introduction of the concept of social immunity, we have gained significant insight into how cooperative disease defenses function. Yet, we still know little about how they evolve. Some components of social immunity, such as collective hygiene and sanitary care, are present in many forms of social life, including parent–offspring groups where the young disperse, and in communities of unrelated individuals. However, the extent to which individuals engage in collective disease defenses, and so risk contracting infections themselves, will be dependent on their relatedness to others in the group, their own future
Current Biology 29, R425–R473, June 3, 2019
reproductive potential, and the severity of the pathogen in question. Moreover, it is probably unlikely that there will be selection on sick animals to actively signal when they are ill to others – except in groups of relatives – as this risks expulsion from the group. Indeed, there is ample evidence that infected conspecifics are excluded and selection seems, in some cases, to have therefore led to the hiding of illnesses from others. Whilst unconditionally sterile individuals in social insect colonies must behave altruistically to maximize their inclusive fitness, individuals in other types of social groups that currently reproduce, or will in the future, are most likely to act selfishly to maximize their own direct fitness. With the exception of superorganisms, where the conditions for social immunity to evolve are met, predicting when, or if, social immunity can evolve in other social groups is not trivial. This has thus far challenged our comprehension of which social systems, apart from superorganisms, will express the full repertoire of social immunity, including elimination of infected individuals. It is therefore necessary to study collective disease defense in non-superorganismal eusocial, or subsocial insects to learn more about the evolution of social immunity in insects. Conclusion Given their superorganismal biology, which perhaps has more in common with a metazoan body than other animal groups (Figure 3), can social insects still serve as good model systems to understand the principles of disease defense in other societies? The answer appears to be yes. Although superorganisms can only reproduce as a whole, whereas each member in an animal society can reproduce independently (each being a ‘totipotent’ organism; Figure 3), social insects are still groups of freely interacting animals. This means, from an epidemiology perspective, they behave similarly to conventional societies. Therefore, the same principles about how network properties affect pathogen spread between interacting hosts apply as equally to the social insects as they do to humans. Improving our epidemiological understanding is relevant for better
disease management in a variety of fields, including, for example, livestock protection and outbreaks in human societies. To date, epidemiological research relies mostly on correlational data that are gathered during or following outbreaks. Research on insect societies, however, allows for experimental prospective studies with full surveillance of the social interactions and their effect on pathogen spread. This even allows us to test how manipulation of the network properties, or the individual behaviors, affect disease outbreaks. This may provide valuable data to advance current epidemiological modeling, and hence develop more realistic prevention strategies against disease outbreaks in social groups. FURTHER READING Boomsma, J.J., and Gawne, R. (2018). Superorganismality and caste differentiation as points of no return: how the major evolutionary transitions were lost in translation. Biol. Rev. 93, 28–54. Cremer, S., Armitage, S.A.O., and Schmid-Hempel, P. (2007). Social immunity. Curr. Biol. 17, R693– R702. Cremer, S., and Sixt, M. (2009). Analogies in the evolution of individual and social immunity. Phil. Trans. R. Soc. B 364, 129–142. Evans, J.D., and Spivak, M. (2010). Socialized medicine: individual and communal disease barriers in honey bees. J. Invertebr. Pathol. 103, S62–S72. Hughes, W.O.H., Eilenberg, J., and Boomsma, J.J. (2002). Trade-offs in group living: transmission and disease resistance in leaf-cutting ants. Proc. R. Soc. Lond. B 269, 1811–1819. Meunier, J. (2015). Social immunity and the evolution of group living in insects. Philos. Trans. R. Soc. B. 370, 20140102. Pull, C.D., Ugelvig, L.V., Wiesenhofer, F., Grasse, A.V., Tragust, S., Schmitt, T., Brown, M.J.F., and Cremer, S. (2018). Destructive disinfection of infected brood prevents systemic disease spread in ant colonies. eLife 7, e32073. Rosengaus, R.B., Traniello, J.F.A., and Bulmer, M.S. (2010). Ecology, behavior and evolution of disease resistance in termites. In Biology of Termites: a Modern Synthesis, D. Bignell, Y. Roisin and N. Lo, eds. (Dordrecht: Springer). Traniello, J.F.A., Rosengaus, R.B., and Savoie, K. (2002). The development of immunity in a social insect: evidence for the group facilitation of disease resistance. Proc. Natl. Acad. Sci. USA 99, 6838–6842. Schmid-Hempel, P. (1998). Parasites in Social Insects (Princeton: Princeton University Press). Stroeymeyt, N., Grasse, A.V., Crespi, A., Mersch, D.P., Cremer, S., and Keller, L. (2018). Social network plasticity decreases disease transmission in a eusocial insect. Science 362, 941–945. Wilson-Rich, N., Spivak, M., Fefferman, N.H., and Starks, P.T. (2009). Genetic, individual, and group facilitation of disease resistance in insect societies. Annu. Rev. Entomol. 54, 405–423.
