Chemical puppeteers

There's more chemical warfare going on, but you will have to look carefully at ground level to observe them. Large black carpenter ants of the genus Camponotus can be found scurrying around everywhere. Occasionally, a keen and persistent observer will notice some ants abandoning normal colony activities and climbing vegetation to specific heights. They have been attacked by falling fungal spores of Ophiocordyceps unilateralis that can chemically penetrate the formidable insect cuticle. These ants will bite firmly onto leaf veins (called the "death grip") and remain fixed until they die. The fungus multiplies inside the ant and, after death, it sprouts fruiting bodies from the ant's head or thorax, releasing spores onto other ants below. And so, the cycle of chemically creating 'zombies' continues.

These bizarre and gruesome life cycles, which involve manipulating the behaviour of the predator's food, are not uncommon. Cats can shed a protozoan parasite, Toxoplasma gondii, that can infect rodents. The parasite manipulates the mouse's brain to suppress its fear of cat odours, making it an easy prey in a cat-and-mouse game. Some locusts can be infected by nematodes, guiding them to dive underwater. Best known is Paragordius tricuspidatus, commonly known as a horsehair worm or hairworm. The nematode's eggs hatch into larvae in water. These aquatic larvae infect terrestrial insects when they ingest water or aquatic hosts that are infected. Once matured within the terrestrial insect host, the worm manipulates host behaviour. Infected locusts and grasshoppers are compelled to seek water, even diving into ponds or streams, drowning themselves. Adult nematode worms emerge from their host to reproduce, completing their life cycle.

Creating 'zombies' is not a simple process and requires time, even millions of years. Wasps and fungi need to develop surgical and chemical penetration strategies. Their hosts are not passive prey. They mount formidable defences that resist entry, and a strong immune response to deal with the infection. 'Summit' behaviour, where insects climb up vegetation to die and spread fungal spores, is documented in the fossil record as far back as 48 million years ago. Fossilised leaves have been discovered with insect remains showing characteristic "death-grip" biting marks induced by Ophiocordyceps-like fungi. Fungi evolved specialised abilities to infect insects from ancestors that initially infected plants or simpler arthropods. Remarkably, such behavioural manipulation traits (summiting, death-grip) have evolved independently in diverse fungi. This independent evolution of similar behaviours across distantly related fungal groups strongly suggests adaptive pressures favouring this strategy for spore dispersal and survival. Each fungal species has a particular insect host, indicating co-evolutionary arms races between fungi and insect hosts, with fungi constantly adapting to overcome insect immune defences.

While dedicated naturalists have observed and described 'zombie' behaviour and 'summitting', such as seen with the carpenter ant, for about two centuries, it is only recently that the exquisite details of the chemical manipulation of the host's brain have been understood, with the mechanisms of how fungi manipulate the behaviour of the host insect being best studied. This knowledge has come from combining studies from natural history with the use of laboratory-based organisms.

Just as ants can be manipulated by a fungus, the common housefly can be infected by another fungus, Entomophthora muscae, which causes the fly to exhibit characteristic "zombie" behaviour. Like the ants, infected houseflies climb upwards to elevated positions, where they eventually die. It is challenging to investigate the underlying details of the chemistry and brain manipulation of infected houseflies due to the limited tools available for this natural host species. Getting a large number of houseflies to study the chemistry of infection is not feasible. Nor is the brain of the housefly well understood, a prerequisite for understanding how the brain is manipulated by the fungus. Working at the University of California, Berkeley, Michael Eisen and colleagues identified a strain of E. muscae capable of infecting the genetically accessible model organism, the fruit fly Drosophila melanogaster, which has provided researchers with powerful tools to dissect this phenomenon.

Following up on this lead, Benjamin de Bivort and his colleagues at Harvard University used the genetic versatility and well-mapped neural circuitry of fruit flies to uncover neural and molecular mechanisms underlying summiting. They first developed a high-throughput assay to systematically monitor summiting behaviour, identifying that the hallmark of this behaviour is a specific burst of locomotion occurring about 2.5 hours before the fly's death.

SCAN & WATCH

The fruit fly provides us with a versatile toolkit to dissect what parts of the nervous system contribute to specific behaviours. For example, it is possible to very specifically inactivate a chosen set of neurons or nerve cells and observe the effects on behaviour. Similarly, chosen neurons can be activated, and their effects can be observed. Using such extensive screens involving neural silencing and genetic knockdowns, the researchers pinpointed critical circuits within the fly's brain involved in behaviour upon infection by the fungi. Key to this are two aspects of the fly's brain. The first is the circadian clock system, involved in regulating day/night changes in activity. The second pair involves neurons that project to the gland responsible for synthesising juvenile hormone, which is important for many aspects of development and function.

SCAN & WATCH

Silencing these neuronal pathways drastically reduced summiting, while artificially activating them induced a summiting-like locomotor burst. This confirmed the essential role of these neurons in driving this manipulated behaviour. Through targeted genetic ablations of these neurons and pharmacological manipulations, they further confirmed the crucial role of these neurons and juvenile hormone production.

