Zombie Ants Disease In The Amazon Rainforest

In the Amazon Rainforest, a carpenter ant is acting strangely. In the night, the worker ants march in a neat foraging trail and search the forest floor for leaf litter and dead insects. One ant, however, abandons the group and begins to climb a nearby tree. It ignores the pheromones released by its kin as if its mind is being controlled by some nefarious external force. When it reaches the top of a branch, it bites down firmly on a leaf with its mandible and remains there until it dies. The next day, around noon, the ant’s exoskeleton bursts open. A sprout of fungus erupts from its head, releasing millions of spores onto the forest floor below. It sounds like something out of a science fiction novel, but this ant is a victim of Ophiocordyceps unilateralis: the zombie ant disease.

Ophiocordyceps was discovered by Alfred Russel Wallace in 1859 in Indonesia (Wallace 1869) and was first scientifically described by Thomas Petch in 1931 (Petch 1931). Petch described the Cordyceps species as fungi that grew from tropical insects that were found not in their usual soil or wood habitats, but attached to branches or leaves in the jungle undergrowth. Currently, there are over 140 known Ophiocordyceps species that infect a variety of arthropod insects (Kirk et.al 2008), and new species are still being discovered (Araújo el. al. 2015). In this review, I will focus on the species complex of Ophiocordyceps unilateralis which is a parasite of ants (Mongkolsamrit et. al. 2012).

Social insects such as wasps, bees, and ants, represent only 2% of the known insect species on Earth but compose more than half the biomass. Ants are susceptible to many parasites, but Ophiocordyceps are especially well adapted to infecting the abundant arthropods (Araújo et. al. 2015). O. unilateralis infects ants on the humid, moist forest floor: a perfect environment for fungal growth (Mongkolsamrit et. al. 2012). It is an entomopathogenic fungus which infects its ant host when a spore contacts the cuticle of the exoskeleton (Araújo et. al. 2015). The spore germinates and produces an adhesive pad called an appressorium which then develops into a germ tube. This germ tube penetrates the ant’s exoskeleton and inside the ant’s body, the fungal cells rapidly proliferate until the number of fungal cells outnumber the ant’s own cells. Full pathogenesis is characterized by the alteration of the ant’s behavior: when it leaves its normal environment and climbs to an elevated position. The ants use their powerful mandibles to affix themselves to the underside of a leaf or branch as the fungus enters its reproductive stage. Experts in the field refer to this affixation as “the death grip”. In the final act, a fruiting body consisting of a long stalk, the stroma, and an elevated spore-producing structure, the ascoma, erupts from the ant’s head. This conspicuous protrusion releases mature fungal spores, ascospores, which fall widely on the soil below. The entire process, from infection to death, takes only 4-10 days (Mongkolsamrit et. al. 2012).

Ongoing research focuses on several aspects of the O. unilateralis fungi. The prevailing questions in the field are as follows: how did the fungus evolve, how is its metabolism unique, and how does it manipulate the ant’s behavior? One notable researcher who is working to answer these questions is Dr. David P. Hughes. Hughes is a professor of entomology and biology at the Penn State college of agricultural studies (Zimmer 2019). His research focuses on the genomics and transcriptomics of behavioral manipulation in O. unilateralis. Dr. Hughes is an extremely relevant researcher in this field not only because of his academic pursuits but also because he was a professor and mentor for several other important researchers in the field including Dr. Christina de Bekker and Dr. João Araújo. Dr. Hughes’ work has received significant media attention, from televised specials on BBC to headlining articles in the NY Times. Finally, he is also a “zombie expert” and has consulted with Hollywood directors on several zombie shows and movies including World War Z. Dr. Hughes’ media influence has brought more attention and funding to Ophiocordyceps research, although fungal-insect associations are still understudied compared to fungal-plant associations (de Bekker et.al. 2017).

