Systems biology of lactic acid bacteria: a critical review

Abstract

Understanding the properties of a system as emerging from the interaction of well described parts is the most important goal of Systems Biology. Although in the practice of Lactic Acid Bacteria (LAB) physiology we most often think of the parts as the proteins and metabolites, a wider interpretation of what a part is can be useful. For example, different strains or species can be the parts of a community, or we could study only the chemical reactions as the parts of metabolism (and forgetting about the enzymes that catalyze them), as is done in flux balance analysis. As long as we have some understanding of the properties of these parts, we can investigate whether their interaction leads to novel or unanticipated behaviour of the system that they constitute. There has been a tendency in the Systems Biology community to think that the collection and integration of data should continue ad infinitum, or that we will otherwise not be able to understand the systems that we study in their details. However, it may sometimes be useful to take a step back and consider whether the knowledge that we already have may not explain the system behaviour that we find so intriguing. Reasoning about systems can be difficult, and may require the application of mathematical techniques. The reward is sometimes the realization of unexpected conclusions, or in the worst case, that we still do not know enough details of the parts, or of the interactions between them. We will discuss a number of cases, with a focus on LAB-related work, where a typical systems approach has brought new knowledge or perspective, often counterintuitive, and clashing with conclusions from simpler approaches. Also novel types of testable hypotheses may be generated by the systems approach, which we will illustrate. Finally we will give an outlook on the fields of research where the systems approach may point the way for the near future.

Figure 1

Primary metabolism of L. lactis with major players discussed in the main text. Indication of the (positive) regulatory feedback and feedforward loops that involve F16bP in dashed orange line. In red are the enzymes of the las-operon. In green boxes the PTS system and GAPDH, respectively. G6P and PEP pool indicate pools of intermediates that are considered in rapid equilibrium.

Figure 2

Illustration of the difference between measuring flux in systems without and with regulatory constraints. The flux in a pathway with two irreversible Michaelis-Menten enzymes was calculated. The amount of the first enzyme e₁ was varied either independently of e₂, as it should be to measure its control coefficient, or in a system in which the total amount of enzyme e₁ + e₂ is constrained. In the latter system an optimal distribution of the enzymes is observed at e₁ = 0.4. In the neighborhood of that optimum e₁ has no apparent control on the flux (nor does e₂). Rate equations and parameter values used: v₁ = k₁e₁(S / K_S)/(1 + (S / K_S) + (X / K_SK_IX))v₂ = k₂e₂X / (K_X + X) k₁ = 2, k₂ = 1, K_S = 1, K_IX = 5, K_X = 2, and S = 5.

Figure 3

Peptide cross-feeding between Prt⁺ and Prt^- strains in a mixed culture of both variants. The extra cellular protease of Prt⁺ strains cleaves casein into peptides that diffuse away from the cell and can be utilized by invading Prt^- cheating cells (Panel a). Modelling the dynamics between the two variant strains shows that at low cell densities and low Prt⁺ frequencies the fractional gain after one propagation step is high for Prt⁺ cells. If cultures are grown (inoculated) at high cell desities and high frequencies of Prt^- cells the fractional gain of Prt⁺ cells is negative indicating that they are outcompeted by Prt^- cells (Panel b). Reproduced from Bachmann et. al. 2010 with permission from the publisher.