I recently had the opportunity to write a post for Nautilus on a subject that is dear to me - the use of crows and other intelligent members of the corvid family for neuroscience research. Corvid intelligence has been noticed by humans for millennia, and more recently by ethologists and psychologists. The fascinating thing about these animals is that like all birds, they do not have a neocortex - the part of the mammalian brain that has countless times been implicated in intelligence. Now, there is just one lab in the world - Andreas Nieder at the University of Tübingen - that has started peering into the brains of these fascinating creatures to try to understand how crows’ cortex-less brains enable them to perform amazing cognitive feats. You can read the full story on Nautilus.Read More
In my research on the rat visual system, I have been designing an apparatus that would allow me to record neuronal responses to visual stimuli in freely moving rats. Most visual neuroscience experiments are now performed on restrained animals, who are usually treated with different drugs to suppress movement (anesthetics, muscle relaxants). But as anyone who has tried reading while falling asleep knows, just because your eyes are open does not mean that information is getting through to the brain. It makes more sense to study how neurons respond to images when the research subject is awake and paying attention.
While few researchers are studying vision in unrestrained rats today, I was surprised to find that the basic setup I have been working on for my experiments had already been created — in 1980’s Soviet Russia.
Working at the Moscow State University, Sergei Girman wanted to study the visual system in freely moving animals. So Girman chose to perform his experiments on rats, noting two features that made them convenient to use - “the eyes in this animal are relatively immobile,” making it easy to know where they are looking (researchers go through a lot of trouble training a monkey to look at computer monitors in visual experiments), "while the visual analyzer is well developed” (analyzer being perhaps the fashionable word of the time to refer, in this case, to the visual areas of the brain).
The goal of Girman’s 1985 paper was to compare the responses of neurons in primary visual cortex in awake and attentive rats compared to those in restrained, anesthetized rats.
He outfitted the rats with a metal platform, glued to the head, that contained electrodes to record neural activity. He then trained the animals to enter a vestibule that would constrain the head, so that when the visual stimuli were presented on a computer screen, the head would always be in the same position. While this may sound gruesome, by today’s experimental standards Girman’s procedure was quite original.
At the time, most experiments in visual neurophysiology were done in one long session - researchers would prepare the animal, insert electrodes and present visual stimuli for hours on end; at the end of the session, the electrodes were removed and the animal euthanized (in some cases the experiment could be repeated, but the electrodes would be placed in different brain areas each time). Girman was interested, in part, in recording the activity of neurons for long periods of time. If a neuron responds to particular visual stimuli today, would it respond to the same stimuli in the same way again, tomorrow? To be able to answer questions like this, Girman needed to construct an apparatus that would stay on the animal’s head for months without causing so much damage that the body would reject the implant.
Today, researchers routinely record neuronal activity in rodents for months at a time (techniques such as calcium imaging allow one to examine the activity of the same neurons from one day to the next), but the surgical procedure of attaching head implants is quite drastic. In most cases, the animals are scalped (nonviolently, of course), holes are drilled in the skull, and after the electrodes are inserted into the brain, the scalp is replaced with glue (usually dental cement). Researchers take great care to perform such procedures in a sterile environment to reduce the risk of infections. Inevitably, however, after months (or in lucky cases, perhaps a year), the implants fall off.
Girman’s original solution to this problem was to not scalp the animals in the first place. Instead, he would only make holes large enough for the electrodes to pass through (0.12 mm, according to his paper). Then, he would create a platform for the electronic equipment by threading stiff metal wires under the scalp. While this sounds like a less invasive solution, it must be quite difficult to perform (although I haven’t tried it in my own experiments yet).
Girman’s papers are quite fascinating, partially because of his unique methods, which make me wonder if they really do work better than today's established techniques. Is it just the no one read Girman’s work (which was originally published in Soviet Russian journals and translated later)? Or is Girman's idea of keeping the scalp intact really not all that better than removing it?
It is both encouraging and frustrating to learn about obscure research techniques: the wheel does get reinvented over and over, but perhaps we learn something new each time.
(Please tweet us @harvardneurosci if you have trouble accessing the paper)