All models are wrong, some are useful

 Cross referencing different systems can compensate for issues in individual systems and provide novel insight.

The saying about the limitations of models is attributed to a British Statistician, George Box. It speaks to how we use models to understand the world around us. This reflects is a deeper philosophical discussion as to what extent we can ever reach ‘truth’. But for now the point is that the models and observations we make in medicine and biology are often flawed. And we need to be aware of these flaws in order to better utilise the models.

In our recently published paper Comparative cross-species transcriptomics during RSV infection identifies targets to treat RSV disease we combined three different approaches to understand infection with Respiratory Syncytial Virus (RSV), a significant cause of illness in babies less than six months old.

Although natural RSV infection in children is the disease of interest, it is challenging to study directly in babies. It is very difficult to collect samples and it is often unknown when the child was first infected. An alternative is to use human infection challenge studies – where volunteers are deliberately infected with RSV. This has the advantage of being in the same species (the human) but the studies use young, healthy adults not babies. There is an additional challenge that all adults in these studies will have previously been infected with RSV at some time in their lives (probably several times) which will affect the immune response to any subsequent infection. The volunteers typically experience mild-to-moderate disease and so don’t fully recapitulate the disease seen in babies. Another alternative is to use mouse models. As well as the possibility to perform biological repeats in genetically identical individuals, following infection of a mouse, you can access all tissues and you can manipulate the response experimentally. However, there are interspecies differences in physiology, behaviour, viral tropism and genetics which can limit interpretation. All experimental approaches are ethically assessed, but there is a sliding scale – more can be done in adults than babies, more in mice than humans. And so by combining these different approaches we can compensate for limitations inherent to each.

In the published study, we compared the immune response in the blood, the lungs and the nose. As with different models, sampling different sites compensates for limitations of sampling an individual site. Blood is easily accessible and it is possible to collect large volumes of material, multiple times. But for a respiratory virus like RSV, it isn’t the actual site of infection (even though the lungs are highly perfused with blood). Blood can reflect cells moving into or out of the lungs, but not the complete picture. The lungs, as the site of infection tell us what is happening where the virus is and therefore can give us much more information. However, they are much harder to sample – it takes a medical procedure called a bronchoscopy to collect the tissue. Repeat sampling over time is not possible and these kind of samples cannot be collected from babies. The nose represents a good compromise. It is easily accessible and is the entry site for infection. Newer sampling methods have enabled the collection of good quality material from the nose without causing discomfort.

Having collected material from infected individuals, we then measured changes in gene expression. Cells, when they are infected or when they are responding to a local infection produce RNA that encodes the proteins they will use to fight off the virus. Profiling the changes in the RNA in a particular sample gives us a snapshot of how the immune system is working. We used an approach called RNA-Seq which captures all of the RNA in a sample and measures how many copies of each gene have been expressed. When two samples are compared side by side, the relative amounts of RNA can be evaluated; this is then presented as differentially expressed genes (DEG for short). The idea being that genes that change in amount are the ones that are important for the response to the infection.

Overall we evaluated 209 samples. When we pooled the data, we saw a clear increase in immune system genes following infection. This was not unexpected, the question we wanted to address was whether they were beneficial. The immune system has a dual role in disease, it is vital to protect us against viruses, but sometimes it overshoots and being in in excess of that required to clear the virus it can damage the lungs and cause us to feel sick. Within the data from this study, we observed increases in genes from the interleukin 17 family (IL-17). This is a type of signalling molecule called a cytokine which shapes the flavour of the immune response. It triggers a cascade of other genes, one of which is called S100A. Both IL-17 and S100A have been shown to cause enhanced disease following other viral infections, but their role in RSV is not known. Returning to the mouse model, we were able to block the action of S100A and show that when inhibited, there is less disease. Overall the study showed that integrating different data sets provides new insight and may ultimately lead to new treatments.

Bacterial Blockers

 

RNA vaccines have been the breakthrough vaccine technology of the past 5 years. They have been extremely successful in preventing disease following viral infection – particularly against SARS-CoV-2, the virus that causes COVID-19. They have also been licensed for use against another virus, RSV.

However, some questions remain as to whether they can be used against bacterial infections. One of the major questions relates to the way in which RNA vaccines work. The power of RNA vaccines is that they use the common building block of life (RNA), which is universal across all organisms. Because the platform uses the same stock material, the same manufacturing approach can be used to make any vaccine from Anthrax to Zika. The neat trick is that we are relying upon the injected cells to turn the RNA into protein, in a process called translation. However, whilst the code is the same for all living things, translating the letters of the RNA into the letters of amino acid isn’t all that is required to make a functional protein. There are subsequent modifications for example adding sugar molecules. These modifications differ between the different kingdoms, in particular, bacteria use a very different system to modify their proteins than humans. The questions is therefore whether RNA encoding bacterial genes would make properly folded proteins in human cells.

This is an important question because many of the most dangerous pathogens we face are bacterial in origin. This threat is particularly concerning due to the rise of antibiotic resistance. Which is where the frontline drugs we use to treat bacterial infections begin to fail. This is especially a concern in low income countries where rates of extremely antibiotic resistant bacteria are dramatically rising. We were interested in developing a vaccine against a bacteria called Acinetobacter baumannii which causes severe infection in people who have had to have mechanical ventilation, for example following surgery. In Vietnam, strains of this bacteria have been isolated that are resistant to nearly every single antibiotic.

An additional challenge for RNA vaccines for bacteria is choosing thee right part of the bacteria to target. Viruses are (often) simpler organisms encoding fewer proteins overall, so selecting the right one is (relatively) more simple. Bacteria can encode several hundred different proteins, which are often invisible to the human immune system.

In our recent paper, Intranasal delivery of mRNA expressing newly identified Acinetobacter baumannii antigens protects against bacterial lung disease we set out to identify and test a new vaccine targeting this tricksy bacteria. In a previous study we had used another vaccine approach called outer membrane vesicles (OMV). These are little packages that the bacteria spit out, containing a mix of proteins and other biochemicals. We had demonstrated that these could be used to protect against infection. Because OMV contain some, but not all of the proteins that A. baumannii encodes, we reasoned that some of them must protect against infection. We took two approaches to sift the proteins and found a subset of three that were in both datasets, we then generated RNA vaccine constructs of these and tested them. We were able to demonstrate protection against infection with one of the genes – Oxa23. Protection was even better if we delivered the RNA to the lungs.

We have been supported to Bactivac to undertake this work. They are a funding agency with a remit to develop new vaccines against antibiotic resistant bacteria. This is an important first step in the development of a vaccine against A. baumannii.

Bacterial Bubbles blow away the bugs

 

There’s no shame in not knowing the names of every bacteria, there are billions if not trillions of different bacterial species. The number that are pathogenic in humans is significantly lower – probably in the high hundreds. One review estimates that there are 1,400 known species of human pathogens (including viruses and parasites). So, it is still ok to not have heard of all of them.

