Design and analysis of synthetic carbon fixation pathways

Introduction

The process by which carbon dioxide is incorporated into organic compounds is carbon fixation.[1] The existing plant carbon fixation has a main enzyme, Rubisco, but it is known that is a slow catalyst, it assimilates only five or ten CO2 molecules per second.[2] One of the assignments that the authors of “Design and analysis of synthetic carbon fixation pathways”, have given to themselves, is to compare all the known carbon fixation enzymes in the Kyoto Encyclopedia of Genes and Genomes database.[3] And after that, by combination of these known metabolic building blocks, to build synthetically carbon fixation pathways, computationally, and compare their efficiency, kinetics, thermodynamics and topology, with the ones that are already occurring naturally.

Criteria and methods

Up to date, six natural metabolic carbon fixation pathways are known[4]:

- Reductive pentose phosphate (rPP) cycle - Reductive tricarboxylic acid (rTCA) cycle - Oxygen-sensitive reductive acetyl-CoA (rAcCoA) pathway - 3-hydroxypropionate cycle - 3-hydroxypropinate/4-hydroxybutyrate cycle - Dicarboxylate/4-hydroxybutyrate cycle

Despiste that these pathways are occurring in nature with a significantly great optimization, for synthethic biologists and for the biotechnology needs and processes, these pathways are not efficient enough.[2] Furthermore, researchers have developed different approaches based on constraint modeling and defined pathway analysis metrics. [3] There are four main criteria with which comparison between natural and synthetic pathways can be made. The end product for each pathway was chosen to be GA3P, the end product of the rPP cycle and such a common product, in order to enable comparison.

2.1 Superior kinetics

Pathway kinetics, or Pathway specific Activity is the maximal flux sustained by 1mg of pathway total protein. This criterion is analogous to an enzyme’s specific activity. The enzyme cost of the reaction is 1/Vi mg for 1 μmol/min product flux, where Vi stands for the enzyme’s specific activity (μmol/min/mg). Or, 1/Vi is how much mass of an enzyme is needed for 1 μmol/min product flux to be achieved. The enzyme cost of a whole pathway is a sum of the individual enzyme costs. Therefore, for a linear pathway, the pathway specific activity is 1/(Σ(1/Vi)). [3] In addition, there is no correlation between simplicity of the pathway (number of the enzymes included in the metabolic pathway) and pathway specific activity. [3]

2.2 Energetic efficiency There are two different subcategories of the energetic costs: [3] - NADPH cost is the number of moles of NADPH equivalents consumed in the production of one mole of product - ATP cost is the number of moles consumed in the production of one mole of product. The hydrolysis of more ATP molecules will increase the energetic cost. Though, a minimal amount of ATP molecules is needed in order to assure the thermodynamic usefulness (criterion III) of the pathway. The type of electron donors used in the pathway, pH and ionic strength, determinate the minimal ATP quantity that is used for product formation. For example, ferredoxin (Fd) is a one electron carrier, and therefore two ferredoxins are needed in order to replace one NADP molecule. So Fd has a lower reduction potential than NADPH and it shifts the thermodynamic profile with energetic gain. [3]

2.3 Thermodynamics

The third criterion gives information for the free energy change. If the production of one mole gives negative free energy, we can say, that the pathway is thermodynamically favorable. In addition, there is an interest in the thermodynamics under specified pH. So, the determination is of the standard Gibbs energy change, according the following equation: [3] ΔG,0f = ΔG0f (I = 0) + NHRTln(10) . pH – (2.91482(z2 – NH)I0.5)/(1 + 1.6I0.5) I, NH and z refer to the ionic strength, number of hydrogen atoms in the compound and charge of the compound, for a basic net reaction of a certain cycle. According the calculations, almost all carbon fixation pathways are thermodynamically feasible at broad range of pH and ionic strengths, except natural occurring rTCA and rAcCoA cycles. In addition, increased CO2 concentration can make those pathways feasible at all pH and ionic strength values. Such elevated CO2 concentration might be expected in certain environments, in C4 and CAM plants and in algae that use CO2 concentrating mechanisms. On the other hand, rTCA and rAcCoA are мore feasible then the rPP cycle at pH up to 7, but this energetic efficiency might be explained, that the organisms where rTCA and rACCoA are occurring are anaerobes and have energy restrictions, compared to the aerobes. [3]

2.4 Topology

For the topology criterion of great importance is the number of enzymes that the cycle utilizes as an independent unit and the number of enzymes that are included in the whole pathway. Furthermore, of great importance is how the pathways will integrate into the endogenous metabolic network. They have calculated the growth yield supported by each pathway as well as the number of significantly changed fluxes in the modified network compared with the wild type model of algae Chlamydomonas. [3]