IST Austria (Institute of Science and Technology Austria), Am Campus 1, 3400 Klosterneuburg, Austria. E-mail: [email protected]
Primer
Invertebrate allorecognition Matthew L. Nicotra Most colonial marine invertebrates live as surface encrustations in benthic environments. As they grow, these animals frequently encounter other members of their own species. These encounters typically lead to conflict, in which the colonies aggressively compete for space, or co-existence, in which the colonies peacefully border each other. Sometimes, however, interacting colonies will engage in a form of cooperation in which they fuse together and actively share resources. In most species, the decision of whether to compete, co-exist, or cooperate is controlled by a genetically encoded allorecognition system, which gives them the ability to distinguish at a molecular level between self-tissues and tissues of conspecifics. In this Primer, I will introduce allorecognition systems, explain why they have evolved in colonial invertebrates, summarize our nascent mechanistic understanding of how they work, and highlight some open questions in the field. What is allorecognition? Sessile colonial invertebrates include sponges, cnidarians (e.g. corals and hydroids), some tunicates, and nearly all bryozoans. Colonial animals are described as ‘modular’ because they grow by the repeated addition of discrete units, often called zooids, which remain physiologically linked to the original organism, thus creating a colony. An example is a coral, which is made up of many interconnected polyps and adds new polyps to its periphery in order to grow across a surface or up into the water column. The modularity of colonial invertebrates makes it possible for a zooid to generate a new colony if separated from a colony, with the new colonies becoming clones of the original. Colonial invertebrates span an enormous range of morphologies and organizational complexities, from the apparently haphazard mounds of encrusting sponges to the lace-like fans of bryozoans.
Regardless of their growth pattern, all colonial invertebrates gain a fitness advantage as they increase in size. Larger colonies capture more resources and produce more gametes, leading to higher fecundity. Larger colonies are also more likely to survive predation, physical disruption, or disease, since killing just part of the colony does not kill the entire organism. Having enough space in which to grow is therefore critical for the success of the colony. And since there is very little free space in benthic environments, sessile colonial animals spend most of their lives competing for open real estate. This creates a problem. A colony that encounters a neighbor will typically compete with it. But what if this neighbor is the same species? It could actually be a part of the same colony that has grown around a three-dimensional surface or is recovering from damage. In that case, competition would literally be selfdefeating. Co-existence or cooperation would be favored by natural selection. To solve this problem, colonial invertebrates have evolved allorecognition — the ability to distinguish between self-tissues and tissues of conspecifics. Allorecognition is controlled by genetic systems consisting of one or more loci. Colonies that share alleles at these loci typically fuse to create a chimeric colony, while those that are mismatched reject and aggressively compete. This self-recognition strategy is effective because allorecognition loci are among the most polymorphic genetic systems known. Each locus can have hundreds of alleles in a population. It is therefore unlikely that two colonies will share an allele by chance. Since rare alleles are the best markers of self, allorecognition loci evolve under a form of balancing selection called negative frequency-dependent selection, which maintains many divergent allorecognition alleles at low frequency over broad geographic distributions. Costs and benefits of allorecognition One might reasonably ask why colonial organisms have evolved allorecognition systems in the first place, rather than always fusing with conspecifics to gain the immediate fitness benefits of increased size. The answer lies in the costs and benefits of fusion and rejection. The main cost of rejection is that it is often aggressive, with colonies
Current Biology 29, R425–R473, June 3, 2019 © 2019 Elsevier Ltd. R463
Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA 02138, USA. E-mail: [email protected]
R458
Primer
