Read on Twitter
Excited to share our latest, out in @cellhostmicrobe. Great work by @alvalesano. This was a real team effort with major contributions from @mt2fus @famulare_mike many colleagues from @UVA_ID and @icddr_b Funding from @gatesfoundation @BWFUND. A thread 1/x https://www.cell.com/cell-host-microbe/fulltext/S1931-3128(20)30574-6#.X7VVhgTUTYw.twitter
I've worked on poliovirus and evolution for years. More recently, we've done work on how viruses evolve within and between hosts using samples from individuals in clinical trials and observational studies. So this paper has been a real personal and scientific journey. 2/x
When @alvalesano give talks, sometime he's asked the question, "Polio, is that still a thing?" Yes, it is very much still a thing. Polio is a horrible disease and it's still around. We are tantalizingly close to global eradication, but the final steps have been hard. Why? 3/x
Many issues, but the one of them has to do with the vaccines. The Salk vaccine is inactivated (IPV). It was the first and the one still used in much of the world. Of course, need to mention that trial of Salk vaccine was performed at UM by Thomas Francis. 1.8 millions kids. 4/x
Results announced in 1955 at the Rackham Auditorium on campus. Here's @mt2fus and @alvalesano by the plaque.
IPV is good at preventing disease, likely by stimulating IgG in the blood. In contrast, the Sabin vaccine (oral poliovirus vaccine, OPV) is a live, attenuated vaccine. It is given as drops orally. The virus then replicates in the gut (like wild polioviruses do). 6/x
OPV has many advantages. It is stable (cold chain), it doesn't require needles, and it probably stimulates more mucosal immunity. It's highly effective and can prevent transmission. It's been the workhorse of the eradication campaign. There's a problem though. 7/x
Because it replicates, is shed in stool, and transmits, it also evolves. If it establishes chains of transmission in poorly immunized populations, it can evolve for weeks to months. Eventually, it can cause polio outbreaks, indistinguishable from ones caused by wild viruses 8/x
To get a sense of the issue...wild type 2 polioviruses have essentially been eradicated. The only type 2 polioviruses we still see are vaccine derived polioviruses VDPV, evolved from OPV. See e.g. this paper from @VaccineEpi https://www.nejm.org/doi/10.1056/NEJMoa1716677
Lots of work to understand this process. To date, much of it has been done using viruses recovered from VDPV outbreaks. One can then infer the steps the virus took in the months leading up. h/t to @SternLab and @raulandino16. https://www.sciencedirect.com/science/article/pii/S0092867417302921?via%3Dihub
But what are the very first steps that OPV takes on its way to becoming a VDPV? What is being selected for? @mt2fus and @famulare_mike ran an OPV trial which gave us a unique opportunity to look at this. 11/x https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(17)30358-4/fulltext
This was a trial of OPV2 in Matlab, Bangladesh. Again, many thanks to the team @icddr_b and all the participants. Study enrolled hundreds of kids. They were given OPV2 and then the vaccinated individuals and their household contacts were sampled weekly. 12/x
@alvalesano @itswillfitz then worked with @mt2fus and her team to sequence the OPV from these stool samples. We spent a lot of time optimizing our protocols to do deep sequencing of these samples - stool is harder than nasal swabs. Got pretty good data! 13/x
First thing we found was that the three gatekeeper mutations identified by @SternLab are rapidly selected. We were able to actually measure their rate of fixation over the first few weeks following vaccination. 14/x
We then looked across individuals to see if mutations were shared, a trick we learned from @ksxue and @LouiseHMoncla. It was powerful since everyone received the same virus in the trial. We found lots of parallelism, indicating that many mutations were repeatedly selected. 15/x
This much positive selection, this quickly, is pretty uncommon in acute viral infections (in our experience). Probably because wild viruses are already pretty good at what they do, but attenuated vaccines have lots of room to improve. 16/x
What do these mutations do? We found lots across the genome. Many were in the capsid. Some are in antigenic sites. We do not think they are being selected to escape vaccines. Rather, we suspect they make the virus better at replicating within hosts. 17/x
Well, if OPV evolves so quickly why does it take so long for there to be a VDPV outbreak? We think transmission is a big barrier. To get at this, we looked at viruses from vaccinated individuals and their household contacts (to whom they transmitted). 18/x
We estimated a tight transmission bottleneck that restricts the genetic diversity that is transmitted from one host to the next. Less data than our flu work, but it seems pretty tight. 19/x
We then did a secondary analysis, predicting how frequently you would see these mutations in the general population given the bottleneck, the rate of increase of the mutations, and the probability of transmission by week. It actually matched data on mutational prevalence. 20/x
So. OPV2 can evolve pretty quickly within hosts. We think there's selection for within host replication, but (limited data alert) a transmission bottleneck limits their spread. We don't see selection for transmission, but were underpowered to do so. 21/x
Some impacts. Understanding which mutations arise repeatedly and when has implications for poliovirus surveillance. Sewage sampling is in vogue now, but has been used in polio surveillance for years. Looking for these mutations may identify bad VDPV earlier. 22/x
Our work may also inform the development of novel OPV strains. These recently received WHO EUA. The goal of these viruses is to limit the reversion of OPV to VDPV - keeping the good parts of OPV while limiting OPV evolution. 23/x
This was a fun and interesting project. Can't thank @mt2fus and @famulare_mike enough. We also had a bit of scientific exchange. Here is @alvalesano with Kajol Islam in Bangladesh. Kajol also visited my lab to learn some of our approaches. 24/x
