The notion of dark matter, as the name suggests, conjures up a feeling of the ethereal and the mysterious. One probably first thinks of the hypothetical material that makes up 85% of the universe’s density. That is what came to my mind upon seeing a new paper expanding on a theory of the dark matter of the brain. The paper, published last month in Brain Structure and Function, aims to convince the reader that large swaths of neurons in the brain remain “silent” throughout the lifespan. These silent ensembles of neurons make up this so called dark matter.
Historically, this is not the first time a biological phenomenon has been given the label of dark matter. The Human Genome Project discovered that up to 98% of the genome is non-coding DNA — in other words, DNA that does not code for a protein. Even more, the glial cells that make up the majority of cells in the brain were once thought to be merely structural, essentially silent “dark matter” of the brain. We now know that even the non-coding regions of DNA perform cellular functions that scientists are still elucidating. And glial cells have been found to be integral to brain function.
The paper is by Saak Ovsepian, who has multiple academic appointments at more neuroscience institutes in Europe than any author I have ever seen. In the paper, Ovsepian describes a population of neurons in the brain that remain essentially silent — they do not fire action potentials and do not communicate to downstream neurons. He builds this theory on two basic principles: 1) that there is a mismatch between the required energy consumption of the brain given its small size and the actual consumption of energy. Paradoxically, the brain consumes over 20% of total body energy, yet comprises 2% of body weight. He suggests that this mismatch is due more to neurons that inhibit other neurons (inhibitory interneurons), which are more active than their counterpart, excitatory neurons. Thus, if inhibitory neurons make up most of the firing in the brain, then there should be vast populations of neurons that are not firing — because they are being inhibited. 2) Electrophysiological recordings and functional imaging in experimental models reveal that more than half of neurons in the brain are functionally silent. The latter point above stems from the first point, so we’ll focus on that here.
What do we make of this new addition to the list of mysterious, if just currently unknown, phenomena? It definitely hearkens back to the myth that we only use 10% of our brains. However, the problem I have with this theory is the second point from above: that current functional recordings (whether electrophysiological or imaging-based) demonstrate the existence of silent neurons. Ovsepian’s premise, and he provides a significant number studies to support his claim, is that there are populations of neurons in the brain that when recorded in the presence of various stimuli, never fire an action potential to the next neuron in the circuit. So, no matter how many ways you prod them, they do not respond. The obvious problem, I think, is that what if the experimenters simply did not try the correct stimuli for the particular neurons they were testing. It seems likely that on the laundry list of stimuli, they just hadn’t reached the correct stimulus. It would certainly be a long, tedious, and expensive experiment to try out every conceivable stimulus on every single neuron. No graduate student would sign up for that project. The author acknowledges this potential flaw in his theory:
“Another possible explanation for the presence of great numbers of inactive neurons is their narrow tuning to respond only to specific inputs [i.e. sparse coding]… but whether these considerations can explain the perpetual silence of the vast majority of neurons throughout the brain remains to be shown.”
Where I actually think this theory holds some water is in its prediction that we maintain ancient, evolutionarily dormant circuits. Ovsepian describes the acquisition of talents such as in autistic savants or relic behaviors as might be seen in our ancestors living thousands of years before civilization, that surface in patients with schizophrenia. The notion that inhibitory neurons act to keep these talents and behaviors in check seems plausible. The disinhibition that occurs during stressful times could uncover the remarkable abilities of savants as well as the detrimental behaviors seen in schizophrenia.
However, one could argue, and I would agree, that the above phenomena could be the result of being born with too many functioning neurons. The brain starts out with far more connections between neurons than it needs. As the organism develops, the number of connections is trimmed back. It is now thought that these extraneous connections result in the behaviors seen in autism and schizophrenia — no disinhibition of dormant circuits necessary; they are already there and functioning, albeit inappropriately. What is not clear in this model is why the gain-of-function behaviors surface later in life than one might expect from a developmental problem (e.g. schizophrenia does not usually become asymptomatic until early twenties).
Ovsepian’s theory does help explain this progression pattern. Consider a person born with a genetic makeup that predisposes him to schizophrenia. Let’s say the risk genes are involved in the myelination* of inhibitory interneurons. We have evolved so that our brains aren’t fully myelinated until around 30 years old. If this person’s brain myelin is a bit slower to develop than his peers and the brain has evolved so that by the age of 21 it should be functioning with a pre-specified amount of myelin, his brain’s threshold for managing a stressor may be lowered. This seems to be the case. Life stress rises around the age of 20, which is also the age of the highest incidence of schizophrenia. It seems possible then, that loss of inhibitory interneuron function, because of a known defect, could thus disinhibit circuits that in a “normal” person would not be functioning. The dark matter theory helps explain why the psychosis seen in these patients develops at this time, rather than earlier as might be expected.
All in all, I think the data supporting a dark matter — probably better described as a dark neuron — theory of brain function is a bit thin. On the other hand, the idea of ancient neural networks sitting idle adds a fascinating twist to the evolution of the brain. Perhaps one day, we will have tried every plausible stimulus for every circuit to determine if there is in fact a population of neurons lying dormant, waiting to be released.
*Myelin is a fatty layer that wraps around neuronal axons allowing them to fire signals faster and more efficiently.