Social immunity in insects Sylvia Cremer When animals become sick, infected cells and an armada of activated immune cells attempt to eliminate the pathogen from the body. Once infectious particles have breached the body’s physical barriers of the skin or gut lining, an initially local response quickly escalates into a systemic response, attracting mobile immune cells to the site of infection. These cells complement the initial, unspecific defense with a more specialized, targeted response. This can also provide long-term immune memory and protection against future infection. The cell-autonomous defenses of the infected cells are thus aided by the actions of recruited immune cells. These specialized cells are the most mobile cells in the body, constantly patrolling through the otherwise static tissue to detect incoming pathogens. Such constant immune surveillance means infections are noticed immediately and can be rapidly cleared from the body. Some immune cells also remove infected cells that have succumbed to infection. All this prevents pathogen replication and spread to healthy tissues. Although this may involve the sacrifice of some somatic tissue, this is typically replaced quickly. Particular care is, however, given to the reproductive organs, which should always remain disease free (immune privilege). Similarly, when an ant colony is infected, the contaminated or infected ants, along with their healthy nestmates, fight the infection collectively. A colony of social insects is hence protected by both the individual defenses of its members and a systemic, colony-wide response, providing the colony with a protection known as ‘social immunity’. Individual defenses comprise behaviors that prevent pathogen contamination (e.g., pathogen avoidance and selfgrooming) and those that fight infections. The latter utilizes the innate immune system, which is capable of raising specific and long-lasting responses, or immune memory (‘immune priming’). In addition, entry of pathogens into the colony and infection of individuals is quickly detected by nestmates. These
Current Biology 29, R425–R473, June 3, 2019 © 2019 Elsevier Ltd.
ants will try to remove contaminations from the nest, or from contaminated nestmates, to prevent infection. In cases where infections cannot be prevented, sick individuals may either remove themselves from the colony or are isolated by other ants. Similar to immune cells targeting infected cells, social immunity may also involve the sacrifice of infected colony members. This prevents pathogen replication and spread of pathogens to healthy colony members. Additionally, there is extra protection directed at the queen, the sole reproducing member of the colony. Metazoan bodies and the colonies of social insects (the social bees and wasps, the ants and termites) are hence protected from disease by strikingly similar principles, despite the fact that they represent very different types of biological organization. A central feature common to both, however, is the presence of a distinct germline and soma. Both are interdependent and require one another for reproduction. In an insect colony, workers are typically sterile themselves, but rear the eggs laid by the queen. Successful colony reproduction relies on both tasks. Even if cells can survive in cell culture when isolated from the body, and individual ants can survive when removed from their colony, reproduction – that is, the formation of new bodies or new colonies – requires cooperation between the subunits of the whole reproductive entity. Due to these organizational similarities, insect colonies are often referred to as ‘superorganisms’ (Figure 1). Just like organisms, superorganisms fight disease in a stepwise manner, meaning they employ distinct mechanisms to fight pathogens at the different steps of disease progression. Infectious particles enter the body or colony, establish and begin to spread around the host. This occurs by infecting a cell or colony member, replicating inside them, and then producing transmissible stages that spread to others. Host defenses have evolved to try to break this process at each step. If host defense is successful, the cycle is interrupted and the infection will die out; if not, disease progresses and has to be fought at the next step (Figure 2). Collective nest hygiene Even in the absence of pathogens, social insects keep their nests
