Each of our cells contains about 40 million proteins which together perform all the tasks the cell needs to survive. For smooth action, the right proteins must be concentrated in precise amounts, at a precise time and in a precise place. However, establishing such a delicate distribution requires an extremely precise process, occurring at tiny spatial resolutions that standard cell biology tools are often unable to detect. To understand how this mechanism works, researchers from the University of Geneva (UNIGE) have developed a new approach combining experiments in genetics and cell biology with physical modelling. Thanks to specific algorithms, they simulated the formation of protein gradients in 3D and over time and were able to explain these complex mechanisms. Moreover, their innovative model can be adapted to other biological systems to study protein dynamics. These results can be read in the Proceedings of the National Academy of Sciences.
Like a drop of ink in a glass of water, proteins can diffuse and distribute themselves evenly throughout the cell. However, for a number of tasks, proteins must form gradients. “Protein gradients, which result from the uneven distribution of proteins in specific cellular areas, are central to many cellular and body functions,” says Monica Gotta, a professor in the Department of Cellular Physiology and Metabolism and the Center for Translational Research in Onco-Hematology (CRTOH) of the UNIGE Faculty of Medicine, which directed this work. “For example, protein gradients are important for cell differentiation, the process by which the different cell types that make up a complex organism emerge from a single cell, the fertilized egg.”
A use of chance
The PLK-1 protein, a key regulator of cell division, is known to be most concentrated on the anterior side of the embryo. But how to set up this mechanism, and what would be the consequence if the smallest detail went wrong? The usual tools of biology not being sufficient to answer this question, Monica Gotta was happy to welcome in her team a physicist, Sofia Barbieri, post-doctoral researcher in the Department of Cellular Physiology and Metabolism of the Faculty of Medicine of the UNIGE. “By compiling everything that is known about this biological process and new hypotheses on the mechanisms, I developed a statistical model of protein gradient formation based on probabilistic mathematics”, explains Sofia Barbieri. “I used specific calculation algorithms, called Monte Carlo simulations, named after the famous city in the game.” These algorithms are used to model phenomena with a high level of complexity, such as finance, trading or particle physics.
The team was able to simulate protein gradients, not only in 3D, but also over time. Such a model however required several iterations between the optimization of the parameters and the comparison with the biological data. The researchers built a first version of the model integrating all the known physical and biological elements of the system, then introduced specific parameters necessary to test several hypotheses concerning the unknown variables. They simulated possible physical and biological outcomes that reproduced computationally protein dynamics and gradient establishment in the cell, and tested them in real life with in vivo experiments using the embryos of a small worm, the C.elegans nematode.
Complex protein interactions at play
Thanks to the continuous interaction between modeling and cell biology, the UNIGE researchers were able to determine how the PLK-1 gradient was established and maintained over time. Indeed, PLK-1 must bind and unbind dynamically from MEX-5, another protein crucial for development in the C.elegans embryo, to counteract its natural tendency to diffuse homogeneously in the cell. MEX-5 indeed has the ability to change its diffusivity depending on its position in the cell and to interact with other proteins, which is essential to enrich PLK-1 where it is needed. “But surprisingly enough, MEX-5 is not as effective at its task, because a large amount of PLK-1 is not bound to MEX-5!” points out Sofia Barbieri.
This study provides a unique quantitative model for understanding dynamic protein interactions and can be adapted to other cells or proteins whose complex mechanisms cannot be tested with standard cell biology experiments. “Our work shows that interdisciplinary collaborations are increasingly important to advance research! concludes Monica Gotta.