NEW YORK – As scientists have worked throughout the course of the COVID-19 pandemic to identify or develop drugs targeting the virus, protein-protein interactions studies have emerged as a potential tool for such efforts.
Using this approach, researchers have identified dozens of potential agents that could be useful for treating COVID-19, and while few if any have made it into widespread clinical practice, a number are in various phases of clinical trials.
More generally, the efforts have highlighted the potential of protein-protein interaction studies to inform drug development, said Nevan Krogan, director of the Quantitative Biosciences Institute at the University of California, San Francisco School of Pharmacy and lead author on several studies taking such an approach to look at potential COVID-19 treatment targets.
"This is a way we can go now," he said, adding that the method could prove useful beyond COVID-19.
"If you have a mutated protein in cancer, everyone looks at that to target it," Krogan said. "Maybe it's not that druggable. But maybe it is connected to another protein that is druggable. It's conceptually the same thing."
"We were pushing this vision before," he said, but, he noted, the COVID-19 work had drawn substantial new attention to the idea.
Starting with an April 2000 Nature study, Krogan and colleagues numbering more than 100 researchers around the world have published a series of papers using protein interaction data and phosphoproteomics to identify potential COVID-19 drug targets and candidates for drug repurposing.
In the Nature paper, the researchers generated protein expression plasmids for 26 of the 29 proteins that are predicted to be produced by SARS-CoV-2, based on its genome, and transfected these plasmids into human HEK293T cells. After the proteins were expressed, they lysed the cells and pulled out the viral proteins via affinity purification, followed by mass spectrometry to identify human proteins that were bound to them.
They then explored which molecules might disrupt these interactions by looking for ligands of the human interacting proteins using chemoinformatics and literature searches, identifying 69 agents targeting 63 of the human interactors.
Using this information, the researchers performed viral assays, screening 47 of the 69 identified compounds along with an additional 28 molecules identified independently of the protein interaction experiments.
Among the more promising candidates identified by this work was zotatifin, an inhibitor of the translation initiation factor eIF4A that the drug company Effector Therapeutics had been studying as a potential treatment for breast and lung cancer. In July, Effector launched Phase Ib clinical trials in partnership with QBI looking at the safety and tolerability of the drug in outpatients with mild to moderate COVID-19. The trial is being funded with $5 million from the US Defense Advanced Research Projects Agency and Defense Health Agency.
Another drug identified by the work was plitidepsin, a cancer drug candidate marketed by Spanish pharm PharmaMar as Aplidin, that inhibits the eEF1A protein, which Krogan and his collaborators identified as an interactor with SARS-CoV-2 proteins. In May of this year, PharmaMar presented data from a Phase II trial studying the drug in hospitalized patients with mild, moderate, or severe cases of COVID-19 in which it met safety endpoints and showed some clinical efficacy, particularly in patients with moderate disease.
In February, the company launched a Phase III trial to study the efficacy of the drug for treating hospitalized patients with moderate infections. The trial aims to enroll around 600 patients at sites in 12 countries and to compare treatment with plitidepsin at two dose levels versus each country's standard of care.
The researchers also identified hydroxychloroquine as a potential hit. The drug has been one of the most widely studied potential treatments for COVID-19 and the subject of much controversy, but clinical trials involving thousands of patients have found that it has little or no effectiveness at preventing infection or treating infected patients.
Several other groups have also published research using protein-protein interaction studies in combination with other techniques to identify potential targets and drugs for treating COVID-19. In April 2021, a team led by researchers at the Technical University of Munich and the Max Planck Institute of Biochemistry looked at the SARS-CoV-2 proteome and how its proteins impacted the host proteome including measuring a range of post-translational modifications. Using affinity purification mass spec, the team identified 1,801 interactions between host proteins and SARS-CoV-2 and SARS-CoV proteins. Among the most promising drug candidates they identified were the FLT3/AXL inhibitor gilteritinib, the AKT inhibitor ipatasertib, and the matrix metalloprotease inhibitors prinomastat and marimastat, though none of those agents are currently in clinical trials for treating COVID-19 based on a search of clinicaltrials.gov.
In October, a team led by researchers at the University of Helsinki published a SARS-CoV-2-host protein interaction study in Molecular Systems Biology using both affinity purification mass spectrometry and the BioID proximity labeling approach that map interactions between viral and host proteins. The authors also looked at so-called host "hub" proteins that shared interactions with both the viral bait proteins and host bait, identifying 693 such proteins. Using this data they identified 59 drugs targeting 15 proteins as potential candidates for repurposing as COVID-19 therapies.
"The basic strength of this study was really the basic biology and identifying the key interactions between the host and the viral proteins and to go from that then to really what the functions are and to identify possible targets and then look if we could find existing drugs that would target these," said Markku Varjosalo, group leader at the University of Helsinki and senior author on the paper.
He said that he expected the approach would not identify a single drug that by itself would prove highly effective as a COVID-19 therapy but that by looking at a number of targets it could aid development of combination therapies that would disrupt multiple viral processes.
Varjosalo and his colleagues used an in vitro drug screen to look at 10 of the 59 agents they identified in the study as potentially useful, finding that six of them — baicalein, methotrexate, guadecitabine, PX478, mizoribine, and BMS-863233 — showed anti-SARS-CoV-2 activity. Of those, methotrexate is currently being studied in a pair of Brazilian clinical trials looking at the drug's effectiveness for treating mild and severe COVID-19 cases.
Of course, taking agents identified through such work and demonstrating they are actually effective in patients is a challenging process, as shown by the small number of effective antiviral drugs currently in use for fighting COVID-19.
"The human system is very complex," Varjosalo said, noting that even if a compound shows promise in vitro it can prove ineffective in actual patients due to factors like interactions with other molecules in the body or an inability to access in vivo the target protein.
"You really need to have big numbers, tens of thousands of patients, to get a clear view of if a [drug] is beneficial or not," he said, adding, though, that protein-interaction work offered a way to more rapidly identify a wide range of good candidates.
"It helps when you know sort of the search space where you are looking for targets and then the drugs against those targets," Varjosalo said. "And through the same process you learn a lot about the basic biology. It's nice that more and more of these studies are coming out so that we can see what is background and what is real and what are the high-confidence interactions."