< Return to NEB Podcast Home
Transcript
Interviewers: Lydia Morrison, Marketing Communications Writer & Podcast Host, New England Biolabs, Inc.
Interviewee: Joshua Quick, UKRI Future Leaders Fellow, University of Birmingham.
Lydia Morrison:
Welcome to the COVID-19 Researcher Spotlight series. Today I interview
Dr. Josh Quick of the
University of Birmingham, who is integrally involved with the
ARTIC Network, a collaborative effort to understand the genomics of viral outbreaks.
Lydia Morrison:
Josh's experience began with
Ebola and
Zika viral outbreaks. And in January, the ARTIC Network released a protocol for coronavirus sequencing, which has been widely shared around the globe, enabling worldwide genetic sequencing of the COVID-19 outbreak to help deliver epidemiological insights.
Lydia Morrison:
Hi Josh. Thanks so much for joining us today.
Josh Quick:
Oh, it's a pleasure to be here.
Lydia Morrison:
Could you tell our listeners what the ARTIC Network is?
Josh Quick:
Essentially what the ARTIC Network is a
Wellcome-funded Collaborative Award between Birmingham,
Edinburgh, and
Cambridge Universities in the UK. It was awarded to
Nick Loman,
Andrew Rambaut, and
Ian Goodfellow, who began collaborating during the 2015 West African Ebola outbreak and decided to put the proposal together to build the network.
Josh Quick:
Ian established an Ion Torrent sequencing laboratory in Sierra Leone during the West African outbreak, and Nick Loman and myself did the first
Nanopore surveillance sequencing in Guinea. And Andrew played a big part orchestrating all the data that came from those sequencing sites and performing phylogenetics analysis. That was really the genesis of the network.
Josh Quick:
So there's a large group of people, and as a group we're working to develop tools for sequencing viral genomes during outbreaks to generate real-time epidemiological data. We focus mainly on Nanopore sequencing because that's really something to base the research around, because we're really interested in the idea of portability and generating real-time results, because this idea of a very rapid sample-to-answer time and doing sequencing on-site, close to the samples, is really the basis for all this kind of work.
Lydia Morrison:
Why is it important to sequence virus samples during an epidemic?
Josh Quick:
It's important to sequence genomes because there's a lot of useful information contained within genome sequences from viral isolates, and that's because as viruses replicate they introduce errors in their genomes, and those are detected by the genome sequencers as mutations. Those mutations can be used to reconstruct the epidemiology of the outbreak through a field called phylogenetics, which is essentially building trees showing the relatedness of isolates. And that can be combined with metadata, such as the sampling date, to produce a time-calibrated tree as well, which you might have seen. And that can be used to predict the start of an outbreak, the most recent common ancestor. It can also be used to model the evolutionary rate. So there's a lot of useful information, in terms of the phylogenetics.
Josh Quick:
There are a lot of side benefits as well. So you can use genome sequences to identify any signs of adaptation to hosts, treatments, or vaccines, which will also appear as mutations in the genome.
Josh Quick:
But the main reason that we do this is to identify circulating lineages within an outbreak. So for example, during the 2015 Ebola outbreak, I flew to Guinea with a mobile lab and set it up, and within days we were able to show that there were actually two distinct lineages circulating Guinea at that time, in April, 2015. And that wasn't known at the time. And that's because there was a limited amount of genome sequencing available from that region.
Josh Quick:
What we're really trying to do in the network is generate actionable data in real time, which can deliver epidemiological insights. But also we're trying to generate information which is useful to local health authorities, so that they can make interventions, which can hopefully positively influence the course of an epidemic or an outbreak.
Josh Quick:
To give another example, the epidemiological findings that we made during the Zika virus outbreak were able to show that ... So to give a bit of background to that project, we had a mobile lab, which was actually on a bus, doing a thousand mile road trip through northeast Brazil. And we were collecting samples in public health laboratories along the way. But we were able to determine, by identifying the most recent common ancestor using phylogenetics approaches, that the actual virus had been circulating in Brazil for over a year before it was detected. And that was because symptoms of the Zika virus infection were quite similar to other endemic viruses in the region, like chikungunya and dengue.
Josh Quick:
So those are a couple of examples in projects that we've been involved in, in the past. The current project that we're working on is called the
COVID-19 Genomics UK Consortium, or COG-UK for short. And there have been a number of important findings from genome sequencing from this project so far. In the UK, we've generated the largest number of genome sequences of any country so far, of SARS-CoV-2 genomes: nearly 30,000 from the UK. And what we were able to do using that information and flight and travel information was ... one of the important findings from the consortium so far is that there was over 1,350 separate introductions into the UK during March. And that those were mainly from our close European neighbors: France, Spain, and Italy. So it might be different to what people have been expecting, in terms of how the virus came to the UK.
Josh Quick:
For example, was there a sort of patient zero that brought the virus into the UK, and then there was onward spread? It wasn't like that at all. There was massive numbers of separate introductions, and we've been able to show that locking down two weeks earlier probably could have prevented about half of the deaths in the UK.
