Find the latest COVID-19 news and guidance in Medscape's Coronavirus Resource Center.
This transcript has been edited for clarity.
Eric J. Topol, MD: Hello. This is Eric Topol for Medscape One-on-One. I'm really delighted to have the chance to have a conversation with Professor Eddie Holmes, one of the world's great evolutionary virologists, from the University of Sydney. Welcome, Eddie.
Edward C. Holmes, FRS, FAA: Thank you. It's a real pleasure to be here.
Topol: We have plenty to talk about, obviously, with SARS-CoV-2 being the only thing going on in the world. How are things in Australia right now?
Holmes: Good. We are an isolated island, which has made things somewhat easier. Australia, back in late January, instigated a flight and travel ban to China. That actually helped, and it delayed the importation of the virus until late February.
Then we had an outbreak. The first wave came in March and ended in May, and then a second wave came in Melbourne in July. I live in Sydney, and we didn't really get the second wave. The second wave was quite bad in Melbourne. They had quite a serious lockdown, but they've had no cases there for 50-plus days.
In Sydney, just to put it in perspective — and I think your listeners may be shocked by this — we've had maybe a bit over 4500 cases in total in Sydney, New South Wales, for the entire pandemic and our mortality number is around 60 cases. Many of those people have been hotel quarantined.
If you close your borders and you do clamp down hard, you really can control this. At the moment, we've got a small outbreak — hopefully — going on in northern Sydney. We had something like eight cases yesterday, December 22. Again, with quarantine, you close your borders, but occasionally things get through quarantine. This appears to have gotten in, and hopefully they will control that. It's a very, very different picture here than what you're seeing in many countries.
Topol: Are you able to proceed during the pandemic?
Holmes: Yes. We were in lockdown in March and April. My university reopened in May, so we've been open and working since May. We had basically no cases for weeks in New South Wales, and just recently we had a small outbreak. The suburb where it's at, they're in lockdown, but the rest of Sydney is not. There are some restrictions, so it's not pre–COVID-19, but it's certainly a very different landscape here than you'd see in many countries.
Topol: We're envious, as you can imagine.
Before we get into what happened in January, you've been studying different RNA viruses throughout your career. Can you give a bit of background about what was going on before this?
Holmes: I've been doing this for 30 years now. I started off, like many people did, on HIV because that was the virus that introduced us all to emerging viruses. I was living in Edinburgh and I was working on the outbreak in Edinburgh that was largely associated with injection drug use. That was back in the early 1990s.
Then I moved to Oxford and I started doing more comparative evolutionary virology, looking at how different viruses evolve, what the common principles are that underlie virus evolution and emergence. In the past decade or so, I've mainly been doing metagenomics to understand what shapes the virosphere, how big is the virosphere, and how do things move between virus species.
The way I do that is by doing metagenomic sequencing. Normally, it's taking animals and looking at what viruses they have; it's looking at whole ecosystems and what moves between them. All of that work is done in collaboration with Professor Zhang, whom we'll talk about shortly. He was in Beijing; now he's in Shanghai. Together, we did metagenomics sequencing. It was that technique, in fact, that they used to sequence SARS-CoV-2.
Topol: That's a really important point you're getting at, which is that next-generation sequencing — where you use a sequence-everything approach and then figure out what it all is — is very informative.
Holmes: It's incredible. Just in terms of pattern discovery, think back to HIV. It took 2 years from the first description of AIDS until we found the causative virus. That was done using classic virologic techniques. SARS-CoV-1 took a few weeks. And as we'll discuss, it took 40 hours from the sequencing samples arriving in Zhang's lab to getting the viral sequence for SARS-CoV-2.
There are many failings in our approach to emerging diseases, but pathogen identification is not one of them. It's something we can do. The genomics in this outbreak has been absolutely astonishing. We have over 250,000 genomes sequenced, and we may well get a million by the end of the outbreak. That bit has been good.
Topol: I think it's changed the world in terms of the reliance on genomics to understand spatiotemporality, what's going on in an outbreak, and so much more.
The first documented case of COVID-19 was in December, and the samples arrived to Professor Zhang's Shanghai lab on January 3. He, as you said, had this sequenced 40 hours later.
