The reconstruction of a once-living landscape in northern Greenland from 2 million years ago, deduced from bits of DNA bound to minerals…
I’ve long been fascinated with the 1918 influenza pandemic because my grandfather Sam survived it. He married his nurse, lived 103 years, and likely had lifelong B cells that held the memory of his encounter with the flu. I wrote “A 1918 Flu Memoir” about him in 2008 for The American Journal of Bioethics.
We know very little about the 1918 pandemic flu, other than what it did to millions. The virus wasn’t even identified until 1933. Compare that to the deluge of SARS-CoV-2 genome sequences posted daily, nearly 11 million as I write this.
What we do know about the 1918 flu comes from bits of lung tissue from museum specimens or preserved in permafrost.
Sequencing Viral Genetic Material from Preserved Lungs
Much of what we know of the 1918 virus comes from the laboratory of Jeffery Taubenberger, senior investigator at the NIAID. In 1997 his team assembled nine fragments of the virus from samples of a single lung, publishing “Initial Genetic Characterization of the 1918 ‘Spanish’ Influenza Virus” in Science.
The victim was a young soldier whose lung slices were preserved at the Armed Forces Institute of Pathology, where Taubenberger worked in the 1990s. The researchers had meticulously probed four dozen lungs before finding one, from the young man, that was free of the bacteria that killed most victims so swiftly. The prized lung bore only the viruses, frozen in time at the beginning of their assault. Amplifying and sequencing pieces of viral genetic material – RNA – and comparing them to sequences from other strains of flu, revealed bird and pig origins. Flu viruses start in birds but some mingle in pig throats, then can pass to us.
Until recently, the handful of 1918 flu virus samples investigators had to work with came solely from the US and the UK. Now RNA sequences from lung tissue discovered in a German museum have provided information that has sparked reconsideration of how flu viruses evolve.
The new sequences reveal single-RNA-base (nucleotide) mutations that had been under the radar of the easier-to-track reassortment of genome pieces that had been the basis of classifying flu viruses. The findings reveal that seasonal H1N1 flu directly descended from the 1918 strain – that had been thought not to be so.
“Relics in Our Backyard”
Two groups, lead by Robert Koch Institute scientists Sébastien Calvignac-Spencer and Thorsten Wolff, published the new view of flu in Nature. They gave a webinar to journalists when the paper recently came out.
“The 1918 flu pandemic affected more than half of mankind and killed 50 to 100 million people. When we started we only had 18 specimens from which RNA sequences were available, and only two had complete genomes, in London and in the US. There was no genomewide information on the early phase of the pandemic, before the multiple waves in 1918.”
Information from after 1919 is particularly sparse. “We are completely in the dark because we have no sequences at all from the 1920s. We can’t pinpoint any changes that would correspond to after the pandemic. Having any genome sequences, especially when we can document new locations and periods, can add to our knowledge,” Calvignac-Spencer added.
Luck stepped in. The researchers found additional genomes, from German flu victims, in a pathology collection that included the medical archive that the “father of modern pathology” Rudolf Virchow started in the mid-1800s.
“We were crazy lucky to find a handful of specimens in the museum just around the corner. We’d been in touch with 60 pathology collections worldwide and could only locate one or two additional specimens. We were working with a number of other labs to collectively reach a critical mass that could ask more questions and answer with more robustness,” Calvignac-Spencer said.
Added Thorsten Wolff, “Over the past century lots of specimens have been collected, but they are terribly hard to work with. When Sébastien contacted me and said we found relics in our backyard, I was excited and interested to look at the genomes.”
The new material came from 13 lung specimens collected between 1901 and 1931 in Germany and Austria, including 6 from 1918 and 1919. Using new technology to analyze preserved RNA, the team sequenced two partial genomes stored in Berlin in June 1918 and a complete genome from a sample in Munich from November of that year. The genetic diversity of the samples from the two cities, about 363 miles apart, suggests a combination of local transmission of flu as well as long-distance dispersal events, the researchers conclude.
Genetic Changes Small and Large Propelled the 1918 Pandemic
Influenza viruses change in two ways: large “shifts” that reassort surface molecules and smaller “drifts” of single-RNA-base mutations. Until now, new flu strains were thought to arise only through shift.
