Metabolic modeling of Chlamydomonas reinhardtii: energy requirements for photoautotrophic growth and maintenance

Abstract

In this study, a metabolic network describing the primary metabolism of Chlamydomonas reinhardtii was constructed. By performing chemostat experiments at different growth rates, energy parameters for maintenance and biomass formation were determined. The chemostats were run at low irradiances resulting in a high biomass yield on light of 1.25 g mol(-1). The ATP requirement for biomass formation from biopolymers (K(x)) was determined to be 109 mmol g(-1) (18.9 mol mol(-1)) and the maintenance requirement (m(ATP)) was determined to be 2.85 mmol g(-1) h(-1). With these energy requirements included in the metabolic network, the network accurately describes the primary metabolism of C. reinhardtii and can be used for modeling of C. reinhardtii growth and metabolism. Simulations confirmed that cultivating microalgae at low growth rates is unfavorable because of the high maintenance requirements which result in low biomass yields. At high light supply rates, biomass yields will decrease due to light saturation effects. Thus, to optimize biomass yield on light energy in photobioreactors, an optimum between low and high light supply rates should be found. These simulations show that metabolic flux analysis can be used as a tool to gain insight into the metabolism of algae and ultimately can be used for the maximization of algal biomass and product yield. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1007/s10811-011-9674-3) contains supplementary material, which is available to authorized users.

Fig. 1

Schematic front and side view of the photobioreactor setup (not on scale) for the chemostat experiments. MFC mass flow controller for both air and CO₂, WB water bath, F air filter

Fig. 2

Overview of an algal cell in the light, showing the main metabolic processes. Only the chloroplast and the cytosol were modeled as cell compartments; therefore, mitochondrial processes such as respiration through the electron transport chain were placed in the cytosol. Light is fixed in the chloroplast, yielding O₂, ATP, and NADPH. These are needed for the fixation of carbon dioxide in the Calvin cycle into glyceraldehyde 3-phosphate (GAP). GAP can be transported to the cytosol to be converted into building blocks for biomass. Lipids are formed through glycolysis and the tricarboxylic acid (TCA) cycle. Nitrate is taken up by the cell and converted into glutamate which in turn can be converted to protein and chlorophyll. GAP can be converted to glucose 6-phosphate (G6P) from which carbohydrates are formed. G6P can also enter the pentose phosphate pathway (PPP) which yields NADPH, DNA and RNA. Electrons are carried by NADH and FADH₂ to the mitochondrial electron transport chain, yielding ATP by taking up O₂

Fig. 3

Plot of the specific calculated overall ATP production rate q _ATP against the experimentally determined growth rate. q _ATP was calculated with the model for each growth rate. Regression through these points yields a straight line of which the offset gives the ATP required for maintenance m _ATP, 2.85 mmol g⁻¹ h⁻¹. The slope gives the constant which represents the additional amount of ATP needed to make biomass from the biopolymers (K _x), 109 mmol g⁻¹, according to Eq. 4. The error bars represent the minimum and maximum values for q _ATP and the growth rate μ, which follow from the relative errors of biomass measurements

Fig. 4

Simulated respiration rates (mmol O₂ g⁻¹ h⁻¹) at several growth rates (μ, h⁻¹). Regression through these points yields a straight line with an intercept at μ = 0 of 0.53 mmol O₂ g⁻¹ h⁻¹. This value is the oxygen uptake rate through respiration which is necessary for maintenance purposes. The maintenance fraction of the respiration rate is also plotted and shows which part of respiration is used for maintenance purposes. This graph also shows the simulated biomass yields on light energy (Y _xE, g mol⁻¹) for these growth rates. The simulations give an ideal situation since light saturation is not modeled. If light saturation would be taken into account the yield would reach an optimum and decrease as soon as light saturation occurs