He was submitting this sequence to Nature, which was great. You and he have collaborated quite a bit over the years, and you suggested that it would be good for him to put it out on the web somewhere so everybody could start working on it.
Holmes: Yes. What happened was Zhang had seven samples arrive in his lab on January 3. They were all associated with this famous seafood market in Wuhan, from workers there or people who live right next door. They were all lung wash samples; we actually had a lung wash study going on in Wuhan at the time.
Of the seven, one of the lung wash samples was positive on January 3; he sequenced it on January 5. On that same day — this is actually very important — he sent it to GenBank. More important, he told the ministry of health in China that it was a novel coronavirus. It was clearly very closely related to the first SARS virus. In fact, between us, we were calling it SARS; it was kind of obvious at the time. It looked so close. Because of that and because the first SARS coronavirus was respiratory, we thought this was going to be a respiratory virus. We said this was likely to be respiratory and that people should take precautions. That was made apparent very early on.
Remember, this is the first week of January, so it's a very different picture than where we are now. At that point, we had 20-odd cases. You couldn't at that point say that this is going to be a global pandemic that's going to completely affect our lives. It wasn't like that in the first week of January. We thought it was a really interesting outbreak. The case numbers were not that high at that point, so I thought it would be great to put this out there. Most people wouldn't even notice what was going on.
I was encouraging them to do that, but — and I want to be really careful what I say here, because there was some pressure in China not to publish anything at all — I think the Chinese authorities wanted to keep a lid on what was going on. Obviously, the West is a bit different, but Zhang, in China, was under some pressure not to release too much data.
It was on January 8 that the Wall Street Journal, of all places, published an article saying it's a novel coronavirus. It was clear that they had heard from other scientists in China. There were other groups sequencing, so there were many people working on this. Zhang was not the only one or even the first one, but it turned out that we first made it open access.
We submitted a paper on January 7 to Nature, and they were very keen for us to release it. Wall Street Journal announced it on January 8. On January 9, I think the Chinese authorities confirmed that it was a novel coronavirus. At that point, it just seemed to me quite ridiculous because everyone knew this was a coronavirus, but we're not saying what it is.
Of course, as the week was going on, the cases were slowly increasing and it became more and more of an issue. Then, social media people were discussing that it's a coronavirus and we need to know what it is. Jeremy Farrar, director of the Wellcome Trust, certainly was very vocal in that.
I was trying to encourage Zhang to release the sequence, but again, he was under some constraints. It got to the point where I thought, We have to get this out, because it was just crazy. I called him very early Saturday morning my time, even earlier his time. He was on a plane. He was about to fly between Shanghai and Beijing — I can't remember which direction he was going, but he was strapped in his seat.
I said, "Zhang, we have to release the sequence." He said, "Give me a minute to think about it," and he said, "Okay, do it." Luckily, my good friend and colleague Andrew Rambaut, from Edinburgh, was still awake — it was about 1 AM his time. He runs this website called virological.org, which has been a very nice open-access forum. We had some frantic emails, and then we wrote some text for the release, and then I posted the sequence on that website at lunchtime on Saturday here and posted it simultaneously on Twitter. That was the open-access moment.
Topol: It's going to go down in history as one of the most famous tweets ever.
Holmes: I'm not very good at Twitter, Eric.
Topol: It's actually pretty remarkable. In fact, in Nature, Professor Zhang is one of the top 10 of 2020. The story that you just recounted is in there, where you were on the phone with him and the flight attendant was telling him to get off the phone and he was telling you to go ahead and put out the sequence. It's really historic because obviously every minute counts when you're trying to move forward.
The key here is that it went to the National Institutes of Health (NIH) and Moderna very quickly. In fact, what's interesting is to hear how it was just a matter of hours before they started rolling from that sequence to the vaccine. Did you envision that was the next step, the vaccine?
Holmes: No; that's happened so, so quickly. I thought the first thing people would do is use it to design polymerase chain reaction (PCR) tests. I thought that was the obvious thing. And that's what Christian Drosten and Marion Koopmans' people did in Europe, which is very good. The mRNA vaccine, I didn't see that coming. That is a complete game changer, and it's so gratifying.
