Synthetic pandemics

"On reading," by Simon Wain-Hobson, is a weekly discussion of scientific papers and news articles around gain of function research in virology.

Since January 2024, Dr. Wain-Hobson has written weekly essays for Biosafety Now discussing risky research in virology. You can read his entire series here.

Sancerre, early morning

On reading Emerging technology and risk analysis. Synthetic pandemics by Daniel Gerstein, Bianca Espinosa and Erin Leidy. Feb 15, 2024. From the Homeland Security operational analysis center, operated by the RAND Corporation under contract with DHS.

---

Let’s see what this recent Federally commissioned report has to give.

The authors …assessed whether a pathogen could be developed applying “engineering principles to biology” that would be capable of sustained, human-to-human transmission and causing mortality and morbidity. And, if so, how would such an engineered pandemic compare with a naturally occurring spillover event? … our primary focus was on a bioterrorist attack scenario.

Just to be clear, there is no established set of rules. Nobody has ever made a virus from scratch. Fouchier and Kawaoka started out with existing bird flu viruses.

First, virology has had its synthetic pandemic – 1977 Russian flu pandemic, that may well have arisen by poor attenuation of a vaccine strain. Everything’s on the Wikipedia page. However, there’s no mention of it here. Some would argue COVID-19 was #2. Neither are synthetic. Second, by definition, a pandemic is hundreds of millions of infections more important than a spillover event. To year end 2024 there have been 972 human H5N1 and 1568 H7N9 flu spillover infections in humans from birds, mainly chickens.

The six key findings are prosaic to say the least. Take the first, Biotechnology will continue to become more readily available, more capable, easier to use, and less expensive. Gosh!

Or the second: Ribonucleic acid (RNA) viruses… - both naturally occurring and those used to cause a synthetic pandemic - are likelier to infect new host species because of their shorter generation times and their faster evolutionary rates, allowing them to survive and adapt to wider environmental and host-related conditions than deoxyribonucleic acid (DNA) viruses can.

Nonsense. The authors mention bubonic plague in the first paragraph that killed Between 75 and 200 million people died in a few years’ time…” That’s caused by the bacterium Yersinia pestis which has a DNA genome. It has a doubling time in the lung of less than 2 hours. By contrast, one of the fastest growing mammalian RNA viruses around is vesicular stomatitis virus of cows. In the lab you can detect new virus by 5-6 hours after infection.

Yes, RNA viruses have much faster mutation rates compared to DNA viruses. That said, DNA genomes are far more plastic and pick up DNA and play with it. So many examples can be conjured up to make this key finding unhelpful.

By the way, some DNA viruses are hugely successful infecting more than half the globe. Just think of the chickenpox virus, herpesviruses, papillomaviruses or the asymptomatic circoviruses. SARS-CoV-2 is in that macabre league which is open to both RNA and DNA viruses.

The fifth key finding is Furthermore, virus evolution over time would create uncertainties in the effectiveness of the synthetic pandemic pathogen. In other words, the fact that a pathogen has been modified does not mean that it will be more effective than a naturally occurring virus in causing a pandemic. Certainly, most virologists would agree that the Fouchier and Kawaoka engineered flu viruses are way more dangerous for humans that the spillover parental H5N1 viruses.

The point about these experiments is that the H5N1 bird flu viruses were selected to transmit efficiently by the aerosol route, something the natural virus has great difficulty with. That was the intent. Downplaying their capacity, which hopefully will never be put to the test, is silly.

The uncertainties surrounding the transmissibility and virulence of a synthetic pandemic pathogen make the risks of synthetic and naturally occurring pandemics similar. Apart from the unmentioned Russian flu and perhaps the COVID-19 virus where’s the data? The outcomes of human flu pandemics range from less than 1 million dead for the 2009 H1N1 pandemic to around 50 million for Spanish flu. Such a huge range in outcomes sucks out any remaining oxygen in the sentence. What were they trying to say apart from filling space?

They like to show their acquisition of jargon for just below we learn that For this analysis, we did not consider a specific virus but rather a pathogen category: a Baltimore class III virus, which would be a single-stranded RNA (ssRNA) virus that supports human-to-human respiratory transmission - a similar virus to the one that causes COVID-19.

Baltimore class III viruses have double stranded RNA genomes. Those with single stranded genomes like the COVID virus belong to class IV. We’re not told why they selected this category of virus. No doubt all readers had the pandemic in mind, so it was relatable. Do note that influenza viruses belong to the Baltimore class V group and have sparked 4 natural pandemics in the last 100 years.

