Please visit the new BioModels platform to access the latest content. This website is no longer updated and will be retired on 20 July 2020.
BioModels Database logo

BioModels Database


Fung et al. (2005), Metabolator

October 2008, model of the month by Nick Juty
Original model: BIOMD0000000067

Biochemical oscillatory behaviour is well described and includes examples such as yeast glycolytic and calcium oscillations [1]. As we move into the realm of synthetic biology, where we can generate 'plug and play' components, there exists a desire to generate novel behavioural characteristics and control circuits with which to govern them [2]. One such interesting application of novel behaviour allowed cells, for example, to form biofilms in response to DNA-damaging agents [3].

In this work (BIOMD0000000067, [4]) the authors created a de novo circuit in Escherichia coli between two metabolite pools, Acetyl-CoA (M1) and Acetyl phosphate (AcP, M2). Acetyl phosphate itself is a signalling molecule that can induce the expression of acetyl-CoA synthetase (Acs), or through the action of LacI, repress phosphate acetyltransferase (Pta). Both the conceptual and the actually implemented model are shown in Figure 1.

conceptual metabolator metabolator realised

Figure 1: Conceptual (left) and realised (right) metabolator. The yellow boxes depict the 2 metabolite pools, M1 and M2. The grey arrows depict influx (sugars, fatty acids, glycerol), or outflux from the system from M1 (ethanol, TCA cycle) or M2 (acetate). Dashed (green lines - right side) with arrows represent positive transcriptional or translational regulation, while (red - right side) dashed line with blunt bar indicates negative regulation. Figure from [4].

An initial large input flux drives the accumulation of M2, which then represses Pta and upregulates Acs, driving the flux back to M1. A green fluorescent protein (GFP) was used to monitor cell status. Use of carboxy terminal degradation tags reduced the half-life of the proteins, including GFP, involved in the circuit.

Fluorescence in single cells

Figure 2: Arbitrary units depicting fluorescence in single cells at 30 degrees Centigrade. Insets show fluorescence at peak and trough values during oscillation. Red triangles represent cell division events. Figure from [4].

Model characterisation

Figure 3: Model characterisation. Increasing glycolytic rate (Vgly), left to right, increases the tendency for the system to oscillate. Figure from [4].

Oscillatory behaviour in different media

The oscillations were observed in individual cells in both solid and liquid media with glucose. The oscillations were not correlated with cell division times. Figure 2 shows fluorescence oscillations in single cells. A non-linear ODE model was created for further quantitative analysis, where enzyme kinetics were described by Michaelis-Menten, and gene expression by the Hill equation.

The model system operates at a stable steady state at low glycolytic flux, but was shown to have oscillatory tendencies that increased with glycolytic flux above a threshold, as shown in Figure 3. These oscillations were stabilised by the addition of high concentrations of acetate (see Figure 4, lower panel).

To verify model predictions, the glycolytic flux was varied experimentally by using the alternative carbon sources fructose, mannose and glycerol. As predicted, glycerol supplemented media did not support oscillatory behaviour in cells, and 10mM acetate suppressed any oscillations.

While pure transcriptional oscillators show a timescale in the order of hours, metabolic oscillators, based largely on allosteric regulation, show a timescale of around 5 mins. The fusion between the two, exemplified here, showed a period of around 45 mins.

Figure 4: The upper panel shows the apparent period (time between the first and second troughs) plotted against averaged values (of at least 3) of glycolytic flux. The dashed curve aims to plot the this correlation. Region I represents the stable steady state, and region II the oscillatory state. The lower panel shows that 10mM acetate suppressed oscillations, while 0.1mM did not. Figure from [4].

Bibliographic References

  1. A. Goldbeter. Biochemical Oscillations and Cellular Rhythms: The Molecular Bases of Periodic and Chaotic Behaviour. Cambridge Univ. Press, 1996.
  2. M.B. Elowitz, S. Leibler. A synthetic oscillatory network of transcriptional regulators. Nature 403:335-338, 2000.[SRS@EBI]
  3. H. Kobayashi, M. Kaern, M. Araki, K. Chung, T.S. Gardner, C.R. Cantor, J.J. Collins. Programmable cells: Interfacing natural and engineered gene networks. PNAS 101(22):8414-8419, 2004. [SRS@EBI]
  4. E. Fung, W.W. Wong, J.K. Suen, T. Bulter, S.G. Lee, J.C. Liao. A synthetic gene-metabolic oscillator. Nature 435:118-122, 2005. [SRS@EBI]