ABOVE: Ophiocordyceps fungi grow a stalk-like fruiting body out of the infected ant’s corpse to disperse their spores. Charissa de Bekker
To the casual observer, the motivations that drive insect behaviors may appear quite simple: An insect might leave the nest to find food, wander around to seek out potential mates, or move into the sun or shade to maintain an optimal body temperature.
But sometimes the drivers of these behaviours are far more complex—and more sinister—than they first appear. In a surprisingly large number of cases, insects are not acting of their own free will in a way that benefits themselves or even their species. Instead, they have become “zombies,” controlled by barely visible fungal puppet masters that direct the insects’ behaviors, steering them into optimal conditions for dispersing infectious spores. While these fungi were described in the scientific literature as early as the mid-1800s, the extent and precision of the behavioral control that they exert on their unfortunate insect hosts—and the mechanisms they use to do so—are only just starting to be appreciated.1
As they begin to explore the complex molecular dialogue between these fungi and their insect hosts, scientists aren’t sure exactly what they will find. So far, the fungal kingdom as a whole has proven to be a rich source of bioactive metabolites; fungal-derived drugs are currently used as antibiotics, immunosuppressants, cholesterol-lowering agents, and migraine therapeutics, so there may be much to discover in these insect-manipulating species.2 “This is a group of fungi that haven't quite been mined yet, for all the things that they might produce. I'm quite certain that we'll bump into some interesting stuff,” said Charissa de Bekker, a molecular biologist who studies insect-fungi interactions at Utrecht University.
This fascination has spread beyond the scientific community into pop culture, as evidenced by video games and movies like The Last of Us and The Girl with All the Gifts. So, although these fungi cannot literally infect humans, they have certainly extended their mycelia into the hearts of scientists and non-scientists alike.
Ophiocordyceps: The Zombie Fungus Poster Child
 Modified from © istock.com , tatarnikova, sorbetto, vdvornyk, Annandistock, Natalia Misintseva; Designed by Ashleigh Campsall
|
Like many scientists who study the natural world, de Bekker’s career path was influenced by the Planet Earth television series. During graduate school, she studied the ubiquitous fungi Aspergillus niger, but when she watched an episode in which David Attenborough described the fungus Ophiocordyceps—how it directs an ant up a leaf or stem, forces it to anchor itself with its mandibles, and then slowly grows a fruiting body out of the ant’s desiccated corpse—she was captivated.
“My PhD was all about fungal complexity—how different cells in a fungal colony might be [performing] different tasks and doing interesting things,” said de Bekker. “So, for me, a fungus that could infect an insect and change its behavior seemed like the ultimate complex thing that a microbe could do. I was super fascinated by it.”
We got an idea of the fungal genes that turn on during manipulation, and which ant genes turn on or off at the same time. It gives us an idea of how these two organisms might interact.
- Charissa de Bekker, Utrecht University
Scientists already knew that Ophiocordyceps encourages ants to climb up nearby vegetation, a behavior known as summiting, to ensure widespread dispersal of the infectious spores. However, as de Bekker and other researchers explored these ant-fungi dynamics in more depth, they realized that the behaviors orchestrated by the fungi were even more finely tuned to support fungal fitness than they had originally believed.
“We found that infected ants first lose their daily rhythms,” said de Bekker. “Ants—like us—have biological clocks, and many of the species that get infected are actually nocturnal species, so [usually] they're mostly foraging for food during the nighttime, and then during the day, they stay inside the nest.”
Infected ants show no such patterns; instead, they are constantly active. Yet something within them is still keeping track of the time, or at least responding to time-dependent environmental cues: summiting behavior largely occurred at solar noon.3 The fungi also appeared to be quite picky about the conditions in which they choose to anchor their hosts and grow their fruiting bodies: de Bekker found that shading a portion of ant graveyard reduced the number of new infected ant corpses and reduced fungal fruiting body production.4
“It became a super complex, interesting story—even more interesting than it already was,” said de Bekker.
