Jul 2, 2020

Axios Science

Welcome back and thanks for reading Axios Science. This week we look at the immune response to COVID-19, an emerging swine flu, birdsong and more.

Today's newsletter is 1,703 words, a 6-minute read.

1 big thing: The other immune responders to COVID-19

Illustration: Eniola Odetunde/Axios

Scientists are inching closer to understanding how antibodies and immune cells are unleashed by the body in response to the novel coronavirus.

Why it matters: Natural immunity differs from that afforded by vaccination but it offers clues for the design of effective vaccines and therapies, Eileen and I write.

Driving the news: The FDA earlier this week issued guidance on evaluating COVID-19 vaccines, placing an emphasis on assessing safety and efficacy through clinical trials.

  • That is standard procedure in the development of a vaccine, but the agency doubled down on its message in a climate of compressed timelines and as misinformation and skepticism of vaccines spread.
  • The FDA established a benchmark that a COVID-19 vaccine "would prevent disease or decrease its severity in at least 50% of people who are vaccinated."

Where it stands: Vaccine development typically centers on creating an antibody response to prevent infection from a virus. Here's the latest on the antibody response to SARS-CoV-2, the virus that causes COVID-19:

  • Most people who recover produce some antibodies to the virus, some of which are neutralizing antibodies that block the virus from entering cells.
  • The amount and type of antibodies appear to vary by person.
  • People with severe infections appear more likely to develop antibodies. "If you have evidence of more inflammation ... then your antibody levels tend to be higher," says Sanjeev Krishna, of St. George’s, University of London, who is a co-author of a different pre-print study that looked at the levels of a different antibody (IgG) to the virus in 177 people over time.
  • But not everyone produces antibodies, according to several recent studies. Krishna said they found 2%–8% of people infected with the virus didn't test positive for IgG antibodies.

One big question: For those who do develop antibodies, it's too early to say how long the antibodies will last and whether they will confer protection, Daniel Lingwood of Harvard Medical School tells Axios.

  • Krishna's study showed no decline after almost two months.
  • But a recent study published in Nature Medicine looking at 37 people with symptoms and 37 without found a decline in levels of both IgG and neutralizing antibodies after two or three months.

Meanwhile, the body's other immune actors for fending off viruses are increasingly receiving attention.

  • Helper T cells assist the body in remembering the targets for antibodies they deploy. Researchers are finding many people who recover from COVID-19 have helper T cells that target specific proteins on SARS-CoV-2.
  • Another type of T cell — killer T cells — is at the heart of the separate cellular immune response. These longer-term immune actors arrive a bit later in the process and help tamp down an infection and its severity by killing infected cells rather than protecting against infection.
  • A small study of patients admitted to the ICU for COVID-19 complications found the majority had SARS-CoV-2-specific killer T cells.

It's no surprise that T cells are important as they're known to help combat a host of different viral infections, says Mark Poznansky, director of the Vaccine & Immunotherapy Center at Massachusetts General Hospital.

  • Yes, but: It's still unknown how long the T-cell responses last, whether they prevent infection with SARS-CoV-2, and if the virus can evade them (and antibodies).

What to watch: T cells and the proteins they target on viruses should be considered as well as antibodies in vaccine design, Poznansky and other researchers urge.

  • A successful vaccine will depend on it working in a clinical study, not necessarily on the level of antibody response.
  • "The unknowns of immunity are overridden by whether it’s actually clinically working,” Poznansky adds.

Read the entire story.

2. A swine flu threat emerges in China

A masked worker checks the pigs in a hog pen in southwest China in February 2020. Photo: Feature China/Barcroft Media via Getty Images

A flu virus found in pigs in China has pandemic potential and should be "urgently" monitored, researchers warn in a study in the peer-reviewed journal PNAS this week, Eileen reports.

Threat alert: While it has not been transmitted from human to human, the virus is raising concern because it is a genetic mix of strains that have devastated humans before and there are indications it has "acquired increased human infectivity," the researchers say.

Driving the news: NIAID director Anthony Fauci told a Senate hearing on Tuesday there was no immediate threat but they are keeping an eye on it.

Details: The flu's genetics include characteristics from the 2009 and 1918 flu pandemics.

  • "They’re seeing the virus in swine, in pigs now, that have characteristics of the 2009 H1N1, of the original 1918, which many of our flu viruses have remnants ... as well as segments from other hosts, like swine,” Fauci said.
  • The researchers' other concern is that 10% of 338 people who worked with swine tested from 2016 to 2018 were found to be positive for the virus, G4 EA H1N1 (although most had not reported being ill).

What they're saying: "This is one of the flu viruses we'll have to keep track of, and watch, and understand what its potential might be for causing human infections, just like we do for H5N1 and H7N9, which are the other two potential pandemic flu viruses that we've also been tracking," Amesh Adalja, senior scholar at Johns Hopkins Center for Health Security, tells Axios.

What's next: If the virus does start human-to-human infection and moves toward an epidemic or pandemic, scientists will likely work on developing a targeted flu vaccine similar to 2009, but this does re-emphasize the need for a universal flu vaccine, says Adalja.