I can also say that it was actually quite tense in all this, sending the sequencing and editing things. I was under such pressure to get it out that I didn't even check what it was before I posted it. I got the sequence file and I thought, I'm going to put it up. After I posted, I thought I should probably check that it really is a coronavirus. I quickly did a bit of blasting and it really was the coronavirus. I thought, Oh God, I could have put completely the wrong thing up and gone down in history as one of the stupidest tweets ever. Luckily, it was the coronavirus.
Topol: It basically accelerated the whole process. I think the fact that we now have two mRNA-based vaccines out, based on that sequence, is quite remarkable.
I don't want to get into the details, although it's been written about that there certainly was pressure on Professor Zhang with respect to the rectification order and some lab issues. There's obviously his side of the story and whatnot, but it wasn't easy. It was great for moving the field forward, but he may have suffered in part because of his open-science approach.
Holmes: I need to be cautious here, because people's careers are involved in this. Yes, there was pressure, and there were some consequences for what we did. I don't regret it, and I know Zhang doesn't regret for a microsecond what we did. I'd do it again in a heartbeat. It was a moral, scientific imperative for us to do it.
I think, as you said, because of the amazing, rapid development of the vaccines based on that sequence, it shows the absolute power and necessity and beauty of open science. I think that's a good thing. I hope that once our present political issues change — and I hope they'll change in January — that the authorities in China will realize how important open science really is.
The word "SARS" is quite a difficult thing in China. I think they really didn't want to panic people and they wanted to control the message. That was their goal, but it probably stretched too far. I think that's the issue.
Topol: In that Nature recognition of Professor Zhang, he is quoted as saying, "It was a turning point for China. It was a turning point for the world." Clearly, Eddie, your role in this was very important because you helped get it out there, and we're indebted to you.
Holmes: I think I'm just the messenger. I've worked in China for many years. I understand a little bit how China works. They have a lot of national pride. You just have to be wary of the way you do things. I am the only Western scientist involved in many of those early papers because I spent time going to China and working.
When you go to China, you're working with scientists; you're not working with the state. It's just like you and I are talking now. You're with these people, and it's just a normal scientific interaction. They just happen to be Chinese. You're in China. That's the only difference. Everything else works the same way.
I think the critical thing is — this is what worries me most of all — that if the politics gets in the way of sharing the science, we will be in a much worse place. The key thing now is that we have to try to prevent this from ever happening again. Step number one has to be immediate, rapid, open data sharing.
You've seen how speed is of the essence when it comes to a pandemic, and any barrier that gets in the way of us sharing that data, because there are political consequences of that, is going to make this world a much less safe place. We have to work with these people, we have to share data as openly as possible, and that should be the lesson you get from this outbreak.
Topol: That is so important to reinforce. As you say, it wasn't only to develop a vaccine, which you could hardly envision could move with this unprecedented velocity, but also diagnostic testing. There was so much at stake.
Fast-forward: Now we have these two vaccines based on that knowledge of the sequence, with the two consecutive proline substitutions to help stabilize and to get the maximum antibody production. That was a nice tweak, which is historic, too.
Holmes: It was beautifully done.
Topol: It's kind of amazing. Andrew Ward, who's a colleague of mine here at Scripps, and many others participated in that.
A New Problem: Virus Variants
Topol: Now we have a new problem, not that we've gotten this virus squashed in any respect outside of countries like yours, continents like yours. This variant in the United Kingdom — not necessarily born in the United Kingdom — but this B117 variant has cropped up in recent weeks. It's led to a big surge in spread, not only in parts of the United Kingdom; I understand that there's at least one case in Australia, the Netherlands, Italy, and Denmark. It's gotten around a bit. Obviously, there may be other countries that aren't doing sequencing that may have it there. In the United States, we haven't seen that yet or the South African variant, N501Y, which seems to have some thread as far as an important spike mutation. [Editor's note: A patient with the B117 variant was identified in the United States after this interview was recorded.]
What's your sense about this? Where are we headed now?
Holmes: It's obviously the question of the moment. I'm not directly involved in the UK works. I'm seeing it kind of secondhand. From the body of data that I've seen, there are definitely concerns here. I think there are a number of things that are acting as the links in the chain.
