NEW YORK – A team of US researchers has leveraged transcriptomics data sets to identify existing drugs that may help treat COVID-19.
First, researchers in Marina Sirota's lab at the University of California, San Francisco obtained gene expression profiles, including up- and downregulated genes, from three studies of human cells infected with SARS-CoV-2. Then, they compared them to publicly available datasets of drug-induced gene expression in cultured human cells.
"Our hypothesis is that if the gene expression signature for a drug is opposite that of COVID-19, that drug could be therapeutic," Tomiko Oskotsky, a research scientist in the Sirota lab, said in a presentation Wednesday at the virtual American Society of Human Genetics conference.
The algorithm compared each of the three profiles taken from the SARS-CoV-2 studies to more than 1,300 available drugs, creating a gene expression reversal score for each profile-drug combo. Of 102 total drug hits, 25 drugs reversed two of the three profiles and four drugs reversed all three: bacampicillin, an antibacterial drug; haloperidol, an antipsychotic drug used to treat schizophrenia; clofazimine, a drug used to treat leprosy and tuberculosis; and valproic acid, a psychiatric drug used to treat epilepsy and bipolar disorder.
"These aren't the first [drugs] to come to mind for treating COVID-19, let alone a viral infection," Oskotsky said. "So, it's interesting to see analyses from three transcriptomic signatures identifying these as being potentially therapeutic."
The Sirota lab's drug repositioning pipeline has previously returned candidates for therapies for Crohn's disease and preterm birth. With over 44 million total COVID-19 cases worldwide and a reacceleration of the pandemic seen on several continents, the need for therapies remains urgent. Remdesivir, which had previously been used to treat Ebola, has shown promise and neutralizing antibodies are in development, but researchers are still searching for treatments that could help slow the pandemic.
In follow-up validation studies performed in Melanie Ott's lab at the UCSF Gladstone Institute and Adolfo Garcia-Sastre's lab at Icahn School of Medicine at Mount Sinai, researchers tested 16 of the 25 drugs that reversed two of three profiles, chosen for their side effect profile and availability. Between the two labs, 11 of 16 drugs showed they could inhibit viral infection in two human cell line assays, and haloperidol, clofazimine, and cyclosporin all showed viral inhibition effects.
Oskotsky noted that the study was limited in terms of transcriptomic profiles taken from actual COVID-19 patients. Just one of the profiles was generated from patient samples.
The researchers also posted their results in a BioRxiv preprint on Oct. 23, 2020.
The group plans to look at combination therapies as well as leverage electronic medical record data to further explore drug repositioning for COVID-19, Oskotsky said.