Under Challenge 1 we developed novel synthetic cooperative systems using the rice blast fungus Magnaporthe oryzae and Saccharomyces cerevisiae. In combination with mathematical modelling we generated the following discoveries:
1. We hypothesized that while a monomorphic, co-operator only, population maximises virulence in a single-trait cooperative system, a mixture of co-operators and cheats maximises population virulence in a multi-trait system. We reasoned that alongside the first social dilemma, namely public goods production, M.oryzae faces a second dilemma of “tragedy of the commons”. Our mathematical model supported this hypothesis and the theoretical results were subsequently verified experimentally. The results were published in Lindsay et al 2016 eLife, demonstrating their broader significance for anti-virulence disease management strategies.
2. We have classified the impact of variation in population density and spatial structure on the frequency of cooperation in multi-trait cooperative systems. We provided the first experimental evidence that high population density can support co-operators in spatially structured environments. We also showed that prior empirical procedures did not capture the extent of spatial structuring required, in theory, to support cooperation. The work was published in Lindsey et al 2017 ISME J and is central to our understanding of ecological and epidemiological processes from nutrient recycling to antibiotic resistance.
3. Pathogens damage their host when they consume nutrients during infection, so understanding how differing pathogen metabolic strategies influence virulence evolution is crucial to effectively manage disease. To this end we generated a library of M. oryzae strains with a range of invertase expression levels and thus different growth rates. Our in vitro and in vivo competition experiments found that faster growers were promoted when the environment was structured. However, we have also found that growth rate was not necessarily the primary determinant of disease virulence, as theory classically assumes. This study is currently in preparation for publication.
4. We have generated a synthetic community of S. cerevisiae strains that differ in the way they metabolise sucrose. Public-metabolizers digest resources externally while private-metabolizers internalize resources before digestion. We found that privatization of public goods could give strains a competitive advantage so that they could dominate mixed-strategy populations. However, owing to a reduced growth rate, once dominant in the population they left the community prone to population collapse in ephemeral environments. This study was published in Lindsay et al 2019, Nature E&E.
The work under Challenge 2 combines mathematical modelling and Candida species in vitro experiments and has led to the following scientific discoveries:
5. Our pilot work has shown that single species dose response is a poor predictor of multi-species community dynamics because it cannot foresee the tipping points that cause irreversible changes in resistance that persists, even when treatment stops. These results have been published in Beardmore et al 2018 Nature E&E and Reding-Roman et al 2017 Nature E&E.
6. We examined whether rapid or slow fluctuation in nutrient and antimicrobial concentrations select for, or against, resistance. To this end we studied dynamics of two Candida species , one sensitive and the other resistant to an antibiotic drug. We discovered a fundamental difference between resistance in single species populations, the context in which it is usually assayed, and in communities. While fast environmental changes are known to select against resistance in single-species populations, they can promote the resistant species in mixed-species communities. Our findings were published in Nev et al 2020 R. Soc. Interface.
7. We investigated why current models describing microbial growth frequently fail to capture growth dynamics at different nutrient concentrations. We observed that the maximal nutrient uptake rate is a decreasing function of the nutrient concentration and derived the explicit form of this relationship. We then propose an approach that allows us to accurately predict microbial growth and competition outcomes for a range of nutrient concentrations. Our findings were published in Nev et al 2021, PloS Comp Biol.
8. We considered how is the outcome of resource competition determined? We found that classical approaches that use short-term competition measures may be heavily time dependent, especially in the presence of environmental stressors and as such could not reliably be deployed to predict long-term microbial competition dynamics (Jepson et al in revision).
