Author: Abinaya Helbig Edited by: Ruchi Maniar
Dr. Bert Sakmann M.D., PhD. is a German cell physiologist who, with physicist Erwin Neher, was awarded a Nobel Prize in Medicine and Physiology in 1991 for their discoveries concerning the function of single ion channels in cells by the invention of the patch-clamp technique (1). This remarkable breakthrough established a new way of understanding membrane physiology and the ground-breaking invention of the patch-clamp technique enabled the study of individual ion channel proteins in real time. Sakmann and Neher examined cellular functions over a broad range, which led to the discovery of the important roles of ion channels in diseases (e.g. CVDs, NMDs, CF and diabetes) (2). Having concluded the important roles ion channels play in diseases, Sakmann and his colleague opened new concepts of using therapeutics targeting ion channels for disease treatment. Dr. Sakmann is also currently the Inaugural Scientific Director and Research Group Leader for the Max Planck Florida Institute for Neuroscience's (MPFI) (3).
1. Can you briefly describe how you got into your current field of work and did you face any major obstacle(s) working towards your research?
It was during the 1970s when I was a post-doctoral fellow and I was working with Professor Sir Bernard Katz in London. Katz and his colleagues had just discovered what is called the “membrane noise”, which is an increase in the background noise when you activate ion channels, and the big challenge at the time was to find out what was the basis of this membrane noise. I was very lucky to meet Erwin Neher (the co-recipient of the Nobel Prize), whom I met in Munich. We worked together in a very small independent group for about ten years in Gottingen, to identify the determinants of ion flow via ion channels and also to make some educated guesses about its underlying structures.
It took about ten-eleven years to describe the ion channels and the method in great detail. I thought I would do something else, because at the time I had two choices: either to go into the structure or take a look at the function of many ion channels that generate the electrical signals that make our brains think, our muscles move, and our glands secrete. So, I chose the latter of the two, because structure was very much dependent on luck at the time and it was dependent on the use of protein crystallization. I initially went to an institute in Heidelberg where experts conducted research using crystallization techniques, and here I decided that such techniques were not for me, because the whole technique was too static and you had to be very patient to get a crystal.
When it comes to challenges, I faced challenges all the time! For example, we had some initial hints that in principle, the size of the ion currents may be measurable if we improve our techniques, leading to further reduction of the noise. So, during the first two-three weeks, we had some promising indications, but to make my primary work to translate into something significant, it took about ten years of meaningful experimental work.
2. What research are you currently doing, and what obstacles have you faced trying to translate your work into the real world?
I worked in Florida on a detailed reconstruction of a part of the brain, the cortex, where my aim was to understand the representation of a stimulus in relation to anatomical and electrical signals. I started this is in Florida around five-six years ago, and it’s a bit more static because anatomy is something where you can basically stop whenever it pleases you. Physiology is much more exciting, but it is necessary with all the detailed anatomy to understand the cortex and that’s what I was working on.
Currently, I am working in Munich and I have an on-going research project to measure electrical activity of a stimulus representation in an awake animal with another postdoc colleague based in Amsterdam. I am also working on another research project that focuses on the nerves’ structural arrangement in the cortex, and we, as a team at Max Plank Institute, are hoping to create a 3D map of nerve cells to create a foundation to study neurological diseases.
3. How do you think your current work in digital neuroanatomy will impact and benefit the health care system - specifically the diagnostic and therapeutic aspects of medicine? Do you foresee any challenges?
Right now, it is still very cumbersome and it’s not a standard technique, but once we have developed methods to look at the detailed cellular anatomy, the faster we will progress with time. We will first look into diseased brains of experimental animals and this will be possibly the basis for using digital neuroanatomy in the diagnosis of human brain diseases. In the beginning, this can only be done in retrospect, as it cannot work in currently diseased individuals. However, in order to understand signals that you can measure in the human brain like fMRI (functional magnetic resonance imaging) or PET (position emission tomography), you need to have a detailed anatomy. Unless we have a clear understanding of the detailed anatomy this is all rather descriptive. So, I think it is still far but it will be the basis for interpreting signals that you can record from the healthy and diseased human brain. fMRI is basically a collection of colourful pictures but they all appear to reflect a cellular basis which we don’t understand yet. In order to comprehend these signals, we will have to do detailed anatomy, and therefore, it is very cumbersome and not very many people are ready to spend the time because it will be a long time before we gain any meaningful data. There are a few labs and I am very happy to be in a lab like in Munich with Winfried Denk, who has as an aim to reconstruct a mouse brain for the moment, but later for higher animals at EM (electron microscopic) resolution and that’s what is needed. We will take possibly a decade or so, but it all depends on the methodological advances in anatomy.
It is not a problem to have the anatomy; you can cut the whole brain, but you have to what’s called ‘segment it’, which is to say you have to assign to a histological cut the identity of particular cells and their function.
You can’t do this easily; you have to do a 3D reconstruction and that’s what I called digital neuroanatomy, a bottle neck term for what’s called segmentation.
4. For your current work in digital neuroanatomy you are said to be using a program to map the brain and construct a circuit system of the brain. Is this already in place or are you still working towards this and how do you think new software programs affect your work in mapping or illustrating neurons?
We are working on it, but the first thing to remember is that there are different levels of accuracy. I have been working with my collaborator who is at the light microscopic (LM) level, and this will set the framework for more detailed work at the EM level which is going to be the end product. We have developed a number of software tools to facilitate the reconstruction of neurons in a more efficient way, along with databases and tools that we have created to allow us to examine simulations of electrical signals conducting through the reconstructed networks.