Since then, her fascination with this fungus has only grown, and the increasing accessibility of genomics, transcriptomics, and metabolomics technologies has allowed her to dive into the strategies that Ophiocordyceps uses to control its host. “Over the years, we got an idea of the fungal genes that turn on during manipulation, and which ant genes turn on or off at the same time,” said de Bekker. “It gives us an idea of how these two organisms might interact.”
Because infected ants lost their normal activity patterns but still summited at a particular time, de Bekker suspected that circadian clocks played an important role in mediating these ant-fungi interactions. While circadian rhythms are poorly studied in most species of fungi, other researchers had previously characterized a circadian clock in the fungus Neurospora crassa which included genes encoding photoreceptors and other clock genes forming feedback loops to keep the clock running.5
Using Neurospora as a template, de Bekker identified several clock gene homologues in Ophiocordyceps, including putative photoreceptors, and showed that the transcription of many other genes rose and fell predictably during a 24-hour cycle.6 Infected ants, on the other hand, displayed dysregulation in the expression of genes governing circadian rhythms, suggesting that the fungus may hijack host rhythms for its own purposes.7
However, the circadian disruptions are just one element of the fungal manipulation. Given the variety and complexity of the behavioral alterations, de Bekker said it’s unlikely that one single gene or compound controls them all. From their data sets of thousands of differentially expressed genes and metabolites that correlate with various behaviors over the course of the infection, de Bekker and her research team have begun to assemble a list of fungal candidate compounds that might be involved in the manipulation.8,9 But while they are able to determine the identity of many proteins and metabolites secreted by the fungi, “For a lot of them, we have no idea really what they do,” de Bekker said.
Sometimes, however, scientists get lucky and find a compound that’s similar to a well-understood chemical produced by an entirely different species of fungus. For example, de Bekker’s transcriptomic analysis of Ophiocordyceps revealed a massive upregulation of a gene cluster homologous to the one that produces the toxin aflatrem in Aspergillus flavus.10 Scientists have been studying this potent mycotoxin for decades, after early observations that cattle that ate grain infected with A. flavus developed tremors and convulsions.11
de Bekker realized that the staggering behaviors observed in the poisoned cattle were similar to the behaviors in her infected ants, so she wondered if these aflatrem-like compounds might be responsible. And indeed, when the researchers injected aflatrem into the ants, they observed comparable staggering behaviors and many similarities in the transcriptional profiles of aflatrem-treated and Ophiocordyceps-infected ants. Some of these genes are involved in sensory and motor functions, suggesting that this type of compound is likely part of the fungi’s ant-manipulating chemical arsenal.
In other cases, de Bekker and her colleagues use machine learning to generate hypotheses about how fungal effectors might interact with host proteins.12 According to these predictions, said de Bekker, “A lot of them can bind to G-protein-coupled receptors in the ant, [which] have a lot of different functions. Some of them are there to detect odor or detect light—you can imagine that has something to do with behavior.” Other predicted interactions include receptors for neurotransmitters like dopamine and octopamine, the insect homolog of noradrenaline. de Bekker is currently following up on these in silico findings, investigating how the predicted protein-protein interactions could be involved in fungal manipulation of the host.
While de Bekker expects to find molecules that could one day be useful in medicine or industry, there are also more concrete, nearer term applications. For example, de Bekker showed that one species of Ophiocordyceps could infect multiple ant species in a laboratory setting, but was not able to manipulate the behavior of ants that it did not infect in nature, even those in the same genus as its preferred hosts.13 Identifying the mechanisms that drive these species-specific effects could help researchers design more targeted pesticides, said de Bekker. These fungi-inspired pesticides could be biodegradable and potentially specific to a certain species of insect, reducing the collateral damage that conventional pesticides can inflict on pollinators, other beneficial insects, and even humans.
Drosophila-Manipulating Fungi: Fly-ing Under the Radar
While Ophiocordyceps is likely the best-known zombie fungus, it’s certainly not the only one. “This happens so much in nature, maybe more than we [will] ever realize,” said de Bekker. “And now that people are looking into this, we're finding more and more insects that are affected.”
