Taken together – animals need vaccine development, children around the world cannot get access to vaccines, and – to be plain – idiots in the US are ignoring an effective vaccine to measles, allowing for outbreaks of the disease.
GlaxoSmithKline have announced that 300 doses of an experimental Ebola vaccine are on their way to Monrovia, Liberia. The rapidly developed vaccine has gone through phase one trials in 200 healthy volunteers across the globe to test its safety. Having passed this test it will now enter phase two trials in Liberia to test whether the vaccine is actually effective at preventing Ebola infection and disease.
How will this work?
To test vaccines, healthy people are either given the vaccine or a dummy shot. Those two groups are then followed over the course of months to record the number of individuals that go on to catch the virus or develop the disease. If fewer people in the vaccinated group catch Ebola virus than the dummy group, then we can conclude that it works. Obviously, nobody wants anyone to get Ebola in the first place (this is why we’re trying to develop vaccines, after all), but it’s necessary to tell if the vaccine is an effective tool against the virus.1
As a result, the success of the experiment is dependent upon a significant proportion of the studied people coming into contact with the virus. If nobody catches Ebola, you can’t tell whether the vaccine works. You then have nothing to aid those at extremely high risk of disease, such as family members of the already sick, or medical workers.
By all means, these vaccines cannot come too soon. If they are effective, we will have a powerful weapon against a virus that will strike again in the future.2 And clearly, the decision to send these doses to Liberia is months in the planning, it didn’t get decided overnight.
But is sending the vaccine to Liberia actually going to tell us whether the vaccine works?
A senior health minister from the Liberian government has today announced that there are just five Ebola cases left in the country. This is fantastic news and a real testament to both the national and international response to the outbreak. While not every Ebola case may be officially diagnosed in a country, the fact that new cases have been reduced to single digit numbers suggests a virus on the way out.
Taking the CDC’s cumulative case numbers (.csv download) (which can be viewed without download here) and plotting the number of cases per month, it’s quite clear where the Ebola outbreak is headed:
Ebola cases per month (CDC figures). These data are complete up to January 21st 2015.
The virus is outta here, with cases dropping in every country (note, the January data are up until the 21st, not the end of the month). Most importantly, we can see that Sierra Leone is really the final bastion of the disease. If we want to test the efficacy of this GlaxoSmithKline vaccine, and be potentially3 ready to contain future outbreaks with it, surely it would be reasonable to retarget our efforts to Sierra Leone? While the logistics of this aren’t trivial, should we be sending a trial vaccine to a country where the chance of catching the virus is rapidly plummeting to zero? We may as well send the country thousands of sterile syringes, they’d be more useful.
To be clear, the studied group are not then purposefully exposed to the virus to test the vaccine. Individuals already thought to be at a risk of infection, such as medical workers, are included in the study.↩
The virus is thought to persist in insectivorous bats in the wild. When a virus persists in an animal reservoir, the only way to effectively eliminate it is to vaccinate or kill enough of that species. Both are extremely difficult and undesirable courses of action. With the virus persistent in bats, there is always the potential for human infection in the future.↩
In the last article, I wrote about a study which looked at the ability for the flu virus to package all 8 segments of its RNA genome. In short: they often don’t package all 8. And if the process of stuffing flu particles with RNA isn’t that picky (hey, important pieces are supposedly missing here), you could think that any one RNA segment is just as likely as another to be left out.
But work published in 2014 provided a neat example of how this process can be far from random, and can be directly influenced by the virus itself.
It all focuses on a mutant of the flu NP (or NucleoProtein). NP smothers flu RNA, keeping it organised for transcription and copying the viral genome, as well as allowing it to be packaged into the virus. Following previous experimental infections with the mouse-adapted influenza A strain ‘PR8’, a single mutation in NP (F346S) was found to increase virus replication in guinea pig noses. Whatever this change in NP was doing, as far as making new viruses goes, it was doing a good job. But when the same lab group looked at infection in cell culture, they found something wrong with the virus: it didn’t make very much NA protein.