meticulously clean of any potential sources of infection (Figure 2, Step 1). For example, when garden ants move into a new nest, they treat the nest material, including the new brood chambers, with a self-produced disinfectant, a formic acid-rich poison. Other species, like wood ants and honeybees, use antimicrobials collected from the environment to enrich their nest, with antimicrobial tree resin being the most commonly used. Termites incorporate their own feces into their nests, which, at first sight may seem unexpected. Yet, as they contain a lot of microbiota that produce antimicrobials, termite feces are not only a good building material, but also ensure cleanliness. Waste is removed before it can accumulate and form a niche for fungi, bacteria, or parasites such as mites to grow. Similarly, dying nestmates are removed from the nest. All corpses and debris are brought to specific graveyards or middens, located either in peripheral nest chambers or outside the nest. Waste removal is particularly sophisticated in the massive colonies of leaf-cutter ants, which process enormous masses of leaf material on a daily basis, to rear their symbiotic fungus as food. These fungal crops are continually weeded and treated with fungicides to remove errant, parasitic fungi, which would otherwise wipe out the ants’ monoculture. Honeybees, faced with nest contamination, employ a tactic used by our own bodies: they increase the temperature. This social fever is the result of the bees shivering their flight muscles, and their collective effort increases the overall hive temperature. This behavior has been shown to kill heat-sensitive pathogens. Nest hygiene is often a general precaution that is employed prophylactically and non-specifically. It relies on the mechanical removal of potentially infectious material, often combined with the application of antimicrobials with a broad spectrum of activity. Nest hygiene hence plays a similar role to the general grooming and prophylactic use of disinfectants (e.g., medical anointing in monkeys), observed in many species across the animal kingdom. Sanitary care If an individual is contaminated with a pathogen, either due to inefficient
Organism
Superorganism
Soma
Germline Current Biology
Figure 1. Insect colonies as superorganisms. Insect colonies form a reproductive entity consisting of one or a few reproductive individuals (the queen(s), pink), and their sterile workers (blue), which rear the brood and maintain the colony. Colony reproduction is dependent on the interaction of both these functional subunits, similar to the interdependency between the germline (reproductive organs; pink) and the somatic tissue (blue) in a multicellular, metazoan organism. Due to this analogous functional division of insect colony members into germline (queen) and soma (workers), insect colonies are often termed superorganisms. Graphics by ppm-vi.
nest hygiene or from foraging outside, sanitary care functions as the second step in the colony’s social immune response (Figure 2, Step 2). Similar to delousing monkeys, honeybees remove ectoparasites, such as the Varroa mite, by grooming their nestmates. This mite – the scourge of beekeepers – is not only damaging due its feeding on the blood (hemolymph) of the bees, but because it is also a vector of many diseases, including several viral diseases. Ants and termites also groom fungal spores from the body of contaminated nestmates, usually before the spores can penetrate the cuticle and cause internal infection. This allogrooming is more effective at preventing infection than selfgrooming, as additional body parts can be reached, such as the thorax. Garden ants combine the mechanical removal of infectious particles with the application of antimicrobials, increasing their chances of effectively preventing infection. Grooming is one of the most common sanitary care behaviors in social insects, and is very effective. Yet,
does allogrooming increase the risk of infection for groomer? Grooming ants collect the infectious material they remove into pouches inside the mouth, termed the infrabuccal pockets. Here, it is compacted and sterilized using antimicrobial gland compounds. The compacted pellets are eventually ejected and have an extremely reduced ability to germinate. Despite these measures, pathogen transfer can still occur from the contaminated individual to its nestmates. However, although many nestmates contract infections, only a small proportion of these cause disease. This is because only a small fraction of pathogen is transferred to each ant, leading to non-lethal, lowlevel infections. Low-level infections have been shown to trigger a protective immunization, in both ants and termites. This is similar to the use of low-level infections in early human medicine, before modern vaccination was developed. Known as ‘variolation’, it was used to immunize people against diseases such as small pox.
