25 Interviews for the FNP’s 25th Anniversary. The Foundation for Polish Science (FNP) celebrates its 25th anniversary this year. To mark the occasion, we have invited 25 beneficiaries of our programmes to tell us about how they “practise” science. What fascinates them? What is so exciting, compelling and important in their particular field that they have decided to devote a major part of their lives to it? How does one achieve success?
The interviewees are researchers representing many very different fields, at different stages of their scientific careers, with diverse experience. But they have one thing in common: they practise science of the highest world standard, they have impressive achievements to their credit and different kinds of FNP support in their extensive CVs. We are launching the publication of our cycle; successive interviews will appear regularly on the FNP website.
A Simple Desire for Knowledge
Dr Ewelina Knapska, a biologist who studies the cerebral basis of emotions, talks to Patrycja Dołowy.
PATRYCJA DOŁOWY: You study the neurobiological mechanisms of socially transferred emotions. This is very much needed research in today’s world. What exactly do you do?
EWELINA KNAPSKA: Our field is fundamental research. We are interested in how the brain works, and especially how it controls emotions. One such emotion is fear – a very old emotion, which is preserved through evolution and important for survival. It occurs in practically all organisms. The nervous basis of fear seems to be very similar in people, mice and rats. This enables us to work on animal models in which we can manipulate the nervous system or even single cells and study the mechanisms of fear at a very detailed level. First we studied fear and fear extinction in classical models. We taught a rat to fear certain stimuli and then prescribed therapy. It turns out that therapy is effective the next day, but the fear returns after some time. It’s similar with patients. It’s quite a widespread phenomenon: in people who had behavioural therapy, fear returns within 3-4 years. The memory trace of therapy, or the process known as fear extinction, is much weaker than the original trace of fear. It is possible that the neural pathways of extinction become weaker or disintegrate. We know quite a lot about the role of different brain structures in fear conditioning and extinction. Three structures are the most important: the prefrontal cortex, the amygdala, which controls emotions, and the hippocampus, which is a gateway identifying whether a situation is familiar or new. All this works thanks to connections among these structures that control one another. At present we are trying to study these structures and the connections among them in models of socially transferred fear, i.e. in a situation of the contagion of fear. Rats, mice, but also people receive stimuli from other individuals. If one gets scared and another learns about this, it is clearly reflected in the activation of the second one’s brain. We study that activation – we observe how the response to fear and the memory of fear change in social situations.
How do you do this?
In the first model, we conduct observations under conditions of a home cage in which two rats are kept together. If we take one of the rats, scare it and then return it to the home cage, we see that the demonstrator rat, as we call it, communicates fear to the other rat. In other words, the animals exchange information; in particular, a clear signal appears that it is not safe anymore. Contagion of fear can also be observed in people; one excellent example of this is when panic breaks out in a crowd. With rats, contagion of fear is accompanied by a specific behaviour pattern. The demonstrator returns to the cage and runs about, the other rat runs after it, sniffing it where scent messages are secreted, namely the neck and tail regions. The animals also make special sounds. We checked how this situation affected the conditioning of fear, or how they learned from each other. It turns out that rats which receive fear from a fellow rat learn quicker and remember better, whereas fear in the demonstrators weakens. We study this particular phenomenon, called social buffering of fear, in our second model. We observe that the presence of another rat, especially a familiar individual, makes the fear of the first rat decrease. In this case the demonstrator rat is taken away, scared and then placed in an experimental cage. If the rat is alone in this environment, it is clearly scared. If it is placed in a cage with another rat, it is much less afraid. In these tests we measure the fear level with the help of the level of freezing behaviour, a typical rodent response. If you scare two rats separately and then place them in the same cage, the fear level in both of them decreases quite a lot.
Like a form of group therapy…
Fortunately that is quite a remote analogy, because when we test the memory of fear the following day, by measuring the fear level in response to a conditioned stimulus, e.g. a sound, it turns out that this memory is not significantly different in individuals that were scared alone and those that were scared and then ended up in a cage with a partner and stopped being scared more quickly. The stimulus causes a similar fear level in both groups, which means it is not a lasting effect. But what mainly interests us in this situation is how the presence of another rat modifies the activity of the fear learning and extinction pathway, whether it’s the same mechanism or a different one. We look at activity specifically linked to the appearance of the other rat. We check which neurons are active and how they are connected within the structures mentioned earlier: the prefrontal cortex, amygdala and hippocampus.
You observe what is activated in the brain in such situations in both rats?
And we also compare social and non-social models. First a rat undergoes classical fear conditioning (non-social stimulation) and then we try to see what this looks like at the neuronal circuit level. This is possible because we have implemented a technique called optogenetics at our laboratory, thanks to which we can manipulate – activate and deactivate – particular nerve circuits and even single cells. As I haveve mentioned, we are mainly interested in the role of individual connections.
What does the optogenetic technology involve?
