“Neuroscience has been stuck”
Sebastian Seungs quest for the connectome
Alzheimer, autism, schizophrenia – the public expects neuroscience to explain these diseases. But brain researchers don´t even understand how neurons in a fly´s brain are activated by a moving stimulus. Sebastian Seung advocates a way out: connectomics.
An interview with Sebastian Seung, Professor of Computational Neuroscience at the Massachusetts Institute of Technology, MIT, in Cambridge Boston (USA), and author of the book „Connectome: How the Brain’s Wiring Makes Us Who We Are.”
Your TED-Talk was watched by over half a million people. Why do you think the public is so interested in the connectome?
People know that we don´t understand the brain. The idea that we could have this fantastic map with an incredible resolution sounds important and exciting. But I think for the public it is very difficult to understand what a connectome really is.
Is that why you wrote your book?
There were two goals. One goal was to address the public because there is tremendous interest in the brain, and yet almost nothing that is written for the public contains real neuroscience. I wanted to take the public on a journey, so that there is a better understanding of neuroscience and what is really happening at the frontiers. The second reason I wrote the book was to address my colleagues.
What is in there for them?
I think that neuroscience has been stuck, and that many of my colleagues don´t have any idea what it means to understand the brain. Most of neuroscience is just description; we rarely ever explain anything. It reminds me of a naturalist in the 19th century who journeys into the world and describes the wondrous diversity of species, but fails to explain how living things function. That’s the situation neuroscience has been in. I think there is a certain direction we need to go in, and I feel strongly enough about my beliefs, that I laid it out in a book.
Where is the understanding of the brain stuck?
For example, in 1956 the German scientists Werner Reichardt and Bernhard Hassenstein proposed a theory of how the insect visual system detects motion. For decades nobody could test their theory. We could describe the activity of the involved neurons, yet had no idea how motion detection happens. This shows the incredible disconnect between the reality of neuroscience and what the public expects of neuroscience. We are supposed to find cures for autism or schizophrenia, even though we failed to explain how neurons in a fly brain get activated by a moving stimulus. Our field developed a culture so accustomed to failure that we no longer even recognized that we were failing. But now there is hope, because connectomics and other new technologies like genetics are becoming powerful enough to show a way out.
In which way?
In August, Nature had a cover story on connectomics, which presented three papers. Two were about motion detection in the fly brain, and used connectomics and genetics to bring us close to a solution finally. These papers proved that connectomics can lead to important scientific discoveries. I was coauthor of a third paper about a connectomic analysis of the mouse retina. This demonstrated the feasibility of applying connectomics to the nervous systems of mammals, including potentially our own.
How can connectomics help us to better understand the brain in the next decades?
There are two obvious questions, one of them is perception, the other memory. How do we perceive objects and motions? We are starting to answer this. The second question, what changes inside the brain when we form memories, is surprisingly difficult to answer. There are theories, but we would actually like to see how the patterns of connections change inside the brain when we learn new things. The third question addresses mental disorders. That’s the hardest one. Can we see the pathology in mental disorders such as schizophrenia or autism?
Many scientific efforts have failed to explain mental disorders. Aren´t you promising too much?
I don´t promise a cure, but a way of understanding. Most people try to find treatments without understanding the disease. We need to understand schizophrenia to have better means to find therapies. And understanding will start by seeing something that is wrong inside the brain. For over a hundred years neuroscientists have looked for clear and consistent pathology inside the brain of schizophrenics, and they have never found it. Again, there is a backlog of 100 years of failure, and connectomics will offer new insights.
What are the future goals of connectomics?
One proposed project is to map every connection in the mouse brain. If the neuroscience community made a serious investment it could be done within ten to fifteen years. The other project would be to map all the long-range connections in an entire human brain. But I want to emphasize, we don´t need to do an entire brain to learn interesting science. Because we know so little, every little piece is telling us something. In the next few years you are going to see all kinds of discoveries coming out which are informed by connectomics.
Your critics argue your approach is lacking a theory and is therefore wrong.
The counter example to that is the Human Genome Project. We did not need a particular theory in mind to go sequence the genome. Now every researcher in biology uses the genome, and gene therapy is starting to happen now. The genome is simply essential information in every area of biology; it is not driven by a particular theory. The same will be true of the connectome. It will provide so much information that every neuroscientist will need to know about it, and make use of it. I think, the time for a philosophical debate on the importance of connectomics is over, because now there are concrete accomplishments to point at.
The interview took place in August 2013, and was part of an article on the connectome, which appeared in the Neue Zürcher Zeitung.