How a UCSF lab took on the global race to block the coronavirus

In late February, the virus expert Nevan Krogan called an early morning meeting at his UCSF lab in Mission Bay and told 20 fellow scientists that their lives were about to change.

The new coronavirus, which emerged in China, was now spreading from person to person in California. Health authorities had just confirmed it. Soon the virus known as SARS-CoV-2 would be everywhere, Krogan realized, and many were bound to die. But maybe the San Francisco scientists could do something to help, Krogan told the group, if they agreed to work through the night at the lab, achieving in months or weeks what normally would take years, racing to complete one big project.

“I just kind of flipped out,” recalled Krogan, 44, a tall, energetic Canadian native. “We have to do this now. And they’re like: ‘Yeah.’”

Two months later, the results of that quest are in, and they’re likely to draw wide attention, thanks to a striking paper that Krogan and his colleagues published Thursday in the journal Nature.

“We’ve found something about this virus that I hope can help people,” he said.

Krogan directs the Quantitative Biosciences Institute at UCSF, a group of research labs that study disease. QBI has been peering into the molecular world of the virus, trying to understand how it hijacks human cells and how to stop it. They partnered with teams at the Icahn School of Medicine at Mt. Sinai in New York and the Institut Pasteur in Paris, searching for existing drugs and experimental compounds that might block infection and testing them in the lab. And the group uncovered a range of drugs and compounds that showed strong antiviral effects — including five existing drugs that are already approved to fight mental illness, malaria, menopause, coughs and allergies.

None of the five FDA-approved drugs is approved to treat COVID-19, and some have dangerous side effects. The paper cautions that people with the disease should not take the drugs outside of controlled studies. But scientists “can learn from them,” Krogan said.

One is hydroxychloroquine, a malaria treatment made famous by President Trump, who has hawked it as a COVID-19 cure against the advice of health experts. Hydroxychloroquine can be toxic to the heart, and a recent study of coronavirus patients at Veterans Health Administration medical centers found a higher death rate in patients who received the drug.

The other four approved drugs that stopped the virus in the lab tests were haloperidol, often sold as Haldol, a widely available drug to treat schizophrenia; cloperastine, a cough suppressant that has been around since the 1970s; clemastine, an antihistamine; and progesterone, a natural hormone that is also available as a medication.

The rest of the drug candidates flagged in the Nature paper are experimental compounds. Previously, most have been studied as potential weapons against cancer.

The paper is more than a list of drugs. Crucially, Krogan says, the scientists have figured out how the virus infects human cells. They know the mechanism, and they’ve published it. Now any scientist in the world can try to design a better drug.

He even thinks they could repeat the process all over again, successfully, when the next pandemic hits. “In a few years,” he said, “we might be dealing with Covid-24.”

The starting gun

Krogan first heard about the epidemic in China late last year. At the time, SARS-CoV-2 was just another emerging virus, worth keeping an eye on, and for him, intrinsically fascinating.

The white walls of his office at QBI, within the School of Pharmacy and housed in a modern biotech building on UCSF’s Mission Bay campus, are covered with dry-erase notes on past projects about dengue fever, the Zika virus, Ebola, HIV. “That’s the game plan to cure HIV,” he said on a recent morning, pointing to a cluster of scribbles. “It’s all related.”

He wore a surgical mask, jeans, a Canadian Maple Leaf pin and a bright purple blazer made for him by a Ugandan tailor named Muwanguzi Charles; they met years ago when Krogan was doing infectious-disease research there. (Reached via Facebook Messenger, Charles said of Krogan, “He likes bold colors.”) Taking a few brisk steps down the hall, he gave a tour of the Krogan Lab itself, which featured the typical clutter of molecular biology: Flasks, shakers, micro-pipettes for handling fluids, honeycombed plastic plates, incubators for growing cells, science jokes taped to the walls:

KEEP CALM AND PIPETTE ON

CAUTION

MIXED GENDER LAB

NO FALLING IN LOVE OR CRYING PERMITTED

Krogan, who is also an investigator with the Gladstone Institutes, grew up in Regina, Saskatchewan, a city in northwestern Canada better known for producing hockey players than scientists. But he liked the precision of science, and during grad school in Toronto, while studying proteins in yeast, he became passionate about an approach called systems biology. Often, biologists study an organism one piece at a time, zooming in on a single gene or protein. Systems biologists instead begin with the big picture, analyzing large numbers of genes and proteins and combing the data for insight.

