Dr Mateusz Sikora from the Małopolska Centre of Biotechnology of the Jagiellonian University (JU MCB) has developed a novel approach to predict how sugars form shields around proteins. This new computational approach, GlycoSHIELD, was created as a result of Polish-German collaboration within the framework of the Max Planck Dioscuri programme, in a Dioscuri Centre established in 2023 at the MCB as well as collaboration with the National Institute of Health and Medical Research (Inserm) in Paris and the Academia Sinica (Taiwan).
Sugar chains (glycans) that cover the surface of proteins in our body influence their interactions with other molecules. Hence, the knowledge of glycans plays an important role in in several areas, including the creation new drugs. The high mobility and variability of sugars makes experimental research in this field a very complex process. Simulations allowing scientists to show how glycans cover the surface of proteins, performed by supercomputers, require hundreds of thousands of hours to complete, which means that they cannot be used on a routine basis such as in drug discovery.
These challenges can be answered by GlycoSHIELD – a very efficient algorithm which takes only several minutes to predict how sugars cover the protein surface. This work was led by Dr Sikora, leader of the Dioscuri Centre for Modelling of Posttranslational Modifications, established in 2023 at the Małopolska Centre of Biotechnology of the Jagiellonian University (JU MCB). The Dioscuri programme, initiated by the Max Planck Society is intended to help strengthen and expand research excellence in Central and Eastern Europe.
Understanding proteins
To understand how proteins work and impact the development and progress of diseases, researchers study their three-dimensional structures both through laboratory experimentation and AI prediction using tools such as Alphafold based on the amino acid sequence of the protein.
However, more than 75% of proteins consist not only of amino acids but the surface of these proteins is ‘adorned’ with sugar chains that form very dynamic ‘shields’ around them. It is still not fully known how these shields behave and what their impact on drug molecule binding is. Simulating the morphology of these sugars is difficult because of their mobility, and traditionally scientists would have removed these sugars to assist structural studies.
Taking up this scientific challenge, Dr Mateusz Sikora, head of the Dioscuri Centre for Modelling of Posttranslational Modifications at the JU MCB, his team from Kraków, and their collaborators: Dr Cyril Hanus from the Inserm in Paris and Dr Danny Hsu from the Academia Sinica in Taiwan as well as Gerhard Hummer, partner of the Dioscuri centre from the Max Planck Institute of Biophysics in Frankfurt am Main have created the GlycoSHIELD algorythm. This new software allows very quick and accurate prediction of sugar chains on protein surfaces in a process that takes minutes and does not require a supercomputer. The new method was described in the prestigious journal Cell.
Minutes instead of thousands of hours
For example, the sugar layer on the coronavirus protein spike hides it from the immune system, making it difficult for antibodies to attach themselves in order to get rid of the pathogen. Until now, predicting the structure of sugar layers by means of simulations required expert knowledge and millions of hours of calculations by supercomputers.
On the other hand, the GlycoSHIELD algorithm overcomes these limitations, providing not only a faster, but also a more environmentally friendly alternative to the previously used methods. ‘This approach reduces the use of resources, calculation time, energy and the required technical expertise. Now everybody can visualise the distribution and dynamics of behaviour of sugar molecules on proteins on their personal computers or using our app. There is no need for specialist knowledge or highly efficient computers’, stresses Dr Mateusz Sikora. The developed software can prove useful in working out new vaccines, drug discovery and supporting advanced therapies, such as cancer immunotherapy.
Sugar puzzle
How was it possible to make the simulations so much more effective? The authors created and analysed a library of thousands of forms of 3D sugar chains most often found on proteins in humans and microorganisms. Using full-screen simulations and experiments, the scientists have discovered that to reliably predict glycan shields it is enough to ensure that the attached sugars no not collide with membranes or protein parts. The algorithm is based on these conclusions. ‘GlycoSHIELD users must only specify the protein and place where the sugars are attached. Then our software arranges them on the surface in the most probable configuration’, explains Dr Sikora. We could very effectively re-create the spike protein’s glycan shields. They look exactly the same as what we have seen in our experiments’, comments the JU researcher.
GlycoSHIELD already allows the complementation of old and existing glycoprotein structures with information about glycans. The scientists have also used GlycoSHIELD to reveal the pattern of glycans on GABA neurotransmitter receptor, important for anaesthesiology and sedatives’ activity.
Success of the Dioscuri Centre
Dioscuri Centres initiative of the Max Planck Society is aimed to create and support scientific excellence centres in Central and Eastern Europe. Since May 2023 Mateusz Sikora, the former postdoc at the Max Planck Institute of Biophysics, has received financial support as part of this initiative as the head of the Dioscuri Centre for Modelling of Posttranslational Modifications at the Jagiellonian University. His research is carried out together with German partners from the team led by Prof. Gerhard Hummer, Head of the Department of Theoretical Biophysics and Director of the Max Planck Institute of Biophysics in Frankfurt am Main. Dr Sikora’s studies would not be possible without collaboration with the Academia Sinica from Taiwan and the Inserm Institute from France.
The creation and publication of the environmentally friendly GlycoSHIELD algorithm after less than a year after the foundation of the centre is certainly a significant achievement, showing Poland as an attractive place to conduct world-class research.
‘The publication of research results in the prestigious journal Cell is an important event in the academic life of the Jagiellonian University, proving that the Jagiellonian University meets highest international standards in the field of biological sciences. Such a remarkable success would not be possible without a long-term research strategy adopted by the Jagiellonian University, which, to ensure high quality of the conducted research, promotes the most ambitious projects based in scientific excellence centres. By this term I mean teams that run projects funded by prestigious grants from such institutions as the European Research Council or Max Planck Society, whose results are widely recognised and published in prominent research journals. I would like to congratulate Dr Mateusz Sikora and his collaborators on their success achieved as part of the Dioscuri Centre established less than a year ago, specialising in bioinformatic research’, says Prof. Piotr Kuśtrowski, JU Vice-Rector for Research.
Photo: Katarzyna Wrona
Further information
Availability of GlycoSHIELD as web application hosted by the Max Planck Computing and Data Facility:
Websites of the Disocuri Centre for Modelling of Posttranslational Modifications:
https://www.biophys.mpg.de/dioscuri-centre-for-modelling-of-posttranslational-modifications
https://mcb.uj.edu.pl/sikora-lab
Original citation:
Tsai et al., 2024, Cell 187, 1–16
https://doi.org/10.1016/j.cell.2024.01.034
(Text based on a press release of the Max Planck of Biophysics by Katharina Kaefer)