So why single out ssRNA viruses and mention the cell cytoplasm? ALSO, the astute lay science reader, emailed to point out a line in the essay on the UK governments response to COVID - The UK prepared for the wrong pandemic, meaning they placed their bets on an influenza pandemic and were foxed by a coronavirus. The UK COVID report shows the need to plan broadly, yet the RAND authors seem to be going the other way. Methinks they were impressed by COVID.

The present authors dare push aside flu! Utter madness. This alone disqualifies the report.

Then we run into Finally, viruses with the greatest pandemic potential have been shown in studies to be likelier to replicate in the cytoplasm of a cell, which is the case for ssRNA viruses. This is in contrast with DNA viruses, which tend to have a nuclear replication cycle.

It is a fact that most, but not all RNA viruses replicate in the cytoplasm. Strikingly, influenza viruses replicate in the nucleus. This is a corollary and has nothing to do with pandemic potential.

Next up, Regarding developing a chimera, in Biohazard—coauthored by Ken Alibek, former deputy of the Soviet Biopreparat BW program—the authors discuss attempting to develop a chimera of vaccinia and Venezuelan equine encephalitis or of smallpox and Ebola. We assessed that such manipulations to a pathogen are complex and do not necessarily ensure that the chimera virus will have the intended functionality.

A little data first. Although vaccinia and smallpox viruses are DNA viruses and replicate within the cytoplasm, they do so in closed worlds of their own making. Venezuelan equine encephalitis virus (VEEV) is a mosquito-borne pathogen that replicates in the cytoplasm. Use of the word chimera here would indicate cloning a gene from VEEV or Ebola into a poxvirus genome. It's the last sentence that is problematic – it’s easy, not complex, to clone pretty much any genes into vaccinia virus.

As far back as 1988 virologists made a vaccinia VEE chimera and showed they could efficiently vaccinate mice against VEEV. So intended functionality is vague and unhelpful. Do they mean they can morph vaccinia virus into one that causes encephalitis or might be mosquito borne which would be nonsense? You wonder if they understood the simpler concepts. Wasted words.

On reading’s attention started to wane but was roused by their fixation on horsepox. In 2017 a Canadian group succeeded in the complete synthesis of horsepox virus from the genome sequence. This created a stir at the time even though the virus is believed to no longer exist in nature, nor is it known to harm humans nor is it seen as an agricultural threat.

The work wasn’t even a tour de force; by 2008 the 580,000 bp genome of Mycoplasma genitalium had been synthesized. In 2014 the journal Science reported the synthesis of a functional 272,871–base pair designer eukaryotic chromosome, synIII, based on the natural chromosome III of baker’s yeast. The horsepox work was undertaken because of the word pox. It made headlines because of the word pox.

We also expect that the policy, legal, ethical, and regulatory impediments will not present significant obstacles for a determined bioterrorist. Since when has a terrorist bothered about ethical impediments? By definition, a terrorist couldn’t care a damn.

Finding ways to compress the amount of time between identifying a pathogen and developing licensed medical countermeasures, such as vaccines, antibiotics or antivirals, and therapies and protocols, can also greatly affect outcomes. In this regard, time is one of the most important

factors in understanding the consequences of an attack. Everybody knew that.

Alternatively, we assessed, failure to be prepared or act quickly can create the conditions for rapid disease spread and direr outcomes. They assessed a truism.

In the section Our Emerging Technology Risk Assessment we read that The most significant limitation is currently in understanding of the genotype–phenotype relationships of a pathogen. Engineering a synthetic pathogen, even beginning with a natural pathogen in the form of an ssRNA virus, requires understanding these relationships. No and no.

First, dangerous work was done on microbes before the Biological Weapons Convention in Geneva was set up in 1972, which was itself before the first genome ever was elucidated – that of a humble RNA virus of the bacterium E. coli called MS2. The Belgian group of Walter Fiers published the sequence of the 3569 building blocks in 1976.

Second, nobody has ever designed a virus from scratch. It’s far easier to take a known pathogen and engineer it or pull out a mutant virus with altered properties. Neither Kawaoka nor Fouchier fretted about genotype-phenotype relationships. They selected for viruses that had the property they wanted, airborne transmission between ferrets, and once successful went back and looked at the mutations.

This is why scientists do experiments. We learn a huge amount form them, but it takes time. Take a Baltimore class IV ssRNA virus like hepatitis C virus. There is still much we don’t know about this major human pathogen even though it was discovered 36 years ago.