One that probably passes under the radar of many people is Acinetobacter baumannii, not least because it is an absolute nightmare to spell. It is a gram-negative opportunistic pathogen. Most cases of A. baumannii infection are hospital acquired, often catheter or ventilator associated. There are approximately 1 million A. baumannii infections annually and it is associated with a nearly 35% mortality rate. The most pressing problem associated with A. baumannii is antibiotic resistance; it is extremely drug resistant – with nearly half of the infections resistant to the last line antibiotic carbapenem. In fact Acinetobacter is the A in ESKAPE (the priority list of the most important antibiotic resistant bacteria).

Since it is so drug resistant, other strategies are needed to control it. One of which might be vaccines. In our recent study - Intranasal immunization with outer membrane vesicles (OMV) protects against airway colonization and systemic infection with Acinetobacter baumannii we developed a novel vaccine against this important pathogen. The approach we used was to manipulate a product of the bacteria itself to generate protection. A. baumannii produces little bubbles of lipids and proteins called outer membrane vesicles (OMVs). Bacteria use these OMV to communicate with each other and potentially help them better infect us. However, they also contain lots of different bacterial antigens, some of which are potentially protective. The highly effective vaccine against Meningitis B (Bexsero made by GSK) contains OMV.

In our study, we isolated OMV from clinical isolates of A. baumannii. Using clinically derived strains was important – as these are probably closer to the ones in circulation, than the ones used in many studies from older varieties. To test whether the OMV worked, we developed new models of infection using the same strains. We showed the recent clinical isolates were more pathogenic than the standard lab strain. The clinical isolates were able to escape from the lungs into the blood and appeared to stably colonise the upper airways for at least 7 days after the initial infection.

Our first studies tested injecting the vaccine into the muscle – as this is the most common route of vaccination used. OMV injected this way did give a good immune response – leading to the induction of antibodies. However, the immunity raised following this route of immunisation was not very protective against subsequent infection. When area of interest in the vaccine field has been mucosal vaccination – delivering vaccines to the site of infection, in this case the nose and lungs. When Dr Sophie Higham (lead author on the paper) immunised via the nose, protection was significantly improved with a dramatic reduction in bacterial load following infection.

This work shows two things – OMV can be very effective vaccine candidates against bacterial infections and that immunising in the site of infection can be beneficial. The next step is to look how to scale up for human studies.

A Pre-Pandemic vaccine.

 A paper about pandemic vaccines: shaped by a pandemic.

We first started writing the grant application that supported this work in 2017. The work was funded by the Coalition for Epidemic Preparedness Innovations (or CEPI). CEPI is a global partnership that was set up to accelerate the development and delivery of vaccines. The original plan was to look at three viruses – the one I was leading was influenza. We had a remit to develop a vaccine that could provide protection against infection within 6 weeks of the first immunisation.

The approach we used was an RNA vaccine, which unless you have been living under a rock for the last 3 years I am assuming you have heard of. But just in case, what these vaccines do is to take the genetic material that encodes a tiny bit of the virus and inject it into the muscle. Once injected your muscle cells make the viral protein training your immune system to recognise them. This means that when you are exposed to the real virus, you can fight it off better. Specifically we were using a self-amplifying RNA vaccine, this is subtly different; you can in theory get bigger responses for smaller amounts of material. I have described them before here.

One of the first questions we looked at was how best to formulate the vaccine; the work is described in our recently published paper Formulation, inflammation and RNA sensing impact the immunogenicity of self-amplifying RNA vaccines. RNA is quite unstable and needs to be mixed with other compounds in order to get it into cells. We tried three approaches a cationic polymer (pABOL), a lipid emulsion (nano-structured lipid carrier, NLC) and three lipid nanoparticles (LNP). In simple terms to get RNA into a cell, you either need to add positive molecules (the cationic polymer) or some fat bubbles (the LNP and the NLC). We noticed that responses to the LNP were very much better than the other approaches and wanted to understand why.

One of the questions we asked was about the role of inflammation. Inflammation is a cascade of signalling by which the immune system recruits cells to a site of infection or danger. In conventional vaccines it is important because it alerts the cells that there is something foreign to be recognised and trains up the vaccine response. However, there was a question about its role in RNA vaccines, because RNA vaccines need to be made into proteins in the body to work some aspects of the immune system might inhibit this. Surprisingly (which is up there in the go to words of academic writing, alongside interestingly), we saw that vaccine induced inflammation was associated with better, not worse responses. The next step is to further dissect how this inflammation is beneficial.

Coming back to the timing of this project. We started the labwork in 2019, when the idea of a world changing pandemic virus was somewhere in the future. Then of course the events of 2020 caught up with us. The work definitely slowed down – though aspects of it were incorporated into the Imperial College vaccine trial. The formulation that we had showed to work best in the saRNA system was the one used in the clinical trial. As part of the ongoing vaccine research, I was able to come into the lab occasionally, providing a sanity lifeline and also the very odd experience of commuting through an empty London. Where this exciting vaccine technology goes next is still an important research question and one I look forward to continuing to investigate.

Do dirty viruses make you sicker?

 I think we can all safely agree that the immune system is complicated; that viruses are complicated and when you combine the two the overall outcome is more complicated than the sum of its parts. In the current study (Levels of Influenza A Virus Defective Viral Genomes Determine Pathogenesis in the BALB/c Mouse Model) led by Becky Penn in Wendy Barclay’s group at Imperial College London, the question was what happens when the infectious virus (in this study influenza) contains more (or less) junk.

Viruses contain genetic information wrapped up in a protein coat. For a more detailed overview of what viruses are – can I recommend my book Infectious (out now in paperback). The immune system recognises viruses through a number of different ways, but one of them is through recognising the viral genetic material. In the case of influenza this is somewhat more simple because the genetic material is different to human genetic material. Humans pass their genes between different generations using DNA, but influenza uses a related molecule called RNA. Human cells also use RNA, but for different purposes – mostly for transmitting genetic information within the cell. The viral RNA therefore is a clear danger sign to cells that there is an infection. Viruses have evolved to hide their RNA from the immune system. They wrap it around proteins like thread wrapped around a cotton reel. This means the immune system can no longer see the RNA and reduces the response to it.

However, the amount of RNA is not constant. Sometimes individual virus particles fail to hide their RNA, which makes the immune response to them stronger. One factor that affects the packaging of the RNA is how the viruses grow. Becky manipulated the way that she grew her influenza virus, with some of the stocks having much more of these badly packaged viruses and some having neatly packaged genetic material. The theory was that the badly packaged viruses was, because they are more dirty trigger the immune response, what was unknown was how this would effect the outcome. There were 2 possibilities:

1. The immune system would be so fast at seeing the virus, it would kill it before it ever took hold.
2. The immune system would go into overdrive.

Surprisingly, the results were a bit of both. The dirty viruses (also called High DVG because they have more defective viral genomes) were indeed detected quicker by the immune system. This led to the release of signalling molecules called interferon. Interferons trigger an anti-viral state, shutting down the growth of virus within the cells.

