By Nevan Krogan
The more researchers know about how the coronavirus attaches, invades and hijacks human cells, the more effective the search for drugs to fight it. That was the idea my colleagues and I hoped to be true when we began building a map of the coronavirus two months ago. The map shows all of the coronavirus proteins and all of the proteins found in the human body that those viral proteins could interact with.
In theory, any intersection on the map between viral and human proteins is a place where drugs could fight the coronavirus . But instead of trying to develop new drugs to work on these points of interaction, we turned to the more than 2,000 unique drugs already approved by the FDA for human use. We believed that somewhere on this long list would be a few drugs or compounds that interact with the very same human proteins as the coronavirus.
We were right.
Our multidisciplinary team of researchers at the University of California, San Francisco, called the QCRG , identified 69 existing drugs and compounds with potential to treat COVID-19 . A month ago, we began shipping boxes of these drugs off to Institut Pasteur in Paris and Mount Sinai in New York to see if they do in fact fight the coronavirus.
In the last four weeks, we have tested 47 of these drugs and compounds in the lab against live coronavirus. I’m happy to report we’ve identified some strong treatment leads and identified two separate mechanisms for how these drugs affect SARS-CoV-2 infection. Our findings were published on April 30 in the journal Nature .
Every place that a coronavirus protein interacts with a human protein is a potential druggable site. QBI Coronavirus Research Group, CC BY-ND
The testing process
The map we developed and the FDA drug catalog we screened it against showed that there were potential interactions between the virus, human cells and existing drugs or compounds. But we didn’t know whether the drugs we identified would make a person more resistant to the virus, more susceptible or do anything at all.
To find those answers we needed three things: the drugs, live virus and cells in which to test them. It would be optimal to test the drugs in infected human cells. However, scientists don’t yet know which human cells work best for studying the coronavirus in the laboratory. Instead we used African green monkey cells, which are frequently used in place of human cells to test antiviral drugs. They can be readily infected with the coronavirus and respond to drugs very closely to the way human cells do.
These are preliminary findings and have not been tested in people. No one should go out and buy these drugs.
After infecting these monkey cells with live virus, our partners in Paris and New York added the drugs we identified to half and kept the other half as controls. They then measured the amount of virus in the samples and the number of cells that were alive. If the samples with drugs had a lower virus count and more cells alive compared to the control, that would suggest the drugs disrupt viral replication. The teams were also looking to see how toxic the drugs were to the cells.
After sorting through the results of hundreds of experiments using 47 of the predicted drugs, it seems our interaction predictions were correct. Some of the drugs do in fact work to fight the coronavirus, while others make cells more susceptible to infection.
It is incredibly important to remember that these are preliminary findings and have not been tested in people. No one should go out and buy these drugs.
But the results are interesting for two reasons. Not only did we find individual drugs that look promising to fight the coronavirus or may make people more susceptible to it; we know, at a cellular level, why this is happening.
We identified two groups of drugs that affect the virus and they do it two different ways, one of which has never been described.