Engineering a microbial strain for production sometimes entails metabolic modifications that impair essential physiological processes for growth or production. Restoring these functions may require amending a variety of non-obvious physiological networks, and thus, rational design strategies may not be practical. Here we demonstrate that growth and production may be restored by evolution that repairs impaired metabolic function. Furthermore, we use genomics, metabolomics and proteomics to identify several underlying mutations and metabolic perturbations that allow metabolism to repair. Previously, high titers of butanol production were achieved by Escherichia coli using a growth-coupled, modified Clostridial CoA-dependent pathway after all native fermentative pathways were deleted. However, production was only observed in rich media. Native metabolic function of the host was unable to support growth and production in minimal media. We use directed cell evolution to repair this phenotype and observed improved growth, titers and butanol yields. We found a mutation in pcnB which resulted in decreased plasmid copy numbers and pathway enzymes to balance resource utilization. Increased protein abundance was measured for biosynthetic pathways, glycolytic enzymes have increased activity, and adenosyl energy charge was increased. We also found mutations in the ArcAB two-component system and integration host factor (IHF) that tune redox metabolism to alter byproduct formation. These results demonstrate that directed strain evolution can enable systematic adaptations to repair metabolic function and enhance microbial production. Furthermore, these results demonstrate the versatile repair capabilities of cell metabolism and highlight important aspects of cell physiology that are required for production in minimal media.

The absence of orthogonal aminoacyl-transfer RNA (tRNA) synthetases that accept non-l-α-amino acids is a primary bottleneck hindering the in vivo translation of sequence-defined hetero-oligomers and biomaterials. Here we report that pyrrolysyl-tRNA synthetase (PylRS) and certain PylRS variants accept α-hydroxy, α-thio and N-formyl-l-α-amino acids, as well as α-carboxy acid monomers that are precursors to polyketide natural products. These monomers are accommodated and accepted by the translation apparatus in vitro; those with reactive nucleophiles are incorporated into proteins in vivo. High-resolution structural analysis of the complex formed between one PylRS enzyme and a m-substituted 2-benzylmalonic acid derivative revealed an active site that discriminates prochiral carboxylates and accommodates the large size and distinct electrostatics of an α-carboxy substituent. This work emphasizes the potential of PylRS-derived enzymes for acylating tRNA with monomers whose α-substituent diverges substantially from the α-amine of proteinogenic amino acids. These enzymes or derivatives thereof could synergize with natural or evolved ribosomes and/or translation factors to generate diverse sequence-defined non-protein heteropolymers.

U nlike DNA, metabolic systems do not possess a repair mechanism. However, it has been proven that metabolism has the ability to overcome deleterious consequences caused by pathway damage when reactions of a metabolic pathway are blocked 1-4. One means by which blocked enzymes can be complemented is with the use of isozymes, which may serve as spare parts 5. Alternatively, promiscuous enzymes may be present to complement a lost function 2,6,7. Other enzymes are highly specialized and catalyze unique reactions, and when damaged they may compromise growth and alter physiology. The wide range of possible reactions catalyzed by cryptic and promiscuous enzymes have the potential to form pathways that can restore metabolic function when these specialized reactions are blocked 8. Although cells have proven the ability to bypass blocked metabolic function, the diverse mechanisms and extent to which the cell can bypass metabolic blocks have not been thoroughly explored. Here, we use a ΔpanD (encoding for aspartate 1-decarboxylase) strain of E. coli, which is incapable of producing the β-alanine required for synthesis of CoA, as an example to demonstrate the repair capabilities of cell metabolism. Aspartate 1-decarboxylase (PanD) is the only enzyme capable of β-alanine synthesis in E. coli. In bacteria, fungi, and plants, β-alanine is a precursor to pantothenate, which in turn is a required metabolite for the synthesis of coenzyme A (CoA) in all organisms 9. In animals, β-alanine is synthesized as a precursor to carnosine, which is found at high concentrations in skeletal muscle tissue and the central nervous system and is used for various physiological purposes 10. Without CoA, the cell is incapable of carrying out essential cellular processes including the TCA cycle, fatty acid biosynthesis, and acetyl-CoA synthesis, which is used as a building block for many essential compounds 9. Therefore, unless β-alanine or pantothenate are supplemented, a ΔpanD strain cannot grow on minimal media alone. Unlike most decarboxylases that use pyridoxal-5′-phosphate (PLP) as a cofactor, PanD uses a covalently bound pyruvoyl group 11. PanD is first translated as an inactive protoenzyme that is cleaved into α-and β-subunits, which is triggered by the activator PanZ 12. This likely serves as an additional regulatory element to control intracellular levels of pantothenate. Several other pathways are believed to exist in other organisms to supply β-alanine: degradation of propionate into malonic semialdehyde (MSA) and subsequent transamination 13 , a reductive uracil degradation pathway using dihydrouracil as an intermediate 14 , and an oxidative degradation of spermine into β-alanine using 3-aminopropanal as an upstream precursor 15. However, within E. coli, these pathways are not known to exist. Here, we show that to overcome damage to the β-alanine pathway, E. coli metabolism can be repaired through the emergence of at least three novel metabolic pathways to produce β-alanine. Our results demonstrate the intrinsic pliability of biological...

The absence of orthogonal aminoacyl-tRNA synthetases that accept non-L-a-amino acids is the primary bottleneck hindering the in vivo translation of sequence-defined hetero-oligomers. Here we report PylRS enzymes that accept a-hydroxy acids, a-thio acids, N-formyl-L-a-amino acids, and a-carboxyl acid monomers (malonic acids) that are formally precursors to polyketide natural products. These monomers are all accommodated and accepted by the translation apparatus in vitro. High-resolution structural analysis of the complex between one such PylRS enzyme and a meta-substituted 2-benzylmalonate derivative reveals an active site that discriminates pro-chiral carboxylates and accommodates the large size and distinct electrostatics of an a-carboxyl acid substituent. This work emphasizes the potential of PylRS for evolving new enzymes capable of encoding diverse non-L-a-amino acids in synergy with natural or evolved ribosomes.

Aminoacyl-tRNA synthetases (aaRSs) provide the functional and essential link between the sequence of an mRNA and the protein it encodes. aaRS enzymes catalyze a two-step chemical reaction that acylates specific tRNAs with a cognate α-amino acid. In addition to their role in translation, acylated tRNAs contribute to non-ribosomal natural product biosynthesis and are implicated in multiple human diseases. From the standpoint of synthetic biology, the acylation of tRNAs with a non-canonical α-amino acid (ncAA) or more recently, a non-α-amino acid monomer (nαAA) is a critical first step in the incorporation of these monomers into proteins, where they can be used for fundamental and applied science. These endeavors all demand an understanding of aaRS activity and specificity. Although a number of methods to monitor aaRS function in vitro or in vivo have been developed, many evaluate only the first step of the two-step reaction, require the use of radioactivity, or are slow, difficult to generalize, or both. Here we describe an LC-MS assay that rapidly, quantitatively, and directly monitors aaRS activity by detecting the intact acyl-tRNA product. After a simple tRNA acylation reaction workup, acyl- and non-acyl-tRNA molecules are resolved using ion-pairing reverse phase chromatography and their exact masses are determined using high-resolution time-of-flight mass spectrometry. The intact tRNA assay we describe is fast, simple, and quantifies reaction yields as low as 0.23%. The assay can also be employed on tRNAs acylated with flexizyme to detect products that are undetectable using standard techniques. The protocol requires basic expertise in molecular biology, mass spectrometry, and RNAse-free techniques.