Role of energetic coenzyme pools in the production of L-carnitine by Escherichia coli

Autodisplay of glucose-6-phosphate dehydrogenase for redox cofactor regeneration at the cell surface

Inherent cofactor regeneration is a pivotal feature of whole cell biocatalysis. For specific biotechnological applications, surface display of enzymes is emerging as a tool to circumvent mass transfer limitations or enzyme stability problems. Even complex reactions can be accomplished applying displayed enzymes. Yet, industrial utilization of the technique is still impeded by lacking cofactor regeneration at the cell surface. Here, we report on the surface display of a glucose-6-phoshate dehydrogenase (G6PDH) via Autodisplay to address this limitation and regenerate NADPH directly at the cell surface. The obtained whole cell biocatalyst demonstrated similar kinetic parameters compared to the purified enzyme, more precisely K_M values of 0.2 mM for NADP⁺ and calculated total turnover numbers of 10⁷ . However, the K_M for the substrate G6P increased by a factor of 7 to yield 1.5 mM. The whole cell biocatalyst was cheaper to produce, easy to separate from the reaction mixture and reusable in consecutive reaction cycles. Furthermore, lyophilization allowed storage at room temperature. The whole cell biocatalyst displaying G6PDH was applicable for NADPH regeneration in combination with soluble as well as surface displayed enzymes and model reactions in combination with bacterial CYP102A1 and human CYP1A2 were realized. Biotechnol. Bioeng. 2017;114: 1658-1669. © 2017 Wiley Periodicals, Inc.

On the in vivo redox state of flavin-containing photosensory receptor proteins

Measured values of the redox midpoint potential of flavin-containing photoreceptor proteins range from physiologically very negative values, i.e., < -300 mV (compared to the calomel electrode) for some LOV domains, to slightly positive values for some cryptochromes. The actual intracellular redox potential of several key physiological electron-transfer intermediates, like the nicotinamide dinucleotides, particularly in chemoheterotrophic bacteria, may be varying beyond these two values, and are subject to physiological- and environmental regulation. The photochemical activity of photoreceptor proteins containing their flavin chromophore in the reduced, and in the fully oxidized form, is very different. We therefore have addressed the question whether or not the functioning of these flavin-containing photosensory receptors in vivo is subject to redox regulation. Here we (1) provide further evidence for the overlap of the ranges of the redox midpoint potential of the flavin in a specific photoreceptor protein and the redox potential of key intracellular redox-active metabolites, and (2) demonstrate that the redox state and photochemical activity of LOV domains can be recorded in vivo in Escherichia coli. Significantly, so far in vivo reduction of LOV domains under physiological conditions could not be detected. The implications of these observations are discussed.

Metabolic engineering for high yielding L(-)-carnitine production in Escherichia coli

Background

L(-)-carnitine production has been widely studied because of its beneficial properties on various diseases and dysfunctions. Enterobacteria possess a specific biotransformation pathway which can be used for the enantioselective production of L(-)-carnitine. Although bioprocesses catalyzed by enzymes or whole cells can overcome the lack of enantioselectivity of chemical methods, current processes for L(−)-carnitine production still have severe disadvantages, such as the low yields, side reactions and the need of high catalyst concentrations and anaerobic conditions for proper expression of the biotransformation pathway. Additionally, genetically engineered strains so far constructed for L(-)-carnitine production are based on plasmids and, therefore, suffer from segregational unstability.