Current Biology 29, R425–R473, June 3, 2019
R459
Care–kill dichotomy Together, nest hygiene and sanitary care are effective measures to prevent pathogens from establishing within the colony. Most studies of social immunity describe examples of these early steps to stop disease, but they are not always fully effective. What happens when these initial defenses fail to prevent infection and disease takes hold? In these situations, social immunity switches from preventative measure aimed at protecting the whole colony and all its members, to preventing the replication and transmission of disease. These latter measures focus on the exclusion or elimination of infected individuals, akin to immune cells targeting infected cells in a body. Hence, once an infection has developed to the point of no return, social immunity switches from a ‘care’ to a ‘kill’ strategy. For the colony, the outcome is similar — for the infected individual, they face being either rescued or sacrificed for the colony. How do social insects resolve this apparent conflict of interest between the individual and the colony? In superorganisms, fitness is not measured at the level of the individual, as no single ant can reproduce independently. Instead, the fitness of each individual depends entirely on the overall fitness of the colony. If removing infected individuals to prevent disease
Collective hygiene Step 1: Contamination
Sanitary care
Step 2: Infection
Infection elimination
Care–kill breakpoint Step 3: Replication
Network modulation
Step 4: Transmission
Reproductive entity
Pathogen
Subunit
Cooperative defense Current Biology
R460
Current Biology 29, R425–R473, June 3, 2019
Figure 2. Disease stage-dependent course of social immunity. Social immunity acts to interfere at each possible step of disease progression (steps 1–4). To protect the reproductive entity from disease, distinct cooperative defense mechanisms are employed at each step (green upward-bent arrows). If they fail, the disease progresses to its next step (black downward-bent arrows), and the pathogen (blue diamonds) may be able to infect individuals (green circles), replicate (multiple diamonds), and transmit to new colony members (small green arrows). Following disease progression, the mechanisms of social immunity involve collective nest hygiene (e.g., ants spraying antimicrobial poison into their nest), sanitary care of contaminated individuals (e.g., removal of infectious particles by allogrooming), elimination of the infection (e.g., application of disinfectant after piercing the cuticle), and modulation of the social interaction network of the colony (by behavioral changes of colony members). Note that social immunity actions progress from initial steps that aim to protect the colony by protecting its individual members, to protecting the colony by preventing disease transmission. The latter involves the sacrifice of infected individuals, indicated by the ‘care–kill breakpoint’. Graphics by ppm-vi.
spread can maximize fitness, it will be to the benefit of not only the healthy portion of the colony, but also that of the sacrificed individual. Being sterile, social insect workers cannot reproduce directly, and hence they gain their fitness through caring for their relatives, with whom they share a significant portion of the same genes. Any action that increases the chance of their own genes being passed on into the next generation – be that sanitary care or being sacrificed due to infection – will indirectly increase their own inclusive fitness. We should therefore expect social insect workers to signal honestly about their health status, regardless of disease progression. That is, they should signal for help when contaminated and signal for elimination when infected. As in the metazoan body, where cells signal about their status to the immune system, we should expect superorganisms to have evolved similar mechanisms that protect the reproductive entity over its individual parts. What are these mechanisms (Figure 2, Step 3)? Infection elimination It has been known since the 1960s that honeybees can detect infected brood through the sealed brood comb, uncap it, and then drop the brood to the ground away from the hive. The bees performing this behavior have highly sensitive olfaction and have even been artificially selected for this hygienic behavior, in the hopes of combatting some honeybee diseases. More recently, this removal behavior has also been described for a number of ant species. Unlike bees, ants pool their brood in open piles, where they can single out infected individuals and remove them from the brood chamber. However, unlike the winged honeybee, simply dropping infectious brood outside the nest would run the risk of re-infection, since ants forage in the same territories they place their dead, if not even placing them in chambers inside the nest itself. To overcome this problem, garden ants have evolved a complex multicomponent behavior to rid colonies of brood infections. First, they slice open an infected pupa’s cocoon, and then bite and pierce its soft cuticle. Having made a wound, the ants then bend their abdomen over the pupae and emit a poisonous spray, rich in formic acid. Together, this behavior ensures the
