Staphylococcus aureus is an aggressive pathogen, responsible for an array of infections that can affect virtually any organ or tissue within the human body. This is enabled by a raft of virulence factors, almost all of which are controlled by a single quorum-sensing system known as Agr. One of the mysteries of staphylococcal biology is that invasive and persistent infections are associated with strains that possess decreased Agr function. This begs the question: “Why are the most serious infections caused by strains with the weakest production of virulence factors?” Not one to shy away from a challenge, this is what I decided to try and answer when I began as a PI in the MRC CMBI at Imperial College London, initially teaming up with fellow CMBI member Ramesh Wigneshweraraj, who was developing a keen interest in Agr regulation.
Persistent bacterial infection implies failure of both the host immune system and antibiotic therapy to clear pathogens from the tissues. Therefore, I decided to study how Agr activity influenced staphylococcal survival of host defences and antibiotics. The first PhD student to join my lab in this endeavour was Kimberley Painter, who made a panel of agr mutants and started to investigate how these survived exposure to host defences. All I needed then was someone to look at Agr and antibiotics. Fortunately I had the perfect opportunity to get this project up and running when Vera arrived at the lab in October 2014. Vera had done an MRes project rotation with me the year before and was obsessed with quorum-sensing. When she asked if she could come back and do a PhD in my lab, all of the pieces fell into place.
Using Kim’s agr-mutants, Vera began to look at whether Agr activity influenced susceptibility to various antibiotics. It was mostly uninteresting until she looked at daptomycin – one of the very few drugs available to combat serious MRSA infections. I still remember looking at her agar plates from those first assays in amazement. As expected, wild-type bacteria were rapidly killed by the antibiotic but the agr mutants survived! If I’m honest, the difference in survival was so great that I didn’t believe it at first; but, several replicates were in agreement and we saw the same phenotype in another genetic background. Game on.
To our great surprise, Vera subsequently discovered that agr-mutants inactivated daptomycin, and we started to think about the mechanism responsible. “Could it be a lipid?” Vera asked. “It’s definitely not a lipid” I responded. “Antibiotics are inactivated by proteins”. Fortunately, Vera had by then learnt a crucial lesson that comes to all PhD students sooner or later: supervisors can be wrong!
So, Vera continued to plug away figuring out what was inactivating daptomycin. She ruled out an enzyme and then showed that phospholipase prevented daptomycin inactivation. She also showed that lipid was present in culture supernatants from bacteria exposed to the antibiotic and that this bound to, and inactivated, daptomycin. No getting away from it, I was wrong, it was a lipid! Reinforcements then came in the form of Sanika, an MRes student who found additional evidence that S. aureus releases phospholipids in response to daptomycin and that these inactivate the antibiotic. Tom Clarke, a Wellcome Trust/Royal Society Sir Henry Dale Fellow was also roped in to test whether loss of agr activity could lead to daptomycin treatment failure - it did.
Feeling a little bruised by my error over the lipid, but eager to play my part in this emerging story, I started to think about the second part of the puzzle – why was Agr activity making S. aureus susceptible to daptomycin? Vera had already figured out it was something small, secreted and heat sensitive but progress had started to slow. However, as is so often the case, the answer was right in front of me. Figure 1 in a Trends in Microbiology article that Kim and I had written with Ramesh and his student Aisha the previous year was a diagram of the Agr system, highlighting its direct regulation of PSM toxins. PSMs are small, secreted peptides with surfactant activity – could these be responsible for the sensitivity of Agr-functional strains to daptomycin?
We were able to rapidly test this hypothesis thanks to the generosity of investigators from the US (Michael Otto) and the Netherlands (Jan Maarten Van Dijl), who each sent us their panels of psm mutants. Vera tested them and, to my enormous relief, the psmα mutants had an identical phenotype to the agr mutants. Bingo! Follow-up studies with synthesised PSM peptides confirmed our finding – PSMα peptides prevented the phospholipid from binding to and inactivating daptomycin. PI pride restored.
Looking back, I think I learnt as much about PhD supervision during the project as I did about daptomycin. Supervisors frequently tell their students that they are the experts on their projects. This supervisor will be listening to his team of experts a little more closely in the future….