Friday, May 25, 2012

Neuroscientists should study Zombie Ants

Zombie ant controlled by fungus (source)
The fungus-controlled zombie ant is one of nature's greatest wonders. A fungus (e.g. O. Unilateralis) is inhaled by an ant (e.g. Camponotus Leonardi), and begins to grow inside its body.  Eventually the fungus infests the brain of the ant, causing it to drunkenly wander, periodically convulse, climb up a leaf and clamp down on its ridge. Once the ant is securely in place, the fungus devours the brain and innards of the ant and grows out the back of its head often (but not always) releasing its spores onto the ground below. Un-freaking-believable, right?

As if this wasn't amazing enough, it's not like it is only one fungus species that infects only one ant species. There are many of these fungi and they infect many different kinds of insect, but somehow maintain a species specificity. In other words, fungus#1 can infect SpeciesX, but not SpeciesY, and Fungus#2 infects SpeciesY, but not SpeciesQ, and so forth. 

So WHY does this happen? and HOW has no one looked at the brain cells of these ants? 

Though no one has looked at the brains of these ants, Last year a paper painstakingly characterized their behavior under 'fungi control'. The most interesting characteristics are:
  1. The ants display a 'drunkard's walk' (the author's words)
  2. The ants periodically spasm and fall down (if they are above ground level)
  3. The ants clamp down on the underside main vein of a leaf (never the side of the leaf, never the top) Interestingly they all bite down on the leaf around solar noon.

Figure 1, Hughes et al., 2011

This figure shows the behavior of several ants.  Each ant was observed during the time of the horizontal blue bar.  The black vertical lines and 'spasms' which caused the ants to fall down (gray stars), and the red triangles are when the ant bit down on the leaf ridge. 

Because we have no idea how the fungus is manipulating the ant, let's wildly speculate.

1. The Drunken Walk:
Why: The reason for this is not clear.  The ant doesn't go far, so the non-directional walking could be to keep it close to more ants.
How: The mechanism is also not clear, but usually an ants directional walking could be following a pheromone trail. The fungus could presumably cause random walking by confusing the ants ability to sense pheromones. It could possibly even cause 'hallucinatory' pheromone sensing.

2. The Periodic Spasms:
Why: The authors speculate that the purpose of these spasms is to keep the ant near the ground.  The infect ants spend much more time on the ground level than the uninfected ants, and the spasms are often followed by a fall.
How:A fungus could essentially cause a seizure in the ants brain by manipulating potassium or calcium channels. On the other hand, I suppose the fungus could be acting directly on the muscles, causing them to twitch in an uncontrolled way. 

3. The Clamping:
Why: This has an obvious function, to root the ant for ultimate fungal growth and dispersion. 
How: First of all, biting and even walking on leaves is not something these ants normally do. So the fungus isn't just hijacking a behavior that the ant already has, it's basically creating a new one.  The correlation with solar noon indicates that a light or heat signal could contribute to the trigger, but basically nothing else is known about it. Interestingly, the clamping does not always have to be one single event either.  A few of the ants clamped down on the leaf vein more than once. The authors of this paper spend time discussing fungi's direct effect on the mandible muscles of the ant.

Figure 3 Hughes et al., 2011
They show that the mandible muscles of the normal ant are fat and healthy (B), but the muscles of the infected ant are separated and reduced in size (C). Though this image is of an ant at the moment of biting, the authors suggests that the deterioration of the mandible muscle might be to prevent re-opening of the clamp. They do not speculate on how the clamp is initiated in the first place, or why it occurs at noon.

So please, fellow neuroscientists, somebody stain these brains! It's just too fascinating to resist exploration. What proteins are altered? What is the receptor composition of behaviorally-specific neurons? Are the dendrites differently shaped?
And who knows what sort of great advances might be hidden in these brain-controlling fungi. The magic of optogenetics comes from lowly light-sensitive bacteria, just think of the possibilities hidden in brain-controlling fungus. 

To be fair, some neuroscience has been done on parasitic brain control, but it is very limited.  In fact it is limited to basically one histological study about parasitic worms who infest crickets and cause them to drown themselves (the subject of a future blog post). However, suicide-crickets are no zombie-ants and the exact mechanisms of the interaction is not likely the same.

© TheCellularScale Hughes DP, Andersen SB, Hywel-Jones NL, Himaman W, Billen J, & Boomsma JJ (2011). Behavioral mechanisms and morphological symptoms of zombie ants dying from fungal infection. BMC ecology, 11 (1) PMID: 21554670


  1. Many years ago, I learned in a parasitology class about a sheep tapeworm that uses ants as an intermediate host. The story goes (I haven't looked this up to confirm it, but this is just an extreme example of a pretty common story in parasites) that once a threshold population of worm larvae have built up, a few of them migrate into the ant's head, eat FOUR PARTICULAR CELLS there, and reverse the ant's aversion to light, which makes many of them climb to the top of nearby grass stalks, where they have a good chance of being eaten by the sheep that are the definitive host of the tapeworms. Acanthocephalans achieve a similar response in their hosts, but appear to do it chemically. The last I heard, they were being studied by pesticide company researchers trying to develop a "pre-spray" that would get a lot of targeted pests into the open just before actual spraying.

    There's a lot of stuff to study out there...

  2. this means new adavances in biological warfare

  3. Michael - Fascinating that the specific destruction of 4 cells might lead to such a distinct behavior. I have heard about that sheep-eating-grass-getting-parasites thing, but I don't know a lot about it. I'll have to look further into it.
    Anonymous - I don't know if this could be applied to biological warfare (yet) because it is so specific. A fungus that creates a zombie ant can't act on a cricket or moth. Each fungus is species-specific. I suppose it could be possible to study the fungus and target it to a particular species, but I don't think enough is known yet.