Scientists reveal unprecedented “live” insight into the complexity of the brain

Researchers have developed a new imaging and virtual reconstruction technology called LIONESS, which provides high-resolution images of living brain tissue, displaying it in real-time 3D nanoscale detail . LIONESS integrates advanced optics, artificial intelligence and a collaborative interdisciplinary approach, overcoming the limitations of previous imaging methods and paving the way for a better understanding of dynamics and complexity. impurities of brain tissue.

Collaborative efforts at ISTA provide unprecedented “live” insight into the complexity of the brain.

The human brain, with its intricate network of some 86 billion neurons, is said to be one of the most complex specimens scientists have ever encountered. It holds a huge but currently unmeasurable amount of information, positioning it as the pinnacle of computing devices.

Grasping this level of complexity is challenging, making it necessary to use advanced technologies that can decipher the minute, complex interactions that occur in the brain at the microscopic level. Imaging thus emerged as a key tool in the field of neuroscience.

The new imaging and virtual reconstruction technology developed by Johann Danzl’s team at ISTA is a major leap forward in brain imaging and is aptly named LIONESS – Optimized Nanoendoscopy Direct information allows for saturated segmentation. LIONESS is a system for imaging, reconstructing and analyzing living brain tissue with a spatial and comprehensive resolution that has not been possible until now.

LIONESS describes the complexity of dense brain tissue. a: Complex neuronal environments b: LIONESS can image and reconstruct samples in a way that illuminates the many dynamic structures and functions in living brain tissue. Credit: Johann Danzl

“With LIONESS, for the first time, it is possible to fully and densely regenerate living brain tissue. By imaging tissue repeatedly, LIONESS allows us to observe and measure dynamic cell biology in the brain,” said first author Philipp Velicky. “The output is a reconstructed image of the cell arrangement in three dimensions, with time making up the fourth dimension, as the sample can be imaged in minutes, hours or days,” he added.

Collaboration and AI are key

LIONESS’s power lies in its exquisite optics and in the two levels of deep learning – an Artificial Intelligence method – that make up its core: the first level enhances image quality, and the second level identify different cellular structures in dense neuronal environments.

The pipeline is the result of a collaboration between the Danzl team, the Bickel team, the Jonas team, the Novarino team and ISTA’s Scientific Services Unit as well as other international collaborators. “Our approach is to bring together a dynamic team of scientists with uniquely combined expertise across disciplinary boundaries, who work together to narrow Technology gap in brain tissue analysis”.

A pipeline to regenerate living brain tissue. Microscopy acquisition with optimized laser focus – Image processing (DL) – Segmentation (DL) – 3D image analysis. Credit: Johann Danzl

Obstacle

It was previously possible to regenerate brain tissue using an Electron Microscope. This method captures images of a sample based on its interaction with electrons. Although capable of capturing images at a resolution of a few nanometers – millionths of a millimeter, Electron Microscopy requires the sample to be fixed in a biological state, which needs to be physically sliced ​​to obtain 3D information. Therefore, dynamic information cannot be obtained.

Another previously known technique is Light Microscopy which allows the observation of living systems and the recording of intact tissue volumes by slicing them “optically” instead of physically. However, the Light Microscope is seriously hampered in its resolving power by the very properties of the light waves it uses to produce images. Its best resolution is a few hundred nanometers, too coarse to capture important cellular details in brain tissue.

Using Super Resolution Light Microscopy, scientists can break this resolution barrier. Recent work in this field, called SUSHI (Super Resolution Shadow Imaging), shows that applying dye molecules into the space around cells and applying STED super-resolution microscopy (Nobel Prize-winning Stimulated Emission Depletion) will reveal ‘super-resolution’ shadows of all cellular structures and thus visualize them in tissue.

LIONESS can image and reconstruct samples in a way that reveals the many dynamic structures and functions in living brain tissue. Image supplier: Julia Lyudchik ISTA

However, it is not possible to image an entire volume of brain tissue with enhanced resolution consistent with the complex 3D architecture of brain tissue. This is because increasing resolution also requires high amounts of imaging light on the sample, which can damage or ‘fry’ delicate living tissues.

According to the authors, this is the power of LIONESS, which has been developed for “fast and light” imaging conditions, thus keeping the sample alive. This technique does so while providing super-isotropic resolution—meaning it is equally good in all three spatial dimensions—allowing the cellular components of the tissue to be visualized in 3D nano size details solved

LIONESS collects only the necessary amount of information from the sample during the imaging step. This is followed by a first deep learning step to fill in additional information about the structure of the brain tissue in a process called Image Restoration. In this innovative way, it achieves a resolution of about 130 nanometers, while being gentle enough to image living brain tissue in real time. Together, these steps enable the second step of deep learning, this time to understand extremely complex image data and identify neuronal structures automatically.

ISTA scientist Johann Danzl in his laboratory at the Austrian Institute of Science and Technology. Credits: Nadine Poncioni | ISTA

Return

“Our interdisciplinary approach allows us to break through the intertwined limitations of addressing energy and light exposure to living systems, understanding complex 3D data, and incorporating structural Tissue cytoarchitecture with molecular and functional measurements”.

For the virtual reconstruction, Danzl and Velicky collaborated with visual computing experts: the Bickel group at ISTA and the group led by Hanspeter Pfister at Harvard University, who contributed their expertise on automated segmentation—the process of automatically identifying cellular structures in tissue— and visualization, with additional support from ISTA image analysis staff scientist Christoph Sommer. Neuroscientists and chemists from Edinburgh, Berlin and ISTA contributed to the complex labeling strategies.

It is thus possible to connect functional measurements, i.e. readings of cellular structures along with biological signaling activity in the same living neural circuit. This was done by imaging the flow of Calcium ions into the cell and measuring the cell’s electrical activity in collaboration with the Jonas group at ISTA. Novarino’s team contributed human brain organoids, commonly known as mini-brains that mimic human brain development. The authors emphasize that all of this is facilitated by the expert support of ISTA’s leading scientific services units.

The brain’s structure and activity are dynamic; Its structure evolves as the brain performs and learns new tasks. This aspect of the brain is often called “plasticity.” Observing changes in the brain’s tissue structure is therefore essential to decipher the secrets behind its plasticity. New tool developed at ISTA shows potential for understanding the functional structure of brain tissue and potentially other organs by revealing subcellular structures and capturing how they may change over time .

Reference: “Dense 4D Nanoscale Living Brain Tissue Regeneration” by Philipp Velicky, Eder Miguel, Julia M. Michalska, Julia Lyudchik, Donglai Wei, Zudi Lin, Jake F. Watson, Jakob Troidl, Johanna Beyer, Yoav Ben-Simon, Christoph Sommer, Wiebke Jahr, Alban Cenameri, Johannes Broichhagen, Seth GN Grant, Peter Jonas, Gaia Novarino, Hanspeter Pfister, Bernd Bickel and Johann G. Danzl, July 10, 2023, Natural method.
DOI: 10.1038/s41592-023-01936-6

The research was funded by the Austrian Science Foundation, Gesellschaft für Forschungsförderung NÖ (NFB), H2020 Marie Skłodowska-Curie Actions, H2020 European Research Council, Human Frontiers Science Programme, Simons Foundation, Wellcome Trust and the Foundation national science.