Dr. Hughes former student, Dr. Christina de Bekker, heads a lab at the University of Central Florida which performs controlled experiments to demonstrate the importance of synchronized events to O. unilateralis infections. De Bekker observed how fruiting bodies tended to erupt from the ant exoskeleton around noon and hypothesized that circadian rhythms may play a role in infection (de Bekker et al. 2014). In 2014, de Bekker and her colleagues designed an experiment to establish the importance of heterogeneous fungal cell activities and temperature/light rhythms in O. unilateralis infections. The team injected individuals in a Camponotus ant population with O. unilateralis and kept the colony in a controlled environment with strict 24-hr light-dark and warm-cool rhythms. Tissues were extracted from infected ants’ brain and muscle tissues at various times during the cycle and were analyzed using RNA sequencing. De Bekker and her team found that in the absence of fixed light and temperature cycles, full pathogenesis was never reached and the ant’s behavior was not manipulated. However, the brain chemistry of the infected ants was similar with or without circadian rhythms. This indicates that circadian rhythms do not affect the fungal metabolism, but potentially act as some kind of trigger to behavior manipulation, death, and spore release. This experiment also found that fungal cells had heterogeneity in different tissues, secreting differing metabolites depending on their location. This finding was significant because it showed that the colonizing parasites were organizations with divisions of labor. In future experiments, the complex interactions between the parasitic cells and the ant’s variable tissues were considered.

In 2017, de Bekker and her team conducted another experiment where they compared the genomes of five Ophiocordyceps species, including O. unilateralis (de Bekker et. al. 2017). They compared the overlap of orthologous clusters between the Ophiocordyceps species and other ascomycetes to develop an understanding of the pan and core genomes. Using RNAseq and BLASTp, they predicted the genes and hypothetical proteins that were unique to the Ophiocordyceps. They found a considerable number of unique small secreted proteins or enterotoxins. However, these proteins were not highly conserved across the five groups, suggesting the possibility of convergent evolution of similar proteins. They also performed biological assays to determine the concentration of these enterotoxins in ant tissues during manipulated biting behavior. Amazingly, some of the enterotoxin orthologs were upregulated over 3,000-fold in mandibular tissue during behavior manipulation. De Bekker hypothesizes that these enterotoxins could be the factor that modulates the ant’s behavior, but more research would be needed to support that conclusion.

Dr. João Araújo, another former student of Dr. David P. Hughes, focuses primarily on fieldwork, researching most prolifically in the Brazilian Amazon. He is responsible for identifying several O. unilateralis species. His most recent work provides insight into the evolutionary history of Ophiocordyceps via ancestral character state reconstruction analysis or ACSR (Araújo and Hughes 2019). In their 2019 experiment, Araújo worked with Dr. Hughes collecting samples of fungal pathogens from ants and other related species such as beetles and caterpillars. Their work indicated that the zombie-ant fungus likely evolved from an ancestral species that infected beetles. There are no known beetle fungal pathogens that manipulate host behavior in any way. Araújo and Hughes hypothesize that Ophiocordyceps evolved to manipulate host behavior because of the selection that occurred in the ant’s environment. Since ants are social insects, they are very susceptible to fungal infections as they live in close proximity to each other. All known Ophiocordyceps species can only enter their reproductive stage in a dead host. The issue here is that when an ant dies it is quickly removed from the colony by other ants and placed in a secluded area away from the colony referred to as an “ant cemetery”. The previous research of de Bekker demonstrated that the fungal spores erupt around noon, generally over an hour after the ant’s death (de Bekker et. al. 2014). To overcome this problem, Araújo and Hughes hypothesize that selection favored behavioral manipulation of the ant (Araújo and Hughes 2019). By manipulating the ant’s chemosensory recognition, the fungus could separate the ant from its colony before death, ensuring that the fungus would not be placed in an ant cemetery before releasing its spores. Also, the ant is manipulated to secure itself high above the ground so when the spores are released they can coat a wide area of moist soil where other ants are likely to pass over.