Indeed, “summit disease”—in which the affected organism climbs some distance above the ground, strikes a characteristic death pose, and rains infectious particles down onto unfortunate passerby—has been reported throughout the insect family tree, in species like house flies, grasshoppers, crickets, and soldier beetles.14,15 However, many of these species are difficult to study, so an infection in an insect model organism that is amenable to genomic tinkering with established approaches and tools, would be extremely useful—if researchers could find it.
In fact, during the 1900s, there had been a few reports published in relatively small journals, describing summit disease in wild fruit flies, or Drosophila melanogaster, the canonical model insect. Yet these reports had largely fallen into obscurity, and this fungus-fly pair would await rediscovery until the mid-2010s by graduate student Carolyn Elya.
Today, Elya is a self-described “zombiologist” at Harvard University, but at the time, she was working in molecular biologist Michael Eisen’s laboratory at the University of California, Berkeley (UC Berkeley), studying the fruit fly microbiome. She wanted to compare the microbes in her laboratory flies to those in wild flies. Fortunately—as anyone who has ever left an overripe peach on a kitchen counter has discovered—wild Drosophila are not difficult to find.
Elya set up a bait in her Berkeley backyard and waited for the flies to arrive. She easily captured many flies, but she also noticed something strange: Some of the dead flies at the bottom of the trap had died in an unusual, but recognizable pose. “My interest was piqued,” she recalled, “because I knew that having the wings raised at a 90-degree angle to the body axis was something that the fungus Entomophthora muscae [could cause].”
“And I got really excited, and I ran the specimens to the lab, and looked at them under a microscope, and I saw spores, and [thought] ‘Oh my god!’ I got really, really excited that this behavior-manipulating fungus had just sort of shown up out of the blue,” she said. “I decided that I was going to catch it and work with it, because that was way cooler than what I was working on.”
Elya has been devoted to studying Entomophthora ever since. After a lengthy trial-and-error process, Elya and her fellow researchers at UC Berkeley managed to culture the fungus in the lab and reliably establish infections in lab-reared flies.16 Finally, researchers would be able to use the enormous wealth of Drosophila genetic tools and widely shared stocks of genetically-altered flies to investigate the molecular mechanisms of fungal behavior manipulation.
Like Ophiocordyceps, Entomophthora alters its host’s activity levels, provoking a burst of activity shortly before death. It coaxes the host to climb up some nearby vegetation and anchor itself in place before expiring. In both, there is a circadian element at play: Ophiocordyceps kills at solar noon, while Entomophthora kills near sunset. After the insect dies, the fungus consumes the host tissues to fuel its growth and the production of spores to infect the next generation of victims.
Although Ophiocordyceps and Entomophthora both drive similar behaviors in their hosts, the evolutionary distance between them is surprisingly large. “[They’re] about as distantly related as I am from a fruit fly,” said Elya. As such, scientists don’t yet know whether these fungi have evolved resemblant insect-manipulation mechanisms or whether each has a unique strategy for achieving a common result.
To determine how her fungus hijacked the fly brain, Elya and her colleagues screened hundreds of flies in which the functions of specific neurons or specific genes were disrupted and measured how this affected summiting behavior.17 Their screen highlighted the importance of two neural systems. “One was that when we mutated circadian genes in the fly, we tended to get poor summiting,” said Elya. “The second was that when we silenced neurons within this region of the brain called the pars intercerebralis (PI), which is this neurosecretory region…we also saw a reduction in summiting behavior.”
To be perfectly honest, we don't know what studying this weird fungus is going to get us… Fungi are amazing chemists, so there's probably some really cool natural products that are just waiting for us to discover.
- Carolyn Elya, Harvard University
By silencing different populations of neurons in the PI, they identified a tiny subpopulation that seemed to have an outsize effect on summiting behavior. They dubbed these cells PI-CA neurons, since they project from the PI to the corpus allatum (CA), an endocrine gland that releases a substance called juvenile hormone into the fly circulatory system.
This hormone, said Elya, “maintains the juvenile state, but it also has a ton of different roles beyond its role in development. It's this really multifaceted and interesting hormone; we're still figuring out all the things that juvenile hormone does.”
