In March, a team of scientists reported that they had run and analyzed a computational screen that helped them pinpoint 69 compounds that might treat COVID-19, the disease caused by the novel coronavirus SARS-CoV-2. They now have new data from lab experiments showing that some of those compounds can stop the virus from replicating in cells. The hits include a cancer drug currently in clinical trials, an over-the-counter antihistamine, and a compound that’s never been tested in humans, but out-performed hydroxychloroquine in the cell studies. The researchers also found that dextromethorphan, the active ingredient in many cough suppressants, promoted the growth of the virus in cells (Nature 2020, DOI: 10.1038/s41586-020-2286-9).
The international group, led by molecular biologist Nevan Krogan of the University of California, San Francisco, identified the original 69 compounds by running a screen to look for human proteins that might interact with the virus’s proteins. Their program then searched for molecules that could disrupt those potential interactions. To get the new data, part of the team at Mount Sinai Hospital in New York and the Pasteur Institute in Paris tested 47 of these compounds, about two-thirds of the 69, to see how the drugs interact with the virus in African green monkey cells. They chose these cells because SARS-CoV-2 replicates at high levels in them, and the results translate well to human cells. The researchers are now studying the remaining compounds.
The team found that two kinds of compounds seemed to block the virus’s replication in the cells: those that inhibit the translation of the viral RNA into proteins, and molecules that modulate Sigma1 and Sigma2 receptors, which play a role in cell stress signaling. The scientists also found that these two types of molecules interact with the virus in different ways, suggesting that a combination of two or more drugs could be an effective approach to treating COVID-19.
The new data in the Nature paper provides complementary information to what was published in the March pre-print (bioRxiv 2020. DOI: 10.1101/2020.03.22.002386), says Dennis Liotta, executive director for the Emory Institute for Drug Development. “It points out something we’ve known about viral disease for a long time, which is the really effective therapies are always going to be combinations.”
“We’re particularly excited about PB28, which seemingly has the most potent antiviral activity out of all of them,” Krogan said during an April 30 press briefing about the new data. PB28 is a preclinical antipsychotic compound, and has about 20 times the antiviral activity of hydroxychloroquine in the cell experiments. This compound has never been studied in humans, so “it’s a little bit further away from being used,” said Brian Shoichet, a computational biologist at UCSF and part of the research team.
Zotatifin, a compound that’s currently in clinical trials as a possible treatment for solid tumor cancers, also showed high antiviral activity, as did two antihistamines, clemastine and cloperastine. Clemastine is available over the counter in the US as an allergy medication, while cloperastine is available in some countries as a cough suppressant.
Hydroxychloroquine, a compound that’s currently in multiple clinical trials as a possible treatment for COVID-19, could restrict replication of the virus in cells; however the compound was toxic to the cells, a finding that mirrors some of what’s being found when treating people. Some clinical trials have been halted due to cardiac side effects. The team found that hydroxychloroquine binds to the hERG ion channel, which helps regulate the heart’s electrical activity. Hydroxychloroquine binds more tightly to this heart ion channel than it does to the Sigma receptors, Krogan says. The team predicts that clemastine and PB28 would not have the same heart toxicity issues as hydroxychloroquine because they do not bind as tightly to the hERG channel.
The female hormone progesterone also had some antiviral activity in the experiments. Progesterone concentration is higher in women than in men, but not by much, Shoichet says, which could explain some of the difference in the rates of infection and death in women and men. But he points out that the antiviral activity of progesterone isn’t strong, and the hormone has multiple effects in the body, so what effects it does have on COVID-19 are going to be complex. “It’s an attractive hypothesis, but it needs a lot more study,” Shoichet says.
The team also found a compound that may help SARS-CoV-2 replicate. Dextromethorphan, found in several over-the-counter cough medicines such as Children’s Robitussin and Delsym, promoted viral infection of cells in the lab experiments. The researchers aren’t sure of the mechanism, but hypothesize that the compound activates Sigma1, which may help turn on the stress response that the virus hijacks. “It’s speculative at this point,” Shoichet says. He stresses that these are cell studies in a lab, not in a person with COVID-19, so the results don’t necessarily mean that dextromethorphan and medicines that contain it are harmful. “More more work needs to be done there for sure, but it’s something to look out for,” he says.