After joining UCSF in 2006, Krogan began to tackle viruses, and over the next 13 years, his lab developed an unusual skill set that would, in the midst of a global pandemic, prove very handy.

A virus can’t exist by itself; it needs our proteins to replicate, to subvert the body. Krogan and colleagues created a system for revealing and mapping those human proteins — the ones a virus really needs.

Ultimately, he believed, those maps could lead to powerful new antiviral drugs. Many scientists searching for antivirals choose to attack the virus itself, designing compounds to disrupt its machinery. (This is the story with remdesivir, a candidate drug for COVID-19 developed by Gilead Sciences and now being tested in patients.) But viruses can often escape those attacks by mutating their own shape, leading to resistance. The virus can’t change the host cells, though, which means that drugs focusing on the host can remain effective for longer and can potentially work against multiple types of viruses.

Krogan’s lab created maps for Ebola, Zika and HIV, among other viruses. Each took a year or more of painstaking work. In terms of drugs, nothing much came of these maps; outbreaks receded, funding dried up.

On Jan. 24, when there were only two confirmed coronavirus cases in the U.S., Krogan talked about a new project with two close colleagues in his lab, David Gordon and Gwendolyn Jang. They were all curious about SARS-CoV-2, and it was natural to start looking at it.

On a website called GenBank, scientists in Washington had shared the genetic code of a SARS-Co-V-2 strain found in a patient there. Gordon and Jang began to analyze the virus’ gene sequence.

Using cloning techniques, they created copies of the genes, one by one. Then they fashioned each piece of the virus into a kind of hook, lowering the hooks into human cells to see which human proteins got stuck to them. Those were probably the proteins that the virus needed most — and the ones most relevant to disease.

In the midst of this early research, Krogan flew to New York City the first week of February, and visited a longtime friend, Adolfo Garcia-Sastre, director of the Global Health and Emerging Pathogens Institute at Mt. Sinai.

A bearded, 55-year-old Spaniard, Garcia-Sastre is one of the world’s leading experts in flu viruses and vaccines. The two men first met years ago when Krogan took a sabbatical at Mount Sinai, and that night in February, they were having dinner at Garcia-Sastre’s Manhattan apartment. Krogan was feeling horrible, fighting a dry cough and chills; he now wonders if he had COVID-19 and didn’t know it. Garcia-Sastre gave him some tomato soup and wine, and they got to talking about the new coronavirus in China.

“I was telling him, this thing is going to go all around the world,” Garcia-Sastre recalled.

Economic collapse. Death. It was going to be “very nasty.”

“He just said it kind of matter-of-fact,” Krogan recalled. “‘Do you want some more wine?’”

When Krogan replied with a joke, saying maybe they should go hide up in Canada, Garcia-Sastre was having none of it. This was a time “for us to do some science,” he said.

Krogan agreed, and he kept an eye on the news as the virus spread and the case counts crept upward. At the Feb. 28 meeting in the lab, once he’d rallied the UCSF scientists to drop everything and focus on the map, the group came up with a strategy for upending their usual working methods and moving forward fast.

They realized that if they could complete the map, they could use it to dramatically accelerate the usual process of drug discovery. While it can take 10 years to design an entirely new drug, “we didn’t have time to start from scratch,” said Brian Shoichet, a computational chemist at UCSF and a longtime Krogan colleague. Instead, they would search libraries of existing drugs for something that worked against the virus.

