Building and dismantling polymicrobial pathogen communities in cystic fibrosis.

People whose immune defences are weakened are easy targets for infection by bacteria: they can suffer from chronic infections that last for months or even years. For example, people with the genetic disorder cystic fibrosis (CF) pick up lung infections which they cannot clear because their airways are blocked by sticky mucus. 90% of people with CF die from lung failure as a result, 50% of them before the age of 41. People with chronic obstructive pulmonary disease, HIV/AIDS or asthma, and hospital patients on ventilators, are also very vulnerable to chronic lung infection. These infections are typically resistant to many antibiotics, and it can be hard for doctors to predict which antibiotics could help a particular patient.

We have only recently understood that chronic infections are not simple infections of one type of microbe - rather, they are complex systems where different species of bacteria, fungi and viruses live together and interact inside the infected organ. The interactions are probably what make chronic infections so unpleasant and so hard to treat with antibiotics. Bacteria are especially interactive and sociable. Cells of the same and different species can glue themselves together in sticky biofilms that protect them from stress, immune clearance and antibiotics, and they can communicate with one another using chemical signals to co-ordinate the release of molecules that damage the host tissues.

There is another layer of complexity in chronic infection: the long time periods that infections persist for, and the changing environment inside the host due to medical treatment and tissue damage, mean that pathogens evolve. They can adapt to become better at living inside the host and resisting antibiotic treatment, and a clonal founding population of any given species can become very genetically diverse as it spreads into different areas or "niches" inside the infected organ.

This complexity and constant change means that it is very difficult to treat chronic infection. Antibiotic resistance can be hard to recognise when it is a function of two or more bacterial species interacting. Also, if we try to kill one species with an antibiotic, we do not know how the other species in the community will respond: more dangerous bacteria might be able to grow better if their competitors are removed. This probably explains why standard lab tests of antibiotic susceptibility of bacteria from patients - which are conducted only on one bacterial type from a patient - are not very good at predicting which antibiotics will help when they are prescribed.

People have tried to understand how chronic infection ecology drives antibiotic resistance and infection severity in samples tested in a lab by growing the bacteria in a test tube or on an agar plate, but these are not at all like a complex organ, like a lung. Tests have also been performed on mice and insects but these are very different from human lungs and so are of limited value (and there are ethical concerns about using live animals in this way).

I have developed a more realistic and ethical way of mimicking chronic lung infection using lungs from pigs killed for meat and donated human lungs from the NHS biobank. I use culture techniques that closely simulate chronic infection in people with CF. I will use this system to study how bacterial species that commonly cause chronic lung infection (in CF and other conditions), interact with one another to produce highly antibiotic-resistant, persistent infection. I will also develop my model into a new tool for conducting better, more predictive tests of which antibiotics, or which novel therapies, might work in patients.

Chronic infections are caused by diverse communities of microbes that inhabit spatially-structured host environments. Inter-microbe interactions mediate virulence, persistence and antimicrobial resistance (AMR). Thus, improved diagnostic and R&D tests for the likely clinical impact of antimicrobial compounds require testing platforms that recapitulate diverse infection communities in structured tissue. Further, future treatments for chronic infection (e.g. bacteriophage) must be tested in such models to gain reliable predictions of their clinical potential.

I have developed a clinically valid, high-throughput model of polymicrobial infections of the lungs of people with cystic fibrosis (CF): ex vivo lung tissue cultured in artificial mucus. I will characterise infection communities in this model, revealing how biodiversity mediates AMR and virulence. I will study clinical isolates of key lung pathogens (P. aeruginosa, S. aureus, B. cenocepacia, H. influenzae, S. milleri + whole CF sputum samples) in single and polymicrobial infections. I will assess how the model could be developed into a clinically predictive platform for AMR profiling and the activity of novel antimicrobial/antivirulence drugs. I will use hypothesis-driven experiments about microbial communities in structured tissue to reveal key drivers of AMR in CF infection, the likely triggers of acute pulmonary exacerbation, and how these may be neutralised.

My model uses pig lung tissue from a commercial butcher and human lung from an NHS biobank. I will define how well pig lung recapitulates human tissue infections and promote its wider adoption in academia and industry to increase the clinical validity and 3Rs compliance of basic research and R&D testing. Through consultancy work with colleagues, I will explore the potential for ex vivo lung to underpin tests of novel therapeutic agents (antibiotics and phage) on realistic infection communities in structured tissue.

STRATEGIC IMPACT
The project will deliver a clinically valid, high throughput model for studying how microbial interactions drive pathogenesis and antimicrobial resistance (AMR), and for testing antimicrobial susceptibility. This meets the MRC's goal of securing impact from research by maximising the translatability of lab studies. It also supports the cross-council initiative on tackling AMR, specifically the aims of "Understanding resistant bacteria" and "Understanding real world interactions". Beneficiaries include patients; academic, clinical and industry researchers; project staff and the general public.

BENEFITTING PATIENTS BY PROVIDING A NEW RESEARCH TOOL
The ultimate beneficiaries (beyond the lifetime of the grant) will be people with chronic infections. During the grant, I will liaise with the James Lind Alliance to align current and future work with their recommended priorities for cystic fibrosis research. By the end of the grant, we will have determined whether antibiotic resistance profiling of infection isolates could be made more predictive by using ex vivo lung to test polymicrobial samples, and assessed the potential of lung models for testing phage therapy. We will work with Warwick Ventures to identify routes to future commericialisation, validation and regulation of lung models for use in routine clinical testing (health service labs) and in testing pipelines for new infection treatments (academia & industry).

3RS IMPACT
Ex vivo lung models will reduce the number of animals used in infection research, saving time and money and helping researchers meet national aims to Reduce, Refine and Replace animals. Studies of basic in-host microbiology can be carried out in ex vivo lung instead of live rodents, and new clinical interventions can be trialed to rule out cytotoxic/ineffective treatments and narrow down experimental variables (e.g. dose) before moving in vivo. Of 25 P. aeruginosa mouse lung infection papers, 36% dealt with questions that we can investigate in ex vivo lung. These used a median of 36 mice, implying that annually, c.1300 mice are used in experiments that may instead be carried out in ex vivo lung. This number will be larger if unpublished pilot work and work on other bacteria are considered. This aspect of the project is an important societal impact, because the British public feel strongly about this area. A recent poll found that support for animal research is conditional on the lack of alternatives, that 78% of respondents wanted more research into alternatives and that people are interested in hearing about such research (http://bit.ly/1SWdtXN). We will track & publish the uptake and 3Rs impact of the project over the lifetime of the grant (and reassess these at intervals beyond the lifetime of the grant).

STAFF
The post-doc and technician will gain technical (molecular microbiology, genomics, imaging) and general professional (presentation, writing, statistics, communication, management of students or junior staff) skills that are transferrable to careers in academia, industry or policy, benefitting themselves and the economy. They will benefit immediately after the grant by moving to new employment.

PUBLIC
Besides the general benefits of providing dialogue with researchers and higher education outreach, talking about this project will help show that scientists are working to solve AMR and chronic infection - problems that touch many people's lives. It will also show that scientists are actively researching alternatives to animal experiments. During the grant, we will take part in various activities as outlined in the Pathways to Impact, including working in science centres/museums, science cafes and schools. We will engage with national news media by working with MRC/university press officers and using contacts from my previous media experience (e.g. work with BBC Radio 4). We will specifically engage the cystic fibrosis community by liaising with CFUnite (cfunite.org)