Universität Stuttgart
Permanent URI for this communityhttps://elib.uni-stuttgart.de/handle/11682/1
Browse
5 results
Search Results
Item Open Access Enzymatic asymmetric dihydroxylation of alkenes(2016) Gally, Christine; Hauer, Bernhard (Prof. Dr.)The introduction of chirality into C=C double bonds is of special interest in organic synthesis. In particular, the catalytic asymmetric dihydroxylation (AD) of alkenes has attracted considerable attention due to the facile transformation of the chiral diol products into valuable derivatives. By chemical means, the metal-catalyzed AD of olefins provides both stereo- and regiospecific cis-diol moieties. Next to their toxicity, however, these metal catalysts can also lead to byproduct formation as a result of oxidative fission. In nature, Rieske non-heme iron oxygenases (ROs) represent promising biocatalysts for this reaction since they are the only enzymes known to catalyze the stereoselective formation of vicinal cis-diols in one step. ROs are key enzymes in the degradation of aromatic hydrocarbons and can target a wide variety of different arenes. Despite their broad substrate scope, limited data is available for the conversion of unnatural substrates by this class of enzymes. To explore their potential for alkene oxidation, three ROs were tested for the oxyfunctionalization of a set of structurally diverse olefins including linear and cyclic arene-substituted alkenes, cycloalkenes as well as several terpenes. Naphthalene- (NDO), benzene- (BDO) and cumene dioxygenases (CDO) from different Pseudomonas strains where selected as they are amongst the RO enzymes that have already been reported to catalyze the oxidation of a small number of olefins. The majority of compounds from the selected substrate panel could be converted by NDO, BDO or CDO and products were either isolated and identified by NMR analysis or using the authentic standards. Dependent on the substrate, allylic monohydroxylation was found in addition to the corresponding diol products, a reaction which is chemically still most reliably achieved by the use of SeO2 in stoichiometric amounts. However, having been evolved for the dihydroxylation of aromatic compounds, wild type ROs displayed low conversions (< 50%) and modest stereoselectivities (≤ 80% ee/de) for several of the tested olefins. To overcome these limitations, changes in the active site topology of RO catalysts were introduced. A single targeted point mutation that was identified based on sequence and structural comparisons with other members of the RO family proved to be sufficient to generate BDO and CDO variants displaying remarkable changes in regio- and stereoselectivity for various substrates. In particular biotransformations with CDO M232A gave excellent stereoselectivities (≥ 95% ee/de) and good activities (> 90%) also for linear alkenes, which have been reported to be challenging substrates for RO-catalyzed oxyfunctionalizations. Site-saturation mutagenesis at position 232 in CDO revealed a correlation between the steric demand of the amino acid side chain and its influence on regio- and/ or stereoselectivities for styrene and indene. While the wild type enzyme almost exclusively catalyzed the dihydroxylation of the aromatic ring, the regioselectivity was shifted with decreasing side chain size to the terminal vinyl group of styrene, yielding up to 96% of the alkene-1,2-diol. For cis-1,2-indandiol formation, enantiocomplementary enzymes could be generated, a fact further highlighting the importance of position 232 for the engineering of ROs. Moreover, site-saturation mutagenesis of additional residues in the substrate binding pocket of CDO (F278, I288, I336 and F378) identified further positions having an influence on selectivity and product formation for alkene oxidation. To proof the applicability of ROs for organic synthesis, semi-preparative scale biotransformations (70 mg) of selected substrates were performed with CDO M232A. Without further optimization of the reaction set-up, products were successfully isolated in > 30% yield. In addition, up-scaling of (R)-limonene hydroxylation to 4 L in a bioreactor with growing cells gave final isolated product titers of 0.4 g L-1 even though substrate volatility and product toxicity diminished the yield. In conclusion, these examples demonstrated that a single point mutation was sufficient to transform CDO wild type into an efficient catalyst, furthermore constituting the first example of the rational engineering of CDO and BDO enzymes for the oxyfunctionalization of a broad range of alkenes.Item Open Access Metabolismus nicht-physiologischer Substrate in Mikroorganismen(2016) Reznicek, Ondrej; Hauer, Bernhard (Prof. Dr.)Item Open Access Characterization and application of novel imine reductases(2016) Scheller, Philipp; Hauer, Bernhard (Prof. Dr.)Chiral amines are an ubiquitously distributed class of bioactive compounds, what turns them into preferred scaffolds for pharmaceuticals. The high chemical and enantiomerical purities required for such an application are ideally suited for biocatalysis as enzymatic methods routinely display high specificities. The established methods for chiral amine synthesis with lipases, ω-transaminases and amine oxidases, however have considerable limitations regarding their access to pharmaceutically relevant chiral secondary and tertiary amines. Recently the new enzyme class of imine reductases (IREDs) was described, offering an attractive extension to the currently used techniques as the preparation of imines by chemical methods in organic solvents is a well established