The cleaner stocks performed very differently. They were initially able to escape the immune system and set up an infection in the lungs. This infection then, paradoxically, led to the production of the dirty viruses in the lungs of the infected animals. This production in the lungs led to a cycle of inflammation, basically putting fuel on the fire and increasing the severity of disease. This tells us that disease after viral infection is driven by a combination of factors – the damage caused by the virus itself and the immune response to the damage. On a more niche note, it tells us that the quality of experimental virus used is really important in shaping the outcome of the study. Which is one of the almost infinitely complicated variables in doing biological studies with 2 live agents (let’s not even begin talking about time of day, chronobiology and the impact that has!

Can we save lives by deliberately infecting people?

In the middle of the pandemic, scientists intentionally infected healthy volunteers with SARS-CoV-2, the virus that causes COVID-19. John Tregoning, Reader in Respiratory Infections at the Department of Infectious Disease, explains why these experiments, and the volunteers who take part in them, are critical to modern medicine.

In early March 2021, in the middle of the COVID-19 pandemic, a surprising-sounding set of experiments were taking place. Researchers at Imperial College London (and separately at the University of Oxford) were deliberately infecting healthy volunteers with SARS-CoV-2. This was in fact the latest in a long line of controlled human infection studies – where volunteers are deliberately infected with an infectious pathogen under extremely controlled conditions.

Deliberate human infection for health benefit goes back a long way – the earliest evidence of infection for beneficial use is 10th Century China, deliberately inoculating healthy people with smallpox to make them immune to the disease. This practice continued into the 18th century, when an English doctor, Thomas Dimsdale, deliberately infected Catherine the Great and her son with a very low dose of smallpox virus to protect them against the disease.

This idea of infecting people deliberately to protect them from disease led to Edward Jenner’s famous studies inventing the first ever vaccine. Jenner hypothesised that you didn’t need to use material derived from smallpox to be protected, you could use material from a related virus, cowpox. He proved this worked using a human challenge study; he vaccinated James Phipps (his gardener’s son) with cowpox then deliberately exposed him to smallpox repeatedly, showing that the vaccine worked and Phipps was immune to smallpox.

Deliberate infection

The practice of deliberate infection for scientific benefit really took off after the demonstration by Pasteur, Koch and others that microbes cause disease. In the early 1900’s, Walter Reed, the American public health pioneer, was trying to understand where yellow fever came from – he had a suspicion that it came from mosquitos. This was important because identifying the source could alter behaviour and reduce the incidence. To test his hypothesis, Reed recruited 11 volunteers to be bitten by mosquitos that had previously bitten a yellow fever patient; two of the volunteers contracted yellow fever, strongly supporting his idea. One important development in Reed’s infection studies was the concept of ‘informed consent’. The volunteers were informed about the risk to themselves of participation, before they gave their consent to take part. Sadly, later in the 20th century, some human infection studies entered a darker chapter, with the horrific experimentation on prisoners in Nazi Germany and Imperial Japan without their consent.

Informed consent is the bedrock upon which all modern research involving volunteers is built, and it is crucial for infection studies. The landscape of human infection studies has changed dramatically since the middle of the 20th century; now, ensuring the health and safety of participants is of paramount importance and trials are carefully designed to minimise any potential risks. Studies are only performed following extensive ethical review by an external body, for example all human infection studies carried out at Imperial College London have ethical approval from the UK Health Research Authority. There is ongoing debate about whether infection studies can ever be ethical, in terms of deliberately exposing someone to the risk of harm; even in the context of minimising the risk. However, there are many benefits to the studies and when volunteers understand the risk and choose to participate for the greater good, they can achieve important things.

Vaccines

One of the ways in which deliberate human infection studies (or ‘challenge studies’) are most beneficial is in the testing of vaccines. Vaccines are tested in the same way as any drug and the first studies involve a small number of participants who receive a dose of vaccine and are closely monitored to check first and foremost whether the vaccines are safe. These early studies (called Phase I clinical trials) can also inform about whether the vaccine is making an immune response. However, in order to demonstrate that the vaccine can prevent disease, much larger studies are needed. These studies (phase III) are often very large, because you can’t be sure how many of your vaccinated volunteers will then be exposed to the infectious agent. This means that you need huge numbers of subjects to get to statistically meaningful numbers to compare infection rates with and without a vaccine.

Infectious challenge studies can help to overcome this barrier, especially when the pathogen being tested is rare. One example of this is typhoid, a bacterial infection that causes diarrhoea in approximately 10-20 million people a year, mostly in low- and middle-income countries. A research team in Oxford gave volunteers a typhoid infection and tracked them untill they had clinical symptoms before treating them with antibiotics. Again, pausing to think of the volunteers – knowing that you are likely to get a bout of diarrhoea and going ahead with it anyway, for other people’s benefit, takes a special mindset.

Indeed, without volunteers, modern medicine would falter, so we all owe a large debt of thanks to these selfless individuals. Having shown it was possible to infect people in a controlled way, the Oxford group tested whether two new vaccines could reduce disease. They showed that whilst 77% of the volunteers without a vaccine developed typhoid, only 35% of the vaccinated volunteers did. Deliberate infection studies have also been used to support the rollout of vaccines for cholera, malaria and shigella.

Measuring antibodies

Another important benefit of deliberate human infection studies is in understanding how specific viruses cause disease and how we can be protected against them. The Common Cold Unit was a British research centre operating on Salisbury plain between 1946 and 1989. It set out to understand respiratory infections; being somewhat isolated it was able to look at transmission of colds, by infecting one volunteer and then housing them together with other uninfected volunteers. It also provided us with important information about the levels of immunity required to protect against influenza. By measuring antibodies in the blood of people before they were infected it was possible to identify a threshold above which infection was unlikely to occur; this threshold is still in use for the development of influenza vaccines.

Some diseases have more challenges than others in setting up the infections. Whilst respiratory viruses can be grown and dripped into the nose, other infections get into our bodies through a third organism, called a vector. In the case of schistosomiasis (sometimes called bilharzia), the disease-causing parasites live in snails (the vector) before infecting people. To help develop drugs and vaccines for this neglected tropical disease, researchers have had to learn snail husbandry!

Returning to the coronavirus human challenge studies, these look to address both the development of vaccines and improve our understanding about infection. In the earliest results from the Imperial-led study, the team showed symptoms start to develop very fast, on average about two days after contact with the virus. The infection first appears in the throat; infectious virus peaks about five days into infection and, at that stage, is significantly more abundant in the nose than the throat. It was seen that volunteers who had not had COVID-19 before could be infected with an extremely low dose of the virus, which might help to explain why SARS-CoV-2 is so infectious. It can also inform more generally about the behaviour of respiratory viruses. These studies are now progressing to help in the design and testing of the next generation of vaccines and drugs.

As we have seen in the last two years, infections can be extraordinarily disruptive. Studying how they behave, why we get infected and how to prevent this is extremely important – when performed safely and ethically, human infection studies are an important part of our toolkit.