Results

In this work, a stable, high yielding strain for L(-)-carnitine production from low cost substrates was constructed. A metabolic engineering strategy was implemented in a multiple mutant for use in both growing and resting cells systems. The effect of mutations on gene expression and metabolism was analyzed to characterize the productivity constraints of the wild type and the overproducer strains. Precise deletion of genes which encode proteins of central and carnitine metabolisms were performed. Specifically, flux through the TCA cycle was increased by deletion of aceK (which encodes a bifunctional kinase/phosphatase which inhibits isocitrate dehydrogenase activity) and the synthesis of the by-product γ-butyrobetaine was prevented by deletion of caiA (which encodes a crotonobetainyl-CoA reductase). Both mutations led to improve the L(-)-carnitine production by 20 and 42%, respectively. Moreover, the highly regulated promoter of the cai operon was substituted by a constitutive artificial promoter increasing the biotransformation rate, even under aerobic conditions. Resting cells of the BW ΔaceK ΔcaiA p37cai strain produced 59.6 mmol l^-1 · h^-1 of L(−)-carnitine, doubling the productivity of the wild type strain. In addition, almost total conversion was attained in less than two hours without concomitant production of the side product γ–butyrobetaine.

Conclusions

L(-)-carnitine production has been enhanced by strain engineering. Metabolic engineering strategies herein implemented allowed obtaining a robust and high yielding E. coli strain. The new overproducer strain attained almost complete conversion of crotonobetaine into L(-)-carnitine with growing and resting cells, and even under aerobic conditions, overcoming the main environmental restriction to carnitine metabolism expression. So far, this is the best performing L(-)-carnitine production E. coli strain described.

Modelling and analysis of central metabolism operating regulatory interactions in salt stress conditions in a L-carnitine overproducing E. coli strain

Based on experimental data from E. coli cultures, we have devised a mathematical model in the GMA-power law formalism that describes the central and L-carnitine metabolism in and between two steady states, non-osmotic and hyperosmotic (0.3 M NaCl). A key feature of this model is the introduction of type of kinetic order, the osmotic stress kinetic orders (g(OSn)), derived from the power law general formalism, which represent the effect of osmotic stress in each metabolic process of the model.By considering the values of the g(OSn) linked to each metabolic process we found that osmotic stress has a positive and determinant influence on the increase in flux in energetic metabolism (glycolysis); L-carnitine biosynthesis production; the transformation/excretion of Acetyl-CoA into acetate and ethanol; the input flux of peptone into the cell; the anabolic use of pyruvate and biomass decomposition. In contrast, we found that although the osmotic stress has an inhibitory effect on the transformation flux from the glycolytic end products (pyruvate) to Acetyl-CoA, this inhibition is counteracted by other effects (the increase in pyruvate concentration) to the extent that the whole flux increases. In the same vein, the down regulation exerted by osmotic stress on fumarate uptake and its oxidation and the production and export of lactate and pyruvate are reversed by other factors up to the point that the first increased and the second remained unchanged.The model analysis shows that in osmotic conditions the energy and fermentation pathways undergo substantial rearrangement. This is illustrated by the observation that the increase in the fermentation fluxes is not connected with fluxes towards the tricaboxylic acid intermediates and the synthesis of biomass. The osmotic stress associated with these fluxes reflects these changes. All these observations support that the responses to salt stress observed in E. coli might be conserved in halophiles.Flux evolution during osmotic adaptations showed a hyperbolic (increasing or decreasing) pattern except in the case of peptone and fumarate uptake by the cell, which initially decreased. Finally, the model also throws light on the role of L-carnitine as osmoprotectant.

Quantitative analysis of the dynamic signaling pathway involved in the cAMP mediated induction of l-carnitine biosynthesis in E. coli cultures