Another 2019 experiment by Dr. Raquel Loreto and Dr. David Hughes revealed that although the ant’s musculature is extensively atrophied and invaded by fungi, the brain of an ant infected by O. unilateralis is not invaded by any fungal cells (Loreto and Hughes 2019). During experimentation, Loreto and Hughes injected ants with O. unilateralis. At the exact moment the ants initiated their bite behavior, before their death, the ants were flash-frozen in liquid nitrogen. Once sufficient samples were collected, the ants were defrosted and dissected, and metabolites were extracted from the individual brains. Extracted metabolites from all the ant brains were then compared then to a normal metabolic profile which Loreto and Hughes established by performing dissections and analysis on uninfected ants. Histological work showed no invasion of the neural tissue by fungal cells but the significant alteration of several metabolites involved in neural modulation and signaling. Loreto and Hughes hypothesize that these metabolites are excreted by the fungus and pass through the blood-brain barrier to the central nervous system of the ant. Anti-oxidative metabolites were significantly upregulated in the brains of infected ants, suggesting that the fungus is actively preserving the ant’s neural tissues. Loreto and Hughes hypothesize that this allows the fungus to manipulate the ant’s behavior. This experiment is significant for future research because if Loreto and Hughes are correct in their hypotheses, the O. unilateralis is manipulating ant behavior while acting like a microbe outside of the brain, making the zombie ant a useful model to consider in research of other pathogens. There is also importance in this question for human neurological research. If the Ophiocordyceps Unilateralis is in fact producing proteins or metabolites that cross the blood-brain barrier and affect the brain’s chemistry while simultaneously actively preventing neurodegeneration, this could be exploited to develop drugs effective in treating human brain diseases.

At the same time as the aforementioned experiment, Dr. Hughes was also working with researchers from the College of Plant Protection in China. The experiment, led by Shanshan Zheng, compared metabolite concentrations in the mandibular tissues of O. unilateralis infected ants and uninfected ants using HPLC-MS (Zheng et. al. 2019). This experiment, unlike previous experiments, included a positive control in the form of B. bassiana, an ant fungal pathogen that does not alter behavior. The inclusion of a positive control allowed researchers to make more concrete conclusions that metabolite concentrations in ant musculature were due specifically to O. unilateralis infections and not fungal infections in general. Zheng and their team observed significantly elevated levels of sugars, purines, ergothioneine, and hypoxanthine in infected ants. They also observed a downregulation of AMP. The purine pathway was identified as the most significantly altered metabolic pathway, as adenine, guanine, uracil, and hypoxanthine were all notably up-regulated in O. unilateralis ants. In ants that were analyzed during the death grip behavior, purine levels were even higher. Histological data revealed that the number of mitochondria in infected tissues was dramatically lower than in healthy ants or ants infected by B. bassiana. These observations led Zheng and their team to hypothesize that O. unilateralis stimulates ATP breakdown, producing hypoxanthine and AMP. The AMP is then rapidly depleted as it is used for cellular processes. This suggests that O. unilateralis likely uses large amounts of glucose as it grows. However, this experiment could not provide conclusive evidence that one particular metabolite induced the notorious death grip.

Ongoing research has provided important insight into the complex metabolism of O. unilateralis and its manipulation of ant behavior. We understand the importance of circadian rhythms to the fungal life cycle (de Bekker et. al 2014), the possible role of enterotoxins in behavioral manipulation (de Bekker et. al. 2017), and the alteration of metabolic pathways in infected ants (Loreto and Hughes 2019, Zheng et. al. 2019). We have also identified that O. unilateralis may have evolved from a beetle-infecting fungal ancestor (Araújo and Hughes 2019). However, several questions remain. It remains unknown how O. unilateralis effects ant’s behavior and what specific metabolites are responsible for the disruption of chemo-sensation and the death grip. We also have identified that zombie ants may provide a model that could aid in the research of other pathogenic species and neurological research of the human neurodegenerative disease. Finally, Dr. Hughes expertise as a “zombie expert” illustrates the question that has caused a popular fascination with Ophiocordyceps unilateralis: if something like that evolved to control the minds of ants, could something evolve to control the minds of us?