Nevertheless, the researchers were able to demonstrate the importance of this pathway in fungal behavior manipulations: Silencing the PI-CA neurons or genetically ablating the CA itself dramatically reduced summiting behavior in infected flies. “But then the big question is: What the heck is the fungus doing to actually cause any of this? And that's where we're still working to find stuff out,” said Elya.
“One thing that we found was that the fungus invades the brain in a very creepy and stereotyped pattern,” she continued. “In the brains of flies that are executing summiting behavior, fungi are disproportionately congregated in the region of the brain that contains the processes that are coming off of those [PI] neurons. It was really kind of wild to see, because we found the neurons and then we did this completely different experiment, and we found that the fungal cells were sitting basically right next to all of the communication infrastructure.” In other words, the fungal cells were perfectly placed to manipulate the pathway that researchers had previously identified as a major driver of this behavior.
Scientists are beginning to develop a picture of the host targets that are important for these behaviors, but to truly understand these manipulations from the perspective of the fungus, they need to go back to basics and start to untangle its arrestingly weird genome.
Elya’s Entomophthora isolate has one of the largest genomes of any fungus, weighing in at about one billion bases (for comparison, an entire human can be built with the information contained in 3.4 billion bases).18 “These genomes are huge—not because they have a ton of genes, but because they have a ton of transposable elements,” said Elya. “It's like there was just a transposable element party that happened at some point during their evolution…but we don’t know why.” Elya is currently working on developing tools that will allow her to fluorescently tag and genetically manipulate the fungus to investigate the molecular mechanisms by which the fungus manipulates the fly.
“To be perfectly honest, we don't know what studying this weird fungus is going to get us… Fungi are amazing chemists, so there's probably some really cool natural products that are just waiting for us to discover, and maybe some of those have neuromodulatory properties. And fly brains are actually not that different from the brains of other creatures, such as ourselves. You can imagine we might potentially find things that can be used in the context of treating psychiatric disorders,” said Elya.
Elya is also brainstorming some “wonkier” applications inspired by behavior-manipulating fungi adapting to particular quirks of insect biology. For example, when Ophiocordyceps needs to secure its dying host to a leaf or a stem, it can force the ant’s powerful jaws to close around the strand of plant matter. Since fruit flies do not have jaws—their mouthparts are more suited to slurping than biting—Entomophthora has to be slightly more creative.
“[The flies] stick out their proboscises, and there is a droplet of goo that comes out of the proboscis and sticks them in place. And this is an incredibly strong adhesive. I think this could have some really cool biological properties—maybe it’s a glue that could be used in medicine. So, one of the projects in the lab is trying to figure out what's in the goo.”
Massospora: How to Win Friends and Influence Cicadas
Matt Kasson has always been fascinated by the wealth of unappreciated biological interactions taking place within arm’s reach. “I've always done domestic research,” said Kasson, who studies forest pathology at West Virginia University. “I think there are a lot of things that are overlooked, right below our feet…. I like the idea of shining a light on things that are just right outside our doors, that people are seeing, but maybe not appreciating.”
Kasson began his scientific career as a plant pathologist, studying how certain species of fungi might help control the tree-of-heaven, a fast-growing, foul-smelling tree that is highly invasive in many countries throughout Europe, North America, and Oceania. Kasson noticed that as the trees died, they released substances that attracted beetles; he began to wonder whether the beetles played any role in the battle between the invasive tree and the pathogenic fungus.
For Kasson, this curiosity was a gateway into the strange world of arthropod-fungus interactions. “I became really captivated by these obligate parasites, where they absolutely require the host to survive.”
“It's funny, because in graduate school, I had some colleagues that studied rust fungi, which are obligate on their host. They're very specialized, and they're really hard to work with,” said Kasson. “I always thought that I would never want to work on a project like that.”
Now, however, Kasson studies even more intractable fungus-host pairs: fungi from the genus Massospora and their hapless hosts, the periodical cicadas. Like several other behavior-modifying fungi, this species of Massospora is not exactly a new discovery; descriptions appeared in the scientific literature as early as the mid-1800s.19 “But because it's unculturable, and because its host has this 17-year or 13-year life cycle, it didn't really make sense to study it,” said Kasson. “So, it remained kind of a mycological oddity for like 100 years.”
