Somewhat exasperated by the Covid-19 models publicly showcased over the last two months, I decided to ring up virologist Luis Villarreal for his perspective. Luis Villarreal is founding director of University of California, Irvine’s Center for Virus Research and UCI Professor Emeritus, Molecular Biology and Biochemistry.
Villarreal told me he takes these models “with a grain of salt” and that he is “leery of too much mathematical application to biological behavior.” He said further that the models “just assume there is no context or history dependence,” when coronavirus and bat-host—probable origin of the current pandemic—share a long evolutionary relationship.
Villarreal is optimistic that by year’s end we could have a safe, effective Covid-19 vaccine if protocols are adjusted for the promising “new technologies that are much less prone to bad outcomes.”
Our conversation follows.
Suzan Mazur: There’s been considerable misinformation circulating about Covid-19, including a recent alert posing as one from Johns Hopkins Hospital, which describes the virus as “a protein molecule (DNA) covered by a protective layer of lipid (fat),” saying further it is “not a living organism.” Would you comment?
Luis Villarreal: It’s clear to me that quote has an error in it. First of all, I can’t imagine the people at Johns Hopkins saying an RNA virus like coronavirus has a DNA genome. It is true that it does have protein surrounding the nucleic acid and a lipid around it, like pretty much all RNA viruses. And it’s true that once it is in its transmission extracellular form, it is just a biological entity with none of the characteristics of life. However, it’s also true and equally true that once joined with the cell, the virus starts to program that cell to do what the virus wants or needs and is part of the living system of the host itself. So it’s both dead and alive depending on the circumstance. That the virus is not alive is a conservative, old view and not based on our current understanding of an enormous influence in all living entities due to virus activity and virus colonization.
Suzan Mazur: In a previous conversation you told me the following about viruses:
“What I’m asserting is the fundamental importance of the ability of entities to persist not just to replicate. For me the persistence is a big deal. The persistence of Ebola in a bat, for instance, is not an accident. It’s a deep evolutionary relationship associated with the origin of bats.”
Ditto for Covid-19?
Luis Villarreal: Yes, very much so. In fact, this is the can of worms now being argued: the origin of this coronavirus. People have pointed to a lab in China as the origin. But that lab actually has been studying bats and viruses with respect to emergence events. It’s been trying to determine whether this virus is a natural phenomenon like all the other emergent coronaviruses resulting in epidemics over the last 50 years, where viruses mainly arise from persistence in sources, i.e., virus and host have been in apparent long-term relationships.
In this case, this family of coronaviruses has numerous relationships with bats in China and that part of the world. It’s also true for different families of coronaviruses and bats in other parts of the world as well. This long-term relationship—phylogenetic incongruence—means bat and virus frequently co-evolve in the same environment.
Suzan Mazur: And billions of other viruses have found a host in every human—though not usually harmful?
Luis Villarreal: The fact that a virus persists doesn’t mean it is not capable of harm. Consider the context of coronavirus. A very well studied coronavirus is mouse hepatitis virus.
If this hepatitis virus gets introduced by a persistently infected mouse to a colony that has a history of infection, the virus will then move harmlessly from parent to offspring and the offspring will host that virus throughout their lives and transmit it to their offspring. However, if the persistently infected mouse is introduced to a colony that’s never been infected with the virus, that colony will be decimated by virus interaction with a naïve, virgin host. Introducing a persistently infected mouse into a virgin mouse colony pretty much destroys the capacity of that colony to reproduce. So it is consequential. It’s damaging to those who are not infected but not to those persistently infected.
Suzan Mazur: Thank you. In our previous conversation you also said:
“A virus will spit out entities that regulate themselves and are subfunctional.”
Would you comment on the quasispecies aspect of viruses and what it means for the current pandemic?
[Note: Villarreal’s PhD advisor was John Holland, one of the principal scientists to experimentally evaluate quasispecies. Villarreal did postdoctoral research in virology with Nobel laureate Paul Berg at Stanford University.]
Luis Villarreal: Many virologists still mistakenly believe that a pure RNA virus population exists when it’s actually been established for some years now—both theoretically and experimentally—that RNA viruses don’t exist as pure populations but as heterogenous populations known as quasispecies.
Some hold the view that there is a consensus sequence within the virus, which they term the master fittest type. However, experiments contradict this view and indicate that the quasispecies has within it components that oppose replication of the virus itself. It has members not participating in the replication of the infectious virus. That’s pretty much true for all the RNA quasispecies characterized.
The coronavirus we’re talking about is a little narrower in makeup than most RNA viruses. It has polymerase that has an added feature. It has what we call an error and repair function, which limits the ability of the genome to some degree but not absolutely. That is why when this virus is passed to any individual and that individual transmits it to any other individual, the virus takes on a sequence character that’s unique. It’s not necessarily mutation. It’s a variant. It means the coronavirus is introducing variations into its sequence all the time, in this case, to bats in different habitats, different geographies, different populations. It will acquire a different consensus just by the drift of the quasispecies.
That’s why we can map where this virus has been. We can tell where it came from. The virus that infected New York City came from Europe not directly from China—because of this quasispecies.
These aren’t mutants in the sense that mutants have a phenotype associated with them. This is where my analysis differs from others who do consider them mutants, i.e., with a different phenotype. These are variants for the most part with no measurable phenotype.
So when people are arguing that a new or transmissible version of coronavirus has emerged because there’s been a variation—there’s been no measurement of that. Unless there’s evidence, the virus is just a natural variant. I’m referring to the recent study at Los Alamos for which there was no experimental evidence that a phenotype change had occurred.
The character of these viruses is dynamic. They are always changing and they have within them the capacity to limit their own propagation and replication as well.
