Discovery of novel amino acid production traits by evolution of synthetic co-cultures.

BACKGROUND: Amino acid production features of Corynebacterium glutamicum were extensively studied in the last two decades. Many metabolic pathways, regulatory and transport principles are known, but purely rational approaches often provide only limited progress in production optimization. We recently generated stable synthetic co-cultures, termed Communities of Niche-optimized Strains (CoNoS), that rely on cross-feeding of amino acids for growth. This setup has the potential to evolve strains with improved production by selection of faster growing communities. RESULTS: Here we performed adaptive laboratory evolution (ALE) with a CoNoS to identify mutations that are relevant for amino acid production both in mono- and co-cultures. During ALE with the CoNoS composed of strains auxotrophic for either L-leucine or L-arginine, we obtained a 23% growth rate increase. Via whole-genome sequencing and reverse engineering, we identified several mutations involved in amino acid transport that are beneficial for CoNoS growth. The L-leucine auxotrophic strain carried an expression-promoting mutation in the promoter region of brnQ (cg2537), encoding a branched-chain amino acid transporter in combination with mutations in the genes for the Na+/H+-antiporter Mrp1 (cg0326-cg0321). This suggested an unexpected link of Mrp1 to L-leucine transport. The L-arginine auxotrophic partner evolved expression-promoting mutations near the transcriptional start site of the yet uncharacterized operon argTUV (cg1504-02). By mutation studies and ITC, we characterized ArgTUV as the only L-arginine uptake system of C. glutamicum with an affinity of KD = 30 nM. Finally, deletion of argTUV in an L-arginine producer strain resulted in a faster and 24% higher L-arginine production in comparison to the parental strain. CONCLUSION: Our work demonstrates the power of the CoNoS-approach for evolution-guided identification of non-obvious production traits, which can also advance amino acid production in monocultures. Further rounds of evolution with import-optimized strains can potentially reveal beneficial mutations also in metabolic pathway enzymes. The approach can easily be extended to all kinds of metabolite cross-feeding pairings of different organisms or different strains of the same organism, thereby enabling the identification of relevant transport systems and other favorable mutations.

Characteristics of the GlnH and GlnX Signal Transduction Proteins Controlling PknG-Mediated Phosphorylation of OdhI and 2-Oxoglutarate Dehydrogenase Activity in Corynebacterium glutamicum.

In Corynebacterium glutamicum the protein kinase PknG phosphorylates OdhI and thereby abolishes the inhibition of 2-oxoglutarate dehydrogenase activity by unphosphorylated OdhI. Our previous studies suggested that PknG activity is controlled by the periplasmic binding protein GlnH and the transmembrane protein GlnX, because ΔglnH and ΔglnX mutants showed a growth defect on glutamine similar to that of a ΔpknG mutant. We have now confirmed the involvement of GlnH and GlnX in the control of OdhI phosphorylation by analyzing the OdhI phosphorylation status and glutamate secretion in ΔglnH and ΔglnX mutants and by characterizing ΔglnX suppressor mutants. We provide evidence for GlnH being a lipoprotein and show by isothermal titration calorimetry that it binds l-aspartate and l-glutamate with moderate to low affinity, but not l-glutamine, l-asparagine, or 2-oxoglutarate. Based on a structural comparison with GlnH of Mycobacterium tuberculosis, two residues critical for the binding affinity were identified and verified. The predicted GlnX topology with four transmembrane segments and two periplasmic domains was confirmed by PhoA and LacZ fusions. A structural model of GlnX suggested that, with the exception of a poorly ordered N-terminal region, the entire protein is composed of α-helices and small loops or linkers, and it revealed similarities to other bacterial transmembrane receptors. Our results suggest that the GlnH-GlnX-PknG-OdhI-OdhA signal transduction cascade serves to adapt the flux of 2-oxoglutarate between ammonium assimilation via glutamate dehydrogenase and energy generation via the tricarboxylic acid (TCA) cycle to the availability of the amino group donors l-glutamate and l-aspartate in the environment. IMPORTANCE Actinobacteria comprise a large number of species playing important roles in biotechnology and medicine, such as Corynebacterium glutamicum, the major industrial amino acid producer, and Mycobacterium tuberculosis, the pathogen causing tuberculosis. Many actinobacteria use a signal transduction process in which the phosphorylation status of OdhI (corynebacteria) or GarA (mycobacteria) regulates the carbon flux at the 2-oxoglutarate node. Inhibition of 2-oxoglutarate dehydrogenase by unphosphorylated OdhI shifts the flux of 2-oxoglutarate from the TCA cycle toward glutamate formation and, thus, ammonium assimilation. Phosphorylation of OdhI/GarA is catalyzed by the protein kinase PknG, whose activity was proposed to be controlled by the periplasmic binding protein GlnH and the transmembrane protein GlnX. In this study, we combined genetic, biochemical, and structural modeling approaches to characterize GlnH and GlnX of C. glutamicum and confirm their roles in the GlnH-GlnX-PknG-OdhI-OdhA signal transduction cascade. These findings are relevant also to other Actinobacteria employing a similar control process.