The New Weapon in the Covid-19 War

By Michael Lewis

June 22, 2020, 5:00 AM EDT

Aerial view of people picnicing inside circles painted six feet apart on the grass in Dolores Park in San Francisco on May 22

A series of dispatches from America in the age of Covid-19.

No one who passed the young woman on the street that night gave her so much as a second glance. It was a Saturday evening in the South Beach neighborhood of San Francisco and she was heading away from the stadium where the San Francisco Giants were meant to be opening their season and toward the arena where the Golden State Warriors were meant to be ending theirs.

The young woman carried a small, white Styrofoam cooler that looked as if it had made one too many fishing trips. She wore a mask and running shoes. She made this trip almost every night and she might as well have been invisible. And she thought that was funny, because of what she usually had inside her Styrofoam cooler: racks of tubes filled with Covid-19. The buildings in the area housed some of the world’s leading labs studying deadly diseases: Ebola, Zika, MERS, SARS, bubonic plague. They kept it all in refrigerators, but sometimes they needed to move it around. “Hang out here long enough,” the young woman said as she walked, “and you realize you don’t have any idea of what people are carrying.”

Her name was Hanna Retallack. She was a microbiologist and student of Joe DeRisi, the virus hunter who, in early March, had turned the Chan Zuckerberg Biohub into what might be the country’s most interesting coronavirus testing lab. Retallack was just 29 but had established herself as a leading pupil in what I’ve come to think of as the Joe DeRisi School of Badass Virus Hunters.

One brief example, out of many. A few years ago, a great white shark washed up on a Santa Cruz beach — not dead, but madly thrashing about. DeRisi saw the news report, wondered if maybe the shark had a virus, and called up the California marine authorities. A guy there told him that actually hundreds of sharks were washing up on San Francisco Bay Area beaches. People had been calling to complain that some company must have dumped toxins into the bay. DeRisi asked his students if any of them wanted to go shark-virus hunting. Hanna Retallack was the first to raise her hand.

A few days later, Retallack drove herself to a beach just south of San Francisco International Airport and waded out into the water. “You need to collect some samples, and the fresher the better,” she said. She hoped to find a dying shark but instead found a dead one with its eyes intact and no stink — both signs that it had just died. She hauled it into the parking lot and popped open its skull. Its brain looked as if it had been exploded by a grenade.

Hanna Retallack then set out to do for the shark what Joe DeRisi had done for other species, including our own: isolate and understand its killer. DeRisi had helped to invent a technology that made this easy to do. It allowed Retallack to subtract from her shark specimen all the genetic material that naturally belonged to the shark. What remained were the genes that shouldn’t be inside a shark. As it turned out these belonged not to a virus but to a single-cell parasite that plagued fish farms. Miamiensis avidus. The parasite was not usually found in the Bay’s very salty waters — but that season had been unusually rainy. Retallack guessed that the rain reduced the salinity of the Bay to the point where it accommodated the parasite. She went back and asked the California authorities for their records of unexplained shark die-offs in the San Francisco Bay. There had been several, all in unusually rainy years. The sharks hadn’t been killed by some toxin. They’d been killed, in effect, by the weather.

Retallack, like her mentor DeRisi, was now training her virus detective skills on Covid-19. The white Styrofoam cooler that she carried through the streets of San Francisco was packed with plastic tubes filled with both human and coronavirus genes. She and three of Joe DeRisi’s graduate students, all women, had processed the tests themselves inside a lab across the street from Oracle Park. “The Bomb Squad,” DeRisi called them. When they found the virus, they defused it so that it could no longer infect people. But they preserved the genomes and carted them in this little Styrofoam cooler back to the Biohub’s smoking-cold minus-81-degree freezer — where they were about to do the most thrilling detective work in this entire pandemic.

At the moment Hanna Retallack was supposed to be finishing her Ph.D. in biomedical sciences, she was sliding trays from her cooler onto freezer shelves jammed with virus samples. I asked her why she had put her studies on hold to go hunting the coronavirus. “Who else would do it?” she said.

*****

We now have the first case study. It shows how Hanna Retallack and her colleagues inside the Chan Zuckerberg Biohub might dig us out of the mess we have gotten ourselves into. Seven weeks ago, researchers at the University of California San Francisco offered to test for Covid-19 everyone who either lived or worked in a fantastically diverse four-square-block area of San Francisco’s Mission District: census tract number 229.01. It wasn’t a simple thing to do. The tract was heavily Latino and wary of authority. An enterprising community organizer named Jon Jacobo, who had grown up in the neighborhood, made sure that each of its 1,400 front doors got banged on at least five times, and that all 4,087 official adult residents, including the small homeless population, were told about the coronavirus test in their native tongue.

“You talk about a PR battle,” said Jacobo, laughing. “We had people tell us that the virus is a hoax. We had other people tell us ‘It’s an ICE operation,’ or ‘We’re the guinea pigs. They’re going to test us, then they are going to test the vaccine on us too.’” Just banging on the doors had been an eye-opener. “The actual number of people living here is way different from the census count,” said Jacobo. One three-bedroom apartment that was meant to be housing five people contained 30, who were using the bedrooms around the clock, in shifts.

In the end, roughly 3,000 people showed up to be tested over four days in late April, and the Biohub processed their tests. A bit more than 6% of the Latinos were infected by Covid-19, most with high loads of the virus, though many had no symptoms. There were patterns in the test results — for example, the wealthier the person, the less likely he was to be infected. And of the 981 white people tested, zero were positive.