There are tens of thousands of FDA-approved drugs and experimental compounds. Without the map, getting a hit would be like throwing darts in a darkened bar. But with the map, they hoped to quickly zero in on a few dozen promising candidates, then test those against the virus in the lab and learn how they functioned, gathering clues to make even better drugs.

Krogan feared the university would close all the labs if the epidemic worsened. No one in the group, him included, was sure they could finish the map and get it published before they were shut down. Molecular biology is delicate work. All kinds of things can go wrong even in the best circumstances. But only a few labs in the world were deeply familiar with how to do these experiments, and theirs happened to be one of them.

“We knew that if we were to get this data published, it would be huge,” recalled 23-year-old Jeffrey Guo, a recent UC Berkeley graduate who was working as Gordon’s assistant. “We just knew it would be a sprint.”

The new rules

By the first week of March, “People were on almost 24-hour shifts,” recalled Shoichet, who leads his own team in a space two floors above Krogan’s. “All hands on deck.”

The effort spread. Soon 21 other professors within QBI had pledged their labs to the effort. Krogan dubbed it the QBI Covid-19 Research Group, or QCRG. Researchers on every floor and every hallway were dividing up tasks and sharing data, assembling the map piece by piece.

No one working on the project had ever seen such a large collaboration grow so quickly, spurred by the urgency of the crisis. For many, it was a thrill — a vision of how science is supposed to work but often doesn’t.

“I don’t know if this will ever happen again,” Gordon said of the collaboration. “I hope it will.”

They pushed their equipment to the limit. Danielle Swaney, a 36-year-old assistant professor of cellular and molecular pharmacology at UCSF, managed the all-important mass spectrometers. Everyone just called them “the mass specs.” Mass specs are like gene sequencers for proteins, determining their identity and abundances. There were five mass specs, each costing up to $1 million. Normally, Swaney would only use one mass spec per protein sample, but now she was using three, processing two backup versions of each sample for safety’s sake — the project was so important, and they wouldn’t have time to redo it if something went wrong. The mass specs crunched proteins 24 hours a day to run 100 samples, emitting loud hums from their vacuum pumps.

As fast as they moved, the virus was always advancing faster. On March 10, the Golden State Warriors played a home game against the L.A. Clippers at the Chase Center in Mission Bay, a block from the Krogan lab. The next evening, Guo, a basketball fan, arrived at the lab to work a four-hour shift. Eight NBA teams across the country set picks and shot threes while Guo moved fluids and harvested proteins in a sterile biosafety hood. When he was finished, around 9 p.m., he walked outside, looked down at his phone and saw that the NBA had suspended the season. A few hours after that, the World Health Organization declared a pandemic.

“Time was always an issue,” Jang said. “Every day, every week, we were finding that we had less and less time.”

Meanwhile, Krogan was spending a big part of each day worrying about money.

He had basically launched the QCRG on faith. The institute had grants from the National Institutes of Health earmarked for other projects, and Krogan’s lab had received some funding in the past from the Swiss drug company Roche Pharma. But there was no money to study COVID-19, and the growing effort was putting a strain on existing funds. Krogan needed to hire more staff. He needed a cutting-edge electron microscope.

“It would be great if we had $50 million right now,” he told the newspaper that month. Where do you get $50 million?

Months earlier, he had set up a phone call with a UCSF donor he had met at a Giants game once but didn’t know well: Ron Conway, the Silicon Valley venture capitalist. Krogan was starting a company and wanted some advice. The call was set for March 12, Krogan recalled, and when he dialed Conway, “He immediately said, ‘What are you doing on Covid-19?’ And I told him what was going on, and he said, ‘I want to help you.’”

“He said he needed 10 million,” Conway recalled. “I was like, Jesus, thanks a lot. I thought he’d say maybe a million.”

Conway admired how Krogan had put UCSF on a wartime footing to solve a huge problem. “The minute he said, ‘I’m testing drug compounds against the virus,’ I was like, I know exactly what you’re up to, kaboom.” As soon as Conway hung up, he sent an email to his vast list of Silicon Valley contacts. He has since raised close to $5 million for QBI — halfway to the $10 million goal.