5. Regarding your work on 3D mapping of the brain¸ how accurately does it resemble the human real brain? Are there any challenges that you have been faced with in digital neuroanatomy?
Currently we are using mouse brain models, and they are very different from the human brain. Right now, the information obtained from the mouse model cannot be applied to understanding of the human brain, but we learn to sharpen our techniques. The reason why we don’t use higher animals such as chimpanzees is because of ethical barriers. I am working on problems which are so basic that they probably apply to higher brains as well: like what is a structure of a neuron, do you have an excitability of the dendrites etc. Let’s take the example of the genetic code - it was first clarified for bacteria and then it turned out to have many rules that apply to all DNAs. The more basic you are the more general it is, and ion channels are all over the place and dendrites act on all places. The problem comes when you look at the interaction of many neurons and not of a few - few neurons we can perfectly well understand - but if you have thousands of neurons interacting, this is where a different start between a rodent brain and a monkey occurs.
In terms of obstacles, I would say to get a 3D model of the entire brain itself is a huge challenge, because that would be the basis of understanding electrical and also optical signals. It is more of a methodological challenge. The challenge in digital neuroanatomy and specifically circuit analysis can be summarised by the word ‘connectomics’, which means that to elucidate a neural circuit you have to have a detailed anatomical map of all the cells that are participating in this circuit. There are various ways to achieve this: there is a light microscopic level but there's also the electron microscopic level, which will in the future enable us to have a complete 3D model of a pathway or in the somewhat more distant future, of a whole cortical area or of the whole cortex.
6. How do you aim to use the 3D model to understand disease processes of neurological disorders?
Before we start looking at the diseases we have to understand the normal behaviour, which I believe we are far from. To break it down, you have to have the normal structure plus the normal physiology in order to look into diseased models to determine what is different either on the anatomy level, the physiology or both. At the moment, I don’t foresee an immediate application of anatomical techniques. There is currently a large program of 3D reconstruction of cortical cells in epileptic or tumour brains. To create this database, patient that undergo neurosurgery due to an epileptic focus that cannot be controlled or they have a tumour, the surgeons take a bit of the brain tissue - and that is the beginning of the reconstruction of the normal cortex. Nonetheless, in my eyes, this is very far away.
In order to study disease processes of neurological diseases, like epilepsy or Alzheimer’s, many mouse models of neurological diseases were developed. Our work lays out a template for the state of the rodent cortex under normal conditions, and this information can be used to describe changes in the anatomy relevant with these diseases. Reconstruction of animal brains can create a new way of looking at cellular structure of a significant part of the brain, which can open a window to a better understanding of brain function, such as degenerative disease processes. It could also help us understand the mechanisms associated to the brain´s process of learning and how small changes can result in different sensory and cognitive abnormalities.
7. What are your thoughts on implementing advanced technology such as artificial intelligence in the world of neuroscience? How do you think it will help with the better understanding of neuroscience and what do you think will be some drawbacks?
I believe AI (artificial intelligence) is just an automatization of a particular task. The point is that you have to make decisions, and so far, if you look at neural nets that are capable of learning, they simply learn by trial and error and nothing else, until these neural nets can perform one task. So, if you change the condition it won’t work anymore. Then you have to start another learning process. If you change something and then again nothing works. To my understanding, learning to perform a task has nothing to do with intelligence, it is basically an educational process of many different situations which might be useful if you have a limited number of decisions. Nevertheless, if you want to use your intelligence, your insight or your percept to make a decision that is not very strongly defined, it fails. I do hope that with time AI will get better, but we have heard about AI for so many years and every now and then there are big breakthroughs. It is like treating Alzheimer’s, every other year there is a big breakthrough but then nothing happens, but it doesn’t mean it’s not going to work. In my opinion, I can’t see AI coming in the near future except for basic automates where you have a very clear system of things that you have to do.
8. What kept you motivated throughout your studies? What are some suggestions/ tips that you would give to medical students, and what are your words of encouragement for individuals who are interested in research in neuroscience?
First, have a good idea of what you want to do, what you want to find out, and how much time you want to spend on it. Second, look for a lab and judge the size of the lab in terms of of the team (an ideal group is about ten). Also, make sure the lab has good workshops. Nowadays, many students by necessity expect a good salary but you can’t, so you have to think about what you want to do. Perhaps you could start working as a lab assistant to find out whether the atmosphere is suitable for yourself, and very often you may realise that is not the place you want to be in. On the other hand, if you are being awarded a good salary, it may keep you from changing your topic of work.
1. Bert Sakmann - Facts [Internet]. Nobelprize.org. [cited 2017Dec12]. Available from: https://www.nobelprize.org/nobel_prizes/medicine/laureates/1991/sakmann-facts.html
2. Max Planck scientists identify neuron types that mediate different behavioural states [Internet]. Max Planck Society. [cited 2017Dec12]. Available from: https://www.mpg.de/1232069/neural_circuits
3. Mpfneuro. Nobel Laureate Dr. Bert Sakmann awarded 2015 International Ellis Island Medal of Honor [Internet]. EurekAlert! 2015 [cited 2017Dec12]. Available from: https://www.eurekalert.org/pub_releases/2015-05/mpfi-nld051915.php