Specifically, a shift shuffles the variants of two types of glycoproteins festooning the viral surface: HA for hemagglutinin and NA for neuraminidase. The 16 types of HA and 9 types of NA recombine into different strains.
The abbreviations provide familiar nametags. The 1918 pandemic flu and the swine flu of 2009 were “H1N1” while the bird flu of 2004 was “H5N1.” The genetic connection is that RNA encodes the HA and NA protein parts.
Wolff delved into the details. “We compared the new Berlin sequence to ones characterized in the US before. We found 51 nucleotide changes corresponding to 17 amino acid changes in viral proteins. We were surprised to find that the virus surface glycoprotein hemagglutinin was very much conserved from season to season, but 9 of the 17 amino acids in the enzyme (polymerase) that the virus uses to replicate changed.”
The researchers also looked beyond the virus’s surface, at the gene that encodes nucleoprotein, which wraps around the viral genome and protects the delicate RNA from scissor-like enzymes. Mutations in the nucleoprotein gene could have shielded the viral RNA from our immune defenses as the pathogens coursed through human populations unscathed. Looking at the nucleoprotein and polymerase genes, in addition to the conventional HA and NA genes, painted a more complete portrait of the influenza virus.
Furthermore, using mutation rates as a “molecular clock,” the researchers deduced that all genome parts of today’s seasonal H1N1 flu could have directly descended from the initial 1918 pandemic strain through a series of single RNA base mutations. The prevailing idea has long been that seasonal flu arose solely through the large-scale reassortment of their HA and NA surface features, the shifts.
Unfortunately, even the more granular RNA sequence information doesn’t reveal what happened during the pandemic’s aftermath, largely due to the dearth of data. “We know from reconstitution of 1918 flu viruses described in 2005 that this was a very virulent virus compared to later seasonal flu. It was different and whether there was a gradual decline in virulence or an on-and-off effect we cannot tell, due to the lack of data for the 15 years after 1918. Was decreased mortality due to increase of immunity in the population, or to decrease in the virulence of the virus?” said Wolff.
The question of declining virulence or increasing immunity, or both, is one that many are now asking about where the COVID pandemic is headed. But direct comparisons won’t fly, the researchers emphasized.
Flu Versus COVID: No Comparison
The pathogens behind the 1918 flu pandemic and COVID are about as different as a hippo from a hedgehog. The viruses differ in several ways.
“Influenza and coronaviruses come from 2 different families. They are not related at all. Flu viruses found after 1918 are descendants; SARS-CoV-2 is new. Before the COVID pandemic we had fought other coronaviruses in seasonal circulation,” said Wolff.
The choreography of change differs for the pathogens too. For flu viruses, the study suggests, one variant gradually becomes another. In contrast, the COVID pandemic continues to unfold as a series of waves with one variant or subvariant replacing another. “In contrast to what we see with COVID, where different waves go hand-in-hand with new variants, 1918 was probably not like that. The new data do not suggest replacement between the waves,” explained Calvignac-Spencer.
He outlined other distinctions. “These are two very different pandemics, with different viruses and conditions for the virus to spread. And human populations are not connected in the same way and the environments are different. One-to-one comparisons cannot help us. Influenza viruses and SARS-CoV-2 don’t share anything fundamental in their underlying biology.”
The two pandemics share, of course, their scale. And the two types of viruses also share bouts of accelerated evolution. In flu, that led to the dying down of the virulent pandemic strain in 1918 to its tamer seasonal descendants. SARS-CoV-2 also evolves quickly, as tracking their genome sequences daily shows.
A statement at the end of the new article provides a compelling example of the fallacy of the idea of scientific “proof”:
”A major limitation of this research is that we only have limited samples so we remain humble and consider all our results provisional.”
All scientific papers do this, qualify the findings. I’ve been reading technical papers for decades, and currently edit a medical journal. That’s how science is done. But journalists who rely on news releases, perhaps sucked in by a buzzword like “breakthrough” that would rarely, if ever, appear in a research report, without reading the original papers, can pass along the misconception that a scientific finding is the final word.
Science just doesn’t work that way. That’s why the phrase “settled science” is an oxymoron.
So when new SARS-CoV-2 subvariants arise, or a monoclonal antibody treatment no longer works, or normally healthy people become vulnerable to a new infectious disease, it’s not that science or scientists have failed. It’s that the very nature of science is still so widely misunderstood.