protein modeling
Protein Mapping Study Reveals Valuable Clues for COVID-19 Drug Development
Posted on by Dr. Francis Collins
One way to fight COVID-19 is with drugs that directly target SARS-CoV-2, the novel coronavirus that causes the disease. That’s the strategy employed by remdesivir, the only antiviral drug currently authorized by the U.S. Food and Drug Administration to treat COVID-19. Another promising strategy is drugs that target the proteins within human cells that the virus needs to infect, multiply, and spread.
With the aim of developing such protein-targeted antiviral drugs, a large, international team of researchers, funded in part by the NIH, has precisely and exhaustively mapped all of the interactions that take place between SARS-CoV-2 proteins and the human proteins found within infected host cells. They did the same for the related coronaviruses: SARS-CoV-1, the virus responsible for outbreaks of Severe Acute Respiratory Syndrome (SARS), which ended in 2004; and MERS-CoV, the virus that causes the now-rare Middle East Respiratory Syndrome (MERS).
The goal, as reported in the journal Science, was to use these protein “interactomes” to uncover vulnerabilities shared by all three coronaviruses. The hope is that the newfound knowledge about these shared proteins—and the pathways to which they belong—will inform efforts to develop new kinds of broad-spectrum antiviral therapeutics for use in the current and future coronavirus outbreaks.
Facilitated by the Quantitative Biosciences Institute Research Group, the team, which included David E. Gordon and Nevan Krogan, University of California, San Francisco, and hundreds of other scientists from around the world, successfully mapped nearly 400 protein-protein interactions between SARS-CoV-2 and human proteins.
You can see one of these interactions in the video above. The video starts out with an image of the Orf9b protein of SARS-CoV-2, which normally consists of two linked molecules (blue and orange). But researchers discovered that Orf9b dissociates into a single molecule (orange) when it interacts with the human protein TOM70 (teal). Through detailed structural analysis using cryo-electron microscopy (cryo-EM), the team went on to predict that this interaction may disrupt a key interaction between TOM70 and another human protein called HSP90.
While further study is needed to understand all the details and their implications, it suggests that this interaction may alter important aspects of the human immune response, including blocking interferon signals that are crucial for sounding the alarm to prevent serious illness. While there is no drug immediately available to target Orf9b or TOM70, the findings point to this interaction as a potentially valuable target for treating COVID-19 and other diseases caused by coronaviruses.
This is just one intriguing example out of 389 interactions between SARS-CoV-2 and human proteins uncovered in the new study. The researchers also identified 366 interactions between human and SARS-CoV-1 proteins and 296 for MERS-CoV. They were especially interested in shared interactions that take place between certain human proteins and the corresponding proteins in all three coronaviruses.
To learn more about the significance of these protein-protein interactions, the researchers conducted a series of studies to find out how disrupting each of the human proteins influences SARS-CoV-2’s ability to infect human cells. These studies narrowed the list to 73 human proteins that the virus depends on to replicate.
Among them were the receptor for an inflammatory signaling molecule called IL-17, which has been suggested as an indicator of COVID-19 severity. Two other human proteins—PGES-2 and SIGMAR1—were of particular interest because they are targets of existing drugs, including the anti-inflammatory indomethacin for PGES-2 and antipsychotics like haloperidol for SIGMAR1.
To connect the molecular-level data to existing clinical information for people with COVID-19, the researchers looked to medical billing data for nearly 740,000 Americans treated for COVID-19. They then zeroed in on those individuals who also happened to have been treated with drugs targeting PGES-2 or SIGMAR1. And the results were quite striking.
They found that COVID-19 patients taking indomethacin were less likely than those taking an anti-inflammatory that doesn’t target PGES-2 to require treatment at a hospital. Similarly, COVID-19 patients taking antipsychotic drugs like haloperidol that target SIGMAR1 were half as likely as those taking other types of antipsychotic drugs to require mechanical ventilation.
More research is needed before we can think of testing these or similar drugs against COVID-19 in human clinical trials. Yet these findings provide a remarkable demonstration of how basic molecular and structural biological findings can be combined with clinical data to yield valuable new clues for treating COVID-19 and other viral illnesses, perhaps by repurposing existing drugs. Not only is NIH-supported basic science essential for addressing the challenges of the current pandemic, it is building a strong foundation of fundamental knowledge that will make us better prepared to deal with infectious disease threats in the future.
Reference:
[1] Comparative host-coronavirus protein interaction networks reveal pan-viral disease mechanisms. Gordon DE et al. Science. 2020 Oct 15:eabe9403.
Links:
Coronavirus (COVID-19) (NIH)
Krogan Lab (University of California, San Francisco)
NIH Support: National Institute of Allergy and Infectious Diseases; National Institute of Neurological Disorders and Stroke; National Institute of General Medical Sciences
AIDS Vaccine Research: Better By Design?
Posted on by Dr. Francis Collins

Caption: eOD-GT8 60mer nanoparticle based on the engineered protein eOD-GT8. Yellow shows where eOD-GT8 binds antibodies; white is the protein surface outside the binding site; light blue indicates the sugars attached to the protein; dark blue is the nanoparticle core to which eOD-GT8 has been fused.
Credit: Sergey Menis and William Schief, The Scripps Research Institute
A while ago, I highlighted a promising new approach for designing a vaccine against the human immunodeficiency virus (HIV), the cause of AIDS. This strategy would “take the immune system to school” and teach it a series of lessons using several vaccine injections—each consisting of a different HIV proteins designed to push the immune system, step by step, toward the production of protective antibodies capable of fending off virtually all HIV strains. But a big unanswered question was whether most people actually possess the specific type of precursor immune cells that that can be taught to produce antibodies that kill HIV.
Now, we may have the answer [1]. In a study published in the journal Science, a research team, partly supported by NIH, found that the majority of people do indeed have these precursor cells. While the total number of these cells in each person may be low, this may be all that’s needed for the immune system to recognize a vaccine. Based in part on these findings, researchers plan to launch a Phase 1 clinical trial in human volunteers to see if their latest engineered protein can find these precursor cells and begin coaxing them through the complicated process of producing protective antibodies.
LabTV: Curious about Computer Modeling of Proteins
Posted on by Dr. Francis Collins
In many ways, Josh Carter is a typical college student, with a hectic schedule packed with classes and social activities. But when he enters a structural biology lab at Montana State University in Bozeman, Carter encounters an even faster paced world in which molecular interactions can be measured in femtoseconds—that is, 1 millionth of 1 billionth of 1 second.
Working under the expert eye of principal investigator Blake Wiedenheft, Carter is applying his computational skills to X-ray crystallography data to model the structures of various proteins, as well as to chart their evolution over time and map their highly dynamic interactions with other proteins and molecules. This basic science work is part of this NIH-funded lab’s larger mission to understand how bacteria defend themselves from the viruses that try to infect them. It’s a fascinating area of science with a wide range of potential applications, from treating diseases that arise from imbalances in the microbiome (the communities of microbes that live in and on our bodies) to developing new methods for gene editing and programmable control of gene expression.