poison can enter the infected pupae, where it disinfects the host from the inside out. This effectively prevents pathogen replication before new transmissible stages can be produced. This process of ‘destructive disinfection’ is not only functionally similar to the elimination of infected cells in a body, but also mechanistically equivalent. For example, infected cells can be made porous through perforin proteins, which make holes into which cytotoxins are injected. Infected cells emit a ‘find me, eat me’ signal that attracts immune cells towards them. Similarly, the infected pupae have an altered chemical odor that is detected by the workers. In both cases, this allows for the accurate detection of the sick cells or individuals, respectively, and prevention of collateral damage of healthy tissue or colony members (‘immunopathology’). Whilst the brood is immobile and requires active removal by workers, adult insects can leave the nest of their own accord. It has indeed been found that sick or generally moribund ants leave the colony shortly before death. The basis for this behavior is not yet well understood, but seems to include a loss of the insects’ ability to integrate social cues. Nest departure may also be prompted by the healthy nestmates, with honeybees expressing aggression towards infected individuals. In termites, initial grooming by nestmates can sometimes escalate to aggressive grooming, that then leads to cannibalism, and hence the ‘removal’ of the infected individual, via the antimicrobial environment of the termites’ gut. Modulation of the social interaction network Whilst the cells in a body are highly connected and embedded in an extracellular matrix, insects in a colony can move freely. This means they can flexibly adjust their interactions with the other colony members to potentially reduce disease transmission. Epidemiological modeling forecasts disease spread through a network of hosts, which is typically visualized as dots representing individuals (nodes) and interconnecting lines (edges) showing their contacts. These models predict higher disease spread in simulated colonies that lack a structured network, as if all individuals interacted randomly,
compared with those with structure. In fact, social insects express structured interaction networks, due to a clustering according to the tasks they perform, and where these are carried out in the nest. Most strikingly, young workers, the nurses, take care of the brood and the queen in the center of the nest, whereas older workers leave the nest to forage. This leads to an inherent network structure that may have, in part, been shaped by selection to limit infectious disease transmission, as predicted by the ‘organizational immunity hypothesis’. Is it possible for social insects to enhance their constitutively expressed transmission-inhibiting network properties even further upon pathogen exposure? Experimental evidence for induced organizational immunity by behavioral modulation has indeed been recently observed (Figure 2, Step 4). Pathogencontaminated garden ant foragers were found to change their spatial preferences, staying outside the nest for longer and interacting less with the nurse ants in the center of the network. In response to forager contamination, the nurses brought the brood even closer to the nest center. This led to an even stronger disassociation between foragers and the nurses, resulting in a reduced risk of infection for the nurses and queen. Pathogen spread was thus highest towards other individuals of the same age- and task-group – other foragers – whereas most nurses and queens received only low pathogen amounts, which typically cause immunization rather than disease. Host susceptibility Whether or not hosts develop disease when encountering a pathogen depends on the number of infectious particles and the hosts’ disease susceptibility. Hosts can differ greatly in susceptibility, which is an inherent property and can also be altered by social context and infection history. Further, pathogen strains are often more or less virulent against different host genetic backgrounds. In social insects, colony members are usually highly related, resulting in a correlated susceptibility to pathogens. Genetic similarity across colony members is highest in colonies headed by a sole singly mated queen. This means that a pathogen strain able to infect one
Current Biology 29, R425–R473, June 3, 2019
R461
Totipotent society
Superorganismal society
All individuals
Soma
Germline only
Germline
Reproduction Current Biology
Figure 3. Totipotent versus superorganismal societies. In societies of totipotent individuals (nodes: circles; edges: black lines), each society member integrates germline (pink) and soma (blue), so that all individuals can reproduce to form a next generation. In superorganismal societies, members belong either to the germline (the queen(s)) or the soma (sterile workers). At each reproduction event, the network collapses to the single queen state, which then builds up a new society by its sterile worker offspring. Graphics by ppm-vi.