Next up, The challenge would be synthetically engineering a pandemic with a replication-competent, contagious pathogen… As a result, predicting an outcome with great certainty would be challenging. Throughout the text they delight in telling us about the difficulties in genotype-phenotype relationships and the error rate in nucleic acid synthesis, making the outcome of synthetic pathogens, whatever they are, uncertain.

Let’s presume the bioterrorist wants to get hold of a pathogen. There are many nasty bacteria around that can be easily found and grown using methods available on the internet. If they want a nasty virus, by far the simplest way is to download the Fouchier and Kawaoka papers on ferret transmissible H5N1 flu virus and synthetize the genome. Or that of the resurrected Spanish flu virus, curtesy of the US government.

The synthesis wouldn’t be difficult and as the result is known, no worries about genotype-phenotype relationships. These RAND authors have simply forgotten that playing the copycat is the easiest solution. The authors’ reasoning was that the bioterrorist would design one on their own.

This is why the US NSABB voted first not to allow publication of the Fouchier and Kawaoka papers with the novel mutations. They were forced into a U-turn after strong pressure from the NIH top brass who decided unilaterally that GOF research was a risk worth taking. An authoritarian decision that didn’t stand scrutiny (Chilled virology).

This RAND report brings us back to the problems of looking into the exciting world of science. It is hard for all. On reading served on the Scientific Council at the Pasteur Institute for 8 years. Some dossiers in parasitology or bacteriology were hard to appreciate. With experience the good and bad dossiers are easy to distinguish. It’s the ‘maybe’ dossiers that are tough. They require far most time and discussion.

The problem lies with those who read these ultra processed foods, sorry reports. As reports go up the report chain, readers rely on the executive summary, wording and tone to make decisions. Accordingly, if the food is prepared by people who don’t master cuisine and confuse garlic with celery, then what hope have the readers? It is a long-standing problem of specialization.

It's akin to the Peter principle. Information rises to the point where it befuddles those handling it. Scientists are the first to be confused, but it’s our job to do experiments are to sort things out, which we mostly succeed at doing, although it can take time. And yes, On reading is acutely aware that we do not have an HIV vaccine after getting into AIDS research in 1984.

When deciders get ultra processed food, the perception of risks and benefits become a problem. We saw this at Boeing which was in excellent shape until the cost cutting financiers absorbed after the takeover of McDonnell-Douglas started compromising safety. Or the case of the so-called ‘London whale’ which resulted in a trading loss of $6 billion for JPMorgan Chase.

The authors are playing the specialist consultant while discovering the subject and are clearly enamored with some of their personal discoveries - which is fine - but not useful to the reader. The report could have been commissioned from those in the know, but to take it on when you don’t know the field, well there’s a problem. You’d never go to a kidney specialist for an ear complaint.

We’ve seen this lack of modesty before (Off target) or (Some housekeeping). Now ditto with staff from a big think tank. What is disturbing is, given the science training of the present authors, they pontificate.

We need some moderating boron rods of decent science sense and philosophy amidst the often-uncontrolled fission reactions released by dangerous GOF research. Fortunately, the new NIH Director is thinking differently.

Conclusions

The report has nothing to offer. Its subtitle could have been ‘The sound of words.’ It was wasted money that could have induced the sponsor, the Department of Homeland Security, into making poor choices which is dangerous.

---

Aside 1

Viruses adapt to their hosts. Rapidly replicating bacteria that can divide in as little as 15 minutes. Accordingly, their viruses multiply as rapidly if not faster. As an example, take phage lambda which was in the late 1950s the best studied organism around. It has a DNA genome of some 50,000 building blocks, larger than any RNA viral genome and it multiplies faster than its bacterial host. The notion that DNA virus replication is intrinsically slower compared to that of RNA viruses is nonsense.

---

Aside 2

A review article has just emerged from Gerald L Epstein at the RAND corporation entitled The evolution of United States governance policies for research using pathogens with enhanced pandemic potential.

It provides useful timelines and background information to those who may need this. A reliable email correspondent described it thus: the usual dull think tank sort of sort. It is now history given President Trump’s Executive Order stopping dangerous GOF research in the US at the federal level.

---

Simon Wain-Hobson is an emeritus professor at the Institute Pasteur, Paris, from which he retired in 2021. He and his colleagues were the first to sequence the genome of HIV, and Wain-Hobson has published more than 230 papers on virology and cancer.