Identification of activated neurons in the whole mouse brain

It is the aim of this study to visualize those neuronal networks in the brain which play a role during the storage of fear memories. This process plays a pivotal role during the development and maintenance of e.g. post traumatic stress disorder. It is still unclear which brain structures are involved herein. Approaches in humans using imaging modalities like fMRI do not achieve cellular resolution. Animal experiments which allow such resolution concentrate however on small parts of the brain in a dish. Up to now there are no methodological approaches which allow the visualization of neuronal activity with cellular resolution throughout the whole brain. Such resolution can be obtained if strongly activated neurons produce a fluorescent protein that can be visualized. We want to use transgenic mice which allow this to identify the structures active during fear memory storage in the animal model. To achieve this the brains of mice with fear memories will be made transparent after the experiment with a chemical procedure we developed. Thereafter we will record the brains with our new 3D-ultramicroscope. We hope to get this way something like a single shot nuclear magnetic resonance recording of fear memory storage with cellular resolution. The realization of this technology would open up completely new avenues for the investigation of changes in the brain during emotion and behaviour.

If one wants to get a three dimensional impression of the anatomy and the distribution of active neurons in the mouse brain there is since some years a new technology available: Chemical clearing of the mouse brains which means making the brains transparent by putting them in a certain series of chemicals. By these chemicals first the water is extracted, afterwards they are put in a mixture of oily substances. Everybody knows this effect when putting a drop of salad oil on a sheet of paper. At the point of the oil drop the paper gets transparent. Of course the neurons also have to be labelled with a fluorescent substance to be seen. This is possible by using special mice who express the fluorescent Protein GFP in a subset of neurons. Other mice can also be created which contain a red fluorescent marker which only appears if the neuron has been strongly active. This way one can visualize the parts of the brain and the neurons which are active during a certain kind of behaviour or after certain strong stimuli. To be able to map exactly the 3D distribution of the activated neurons one needs an imaging method which is able to record with micrometer resolution the complete volume of a mouse brains which is in the cubic centimetre regime. This is possible with a special kind of microscopy, ultramicroscopy, which we once developed. Ultramicroscopy means the application of light sheet microscopy to cleared samples. The transparent and fluorescent sample is moved slowly through a thin sheet of light creating thousands of optical sections. These optical sections are sequentially recorded with a camera and finally a 3D image is reconstructed from these recordings. A problem with this approach is that the chemical substances used to make the brain transparent can destroy the fluorescent proteins one wants to image. Thus an important part of the project was the improvement of the existing clearing technology so that the fluorescent signal expressed in the activated neurons was well preserved. Another problem was that to be able to record a whole mouse brain one needs microscope objectives with good resolution and a wide field of view. This was not possible with the available low power air objectives so we had to modify and correct them to be suitable for immersion. In ultramicroscopy the axial resolution depends on the thickness of the light sheet used. Therefore we took considerable efforts to create even, thin light sheets with specialized optics. With all these technical developments we were finally able to record subpopulations of neurons in the certain parts of the mouse brain that are active during stress reactions of the mouse.