 This first appeared on the Imperial College, Faculty of Medicine Blog

Ferreting about

 

As has been clearly shown during the COVID-19 pandemic, RNA vaccines are an incredibly powerful tool. We have seen how potent they are against a rapidly emerging infection, with the speed at which they can be produced a major benefit. From the first genome sequence published to the first experimental vaccine dose designed, manufactured and administered was seventy-one days, less time than it takes to make cheese. However, more work is needed to deliver on the promising start that RNA vaccines have had.

One of the challenges has been with the side effects of the vaccine, sometimes called reactogenicity. This is in part caused by the body’s reaction to the vaccine itself: RNA is a trigger for immune reactions, it is how the body sees viral infections. Tweaking the amount of RNA in the vaccine, whilst maintaining its ability to induce a protective antibody response is one way in which we can improve RNA vaccines.

We have been interested in a slightly different approach, called self-amplifying vaccines. We showed (pre-pandemic) that these saRNA vaccines can lead to protection for a much lower dose. These vaccines are based on RNA, but able to make more copies of themselves in the injected muscle, giving you more bang for your buck. Professor Robin Shattock (also at Imperial College London) performed the first in human clinical study using saRNA vaccines during the pandemic. This trial had a mixed result, with not all of the volunteers producing antibodies after vaccination. In parallel with this first in human study, we looked at how the vaccine might work in other model organisms, as a way to understand how these vaccines work.

In our recent study Polymer Formulated Self-Amplifying RNA Vaccine is Partially Protective against Influenza Virus Infection in Ferrets, we tested the immune response to saRNA vaccines in ferrets. This might sound like an odd choice, but ferrets have a number of similarities to humans in they way they respond to infections, especially influenza. We set out to test whether saRNA vaccines could protect ferrets against flu. As with the human study, we had mixed results. The ferrets that responded to the vaccine by making antibody were well protected against infection, however, not all of the ferrets made antibodies. Understanding why this is the case is important in the future development of this extremely promising vaccine platform.

Ever shifting strains

The bacteria Streptococcus agalactiae is more commonly known as Group B Streptococcus or GBS. The Streptococcus groupings were developed by Rebecca Lancefield in the 1930s. They are somewhat redundant now due to other methods of categorising bacteria, but Group A Strep and Group B Strep are still commonly used (Group A strep or GAS is Streptococcus pyogenes, which causes a range of diseases, the most well known of which is scarlet fever).

Anyway back to GBS, which is mostly harmless – until it isn’t. It can colonise people without causing an infection. It mainly lives in the gastro-intestinal (GI) and genito-urinary (GU) tracts. Unfortunately, it can have major and severe impacts during pregnancy. It is the leading cause of neonatal infection, having severe consequences for the newborn baby. It passes from the mother, who has some level of immune protection, to the baby (who doesn’t) during, or shortly after, birth. Whilst antibiotics can reduce the burden of disease, they are not completely effective and alternative preventative approaches are needed. One of which is vaccination, but one of the challenges is that the mechanism by which women are protected against disease is not well understood. More work is needed to understand the interplay between the immune system and the colonising bacteria.

In our latest study, Group B Streptococcus (GBS) colonisation is dynamic over time, whilst GBS capsular polysaccharides-specific antibody remains stable we worked with Professor Kirsty le Doare and her group at St George’s University in London to look at the interplay of GBS colonisation and the immune response. Women were recruited and had samples taken from their GI-GU tracts every 2 weeks for 12 weeks. We then measured whether GBS was present, which strain of GBS it was and whether colonisation led to an increase in the amount of antibody the women made.

We observed that colonisation with GBS was dynamic – there were volunteers that were colonised initially that subsequently cleared the colonisation, there were others that acquired bacterial colonisation, there were some that were never colonised and others that had multiple different strains. However, whilst the bacterial colonisation was variable, the levels of antibody were fairly constant implying that acquisition of bacteria may not be directly affecting whether the women make more antibody – or that they have previously been exposed to the same strains and therefore don’t make new responses. Alternatively, there may be something special about the guts that mean that bacteria that live there (but are not necessarily causing disease) don’t trigger the immune response.

This study was performed during the COVID lockdowns which significantly disrupted the collection of samples. However, thanks to innovations by the study leads – such as home sampling, it was possible to continue. These kinds of approaches may mean larger studies can be performed as they are less intensive on the researchers and participants.

This study demonstrated the complex interplay between host and bacteria. Further investigation is needed to understand how best to utilise a vaccine to protect against this devastating disease.

Don't T me off

 

Vaccines can be made of all sorts of things – the pathogen itself, either killed or in a weakened form; protein or sugar from the coat of the pathogen or nucleic acid derived from the pathogen. This nucleic acid can either be double stranded DNA or single stranded RNA. And whilst RNA has suddenly acquired celebrity status, the DNA based approach is the older (slightly less successful) sibling.

One of the ongoing challenges with nucleic acid-based vaccines is balancing the expression of the vaccine antigen against recruiting the immune cells that are needed for the vaccine response. This is because, unlike older vaccines, nucleic acid vaccines need the cells of the body to make the protein that the immune system will recognise in situ. There are 2 challenges with this, firstly the DNA itself can trigger a cascade of events that mean the cells are less likely to make the DNA encoded protein and secondly, if the cells can be persuaded to make the foreign protein, it can act as a flag for the immune system to target and attack those cells.

We were interested in this second challenge in our recent study: Blocking T-cell egress with FTY720 extends DNA vaccine expression but reduces immunogenicity. We wanted to understand how one arm of the immune system – the T cell impacts the response to DNA vaccines. T cells have a range of roles, including where one T cell subset (CD4 T helper cells) orchestrates the immune response to another T cell subset (CD8 T killer cells) to facilitate killing of infected cells. The CD8 cells recognise target cells because infected cells wave tiny bits of foreign protein on their cell surface.

To explore the role of these different T cells in the context of DNA vaccines, we made use of a protein called luciferase; this comes from fireflies and gives them the ability to produce light. Technically speaking it is an enzyme that catalyses the breakdown of a molecule called luciferin, when luciferin is chopped up by luciferase it becomes unstable and can only relax by releasing light energy in the form of light.

We injected DNA encoding luciferase into the leg muscle of mice and measured how much light their legs emitted. As expected, the injected mice produced light (not loads – you need a special machine to measure it); the amount of light peaked 2 days after injection but disappeared 3 weeks later. To test the role of T cells, we gave one group a drug called FTY720 (or fingolimod, which is why we refer to it as FTY720). FTY720 has the interesting property of preventing T cells moving around the body – so they can’t get to where the luciferase was being made. The mice treated with FTY720 kept on producing light for as long as we injected them with the drug. The FTY720 injections could be staggered, it just needed to be kept over a threshold.

We thought this was then going to be good news in terms of the immune response, reasoning that the longer the vaccine was made in the cells the greater the response. So, we repeated our study, but instead of using DNA making luciferase and measuring light, we used DNA that encoded the surface protein from HIV. In this experiment, the FTY720 significantly reduced the immune response to the vaccine – with the treated mice making much less antibody.

