Thomas Burg loves extremes. The objects he studies are extremely small, extremely cold and extremely difficult to resolve. Burg has been a professor at the department of Electrical Engineering and Information Technology at TU Darmstadt since September 2018 and heads the “Integrated Micro-Nano Systems“ research group. His aim is to investigate new technologies for studying cellular processes with high resolution in space and time, as life is not static, but dynamic. To this end, however, light and electron microscopy must be integrated better than they are today. Burg’s research is worth around two million euros to the European Research Council (ERC) over the coming five years.
Light microscopy can reveal the dynamics of living cells and organisms, and it allows specific molecules to be marked and highlighted with fluorescent dyes. However, the resolution of standard fluorescent light microscopy is limited to some 200 nanometres due to the wave characteristics of light. Although a trick has recently been invented to overcome this limitation, detailed structural analysis in the range 0.1 to 10 nanometres is still only possible with electron microscopy. Cells and organisms have to be fixed in place to this end. This can only be done without damage using cryofixation. Using this method, samples are cooled down rapidly to temperatures below -135o Celsius. This is the only way water can retain its glassy, disordered structure. In this way, cellular structures remain largely unchanged due to the lack of ice formation.
Transfer step causes delay
“But to observe dynamic processes remains a challenge”, says Burg, who led a research group at the Max Planck Institute for Biophysical Chemistry in Göttingen for ten years prior to his move to TU Darmstadt. “The samples observed in the light microscope have to be moved to a different instrument for cryofixation. There is therefore a transfer step. The manual transfer sometimes takes minutes, depending on the operator. Automatic transfer takes one to five seconds. There always remains therefore a gap between the last observation of the dynamic process in the living object under the light microscope and the flashfrozen state in the cryofixed object. Part of the chain of events is therefore always missing“.
Burg’s aim is thus to flash-freeze cells and organisms directly under the light microscope so that they can then be examined in the frozen state using light and electron microscopy. To this end, the heat has to be removed rapidly and the sample cooled down and kept at below -135o Celsius continuously to ensure that the water molecules do not slowly rearrange into ice crystals, destroying the delicate structure.
Burg has crossed an important hurdle on this path. For freezing the objects under the light microscope, he uses components from microsystems technology. These include an electrically heated microchannel, in which the cells or organisms can initially be observed at physiological temperatures. A heating element located under the microchannel has contact to a silicon chip that is cooled with liquid nitrogen. When the heating element is switched off, the stored heat dissipates rapidly via this silicon chip. At the same time, the sample freezes rapidly. This method gives the observer the opportunity of studying the cellular processes up to the point at which the heating element is switched off, causing the temperature difference between the object under examination and the liquid nitrogen bath to collapse and the sample to freeze within milliseconds. The frozen object can then be examined under the electron microscope too following appropriate preparation.
“By eliminating the transfer step, we improve the link between light and electron microscopy“, Burg says. “But we want more. We also want to examine the cryofixed samples directly using light microscopy“. However, this so-called cryo-light microscopy is still in its infancy. One challenge is the inevitable temperature difference between the microscope at room temperature and the sample cooled down to -140o Celsius. For the best-possible resolution, the air gap between the objective lens of the microscope and the sample has to be filled with an immersion liquid. This must be done without damaging the delicate lenses by the low temperature and without transferring too much heat to the frozen sample via the objective. Moreover, Burg and his colleagues need an immersion liquid with the same refractive index at this low temperature as water at room temperature.
Freezing and thawing under the microscope
The physicist and his colleagues have solved the first problem by cooling the very small front lens of a high-quality immersion objective and shielding it from the microscope via a carefully heated ceramic mount. In this way, the cold front lens and the sample are thermally camouflaged to not affect the larger and more delicate lenses of the microscope. Burg and his team have also discovered a suitable immersion liquid. It is called ethoxynonafluorbutane. “It wasn’t easy to find something suitable“, Burg says. “There is very little data on the refractive index of liquids at these low temperatures. There is some data for temperatures down to -80o Celsius, but not lower. We had to try a lot of things“. Burg currently achieves a resolution of 350 nanometres with his cryo-light microscopy set-up. As their next aim, he and his team want to improve the method further, make it suitable for routine application and use it in conjunction with so-called super-resolution methods that make a 10-fold higher resolution possible with the same lens.
Burg wants to use the financial support of the European Research Council to also test whether the rapidly frozen cells can be thawed again without damage. “It would be ideal if the cells had no recollection of the freezing after they’re thawed, so that they resume their activity at the point at which they were stopped by the freezing“. Burg knows that this is a long way off but he is aware of the problems needing to be solved. Extremely high cooling rates are required, for example and the cells and organisms must be frozen with very little cryoprotectants or none at all. But improved cryoprotectants would be desirable too. Perhaps Burg may ultimately succeed in freezing cells under the microscope and thawing them without any damage at all. Dynamic processes could then be observed, stopped and restarted on demand. There is still a long way to go however till that point is reached.