[3IA Young Women Researchers] AI with Vasiliki Stergiopoulou

Published on March 8, 2022 Updated on March 9, 2022

What is your research topic?

The objective of my PhD thesis is to develop new super-resolution fluorescent microscopy techniques, able to output images with fine temporal and spatial resolution at the cell membrane level. The design and development of fine imaging techniques capable of accurately processing images of samples, lying on a small layer above the microscope’s coverslip, will provide precise 3D localization of important sub-cellular biological entities and will allow for the investigation of their functions.

See the poster


Could you briefly explain it?

For all optical microscopes (including fluorescence microscopes) the highest achievable spatial resolution is governed by some fundamental physical laws related to the propagation of light and there is a resolution limit (called diffraction limit) that cannot be overcome by physical means (e.g by changing the objective lens or aperture designs). In standard acquisition settings, the diffraction limit, i.e. the size of the biological entities we are able to distinguish, is approximately equal to 200nm in the lateral plane and to 500nm in the axial one. But several biological entities of interest have size smaller than these values. The implementation of appropriate super-resolution techniques can thus significantly improve the visualisations of invisible sub-cellular entities, by allowing the understanding of the biological functions happening at the molecular level.

Recently, a MA-TIRF (Multi Angle - Total Internal Reflection Fluorescence) microscope has been prototyped (one of 3 in France) in collaboration with the iBV Lab. Along with the microscope there is a dedicated reconstruction algorithm that achieves unprecedented axial resolution much further than the diffraction limit and still unmatched by commercial systems for imaging of living cells. However, the lateral resolution of this system remains limited by diffraction. Therefore, my contribution will be to develop an acquisition protocol and a reconstruction algorithm to obtain super-resolved 3D images, by combining the pre-existing MA-TIRF reconstruction with advanced super-resolution techniques in the lateral plane. 


Can you illustrate with an example?  

Studies on the maturation and trafficking of membrane proteins (e.g. ion channel subunits, receptors, etc.), will allow scientists to analyse the exchanges at the cell membrane level. However, proteins are only a few nanometres in size and indistinguishable when imaged through a conventional fluorescent microscope. In order to retrieve the precise locations of the membrane proteins so as to better study them, super-resolution techniques could be applied. It is important although these techniques to allow for live-cell imaging and to not damage the sample being imaged. For example, they should not require high-power lasers or sample fixation. In this context, our task is to find a super-resolution technique with the above requirements. 


Can you tell us about an important result?

We have proposed a 3D super-resolution method, called 3D-COL0RME. It is applied on images obtained by a MA-TIRF microscope. The MA-TIRF microscope benefits from low photobleaching and low photodamage and has a really short acquisition time... in a few words is ideal for live-cell imaging. In more details, 3D-COL0RME combines the pre-established MA-TIRF reconstruction, that achieves very good levels of axial resolution, with a later super-resolution technique that exploits the independent statistical behaviour of standard fluorescent emitters. 

In the following figure you can see the improvement of the lateral resolution between the pre-existing standard MA-TIRF reconstruction (bottom-right part) and our method, the 3D-COL0RME (upper-left part). In this figure, not membrane proteins but rather microtubules of bovine aortic endothelial (BAE) cells acquired by a MA-TIRF microscope are being imaged. The colour in the following figure quantifies the sample’s depth, for example the filaments displayed with a blue colour are closer to the microscope’s coverslip while the red ones are deeper.

What are the challenges related to this topic?

Choosing the right super-resolution technique is a complex task and someone must consider different criteria (e.g. sample type, structure size, imaging requirements, etc). Especially, in our case, that we would like to perform 3D super-resolution imaging in live-samples using a not very complex prototype, to find an acquisition protocol and a reconstruction algorithm is very challenging. 


What are the real-world impacts, issues?

For many centuries, cell biology has relied on light microscopy, while being limited by its optical resolution. However, several new technologies and algorithms, included ours, have been developed recently to circumvent this limitation, allowing the understanding of cellular functions that require detailed knowledge of all factors (e.g. interactions, modifications, cell distribution, etc). Overall, all these new approaches have created unprecedented new opportunities to study cell structure and function.