NA (or NeurAminidase) is an enzyme that sits on the surface of the flu particle and allows the virus to cut itself free from infected cells (and may also have a role in cleaving through mucus on the way into the host1). By all measures, flu needs NA for infection. Yet, compared to the starting PR8 strain, the mutant virus produced less NA protein, less NA mRNA, and populations of purified virus particles contained fewer copies of the NA RNA segment. In other words, there appeared to be less of the protein because there was a shortage of instructions to make it – and all seemingly because of a single mutation in NP.
Furthermore, the authors show in this work that the mutant flu virus (F346S) produced a greater proportion of “semi-infectious” viruses during replication. Semi-infectious means a flu virus without all 8 RNA segments and incapable of completing an infection cycle all by itself. When equal amounts of fully infectious particles were added to cells, the authors saw that the mutant virus infected 8x as many cells.2This demonstrated that the vast majority of mutant virus particles were semi-infectious. See the following diagram for clarity:
A virus stock is made up of a) fully infectious particles (green circles) and b) semi- or non-infectious viruses (red crosses).Here, both mutant and normal flu samples total 10 particles each, but the proportion of circles and crosses is different for each virus. If you wanted to infect cells with 9 green circles from each stock, you would end up adding 90 total particles from the mutant and just 10 from the normal virus stock. This explains how so many more cells are infected during the mutant infection.
So how can a virus lacking such an important protein replicate better during guinea pig infection? Especially given that so many particles aren’t fully infectious on their own?
Firstly, why isn’t it just worse at replicating? Here, the authors suggest that the virus replicates well in guinea pigs because a large number of cells get infected with multiple viruses. If one virus lacks all 8 segments, then another can help out if they infect the same cell. Here’s this image again from my last article:
Lefty here only has 7 of the 8 pieces of RNA he needs, so this infection is doomed to fail. Righty and friend also only have 7 pieces, but between them they have the full set – they’ve all they need to make more virus.
I think the authors do a cool job of showing how co-infection occurs over time in the guinea pigs. When they took cells washed from infected noses at 9 hours after infection and flow sorted them, they saw that only a small proportion (25–44%) of cells produce both proteins HA and NA. But when they did the same thing at 48 hours after infection, 80–90% of cells contained both. These data suggest that early in infection, cells are generally infected with individual viruses, both fully or semi-infectious. But once more rounds of flu are produced in the guinea pig the majority of cells get infected with multiple viruses. Who cares about carrying all 8 RNA segments when you have a load of mates to help you out?
So while co-infection minimises the detrimental effect of semi-infectious particles, how exactly does this mutant virus replicate better in guinea pigs?
The authors suggest a number of possibilites in the discussion of this work, notably that some unknown beneficial effect of the NP mutation may outweigh the lack of NA, or that in combination with the NP mutation, reduced NA activity could be an advantage. It’s certainly not clear. To finish, I’ll posit a version of the second possibility.
Perhaps, in the absence of NA, the flu particles are simply aggregating together. This idea is supported by work from the Barclay lab at Imperial College London, in a 2013 paper.3 Specifically checking out figure 5, flu virus particles possessing a shorter form of NA aggregated together far more than those with the normal longer length NA. Why does this happen? Because sialic acid (the cell molecule to which the flu virus binds in order to infect cells) gets stuck on to newly formed virus particles and must be cleaved off by the NA protein. NA usually cuts flu free from cells, but it can also sever the tethers between viruses. The short NA is worse at this than the longer form, and you know what, I bet viruses lacking NA altogether are terrible at separating themselves.
On the topmost cell, influenza buds and escapes normally due to the presence of NA (long red lines) on the virus surface, which cleaves the sialic acid molecules on the cell (very long green lines). On the bottom cell, the flu particles are covered in far fewer copies of NA, impairing the break down of sialic acid. The flu HA protein (short black lines) binds sialic acid on both the cell and virus membranes, leading to aggregation of virus particles. If some NA is present, the aggregate could come away from the cell and infect as one mass.
