Limits to life : what controls maximum growth rates? : A proteomic analysis of Bacillus cereus sensu lato across multiple temperatures

Microbiological growth affects many facets of human life. Understanding factors that influence the growth rate of microorganisms, i.e. how fast an organism can reproduce under specific circumstances, is integral to fields such as health and disease, food safety, engineering, ecology and even climate science. Temperature-dependent growth rate is of particular importance because temperature is the most variable factor in many situations (e.g. the temperature at which food is kept), or because temperatures are changing on a global scale and there is a pressing need to understand how microorganisms will respond to that (e.g. whether algae grow faster under higher ocean temperatures). Accordingly, there exists a significant body of research on how to best define and model the temperature-dependent growth rate of an organism. This study explores the biochemical basis of one of a group of models known as ‘thermodynamic models’.

Thermodynamic models posit the existence of a biological "master reaction system" (MRS); a single rate-limiting enzyme reaction whose function is integral to the growth of the organism and whose rate of reaction determines the upper limit of the growth rate. Accordingly, any factor(s) that change that reaction rate will directly alter the growth rate.

All enzymes have a range of temperatures over which they function nearly optimally, beyond which they begin to deviate from their optimum conformation and, thereby, lose catalytic efficiency. There are four temperature-related mechanisms potentially involved in the rate of the master reaction. First, there are the reaction kinetics within the cell, which determine how often the enzyme comes in contact with its substrate. Second is the rate at which the enzyme is produced and replaced, as higher concentrations of the active form of enzyme within the cell will increase the overall rate of reaction. Third is the stability of the enzyme and its ability to resist denaturation, both reversible and irreversible at specific temperatures, which determines the upper and lower temperatures at which the enzyme remains active. Fourth is the presence or absence of isoenzymes; enzymes with the same biochemical function as the master reaction enzyme, but with sequence variations that enable them to function better at higher or lower temperatures. At temperatures where an organism is exposed to temperature stress, either hot or cold, the interplay between these four factors will determine the rate of the growth of the organism.

This thesis employs the thermodynamic model of Corkrey et al (2012) for which the range of temperatures over which the enzyme retains conformational functionality is termed the “thermal stability range”. The midpoint of this range is T_mes, the temperature of maximum enzyme stability. The upper limit of the thermal stability range, before heat causes significant (>10%) denaturation, is termed the upper temperature of maximum enzyme stability, or U_mes. Similarly, Lmes is used to describe the lower temperature of maximum (i.e. ≥90%) enzyme conformational stability.

U_mes is of particular interest, because U_mes is assumed to be the temperature at which the growth rate of an organism is at its highest efficiency. However, U_mes is always lower than the temperature at which the growth rate of an organism is fastest (T_opt) but, at U_mes, the temperature stress on the organism is less.

While the current literature describes the relationship between the MRS and the rate of growth, it has not identified what the master reaction system could be; that, however, is the goal of this thesis. It is hypothesised that U_mes will have lower levels of enzymes associated with stress and protein turnover than temperatures outside of the thermal stability range, coupled with an associated increase in anabolic activity linked to the growth rate at U_mes. In this thesis, proteomic analyses of a group of closely genetically related bacteria (with different temperature adaptations) was used to identify enzymes that differed between the bacteria, to help identify a putative master reaction. The core proteome, i.e. the set of proteins common to all of the bacteria studied, was compared between three temperatures, each of which was chosen to represent of a different level of temperature-based stress.

The organisms examined are a clade of the endospore forming bacteria Bacillus cereus, a pathogen and food spoilage organism that is found in a diverse range of environments and with diverse ecologies. The B. cereus group are effectively the same species with minor differences, but have been given different species names to recognise the diversity of ecological niches they occupy. Collectively, this group is known as B. cereus sensu lato. This clade was chosen for their genetic similarity and because their temperature-dependent growth rates behave predictably across the clade, effectively forming a continuum despite their different preferred temperature ranges. The maximum growth rate of any member of the clade at any specific temperature can be modelled with a single model, as if they were a single organism.

The B. cereus isolates were sourced from prior research, having been sampled from a selection of Tasmanian dairy farms and ricotta cheese. The isolates are not strictly a random sample, having been isolated from similar environments in a single region of the world. However, the isolates were sampled from a range of environmental sites, e.g. soil, cow’s teats, milking equipment, and dairy products, each of which are exposed to different ranges of temperatures. The subset of isolates used for this study were selected for the temperature adaptation they showed. The isolates used were also selected for genetic variation, possessing Average Nucleotide Identity (ANI) similarity values between 85% and 95%, ensuring that all isolates were genetically comparable, but without being duplicates of the same isolate.

To test the hypothesis, several approaches were used. The selected B. cereus isolates were cultured at three temperature points: each isolate's L_mes, each isolate's U_mes, and a third temperature, above the optimum temperature for growth rate of each isolate, which was termed T_sup. Proteins were extracted from each isolate after growth at each temperature point, and label-free quantitative (LFQ) proteomics was used to identify proteins and detect changes in their relative abundance. A number of statistical tools, including cluster analysis and functional annotation were used to determine which proteins differed between the isolates over the three temperature points, as well as protein functions that were associated with temperature adaptation.

When the isolates grown at their respective U_mes were compared, some isolates were found to cluster together. The differences between the isolates were examined using functional annotation and the clusters were found to mainly differ in relative abundance of proteins associated with transcription and DNA replication, as well as proteins associated with general cell metabolism and pyruvate synthesis. It was determined that the isolates with higher relative levels of these transcriptomic and metabolic proteins on average grew faster, and on average had higher U_mes values. It was also determined that the biological diversity among isolates was most closely correlated with each isolate's individual U_mes, and the source of the isolates. As noted, there is a known link between the environmental source of each isolate and the range of temperatures which it might be expected to experience. The results imply a further link between the thermal stress an isolate is exposed to, its ecological niche and its growth rate.

The analysis of the proteomes from isolates grown at their U_mes provided a baseline on the natural variation between the isolates that was then compared to proteomes of the respective isolates grown at their Lmes and Tsup. The clusters observed at U_mes were not present at either L_mes or T_sup, indicating that the U_mes clusters were not good predictors of trends in proteome responses at the other temperatures. Comparison of the relative protein abundances for Lmes and U_mes samples demonstrated that the Lmes samples had comparatively higher levels of ribosomal proteins, as well as proteins responsible for ribosomal subunit assembly and tRNA formation and modification. On the other hand, the functions of the predominant proteins that were increased at U_mes over Lmes were primarily related to cellular energy metabolism, particularly relating to pyruvate and acetyl-CoA synthesis, glycolysis, and the TCA cycle. When the proteomes of isolates grown at Tsup were compared to those at U_mes, the U_mes samples showed comparatively increased carbon metabolic proteins, as well as a TCA cycle proteins and ribosomal proteins.

See Notes field below for the remainder of the abstract.