Josh Quick:
The other thing that we're doing is monitoring the lineages that are still circulating now, even though we're on the downward epidemic trend. We're monitoring the lineages that are circulating, and many of those are now going extinct, which we can determine from the sequencing results from the current samples.
Lydia Morrison:
Why is it so important to do the sequencing close to the source of the viral outbreak?
Josh Quick:
Well, there are lots of good reasons to do that, but one of the main advantages of doing that is speed really, because we're trying to focus on real-time prospective sequencing. If you can take the lab to the samples, you don't have any of the delays associated with shipping the samples to other labs. The turnaround time that you can achieve in-country can be as little as 24 hours from sample collection to a genome sequence being available. And that just wouldn't be possible if you had to ship the sample abroad, or to another lab elsewhere in the country. So it really makes a lot of sense to do sequencing at the point of need.
Josh Quick:
And I think that one of the reasons that this is not more popular is that people worry about not having that sort of lab infrastructure available, but we've been developing what we call a Lab-In-A-Suitcase since our first outbreak response, and it's developed and improved over time, but it includes sort of portable isolation cabinets for keeping amplicon contamination to an absolute minimum. And even though it's a portable system, which is actually built from repurposed hydroponics tents, it's very effective at keeping contamination at bay in a portable form factor. So there's no really serious drawbacks to doing sequencing in-country, especially if you have a Lab-In-A-Suitcase available.
Lydia Morrison:
Yeah. It seems like you've had a lot of experience in sequencing viral genomes. Could you tell our listeners how you got started in that?
Josh Quick:
Yeah, so you have to go back quite a long way to the 2015 West African Ebola outbreak, which was when-
Lydia Morrison:
Josh goes on to say that they began a long shot project to perform real-time sequencing of Ebola in Guinea, by using the minION to create the Lab-In-A-Suitcase, which was moved several times during the study, demonstrating the advantages of a mobile lab.
Lydia Morrison:
Shortly after, the Zika outbreak in central America necessitated a change to the singleplex RT-PCR and pooling protocol that had been employed for Ebola. They decided to use multiplex PCR coupled with native barcoding to achieve high throughput sequencing. But the sequence wasn't as successful, which led Josh to create his own primer design program.
Josh Quick:
Which I called
PrimalScheme, which is a pun based on a Scottish band, Primal Scream. I still develop that now with another developer, Andrew Smith. But it was inspired by AmpliSeq workflow, which is a commercial multiplex PCR panel. And I was convinced that the success lay in the design of the primers themselves, and that there wasn't actually any fancy chemistry going on. The one notable thing about it was that it had an unusual two-step PCR program, which included a 15 minute 65 degree annealing extension step.
Josh Quick:
So I went to the lab, started to fiddle with a toy multiplex panel, testing different conditions and primer designs. And that was when I started using NEB
Q5 polymerase because it allowed for higher Tm primers to be used. It has a DNA-binding domain, which allows very high annealing temperatures to be used, and that was ideal, I thought, for the application, because I wanted a very, very, highly specific PCR with very little off-target amplification. And I felt that one of the ways to do that would be to use high annealing temperature primers.
Josh Quick:
So yes, but a number of the things that we still use in the protocol today, were locked down during that time, for example, the two-step cDNA synthesis using random hexamer priming, the Q5 polymerase, which I mentioned, the primer concentration, which I determined in the lab to be the best, and the primer design tool itself. So we were able to put all those things together and go back to Brazil, another two times actually, and successfully sequence a number of genomes from Brazil. And luckily it worked well enough to complete the project.
Josh Quick:
Another thing that we did during that time was to develop the one-pot ligation protocol, which was based on the Ultra II Illumina Library prep workflow from NEB, which we've been developing and supporting for a long time, because it makes such a big difference to the hands-on time when you use that in the lab, which is quite important when we're working in the field, because it obviously directly impacts the number of samples that we can process. And we still use that now in a yet improved fashion.
Josh Quick:
So during the COVID 19 outbreak, development on the protocols really started up again, and we've just finished writing up a paper on the work. We made a number of improvements to the protocol to improve the performance and streamline it. The double barcoding rate has now been improved to 70% using this one-pot barcode ligation protocol, through tweaks to the conditions ... that's a really, really high efficiency of ligation that we're managing to get now. And that's while reducing incorrect assignments down to 0.001%. So very, very pure binning using this protocol.
Josh Quick:
Another thing that we were able to do was to remove the post-PCR cleanup after, and replace that with simply a one in 10 dilution, which removes a lot of hands-on time and a lot of manual bead cleanups. So we're now able, with this protocol, to go all the way from cDNA synthesis to barcoding without any cleanups.
Lydia Morrison:
Yeah. Can you walk us through sort of the
basic ARTIC protocol?
Josh Quick:
I briefly alluded to the fact that it's a multiplex PCR followed by barcoding and sequencing, but to go into it in more detail: you start with extracted RNA, and then we do a cDNA synthesis using the
LunaScript SuperMix from NEB. We were originally using SuperScript, but we've moved to LunaScript because if you factor in the cost of the RNAs out and the random primers, it's actually cheaper to use the SuperMix. It also comes as a 5X Master Mix, which means you can get more RNA into the reaction. In addition, it doesn't require a separate denaturation step, which saves hands-on time in the lab. It also contains a dye which makes it easier to use and makes errors less likely to happen.
