Search

Arabidopsis thaliana Carotenoid Cleavage Dioxygenase 1 (AtCCD1) is the role model for CCDs, a class of enzymes that cleave highly specific C=C double bonds in carotenoids and apo-carotenoids. CCD1s cleave carotenoids symmetrically and regio-selective at the 9-10 and 9’-10’ position. The active site of CCDs contains a central, heme-free Fe2+, which is coordinated by four histidines. The enzymatic oxidation utilizes molecular oxygen and does not require further cofactors. This unique cleavage step is used in nature to create a huge variety of colorants, flavors and fragrances. It occurs in all branches of life from protozoa, metazoa to plants and fungi. In vitro, organic solvents show an activation effect towards CCD1 cleavage. The understanding of this mechanism has not been resolved up till now. Vast numbers of cleaved carotenoids show a complex chemistry and cannot be synthetized by chemical methods. Utilization of carotenoids as substrates to produce apo-carotenoids is a promising and novel approach. However, recombinant expression of AtCCD1 occurs predominantly as inclusion bodies and thus low enzymatic activity is achieved. CCD1 can be refolded from inclusion bodies, or purified by N- and C-terminal tags. Both methods yield a modified or tagged enzyme, and are limited in their application to characterize CCDs. A tag-free purification protocol was developed and permits investigation of the native AtCCD1 in vitro. High throughput screening (HTS) of AtCCD1 was possible by establishing the expression in a 96 well micro-titer-plate (MTP) format. The activity and stability of AtCCD1 with organic co-solvents was of special interest because it aided solubility characteristics of the hydrophobic carotenoids. In addition, the conversion of different carotenoid substrates was of key interest because of the regio-specificity of the cleavage reaction. Four complimentary mutagenesis methods were used to generate libraries of AtCCD1 in iterative rounds of directed evolution. Development of recombinant MTP expression in combination with novel HTS methods allowed; (I) screening conversion of carotenoids by substrate depletion, (II) fluorescence of the central cleavage product of carotenoids, (III) the conversion of apo-8’-retinal in the presence of organic co-solvents. The methods were applied to evolve AtCCD1 towards; (a) stability in the presence of organic solvents and (b) conversion of carotenoids. Several key-residues for AtCCD1 were determined, including serine 363, which is involved in substrate determination.

Benzaldehyde lyase from Pseudomonas fluorescens (BAL) and carbonylreductase from Candida parapsilosis (CPCR) are versatile catalysts for the production of various chiral intermediates, which are of interest for the pharmaceutical industry. Due to the mostly hydrophobic reactants only low productivities can be achieved in aqueous reaction systems and thus the use of alternative reaction media is desirable. Especially the use of aqueous-organic biphasic systems is promising, since the extraction of products and the spatial separation from the enzyme facilitate efficient processes. However, both enzymes show only insufficient stability in these systems and therefore a stabilisation is indispensable. Stability of both enzymes was determined in three different biphasic systems as well as the respective solvent-saturated aqueous monophasic systems (BAL: MTBE, 2 octanone, toluene; CPCR: MTBE, cyclohexane, n heptane). It was shown that BAL was already significantly deactivated by the solvent molecules within the aqueous phase. In the respective biphasic systems especially the interfaces towards 2-octanone and toluene led to considerable additional deactivation. On the molecular level CPCR was deactivated mainly by dissolved MTBE-molecules. In contrast, dissolved cyclohexane and n-heptane molecules only had minor effects on CPCR stability, whereas the respective biphasic systems led to significant deactivation of CPCR. The deactivation by the solvents is probably a result of interactions between the solvents and the proteins’ surfaces. Therefore the surface characteristics of both enzymes were altered in order to prevent these interactions. Chemical modification of free amino groups was applied to introduce new chemical functionalities to the enzymes’ surfaces. Acetylation with cyclic anhydrides decreased activity of both enzymes significantly. In contrast, modification with mPEG750 and mPEG2000 resulted in good residual activities of BAL (60-70%), while CPCR retained less than 10% residual activity. Gel electrophoresis showed that apparently only few amino groups of BAL were accessible for mPEG-modification, whereas CPCR was modified to great extent. Storage stability measurements with the mPEG-BAL-variants revealed a considerable increase of stability towards dissolved molecules in the solvent systems with MTBE and toluene. Site-directed mutagenesis was applied to exchange unpolar surface amino acid residues against polar amino acid residues in order to decrease interactions between the enzymes and the solvents. Most of the resulting BAL-variants showed similar activities as the native BAL, whereas many CPCR-variants lost most of their activity. For both enzymes enantioselectivity was unaffected by the introduced mutations. Kinetic constants of BAL-variants differed only marginally from the ones of native BAL. CPCR-variants exhibited higher kcat-values than native CPCR in both reduction- and oxidation reaction. Determination of storage stability in the different mono-and biphasic systems showed that most variants exhibited decreased stability against dissolved solvent molecules compared to native enzymes. However, additional deactivation at the interface towards toluene (BAL) and MTBE, cyclohexane und n-heptane (CPCR) respectively, was decreased. For the technical application enzyme stability under relevant process conditions is of importance. Therefore a reactor set-up was established, that allows stability determination in reactive, aqueous-organic biphasic systems. Respective measurements with BAL indicated that deactivation by the aldehyde substrate superimposed the deactivation by the solvents, thus resulting in similar BAL stability in all tested solvent systems. Neither the mPEG-modification nor the introduced mutations prevented this deactivation by the substrate. CPCR showed a strong stability decrease under process conditions in the MTBE-system, probably due to a combined effect of dissolved MTBE-molecules and isopropanol which was used for substrate-coupled cofactor regeneration. CPCR stability in reactive systems with cyclohexane and n-heptane was comparable to stability under non-reactive conditions. Variant CPCR-I79T,V83S showed increased process stability in all three biphasic systems.

