An Integrative Systems Biology approach to define the divergent kinetic responses of S. cerevisiae and C. albicans to amino acid starvation

Microbes must adapt to rapid changes in their environment if they are to survive these changes. For example, microbes must be able to adapt their metabolism to use the available nutrients, and they must adapt to nutrient limitation as these nutrients become exhausted. We are comparing how two different yeasts adapt to a particular type of nutrient limitation (amino acid starvation). We are studying bakers' yeast (Saccharomyces cerevisiae) because it is one of the best model organisms available, and because there is already a strong platform of knowledge about the amino acid starvation response in this yeast that has allowed us to build a mathematical model of this response. We are comparing bakers' yeast to Candida albicans because this is a medically important pathogen of humans that frequently causes infections in the mouth and vagina (thrush) and causes life-threatening bloodstream infections in intensive care patients. Clearly these yeasts have evolved in very different niches. Nevertheless we have shown that the pathogenic yeast C. albicans responds in roughly the same way as bakers' yeast to amino acid starvation. However, there are significant differences in the way their responses are regulated. Hence these yeasts appear to have retained a similar solution to the problem (they both make more amino acids via metabolism to overcome the shortage of amino acids), but there are differences in the control systems that regulate their adaptive responses. Both yeasts must respond rapidly to the initial nutrient starvation, but slowly turn off this starvation response as amino acids become available through metabolism. Therefore, the responses in these two yeasts must be effectively managed over time, even though their control systems differ. Our aim is to characterise these interesting differences because they will tell us about how such control systems have evolved in these yeasts. Our approach includes the building of a mathematical model that can describe the amino acid starvation response quantitatively, and that can accurately predict responses to novel experimental conditions. We have built a preliminary model. In this project we will optimise this model for bakers' yeast, and then build an equivalent model for the pathogenic yeast. These mathematical models will be very useful because they will allow us to rapidly simulate (on the computer) large numbers of experiments that are impractical to perform in the lab. This will allow us to focus our efforts in the laboratory on those experiments that are likely to be most interesting and informative. In this way we will characterise the differences between the control systems in these two yeasts. This will generate information about how these control systems have evolved, which will provide valuable messages about the evolution of microbial control systems in general.

We are comparing nutrient responses in model and pathogenic yeasts using an integrative systems biology approach because this will teach us about the evolution of kinetic systems in microbes. S. cerevisiae up-regulates most amino acid biosynthetic genes in response to amino acid starvation (the GCN Response). This is one of the best-characterised systems in yeast at the molecular level. We have built a quantitative mathematical model of this GCN system that reflects the limited kinetic data available. We have also studied the GCN response in Candida albicans (the major fungal pathogen of humans) because amino acid metabolism is differentially regulated in host niches, and because this response is mechanistically linked to morphogenesis (a known virulence attribute). Key regulators are conserved in C. albicans (Gcn2, Gcn4), but there are significant differences in their contributions to the regulation of this response. Therefore, we hypothesise that the general structure of the GCN System has been conserved in C. albicans and S. cerevisiae, but that the kinetic behaviour of the GCN System has diverged in these yeasts. Our overall aim is to test hypothesis this by comparing the kinetic behaviour of the GCN Systems in these pathogenic and benign yeasts. To achieve this we will: (a) optimise our S. cerevisiae GCN Model by experimentally defining the kinetic behaviour of this system during activation and down-regulation; (b) generate analogous datasets for C. albicans, and use them to build a C. albicans GCN Model; (c) test these models by comparing simulations with experimental data for specific S. cerevisiae and C. albicans mutants; and (d) determine the molecular basis for the divergent kinetic behaviours of the GCN Systems in these yeasts. This will provide major new insights into the evolution of the kinetic systems that drive adaptive responses in microbes. All of the technologies required for this project are in place in our labs.