Over the next few days, the final pieces of the map fell into place, and with only hours to spare: On March 15, the university said it would shut down all research labs except for a few essential personnel. But the scientists had gotten what they needed to move forward: a blueprint of the virus’ “interactome,” consisting of 332 interactions between viral and human proteins.

The team could now move to a crucial next phase: Using the map to find drugs that might kill the virus.

First, they made a list of candidates. Shoichet and Kevan Shokat, a UCSF chemist, led a team that pored through databases for chemical compounds that act on the same human proteins that the virus seems to need, and ended up with a list of 69 possible drugs, both approved and experimental — everything from sedatives to cancer drugs to inflammation treatments to hair-loss medications.

Now they needed to test those candidates against the virus, placing each compound into a plate swimming with cells and plenty of SARS-CoV-2.

Only a handful of labs in the world are currently equipped to run such experiments with the live virus. Fortunately for Krogan, two of those labs are run by longtime friends.

One is Garcia-Sastre at Mt. Sinai.

In mid-March, Garcia-Sastre’s team, located on New York’s Upper East Side, began testing the drug candidates inside an enhanced Biosafety Level 3 laboratory. A BSL3 is required to work safely with the coronavirus; technicians wear full-body protective equipment, and the rooms are negatively pressurized, to prevent pathogens from escaping.

A few days later, across the Atlantic, the virologist Marco Vignuzzi joined the effort from Paris, agreeing to test the same set of drug candidates in the BSL3 labs of the Institut Pasteur, a mile from the Eiffel Tower.

“How nice would it be to know that I put a dent in what viruses do?” Vignuzzi told The Chronicle in a Zoom call in late March.

Vignuzzi got involved because he admired the “Californian go-for-it” vibe of the project — he has known Krogan and Shoichet since the 2000s, when he did his postdoctorate at UCSF. And the urgency was now clear. The pandemic was moving across Europe, with cases skyrocketing in Italy and rising in France. It was like living inside the 2011 movie “Contagion,” about the rampage of a killer virus and the scientists trying to stop it. Vignuzzi has seen the movie 20 times, he says, and always cries at the same scene: A virologist walks into a room of monkeys given experimental vaccines, and one monkey is alive.

“To me it’s a tear-jerker,” he said. “I know a lot of people don’t watch it that way.” (He added that he hasn’t rewatched “Contagion” since the pandemic started: “I’m good now.”)

While Paris and New York were busily testing the drugs, Krogan expanded the collaboration further. He tried to get the whole world involved.

On March 22, he and his colleagues published their protein map and the list of 69 drug candidates on biorxiv.org, a site where new research can be shared before it’s peer-reviewed. The idea was to get the information out quickly, so other scientists could test their own theories about the virus. Talking it over, the group also decided not to patent any of their techniques or discoveries: Sharing the data with no strings attached, they felt, would “ultimately expedite finding the treatment or the cure,” Krogan said.

Krogan not only tweeted a link to the paper as soon as it went online, he offered to mail ingredients for growing the proteins to any labs that were interested.

Ultimately, 318 labs in 38 countries took him up on the offer.

Gordon, who still had access to the UCSF lab, assembled the packages, squeezing drops of DNA into plastic plates, sealing the plates with wax and slipping them into FedEx envelopes destined for Quebec, Saudi Arabia, Boston, Toledo, Ohio.

Meanwhile, Krogan was mostly working from home in Bernal Heights, where he lives with his family. On a typical day, he woke up at 4 a.m., made a quick Zoom call to speak with the Paris researchers, gulped down a bowl of cereal while watching Edmonton Oilers clips on YouTube, and spent the next 18 hours checking in with scientists and funders. Sometimes he conducted a Zoom chat with 150 biologists and chemists all at once. Sometimes he needed to ask his Fortnite-playing son to stop hogging the family Wi-Fi.