We start with the epidemiology. As you've mentioned, this variant does appear to be growing very rapidly in the United Kingdom. That's not just because the south of England had fewer restrictions, because if that was the case, then all the variants would increase in frequency. It's a particular variant, one lineage, that's increasing compared with any others. It's growing quite, quite quickly. The same appears to be true in South Africa. It's a different lineage, but it has at least one of the same mutations. That's a worry.
The second is they're reporting now that the virus has evidence of a higher viral load of infection, measured by lower-on-average cycle threshold (Ct) values and more sequence reads on sequencing. That's saying there's more virus, which would explain the faster growth rate.
If you boil down to the actual biology, the virus, as you mentioned, has this mutation in amino acid 501 in the spike protein in the receptor binding domain, which is one that we'd already flagged as being a really key site. Other labs have shown that this mutation is critical for receptor binding. That's the same mutation that's come in South Africa as well. That would be the molecular explanation for the higher viral load, which then explains the growth rates. All those things move together.
I think what they felt they had to do in the United Kingdom, because they've had a very difficult epidemic, was to act sooner rather than later. Even though not all of the I's have been dotted and the T's crossed on the biology of this virus, it's important to act now to stop it spreading. That was what made the UK government do what it did.
Obviously, we're still waiting a little bit on complete functional characterization. We want to know exactly what's going on. I think the interesting thing about South Africa and the United Kingdom is that it's not just this one mutation — the amino acid 501 change; it's that those lineages have a huge number of changes compared with other ones, which is really fascinating.
There's one suggestion that makes a lot of sense, which is that they evolved in somebody who may be immunocompromised, because you have chronic infection for a much longer time in a single individual. You have a partial kind of immune response there. Maybe that's allowed this virus to evolve and select in an unusual way. It's a very unique set of selective pressures on the virus in a patient like that that's led to these changes. That's not proven by any means at all, but it's an interesting theory.
Topol: There have been a few cases of immunocompromised patients in which they showed rapid evolution, so your point there is well taken.
The other question, Eddie, is that in the United States, convalescent plasma is being used highly without the evidence in hundreds of thousands of people. Could that also lead to more evolution of the virus?
Holmes: I think it depends on the number of people being used relative to the proportion of the population. Selection is kind of a numbers game. There needs to be selection pressure for the virus to evolve in that way. If most people in the population are using convalescent plasma, then you see a selection pressure. If it's only a small proportion, I don't think it would select that much.
This case is different, though, because it's in a single individual. A single patient has a chronic infection with a lot of viral replication. So you're replicating a whole population's worth of evolution in a single patient. That's why it's so unusual. As you mentioned, in some other SARS-CoV-2 cases within immunocompromised hosts, you see mutations, as you also do in norovirus infections and influenza virus infections. I've seen it in my past work. It's certainly a very interesting theory.
This current variant, we need to monitor it closely. We need to see more on the basic functional biology. We need to see how the vaccines will cope with this. That is a key question. That work is being done at the moment in the United Kingdom, I think, and fingers crossed that the vaccines are still going to be tip-top. If they are, we can breathe a small sigh of relief, I think.
A Need for Booster Shots in the Future
Topol: Let's say the virus variant is not a problem for this vaccine, but it shows you that while the virus was so slow in evolving in terms of any meaningful functional variant except for this D614G, that became the dominant one. Now we have a different look at this virus.
Does that make you think that even if the current vaccines hold up well, we're going to be looking at booster shots adjusted to the continued evolution of the virus for the years ahead?
Holmes: That's a great question. We need to think about evolution in different phases. During the first phase, basically most of this year, what you've had is a virus spreading in a population where there's no immunity. That's been the key marker of this outbreak, really, because everyone is susceptible. In those circumstances, there's no immune selection pressure on the virus whatsoever. Any lineage can find a susceptible host to infect; it's actually really easy. The virus spreads, and it just infects people. It's kind of a free-for-all.
As immunity rises in the population, hopefully by vaccination — although some countries, such as the United States and the United Kingdom, are really trying hard to do this without vaccination — as immunity rises in the population, that's going to change the selective landscape.