These studies reflect the fact that T cells have 2 roles; the killer CD8 T cells which can kill the cells that have taken up the DNA vaccine and are making the foreign protein but also the helper CD4 T cells which boost the immune response. FTY720 was a blunt instrument shutting down both types, with a resultant negative impact. This study showed us that T cells can limit expression of DNA vaccines in the host cell, but we need to do more research to extend that expression without impacting other arms of the immune response.

The complex world of the child’s nose.

As any of you who have children, in fact anyone who has ever met, seen or been a child, there are all sorts of things up their noses, fingers, Lego pieces, but more importantly viruses. They are awash with different viruses. We are particularly interested in the effect of one viral infection on another.

We investigated this in the context of a vaccine study. There is an influenza vaccine that is live, but genetically weakened; called live attenuated influenza vaccine or LAIV. This vaccine is sprayed up the noses of children where it causes a limited infection that can train the immune system and protect against future influenza infections. This gives us a safe way to explore how flu infections happen in children and what factors affect infections particularly the immune response. This study was performed in The Gambia by Dr Thushan de Silva.

In our recent study: Prior upregulation of interferon pathways in the nasopharynx impacts viral shedding following live attenuated influenza vaccine challenge in children we looked at the influence of the presence of other infections on LAIV infection. The first thing we measured was the presence of viruses in children – remarkably 42% of children had detectable viral RNA. This reflects other studies we have done where we saw that 20% of children had another viral infection. For the analysis we then compared the responses in children with and without another infection and what impact this had on how well the LAIV viruses could establish a local infection in the nose .

In particular, we were interested changes in the genes in the nose due to infection with other viruses. These changes in genes give us a more global sense of the response to infection. What we are actually measuring are changes in RNA levels, called the transcriptome. We can group the individual genes into families based on their function. We are particularly interested in genes associated with the immune response.

What we observed in our study was that children who had another infection at the time of LAIV immunisation had far greater levels of genes associated with signalling through a pathway called type I interferons. These genes are part of the innate immune response to infection and switch on an antiviral state, which makes it more difficult for other viruses to infect them. In line with this, children with higher levels of the anti-viral genes had lower LAIV viral replication.

These findings might help us understand differences in outcome following infections such as flu and how other viral infections immediately before exposure could influence this.

Signalling failure delays training of antibody in early life

 

This may come as a surprise, but there are other viruses that infect our lungs than COVID! One of them is called Respiratory Syncytial Virus (RSV). This innocuously named virus is THE leading cause of hospitalisation in children during winter months, but somehow doesn’t get the same level of attention as its sexier relatives influenza and SARS. One of my colleagues Prof Peter Openshaw has previously suggested it be called ‘deadly killer virus’ to get it the attention it deserves.

One of the interesting aspects about RSV is that it is possible to get re-infected with the same virus. This is unusual because the assumption is mostly that once the immune system has seen a virus once it is then better trained to deal with it in the future, preventing further infections. However, some viruses such as RSV (and also coronaviruses, like the one that causes COVID) can reinfect. We don’t fully understand why this is the case, but one contributing factor is a protein in the blood called antibody. Antibodies are highly specific molecules made by the immune system that can bind and kill viruses.

Antibody molecules are produced by a white blood cell called the B cell, but in order to produce the best possible antibody, B cells need help from another type of cell called the T cell. The conversation between T and B cells happens in the lymph nodes – which is why you get swollen glands after infection or immunisation. In our recent work, ‘Enhanced IL-2 in early life limits the development of TFH and protective antiviral immunity‘ recently published in the Journal of Experimental Medicine we explored this interaction. Specifically, we asked the question are there differences between the T-B cell crosstalk in early life – the time of greatest susceptibility to RSV.

We found that baby mice infected with RSV produce less antibody than adult mice infected with the same dose. This lack of antibody left the mice susceptible to re-infection with the virus. Side by side with the reduction of antibody, there were fewer of the helpful T cells needed to train the B cells. If we specifically removed those helpful T cells (called T follicular helper cells or Tfh) from adult mice before RSV infection, we saw a very similar effect – re-infection.

We dug deeper as to why these handy Tfh cells are not so active in early life. We identified a role for a molecule that cells use to talk to each other called interleukin two (the sequel to the commercially more successful, but less interesting interleukin one). There is more of this molecule sloshing around in early life and it may play a role in training the early immune system what is good and what is bad.

Overall, out findings may help us to develop better vaccines that work from the moment babies are born and stop them catching viruses such as RSV.

Double Trouble: IFI44 and IFI44L

An important component of host defence against viral infection is cell intrinsic immunity. This type of immunity is mediated at a cellular level rather than requiring recruitment of other cells to restrict the infection. It is characterised by the induction of an anti-viral state, which limits the ability of viruses to enter cells, make copies of themselves within the cell or exit the cell having replicated. The induction of this anti-viral state is triggered by a signalling molecule called interferon. Interferon signalling leads to the expression of a multitude of interferon stimulated genes (ISG). Many of these ISG are uncharacterised in terms of function.

Technological developments over the last twenty years have changed the way that we investigate how cells work. In particular, the use of transcriptomics, where the messenger RNA (mRNA) in a sample is measured. mRNA is important because it is the intermediary between the cell nucleus, where the genetic information is stored and the ribosome, where proteins are made. Transcriptomics gives an overview of what the cell is doing. However, transcriptomics is a broad-brush tool that does not necessarily give the fine detail of what individual genes do in the prevention of infection.

Over the last few years we have undertaken a program of work to understand the role of individual ISG in the control of viral infection. In particular we are interested in respiratory syncytial virus (RSV). RSV infects the lungs of children – all children will be infected with it before the age of 2 years old, most before 6 months of age. Some of these children will get extremely sick with RSV infection and we hypothesized that this is because they fail to control the virus early on during infection. However, prioritising which ISG to investigate was an issue, especially given the large amount of data available. We therefore used a screening process to identify those genes which are more commonly associated with RSV infection (https://doi.org/10.1128/mSystems.00051-16).

This screening process led us to work on a pair of genes called Interferon-induced protein 44 (IFI44) and interferon-induced protein 44-like (IFI44L) which we published in the Journal of Virology (https://jvi.asm.org/content/early/2020/06/26/JVI.00297-20). We confirmed that the genes were induced following RSV infection and then set about exploring whether they played a role in the control of infection. The first question was what would happen in the absence of either gene. Using two different gene-knockout approaches, CRISPR-cas and siRNA, we showed that when you reduce expression of either gene, the virus replicated better. We then did the opposite experiment, increasing the amount of both genes in the cells, this led to decreased viral replication. These initial findings were supported by studies in mice and children. Mice lacking the IFI44 gene were more susceptible to RSV infection and children with lower expression levels of the gene, as determined using transcriptomics on their blood, were more likely to have a more severe infection – though this was a weak association.

The question remains as to how IFI44 and IFI44L prevent viral infection. One of our observations was that altering the levels of the two genes altered the ability of cells themselves to replicate. When there was more IFI44, the cells replicated more slowly, when it was removed they replicated faster. We think that this gives us a clue as to their function – somehow they limit resources that both the cells and the virus need to make more copies of themselves. We are now looking to understand exactly how this happens. What is fascinating is that there are so many different genes involved in the prevention of viral infection and an important question is how do they interact to protect us.