L-(-)-carnitine can be synthesized from waste bioprecursors in the form of crotonobetaine. Such biotransformation is carried out in E. coli by the enzymes encoded by operons regulated by the cAMP receptor proteins. Non-phosphorylated sugars, such as glycerol are used as energy and carbon source since glucose inhibits cAMP synthesis. Until now little attention has been paid to the regulatory signaling structure that operates during the transition from a glucose-consuming, non-l-carnitine producing steady state, to a glycerol-consuming l-carnitine producing steady state. In this work we aim to elucidate and quantify the underlying regulatory mechanisms operating in the abolition of the glucose inhibiting effect. For this purpose we make use of the systemic approach by integrating the available information and our own experimentally generated data to construct a mathematical model. The model is built using power-law representation and is used as a platform to make predictive simulations and to assess the consistency of the regulatory structure of the overall process. The model is subsequently checked for quality through stability and a special, dynamic sensitivity analysis. The results show that the model is able to deal with the observed system transient phase. The model is multi-hierarchical, comprising the metabolic, gene expression, signaling and bioreactor levels. It involves variables and parameters of a very different nature that develop in different time scales and orders of magnitude. Some of the most relevant conclusions obtained are: (i) the regulatory interactions among glucose, glycerol and cAMP metabolism are far stronger than those present in the l-carnitine transport, production and degradation processes; (ii) carnitine biosynthesis is very sensitive to the cAMP signaling system since it reacts at very low cAMP receptor concentrations, and (iii) ATP is a critical factor in the transient dynamics. All these model-derived observations have been experimentally confirmed by separate studies. As a whole, the model contributes to our general understanding of the transient regulation through the signal regulatory structure, thus enabling more accurate optimization strategies to be used.

Production of L-carnitine by secondary metabolism of bacteria

The increasing commercial demand for L-carnitine has led to a multiplication of efforts to improve its production with bacteria. The use of different cell environments, such as growing, resting, permeabilized, dried, osmotically stressed, freely suspended and immobilized cells, to maintain enzymes sufficiently active for L-carnitine production is discussed in the text. The different cell states of enterobacteria, such as Escherichia coli and Proteus sp., which can be used to produce L-carnitine from crotonobetaine or D-carnitine as substrate, are analyzed. Moreover, the combined application of both bioprocess and metabolic engineering has allowed a deeper understanding of the main factors controlling the production process, such as energy depletion and the alteration of the acetyl-CoA/CoA ratio which are coupled to the end of the biotransformation. Furthermore, the profiles of key central metabolic activities such as the TCA cycle, the glyoxylate shunt and the acetate metabolism are seen to be closely interrelated and affect the biotransformation efficiency. Although genetically modified strains have been obtained, new strain improvement strategies are still needed, especially in Escherichia coli as a model organism for molecular biology studies. This review aims to summarize and update the state of the art in L-carnitine production using E. coli and Proteus sp, emphasizing the importance of proper reactor design and operation strategies, together with metabolic engineering aspects and the need for feed-back between wet and in silico work to optimize this biotransformation.

Redirecting metabolic fluxes through cofactor engineering: Role of CoA-esters pool during L(-)-carnitine production by Escherichia coli

Cofactor engineering, defined as the purposeful modification of the pool of intracellular cofactors, has been demonstrated to be a very suitable strategy for the improvement of L(-)-carnitine production in Escherichia coli strains. The overexpression of CaiB (CoA-transferase) and CaiC (CoA-ligase), both enzymes involved in coenzyme A transfer and substrate activation during the bioprocess, led to an increase in L(-)-carnitine production. Under optimal concentrations of inducer and fumarate (used as electron acceptors) yields reached 10- and 50-fold, respectively, that obtained for the wild type strain. However, low levels of coenzyme A limited the activity of these two enzymes since the addition of pantothenate increased production. Growth on substrates whose assimilation yields acetyl-CoA (such as acetate or pyruvate) further inhibited L(-)-carnitine production. Interestingly, control steps in the metabolism of acetyl-CoA of E. coli were detected. The glyoxylate shunt and anaplerotic pathways limit the bioprocess since strains carrying deletions of isocitrate lyase and isocitrate dehydrogenase phosphatase/kinase yielded 20-25% more L(-)-carnitine than the control. On the other hand, the deletion of phosphotransacetylase strongly inhibited the bioprocess, suggesting that an adequate flux of acetyl-CoA and the connection of the phosphoenolpyruvate-glyoxylate cycle together with the acetate metabolism are crucial for the biotransformation.