Nicotinamide coenzyme biomimetics for biocatalysis

In the last century, the use of enzymes for manufacturing of bulk and fine chemicals has spiked. Enabling redox reactions, oxidoreductases are mostly employed for the generation of chiral centres due to their exquisite enantioselectivity. These biocatalysts demand the presence of coenzymes to mediate the electron transfer, with nicotinamide adenine dinucleotide being the most common redox coenzyme. Both its unphosphorylated (NAD) and phosphorylated (NADP) forms are expensive and unstable, particularly in their reduced state. Born as NAD(P) models to be used for mechanistic studies, nicotinamide coenzyme biomimetics (NCBs) retain the dihydropyridine core of the natural coenzymes but lack of the dinucleotide moiety. While being a simpler version of NAD(P), NCBs maintain the redox capacity of the natural pyridine coenzyme and, in the last decade, gained popularity as alternative and cheaper coenzymes for oxidoreductases. In this thesis we further explore the use of NCBs with multiple oxidoreductase classes and debate their future in the applied biocatalysis sector.Chapter 1 provides the background for this Ph.D. thesis. Redox biocatalysts, their chemistry and their coenzyme requirements are here described. A more in depth description of nicotinamide adenine dinucleotide coenzymes (NAD(P)) and the development of their synthetic biomimetics (NCBs) is included. Moreover, the relevance of this project is outlined.Chapter 2 revolves around coenzyme alternatives to the natural nicotinamide, flavin and S-adenosyl-L-methionine coenzymes. An overview of the application of the artificial coenzymes with different types of redox enzymes is given by reviewing recent literature. As described, artificial coenzymes can be cost effective and can be employed to tune reaction rates or the type of reaction. Special attention is given to NCBs, as these biomimetics have been proven to be particularly effective with flavin-dependent oxidoreductases.Chapter 3 offers a detailed picture of the diverse catalytic mechanisms and chemistries performed by flavoprotein monooxygenases (FPMOs) and halogenases. It is also explained which FPMOs can be coupled to NCBs for the regeneration of their flavin coenzyme.In Chapter 4 we show the coupling of four group A FPMOs with a panel of NCBs for the regioselective hydroxylation of phenolic compounds. Because of product inhibition, one of the FPMOs showed meagre conversions with NCBs, whereas the other three enzymes accepted NCBs as electron donors with 100% product yield. Intriguingly, each enzyme preferred a specific NCB as hydride donor and formed only minor amounts of the by-product hydrogen peroxide. Steady state kinetics revealed that the biomimetics do have weak binding and relatively low catalytic efficiency compared to the preferred natural coenzymes. Based on this behaviour we propose a “kiss and ride” mechanism for these FAD-dependent aromatic hydroxylases.Oxidoreductases that contain a Rossmann-like domain for NAD(P) binding poorly accept NCBs as coenzymes. This is particularly true for solely NAD(P)-dependent oxidoreductases, of which only a few were reported to accept the artificial coenzymes, although with low activity. To solve this scarcity, Chapter 5 and Chapter 6 focus on the discovery of flavin-independent redox biocatalysts that can recognize NCBs as their coenzymes.Chapter 5 deals with flavin- and zinc-independent double bond reductases (DBRs). DBRs are a promising class of biocatalysts for the asymmetric hydrogenation of activated alkenes, and their catalytic mechanism as well as their substrate scope and applications are reviewed in this Chapter. Starting from a previously reported activity of a DBR with several NCBs, sequence alignment was employed to identify two putative DBRs, which are presumably also able to catalyze their reaction in presence of the biomimetics. Together with the previously reported DBR, we screened these enzymes for the reduction of phenylpropenal substrates with three NCBs to find that moderate conversions were achieved in presence of NCBs (≤ 10%), with an underlying preference for N1-benzyl substituted biomimetics. Intriguingly, the activity of these DBRs with NCBs depended also on the substrate used in the reaction.Chapter 6 follows a different computational approach, based on catalophores, to find putative NCB-dependent oxidoreductases. By mapping the active sites of biocatalysts reported to use NCBs as their hydride donors, and subsequently mining the Protein Data Bank for enzymes showing similar active sites in terms of physical-chemical properties, we were able to identify four NADP-dependent enzymes and test them for activity with NCBs. The best result was obtained with an aldehyde dehydrogenase, which yielded 6% conversion over 24 h. Although product formation was confirmed, a ketoreductase gave very meagre activity in presence of NCBs. Interestingly, a possible new reactivity was found for a double bond reductase.Although the achieved conversions were not optimal, Chapter 5 and Chapter 6 serve as a starting point for the future optimization of the interactions between the newly identified NCB-accepting oxidoreductases and the synthetic coenzymes, as well as for the future discovery of other NCB-dependent enzymes.Finally, by merging the results described in the previous Chapters, this thesis is concluded by a general discussion (Chapter 7) that places our findings in a broader context and includes unpublished data. At first, an analysis regarding the multiple uses of NCBs with redox enzymes is given. Further accent is put on the crucial challenges that NCBs have to face in order to be widely applied for biocatalysis such as their stability and their in vitro recycling. Lastly, the use of NAD(P)H and NCBs for photobiocatalysis is introduced and preliminary data are shown together with future suggestions for this exciting new branch of biocatalysis.