Exploiting bifunctional fusion enzymes from biosynthetic pathways as biocatalysts

Abstract

Biocatalysis is anenvironmentally friendly alternative to traditional synthetic techniques andleads to the production of products with high enantio-, stereo- and chemo-selectivity. Conducted at ambient temperature and pressure, enzymes aredesirable catalysts that can produce a product of interest with limited wasteand side-products. Enzymes exist as either single domains, with a singlefunction, or can be multifunctional, composed of numerous domains which eachcatalyse a different reaction. These enzymes are known as fusions as eachdomain is fused together by a linker region. Fusions are gaining interest dueto their ability to transform a molecule through consecutive reactions inone-pot. This multifunctionality has been evolutionarily optimised, and they canoften outperform synthetic methods and biocatalytic cascades to producesynthetically complex molecules. Found within the biosynthetic pathways of manykey natural products, fusion enzymes remain an interesting target forbiocatalysis and would be a valuable addition to the biocatalytic toolbox. Aseries of fusions from two biosynthetic pathways in particular (tambjamine YP1and biotin production) will be explored.

Tambjamines are a class ofsecondary metabolite natural products (NPs) that are closely related to theprodiginines through a shared 4-methoxy-2-2’-bipyrrole-5-carbaldehyde (MBC)core structure. These yellow pigmented molecules have attracted interest due totheir bioactivity. Originally isolated from the marine microbe Pseudoalteromonastunicata, the most well-known is tambjamine YP1. The predicted biosyntheticpathway comprises two convergent routes, and due to the conserved nature of MBCbiosynthesis, the formation of the YP1 bipyrrole can be predicted withconfidence. In contrast, the production and attachment of the fatty amine tailto the MBC core is less well understood. Analysing enzymes encoded in the P.tunicata ‘tam’ biosynthetic gene cluster (BGC) showed TamA to be anadenosine triphosphate (ATP)-dependent didomain enzyme. TamA comprises a classI adenylation (ANL) domain fused to an acyl carrier protein (ACP) domain andcatalyses activation and attachment of C12 lauric acid to the4’-phosphopantetheine (4’-PP) arm of the ACP. Truncation of the enzyme byremoving the ACP domain, enabled the formation of a number of amides withvarying chain lengths (C6-C14) and amino acid or aminesubstrates. Within the BGC another unusual uncharacterised didomain enzyme(TamH) was discovered and predicted to be involved in processing the C12-boundTamA intermediate to the free amine. Sequence analysis predicts TamH iscomprised of an N-terminal, pyridoxal 5’- phosphate (PLP) dependenttransaminase domain (TA) fused to a NAD(P)H-dependent C-terminalthioester reductase (TR) domain. It is therefore proposed that TamH catalyses reductionof the TamA ACP-bound C12 thioester to release the aldehyde, whichthen undergoes transamination to generate the C12 amine. Here, thisunique TamH fusion was studied for the first time and revealed the acyl-ACPdependence of the TR domain, as opposed to a free acyl-CoA thioester. The TAdomain was shown to utilise ʟ-Glu as an amine donor and exhibited promiscuitywith aldehyde chain lengths of C7-C14 being transformedthrough to the corresponding amine. Furthermore, both the TamA and TamH fusionswere successfully coupled together to generate a biocatalytic cascade thatsmoothly converts fatty acids to amines in one-pot. This work not onlycharacterises the formation of the YP1 tail for the first time, but alsosuggests that TamA and TamH could be useful biocatalysts.

In a separate project, thebiosynthesis of the essential NP biotin was explored. Biotin is an importantvitamin (vitamin H) utilised as a cofactor by a number of carboxylase enzymesin essential metabolic processes such as fatty acid synthesis, amino acidmetabolism and gluconeogenesis. Microbial biotin biosynthesis is divided intotwo stages, firstly, the synthesis of the DC7 dicarboxylate pimelatemoiety as a free pimeloyl-CoA thioester or bound to an acyl carrier protein(ACP). Secondly, the biotin bicyclic ring structure is formed, which is believedto be evolutionary conserved across different organisms. The biotinbiosynthetic pathway has been studied in various organisms, with two distinctpimeloyl-CoA/ACP pathways; the more common BioC-BioH pathway found in E. coliand the much rarer BioW-dependent route which was discovered in B. subtilis.BioW is an adenylating enzyme responsible for the production of the DC7pimeloyl-CoA thioester in an ATP-dependent manner. This is followed by thereaction of the pimeloyl-CoA intermediate with ʟ-Ala through a decarboxylative,C-C bond forming, claisen-like condensation. This step produces the keyintermediate 8-amino-7-oxononanioc acid (AON) and is catalysed by thePLP-dependent, class II aminotransferase AON synthase (BioF). While searchingfor novel biocatalysts, an unusual BioWF didomain fusion was identified in thebacterium Corynebacterium amycolatum. In contrast to the highly selective BsBioW,the BioW domain of the CaBioWF fusion displayed unexpected promiscuity,accepting a range of mono and di-fatty acid substrates. Sequence and structuralanalysis revealed a natural divergence in a key active site residue shown tohydrogen bond the terminal carboxy group in the BsBioWpimeloyl-adenylate structure (PDB:5FLL). In the predicted active site of CaBioWF,this residue was present as a phenylalanine in place of the highly conservedtyrosine. Based on previous BsBioW engineering, and confirmed bymutagenesis studies, this substrate promiscuity is due to thetyrosine/phenylalanine swap. This promiscuity extends to the BioF domain, withother amino acids, in addition to ʟ-Ala, being converted into the correspondingAON product. Therefore, the CaBioWF fusion can catalyse the formation ofvarious AON derivatives. Such unexpected broad substrate scope suggests that CaBioWFis a naturally promiscuous biocatalyst that can be utilised to prepare a numberof AON analogues.

In conclusion, this thesisdescribes functional characterisation of novel enzyme fusions that play keyroles in their respective NP biosynthetic pathways. Moreover, detailed analysisprovides strong supporting evidence that these fusions display promiscuity thatsuggests they can be used as versatile biocatalysts for synthetic chemistry.