Suzan Mazur: You’ve also said that viruses move in a non-linear path. Bogdan Dragnea, in fact, has been trying to harness this movement of viruses for use in laser light for surgery. Would you say the non-linear path of the virus is being factored into the Covid-19 models being presented to the public?
Luis Villarreal: I take these models with a grain of salt. I’m leery of too much mathematical application to biological behavior.
One of the very curious features of trying to model and think about the progression of these entities is that they have this fundamentally stochastic and conditional quality to them. How they behave and what they are capable of to a large part depends on the context and history of each particular entity and its host.
This is easy to demonstrate. For example, if we had all been infected with a variant human coronavirus that had a spike protein that cross-reacted with the current Covid-19 spike protein, then the human population would be essentially immune to this coronavirus and the consequence of being infected would be completely different. Therefore, you can’t predict biological behavior outside of the context that you’re talking about at a particular time with a particular host. This gives the virus a very stochastic character that essentially approaches unpredictability, particularly by any mathematical model.
For example, try to imagine how we could have modeled the emergence events that happened in South Korea where you had a large infection curve due to behaviors associated with a particular religious group, which allowed the virus to first get a foothold in one city and then drive infection in the country to some degree until it was mitigated. Those kinds of things may or may not happen. They’re completely stochastic. Whether they do or don’t occur doesn’t really lend itself to statistical analysis.
Think about the very same thing regarding the initial emergence of the coronavirus from its bat host. There’s a good chance the virus wasn’t fully in sync with human infection when it emerged from its bat host, that it did not have the disease capacity it has now acquired. So what are the probabilities that that virus is going to make its way from a bat to a human to initiate the pandemic that we currently have? That’s not really a predictable event. We can say the possibility exists but we can’t put a number on it.
I would add that this coronavirus is a 10,000-letter instruction from the epigenome of a bat trying to become part of the epigenome of a human. And this instruction happens to be very effective and consequential because of the context in which it has occurred. That context being humans have not been exposed to a virus similar to this coronavirus and we don’t have an endogenous virus that’s similar enough to interfere or affect the outcome of the infection. So that’s a historical circumstance. And if you try to model without taking that into account, you can’t because there is no history. That’s my main complaint about modeling. It just assumes there is no context or history dependence.
Suzan Mazur: How soon do you anticipate that we will have an effective Covid-19 vaccine considering that the virus keeps changing with every infection?
Luis Villarreal: Yes. All RNA viruses have quasispecies and they all drift and change to some degree. But this does not prevent us from making effective vaccines for the majority of them. A few represent more challenging circumstances like influenza, which has a segmented RNA genome and it’s picking up RNA pieces from other sources and shuffling its surface proteins with some regularity. That doesn’t apply to this family of viruses. This is a large single-stranded RNA virus—about 30 kilobases of RNA—and the spike protein. This spike protein is the main target for a vaccine. The virus RNA is drifting, like I said, undergoing variation but not mutating. So I don’t see Covid-19 vaccine development being fundamentally any different from that for a vaccine to combat measles, mumps or other RNA viruses we’ve tackled in the past.
Suzan Mazur: What timeline do you see?
Luis Villarreal: Three months is a theoretical possibility for a vaccine but it would require a different protocol and procedure for evaluating safety. There’s a whole array of issues here. I’ve been thinking about this for a very long time, since when I set up the research center at UC-Irvine back in the late 1990s.
One of the main objectives was to anticipate the emergence of pandemic strains of viruses and do something about it. At the time, there were various issues that had to be resolved such as rapid identification and isolation of the agent. Now that can be done in a matter of weeks. And it was done that quickly in the context of this particular coronavirus. So that’s one issue that’s been solved.
Another issue was to generate vaccines that are fundamentally safer than vaccines we’ve traditionally developed for use, vaccines that have historically required long protocols to address safety concerns.
Suzan Mazur: So you think it’s feasible we could have a safe, effective Covid-19 vaccine by year’s end?
Luis Villarreal: Yes. In fact, I see no scientific reason why that couldn’t happen. But there are some things that have to happen that are not allowed to happen right now. For example, Ian Lipkin offered this recommendation that rather than going through the usual slow steps of Phase I, Phase II, Phase III clinical trials, we could accelerate the process by showing the virus ineffective in primate models. You challenge monkeys with live virus and protect them with the vaccine. You show that the virus actually protects against infection. That can be done while the safety trials are occurring.
Suzan Mazur: Can you do those experiments with monkeys in the US?
Luis Villarreal: An Oxford University group, I believe, is proposing to do this in the context of a recombinant vaccine they’re developing. I understand the experiment has already been done, in fact, by Chinese scientists using one of their kill-virus vaccines and that the vaccines were protective in those monkeys.
I suspect all the strategies being pursued now are likely to work. It’s just a matter of what’s going to be safest, quickest and cheapest. That’s where the variation comes in. But I don’t doubt one of these strategies will work.
As for the timescale—for example, we have now an array of nucleic acid vaccines, such as Moderna Therapeutics’ mRNA-based vaccine with an RNA messenger that corresponds with the spike protein. Fundamentally, that approach is much safer than any in the past. I would suggest protocols be adapted to these new technologies that are much less prone to bad outcomes.
The same probably applies to the skin patch concept developed at the University of Pittsburgh School of Medicine. Using a skin patch [PittCoVacc] to present antigens to the immune system, resulting in a very powerful reaction without introducing antigens directly into the bloodstream and risking potential problems.
What I’m saying is there are promising methods that appear a lot safer than those we used to use. Protocols haven’t been adapted to take advantage of this increase in safety.