Faster Drug Discovery with GlycoSHIELD

An international research team, which includes members of the University of Bremen, has developed a novel, environmentally friendly algorithm that can predict the morphology of sugar coats on clinically relevant proteins within minutes.

The proteins on the surface of the cells of our body are coated with sugars that influence how cells interact with their environment and with pathogens. Therefore, they play an important role in medical drug development. The new, environmentally friendly, fast and open-source software toolkit GlycoSHIELD now enables quick and efficient analyses of protein sugar shields. The new approach was developed as part of the Max Planck Society’s German-Polish Dioscuri program in cooperation with the University of Bremen, the Inserm research institute in France, and the Academia Sinica in Taiwan.

In order to understand how proteins function and how they influence the development and progression of diseases, researchers need to know their three-dimensional atomic structure. In addition to experimental methods, they also use computer simulations. The artificial intelligence Alphafold, for example, reliably predicts the spatial structure of a protein from the sequence of individual protein building blocks, the amino acids. However, over 75% of the proteins present on the surface of our cells do not consist solely of the amino acid building blocks, but are covered by sugars, known as glycans. These form very dynamic “shields” around the proteins. At present, it is not well understood how exactly these shields behave, or how they influence binding of drug molecules. Due to their mobility and variability, predicting the morphology of sugars has been difficult and resource-intensive, until now. Dr. Isabell Louise Grothaus of the Hybrid Materials Interfaces Group, under the direction of Professor Lucio Colombi Ciacchi at the University of Bremen, has been working on computer simulations of glycans for the past four years and has developed a software called GlyCONFORMER for the analysis and classification of their mobility and variability. Her expertise and software have been pivotal to the development of the new GlycoSHIELD toolkit, which enables a fast and realistic modeling of the sugar chains present on protein surfaces, as has just been reported in the renowned Cell journal.

From Thousands of Hours to a Few Minutes

Sugars strongly influence how proteins interact with other molecules. The sugar shield on the spike protein of the SARS-Cov-2 coronavirus, for example, makes it difficult for antibodies to recognize the virus, thus, hiding it from our immune system. The sugar shields therefore play an important role in drug development – pharmaceutical research could benefit from routinely predicting their morphology and dynamics. Until now, however, predicting the structure of sugar layers with the help of computer simulations has only been possible with expert knowledge and special supercomputers. In many cases, this required thousands or even millions of computing hours. With GlycoSHIELD, the team of researchers led by Dr. Mateusz Sikora, head of the Dioscuri Center for Modeling Post-Translational Modifications at the Jagiellonian University in Krakow, has created an environmentally friendly open source alternative. “Our approach reduces resources, computing time, and the necessary technical expertise,” says Sikora. “Anyone can now calculate the arrangement and dynamics of sugar molecules on proteins on their personal computer within minutes.” The software could be helpful for the development of drugs or vaccines, for example in immunotherapy against cancer cells, which present a very distinctive profile of glycans on their surface.

A Jigsaw Puzzle Made of Sugars

How did the team manage to achieve such a high increase in efficiency? The researchers created and analyzed a library of thousands of the most likely 3D poses of the most common forms of sugar chains on proteins found in humans and microorganisms. Using long simulations and experiments, they found that it is possible to reliably predict how certain glycans behave. First of all, it is not necessary to include the interaction between the shells of cells or parts of proteins with the attached sugar. The algorithm is based on these findings. “GlycoSHIELD users only need to specify the protein and the locations where the sugars are attached. Our software then puzzles them onto the surface in the most likely arrangement,” explains Sikora. “We were able to reproduce the sugar shields of the spike protein very well; they look exactly as we see them in our experiments!” With GlycoSHIELD, it is now possible to supplement both new and existing protein structures with sugar information. The scientists also used GlycoSHIELD to reveal the pattern of the sugars on the GABA_Areceptor, an important target for sedatives and anesthetics.

Further Information:

Y.-X. Tsai, N.-E. Chang, K. Reuter, H.-T. Chang, T.-J. Yang, S. von Bülow, N. Zerrouki, M. Gecht, C. Penet, I.L. Grothaus, L. Colombi Ciacchi, K.-H. Khoo, G. Hummer, S.-T.D. Hsu; C. Hanus, M. Sikora.
Rapid simulation of glycoprotein structures by grafting and steric exclusion of glycan conformer libraries.
Cell, DOI: 10.1016/j.cell.2024.01.034 (2024). https://doi.org/10.1016/j.cell.2024.01.034

About the Bremen research group: https://www.hmi.uni-bremen.de/index.html

About the coordinating group: https://www.biophys.mpg.de/dioscuri-centre-for-modelling-of-posttranslational-modifications

News from the MPI for Biophysics: https://www.biophys.mpg.de/de/news