Josh Quick:
So following that we have a multiplex PCR using the ARTIC v3 primer panel. So we use 10% of the PCR reaction in cDNA. We use Q5 polymerase for the multiplex PCR, as I talked about earlier, which allows us to use very high annealing temperature primers, and works really well with this two-step PCR thermocycling program we use. And we do two reactions per sample that we're sequencing. And those are effectively, the alternating regions, amplicons which tiled down the genome, we put into two different pools. And that's that's because you can't do them in the same pool because they overlap, which allows us to determine the sequence all the way along. But if you do that in one reaction, then you would produce a short overlap product preferentially. So then after we've done those PCR reactions, we pool them together to generate effectively pools covering the whole genome in one tube, and to dilute them one in 10.
Josh Quick:
And what we discovered during this work is that the PCR acts as a normalization, no matter what the input; if you do 35 cycles of PCR, the PCR plateaus at about a hundred nanograms per microliter, post-PCR concentration. And that allows you to effectively crudely normalize the samples from one to the other, which means that we can take a fixed volume of that one in 10 diluted amplicon pools and we go directly into the end prep using the Ultra II end preparation kit without any cleanups. This was what I was talking about earlier, replacing the cleanup there with a one in 10 dilution.
Josh Quick:
And we found that end preparation is compatible with that dilution of PCR reactions. That's a really helpful time-saving there. And it's been very popular with lots of groups that are doing this kind of sequencing because it saves so much time.
Josh Quick:
So really now, by taking that cleanup out, we've gone from RNA all the way through cDNA, PCR, and preparation barcode ligation with no cleanups, just dilution. So that's really an impressive workflow really, of sequential enzymatic reactions all working together.
Josh Quick:
Another thing that we always have done is instead of using a cleanup after the barcoding reaction, we use a heat denaturation to destroy the ligation activity before pooling them together, just so we can do a single clean up there. And that's something we've done since the days of Zika.
Josh Quick:
So then we adapt those barcoded pools by ligating on the Nanopore adapters, and then localize to the flow cell. And James Hadfield from the ARTIC network has developed an application-specific web tool called
RAMPART, and it's designed to monitor these kinds of amplicon runs, so you can see the amplicon coverage by barcode in real time on the laptop you're using to run the sequencing, either the
minION or the
GridION, so that's very useful.
Lydia Morrison:
So it sounds like a really streamlined workflow. What countries are currently using this protocol?
Josh Quick:
Well, it's actually hard to keep track, but I have shipped out over 130 packages of primers to people that have got in contact and requested them, over huge variety of countries, 30 or 40 different countries. So we were able to distribute these primers just by sending out packages via DHL to lots of different other countries.
Josh Quick:
And I think that's helped people to get up and running quite quickly, because all you really need to generate these amplicon pools is Q5 polymerase and the primer pools, and you can try that on your cDNA to see if you can amplify it. And then once you've got those amplicon pools, you can sequence them using any next generation sequencing technology that you like. We generate 400 base pair amplicons by default, so you can add the sequence by using the 2X250 on Illumina or on Nanopore sequencing.
Lydia Morrison:
So I just wanted to ask: as you were looking at the genetics of each of these individual viral strains, and you're seeing these mutations, I imagine some of them are fleeting and some of them are more persistent, and you might see multiple sort of major strains that emerge from this work. I'm curious as to what the implications of those genetic differences in those major strains are? Can you then correlate those genetic differences to things like the rate of contagion, or perhaps the degree to which people become sick from the virus? Could you speak on that a little bit?
Josh Quick:
Yeah. It's a good question. And you rightly point out that many of the mutations which occur are neutral, have no effect phenotypically. And even though we can detect them with genome sequencing, they appear as mutation in the genome. But obviously one of the things you can do by doing this type of surveillance is monitor for these kind of changes, which do have an effect. And within the consortium, it was reported that these two genotypes ... they're called 614D and 614G for the amino acid changes ... had differences in transmission rate and pathogenicity.
Josh Quick:
And one of the things we were able to do was to investigate that by looking at the prevalence of those two genotypes within the UK dataset, and show that actually there is signal there showing that they do have a slightly different epidemic growth rates. The other way of confirming that finding is in a lab looking at that, but it's also possible to observe that using the national scale genome sequencing data.
Lydia Morrison:
Thanks so much for joining me today, Josh.
Josh Quick:
It's been great to talk to you. Thanks for inviting me.
Lydia Morrison:
Thanks for joining us for this episode of the COVID-19 Researcher Spotlight Series. Join us next time when I interview three scientists who are members of the
Chan-Zuckerberg Biohub in the San Francisco Bay area. Tune in to hear how Eric Chow, Amy Kistler, and Emily Crawford are working together to bring reliable testing and genomic sequencing to Californians.