This project's purpose was mainly to explore SARS-CoV-2's biology and find something that would spark my interest and it would be worth to visualize. After a lot of back and forth and a lot of unused ideas and sketches, I settled up for a concept which was quite interesting to me and didn't seem to get a lot of attention; how the virus uses glycans to defend from antibodies and where its Achille's heel is.
Client: Evelyn Maizels (UIC)
Audience: People in the healthcare sciences
Software: VMD, UCSF Chimera, ZBrush, 3ds Max, tyFlow, VRay, Photoshop, Illustrator
Format: Educational poster
I devoted a lot of time to research in depth the virus and flesh out concepts worth to visualize.
Initially, I was planning to showcase how the viral fusion mechanism works and the dramatic conformational change the S protein undergoes. However, I felt that an animation rather than a still image would serve this point better and this already existed (created by Jonathan Khao and Gaël McGill).
I ended up finding a paper that exposed the Achille's heel of the virus. The S protein of the virus which is needed for the attachment to cells is covered with glycans and thus it's mostly inaccessible by antibodies. However, the small receptor binding domain (RBD) of the S protein has low level of glycosylation and thus it appears to be the target of 90% of the neutralizing activity present in
SARS-CoV-2 immune sera.
To create an accurate represenation of the interaction of the virus with ACE2 I had to utilize a plethora of protein models and molecular tools. The first step in this process was to create a complete model of the S protein because even though the existing structures at the time (November 2020) had critical information, they were incomplete (missing loops, missing glycosylation etc). I used de novo modeling with MODELLER to model the missing loops.
Then the next step was to orient ACE2 and the S protein in the membrane of the cell and the membrane of the virus respectively. To do that I used PPM server from OPM database.
Once I had the proteins oriented in the membranes I had to 1. model the glycans to the S protein and 2. to build the membranes of the cell and the virus according to their lipoprofiles. To do both of these, I used CHARMM-GUI.
Once I had the complete ACE2 and S protein -positioned on their respective membranes- the only missing thing was their interaction. Thankfully, 6M17 entry on RCSB-PDB provided the needed information having the RBD region of the S protein bound to the ACE2 receptor. It was only a matter of aligning the models created previously to the corresponding structures of 6M17. Many of the previously described steps are explained in more detail in Dr. Amaro's paper.
During this stage, and while I was debating on whether I should include the rest of the viral membrane proteins (E and M proteins) or not, I found evidence that the orientation of the E protein in the membrane might have been erroneously visualized being upside down (due to the lack of experimental data up until that point and mixed in silico results). After digging up more this discrepancy with the help of Dr. Maizels, I came to realize that indeed the N-terminus is extravirion-lumenal while the C-terminus is intravirion-cytoplasmic. Soon after that, I made sure that the medical illustration community knows about this and I was honored to see my name brought up by Dr. Fairman in her webinar with title COVID-19: Visualizing a Moving Target. For the record, I decided to avoid showing E and M proteins.
Once I had all my molecular models ready, I moved in ZBrush to create low-poly versions of my models. Then I went into 3ds Max to set up my scene and I used tyFlow to create the viral and cellular membranes and also to distribute the surface proteins.
Next step of the process was to composite my render in the post-production phase to make the illustration dynamic and immersive.