And if you’re a flu particle looking for a friend to go co-infecting with, what better way than to hold hands as you both enter the cell at the same time?
Whilst this hypothesis isn’t complete, as I have no answer as to how NP affects NA or what else it may be doing, I figure less NA on viruses = better virus aggregation and thus more efficient co-infection. Greater co-infection could allow for greater reassortment of novel mutations, and could aid the virus in the battle against intracellular immunity by overwhelming the cell before immune signalling is fully established.
Big clumps of virus sounds ideal for infection, but viruses don’t just ‘worry’ about what to do inside a host. What about getting between hosts?
Note: a virus doesn’t have to be fully infectious to get into a cell. So long as it can enter a cell and produce some virus proteins, that’s enough. The differentiation between semi and fully infectious is just about whether the virus can get in, copy itself AND get out again to repeat it all.↩
Influenza viruses are a common winter scourge. As the air turns cold and we spend more time indoors with other people, flu spreads easily throughout the human population. Like many virus infections, your immune system finally wins the battle and protects you from getting infected a second time. The problem with this is that the flu comes in many flavours. New mutant flu strains that our immune systems cannot recognise begin to successfully infect people, and by the time winter reappears next year, the flu season strikes again. If your job is to develop the flu vaccine, this is a nightmare. Every year the virus is different, and so every year the vaccine must be too. Making the problem worse, producing enough flu vaccine doses for the global population takes most of the year. The result? The strains of flu targeted by each year’s flu vaccine must be decided around the same time as the previous flu season is just ending. This article takes a frank look at how this year’s predictions haven’t lined up with the major flu strain (H3N2) striking this season, and explores how difficult getting the predictions correct can be. Despite our struggle to combat the flu, there is no doubt that the flu vaccine saves lives every year. Getting an annual flu shot is the right way to protect both yourself and your loved ones from this nasty virus.
I missed this episode the first time round, but it came up in a review of 2014 podcast from the TWiV crew. In this podcast (2hrs 19min) from March last year, TWiV met up with Eugene Koonin at the National Center for Biotechnology Information to discuss his work with comparative genomics, the role of viruses in the evolution of all life and what we can understand from virus diversity. It’s fantastic big-picture evolution, and I’ll do no more than tell you that Eugene Koonin and this episode are fascinating to listen to. If video is your jam, you can also find the podcast over at Youtube.
Alongside the medical teams that directly treat the sick, scientific support is critical when controlling outbreaks of infectious disease. The Ebola outbreak in West Africa is no different. During November and December last year, Professor Ian Goodfellow from the Division of Virology at the University of Cambridge volunteered to help establish and work in an Ebola diagnosis lab in Makeni, Sierra Leone. In this article, Ian describes his average day working in Makeni and what it’s like to set up a molecular biology lab in difficult conditions. This article explains what the lab was set up to do and also provides an insight on what it’s like to actually work there. Labs like these are drastically decreasing the time taken to diagnose a patient with Ebola: a factor that speeds up isolation and treatment, ultimately containing the spread of the disease in the community and saving lives.
Fascinating window into the viruses that exist in our food. Some polyomaviruses are carcinogens, including Merkel Cell polyomavirus in humans. Interestingly, the ability to cause tumours may result from the recent introduction of such viruses from a natural host species into a brand new one. Whilst a link to cancer in humans is speculative right now, it’s a link worth investigating.
A great podcast (35min) covering vaccine hesitancy, the damage to international healthcare efforts caused by the CIA campaign to assassinate Osama Bin Laden, more US-centric reasons for avoiding the flu vaccine, how to increase the number of people getting shots and self-interested vs altruistic vaccination.
An interesting reason for avoiding the vaccine is the perception that it’s not necessary. There’s a paradox of public opinion and public action: when the vaccine is working the disease is less of a problem, which in turn reduces the disease’s newsworthiness, which in turn reduces the public perception of flu being a problem. The end result? “Who worries about getting the shot if flu isn’t a problem?”
