As the COVID-19 pandemic engulfs our world, many scientists have dropped what they’re doing to fight the virus. Now, a group of researchers are searching for a new molecule to disrupt SARS-CoV-2, before it can enter the human cell. Finding the right one could lead to a vital intervention for infected patients. Researchers from the Pentelute lab at MIT recently discovered a peptide that strongly binds the spike protein in SARS-CoV2 [1]. Currently, they collaborate with Prof. Rafael Gomez-Bombarelli (MIT Department of Materials Science and Engineering) to continue to hone in on promising peptides with machine-learning. SARS-CoV-2 binds to a specific enzyme found on the surface of human cells. The enzyme, named angiotensin-converting enzyme 2, or ACE2 for short, lines endothelial cells in the lungs’ blood vessels. The enzyme happens to be the same target for SARS-CoV, the virus that caused SARS. Like other kinds of coronaviruses, SARS-CoV-2 forms a series of “spike” proteins which protrude from its core, giving it the shape of the sun’s corona. Those spike proteins latch onto ACE2, which allows the virus to fuse its membrane to the cell membrane. Once inside, SARS-CoV-2 acts like any other virus - it hijacks the human cell, forcing the cell to produce many copies of the virus, spurring an infection. The group of researchers began their work by looking at the structure of the spike protein and ACE2 - specifically where the two meet. They found a large, elongated surface area, which is common for interactions between two proteins. A smaller molecule would be unlikely to bind to this large surface. Peptides, on the other hand, can cover a larger area and have the chemical properties needed to apprehend the virus before it sticks to a cell. The success of existing peptide-based drugs, like the clinically-approved Fuzeon to treat HIV, hint that this direction could be promising. Next they took on the challenge of synthesizing a winning peptide. To do this, they analyzed the structure of ACE2, in order to find a peptide that could mimic the enzyme and bind to the virus. Eventually, they honed in on SBP1, a peptide that is 23 amino acids long. In nature, ribosomes produce peptides inside the cell. But in this case, the team used a fast-working machine to churn out SBP1. Using a molecular dynamics simulation, they showed SBP1 could stick to the spike proteins of the coronavirus in a way that would potentially disrupt its ability to attack human cells. Since the SBP1 has an amino acid sequence derived from ACE2, it is considered endogenous - in other words, it’s something naturally found in the human body, and therefore assumed to be tolerated by the immune system. It makes a promising candidate for treatments. Pentelute and the group are continuing their work to tweak the length and sequence of amino acids. In collaboration with Rafael Gomez-Bombarelli, they utilize machine learning to create a peptide with an even higher affinity for SARS-CoV-2. The team is sharing their current peptide - and any future versions - to other researchers working towards a treatment for COVID-19. It’s one step towards understanding the virus - and hopefully towards treating the disease.
[1] The first-in-class peptide binder to the SARS-CoV-2 spike protein
G. Zhang, S. Pomplun, A. R. Loftis, A. Loas, B. L. Pentelute
bioRxiv, 2020.
[2] Xenoprotein engineering via synthetic libraries
Gates, Zachary P., et al.
PNAS, 2018.