Systematic engineering enables efficient biosynthesis of L-phenylalanine in E. coli from inexpensive aromatic precursors.

BACKGROUND: L-phenylalanine is an essential amino acid with various promising applications. The microbial pathway for L-phenylalanine synthesis from glucose in wild strains involves lengthy steps and stringent feedback regulation that limits the production yield. It is attractive to find other candidates, which could be used to establish a succinct and cost-effective pathway for L-phenylalanine production. Here, we developed an artificial bioconversion process to synthesize L-phenylalanine from inexpensive aromatic precursors (benzaldehyde or benzyl alcohol). In particular, this work opens the possibility of L-phenylalanine production from benzyl alcohol in a cofactor self-sufficient system without any addition of reductant. RESULTS: The engineered L-phenylalanine biosynthesis pathway comprises two modules: in the first module, aromatic precursors and glycine were converted into phenylpyruvate, the key precursor for L-phenylalanine. The highly active enzyme combination was natural threonine aldolase LtaEP.p and threonine dehydratase A8HB.t, which could produce phenylpyruvate in a titer of 4.3 g/L. Overexpression of gene ridA could further increase phenylpyruvate production by 16.3%, reaching up to 5 g/L. The second module catalyzed phenylpyruvate to L-phenylalanine, and the conversion rate of phenylpyruvate was up to 93% by co-expressing PheDH and FDHV120S. Then, the engineered E. coli containing these two modules could produce L-phenylalanine from benzaldehyde with a conversion rate of 69%. Finally, we expanded the aromatic precursors to produce L-phenylalanine from benzyl alcohol, and firstly constructed the cofactor self-sufficient biosynthetic pathway to synthesize L-phenylalanine without any additional reductant such as formate. CONCLUSION: Systematical bioconversion processes have been designed and constructed, which could provide a potential bio-based strategy for the production of high-value L-phenylalanine from low-cost starting materials aromatic precursors.

Engineering a Novel Acetyl-CoA Pathway for Efficient Biosynthesis of Acetyl-CoA-Derived Compounds.

Acetyl-CoA is an essential central metabolite in living organisms and a key precursor for various value-added products as well. However, the intracellular availability of acetyl-CoA limits the efficient production of these target products due to complex and strict regulation. Here, we proposed a new acetyl-CoA pathway, relying on two enzymes, threonine aldolase and acetaldehyde dehydrogenase (acetylating), which can convert one l-threonine into one acetyl-CoA, one glycine, and generate one NADH, without carbon loss. Introducing the acetyl-CoA pathway could increase the intracellular concentration of acetyl-CoA by 8.6-fold compared with the wild-type strain. To develop a cost-competitive and genetically stable acetyl-CoA platform strain, the new acetyl-CoA pathway, driven by the constitutive strong promoter, was integrated into the chromosome of Escherichia coli. We demonstrated the practical application of this new acetyl-CoA pathway by high titer production of ß-alanine, mevalonate, and N-acetylglucosamine. At the same time, this pathway achieved a high-yield production of glycine, a value-added commodity chemical for the synthesis of glyphosate and thiamphenicol. This work shows the potential of this new acetyl-CoA pathway for the industrial production of acetyl-CoA-derived compounds.

A novel strategy for L-arginine production in engineered Escherichia coli.

BACKGROUND: L-arginine is an important amino acid with applications in diverse industrial and pharmaceutical fields. N-acetylglutamate, synthesized from L-glutamate and acetyl-CoA, is a precursor of the L-arginine biosynthetic branch in microorganisms. The enzyme that produces N-acetylglutamate, N-acetylglutamate synthase, is allosterically inhibited by L-arginine. L-glutamate, as a central metabolite, provides carbon backbone for diverse biological compounds besides L-arginine. When glucose is the sole carbon source, the theoretical maximum carbon yield towards L-arginine is 96.7%, but the experimental highest yield was 51%. The gap of L-arginine yield indicates the regulation complexity of carbon flux and energy during the L-arginine biosynthesis. Besides endogenous biosynthesis, N-acetylglutamate, the key precursor of L-arginine, can be obtained by chemical acylation of L-glutamate with a high yield of 98%. To achieve high-yield production of L-arginine, we demonstrated a novel approach by directly feeding precursor N-acetylglutamate to engineered Escherichia coli. RESULTS: We reported a new approach for the high yield of L-arginine production in E. coli. Gene argA encoding N-acetylglutamate synthase was deleted to disable endogenous biosynthesis of N-acetylglutamate. The feasibility of external N-acetylglutamate towards L-arginine was verified via growth assay in argA- strain. To improve L-arginine production, astA encoding arginine N-succinyltransferase, speF encoding ornithine decarboxylase, speB encoding agmatinase, and argR encoding an arginine responsive repressor protein were disrupted. Based on overexpression of argDGI, argCBH operons, encoding enzymes of the L-arginine biosynthetic pathway, ~ 4 g/L L-arginine was produced in shake flask fermentation, resulting in a yield of 0.99 mol L-arginine/mol N-acetylglutamate. This strain was further engineered for the co-production of L-arginine and pyruvate by removing genes adhE, ldhA, poxB, pflB, and aceE, encoding enzymes involved in the conversion and degradation of pyruvate. The resulting strain was shown to produce 4 g/L L-arginine and 11.3 g/L pyruvate in shake flask fermentation. CONCLUSIONS: Here, we developed a novel approach to avoid the strict regulation of L-arginine on ArgA and overcome the metabolism complexity in the L-arginine biosynthesis pathway. We achieve a high yield of L-arginine production from N-acetylglutamate in E. coli. Co-production pyruvate and L-arginine was used as an example to increase the utilization of input carbon sources.

Single Image De-Raining via Improved Generative Adversarial Nets.

Capturing images under rainy days degrades image visual quality and affects analysis tasks, such as object detection and classification. Therefore, image de-raining has attracted a lot of attention in recent years. In this paper, an improved generative adversarial network for single image de-raining is proposed. According to the principles of divide-and-conquer, we divide an image de-raining task into rain locating, rain removing, and detail refining sub-tasks. A multi-stream DenseNet, termed as Rain Estimation Network, is proposed to estimate the rain location map. A Generative Adversarial Network is proposed to remove the rain streaks. A Refinement Network is proposed to refine the details. These three models accomplish rain locating, rain removing, and detail refining sub-tasks, respectively. Experiments on two synthetic datasets and real world images demonstrate that the proposed method outperforms state-of-the-art de-raining studies in both objective and subjective measurements.