Vincent Racaniello’s blog post about a cool study looking at the ability of the inactivated polio vaccine (IPV) to boost mucosal immunity in children previously receiving oral polio vaccination (OPV).
IPV is a killed preparation of virus that stops paralytic polio disease, but doesn’t prevent wild polio infecting the human gut. With the aim of eradicating polio, stopping the spread of the virus is crucial.
OPV is great at stopping virus spread but is itself a live virus preparation that can (on very rare occasion) mutate and spread through the human population.
Neither vaccine alone appears to hold the key to eradicating the virus, but this new study shows a third way: follow one with the other.
A pretty cool (pun intended) story’s cropped up about the influence of cold temperatures on the body’s response to Rhinovirus infection – one of the viruses that leads to the common cold.
In brief, the interferon response pathway (which counteracts virus infection inside cells) doesn’t work as well at temperatures colder than 37°C. When the research team at Yale University School of Medicine compared interferon production in mouse cells at 33°C and 37°C, they found far more at the hotter temperature. Conversely, when they looked at Rhinovirus growth, surprise surprise, the virus grew better in the low temperature / low interferon environment.
This is neat, because it suggests that cold temperature may actually increase your susceptibility to getting a cold, just like your family always told you. I say may, because this study was conducted in mouse cells using a mouse-adapted strain of Rhinovirus, but this is intriguing evidence that your nose may be a frosty gateway to virus misery.
This is also interesting given a wonderful talk I listened to by Cardiff University’s Prof. Ron Eccles at the Society of General Microbiology conference in Liverpool last year, in which he discussed the symptoms behind the common cold. Most notable to me was his mention of how the insides of our nostrils reciprocally expand and relax, due to the dilation of large veins in our nose. This expansion becomes exaggerated during infection due to inflammation, which results in a horrible blocked nose (here’s a cool paper – not Open Access, sorry).
Not only was this interesting enough, but he hypothesised that this could be a defence against virus infection. How? The blocked nostril steadily increases in temperature and the poor sucker of a virus has a hard time growing as it finds itself out of its comfort-zone. If this is true, perhaps it also gives the innate immune response the jump-start it needs to get to work.
This holiday season, many of us will have enjoyed the convenience of ordering presents online. Want gift, find gift, buy. Behind the scenes, your gift was found in a warehouse full of other stuff, selected by a person or machine, labelled with a delivery address and transported to the right place (hopefully your door!). But what if you’re a flu virus, hanging out inside a lung cell with only the gift of more flu on your wish list?
Down at the size of molecules, nobody is going to take your order. Influenza has a genome made out of 8 separate segments of RNA, each containing genes essential for building the virus. Imagine one long set of instructions, but chopped into eight pieces. If you have one complete piece of genetic material – like Poliovirus, for example – you have the easy task of sticking exactly one copy of RNA into one empty virus shell before leaving the infected cell. If you’re flu, you need to collect eight pieces: eight different pieces.
Headache. But just how good is flu at stuffing itself with RNA?1
According to an article published in 2013, flu doesn’t care much for finishing the full set. When Brooke and colleagues2 looked at infected cells using fluorescent antibodies to see the flu virus’ proteins, they found many in which at least one protein was missing. Some of these proteins included HA and NA, proteins sitting on the outside of the virus that act as both the key into cells and the knife to cut the virus free – in short, proteins the virus needs to spread between cells and people.
This observation is weird because it suggests that some flu viruses either don’t carry all eight RNA segments (after all, no HA gene segment in the incoming virus = no HA protein in the infected cell), or that individual genes just randomly fail to be read during infection.
To the last point, the authors infected the same type of cells with Vesicular Stomatitis Virus (VSV), which has a genome made from just one piece of RNA. When they looked at the production of two VSV proteins during infection, they couldn’t find cells in which one or the other was missing – failing to make proteins at random doesn’t look like a regular foible of virus infection. The authors are also trying to make the point here that unlike flu, VSV can’t misplace any of its genome, and perhaps this explains the difference between the viruses when it comes to gene expression3. Maybe. But these data are a bit weak for me – they’re different viruses. You wouldn’t infer dog behaviour by studying cats.