We introduce photosensitive proteins into cells that are active – in our case, during social interaction – and these enable us to stimulate or dampen them later by inserting an optical fibre and directing light at those cells. In this way, we try to find out which cells control specific behaviours, in which brain structures they are found, and how they affect one another. We also watch how the animals behave during interaction with another individual or in a new environment when we stimulate or dampen those cells. We have found that stimulating cells in the amygdala central nucleus interrupts social interaction. The rat starts being interested in its surroundings, which is sensible because it feels threatened and is trying to find any dangerous stimuli. Contagion of fear, which we are studying, is considered to be the simplest form of empathy that can be observed even in very primitive animals.
So you analyse simple empathy models?
We need to start from how we define empathy. Some psychologists define it as a purely human thing. Right now, though, we have many studies on animals that suggest it is more widespread. The first research, on monkeys, was conducted in the 1960s, and quite recently we have seen projects involving rodents. It turns out that they, too, display empathetic behaviours. Hypotheses have been put forward that there is evolutionary continuity in these behaviours. I’d like to briefly outline one such study. A laboratory in Chicago studied rats, one of which was kept in a small, see-through box. Rats hate being in such tight, bright spaces. The box had a clever latch that a second rat, outside the box, was able to open. The study showed that one of the first things the free rat did was to set free its imprisoned fellow. The experimenters even offered the free rat pieces of chocolate, which rats really like, but even this did not distract the rat – it saved the other rat first and only then went for the chocolate. Sometimes it would even share the chocolate with the other rat. Rats are more willing to help other rats they know. They help strangers, too, but then the decision to help takes them longer. Experiments were conducted on rats that looked different, e.g. a white rat was meant to free a rat with black spots. It turned out that different-looking rats were freed the least willingly. But the situation changed completely if the rats had been kept together earlier. In such cases, spotted rats were freed just as quickly as other familiar rats. The empathetic behaviours of monkeys are well described in the books of Frans de Waal. Among chimpanzees and bonobos, healthy individuals were observed taking care of apes with Down syndrome (with trisomy). These behaviours are very complex and would have been hard to control under laboratory conditions. Our models are much simpler, which has two main merits. First of all, we can precisely control the influence of individual factors, e.g. one gene, and secondly, the point is to obtain repeatable results. We focus on how neuronal circuits affect empathetic behaviours. Optogenetics enables us to precisely manipulate neuron activity. This is the first step in seeing how social behaviours could be controlled. We can find out which structures in the brain, and even which neurons are responsible for specific responses.
You discovered not so long ago that a specific part of the brain takes part in the process of remembering pleasant sensations in mice.
The central nucleus, which is part of the amygdala, is a very interesting structure that controls many aspects of behaviour. We started by studying mice that didn’t have the gene conditioning the MMP-9 extracellular matrix protein. This enzyme takes part in learning. To our surprise, we found that mice without this gene learn defence responses quite well, e.g. when air is blown into their noses in the IntelliCage system. This is a large cage where mice live in a group. The mice are marked, so every mouse is identifiable when it goes into one of the corners. There are two bottles in these corners, one with plain water, the other with sweetened water. To get at them, a mouse has to poke its nose in a hole, which opens a door enabling the mouse to have a drink. In the other corners, air can be blown at the mouse, which mice don’t like. It turned out that mice without the aforementioned protein learned much more slowly than control mice how to find the sugar water. Meanwhile, they learned just as quickly as other mice where the air was blown at them. In a more difficult option, there was sweet water in one bottle, plain water in another, and bitter water (with quinine) in the third. The mice learned very quickly where the quinine water was but had serious trouble learning where the sugar water was. We thought that perhaps they didn’t like sugar, but their thirst for sugar water once they got it was just as great as in the control mice. This allowed us to conclude that the animals simply had trouble learning where to find the sweet water. We began to suspect it had something to do with the lack of a specific protein. So, we developed a rather subtle tool: nanomolecules, which – after being injected into the brain – release the MMP-9 protein inhibitor, i.e. the substance that inhibits the production of this protein. When we injected the nanomolecules into ordinary mice, the effect was the same as in the mice without the gene – they started having trouble learning pleasant sensations. In this case we were very lucky that a single structure was responsible for this. Usually the circuits controlling positive and negative emotions are very close to each other; you can distinguish them, but the cells are mixed, so the above-mentioned technology doesn’t work. We also have mice with fragile X syndrome at our laboratory. This disorder manifests itself, among other things, in disrupted social interactions. These mice, like mice without the MMP-9 protein, have trouble learning pleasant emotions because their genetic defect affects MMP-9 levels. Specifically manipulating the MMP-9 in the central nucleus could help them. We want to test this further, to check to what extent we can help mice function in different social situations. It’s still a long way to applications involving humans, but if a therapy based on this idea is developed one day, the appropriate nanomolecules could be administered intravenously and used, for example, in treating autism.
So mice can have autism?