and widely applicable method. As the number of IREDs known initially was limited to only three enzymes, this project started with a database search for the discovery of novel enzymes. For the first time it was shown that the IRED family is much larger than assumed and over 350 novel, putative IREDs were identified. A sequence analysis of the database members revealed (R)- type and (S)- type superfamilies and led after an update to the identification of IRED specific sequence motifs. These criteria allowed to define this new enzyme family on a sequence level and discriminate them from the closest related homologues. Based on the biochemical information about the three published IREDs and a conservation analysis of the database members, three new enzymes from Streptosporangium roseum DSM43021, Streptomyces turgidiscabies and Paenibacillus elgii were selected for characterization. The enzymes were shown to encode for functional IREDs with much higher activity than the previously known IREDs. By site directed mutagenesis the mechanism of the IREDs was probed and the importance of a conserved Tyr for catalysis of an (S)- type IRED shown, while the crucial role of the proposed Asp residue for catalysis in the (R)- type IREDs was questioned. The characterization of the new IREDs revealed their pH optima and confirmed the suspected dimerization. The thermostability of the IREDs was investigated and the selected (S)- type IRED identified as the most stable enzyme known to date. Further the activity in the presence of water miscible organic solvents was tested and high tolerance versus MeOH found. In biotransformations all IREDs showed high activity and a broad panel of cyclic imines was fully converted to piperidines and tetrahydroisoquinolines with enantioselectivities up to 99% ee. With purified IREDs kinetic constants for these substrates were recorded and their substrate preference investigated. This indicated a preference of the (S)- type IRED for more bulky substrates, compared to the (R)- type IREDs. After optimization of the reaction conditions, with purified IREDs also high activities and chemo- as well as enantioselectivities for very labile exocyclic imines were detected. The possibility to effectively reduce already low levels of such imines led to the application of one (R)- type IRED for the generation of novel C-N bonds by reductive aminations. The established methodology revealed the crucial influence, conditions that favor imine formation (high molar excess of the amine nucleophile and high pH) display on the conversion rates. Under optimized conditions, different carbonyls could efficiently be transformed with a variety of amines in the aqueous buffer system with moderate to good conversions into primary and secondary (chiral) amines with very high selectivities (ee up to 98%). Finally, an application of IREDs in cascade reactions to produce saturated N-heterocyclic compounds was envisioned. A microbial putrescine oxidase (PuO) was chosen to selectively oxidize polyamines to aminoaldehydes, thereby triggering their spontaneous cyclization to an imine. To target a broad range of heterocycles, PuO was characterized with a range of unnatural polyamines. The results indicated a narrow substrate scope and low activity for these compounds. To enhance the activity for such substrates directed evolution of PuO with epPCR was performed and led to the identification of a Glu residue, representing a hotspot for mutagenesis. This residue aligns to one of the multiple channels that lead into the deeply buried active site of PuO and it is located in the second shell around the active site. By site-saturation mutagenesis further mutants in the active site and this channel were generated and many mutants with smaller amino acids demonstrated the influence of this hotspot position to increase the activity of the enzyme. The best mutant exhibited considerably increased activities (up to 25-fold) for unnatural polyamines and also for natural polyamines the substrate spectrum was strongly shifted from putrescine towards longer polyamines like spermidine which is now transformed with a 10-fold increased catalytic efficiency (kcat/KM). The combination of both enzymes in purified form as well as in whole cells enabled the production of heterocyclic amines relying on the consecutive transformation of the substrate by both enzymes. While with the whole cell system only low amounts of the N-heterocyclic compounds were produced, the utilization of purified enzymes led in case of all three IREDs to high conversions of different polyamines into pyrrolines and piperidines.Item Open Access Substrate characterization and protein engineering of bacterial cytochrome P450 monooxygenases for the bio-based synthesis of omega-hydroxylated aliphatic compounds(2013) Honda Malca, Sumire; Hauer, Bernhard (Prof. Dr.)The selective oxyfunctionalization of alkanes and fatty acids is a challenging task in basic and applied chemistry. Biocatalysts belonging to the superfamily of cytochrome P450 monooxygenases (CYPs) can introduce oxygen into a wide variety of molecules in a very regio- and stereospecific manner, which can be used for the synthesis of fine and bulk chemicals. CYPs from the bacterial CYP153A subfamily have been described as alkane hydroxylases with high terminal regioselectivity. In the present work, CYP153A monooxygenases were screened for the synthesis of industrially relevant omega-hydroxylated aliphatic compounds, such as primary alcohols, alpha,omega-diols, omega-hydroxyfatty acids (omega-OHFAs) and alpha,omega-dicarboxylic acids (alpha,omega-DCAs). One enzyme candidate was tailored by rational design and applied in whole cell biotransformations with recombinant E. coli or Pseudomonas strains. The biocatalytic systems were further improved by utilizing a fusion enzyme construct for increased coupling efficiency. In summary, this work constitutes the first example of rational engineering of a CYP153A enzyme, which allowed the identification of key residues for activity and substrate specificity within the enzyme subfamily. CYP153A enzymes have also been applied for the first time in the omega-hydroxylation of fatty acids. Comparative in vivo studies with recombinant E. coli and P. putida cells provided information on the effects of host strains on product yields, eventually leading to the generation of an efficient bacterial whole cell biocatalyst for the synthesis of selected omega-hydroxylated aliphatic compounds.Item Open Access Novel route to vanillin - an enzyme-catalyzed multi-step cascade synthesis(2016) Klaus, Tobias; Hauer, Bernhard (Prof. Dr.)The selective hydroxylation of aromatic compounds is one of the most challenging chemical reactions. As an alternative to traditional chemical catalysis, biocatalysis emerged during the past decades. Hence, in the present work, a number of biocatalysts was investigated with regard to the realization of a novel synthesis route to the valuable aromatic compound vanillin, starting from the simple low-cost aromatic substrate 3-methylanisole via the intermediate products 3-methoxybenzyl alcohol or 4-methylguaiacol and via vanillyl alcohol, as an example of consecutive enzyme-catalyzed oxidation reactions accomplished in a multi-enzymatic three-step cascade reaction. For this reason a preselected set of enzymes, namely the m-hydroxybenzoate hydroxylase MobA from Comamonas testosteroni GZ39 and the cytochrome P450 monooxygenases CYP116B3 from Rhodococcus ruber DSM 44319 and CYP102A1 from Bacillus megaterium ATCC 14581, was investigated towards the selective hydroxylation of the substrate 3-methylanisole. Beside the wild type enzymes, a variant of MobA, which was created by rational protein design, and an existing focused minimal mutant library of CYP102A1 were applied in initial biotransformation reactions, combined with an efficient cofactor recycling system. Though the wild type enzymes of CYP116B3 and CYP102A1 displayed only a basic level of activity towards 3-methylanisole, highly increased activity was detected for many of the CYP102A1 variants with a maximum of 59% total conversion for the double mutant F87V/A328L. With 3-methoxybenzyl alcohol and 4-methylguaiacol both intermediate compounds of the intended cascade synthesis were generated, though 4-methoxy-2-methylphenol was the main product in most of the reactions. However, none of the so far investigated variants accepted any of the intermediate compounds as substrate. As CYP116B3 was a good candidate for further protein engineering approaches, as a basic level of activity towards the substrate of interest was already present in the wild type enzyme, a focused mutant library of 20 single mutant variants of CYP116B3 was created based on literature, sequence and structure information in order to improve the enzymes activity and selectivity towards conversion of the substrate 3-methylanisole and in order to find variants for the conversion of 3-methoxybenzyl alcohol and/or 4-methylguaiacol. Therefore a homology model of the monooxygenase domain of CYP116B3 was generated. Though, compared to the wild type, variants with up to almost six time increased activity towards the model substrate 7-ethoxycoumarin were found, total activity towards 3-methylanisole was still much lower compared to the best CYP102A1 variants. In addition, none of the variants displayed appropriate conversion of the intermediate compounds 3-methoxybenzyl alcohol and 4-methylguaiacol. Moreover, additional mutation in the literature known amino acid position 437 of CYP102A1 variant F87V/A328L revealed no benefit towards conversion of any of the substrates, too. Molecular dynamics simulations of a CYP102A1 variant with 4-methylguaiacol as substrate revealed the bottleneck in the conversion of this compound. 4-Methylguaiacol was shown to be stabilized at the entrance of the substrate access channel by the polar amino acid residues R47 and Y51. Replacement of these residues by the hydrophobic residues leucine and phenylalanine, respectively, resulted in successful conversion of 4-methylguaiacol to vanillyl alcohol, the precursor of vanillin in the intended cascade synthesis. Though, as the yield of vanillyl alcohol synthesized from 4-methylguaiacol with CYP102A1 variants was rather low, a vanillyl alcohol oxidase from Penicillium simplicissimum and rationally designed variants thereof, described in literature, were investigated. As a result, not only vanillyl alcohol but also 4-methylguaiacol was converted in high yield to vanillin. Finally, a combination of the best 4-methylguaiacol producing variant, CYP102A1 variant A328L, with the best 4-methylguaiacol converting variant, VAO variant F454Y, in one reaction system both in vitro and in vivo yielded vanillin from 3-methylanisole with a maximal product formation of 2.0% and 1.1% vanillin, respectively. We demonstrated as a proof-of-principle the establishment of the proposed multi-enzymatic three-step cascade reaction pathway. Though further optimizations concerning increase of enzyme activity and improvement of enzyme selectivity are required, the above mentioned exemplary synthesis of vanillin illustrates the capability of biocatalysis.