How does flu affect poo?

There are lots of motivations to do science – to solve a grand global challenge, to answer a fundamental question about how the world works, because it’s a job and you’ve got to do something before pub opening times. For me, it is about solving puzzles – piecing together a coherent story from a jumble of observations.

In our most recently published paper, we got to do exactly that, and best of all, we got to draw it all on a white board, like in CSI.

Solving science problems, one whiteboard at a time

We started with an observation (from a previously published paper). When mice get viral lung infections, the bacteria in their guts (the gut microbiome) changes. This was quite surprising. So in the follow up we wanted to understand why, but also to understand if the changes in the microbiome affected anything.

But first of all, why we did care at all. As a brief reminder, it is entirely normal to have gut bacteria, in fact various estimates put the number of cells of gut bacteria as higher than the cells of person surrounding them. The gut bacteria do a whole range of important things, from breaking down your food for you to regulating your immune system.

Back to the story. One other thing we did know was that when infected with a virus, mice lose weight. In fact, relative to body size they lose quite a lot of weight, up to 25%, which they can put back on very quickly. This was our starting point and we speculated it could be due to a number of reasons, including the increased effort of breathing. Helen (the student working on the project) made the very simple step of measuring the amount of food eaten. She found that over the course of the infection, the mice stopped eating and it was this that caused the weight loss. When we mimicked this reduction in eating by reducing the amount of food the mice had per day, we saw an identical change in the gut microbiome. So problem one solved.

However, this then led to the next question – why do the mice stop eating? We drew on our previous studies, in which we had showed that the immune response to infection was associated with weight loss after infection. Specifically we were interested in an immune cell called the CD8 T cell and a signalling molecule called Tumour Necrosis Factor (TNF), which the immune system releases to activate other cells. We used antibodies (molecules that are recognise other proteins very specifically) to block CD8 T cells or TNF to see if the immune response was responsible for weight loss. To cut the story short, T cells were important, TNF not so much. It still begs the question, why does the immune system cause weight loss. At this point the simple answer is, we don’t know.

We then switched our attention to the downstream effects. One thing we wanted to see was if the change in the gut bacteria, changed susceptibility to other gut infections. In our hands, they didn’t – when we infected mice with a lung virus, there was no effect on a subsequent gut infection. However, we did observe one difference after infection. There is a complex network of chemicals in your guts, called the metabolome. After infection, these biochemicals in the guts changed. What was intriguing was that some of the chemicals that increased have previously been described as anti-inflammatory. One crazy speculation from this is that by stopping eating the mice actually get better, but we have no proof that this is true and it might just be a correlation.

So what did we learn? When mice are infected with a respiratory virus (in their lungs) their gut microbes change. The gut microbes change because they lose weight. They lose weight because they stop eating. And they stop eating because of something to do with the immune system. So problem solved, sort of.

Just hanging about

Presumably the question you are asking yourself is what determines persistence in acute RNA viruses? If not, why not?

Viruses have been shown to persist – stay present in the body, potentially after the symptoms of infection have passed. Most of the evidence and mechanism for viral persistence has been collected for DNA viruses and retroviruses (that is RNA viruses that convert their RNA genome into DNA and insert it into the host). However, there is clear evidence that non-retroviral RNA viruses can persist (see the review here). We normally think of these acute RNA viral infections as being short lived, cleared by the host and only succeeding if they can transmit to a new individual. However, this strategy has limitations, particularly if there are no new individuals who haven’t been infected by the virus. Therefore, viruses need to have evolved a way to maintain a reservoir, this is particularly important when we consider that viruses are obligate parasites – they have to use host cells to replicate and survive. It is particularly interesting to think that the viruses can persist in spite of selective pressure from the host immune response which is trying to clear the virus.

The question we set out to answer in a recently published study, led by Prof Rick Randall and Prof Steve Goodbourn, was how RNA viruses can switch between an acute and persistent state. The work focused on parainfluenza virus (PIV), which is a member of the paramyxovirus family. Viruses require specific proteins to make copies of their genetic information, which is described as the polymerase complex. The imaginatively named P protein of parainfluenza virus is a core part of the viral polymerase. If the P protein was phosphorylated (a mechanism by which cells can control protein activity), then the virus no longer replicated in the cells, but and this is important, the viral RNA was maintained within the cell. We then demonstrated that the phosphorylation status of the P protein and was determined by a single amino acid within the protein, if this changed then the protein could be activated or de-activated. Since amino acids are determined by the genetic code of the virus, specifically by 3 nucleotides, a single nucleotide change can alter the amino acid sequence, in turn affecting the phosphorylation of the protein and whether it is active or not. So in essence there is a switch that can control whether the virus makes copies of itself within the cell, given that RNA viruses have leaky polymerases (they make inaccurate copies of their own genes), this flip between active and inactive states can occur readily during the infection/ replication cycle. The switch may be driven by immune pressures, we demonstrated that lytic viral variants replicated to higher levels in a mouse model but were cleared much faster, whereas the persistent variant led to a prolonged infection. We proposed that the virus may start in an active state producing lots of copies of itself, before switching to a persistent state to develop a reservoir.

This was in essence a piece of basic research addressing a fundamental question in virology, but it does have broader impact, understanding why and how RNA viruses persist has implications for infection epidemiology as well as potential for developing novel vaccine platforms.

Breathe it in

Influenza is a serious cause of death and disease, contributing to the winter healthcare burden. One approach to reduce this is vaccination. In addition to the injectable influenza vaccine, which is given as an intramuscular injection there is an intranasal vaccine. This vaccine is also referred to by its initials LAIV – live attenuated influenza vaccine. The vaccine is a live vaccine, that has been adapted to reduce its pathogenicity. Specifically, the vaccine virus was adapted so that it can only replicate at lower temperatures. This is important because there is a temperature differential across the airways: the nose, because it is drawing in cold air is cooler than the lungs. The nose is at approximately 30°C, compared to the lungs which are at 37°C. This means that viruses that can replicate at 30°C are restricted to the upper airways and therefore cannot cause severe disease. The vaccine virus is then administered by a nasal spray syringe, once it gets into the nose, it replicates and this replication is important in the induction of an immune response.

However, one of the problems with influenza is that the virus changes season on season, sometimes in small steps (antigen drift) but sometimes in much bigger jumps (antigen shift). This changing of the viral strain necessitates new vaccines each influenza season. Most of the variation comes through the surface antigens, haemagglutinin and neuraminidase, which are the H and N of influenza virus nomenclature. Luckily the same temperature sensitive attenuated vaccine virus strain can be used as a backbone into which different H and N genes can be substituted. However, to achieve greater coverage three or four (depending on manufacturer) different virus strains in the vaccine, normally two A strains (H1N1 and H3N2) and two B strains.

