A quantitative description of light-limited cyanobacterial growth using flux balance analysis

The metabolism of phototrophic cyanobacteria is an integral part of global biogeochemical cycles, and the capability of cyanobacteria to assimilate atmospheric CO2 into organic carbon has manifold potential applications for a sustainable biotechnology. To elucidate the properties of cyanobacterial metabolism and growth, computational reconstructions of genome-scale metabolic networks play an increasingly important role. Here, we present an updated reconstruction of the metabolic network of the cyanobacterium Synechocystis sp. PCC 6803 and its quantitative evaluation using flux balance analysis (FBA). To overcome limitations of conventional FBA, and to allow for the integration of experimental analyses, we develop a novel approach to describe light absorption and light utilization within the framework of FBA. Our approach incorporates photoinhibition and a variable quantum yield into the constraint-based description of light-limited phototrophic growth. We show that the resulting model is capable of predicting quantitative properties of cyanobacterial growth, including photosynthetic oxygen evolution and the ATP/NADPH ratio required for growth and cellular maintenance. Our approach retains the computational and conceptual simplicity of FBA and is readily applicable to other phototrophic microorganisms.


→ We have added another column in Table 1 and changed the format.
There are small typos, like "and and ranges from 1.84 to 1.97 gCDM/mol photons for the light-dependent BOF" in page 6, "was shown to less efficient" in page 11, "prepheneate" in page 22.
Although it is detailed that all the experimental data comes from Zavřel et al, I think that "Specifically, in addition to the O2 evolution as a function of the light intensity, the O2 consumption shortly after onset of darkness was measured" on page 9 still need to be referenced.
→ We added the reference after the sentence.I think it is a stylistic choice, but I would prefer "free parameter" rather than "adjustable parameter" (as I noted a change in page 10: "with KL as an additional free parameter").Other choices such as "GSRM" instead of "GSMM" can be settled by the editor.
→ We agree and changed all occurrence of "adjustable parameters" to "free parameters".
In page 10, values should match the order of the parameters in the text: "The estimated parameters for the terminal oxidase and the NGAM reaction are vmin OX = 9.1 ± 9.04 [%O2] and vmin NGAM = 1.45 ± 0.48 [mmol/gCDM/h], respectively" → We have changed the order.
Another choice is "Synechocystis 6803" rather than "Synechocystis sp.PCC 6803".The former is only used two times (page 10 and 22), while the second should be used throughout the manuscript.
→ We changed all occurrences to Synechocystis sp.PCC 6803.
On page 22, "N" is not defined.Usually, "S" is employed to denote the Stoichiometric matrix.
→ We now also define N ("FBA is based on knowledge of the stoichiometry of metabolism, as given by a stoichiometric matrix N, … ").We are aware that both N and S are used to denote the stoichiometric matrix.Here, we use N, following the works of Stefan Schuster, Reinhart Heinrich and others.Is it not better to combine Figure 6 and Figure S2 to improve the comparison of model simulations?Also, the description of Figure 6 reads "an offset of ≈ 12%" while Figure S2 says "deviation of 12%" (missing the "≈" sign) → We now have combined Figure 6 and S2 into a single figure (as Fig 6 in the main text).The description was changed to an "an offset of 12%" (the experimental expectation is approximately 12%, but the line is, of course, exactly 12%).
Is there a missing value in Figure 6 and Figure S2?Or oxygen exchange and dark respiration were not measured at 1100 µE/m2/s?Would these missing values affect the parameterization?→ Indeed, the oxygen exchange was not measured at all values for which growth was measured.However, the missing values have no (significant) effect on the parameterization (the precise numerical values will be slightly different, of course).In particular, the missing values follow the expectations given by measured values.This fact was also ascertained by comparing the values to earlier measurements (reproducibility of the setup was one of the major themes in reference [1] from which the data was obtained.The figure below is sourced from Figure 1 from reference [1]).For the current analysis, we preferred to stay within a single dataset.
It might be helpful to add markers to Figure4, showing exactly the tested light intensities.This was done for FigureS4and others.Similar comment for Figure 5. → We have changed the figures accordingly and now indicate measured light intensities with a bar.