Next, using real-time machine learning classifiers, they precisely identified and isolated summiting flies for molecular and anatomical studies. Surprisingly, despite fungal invasion and substantial host tissue damage elsewhere in the fly body, crucial neural structures, including the circadian rhythm pathway and juvenile hormone pathway, remained structurally intact during summiting. This suggests that the fungus strategically preserves specific neural circuits critical for its own dispersal strategy.

To investigate how the fungus may influence neural circuitry from a distance, researchers examined the integrity of the blood-brain barrier. They found increased permeability during late infection stages, suggesting that fungal factors circulating in the fly's haemolymph (blood equivalent) might enter the brain to trigger neural changes. Indeed, analyses of the components of the haemolymph showed distinct profiles in summiting flies compared to controls. Importantly, transfusing haemolymph from summiting flies into non-summiting recipients triggered a summiting-like increase in locomotion, providing strong evidence that circulating fungal factors directly induce this manipulated behaviour.

Such studies leverage the extraordinary genetic, molecular, and neurobiological toolkit available in fruit flies, combined with sophisticated behavioural assays, machine learning classification, and metabolomics. These approaches collectively provide unprecedented insight into how E. muscae manipulates host behaviour. Specifically, the fungus appears to alter the host's haemolymph chemistry, which then affects circadian and neuroendocrine pathways in the brain, ultimately causing the host to exhibit behaviours advantageous to the fungus. The use of Drosophila genetics thus transforms what was once a mysterious natural phenomenon into a clearly defined model system for unravelling complex host-parasite interactions at a mechanistic level.

Fungal manipulation of insect behaviour represents an extraordinary evolutionary adaptation, exemplified dramatically by parasites such as O. unilateralis and E. muscae. These fungi are capable of altering the neural functions and behavioural patterns of their hosts with remarkable specificity, effectively turning insects into living vehicles to ensure their transmission and reproductive success. This intricate manipulation requires the production of a suite of specialised bioactive compounds that precisely target insect neuronal circuits and penetrate protective biological barriers, analogous to the mammalian blood-brain barrier.

This natural phenomenon presents exciting possibilities for biomedical research, particularly in the fields of neurology and pharmacology. The neuroactive compounds secreted by these fungi, which are capable of breaching the insect equivalent of the blood-brain barrier and modulating specific neurotransmitter pathways, hold significant promise as sources of novel bioactive molecules. Indeed, because neurotransmitter systems such as dopamine, serotonin, and GABA (Gamma-Aminobutyric Acid) are evolutionarily conserved across insect and mammalian taxa, compounds that manipulate these systems in insects could provide valuable molecular templates for developing therapeutic agents targeting human neurological disorders.

One of the most significant challenges in neuropharmacology is delivering therapeutic agents selectively and efficiently across the human blood-brain barrier — a formidable barrier that protects the brain but also severely limits the delivery of drugs to central nervous system (CNS) targets. Fungal-derived molecules, evolved explicitly to overcome similar barriers in insect hosts, could inspire novel strategies or chemical scaffolds capable of penetrating or temporarily modulating the human blood-brain barrier without adverse effects. Such a strategy would significantly enhance our ability to treat neurodegenerative diseases, psychiatric conditions, and disorders requiring highly targeted modulation of neuronal populations. The insect blood-brain barrier has, remarkably, many features similar to that of humans. 

However, despite their enormous potential, translating these fungal neuroactive substances into clinically viable drugs is not straightforward. Evolutionary divergence between insects and mammals necessitates the rigorous study of the precise receptor interactions, pharmacodynamics, and safety profiles of these compounds. What is effective and safe in an insect model may not directly translate to human applications without extensive modification and optimisation. Additionally, the inherent toxicity or unintended immune responses to fungal metabolites must be thoroughly characterised to ensure their safety and efficacy in human therapeutic contexts. Whatever the outcome of such efforts at application, the understanding of the wonders of natural history at the level of chemical biology of the brain is a wonder in itself.

ALSO READ
Elya, C., De Fine Licht, H.H., and Eisen, M.B. The fungus Entomophthora muscae hijacks Drosophila circadian rhythms to enhance pathogen transmission. Current Biology, 31, 4889–4902 (2021).

Elya, C., Lavrentovich, D., Lee, E., Pasadyn, C., Duval, J., Basak, M., Saykina, V., and de Bivort, B. Neural mechanisms of parasite-induced summiting behavior in 'zombie' Drosophila. eLife 12:e85410 (2023). bit.ly/zombie-summiting

Fredericksen, M.A., Zhang, Y., Hazen, M.L., Loreto, R.G., Mangold, C.A., Chen, D.Z., and Hughes, D.P. Three-dimensional visualization and a deep-learning model reveal complex neural and muscular interactions in zombie ant behavior induced by fungal infection. J. Exp. Biol. 220, 2892–2900 (2017). bit.ly/3D-parasite