And so the big takeaway seemed to be what everyone in the past few weeks has figured out: The virus is now disproportionately attacking poor people of color, and lots of infectious people are walking around without a clue about their condition. But neither of those, it turns out, is the biggest takeaway. The biggest takeaway is this chart.

Source: Chan Zuckerberg Biohub
Note: People not connected with a line to a specific strain have a version of the virus with an undetermined genome and are placed according to best-guess assessments. An older infection is indicated by a positive antibody test result. Three workers who tested positive for Covid-19, but didn’t receive an antibody test to determine whether their infection was older or recent, are shown in the graphic as recent infections.

What you see here is a portrait of the genetic relationships of all the Covid-19 found in a four-square-block area of San Francisco in late April. A few weeks after the results were gathered, I sat in a room at the Biohub staring at a version of this image, trying to figure out what it meant. DeRisi and eight of his protégées — eight Hanna Retallacks — were beside me, turning their minds to this new problem.

The first thing they explained to me was that viruses mutate, in ways big and small, but they make errors in themselves at different rates. A perfectly stable virus — that is, a virus that does not mutate — would leave no meaningful fingerprints. Everyone infected by it would have the exact same virus. The virus genome would be able to say nothing about where they got it, or how, or from whom. On the other hand, if the virus was a lot less stable — if it mutated, as some do, many times inside a single person — its fingerprints would be smudged beyond recognition. From the point of view of a virus hunter, Covid-19 is, as DeRisi put it, “in the sweet spot. It mutates once every two or so transmissions.” Its fingerprints are rife with meaning and easy to track. You can tell where it’s been and follow its journey along the way, judging by how it’s changed.

The starting point for the diagram is not in the San Francisco neighborhood but in Wuhan, where the virus originated in early December. Zero means zero mutations. Drop in on any cluster of figures in the chart and you begin to see stuff that’s nearly impossible for virus hunters to see with the naked eye. Look at the household with the middle annotation. Three people in the same household were infected with the same virus. No big news there. One of them likely gave it to the others.

The news is how it entered the household — likely from the Mission resident on the same line. He or she has the exact same virus but contracted it earlier, which is why he or she has antibodies (marked with yellow). Without the genetic connection, you might never have any idea that these people had any sort of relationship at all. Even if a test had identified the person who infected the household, and that person was grilled by teams of contact tracers, the connection with the household might never have been made. The person might not even be aware of the connection, or might know it and want to hide it.

But when we know these people are surely connected, then the question becomes: How? Do they ride the same bus? Are they having an affair? Is it a little kid who happened to have played in the park near the child of the household? “How are these things linked?” asked DeRisi as he moved the cursor back and forth. “Are they linked by the building they are in? By the church they go to? All these things can now be deduced.”

Before my eyes, the novel coronavirus was being turned into a work of nonfiction. “It’s amazing that all these stories fall into place,” said DeRisi as he scrolled through the chart, moving, basically, from one week to the next. “This same resident who likely infected the household may have also infected two workers who don’t even live in the Mission (the household of workers one step further from those residents). It’s possible there is one degree of separation between them — that he gave it to someone who gave it to them. But no more than that. So how did that person infect those people?”

You have to stare at this image a bit to see its power. But look hard enough and you’ll see that, from the virus’s point of view, it’s a terrifying new weapon to slow the pandemic. By revealing the genetic relationship between the viruses, the diagram exposes the social relationships between people it infects. “It’s basically a spotlight that tells you where to dig,” said DeRisi. “In all of history, we’ve never had a really clear picture of the spread of the virus. That just changed.”

Back in 2003, when the original SARS virus started killing people in Hong Kong at a frightening rate, DeRisi sequenced its genome. But the process was too slow and expensive to be of practical use. “It’s 50,000-fold cheaper now than it was for SARS,” he told me. “What cost me $10,000 to do in 2001 now costs a penny.” And so we might now test for the virus in a way that gives us a picture that you can’t get from more conventional random sampling. Explore how the virus works in one neighborhood and you can apply what you learn to others. “Our state government should be doing this,” said DeRisi. “It should be asking: What are our social relationships and which ones lead to the transmission of disease? That’s what you would do in a rational society.”

It’s growing clearer that the coronavirus does not spread in an orderly way. Each infected person might infect two others on average but most people who get it infect no one. On the diagram, where DeRisi’s cursor lingers, he highlights a person with a particular talent for spreading the disease. The genetic information shows you the urgency of getting that person into quarantine, but it does more than that: It has the potential to lead you more generally to the social activity that’s spreading the disease.

It works the other way, too. The approach DeRisi has developed can be used not just to shut things down but to open them up. Last week, in Northern California, a pair of workers at a fish-packing plant came down with symptoms of Covid-19. The Biohub processed their tests and found both workers had the virus. In an age not all that distant from ours, the fish-packing plant, which believed it had taken the measures to keep its workers safe, would have been forced to close, as it would have had to assume that one of the workers had infected the other on the job. But then Joe DeRisi’s Badass Virus Hunters sequenced the two viruses and showed they were genetically far apart: The two workers had contracted the virus independently and outside of work. The fish-packing plant was able to stay open — and its workers were able to stay on their jobs.

The New Weapon in the Covid-19 War

Genetic Clues for Tracing Covid-19 Transmission