Autism is complex. We do not know what underlies most of the cases observed in people. We do know, however, that in a few percent of all those affected autism is a consequence of damage to a single gene. Thus, we are talking about a hereditary mutation that causes characteristic social disorders. This has been the basis for mouse models of autism, where that one gene is knocked out – it’s not there and these mice really do show signs of social disorders. There are also “environmental” models in which you influence mouse brain development, e.g. by administering valproic acid, an active ingredient of a seizure medicine (this is the result of the observation that children with autism are born more often in the population of women who took this medicine during pregnancy). We look for specific disturbances in social behaviours that are observed in autism. Many autistics do not socialize well, function poorly in a group, have problems with social punishments and rewards. On the other hand, they respond well to behavioural (“traditional”) rewards or punishments. By the way, when I was on a fellowship in the United States, there was an ongoing discussion about therapies for autistic children, where electric stimuli from a special bracelet had been used earlier. This therapy had been banned, causing parents to protest because they claimed it was a very good method for limiting children’s dangerous behaviours, such as self-mutilation. At the cerebral level, to some extent separate mechanisms could be responsible for behavioural and social stimuli. They are what we’re looking for.
…using mouse models.
We build cages for automated studies. Contrary to rats, mice are very sensitive to the presence of an experimenter. People can be a very strong stress-inducing factor for them. We are testing cages in which mice can be marked with special chips so that we know where a mouse is and what it’s doing without touching it or looking inside the experimental room. We have a system of cages connected by corridors, and special antennae follow the movements of the mice. The mice have food and drink in the two extreme cages and another two cages have partitions behind which we can place a scent stimulus, for example. The mice live in a group while we check which stimuli interest them. It is all computer-controlled. The development of these techniques was made possible by a collaboration with physicists – a lot of the equipment is designed and built by scientists from the Warsaw University of Technology.
Interdisciplinarity can be useful.
I have two degrees – I graduated in biology and psychology. We work with geneticists – the transgenic models we use come from them – and molecular biologists, who study responses at the cellular level. We also work with physicists because we need new technologies like optogenetics, and with human psychologists – that project is due to be launched in autumn. We have started a collaboration with a researcher from Sweden’s Karolinska Institutet, where neuroimaging scans on people are performed. Of course we collaborate with animal psychologists. This interdisciplinarity is also the effect of experience acquired at different laboratories. Thanks to taking part in the FNP’s Kolumb programme, I myself was able to spend some time at the University of Michigan in Ann Arbor (USA), returning from there with contacts and ideas that I later developed under another FNP programme: Homing. Working like this, above strictly defined divisions into disciplines and institutions, is extremely inspiring.
That’s how science is practised today, isn’t it?
Yes, we have to process a growing amount of information from different levels of biological system complexity corresponding to different fields of learning. Moreover, having equipment and knowing how to use it goes beyond what you can do within a single laboratory. If you want to use more techniques, collaboration is the only option. We are seeing great technological progress; in my field there has been a huge leap over the past 2-3 years thanks to the development of small microscopes that you can place in a brain and see directly what structures are activated. Researchers who use this method enter the hippocampus and see a few hundred, a few dozen active cells, whereas the animal they are studying just has a small, practically imperceptible device on its head. To keep up, you have to keep developing. Without collaboration this could be difficult.
At the beginning you said straight out that you conducted fundamental research. That’s important, because it is the trend now to keep underlining what application everything will have.
When we look at the history of science, we see that most discoveries and inventions resulted from a simple desire for knowledge. Someone was doing something and discovered something else in the process. This is an important lesson. We have to keep reminding people who decide about financing science about it. Fundamental research has potential. I attended a biomedical congress last week. We know very little about mental disorders. Neither depression nor autism have a uniform underlying cause. The only way to expand our knowledge is through fundamental research, and not research aimed at inventing a medicine. There are a lot of studies that don’t seem to lead to any specific applications –they involve bugs, or bacteria… A good example of how important it is to support fundamental research is glowing bacteria. The technology that uses glowing proteins, which revolutionized contemporary biology, originated from a man at a laboratory who studied glowing bacteria. He manipulated proteins, and later it turned out this could be put to use. One woman studied crustaceans. She was accused of wasting taxpayers’ money. Meanwhile, her research turned out to have a serious effect on the construction of aircraft engines: crustaceans have a mechanism enabling them to expel air at high pressure. When we apply for a grant, by the time the application review procedure is over and funding is awarded for research, we are usually in a different place than when we wrote the application. We get the first results, often different than we had expected. Two strategies are possible: either we sweep it under the rug and research only that one small segment we had predicted earlier (which is not very promising), or we change our objectives and follow the results we have. I apply the latter strategy. In our case, the research on buffering of fear started from a completely different study on mice. We observed that when we added a second mouse to one that had stopped being scared, the latter’s fear suddenly returned. We wanted to repeat this model, only in a rat because it was easier. And that is when it turned out that rats stopped being scared. We conducted some very interesting research that turned our earlier plans upside down. Scientists sometimes have qualms about taking such a route, because what will the grant provider think? I think that being flexible pays off quite well. It is also important to attend conferences and to get to know what others are doing. That is the value of networking. In our field, as opposed to those that bring direct financial profits, people usually share their ideas, because we are driven mainly by cognitive motivation: prestige and scientific curiosity. This means we have cooperation and not competition, and we can join forces.
Dr EWELINA KNAPSKA, heads the Laboratory of Cellular Neurobiology at the Institute of Experimental Biology of the Polish Academy of Sciences in Warsaw. A beneficiary of FNP programmes: KOLUMB (2006) and HOMING (2008).