Searching for principles of microbial physiology

Abstract

Why do evolutionarily distinct microorganisms display similar physiological behaviours? Why are transitions from high-ATP yield to low(er)-ATP yield metabolisms so widespread across species? Why is fast growth generally accompanied with low stress tolerance? Do these regularities occur because most microbial species are subject to the same selective pressures and physicochemical constraints? If so, a broadly-applicable theory might be developed that predicts common microbiological behaviours. Microbial systems biologists have been working out the contours of this theory for the last two decades, guided by experimental data. At its foundations lie basic principles from evolutionary biology, enzyme biochemistry, metabolism, cell composition and steady-state growth. The theory makes predictions about fitness costs and benefits of protein expression, physicochemical constraints on cell growth and characteristics of optimal metabolisms that maximise growth rate. Comparisons of the theory with experimental data indicates that microorganisms often aim for maximisation of growth rate, also in the presence of stresses; they often express optimal metabolisms and metabolic proteins at optimal concentrations. This review explains the current status of the theory for microbiologists; its roots, predictions, experimental evidence and future directions.

Keywords: biophysics; constrained biosynthetic resource allocation; evolutionary biology; mathematical modelling; microbial physiology; systems biology.

Figure 1.

Experimental illustration of balanced growth. A fluorescent-protein expressing B. subtilis strain was grown in mineral medium on glucose in shake flask. Samples of it were measured in a flow cytometer. This data was reproduced from Nordholt et al. (2017).

Figure 2.

Cellular compartments have finite protein storage capacities. In the theory, protein compete for biosynthetic resources, like RNA polymerases, sigma factors, nucleic acids, ribosomes, amino acids, etc. and space, as shown in this figure.

Figure 3.

Illustration of the fitness cost (growth-rate reduction) due to the expression of an unneeded enzyme. The black line illustrates a fit with slope −2.7, indicating that growth rate is zero at an unneeded protein expression of 37% (Bentley et al. ; Dong, Nilsson and Kurland ; Snoep et al. ; Scott et al. 2010C).

Figure 4.

Illustration of enzyme titration and optimal protein-expression by a wild type L. lactis (A) and E. coli (B) strain. A. Three glycolytic enzymes and an operon display optimal expression levels in L. lactis (Koebmann et al. ; Koebmann Solem and Jensen ; Koebmann, Solem and Jensen ; Solem et al. ; Solem, Koebmann and Jensen 2008). B. H⁺-ATPase of E. coli is optimally expressed in two growth environments (Jensen, Michelsen and Westerhoff 1993).

Figure 5.

Linear relation between ribosomal activities and growth rate in E. coli and S. cerevisiae. Data is from Scott et al. (2010) and Metzl-Raz et al. (2017).

Figure 6.

Illustration of a stoichiometric model of metabolism and enzyme kinetics. The fermentation of glucose into ethanol via glycolysis, as it occurs in S. cerevisiae, is shown as an example reaction network, together with its formulation in terms of stoichiometric matrix. Each metabolite is denoted by the number of carbon atoms it contains. An example of an enzyme rate equation is also shown. This network has 15 reactions, 18 intracellular, variable metabolites concentrations and 4 conservation relations of chemical moieties, i.e. total nicotinamide adenine dinucleotide (NAD), total phosphate (P), total adenosine (A) and electrons (e.g. AP₂ + AP₃ = constant = A_total). Therefore, 18 steady-state flux relationships exist, deriving from the requirement that the 18 concentrations are at steady state. Four of these relationships are redundant—they are linear combination of the remaining 14, due to the moiety conservation. Thus, 15 unknown reaction rates and 14 linear relationships between them exist in this network. This we means that we need to known only 1 value to determine all values—thus this network is an elementary flux mode. Say we know the value of reaction rate 5 then all there equal the values shown in the steady-state flux vector.

Figure 7.

Elementary flux modes can be complicated metabolic networks, including branches and cycles. A schematic overview of the aerobic respiration of glucose into carbon dioxide and water is shown by glycolysis and the citric acid cycle. This network has 27 reactants, 23 reactions and 5 conservation relationships of chemical moieties (i.e. of CoA (B), ubiquinone (U), nicotinamide adenine nucleotide (N), phosphate (P) and adenosine (A)). Thus, we have 22 (linearly independent) relationships between reaction rates and 23 unknown reactions; thus, we need to 1 flux value to determine them all. This illustrates that this network is an elementary flux mode.

Figure 8.

Illustration of common microbial physiology: overflow metabolism occurs after a critical growth rate and linear flux-growth rate relations. Values of the glucose (A) and oxygen (B) uptake fluxes in glucose-limited chemostats from different studies (Postma et al.; Holms ; Van Hoek, Van Dijken and Pronk ; Nanchen, Schicker and Sauer ; Fonseca et al. 2007). The figures shows on the right are examples, shown also in the left figure.

Figure 9.

Occurrence of linear relations between flux and growth rate and a critical dilution after which fermentation starts in a glucose-limited chemostat of S. cerevisiae (Van Hoek et al. 1998). The dots are the experimental data and the lines are linear fits.

Figure 10.

The simplest, biological self-fabricator. For a cell to grow, it needs to make its own components out of nutrients, which it takes up from its environment. For it to grow at a constant rate, it needs to make all its components in constant proportions, such that their concentrations remain constant and, therefore all biosynthetic fluxes. Then, self-fabrication occurs as balanced growth. The nutrient is the growth substrate, is a metabolite, is a membrane-embedded uptake enzyme, and is the ribosome that makes and itself. The regulation of the synthesis of the and are regulated by the concentration of (Price et al. 2019).