Go deeper

3. Math to root out AI bias

Illustration: Sarah Grillo/Axios

A team of mathematicians is offering a strategy to prevent algorithms used in business from pushing unethical policies, my Axios colleague Bryan Walsh writes.

Why it matters: Machine-learning algorithms are increasingly being deployed in commercial settings. If they are optimized only to seek maximum revenue, they can end up treating customers in unethical ways, putting companies at reputational or even regulatory risk.

How it works: In a paper published in Royal Society Open Science, researchers formulated what they call the "Unethical Optimization Principle."

  • It essentially boils down to the idea that "if there is an advantage to something that will be perceived as unethical, then it is quite likely the machine learning is going to find it," says Robert MacKay, a mathematician at the University of Warwick and an author of the paper.
  • MacKay uses the example of an algorithm that prices insurance products. If it is optimized only to maximize revenue, it's likely to treat customers unfairly and even unlawfully, selecting a higher price for users whose names code as non-white.
  • In their paper, MacKay and his colleagues lay out complex mathematics that can help businesses and regulators detect the unethical strategies an algorithm might pursue in a given space and identify how the AI should be modified to prevent that behavior.

The big picture: As increasingly sophisticated algorithms take more decisions out of the hands of humans, it becomes even more important for programmers to set initial clear limits.

  • Unfortunately, as a new survey from the data science platform Anaconda demonstrates, while data scientists are increasingly concerned about the ethical implications of their work, 39% of those polled say their team has no plans in place to address fairness or bias.
  • "Businesses using algorithms need to ask questions of 'ought' rather than just 'can,'" says Peter Wang, Anaconda's CEO.
4. Worthy of your time

An illustration of the disappearing star. Credit: ESO/L. Calçada

Stars aren't supposed to go out like this (Marina Koren — The Atlantic)

  • “If indeed the star turned into a black hole with no supernova at all, then it’s a case of ‘gone without a bang’ that astronomers have been searching for for a while now,” astronomer Iair Arcavi told Koren.
  • Sometimes finding nothing is the best finding.

Coronavirus antibody tests have a mathematical pitfall (Sarah Lewin Frasier — Scientific American)

  • "[E]ven with a very accurate test, the fewer people in a population who have a condition, the more likely it is that an individual's positive result is wrong."

Elk return to Kentucky, bringing economic life (Oliver Whang and Morgan Hornsby — NYT)

  • 13,000 elk are now living on reclaimed mine land in the state, drawing tourists, hunters and, so far, about $5 million each year to local economies in one of the country's poorest regions, they report.

Researchers use a nanosatellite to produce quantum signals (Axios)

  • Constellations of small, relatively less-expensive satellites that beam the signals from space to receivers on Earth may help researchers realize their vision for global quantum communications networks.
5. Something wondrous

White-throated sparrow. Photo: Scott M. Ramsay/Wilfrid Laurier University

Over 19 years, a once rare song sang by sparrows in western Canada spread across North America, replacing a traditional song along the way, according to new research.

The big picture: Birdsongs, like human languages, have dialects that can evolve, and birds and humans learn their languages in similar ways and time frames.

  • Studying birdsongs might help scientists to understand how humans develop dialects, says Angelika Nelson, an ornithologist at the Landesbund für Vogelschutz in Bavaria, Germany, who studied the white-throated sparrow song and wasn't involved in the study.

Birdsongs can change over time, but those changes are typically limited to a region and its dialect. In the case of the white-throated sparrows of North America, their song changed across the continent.

  • The traditional three-note-ending song, which dominated the repertoire of sparrows west of the Rockies, was abandoned for one that ends in two notes. (The mnemonic is Oh my sweet, Can-a-da, Can-a-da, Can-a-da! for the three-note tune versus Oh my sweet, Can-a, Can-a, Cana-da!)
  • As of 2019, only birds in the easternmost regions of the continent continue to sing the triplet-ending song.
"This would be like, if you were from Kentucky, and you move to Seattle, and everybody starts thinking, 'Hey, this Kentucky accent sounds awesome.' And suddenly 10 years later, everybody in Seattle has a Kentucky accent."
— Ken Otter of the University of Northern British Columbia

What they did: Otter and his colleagues used 1,785 birdsong recordings of male white-throated sparrows, Zonotrichia albicolis, collected by citizen scientists to show the spread of the new song, they report in Current Biology.

  • The team also tracked the location of sparrows with backpack geolocators and found those from the west, where the song was first observed, were overwintering with birds from farther east, where the song was later observed.
  • That suggests the younger birds were learning the new song from their seasonal companions and taking it back to their breeding grounds.

The intrigue: It's unclear why the new song is preferable.

  • The researchers think females may have a preference for mates who sing novel songs — but not too novel.
  • What to watch: Evidence for that comes from another song that has emerged — and spread — in the west. (This one is a doublet too, but the first note is modulated.)
  • "As of this year, it's completely replaced the traditional doublet. And so there was nothing super special about the doublet," says Otter. "It was just something different."