In the UK, LAIV has been recommend for all primary school age children (up to 11), and some other high risk groups. This decision is based in part on the herd protection that this vaccine could potentially have, protecting the elder generation by reducing the infectious reservoir. However, in recent years there have been some concerns with the efficacy of the LAIV – particularly in the USA where efficacy dropped from 85% before 2009 to 17% in 2013-14 which led to a reversal of the American  Advisory Committee on Immunization Practices to recommend suspension of LAIV between 2016 and 2018. We wanted to understand factors that affected the immunogenicity of this vaccine.

In a previous study, we had described how nasal antibodies, specifically of the IgA type were associated with reduced viral shedding after influenza infection (https://www.frontiersin.org/articles/10.3389/fmicb.2017.00900/full). So  now we wanted to look into the effect of immunisation of children with LAIV on IgA. In our recently published study (https://onlinelibrary.wiley.com/doi/full/10.1111/cei.13395) we saw that three out of the four strains in the vaccine were able to induce a significant increase in IgA. Interestingly the only strain not to induce an increase in IgA – H1N1 – was the one for which concerns have been raised for protective efficacy. Though it was not clear in this study why the H1 strain might behave differently.

In a separate study (https://www.thelancet.com/journals/lanres/article/PIIS2213-2600(19)30086-4/fulltext) in collaboration with Dr Thushan da Silva in Sheffield, we looked into potential reasons for the differences. By happy coincidence, Thushan ran clinical trials with LAIV over 2 seasons and the H1 vaccine strain was changed between the years from A/17/California/2009/38 (Cal09) to A/17/New York/15/5364 (NY15). This gave us an opportunity to understand a bit more about how vaccine strain changes can affect immunogenicity. Strikingly the change in vaccine led to a significant increase in vaccine response and this was linked to how good the vaccine was at replicating – the newer strain (NY15) replicated better both in vivo and in vitro and this was associated with stronger immune responses.

Based on these studies, we want to look at how vaccine replication is associated with immunogenicity and what viral factors enable enough replication to work as a vaccine without causing infection themselves.

Postscript
Interestingly the uptake rate for the free LAIV vaccination in schools in the UK is only 30% (as at 16/12/19). This vaccine has been opt-in rather than opt out. One question is if it was made the default and then people had to opt-out would uptake be higher?

Protective protozoa

Flu, caused by the influenza virus is unpleasant. Even in non-pandemic years, it causes 290,000 to 650,000 deaths. In the absence of a ‘universal’ vaccine that could provide protection against all possible variants of the virus, new vaccines need to be selected and manufactured each year. The majority of these vaccines are manufactured using eggs. Influenza virus is grown in chicken embryos inside the eggs which are then cracked open prior to purify and inactivate the virus for vaccine use.

There are a number of limitations to this approach. Firstly, it is complex to scale up, for example during a pandemic. It can also induce a selective pressure upon the vaccine virus – chicken cell and human cells have slightly different receptors for influenza on their cell surface and co-factors within the cell. This means that in order to replicate efficiently in egg cells the virus may undergo some slight changes. If these changes are in regions of viral proteins recognised by the immune system,  for example haemagglutinin, then the vaccine virus might induce a memory immune response which does not recognise the virus that is actually circulating in the wild. The final problem is that some viruses, in particular the highly pathogenic ones (H5N1 and H7N9) are deadly to birds and kill the chicken embryos before enough virus is made for the vaccine.

Therefore alternative manufacturing approaches are required. One tool that has been widely applied across all fields of biological drug manufacture is the use of recombinant cell culture – where genes from one organism are expressed in cells of another. There is a licensed influenza vaccine (Flucelvax) which is manufactured using the MDCK cell line. These cells were originally isolated in the 1950s from a dog kidney, specifically a cocker spaniel, by S.H. Madin and N.B. Darby – hence Madin-Darby Canine Kidney (MDCK) cells. Growing cells from mammals has advantages compared to embryonated chickens, but there is value in developing alternative methods.

We investigated an alternative manufacturing approach in our recently published study Recombinant Haemagglutinin Derived From the Ciliated Protozoan Tetrahymena thermophila Is Protective Against Influenza Infection in Frontiers in Immunology. Working with a biotech company based in Germany (Cilian, AG) who use a protozon ciliate called Tetrahymena thermophila for the manufacture of biologics. This system has a number of potential advantages, it uses conventional manufacturing equipment, the same as that used for both bacterial and yeast based manufacturing systems.

However, it was possible that viral proteins manufactured using a protozoan might not induce a good vaccine response. We therefore set out to test the immunogenicity of the ciliate derived material. We demonstrated that immunisation with recombinant haemagglutinin could protect against an infection with a matched influenza virus. We saw this with haemagglutinin derived from either influenza A or influenza B viruses.

This proof of principle study therefore opens that path for further development of the Tetrahymena thermophila platform for vaccines. The major next step will be to work the platform up to a good manufacturing practice (GMP) grade material so it can be tested in clinical trials.

Natural antivirals

Before the immune system is activated, cells have their own intrinsic defence against viral infection. This comes in the form of proteins that can inhibit various stages of the viral life cycle. Some of these proteins are constitutively expressed and others are inducible – activated by signals from other infected cells as part of the early response to infection. Many of the proteins that provide this cell intrinsic immunity are expressed in response to signalling by a family of cytokines called interferons, in particular interferons alpha, beta and lambda. These antiviral genes are collectively known as interferon stimulated genes (ISGs). There are a large number of these genes (upwards of 300), but knowledge of what viruses they restrict and how they do it is limited, many genes have unknown functions.

In our recently published study we investigated the role of a specific ISG called IFITM1 (short for interferon induced transmembrane protein 1). IFITM1 is one of 3 IFITM proteins expressed by human cells and they appear to have a role in restricting early events in viral ifnection. We followed up previously published work, investigating where in the cell the protein was expressed and showed that unlike the better characterised IFITM3, IFITM1 was found on the plasma membrane, suggesting it prevents direct viral entry. We demonstrated that for a wide range of RNA viruses that infect the respiratory tract, including Respiratory Syncytial Virus, Influenza and Measles, increasing the level of IFITM1 in the cell reduced the level of viral infection. Interestingly we showed that IFITM1 was able to prevent infection with a virus with a DNA genome (Herpes Simplex Virus), suggesting IFITM1 function was associated with its cellular location rather than an effect on specific viral families. The importance of the location of the protein with in the cell was supported by studies that reduced the ability of IFITM1 to localise to the cell surface, leading to increased infection. These studies were supported by increased levels of infection in mice lacking the IFITM1 gene.

Understanding more about the function of interferon stimulated genes can help us to understand how viruses infect cells and may provide insight into strategies to prevent viral infections. In the case of IFITM1, we have shown that human cells make a robust anti-viral response at the cell surface and this can help to reduce viral infections.

What is up your nose?

We are interested in the contents of your nose, not at the level of hair, bogey and the occasional finger, but at the level of the complex microbial community that lives there and how it is associated with respiratory infection. The makeup of this community has been interrogated through sequencing (the airway microbiome) with indications that some bacterial communities may be associated with health and others with disease.