Much better evidence comes from flow cytometry experiments analysing many thousands of cells for the presence and absence of individual virus proteins. Not only is this approach higher throughput, but it also allowed the authors to study the effect of increasing the amount of flu used to infect the cells. When flu virus was outnumbered by cells, more than half of all infected cells were shown to have at least one flu protein missing. But by increasing the amount of flu so as to outnumber the cells, the vast majority then contained a full set of virus proteins4. Whilst a large number of individual flu viruses are sloppy enough to forget segments of their genome, when they gang up in the same cell, together they have everything they need.
Lefty here only has 7 of the 8 pieces of RNA he needs, so this infection is doomed to fail. Righty and friend also only have 7 pieces, but between them they have the full set – they’ve all they need to make more virus.
Looking for proteins is cool, but viruses live and die on their ability to transmit themselves to new cells. When the authors looked at flu’s ability to spread to new cells5 they found that this happened just over a tenth of the time. By all accounts, this suggests that influenza is rubbish at its job of making more flu. And yet yearly flu epidemics suggest otherwise.
I think we virologists6 get hung up on a faux “physiological relevance” when we study viruses in a dish. We have the ability to infect cells with billions of virus particles, but as most people don’t get infected by a scientist wielding a pipette, we imagine the individual invaders staking their claim to just a few cells in our body. Therefore, when it comes to scientific thinking, experimenting with low virus concentrations appears to be inherently more “realistic” than high concentrations. This thinking is probably correct when a host is first infected, but once each hijacked cell begins to spew out hundreds of new viruses into the neighbourhood, co-infection of nearby cells with 10’s to 100’s of viruses is probably common. All this is to say, if flu needs more than one virus to infect a cell successfully, no problem.
That aside, what I find most interesting about this work is the idea that influenza infects as a swarm of viruses by necessity. Flu, like many RNA viruses, is renowned for the sloppy reproduction of its genetic code. What we have here is not only a virus that creates a population of mutants in its host, but a virus that appears to be critically dependent on co-infection of cells to successfully complete its life cycle.
Some final thoughts on this study. I’d have liked to see some effort to address the impact of defective interfering particles (DIPs) on these results. DIPs are viruses with a large deletion in one or more of their genome segments and crop up in lab stocks of flu. Cells infected with DIPs would appear just the same as the cells studied in this work: missing one or more flu proteins. While I doubt DIPs can explain the high frequency of cells observed to lack flu proteins, the suggestion that influenza often fails to package its entire genome is extraordinary and thus needs all the evidence it can get – including ruling out a significant contribution from other known phenomena.
Also, regarding the infectious foci assays, it would be interesting if the relatively rare clusters of infected cells the authors observed were actually the result of co-infection themselves – just how often can a flu virus go in alone and be successful?
This study looked at flu infection of cells in a dish, but what’s the story in a complex species like a rodent or human? More on this next time.
I’m only covering one article today (and another related piece soon!), but I’ve no desire to sell the field short: there’s a wealth of research into flu genome packaging that I can’t get into now. Here’s a great open access review if you want to get your teeth into this area: Hutchinson et al. 2010.↩
Authors! This would be way more compelling if you could have looked at more than 2 VSV proteins, and also quantified the analysis with flow cytometry to make a better point. #nitpick↩
I’d like to bring you a number here, but this isn’t directly quantified in the paper. Check out the flow cytometry in Figure.3D to see how the top right hand quadrant of the graph (NA and HA co-expression) increases in intensity as multiplicity of infection (the number of viruses per cell) increases.↩
The authors used an infectious foci assay here, whereby they looked at cell layers with fluorescent antibodies to flu proteins again. After 15 hours they could see either single infected cells or clusters. In the clusters, the virus had colonised the neighbouring cells around the original victim.↩
MikeNicoll.blog 