You will see that the virus will evolve away from that. I think that's an absolute certainty. Now, it does evolve a little bit more slowly than some RNA viruses — maybe three times more slowly than the influenza virus. It's not an abnormally low rate of evolution; it's actually pretty average, but lower than flu.
My guess is that as immunity rises in the population, hopefully by vaccination, you will start to see immune escape gradually. That will happen. That's an inevitable consequence of natural selection. It's been played out for millennia, and it's going to happen again. We will very likely need to update these vaccines at some point. That may take 2 years or 5 years or 1 year; I don't know.
To me, it's a racing certainty that immune selection pressure is going to push the virus in a certain way. You'll probably start to see more direct evolution than you have done in the past, because now it's harder for the virus to find a susceptible host because people are immune. Only the fittest strain is going to make it through, and that fitness is going to depend on a particular antigenic configuration.
Rising immunity will completely change selection pressure. I think it'll become even more seasonal, too. Early on, the virus didn't need to be seasonal because everyone was certainly susceptible to any lineage at any time. As immunity rises and susceptible persons become fewer in the population, the right kinds of conditions for spread become more important. The virus will change in its behavior because of rising immunity.
Topol: This is a central point you're making, in that the race toward population-level herd immunity, vaccine-induced, is countered by the virus evolving. We're not seeing the end of this virus just because you get 80% of the world vaccinated. This is an endemic story, would you say?
Holmes: I would put money on this being an endemic respiratory virus. Absolutely. Even if we rolled out the best vaccine coverage program ever, we're not going to vaccinate everybody. We can't do it simultaneously. The virus will evolve fast enough to keep itself going, and they'll reenter the susceptible class. I think it's endemic. Absolutely.
Topol: Well, that's exciting, isn't it? It's kind of interesting to speak to a leading evolutionary virologist because you get a different perspective about it. This is what you spend your whole life on, and you really understand the context of what we're looking at right now.
Holmes: It's also thinking comparatively. Talking about evolution, I work on many different viruses, and I see the patterns between them. SARS-CoV-2, in a comparative way, is not mysterious. It's not a magical virus. It has the standard properties that respiratory viruses do, and it's subject to the same rules of epidemiology and evolution. They work pretty well.
I can't predict what mutations will appear in what order or at what time, but I think I can make a pretty strong prediction that it is going to evolve and is going to escape immunity like everything always does. I think that's a pretty safe prediction.
Topol: Would you say it's fortunate that the spike protein gave us this ability to get a potent vaccine? Would you have predicted, for example, this 95% efficacy?
Holmes: No. I was optimistic that we would get a vaccine. I was thinking more around 60%, maybe 70% if we were lucky. Certainly not 95%, which is absolutely spectacular. And more than once — multiple vaccines can do the same thing. Some months ago, I did a little exercise with the Wellcome Trust, which was very interesting, on horizon scanning, playing out what the future might be like in 5-10 years' time.
We've looked at vaccination, immunity, and antivirals. In our vaccine horizon scanning future predictions, a vaccine of this efficacy was our absolute best-case scenario. Our middle-case, most likely scenario was much worse. I think with the vaccinations, we're in a very good place, with incredible speed. These people did an amazing job.
Topol: Well, you helped them — you and Professor Zhang.
Monitoring the Fault Lines
Topol: How do we get smarter for the next pandemic? We're going to have another one. Your friends, these viruses, are going to haunt us in the future. How do we avoid having the toll here of harm, of deaths, and long COVID? How can we be smarter?
Holmes: I think there are three things we can do, each with increasing difficulty. The simplest thing we try to do is to somehow distance ourselves more from the animal world. There are clearly practices that we do today, such as live animal markets, the wildlife trade, not zoning (we build on these wildlife areas where we're exposed) — all those things increase our proximity to wildlife that carry viruses, some of which can infect us. We need to be much smarter in how we regulate our exposure to the natural world. That's a relatively easy thing to do, just to regulate those practices more.