Kasson said he’s come to appreciate the challenge of working with these fiendishly difficult systems. “I recognize that there's great opportunity here. Sure, there's a lot of risk in trying to study these unculturable, obligate pathogens and parasites—they have all these strange quirks. But with that comes the chance for novel discovery.”
Periodical cicadas have a drastically different lifestyle than ants or flies. They spend the vast majority of their lives several meters underground, feeding on tree roots and storing up energy before emerging in a fashion that scientists have called “spectacular [and] recklessly theatrical,” to mate, lay eggs, and expire.20 Accordingly, Massospora uses different—arguably even more horrifying—strategies to ensure its propagation among its cicada hosts.
Unlike Ophiocordyceps or Entomophthera, which wait until their hosts are dead before replacing the insect tissues with spore-producing structures, Massospora shows no such restraint. Well before death, an infected cicada will lose the back portion of its abdomen, which is replaced with a whitish, spore-shedding fungal “plug.”
Fungus-insect interactions might represent the next frontier for drug discovery.
- Matt Kasson, West Virginia University
Infected cicadas display two especially noticeable behavior modifications. First, they appear relatively unbothered by the loss of a significant portion of the body. As Kasson noted, “If you were to have your abdomen ripped open by a fungus, you would probably not be very motivated to walk around or engage in any of the normal stuff that humans do. Yet the cicadas that are infected continue to fly around, as if nothing were wrong.” A more active cicada—one that’s interacting with lots of other cicadas—is likely beneficial for a fungus seeking to maximize infections.
The second, equally upsetting modification is this: Massospora-infected cicadas are not just zombies—they are zombies that are especially enthusiastic about mating. Or rather, attempting to mate, as they no longer have genitalia. In a normal cicada mating ritual, explained Kasson, the male will sing (or scream) to attract nearby females and, if a female is receptive, she will perform a characteristic wing tap. Infected males will scream to attract females, but they will also wing tap in response to another male’s song.21 “In a way, the fungus is optimizing dispersal by doubling the number of cicadas it could potentially come in contact with,” said Kasson.
To begin their investigations into how the fungus might cause these behaviors, Kasson and his colleagues analyzed the chemical makeup of the fungal plugs from periodical and annual cicadas infected with different species of Massospora. While many of the small molecules had unknown functions, two chemicals with notable psychoactive properties popped up: cathinone, an amphetamine produced by plants, and psilocybin, a psychedelic produced by mushrooms.22
There may be other interesting psychoactive molecules as well, but it’s difficult to identify them, said Kasson. “There are a lot of barriers to advancing our understanding of a majority of these molecules, partly because of databases being [incompletely] populated and also the fact that fungi, especially in these unique lifestyles, that are so highly specialized, are going to produce a suite of novel compounds.”
However, given the psychoactive molecules that the researchers have already identified, “Fungus-insect interactions might represent the next frontier for drug discovery,” said Kasson.
Now, said Kasson, the researchers are using transcriptomics on a fresh batch of fungus to determine which genes it switches on during infection. In addition to collecting these fungal samples, Kasson and his research team carefully gathered and preserved cicada heads to send to Meghan Barrett, an insect neuroscientist at Indiana University Indianapolis and likely one of the few individuals who would be excited to receive 200 cicada heads in the mail.
Although Barrett has worked with many species of insects before, this is her first foray into cicada brains. It’s also one of the first times that anyone has looked especially closely at cicada brains, at least in the present century. “I went looking for literature about cicada neuroanatomy, and there's actually almost nothing,” said Barrett. “So, part of what we'll be doing on this project is actually describing the nervous system of cicadas—especially these species of cicadas—for the first time.”
Once she establishes what a healthy cicada brain looks like, Barrett will determine how the fungus affects the volume and the number of brain cells in different regions over the course of the infection. While Barrett herself is interested in studying cicada brains for their own sake, she said this research addresses bigger picture questions as well, shedding light on the ways that parasites impact the nervous systems of their hosts.