A bacterial diet

However, what the bacteria in the airways eat to survive is less well understood. One tool that may help us to characterise which biochemicals in the airways bacteria can use as food is called metabolomics. This uses liquid chromatography, to separate the biochemicals, and then mass spectrometry, to interrogate what they are. Comparing the mass spectrometry data against a curated library, we can then determine which individual biochemicals are present and their relative abundance. This tool has been used widely to investigate changes in the blood but has not been used much to interrogate the airway.

Blotting paper 2.0

The aim of our recently published study was to compare methods for sampling the airway metabolome. We looked at two standard techniques – nasal lavage (flushing a millilitre of saline through the nose and recovering whatever you can – quite a lot never comes back!) and induced sputum (getting people to breathe in an expectorant and then spit in a cup). We also used a newer technique, that had never been used for metabolomics sampling, called Synthetic Absorptive Matrix (SAM) strips. These are hi-tech blotting paper and have been used to recover other types of sample from the airways, including antibodies and cytokines. You can watch a video of their inventor having them put up his nose here. We looked at the use of these SAM strips in both the upper airway (via the nose) and the lower airway (via a bronchoscope). In the traditions of Barry Marshall (though I doubt I will get a Nobel prize for this), I volunteered to be one of the subjects for the sampling; the nasal wash, induced sputum and upper airway SAM were all fine, but having a bronchoscopy was fairly unpleasant.

It’s good to share

Having collected the samples, we then outsourced the running of the metabolomics to a company, called Metabolon in the US. This choice had mixed reviews, but I think it is ok to outsource, increasingly labs are outsourcing some of the more specialist analysis approaches – sequencing, transcriptomics, metabolomics. This makes sense in terms of time, expertise and access to equipment. Specifically in the case of metabolomics, outsourcing gave us access to a much larger curated library of samples, giving us more information from our samples, the biochemicals were also grouped into families, enabling us to interrogate the data more easily.

Sooo much data

From the point of performing the study to publishing it has been a lengthy process. In part this was due to the complexity of the dataset. We had approximately 14,000 data points – which may be small compared to some types of project, but when you are used performing focussed studies on individual mediators it was quite a step change. This was combined with a bewildering list of biochemicals, most of which we had never heard of – 1-stearoyl-2-arachidonyl-GPC anyone? In the end, through the power of the R programming platform and a very talented PhD student, we have ended up with a paper that uses a wide range of graph types, all of which aimed to compress the data into a meaningful form.

Food for the Pseuds

So what did we find? In total, 581 biochemicals were recovered from the airways belonging to a range of different families. When we compared the relative abundance of the these biochemicals between the different sampling techniques, we saw that the SAM strips gave us a much greater recovery of biochemicals than the other approaches. Since we were interested in how the airway metabolome enables bacterial colonisation, we screened some of these biochemicals for their ability to support bacterial growth. 35 of these biochemicals were able to support growth of the opportunistic airway bacteria Pseudomonas aeruginosa, including a number of sugars and amino acids.

A microcosm in a nostril

The airways represent a fascinating ecosystem because they are nutritionally more restricted in terms of the range and specific concentrations of any one biochemical compared to say the gut, but at the same time the nutrients are constantly refreshed. The balance of biochemicals in the airways shapes the bacteria that can live there, and we believe that this could be dysregulated in disease. By developing the tools to sample the airway metabolome, we are now one step closer to understanding how changes in airway biochemistry affects infection.

Buy one, get one free: Vaccinate the mother, protect the child

Flu vaccine the best way to protect

Infection with influenza virus, the causative agent of flu, is particularly severe in pregnant women and newborn children. If they do get infected, they are far more likely to get severe disease leading to hospitalisation. For the pregnant mother, the simplest approach to avoid this is to get the flu vaccine as soon as it becomes available. However, these vaccines are not licensed for children under 6 months of age – leading us to ask the question, how do we protect newborn children against influenza infection?

Maternal vaccination, protecting both mother and child

Luckily, the body has already come up with a solution. During pregnancy, mothers pass on immunity to their children. This passive protection is transferred in the form of antibodies, which are proteins made by the immune system that are highly specific for the molecules that make up the coats of viruses and can prevent the viruses from infecting our cells. This antibody transfer occurs in the second and third trimester of pregnancy and has evolved so that the newborn child has some early protection against whatever infections the mother has been exposed to. We can utilise this system with maternal vaccination. If we vaccinate the mother, she will make antibodies that recognise the virus in the vaccine and some of these antibodies will pass from her to her baby. This maternal immunisation approach has been seen to be very effective in reducing the burden of infection with tetanus and pertussis (whooping cough) in babies. Maternal immunisation has also been recommended as a method of reducing influenza infection in babies since 2005.

When is the best time to vaccinate?

One important question is when is the best point during pregnancy to vaccinate the mother to ensure the maximum transfer of antibody to the baby. It was originally thought that early in the third trimester (weeks 25-36 of pregnancy) was best as this was the peak of antibody transfer, but recent studies investigating pertussis vaccination of mothers saw higher levels in babies if the mothers were vaccinated in the second trimester (weeks 13-24 of pregnancy). We wanted to explore the best time to immunise mothers with influenza vaccine. In our latest paper, we measured the level of influenza virus specific antibodies in both mothers and babies at the time of birth. We compared babies born to mothers who were vaccinated in the first, second or third trimesters with babies born to unvaccinated mothers. We saw that there was significantly more influenza specific antibody in babies born to vaccinated mothers than in those born to unvaccinated mothers – demonstrating that maternal flu vaccination is highly effective at boosting the protection against influenza infection in the baby. We then investigated timing and observed that the high levels of antibody were seen in children born to mothers vaccinated in either the second or the third trimester, suggesting that either timepoint was equivalent, though there was less antibody transferred if the gap between vaccination and birth was less than four weeks.

Flu the ever changing

However, there is a complication with influenza virus; unlike the other pathogens for which maternal immunisation is recommended – pertussis and tetanus, the influenza virus changes. These changes in virus necessitate a new flu vaccine each year to match the viruses that are circulating. Flu is also seasonal – you are much more likely to get flu in winter months (in temperate climates). This seasonality had an effect on the levels of immune protection in our study: children born during the flu season had higher levels of antibody than those born outside it. 

The time is now

When we put the seasonality of influenza together with the best time to vaccinate mothers to pass antibody to children, we see that the current practice of offering flu vaccine to mothers as soon as it becomes available gives the best balance of protection to both mothers and their babies at the times when they need it most. This is because the flu season is 6 months long and pregnancy is nine months long. Whilst immunising mothers in the first trimester does not pass on the most antibody to the baby, immunising the mother at the start of the flu season gives the mother maximal protection for the whole flu season and they will give birth outside the flu season, so the baby requires less protection. Mothers who are in the second or third trimester at the start of the flu season will benefit from the protection of the vaccine themselves and pass antibody protection to their baby.

Therefore our study supports the current practice of offering influenza vaccine to mothers as soon as it becomes available.