colony member is likely able to infect all others in a colony. Multiple mating in queens or multiple queens reproducing in a colony has been suggested to evolve so as to reduce susceptibility to disease, by increasing a colony’s genetic diversity. Still, the overall high relatedness in social insects increases the risk of colony collapse. Disease susceptibility can change dynamically over an individual’s lifetime. It can depend on an animal’s overall condition (e.g., nutrition, stress), but also its infection history. As insects show both individual and transgenerational immune priming, previous infections earlier in life or in the parents can induce protective immunization. Interestingly, in social insects, changes in disease susceptibility induced by previous pathogen contact seem to affect how they care for nestmates in the future. For example, garden ants adjust their grooming and disinfection behaviors in a risk-averse manner, reducing their risk of contracting diseases they have an increased susceptibility to. R462
Overall, social and individual defenses are tightly interconnected and the overall outcome of disease dynamics in the colony will be shaped by how effectively sanitary care and infection elimination reduce pathogen numbers in the colony, how easily the remaining pathogens can spread across the social interaction network, and the disease susceptibility of the colony. Evolution of social immunity Since the introduction of the concept of social immunity, we have gained significant insight into how cooperative disease defenses function. Yet, we still know little about how they evolve. Some components of social immunity, such as collective hygiene and sanitary care, are present in many forms of social life, including parent–offspring groups where the young disperse, and in communities of unrelated individuals. However, the extent to which individuals engage in collective disease defenses, and so risk contracting infections themselves, will be dependent on their relatedness to others in the group, their own future
Current Biology 29, R425–R473, June 3, 2019
reproductive potential, and the severity of the pathogen in question. Moreover, it is probably unlikely that there will be selection on sick animals to actively signal when they are ill to others – except in groups of relatives – as this risks expulsion from the group. Indeed, there is ample evidence that infected conspecifics are excluded and selection seems, in some cases, to have therefore led to the hiding of illnesses from others. Whilst unconditionally sterile individuals in social insect colonies must behave altruistically to maximize their inclusive fitness, individuals in other types of social groups that currently reproduce, or will in the future, are most likely to act selfishly to maximize their own direct fitness. With the exception of superorganisms, where the conditions for social immunity to evolve are met, predicting when, or if, social immunity can evolve in other social groups is not trivial. This has thus far challenged our comprehension of which social systems, apart from superorganisms, will express the full repertoire of social immunity, including elimination of infected individuals. It is therefore necessary to study collective disease defense in non-superorganismal eusocial, or subsocial insects to learn more about the evolution of social immunity in insects. Conclusion Given their superorganismal biology, which perhaps has more in common with a metazoan body than other animal groups (Figure 3), can social insects still serve as good model systems to understand the principles of disease defense in other societies? The answer appears to be yes. Although superorganisms can only reproduce as a whole, whereas each member in an animal society can reproduce independently (each being a ‘totipotent’ organism; Figure 3), social insects are still groups of freely interacting animals. This means, from an epidemiology perspective, they behave similarly to conventional societies. Therefore, the same principles about how network properties affect pathogen spread between interacting hosts apply as equally to the social insects as they do to humans. Improving our epidemiological understanding is relevant for better
disease management in a variety of fields, including, for example, livestock protection and outbreaks in human societies. To date, epidemiological research relies mostly on correlational data that are gathered during or following outbreaks. Research on insect societies, however, allows for experimental prospective studies with full surveillance of the social interactions and their effect on pathogen spread. This even allows us to test how manipulation of the network properties, or the individual behaviors, affect disease outbreaks. This may provide valuable data to advance current epidemiological modeling, and hence develop more realistic prevention strategies against disease outbreaks in social groups. FURTHER READING Boomsma, J.J., and Gawne, R. (2018). Superorganismality and caste differentiation as points of no return: how the major evolutionary transitions were lost in translation. Biol. Rev. 93, 28–54. Cremer, S., Armitage, S.A.O., and Schmid-Hempel, P. (2007). Social immunity. Curr. Biol. 17, R693– R702. Cremer, S., and Sixt, M. (2009). Analogies in the evolution of individual and social immunity. Phil. Trans. R. Soc. B 364, 129–142. Evans, J.D., and Spivak, M. (2010). Socialized medicine: individual and communal disease barriers in honey bees. J. Invertebr. Pathol. 103, S62–S72. Hughes, W.O.H., Eilenberg, J., and Boomsma, J.J. (2002). Trade-offs in group living: transmission and disease resistance in leaf-cutting ants. Proc. R. Soc. Lond. B 269, 1811–1819. Meunier, J. (2015). Social immunity and the evolution of group living in insects. Philos. Trans. R. Soc. B. 370, 20140102. Pull, C.D., Ugelvig, L.V., Wiesenhofer, F., Grasse, A.V., Tragust, S., Schmitt, T., Brown, M.J.F., and Cremer, S. (2018). Destructive disinfection of infected brood prevents systemic disease spread in ant colonies. eLife 7, e32073. Rosengaus, R.B., Traniello, J.F.A., and Bulmer, M.S. (2010). Ecology, behavior and evolution of disease resistance in termites. In Biology of Termites: a Modern Synthesis, D. Bignell, Y. Roisin and N. Lo, eds. (Dordrecht: Springer). Traniello, J.F.A., Rosengaus, R.B., and Savoie, K. (2002). The development of immunity in a social insect: evidence for the group facilitation of disease resistance. Proc. Natl. Acad. Sci. USA 99, 6838–6842. Schmid-Hempel, P. (1998). Parasites in Social Insects (Princeton: Princeton University Press). Stroeymeyt, N., Grasse, A.V., Crespi, A., Mersch, D.P., Cremer, S., and Keller, L. (2018). Social network plasticity decreases disease transmission in a eusocial insect. Science 362, 941–945. Wilson-Rich, N., Spivak, M., Fefferman, N.H., and Starks, P.T. (2009). Genetic, individual, and group facilitation of disease resistance in insect societies. Annu. Rev. Entomol. 54, 405–423.