Studying a variety of parasites and hosts can help researchers understand the nature of host-parasite interactions more broadly. From contributing to the fundamental understanding of the stranger corners of biology to potentially identifying new natural products that could function as super-powered adhesives, insect-specific pesticides, or even neurotropic drugs, scientists are eager to keep exploring these insect-manipulating fungal oddities.
- Fawcett W. An entomogenous fungus. Ann Mag Nat Hist. 1886;18(106):316-318.
- Prescott TAK, et al. Fungal drug discovery for chronic disease: History, new discoveries and new approaches. Biomolecules. 2023;13(6):986.
- Hughes DP, et al. Behavioral mechanisms and morphological symptoms of zombie ants dying from fungal infection. BMC Ecol. 2011;11:13.
- Andriolli FS, et al. Do zombie ant fungi turn their hosts into light seekers? Behav Ecol. 2019;30(3):609-616.
- Cheng P, et al. WHITE COLLAR-1, a multifunctional neurospora protein involved in the circadian feedback loops, light sensing, and transcription repression of wc-2. J Biol Chem. 2003;278(6):3801-3808.
- de Bekker C, Das B. Hijacking time: How Ophiocordyceps fungi could be using ant host clocks to manipulate behavior. Parasite Immunol. 2022;44(3):e12909.
- Will I, et al. Genetic underpinnings of host manipulation by Ophiocordyceps as revealed by comparative transcriptomics. G3. 2020;10(7):2275-2296.
- de Bekker C, et al. Gene expression during zombie ant biting behavior reflects the complexity underlying fungal parasitic behavioral manipulation. BMC Genom. 2015;16(1):1-23.
- Will I, et al. Multiomic interpretation of fungus-infected ant metabolomes during manipulated summit disease. Sci Rep. 2023;13(1):14363.
- Beckerson WC, et al. 28 minutes later: Investigating the role of aflatrem-like compounds in Ophiocordyceps parasite manipulation of zombie ants. Anim Behav. 2023;203:225-240.
- Wilson BJ, Wilson CH. Toxin from Aspergillus flavus: Production on food materials of a substance causing tremors in mice. Science. 1964;144(3615):177-178.
- Will I, et al. Using machine learning to predict protein-protein interactions between a zombie ant fungus and its carpenter ant host. Sci Rep. 2023;13(1):13821.
- de Bekker C, et al. Species-specific ant brain manipulation by a specialized fungal parasite. BMC Evol Biol. 2014;14:166.
- Lovett B, et al. Going gentle into that pathogen-induced goodnight. J Invertebr Pathol. 2020;174:107398.
- Casique-Valdes R, et al. A PCR-based method to identify Entomophaga spp. infections in North American grasshoppers. J Invertebr Pathol. 2012;109(1):169-171.
- Elya C, et al. Robust manipulation of the behavior of Drosophila melanogaster by a fungal pathogen in the laboratory. Elife. 2018;7:e34414.
- Elya C, et al. Neural mechanisms of parasite-induced summiting behavior in ‘zombie’ Drosophila. eLife. 2023;12:e85410.
- Stajich JE, et al. Signatures of transposon-mediated genome inflation, host specialization, and photoentrainment in Entomophthora muscae and allied entomophthoralean fungi. Elife. 2024;12:RP92863.
- Speare AT. Massospora cicadina Peck: A fungous parasite of the periodical cicada. Mycologia. 1921;13(2):72-82.
- Simon C, et al. Advances in the evolution and ecology of 13- and 17-year periodical cicadas. Ann Rev Entomol. 2022;67(Volume 67, 2022):457-482.
- Cooley JR, et al. A specialized fungal parasite (Massospora cicadina) hijacks the sexual signals of periodical cicadas (Hemiptera: Cicadidae: Magicicada). Sci Rep. 2018;8(1):1432.
- Boyce GR, et al. Psychoactive plant- and mushroom-associated alkaloids from two behavior modifying cicada pathogens. Fungal Ecol. 2019;41:147-164.