Gaining virtual insights into the molecular structure of the brain

By Sandra Clark

January 26, 2018


Using an advanced computing network to visualize interactions between two key proteins has brought researchers in the Department of Molecular and Cellular Biology one step closer to understanding the molecular structure of the white matter of the brain – findings that could also lead to important insights regarding the onset of multiple sclerosis (MS).

 A team led by Prof. George Harauz used the Shared Hierarchical Academic Research Computing Network (SHARCNET) facility at the University of Guelph to better understand how myelin – that is, the protective coating of the central nervous system that breaks down in MS – develops. The advanced molecular modelling techniques enabled by SHARCNET allowed the team to visualize the interaction between two proteins: myelin binding protein (MBP) and Fyn-SH3, each essential to the health of myelin. 

Computer visualization of proteins involved in myelin formation (image courtesy of G. Harauz)
Computer visualization of proteins involved in myelin formation
(image courtesy of G. Harauz)

“Our work has been driven by the desire to understand better at the molecular level how myelin is assembled in the healthy brain and how it degenerates in the MS brain,” says Harauz. 

MS ultimately involves an autoimmune attack where myelin is depleted and can’t fully repair itself. Because myelin upholds the body’s communication system by enabling nerves to transmit properly, its breakdown leads to MS symptoms such as pain, numbness, and loss of coordination. Harauz’s team believes it is crucial to first understand how myelin develops, in order to then begin to understand why it can’t regenerate itself fully in MS.

Former University of Guelph research associates Dr. Kyrylo Bessonov and Dr. Kenrick Vassal led the modeling efforts, spending almost four years delving into the complex molecular world of myelin formation.   

The team had previously carried out cell-based research that showed the central region of MBP is the target of phosphorylation – a process that modifies the protein’s structure and interactions with other cellular components. Knowing that phosphorylation would change MBP’s interaction with Fyn-SH3, they investigated whether improper signalling between these proteins is the reason why remyelination ultimately fails in the MS brain.

The virtual modelling program available on SHARCNET allowed them to develop a protein model to visualize, in 3D, how phosphorylation can change the interactions between MBP and Fyn-SH3.

The importance of using molecular modelling becomes clear: while experimental data that comes from cellular and biochemical research is essential to understanding these interactions, it can be costly and time-consuming, and in general is only indirectly able to contribute to visualizing these interactions. Publicly available computing infrastructure such as SHARCNET allows researchers all over Canada, including the Harauz lab, to run longer simulations while focusing on specific questions.

“This is one step towards modelling more complex systems that reflect more closely the true composition of brain myelin, which would be really exciting,” says Harauz.

Harauz and his team hope the study can guide efforts to model these interactions further in an increasingly complex environment more closely matched to that of the human brain, and to better understand the implications for diseases such as MS.


The study was funded by the Natural Sciences and Engineering Research Council of Canada.


Read the full article in the journal Proteins.

Read about other CBS Research Highlights.