IST Austria (Institute of Science and Technology Austria), Am Campus 1, 3400 Klosterneuburg, Austria. E-mail: [email protected]
Primer
Invertebrate allorecognition Matthew L. Nicotra Most colonial marine invertebrates live as surface encrustations in benthic environments. As they grow, these animals frequently encounter other members of their own species. These encounters typically lead to conflict, in which the colonies aggressively compete for space, or co-existence, in which the colonies peacefully border each other. Sometimes, however, interacting colonies will engage in a form of cooperation in which they fuse together and actively share resources. In most species, the decision of whether to compete, co-exist, or cooperate is controlled by a genetically encoded allorecognition system, which gives them the ability to distinguish at a molecular level between self-tissues and tissues of conspecifics. In this Primer, I will introduce allorecognition systems, explain why they have evolved in colonial invertebrates, summarize our nascent mechanistic understanding of how they work, and highlight some open questions in the field. What is allorecognition? Sessile colonial invertebrates include sponges, cnidarians (e.g. corals and hydroids), some tunicates, and nearly all bryozoans. Colonial animals are described as ‘modular’ because they grow by the repeated addition of discrete units, often called zooids, which remain physiologically linked to the original organism, thus creating a colony. An example is a coral, which is made up of many interconnected polyps and adds new polyps to its periphery in order to grow across a surface or up into the water column. The modularity of colonial invertebrates makes it possible for a zooid to generate a new colony if separated from a colony, with the new colonies becoming clones of the original. Colonial invertebrates span an enormous range of morphologies and organizational complexities, from the apparently haphazard mounds of encrusting sponges to the lace-like fans of bryozoans.
Regardless of their growth pattern, all colonial invertebrates gain a fitness advantage as they increase in size. Larger colonies capture more resources and produce more gametes, leading to higher fecundity. Larger colonies are also more likely to survive predation, physical disruption, or disease, since killing just part of the colony does not kill the entire organism. Having enough space in which to grow is therefore critical for the success of the colony. And since there is very little free space in benthic environments, sessile colonial animals spend most of their lives competing for open real estate. This creates a problem. A colony that encounters a neighbor will typically compete with it. But what if this neighbor is the same species? It could actually be a part of the same colony that has grown around a three-dimensional surface or is recovering from damage. In that case, competition would literally be selfdefeating. Co-existence or cooperation would be favored by natural selection. To solve this problem, colonial invertebrates have evolved allorecognition — the ability to distinguish between self-tissues and tissues of conspecifics. Allorecognition is controlled by genetic systems consisting of one or more loci. Colonies that share alleles at these loci typically fuse to create a chimeric colony, while those that are mismatched reject and aggressively compete. This self-recognition strategy is effective because allorecognition loci are among the most polymorphic genetic systems known. Each locus can have hundreds of alleles in a population. It is therefore unlikely that two colonies will share an allele by chance. Since rare alleles are the best markers of self, allorecognition loci evolve under a form of balancing selection called negative frequency-dependent selection, which maintains many divergent allorecognition alleles at low frequency over broad geographic distributions. Costs and benefits of allorecognition One might reasonably ask why colonial organisms have evolved allorecognition systems in the first place, rather than always fusing with conspecifics to gain the immediate fitness benefits of increased size. The answer lies in the costs and benefits of fusion and rejection. The main cost of rejection is that it is often aggressive, with colonies
Current Biology 29, R425–R473, June 3, 2019 © 2019 Elsevier Ltd. R463