Designer Protein Inhibitors
Until an effective vaccine becomes available, the best method to counter SARS-CoV-2 infection is by prophylactic inhibitors. The viral machinery needs to engage cells through their surface proteins, and these are often the sites the antibodies target as part of our immune response. The antibodies that can successfully neutralize the pathogen often interact with critical viral proteins.
At the time of this writing, there are no reported SARS-CoV-2-specific neutralizing antibodies (nAbs). There are on-going research on isolating antibodies from infected patients, but their development and characterization takes time. While it is possible to use convalescent plasma to deliver antibodies as treatments, alternative strategies that the Protein Design lab at Stanford Bioengineering, led by Professor Possu Huang, are developing are: (1) designing small protein inhibitors and (2) directly designing neutralizing antibodies. We have the ability to accelerate the development of these agents.
Using the state-of-the-art design methodology, we specialize in the creation of de novo proteins. As a countermeasure to COVID-19, we can take the known structures of the spike protein and develop novel proteins directed at the functional sites of the spike to neutralize the viral function. The tools originally developed by our lab had been applied to create inhibitors for influenza1, and we are now applying the same methodology to develop a broad range of inhibitory proteins in the urgent response to COVID-19. Earlier designs of these small proteins have exhibited exceptional stability (long shelf life), low immunogenicity, and great versatility in application.
These designer small proteins specific for SARS-CoV-2 can be implemented as aerosol inhalers for prophylactic protection, and they can also be use as urgent remediation to the lung to reduce active viral loads. They can also be implemented into test-kits. We are collaborating with Prof. Alice Ting’s group on the test kit development.
For intravenous treatment, antibody-based biologics have well-established history. Our group is pioneering a neural network design pipeline2 that specializes in creating antibody structures. The neural network can produce vast variations of antibodies in silico for screening against the desired neutralizing sites on the viral spike protein.
We can potentially bypass the need to isolate nAbs from patients and expedite the creation of nAbs directly in the lab. The antibody-based designs will potentially encounter fewer immunogenicity issues as injectable drugs.
Relevant Publications or More Information:
1. Chevalier, A. et al. Massively parallel de novo protein design for targeted therapeutics. Nature 550, 74–79 (2017).
2. Anand, N. & Huang, P.-S. Generative modeling for protein structures. Advances in Neural Information Processing Systems 31, 7494-7505. (2018)