Results and conclusions

By combination of the different, known, enzymes, and by using the developed algorithm which for a given substrate and product, finds the shortest cycle, different carbon fixation cycles can be build and, after that, compared between, by the criteria given before. [3] This study, is based, on building new pathways that will produce glyoxylate. Glyoxylate can be metabolized to glycerate 3-phosphate (GA3P) and each pathway it is consisted of a cycle and an assimilation sub-pathway which converts the cycle’s product into GA3P. The reason for this is that this is the simplest molecule that leads to possible bio-synthesis of other, larger molecules and sugars.[3]

So, taken in consideration all the previous data, and making the shortest cycle construct, that will produce glyoxylate, brought to a conclusion that, even though, this pathways are the simplest, they can not be thermodynamically feasible, when taking into account the physiologically relevant glyoxylate concentrations. These kind of metabolic pathways, which are shortcuts of the naturally occurring pathways, are having less pathway specific activity, compared to their “longer, original versions”.[3] But if the focus is turned on the kinetic aspects, while designing synthetic carbon fixation pathways, the use of carboxylating enzymes with high rate and high specific activities is unnecessary, like, for example, phosphoenolpyruvate (PEP), pyruvate and acetyl-CoA carboxylase. These enzymes have high affinity to HCO3- and high specific activity after saturating with CO2/ HCO3-. The shortest pathways that are consisted of different favorable carboxylating enzymes that will produce the same product, have the same metabolic structure. This type of synthetic metabolic pathways were named malonyl-CoA-oxaloacetate-glyoxylate (MOG) pathways.[3] MOG pathways, compared to the naturally occurring rPP cycle, have higher specific activity. In this pathways there can be included unique reactions, like alanine bypass and lactate bypass for converting malonyl-CoA to pyruvate. Among the few versions of MOG pathways, may be find pathways that use only one carboxylating enzyme: PEP carboxylase, which is in the group of the most favorable carboxylating enzymes. PEP is light regulated and can serve in switching the cycle activity according to the light exposure. They employ the same electron donors as rPP cycle, hydrolyze up to twelve ATP molecules and are thermodynamically favorable under different pH and ionic strengths. These pathways, are also called C4-glyoxylate cycles, because of their overlapping with the mechanism occurring in the C4 plants.[3] However, MOG cycles produce glyoxylate, the assimilation of which into central metabolism – via the bacterial glycerate pathway is rather inefficient as it involves a decarboxylation step.[3] Moreover, these cycles are very long and highly complicated. While implementing these pathways in microbes might be possible, their integration into the plant metabolic network seems highly unlikely. If one considers also unnatural reactions, simpler, more direct pathways can be proposed.

Synthetic carbon fixation pathways in the present and their future perspectives

Up to date, there are no registered in vivo applications of the MOG pathways. There is no evidence that this cycle is also operating, beside that it is good in theory. Though, this concept of comparing enzymes and building new carbon fixation cycles is a good starting point for several research directions, for instance, bio-based industry. After the optimization of the invented cycles, the next step is to use these cycles to implement in cell and cell structures. To make artificial cells, membranes, lyposomes or chloroplasts, where the system can be energized and the metabolic module could be used to fix carbon dioxide.[5] Another thing that can be done is to implement and try to transplant the cycle into living systems.[5] In vivo transplantation is a big challenge because MOG are a complex network cycles consisted of different enzymes (up to sixteen), and the implementation into living system, which has up to three thousand different reactions, can be really hard. [3]

 

Literature

[1] I. Tikh and C. Schmidt-Dannert, “Towards Engineered Light–Energy Conversion in Nonphotosynthetic Microorganisms,” Synth. Biol., pp. 303–316, Jan. 2013. [2] T. J. Erb and J. Zarzycki, “A short history of RubisCO: the rise and fall (?) of Nature’s predominant CO2 fixing enzyme,” Curr. Opin. Biotechnol., vol. 49, pp. 100–107, Feb. 2018. [3] A. Bar-Even, E. Noor, N. E. Lewis, and R. Milo, “Design and analysis of synthetic carbon fixation pathways,” Proc. Natl. Acad. Sci., vol. 107, no. 19, pp. 8889–8894, 2010. [4] F. Gong, Z. Cai, and Y. Li, “Synthetic biology for CO2fixation,” Sci. China Life Sci., vol. 59, no. 11, pp. 1106–1114, 2016. [5] D. L. Trudeau et al., “Design and in vitro realization of carbon-conserving photorespiration,” Proc. Natl. Acad. Sci., vol. 115, no. 49, p. E11455 LP-E11464, Dec. 2018.