The pandemic was still placing new obstacles in the team’s path. As March turned to April, America was well on its way to 2,000 daily COVID-19 deaths. Four members of Garcia-Sastre’s team came down with symptoms and needed to self-quarantine. And Trump started talking a lot about hydroxychloroquine, one of the drugs they were testing, portraying it as a kind of miracle cure.

It made the scientists nervous. They wondered what might happen when they released their results. If they found FDA-approved drugs that stopped the virus in the lab, would people get the wrong message and ingest them?

“Yeah, we’re terrified,” Shoichet said in early April.

What if people got hurt from the side effects?

What if the president said something wrong?

They realized they would need to write up their results with extreme care, making it clear they weren’t endorsing any drugs for COVID-19 patients.

“We talk about it all the time,” Krogan said.

But by now, their nervousness was giving way to excitement, as they saw data on the drug tests from New York and Paris.

“I think we have something here,” Krogan said in the first week of April. “I’m 99% sure.”

A few days later, after looking at more data, he said, “99.8%.”

It was clear that some of the drugs were really working. New York and Paris tested almost 50 of the 69 candidates; 10 of them showed “strong killing potential.” Some were many times more potent than remdesivir, at least in the lab setting. And the results were consistent between the two labs in multiple tests.

More than that, the tests were revealing secrets of the virus that could help scientists design improved versions of those drugs in the future.

“You have a mechanism of action,” said Garcia-Sastre. “You understand a little bit about how it works. And so you can make a better drug, sooner.”

For instance, the research was drawing a bull’s-eye on a pair of human proteins known as Sigma R1 and Sigma R2. These are like complicated doorknobs that control access to important machinery inside human cells. The virus knows how to turn them.

Seven of the drugs that were stopping the virus in the lab tests were doing it by manipulating Sigma R1 and R2. The antipsychotics, the antihistamines, hydroxychloroquine, the female hormone progesterone — they were all targeting these same two proteins. One promising experimental compound in this group was called PB28. It was 20 times more potent than hydroxychloroquine in the lab tests and appeared less toxic as well, because it didn’t bind as strongly to proteins that affect the heart.

For the scientists, this was a suitcase full of clues. Krogan said he is convinced that Sigma R1 and R2 are the key receptors that the virus needs to wreak havoc in the body and, luckily, the receptors are “very druggable” and can be tweaked in various ways. “From a scientific point of view, that’s huge, and we’re digging into this deep,” he said. “This is a way to manipulate the virus.”

The scientists also found a second class of potential drugs that interfere with SARS-CoV-2 in a different way, by foiling its ability to create proteins inside the cell. One of the most effective of these compounds, dA3-Ternatin 4, an experimental cancer agent, is similar to an FDA-approved drug called plitidepsin that fights multiple myeloma.

In coming weeks, some of the drug candidates will probably be given to COVID-19 patients in controlled settings. On Thursday morning, a drug company that owns one of the compounds, eFFECTOR, said it hoped to start a clinical trial. Roche Pharma also says it will fund some follow-up biology studies.

Depending on how these go, new results could be available in three to four months, according to John Young, head of infectious disease research for Roche, and Krogan anticipates new treatments for COVID-19, “hopefully by the end of the year.” He says that, ultimately, a cure for the disease might come from a combination of different drugs — a cocktail therapy, similar to how HIV is treated.

In the last few days, Krogan has continued working with New York and Paris to test more drugs and compounds; he says they’re still discovering new ones that kill the virus. He’s also looking ahead, trying to figure out how to keep the spirit of the project alive, to create some sort of lasting infrastructure to fight pandemics of the future.

What if the project was the skeleton of some bigger animal? After all, in three quick months, the scientists built a global team, mapped an unknown virus, searched for drugs, set up tests, got results. “The question is, why can’t we do this normally, with other viruses and diseases?” Krogan said.

“I’m going to be working on HIV like that,” he said. “I’m going to be working on TB. I want to be working on breast cancer and Parkinson’s,” he said. “So I just keep thinking about that.”

He has also been thinking about something more mundane: sleep.

“I’m exhausted,” Krogan said.