Second, we need much better global surveillance. By that, I think the people who work at the human/animal-level interface are the sentinels. They're the canaries in the coal mine because they're going to get exposed more than anyone else. Those sorts of people maybe will need regular virologic screening, something like VirScan, which is a cool technique. I've heard that there is a global observatory looking at blood samples globally; maybe metagenomics should be performed occasionally of people who work in abattoirs or live animal markets on a regular basis.
They are the front line and are like the fault line. I like to think of it as an earthquake analogy. They are where the tremors take place, so they need to be monitored really closely. Those data have to be shared absolutely freely and as quickly as possible globally. There shouldn't be local governments holding onto it, saying, "We're handling it ourselves." That's a barrier to permanent prevention. We need surveillance of the frontline people at the human-animal interface and data sharing.
Finally — and this is really difficult, like an Apollo project— we need to have stockpiled in our freezers broad-acting antivirals and potentially vaccines that can recognize a whole span of coronaviruses or influenza viruses. I'm not into prediction, but I think it's pretty obvious that there is a set of viruses that are particularly jumpy and that are likely to emerge in the future.
I would say the top three are coronaviruses — this is number five in the past 20 years in humans — so it's coronaviruses, influenza viruses, and paramyxovirus that seem to be the most likely to emerge. For those three, are there ways — this is a really big science project — that we can develop antivirals that can recognize several of these, or vaccines that can recognize multiples, and have those ready rather than having to wait? Even a year is really quick, but it's time. Rather than having to wait for that, we have them there that we can roll out. That requires a massive investment in basic science, with many smart people working on it.
Topol: That's really helpful to kind of get a sense of what lurks ahead, particularly your ranking of the virus families that need special attention. I love the concept of broad preparation with antibodies and structure-based vaccines that have that broad capability.
Holmes: Even now, I think the coronaviruses we know, if you look at the evolution history of coronavirus, you can see that some lineages appear — like in the beta coronaviruses that jump most often — I think we know what they are. I think we can certainly start to plan around the likely ones. If they have any structural features in common that we can now utilize, I think we can start on that now.
Topol: Excellent point. I have to tell you, Eddie, this has been a fascinating discussion. We hadn't met before, but to get your sense of the world is just invaluable. Having seen the historic tweet and some of the story about what all happened back in January, as well as some of your extraordinary work in the past, it's really a privilege to have this conversation with you. Any parting words of wisdom?
Holmes: It's a pleasure to talk to you, because the way you convey the message on Twitter for people to understand is absolutely invaluable. I think that's been a huge thing. That does lead me to one of the things that I have noticed and that you're part of as well: In regard to social media, its power in the pandemic is absolutely amazing because it's so rapid, so immediate. You can get your message out extremely quickly. Unfortunately, sometimes that's led to confusion, but normally it's been a phenomenal way of rapidly passing on what needs to be known.
It's more efficient than the other standard channels that we've built since World War II to convey information about pandemics. It really is. I think in the future, another thing we need to do is to enhance those sorts of social media things because they are so direct and so rapid. Because of the pace of pandemics, that has to be the way. We can't wait for these official committees to meet and have everyone sign off. As valuable as they are, sometimes it's going to be quick. Social media is just fantastic, and that has been an absolute game changer too, I think.
Topol: Well, there's no question about the open science, as well as having that ability to get the word out through Twitter. One of those things, just to mention, is that of all the different parts of life sciences or medicine that I'm familiar with, the genomics community has really led the charge to be open like this.
You have done this, of course, throughout your career, and now we saw how it paid off, because it could have taken a lot longer to get where we are in terms of a remedy. Thank you for that. Thanks for the chance to visit with you. I look forward to following you closely and checking in with you in the times ahead as we deal with this endemic mess.
Holmes: My absolute pleasure. Thank you so much.
Eric J. Topol, MD, is one of the top 10 most cited researchers in medicine and frequently writes about technology in healthcare, including in his latest book, Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again.
Edward C. Holmes, PhD, is an evolutionary biologist and virologist. Since 2012, he has been a professor at the University of Sydney and a National Health and Medical Research Council Australia Fellow. He also has had an appointment as a guest professor at the Chinese Center for Disease Control and Prevention, Beijing, China, since 2014.
Follow Medscape on Facebook, Twitter, Instagram, and YouTube