Despite the broad use of lipases in ester synthesis procedures, crucial aspects on optimal pH and enzyme behaviour in biphasic reaction media are still not well-known. Such information is highly desired for the rational process design. Therefore, the influence of the pH in lipase-based esterifications in biphasic media was studied. For Candida rugosa and Thermomyces lanuginosa lipases the pH-optima of lipase-catalysed esterifications were pH 3.5 and 4.25, respectively, what significantly differ from the optima of the hydrolysis reaction. Enzyme activity increased with increasing concentrations of protonated acid, allegeable by the reaction mechanism: Different from the protonated, the deprotonated acid cannot react with the nucleophilic serin residue due to its delocalized negative charge. Hence the protonated acid is the substrate for esterification. Conclusively, high synthetic activity is only accessible at lower pH. The importance and generality of this conclusion was verified by testing of other lipases under the same reaction conditions. The pH drop in biphasic systems – caused by enrichment of the acid in the aqueous phase – is not a severe problem as assumed by many researchers and efforts for pH-control are dispensable. Moreover, a strong linear correlation between the enzyme activity and the specific interface was shown for esterifications. An indepth analysis of the experiments in consideration of the thermodynamic facts resulted in the development of a detailed hypothesis about the nature of lipase-catalysed esterification in biphasic systems. The ester synthesis predominantly takes places at the interface, thus in interest of high productivity the specific interface needs to be maximized. Based on this hypothesis, the application of lipase containing hydrogels is useless and an alternative concept for the immobilisation of lipases in biphasic systems was developed. The innovative concept – the static emulsion – consists in the emulsification of an aqueous lipase solution within a hydrophobic silicone elastomer. Spherical immobilisates were obtained with a gum-like consistency. The catalytic and mechanic properties of the static emulsion were closely characterised. According to their mechanical properties, the static emulsion is superior to hydrogels and sol-gels. For all investigated lipases it was shown that the immobilisation comes along with a strong enhancement of the catalytic activity, which was so far only accessible in a similar extent by immobilization in alkylsubstituted sol-gels. The practical suitability was demonstrated by the solvent-free synthesis of fatty acid ethyl hexyl ester, propyllaurate and the kinetic resolution of racemic benzoine in tetrahydrofuran and makes clear that the static emulsion is a method with high future potential, since it enables an improved performance of immobilised lipases.